Intense physiological angiogenesis occurs during the early stages of luteal development, providing a model in which the complex processes regulating the angiogenic pathway may be studied. Here, a working hypothesis is presented to explain the diverse changes in the vasculature of the corpus luteum that occur over a short period, based around changes in vascular endothelial growth factor, the angiopoietins and matrix metalloproteinases. An illustration is given of how angiogenesis can be monitored in a primate model and how the role of individual angiogenic factors such as vascular endothelial growth factor may be explored in vivo. Because of the marked effect of inhibition of angiogenesis on luteal function, it is predicted that the normal processes of follicular development, ovulation and luteal function could all be profoundly influenced by the manipulation of angiogenesis.
HM Fraser and SF Lunn
AJ Rowe, C Wulff and HM Fraser
Implantation of a blastocyst into a receptive endometrium is a prerequisite for successful pregnancy. Angiogenesis is a key event in this process but the mechanisms by which localized changes in vascular permeability and angiogenesis occur have yet to be elucidated. Vascular endothelial growth factor (VEGF) and its receptors VEGFR-1 and VEGFR-2 have been implicated as key players in vascular remodelling and placentation. Angiopoietins also appear to have a significant role in regulation of blood vessel growth, maturation and regression. The aim of this study was to describe the molecular regulation of angiogenesis in the first month of pregnancy in marmosets and to address the putative physiological roles for these factors. Uteri were studied at weeks 2, 3 and 4 of pregnancy and compared with late secretory non-pregnant endometrium. Implantation in marmosets occurs at day 11 of pregnancy; hence, these time points were chosen so that the peri-implantation period and very early pregnancy could be studied. mRNAs for VEGF, VEGFR-1 and VEGFR-2, angiopoietin 1, angiopoietin 2 and their receptor Tie-2 were localized and quantified by in situ hybridization. Endothelial cells were identified by CD31 immunocytochemistry. VEGF mRNA was present in all compartments except endothelial cells, and its expression generally increased throughout pregnancy except in upper zone glandular epithelium and luminal epithelium, where a decrease in expression was observed. VEGF receptor mRNAs were found in endothelial cells of the upper zones immediately surrounding glandular epithelium. Angiopoietin 1 mRNA was localized to glandular epithelium of the upper and lower zones throughout pregnancy, and increased in stroma at week 4. Expression of angiopoietin 2 mRNA was localized exclusively to endothelial cells of large luminal vessels and was higher in endometrium from marmosets at week 4 of pregnancy than in endometrium from all other stages. These data provide comprehensive evidence that VEGFR-1 and -2, and angiopoietin 1, angiopoietin 2 and Tie-2 interactions may be involved in the preparation of endometrium for implantation, remodelling of the maternal vasculature and trophoblast invasion during the peri-implantation period in this primate species.
C Wulff, M Weigand, R Kreienberg and HM Fraser
Normal embryonic development is dependent upon a sufficient oxygen, nutrient and waste exchange through the placenta. In primates including humans, this exchange is attained by successful haemochorial placentation which requires the transformation of maternal intramyometrial spiral arterioles by trophoblast invasion to gain uteroplacental circulation, and establishment and maintenance of a competent fetoplacental vasculature. Thus, trophoblast and endothelial cell differentiation, proliferation and invasion occurring during placentation have to be tightly regulated. This review focuses on the diverse developmental steps during haemochorial placentation in humans and other primates and the possible involvement of angiogenic growth factors (vascular endothelial growth factor (VEGF) and angiopoietins (Ang)) in these processes, highlighting the importance of specific actions of angiogenic ligand-receptor pairs. It is hypothesized that VEGF/VEGF-R1 and Ang-1/Tie receptor 2 (Tie-2) may regulate trophoblast differentiation and invasion; VEGF/VEGF-R2 and Ang-1/Tie-2 may promote fetoplacental vascular development and stabilization; and Ang-2/Tie-2 may be involved in maternal vascular remodelling. The importance of a tight regulation of angiogenic factors and their endogenous antagonists for normal development of the placenta is demonstrated by failure of this system, resulting in abnormal placenta vascularization and trophoblast invasion associated with intrauterine growth retardation or pre-eclampsia.
MO Al-Zi'abi, HM Fraser and ED Watson
In mares, little information is available on the type of cell death that occurs during natural and induced luteal regression. Corpora lutea were collected from mares in the early luteal phase, days 3-4 (n = 4); mid-luteal phase, day 10 (n = 5); early regression, day 14 (n = 4); late regression, day 17 (n = 4); and 12 and 36 h (n = 3 per group) after PGF2alpha administration on day 10. Histological and ultrastructural sections were examined and TUNEL was used to detect DNA fragmentation. In early luteal regression, there were more pyknotic luteal cells and extracellular round dense bodies compared with the mid-luteal phase. By late regression, there was a significant decline (P < 0.01) in the number of round dense body clusters and a marked accumulation of lipid. Twelve and 36 h after PGF2alpha administration, changes were similar to those seen in natural regression, but there was also a marked infiltration of neutrophils. Accumulation of lipid was not apparent until 36 h after PGF2alpha administration. Ultrastructural examination revealed rarefaction and distortion of the mitochondrial cristae in most of the luteal cells by the mid-luteal phase. Luteal cells showed shrinkage, accumulation of lipid with foamy appearance, and disruption in both smooth endoplasmic reticulum and mitochondria during natural and induced regression. Some luteal cells showed fragmented or pyknotic chromatin characteristic of apoptosis. Other luteal cells showed crenation of the nuclear membrane and shrinkage of the nucleus, features not characteristic of apoptotic cell death. In late regression, capillaries were obstructed by swollen endothelial cells and round dense bodies. These results show that structural regression may be initiated as early as the mid-luteal phase, and is clearly visible by day 14 in natural regression and 12 h after induced regression. Apoptosis did appear to be involved in luteolysis in the equine corpus luteum, but non-apoptotic changes were also observed in some luteal cells during regression. Accumulation of lipid was a late feature of luteal regression.
