What Are IFITM Proteins and How Do They Fight Viruses?

IFITM proteins are small membrane-embedded proteins that cells produce in response to interferon signaling, and their primary job is to block viruses from entering cells. They do this by physically altering the membranes of cellular compartments so that a virus’s outer envelope cannot fuse with them. The family includes several members, each stationed at slightly different locations in the cell, and together they form one of the earliest lines of defense against a remarkably wide range of pathogens. But the biology turns out to be more interesting than a simple shield: these proteins can sometimes help certain viruses get in, they carry genetic variants linked to severe illness in pandemics, and they play unexpected roles in bone formation and pregnancy.

The IFITM Family Members

Humans have three well-studied immunity-related IFITM proteins: IFITM1, IFITM2, and IFITM3. All three are switched on when cells detect interferons, the signaling molecules the immune system releases upon sensing a viral threat. A fourth family member, IFITM5, is expressed almost exclusively in bone-forming cells and is not involved in antiviral defense. A fifth, IFITM10, remains poorly understood.

Each immunity-related IFITM protein is tiny by protein standards. Structurally, they share a common layout: the portion that does most of the work sits embedded in or draped across the cell’s lipid membranes, with the tail end (the C-terminus) poking out into the space outside the cell or inside a membrane-bound compartment, and the front end (the N-terminus) facing the cell’s interior. Studies using antibody staining and enzyme-cleavage experiments confirmed this orientation, though earlier predictions had the protein threaded through the membrane twice, which turned out to be incorrect for most cell types.1PubMed Central. Interferon-induced transmembrane protein 3 is a type II transmembrane protein The same general layout holds across all the human immunity-related IFITMs.2PLoS ONE. A Membrane Topology Model for Human Interferon Inducible Transmembrane Protein 1

Within this shared architecture, the proteins differ in where they concentrate inside the cell. IFITM2 and IFITM3 accumulate in endosomes and lysosomes, the compartments deep inside the cell where many viruses are trafficked after they hitch a ride into the cell. IFITM1 tends to sit closer to the cell surface. This division of labor matters because different viruses enter cells at different depths: some fuse at the plasma membrane, and some wait until they are carried into acidic endosomes before attempting to merge with the host cell’s membrane.

How IFITM3 Gets to Where It Needs to Be

IFITM3 does not start out in the endosomes. It first arrives at the cell’s outer membrane, then gets pulled inward through a process called endocytosis. Researchers identified a specific sorting signal in IFITM3, a short stretch of amino acids near its front end, that acts like a zip code telling cellular machinery to internalize the protein and route it to late endosomes.3PubMed Central. Identification of an endocytic signal essential for the antiviral action of IFITM3 When researchers mutated this signal or disrupted the cellular machinery that reads it, IFITM3 got stuck on the cell surface and lost its ability to block viruses that enter through endosomes.

IFITM3 also plays a trafficking role for its sibling, IFITM1. When IFITM3 is present, it helps pull IFITM1 into the same deep endosomal compartments. When researchers silenced IFITM3 expression, IFITM1 drifted back toward the plasma membrane.4PubMed Central. IFITM1 and IFITM3 cooperate to restrict virus entry in endolysosomes So the family members do not work in isolation; they cooperate, and IFITM3 acts as something of a gatekeeper for the others.

The Physical Trick That Stops Viruses

Most enveloped viruses, the kind wrapped in a stolen piece of host-cell membrane, get into cells by fusing their outer envelope with a host membrane. This fusion event requires the two membranes to bend toward each other, merge partway (a stage called hemifusion), and then open a pore that releases the virus’s genetic cargo into the cell. IFITM proteins disrupt this process by making the host membrane physically resistant to the bending and merging a virus needs.

