[["sea". Although simplistic and incomplete, this model is still widely referenced today.
Contents |
Since the invention of the microscope in the seventeenth century it has been known that plant and animal tissue is composed of cells. The plant cell wall was easily visible even with these early microscopes but no similar barrier was visible on animal cells, though it stood to reason that one must exist. By the mid 19th century, this question was being actively investigated and Moritz Traube noted that this outer layer must be semipermeable to allow transport of ions.[1] Traube had no direct evidence for the composition of this film, though, and incorrectly asserted that it was formed by an interfacial reaction of the cell protoplasm with the extracellular fluid.[2]
The lipid nature of the cell membrane was first correctly intuited by Quincke, who noted that a cell generally forms a spherical shape in water and, when broken in half, forms two smaller spheres. The only other known material to exhibit this behavior was oil. He also noted that a thin film of oil behaves as a semipermeable membrane, precisely as predicted.[3] Based on these observations, Quincke asserted that the cell membrane comprised a fluid layer of fat less than 100 nm thick.[4] This theory was further extended by evidence from the study of anesthetics. Hans Meyer and Ernst Overton independently noticed that the chemicals which act as general anesthetics are also those soluble in both water and oil. They interpreted this as meaning that to pass the cell membrane a molecule must be at least sparingly soluble in oil, their “lipoid theory of narcosis.” Based on this evidence and further experiments, they concluded that the cell membrane might be made of lecithin (phosphatidylcholine) and cholesterol.[5] Interestingly, one of the early criticisms of this theory was that it included no mechanism for energy-dependent selective transport.[6] This “flaw” would remain unanswered for nearly half a century until the discovery that specialized molecules called integral membrane proteins can act as ion pumps.
Thus, by the early twentieth century the chemical, but not the structural nature of the cell membrane was known. Two experiments in 1925 laid the groundwork to fill in this gap. By measuring the capacitance of erythrocyte solutions Fricke determined that the cell membrane was 3.3 nm thick.[7] Although the results of this experiment were accurate, Fricke misinterpreted the data to mean that the cell membrane is a single molecular layer. Because the polar lipid headgroups are fully hydrated, they do not show up in a capacitance measurement meaning that this experiment actually measured the thickness of the hydrocarbon core, not the whole bilayer. Gorter and Grendel approached the problem from a different perspective, performing a solvent extraction of erythrocyte lipids and spreading the resulting material as a monolayer on a Langmuir-Blodgett trough. When they compared the area of the monolayer to the surface area of the cells, they found a ratio of two to one.[8] Later analyses of this experiment showed several problems including an incorrect monolayer pressure, incomplete lipid extraction and a miscalculation of cell surface area.[9] In spite of these issues the fundamental conclusion- that the cell membrane is a lipid bilayer- was correct.
A decade later, Davson and Danielli proposed a modification to this concept. In their model, the lipid bilayer was coated on either side with a layer of globular proteins.[10] According to their view, this protein coat had no particular structure and was simply formed by adsorption from solution. Their theory was also incorrect in that it ascribed the barrier properties of the membrane to electrostatic repulsion from the protein layer rather than the energetic cost of crossing the hydrophobic core. A more direct investigation of the membrane was made possible through the use of electron microscopy in the late 1950s. After staining with heavy metal labels, Sjöstrand et al. noted two thin dark bands separated by a light region,[11] which they incorrectly interpreted as a single molecular layer of protein. A more accurate interpretation was made by J. David Robertson, who determined that the dark electron-dense bands were the headgroups and associated proteins of two apposed lipid monolayers.[12][13] In this body of work, Robertson put forward the concept of the “unit membrane.” This was the first time the bilayer structure had been universally assigned to all cell membranes as well as organelle membranes.
Around the same time the development of the first model membrane, the painted bilayer, allowed direct investigation of the properties of a simple artificial bilayer. By “painting” a reconstituted lipid solution across an aperture, Mueller and Rudin were able to determine that the resulting bilayer exhibited lateral fluidity, high electrical resistance and self-healing in response to puncture.[14] This form of model bilayer soon became known as a “BLM” although from the beginning the meaning of this acronym has been ambiguous. As early as 1966, BLM was used to mean either “black lipid membrane” or "bimolecular lipid membrane".[15][16]
This same lateral fluidity was first demonstrated conclusively on the cell surface by Frye and Edidin in 1970. They fused two cells labeled with different membrane-bound fluorescent tags and watched as the two dye populations mixed.[17] The results of this experiment were key in the development of the "fluid mosaic" model of the cell membrane by Singer and Nicolson in 1972.[18] According to this model, biological membranes are composed largely of bare lipid bilayer with proteins penetrating either half way or all the way through the membrane. These proteins are visualized as freely floating within a completely liquid bilayer. This was not the first proposal of a heterogeneous membrane structure. Indeed, as early as 1904 Nathansohn proposed a “mosaic” of water permeable and impermeable regions.[19] But the fluid mosaic model was the first to correctly incorporate fluidity, membrane channels and multiple modes of protein/bilayer coupling into one theory.
Continued research has revealed some shortcomings and simplifications in the original theory.[20] For instance, channel proteins are described as having a continuous water channel through their center, which is now known to be generally untrue (an exception being nuclear pore complexes, which have a 9 nm open water channel).[21] Also, free diffusion on the cell surface is often limited to areas a few tens of nanometers across. These limits to lateral fluidity are due to cytoskeleton anchors, lipid phase separation and aggregated protein structures. Contemporary studies also indicate that much less of the plasma membrane is “bare” lipid than previously thought and in fact much of the cell surface may be protein-associated. In spite of these limitations, the fluid mosaic model remains a popular and often referenced general notion for the structure of biological membranes.