cell structure research paper

• Journal home
• About the journal
• J-STAGE home
• Cell Structure and Function
• Lysosomal membrane integrity in fibroblasts derived from patients with Gaucher disease Asuka Hamamoto, Natsuki Kita, Siddabasave Gowda B. Gowda, Hiroyuki Takatsu, Kazuhisa Nakayama, Makoto Arita, Shu-Ping Hui, Hye-Won Shin
• In vivo evidence for homeostatic regulation of ribosomal protein levels in Drosophila Daiki Kitamura, Kiichiro Taniguchi, Mai Nakamura, Tatsushi Igaki

Ubiquitin-like 3 as a new protein-sorting factor for small extracellular vesicles

Released on J-STAGE: February 23, 2022 | Volume 47 Issue 1 Pages 1-18

Yusuke Takanashi, Tomoaki Kahyo, Sae Kamamoto, Hengsen Zhang, Bin Chen, Yashuang Ping, Kiyomichi Mizuno, Akikazu Kawase, Kei Koizumi, Masanori Satou, Kazuhito Funai, Norihiko Shiiya, Mitsutoshi Setou

Two coral fluorescent proteins of distinct colors for sharp visualization of cell-cycle progression

Released on J-STAGE: July 29, 2023 | Volume 48 Issue 2 Pages 135-144

Ryoko Ando, Asako Sakaue-Sawano, Keiko Shoda, Atsushi Miyawaki

GFP-like Proteins Stably Accumulate in Lysosomes

Released on J-STAGE: March 27, 2008 | Volume 33 Issue 1 Pages 1-12

Hiroyuki Katayama, Akitsugu Yamamoto, Noboru Mizushima, Tamotsu Yoshimori, Atsushi Miyawaki

Lysosomal membrane integrity in fibroblasts derived from patients with Gaucher disease

Released on J-STAGE: January 23, 2024 | Volume 49 Issue 1 Pages 1-10

Asuka Hamamoto, Natsuki Kita, Siddabasave Gowda B. Gowda, Hiroyuki Takatsu, Kazuhisa Nakayama, Makoto Arita, Shu-Ping Hui, Hye-Won Shin

In vivo evidence for homeostatic regulation of ribosomal protein levels in Drosophila

Released on J-STAGE: February 16, 2024 | Volume 49 Issue 1 Pages 11-20

Daiki Kitamura, Kiichiro Taniguchi, Mai Nakamura, Tatsushi Igaki

• Volume Vol 49 Vol 48 Vol 47 Vol 46 Vol 45 Vol 44 Vol 43 Vol 42 Vol 41 Vol 40 Vol 39 Vol 38 Vol 37 Vol 36 Vol 35 Vol 34 Vol 33 Vol 32 Vol 31 Vol 30 Vol 29 Vol 28 Vol 27 Vol 26 Vol 25 Vol 24 Vol 23 Vol 22 Vol 21 Vol 20 Vol 19 Vol 18 Vol 17 Vol 16 Vol 15 Vol 14 Vol 13 Vol 12 Vol 11 Vol 10 Vol 9 Vol 8 Vol 7 Vol 6 Vol 5 Vol 4 Vol 3 Vol 2 Vol 1
• Add to favorites
• Announcement alert
• New arrival alert
• This publication was supported by JSPS KAKENHI Grant Number JP17HP2012.
• Web site of Japan Society for Cell Biology: http://www.jscb.gr.jp/
• Web site of Cell Structure and Function http://www.jscb.gr.jp/csf/en/

Register with J-STAGE for free!

Already have an account? Sign in here

The Pathophysiologic Basis of Nuclear Medicine pp 1–28 Cite as

The Cell: Structure, Function, and Molecular Biology

• Shankar Vallabhajosula &
• Seham Mustafa  

1622 Accesses

This is a preview of subscription content, log in via an institution .

Buying options

• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Unable to display preview.  Download preview PDF.

Alberts B, Bray D, Lewis J, et al (1994) Molecular biology of the cell, 3rd edn. Garland Publishing, New York

Google Scholar  

Kumar V, Abbas A, Faust N (2004) Robbins and Cotran pathologic basis of disease, 7th edn. W.B. Saunders, Philadelphia

Devin TM (1997) Textbook of biochemistry with clinical correlates, 4th edn. Wiley-Liss, New York

Guyton AC, Hall JE (1997) Human physiology and mechanisms of disease, 6th edn. W.B. Saunders, Philadelphia

Junqueira LC, Carneiro J, Kelley R (1995) Basic histology, 8th edn. Prentice-Hall, NJ, USA

McCance KL, Huether SE (1998) Pathophysiology. The biologic basis for disease in adults and children, 3rd edn. Mosby Year Book, St. Louis

Pollard TD, Earnshaw WC (2002) Cell biology, 1st edn. Saunders, Philadelphia

Raven PH, Johnson GB (1992) Biology, 3rd edn. Mosby Year Book, St. Louis

Sumner AT (2003) Chromosomes organization and function, 1st edn. Blackwell, Oxford

Yin XM, Dong Z (2003) Essentials of apoptosis: A guide for basic and clinical research, 1st edn. Humana Press, New Jersey, USA

Download references

You can also search for this author in PubMed   Google Scholar

Editor information

Editors and affiliations.

Department of Nuclear Medicine, Faculty of Medicine Kuwait University Health Sciences Center, PO Box 24923, 13110, Safat, Kuwait

Abdelhamid H. Elgazzar MD, FCAP ( Diplomate, American Board of Pathology Diplomate, American Board of Nuclear Medicine Professor and Chairman ) ( Diplomate, American Board of Pathology Diplomate, American Board of Nuclear Medicine Professor and Chairman )

Rights and permissions

Reprints and permissions

Copyright information

© 2006 Springer-Verlag Berlin Heidelberg

About this chapter

Cite this chapter.

Vallabhajosula, S., Mustafa, S. (2006). The Cell: Structure, Function, and Molecular Biology. In: Elgazzar, A.H. (eds) The Pathophysiologic Basis of Nuclear Medicine. Springer, Berlin, Heidelberg . https://doi.org/10.1007/978-3-540-47953-6_1

Download citation

DOI : https://doi.org/10.1007/978-3-540-47953-6_1

Publisher Name : Springer, Berlin, Heidelberg

Print ISBN : 978-3-540-23992-5

Online ISBN : 978-3-540-47953-6

eBook Packages : Medicine Medicine (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

• Search Menu
• Browse content in Arts and Humanities
• Browse content in Archaeology
• Anglo-Saxon and Medieval Archaeology
• Archaeological Methodology and Techniques
• Archaeology by Region
• Archaeology of Religion
• Archaeology of Trade and Exchange
• Biblical Archaeology
• Contemporary and Public Archaeology
• Environmental Archaeology
• Historical Archaeology
• History and Theory of Archaeology
• Industrial Archaeology
• Landscape Archaeology
• Mortuary Archaeology
• Prehistoric Archaeology
• Underwater Archaeology
• Urban Archaeology
• Zooarchaeology
• Browse content in Architecture
• Architectural Structure and Design
• History of Architecture
• Residential and Domestic Buildings
• Theory of Architecture
• Browse content in Art
• Art Subjects and Themes
• History of Art
• Industrial and Commercial Art
• Theory of Art
• Biographical Studies
• Byzantine Studies
• Browse content in Classical Studies
• Classical History
• Classical Philosophy
• Classical Mythology
• Classical Literature
• Classical Reception
• Classical Art and Architecture
• Classical Oratory and Rhetoric
• Greek and Roman Papyrology
• Greek and Roman Epigraphy
• Greek and Roman Law
• Greek and Roman Archaeology
• Late Antiquity
• Religion in the Ancient World
• Digital Humanities
• Browse content in History
• Colonialism and Imperialism
• Diplomatic History
• Environmental History
• Genealogy, Heraldry, Names, and Honours
• Genocide and Ethnic Cleansing
• Historical Geography
• History by Period
• History of Emotions
• History of Agriculture
• History of Education
• History of Gender and Sexuality
• Industrial History
• Intellectual History
• International History
• Labour History
• Legal and Constitutional History
• Local and Family History
• Maritime History
• Military History
• National Liberation and Post-Colonialism
• Oral History
• Political History
• Public History
• Regional and National History
• Revolutions and Rebellions
• Slavery and Abolition of Slavery
• Social and Cultural History
• Theory, Methods, and Historiography
• Urban History
• World History
• Browse content in Language Teaching and Learning
• Language Learning (Specific Skills)
• Language Teaching Theory and Methods
• Browse content in Linguistics
• Applied Linguistics
• Cognitive Linguistics
• Computational Linguistics
• Forensic Linguistics
• Grammar, Syntax and Morphology
• Historical and Diachronic Linguistics
• History of English
• Language Evolution
• Language Reference
• Language Acquisition
• Language Variation
• Language Families
• Lexicography
• Linguistic Anthropology
• Linguistic Theories
• Linguistic Typology
• Phonetics and Phonology
• Psycholinguistics
• Sociolinguistics
• Translation and Interpretation
• Writing Systems
• Browse content in Literature
• Bibliography
• Children's Literature Studies
• Literary Studies (Romanticism)
• Literary Studies (American)
• Literary Studies (Asian)
• Literary Studies (European)
• Literary Studies (Eco-criticism)
• Literary Studies (Modernism)
• Literary Studies - World
• Literary Studies (1500 to 1800)
• Literary Studies (19th Century)
• Literary Studies (20th Century onwards)
• Literary Studies (African American Literature)
• Literary Studies (British and Irish)
• Literary Studies (Early and Medieval)
• Literary Studies (Fiction, Novelists, and Prose Writers)
• Literary Studies (Gender Studies)
• Literary Studies (Graphic Novels)
• Literary Studies (History of the Book)
• Literary Studies (Plays and Playwrights)
• Literary Studies (Poetry and Poets)
• Literary Studies (Postcolonial Literature)
• Literary Studies (Queer Studies)
• Literary Studies (Science Fiction)
• Literary Studies (Travel Literature)
• Literary Studies (War Literature)
• Literary Studies (Women's Writing)
• Literary Theory and Cultural Studies
• Mythology and Folklore
• Shakespeare Studies and Criticism
• Browse content in Media Studies
• Browse content in Music
• Applied Music
• Dance and Music
• Ethics in Music
• Ethnomusicology
• Gender and Sexuality in Music
• Medicine and Music
• Music Cultures
• Music and Media
• Music and Religion
• Music and Culture
• Music Education and Pedagogy
• Music Theory and Analysis
• Musical Scores, Lyrics, and Libretti
• Musical Structures, Styles, and Techniques
• Musicology and Music History
• Performance Practice and Studies
• Race and Ethnicity in Music
• Sound Studies
• Browse content in Performing Arts
• Browse content in Philosophy
• Aesthetics and Philosophy of Art
• Epistemology
• Feminist Philosophy
• History of Western Philosophy
• Metaphysics
• Moral Philosophy
• Non-Western Philosophy
• Philosophy of Language
• Philosophy of Mind
• Philosophy of Perception
• Philosophy of Science
• Philosophy of Action
• Philosophy of Law
• Philosophy of Religion
• Philosophy of Mathematics and Logic
• Practical Ethics
• Social and Political Philosophy
• Browse content in Religion
• Biblical Studies
• Christianity
• East Asian Religions
• History of Religion
• Judaism and Jewish Studies
• Qumran Studies
• Religion and Education
• Religion and Health
• Religion and Politics
• Religion and Science
• Religion and Law
• Religion and Art, Literature, and Music
• Religious Studies
• Browse content in Society and Culture
• Cookery, Food, and Drink
• Cultural Studies
• Customs and Traditions
• Ethical Issues and Debates
• Hobbies, Games, Arts and Crafts
• Lifestyle, Home, and Garden
• Natural world, Country Life, and Pets
• Popular Beliefs and Controversial Knowledge
• Sports and Outdoor Recreation
• Technology and Society
• Travel and Holiday
• Visual Culture
• Browse content in Law
• Arbitration
• Browse content in Company and Commercial Law
• Commercial Law
• Company Law
• Browse content in Comparative Law
• Systems of Law
• Competition Law
• Browse content in Constitutional and Administrative Law
• Government Powers
• Judicial Review
• Local Government Law
• Military and Defence Law
• Parliamentary and Legislative Practice
• Construction Law
• Contract Law
• Browse content in Criminal Law
• Criminal Procedure
• Criminal Evidence Law
• Sentencing and Punishment
• Employment and Labour Law
• Environment and Energy Law
• Browse content in Financial Law
• Banking Law
• Insolvency Law
• History of Law
• Human Rights and Immigration
• Intellectual Property Law
• Browse content in International Law
• Private International Law and Conflict of Laws
• Public International Law
• IT and Communications Law
• Jurisprudence and Philosophy of Law
• Law and Politics
• Law and Society
• Browse content in Legal System and Practice
• Courts and Procedure
• Legal Skills and Practice
• Primary Sources of Law
• Regulation of Legal Profession
• Medical and Healthcare Law
• Browse content in Policing
• Criminal Investigation and Detection
• Police and Security Services
• Police Procedure and Law
• Police Regional Planning
• Browse content in Property Law
• Personal Property Law
• Study and Revision
• Terrorism and National Security Law
• Browse content in Trusts Law
• Wills and Probate or Succession
• Browse content in Medicine and Health
• Browse content in Allied Health Professions
• Arts Therapies
• Clinical Science
• Dietetics and Nutrition
• Occupational Therapy
• Operating Department Practice
• Physiotherapy
• Radiography
• Speech and Language Therapy
• Browse content in Anaesthetics
• General Anaesthesia
• Neuroanaesthesia
• Clinical Neuroscience
• Browse content in Clinical Medicine
• Acute Medicine
• Cardiovascular Medicine
• Clinical Genetics
• Clinical Pharmacology and Therapeutics
• Dermatology
• Endocrinology and Diabetes
• Gastroenterology
• Genito-urinary Medicine
• Geriatric Medicine
• Infectious Diseases
• Medical Toxicology
• Medical Oncology
• Pain Medicine
• Palliative Medicine
• Rehabilitation Medicine
• Respiratory Medicine and Pulmonology
• Rheumatology
• Sleep Medicine
• Sports and Exercise Medicine
• Community Medical Services
• Critical Care
• Emergency Medicine
• Forensic Medicine
• Haematology
• History of Medicine
• Browse content in Medical Skills
• Clinical Skills
• Communication Skills
• Nursing Skills
• Surgical Skills
• Browse content in Medical Dentistry
• Oral and Maxillofacial Surgery
• Paediatric Dentistry
• Restorative Dentistry and Orthodontics
• Surgical Dentistry
• Medical Ethics
• Medical Statistics and Methodology
• Browse content in Neurology
• Clinical Neurophysiology
• Neuropathology
• Nursing Studies
• Browse content in Obstetrics and Gynaecology
• Gynaecology
• Occupational Medicine
• Ophthalmology
• Otolaryngology (ENT)
• Browse content in Paediatrics
• Neonatology
• Browse content in Pathology
• Chemical Pathology
• Clinical Cytogenetics and Molecular Genetics
• Histopathology
• Medical Microbiology and Virology
• Patient Education and Information
• Browse content in Pharmacology
• Psychopharmacology
• Browse content in Popular Health
• Caring for Others
• Complementary and Alternative Medicine
• Self-help and Personal Development
• Browse content in Preclinical Medicine
• Cell Biology
• Molecular Biology and Genetics
• Reproduction, Growth and Development
• Primary Care
• Professional Development in Medicine
• Browse content in Psychiatry
• Addiction Medicine
• Child and Adolescent Psychiatry
• Forensic Psychiatry
• Learning Disabilities
• Old Age Psychiatry
• Psychotherapy
• Browse content in Public Health and Epidemiology
• Epidemiology
• Public Health
• Browse content in Radiology
• Clinical Radiology
• Interventional Radiology
• Nuclear Medicine
• Radiation Oncology
• Reproductive Medicine
• Browse content in Surgery
• Cardiothoracic Surgery
• Gastro-intestinal and Colorectal Surgery
• General Surgery
• Neurosurgery
• Paediatric Surgery
• Peri-operative Care
• Plastic and Reconstructive Surgery
• Surgical Oncology
• Transplant Surgery
• Trauma and Orthopaedic Surgery
• Vascular Surgery
• Browse content in Science and Mathematics
• Browse content in Biological Sciences
• Aquatic Biology
• Biochemistry
• Bioinformatics and Computational Biology
• Developmental Biology
• Ecology and Conservation
• Evolutionary Biology
• Genetics and Genomics
• Microbiology
• Molecular and Cell Biology
• Natural History
• Plant Sciences and Forestry
• Research Methods in Life Sciences
• Structural Biology
• Systems Biology
• Zoology and Animal Sciences
• Browse content in Chemistry
• Analytical Chemistry
• Computational Chemistry
• Crystallography
• Environmental Chemistry
• Industrial Chemistry
• Inorganic Chemistry
• Materials Chemistry
• Medicinal Chemistry
• Mineralogy and Gems
• Organic Chemistry
• Physical Chemistry
• Polymer Chemistry
• Study and Communication Skills in Chemistry
• Theoretical Chemistry
• Browse content in Computer Science
• Artificial Intelligence
• Computer Architecture and Logic Design
• Game Studies
• Human-Computer Interaction
• Mathematical Theory of Computation
• Programming Languages
• Software Engineering
• Systems Analysis and Design
• Virtual Reality
• Browse content in Computing
• Business Applications
• Computer Security
• Computer Games
• Computer Networking and Communications
• Digital Lifestyle
• Graphical and Digital Media Applications
• Operating Systems
• Browse content in Earth Sciences and Geography
• Atmospheric Sciences
• Environmental Geography
• Geology and the Lithosphere
• Maps and Map-making
• Meteorology and Climatology
• Oceanography and Hydrology
• Palaeontology
• Physical Geography and Topography
• Regional Geography
• Soil Science
• Urban Geography
• Browse content in Engineering and Technology
• Agriculture and Farming
• Biological Engineering
• Civil Engineering, Surveying, and Building
• Electronics and Communications Engineering
• Energy Technology
• Engineering (General)
• Environmental Science, Engineering, and Technology
• History of Engineering and Technology
• Mechanical Engineering and Materials
• Technology of Industrial Chemistry
• Transport Technology and Trades
• Browse content in Environmental Science
• Applied Ecology (Environmental Science)
• Conservation of the Environment (Environmental Science)
• Environmental Sustainability
• Environmentalist Thought and Ideology (Environmental Science)
• Management of Land and Natural Resources (Environmental Science)
• Natural Disasters (Environmental Science)
• Nuclear Issues (Environmental Science)
• Pollution and Threats to the Environment (Environmental Science)
• Social Impact of Environmental Issues (Environmental Science)
• History of Science and Technology
• Browse content in Materials Science
• Ceramics and Glasses
• Composite Materials
• Metals, Alloying, and Corrosion
• Nanotechnology
• Browse content in Mathematics
• Applied Mathematics
• Biomathematics and Statistics
• History of Mathematics
• Mathematical Education
• Mathematical Finance
• Mathematical Analysis
• Numerical and Computational Mathematics
• Probability and Statistics
• Pure Mathematics
• Browse content in Neuroscience
• Cognition and Behavioural Neuroscience
• Development of the Nervous System
• Disorders of the Nervous System
• History of Neuroscience
• Invertebrate Neurobiology
• Molecular and Cellular Systems
• Neuroendocrinology and Autonomic Nervous System
• Neuroscientific Techniques
• Sensory and Motor Systems
• Browse content in Physics
• Astronomy and Astrophysics
• Atomic, Molecular, and Optical Physics
• Biological and Medical Physics
• Classical Mechanics
• Computational Physics
• Condensed Matter Physics
• Electromagnetism, Optics, and Acoustics
• History of Physics
• Mathematical and Statistical Physics
• Measurement Science
• Nuclear Physics
• Particles and Fields
• Plasma Physics
• Quantum Physics
• Relativity and Gravitation
• Semiconductor and Mesoscopic Physics
• Browse content in Psychology
• Affective Sciences
• Clinical Psychology
• Cognitive Psychology
• Cognitive Neuroscience
• Criminal and Forensic Psychology
• Developmental Psychology
• Educational Psychology
• Evolutionary Psychology
• Health Psychology
• History and Systems in Psychology
• Music Psychology
• Neuropsychology
• Organizational Psychology
• Psychological Assessment and Testing
• Psychology of Human-Technology Interaction
• Psychology Professional Development and Training
• Research Methods in Psychology
• Social Psychology
• Browse content in Social Sciences
• Browse content in Anthropology
• Anthropology of Religion
• Human Evolution
• Medical Anthropology
• Physical Anthropology
• Regional Anthropology
• Social and Cultural Anthropology
• Theory and Practice of Anthropology
• Browse content in Business and Management
• Business Ethics
• Business Strategy
• Business History
• Business and Technology
• Business and Government
• Business and the Environment
• Comparative Management
• Corporate Governance
• Corporate Social Responsibility
• Entrepreneurship
• Health Management
• Human Resource Management
• Industrial and Employment Relations
• Industry Studies
• Information and Communication Technologies
• International Business
• Knowledge Management
• Management and Management Techniques
• Operations Management
• Organizational Theory and Behaviour
• Pensions and Pension Management
• Public and Nonprofit Management
• Strategic Management
• Supply Chain Management
• Browse content in Criminology and Criminal Justice
• Criminal Justice
• Criminology
• Forms of Crime
• International and Comparative Criminology
• Youth Violence and Juvenile Justice
• Development Studies
• Browse content in Economics
• Agricultural, Environmental, and Natural Resource Economics
• Asian Economics
• Behavioural Finance
• Behavioural Economics and Neuroeconomics
• Econometrics and Mathematical Economics
• Economic History
• Economic Systems
• Economic Methodology
• Economic Development and Growth
• Financial Markets
• Financial Institutions and Services
• General Economics and Teaching
• Health, Education, and Welfare
• History of Economic Thought
• International Economics
• Labour and Demographic Economics
• Law and Economics
• Macroeconomics and Monetary Economics
• Microeconomics
• Public Economics
• Urban, Rural, and Regional Economics
• Welfare Economics
• Browse content in Education
• Adult Education and Continuous Learning
• Care and Counselling of Students
• Early Childhood and Elementary Education
• Educational Equipment and Technology
• Educational Strategies and Policy
• Higher and Further Education
• Organization and Management of Education
• Philosophy and Theory of Education
• Schools Studies
• Secondary Education
• Teaching of a Specific Subject
• Teaching of Specific Groups and Special Educational Needs
• Teaching Skills and Techniques
• Browse content in Environment
• Applied Ecology (Social Science)
• Climate Change
• Conservation of the Environment (Social Science)
• Environmentalist Thought and Ideology (Social Science)
• Natural Disasters (Environment)
• Social Impact of Environmental Issues (Social Science)
• Browse content in Human Geography
• Cultural Geography
• Economic Geography
• Political Geography
• Browse content in Interdisciplinary Studies
• Communication Studies
• Museums, Libraries, and Information Sciences
• Browse content in Politics
• African Politics
• Asian Politics
• Chinese Politics
• Comparative Politics
• Conflict Politics
• Elections and Electoral Studies
• Environmental Politics
• European Union
• Foreign Policy
• Gender and Politics
• Human Rights and Politics
• Indian Politics
• International Relations
• International Organization (Politics)
• International Political Economy
• Irish Politics
• Latin American Politics
• Middle Eastern Politics
• Political Behaviour
• Political Economy
• Political Institutions
• Political Methodology
• Political Communication
• Political Philosophy
• Political Sociology
• Political Theory
• Politics and Law
• Public Policy
• Public Administration
• Quantitative Political Methodology
• Regional Political Studies
• Russian Politics
• Security Studies
• State and Local Government
• UK Politics
• US Politics
• Browse content in Regional and Area Studies
• African Studies
• Asian Studies
• East Asian Studies
• Japanese Studies
• Latin American Studies
• Middle Eastern Studies
• Native American Studies
• Scottish Studies
• Browse content in Research and Information
• Research Methods
• Browse content in Social Work
• Addictions and Substance Misuse
• Adoption and Fostering
• Care of the Elderly
• Child and Adolescent Social Work
• Couple and Family Social Work
• Developmental and Physical Disabilities Social Work
• Direct Practice and Clinical Social Work
• Emergency Services
• Human Behaviour and the Social Environment
• International and Global Issues in Social Work
• Mental and Behavioural Health
• Social Justice and Human Rights
• Social Policy and Advocacy
• Social Work and Crime and Justice
• Social Work Macro Practice
• Social Work Practice Settings
• Social Work Research and Evidence-based Practice
• Welfare and Benefit Systems
• Browse content in Sociology
• Childhood Studies
• Community Development
• Comparative and Historical Sociology
• Economic Sociology
• Gender and Sexuality
• Gerontology and Ageing
• Health, Illness, and Medicine
• Marriage and the Family
• Migration Studies
• Occupations, Professions, and Work
• Organizations
• Population and Demography
• Race and Ethnicity
• Social Theory
• Social Movements and Social Change
• Social Research and Statistics
• Social Stratification, Inequality, and Mobility
• Sociology of Religion
• Sociology of Education
• Sport and Leisure
• Urban and Rural Studies
• Browse content in Warfare and Defence
• Defence Strategy, Planning, and Research
• Land Forces and Warfare
• Military Administration
• Military Life and Institutions
• Naval Forces and Warfare
• Other Warfare and Defence Issues
• Peace Studies and Conflict Resolution
• Weapons and Equipment

• < Previous chapter
• Next chapter >

2 (page 17) p. 17 The structure of cells

• Published: September 2011
• Cite Icon Cite
• Permissions Icon Permissions

‘The structure of cells’ examines detailed cell structure. The plasma membrane is the cell's boundary structure and provides a barrier to other cells and the external environment. These membranes are an incredibly complicated and dynamic mixture of lipid (fat) molecules and proteins. The features of cell structure such as mitochondria, cytoplasm, organelles, endoplasmic reticulum, cilia, flagella, microtubules, and cytoskeleton are described. The principles of energy and heat generation, protein and lipid production, and the processes of phagocytosis and exocytosis are all explained. The history of cell structure discovery is provided along with descriptions of phase contrast microscopy and time-lapse microscopy used to view these structures.

Signed in as

Institutional accounts.

• Google Scholar Indexing
• GoogleCrawler [DO NOT DELETE]

Personal account

• Sign in with email/username & password
• Get email alerts
• Save searches
• Purchase content
• Activate your purchase/trial code

Institutional access

• Sign in with a library card Sign in with username/password Recommend to your librarian
• Institutional account management
• Get help with access

Access to content on Oxford Academic is often provided through institutional subscriptions and purchases. If you are a member of an institution with an active account, you may be able to access content in one of the following ways:

IP based access

Typically, access is provided across an institutional network to a range of IP addresses. This authentication occurs automatically, and it is not possible to sign out of an IP authenticated account.

Sign in through your institution

Choose this option to get remote access when outside your institution. Shibboleth/Open Athens technology is used to provide single sign-on between your institution’s website and Oxford Academic.

• Click Sign in through your institution.
• Select your institution from the list provided, which will take you to your institution's website to sign in.
• When on the institution site, please use the credentials provided by your institution. Do not use an Oxford Academic personal account.
• Following successful sign in, you will be returned to Oxford Academic.

If your institution is not listed or you cannot sign in to your institution’s website, please contact your librarian or administrator.

Sign in with a library card

Enter your library card number to sign in. If you cannot sign in, please contact your librarian.

Society Members

Society member access to a journal is achieved in one of the following ways:

Sign in through society site

Many societies offer single sign-on between the society website and Oxford Academic. If you see ‘Sign in through society site’ in the sign in pane within a journal:

• Click Sign in through society site.
• When on the society site, please use the credentials provided by that society. Do not use an Oxford Academic personal account.

If you do not have a society account or have forgotten your username or password, please contact your society.

Sign in using a personal account

Some societies use Oxford Academic personal accounts to provide access to their members. See below.

A personal account can be used to get email alerts, save searches, purchase content, and activate subscriptions.

Some societies use Oxford Academic personal accounts to provide access to their members.

Viewing your signed in accounts

Click the account icon in the top right to:

• View your signed in personal account and access account management features.
• View the institutional accounts that are providing access.

Signed in but can't access content

Oxford Academic is home to a wide variety of products. The institutional subscription may not cover the content that you are trying to access. If you believe you should have access to that content, please contact your librarian.

For librarians and administrators, your personal account also provides access to institutional account management. Here you will find options to view and activate subscriptions, manage institutional settings and access options, access usage statistics, and more.

Our books are available by subscription or purchase to libraries and institutions.

• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Rights and permissions
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

• Open access
• Published: 19 December 2014

Once upon a time the cell membranes: 175 years of cell boundary research

• Jonathan Lombard 1  

Biology Direct volume  9 , Article number:  32 ( 2014 ) Cite this article

51k Accesses

105 Citations

7 Altmetric

Metrics details

All modern cells are bounded by cell membranes best described by the fluid mosaic model. This statement is so widely accepted by biologists that little attention is generally given to the theoretical importance of cell membranes in describing the cell. This has not always been the case. When the Cell Theory was first formulated in the XIX th century, almost nothing was known about the cell membranes. It was not until well into the XX th century that the existence of the plasma membrane was broadly accepted and, even then, the fluid mosaic model did not prevail until the 1970s. How were the cell boundaries considered between the articulation of the Cell Theory around 1839 and the formulation of the fluid mosaic model that has described the cell membranes since 1972? In this review I will summarize the major historical discoveries and theories that tackled the existence and structure of membranes and I will analyze how these theories impacted the understanding of the cell. Apart from its purely historical relevance, this account can provide a starting point for considering the theoretical significance of membranes to the definition of the cell and could have implications for research on early life.

This article was reviewed by Dr. Étienne Joly, Dr. Eugene V. Koonin and Dr. Armen Mulkidjanian.