Biophysical experiments showed that IFITM3 stiffens the membrane it sits in and changes its curvature. A short helical segment of the protein, called the amphipathic helix, embeds in the inner leaflet of the membrane and increases lipid order while inducing curvature that opposes the transition from hemifusion to full fusion.5PubMed Central. Interferon-Induced Transmembrane Protein 3 Blocks Fusion of Diverse Enveloped Viruses by Altering Mechanical Properties of Cell Membranes Think of it as making the membrane too rigid and wrongly curved for a virus to punch through. A separate line of experiments confirmed that IFITM-expressing membranes are more molecularly ordered and less fluid, and that these proteins suppress fusion even before the hemifusion stage is reached.6PLoS Pathogens. IFITM Proteins Restrict Viral Membrane Hemifusion

The transmembrane domain of IFITM3 can also directly interact with certain viral fusion proteins. Work on influenza showed that IFITM3’s transmembrane region physically contacts the HA2 subunit of the influenza hemagglutinin protein, the part responsible for driving membrane fusion.7Virologica Sinica. Transmembrane domain of IFITM3 is responsible for its interaction with influenza virus HA2 subunit So IFITM proteins may fight viruses through both a generalized physical barrier, stiffer and less cooperative membranes, and a more specific protein-to-protein confrontation with viral machinery.

Fine-Tuning by Chemical Tags

Cells do not leave IFITM proteins unsupervised. Two chemical modifications act as opposing switches on IFITM3’s antiviral potency. S-palmitoylation, the attachment of fatty-acid chains to specific sites on the protein, enhances its membrane association and its ability to block viruses. The palmitoylation sites are highly conserved across vertebrate species, suggesting this regulatory mechanism is ancient.8PubMed Central. Palmitoylome profiling reveals S-palmitoylation-dependent antiviral activity of IFITM3

Working in the opposite direction, ubiquitination, the attachment of a small tag that often marks proteins for degradation, pulls IFITM3 away from the endolysosomal compartments where it needs to be and weakens its antiviral function.9PubMed Central. S-palmitoylation and ubiquitination differentially regulate interferon-induced transmembrane protein 3 (IFITM3)-mediated resistance to influenza virus The balance between these two modifications gives the cell a way to dial IFITM activity up or down. Too much IFITM activity could cause problems of its own, as we will see, so cells need to keep the system calibrated.

Which Viruses They Block, and Which They Do Not

The list of viruses restricted by IFITM proteins is long. Influenza A virus was among the first to be characterized, and it remains the most studied example. But experiments have confirmed restriction of HIV-1, dengue virus, West Nile virus, Ebola virus, Marburg virus, and SARS coronavirus, among others.10Journal of Biological Chemistry. IFITM proteins: Understanding their diverse roles in viral infection, cancer, and immunity Work on filoviruses and SARS-CoV showed that IFITM proteins differentially restrict these pathogens and can modulate which cell types a virus can infect, independent of whether the cell carries the virus’s receptor.11PLoS Pathogens. Distinct Patterns of IFITM-Mediated Restriction of Filoviruses, SARS Coronavirus, and Influenza A Virus

The general pattern is that IFITM proteins are effective against enveloped viruses that enter through endosomes. Viruses that bypass this route, or that are non-enveloped, tend to be less affected. Even among enveloped viruses, though, the degree of restriction varies: IFITM3 is especially potent against influenza and flaviviruses, while IFITM1 may contribute more against viruses that fuse at or near the plasma membrane.

The restriction also appears limited to early entry events. Experiments with respiratory viruses showed that while IFITM proteins potently blocked viral entry, they did not significantly influence later stages of the virus’s replication cycle once it was already inside the cell.12PubMed Central. IFITM Proteins That Restrict the Early Stages of Respiratory Virus Infection Do Not Influence Late-Stage Replication IFITM proteins are bouncers at the door, not internal security.

When IFITMs Help Viruses Instead

Here is where the story gets strange. While IFITM proteins block most viruses they encounter, certain human coronaviruses actually exploit them to get in. The common cold coronavirus OC43 uses IFITM2 and IFITM3 as entry factors, essentially co-opting the very molecules meant to stop it. Interferon treatment, which ramps up IFITM production, actually enhanced OC43 infection in lab experiments.13PubMed Central. Identification of Residues Controlling Restriction versus Enhancing Activities of IFITM Proteins on Entry of Human Coronaviruses

This enhancement is not a universal coronavirus phenomenon. SARS-CoV-2, for instance, is inhibited by IFITM overexpression in some experimental setups. But the picture with coronaviruses is consistently more complicated than with, say, influenza. Specific mutations in IFITM3 can flip it from a restriction factor to an enhancer for SARS-CoV or MERS-CoV entry. Three distinct mutation sites in the protein have been identified that can cause this switch: changes at positions near the front of the protein, or deletion of a short stretch at the tail end, converted the protein from virus-blocker to virus-helper for certain coronavirus spike proteins.

