ARA-290 — the erythropoietin fragment that doesn't make red blood cells

Because EPO also makes blood thicker. And thicker blood, in sick people, is dangerous.

That tension — a hormone with two distinct biological personalities — is what eventually led to ARA-290. Understanding how that compound came to exist requires understanding what erythropoietin actually does, which turns out to be considerably more than its name suggests.

EPO's classical story is straightforward. Oxygen drops in the blood, the kidneys sense it, and a glycoprotein hormone gets secreted into circulation. That hormone reaches the bone marrow and tells erythroid precursor cells to mature into red blood cells. More red blood cells mean more oxygen-carrying capacity, and the system corrects itself. This is why athletes at altitude adapt over weeks — EPO rises, hematocrit rises, performance improves. It's also why synthetic EPO became one of the most abused compounds in endurance sports: it does exactly what it says it does, reliably and measurably.

The problem with using EPO therapeutically for tissue protection is the same problem that makes it useful for doping. More EPO means more red blood cells. More red blood cells mean higher hematocrit — the proportion of blood that's cells rather than plasma. And higher hematocrit means blood that clots more readily, flows less freely through small vessels, and raises the risk of stroke, pulmonary embolism, and death. Several large clinical trials that attempted to use EPO for heart and kidney protection had to be abandoned or reported net harm. The hormone that protected tissue in animal models killed people in clinical trials when given at doses high enough to produce the effect. The erythropoietic problem couldn't be disentangled from the tissue-protective one — at least not with EPO itself.

Anthony Cerami, a researcher who had spent decades at Rockefeller University studying the biology of inflammation and glycosylation before founding Araim Pharmaceuticals, approached the problem from an unusual angle. His starting point wasn't "how do we give EPO safely" but "why does EPO protect tissue at all." The answer pointed toward a receptor complex that had nothing to do with erythropoiesis.

The classical EPO receptor — the one that drives red blood cell production — is a homodimer, meaning it consists of two identical subunits of the EPOR protein sitting side by side. When EPO binds it, the homodimer activates intracellular signaling cascades that tell precursor cells to differentiate into red blood cells. This receptor is what you're targeting when you use EPO for anemia. It's also what you're activating, harmfully, when you try to use EPO for tissue protection.

But tissue cells — neurons, cardiac myocytes, vascular endothelium — express a different receptor configuration. They have a heterotrimeric complex: two EPOR subunits paired with one subunit of the beta-common receptor, also called CD131 or the beta-c chain, which is shared with several other cytokine receptors including those for IL-3, IL-5, and GM-CSF. This heterotrimeric configuration is sometimes called the innate repair receptor, or IRR, because it appears to be the receptor that mediates EPO's tissue-protective effects. When EPO is present in high concentration during injury or hypoxia, it activates not only the classical homodimeric receptor but also this heterotrimeric one, and the signaling downstream of that heterotrimeric activation is anti-inflammatory, anti-apoptotic, and repair-promoting.

The insight that separated EPO's two functions was that the binding geometry for each receptor is different. EPO's interaction with the homodimeric EPOR uses a particular domain of the protein — a specific structural face of the molecule that fits into the symmetrical groove of the two identical subunits. The heterotrimeric IRR has different geometry, different subunit spacing, and different binding requirements. This meant, in principle, that you could design a molecule that fit the IRR without fitting the homodimeric EPOR — that retained the tissue-protective signaling without triggering erythropoiesis.

Cerami's team at Araim spent years characterizing the structure-activity relationships, working out which portions of EPO were necessary for IRR binding and which were necessary for EPOR homodimer binding. What they arrived at was an eleven-amino-acid peptide. ARA-290 — also called cibinetide — is short enough that it can't span the structural distance required to activate the homodimeric receptor. It fits the heterotrimeric IRR but not the classical EPO receptor. In preclinical models, it produced no increase in hematocrit and no erythropoietic effect. The tissue-protective, anti-inflammatory signaling downstream of IRR activation was intact.

