Angiogenesis in the human corpus luteum: simulated early pregnancy by HCG treatment is associated with both angiogenesis and vessel stabilization

Christine Wulff1,3, Sarah E. Dickson1, W. Colin Duncan2 and Hamish M. Fraser1

1 Medical Research Council, Human Reproductive Sciences Unit and 2 Department of Reproductive and Developmental Sciences of the University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9ET, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: This study examined changes in the luteal vasculature throughout the menstrual cycle and during simulated pregnancy with human chorionic gonadotrophin (HCG) in the human. METHODS: Endothelial cell and pericyte area were assessed by quantitative immunocytochemistry for CD34 and {alpha}-smooth muscle actin respectively, taking into consideration the dynamics of lutein cell hypertrophy and atrophy throughout the cycle and after HCG treatment. Endothelial cell proliferation was detected by Ki-67/CD34 dual staining and a proliferation index was obtained. The molecular regulation of angiogenesis was studied by examining changes in vascular endothelial growth factor (VEGF) immunostaining. RESULTS: The early luteal phase is associated with intense angiogenesis, as indicated by high endothelial cell proliferation, and by the mid-luteal phase a mature vasculature was apparent, as shown by maximal endothelial cell and pericyte areas. During the late luteal phase, decreased endothelial proliferation, endothelial cell and pericyte area indicated vascular regression. HCG treatment induced a second burst of total and endothelial cell proliferation and a concomitant increase in endothelial cell and pericyte areas. VEGF protein was expressed throughout the luteal phase and a significant increase was found after HCG treatment. CONCLUSION: Luteal rescue with HCG is associated with a second wave of angiogenesis and vascular stabilization.

Key words: angiogenesis/corpus luteum/endothelial cells/VEGF/simulated early human pregnancy


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Angiogenesis is a dynamic process of endothelial cell proliferation and capillary tubule formation followed by recruitment of pericytes to confer stabilization to the newly formed vessels. Luteal angiogenesis is tightly regulated, occurring in a spontaneous cycle of vascular development, stabilization and regression. For quantification of such a dynamic process in vivo, it is necessary to examine particular time points in the lifespan of the corpus luteum: the early, developing corpus luteum, the mid-luteal stage, when maturity of the gland is reached and the late luteal phase, which is associated with functional and structural regression. In order to explore the regulation of angiogenesis, it is essential to quantify accurately changes in vascular density. Recently, it has been shown that changes in vascular density may be influenced not only by angiogenesis but also by changing tissue volume (Gaytan et al., 1999Go). For example, the increase in luteal volume from early to mid-luteal phase and the decrease during luteal regression influence vascular density in the human corpus luteum (Gaytan et al., 1999Go). When changes in a particular cell content are quantified per unit area of luteal tissue, factors such as the change in luteal volume (translated to area in tissue sections) over the lifespan of the corpus luteum have to be addressed, or results will reflect the effects of expanding or contracting tissue in addition to increases or decreases in content of the particular cell in question. Of the factors involved in determining the volume of the corpus luteum throughout the luteal phase, it is the change in size of the large steroidogenic cells that is of prime importance.

