Neutrophil Extracellular Traps Directly Induce Epithelial and Endothelial Cell DeathA Predominant Role of Histones

Picture the lung in crisis. Infection lights up the airways, neutrophils rush in, and then—like a superhero move and a shrapnel blast rolled into one—they cast out webs. These are neutrophil extracellular traps, or NETs: sticky lattices built from DNA and studded with proteins like histones, neutrophil elastase, and myeloperoxidase. The idea is brilliant. NETs ensnare bacteria and fungi right where they land. But there’s a catch. Those same proteins that shred microbes don’t carry business cards. They don’t know friend from foe. And in the lung, where the alveolar-capillary barrier is a whisper-thin interface that keeps air and blood in their lanes, that can be deadly. In acute respiratory distress syndrome, or ARDS, that barrier fails; mortality often sits somewhere between forty and sixty percent. So the question becomes very concrete, very fast: are NETs helping, or are they punching holes in the wall they’re supposed to defend? Saffarzadeh and colleagues set out to answer a deceptively simple version of that question back in 2012. Do NETs directly kill the barrier cells of the lung—epithelial and endothelial—and if so, which parts of the NET do the damage? Is it the DNA scaffold itself, the granular enzymes, or is it the histones, the core proteins that wrap DNA inside our cells and make up a huge share of the NET mass? Start with the concept, then test it piece by piece. They began at the source. Human neutrophils were purified from healthy donors with old-school density gradients—clean preparations, more than ninety-six percent pure and about ninety-eight percent viable. Then they flipped the NETosis switch with a classic stimulus, phorbol myristate acetate, or PMA, at fifty nanomolar for four hours, and collected the expelled traps. That gave them raw NETs they could manipulate: leave the DNA alone, chew it up completely with DNase, or nibble it into fragments with micrococcal nuclease. They also boiled some preparations and, crucially, measured DNA content with a fluorescent dye so they could standardize doses. DNA alone served as a control. If the backbone was the toxin, it would show up. On the receiving end, they used a roster of lung barrier cells. Human A549 epithelial cells. Mouse MLE-12 epithelial cells. Endothelial cells from human pulmonary artery and umbilical vein. Even primary-like murine type II alveolar cells. They grew them to confluence—cells touching, forming a sheet the way an epithelium does—and then applied NETs at two doses that corresponded to about three point four and ten point one micrograms per milliliter of NET DNA. Some exposures were short, four hours. Others stretched to sixteen. Cell injury was read out in multiple ways: lactate dehydrogenase release into the medium, which flags membrane damage; markers of apoptosis like caspase activity and annexin V; and nuclear dyes that light up dead cells. Not just a single assay, but a pattern. Across that panel, the pattern was clear. NETs killed. The effect was dose-dependent—more NET, more death—and it spanned both epithelial and endothelial cells. A549 cells were classic canaries in this mine; they took a hit at the lower dose and a bigger one at the higher dose. Similar losses showed up in the endothelial monolayers. Pause here and tie it back to the lung. If a sheet of epithelial cells starts to slough when NETs wash over it, that’s the beginning of fluid seeping into alveoli, of oxygen struggling to cross into blood. It’s not the whole story of ARDS, but it’s one mechanistic piece you can hold in your hand. Then came the twist. If DNA were the villain, digesting it should rescue the cells. It didn’t. Whether the NETs were intact, sliced into oligonucleotides with micrococcal nuclease, or completely digested with DNase, the cytotoxic punch didn’t let up. And when they put DNA alone on the cells at the same concentrations used to dose NETs, nothing close to the same injury happened. The scaffold wasn’t the killer. That shifts the spotlight to proteins. Histones took center stage quickly. When the team incubated cells with purified histone type II-A—the bundled set of core histones used in many toxicity assays—cells died in a clean, dose-dependent arc. At one hundred to two hundred micrograms per milliliter, histones alone were enough to stop growth and compromise membranes. That’s important twice over. First, it shows histones can be lethal on their own. Second, it explains why chopping away the DNA backbone didn’t solve the problem: you didn’t remove the teeth; you just pulled out the wire. If histones are the problem, can you blunt them? Activated protein C, or APC, is a natural anticoagulant with protease activity that had been reported to clip histones and soften their toxicity. In this study, APC at one hundred nanomolar did exactly that when mixed with purified histones: histone toxicity declined. But when APC was pre-incubated with NETs, even as the ratio of APC to NET protein rose from one to five up to one to one, the NETs remained just as toxic. Gel analyses backed that up—no evidence that APC had chewed up histones within the NET complexes. Inside the trap, histones looked protected, tucked into a structure that kept proteases at bay. Other perturbations told the same story from different angles. Antibodies against particular histones reduced NET-mediated damage. The strongest protection came from antibodies to DNA and histone types H1, H2A, H2B, and H4. Antibodies to H3 or its citrullinated form didn’t shift the needle, suggesting not all histone surfaces are equally accessible or equally culpable inside NETs. And a simple, charge-based approach—adding polysialic acid, a long, negatively charged sugar—made a big dent in toxicity from both purified histones and intact NETs. Histones are intensely cationic; coat