What is Dr. Zubair Khalid's research focus?

Dr. Zubair Khalid specializes in molecular virology, mRNA vaccine development, and computational biology, with a focus on avian pathogens like IBDV and Avian Reovirus.

Where is Dr. Zubair Khalid currently working?

Dr. Zubair Khalid is a Postdoctoral Research Associate at the University of Maryland (UMD), specifically within the Department of Animal and Avian Sciences.

Integral Proteins: Structure and Function

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-19

An integral membrane protein is a protein that is permanently embedded within or covalently anchored to the lipid bilayer, so that it cannot be released from the membrane by high salt or a change in pH. Integral proteins carry out the membrane's most specialized jobs, including selective transport, signal reception, and cell-to-cell recognition, which is why they make up a large fraction of the targets studied in cell biology and pharmacology.

Membranes are not passive grease. The lipid bilayer is a two-dimensional fluid that separates the cytosol from the extracellular space, and almost every function that depends on that separation is performed by proteins sitting in or across it. A cell that loses its integral membrane proteins loses its ability to sense the outside world, import nutrients, and maintain ionic gradients. This article explains what integral proteins are, how their transmembrane domains are built, how they are distinguished from peripheral and lipid-anchored proteins, and how researchers detect and extract them at the bench.

What Is an Integral Membrane Protein?

The defining property of an integral membrane protein is thermodynamic, not anatomical. These proteins partition into the hydrophobic core of the bilayer and remain associated with it after the membrane has been washed with solutions that strip away loosely bound proteins. To remove them, the membrane itself must be dissolved with a detergent, or the protein must be cleaved from its anchor.

Two broad classes fall under the integral umbrella. Transmembrane proteins physically span the bilayer, with portions exposed on both sides. Lipid-anchored proteins are tethered to the membrane by a covalently attached fatty acid or lipid group rather than by a membrane-spanning segment. Both classes are integral because both require detergent or enzymatic cleavage for release.

The biogenesis of these proteins is a major logistical problem for the cell. Integral membrane proteins must be inserted into the correct membrane with the correct orientation, and cells devote a large set of dedicated machinery to their insertion, folding, and quality control [1]. The signal recognition particle (SRP) recognizes hydrophobic signal sequences on nascent chains as they emerge from the ribosome, docks the ribosome at the endoplasmic reticulum (ER), and hands the growing chain to the translocation channel. The ER membrane itself helps time this handoff, promoting a prehandover conformation in which delayed GTP hydrolysis optimizes the transition from targeting to translocation [2].

Integral Proteins Function: What They Actually Do

The functional repertoire of integral membrane proteins is easiest to remember by category.

Transporters and channels. These proteins move ions, metabolites, and water across the bilayer. The facilitative glucose transporter family is a classic example. CsGTP5, a glucose transporter-related protein from the liver fluke Clonorchis sinensis, has 12 predicted transmembrane domains and mediates sodium-independent uptake of deoxy-D-glucose when expressed in Xenopus oocytes, with an apparent Km of 4.0 mM and a Vmax of 300 pmol per oocyte per hour [3]. Aquaporins are the dedicated water channels, and ion channels such as potassium and sodium channels are responsible for electrical signaling.

Receptors. G protein-coupled receptors (GPCRs) form the largest family of integral membrane receptors. Each GPCR crosses the bilayer seven times as a bundle of alpha helices, which is why the family is also called the seven-transmembrane receptor family. The transmembrane helices form the ligand-binding pocket and the conformational switch that relays the signal inward.

Enzymes. Many catalytic domains sit on one face of the membrane and are held there by a transmembrane anchor. The serine protease matriptase is expressed on the plasma membrane of many cell types and depends on its transmembrane domain for correct localization [4].

Adhesion and recognition molecules. Surface proteins mediate attachment and immune recognition. A proteomic survey of bovine spermatozoa identified beta-defensins, SPACA1, BSP-family members, ADAM-family members, and SLC-family transporters among the prominent surface-associated proteins involved in immune protection, adhesion, and fusion [5].