Introduction

Modern descriptions of the cell are intimately related to the notion of cell membranes. The cell membrane is not only the boundary of the unit of life, it is also a specific compartment that harbors many essential cell functions including communication with the environment, transport of molecules and certain metabolic functions. Nowadays, the consensual model to depict the membrane structure and functions is called the “fluid mosaic model” [ 1 ].

The fluid mosaic hypothesis was formulated by Singer and Nicolson in the early 1970s [ 1 ]. According to this model, membranes are made up of lipids, proteins and carbohydrates (Figure  1 ). The main lipid membrane components are phospholipids. These molecules are amphiphilic, i. e. they have one polar part attracted by water (hydrophilic) and one apolar component repelled by water (hydrophobic). When they are diluted in water, amphiphiles spontaneously adopt the most thermodynamically stable molecular structure, namely the one that maximizes both hydrophilic and hydrophobic interactions [ 2 ]. These interactions may be affected by several parameters, such as the chemical nature of the molecules, their size, the salinity and pH of the solution. In biological conditions, cell phospholipids form a bilayer in which hydrophobic tails face each other in the core of the structure whereas the hydrophilic heads interact with the surrounding water (Figure  1 ). Since proteins are also amphiphilic molecules, the same constraints apply to them. Some proteins (called intrinsic or integral) are embedded in the lipid bilayer matrix where they are able to establish hydrophobic and hydrophilic interactions with their respective lipid counterparts. Other proteins, called extrinsic or peripheral proteins, can also be transiently associated with membrane surfaces through weaker interactions (Figure  1 ). Finally, carbohydrates can be linked to either proteins or lipids, resulting in glycoproteins or glycolipids.

Fluid mosaic model. Schematic view of biological membrane structure as currently depicted.

The “mosaic” term of this model refers to the mixture of lipids and intrinsic proteins in the membrane. These boundaries are also “fluid” because their components can move laterally, allowing both diffusion of components and local specific gatherings. Other lipids, such as cholesterol, act as membrane fluidity regulators. Phospholipid movements are generally restricted to lateral drift, because the cross of the membrane from one side to the other requires the energetically unfavorable transient contact of their hydrophilic head with the hydrophobic membrane core. Thus, the transfer of molecules from one side of the membrane to the other generally involves the activity of some specific integral membrane proteins, called flippases [ 3 ]. For the same reasons, integral proteins can diffuse within the lipid matrix but they seldom switch their polarity from one membrane side to the other. As a result, lipid, protein and carbohydrate composition are different between the two monolayers, a characteristic that is referred to as membrane asymmetry.

Membrane functions are extremely diverse. As cell borders, membranes control the molecular exchanges with the environment, resulting in cell pH regulation and osmotic homeostasis. Membranes are “selective barriers”: They concentrate nutrients within the cell, exclude the cellular waste products, keep the ionic gradients and transform them into chemical energy. Since they allow the transduction of many external stimuli into cell signals, they are also major actors in the responses of the cell to their environment. In addition, their composition also turns membranes into the main apolar compartment of the prominently aqueous cell medium, thus concentrating most lipid pigments (e.g. chlorophyll) and hydrophobic proteins. The presence of these molecules in the membranes doubles their bounding function with essential metabolic and bioenergetic activities.

Except for some rare authors who still envisage the cell as a naked colloid network [ 4 ], there is nowadays little disagreement that membranes are essential parts of all contemporary cells. Despite this basic acceptance concerning modern cells, we have witnessed in recent years a strong debate questioning the presence of similar membranes in the last common ancestor of living organisms, namely the cenancestor. Arguing about the presence or absence of membranes in early organisms–not only the cenancestor, but also previous organisms closer in time to the origins of life–challenges what we consider to be the basic unit of life, i.e. the cell. Unfortunately, because the lack of membranes is generally unquestioned in modern organisms, it is nowadays difficult to come across discussions about the theoretical importance of membranes.

The limited attention currently given to membranes in defining the cell concept contrasts greatly with the importance that this issue had in early cell studies. Indeed, when the Cell Theory was formulated 175 years ago in the XIX th century, the reality of the membrane was unknown. Its universal character was not generally acknowledged until well into the XX th century, and even when the cells were assumed to be bounded by some kind of membrane, the fluid mosaic model was not accepted until as late as the 1970s. The natural question then is: how were the cell boundaries envisioned between the formulation of the Cell Theory around 1839 and the final predominance of the fluid mosaic model in 1972? In this review, I will provide some answers to this question that I think will be useful in three different ways. First, it will considerably extend the range of some recent publications [ 5 ],[ 6 ] in order to provide a more complete account of the discovery of membranes and their structure; Second, I will suggest that, contrary to the ideas favored in some articles, the discovery of biological membranes was not quite as a linear cumulative process as it has been generally depicted; And third, I expect that the acknowledgment of the importance of the cell boundary concept on modern conceptions of the cell will provide a fertile ground for discussions about the membranes in ancestral organisms. This, I hope, will open new perspectives for the stimulating field of the origins of life.

In the next sections, I will review the main discoveries that led to our current model of biological membranes: (1) the long path from the original assumptions about cell boundaries in the early Cell Theory to the first evidence that supported the existence of membranes; (2) early studies on cell membrane structure; (3) how evidence from permeability studies progressively built an alternative vision of cell boundaries–distinct from the model favored in the field of membrane studies; and (4) how the fluid mosaic model came into being. As shown in the sections below, many authors coming from different fields contributed to our understanding of membranes along the centuries. The reader will find a timeline summarizing the most dramatic contributions to membrane knowledge up to 1972 in Figures  2 and 3 .

Timeline 1665–1925. Summary of the main contributions related to cell membrane discovery between the coining of the term “cell” in biology and the first studies on cell membrane structures. The events are approximatively ordered from top to bottom from the earlier events to the most recent. Although the studies are sometimes difficult to classify, the colors of the boxes reflect some major research axes influent to this history: dark blue, doubts about the existence of cell membranes; orange, osmotic studies; red, studies with artificial membranes; purple, electrophysiology works; dark green, direct description of membranes. Although most of these contributions were highly interconnected, full lines between boxes highlight particularly important relationships and dashed lines point out to contradictory views in major controversies.

Timeline 1925–1972. Summary of the main contributions related to cell membrane discovery between the first studies on cell membrane structures and the formulation of the fluid mosaic model. The events are approximatively ordered from top to bottom from the earlier events to the most recent. Although the studies are sometimes difficult to classify, the colors of the boxes reflect some major research axes influent to this history: orange, osmotic studies; red, studies with artificial membranes; purple, electrophysiology works; dark green, direct description of membranes; pink, some transporter theories; light blue, asymmetric ion distribution debate; light green, electron microscopy studies. Although most of these contributions were highly interconnected, full lines between boxes highlight particularly important relationships and dashed lines point out to contradictory views in major controversies.

A time before the cell membrane

It is surprising to note that most short accounts of cell membrane discovery barely discuss the early controversies on the existence of membranes around the cells [ 5 ]-[ 10 ]. On the one hand, most papers dealing with the characterization of membranes assume that cell membranes were a corollary of the Cell Theory. On the other hand, many authors studying the Cell Theory consider the mere osmotic or permeability studies of the late XIX th century to be sufficient in extinguishing previous reluctances to the existence of membranes. In this section, we will see that the existence of plasma membranes seemed to be an unnecessary postulate for most of the XIX th century. Only at the turn of the XX th century, the existence of membranes became a convenient assumption for the study of most cell processes. The issue was not settled until molecular descriptions became more precise in the mid-XX th century.

The cell walls at the time of the cell theory proposal

The construction of the cell concept was a complex process that spanned the work of a large number of naturalists from the XVII th to the XIX th centuries [ 11 ],[ 12 ]. Since it is not my intention here to discuss the history of the Cell Theory, I will only deal with those authors whose work was particularly relevant for their conceptions of cell membranes (see dark blue boxes in Figure  2 ). Because many authors of this period wrote in German, the citations in the following sections will include both the primary and the secondary documents that I used to prepare this review.

In 1665, Hooke observed a piece of cork with his microscope and saw cavities that he compared to honey combs [ 13 ]. He named these cavities “cells”. This name is revealing because from the start it suggests the existence of some borders limiting an empty space. Yet, the cell boundaries that Hooke and his contemporaries could easily observe with their microscopes were the plant cell walls. Nowadays, we know that many cells can be surrounded by hard cell walls which are different from the universal cell membranes. However, since only cell walls could be easily observed at that time, early debates among microscopists focused on these structures for over 150 years.

During that period, two different conceptions of the microscopic observations competed with each other. On the one hand, some authors thought that the cell walls were continuous structures spanning the plant organism. Although he had first described the parenchyma of plants as a “mass of bubbles” [ 14 ], Grew was later the first to embrace the opinion that cell walls were made up of fibers woven together in a structure comparable to a textile fabric [ 15 ] –incidentally, giving rise to the introduction of the term “tissue” in biology [ 11 ]. This line of thought lasted until the early XIX th century; its last proponent was Mirbel, who assumed that the whole plant organism was made of a unique membranous structure (note here that the terms “membrane” and “cell wall” were indistinctly used at that time). From his point of view, the “cells” that were observed among the “membranes” were also thought to be parts of a continuous cavity [ 16 ]. To quote one of his opponents, Mirbel’s cells were like “the bubbles in the bread crumb” [ 17 ]. On the other hand, many authors, the first of whom was Malpighi, envisioned the cells not just as the space between the “membranes” but as discrete structures bounded by cell walls [ 11 ],[ 18 ]. The latter hypothesis was eventually accepted in the early XIX th century when Treviranus, Moldenhawer and Dutrochet managed to separate the cells from the plant tissue using different methods [ 11 ],[ 17 ],[ 19 ],[ 20 ]. Link’s demonstration that pigments from one cell did not pass into neighboring cells unless the cell walls were broken also contradicted Mirbel’s assumption that cavities formed a continuous compartment [ 11 ],[ 21 ]. By the first quarter of the XIX th century, plant cells were widely acknowledged as unconnected utricules bounded by separate cell walls [ 22 ]. Yet, the distinction between cell walls and cell membranes remained impossible.

The finding that plant cells could be separated from plant tissues contributed in shaping the increasingly popular idea that all organisms were made up of cells, namely the Cell Theory. Many biology manuals credit Schleiden and Schwann for the formulation of this theory. More thorough historical analyses actually show that the idea that cells were universal structures predated these authors and most of the features that we now recognize as cell-defining were discovered after Schleiden and Schwann [ 11 ],[ 12 ]. Nevertheless, Schleiden and Schwann’s contributions were highly influential because they were among the first to intrinsically relate the idea of the universality of cells to the universality of their multiplication and growth. Their point of view on cell development deserves specific attention from us because it impacted the way people thought about cell membranes for the rest of the XIX th century.

In 1837, Schleiden postulated a common development mechanism for all plant cells [ 23 ],[ 24 ]. Two years later, in 1839, Schwann enriched and extended Schleiden’s hypothesis to animal cells, thus suggesting that there was an universal mechanism for cell development [ 25 ],[ 26 ]. Their hypothesis was as follows (Figure  4 ): All living cells were made up of an amorphous substance called cytoblastema from which cells originated. The main difference between their respective hypotheses was that Schleiden thought that new cells always grew inside other cells, whereas Schwann acknowledged the possibility that cells could grow from any cytoblastema— whether internal or external. According to both authors, the first step for the formation of a new cell would have been the coagulation of a part of a preexisting cytoblastema into a nucleolus. The nucleolus would have acted as a nucleation center that would incorporate other molecules from the cytoblastema in a process similar to mineral crystallization. During growth, a differentiation process would have allowed the separation of the nucleus from the rest of the cell. Hardened membranes around the nucleus and the cell emerged as the result of the contact between two “phases”, i.e. the nucleus/cytoplasm or cytoplasm/environment, respectively. Although Schleiden did not discuss membranes much, Schwann considered them to be important structures responsible for separating the cell from its environment, and to be the place where “fermentation” (metabolism) took place. He assumed that membranes always limited the cells, even when they were invisible, and he suggested that the existence of membranes could be inferred from the internal Brownian movement of cell components, which did not cross the cell borders.

The development of cells according to Schleiden. This figure has been drawn for clarity from descriptions by Schleiden and Schwann, but these authors never tried to provide such a synthetic depiction in their work. Schwann’s model was very similar, except for his opinion that new cells could also crystallize from cytoblastema outside previous cells.

Despite the fact that Schleiden and Schwann’s models proved to be wrong, they acted as catalysts to foster hot debates about cell multiplication and organization that dominated histology for the rest of the century. Regarding the cell membranes, there are several points to keep in mind for the next section: First, even though Schwann assigned them essential roles, his conception of cell membranes as hardened interface structures is completely different from our current knowledge of the subject. Second, it must be recalled that, at the time, it was still impossible to make the distinction between the cell membranes and the cell wall; some people tried to look for cell walls in animal tissues and make comparisons between plant cell walls and other animal external structures, but their results were confusing [ 11 ]. Finally, many of his contemporaries called into question the assumption that, even when invisible, membranes always bounded the cells: this seemed to be theoretically unnecessary and hard to prove. As it will be developed in the next section, those who–correctly–recognized the lack of an animal equivalent to the plant cell wall predominantly assumed that membranes were not a mandatory characteristic of all cells.

Cells without membranes

The second half of the XIX th century was a period of many fascinating biological debates and discoveries related to the Darwinian evolution, the physiology of both animal and plants and, even in histology, the discovery of mitosis. Within this context, the cell boundaries received relatively limited attention. By the early 1890s, cell membranes were often thought to be unessential secondary structures [ 27 ].

Although some authors had already thought of the cell membranes as optional secondary structures [ 11 ],[ 28 ],[ 29 ], it seems that the first author who explicitly dismissed the existence of cell membranes was Leydig in 1857 [ 11 ],[ 30 ]. He based his opinion in the fact that membranes were not always observable and depicted the cell simply as a “substance primitively approaching a sphere in shape and containing a central body called a kernel [nucleus]” [ 11 ]. He recognized the existence of membranes as secondary structures resulting from the hardening of the cell surface. Later on, studies on amoebas reinforced the opinion that cell walls were not necessary characteristics of cells: De Bary, who understood the cell boundaries to be a solid structure like the cell walls, observed in Plasmodium several nuclei with no partition around each of them. He concluded that membranes may exist or not depending on different cell types [ 12 ],[ 31 ]. He also reasonably argued that the presence of a rigid cell wall would have prevented the protoplasm contraction that allowed the amoeboid movements in his model organisms. In a similar way, Haeckel agreed that bounding membranes were facultative in protists [ 12 ],[ 32 ].

In the second half of the XIX th century, the main opponent to the existence of membranes around cells was the protistologist Max Schultze [ 11 ],[ 12 ],[ 27 ],[ 33 ],[ 34 ]. This author described the cells as small lumps of contractile protoplasm that held together because of their inability to mix with water. In his opinion, membranes were only secondary structures resulting from the hardening of the cell surface. Their appearance was an artifact that marked the beginning of the degeneracy of protists. When membranes existed, they impeded the cell division and the internal protoplasmic movements, resulting in a loss of cellular activity. Beale accordingly viewed the appearance of membranes as a mark of natural degeneracy that differentiated the active, living protoplasm from the inactive, dead material produced by the cell [ 35 ]. In 1890, Turner published a review that explored the history and updates to the Cell Theory [ 27 ]. He described plasma membranes as being secondary structures and extended this idea to their intracellular counterparts based on the fact that the nuclear membrane disappeared during mitosis. In this context, it is not surprising that Schultze and later Sachs argued for the absurdity of the very term “cell” [ 11 ],[ 36 ],[ 37 ]. The original term coined by Hooke stressed the existence of the cell walls, whereas these authors acknowledged the protoplasm (the protoplast, according to Hanstein, [ 11 ],[ 38 ]) as the seat of biological activities.

Interestingly, although many authors from this period took for granted the absence of cell membranes, this was also the era of the first osmotic studies. Speculations about cell borders remained intact because the histological observations could not find a difference between specific cell membranes and the simple edge of the protoplasm.

Early osmotic studies and the cell boundaries

Osmosis studies (orange boxes in Figure  2 ) had an ambiguous relationship with the early understanding of cell membranes. From the earliest studies, water movement across semipermeable membranes was explicitly related to the volume changes of the cell. Osmosis can hardly be understood without the concept of membrane semipermeability and, as a result, osmotic studies have been relevant to theoretically acknowledge the cell membranes as selective barriers. Nevertheless, the first studies using artificial membranes were difficult to compare to the complexity of natural cell membranes and the analogy between the two types of membranes remained obscure for a long time.

In 1748, while he was trying to preserve some alcohol from the exposition to air, Nollet immersed a vial full of ethanol in a water container and covered it with a bladder membrane. After some hours, the bladder membrane had significantly swelled. Confronted to this observation, he carried out the opposite experiment. He put the water in the vial covered with the bladder membrane and the alcohol in the exterior container; the membrane sank. He concluded that the bladder membrane was permeable to water but not to ethanol [ 39 ]. Some years later, Hewson reported what could be considered as the first osmotic observations on living cells: He studied the shape of erythrocytes and noticed that these cells shrank or swelled depending on the salt concentration of the medium [ 40 ]. Although inspiring, these first reports went mostly unnoticed at first.

The importance of osmosis was not entirely recognized until Dutrochet’s work between 1826 and 1838. Dutrochet rediscovered the osmosis phenomena and carried out many experiments using different solutions and membranes that settled both the physical description of the phenomena and their physiological relevance. From his early works using animal membranes, he concluded that water moved from the compartment where the solutions were the less dense, acidic or positively charged to the compartment where the substances were more dense, alkaline or negatively charged [ 41 ]. He was influenced by previous work by Porret, who showed that an otherwise impermeable animal membrane could become water-permeable when an electric current was applied [ 42 ]. Dutrochet repeated Porret’s experiments and first thought that the water movement across the membrane could be somehow related to electricity [ 41 ]. His contemporaries also suggested that capillarity or viscosity differences between the solutions may account for the observed phenomena [ 43 ],[ 44 ]. However, a more systematic analysis allowed Dutrochet to discard all these hypotheses, including his own [ 45 ]. He concluded that the reason for the water movement was the heterogeneity of the liquids in the two compartments, but the underlying nature of the heterogeneity remained unknown to him [ 44 ]. It may be surprising to notice how hard it was for Dutrochet to explain the phenomena that he observed, but this has to be considered in its context: Diffusion was qualitatively described by Graham in the 1830s and Fick did not provide his quantitative equations for diffusion until 1855 [ 46 ],[ 47 ].

From the beginning of his experiments, Dutrochet extended his observations on osmosis to physiology [ 41 ]. He explained plant turgescence by the fact that plant cells used osmosis to accumulate water. His work on osmosis certainly influenced his opinion that cells were surrounded by essential cell membranes, though his way of describing them looks alien to us today. He suggested that cell borders acted as “chemical sieves”, which we would now describe as semipermeable, although he did not use that term. The chemical sieve-membrane would have been able to change the composition of the cell medium, resulting in the “secretion” (~metabolism) of substances to both the exterior and the interior of the cell [ 41 ].

Later osmotic studies and artificial membranes

In 1844, Von Mohl treated plant tissues with alcohol and different acids and described the detachment of the protoplasm from the interior of the cell walls. He named the shrinking vesicle that separated from the cell walls the “primordial utricule” [ 11 ],[ 48 ]-[ 50 ]. In addition to this chemical method, in the 1850s Nägeli and other authors put plant cells in hypertonic media and observed the contraction of a vesicle within the cell walls [ 11 ],[ 12 ],[ 51 ]-[ 54 ]. Over time, such osmotic studies became more quantitative, and by 1884, De Vries and Hamburguer among other authors were able to use plant and animal cell models to show that, except for electrolytes, most solutions applied equal osmotic pressures at equal concentrations [ 55 ],[ 56 ].

Although today these results may seem quite straightforward to analyze, at their time they were not decisive because the membranes remained invisible. It is unclear if Von Mohl thought that the primordial utricule was surrounded by an envelope or was a naked portion of protoplasm [ 11 ]. Nägeli believed that the semipermeable membrane resulted from the hardening of the exterior layer of protoplasm in contact with water, thus supporting the idea that membranes were not different from the rest of the cell [ 54 ]. Jacobs has argued that, even after their detailed osmotic studies, de Vries and Hamburguer did not assume that cells were necessarily bounded by membranes [ 57 ]; de Vries was actually aware of the fact that the osmotic phenomena he was measuring reflected volume changes in the massive plant vacuole, thus diverting the attention from membranes [ 11 ],[ 55 ]. Finally, the opponents to cell membranes were comforted by Nägeli’s experiments on the idea that the naked protoplasm was the active component of the cell, whereas the rigid cell wall was an unessential secondary element that could even be removed from the cell [ 27 ].

Two concepts are important here to understand how the cell was portrayed in the late XIX th century and early XX th century: the colloid and the precipitation membranes (red boxes in Figure  2 ). In 1861, Graham separated the water-soluble molecules into two sorts according to their ability to cross a parchment paper: Inorganic salts and sugars easily crossed the membrane and were called crystalloids, whereas gelatin-like compounds were unable to do so and were named colloids [ 58 ]. Biologists rapidly adopted the term “colloid” to refer to the structure of the protoplasm, probably because it seemed more precise in describing the viscous jelly-like interior of the cell. The open question then was to determine if the surface of the cell was made up of the same substance as the rest of the protoplasm. A possible answer to that question was provided by studies on precipitation membranes.

Precipitation membranes were the first artificial membranes to be synthesized. They were first developed by Traube in 1867 and so named because they were obtained by the precipitation of molecules at the interface between one solution of potassium ferrocyanide and another of copper sulfate [ 59 ],[ 60 ]. These membranes were permeable to water but not to other molecules, thus becoming important tools for osmotic studies. For instance, Pfeffer was able to produce sturdier precipitation membranes in 1877 and carried out several experiments that established the correlation between the osmotic pressure and the solution concentration and temperature [ 60 ],[ 61 ]. Precipitation membranes were highly influential in the debate about the existence of cell membranes because most authors, including those working on osmosis, assimilated the cell membranes to a precipitation membrane. According to that point of view, the protoplasmic colloid precipitated when it was in contact with the aqueous medium, but this did not require the membrane to be any different from the rest of the protoplasm (Figure  5 ).

XIX th century doubts about the existence of membranes. A . In this vision, the cell is devoid of any membrane and all the properties of the cell are defined by the activity of the protoplasmic colloid. B . The cell is surrounded by an external layer (membrane) of which the nature is distinct to the rest of the protoplasm. Yet, in this view, the inside of the cell remains a colloid.

To summarize, despite the important developments in osmosis that were taking place during the second half of the XIX th century and the intrinsic necessity of membranes to explain this phenomenon, it would be misleading to think that cell membranes were considered mandatory cell structures at the time. Even those who acknowledged the fact that osmosis requires semipermeable membranes to take place, envisioned cell borders as precipitation membranes at the surface of a protoplasmic colloid–a point of view irreconcilable with our current understanding of cells.

Late membrane-less hypotheses

As it has been shown so far, during most of the XIX th century, cell membranes attracted limited attention and the dominant opinion was that cell membranes (often mistaken for cell walls) were secondary structures that resulted from the contact between the protoplasmic colloid and the environment. This vision of things changed at the turn of the XX th century but marginally remained until well into that century. In this regard, Fischer’s work in 1921 is noteworthy because he called into question the concept of cell membranes at a time when their existence was generally taken for granted but direct evidence remained scarce. His arguments, which can be considered as the culmination of the XIX th century point of view on membranes, were the following [ 62 ]: 1) Cell membranes were invisible using optical microscopy; even when the edge of the cell was visible, it did not prove the existence of membranes with characteristics different from the rest of the protoplasm; 2) When cells were immersed in a hypertonic medium, they shrank less than would be expected from strict osmotic criteria; 3) Cell permeability to different molecules seemed to change depending on many factors—an observation seemingly incompatible with contemporary descriptions of the membrane as sieves or apolar solvent layers; 4) Fischer claimed that cell fragments behaved similarly with solutes than whole cells, although he did not provide a precise account of the experiments that made him say so; and 5) Since he assumed that the interior of the cell was a colloid, he argued that cell membrane models still had to explain how the molecules moved within the colloid (the cell) once they had crossed the membrane. As a result of all these criticisms, he concluded that the external layer of the protoplasm could only be conceived as a “surface tension film” made up of the same compounds as the protoplasm and lacking any osmotic value.

Fischer’s opinions were certainly not dominant at the time, but the colloid hypothesis did not die out until progress in enzymology and molecular biology replaced the homogeneous, gelatin-like description of the protoplasm by studying discrete compounds of life. The existence of membranes was vindicated as late as the 1950s and there are even some authors who still call it into question [ 4 ],[ 63 ].

The birth of cell membranes

The existence of cell membranes did not become popular until the turn of the XX th century. In this section I will present the different pieces of evidence that led to the general acknowledgement of the existence of cell membranes in its modern sense.

The permeability of molecules according to their polarity

In the 1890s, two confronting views competed each other to explain how semipermeable membranes operated: Traube had suggested that precipitation membranes had small pores that allowed them to behave like sieves, whereas Nernst introduced the idea that permeating substances were those that could dissolve in the membranes [ 59 ],[ 64 ],[ 65 ]. From 1895 to 1899, Overton carried out a series of experiments in which he immersed cells in solutions of over 500 different substances at the same concentration in order to study their permeability with different molecules [ 66 ]-[ 68 ]. He noticed that solutions of ether-soluble (apolar) molecules did not result in the shrinking of cells, contrary to solutions of water-soluble (polar) substances. He concluded that apolar molecules entered the cells with less difficulty than polar substances, and he showed that this was irrespective of their molecular size. Since solubility, not molecular size, was the best predictor of the entry of substances in the cell, Overton favored Nernst’s hypothesis for membrane permeability [ 68 ]. Based on the observation that not all molecules could enter the cells with the same ease, but also aware that the cellulose cell wall could not be involved in the phenomenon, he suggested that there was a cell membrane distinct from the cellulose cell wall [ 11 ],[ 66 ] and that these cell membranes were made up of ether-soluble components [ 67 ],[ 68 ]. Looking for specific polar candidates that could make up the membranes, he ruled out the triglycerides because they would be subject to saponification in the regular living conditions of cells. He suggested that cholesterol and phospholipids could be the main components of cell membranes even though little was known at the time about the cellular functions of these molecules. He also recognized the difficulty that the membrane solubility theory may introduce in explaining the movement of water and other hydrophilic substances across the cell boundaries. He tried to solve this paradox by recalling that, in spite of their hydrophobic nature, cholesterol esters and cholesterol-lecithin mixtures were known to absorb large water volumes [ 68 ]. He also suggested that some kind of active property of the protoplasm could allow the active transport of molecules into the cell [ 66 ],[ 69 ]. Similar, lesser-known observations were also reported in bacteria and analyzed in a similar way [ 70 ],[ 71 ]. Yet, Overton remained the authoritative figure to which most future cell membrane works would refer (dark green boxes in Figure  2 ).

The permeability to dyes

In 1855, Nägeli had already made some interesting observations about the permeability of dyes in the plant cell [ 52 ],[ 54 ]. First, he noticed that when plant vacuoles were filled with a pigmented solution, the osmotic changes could modify the volume of the vacuole but the pigment did not leak outside the vacuole unless it was artificially damaged. He also showed that when plant cells were immersed in hypertonic colored solutions, the protoplasm shrank and the colored solutions could be observed in the space between the protoplasm and the cell wall, but the pigments did not enter the protoplasm. As a result of both observations, he concluded that the vacuole boundaries and the protoplasmic surface were barriers to penetration [ 52 ],[ 54 ]. Still, as we have seen previously, Nägeli did not think that the cell was bounded by a differentiated membrane so he assumed that the resistance to pigmentation was a general characteristic of the whole protoplasm rather than the consequence of the activity from a specific part of the cell. Overton revived the interest for dyes and, in accordance with his previous work, visually confirmed that lipid-soluble dyes entered the cells more easily than water-soluble dyes [ 11 ],[ 72 ]. Yet, these observations did not definitively prove that cell membranes were chemically different from the rest of the protoplasm. It was not until 1922 that the improvement in the microinjection techniques provided a crucial answer to this issue. Chambers used this technique to apply a water-soluble cytolysogenic (i.e. able to digest the cytoplasm) solution in different parts of the cell. He showed that he could apply the hydrophilic cytolysogenic substance on the surface of starfish eggs without damaging them. Then, he injected a small amount in the interior of the cell and he observed the cytolysis of the protoplasm. When the injection was made close to the cell borders, the cytolysis spread in the protoplasm but did not impact the membrane until the rest of the cell had been massively damaged [ 73 ]. This was the first unavoidable evidence that the nature of the cell surface was different from the rest of the protoplasm, supporting the existence of the cell membrane.

The electrophysiology of excitable cells

The electrophysiology is a domain unto itself, so here I will only briefly summarize the main indirect contributions to the field of cell membranes at the turn of the XX th century (purple boxes in Figure  2 ). There are two components in this story. The first is the primal electrophysiology debate incarnated in the opposition between du Bois-Reymond and Hermann [ 74 ]: In 1848, the former had reported an electric current and action potential in muscles and nerves and tried to explain them by the preexisting charge differences between the interior and the exterior of the tissues [ 74 ],[ 75 ]; In 1867, the latter assumed that the currents measured by du Bois-Reymond were an artifact and that the electrolytes found in the external medium resulted from the chemical decomposition of the samples [ 74 ],[ 76 ]. The second element of this story is the field of electrochemistry, which was emerging very fast thanks to the development of more precise mathematical models. It began when de Vries attracted Von’t Hoff’s attention to the problem of osmotic pressure. Von’t Hoff first used the osmotic measurements from Pfeffer, De Vries, Hamburguer, Donders and Raoult to suggest the theoretical and experimental equivalence between the laws of ideal gases and those that ruled the behavior of dilute solutions [ 77 ],[ 78 ]. The main exception that did not seem to fit into Von’t Hoff’s hypothesis were electrolytes. Arrhenius contacted Von’t Hoff in a personal communication and put him on the track to explain the observed anomalies based on the dissociation hypothesis. Eventually, Nernst, who was studying the relationship between electricity and electrolyte movements, developed Von’t Hoff’s equations for the calculation of the electric potential and electromotive force in galvanic cells [ 79 ],[ 80 ].

