Is There More than One Erythropoietin Receptor? Can the Hematopoietic Effects of EPO Be Dissociated from the Organ-Protective Effects by Carbamylated Erythropoietin?: Derivatives of erythropoietin that are tissue protective but not erythropoietic. Science 305: 239-242, 2004.

Abstract

Beyond any doubt in the past two decades, the introduction of erythropoietin (EPO) into the therapeutic armamentarium of the nephrologist has brought about the single most important progress for treatment and rehabilitation of uremic patients. What has been completely unexpected, however, is the recent finding that EPO is not only a renal hormone promoting hematopoiesis but also, presumably as a result of its antiapoptotic properties, at least in extremely high doses, a hormone that protects against parenchymatous organ injury. The spectrum ranges from neuroprotection in various models of central nervous system damage (1) and cerebral infarction in humans (2) to retinal ischemia (3), neuropathic lesions in diabetes (4,5), and experimental myocardial infarction (6). One dilemma, if one considers transferring the insights from these models to patient treatment, is the predicament that administration of the extremely high doses of EPO required for this purpose will necessarily cause unwanted polyglobuly. One way to circumvent this problem was desialylation of the molecule (7). The resulting asialo-EPO crosses the blood–brain barrier but has an extremely short half-life that mitigates—but certainly does not abrogate—the risk of side effects in the periphery. This field has been given a new twist by the recent work of Leist et al. (8). They started with the postulate that the homodimeric EPO receptor [(EPOR)2] that mediates the neuroprotective effects of EPO differs from the receptor that mediates the effects on erythropoiesis. They concluded that it therefore should be feasible to construct EPO analogues devoid of stimulatory effects on erythropoiesis. To this end, the lysine molecules in EPO were transformed to homocitrullin by carbamylation, thus altering protein conformation and hopefully function. The resulting product, named carbamylated EPO (CEPO), was inactive in an hematopoiesis bioassay yet was neuroprotective in various in vitro assays and in vivo models of neural damage. CEPO failed to bind at 10,000 pM concentration to cellular models of erythropoiesis expressing the classical receptor, whereas CEPO bound to EPO receptors on neural cells and activated downstream signaling pathways via STAT-5 or Jak2. CEPO was shown to cross readily the blood–brain barrier, yet chronic administration of high doses of CEPO, in contrast to low doses of EPO, failed to increase the hematocrit. The therapeutic potential of this compound is obvious. The authors deduce from their findings that CEPO engages an alternative receptor that selectively triggers signaling pathways involved in organoprotection. This paradigm will open an entirely new window on organoprotection by EPO.