Could the Functional Expression of HLA-G be Exploited for Successful Stem Cell Transplantation and Engraftment?

757 AS ASSERTED BY Drs. Agrawal and Pandey in their paper in this issue, “The Potential Role of HLA-G Polymorphism in Maternal Tolerance to the Developing Fetus,” the mechanism of maternal tolerance to the developing fetus has remained a major medical puzzle for more than fifty years. Since the fetal genome is half-paternal and half-maternal, how is it that the fetus escapes the maternal immune system? A number of modified hypotheses have been proposed to account for the apparent maternal tolerance of the fetus. However, there is no consensus hypothesis, nor has any compelling data been offered. The answer is indeed important for understanding this unique if not unprecedented response. This may in turn provide clues to devise methods to prevent haplotype disparate transplants from the inevitable and devastating immunological rejection. Every recipient would then have a willing donor whose organs and tissues would not consequently be rejected. This in the absence of immuno-suppression almost appears to be too good to be a viable approach. But, if evolution did not provide a simple, albeit hitherto undefined mechanism for the fetus to escape maternal killer T cells, the species would not have developed by maternal gestation. Was this design an evolutionary exception, or is the answer somewhat simpler? I take this opportunity to respond to the thought-provoking points raised by Drs. Agrawal and Pandey with the objective of not being critical but to seek additional answers to this question of profound importance. Many hypotheses have been advanced to account for maternal tolerance of the antigenically foreign fetus (1,2). Some, perhaps most, of these have been rejected following years of extensive research. Of particular interest are the possibilities that the fetus lacks immunogenicity, that there is a marked attenuation of immune reactivity during pregnancy, or that there exists a status of immunological privilege afforded by the uterus. As Agrawal and Pandey, as well as others, point out, the role of HLA-G expression by the trophoblast and the potential role of HLA-G gene polymorphisms in maternal tolerance to the developing fetus have attracted much attention (3–8). HLA-G can be expressed as seven distinct protein isoforms, each encoded by a specific, alternatively spliced transcript from the HLA-G gene (3). Three isoforms are in soluble form (HLA-G5, -G6, and G-7), while four isoforms are membrane-associated proteins (HLA-G1, -G2, -G3, and G4). Soluble HLA-G isoforms are abundant in the maternal circulation throughout pregnancy. HLA-G is unique among the HLA class I molecules due in part to its quasimonomorphic nature and restricted tissue distribution. However, it remains controversial whether tolerance to HLA-G develops during fetal life. Moreover, proof of tolerance to HLA-G has hitherto not been shown although the expression of HLA-G is present in fetal thymic epithelial cells and in amniotic fluid. This suggests that classical thymic selection could be involved in fetal tolerance to HLA-G. Of particular interest are studies that have shown that HLA-G is expressed in peripheral tissues during certain pathological conditions, such as in cancer, viral infections, and organ transplantation. HLA-G is also expressed during chronic cutaneous inflammatory conditions such as atopic dermatitis and psoriasis. The tolerogenic function of HLA-G is apparent when one considers the magnitude of its effects in immunological processes. For example, HLA-G can inhibit both the transendothelial migration of natural killer (NK) cells as well as NK cell-mediated and antigen-specific cytotoxic T cell (CTL)-mediated cytolysis. HLA-G also shifts the cytokine balance to that of Th2 type effect and suppresses the proliferation of helper alloreactive T-cells. Soluble counterparts of HLA-G may also induce programmed cell death of activated CTL and suppresses the cytolytic activities of NK cells. It has been proposed that the expression of HLA-G exhibits potential for selective inhibition of graft rejection. By inference, quantitative immunological monitoring for the expression of HLA-G in transplantation patients could be important. Others have shown that alloreactive CD41 helper T cells secrete soluble HLA-G5, and that membrane-associated HLA-G1 or soluble HLA-G5 both can inhibit CD41 T cell alloreactivity. Moreover, HLA-G has been shown to be present in the serum as well as in the transplant tissue, and significantly this was associated with improved acceptance of the transplant. In a transgenic animal model, HLA-G inhibited antigen-presenting cell function and maturation. This important obser-