Experiments with endogenous (naturally present) versus overexpressed IFITM proteins revealed another layer. Silencing the natural level of IFITM2 in cells dramatically reduced SARS-CoV-1 replication, cutting viral output by more than 50-fold, suggesting that for some coronaviruses the baseline level of IFITM2 actually supports the virus’s lifecycle.14iScience. Endogenous and overexpressed IFITM proteins exert opposing effects on coronavirus infection The structural motifs a coronavirus uses to hijack IFITMs appear to be distinct from the motifs other viruses encounter when they are blocked, which may explain why the same protein can act in opposite directions depending on the pathogen.

Genetic Variants and Pandemic Severity

Not everyone’s IFITM3 works equally well. A common genetic variant known as rs12252-C produces a version of IFITM3 with reduced antiviral function. This variant has been linked to more severe outcomes during pandemic infections. In a study of Chinese patients during the 2009 H1N1 influenza pandemic, those carrying two copies of the C variant had roughly six times the risk of developing severe illness compared to those carrying the T version.15Nature Communications. Interferon-induced transmembrane protein-3 genetic variant rs12252-C is associated with severe influenza in Chinese individuals Ten of the thirteen patients who died in that cohort carried the CC genotype.

A meta-analysis looking across COVID-19 studies found a similar pattern: the CC genotype was associated with roughly double the risk of severe disease and about a fourfold increase in mortality risk overall. The association was strongest in Chinese populations, where the C allele is more common, while in people of European descent the relationship was weaker for severity but still present for mortality when looking at anyone carrying at least one copy of the C allele.16PubMed Central. IFITM3 rs12252 polymorphism and coronavirus disease 2019 severity: A meta‑analysis

A second variant, rs34481144-A, also reduces IFITM3 activity through a different mechanism. A review of the literature across multiple outbreaks found that IFITM3 variants are primarily associated with worse outcomes during infections with emergent viruses, pandemic H1N1 and zoonotic H7N9 influenza, SARS-CoV-2, Hantaan virus, and HIV, rather than seasonal strains the population has encountered for decades.17PubMed Central. IFITM3 variants point to a critical role in emergent virus infections This makes intuitive sense: during a novel pandemic, the innate immune barriers like IFITM3 bear a disproportionate share of the defensive burden because the adaptive immune system has no pre-existing memory to draw on.

Beyond Viruses

IFITM proteins were identified as antiviral factors, but their biological reach extends further. Experiments showed that IFITM1, IFITM2, and IFITM3 also restrict the intracellular growth of Mycobacterium tuberculosis, the bacterium that causes TB. When all three IFITMs were knocked down simultaneously, TB bacteria grew significantly better inside human immune cells. IFITM3 appeared to enhance the acidification of the compartments containing TB bacteria, essentially making the cellular traps more hostile.18PubMed Central. A Role for IFITM Proteins in Restriction of Mycobacterium tuberculosis Infection This finding expanded the IFITM story from purely antiviral to broadly antimicrobial.

The family’s non-immune member, IFITM5, plays a role in bone mineralization. It is expressed specifically in osteoblasts, the cells that build bone. While mice lacking IFITM5 do not have dramatic skeletal problems, specific mutations in the human IFITM5 gene cause osteogenesis imperfecta type V, a form of brittle bone disease. An additional mutation in the coding region produces even more severe symptoms.19PubMed. IFITM5 mutations and osteogenesis imperfecta Why a protein in this family would be co-opted for bone formation remains an open question.