These two lines of research met through Bernstein’s work in 1902. Bernstein used a physiological model to corroborate some of the electrochemical predictions made by Nernst. In particular, he showed that temperature changes impacted the electromotive forces in muscles according to Nernst’s predictions [ 81 ],[ 82 ]. The reconciliation of electrophysiology with electrochemistry allowed Bernstein to formulate the “membrane theory of electrical potentials”. Bernstein’s membrane theory postulated that (1) nerves consisted of a conducting electrolyte bounded by thin membranes impermeable to ions; (2) in the resting state, the membrane kept an electric potential with internal negative charges and external positive charges; and (3) in the activity period, potassium permeability increased and the electric potential consequently dropped [ 81 ],[ 82 ].

In the 1910s, Höber carried out a series of experiments that corroborated Bernstein’s theory and coincidentally provided supplementary evidence in favor of the existence of cell membranes. Höber showed that conductivities of muscle or compacted erythrocytes were higher at high electrical frequencies than at low frequencies. He suggested that membranes were impermeable at low electrical frequencies but became less resistant at high frequencies because the cells themselves were disrupted. In order to test this hypothesis, Höber measured the internal conductivity of the cells. The values that he obtained were not compatible with the attachment of electrolytes to the protoplasmic colloid, but they supported their solution in the internal medium. The corollary of these results was that the only way to prevent the electrolytes from diffusing out of the cell was to present an impermeable boundary with properties different from the rest of the protoplasm [ 83 ]-[ 86 ].

The existence of cell membranes

In short, in the early XX th century the presence of membranes was becoming widely accepted and was supported by three lines of evidence: Overton’s permeability studies, Chamber’s microinjection experiments and Höber’s electrical measures. Nonetheless, it should be noted that for some organisms, especially bacteria, this debate was not definitively closed until several decades later, when cell membranes could be directly observed using the electron microscope [ 87 ]-[ 89 ].

The first membrane structures

In the previous section, I have shown that many authors in the XIX th century thought that membranes were not essential parts of cells and even those who recognized their importance did not conceive membranes as we do today. By the turn of the XX th century, cell membranes had become a convenient assumption supported by a few direct experiments. The next decades would witness an increasing interest in describing membrane structure (dark green boxes in Figure  3 ).

Gorter and Grendel: a relative breakthrough

In the early XX th century it appeared clear that, if cell membranes existed, they would likely be at least partially lipid-based. The opinion that the cell surface could be covered by a thin lipid layer goes back to the 1880s [ 65 ],[ 90 ] but it was not popularized until Overton’s publications in 1895–1899 (see previous section). The molecular structure of membranes remained unexplored until the major breakthrough made by Gorter and Grendel in 1925.

The genius of this paper was to compare the surface that cell lipids were able to occupy to the total surface of cells. Gorter and Grendel extracted the lipids from an erythrocyte sample; since these cells were known to lack internal membranes, they assumed that all lipids should come from the cell envelopes. Measuring the spread of lipids on water was done using a Langmuir’s trough (Figure  6 ). This device had first been developed by Pockels as a way to precisely measure the surface covered by lipid monolayers at the interface between water and air [ 91 ], but it was named after Langmuir’s version more than 25 years later [ 92 ]. When Gorter and Grendel compared the surface covered by the lipids to the estimated sum of cell surfaces, they found a 2:1 ratio (Figure  7 ). As a result, they concluded that cells were surrounded by a lipid membrane two molecules thick–a lipid bilayer–with the hydrophobic components in the internal part of the membrane and the hydrophilic components in the external part [ 93 ].

Oil at air/water interfaces. A . Oil molecules spontaneously spread on the air/water interface until they form a layer one molecule thick. B . The Langmuir trough allows to precisely measure the surface that these monolayers can spread depending on the applied pressure.

Surface measurement of membrane lipid monolayers as a way to determine membrane structure. A . Summary of the method, consisting in the comparison between the surface occupied by lipids extracted from membranes and the estimated surface of cells B . Different results and interpretations.

This study has been commonly cited as the most conclusive argument in favor of the lipid bilayer nature of cell membranes. In spite of the elegance of this work, it is important to balance its contribution to the field because it is subject to criticisms from both technical and theoretical grounds. First, the extraction technique employed could not isolate the totality of erythrocyte lipids from the samples [ 94 ]. In addition, the equation that Gorter and Grendel used to estimate the surface of the erythrocytes also underestimated the cell area [ 94 ]. Some historical reviews on membrane discovery have argued that it was fortunate that the two errors neutralized each other in order to give credit to the lipid bilayer hypothesis that we now recognize in current membrane models [ 5 ]. But it should be recalled that the lipid bilayer hypothesis of 1925 did not leave room for anything else than lipids to be located in the membrane plane–in contrast to past and current mosaic hypotheses. Some parallel reports even contradicted Gorter and Grendel’s values (Figure  7 ). Dervichian and Macheboeuf carried out a similar analysis and obtained a 1:1 ratio; as a result, they assumed that the cell membrane was a lipid monolayer [ 95 ]. Although this second work also had lipid extraction problems, the major difference between the two studies was that Gorter and Grendel measured the surface covered by lipids at the first detected pressure (i.e. the maximal continuous surface covered by an amount of lipid) whereas Dervichian and Machebouef measured the surface right before the collapse pressure (i.e. the minimal surface before the monolayer collapsed). In the 1960s, a more accurate ratio for the lipid surface at the collapse pressure with respect to the cell surface estimates was calculated to be 1,3:1 [ 94 ]; the authors of this later study suggested that their ratio conformed to loose lipid bilayers, whereas the modern interpretation of the fluid mosaic model supports the idea that membrane lipids are tightly packed and the “excess space” is actually occupied by membrane proteins.

Apart from the technical issues, it is worth noting the theoretical bases on which Gorter and Grendel founded their lipid bilayer concept. Some years earlier, in the 1910s, some pioneer papers had studied the behavior of amphiphilic molecules at the interface between water and air [ 92 ],[ 96 ],[ 97 ]. For instance, in 1917 Langmuir used the valence bond theory suggested by Lewis the previous year [ 98 ] to explain molecular hydrophily on the grounds of “secondary valence”, i.e. chemical polarity [ 92 ]. Langmuir put forward this hypothesis in the same paper in which he presented the trough that, somewhat unfairly, was named after him. Notwithstanding, Gorter and Grendel only cited Langmuir’s paper in a very superficial way related to the trough [ 93 ]. Instead of reasoning in terms of hydrophily and hydrophobicity, they suggested the bilayer structure based on crystallographic studies [ 99 ] and soap bubble observations, which were only distantly related to their subject [ 100 ]. Gorter and Grendel are not really to blame because hydrophobic interactions were very poorly understood at the time. Subsequent authors, like Danielli, discussed the importance of the amphipathic nature of lipids and proteins to account for their respective structural hypotheses, but they disregarded the importance of hydrophobic interactions [ 101 ]. Even Langmuir, who extended his explanation of amphipatic molecules to protein monolayers [ 102 ], overlooked the hydrophobic interactions between proteins and lipids [ 103 ] when he came to envision the cell membrane.

In summary, although Gorter and Grendel’s work was decisive for making the lipid bilayer concept popular in 1925, its actual contribution to current membrane models can only be appreciated in the light of later progress in membrane studies and hydrophobic interaction understanding [ 6 ].

First direct studies on membrane structure

The formulation of the lipid bilayer hypothesis had opened the door to the molecular description of cell membrane structure. In an attempt to confirm or refute the lipid bilayer postulate, one of the first lines of research to be explored was the measurement of membrane thickness. The first attempt to estimate the thickness of cell membranes was directly related to Höber’s research on cell conductivity (see above). In 1925, Fricke measured the static capacitance per surface unit with an estimation of the cell membrane dielectric constant. He extrapolated that the thickness of erythrocyte and yeast cell membranes was in a range between 3.3 and 4 nm [ 104 ],[ 105 ]. This thickness was compatible with a lipid monolayer but not with a bilayer, thus providing support to the monolayer membrane proponents [ 95 ]. The choice of the dielectric constant used in these studies was called into question but the subsequent tests could not refute the estimation by Fricke [ 106 ]. Independently, the leptoscope was invented in order to measure very thin membranes by comparing the intensity of light reflected from a sample to the intensity of a membrane standard of known thickness [ 107 ]. This device measured thicknesses that depended on pH and the presence of membrane proteins and ranged from 8.6 to 23.2 nm. As a result, the lower values supported the lipid bilayer hypothesis whereas the higher ones could support the presence of supplementary superimposed layers [ 108 ]. The thickness of the membrane would only become really accessible two decades later, when the observation of membrane sections using electron microscopy established the now accepted ~8 nm value for standard cell membranes [ 109 ].

The 1930s were important in this field because the decade introduced the most influential membrane structure model until the general agreement on the fluid mosaic model, namely the paucimolecular model (Figure  8 ). The genesis of this model relied on studies of surface tension between oils and echinoderm/teleostei eggs [ 110 ]-[ 112 ]. As the surface tension values appeared to be much lower than would be expected for an oil–water interface, it was assumed that some substance was responsible for lowering the interfacial tensions in the surface of cells [ 112 ]. At this time, membranes were known to contain substantial quantities of proteins [ 113 ],[ 114 ] but little had been said about their position in cell membranes. Therefore, in 1935 Danielli and Davson suggested that the lipid bilayer was sandwiched between two thin protein layers [ 115 ]. Although the paucimolecular model was characterized by the superposition of protein and lipid layers (Figure  8 ), the authors were aware of the contemporary debates about membrane permeability (discussed later), so they admitted the possibility that some proteins could span the membrane.

Membrane structure hypotheses in the 1930’s. A . Paucimolecular model, with a lipid bilayer coated with proteins in both sides. B . Höber’s mosaic model in which membranes behaved both as solvents and sieves.

The paucimolecular model immediately became popular and it dominated cell membrane studies for the following 30 years. From the beginning, it was confronted with a balance of supportive and critical observations. Among the so-called supportive evidence, there were some light polarization and X-ray diffraction studies. In order to provide insights on membrane structure, these methods required repeated structures, so the samples used were myelinized axons. These analyses showed an alternation of protein and lipid layers in support of the paucimolecular model [ 116 ]-[ 118 ]. Unfortunately, we know now that this biological material is physiologically very specific and can hardly be compared to a regular cell membrane.

The hypotheses that challenged the paucimolecular model will be further developed below, but first I will present some details regarding the aforementioned studies on surface tension. The surface tension experiments that led to the paucimolecular model used triglyceride oils and other non-miscible lipids. These substances are appreciably different from most natural amphipathic cell membrane components and, therefore, they are not suitable for a realistic comparison to cell membranes. Danielli and other authors showed soon after the postulation of the paucimolecular model that the addition of amphipatic cell membrane components like fatty acids, cholesterol or phospholipids to non-miscible mixtures was very effective in dropping the interfacial tension between water and highly hydrophobic substances [ 119 ],[ 120 ]. As a result, shortly after the suggestion of the paucimolecular model, the main argument that motivated it, namely the requirement of proteins to lower the superficial tension of the cell, had been dismissed. Yet, the hypothesis remained popular for 30 more years.

Contemporary competitors: the mosaic models

Despite of the prevalence of the paucimolecular model in the mid-XX th century, this hypothesis was not devoid of competitors. Direct experiments on cell surfaces were scarce but permeability studies provided an outstanding playground for indirect speculations about cell membrane structure. Permeation studies suggested what would to be called the “mosaic models”.

As it was previously pointed out, Overton’s hypothesis that apolar molecules easily entered the cell because they could get dissolved in the cell lipid membranes automatically raised the problem of how polar molecules accessed it. In an attempt to circumvent this issue, Nathansohn suggested in 1904 that the cell surface could be a mosaic combining fat-like parts with protoplasmic-like parts [ 62 ],[ 121 ]. In the early XX th century, the “mosaic” term was recycled to refer to membranes with heterogeneous parts [ 122 ]-[ 124 ]. In the 1930s Höber amended it to fit the idea that membranes were a mixture of sieve-like and solvent (i.e. lipid) parts [ 125 ]-[ 128 ]. These mosaic hypotheses were the result of the combination of the two classic ways of understanding the permeation through membranes: Traube’s precipitation membranes and Nernst/Overton lipid membranes. In the first case the molecular and pore sizes were predicted to best account for molecule transport; in the second, the hydrophobicity was the best predictor of molecular permeability. The compromise reached described the cell membranes as lipid layers interrupted by pores (Figure  8 , [ 126 ]). Since the permeability of some polar molecules had been shown to change according to different conditions [ 129 ],[ 130 ], it was also suggested that the pore diameter could change according to the hydration of the pore, the pH, its obstruction by some particular molecule, the membrane stretching, the metabolic activity and the cell type [ 126 ]. Most descriptions did not specify the type of molecules that could form these pores, but proteins were among the best candidates [ 101 ],[ 131 ].

Apart from the mosaic hypotheses proponents, other authors also called into question the paucimolecular model. For example, in their presentation to the very influential 1940 symposium on the permeability of cell membranes, Parpart and Dziemian reported the chemical composition that could be analyzed from cell extracts [ 132 ]. Although these authors did not specifically support the mosaic models, they noted that lipases in contact with cells modified the cell permeability, which suggested that surface phospholipids were naked instead of coated with proteins. The discussion that followed their talk is remarkable because several authors exposed their visions for membrane protein structure: Ponder imagined the proteins adopted a spaghetti-like shape on top of the membrane whereas Davson suggested that proteins could span the whole erythrocyte, not just the membrane.

In 1936, Danielli, who was probably the most influential author in the field at the time, discussed a complete catalog of possible membrane structures in addition to his paucimolecular model [ 101 ]. He excluded all membrane models that were much thicker than 8 nm because he thought it was the most plausible cell membrane thickness. He classified the membrane models in three types: continuous lipid membranes, mosaic membranes and lipo-protein membranes (Figure  9 ). In the first type, he imagined all the possible combinations of lipid monolayers and bilayers coated with proteins. He concluded that a lipid bilayer with the polar parts of the lipids in the exterior would be the most stable structure because it maximized the contact of the hydrophilic lipid parts with water. He assumed that the proteins were subject to the same amphipatic constraints as lipids but in their own layers. In the mosaic-like models, he considered different distributions of proteins and lipids, but he ruled all of them out because he assumed that lateral interactions between lipid hydrophobic parts and proteins would not have been stable. He also considered that if the lipid bilayer had not been covered by proteins, it would not have been solid enough to provide a reliable impermeable barrier to resist to cell deformation. Finally, he did not go into the detail of the lipoprotein membranes because little was known about such kind of molecules. Concerning ion permeability, he acknowledged the three popular possibilities of his time: pores, simple diffusion and the existence of some kind of transporter in the membrane [ 133 ].

Possible molecular arrangements of biological membranes redrawn from Danielli in 1936 [ [ 101 ] ] . A-E . Cross-section of hypothetical membranes with internal lipids and coating proteins. F . Cross-section of an hypothetical membrane made up of lipoprotein subunits. G-I . Surface of mosaic membranes.

In summary, many hypotheses on cell membrane structure were under discussion in the late 1930s. Among these, the paucimolecular and the mosaic models were certainly the most iconic. The next developments would be highly influenced by the evidence from independent, though membrane-related, fields.

Insights from the movements across the membrane

The direct characterization of membrane structure did not progress much until new techniques allowed dramatic discoveries in the late 1950s and 1960s. Before these new methods became available, cell membranes attracted the attention from authors who were studying membrane roles in physiology and metabolism. Although these contributions were not always immediately recognized by the community working on membrane studies, it is important to take them into account because they illustrate how works of this period (1940s-1950s) indirectly changed membrane understanding.

Most of the studies presented in this section are related to molecular transport across membranes and did not directly address the question of membrane structure. Therefore, I will not go into detail for discoveries in these fields but I will just provide some general clues to illustrate how the contextual research impacted cell membrane conceptions. For more complete accounts on the history of transport, excitability and membrane metabolism I invite the interested reader to refer to the insightful works by Robinson and Kleinzeller and colleagues [ 134 ],[ 135 ].

Asymmetric ion distributions

In the early XX th century, Na + , K + and other ion concentrations were already known to be different between the interior of the cell and their environment [ 136 ],[ 137 ]. Three classic hypotheses competed to explain these observations: (1) some ions were stably bound to the cell colloid; (2) the membranes were totally impermeable for some ions; and (3) ion concentrations were kept at the expense of an energy-consuming transport [ 134 ],[ 138 ].

The first hypothesis is by definition related to the attachment or dissolution of molecules in the protoplasmic colloid. As we have seen previously, the electric measures carried out by Höber and Fricke suggested that the cell internal medium could be compared to a conductor solution of free electrolytes [ 85 ],[ 86 ],[ 104 ]. References to the so-called interactions between ions and the colloid lasted for some decades [ 139 ],[ 140 ], but the colloid concept progressively became outdated with the developments of enzymology and molecular biology [ 63 ]. As a result, ion asymmetry was mainly debated as an opposition between the two other possibilities: membrane impermeability or energy consumption (light blue boxes in Figure  3 ).

In 1910, Donnan had shown that, provided a membrane was permeable to some ions and impermeable to others, an ion distribution asymmetry was expected to spontaneously arise according to the second law of thermodynamics (Figure  10 , [ 141 ]). In 1941, Conway and Boyle suggested that the membrane impermeability to negatively-charged proteins and one or several electrolytes could generate a complicated Donnan effect that would account for the observed ion asymmetry [ 142 ]. Nevertheless, the simultaneous accumulation of evidence–especially from diet studies and radioactive cation labeling–supported the hypothesis that physiological ions could effectively cross the membranes [ 143 ]-[ 147 ]. Moreover, the increased interest in blood conservation during the Second World War demonstrated that K + loss from erythrocytes was related to the slowdown of metabolism [ 148 ],[ 149 ]. Finally, cation fluxes seemed able to restore ion asymmetries in the recovery period after muscle stimulation [ 150 ].

Donnan’s equilibrium. Two solutions containing two different initial concentrations of different salts are separated by a membrane. In this case, the membrane is impermeable to anions but permeable to cations. Donnan thermodynamic calculations and experiments showed that, contrary to what could be initially thought, the two cations do not just interchange with each other until they are equally distributed in the two compartments. Instead, equivalent quantities of both cations cross the membrane; as their initial concentrations are different, the cation which was initially less concentrated proportionally crosses the membrane more than the initially highly concentrated cation.

This evidence led Dean to postulate in 1941 that the agent of ion movement against the gradient could result from the activity of some “pump” located in the muscle fiber membrane [ 151 ]. The Croonian lecture by Krogh in 1945 was also an influential opinion in favor of the hypothesis that Na + /K + asymmetries were the result of active transport [ 152 ]. Conway replied to this hypothesis by attacking some aspects of the experimental designs and interpretations of his opponents (see [ 134 ] for details), but his most commented objection was that his own calculations on the energy necessary to extrude all the Na + from the muscle was larger than the actual available energy in the resting muscle [ 153 ]. He admitted that some small amount of Na + transport could be possible, but in derisory proportions. Nonetheless, the impermeability hypothesis was progressively abandoned when new works measured the outflow of radioactive cations more precisely and calculated a more reasonable amount of energy to account for the active cation transport–the active transport hypothesis became predominant [ 154 ]-[ 157 ].

The idea that cell membranes hosted important metabolism-related carriers or transporters was not totally new (pink boxes in Figure  3 ). Already in the 1930s, ions had been suggested to cross the membranes thanks to the interaction with some membrane components (Figure  11  F, [ 158 ] Ion transport had also tentatively been related to metabolism and respiration [ 159 ]-[ 161 ]. In 1947, one early expression of this idea suggested that the membrane fixed the ions to the cell, where the internal respiration was responsible for exchanging the external ions with some other internal ions linked to the cell colloid through the so-called “ion tracks” [ 128 ]; it is worth noting that this hypothesis still favored the continuous nature of the cell colloid instead of the diffusion in the interior of the cell (Figure  5 ). Another transport suggestion was that the membrane had components to which ions could be linked and which could change their conformation to allow the ions crossing the membrane [ 162 ]. It was soon suggested that proteins may be the agents of this ion transport [ 163 ], although this option was not immediately accepted [ 164 ].

Redrawings of some examples of transport across membranes in the 1950’s and 1960’s. A . Eccles depicts in 1963 the coupling between a metabolic-driven ion pump and several different channels [ 165 ]. B . Burgen suggests in 1957 that molecules cross pores thanks to specific and dynamic interactions with them [ 166 ]. C . Mitchell describes in 1957 an enzymatic-like protein transporter embedded in the membrane [ 167 ]. D . Mitchell’s chemiosmotic hypothesis in 1961 is based in the existence of structures embedded in the biological membranes [ 168 ]. E . From an early date, Danielli and collaborators considered the possibility that channels may have existed within their paucimolecular hypothesis (redrawn from [ 169 ]). F . Danielli’s summary of different transporter models in 1954 (redrawn from [ 134 ]).

Still, in the early 1950s, very little was known about the active transporters themselves. Important progress came from studies on excitable membranes and the transport of non-electrolytes.

Excitable membranes

Bernstein’s membrane theory explained action potentials by assuming that transient membrane breakdowns would be responsible for an increased permeability and the subsequent abolishment of the ion gradient ([ 81 ],[ 170 ], purple boxes in Figures  2 and 3 ). For some decades, the membrane breakdown was a popular mechanism to account for ion crossing even among those authors who thought that membranes had pores [ 139 ],[ 142 ]. Of course, according to this hypothesis, the action potential could not possibly surpass the resting potential. Yet, when direct measures between an internal and an external electrode became possible in 1939, it appeared that the action potential was larger than the resting potential [ 171 ]: A mechanism different from the membrane breakdown was necessary to explain these observations. In 1952, after unsuccessful attempts to explain the overshoot based only on K + permeability and membrane changes, Hodgkin and Huxley published in a series of papers that established the sequence of Na + inflow and K + outflow responsible for depolarization and hyperpolarization in axon membranes [ 172 ],[ 173 ].

In his lecture for the Nobel prize, Hodgkin highlighted the fact that, in spite of their radical discovery, little was still known about how these ions flows took place. Since pores were already assumed to exist for the downhill movement of K + across the membrane, this issue drew relatively little attention. On the contrary, since hydrated Na + is larger than hydrated K + , how could a pore specifically select for Na + flow and simultaneously avoid K + movement? It was first suggested that Na + used a specific lipid carrier that could cross the membrane when it was depolarized [ 174 ], but this track was ruled out because it was contradicted by the reported kinetics of Na + movement [ 172 ]. Another related issue also attracted much attention: How could the ions cross the membrane against their gradient after the action potential to recover the values observed at the resting state? This question had already been asked by Overton 50 years before [ 136 ],[ 170 ] and the proceeding answers had been the same as those trying to explain ion distribution asymmetry (see previous section). For instance, Conway suggested that his mechanism based in the Donnan equilibrium could also explain the resting potentials in the nerve [ 142 ], but attention progressively moved to active transport as this hypothesis became dominant [ 157 ].

Excitable tissues became one of the favorite models for the study of active transport. On one hand, Na + or Na + /K + transport in axons was shown to depend on glucose availability and to be affected by several inhibitors of oxydative phosphorylation [ 175 ]-[ 177 ] On the other hand, some preliminary reports pointed out to the existence of an ATPase activity located in the lipid fraction of axons [ 178 ],[ 179 ]. In 1957, Skou found a connection between the ATPase and the ion-dependent activity present in the hydrophobic fraction–membrane–of the nerve (for which he also earned a Nobel prize) [ 180 ]. This observation launched an overwhelming interest on ATPases that led to a fast accumulation of data covering ATPases from different species, tissues and functions [ 181 ]. By 1965, some ATPase features relevant in understanding membrane structure had become conventional wisdom: (1) the active transporters were located in the cell membrane (Figure  11 A); (2) they spanned the membrane and were asymmetric, i.e. they showed different affinities in each membrane side; and (3) as they enzymatically hydrolyzed ATP, they were an intrinsic part of cell metabolism.

As we will see next, this progress in transport understanding was paralleled by simultaneous studies on non-electrolyte uptake. The combination of all these different studies reinforced the notion that membrane proteins were enzymes strongly related to metabolism and cell bioenergetics.

Membrane-located metabolism and enzyme-like transporters

How non-electrolytes entered the cell was also a matter of speculation for a long time [ 182 ]. Initially, if non-electrolyte molecules had crossed the membranes either through their lipid component or through a putative pore, regular diffusion should have been sufficient in predicting their permeability rates. On the contrary, several puzzling observations started to accumulate in the first quarter of the XX th century. For instance, it was shown that the intestine absorbed some sugars more easily than others, even when stereoisomers were compared [ 182 ]-[ 185 ], and that some sugars entered the erythrocytes faster when the external concentrations were lower [ 182 ],[ 186 ]. In 1935, Jacques’ precise kinetic analyses showed that permeation was a saturable process ([ 187 ] pink boxes in Figure  3 ). This result advanced the involvement of transporters in non-electrolyte permeation [ 164 ]. In the 1950s, a revitalized interest in transporter kinetics revealed that some molecules acted as transport inhibitors [ 182 ],[ 188 ] and that transporters were regulated by their own substrates [ 189 ]. These results bolstered the connection between transporters and enzymes not only in eukaryotes but also in bacteria [ 63 ],[ 71 ],[ 166 ],[ 189 ]-[ 191 ].

As a result of the accumulation of kinetic, genetic and energetic data, new transporter hypotheses emerged in the 1950s. The classical view was that the transporter was a molecule present in equivalent amounts on both sides of the membrane and able to simply shuttle the attached molecule from one side to the other [ 162 ],[ 191 ]. Figure  11 summarizes the many new transporter modalities envisioned in the 1950s and 1960s [ 134 ],[ 192 ]: (1) a mechanical small transporter propelled from one side of the membrane to the other [ 193 ]; (2) a membrane-spanning carrier able to flip-flop; (3) a division of the membrane into rotating segments; or (4) a channel-like protein in which the substrates could specifically interact with different amino-acids along the pore [ 166 ]. An additional, provocative and fruitful hypothesis was added by Mitchell in 1957 [ 167 ]. His model (Figure  11 C) is striking at first because it clearly assumed the transporter to be a protein embedded in the membrane; such a protein was metabolically-driven, enzyme-like and able to swing its attaching site from one side of the membrane to the other according to its conformational changes.

Although indirect, Mitchell’s contribution to membrane understanding should be stressed. As a former student of Danielli, he actively tried to fill the gap between the community working on cell membranes and those who studied the metabolism [ 194 ]. In addition to his suggestion for a transporter, his chemiosmotic hypothesis certainly marked a turning point in the way membranes were envisioned. Here too, it would be inappropriate to trace back a detailed account on bioenergetics from their origin (see [ 134 ],[ 195 ] for details). Suffice it to recall that in 1961 the formulation of the chemiosmotic hypothesis accounted for the inclusion of the respiratory chain within the proton-impermeable membrane. According to this hypothesis, the membrane-located respiratory chain employed the energy liberated from redox reactions to translocate protons across the proton-impermeable membrane, and the resulting proton gradient was then available for use by membrane-embedded ATPases to synthesize ATP [ 168 ]. Although this suggestion first encountered a vigorous opposition, it became progressively accepted as supplementary studies refined it and some of its predictions were confirmed [ 134 ],[ 194 ],[ 196 ]. Of particular interest for us was the demonstration that uncoupling agents of ADP phosphorylation and electron transport did not mediate their effect through a direct enzymatic inhibition but through the increase of proton permeability across lipid bilayers [ 197 ],[ 198 ]. This experiment supported the hypothesis that lipids were accessible in the cell surface while contradicting the dominant paucimolecular model in which phospholipid bilayers were insufficient to keep an ionic gradient.

An important input from transport studies

In summary, the 1950s and 1960s were full of discoveries seemingly tangential but actually tightly related to cell membranes. These debates concerned the cell permeability, the formation of gradients or the connection with metabolism. The sound transformation that took place in these fields during those years improved knowledge of many membrane components, especially proteins. The pumps, transporters, respiratory chains and ATPases studied in these lines of research required that membrane proteins had access to both sides of the membrane. Although the existence of transmembrane proteins was far from being totally accepted, these hypotheses certainly impacted contemporary ideas on membrane structure at a time not yet dominated by the fluid mosaic model.

Towards a new membrane model

First insights from electron microscopy.

Electron microscopy emerged in the 1930s and the first attempts to apply it to the elucidation of cell structure rapidly followed. However, it was not until the 1950s that sharper resolutions allowed the direct observation of cell membranes [ 88 ],[ 199 ],[ 200 ]. In addition to corroborating their existence, the visualization of cell membranes was expected to provide a powerful tool to investigate their structure. Paradoxically, instead of making things clearer, electron microscopy images launched 15 years of a passionate debate over cell membrane structure (light green boxes in Figure  3 ).

Indeed, the interpretation of the pictures obtained was difficult and naturally influenced by former conceptions of membrane structure. For instance, Hillier and Hoffman observed the surface of erythrocytes in 1953 and described their membrane as made up of superficial “plaques” attached to an underlying, internal fibrous material [ 199 ]. This hypothesis tried to reconcile both the paucimolecular model and the pore theory: on the one hand, the plaques and fibers were thought to correspond to the protein envelope of the membrane of the paucimolecular model; on the other hand, the variable space left among the plaques was suggested to account for the variability of pore diameter in the mosaic model. The same year, 1953, Frey-Wyssling and Steinmann examined the thylakoid surface of plant chloroplasts with the electron microscope [ 201 ]. They saw a granular surface and suggested that thylakoid membranes were made up of globular lipoprotein subunits–a possibility that would be extensively explored in the next decade.