Energy transducers. Bacteriorhodopsin, the light-driven proton pump of Halobacterium salinarum, converts absorbed photons into a proton gradient across the purple membrane [6]. This is a rare case where a single integral protein performs an entire energy-conversion cycle.

The Lipid Bilayer and How Proteins Sit in It

The bilayer is a sheet roughly 5 nm thick, with hydrophilic phosphate head groups facing the aqueous compartments and hydrophobic acyl chains buried in the middle. This geometry sets a hard constraint on any segment of protein that crosses the membrane. A polypeptide must present a surface of nonpolar side chains to the acyl chains, because polar or charged residues in the hydrophobic core are energetically expensive to bury.

The membrane is also asymmetric. The outer leaflet and inner leaflet differ in lipid composition, and integral proteins are inserted with a fixed orientation that is preserved for the life of the protein. A receptor's ligand-binding domain stays outside. Its signaling domain stays inside. This asymmetry is established at the ER during insertion and is maintained through every subsequent membrane-trafficking step.

Alpha-Helical versus Beta-Barrel Transmembrane Domains

Two structural solutions dominate for spanning a membrane.

Alpha-helical transmembrane domains

Most integral membrane proteins of the plasma membrane and internal organelles cross the bilayer as alpha helices. A single stretch of about 20 to 25 hydrophobic amino acids is long enough to form an alpha helix that spans the approximately 30 angstrom hydrophobic thickness of the bilayer. The side chains spiral around the helix, and the nonpolar ones face outward toward the lipid while polar backbone groups pair with each other inside the helix. This satisfies backbone hydrogen bonding without exposing it to the hydrophobic environment.

Proteins can cross the membrane once (single-pass) or many times (multipass). The 12-transmembrane architecture seen in CsGTP5 and in the bacterial MATE-family efflux protein EmmdR is one of the most common multipass topologies [3][7]. EmmdR's 12-transmembrane-segment architecture was confirmed experimentally, and its insertion followed the "positive inside" rule, in which positively charged residues are enriched on the cytoplasmic side of the membrane. Deleting critical positive residues from a cytoplasmic loop altered the protein's topology [7].

Beta-barrel transmembrane domains

Beta-barrel proteins cross the membrane as a closed cylinder of antiparallel beta strands. Each strand contributes a hydrophobic face to the outside of the barrel and a hydrophilic face to the inside, creating a water-filled pore. Beta barrels are found in the outer membranes of bacteria, mitochondria, and chloroplasts, and they are the structural basis of many porins and outer-membrane transporters. They are rare in the plasma membrane of animal cells.

Hydropathy Plots and How Transmembrane Segments Are Predicted

A hydropathy plot is a graph that scores each window of a protein sequence for its average hydrophobicity. The most widely used scale is the Kyte-Doolittle scale, which assigns each amino acid a numeric value based on how favorably it partitions into a nonpolar solvent. The sequence is scanned with a sliding window, typically 19 residues long, and the average score is plotted against residue position.

A peak of sufficient height and width in a hydropathy plot is a strong predictor of a transmembrane segment. Because an alpha helix needs roughly 20 residues to span the bilayer, a hydrophobic stretch shorter than about 18 to 20 residues is usually not long enough to be a stable transmembrane domain. This is why the 20 to 25 residue range is the working rule.

Hydropathy prediction is a starting point, not proof. Modern topology tools such as TMHMM and DeepLoc are used alongside experimental methods, and cross-source comparison of membrane protein annotations shows that automated predictions disagree with expert curation in a meaningful fraction of cases [8]. A platform that harmonized structures from MPstruc, the RCSB PDB, the Orientations of Proteins in Membranes database, and UniProt identified 121 broad-group conflicts between two major resources, and expert review was needed to resolve them [8]. The practical lesson is that a predicted topology should be confirmed experimentally whenever the conclusion matters.

How Integral Proteins Are Studied at the Bench

Topology reporters. Dual-reporter PhoA-LacZ fusions exploit the fact that alkaline phosphatase is active only in the periplasm while beta-galactosidase is active only in the cytoplasm. By fusing the reporter to different loops of a membrane protein, researchers can map which loops face which side. This strategy was used to validate the 12-transmembrane architecture of EmmdR, together with immunofluorescence of spheroplasts to confirm periplasmic loop localization [7].

