H5N1 influenza binding and cell entry via human class II MHC, and blocking by cross-reactive antibodies

Emerging H5N1 Clade 2.3.4.4b Uses Human HLA-DR as a Secondary Cell Entry Receptor

Scientists have discovered that new H5N1 bird flu viruses can enter human cells using a protein called HLA-DR. This protein is part of the MHC-II complex (a group of molecules used by the immune system to present antigens to T cells). This is different from how most flu viruses work. Most typically rely on sugars called sialic acids. This discovery provides a new potential mechanism for understanding host range and spillover risk.

For over a century, the prevailing model for influenza infection has been relatively simple. The virus uses its hemagglutinin (HA) protein to grab onto sialic acid receptors on the surface of respiratory cells. Think of sialic acid as a specific type of lock. The viral HA acts as a key designed to turn it. While some specialized viruses, like those found in bats, use MHC-II molecules as alternative locks, this was not thought to be a major driver for contemporary H5N1 outbreaks.

However, the emerging H5N1 clade 2.3.4.4b has shown an unprecedented ability to jump between species. It has infected everything from poultry to dairy cattle and humans. A new study from researchers at Stanford and the University of Pennsylvania suggests the "sialic acid-only" model may be incomplete.

Beyond the Sialic Acid Paradigm

The researchers began with a surprising observation. A mutant version of the H5 hemagglutinin was engineered to lack sialic acid binding. Yet, this mutant still attached to a large portion of human B cells in blood samples [Figure 1a]. This suggested the virus was finding a different way into the cell.

To confirm this, the authors employed a systematic approach to isolate the true receptor. First, they used neuraminidase—an enzyme that acts like molecular scissors to snip off sialic acids—to strip away the traditional "locks" [Figure 1b]. Even after this treatment, the clade 2.3.4.4b H5 virus continued to bind to cells.

The study found that this residual binding could be stopped by using antibodies that block MHC-II molecules. Specifically, the researchers targeted the HLA-DR subtype. By testing various blockers, the authors demonstrated that HLA-DR blockade effectively prevented the virus from attaching. Blocking other MHC subtypes like HLA-DQ or CD74 had no effect [Figure 1f]. This established that HLA-DR is a functional docking station for this specific strain of bird flu.

Mapping the Dual-Entry Mechanism

Once the authors identified HLA-DR as a potential receptor, they needed to prove it actually facilitated viral entry. Viral entry is the process of the virus penetrating the cell membrane to begin replication. They used GFP (Green Fluorescent Protein) reporter virus particles to test this. These particles glow green if they successfully infect a cell.

The mechanism follows a two-step logic: 1. Sialic acid-dependent entry: The virus uses its primary receptor to attach to common cell surfaces. 2. HLA-DR-mediated entry: In cells where sialic acids are depleted, the virus utilizes HLA-DR to bypass the missing sugar receptors.

The authors report that in HEK-293T cells where sialic acids were removed, the presence of HLA-DR "rescued" viral entry. For the H5/DairyCattle/2024 strain, they observed a 40.7% increase in successful entry when HLA-DR was present [Figure 2f]. This means nearly half of the entry events in those cells were mediated by the HLA-DR pathway. Conversely, adding an anti-HLA-DR antibody dropped entry levels back down to baseline [Figure 2g]. This provides evidence that clade 2.3.4.4b H5 viruses can utilize a dual-receptor strategy.

Genetic Determinants of Susceptibility

If HLA-DR is a gateway for the virus, then human genetic diversity should dictate susceptibility. The human HLA system is famously polymorphic (meaning it varies widely between individuals). The authors investigated this by analyzing the genomes and cell surfaces of 48 deceased organ donors.

The study found a massive range in how much H5 protein bound to different people's immune cells. Binding frequencies varied from 0.45% to 52.8% [Figure 4b]. This means some individuals showed over 100 times more binding than others. The researchers report a strong correlation between the amount of MHC-II on a person's cells and how much the virus binds [Figure 4d].

More importantly, they identified specific genetic "highways" for the virus. Certain HLA-DRB1 alleles, such as 04:07, 03:01, and 04:04, showed much stronger associations with viral binding than others [Figure 4e]. Interestingly, the response was highly strain-dependent. For example, the allele HLA-DRB111:01 showed almost no binding for the new H5 strain. Yet, it was one of the most responsive alleles for the older H2 influenza strain [Figure 4g, h]. This suggests that as the virus evolves, it reconfigures its interface with the human immune system.

Evolutionary Trade-offs and Immune Defense

One might assume that mutations helping a virus adapt to mammals would also improve its ability to use HLA-DR. However, the paper finds a complex trade-off. The authors examined known mammalian-adaptive mutations, such as E190D and Q226H. These are often seen when flu moves from birds to humans.

The study reports that these specific mutations actually reduce the virus's ability to bind to HLA-DR [Figure 3e, f]. This implies that the evolutionary pressure to optimize sialic acid binding might come at the cost of the HLA-DR entry pathway. This creates a landscape where the virus must balance multiple ways to enter a host.

Despite this complexity, the authors found a potential defense in the human immune response. By isolating B cells from individuals who had never been exposed to H5N1, they identified several monoclonal antibodies. These antibodies could block the HA–MHC-II interaction [Figure 5d]. Some of these were cross-reactive. This means they were likely produced by the body in response to older, seasonal flu viruses. This suggests that existing immune memory might provide some protection against H5N1 spillover by interfering with this secondary entry route.

Assessing the Risk

The findings presented here come with notable technical caveats. Because of biosafety constraints, the authors could not test actual infectious clade 2.3.4.4b H5 viruses. Instead, they relied on safer laboratory surrogates. Additionally, while they used neuraminidase to strip sialic acids, some residual sugars might remain. This could complicate the measurement of purely HLA-DR-dependent entry. Finally, they did not detect direct binding between purified, soluble HA and soluble HLA-DR in certain assays. This suggests the virus might require the physical structure of a living cell membrane to facilitate this interaction.

The verdict is that the "dual-receptor" capability is a biological feature of contemporary H5N1. It complicates our understanding of host range and susceptibility. For researchers and public health officials, the practical takeaway is clear. Pandemic preparedness and vaccine development should account for the fact that human genetic variation may create pockets of vulnerability. We can no longer treat influenza entry as a one-dimensional problem of sugar-binding. It is a multidimensional interaction with the very proteins meant to defend us.

Figures from the paper