The improvement in electron microscopy resolution and sample preparation also allowed for the observation of membrane cross-sections. These pioneering works raised the question of what visual structure should be assumed as the limit of the cell [ 8 ]. This question was very challenging given the biological materials that were often used for these observations: Muscle, nerves and microorganisms displayed complicated external structures that made it difficult to determine which layer corresponded to the cell membrane. Even when extracellular structures had been discarded, the interpretation of the remaining superficial layer of the cell was still not self-evident. Cross-sections of that superficial layer revealed the structure known as the “railroad track”: two dense lines separated by a middle, lighter space. Yet, which one among the whole structure or one of the dense parts should be considered as the quintessential cell membrane? In 1959, Robertson compared a collection of cross-section pictures and observed that the whole railroad track was consistently observed in a variety of cells. He thought that the railroad track fit the paucimolecular model, assuming that the dark parts were the protein layers sandwiching and the lighter, the lipid bilayer. As a result, he considered the whole structure to be the cell membrane. Although his hypothesis was not significantly different from the paucimolecular model, he renamed this model the “unit membrane” in order to stress two points: First, that the three layers observed in the electron microscope cross-section shots were part of the same structure, the cell membrane, regardless of the other cell envelopes that might exist; and second, that this structure was universally shared among all biological membranes [ 202 ].

The diversity of biological membranes

The controversial formulation of the unit membrane hypothesis announced a relentless confrontation in the 1960s between the predominant model at the time (the paucimolecular hypothesis) and the plethora of other membrane explanations suggested by the increasing amount of contradictory information.

Many authors called into question the idea that one unique membrane model could account for all biological membranes because it seemed contradictory to the large membrane diversity that was being discovered at the time [ 9 ],[ 203 ]. For example, it had been observed that the protein-lipid ratio from different membranes could vary from 1:4 to 5:1, suggesting that the amount of membrane proteins was not always enough to entirely cover the cell surface [ 9 ]. In addition, the lack of resolution prevented the observation of the railroad track in some bacteria, thus implying that the three-layer structure was either not always visible or not universal [ 204 ]. The wide functional diversity of membranes also seemed to contradict the unit membrane: How could the insulating membrane of myelinized nerves have the same structure as the metabolically active membranes responsible for oxidative phosphorylation and photosynthesis [ 9 ]?

For some time, the idea that biological membranes could be made up of specific subunits seemed appealing. These hypotheses were natural extensions of the previous thylakoid electron microscopy images by Frey-Wyssling and Steinmann and were popularized by the observation of similar repetitive structures in the chloroplast [ 205 ],[ 206 ], mitochondria [ 207 ],[ 208 ] or even plasma membranes [ 209 ]. These hypotheses generally implied that the subunits could be different from one membrane to another and account for membrane functional diversity [ 9 ]. As it was becoming more generally accepted that lipids and proteins mainly interacted through their hydrophobic parts, some authors imagined the subunits as dynamic micellar mixes of proteins and lipids [ 210 ],[ 211 ]. Electron microscopy images from viral capsides also lent credit to the view that membranes could be made up of subunits [ 212 ]. From an historical perspective, it is noteworthy that these models emphasized the dominant role of proteins as the major structural components of membranes–a fashionable idea at a time when radical developments were being made in molecular biology [ 213 ]. The so-called subunits that were being observed most likely corresponded to the functional proteins that dominated the mitochondrial and chloroplastic membranes, for example, ATPases or photosystems.

Given the criticisms, even advocates of the paucimolecular model were compelled to adopt some modifications. For example, in a 1964 review, Brady and Trams incorporated some characteristics that prefigured the modern formulation of the fluid mosaic model: (1) the proteins also penetrated the membrane and, therefore, the membrane was a lipid-protein mosaic; (2) membranes were not homogeneous but could have membrane segments specific for some permeability functions; (3) the lipid components of the membrane were fluid; and (4) membrane formation was ruled by the thermodynamic search of the lowest energy structure [ 214 ].

Thus, in spite of Robertson’s unit membrane, the debate on membrane structure steadily grew in the 1960s. Although the debates mostly crystallized around the unit membrane and the subunit-based hypotheses, the change in perspective that allowed the prevalence of the fluid mosaic model was highly impacted by the gathering of conclusive evidence from new techniques.

Evidence for a new membrane hypothesis

Artificial lipid bilayers.

Most studies on amphipathic molecules so far had been focused on molecular monolayers. The first lipid bilayers were artificially prepared in the 1960s by Mueller and collaborators (red boxes in Figure  3 ), thus providing a much more suitable model to be compared to the biological membranes [ 215 ]. Synthetic membranes showed that lipid bilayers were stable even when proteins were totally absent. This observation did not refute the paucimolecular model, but it was an important contradiction to Danielli’s assumption that naked lipid bilayers were too delicate to act as boundaries [ 216 ]. The addition of proteins to artificial lipid membranes gave some insight into the capacity of peptides to confer permeability and excitability in membranes [ 215 ],[ 217 ]. Moreover, when artificial lipid bilayers were examined with the electron microscope, it revealed a railroad track similar to biological membranes [ 218 ]; this result was at odds with the idea that dense layers corresponded to the proteins coating the membrane in the paucimolecular model. Finally, a better understanding of the parameters that ruled lipid interaction allowed Bangham and collaborators to prepare the first liposomes (i.e. artificial vesicles bounded by a lipid bilayer [ 219 ]) in 1965. In fact, liposomes had been observed as early as 1854 by Virchow [ 220 ],[ 221 ]; they had been studied by Lehmann and Reinitzer at the turn of the century [ 221 ],[ 222 ] and lipid suspensions had been made throughout the century, but previous authors had failed to understand that liposomes enclosed an aqueous phase [ 221 ]. Now the liposomes would rapidly become important structures in the study of membrane permeability and allowed a fruitful comparison to the biological membranes.

Fracture of frozen membranes

The first electron microscopy results had been criticized because the native structures were suspected to be modified by the chemical fixation of the biological material [ 9 ]. This criticism was circumvented by the development of the freeze-etching method by Moor and Mühlethaller in 1963 [ 223 ]. In this technique, the sample is frozen for fixation and fractured before it is examined with the electron microscope (light green boxes in Figure  3 ). Branton and collaborators carried out further analyses both with natural and artificial lipid membranes in the late 1960s. They found that the freezing technique kept the hydrophilic interactions at the surface but canceled the hydrophobic forces in the interior of the membrane, thus allowing the membrane to be broken apart between the two lipid bilayers [ 224 ],[ 225 ]. As the interior of the natural membrane became visible, some protuberances were reported on the internal side of each monolayer that were mirrored by depressions in the opposite monolayer. According to these results, the membranes were understood as lipid bilayers intercalated with “globules” that would soon be assimilated into the membrane proteins [ 226 ].

Membrane fluidity

One of the earliest questions of membrane studies was if the membranes were better depicted as liquids or solids [ 65 ],[ 106 ],[ 227 ]. A giant step forward was made in 1970 by Frye and Edidin when they fused two cells (one human, one mouse) together in order to monitor the fate of their membranes ([ 228 ] dark green boxes in Figure  3 ). Each cell carried different surface antigens whose movements could be monitored using a fluorescent antibody. After the cell fusion, the two membrane fluorescences became progressively intermixed, thus suggesting that the membrane components were able to freely diffuse in the membrane plane.

Protein-lipid interactions

As previously explained, the first definitions of amphipathy, hydrophobicity and hydrophily were sought in the 1910s (dark green boxes in Figure  2 and 3 ). However, it was not until the 1960s that works like those from Haydon and Taylor emphasized the dramatic role of thermodynamics to determine the structure of biological membranes [ 229 ]. Even once the thermodynamic argument had entered the discussion of biological membranes, the first attempts to account for these constraints failed to fully assess their intensity. Hydrophobic and hydrophilic interactions were shown to rule the contact between lipids and proteins [ 230 ], yet the early hypotheses often assumed that micellar mixes of lipids and proteins would have been the more stable structures [ 231 ], thus favoring the subunit-based models.

A significant breakthrough was achieved when the study of membrane protein conformation became possible. The early paucimolecular model had depicted the coating proteins as globular [ 115 ], but in the 1960s the evolution of the model had led to the assumption that the proteins were unwrapped in a conformation similar to a beta sheet [ 108 ]. In 1966, several studies showed that the membrane proteins had an alpha or globular conformation rather than a beta structure [ 232 ]-[ 234 ]. The alpha helices were suggested to cross the membrane, thus providing a structural framework to the transmembrane proteins that had been predicted in the permeability and transport studies. These works also acknowledged the importance of the thermodynamic constraints to determine the lipid-protein interactions and were especially influential because some of their authors, especially Singer, took part in the formulation of the current version of the fluid mosaic model.

Birth and life of the fluid mosaic

In 1971 and 1972, Singer and Nicolson presented their fluid mosaic model of cell membrane structure. The basics of the model have remained the same ever since: the membrane is a lipid bilayer with hydrophilic parts on the sides and hydrophobic parts in the interior; proteins can interact with the surface through transient polar contacts, but a lot of proteins are partially or totally embedded in the lipid bilayer where their hydrophobic parts also interact with the hydrophobic parts of lipids (Figure  1 ).

In the light of the historical account on cell membrane discovery that I have reported, it is apparent that the success of the fluid mosaic model lay not so much in its originality as in its timeliness and scope: It accommodated most of the evidence available at its time and made predictions that would be demonstrated later.

On the one hand, the model was supported by evidence from different origins: (1) the permeability and transport studies that predicted enzyme-like transmembrane proteins [ 167 ]; (2) the apparent lack of lipids to make up complete bilayers [ 94 ], thus pointing out to the participation of proteins in the membrane plane; (3) electron microscopy pictures, including freeze-etching studies that suggested the presence of proteins within the membranes [ 224 ]; (4) the stability of artificial lipid bilayers that supported them as suitable and sufficient components to make up structures similar in the biological membranes [ 219 ]; and (5) the favorable conformations predicted for the membrane proteins [ 233 ].

On the other hand, the model was even more influential owing to the assumptions that it highlighted or newly predicted. First, as it was soundly established on thermodynamic grounds, the model enhanced the study of hydrophobic forces, which would subsequently become one of the major explanatory parameters to describe the biological macromolecules [ 2 ]. It is important to point out that, more than any generalization from biological observations–as was the case in the unit membrane, for example–the fact that the model is based in universal physico-chemical constraints is the most convincing argument for its general application in biology. Moreover, the acknowledgement of the thermodynamic hydrophobic constraints improved our understanding of membrane proteins, which in turn significantly improved our picture of membranes [ 6 ]. Some dramatic landmarks in membrane protein depiction were the early resolution of the first tridimensional structure of a transmembrane protein (the archaeal bacteriorhodopsin, [ 235 ]); the development of the patch-clamp technique, which allowed the understanding of single ion channels [ 236 ],[ 237 ]; the discovery of the rotatory catalysis that allows the ATP synthesis by ATPases [ 238 ]; and the late discovery of the aquaporins [ 239 ],[ 240 ], which are water channels essential to understanding the water movements that have intrigued cell biologists for more than a century [ 241 ]. Interestingly, now that our knowledge on membrane proteins has developed, the importance of lipid interaction for protein folding is becoming clearer and is still a promising line of research for the future [ 242 ],[ 243 ].

Second, since this model is intrinsically fluid, it predicted that the distribution of most molecules in the lateral range would be essentially random, but it also suggested that specific clusters (i.e. microdomains) may form [ 1 ],[ 244 ]. Although mainly ignored for some time, these microdomains have been the subject of intensive research in the last 20 years and the recent introduction of new techniques should continue to improve our understanding of interactions among membrane components [ 245 ],[ 246 ]. Another proof that the fluid nature of membranes remains a subject to be explored further has been the recent interest in state transitions in membranes: In the last decade some biophysicists have studied how membrane transitions from fluid to gel under compression can allow revisiting Hodgkin and Huxley’s mechanism of pulse propagation in excitable membranes [ 247 ].

Finally, the asymmetry of membranes has also proven to be a fruitful characteristic to explore. The idea that membranes had different components in the inner and outer sides of the membrane was not new, but this hypothesis was taken one step further because the model provided an explanation: The high, free energy of activation necessary for the hydrophilic part of a membrane component to cross the hydrophobic membrane core prevented the random tumbling [ 1 ]. Hence, the asymmetry which was already suspected for the oligosaccharides [ 248 ],[ 249 ] was rapidly extended in the 1970s to lipids, transmembrane proteins or peripheral proteins–for instance those related to the cytoskeleton [ 235 ],[ 250 ]-[ 253 ].

In summary, since its formulation in the 1970s, the fluid mosaic model has been modernized to account for further observations, but it has barely been altered. It remains the most explanatory hypothesis to understand biological membranes.

A revision of the cell membrane historiography

Although the subject of cell membrane discovery could certainly be developed further, the information reviewed here should already be enough to point out the major limitations of the majority of previous, short historic accounts on this topic [ 5 ]-[ 7 ],[ 9 ],[ 10 ],[ 202 ]. I will try to base my critical discussion on some arguments that Kuhn used to criticize what he called the “sources of authority”. These criticisms are not necessarily related to his well-known concept of scientific revolutions, but I will also briefly tackle the question in order to broaden the perspective of this review.

Kuhn’s concept of scientific revolutions (or paradigm shifts) postulates that science does not only evolve through the gradual accumulation of evidence, as was previously thought, but also through dramatic revisions of previous data that radically change the overall opinion of a given scientific community on a subject [ 254 ]. Since its formulation, the paradigm concept has been extensively misused [ 255 ] and Kuhn’s ideas have been frequently decried in biology owing to the difficulty in identifying examples of scientific revolutions from specific historical accounts [ 256 ]–including the field of membrane transport [ 134 ],[ 135 ]. Interestingly, the formulation of the modern fluid mosaic model has recently attracted some epistemological interest [ 213 ]. This recent work suggested that the fluid mosaic model did not drive out its predecessor through a complete revision of available data in a Kuhnian sense; instead, it was the result of a synthesis effort between new pieces of evidence and different models. My historical account supports this synthetic interpretation, as it shows that several mosaic models predated Singer and Nicolson’s hypothesis; the final acceptance of the fluid mosaic did not result from a novel change of perspective, but from the accumulation of supportive data from diverse experiments. Of course, this does not deny the radical contribution of Singer and Nicolson’s model to current biology.

According to this analysis, I think that if the history of the discovery of cell membranes can illustrate some dramatic change in perspective in biology, it should be the transition from the understanding of the cell as a colloid to the bounded, highly concentrated solution currently in use. I think that this transition could be understood in Bachelardian terms as a discontinuity between the pre-scientific era of biology and modern science [ 257 ]. Yet, it is not in the scope of this review to carry out a detailed epistemological analysis on the history of the Cell, so I will now move on to the other Kuhnian arguments that I think to be directly relevant to the critical analysis of the cell membrane historiography.

When Kuhn tried to explain the difficulty in accounting for scientific revolutions, he made a particular case for the analysis of the sources of authority, i.e. science textbooks, philosophical works and some outreach presentations of science [ 254 ]. He criticized that most of these sources did not provide a comprehensive historical account of the actual events in the way they were understood at the time of their discovery. These texts instead presented individual experiments or thoughts that could be easily viewed as explanatory contributions to the current paradigm. As the aim of those texts was not to provide a detailed account of historical events, they tended to stress some scattered “great heroes of an earlier age” in their relationship to the modern paradigm and to omit contemporary interpretations and opposing ideas to preserve clarity. Although these devices may be adequate for pedagogical purposes, such a presentation distorts the actual historical reconstruction. The stake in this strategy is not only that it may contribute to hide a scientific revolution, as Kuhn feared, but also to make an inaccurate and oversimplified historiography become repeated and established.

If we return to the history of membrane discovery, it is alarming to see how well Kuhn’s criticisms of the “sources of authority” match some insufficiencies in the short historical accounts [ 5 ]-[ 7 ],[ 9 ],[ 10 ],[ 202 ]. Some examples of this are: (1) the fact that membrane discovery is systematically presented as a very linear process which obviates the many simultaneous lines of research and debates that have existed throughout history–for example the omission of the coexistence between the paucimolecular and the mosaic models in the first half of the XX th century; (2) the presentation of only a few flagship experiments that supposedly settled the current paradigm, whereas a less hasty analysis shows that the scope of these experiments was repeatedly revisited according to the discovery of new facts, such as different perspectives on osmosis and Overton’s experiments; and, most startling, (3) the incomplete (and misleading) report of some experiments to fit in our current understanding, as in the Gorter and Grendel’s case. These inaccuracies have probably little impact on our current membrane research, but we should be aware of their existence in order to avoid perpetuating a historically questionable timeline.

I hope that this review will encourage other scholars to critically revise our short accounts on fields traditionally underrepresented in the History of Science studies as, for example, cell biology, microbiology and biochemistry.

Perspectives from a historical background

The historical analysis of the formulation of the current membrane model is not only relevant to those interested in membranes: it may also provide some lines of reflection on early evolution, the minimal cell concept, the origins of life and synthetic life. All these subjects question our vision of the cell, and the membrane is arguably one of the most essential components of the unit of life concept.

To begin with, we can step back to ask the question of the very necessity of the unit of life notion. It has been argued that the Cell Theory stands on the biological atomism, which postulated the existence of a basic indivisible unit of life well before any precise description of this unit could be made [ 258 ]. According to this appealing analysis, the atomistic idea remained implicit along with the new discoveries that led to and established the Cell Theory. Even the past and present opponents to the Cell Theory seem to agree with the atomistic idea, as their arguments challenge the relevance of the Cell as the unit of life, but not the existence of a “unit” itself [ 258 ]. Determining the appropriateness of biological atomism is a deep epistemological question, which is not the matter here. Yet, if we accept that the unit of life is a reality that we can study in spite of the diversity of opinions on the identity of this unit, the current preferred candidate for this position among biologists remains the Cell. Therefore, as all known cells are bounded by cell membranes, understanding the importance of these structures becomes crucial to our current definition of the unit of life.

The historical analysis supports the progressive acknowledgement in the last decades of the importance of cell boundaries in the fields of early evolution and the origins of life [ 259 ]-[ 261 ]:

As the unit of life, the cell entails some kind of identity that differentiates it from other cells and from the environment. Since it also has a composite structure, the cell requires a mechanism to keep all its components together. Historically, two mechanisms have been envisioned: either all the components remain together because they establish direct interactions in a physical network (the colloid chemistry) or they are compartmentalized by some structure. It is important to remember that even after the discovery of the cell membrane, the colloid hypothesis survived many years and was only replaced when the biological macromolecules started to be analyzed as discrete structures. This is relevant to the origins of life, as well as synthetic life studies, because it supports current thinking that the compartmentalization is one of the very basic characteristics of any cell [ 259 ]-[ 261 ], no matter how primitive or minimal it may be.

The membrane embodies one of the main paradoxical characteristics of life: a cell is a system dependent on external compounds and energy to keep the differences that it maintains with the same environment where it gets its raw material. Although membranes were thought for a long time to be passive structures that just allowed solutes to diffuse across them, we now know that modern membranes are necessarily endowed with the ability to control the entry and exit of molecules depending on their needs, even sometimes against the chemical gradients. According to this observation, it seems important to include a thought about (active) transport mechanisms in all works trying to describe the nascent life. This does not mean that complicated structures, like proteins, had to be present from the very start of compartmentalization. For instance, transmembrane gradient formation based on membrane dynamics and alternative transporter molecules (e.g. RNA molecules) have been studied in recent years [ 262 ]-[ 265 ]. We can expect that the awareness of the importance of (active) transport for all cells will soon attract more attention to this fascinating subject from the researchers working on the origins of life.

Contrary to early assumptions about membranes, one of the major foundations upon which the fluid mosaic model is built is their ever-changing dynamic structure. This allows modern membranes to constantly change their activities according to the requirements of the cell and it is likely that the same could have occurred in early membranes. Such a hint is promising because it could intersect with the increasing interest of the origins of life field in studying the changing abilities of membranes made up from mixed amphiphile solutions [ 266 ]-[ 268 ].

Finally, there are at least two fundamental aspects of membranes that have not been discussed in this review because their contribution to the understanding of membranes was low, but they cannot be neglected when referring to membrane contributions to cells. These are the division of membranes and their role as transducers of messages from the environment. Although membrane division has already attracted some attention in the context of the origins of life [ 269 ], very little is known about the interactions among early cells. Hopefully, both subjects will be further explored in the near future.

Revierwers’ comments

I thank the reviewers for their comments. The manuscript has been revised twice taking into account their remarks.

First round

Referee 1, dr. étienne joly.

This manuscript by Jonathan Lombard provides a very thorough and detailed historical account of the evolution of the notion of cell boundaries over the 300 years that spanned between the initial recognition that living organisms were comprised of cells, in the middle of the 17th century, and 1972, with the advent of the fluid mosaic model, and the now generally accepted view that all living cells are surrounded by biological membranes made of lipid bilayers. Although I am not competent to judge the accuracy of this historical report, and would not know if equivalent works have been published previously, I feel that this manuscript should represent a valuable addition to the field, and that the final parts of the manuscript, and the discussion in particular, raise several interesting questions and prospects.

All these good things being said, despite the tremendous amount of historical work which has clearly gone into assembling this manuscript, I must admit that I have found the reading of this manuscript to be rather cumbersome, and even very hard work for the early historical parts. I have communicated numerous corrections and editorial suggestions to the author directly, and hope that this will help him to produce a revised manuscript that will be easier to read, and thus more useful for the scientific community.

Author’s response: I thank the referee for his constructive suggestions to make the review easier to read. The overall structure of the new version of the manuscript has remained unchanged, but I have rewritten many paragraphs and shuffled some sections in order to clarify their message. I have also tried to make the transitions between paragraphs more fluent and I have removed redundant information to make more obvious the common thread of the text. The new version of the paper has been checked by a professional journalist native in English who helped me to make the reading smoother. Thus, I think that the current version of the manuscript should be more easily readable than the first version.

Referee 2, Eugene V. Koonin

In this long paper, the author describes, in considerable detail, the history of biological membrane research, with an emphasis on the role of the membrane as the active cell boundary that determines what gets into or out of the cell and what remains inside or outside. It is rather surprising to read, as a submission to a biology journal, an article that earnestly addresses the intricacies of the history of research in a particular field, without making much effort to formulate any new concept on the functions or evolution of membranes. This is not a criticism, the history of concepts and misconceptions is useful and interesting in itself.

What is missing, from my perspective, in this article, is any discussion of organellar and other intracellular membranes as well as membranes found in virions. Even if the main emphasis is on cell boundary, contrasting the features of plasma membranes to those of these distinct membranes, especially in the context of the “active membrane” concept, could be useful.

Author’s response: I thank the reviewer for this comment. The intracellular membranes are indeed a fascinating subject of study and I would have been glad to introduce them in my review. Nevertheless, I think that this manuscript is already very long and, as the first referee has noted, the main challenge here is to remain pertinent in order to keep the readers’ attention. Therefore, I have preferred to stick to the core of the subject, namely the origins of the membrane concept and the fluid mosaic model. As for other fields, I referred to intracellular membranes only when their study directly contributed to the storyline that I was trying to highlight in this paper. But I will consider the possibility of preparing a separate review about the history of intracellular membranes.

To me, the following point: “As the membrane is the cell element which decides what may or may not enter the cell, these entities require active membranes. From my point of view, this active membrane vision contradicts some hypotheses on the origins of life that favor a progressive increase in membrane tightness [ 264 ],[ 265 ]” looks like a non sequitur. I do not see how these hypotheses clash with the “active membrane” concept and furthermore, have a difficulty imagining the origin of protocells without gradual tightening of the membrane.

Author’s response: This sentence has been totally removed from the new version of the manuscript. In order to account for the suggestions made by the third reviewer, the final section of the conclusion has been considerably changed.

Referee 3, Armen Mulkidjanian

The review by Lombard is a nice survey on the evolution in understanding the nature of cell envelopes in the course of past three centuries. It is an entertaining reading indeed and there is not much to comment.

Still several points deserve mentioning:

Introduction, line 31ff : The reader can get an impression that any lipid molecules tend to join into a membrane bilayer as the most thermodynamically stable structure. This is not the case. One of conditions of forming a bilayer is a match between the sizes of the hydrophobic and hydrophilic parts of the molecules involved [1]. A bilayer is formed when these parts are of approximately the same “width”. If the hydrophilic head is larger, then the most stable structure is not a bilayer, but a micelle. If, on the contrary, the hydrophobic part is “broader”, then a reverse micelle or a so-called hexagonal phase are the thermodynamically most stable structures.

Author’s response: The text has been changed to account for this comment. Now it says: “When they are diluted in water, amphiphiles spontaneously adopt the most thermodynamically stable molecular structure, namely the one that maximizes both hydrophilic and hydrophobic interactions [2]. These interactions may be affected by several parameters, such as the chemical nature of the molecules, their size, the salinity and pH of the solution. In biological conditions, cell phospholipids form a bilayer in which hydrophobic tails face each other in the core of the structure whereas the hydrophilic heads interact with the water molecules in the sides (Figure 1 ).”

Section “Osmotic studies and artificial membranes”, line 290 ff . The story of “precipitation membranes” is quite exciting. It might be interesting to know whether they remained a historic curiosity or there is a historic connection with modern chemical engineering of nanoscale systems.

Author’s response: I thank the referee for this comment. I had never thought about precipitation membranes in the perspective of nanostructures. Unfortunately, I do not know anything about chemical engineering of nanoscale systems, and this would be a subject too distant from the rest of the review to be included in it. But I appreciate the comment and I will keep it in mind for the future.

Sections “ Membrane-located metabolism and enzyme-like transporters” and “ An important input from transport studies” : It should be noted that some results directly related to the membrane structure, apparently, were not recognized by the community right away because their authors were interested in quite different subjects and therefore did not put emphasis on these results. While Singer, as it can be followed from his publications, was particularly interested in understanding the nature of biological membranes, Peter Mitchell was more interested in the processes of membrane transport and mechanisms of energy conversion. Thereby Mitchell - and his colleagues - needed (implicitly) some working model of a biological membrane. The Figure from the Mitchell’s paper of 1957 [2], which is kindly redrawn by the author as Figure  10 C, provides a quite modern presentation of a biological membrane with integral membrane proteins embedded, without any protein layers that flank the lipid bilayer, not to mention that the hypothetical mechanism for a membrane permease, as shown in the figure, has been later shown to be valid for the vast majority of membrane permeases indeed. Thereby Mitchell did not make any statements on the nature of biological membranes. On the other hand, Singer, in his PNAS paper published nine (!) years later [3], still confronted the scheme with protein layers covering the lipid bilayer and provided evidence on existence of alpha-helical protein that cross the bilayer. Furthermore, the very fact that chemically quite different, small proton-carrying molecules could uncouple oxidative phosphorylation by diffusing across membrane bilayer (as shown first by Skulachev and co-workers [4], this reference should be included), implies the presence of free lipid patches accessible from the water phase for these small molecules; it is across such patches that the uncoupling molecules could diffuse. Again, Skulachev and co-workers did not discuss the presence of these free patches because they were interested in understanding the mechanism of energy conversion. Still, their working model of the membrane should have been that of a mosaic membrane. Hence, the bioenergeticists, particularly the “Mitchellian” community, took the mosaic nature of the membrane and the existence of asymmetrically organized integral membrane proteins as granted; this happened several years before Singer and Nichols published their seminal paper. However, the studies of energy conversion did not provide - at least at that time - any information on the lateral motility of proteins in the membrane. It is not incidental that the paper of Singer and Nicolson [5], in addition to an extended analysis of literature data, also provided experimental evidence of the protein mobility in native membranes. This was the truly new piece of evidence that helped to compile a whole picture of a fluid, mosaic membrane.

Because of this reasoning, I would suggest to move the sections “ Membrane-located metabolism and enzyme-like transporters” and “ An important input from transport studies” to the next chapter (“Towards a new membrane model”).

Author’s response: As the referee says, Mitchell and other authors interested in bioenergetics did not directly address the issue of membrane structure. Their work certainly influenced the way membranes were considered, but it was not used as a piece of evidence to directly oppose the predominant paucimolecular model. That is the reason why I have included Mitchell and others’ work in the chapter “Insights from the movements across the membrane” together with other researches on transport that also indirectly influenced the perspective on membranes. I think that moving this section to the next chapter (“Towards a new membrane model”) would be difficult to fit in my account, so I have preferred to keep it as it originally was. The main modification that I have made in the “Membrane-located metabolism” section has been reducing it in order to make the reading more fluent, according to the advice of Referee 1. The citation to Skulachev and colleagues has been included in the new version of the manuscript.

Section “A revision of the cell membrane historiography” . I am not quite happy with invoking the Kuhn’s concept of scientific revolution in relation to the historic narrative on the evolution of membranes. Thomas Kuhn has developed his theory based on the history of physics, therefore his model does not work as nicely with less exact subjects. There is an extensive sociological study on the “scientific revolution” in bioenergetics [6]; the book exemplifies that the paradigm change happened in a rather chaotic, “non-Kuhnian” manner. I suspect that the same can be said about the revolution in “membranology”. As argued above, many scientists, particularly the first partisans in the field of membrane bioenergetics, have used an almost correct “working model” of a membrane already a decade before it was formally presented by Singer and Nicolson. This kind of development can be hardly imagined in physics. I would suggest skipping or modifying this section.

Author’s response: I agree with the referee in the fact that the acceptance of the fluid mosaic model should not be considered as a scientific revolution in Kuhn’s terms. Apparently the message was not clear enough in the first version of the paper, so I have modified the text to emphasize more this point. I would like to stress that my main objective in mentioning Kuhn’s work was using his criticisms about the “sources of authority” in science, not saying that the historical process described here was a Kuhnian revolution.

