Exposure to HIV-1 Directly Impairs Mucosal Epithelial Barrier Integrity Allowing Microbial Translocation

Picture a tight junction. Not as a term in a textbook, but as a physical thing — a molecular seal, a belt of interlocking proteins cinched around the top of every epithelial cell, locking it to its neighbors. That seal is what stands between the bacteria in your gut, the viruses in the genital tract, and the tissue underneath. It is, in the most literal sense, the last wall. Now consider this: HIV can dissolve it. Not by infecting the epithelial cells, just by touching them. That finding from Nazli and colleagues reframes something clinicians had noticed for years without being able to explain. People with HIV-1 infection show signs of intestinal barrier dysfunction — leaky guts, bacterial products in the blood, chronic immune activation. The assumption was that this leakiness came later, downstream of immune collapse or opportunistic infection. Nazli and colleagues asked a harder question: what if HIV itself is directly responsible? What if the virus walks up to the mucosal wall and tears it open before any of that? To test that, the team grew polarized, confluent monolayers of primary female genital epithelial cells and a standard intestinal cell line called T84, then measured transepithelial resistance, or TER — the electrical resistance across a sealed epithelial layer, which is essentially a readout of how tightly those junctions are holding. Monolayers above one thousand ohms per square centimeter were considered properly sealed and ready to use. Then they added HIV-1 to the apical surface — the luminal side, the side that faces the outside world — and watched what happened. What happened was fast and dramatic. TER dropped by 30 to 60 percent within 24 hours. The effect showed up across multiple viral strains: laboratory R5 and X4 strains, and clinical isolates, all produced significant TER decreases. One clinical isolate, strain 4648, reduced TER to just 33 percent of its pre-treatment value. The kinetics were sharp — a reduction was already visible at 2 hours, and by 24 to 48 hours, TER had fallen to roughly 40 percent of baseline in genital epithelial cells. The molecular correlates were equally clear: seven tight junction genes — claudin-1 through claudin-5, occludin, and ZO-1 — were downregulated by 2 to 17 fold. Immunofluorescence imaging showed that ZO-1, which normally traces a continuous lacey outline around each cell, was already disrupted at 4 hours and virtually absent from the apical junctions by 24 hours. The number of intact ZO-1 inter-nodal contacts collapsed from 203 in control monolayers to 17 in HIV-treated ones. Here is the result that changes the interpretive frame entirely. When Nazli and colleagues used UV-inactivated HIV — virus rendered non-replicating by ultraviolet light — they got essentially the same TER drop. Same tight junction disruption. Same loss of ZO-1 localization. Productive infection was not required. And when they added HIV to the basolateral surface — the tissue side — nothing happened. The barrier stayed intact. So this is not a systemic effect, not a consequence of cells dying, not a byproduct of viral replication. It is something specific to apical contact with the virus. Cell viability assays confirmed the cells were not dying: MTT assays showed no reduction in viability 24 hours post-exposure. The next question is obvious: what part of the virus is doing this? The paper tested two candidates — gp120, the envelope glycoprotein that studs the outer surface of every HIV particle, and Tat, a regulatory protein released by infected cells. Purified gp120 alone, at 0.1 micrograms per milliliter, reproduced the TER drop and disrupted ZO-1 staining. Purified Tat, at 1.4 micrograms per milliliter, did nothing. Then the team ran three orthogonal confirmation experiments. They pre-treated intact HIV with a neutralizing antibody against gp120 — clone 2G12 at 35 micrograms per milliliter — and the barrier-disrupting effect was significantly blocked. They exposed monolayers to an Env-defective HIV mutant, stripped of functional envelope protein but matched for p24 viral protein content at 79 nanograms per milliliter, and TER stayed flat. Wild-type virus at the same p24 level produced a significant TER decrease. Three different approaches, one consistent answer: gp120 is the trigger. But gp120 is not punching through the barrier directly. It is recruiting the cell's own inflammatory machinery to do it. After 24 hours of HIV-1 exposure, both T84 intestinal cells and primary endometrial cells upregulated a panel of cytokines. T84 cells elevated tumor necrosis factor-alpha, or TNF-alpha, along with interleukin-6, interleukin-8, and MCP-1. Endometrial cells elevated TNF-alpha, interleukin-6, MCP-1, interleukin-10, and interleukin-1 beta. To test whether TNF-alpha was doing the actual barrier work, the authors added recombinant TNF-alpha directly to T84 monolayers and reproduced the TER drop. Then they added a TNF-alpha neutralizing antibody before HIV exposure and found that the TER decrease was prevented. Normal mouse serum, used as a control, did not prevent it. The causal chain, experimentally supported at each link, runs: gp120 engages the epithelial surface, the cell upregulates TNF-alpha, TNF-alpha dismantles the tight junctions, and the barrier falls. Once the barrier falls, what crosses it? Nazli and colleagues ran translocation experiments using non-pathogenic E. coli added to the apical chamber six hours after treatment. Bacterial colonies appeared in the basolateral supernatant after 24 hours of HIV-1 exposure and after six hours of TNF-alpha treatment. The magnitude of bacterial crossing after HIV exposure was roughly 50 percent of what TNF-alpha alone produced. They also added lipopolysaccharide — LPS, the inflammatory fragment shed from bacterial cell walls and a major driver of systemic immune activation — to apical surfaces of HIV-exposed endometrial monolayers. Basolateral LPS levels were elevated by 47 percent compared to controls. And HIV itself crossed: infectious virus appeared in the basolateral compartment by six hours, representing 0.03 percent of the inoculum, accumulating to 0.08 percent by 48 hours. Nazli and colleagues call this small but significant — and those words are doing real work. The absolute numbers are modest, but the point is that anything crossing a barrier that is supposed to be impermeable represents a mechanistic failure, not just a statistical fluctuation. Pull back and look at the full picture. The mucosal epithelium lines the intestinal tract and the genital tract — the two sites responsible for more than 90 percent of HIV transmission globally. For years, the leakiness observed in HIV patients was treated as a downstream consequence of disease. This work shows it can be an upstream event: HIV, via its envelope glycoprotein gp120, triggers an epithelial inflammatory response that dismantles tight junctions and opens the gate, without needing to infect or kill a single epithelial cell. The replication-independent nature of this effect means it could begin at the moment of exposure, before any productive infection is established. That reframing has practical edges. Because gp120 initiates the cascade and TNF-alpha executes it, both represent potential intervention points — blocking the gp120-epithelium interaction or dampening the TNF-alpha response could preserve barrier integrity at the site of exposure. Neither of those strategies requires viral replication to be occurring. The virus just has to arrive. What Nazli and colleagues showed, at its core, is that HIV carries its own key to the mucosal wall. It doesn't need to break in. It knocks on the surface with gp120, and the wall dismantles itself from the inside. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.