While in the normal menstrual cycle the lifespan of the corpus luteum and its vasculature is limited to two weeks, after conception the corpus luteum is rescued by endogenous human chorionic gonadotrophin (HCG) and survives for several months. It has been shown that exposure to increasing concentrations of exogenous HCG simulates the luteal rescue associated with early pregnancy (Illingworth et al., 1990Go). Investigating the changes in the vasculature after HCG rescue is of particular clinical interest because failure of the gland due to a malfunction of the vasculature could possibly lead to miscarriage or infertility. Thus, the aim of this investigation was to quantify angiogenesis with particular reference to changes after HCG treatment in the human corpus luteum, taking into consideration the dynamic changes of lutein cell volume associated with luteal formation, regression and simulated early pregnancy. The lutein cell size was measured and a conversion factor designed to take account of changing volume. This was used to adjust the measurements for the endothelial cell area quantified by CD-34 immunocytochemistry and pericyte areas quantified by {alpha}-smooth muscle actin immunostaining. Cell proliferation during the diverse processes of luteal formation, regression and rescue was assessed by Ki-67 immunocytochemistry and endothelial cell proliferation index by Ki-67/CD34 dual staining. The molecular regulation of angiogenesis was investigated by detection of a principal angiogenic factor, vascular endothelial growth factor (VEGF), a secreted angiogenic mitogen (Leung et al., 1989Go) known to be a major requirement for luteal angiogenesis (Dickson et al., 2001)Go.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Treatments and collection of tissue
Corpora lutea were enucleated at the time of hysterectomy as described previously (Duncan et al., 1996Go; Wulff et al., 2000Go). All women were healthy and aged 32–45 years. Only women with regular menstrual cycles who had not received any form of hormonal treatment during the previous 3 months took part in the study. The date of the pre-ovulatory LH surge was determined by measuring LH concentrations in serial early morning urine samples collected before operation (Duncan et al., 1998Go). On this basis, five corpora lutea were classified as early (LH+1 to LH+5 days), five as mid-luteal (LH+6 to LH+10), and three as late luteal (LH+11 to LH+14 days). In addition, six women were given i.m. injections of HCG (Profasi; Serono Laboratories, Welwyn Garden City, UK) from LH+7 in daily doubling doses, starting at 125 IU and continuing for 6–8 days until surgery to rescue the corpus luteum and reproduce the hormonal changes of early pregnancy (Illingworth et al., 1990Go). There was no difference in average age between any of the groups.

Corpora lutea were enucleated from the ovary by blunt dissection. The tissue was immediately divided into radial blocks and a portion was fixed in 4% paraformaldehyde for 24 h. These corpora lutea were further used in studies to determine changes in other factors associated with control of luteal cell function described elsewhere (Rodger et al., 1997Go; Duncan et al., 1998Go; Wulff et al., 2000Go). An endometrial biopsy was also obtained to assist luteal staging by tissue morphology. In all cases, morphological dating of the luteal-phase endometrium (Li et al., 1988Go) was used to confirm the luteal-phase classification. The study was approved by the Reproductive Medicine Branch of the South-East of Scotland Medical Ethics Committee, and informed consent was obtained from all patients before tissue collection.

Immunocytochemistry
Luteal angiogenesis was assessed by (i) quantifying the total number of mitotic cells by staining for Ki-67, (ii) examining the establishment of the microvasculature using CD34, a specific human endothelial cell antigen to identify endothelial cells, (iii) carrying out double labelling for CD34 and Ki-67 to obtain a proliferation index for endothelial cells, (iv) using vascular {alpha}-smooth muscle actin immunocytochemistry to identify and quantify pericyte recruitment, and (v) examining the expression of the main regulator of angiogenesis, VEGF. Sections of 5 µm were cut onto Tespa-coated slides. Sections were dewaxed in Xylene and rehydrated in descending concentrations of ethanol. Negative controls were performed for all antibodies. For Ki-67, CD34 and vascular {alpha}-smooth muscle actin immunocytochemistry, negative controls were used by replacement of the first antibody with immunoglobulin G from the same species and at the same concentration as the primary antibody was employed. For VEGF immunocytochemistry, a negative control was carried out by pre-absorbing the primary antibody with the peptide it was raised against.

Ki-67 immunocytochemistry
Antigen retrieval for Ki-67 immunostaining was achieved by microwaving. Sections were exposed to two 10 min cycles and one 5 min cycle of microwave irradiation at 700W in citrate buffer 0.01 mol/l, pH 6.0. After cooling sections in Tris-buffered saline (TBS), a normal rabbit serum (NRS, diluted 1:5 in TBS) block was applied for 30 min at room temperature. Monoclonal anti-human Ki-67 antibody (Novacastra, Peterborough, UK) was applied at a dilution of 1:50 in TBS, for 2 h at 37°C. Detection was carried out using a biotinylated rabbit anti-mouse antiserum (Dako Ltd, Cambridgeshire, UK) diluted 1:500 in TBS for 30 min at room temperature, followed by pre-conjugated avidin-biotin-alkaline phosphatase complex (ABC-AP, Dako) in 0.05 mol/l Tris, pH 7.4, for 30 min at room temperature. Nitro Blue tetrazolium (NBT) was used for visualization, and sections were counterstained in haematoxylin.