Section “Perspectives from a historical background”: Here the author writes that “ one of the dramatic historical changes in thought about the cell membranes was the acknowledgement that biological membranes were not inert but very dynamic structures .” Subsequently, it is stated that “ As the membrane is the cell element which decides what may or may not enter the cell, these entities require active membranes. From my point of view, this active membrane vision contradicts some hypotheses on the origins of life that favor a progressive increase in membrane tightness [ 264 ],[ 265 ]”. Finally, it is pointed out that “ historical analysis reminds us that totally passive membranes do not seem very suitable to explain the active processes characteristic of living cells, apart from the transport of gases and the tiniest molecules. This does not imply that early membranes had to have a developed protein apparatus for selective transport, but whatever the nature of the specific transporters present in the primordial cells, their functions are arguably very ancient. I wonder if the presence of a dynamic membrane able to actively transport molecules and maybe metabolize them should not be considered as one of the requirements for a primordial entity to be considered as alive” .

From my opinion, several different subjects are mixed up in this passage. They deserve being sorted out.

The very fact that proteins and lipids diffuse within the liquid matrix of the bilayer means that biological membranes are dynamic . As far as I know, the dynamic nature of biological membrane has not been challenged neither by authors of the mentioned references 262 and 263 [7, 8]. nor by anyone else in recent years.

According to the author, the membrane is active because “membrane is the cell element which decides what may or may not enter the cell”. Here, it is easy to envision a solid state membrane with pores of particular size that will “decide” what may or may not pass. This membrane, although non-dynamic , could be considered as active using the author’s own definition.

The apparent usage of dynamic and active as interchangeable terms by the author makes this passage even more confusing.

The models in the mentioned references 262 and 263 [7, 8], which are accused in implying passive membranes, in fact, build on an assumption that the very first membranes could be impermeable to large polymeric molecules but leaky to small molecules. Accordingly, these membranes could already decide “what may or may not enter the cell” and should be categorized as active according to Lombard. This vision of primordial membranes is not quite original and could be traced to the studies of Deamer, Luisi, Szostak, Ourisson, Nakatani and their co-workers [9-17]. These authors argued that abiotically formed amphiphilic molecules, most likely, fatty acids of phosphorylated, branched polyprenol-like compounds [10, 12, 14, 18, 19], which may have enveloped the first cells, could not be as sophisticated as modern two-tail lipids. Vesicles formed from such molecules are million times more leaky that vesicles from modern, two-tail lipids [18]. Hence, such vesicles could trap large polymers but not small molecules and ions. This leakiness, however, could have been a key advantage. In the absence of membrane-embedded transport proteins, which apparently could emerge only on a relatively late step of evolution, after the emergence of water soluble proteins [20], the membrane leakiness should have enabled the “consumption” of diverse small molecules by the first cells. In turn, this would favor the development of systems that could trap small molecules by attaching them to intracellular polymers - and thus preventing their escape. Hence, leaky membranes could have driven the emergence of different polymerases, including the translation system. The feasibility of such a mechanism has been experimentally shown [13]. Accordingly, even the first abiotically formed membranes mentioned in [7, 8] should have been both dynamic and active (according to Lombard).

It seems that the author, besides confusing the terms “active membrane” and “dynamic membrane”, falls victim to the popular, albeit erroneous, notion that an ionic disequilibrium across the cell membrane is an essential requirement for life. The leakiness to ions per se neither makes a cell membrane passive nor obligatory kills the cell. Even a leaky membrane will faithfully maintain all disequilibria that concern large polymeric molecules. Modern cells are quite robust concerning the tightness of their membranes. The well-known study by Harold and Van Brunt of 1977 has shown that collapsing of the transmembrane electrochemical potential in bacteria by adding diverse ionophores did not affect the growth rate of these bacteria as long as they were in a rich, K + -containing medium [21]. The same treatment, however, blocked the growth of the bacteria in a rich, but Na + -containing medium. These data show that the high tightness of modern cell membranes is crucial only for one particular function, namely keeping the K + /Na + ratio within the cell above one and providing energy for the respective ion pumps. The prevalence of K + over Na + within the cell is crucial indeed for protein synthesis and some other cell functions [8, 22]. When prevalence of K + over Na + is taken care of, bacterial cells can grow even with “bad”, leaky membranes.

I would suggest a truncation or a major modification of the section “ Perspectives from a historical background”.

Author’s response: I thank the referee for pointing out the distinction between “dynamic” and “active” membranes. In a previous version of this manuscript, I used both terms indistinctly to refer to what should only be considered as active transport. I have corrected this error and, according to the comments of this reviewer, I have shuffled the section “perspectives from a historical background”. I have removed all mentions to the hypothesis of the progressive “tightening” of membranes and tried to only discuss points that could easily be related to recent experiments in the field of the origins of life.

Second round

Although I had found the reading of the initial version of this manuscript to be rather cumbersome, I must say that I am impressed by how improved and easier to read the revised version has become, and I am now confident that it represents a very useful addition to the scientific literature.

Author’s response: I thank the referee for his previous comments, which were very useful to prepare the second version of the manuscript.

I have no further comment and believe that the paper is ready for publication.

Author’s response: I thank the referee for his appreciation and previous comments.

I appreciate the author’s efforts on improving the paper in response to the suggestions of the reviewers. Specifically, in response to my comment #5, the author has dramatically truncated and rewritten the section “Perspectives from a historical background”. Still, the revised version of the manuscript contains the following paragraph, which in my opinion is bound to mislead the reader: “Although membranes were thought for a long time to be passive structures that just allowed solutes to diffuse across them, we now know that modern membranes are necessarily endowed with the ability to actively transport molecules according to their needs. According to this observation, it seems important to include a thought about active transport mechanisms in all works trying to describe the nascent life. This does not mean that complicated structures, like proteins, had to be present from the very start of compartmentalization. For instance, recent works have shown that transmembrane gradients across lipid bilayers could spontaneously arise in prebiotic conditions [262,263]. It has even been argued that RNA molecules could have acted as transporters [ 264 ],[ 265 ]. We can expect that the growing awareness of the importance of active transport in early life will soon attract more attention for this fascinating subject”.

From this wording, the reader can get an impression that refs. 260–263 (in the revised manuscript) contain experimental evidence of a spontaneous formation of transmembrane gradients across lipid bilayers in prebiotic conditions. However, this is not the case. The word “gradient” - in the given context - implies a transmembrane difference in concentrations. Such a difference could not arise spontaneously, i.e. without any free energy input (the second law of thermodynamics). Accordingly refs. 260–263 do not provide any experimental evidence for spontaneous emergence of transmembrane gradients. Refs. 261 [23] and 262 [24] contain some interesting speculations, but no experimental data. The ref. 263 [25] contains experimental evidence that a particular RNA aptamer could facilitate the passive translocation of tryptophan molecules across a lipid bilayer; it is important to note that upon such a translocation a concentration gradient can only decay. And, finally, the ref. 260 by Chen and Szostak [26] describes the formation of proton gradient across fatty acid vesicles upon addition of “new” fatty acid molecules. However, Chen and Szostak themselves emphasized that the formation of the proton gradient was driven by thermodynamically favorable incorporation of the new hydrophobic fatty acid molecules into the membranes of the vesicles. Hence, the formation of the proton gradient was not spontaneous. Chen and Szostak observed the formation of the transmembrane pH difference under very special conditions where polar, charged, and bulky arginine molecules were used as the only cations in the medium. Otherwise the pH difference could not arise because the fatty acid membranes were leaky both to protons and to such cations as K+ or Na+. Conditions where arginine molecules are the only positively charged ions in the medium are not “prebiotic” conditions because they cannot be related to any imaginable geological settings. The seminal article of Chen and Szostak, in fact, shows that:

a bilayer made of primitive, single-chain lipids is leaky both to protons and to monovalent cations.

such a bilayer represents a barrier only for large polar molecules carrying several charges.

To summarize, the author of the review fails to provide any experimental evidence of a spontaneous formation of transmembrane gradients across primitive lipid bilayers in prebiotic conditions. Because of the ability of small ions, such as K+, Na+, Cl-, to diffuse freely in and out of nascent cells, active transport of such ions, if any, should have remained futile until the emergence of two-tail phospholipids and ion-tight bilayers, which seemingly happened independently in bacteria and archea [27]. Active transport of small ions in early life may indeed be a “fascinating subject” but so far it remains only an author’s speculation.

Author’s response

I thank the reviewer for his careful lecture of the article. I did not mean this part of the paper to be as controversial as the reviewer seems to find it to be. The main objective of my review is to provide a new survey on the history of the discovery of cell membranes. I think that this subject is even more exciting when it is placed in the context of modern debates about early membranes. But the issue of early membranes is a wide and hot topic and this review is not the place to discuss it into detail. This review was never intended to provide experimental data about transmembrane gradient formation in early membranes.

Although I don’t agree with all the arguments of the reviewer (especially, I don’t understand why there should be any contradiction for a reaction between being thermodynamically driven and spontaneous), I don’t think this article is the place to go into the specifics of the hypotheses about the early evolution of membranes. As a result, I have reworded the problematic paragraph as follows: “Although membranes were thought for a long time to be passive structures that just allowed solutes to diffuse across them, we now know that modern membranes are necessarily endowed with the ability to control the entry and exit of molecules depending on their needs, even sometimes against the chemical gradients. According to this observation, it seems important to include a thought about (active) transport mechanisms in all works trying to describe the nascent life. This does not mean that complicated structures, like proteins, had to be present from the very start of compartmentalization. For instance, transmembrane gradient formation based on membrane dynamics and alternative transporter molecules (e.g. RNA molecules) have been studied in recent years [ 262 ]-[ 265 ] . We can expect that the awareness of the importance of (active) transport for all cells will soon attract more attention to this fascinating subject from the researchers working on the origins of life.”. I hope that the reviewer will find these changes satisfactory.

Reviewer’s request

I request explicitly that my comments to the manuscript should be published together with the complete reference list [that I provided].

The following are the references cited by the reviewer in his comments:

Israelachvili JN, Marcelja S, Horn RG: Physical principles of membrane organization. Q Rev Biophys 1980, 13(2):121–200.

Mitchell P: A general theory of membrane transport from studies of bacteria. Nature 1957, 180(4577):134–136.

Lenard J, Singer SJ: Protein conformation in cell membrane preparations as studied by optical rotatory dispersion and circular dichroism. Proc Natl Acad Sci U S A 1966, 56(6):1828–1835.

Skulachev VP, Sharaf AA, Liberman EA: Proton conductors in the respiratory chain and artificial membranes. Nature 1967, 216(5116):718–719.

Singer SJ, Nicolson GL: The fluid mosaic model of the structure of cell membranes. Science 1972, 175 (4023):720–731.

Gilbert GN, Mulkay M: Opening Pandora’s Box: Cambridge University Press; 1984.

Szathmary E: Coevolution of metabolic networks and membranes: the scenario of progressive sequestration. Philos Trans R Soc Lond B Biol Sci 2007, 362(1486):1781–1787.

Mulkidjanian AY, Bychkov AY, Dibrova DV, Galperin MY, Koonin EV: Origin of first cells at terrestrial, anoxic geothermal fields. Proc Natl Acad Sci U S A 2012, 109(14):E821-830.

Hargreaves WR, Mulvihill SJ, Deamer DW: Synthesis of phospholipids and membranes in prebiotic conditions. Nature 1977, 266(5597):78–80.

Deamer DW: The first living systems: a bioenergetic perspective. Microbiol Mol Biol Rev 1997, 61(2):239–261.

Deamer DW: Origins of life: How leaky were primitive cells? Nature 2008, 454(7200):37–38.

Szostak JW, Bartel DP, Luisi PL: Synthesizing life. Nature 2001, 409(6818):387–390.

Mansy SS, Schrum JP, Krishnamurthy M, Tobe S, Treco DA, Szostak JW: Template-directed synthesis of a genetic polymer in a model protocell. Nature 2008, 454(7200):122–125.

Nakatani Y, Ribeiro N, Streiff S, Desaubry L, Ourisson G: Search for the most primitive membranes: some remaining problems. Orig Life Evol Biosph 2012, 42(5):497–501.

Streiff S, Ribeiro N, Wu Z, Gumienna-Kontecka E, Elhabiri M, Albrecht-Gary AM, Ourisson G, Nakatani Y: “Primitive” membrane from polyprenyl phosphates and polyprenyl alcohols. Chem Biol 2007, 14(3):313–319.

Ourisson G, Nakatani Y: The terpenoid theory of the origin of cellular life: the evolution of terpenoids to cholesterol. Chem Biol 1994, 1(1):11–23.

Gotoh M, Sugawara A, Akiyoshi K, Matsumoto I, Ourisson G, Nakatani Y: Possible molecular evolution of biomembranes: from single-chain to double-chain lipids. Chem Biodivers 2007, 4(5):837–848.

Mansy SS: Membrane transport in primitive cells. Cold Spring Harb Perspect Biol 2010, 2(8):a002188.

Dibrova DV, Chudetsky MY, Galperin MY, Koonin EV, Mulkidjanian AY: The role of energy in the emergence of biology from chemistry. Orig Life Evol Biosph 2012, 42(5):459–468.

Mulkidjanian AY, Galperin MY, Koonin EV: Co-evolution of primordial membranes and membrane proteins. Trends Biochem Sci 2009, 34(4):206–215.

Harold FM, Van Brunt J: Circulation of H+ and K+ across the plasma membrane is not obligatory for bacterial growth. Science 1977, 197(4301):372–373.

Mulkidjanian AY, Bychkov AY, Dibrova DV, Galperin MY, Koonin EV: Open questions on the origin of life at anoxic geothermal fields. Orig Life Evol Biosph 2012, 42(5):507–516.

Lane N, Martin WF: The origin of membrane bioenergetics. Cell 2012, 151(7):1406–1416.

Vlassov A: How was membrane permeability produced in an RNA world?Orig Life Evol Biosph 2005, 35(2):135–149.

Janas T, Yarus M: A membrane transporter for tryptophan composed of RNA. RNA 2004, 10(10):1541–1549.

Chen IA, Szostak JW: Membrane growth can generate a transmembrane pH gradient in fatty acid vesicles. Proc Natl Acad Sci U S A 2004, 101(21):7965–7970.

Koonin EV, Martin W: On the origin of genomes and cells within inorganic compartments. Trends Genet 2005, 21(12):647–654.

Author’s contributions

JL is responsible for the conception of the review, the bibliographic analysis and the manuscript.

Singer SJ, Nicolson GL: The fluid mosaic model of the structure of cell membranes. Science. 1972, 175: 720-731.

PubMed   CAS   Google Scholar  

Tanford C: The Hydrophobic Effect. Formation of Micelles and Biological Membranes. 1973, Wiley, New York

Google Scholar  

Daleke DL, Lyles JV: Identification and purification of aminophopholipid flippases. Biochem Biophys Acta. 2000, 1486: 108-127.

Ling GN: Debunking the alleged resurrection of the sodium pump hypothesis. Physiol Chem Phys Med NMR. 1997, 29: 123-198.

Edidin M: Lipids on the frontier: a century of cell-membrane bilayers. Nat Rev Mol Cell Biol. 2003, 4: 414-418.

Singer SJ: Some early history of membrane molecular biology. Annu Rev Physiol. 2004, 66: 1-27.

Davson H, Danielli JF: The Permeability of Natural Membranes. 1943, Cambridge University Press, London

Robertson JD: The ultrastructure of cell membranes and their derivatives. Biochem Symp. 1959, 16: 3-43.

CAS   Google Scholar  

Branton D, Park RB: Introduction. Papers on Biological Membrane Structure. Edited by: Branton D, Park RB. 1968, Little, Brown and Company, Boston, 1-22.

Robertson JD: Membrane structure. J Cell Biol. 1981, 91: 189-204.

CAS   PubMed Central   Google Scholar  

Baker JR: The Cell Theory. A Restatement, History and Critique. 1988, Garland Publishing Inc., New York & London

Harris H: The Birth of the Cell. 1999, Yale University Press, New Haven and London

Hooke R: Micrographia, or Some Physiological Descriptions of Minute Bodies Made by Magnifying Glasses. London: Royal Society; 1665.

Grew N: The Anatomy of Vegetables Begun. London: Hickman; 1672.

Grew N: The Anatomy of Plants. London: Rawlins, W; 1682.

Brisseau-Mirbel CF: Exposition et Défense de Ma Théorie de L’organisation Végétale. 1808, Van Cleef, La Haye

Dutrochet H: Recherches Anatomiques et Physiologiques. 1824, J. B. Baillière, Paris

Malpighi M, Malpighi M: Anatome Plantarum. Cui Subjungitur Appendix, Iteratas et Auctas Ejustem Authoris de Ovo Incubato Observationes Continens. London: Royal Society; 1675.

Treviranus LC: Beyträge Zur Pflanzenphysiologie. 1811, Dieterich, Göttingen

Moldenhawer JJP: Beyträge Zur Anatomie Der Pflanzen. 1812, Wäser, Kiel

Link DHF: Grundlehren Der Anatomie Und Physiologie Der Pflanzen. 1807, Danckwerts, Göttingen

Mirbel M: Recherches anatomiques et physiologiques sur le Marchantia polymorpha, pour servir à l’histoire du tissu cellulaire, de l’épiderme et des stomates. Mem Acad Roy Sci Inst de France. 1835, 13: 337-374.

Schleiden MJ: Beiträge zur Phytogenesis . Arch Anat Physiol wiss Med. 1838, 1: 137-176.

Schleiden MJ: Contributions to Phytogenesis. 1847, The Sydenham Society, London

Schwann T: Mikroskopische Untersuchungen Über Die Uebereinstimmung in Der Struktur Un Dem Waschstum Der Thiere Und Pflanzen. 1839, Reimer, G.F, Berlin

Schwann T: Microscopical Researches into the Accordance in the Structure and Growth of Animals and Plants. 1847, The Sydenham Society, London

Turner W: The cell theory, past and present. J Anat Physiol. 1890, 24: 253-287.

PubMed   CAS   PubMed Central   Google Scholar  

Simon J: A Physiological Essay on the Thymus Gland. 1845, Renshaw, H, London

Cohn FJ: Nachträge zur Naturgeschichte des Protococcus pluvialis Kützing. Nova Acta Acad Caes Leop-Carol Ger Nat Cur Bonn. 1850, 22: 605-764.

Leydig F: Lehrbuch Der Histologie Des Menschen Und Der Pflanzen. 1857, Meidinger, Frankfurt

De Bary A: Die Mycetozoen. Zeit wiss Zool. 1860, 10: 88-

Haeckel EHPA: Allgemeine Anatomie Der Organismen. 1866, Reimer, G.F, Berlin

Schultze MJS: Die Gattung Comuspira unter den Monothalamien, und Bemerkungen über die Organisation und Fortplanzung der Polythalamien. Arch f Naturges. 1860, 26: 287-310.

Schultze MJS: Das Protoplasma Der Rhizopoden Und Der Pflanzenzellen. 1863, W. Engelmann, Leipzig

Beale LS: On the Structure of the Simple Tissues of the Human Body. 1861, Churchill, J, London

Schultze MJS: Über Muskelkörperchen und das, was man eine Zelle zu nennen habe. Arch Anat Physiol wiss Med. 1861, 1: 1-27.

Sachs J: Beiträge zur Zellentheorie. Energiden und Zellen Flora. 1892, 75: 57-67.

Hanstein JV: Das protoplasma als Träger der pflanzlichen und thierischen Lebensverrichtungen. Für Laien und Sachgenossen dargestellt. Sammlung von Vortragen fur das deutche Volk. Edited by: Frommel W, Pfaff F. 1880, Winter, Heidelberg, 125-

Nollet JA: Recherches sur les causes du bouillonnement des liquides. Hist Acad Roy Sci 1748, 1:57–104.,

Hewson W: On the figure and composition of the red particles of the blood, commonly called the red globules. Philos Trans R Soc Lond B Biol Sci 1773, 63:303–323.,

Dutrochet H: L’agent Immédiat Du Mouvement Vital Dévoilé Dans Sa Nature et Dans Son Mode D’action Chez Les Végétaux et Les Animaux. 1826, Paris, Dentu

Porret R: Expériences galvaniques curieuses. Annal Chim Phys. 1816, 2: 137-140.

Dutrochet H: Recherches sur l’endosmose et sur la cause physique de ce phénomène. Annal Chim Phys. 1832, 49: 411-437.

Dutrochet H: Mémoire sur l’endosmose des acides. Mem Acad Sci. 1838, 15: 281-311.

Dutrochet H: Nouvelles observations sur l’endosmose et l’exosmose, et sur la cause de ce double phénomène. Annal Chim Phys. 1827, 35: 393-400.

Fick A: Über diffusion. Poggendorff Annal Phys Chem. 1855, 94: 59-86.

Fick A: On Liquid diffusion. Lond Edim Dubl Phil Mag J Sci. 1855, 10: 30-39.

Von Mohl H: Einige Bermerkungen über den Bau der Vegetabilischen Zelle. Bot Zeit. 1844, 2: 273-

Von Mohl H: Grundzüge Der Anatomie Und Physiologi Der Vegetabilischen Zelle. 1851, Vieweg, Braunschweig

Von Mohl H: Principles of the Anatomy and Physiology of the Vegetable Cell. 1852

Nägeli C: Zellenkerne, Zellenbildung und Zellenwachsthum bei den Pflanzen. Z wiss Bot. 1844, 1: 34-133.

Nägeli C, Cramer KE: Pflanzenphysiologische Untersuchungen. 1855, Zurich, Bel Friedrich Schluthess

Pringsheim N: Untersuchungen Über Den Bau Und Die Bindung Der Pflanzenzelle. 1854, Hirschwald, Berlin

Smith HW: The plasma membrane, with notes on the history of botany. Circulation. 1962, 26: 987-1012.

De Vries H: Eine Methode zur Analyse der Turgorkraft. Pring Jahrb wiss Bot. 1884, 14: 427-601.

Hamburguer HJ: De invloed van scheikundige verbindingen op bloedlichaampjes in verband met hare molekulair-gewichten. Orderz Ged Physiol Lab Utrecht Hoogs. 1884, 9: 26-42.

Jacobs MH: Early osmotic history of the plasma membrane. Circulation. 1962, 26: 1013-1021.

Graham T: Liquid diffusion applied to analysis. Philos Trans R Soc Lond B Biol Sci. 1861, 151: 183-224.

Traube M: Experimente zur Theorie der Zellenbildung und Endosmose. Arch Anat Physiol wiss Med. 1867, 1: 87-165.

Pfeffer W: Osmotic investigations. The modern theory of solution. Edited by: Jones HC. 1899, Harper brothers publishers, New York & London

Pfeffer W: Osmotische Untersuchungen. 1877, Engelmann, Leipzig

Fischer MH: Oedema and Nephritis. 1921, Wiley and Sons, New York, 3

Cohen GN, Monod J: Bacterial permeases. Bacteriol Reviews. 1957, 21: 169-194.

Nernst W: Ein osmotischer Versuch. Zeit phys Chem. 1890, 6: 37-40.

Loeb J: The general physical constitution of living matter. The dynamics of living matter. Edited by: Loeb J. 1906, The Columbia University Press, New York, 29-52.

Overton E: Ueber die osmotischen Eigenschaften der lebenden Pflanzen und Thierzelle. Vierteljahrschr Naturf Ges Zurich. 1895, 40: 159-184.

Overton E: Ueber die allgemeinen osmotischen Eigenschaften der Zelle. Vierteljahrschr Naturf Ges Zurich. 1899, 44: 88-135.

Overton E: The probable origin and physiological significance of cellular osmotic properties. Papers on biological membrane structure. Edited by: Branton D, Park RB. 1968, Little, Brown and Company, Boston, 45-52.

Kleinzeller A: The postulate of the cell membrane. A history of biochemistry Exploring the cell membrane: conceptual developments. 1995, Elsevier Science, Amsterdam, 27-90.

Fischer A: Vorlesungen Über Bakterien. 1903, Verlag Von Gustav Fischer, Jena, 2

Mitchell P, Moyle J: Permeation mechanisms in bacterial membranes. Discuss Faraday Soc. 1956, 21: 258-

Overton E: Studien über die Aufnahme der Anilinfarben durch die lebende Zelle. Jahrb f wiss Bot. 1900, 44: 88-114.

Chambers R: A microinjection study on the permeability of the starfish egg. J Gen Physiol. 1922, 5: 189-193.

Finkelstein G: Emil du Bois-Reymond vs Ludimar Hermann. C R Biologies. 2006, 329: 340-347.

PubMed   Google Scholar  

Du Bois-Reymond E: Untersuchungen Über Thierische Elektricität. 1884, Verlag Von Georg Reimer, Berlin

Hermann L: Untersuchungen Über Den Stoffwechsel Der Muskeln, Ausgehend von Gaswechsel Derselben. 1867, Hirschwald, Berlin

Van’t Hoff JH: Die Rolle osmotischen Drucks in der Analogie zwischen Lösungen und Gasen. Zeit phys Chem. 1887, 1: 481-508.

Van’t Hoff JH: The role of osmotic pressure in the analogy between solutions and gases. In The modern theory of solutions. Edited by Jones HC. New York and London: Harper & Brothers Publishers; 1899.

Nernst W: Die elektromotorische Wirksamkeit der lonen. Zeit phys Chem. 1889, 4: 129-181.

De Palma A, Pareti G: Bernstein’s long path to membrane theory: radical change and conservation in nineteenth-century German electrophysiology. J Hist Neurosci. 2011, 20: 306-337.

Bernstein J: Untersuchungen zur Thermodynamik der bioelektrischen Ströme. Pflugers Arch. 1902, 92: 521-562.

Seyfarth EA: Julius Bernstein (1839–1917): pioneer neurobiologist and biophysicist. Biol Cybern. 2006, 94: 2-8.

Höber R: Eine Methode, die elektrische Leitfähigkeit im Innern von Zellen zu messen. Pflugers Arch. 1910, 133: 237-253.

Höber R: Ein zweites Verfähren die elektrische Leitfähigkeit im Innern von Zellen zu messen. Pflugers Arch. 1912, 148: 189-221.

Höber R: Messungen der inneren Leitfähigkeit von Zellen. Pflugers Arch. 1913, 150: 15-45.

Pethig R, Schmueser I: Marking 100 years since Rudolf Höber’s discovery of the insulating envelope surrounding cells and of the beta-dispersion exhibited by tissue. J Electr Bioimp. 2012, 3: 74-79.

Chapman GB, Hillier J: Electron microscopy of ultra-thin sections of bacteria: Cellular division in Bacillus cereus. J Bacteriol. 1953, 66: 362-373.

Chapman GB, Kroll AJ: Electron microscopy of ultrathin sections of Spirillum serpens. J Bacteriol. 1957, 73: 63-71.

Chapman GB: Electron microscope observations on the behavior of the bacterial cytoplasmic membrane during cellular division. J Biophys Biochem Cytol. 1959, 6: 221-224.

Quincke G: Ueber periodische Ausbreitung an Flüssigkeitsoberflächen und dadurch hervorgerufene Bewegungserscheinungen. Ann Phys. 1888, 35: 580-642.

Pockels A: Surface tension. Nature. 1891, 43: 437-439.

Langmuir I: The constitution and fundamental properties of solids and liquids. II Liquids J Am Chem Soc. 1917, 39: 1848-1906.

Gorter E, Grendel F: On bimolecular layers of lipoids on the chromocytes of the blood. J Exp Med. 1925, 41: 439-443.

Bar RS, Deamer DW, Cornwell DG: Surface area of human erythrocyte lipids: reinvestigation of experiments on plasma membrane. Science. 1966, 153: 1010-1012.

Dervichian DG, Macheboeuf M: Sur l’existence d’une couche monomoléculaire de substances lipoïdiques à la surface des globules rouges du sang. C R Hebd Seances Acad Sci. 1938, 206: 1511-1514.

Hardy WB: The tension of composite fluid surfaces and the mechanical stability of films of fluid. Proc R Soc Lond A. 1912, 86: 610-635.

Harkins WD, Davies ECH, Clark GL: The orientation of molecules in the surfaces of liquids, the energy relations at surfaces, solubility, adsorption, emulsification, molecular association and the effect of acids and bases on interfacial tension. J Am Chem Soc. 1917, 39: 541-596.

Lewis GN: The atom and the molecule. J Am Chem Soc. 1916, 38: 762-785.

Bragg WH, Bragg WL: X-rays and Crystal Structure. 1924, The Macmillan Co., New York, 4

Perrin J: La stratification des lames liquides. Ann Phys. 1918, 10: 160-184.

Danielli JF: Some properties of lipoid films in relation to the structure of the plasma membrane. J Cell Comp Physiol. 1936, 7: 393-408.

Langmuir I: Protein Monolayers. Cold Spring Harb Symp Quant Biol. 1938, 6: 171-189.

Langmuir I, Waugh DF: The adsorption of proteins at oil–water interfaces and artificial protein-lipoid membranes. J Gen Physiol. 1938, 21: 745-755.

Fricke H: The electric capacity of suspensions with special reference to blood. J Gen Physiol. 1925, 9: 137-152.

Fricke H, Curtis HJ: Electric impedance of suspensions of yeast cells. Nature. 1934, 134: 102-103.

Danielli JF: The thickness of the wall of the red blood corpuscle. J Gen Physiol. 1935, 19: 19-22.

Waugh DF, Schmitt FO: Investigations of the thickness and ultrastructure of cellular membranes by the analytical leptoscope. Cold Spring Harb Symp Quant Biol. 1940, 8: 233-241.

Davson H: Growth of the concept of the paucimolecular membrane. Circulation. 1962, 26: 1022-1037.

Robertson JD: New observations on the ultrastructure of the membranes of frog peripheral nerve fibers. J Biophys Biochem Cytol. 1957, 3: 1043-1048.

Cole KS: Surface forces of the arbacia egg. J Cell Comp Physiol. 1932, 1: 1-9.

Harvey EN, Shapiro H: The interfacial tension between oil and protoplasm within living eggs. J Cell Comp Physiol. 1934, 5: 255-268.

Danielli JF, Harvey EN: The tension at the surface of mackerel egg oil, with remarks on the nature of the cell surface. J Cell Physiol. 1935, 5: 483-494.

Halliburton WD, Friend WM: The stromata of the red corpuscules. J Physiol. 1889, 10: 532-549.