MO Al-zi'abi, ED Watson and HM Fraser
Precise pharmacological control of the corpus luteum is important in the manipulation of the oestrous cycle in mares. Angiogenesis plays a key role in the growth and regression of the corpus luteum; therefore, influencing the vasculature of the corpus luteum may offer a novel method for controlling its lifespan. In the present study, changes in angiogenesis and vascular expression of endothelial growth factor (VEGF) were evaluated throughout the luteal phase and after PGF(2alpha)-induced luteolysis. Corpora lutea were collected from mares in the early luteal phase (days 3-4), mid-luteal phase (day 10), early regression (day 14), late regression (day 17), and at 12 and 36 h after administration of PGF(2alpha) on day 10 of the oestrous cycle. Immunohistochemistry was used to localize Von Willebrand factor and Ki67 in endothelial and proliferating cells, respectively. VEGF mRNA and protein were localized by in situ hybridization and immunohistochemistry. The proliferation index of endothelial cells was intense in the early luteal phase. The early and mid-luteal phases were characterized by a dense network of capillaries. The microvasculature started to regress by day 14. After administration of PGF(2alpha), vasodilation was observed after 12 h, but after 36 h, luteal degeneration was accompanied by a significant decrease in vascularity. VEGF mRNA and protein were expressed mainly in the luteal cells during the early and mid-luteal phases and expression declined at early regression (day 14). However, immunostaining for VEGF protein was high in late luteal regression (day 17) and 36 h after PGF(2alpha) administration. These findings indicate a close temporal association between VEGF expression and angiogenesis in the equine corpus luteum during its functional lifespan.
C McKinnell, PT Saunders, HM Fraser, CJ Kelnar, C Kivlin, KD Morris and RM Sharpe
The aims of this study were: (i) to investigate the cellular immunoexpression of androgen receptor and oestrogen receptor beta in the testes of the common marmoset (Callithrix jacchus) during neonatal life compared with their expression at later ages; (ii) to establish whether neonatal marmoset Sertoli cells are targets for androgens or oestrogens or both; and (iii) to investigate the relationship between neonatal plasma testosterone concentrations and androgen receptor immunoexpression by abolishing the neonatal testosterone surge with a potent GnRH antagonist. Androgen receptor and oestrogen receptor beta immunoexpression were evaluated in neonatal animals aged 1-4 days, 4 weeks and 6 weeks, and compared with immunoexpression in animals aged 18-22 weeks (early infancy), 35 weeks (late infancy), 58-62 weeks (late pubertal) and > 100 weeks (adult). Immunoexpression of androgen receptor in the reproductive tract was also evaluated at each age. Sertoli cell immunoexpression of androgen receptor was weak or absent in neonatal animals, but increased substantially in infant animals, reaching adult levels by the end of infancy. In contrast, immunoexpression of androgen receptor during the neonatal period was strong in testicular interstitial cells and very strong in epithelial cell nuclei throughout the reproductive tract, and did not change greatly with age in these cells or tissues. Similarly, immunoexpression of oestrogen receptor beta was prominent in many Sertoli cells and in the germ cells of neonatal animals, and was relatively constant throughout life. Weak immunoexpression of androgen receptor in neonatal Sertoli cells was associated with high plasma testosterone concentrations (2.7-5.5 ng ml(-1)), whereas strong Sertoli cell immunoexpression was associated with baseline (approximately 0.12 ng ml(-1)) testosterone concentrations in infant animals and with > 10 ng ml(-1) in late pubertal and adult animals. Immunoexpression of androgen receptor and oestrogen receptor beta was also evaluated in co-twin males aged 4 and 35 weeks, after treatment from birth to 4 weeks or from week 25 to week 35, respectively, with either vehicle or with GnRH antagonist at a dose known to suppress the neonatal testosterone surge completely. Only GnRH antagonist treatment during weeks 25-35 reduced androgen receptor immunoexpression, whereas immunoexpression of oestrogen receptor beta was unaffected by treatment during either period. On the basis of these findings it is suggested that: (i) neonatal marmoset Sertoli cells may be targets primarily for oestrogens rather than androgens; (ii) androgen receptor expression in the testes of neonatal and infant marmosets is not regulated in a straightforward way by testosterone; and (iii) high neonatal concentrations of plasma testosterone are not absolutely necessary for expression of androgen receptor in marmoset testes at this time.