CD34 immunocytochemistry
Endogenous peroxidase activity was quenched with a 30 min incubation in 3% hydrogen peroxide in methanol at room temperature. A normal goat serum (NGS, 1:5 dilution) block was used and sections were incubated overnight at 4°C in mouse monoclonal CD34 antibody (Serotec, Oxford, UK) applied at a 1:25 dilution in TBS. Immunolocalization was undertaken using the mouse EnVision kit (Dako). Conjugated goat anti-mouse-avidin-biotin-horse radish peroxidase complex (ABC-HRP, Dako) 100 µl was applied for 30 min at room temperature. Sections were washed in TBS and liquid diaminobenzidine substrate (DAB, EnVision kit) was used for detection. Sections were not counterstained so as to enable image analysis to be performed.

Double labelling for CD34 and Ki-67
For double labelling, the same antibodies and concentrations were used as described above. Double labelling began with staining for CD34. After blocking unspecific binding with NRS:TBS (1:5), CD34 antibody was applied overnight at 4°C. The alkaline phosphatase anti-alkaline phosphatase (APAAP) method was used for detection. Secondary/bridging antibody, rabbit anti-mouse (Dako), was applied at a concentration of 26.6 µg/ml in NRS:TBS (1:5) for 40 min at room temperature. After washing to remove unbound antibody, slides were incubated with mouse APAAP (Dako) 1:100 in NRS:TBS (1:5) for 40 min at room temperature. For signal detection, Fast Red (Sigma, St Louis, MO, USA) was used. Fast Red substrate was applied at 1 mg/ml Fast Red buffer (20 mg naphtol AS-MX phosphate, 2 ml dimethyl formamide, 98 ml 0.1 mol/l Tris, pH 8.2). Development of CD34 was followed by immunostaining for Ki-67. Sections were microwaved as described above. Overnight incubation with Ki-67 antibody was carried out at 4°C and the APAAP method, as described for above, was used for signal detection. Visualization of Ki-67 was obtained with NBT. Sections were not counterstained.

{alpha}-smooth muscle actin immunocytochemistry
For antigen retrieval, sections were pressure cooked (Clypso pressure cooker, Tefal, Essex, UK) in 0.01 mol/l citrate buffer, pH 6.0 for 5 min. Sections were left for 20 min in hot buffer before cooling in TBS. Non-specific background was blocked with NRS for 30 min at room temperature. Mouse anti-human {alpha}-smooth muscle actin (Dako) was applied (4.3 µg/ml) and incubated overnight at 4°C. The APAAP method was used for detection. Secondary/bridging antibody, rabbit anti-mouse (Dako) was applied at a concentration of 26.6 µg/ml in NRS (1:60 dilution) for 30 min at room temperature. After washing to remove unbound antibody, slides were incubated with mouse APAAP (Dako) in NRS (1 µg/ml, 1:100 dilution) for 30 min at room temperature. Slides were placed in TBS before NBT detection. Adjustment to NBT substrate was achieved by pre-incubation in NBT buffer (0.1 mol/l Tris, 0.1 mol/l NaCl, 0.05 mol/l MgCl2, pH 9.7) for 5 min. Sections were incubated in NBT solution containing 45 µl 75 mg/ml NBT in 70% dimethylformamide; 35 µl X-phosphate/5 bromo-4 chloro-3 inodyl-phosphate in 100% dimethylformamide; and 1 mmol/l levamisole; (all in 10ml NBT buffer) for 30 min. Colour reaction was stopped in tap water.

VEGF immunocytochemistry
Sections for VEGF immunocytochemistry were dewaxed and rehydrated. VEGF antigen was retrieved by pressure cooking slides on full power in 3 mol/l glycine, 0.1% EDTA buffer, pH 3.5, for 5 min. The slides remained in hot buffer for a further 20 min before washing in TBS. Endogenous peroxidase activity was blocked as described above. Rabbit polyclonal VEGF antibody, specific for all isoforms of VEGF-A (2 µg/ml pre-diluted in NGS; VEGF A-20, Santa Cruz Biotechnology, Santa Cruz, CA, USA), was added and sections incubated at 4°C overnight. Negative controls were incubated in primary antibody pre-absorbed with VEGF-A peptide in a ratio of 1:5, VEGF-A antibody: blocking peptide (Santa Cruz). Immunolocalization was undertaken using the rabbit EnVision kit (Dako) in the same way as the mouse was used for CD34 immunostaining. Sections were counterstained and mounted.