Jorpes E: The protein component of the erythrocyte membrane or stroma. Biochem J. 1932, 26: 1488-1503.

Danielli JF, Davson H: A contribution to the theory of permeability of thin films. J Cell Comp Physiol. 1935, 5: 495-508.

Schmidt WJ: Doppelbrechung und Feinbau der markscheide der nervenfasern. Z Zellforch Mikrosk Anat. 1936, 23: 657-676.

Schmidt WJ: Birefringence and fine structure of nerve myelin. Papers on biological membrane structure. Edited by: Branton D, Park RB. 1968, Little, Brown and Company, Boston, 83-94.

Schmitt FO, Bear RS, Clark GL: X-ray diffraction studies on nerve. Radiology. 1935, 25: 131-151.

Danielli JF: The relations between surface pH, ion concentrations and interfacial tension. Proc R Soc Lond B. 1937, 122: 155-174.

Harkins WD: The nature and energetics of surfaces. The physical chemistry of surface films. 1952, Reinhold, New York, 1-93.

Nathansohn A: Ueber die Regulation der Aufnahme anorganischer Salze durch die Knollen von Dahlia. Jahrb f wiss Bot. 1904, 39: 607-644.

Höber R, Hoffmann F: Über das elektromotorische Verhalten von künstlichen Membranen mit gleichzeitig selektiv kationen und selektiv anionendurchlässingen Flachenstücken. Arch ges Physiol. 1928, 220: 558-564.

Brooks SC: The accumulation of ions in living cells-a non-equilibrium condition. Protoplasma. 1929, 8: 389-412.

Brooks SC: The intake of radioactive isotopes by living cells. Cold Spring Harb Symp Quant Biol. 1940, 8: 171-180.

Höber R: Permeability. Annu Rev Biochem. 1932, 1: 1-20.

Höber R: Membrane permeability to solutes in its relations to cellular physiology. Physiol Rev. 1936, 16: 52-102.

Collander R: The permeability of plant protoplasts to non-electrolytes. Trans Far Soc. 1937, 33: 985-990.

Lundegardh H: Mineral nutrition of plants. Annu Rev Biochem. 1947, 16: 503-528.

Michaelis L: Contribution to the theory of permeability of membranes for electrolytes. J Gen Physiol. 1925, 8: 33-59.

Jacobs MH: Diffusion processes in non-living and living systems. Proc Am Philos Soc. 1931, 70: 167-186.

Höber R, Hitchcock DI, Bateman JB, Goddard DR, Fenn WO: Chemistry and physics of the plasma membrane. In Physical chemistry of cells and tissues. Philadelphia: The Blakiston Company; 1945:271–284.

Parpart AK, Dziemian AJ: The chemical composition of the red cell membrane. Cold Spring Harb Symp Quant Biol. 1940, 8: 17-24.

Harvey EN, Danielli JF: Properties of the cell surface. Biol Rev. 1938, 13: 319-341.

Robinson JD: Moving Questions. A History of Membrane Transport and Bioenergetics. 1997

Kleinzeller A, Goldman DE, Hollenberg MD, Kinne RKH, Wilson DF: A History of Biochemistry. Exploring the Cell Membrane: Conceptual Developments. 1995, Elsevier, Philadelphia

Overton E: Beiträge zur allgemeinen Muskel- und Nervenphysiologie, II. Mitt. Ueber die osmotischen Eigenschaften der Muskeln. Pflugers Arch. 1902, 92: 346-386.

Macallum AB: On the distribution of potassium in animal and vegetable cells. J Physiol. 1905, 32: 95-128.

Fenn WO: Electrolytes in muscle. Physiol Rev. 1936, 16: 450-487.

Lillie RS: Protoplasmic Action and Nervous Action. 1923, University of Chicago Press, Chicago

Mazia D: The binding of ions by the cell surface. Cold Spring Harb Symp Quant Biol. 1940, 8: 195-203.

Donnan FG: The theory of membrane equilibria. Chem Rev. 1924, 1: 73-90.

Boyle PJ, Conway EJ: Potassium accumulation in muscle and associated changes. J Physiol. 1941, 100: 1-63.

Heppel LA: The diffusion of radioactive sodium into muscles of potassium-deprived rats. Am J Physiol. 1940, 128: 449-454.

Steinbach HB: Electrolyte balance of animal cells. Cold Spring Harb Symp Quant Biol. 1940, 8: 242-254.

Cohn WE, Cohn ET: Permeability of red corpuscles of the dog to sodium ion. Exp Biol Med. 1939, 41: 445-449.

Dean RB, Noonan TR, Haege L, Fenn WO: Permeability of erythrocytes to radioactive potassium. J Gen Physiol. 1940, 24: 353-365.

Hahn L, Hevesy G: Rate of penetration of ions into erythrocytes. Acta Physiol Scand. 1942, 3: 193-223.

Harris EJ: The reversible nature of the potassium loss from erythrocytes during storage of blood at 2-5°C. Biol Bull. 1940, 79: 373-

Danowski T: The transfer of potassium across the human blood cell membrane. J Biol Chem. 1941, 139: 693-705.

Fenn WO: The role of potassium in physiological processes. Physiol Rev. 1940, 20: 377-415.

Dean RB: Theories of electrolyte equilibrium in muscle. Biol Symp. 1941, 3: 331-348.

Krogh A: The active and passive exchanges of inorganic ions through the surfaces of living cells and through living membranes generally. Proc R Soc Lond B. 1946, 133: 140-200.

Conway EJ: Ionic permeability of skeletal muscle fibres. Nature. 1946, 157: 715-717.

Ussing HH: Interpretation of the exchange of radio-sodium in isolated muscle. Nature. 1947, 160: 262-263.

Levi H, Ussing HH: The exchange of sodium and chloride ions across the fibre membrane of the isolated frog sartorius. Acta Physiol Scand. 1948, 16: 232-249.

Harris EJ, Burn GP: The transfer of sodium and potassium ions between muscle and the surrounding medium. Trans Far Soc. 1949, 45: 508-528.

Keynes RD: The ionic movements during nervous activity. J Physiol. 1951, 114: 119-150.

Osterhout WJV: How do electrolytes enter the cell. Proc Natl Acad Sci U S A. 1935, 21: 125-132.

Steward FC: Mechanism of salt absorption by plant cells. Nature. 1935, 135: 553-555.

Hoagland DR: Salt accumulation by plant cells, with special referece to metabolism and experiments on barley roots. Cold Spring Harb Symp Quant Biol. 1940, 8: 181-194.

Lundegardh H: Salt absorption of plants. Nature. 1940, 145: 114-115.

Ussing HH: Transport of ions across cellular membranes. Physiol Rev. 1949, 29: 127-155.

Irwin M: Can protein Act as a carrier in penetration?. Exp Biol Med. 1931, 29: 342-342.

Osterhout WJV: The mechanism of accumulation in living cells. J Gen Physiol. 1952, 35: 579-594.

Eccles JC: Ionic mechanism of postsynaptic inhibition. In Nobel prize lecture Volume. 1963, 145: 6-27.

Burgen ASV: The physiological ultrastructure of cell membranes. Can J Biochem Physiol. 1957, 35: 569-578.

Mitchell P: A general theory of membrane transport from studies of bacteria. Nature. 1957, 180: 134-

Mitchell P: Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature. 1961, 191: 144-148.

Stein WD, Danielli JF: Structure and function in red cell permeability. Discuss Faraday Soc. 1956, 21: 238-251.

Hodgkin AL: The ionic basis of nervous conduction . Nobel prize lecture. 1963, 1: 32-48.

Hodgkin AL, Huxley AF: Action potentials recorded from inside a nerve fibre. Nature. 1939, 144: 706-707.

Hodgkin AL, Huxley AF: A quatitative description of membrane current and its application to conduction and excitation in nerve. J Physiol. 1952, 117: 500-544.

Keynes RD, Lewis PR: The sodium and potassium content of cephalopod nerve fibers. J Physiol. 1951, 114: 151-182.

Hodgkin AL, Katz B: The effect of sodium ions on the electrical activity of the giant axon of the squid. J Physiol. 1949, 108: 37-77.

Hodgkin AL, Keynes RD: Active transport of cations in giant axons from Sepia and Loligo. J Physiol. 1955, 128: 29-60.

Caldwell PC: The effects of certain metabolic inhibitors on the phosphate esters of the squid giant axon. J Physiol. 1956, 132: 35-6P.

Glynn IM: Sodium and potassium movements in Human red cells. J Physiol. 1956, 134: 278-310.

Abood LG, Gerard RW: Enzyme distribution in isolated particulates of rat peripheral nerve. J Cell Comp Physiol. 1954, 43: 379-392.

Skou JC: The identification of the sodium pump. Biosci Rep. 1998, 24: 436-451.

Skou JC: The influence of some cations on an adenosine triphosphate from peripheral nerves. Biochim Biophys Acta. 1957, 23: 394-401.

Skou JC: Enzymatic basis for active transport of Na + and K+ across cell membrane. Physiol Rev. 1965, 45: 596-618.

Kleinzeller A: The concept of a membrane transport carrier. A history of biochemistry Exploring the cell membrane: conceptual developments. 1995, Elsevier Science, Amsterdam, 91-132.

Nagano J: Zur Kenntniss der Resorption einfacher, im besonderen stereoisomer Zucker im Dünndarm. Pflugers Arch. 1902, 90: 389-404.

Cori CF: The fate of sugar in the animal body. I. Th rate of absorption of hexoses and pentoses from the intestinal tract. J Biol Chem. 1925, 66: 691-715.

Westenbrick HGK: Über die Spefizität der Resorption einiger Monosen aus dem Darme der Ratte und der Taube. Arch Neerl Physiol. 1937, 21: 433-454.

Ege R: Variations in the permeability of the red blood cells in man, with particular reference to the conditions obtaining in pernicious anemia. J Clin Invest. 1919, 16: 279-286.

Jacques AG: The kinetics of penetration: X. Guanidine Proc Natl Acad Sci USA. 1935, 21: 488-492.

Jacobs MH, Corson S: The influence of minute traces of copper on certain hemolytic processes. Biol Bull1. 1934, 67: 325-326.

Lefevre PG, Lefevre MEL: The mechanism of glucose transfer into and out of the human red cell. J Gen Physiol. 1952, 35: 891-906.

Davson H, Reiner JM: Ionic permeability: an enzyme-like factor concerned in the migration of sodium through the cat erythrocyte membrane. J Cell Physiol. 1942, 20: 325-342.

Widdas WF: Inability of diffusion to account for placental glucose transfer in the sheep and consideration of the kinetics of a possible carrier transfer. J Physiol. 1952, 118: 23-39.

Danielli JF: Morphological and molecular aspects of active transport. Symp Soc Exp Biol. 1954, 8: 502-516.

Goldacre RJ: The folding and unfolding of protein molecules as a basis of osmotic work. Int Rev Cytol. 1952, 1: 135-164.

Mitchell P: David Keilin’s respiratory chain concept and its chemiosmotic consequences . Nobel prize lecture. 1978, 1: 295-330.

Wilson DF: Energy metabolism in cellular membranes. A history of biochemistry Exploring the cell membrane: conceptual developments. 1995, Elsevier Science, Amsterdam, 231-278.

Mitchell P: Chemiosmotic Coupling in Oxidative and Photosynthetic Phosphorylation. 1966, Glynn research laboratories, Bodmin

Hopfer U, Lehninger AL, Thompson TE: Protonic conductance across phospholipid bilayer membranes induced by uncoupling agents for oxidative phosphorylation. Proc Natl Acad Sci U S A. 1968, 59: 484-490.

Skulachev VP, Sharaf AA, Liberman EA: Proton conductors in the respiratory chain and artificial membranes. Nature. 1967, 216: 718-719.

Hillier J, Hoffman JP: On the ultrastructure of the plasma membrane as determined by the electron microscope. J Cell Comp Physiol. 1953, 42: 203-247.

Murray RGE: Direct evidence for a cytoplasmic membrane in sectioned bacteria. Can J Microbiol. 1957, 3: 531-534.

Frey-Wyssling A, Steinmann E: Fine structure analysis of chloroplasts. In Papers on biological membrane structure. Edited by Branton D, Park RB. Boston: Little, Brown and Company; 1953:210–218.

Robertson JD: The molecular structure and contact relationships of cell membranes. Prog Biophys Mol Biol. 1960, 10: 343-418.

Korn ED: Structure of biological membranes. The unit membrane theory is reevaluated in light of the data now available. Science. 1966, 153: 1491-1498.

Robinow CF, Marak J: On the plasma membrane of some bacteria and fungi. Circulation. 1962, 26: 1092-1104.

Park RB, Pon NG: Correlation of structure with function in Spinacea oleracea chloroplasts. J Mol Biol. 1961, 3: 1-10.

Park RB: Substructure of chloroplast lamellae. J Cell Biol. 1965, 27: 151-161.

Fernandez-Moran H: Cell-membrane ultrastructure: low-temperature electron microsopy and X-Ray diffraction studies of lipoprotein components in lamellar systems. Circulation. 1962, 26: 1039-1065.

Fernández-Morán H, Oda T, Blair PV, Green DE: A macromolecular repeating unit of mitochondrial structure and function. J Cell Biol. 1964, 22: 63-100.

PubMed   PubMed Central   Google Scholar  

Sjöstrand FS: A new repeat structural element of mitochondrial and certain cytoplasmic membranes. Nature. 1963, 199: 1262-1264.

Kavanau JL: Membrane structure and function. Fed Proc. 1966, 25: 1096-1107.

Benson AA: On the orientation of lipids in chloroplast and cell membranes. J Am Oil Chem Soc. 1966, 43: 265-270.

Hechter O: Role of water structure in the molecular organization of cell membranes. Fed Proc. 1965, 24: 91-102.

Morange M: What history tells us. XXX: the emergence of the fluid mosaic model of membranes. J Biosci. 2013, 38: 3-7.

Brady RO, Trams EG: The chemistry of lipids. Annu Rev Biochem. 1964, 33: 75-100.

Mueller P, Rudin DO, Tien HT, Wescott WC: Reconstitution of cell membrane structure in vitro and its transformation into an excitable system. Nature. 1962, 194: 979-980.

Tien HT, Diana AL: Some physical properties of bimolecular lipid membranes produced from new lipid solutions. Nature. 1967, 215: 1199-1200.

Bean RC, Shepherd WC, Chan H, Eichner J: Discrete conductance fluctuations in lipid bilayer protein membranes. J Gen Physiol. 1969, 53: 741-757.

Henn FA, Decker GL, Greenawalt JW, Thompson TE: Properties of lipid bilayer membranes separating two aqueous phases: electron microscope studies. J Mol Biol. 1967, 24: 51-54.

Bangham AD, Standish MM, Watkins JC: Diffusion of univalent ions across the lamellae of swollen phospholipids. J Mol Biol. 1965, 13: 238-252.

Virchow R: Über das ausgebreitete Vorkommen einer dem Nervenraark analogen Substanz in den thierischen Geweben. Virchows Arch Pathol Anat Physiol. 1854, 6: 416-426.

Lasic DD, Barenholz Y: Liposomes: Past, Present and Future. Handbook of nonmedical applications of liposomes From gene delivery and diagnosis to ecology Volume IV. Edited by: Lasic DD, Barenholz Y. 1995, CRC Press, Boca Raton, New York, London, Tokyo, 299-315.

Lehmann O: Die Flussigen Kristalle. 1911, akademische Verlagsgesellschaft m. b. h., Leipzig

Moor H, Mühlethaler K: Fine Structure in Frozen-Etched Yeast Cells. J Cell Biol. 1963, 17: 609-628.

Branton D: Fracture faces of frozen membranes. Proc Natl Acad Sci U S A. 1966, 55: 1048-1056.

Deamer DW, Branton D: Fracture planes in an ice-bilayer model membrane system. Science. 1967, 158: 655-657.

Pinto da Silva P, Branton D: Membrane splitting in freeze-etching. J Cell Biol. 1970, 45: 598-605.

Kopac MJ, Chambers R: The coalescence of living cells with oil drops. J Cell Comp Physiol. 1937, 9: 345-361.

Frye LD, Edidin M: The rapid intermixing of cell surface antigens after formation of mouse-human heterokaryons. J Cell Sci. 1970, 7: 319-335.

Haydon DA, Taylor J: The stability and properties of bimolecular lipid leaflets in aqueous solutions. J Theor Biol. 1963, 4: 281-296.

Richardson SH, Hultin HO, Green DE: Structural proteins of membrane systems. Proc Natl Acad Sci U S A. 1963, 50: 821-827.

Green DE, Fleischer S: The role of lipids in mitochondrial electron transfer and oxidative phosphorylation. Biochim Biophys Acta. 1963, 70: 554-582.

Maddy AH, Malcolm BR: Protein conformations in the plasma membrane. Science. 1965, 150: 1616-1618.

Lenard J, Singer SJ: Protein conformation in cell membrane preparations as studied by optical rotatory dispersion and circular dichroism. Proc Natl Acad Sci U S A. 1966, 56: 1828-1835.

Wallach DFH, Zahler PH: Protein conformations in cellular membranes. Biochemistry. 1966, 56: 1552-1559.

Henderson R, Unwin PNT: Three-dimensional model of purple membrane obtained by electron microscopy. Nature. 1975, 257: 28-32.

Neher E, Sakmann B: Single-channel currents recorded from membrane of denervated frog muscle fibres. Nature. 1976, 260: 799-802.

Neher E: Ion channels for communication between and within cells . Nobel prize lecture. 1991, 1: 10-25.

Walker JE: ATP synthesis by rotatory catalysis. Nobel prize lecture. 1997, 146-168.

Benga G, Popescu O, Pop VI: p-(Chloromercuri)benzenesulfonate binding by membrane proteins and the inhibition of water transport in human erythrocytes. Biochemistry. 1986, 25: 1535-1538.

Preston GM, Carroll TP, Guggino WB, Agre P: Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science. 1992, 256: 385-387.

Lillie RS: Increase of permeability to water following normal and artificial activation in sea-urchin eggs. Am J Physiol. 1912, 40: 249-266.

Phillips R, Ursell T, Wiggins P, Sens P: Emerging roles for lipids in shaping membrane-protein function. Nature. 2009, 459: 379-385.

Zhang YM, Rock CO: Membrane lipid homeostasis in bacteria. Nat Rev Microbiol. 2008, 6: 222-233.

Singer SJ: The molecular organization of membranes. Annu Rev Biochem. 1974, 43: 805-833.

Simons K, Ikonen E: Functional rafts in cell membranes. Nature. 1997, 387: 569-572.

Boekema EJ, Scheffers D-J, van Bezouwen LS, Bolhuis H, Folea IM: Focus on membrane differentiation and membrane domains in the prokaryotic cell. J Mol Microbiol Biotechnol. 2013, 23: 345-356.

Appali R, Petersen S, van Rienen U: A comparison of Hodgkin-Huxley and soliton neural theories. Adv Radio Sci. 2010, 8: 75-79.

Eylar EH, Madoff MA, Brody OV, Oncley JL: The contribution of sialic acid to the surface charge of the erythrocyte. J Biol Chem. 1962, 237: 1992-2000.

Nicolson GL, Singer SJ: Ferritin-conjugated plantagglutinins as specific saccharide stains for electron microscopy. Application to saccharides bound to cell membranes. Proc Natl Acad Sci U S A. 1971, 68: 942-945.

Bretscher MS: Asymmetrical lipid bilayer structure for biological membranes. Nature. 1972, 236: 11-12.

Gordesky SE, Marinetti GV: The asymmetric arrangement of phospholipids in the human erythrocyte membrane. Biochem Biophys Res Comm. 1973, 50: 1027-1031.

Nicolson GL: Transmembrane control of the receptors on normal and tumor cells. Biochim Biophys Acta. 1976, 457: 57-108.

Bennett V, Stenbuck JP: The membrane attachment protein for spectrin is associated with band 3 in human erythrocyte membranes. Nature. 1979, 280: 468-473.

Kuhn TS: The Structure of Scientific Revolutions. 1962, The University of Chicago Press, Chicago & London

Marris E, Pearson H, Waldrop MM, Hayden EC, Schiermeier Q, Baker M, Brumfiel G, Ledford H: Disputed definitions. Nature. 2008, 455: 1023-1028.

Wilkins AS: Are there Kuhnian revolutions in biology?. Bioessays. 1996, 18: 695-696.

Bachelard G: La Formation de L’esprit Scientifique. 1938, Vrin, Paris

Nicholson DJ: Biological atomism and cell theory. Stud Hist Philos Sci Part C. 2010, 41: 202-211.

Deamer DW: The first living systems: a bioenergetic perspective. Microbiol Mol Biol Rev. 1997, 61: 239-261.

Szostak JW, Bartel DP, Luisi PL: Synthesizing life. Nature. 2001, 409: 387-390.

Schrum JP, Zhu TF, Szostak JW: The origins of cellular life. Cold Spring Harb Symp Quant Biol. 2010, 2: a002212-

Chen I, Szostak JW: Membrane growth can generate a transmembrane pH gradient in fatty acid vesicles. Proc Natl Acad Sci U S A. 2004, 101: 7965-7970.

Lane N, Martin WF: The origin of membrane bioenergetics. Cell. 2012, 151: 1406-1416.

Vlassov A: How was membrane permeability produced in an rna world?. Orig Life Evol Biosph. 2005, 35: 135-149.

Janas T, Janas T, Yarus M: A membrane transporter for tryptophan composed of RNA. RNA (New York, NY). 2004, 10: 1541-1549.

Deamer DW: Role of amphiphilic compounds in the evolution of membrane structure on the early Earth. Orig LIfe Evol Biosph. 1986, 17: 3-25.

Monnard P-A, Deamer DW: Membrane self-assembly processes: steps toward the first cellular life. Anat Rec. 2002, 268: 196-207.

Szostak JW: An optimal degree of physical and chemical heterogeneity for the origin of life?. Philosophical transactions of the Royal Society Series B, Biological sciences. 2011, 366: 2894-2901.

Hanczyc MM, Szostak JW: Replicating vesicles as models of primitive cell growth and division. Curr Opin Chem Biol. 2004, 8: 660-664.

Download references

Acknowledgements

I thank D. Moreira, M. Morange, P. Capy, S. Gribaldo, V. Daubin and O. Lespinet for their encouragements to publish this work. I also thank D.M., C. Petitjean, D. Naïm, N. Duncan, two previous anonymous referees and the three reviewers from Biology Direct for their comments on the manuscript. This work was supported by the National Science Foundation through the National Evolutionary Synthesis Center (NESCent) under grant number NSF #EF-0905606.

Author information

Authors and affiliations.

National Evolutionary Synthesis Center, 2024 W. Main Street Suite A200, Durham, 27705, NC, USA

Jonathan Lombard

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Jonathan Lombard .

Additional information

Competing interests.

The author declares that he has no competing interests.

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Authors’ original file for figure 2, authors’ original file for figure 3, authors’ original file for figure 4, authors’ original file for figure 5, authors’ original file for figure 6, authors’ original file for figure 7, authors’ original file for figure 8, authors’ original file for figure 9, authors’ original file for figure 10, authors’ original file for figure 11, rights and permissions.

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/4.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Cite this article.

Lombard, J. Once upon a time the cell membranes: 175 years of cell boundary research. Biol Direct 9 , 32 (2014). https://doi.org/10.1186/s13062-014-0032-7

Download citation

Received : 11 July 2014

Accepted : 03 December 2014

Published : 19 December 2014

DOI : https://doi.org/10.1186/s13062-014-0032-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Cell membrane discovery
• Cell membrane structure
• Cell Theory
• History of Science
• Cell definition
• Origins of life
• Early evolution
• Cenancestor

Biology Direct

ISSN: 1745-6150

• General enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts

Hypoxia induces mitochondrial protein lactylation to limit oxidative phosphorylation

• Jiaojiao Zhang
• Shimin Zhao

System-level time computation and representation in the suprachiasmatic nucleus revealed by large-scale calcium imaging and machine learning

• Zichen Wang
• Heping Cheng

Discovery and structural mechanism of DNA endonucleases guided by RAGATH-18-derived RNAs

• Fengxia Zhou
• Zhiwei Huang

Targeting pro-inflammatory T cells as a novel therapeutic approach to potentially resolve atherosclerosis in humans

• Jian’an Wang

Multiple cancer cell types release LIF and Gal3 to hijack neural signals

Current issue, sensing the toxic aluminum cations in acidic soils.

• Peter R. Ryan
• Jianli Yang

CCHFV entry via LDLR keeps it ‘ticking’?

• Ahmed K. Oraby
• David J. Marchant

Enabling whole pathway reconstruction using artificial chromosomes

• Jens Nielsen

The LRR receptor-like kinase ALR1 is a plant aluminum ion sensor

• Zhong Jie Ding
• Shao Jian Zheng

B cell-reactive triad of B cells, follicular helper and regulatory T cells at homeostasis

Convenient synthesis and delivery of a megabase-scale designer accessory chromosome empower biosynthetic capacity.

• Ying-Jin Yuan

Announcements

Behind the Paper

In these Behind the Paper posts, our authors share the real stories behind the latest research papers, from conception to publication, the highs and the lows.

Center for Excellence in Molecular Cell Science

Guide to Authors

Research Data

Advertisement

Browse articles

Human tumor suppressor PDCD4 directly interacts with ribosomes to repress translation

• Zixuan Huang
• Jingdong Cheng

Structural basis for the recognition of HCoV-HKU1 by human TMPRSS2

• Lingyun Xia
• Yuanyuan Zhang

Peripheral cancer remodeling of central neural system

• Gray Umbach
• Shawn L. Hervey-Jumper

APOE3ch alleviates Aβ and tau pathology and neurodegeneration in the human APP NL-G-F cerebral organoid model of Alzheimer’s disease

Structural basis for human Ca v 3.2 inhibition by selective antagonists

Immunotherapy for atherosclerosis by targeting pro-inflammatory t cells.

Persistent in vivo epigenetic silencing of Pcsk9

• Isabella R. Gengaro

For Authors & Referees

For authors, for referees, collections.

Editor’s selections: top articles from the past 3 years

Trending - Altmetric

A variant-proof SARS-CoV-2 vaccine targeting HR1 domain in S2 subunit of spike protein

Structural organization of an intact phycobilisome and its association with photosystem II

mTORC2 promotes type I insulin-like growth factor receptor and insulin receptor activation through the tyrosine kinase activity of mTOR

This journal is a member of and subscribes to the principles of the Committee on Publication Ethics.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

EDITORIAL article

This article is part of the research topic.

Molecular Links between Mitochondrial Damage and Human Neurodegenerative Disorders - Volume II

Editorial: Molecular Links between Mitochondrial Damage and Human Neurodegenerative Disorders -Volume II Provisionally Accepted

• 1 Department of Research for Parkinson’s Disease, Juntendo University Graduate School of Medicine, Japan

The final, formatted version of the article will be published soon.

Mitochondria have various functions in cells, including energy production, iron and lipid metabolism, regulation of Ca2+ dynamics, and cell death and inflammatory signaling. Dysregulation and degeneration of mitochondria have been shown to cause a variety of neurodegenerative diseases. In 2021, we launched a research topic entitled 'Molecular Links between Mitochondrial Damage and Human Neurodegenerative Disorders', to which many excellent papers were contributed. This topic consists of two original articles and two review articles that were submitted after 2022 on the same research topic. These manuscripts cover mitochondrial metabolism, mitochondrial DNA, membrane potential, and mitochondrial membrane structure. Fragile X syndrome, one of the triplet repeat diseases, is a disease characterized by developmental delay/intellectual disability beginning in childhood due to mutations in the Fragile X messenger ribonucleoprotein 1 (FMR1) gene. The triplet repeat sequence (CGG) of the FMR1 gene on the long arm of the X chromosome is extended with each generation. When the number of CGG repeats exceeds 200, Fragile X syndrome develops. Reduced expression of the FMR1 gene product Fragile X Messenger Ribonucleoprotein (FMRP) is believed to be the core pathophysiology of fragile X syndrome. FMRP, an RNA-binding protein, binds to mRNA and negatively regulates local translation of mRNA in neurons. Using primary cortical neuronal cultures from FMR1 knockout mice, Bülow et al. report that neuronal activity-dependent changes in mitochondrial morphology and membrane potential differ in dendrites and axon initial segment. The mechanism by which FMRP regulates mitochondria has yet to be determined. FMR1 knockout mice exhibit elevated protein synthesis (Bhattacharya et al., 2012). Metformin inhibits protein translation by blocking mammalian/mechanistic target of rapamycin complex 1 (mTORC1) and suppresses an autism spectrum disorder-like phenotype in FMR1 knockout mice (Gantois et al., 2017). Metformin also inhibits mitochondrial respiratory chain complex I and activates AMP-activated protein kinase signaling (Foretz et al., 2014). This mechanism may provide clues for the elucidation of mitochondrial regulation by FMRP. Mitochondrial DNA deletion syndrome (MDS) is a group of diseases characterized by defects in mitochondrial DNA (mtDNA) maintenance (Basel, 2020). The mitochondrial genome encodes 13 proteins, all of which are important catalytic subunits of the electron transfer chain. Thus, depletion of mitochondrial DNA results in impaired oxidative phosphorylation. Dombi et al. determine the most effective nucleoside supplementation method (optimal concentration and nucleoside combination) based on improving mtDNA content, cell viability, and mitochondrial functions using MDS fibroblasts harboring mutations in POLG, DGUOK, and TWNK (genes encoding alpha subunits of mtDNA γ polymerase, deoxyguanosine kinase, and Twinkle mtDNA helicase, respectively) (Bulst et al., 2009). The findings of this study will be useful for future studies of nucleoside supplementation methods in MDS animal models. Mitophagy, autophagy of mitochondria, is a physiological phenomenon that is highly conserved from yeast to humans. Mitophagy is observed in metabolic changes during development and differentiation and in the quality control of defective mitochondria. The relationship between neurodegeneration and mitophagy was first discovered in a series of studies that showed that PRKN and PINK1, the genes responsible for juvenile Parkinson's disease (PD), are involved in mitophagy (Imai, 2020). Since then, many molecules associated with mitophagy have been identified and PINK1-Parkin (the gene product of PRKN)-independent mitophagy pathways have also been reported. Furthermore, mitophagy research continues to expand with reports of its involvement in neurodegenerative diseases such as Alzheimer's disease and amyotrophic lateral sclerosis (ALS), as well as Parkinson's disease. Jetto et al. provide an overview of mitophagy signaling in neurodegenerative diseases, including some unresolved issues. CHCHD2 and CHCHD10 have been identified as causative genes for PD and ALS/ frontotemporal lobe dementia (FTD), respectively (Bannwarth et al., 2014; Funayama et al., 2015). These two proteins are paralogous to each other, forming a complex in the mitochondrial intermembrane space. CHCHD2 and CHCHD10 have been implicated in maintaining mitochondrial respiratory chain complex function and matrix structure, as well as in the regulation of stress response and cell death. Despite this twin-like relationship, missense mutations in each gene cause different diseases. Ikeda et al. outlined the CHCHD2/CHCHD10 studies reported to date and aimed to provide clues to unravel this mystery. The multifaceted roles of mitochondria and the involvement of mitochondria in various diseases remain a fascinating research topic. Why do specific genes in ubiquitously distributed mitochondria cause specific diseases? To address this question, many researchers are expected to continue work in this field.

