Antiidiotype Antibody against Platelet Anti-Gpiiia Contributes to the Regulation of Thrombocytopenia in HIV-1–Itp Patients

Patients with human immunodeficiency virus 1–associated immunological thrombocytopenia (HIV-1–ITP) have markedly elevated platelet-bound immunoglobulin (Ig)G, IgM, and C3C4, as well as serum circulating immune complexes (CICs) composed of the same. Affinity purification of IgGs from their CICs with fixed platelets reveals high-affinity antibody (Ab) against platelet glycoprotein (GP)IIIa 49–66, which correlates inversely with their platelet count. However, sera from these patients have little to no anti-GPIIIa activity. To investigate this, we assayed serum, purified serum IgG, and CIC-Ig from these patients. This revealed ∼150-fold greater Ab activity in purified serum IgG, and ∼4,000-fold greater reactivity in CIC-IgG. This was shown to be associated with the presence of antiidiotype Ab2 (both IgG and IgM) sequestered in the CIC-IgG. The IgM antiidiotype was predominantly blocking Ab, as demonstrated by specificity for F(ab′)(2) fragments of anti–GPIIIa 49–66 of HIV-1–ITP patients and inhibition of reactivity with peptide GPIIIa 49–66, not with a control peptide. The IgM antiidiotype was not polyreactive. Similar measurements were made in nonthrombocytopenic HIV-1–infected patients. Their serum reactivity was not measurable, but serum Ig and CIC-IgG against platelet GPIIIa 49–66 was present, although considerably lower than that found in HIV-1–ITP patients (26- and 35-fold lower, respectively). In addition, their IgM antiidiotype reactivity was 12-fold greater than that found in HIV-1–ITP patients. The IgM antiidiotype Ab titer of both cohorts correlated with in vivo platelet count (r = 0.7, P = 0.0001, n = 32). To test the in vivo effectiveness of the IgM antiidiotype, thrombocytopenia was induced in mice with 25 μg of affinity-purified anti–GPIIIa 49–66 (mouse GPIIIa has 83% homology with human GPIIIa and Fc receptors for human IgG1). Maximum effect was obtained at 4–6 h after intraperitoneal injection into Balb/c mice with a platelet count of ∼30% baseline value. Preincubation of the anti-GPIIIa Ab with control IgM at molar ratios of IgM/IgG of 1:7 before intraperitoneal injection had no effect on the in vivo platelet count, whereas preincubation with patient IgM antiidiotype improved the platelet count to 50–80% of normal. Thrombocytopenia could be reversed after addition of IgM antiidiotype 4 h after induction of thrombocytopenia. Thus, CICs of HIV-1–infected patients contain IgM antiidiotype Ab against anti-GPIIIa, which appears to regulate their serum reactivity in vitro and their level of thrombocytopenia in vivo.