Quantification
Quantitative analysis of immunostaining was performed using an image analysis system linked to an Olympus camera, and the data processed using the Image-Pro Plus Version 3.0 for Windows computer programme.

Quantification of the granulosa lutein cell area and estimation of a conversion factor
Haematoxylin and eosin stained sections were examined under a x40 lens and images captured. Granulosa lutein cells were identified according to their morphological appearance (large polygonal cells with abundant cytoplasm containing a centrally located nucleus) and by comparison with results of a previous study in which steroidogenic cells in the same tissue were stained for 3-ß-hydroxysteroid dehydrogenase (Duncan et al., 1999Go). A stratified procedure was used to make sure that samples were randomly chosen from all parts of the cross-section of each corpus luteum. Ten cells arranged in a cross formation, in 10 fields of view, were outlined and the area of each measured. The mean was calculated per field and per corpus luteum. Changes in lutein cell size during the cycle influence the proportion of the specific cell of interest per unit area of tissue (Figure 1Go). To adjust for this effect, a conversion factor from the measurement of lutein cell area throughout the cycle was used to quantify immunostaining for CD34 and {alpha}-smooth muscle actin. The conversion factors were calculated to give the stage with the largest lutein cell area a value of 1. Thus, the means of the early, mid- and late luteal phase values were divided by the mean of the rescued lutein cell area respectively. Conversion factors are shown in Table IGo and were used to convert immunostaining measurements after quantification which was performed using the same sized area of interest for all corpora lutea. The adjustment of the measurements for CD34 and {alpha}-smooth muscle actin was done by multiplying the mean area of immunostaining by the conversion factors.



View larger version (111K):
[in this window]
[in a new window]
 
Figure 1. Haematoxylin and eosin staining of a corpus luteum of the early luteal phase (a) and after HCG rescue (b) at same magnification (bar = 100 µm) demonstrating lutein cell size differences in vivo. The squares in both sections indicate the area of interest. Note that because of the enlargement of lutein cells during luteal rescue, less cells can be counted in the same sized square of (b) as compared with (a). (c) and (d) show the situation as a schematic diagram to illustrate the influences of changes in cell size of one cell type on the number of another cell type whose size remains constant. In (c) representing early luteal phase, the grey polygonal lutein cells are relatively small, so that in a given area the number of black endothelial cells, or pericytes, is relatively high. (d) represents the situation, e.g. during luteal rescue, in which lutein cell size is at least two fold increased compared with the early luteal phase. By enlargement of the grey cell size, the number of black cells in the same field as used in (c) become automatically reduced. Without adjustment of this effect the endothelial cell area or pericyte area would be underestimated.

 

View this table:
[in this window]
[in a new window]
 
Table 1. Derivation of a conversion factor for immunostaining results taking into account the change in granulosa lutein cell area throughout the luteal phase of the cycle. Values for the lutein cell area from 100 measured cells per corpus luteum are given as mean ± SEM (n = 4 per cycle stage). The largest cell size was measured during luteal rescue. Therefore, the conversion factor for luteal rescue was given a value of 1. The conversion factors for the other stages were calculated by dividing the lutein cell area for early, mid- and late luteal phase respectively by the value for luteal rescue.
 
Quantification of Ki-67 immunostaining
Sections were examined using a x20 lens and 6–10 randomly chosen fields of view analysed per corpus luteum. The image analysis system was set to measure the number of dark stained nuclei only (Ki-67 positive), and the number of dark and light stained nuclei (total cells) per field of view. The set-up of the programme was validated first by manual counting of samples from all luteal stages. To obtain a proliferation index, the number of Ki-67 positive cells was expressed as the percentage of total cell nuclei per field of view. The mean of 6–10 fields was taken as representative for each corpus luteum.