The mechanism downstream of IRR activation is worth understanding because it's what makes ARA-290 interesting for applications far removed from anemia. When the IRR is activated in tissue, the signaling cascade suppresses pro-inflammatory cytokines — particularly TNF-alpha, IL-6, and IL-1beta — while promoting cell survival pathways including PI3K/Akt. It reduces apoptosis in stressed cells. In neurons, it appears to support axon survival and regeneration signaling. In vascular endothelium, it promotes microvascular integrity and reduces endothelial dysfunction. In macrophages and monocytes, it shifts the inflammatory tone toward resolution. The picture that emerges from the preclinical literature is a compound that broadly promotes tissue maintenance in the context of injury or inflammation, working through a receptor that was apparently designed by evolution to help tissues survive under stress.

The most substantive human research has focused on small-fiber neuropathy in the context of sarcoidosis. Sarcoidosis is a systemic inflammatory disease that causes granuloma formation across multiple organs; a meaningful proportion of patients develop neuropathy affecting the small nerve fibers that transmit pain, temperature, and autonomic signals. This neuropathy is notoriously difficult to treat — it doesn't respond well to standard analgesics, it doesn't respond to immunosuppression in many cases, and it produces a quality of pain and sensory disturbance that is both severe and resistant to conventional approaches. The damage is to the nerve fibers themselves, not to large myelinated axons, which is why standard nerve conduction studies often come back normal even when patients are in significant distress.

Dahan and colleagues at Leiden University Medical Center conducted a series of trials in sarcoidosis-associated small-fiber neuropathy using ARA-290. What they found across their work included improvements in intraepidermal nerve fiber density — the count of small nerve fibers in skin punch biopsies, which is a direct structural measure of small-fiber health — and reductions in neuropathic pain scores. These are not surrogate endpoints; nerve fiber density is the actual tissue you're trying to repair, and patients who show increases in that density after treatment with a compound believed to activate repair signaling have provided meaningful mechanistic evidence, not just reported feeling better. The trials were small and the compound remains investigational, but the signal was real enough to sustain continued research interest.

Diabetic neuropathy represents a larger potential application, and the biology connects cleanly. Diabetes destroys small fibers through a combination of microvascular compromise and direct metabolic toxicity to axons. The same repair signaling that ARA-290 appears to activate in sarcoidosis-associated neuropathy is presumably relevant — perhaps more so, given that the microvascular endothelial dysfunction that underlies much of diabetic neuropathy is exactly the tissue EPO's tissue-protective receptor complex was designed to address. Early investigational work has proceeded here, though the evidence base is less developed than in the sarcoidosis context.

The cardiovascular work touches on a related theme. Microvascular dysfunction — the failure of small vessels to dilate and contract appropriately, to maintain endothelial integrity, to support tissue perfusion — underlies much of the cardiac risk in metabolic disease and is one of the mechanisms that makes heart failure with preserved ejection fraction so difficult to treat. ARA-290's effects on vascular endothelium in preclinical models have attracted interest from researchers working on this problem, though the clinical evidence here is earlier stage.

What's worth being precise about: ARA-290, cibinetide, is an investigational compound. It is not FDA-approved. It has not cleared a Phase III trial and reached the market. The clinical research that exists is real and mechanistically coherent, and it provides a stronger foundation than most compounds in this class have — the Leiden trials in particular are methodologically careful and produced meaningful outcomes. But "meaningful Phase II signal in a specific neuropathic condition" is a long distance from "established therapeutic." The research earned the compound serious attention. It has not yet earned a definitive clinical claim.

What it does represent is a genuinely novel approach to an old problem — that certain biological systems, including erythropoietin's tissue-protective function, were designed to do something useful and have been inaccessible as therapies because they were tangled up with side effects in the same molecule. The logic of separating those functions, of engineering a fragment that speaks to one receptor complex and not the other, is not unique to ARA-290 — it's a general principle of rational drug design. What's distinctive here is that the tissue-protective receptor had been sitting in the biology for years before anyone figured out how to talk to it directly. The molecule Cerami's team built is the first serious attempt to have that conversation at scale.

The implication of a functional innate repair receptor — a receptor that appears to be specifically designed to receive signals about tissue distress and initiate repair — is that the body has more capacity for organized self-repair than was understood even recently. What ARA-290 does is attempt to speak the language that receptor already knows.