Patient conditioning is a critical initial step in hematopoietic stem and progenitor cell (HSPC) transplantation procedures to enable marrow engraftment of infused cells. Preparative regimens have traditionally been achieved by delivering cytotoxic doses of chemotherapeutic agents, with or without radiation. However, these regimens impair host immune function and are associated with significant morbidity. The use of monoclonal antibodies, either alone or conjugated to an internalizing toxin, to target specific antigens on hematopoietic cells has been proposed as a tractable alternative, especially in contexts, such as ex vivo autologous gene therapy, where preservation of immunity is desired. Efficient clearance of marrow has been demonstrated in preclinical models using CD45- or CD117-targeting antibodies conjugated to the plant toxin Saporin. However, this approach still awaits demonstration of long-term safety and efficacy in humans. In this study, we investigated whether toxin-conjugated antibodies targeting the cMPL receptor on HSPCs can provide the basis for a conditioning regimen prior to transplant. Thrombopoietin (TPO) and its receptor cMPL act as primary regulators of HSPC self-renewal and survival. The TPO:cMPL axis also regulates megakaryopoiesis and platelet production but, unlike CD45 and CD117 proteins, cMPL is otherwise not expressed in other blood cell types or in non-hematopoietic tissues. Hence, this approach may uniquely allow effective and specific depletion of host HSCs while sparing most hematopoietic progenitors and mature blood cells.
To investigate cMPL as an antigen for targeted depletion of human HSPCs, we produced a recombinant bivalent anti-cMPL single-chain variable fragment (sc(FV) 2) (Orita et al. Blood 2004) fused with diphtheria toxin truncated at residue 390 (DT390) to prevent toxin internalization in off-target cells. We first confirmed the cMPL receptor-dependent cytotoxic effects of the anti-cMPL-DT390 conjugate in a HEK293A cell line engineered to express the human cMPL receptor. We observed marked cellular killing in vitro (IC50 = 21 pM) compared to the cMPL-negative control HEK293A cell line (Fig. A). Next, we assessed anti-cMPL-DT390 for its ability to inhibit growth of human CD34+ cells in vitro. G-CSF mobilized peripheral blood (PB) CD34+ cells were obtained from five healthy individuals. Surface expression of cMPL was compared by flow cytometry in subsets increasingly enriched in cells with long-term repopulating activity, including bulk CD34+, CD34+CD38- and CD34+CD38-CD90+CD45RA-CD49f+ cells. Levels of cMPL expression increased congruently with levels of HSC purity (Fig. B). Consistent with a cMPL dependent cytotoxic effect, increased cellular death was measured in populations expressing higher densities of cMPL receptors (IC50 = 104 nM), suggesting preferential targeting of the most primitive hematopoietic compartment (Fig. C). We then assessed whether anti-cMPL-DT390 could safely target and deplete human HSPCs in vivo in humanized NBSGW immunodeficient mice. At 12 weeks post-transplantation, engrafted animals (mean 19.8% CD45+ cells in PB) received a single maximum tolerated dose of 1.2 mg/kg anti-cMPL-DT390 (n=7) or vehicle control solution (n=7) by tail vein injection. HSPC depletion was assayed by measuring human myeloid (CD45+CD13+) chimerism in the mouse PB after antibody administration. We observed a gradual decline in HSPC activity, as represented by the decreased production of human myeloid cells following administration of anti-cMPL-DT390, peaking at 6 weeks with a 2.6-fold reduction in frequency of human CD45+CD13+ cells compared to untreated animals (p = 0.003) (Fig. D).
Overall, our study provides proof-of-concept that bivalent anti-cMPL immunotoxin can effectively target and deplete human HSPCs, and may thus provide a novel nontoxic preparative approach to improve HSPC engraftment in transplantation for genetic and other nonmalignant disorders.
No relevant conflicts of interest to declare.