Keywords: Mitochondria, mtDNA, mitophagy, CHCHD2, CHCHD10, Fragile X Syndrome, Parkin (PARK2), FMR1

Received: 15 Apr 2024; Accepted: 19 Apr 2024.

Copyright: © 2024 Imai. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Mx. Yuzuru Imai, Department of Research for Parkinson’s Disease, Juntendo University Graduate School of Medicine, Tokyo, Japan

People also looked at

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Portland Press Opt2Pay

Understanding biochemistry: structure and function of nucleic acids

Steve minchin.

School of Biosciences, University of Birmingham, Birmingham, United Kingdom

Julia Lodge

Nucleic acids, deoxyribonucleic acid (DNA) and ribonucleic acid (RNA), carry genetic information which is read in cells to make the RNA and proteins by which living things function. The well-known structure of the DNA double helix allows this information to be copied and passed on to the next generation. In this article we summarise the structure and function of nucleic acids. The article includes a historical perspective and summarises some of the early work which led to our understanding of this important molecule and how it functions; many of these pioneering scientists were awarded Nobel Prizes for their work. We explain the structure of the DNA molecule, how it is packaged into chromosomes and how it is replicated prior to cell division. We look at how the concept of the gene has developed since the term was first coined and how DNA is copied into RNA (transcription) and translated into protein (translation).

The structure of deoxyribonucleic acid

Deoxyribonucleic acid (DNA) is one of the most important molecules in living cells. It encodes the instruction manual for life. Genome is the complete set of DNA molecules within the organism, so in humans this would be the DNA present in the 23 pairs of chromosomes in the nucleus plus the relatively small mitochondrial genome. Humans have a diploid genome, inheriting one set of chromosomes from each parent. A complete and functioning diploid genome is required for normal development and to maintain life.

Discovery and chemical characterisation of DNA

DNA was discovered in 1869 by a Swiss biochemist, Friedrich Miescher. He wanted to determine the chemical composition of leucocytes (white blood cells), his source of leucocytes was pus from fresh surgical bandages. Although initially interested in all the components of the cell, Miescher quickly focussed on the nucleus because he observed that when treated with acid, a precipitate was formed which he called ‘nuclein’. Almost all molecular bioscience graduates would have repeated a form of this experiment in laboratory classes where DNA is isolated from cells. Miescher, Richard Altmann and Albrecht Kossel further characterised ‘nuclein’ and the name was changed to nucleic acid by Altmann. Kossel went on to show that nucleic acid contained purine and pyrimidine bases, a sugar and phosphate. Work in the 1930s from many scientists further characterised nucleic acids including the identification of the four bases and the presence of deoxyribose, hence the name deoxyribonucleic acid (DNA). Erwin Chargaff had found that DNA molecules from a particular species always contained the same amount of the bases cytosine (C) and guanine (G) and the same amount of adenosine (A) and thymine (T). So, for example, the human genome contains 20% C, 20% G, 30% A and 30% T.

DNA is a polymer made of monomeric units called nucleotides ( Figure 1 A), a nucleotide comprises a 5-carbon sugar, deoxyribose, a nitrogenous base and one or more phosphate groups. The building blocks for DNA synthesis contain three phosphate groups, two are lost during this process, so the DNA strand contains one phosphate group per nucleotide.

( A ) A nucleotide (guanosine triphosphate). The nitrogenous base (guanine in this example) is linked to the 1′ carbon of the deoxyribose and the phosphate groups are linked to the 5′ carbon. A nucleoside is a base linked to a sugar. A nucleotide is a nucleoside with one or more phosphate groups. ( B ) A DNA strand containing four nucleotides with the nitrogenous bases thymine (T), cytosine (C), adenine (A) and guanine (G) respectively. The 3′ carbon of one nucleotide is linked to the 5′ carbon of the next via a phosphodiester bond. The 5′ end is at the top and the 3′ end at the bottom.

There are four different bases in DNA, the double-ring purine bases: adenine and guanine; and the single-ring pyrimidine bases: cytosine and thymine ( Figure 1 B). The carbon within the deoxyribose ring are numbered 1′ to 5′. Within each monomer the phosphate is linked to the 5′ carbon of deoxyribose and the nitrogenous base is linked to the 1′ carbon, this is called an N-glyosidic bond. The phosphate group is acidic, hence the name nucleic acid.

In the DNA chain ( Figure 1 B), the phosphate residue forms a link between the 3′-hydroxyl of one deoxyribose and the 5′-hydroxyl of the next. This linkage is called a phosphodiester bond. DNA strands have a ‘sense of direction’. The deoxyribose at the top of the diagram in Figure 1 B is not linked to another deoxyribose; it terminates with a 5′ phosphate group. At the other end the chain terminates with a 3′ hydroxyl.

DNA is the genetic material

Although many scientists, including Miescher, had observed that prior to cell division the amount of nucleic acid increased, it was not believed to be the genetic material until the work of Fredrick Griffith, Oswald Avery, Colin MacLeod and Maclyn McCarty. In 1928, Griffith showed that living cells could be transformed by extracts from heat-killed cells and that this transformation had the potential to permanently change the genetic makeup of the recipient cell. Griffith was working with two strains of the bacterium Streptococcus pneumoniae. The encapsulated so-called S strain is virulent, whereas the non-capsulated R strain is nonvirulent. If the S strain is injected subcutaneously into mice, the mice die, whereas, if either live R strain is injected or heat-killed S strain is injected, the mouse lives. However, if a mixture of live R strain and heat-killed S strain is injected into a mouse, the mouse will die, and live S strain can be isolated from the blood. So, in the Griffith experiment a component of the heat-killed S strain is transforming the R strain. In 1944, Avery, MacLeod and McCarty went on to show that it was DNA that could transform the avirulent bacterium. They isolated a crude DNA extract from the S strain and destroyed any protein, lipid, carbohydrate and ribonucleic acid (RNA) component and showed that this purified DNA could still transform the R strain. However, when the purified DNA was treated with DNAse, an enzyme that degrades DNA, transformation was lost.

Alfred Hershey and Martha Chase confirmed that DNA was the genetic material. They used a virus that infects bacteria called a bacteriophage. The bacteriophage contains a protein capsid surrounding a DNA molecule. They showed that when bacteriophage T2 infects Escherichia coli , it is the phage DNA, not protein, that enters the bacterial cell.

Determining the structure of DNA

Once it had been shown that DNA was the genetic material, there was a race to determine the three-dimensional structure of the DNA molecule. At King’s College London, Rosalind Franklin and Maurice Wilkins, having obtained data using X-ray diffraction, had proposed that DNA had a helical structure and Franklin had obtained a particularly good X-ray diffraction pattern. In Cambridge, James Watson and Francis Crick used model building together with data from a variety of sources including Franklin’s X-ray diffraction pattern and Chargaff’s base composition data to work out the now well-known double helix structure of DNA. Their work was published in Nature in 1953. The Watson–Crick structure is shown in Figure 2 A.

( A ) The DNA double helix, with the sugar phosphate backbone on the outside and the nitrogenous bases in the middle. ( B ) An A:T and a G:C base pair with the C1′ of the deoxyribose indicated by the arrow. Note that the C1′ of the deoxyribose is in the same position in all base pairs. In this figure, the atoms on the upper edge of the base pair face into the major groove and those facing lower edge face into the minor groove. The hydrogen bonds between the base pairs are indicated by the dotted line.

DNA is a two-stranded helical structure, the two strands run in opposite directions. In Figure 2 A, one strand is running 5′ to 3′ top to bottom, whereas the other strand is running 3′ to 5′ top to bottom. The helix is right-handed which means that if you are looking down the axis, the helix turns clockwise as it gets further away from you. The two chains interact via hydrogen bonds between pairs of bases with adenine always pairing with thymine, and guanine always pairing with cytosine. The Watson–Crick structure therefore accounts for and explains the Chargaff data which showed that there was always an equal amount of C and G and of A and T. The regular nature of the double helix comes about because the distance between the 1′ carbon of the deoxyribose on one strand and 1′ carbon of the opposite deoxyribose is always the same irrespective of the base pair ( Figure 2 B). The 1′ carbons of the deoxyribose opposing nucleotides do not lie directly opposite each other on the helical axis, this means that the two sugar–phosphate backbones are not equally spaced along the helical axis resulting in major and minor grooves.

The diameter of the helix is 2 nm, adjacent bases are separated by 0.34 nm (0.34 × 10 −9 m) and related by a rotation of 36°, this results in the helical structure repeating every 10 residues. DNA molecules are normally very long and the sequence of bases along the DNA chain is not restricted. For example, the genome of the bacterium E. coli is a single circular chromosome which contains 4.6 million base pairs (4.6 × 10 6 bp), this is therefore 1.6 mm long (4.6 × 10 6 × 0.34 × 10 −9 m). The human genome is made up of 24 distinct chromosomes, chromosomes 1–22 and the X and Y chromosomes present in the nucleus plus mitochondrial DNA. The nuclear chromosomes vary in size from approximately 50–250 × 10 6 bp, the mitochondrial DNA is 17 × 10 3 bp. The total length of a haploid human genome is 3 × 10 9 bp. Within a single human diploid cell, which contains 23 chromosome pairs there is 2 m of DNA. Based on the assumption that humans contain 3 trillion cells with a nucleus, if all the DNA from a single human individual was put end to end, it would reach to the sun and back approximately 20 times.

Another important class of nucleic acids is RNA, the roles of RNA molecules in the cell will be discussed below. Chemically RNA is similar to DNA, it is a chain of similar monomers. The building blocks are nucleotides containing the 5-carbon sugar ribose, a phosphate and a nitrogenous base. The phosphate is attached to the 5′ carbon of the ribose and the nitrogenous base to the 1′ carbon ( Figure 3 ). RNA contains four bases adenine, guanine, cytosine and uracil. RNA is more labile (easily broken down) than DNA and most RNA molecules do not form stable secondary structures, some notable exceptions will be discussed below. The properties of RNA make it ideal as a genetic messenger during protein synthesis, the idea of this genetic messenger, mRNA, was proposed by François Jacob and Jacques Monod.

An RNA strand containing the four nucleotides with the nitrogenous bases: adenine (A), cytosine (C), guanine (G) and uracil (U) respectively. The 3′ carbon of the ribose of one nucleotide is linked to the 5′ carbon of the next via a phosphodiester bond. The 5′ end on the left and the 3′ end on the right.

Packaging of DNA into eukaryotic cells

DNA has to be highly condensed to fit into the bacterial cell or eukaryotic nucleus. In eukaryotes, histone proteins are used to condense the DNA into chromatin. The basic structure of chromatin is the nucleosome, a nucleosome contains DNA wrapped almost two times around the histone octamer (comprising two copies each of the histone proteins H2A, H2B, H3 and H4) ( Figure 4 ). Further levels of compaction are required to fit the DNA into the nucleus ( Figure 4 ), the nucleosomes are folded upon themselves to form the 30-nm fibre, this is then folded again to form the 300-nm fibre and during mitosis further compaction can occur forming the chromatid which is 700 nm in diameter.

Histone proteins (H2A, H2B, H3 and H4) associate to form a histone octamer. Approximately 147 bp of DNA wraps around histone octamer to form a nucleosome, generating a ‘beads on a string’ structure, the nucleosome together with histone H1 condense into the 30-nm fibre, there is further condensation to form the 300-nm fibre. During mitosis there is further compaction (not shown).

Processes such as DNA replication and DNA transcription need to occur in the chromatin environment and because of the level of compaction, this acts as a barrier to proteins that need to interact with DNA. Therefore, chromatin structure plays an important role in processes such as regulation of gene expression in eukaryotes. DNA and the histone proteins can be chemically modified, these are called epigenetic modifications as they do not change the DNA sequence, however, they can be passed on during cell division and to subsequent generations, a process known as epigenetic inheritance. As these epigenetic modifications can alter the chromatin structure they regulate gene transcription and can affect the phenotype. Epigenetics plays key roles in many processes, including development, cancer and behaviour and addiction. This will be discussed further later in this article.

Nuclear organisation plays an important role in many biological processes including regulation of gene transcription. In recent years the development of several techniques, including microscopy, have allowed us to gain an understanding of the way the genome is organised in 3D. Individual chromosomes are not randomly spaced within the nucleus; each chromosome has a distinct territory. Actively transcribed regions from different chromosomes are often close to each other and near the interior of the nucleus, whereas, inactive genes are on the periphery or near a special area called the nucleolus where ribosomal RNA is transcribed.

DNA replication

Whenever a cell divides there is a need to synthesise two copies of each chromosome present within the cell. For example in a human, prior to cell division, all 23 pairs of chromosomes need to be replicated to form 46 pairs, so that following cell division each daughter cell has a full complement (23 pairs) of chromosomes. The structure of DNA gives us a clue to how it is replicated, this was eloquently postulated by Watson and Crick in their 1953 paper: “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material”. Each strand can act as a template for the synthesis of the complementary strand, so the replication machinery would ‘unzip’ the double helix and read along the two existing ‘parent’ strands, synthesising a complementary new ‘daughter’ strand with A opposite T, C opposite G etc. This is described as semi-conservative, since each ‘new’ double-stranded DNA molecule has one original parent strand and one newly made daughter ‘strand’.

The evidence that DNA replication was semi-conservative came from an elegant experiment completed by Matthew Meselson and Franklin Stahl. They labelled the parental DNA with a heavy isotope of nitrogen ( 15 N) by growing bacteria in a growth medium that contained 15 NH 4 Cl. They then grew the bacteria, in a medium that contained 14 NH 4 Cl, in conditions such that any newly synthesised DNA would contain 14 N. Since DNA replication is semi-conservative, after one round of DNA replication, each cell would have a DNA molecule that contains one ‘old’ parental strand labelled with 15 N and one ‘new’ daughter strand labelled with 14 N. This was shown by analysing the density of the DNA using density-gradient centrifugation. As predicted, they observed that the new daughter DNA molecule had a density consistent with the fact that it contained both 15 N and 14 N and that this daughter DNA contained one strand with 15 N and another strand with 14 N.

DNA polymerase and DNA synthesis

The enzyme, DNA polymerase, is responsible for DNA synthesis. DNA polymerase is a template-driven enzyme, so it will use the parental DNA strand as a template. It cannot synthesise DNA in the absence of a template. In addition, it will only add nucleotides on to the 3′ end of an existing nucleic acid chain. The building blocks for DNA synthesis are deoxynucleoside triphosphates (dATP, dTTP, dCTP and dGTP). During DNA synthesis, the base within the incoming deoxynucleoside triphosphate pairs with the complementary base on the template strand, a phosphodiester bond is formed between the 5′ phosphate on the incoming nucleotide and the free 3′ hydroxyl on the existing nucleic acid chain; pyrophosphate is released ( Figure 5 ).

( A ) DNA polymerase binds the template DNA and the new strand. The next nucleotide to be added to the 3′ end of the growing chain will contain guanine (G), this is complementary to the C on the template strand. DNA polymerase catalyses the formation of a phosphodiester bond. ( B ) The chemical reaction during the formation of a phosphodiester bond, showing the addition of a nucleotide containing guanine and the release of pyrophosphate.

Pyrophosphate is the two phosphate residues within the deoxynucleoside triphosphate building block that are not incorporated into the DNA chain. DNA polymerase synthesises DNA in the 5′ to 3′ direction, because it can only add nucleotides on to the 3′ end of the chain. DNA polymerase has proofreading activity, so after the phosphodiester bond has been formed, the base pairing is checked and if a nucleotide with an incorrect base has been added, DNA polymerase will remove the nucleotide using a 3′ to 5′ exonuclease activity. Exonucleases are enzymes that can remove nucleotides from the ends of a DNA molecule, 3′ to 5′ exonucleases remove nucleotides from the 3′ end of a DNA molecule and therefore can remove the last nucleotide that was added during DNA replication. This is analogous to using the delete key to remove a letter that you have typed incorrectly before adding the correct one and continuing typing.

DNA polymerase requires a short double-stranded region with a free 3′ hydroxyl in order to start making a copy of the template; this ensures that DNA is synthesised in a controlled way. Initiation of DNA synthesis uses a small RNA primer (8–12 bases) made by the enzyme primase. DNA polymerase will then extend from the primer copying the template and synthesising the daughter DNA strand. This means that when DNA synthesis first starts each DNA molecule actually contains a small piece of RNA at its 5′ end. This RNA will ultimately be replaced with DNA, how this is done is discussed below.

The origin of replication and the replisome

A large multiprotein complex, called the replisome, is responsible for DNA replication. In prokaryotes, two replisomes form at a specific point on the chromosome called the Origin of Replication ( ori ). The DNA in this region will be opened up, ‘unzipped’ so that the replication machinery can gain access to single-stranded parental DNA, which will act as template for synthesis of the new daughter strands. The two replisomes then travel in opposite directions around the circular prokaryotic chromosome, each replisome forming a replication fork, a schematic representation of one replication fork is shown in Figure 6 .

A single replication fork showing the leading and lagging strands. The leading strand is synthesised continuously, reading the template 3′ to 5′, synthesising DNA in the 5′ to 3′ direction. The lagging strand is synthesised discontinuously, in short Okazaki fragments (1000 bases in prokaryotes and 100 bases in eukaryotes).

The replication fork

Within the replication fork, on the so-called leading strand, DNA polymerase moves 3′ to 5′ with respect to the template and synthesises DNA in the 5′ to 3′ direction as it moves in the same direction as the replication fork. Although overall the lagging strand is synthesised in the 3′ to 5′ direction, it is actually synthesised discontinuously in small segments called Okazaki fragments, which are synthesised 5′ to 3′ ( Figure 6 ). Each Okazaki fragment will be started with an RNA primer and is synthesised in the opposite direction to the movement of the replication fork. In prokaryotes, Okazaki fragments are 1000–2000 bases in length. In Figure 6 you will see that the DNA polymerase synthesising the Okazaki fragment will eventually reach the primer for the previous Okazaki fragment. When this happens the primer for the previous fragment is removed by a DNA polymerase using 5′ to 3′ exonuclease activity. DNA polymerase then replaces the missing nucleotides by adding them to the 3′ end of the last Okazaki fragment. When all the primer has been removed, there will be two DNA strands adjacent to each other but not joined by a phosphodiester bond, these two strands are joined together by the enzyme DNA ligase.

The replisome contains a number of other important proteins required for DNA replication. The double-stranded DNA needs to be separated, ‘unzipped’, by a helicase to generate the single-stranded DNA templates for DNA polymerase. As the replication fork moves along the helical DNA, the coils in the DNA in front of the fork become compressed so the DNA is described as being overwound; a topoisomerase is required to ‘relax’ it by remove the over-winding. Single-stranded binding proteins (SSBs) bind the lagging strand template to stabilise and protect the single-stranded DNA.

The two replication forks that form at the ori will move in opposite directions around the circular prokaryotic genome until they reach the terminator sequence, ter , which is on the opposite side of the genome compared with the ori , i.e. it is at 6 o’clock compared with 12 o’clock. This results in the complete replication of the genome. Once DNA replication has been completed a post-replication DNA repair process will correct errors that were not corrected by the proofreading activity of DNA polymerase. The fidelity of DNA replication is extremely high, resulting in an error rate of 1 mistake per 10 9 –10 10 nucleotides added.

DNA replication in eukaryotes

DNA replication is essentially the same in eukaryotes and prokaryotes. In both cases two replisomes form at an ori and generate two replication forks moving in opposite directions away from the origin. In each replication fork there are leading and lagging strands. There are two major differences. The first is that, due to the larger genome size, each chromosome has multiple origins of replication, so there will be a large number of replication forks on each chromosome.

The second difference is that, with the exception of mitochondrial DNA, eukaryotic chromosomes are linear and this results in an issue because of lagging strand synthesis. Replication of a linear chromosome results in shortening of one 5′ end of each daughter DNA molecule. This is because when the primer required for the last Okazaki fragment is removed, DNA polymerase cannot fill the gap ( Figure 7 A). Repeated rounds of DNA replication results in shorter and shorter DNA molecules. If this is not corrected, eukaryotes would have become extinct as their chromosomes get shorter with each generation. Eukaryotes have a mechanism to preserve the ends of chromosomes when it counts; that is in the gametes. The terminal ends of chromosomes, telomeres, contain a highly repeated sequence, for example, in humans the sequence TTAGGG is repeated in tandem 100 to over 1000-times. Repeated rounds of DNA replication will result in the shortening of these telomeric sequences that is the number of repeats will reduce. Telomerase, an RNA containing enzyme, can add additional copies of the repeat sequence to the 3′ end, replacing those lost during DNA replication (see Figure 7 ).

( A ) Following DNA replication and removal of the primer for the last Okazaki fragment of the lagging strand, there will be a region at the 3′ end that is not base paired, called a 3′ overhang. ( B ) Telomerase binds and uses the RNA it contains to act as a template to extend the 3′ overhang. This extends the 3′ end sufficiently for a new RNA primer to bind and the final Okazaki fragment to be made.

This actually extends the 3′ end of the telomere rather than extending the 5′ that is initially lost during DNA replication. The RNA sequence within telomerase is complementary to the 3′ telomeric sequence and so can bind and act as a template for synthesis of a short DNA sequence. Telomerase then moves along the newly synthesised strand and the process is repeated. Multiple rounds of elongation and translocation ultimately results in the 3′ end being extended so that it is long enough for it to act as template for synthesis of another Okazaki fragment, hence extending both strands of the telomere. Only germ cells and a few other actively dividing cells (e.g. haematopoietic cells) have sufficient levels of telomerase activity to counteract the loss of repeat sequences during DNA replication. At birth, telomeres are over 10000 base pairs in length and there are enough repeats to allow DNA replication and somatic cell division during the lifetime of the organism. If telomeres become too short this will trigger programmed cell death (a process called apoptosis). The lack of telomerase activity in somatic cells limits the number of cell divisions that can occur, and this is a ‘problem’ that needs to be overcome by cancer cells. Telomerase activity is reactivated in most cancers, allowing these cells to divide indefinitely and therefore this activity is a potential target for cancer therapies.

An understanding of DNA synthesis is central to many experimental approaches in molecular biosciences, it allows us to determine DNA sequences including that of the human genome, to analyse environmental samples to better understand the living world around us and to analyse minute biological samples from crime scenes to identify offenders. It is exploited in medicine, for example several drugs used to treat HIV infection or exposure are nucleoside analogues that inhibit DNA synthesis. Many chemotherapy agents used to treat cancer target DNA replication.

The genetic code and the concept of a gene

As we have seen in the previous two sections, the genetic material in a cell is made of DNA and can be copied and passed on to progeny through DNA replication allowing for inheritance of the information that it carries. A large proportion of the information on the DNA is first transcribed into mRNA and then translated into proteins. However there are some RNAs that are never translated into proteins and these have important functions too. Phrases like ‘it is in my genes’ or ‘in my DNA’ are used in common speech to mean to be an important part of who someone is.

The term gene was coined in the early 1900s to describe the basic unit of heredity. Genes were thought of as distinct loci arranged lineally on chromosomes. Breeding experiments with the fruitfly Drosophila supported this view and showed that if two genes are close together on a chromosome they are more likely to be inherited together. The observation that mutations in genes could give rise to altered phenotypes gave rise to the ‘one gene one polypeptide’ hypothesis. Once it became clear that genes were made of DNA, what is referred to as the central dogma of molecular biology was coined. This describes a two step process in which the genes on the DNA are transcribed into RNA and then translated into a sequence of amino acids that makes up a protein. The information flow is from DNA to RNA and then to protein ( Figure 8 ).

The arrows represent steps where DNA or RNA is being used as a template to direct the synthesis of another polymer, either RNA or protein.

However there are exceptions to this, firstly some viruses have RNA genomes and in some cases these are reverse transcribed into DNA before the genes can be expressed. The retrovirus HIV is an example of this. The other exception is that not all functional RNAs are translated into proteins (see non-coding RNAs below).

The genetic code

The genetic code is the set of rules used by living cells to translate the information encoded within genetic material into proteins. When DNA and RNA were first discovered, the relative simplicity of nucleic acids led many scientists to doubt that it carried the genetic information. DNA only has four different kinds of bases; the question was how it could code for 20 amino acids. If there were a 1:1 correlation between bases and amino acids DNA could only encode four amino acids. Pairs of bases would give 16 possible combinations which is still not enough. However if you consider a triplet code you have 64 possibilities, which is more than enough. This is the code that we are familiar with where each codon, a sequence of three nucleotides, specifies a particular amino acid. This triplet code still did not seem logical because now you have far more codons than you need. There are some other important questions about the genetic code too; are the spare codons used? Is the code overlapping? And is it continuous or are there spacers indicating the end of each codon?

Table 1 shows the genetic code as we now understand it. It is written as RNA with a U rather than a T because it is RNA that cells translate into amino acids. The code is said to be redundant or degenerate because a single amino acid is often coded for by more than one codon. In most cases it is the third nucleotide in the codon that differs; this is often referred to as the degenerate position.

Evidence for the triplet code

The experiments that allowed scientists to decipher the genetic code were carried out long before we were able to determine the sequence of DNA. While it was possible at that time to determine the proportions of each different amino acid in a protein, it was not yet possible to work out the order in which they occurred. Francis Crick and Sydney Brenner answered some key questions with an experiment using mutants of a virus that infects bacteria called bacteriophage. The normal or wild-type phage will infect E. coli and grow. Crick and Brenner investigated mutants that would not grow on some strains of E. coli .

Mutants which are insertions or deletions cause what are called frameshifts. Inserting a single adenine base into the DNA sequence not only changes the amino acid at the position of the insertion but all subsequent amino acids translated from that sequence (Compare Figures 9 A and B); the reading frame has been shifted by one base and it results in a protein that is non-functional. However if you insert three nucleotides you often get a wild-type or near wild-type phenotype. This is because you have inserted a whole triplet codon, you will get one or two amino acids that were not in the original sequence but the reading frame is not shifted ( Figure 9 C) and the rest of the sequence is normal.

( A ) Wild-type sequence, ( B ) a single base insertion (shown in red) causes a frameshift so all subsequent amino acids are different from the wild-type, ( C ) insertion of three base pairs (shown in red) causes two incorrect amino acids to be incorporated into the protein but there is no frameshift so the rest of the protein has the wild-type sequence.

Crick and Brenner were looking for what they called suppressor mutations that would rescue the mutant and allow it to grow normally. They showed that their suppressor mutants did not simply reverse the original mutation; they often added or subtracted one or more bases. They worked out that if you insert or delete one, two or four nucleotides then you see a mutant phenotype. However, if you insert or delete three nucleotides, this has little or no effect. This was strong supporting evidence for a triplet code. This is also evidence for a redundant code where the same amino acid can be coded in more than one way. If the code were non-redundant there would be 20 codons that code for amino acids and 44 that are ‘nonsense’ codons. In this case inserting three nucleotides would be most likely to introduce a nonsense codon and not restore the wild-type. Crick and Brenner proposed correctly that the genetic code is read from a fixed starting point and the bases are read in groups of three.

Cracking the code

At about the same time two American scientists Marshall Nirenberg and Heinrich Matthaei had developed a cell-free system which could synthesise proteins in a test tube when provided with an RNA molecule. They showed that when provided with an artificial RNA chain composed only of uracil (polyuracil) the system made a polypeptide composed entirely of phenyalanine residues. They now had a tool that they could use to crack the genetic code. RNA composed of cytosine (C) residues directed the synthesis of polyproline and RNA composed of adenosine (A) made polylysine. Experiments with combinations of nucleotides demonstrated that, for example, if you make RNA from A and C you produce proteins containing only six amino acids: asparagine, glutamine, histidine, lysine, proline and threonine. There are eight possible triplet codons that can be made from A and C, two of these we know encode proline and cysteine. The remaining four amino acids must be encoded by other combinations of A and C. This of course provides additional evidence for the redundancy of the genetic code.

These experiments using RNA molecules composed of random combinations of two or three bases were not enough to fully crack the genetic code. The use of chemically synthesised RNA molecules of known repeating sequence added some more important information. For example a synthetic RNA of alternating A and G residues (AGAGAGAGAG…) can be read as two alternating codons CAC and ACA. It encodes a protein of alternating histidine and threonine residues.

In the last section, we will discuss how tRNAs and ribosomes decode the genetic code and synthesise proteins. The final detail of the genetic code was determined by a technique using ribosome-bound tRNAs. Pieces of RNA as short as a single codon will bind to ribosomes and if amino acids attached to tRNA are added they will associate with the complementary RNA. If you then filter the solution you trap only the tRNAs that are bound to the ribosome, these are the ones specified by the codon in your RNA.

