Biology of Reproduction, Vol 53, 1373-1384, Copyright © 1995 by Society for the Study of Reproduction

ARTICLES

Is corpus luteum regression an immune-mediated event? Localization of immune system components and luteinizing hormone receptor in human corpora lutea

A Bukovsky, MR Caudle, JA Keenan, J Wimalasena, NB Upadhyaya and SE Van Meter
Department of Obstetrics and Gynecology, University of Tennessee Medical Center at Knoxville 37920, USA.

Factors determining the life span of the human corpus luteum (CL) are not known. In addition to being determined by hormonal factors, such as hCG, the life of luteal cells may be determined by the preservation of luteal vascularization. Furthermore, the CL represents an immunologically unique tissue, as it is formed after menarche, long after adaptation of the immune system toward self. Thus, CL regression may be immunologically mediated. To determine what role the vasculature and immune system play in human CL development and regression, we examined immunohistochemically 1) the expression of Thy-1 differentiation protein by vascular pericytes, 2) the expression of major histocompatibility complex (MHC) class I and class II molecules in granulosa lutein cells (GLC), and 3) infiltration of the CL by macrophages and T lymphocytes. LH receptor (LHR) and cytokeratin 18 expression were also studied. In developing CL, the pericytes of luteal microvasculature released Thy-1 differentiation protein among the endothelial cells of proliferating vessels. In mature CL, Thy-1 released from vascular pericytes accumulated on the surface of GLC, and these cells exhibited LHR immunoreactivity (LHRI). Overall LHRI increased during the luteal phase and was strongest at the beginning of the late luteal phase. Although vascular pericytes showed strong LHRI, no staining of endothelium was detected during the luteal phase. GLC exhibited strong cytokeratin staining and moderate staining for MHC class I and MHC class II antigens; numerous macrophages were detected in luteal tissue. During pregnancy, the staining pattern was similar to that seen in the mature CL at the end of the midluteal phase. During the late luteal phase, surface expression of MHC class I and MHC class II antigens by GLC was substantially enhanced, and some T cells invaded among luteal cells. By the end of the cycle, an acute regression of vasculature and luteal tissue was observed along the fibrous septa. The remaining GLC showed only surface and no cytoplasmic LHRI. During the subsequent cycle, in the presence of numerous T cells, regressing GLC exhibited strong surface expression of various macrophage markers, such as CD4, CD14, CD68, and leukocyte common antigen, a feature not detected in the CL during the luteal phase nor described in other tissues. A complete loss of cytokeratin staining in GLC was observed. In regressing CL, strong LHRI was present in the endothelium of small and large luteal vessels. In conclusion, vascular pericytes and macrophages may stimulate the development and senescence of luteal tissue. The senescence of GLC may be inconsistent with preservation of luteal vasculature, and T lymphocytes appear to participate in terminal regression of the CL. Regression of luteal tissue therefore resembles immunologic rejection of a transplant. During pregnancy, the aging process of GLC appears to be interrupted, possibly due to the temporary acceptance of the CL "graft."