HIV-1 infection induces neutralizing (Nt) antibodies (Abs) against one's infecting virus, yet rarely do such Abs neutralize a broad viral spectrum. A few broadly(b)-Nt Abs have been discovered: 4E10, 2F5, b12, 447-52D and 2G12;all but 2G12 have longer-than-usual CDR-H3 hypervariable loops (H3s). Typically, H3 is crucial to antigen contact;yet the crystal structures of 4E10, 2F5 and 447-52D show that, while prominent, their H3s either contact marginal residues of their peptide-epitopes or do not contact these epitopes at all. Haynes et al. (2005) reported that most of the bNt MAbs bind to self-antigens (e.g., 2F5 and 4E10 bind cardiolipin;CL), and speculated they are autoAbs that arose through a process of positive selection on autoantigen followed by recruitment by Env into the adaptive Ab response. Under this hypothesis, bNt Abs arise when tolerance is broken, allowing self-reactive B cells to survive, and Abs with long H3s to emerge. Thus, it was proposed that, to produce bNt Abs, HIV-1 vaccines must break tolerance and produce self-reactive Abs. Our preliminary studies indicate that if 4E10 and 2F5 are autoAbs they are unusual, as their affinity for CL is 100-1000-fold weaker than for their HIV-1 peptide epitopes. We also did not detect increased CL reactivity in 12 bNt sera from HIV+ donors nor in 20 other HIV+ sera, compared to HIV- controls. As it is claimed that long H3s reflect a loss of tolerance, we produced a DNA sequence database of 450 human MAbs of known specificity, obtained mainly from people with autoimmune disease, chronic and acute infection, and found that long H3s exist mainly among anti-protein subsets of the autoimmune and chronic Abs;among the latter, long-H3 Abs are not restricted to HIV infection. From this we propose 3 hypotheses: (1) Long-H3 Abs are produced during HIV-1 infection through loss of tolerance during B-cell development. This predicts that long-H3 Abs accumulate in the IgM+/IgD+, mature, naive B cell compartment, and among IgG+ memory B cells;bNt sera should have significant autoreactivity. (2) Long-H3 Abs arise during antigen-driven processes. These Abs should be present only in the IgG+ memory B cell and plasma cell compartments, and bNt sera should not to be autoreactive. (3) HIV antigens select long-H3 Abs from normal Ab repertoires;this has the same predictions as (2). To test these alternatives, we plan to analyze sera and PMBCs collected from 12 untreated HIV+ donors whose sera are bNt, 12 HIV+ donors whose sera are not broadly neutralizing, 12 patients with SLE, and 14 healthy donors. B cells will be sorted by phenotype into naive, memory, and plasma cell pools and if possible into single, HIV-specific B and plasma cells;expressed mu or gamma VH genes from each B cell pool and from single cells will be sequenced and analyzed for H3 length, somatic mutations, and gene usage. Sera and Abs produced by antigen-specific cells will be analyzed on microarrays bearing autoantigens, HIV-1 antigens, and markers of polyreactivity. We expect to determine if long-H3 Abs are present in B-cell subsets in HIV-1 infection vs. a truly autoimmune state, and whether distinct autoreactivities or generalized polyreactivity is present in bNt sera. The results should help to clarify if HIV vaccines should be designed to break tolerance.
One approach to AIDS vaccine design uses HIV-1-neutralizing antibodies. These antibodies are rare and their structures are considered unusual;moreover, questions have arisen concerning their ability to react with self . The proposed work will address whether HIV-neutralizing antibodies are unusual and/or self-reactive , and should clarify whether induction of unusual, self-reactive antibodies is a necessary consideration in AIDS vaccine design.
| Almagro, Juan Carlos; Gilliland, Gary L; Breden, Felix et al. (2014) Antibody engineering and therapeutics, The Annual Meeting of the Antibody Society: December 8-12, 2013, Huntington Beach, CA. MAbs 6:577-618 |