them with negative charge and they stick less to cell membranes. That’s chemistry you can feel. What about the classic granular enzymes that ride along on NETs, like elastase and myeloperoxidase? Here the picture got more nuanced. When the team digested NET DNA, elastase activity in the preparation actually increased—likely because cutting the web unfurled more enzyme. You could block elastase’s activity with an inhibitor, and it worked on the enzyme readout. But that didn’t translate into protection from NET-induced cell death. In other words, elastase is there, it’s active, but it wasn’t the primary driver of the cytotoxicity they saw in these assays. Myeloperoxidase, or MPO, was different. Inhibiting MPO gave a modest but real reduction in NET-induced epithelial damage. Not a cure-all, but a sign that oxidant chemistry layered on top of histone charge contributes to how NETs injure membranes. So far, that’s a lab bench story. The natural question is whether this holds water in a lung that’s inflamed, full of cytokines and surfactant and flowing cells. To get at that, Saffarzadeh’s group turned to a straightforward mouse model of acute lung injury. They dripped ten micrograms of Escherichia coli lipopolysaccharide into the trachea in a fifty-microliter volume and waited. At twenty-four hours, the bronchoalveolar space was teeming with neutrophils—the peak of recruitment—and the tissue told the tale. Immunofluorescence lit up extracellular chromatin co-localized with neutrophil elastase in the parenchyma. Staining for myeloperoxidase overlapped with citrullinated histone H3, a hallmark of NETosis. In places where these signals clustered, membranes looked disrupted; you could see patches where the epithelial lining wasn’t intact. They didn’t stop at pretty pictures. Bronchoalveolar lavage fluid, spun down and split into supernatant and pellet, gave them a way to separate free enzymes from those bound up in NETs. Digest the pellet with micrococcal nuclease and now you can measure elastase that had been trapped on chromatin. Over a time course from three to forty-eight hours after lipopolysaccharide, free elastase activity rose, but what grew more tellingly was the NET-associated fraction. By around twelve hours, the pellet—those chromatin-bound complexes—was carrying a larger share of elastase activity, and that persisted into the twenty-four hour peak. They even pulled neutrophils from that lavage and showed they could be driven to form NETs ex vivo with a PMA push. The traps are in the space, on the tissue, and they carry active cargo. Pulling all of that together, you get a coherent picture. In vitro, NETs kill lung epithelial and endothelial cells in a dose-dependent way, and that killing doesn’t go away when you chop up the DNA. It does lessen when you shield histones with antibodies or neutralize their charge with polysialic acid. It softens a bit when you damp down MPO. In vivo, in a classic model of lung injury, you see NETs in the right place at the right time, with enzymes and histone marks that match what you’d expect from activated neutrophils. None of that proves NETs are the sole cause of barrier failure in acute respiratory distress syndrome—real lungs are much messier than dishes—but it puts NETs, and especially their protein payload, squarely in the causal path. There are caveats, and the authors are candid about them. There’s no one way to collect NETs; how you stimulate, wash, and shear them changes what you get. Tissue NETs don’t always look like the swooping webs in textbook images; they can be short, rubbed-down chromatin fragments that still bind proteins. And translating a clean dosing experiment in a plate to the swirl of an injured lung requires humility. Surfactant, protease inhibitors in the lining fluid, and DNase present in vivo—all of that shapes what NETs can do and how quickly they’re cleared. Still, the mechanistic themes stand up. Protein components dominate NET-driven cytotoxicity. Histones are the heavy hitters. Myeloperoxidase adds oxidative stress on top. The DNA scaffold is a vehicle, not the warhead. And some interventions that sound good on paper behave very differently in the full NET context. Activated protein C can disarm purified histones but couldn’t defang histones locked in a NET. By contrast, approaches that either block specific histone surfaces—like antibodies to H2A, H2B, H4, or H1—or smear their charge, like polysialic acid, cut toxicity substantially in vitro. Elastase, despite being a poster enzyme for neutrophils, looked more like a bystander in these killing assays: active, measurable, but not the switch that spared or doomed cells. What does that mean for the bedside? It’s tempting to jump straight to therapies—histone scavengers, myeloperoxidase inhibitors, smarter protease control. And there is a through line here to ARDS, where the alveolar-capillary barrier is already under siege. But the lesson of this work is as much about specificity as it is about enthusiasm. Target the piece that does the damage. If you want to neutralize the toxic face of NETs without throwing away their antimicrobial grip, you might aim at histone charge or accessibility and accept that enzymes like elastase aren’t the main event in barrier injury. That’s a very different strategy than just degrading NET DNA and hoping the problem dissolves. So the next time you hear that the immune system "goes too far," imagine these webs—sticky, bristling with proteins—unfurling across the delicate air-blood interface. The science from Saffarzadeh and colleagues gives us a way to think about that image with precision. It tells us what, inside those nets, does the most harm to lung cells. And it hints at how, with careful chemistry, we might keep the good and curb the bad. That’s the balance acute respiratory distress syndrome desperately needs, and it starts by seeing the trap for what it is.