Platelet Factor 4 Regulates Platelet Count In Vivo: Implications for Platelet Recovery after Cytotoxic Therapy.

Platelet factor 4 (PF4), a platelet-specific CXC chemokine, was the first reported negative autocrine regulator of megakaryopoiesis in vitro. To define the physiological role(s) of PF4, we established mice that either were deficient in murine (m) PF4 (mPF4^−/−) or that over-expressed human (h) PF4 (hPF4^+/+). These mice had a level of PF4 ~6-fold greater than that present in human platelet controls. All lines studied had been backcrossed onto a C57Bl/6J background for >10 generations. Platelet counts in these animals correlated inversely with PF4 determined expression, beginning with a low platelet count of 702 ± 57 x 10³/μL in the hPF4^+/+ mice < hPF4⁺ < wildtype (WT) < mPF4^+/− < mPF4^−/− mice in which the platelet count was 1,404 ± 117 x 10³/μL. The half-life of the platelets from the hPF4^+/+ was identical to that of WT mice. Cultured bone marrow mononuclear cells (BMMNC) in serum-free media showed that each line had identical efficiency in growing megakaryocyte colonies, suggesting that megakaryocyte progenitor cells in these different genetic lines were intrinsically normal. Megakaryocyte colony numbers derived from WT BMMNC were reduced by the addition of recombinant PF4 or supernatant from irradiated bone marrow of hPF4⁺ mice, but not from mPF4^−/− mice, suggesting that megakaryocyte lysis in vivo during cytoreductive therapy may contribute to the subsequent thrombocytopenia by releasing PF4. Additionally, a rabbit polyclonal anti-mPF4 antibody (Ab) was able in culture to significantly reverse this inhibitory effect of PF4 on megakaryopoiesis. Preliminary cytoreductive studies using either 600 cGy or 150 mg/kg of 5-fluorouracil (5-FU) intraperitoneally (IP) were performed. In irradiation studies, mPF4^−/− mice began to recover on the same day as WT littermates, but they clearly had higher platelet counts at their nadir, with a drop to only 42 ± 7% of baseline vs. 32 ± 6% in the WT mice (n =12 in each arm, p = 0.06). By Day 13, 9 of 12 mPF4−/− mice had recovered to >75% of baseline, while only 3 of 12 WT mice had recovered (p <0.001). hPF4⁺ mice (n = 7) were studied after 5-FU treatment. Compared to WT littermates (n = 9), the hPF4⁺ recovered later (15.6 ± 2.2 vs. 11.2 ± 1.5 days, p < 0.0003), and clearly had significantly greater drop to 30 ± 6% vs. 56 ± 9% of baseline (p < 0.00001). By day 15, all of the WT mice had recovered, but only 43% of hPF4+ mice had returned to >75% of baseline platelet count (p = 0.009). To examine if anti-mPF4 Ab was protective of cytotoxic therapy-induced thrombocytopenia, WT mice were treated with 180 mg/kg of 5-FU and were given either anti-mPF4 Ab (25 mg/kg, IV, x 2) or an equal volume of vehicle. By day 5, the Ab-treated group had a platelet count of 45 ± 6% vs. 32 ± 4% in the untreated (n > 13 per arm, p = 0.015). Platelet counts remained higher in the Ab-treated arm throughout the study. By day 10 after intervention, 9 of 16 mice of the Ab-treated arm had platelet counts over 75% of the baseline, while only 3 of 13 control mice did (p < 0.001). Thus, it appears that PF4 is an important negative autocrine regulator of platelet count in vivo. Excessive release of PF4 following cytotoxic therapy may be a mediator of treatment-related thrombocytopenia. Strategies directed to alleviate the consequence of released PF4 may have clinical benefit in ameliorating this thrombocytopenia.