Quantification of Ki-67 and CD34 double labelling
The total number of Ki-67-positive cells and dual stained cells with CD34 was counted manually in 10 fields per corpus luteum at x400 magnification using a Zeiss microscope with a grid overlay. The percentage of proliferating endothelial cells was calculated for the total number of proliferating cells. The mean for the 10 fields was taken as representative for each corpus luteum.

Quantification of CD34/{alpha}-smooth muscle actin
Since immunostaining was either located to the cell cytoplasm or plasma membrane, visualization of individual cells was difficult. It was therefore not possible to measure absolute endothelial or pericyte numbers in the corpus luteum, so quantification based on area of immunostaining per unit area of the corpus luteum proved optimal. The area of expression of these antigens was measured in four randomly chosen fields using a x10 lens for each corpus luteum. The captured grey scale image was thresholded and converted to a binary image, and the area was quantified. The area was expressed as a mean value of the number of fields assessed. A conversion factor, as described above, was used for CD34/{alpha}-smooth muscle actin to adjust measurements in relation to the varying luteal cell size throughout the cycle.

Quantification of VEGF
Quantification of VEGF was performed using Adobe Photoshop 5.0 as described previously (Dickson et al., 2001Go). To ensure that only the VEGF immunostaining would be quantified and not the counterstaining, the colour contrast of the image was enhanced by adjusting the Hue/Saturation tools in the Image Adjust menu to constant levels. The area of expression of brown immunostaining for VEGF was measured using a x10 lens for each corpus luteum in as many fields of view as the size of the cross section of the corpus luteum would allow. The mean of all fields was taken as representative for each subject.

Statistical analyses
Data from each method of quantification was statistically analysed. Separate one way analysis of variance tests were carried out for each set of data, and a factorial Fisher's protected least significant difference post hoc test performed. P < 0.05 was taken as the level of significance for each test. Tests were performed using Stat View Version 4.0 and Statistics Package for Social Sciences 6.1 for Macintosh.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Granulosa lutein cell area
In the human corpus luteum, granulosa- and theca-derived lutein cells are distinguishable. Theca lutein cells are relatively small and do not undergo noticeable hypertrophy during the luteal phase. In contrast, granulosa lutein cells do hypertrophy, thus the change in granulosa cell cytoplasmic volume over the luteal phase contributes largely to the overall change in luteal volume (Figure 1Go). Therefore, the area of these cells was measured in luteal sections from each stage of the cycle. Measurement of granulosa lutein cell area demonstrated a marked increase from early to mid- (P < 0.001), and decrease from mid- to late (P < 0.01), luteal values. Following luteal rescue, these cells were significantly larger (P < 0.03) than at any other stage of the cycle. The conversion factors used to convert measurements after quantification, which were performed using the same sized area of interest for all corpora lutea, and lutein cell size are shown in Table IGo.

Ki-67 immunostaining and Ki-67/CD34 double labelling
Ki-67 immunostaining indicating proliferating cells was readily observable in all corpora lutea. A high level of Ki-67-positive cells was seen in sections from the early luteal phase (Figure 2aGo) which was markedly reduced in mid- (Figure 2bGo) and late luteal sections (Figure 2cGo). A similar pattern of staining as demonstrated in the early luteal phase was seen in sections after HCG treatment (Figure 2dGo). Double labelling for Ki-67/CD34 (Figure 2eGo) indicates proliferating endothelial cells. When quantified (Figure 2fGo), the proliferation index revealed a significant decrease (P < 0.05) in the percentage of Ki-67-positive cells from early luteal phase levels in the mid- and late corpus luteum and a marked increase in proliferation (P < 0.05) similar to early luteal phase levels, after HCG rescue.



View larger version (156K):
[in this window]
[in a new window]
 
Figure 2. Low power photomicrographs of Ki-67 immunostaining (dark nuclei) in (a) an early corpus luteum; (b) mid-luteal phase corpus luteum; (c) late luteal phase corpus luteum; and (d) after HCG rescue. Note the marked decrease from early to mid-luteal phase and the increase of proliferation after HCG treatment (Scale bar = 100 µm). (e) shows double labelling for Ki-67 (black nuclei) and CD34 (red stained cells) in an overview during luteal rescue and at higher magnification (inset, bar = 20 µm) to show how proliferating endothelial cells (arrows) and non-endothelial cells (arrowheads) may be distinguished. (f) Proliferation index, for Ki-67-positive cells (total cell proliferation, grey columns) and Ki-67/CD34 double labelled cells (endothelial cell proliferation, black columns) throughout the lifespan of the corpus luteum. Values are means of 3–6 corpora lutea ± SEM. Different upper cases denote significant differences (P < 0.05) of total cell proliferation between different cycle phases; different lower cases indicate significant differences (P < 0.05) of endothelial cell proliferation between different cycle stages.