Cell-surface biotinylation. Membrane-impermeant biotin reagents label only proteins exposed on the cell surface. After lysis, streptavidin purification pulls down the labeled fraction. This approach demonstrated that a matriptase variant lacking its canonical start codon is still present on the plasma membrane and retains enzymatic activity, because translation initiates from a cryptic downstream signal peptide that restores membrane topology [4].

Heterologous expression. Expressing a candidate transporter in Xenopus oocytes and measuring radiolabeled substrate uptake is the standard functional assay for transport proteins. The CsGTP5 characterization used tritiated deoxy-D-glucose and showed saturable, sodium-independent uptake [3].

Split fluorescence topology reporters. These reporters fluoresce only when a fragment is delivered to a specific compartment, allowing single-molecule-level topology assignment in live cells. This method revealed that a catalytically dead mutant of the metalloprotease ZMPSTE24 traps a subpopulation of the tail-anchored protein IFITM3 in an atypical topology with a cytosolic C-terminus instead of the normal lumenal orientation [1].

Proteomics with membrane enrichment. Quantitative mass spectrometry on membrane-enriched fractions, combined with in silico topology tools, identifies surface-associated proteins at scale. A bovine sperm study identified 558 high-confidence proteins in the membrane-enriched fraction versus 2,311 in the total protein fraction, with 43 predicted surface-associated proteins in the membrane fraction [5].

Molecular dynamics simulation. All-atom simulations compare how a membrane protein behaves in its native bilayer versus in contact with synthetic materials. Simulations of bacteriorhodopsin showed preferential adsorption on the extracellular side of the membrane and preserved native electrostatic contacts on graphene, whereas a polyvinyl alcohol matrix produced a water-rich interface that disrupted key electrostatic interactions [6].

Extraction: Why Detergents Are Required

The single most useful diagnostic difference between integral and peripheral membrane proteins is how you get them off the membrane.

Peripheral membrane proteins attach by electrostatic interactions and hydrogen bonds to the polar head groups of lipids or to the exposed domains of integral proteins. They do not enter the hydrophobic core. A high-salt wash (typically 0.5 to 1 M NaCl) or a pH shift that changes surface charge disrupts these interactions and releases the protein into solution. The membrane itself stays intact.

Integral membrane proteins cannot be removed this way. Their transmembrane segments are buried in the acyl chain region, and no change in ionic strength or pH will pull a hydrophobic helix out of a hydrophobic bilayer. The membrane must be solubilized with a detergent. Detergents are amphipathic molecules that form micelles above their critical micelle concentration, and they displace lipids from around the transmembrane domain, forming a protein-detergent complex that stays in solution. Nonionic detergents such as Triton X-100 and nonyl glucoside, and zwitterionic detergents such as CHAPS, are common choices. The choice of detergent matters because some detergents strip essential lipids or inactivate the protein.

Lipid-anchored proteins occupy a middle position. They have no transmembrane segment, but they carry a covalent lipid attachment that inserts into the bilayer. They are not released by high salt, so they behave like integral proteins in extraction assays, but they can be released by treating with the enzyme that cleaves the anchor, such as phosphatidylinositol-specific phospholipase C for GPI-anchored proteins.

Integral versus Peripheral versus Lipid-Anchored Proteins

Property	Integral (transmembrane)	Integral (lipid-anchored)	Peripheral
Membrane association	Spans the bilayer with one or more hydrophobic segments	Covalently attached lipid inserts into one leaflet	Electrostatic and hydrogen bonding to lipid head groups or to other proteins
Contact with hydrophobic core	Yes, extensive	Yes, via the lipid moiety	No
Extraction by high salt or pH shift	No	No	Yes
Extraction by detergent	Yes	Yes	Not required
Extraction by enzymatic anchor cleavage	Not applicable	Yes, for GPI-anchored proteins	Not applicable
Typical examples	GPCRs, aquaporins, ion channels, glucose transporters, bacteriorhodopsin	GPI-anchored surface proteins, Src-family kinases, Ras	Spectrin, ankyrin, cytochrome c, many peripheral enzymes
Typical transmembrane segment length	20 to 25 hydrophobic residues per alpha helix	Not applicable	Not applicable

Note that GPI-anchored proteins are sometimes grouped with peripheral proteins in older textbooks because they can be released by an enzyme rather than a detergent. The modern and more accurate classification places them with integral proteins, because they are covalently attached to the membrane and are not removed by salt or pH alone.