Start and stop codons

Of the 64 possible codons, 61 encode amino acids. The three remaining codons: UAA, UAG and UGA do not code for an amino acid, they are sometimes called nonsense codons. They are stop codons; when the ribosome encounters these protein synthesis stops. The AUG codon encodes the amino acid methionine but it is also the most common start codon. As you will see in the last section, the first residue in eukaryotic proteins is always a methionine and in prokaryotes it is a modified amino acid N-Formylmethionine.

Expanding the genetic code

Nature uses a small set of amino acids to make proteins, however if we were able to engineer cells that could use a wider range of building blocks with different physical and chemical properties it would be possible to make novel materials some of which could have useful therapeutic properties; this is one of the aims of synthetic biology. To do this successfully we need to reprogramme the genetic code and to engineer the translation machinery (see later section) to use these new combinations. Some progress has been made, for example in using both the UAG stop codon and the AAG codon for arginine to code for amino acids not normally found in proteins.

Current concept of the gene

Once the genetic code was cracked it was clear that a gene is a sequence of bases on a DNA molecule that codes for a sequence of amino acids in a polypeptide chain or for an RNA molecule with a specific function. The availability of DNA sequences (see ‘Recombinant DNA Technology and DNA Sequencing’ in this issue of Essays in Biochemistry ) of individual genes made it possible to look for patterns characteristic of genes. A gene that codes for a protein has a start codon followed by a series of codons that encode the amino acid sequence and then a stop codon; this is called an open reading frame.

Whole genome sequencing has provided biological data on an unprecedented scale. The need to analyse sequence data has led to the development of the field of bioinformatics; the analysis of these data to answer biological questions. One key concept used in bioinformatics is that of homology. Two organisms that have a common ancestor are said to be homologous and the same can be said of a structure or of a gene. For example limbs with five digits (the pentadactyl limb) are found not only in humans and other mammals but also in birds, reptiles and amphibians. The limbs are homologous, and this is evidence of a common evolutionary ancestor of all of these groups of animals. The same is true of genes. All vertebrates have red blood cells that contain haemoglobin, adult human haemoglobin is made from two α and two β globin molecules. The DNA sequence of the genes that encode globin molecules in vertebrates are all similar to each other and you can estimate how long ago two animals shared a common ancestor by looking at how similar their globin genes are. This principle can also be used to find genes in a new piece of DNA sequence; if there is a section of sequence that is similar to a known gene then it is likely to encode a homologous gene.

A gene is more than just the sequence that encodes the protein; it also includes sequences involved in regulation of gene expression such as promoter sequences that define where transcription starts and are the sites where proteins involved in transcription bind to the DNA. In bacteria, almost all genes are a single uninterrupted sequence of DNA. In eukaryotes the situation is more complicated because the coding region is usually interrupted by introns. The primary transcript is referred to as precursor or pre-mRNA, this contains both exons and introns. The introns are removed when the pre-mRNA is processed before it leaves the nucleus ( Figure 10 ) leaving the exons which are spliced together to make the mature mRNA. Eukaryotic mRNAs have a 5′ cap which is a methylated guanosine nucleotide added to the 5′ end of the mRNA by an unusual 5′ to 5′ linkage; this is important in initiating translation. At the 3′ end is the poly A tail, this is a chain of between 100 and 250 adenine residues added to the mRNA to increase its stability. Analysis of the human genome sequence suggests that there are approximately 20000–25000 protein-coding genes, however there are far more different proteins. This is because many genes are capable of encoding several variants of a protein. Alternative splicing allows for different combinations of exons to be included in the mature mRNA and genes can also have several alternative promoters and alternative poly A sites. It is thought that 95% of human genes are alternatively spliced.

The DNA includes an untranslated region at both the 5′ and 3′ ends as well as introns and exons. The codon where translation starts (green) and the stop codon (red) are shown. The DNA is transcribed into mRNA and is processed by addition of the 5′ cap, splicing out the introns and addition of the poly A tail. This mature mRNA is exported from the nucleus into the cytoplasm.

Non-coding RNAs

Only approximately 1.2% of the human genome codes for protein. However, if you compare the genomes of the human, the mouse and the dog you can see that much more of the genome is under what is called negative selection since the species diverged. Negative selection means that mutations which are disadvantageous are selected against. This suggests that more than just the protein coding regions affect the fitness of the organism carrying the DNA. Some of these are DNA sequences that are important in controlling gene expression (next section). However systematic screens are revealing large numbers of RNA transcripts that are processed but do not encode proteins. The most well-known are transfer RNAs and ribosomal RNAs both of which as we will see in a later section are fundamental to protein synthesis. However we are beginning to understand that there are other non-coding RNAs that carry out important cellular processes.

Two types of non-coding RNA, small inhibitory RNAs (siRNAs) and microRNAs (miRNAs) have a role in reducing gene expression after the mRNA has been transcribed from the DNA. They work by targeting a protein complex called RISC to specific mRNAs which it then degrades. Expression of the gene is specifically knocked out or reduced and the phenotypic effect of this can then be observed. Another group of non-coding RNAs play an important role in increasing the stability and correct folding of ribosomal RNAs. This process takes place in a compartment within the nucleus called the nucleolus; the RNAs are called small nucleolar RNAs (snoRNAs). These are mostly generated from intron RNA after it has been spliced out of the precursor mRNA and they function in association with proteins.

Modern concept of a gene

The modern concept of the gene has to take into account all of the complexity of mRNA processing including alternative splicing, regulatory sequences and polyadenylation sites as well as the plethora of non-coding RNAs. A definition of a gene that takes these factors into account would be that a gene codes for one or more transcripts that can function as an RNA or can be translated into one or more proteins.

Transcription

We have seen that a gene can encode either an RNA product or a protein sequence. The production of both requires the gene to be transcribed into RNA, either because the RNA is the final product or because the RNA will need to act as template for protein synthesis. RNA synthesis is very similar in prokaryotes and eukaryotes, being catalysed by the enzyme RNA Polymerase. However, of the processes discussed in this article it is arguably the one that differs most between prokaryotes and eukaryotes. One difference is that in eukaryotes the whole process needs to occur in a chromatin context, so access to the DNA template is limited. Regulation of gene expression is a major facilitator of cell differentiation, homoeostasis and speciation. Different cell types turn on transcription of different genes giving rise to their differentiated phenotypes. If we look at mammals as an example of speciation, they all have roughly the same gene content; it is how transcription is regulated that has changed as mammals have evolved. For example, if you compare humans and mice, the important changes to the human and mouse genome sequence that have occurred since they diverged from a common ancestor, are predominantly in the sequences that control transcription rather than in protein coding sequences.

RNA polymerase

DNA-dependent RNA polymerases are responsible for transcription of DNA into RNA. Like DNA Polymerase, RNA polymerase requires a DNA template and nucleoside triphosphate precursors. RNA polymerase does not require a primer. During RNA synthesis, the base within the incoming nucleoside triphosphate pairs with the base on the DNA template, a phosphodiester bond is formed, and pyrophosphate is released. RNA polymerase synthesises RNA in the 5′ to 3′ direction, because it can only add nucleotides on to the 3′ end of the chain. During transcription only one DNA strand is transcribed into RNA.

Gene transcription

When a gene is transcribed, RNA polymerase will bind upstream from the start of the gene, it will unwind almost two turns of the DNA helix to form a transcription bubble, it will add nucleotides on to the growing RNA chain, the last 12 nucleotides to be added to the RNA chain will base pair with the DNA template, forming a DNA–RNA heteroduplex ( Figure 11 ).

As each nucleotide is added to the growing chain, the transcription bubble and the heteroduplex moves with respect to the DNA template. So, as RNA polymerase synthesises RNA, there is unwinding of the DNA template in front of the site of synthesis and rewinding of DNA once RNA polymerase has passed through. Once RNA polymerase has transcribed the gene, transcription will terminate. For some genes, transcription termination is signalled by a particular sequence within the DNA, a terminator sequence, which RNA polymerase recognises. In some cases, RNA polymerase requires the help of other protein factors to recognise the terminator sequence. Finally, many eukaryotic genes do not contain a specific terminator sequence; instead, termination of transcription is linked to other events, for example cleavage of the RNA prior to addition of the polyA tail. Termination of transcription, leads to the dissociation of RNA polymerase from the DNA template and release of the RNA product. In prokaryotes, mRNA does not need processing before it can be translated, in fact, as will be discussed below, mRNA is translated as it is being made. However, the initial transcript in eukaryotes does need to be processed to produce a functional mRNA that can be exported to the cytoplasm for translation.

Control of transcription in prokaryotes

At many genes in prokaryotes, RNA polymerase can bind to the gene and initiate transcription without other protein factors. However, for most prokaryotic genes, the binding of RNA polymerase to the gene is controlled by transcription factors to ensure the correct genes are transcribed at the correct level within the cell. Upstream from the transcription start there will be a ‘promoter’ which contains specific DNA sequences that are recognised by RNA polymerase and transcription factors. Each gene will have a different promoter sequence and can be controlled by different transcription factors. A good example of this type of promoter is the promoter that controls the lac operon in E. coli ( Figure 12 ). Transcription factors that up-regulate transcription are called activators and those that down-regulate transcription are called repressors. In this example RNA polymerase on its own can bind the promoter and drive low levels of transcription. If the repressor binds it will stop all transcription and would override RNA polymerase and the activator. In the absence of the repressor, if the activator is present then it can drive high levels of transcription.

The different binding sites for transcription factors are shown on the DNA; ABS, activator binding site, RBS, repressor binding site. The left-hand panel indicates the presence of lactose and/or glucose in the environment, the right-hand panel indicates transcription levels.

The lac operon codes for genes required to use lactose and needs to be controlled in response to glucose and lactose concentrations. A repressor protein is responsible for responding to lactose concentration and an activator is responsible for responding to glucose ( Figure 12 ).

In the absence of lactose, the lac operon is kept in an off state by the repressor protein binding to the promoter and stopping transcription. If lactose is present in the cell, it will bind to the repressor and this stops the repressor binding the promoter, RNA polymerase can bind and drive low levels of transcription. If the cell is starved of glucose, the activator is turned on and this binds the promoter and helps RNA polymerase to initiate transcription, resulting in high rates of transcription.

In the examples above, RNA polymerase on its own drives low levels of transcription. This might not be the case for all promoters, at some promoters RNA polymerase may not be able to bind and drive transcription without an activator protein. At other promoters RNA polymerase on its own will be able to drive high levels of transcription and a repressor protein would be needed to turn off transcription.

Control of transcription in eukaryotes

Control of transcription in eukaryotes has to occur on a chromosome which is condensed into chromatin ( Figure 13 ). In addition, transcription requires the assembly of a large multiprotein complex at the gene. This complex will contain RNA polymerase and several other general transcription factors (GTFs). The core promoter is a region that overlaps the transcription start and is the binding site for RNA Polymerase and the GTFs. In addition, there will be further control sequences, enhancers, that can be just upstream or several 1000 base pairs away from the core promoter. In the absence of activator proteins the chromatin structure will stop RNA polymerase and the GTFs binding to the core promoter. Here histone proteins act as generic repressors of transcription. In order for a transcription to be turned on activators will bind the enhancers and recruit co-activators which open up the chromatin structure and ensure the core promoter is not blocked by histone proteins. The activators and co-activators will then assemble RNA polymerase and the GTF at the core promoter and drive transcription initiation. Transcription factors will also ensure the chromatin structure across the whole gene is in a conformation that is suitable for transcription.

( A ) When a gene is in a silent state the surrounding DNA will be in condensed chromatin and the histones will epigenetic modifications which facilitate gene repression (red spheres). ( B ) A gene that is being transcribed will have activators bound to enhancer sequences, the activators recruit co-activators that acetylate the histone and add other epigenetic modifications that facilitate gene transcription (green spheres). The activator and co-activators will recruit RNA Polymerase and the GTFs to the core promoter.

Repressors are not normally required to block assembly of transcription complex at the core promoter, however, they are important in the regulatory patterns needed in complex multicellular organisms. Eukaryotes have repressor proteins which can block the action of a specific activator and ensure the activator is only active when required. Repressors can work in a number of ways including binding to DNA and blocking the binding of the activator to the DNA, stopping the activator interacting with other proteins required for transcription or by binding to the activator and keeping the activator in the cytoplasm.

Epigenetics

As discussed above transcription initiation in eukaryotes requires the opening up of the chromatin structure. This is facilitated by co-activator proteins that can move the relative position of the nucleosomes ( Figure 4 ) with respect to the DNA and hence make certain regions of the DNA more accessible. They can also add chemical tags to both the histone proteins and DNA ( Figure 13 ). These epigenetic modifications can affect whether a gene or genomic region is available for transcription or is transcriptionally silenced. Histones are acylated by enzymes which transfer an acetyl functional group to from acetyl-coenzyme A to lysine residues in the histone protein. This is linked to activation of transcription because it reduces the positive charge on histones and therefore reduces their affinity for the negatively charged DNA. Acetylation can also act as a tag that is recognised by other proteins that drive gene transcription. This modification of the DNA is described as epigenetic because it affects gene expression rather than the genetic code itself. Conversely some repressor proteins will recruit co-repressors that deacetylate histones, increasing their affinity for DNA causing the chromatin to be highly condensed and leading to transcriptional silencing. Methylation of lysine residues is another epigenetic tag, a single lysine residue can have 1, 2 or 3 methyl groups added. Unlike acetylation, methylation of lysine residues does not change the positive charge. The consequences of histone methylation are more complex because depending on which lysine residue is methylated and the level of methylation, the tag may mark that region of the genome for transcription activation or repression.

DNA methylation is another important epigenetic modification which leads to transcriptional silencing of the genomic region that has been methylated. During differentiation in the developing embryo whole regions of the genome will be methylated and therefore transcriptionally silenced. The DNA methylation patterns are maintained during cell division and future generations of that cell.

Analysing transcription on a global scale

For many years individual scientists would study the transcriptional regulation of their ‘favourite’ gene and so we gained an understanding of how individual genes were regulated in response to different development or environmental signals, for example the control of the lac operon in response to lactose and glucose. In the last 15 years, many techniques have been developed to allow us to study transcriptional control of genes within a cell. Using techniques such as ‘RNA-Seq’, we can isolate the total RNA from a cell and use high-throughput sequencing to catalogue the level of transcription of all genes. In the case of eukaryotes this will also show how they have been spliced. It is also possible to analyse the binding of transcription factors and study epigenetic changes within histone proteins across the genome using techniques such as ChIP-Seq. So, combining techniques such as RNA-Seq and ChIP-Seq we can determine when and where a protein factor is bound to DNA and study epigenetic changes in a particular cell type and the consequences in terms of gene transcription. In combination these techniques give a detailed picture of the factors that affect transcription; this has been used, for example to look at differences between cancer cells and normal cells from the same patient.

Transcription and disease

Transcription factors and promoters play major roles in health and disease, below are just a few examples to give an idea of their role in health and disease.

• The transcription factor p53 is a tumour suppressor protein, it guards against cancer and some human cancers have mutations that knock out p53 function.
• The drug Tamoxifen used in the treatment of breast cancer binds the oestrogen receptor inhibiting its function. The oestrogen receptor is a transcription factor that turns on the transcription of genes in response to oestrogen.
• Rett Syndrome is a neurodevelopmental disorder that affects approximately 1 in 15000 female births. It is due to mutations in a transcription factor that would normally repress transcription of specific genes, the mutations lead to inappropriate transcription of these genes.
• Cocaine use results in changes in expression of many genes, this can include epigenetic changes within genes involved in cognition and brain function. These epigenetic changes can be inherited and there is evidence that cocaine use by a father can result in epigenetic changes that result in male, but not female, offspring being cocaine resistant.

Translation of RNA into proteins

The key player in protein synthesis is the ribosome, a complex structure composed of RNA and proteins. The ribosome provides a framework that ensures that the mRNA and tRNA are correctly positioned enabling the deciphering of the genetic code. There are many other proteins that are important in protein synthesis; some of these are part of the ribosome and some are again correctly positioned by the framework of the ribosome. As we will see, the small subunit ribosomal RNA is a ribozyme; an RNA molecule with catalytic properties similar to those of enzymes. Ribosomal RNA can form a peptide bond between two amino acids.

Transfer RNA

The other nucleic acid that you need for protein synthesis is the tRNA. The tRNA molecule is single stranded and folds up into a characteristic structure by base pairing ( Figure 14 ). These act as adaptor molecules, each has an anticodon for a specific mRNA codon and each carries the amino acid specified by that codon. The anticodon has a complementary sequence to the codon on the mRNA.

( A ) Tertiary structure of the phenylalanine tRNA from yeast showing the anticodon (grey), the acceptor stem (violet) with the nucleotides CAA at the 3′ OH end (yellow). Image modified from ‘TRNA-Phe yeast’ Yikrazuul (licensed under CC BY-SA 3.0). ( B ) Clover leaf representation of the secondary structure of tRNA.

The enzymes which attach amino acids to tRNAs are called aminoacyl tRNA synthetases; they recognise a specific amino acid and the corresponding tRNA. The reaction also requires ATP, it is carried out in two steps:

In the first step the enzyme hydrolyses ATP releasing pyrophosphate (PP) and in the second it attaches the amino acid to the 3′ hydroxyl of the tRNA. Aminoacyl tRNA synthetase enzymes are highly specific, they recognise specific amino acids and will only attach them to the correct tRNA. This ensures correct coupling of amino acids and tRNA molecules which is just as important in ensuring the fidelity of protein synthesis as the matching of the anticodon to the codon by the ribosome. In addition this step is said to activate the aminoacyl tRNA as it not only produces the correct substrate for the ribosome but also provides much of the energy required for peptide bond formation during protein synthesis.

Structure of the ribosome

All living things contain ribosomes. The ribosomes in bacteria are slightly smaller than those found in eukaryotic cells ( Table 2 ) but the overall structure and the way in which they work are essentially the same. The 2009 Nobel Prize for Chemistry was awarded to three scientists, Ada Yonath, Thomas Steitz and Venkatraman Ramakrishnan, who used X-ray crystallography to solve the three-dimensional structure of the bacterial ribosome. The ribosome is composed of two subunits, the small subunit which reads the messenger RNA and the large subunit which forms the bonds between amino acids, adding them to the growing polypeptide chain. There are three important binding sites for tRNAs in the ribosome which are at the interface between the two subunits and only formed when the two subunits come together. These sites are shown on the image in Figure 15 , they are referred to as the acceptor or aminoacyl (A) site, the peptidyl (P) site where the peptide bond between amino acids is formed and the exit (E) site from which spent tRNAs leave the ribosome.

In ( A ) the new tRNA is delivered to the ribosome by elongation factor EF-Tu (purple). In ( B ) the amino acid on the incoming tRNA is brought close to the amino acid on the tRNA in the peptidly site to facilitate peptide bond formation (bright green) (Adapted from Goodsell 2010 , licensed under CC-BY-4.0 licence).

In addition to the ribosome, the mRNA and tRNA, there are a number of small proteins that are not part of the structure of the ribosome, but are required for protein synthesis: initiation factors, elongation factors and termination factors. The importance of these factors is illustrated by the inherited condition Vanishing White Matter Disease (VWM). This serious neurodegenerative disease which results in lesions in the white matter in the brain is due to mutations in one of the initiation factors.

Protein synthesis

During protein synthesis the ribosome brings together the amino acid charged tRNA and the mRNA, the codon and anticodon are matched and the amino acids are joined together in the correct sequence. There are three phases to this process: initiation where the ribosome assembles on the mRNA, elongation where the triplet code is read and amino acids are added to the growing peptide chain and termination where protein synthesis stops.

A complex of proteins called the cap-binding complex bind to the 5′ cap of the mRNA ( Figure 10 ) in the nucleus. The mRNA is then exported to the cytoplasm where it recruits initiation factors, tRNA charged with a methionine and the small (40S) ribosomal subunit. Initiation factors also bind and the small subunit scans along the 5′ untranslated region of the mRNA until it encounters the first AUG start codon (Figure 16A). This is recognised by the anticodon codon of the initiator tRNA, the large subunit then docks to give the translation complex. The 80S ribosome with the tRNA charged with methionine at the P site is now ready to accept the next tRNA ( Figure 16 B).

( A ) During initiation, the mRNA recruits a tRNA charged with a methionine and the small ribosomal subunit, ( B ) the large subunit then docks to give the translation complex, ( C ) a tRNA with an amino acid attached enters the A site, ( D ) the peptide bond is formed between the amino acid in the P site and the one in the A site. The effect is that the growing peptide chain is transferred to the incoming aminoacyl tRNA in the A site leaving an empty tRNA in the P site. ( E ) Finally, everything moves along the mRNA by one codon in a process called translocation so the peptidyl tRNA with the growing peptide chain attached moves to the P site and the spent tRNA to the E site from where it leaves the ribosome. ( F ) When a stop codon is in the A site, a termination or release factor enters the A site, ( G ) the peptide is released from the ribosome and ( H ) the two subunits of the ribosome disassociate and are recycled.

With initiation complete, the mRNA is in the correct reading frame with the A site empty and the next codon exposed. In the elongation phase an aminoacyl tRNA, one charged with an amino acid, is brought to the ribosome in a complex with an elongation factor and enters the A site. If the anticodon it carries is complementary to the exposed codon it is correctly positioned in the acceptor site and GTP is hydrolysed on the elongation factor ( Figure 16 C). A peptide bond ( Figure 17 ) is then formed between the C terminus of the amino acid in the P site and the N terminus of the amino acid in the A site, this reaction is catalysed in the peptidyl transfer centre of the large subunit of the ribosome. The effect is that the growing peptide chain is transferred to the incoming aminoacyl tRNA in the A site leaving an empty or spent tRNA in the P site ( Figure 16 D). Finally, the peptidyl tRNA with the growing peptide chain attached moves to the P site. This step is called translocation and the energy being provided by hydrolysis of GTP by the elongation factor EF-G. The spent tRNA moves to the exit site from where it can leave the ribosome. The mRNA moves so that the next codon is exposed in the A site ( Figure 16 E) ready to accept a new aminoacyl-tRNA charged with another amino acid. During the elongation phase the ribosome cycles through this process, adding amino acids to the growing peptide chain until a stop codon is exposed in the A site. The new protein emerges from the ribosome through an exit tunnel in the large subunit.

( A ) Amino acids consist of a carbon atom with an amine group (the N terminus), a carboxylic acid group (the C terminus) and a variable R group. The simplest R group is a methyl group giving the amino acid alanine. ( B ) When two amino acids are joined together a peptide bond is formed between the N terminus of one amino acid and the C terminus of another. This is a condensation reaction releasing one molecule of water.

Termination

The stop codon is not decoded by being recognised by an anticodon on a tRNA. Instead it is detected by proteins called termination or release factors. In eukaryotes there is a single release factor (RF1) that recognises all three stop codons enters the A site ( Figure 16 F). The ester bond linking the peptide chain to the tRNA in the P site is broken and the peptide is released from the ribosome ( Figure 16 G) The two subunits of the ribosome disassociate and are recycled ( Figure 16 H).

The structure and function of ribosomes are highly conserved with a large core of structurally conserved proteins and rRNAs found in both eukaryotic and prokaryotic ribosomes. However, there are some differences both in the rRNAs and in some of the additional proteins involved in translation ( Table 2 ). The elongation phase is highly conserved but there are important differences in how protein synthesis is initiated. Bacterial mRNAs have a specific sequence called the ribosome binding site or Shine–Dalgarno sequence. In order to ensure that the mRNA is correctly positioned in the ribosome the Shine–Dalgarno sequence binds to a complementary sequence of the 16S rRNA in the small subunit. In bacteria the initiator tRNA is charged with a modified amino acid N-Formylmethionine.

Differences between the structure of bacterial and eukaryotic ribosomes can be exploited by antibiotics which are selective in that they affect protein synthesis in bacteria but not in mammalian cells. Macrolide antibiotics like erythromycin, block the exit tunnel in the large subunit of bacterial ribosomes and halt protein synthesis. The exit tunnel in eukaryotic ribosomes is slightly narrower which means that eukaryotic ribosomes are not affected. Streptomycin, an important antibiotic in the treatment of tuberculosis binds to the 16S of bacterial ribosomes. This distorts the structure of the decoding site and results in misreading of the mRNA.

Polyribosome

Protein synthesis can proceed very quickly, particularly in rapidly growing cells or those that are differentiating. In bacteria between 15 and 20, new peptide bonds can be formed per second. In eukaryotes it is slower, more like five peptide bonds per second. A small human protein like insulin would take only 10 seconds to make whereas the largest human protein titin, which is found in human muscle cells, takes about an hour and a half per molecule. One of the mechanisms that ensures that protein synthesis is carried out efficiently is the polyribosome. As soon as one ribosome has started translation another ribosome binds to initiate synthesis of another protein copy. This gives rise to polyribosomes or polysomes which can be seen by electron microscopy. Recent cryo-EM images show that ribosomes can be arranged very closely on the mRNA with the mRNA entry and exit channels aligned to allow the smooth passage of mRNA between them ( Figure 18 ). Sometimes these polyribosomes can form circular structures so that, as soon as the ribosome has finished synthesis of one polypeptide it can rebind the same mRNA molecule and start synthesis of another copy of the protein.

Cyro-electron micrograph reconstruction of eukaryotic polyribosome. Reprinted from ( Myasnikov 2014 ) by permission.

Closing remarks

The study of nucleic acids, from their first identification as the genetic material is littered with landmarks in molecular biosciences, many of them marked with Nobel Prizes. Since Watson and Crick proposed their structure of DNA our knowledge about DNA and how it works has expanded almost exponentially. The topics introduced in this article are important topics covered in all bioscience programmes; understanding them is key to all areas of biosciences from evolution and animal diversity to health and disease. Recent developments in the techniques that we can use to study DNA, often in living cells means that new and exciting developments in our understanding of the way nucleic acids work are occurring all the time. Given the scope of this article we have barely scratched the surface of the topic, however, the reader can find more detail from the articles in the bibliography below and even more detail from a few minutes searching on the internet.

Abbreviations

Competing interests.

The authors declare that there are no competing interests associated with the manuscript.

Recommended reading and key publications Nobel lectures

• Blackburn E.H. (2010) Telomeres and Telomerase: The Means to the End (Nobel Lecture) 49 , Int. Ed., pp. 7405–7421, Angewandte Chemie [ PubMed ] [ Google Scholar ]
• Ehrenberg M. (2009) Scientific Background on the Nobel Prize in Chemistry 2009 Structure and Function of the Ribosome , The Royal Swedish Academy of Sciences, https://www.nobelprize.org/uploads/2018/06/advanced-chemistryprize2009.pdf [ Google Scholar ]
• Kornberg R.D. (2007) The Molecular Basis of Eukaryotic Transcription (Nobel Lecture) 32 , Int. Ed., pp. 12955–12961, Angewandte Chemie [ PubMed ] [ Google Scholar ]

Review articles

• Afonina Z.A. and Shirokov V.A. (2018) Three dimensional organization of polyribosomes–a modern approach . Biochemistry (Moscow) 83 , S48–S55 10.1134/S0006297918140055 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Gerstein M.B, Bruce C., Rozowsky J.S., Zheng D, Du J., Korbel J.O.. et al. (2007) What is a gene, post-ENCODE? History and updated definition . Genome Res. 17 , 669–681 10.1101/gr.6339607 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kruglyak L. and Stern D.L. (2007) An embarrassment of switches . Science 317 , 758–759 10.1126/science.1146921 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Minchin S.D. and Busby S.J.W. (2013) Transcription factors . In Brenner’s Encyclopedia of Genetics (Maloy S. and Hughes K., eds), Elsevier, U.S.A. [ Google Scholar ]
• Roberts M. (2019) Recombinant DNA technology and DNA sequencing . Essays Biochem. 63 , 10.1042/EBC20180039 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Roeder R.G. (2003) The eukaryotic transcriptional machinery: complexities and mechanisms unforeseen . Nat. Med. 9 , 1239–1244 10.1038/nm938 [ PubMed ] [ CrossRef ] [ Google Scholar ]

Historical perspectives

• Dahm R. (2008) Discovering DNA: Friedrich Miescher and the early years of nucleic acid research . Hum. Genet. 122 , 565–581 10.1007/s00439-007-0433-0 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• McCarty M. (2003) Discovering genes are made of DNA . Nature 421 , 406 10.1038/nature01398 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Maddox B. (2003) The double helix and the “wronged heroine” . Nature 421 , 407–408 [ PubMed ] [ Google Scholar ]
• Kemp M. (2003) The Mona Lisa of modern science . Nature 421 , 416–420 10.1038/nature01403 [ PubMed ] [ CrossRef ] [ Google Scholar ]

Original research papers

• Crick F.H.C., Barnett L., Brenner S. and Watts-Tobin R.J. (1961) General nature of the genetic code for proteins . Nature 192 , 1227–1232 10.1038/1921227a0 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Franklin R.E. and Gosling R.G. (1953) Molecular configuration in sodium thymonucleate . Nature 171 , 740–741 10.1038/171740a0 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Meselson M. and Stahl F.W. (1958) The replication of DNA in Escherichia coli . Proc. Natl. Acad. Sci. U.S.A. 44 , 671–682 10.1073/pnas.44.7.671 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Watson J. and Crick F. (1953) Molecular structure of nucleic acid. A structure for deoxyribose nucleic acid . Nature 171 , 737–738 10.1038/171737a0 [ PubMed ] [ CrossRef ] [ Google Scholar ]

Citations for figures

• Goodsell D. (2010) Molecule of the month: ribosome . https://pdb101.rcsb.org/motm/121 [ Google Scholar ]
• Myasnikov A.G. (2014) The molecular structure of the left-handed supra-molecular helix of eukaryotic polyribosomes . Nat. Commun. 5 , 5294 10.1038/ncomms6294 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Yikrazuul X.X. (2010) tRNA-Phe yeast . https://commons.wikimedia.org/wiki/File:TRNA-Phe_yeast_1ehz.png [ Google Scholar ]