 
The proliferation index for endothelial cell proliferation (Figure 2fGo) in relation to the total number of cells was calculated from the results of Ki-67 single and Ki-67/CD34 double labelling. Quantification revealed that during the early luteal phase 70 ± 4.6% of all proliferating cells are of endothelial cell origin. A significant decrease to 40 ± 2.9% (P < 0.05) endothelial cell proliferation was found during mid-luteal phase, and a further significant drop to 10 ± 3.5% (P < 0.05) endothelial cell proliferation was found during the late luteal phase. Total cell proliferation did not change from mid- to late luteal phase indicating, that proliferation of non-endothelial cells increases relatively during the late luteal phase. After HCG rescue endothelial cell proliferation was 59 ± 5%, which was significantly higher (P < 0.05) than values of mid- and late luteal phase.

CD34 immunostaining
Measurement of CD34 immunostaining was used to quantify endothelial cell area. Endothelial cells associated with small capillaries were apparent in the developing corpus luteum (Figure 3aGo) and by the mid-luteal phase the microvascular tree was extensive with staining associated with larger luminal vessels and small capillaries (Figure 3bGo). In the late luteal phase, vessels appeared smaller and more sparsely distributed throughout the corpus luteum (Figure 3cGo). After luteal rescue, vessels remained small but were more tightly packed (Figure 3dGo) signifying an increase in luteal microvasculature. Quantification (Figure 3eGo), confirmed a significant increase (P < 0.05) in endothelial cell area from the early to the mid-luteal phase, and a significant decrease in the late corpus luteum (P < 0.05). Following luteal rescue, endothelial cell area significantly increased (P < 0.05) compared with all other luteal stages.



View larger version (66K):
[in this window]
[in a new window]
 
Figure 3. CD34 immunostaining in the human corpus luteum. Photomicrographs of CD34 immunostaining in (a) an early corpus luteum, showing the presence of small capillaries (arrows) in the developing vasculature, and the absence of staining in the negative control section [inset in (a)]. (b) A mid-luteal phase corpus luteum showing the incidence of larger luminal vessels (LV) in addition to the small capillaries (arrows) in the established vasculature. (c) The late luteal phase, showing the presence of small and large luminal vessels; and (d) the HCG-rescued corpus luteum at the same magnification, in which there is a increased incidence of small luminal vessels throughout the corpus luteum. Scale bars = 100 µm. (e) Quantification of the area of endothelial cell staining. Values are means (3–6 corpora lutea per group) ± SEM. Different letters denote significant differences (P < 0.05) between groups.

 
{alpha}-Smooth muscle actin immunostaining
During the early luteal phase (Figure 4aGo), {alpha}-smooth muscle actin immunostaining was evident in large luminal vessels. During the mid-luteal phase (Figure 4bGo) pericyte staining was also found in smaller vessels and capillaries. A decrease in immunostaining was apparent during the late luteal phase (Figure 4cGo). After HCG rescue (Figure 4dGo), endothelium coverage by pericytes was seen in large luminal vessels as well as in numerous small capillaries. Quantification (Figure 4eGo) of the area of {alpha}-smooth muscle actin staining showed an increase (P < 0.0001) from the early to the mid-corpus luteum and a significant decrease (P < 0.0001) from the mid- to the late corpus luteum. The pericyte area increased to maximal levels (P = 0.02 as compared with mid-luteal levels) following HCG rescue.