Common Mistakes and Limitations

Confusing "integral" with "transmembrane." Transmembrane proteins are a subset of integral proteins. A GPI-anchored protein is integral but does not cross the bilayer.

Assuming a hydropathy peak proves a transmembrane domain. A hydrophobic stretch can also be a signal peptide, a membrane-insertion hairpin, or part of a buried hydrophobic core. Signal peptides are cleaved after targeting, and a non-cleavable cryptic signal peptide can restore membrane topology even when the canonical transmembrane domain is missing [4].

Treating topology as fixed. The IFITM3 example shows that a single protein can exist in more than one topological isoform, and that a chaperone or protease can stabilize the atypical form [1]. Topology is a regulated property in some systems, not a permanent stamp.

Assuming all membrane proteins need the same detergent. Mild detergents preserve activity for some proteins and inactivate others. There is no universal solubilization buffer.

Overrelying on prediction tools. Automated topology and localization predictors are useful, but disagreements between databases are common enough that expert review changes the answer in a substantial number of cases [8].

Forgetting that membrane proteins are hard to crystallize. This is why structural coverage of membrane proteins lags far behind soluble proteins, and why harmonized structural databases matter for the field [8].

Individual experimental systems behave differently, and any specific protein's behavior in a given assay needs to be validated empirically rather than assumed from family membership.

Quick Review

Integral membrane proteins are permanently embedded in or covalently anchored to the bilayer and require detergent or enzymatic cleavage for release.
Peripheral proteins attach by electrostatic and hydrogen-bond interactions and are released by high salt or a pH shift.
Transmembrane alpha helices need roughly 20 to 25 hydrophobic residues to span the bilayer.
Beta barrels are the alternative transmembrane fold and are found mainly in bacterial, mitochondrial, and chloroplast outer membranes.
Hydropathy plots predict transmembrane segments but must be confirmed experimentally.
Major functional classes are transporters, channels, receptors, enzymes, adhesion molecules, and energy transducers.
GPCRs cross the membrane seven times, and aquaporins form dedicated water channels.

Frequently Asked Questions

What is an integral protein?

An integral protein is any protein that is permanently embedded in or covalently anchored to a lipid bilayer and cannot be released by high salt or a pH change. The class includes both transmembrane proteins and lipid-anchored proteins.

What is an integral membrane protein made of?

It is a polypeptide chain whose transmembrane segments are enriched in hydrophobic amino acids such as leucine, isoleucine, valine, and phenylalanine. These nonpolar side chains face the lipid acyl chains, while polar backbone groups hydrogen bond to each other inside the helix or barrel.

How long is a transmembrane segment?

A typical alpha-helical transmembrane segment is about 20 to 25 hydrophobic residues long. This length is enough to span the roughly 30 angstrom hydrophobic core of the bilayer as an alpha helix.

Why do integral proteins need detergent to extract?

Their transmembrane segments are buried in the hydrophobic acyl chain region, so changing ionic strength or pH cannot pull them out. Detergents form micelles that displace lipids and keep the hydrophobic domain soluble as a protein-detergent complex.

Are GPI-anchored proteins integral or peripheral?

They are integral. GPI-anchored proteins are covalently attached to the membrane through a glycosylphosphatidylinositol anchor, so they are not removed by high salt or pH. They can be released by phosphatidylinositol-specific phospholipase C.

What is the difference between an integral protein and a peripheral protein?

The difference is how they attach and how they come off. Integral proteins enter the hydrophobic core or are covalently lipidated and need detergent or enzyme cleavage. Peripheral proteins bind at the membrane surface and come off with high salt or a pH shift.