The Pregnancy Balancing Act

Perhaps the most striking non-immune role involves the placenta. The outer layer of the human placenta, the syncytiotrophoblast, forms when individual placental cells fuse together into a single massive multinucleated structure. This fusion is driven by proteins called syncytins, which are ancient remnants of retroviral genes our ancestors integrated into their genomes millions of years ago. Because IFITM proteins block membrane fusion as their core function, they also inhibit syncytin-mediated trophoblast fusion.

Experiments confirmed that expressing IFITM1, IFITM2, or IFITM3 in placental trophoblast cells blocked their ability to fuse. Conversely, knocking down IFITMs allowed trophoblasts to fuse more readily, even in the presence of interferon.20Journal of Biological Chemistry. Interferon-induced transmembrane proteins inhibit placental syncytiotrophoblast formation and promote fetal demise IFITM1 specifically also inhibited trophoblast invasion, and elevated IFITM1 levels were found in placentas from pregnancies affected by interferon-mediated diseases.21PubMed Central. IFITM1 inhibits trophoblast invasion and is induced in placentas associated with IFN-mediated pregnancy diseases

This creates a genuine biological dilemma. During a viral infection in pregnancy, the body ramps up interferon production to fight the virus, which in turn boosts IFITM levels. Those IFITMs help protect the mother and fetus from infection, but they simultaneously impair the placental fusion that sustains the pregnancy. The system requires precise calibration: enough IFITM to keep viruses out, but not so much that placental development suffers.

Evolutionary Roots of the Family

IFITM genes are found across vertebrates, from fish to primates. A comprehensive survey identified 174 IFITM-related genes and 112 pseudogenes across 27 vertebrate genomes, and the family divides into three evolutionary lineages: the immunity-related IFITMs, IFITM5, and IFITM10.22PubMed Central. Evolutionary dynamics of the interferon-induced transmembrane gene family in vertebrates Each lineage appears to have originated from a different ancestral gene.

The immunity-related branch has expanded aggressively in primates and rodents, with many lineage-specific gene duplications and strong signals of positive selection, the hallmark of an evolutionary arms race with pathogens. IFITM5, by contrast, shows no signs of duplication or positive selection in any lineage, consistent with its conserved role in bone. The IFITM10 branch is most dynamic in aquatic vertebrates, hinting at functions tied to marine environments that remain unexplored.

Therapeutic Possibilities

Because IFITM proteins work against such a broad range of viruses through a physical mechanism rather than targeting a specific viral protein, they are attractive as potential starting points for broad-spectrum antiviral therapies. One promising avenue involves the amphipathic helix of IFITM3, the small segment responsible for stiffening membranes. Research into how cholesterol binds to this helix and regulates its activity has been seen as a stepping stone toward designing therapeutic peptides that could mimic its membrane-altering effects.23PubMed Central. Cholesterol Binds the Amphipathic Helix of IFITM3 and Regulates Antiviral Activity

An entirely different approach involves delivering IFITM3 protein to cells that need it. In a mouse model of Zika virus infection during pregnancy, researchers packaged IFITM3 into tiny membrane vesicles called exosomes and injected them into pregnant mice. The exosome-delivered IFITM3 crossed the placenta and suppressed Zika virus replication in both the mothers and their fetuses. Treated mothers had viral loads roughly 100-fold lower than untreated animals, and in the fetuses the virus was essentially eliminated.24PubMed Central. EVs Containing Host Restriction Factor IFITM3 Inhibited ZIKV Infection of Fetuses in Pregnant Mice through Trans-placenta Delivery This is still early-stage work in animal models, but it illustrates a concept: rather than designing a drug against one virus, you could boost the body’s own broad-spectrum machinery.

The complication, naturally, is the same balancing act the body already manages. Flooding cells with IFITM3 might protect against viruses but could also interfere with normal cell-to-cell fusion events, including in the placenta. And the coronavirus-enhancing effect means that a blanket boost in IFITM levels could theoretically worsen certain infections. Any therapeutic strategy would need to be carefully targeted in terms of timing, dose, and tissue delivery, challenges that are far from solved but that make the biology all the more interesting to watch.