Membrane proteins are classified primarily by how they attach to the cell’s lipid bilayer. The two broadest categories are integral membrane proteins, which are embedded in the membrane, and peripheral membrane proteins, which associate with the membrane surface without being embedded in it. From there, scientists further classify these proteins by their structure, how they’re anchored, and what they do. Understanding these categories matters beyond the classroom: membrane proteins make up more than 60% of current drug targets, making their classification central to modern medicine.
Integral vs. Peripheral: The Core Division
The most fundamental way to classify membrane proteins is by how tightly they’re connected to the lipid bilayer. Integral membrane proteins are permanently embedded in the membrane. They contain one or more regions that sit within the hydrophobic core of the bilayer, and removing them requires detergents that disrupt the membrane itself. Think of them as bricks built into a wall.
Peripheral membrane proteins sit on the membrane’s surface. They attach through weaker interactions: electrostatic attraction to the charged heads of membrane lipids, or shallow hydrophobic contact where the protein nestles into just the outer edge of the bilayer without passing through it. Because their attachment is less permanent, peripheral proteins can often be stripped off by changing the pH or raising salt concentration, no detergents needed. Some peripheral proteins cycle on and off the membrane as part of their normal function, acting more like sticky notes than bricks.
Three Types of Integral Proteins
Integral membrane proteins are further divided into three groups based on topology, meaning how the protein is physically arranged relative to the two sides of the membrane.
- Bitopic proteins cross the membrane exactly once. They have a segment outside the cell, a single stretch passing through the bilayer, and a segment inside the cell. Many cell-surface receptors, including receptor tyrosine kinases, use this design to relay signals from outside to inside.
- Polytopic proteins cross the membrane multiple times. Their protein chain weaves back and forth through the bilayer, sometimes making more than 20 passes. G protein-coupled receptors (a huge family of drug targets) and ion channels are classic examples. This serpentine structure lets them interact with molecules on both sides of the membrane and form pores or binding pockets within the bilayer itself.
- Monotopic proteins are permanently associated with the membrane but do not cross it. They’re anchored in just one leaflet of the bilayer. These are the least well understood structurally, and some sit in a gray zone where it’s debatable whether they’re truly integral or just tightly bound peripheral proteins. Resolving that ambiguity often requires detailed lab work.
Structural Classification: Helices and Barrels
When a protein crosses the membrane, the segment inside the bilayer needs to be compatible with the oily, hydrophobic interior. Proteins solve this problem in two main structural ways.
Alpha-helical transmembrane proteins are by far the most common. The membrane-spanning segment coils into a tight helix made of 17 to 25 amino acids, with hydrophobic (water-repelling) side chains facing outward to interact with the surrounding lipids. A single protein can have anywhere from one helix to more than 20, connected by loops that extend into the watery environment on either side of the membrane. Nearly all membrane proteins in human cells use this design.
Beta-barrel transmembrane proteins take a completely different approach. Instead of helices, multiple flat strands of protein fold into a barrel-shaped tube. These barrels are found almost exclusively in the outer membrane of certain bacteria (gram-negative species). Some barrels form open channels, like porins that let small molecules pass through. Others are filled in by amino acid side chains and function as enzymes or receptors, using the barrel mainly as a rigid scaffold to hold the protein in place.
Lipid-Anchored Proteins
Some proteins don’t span the membrane or even touch it directly. Instead, they’re tethered to the bilayer by a lipid molecule that’s chemically bonded to the protein. These lipid-anchored proteins are sometimes grouped with integral proteins (since they’re covalently attached) and sometimes treated as their own category.
The most well-known type is the GPI anchor (glycosylphosphatidylinositol). GPI-anchored proteins sit on the outer surface of the cell, attached to a lipid in the outer leaflet of the membrane through a complex sugar-containing linker at the protein’s tail end. On the inner surface of the membrane, proteins can be anchored by fatty acid chains added through processes called myristoylation (attachment of a 14-carbon fat) and palmitoylation (attachment of a 16-carbon fat). Each anchoring strategy places the protein on a specific side of the membrane and influences how it moves and functions.
Classification by Function
Membrane proteins can also be grouped by what they do rather than how they’re built. The lipid bilayer determines a membrane’s basic structure, but proteins handle virtually all of its active jobs. The major functional categories include:
- Transport proteins move molecules across the membrane. Channels form selective pores that let specific ions or small molecules flow through. Transporters and pumps use energy to push substances against their concentration gradient.
- Receptors detect signals from outside the cell. These transmembrane proteins bind a molecule (like a hormone) on the extracellular side and trigger a response on the intracellular side, acting as signal transducers rather than physically moving anything across the membrane.
- Enzymes catalyze chemical reactions at the membrane surface or within the bilayer itself. Some beta-barrel proteins in bacteria serve this role.
- Anchors and linkers connect the cell’s internal skeleton to the extracellular matrix or to neighboring cells, giving tissues their mechanical structure.
- Cell identity markers display carbohydrate chains or other molecular signatures on the cell surface that allow the immune system and other cells to recognize them.
Only transmembrane proteins can perform functions on both sides of the bilayer or move molecules across it. Peripheral proteins, by contrast, typically work on just one side.
How Scientists Identify and Classify Them
In the lab, the distinction between integral and peripheral proteins is partly operational. If a protein washes off with high salt concentrations or pH changes, it’s peripheral. If it stays put until you add detergents to dissolve the membrane, it’s integral. Detergents mimic the lipid environment and keep hydrophobic membrane-spanning regions stable once they’re pulled out of the bilayer. Solubilizing integral proteins typically requires detergent concentrations 5 to 100 times higher than needed for later processing steps, which is one reason membrane protein research is expensive and technically demanding.
From a DNA sequence alone, researchers can predict whether a protein will sit in a membrane by looking for stretches of hydrophobic amino acids long enough to span the bilayer. Hydropathy analysis, originally developed by Kyte and Doolittle, scores each amino acid for how water-repelling it is. Transmembrane segments show a characteristic hydropathy peak around 1.7 on the scale, while segments from water-soluble proteins peak near 0.2. Modern prediction tools like TMHMM and Phobius use sophisticated statistical models to scan a protein sequence and predict how many times it crosses the membrane. Phobius improved accuracy significantly over earlier tools by simultaneously predicting transmembrane segments and signal peptides (short tags that direct a protein to the membrane), reducing false classifications of signal peptides from 26% to 4%.
These computational tools allow researchers to classify membrane proteins across entire genomes without needing to purify each protein individually, which has been essential for estimating that roughly 20 to 30% of all genes in most organisms encode membrane proteins.