View larger version (90K):
[in this window]
[in a new window]
 
Figure 4. Pericyte distribution in the corpus luteum. Low power micrograph of {alpha}-smooth muscle actin immunostaining in an early corpus luteum (a), showing that staining is associated with luminal vessels (arrows); (b) mid-luteal phase corpus luteum in which staining is associated with large luminal vessels (arrows) and some small capillaries (arrowheads); similarly in (c) a late corpus luteum (E = erythrocytes). After HCG rescue (d) an increase in pericyte staining is visible (arrow) especially in a increased number of small capillaries (arrowheads). Scale bar = 100 µm. (e) Quantification of pericyte area throughout the lifespan of the corpus luteum. Values are means of3–6 corpora lutea ± SEM. Different letters denote significant differences (P < 0.0001 for a–b and a–c; P = 0.02 for b–c) between groups.

 
VEGF immunostaining
In the developing corpus luteum, intense VEGF staining was seen at the tips of the finger-like processes of granulosa-derived lutein cells lining the former antrum (Figure 5aGo). There was also expression in other granulosa lutein cell areas and in the theca lutein cell compartment. By the mid-luteal phase (Figure 5bGo), expression had spread more evenly across the granulosa lutein cell compartment and decreased in theca lutein cells. VEGF immunostaining declined in the late corpus luteum (Figure 5cGo), with only small areas of punctate expression remaining localized predominantly to granulosa lutein cells. After HCG treatment (Figure 5dGo), an increase in VEGF immunostaining was seen across the corpus luteum. Expression remained localized to lutein cell cytoplasm and was uniformly distributed, similar to the pattern of expression in the mid-luteal phase. Quantification (Figure 5eGo) confirmed there is a significantly higher VEGF area (P < 0.05) in the presence of HCG when compared with the late luteal phase.



View larger version (97K):
[in this window]
[in a new window]
 
Figure 5. VEGF immunostaining in the corpus luteum. Early luteal phase corpus luteum (a) with VEGF (brown staining) in the cytoplasm of granulosa lutein cells. Note that lutein cells near the former antrum (A) in particular are strongly positive (arrows). (b) section of the mid-luteal phase, note that almost every steroidogenic cell is expressing VEGF. A marked decrease is observed during late luteal phase and only single cells (arrows) are stained for VEGF (c). In the rescued corpus luteum (d) a widespread distribution of VEGF staining is found. Scale bar = 100 µm. (e) Quantification of VEGF immunostaining throughout the luteal phase. Note the significant increase (P < 0.05) of VEGF area during luteal rescue compared with the late luteal phase. Values are means of 3–6 corpora lutea ± SEM. Different letters denote significant differences between groups.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The results of this study allow us to propose the following scenario with respect to the changes in angiogenesis and pericyte recruitment throughout the human luteal phase and after HCG-induced luteal rescue. During the early luteal phase in which the initiation of development of the luteal microvasculature is underway, intense angiogenesis is found as indicated by the highest rate of endothelial cell proliferation of the normal cycle. By the mid-luteal phase, endothelial cell proliferation continues at a lower rate. The formation of the microvascular tree is well developed as indicated by CD34 and {alpha}-smooth muscle actin immunostaining, with the presence of small capillaries and mature, luminal vessels. Late luteal phase regression is associated with a further decline in endothelial cell proliferation. However, total proliferative activity was comparable with that of the mid-luteal phase, indicating that non-endothelial cell proliferation has increased, which may be due to infiltration of the corpus luteum by cells such as fibroblasts during the regression process. Decreased area of pericyte staining further indicated that luteal regression is associated with the degradation of stabilised vessels. On the other hand, simulating early pregnancy in the human is associated with increases in cell proliferation comprising both endothelial cell and non-endothelial cell proliferation, endothelial cell content and pericyte recruitment, indicating an increased rate of angiogenesis together with vessel stabilization.

Our finding that cell proliferation was maximal in the early luteal phase and declined significantly in the mid- and late luteal phase in the human corpus luteum is in agreement with previous studies in the human (Rodger et al., 1997Go; Gaytan et al., 1998Go), marmoset monkey (Dickson and Fraser, 2000Go, Young et al., 2000Go), rhesus monkey (Christenson and Stouffer, 1996Go), sheep (Jablonka-Shariff et al., 1993Go) and cattle (Zheng et al., 1994Go). However, in these studies (Jablonka-Shariff et al., 1993Go; Christenson and Stouffer, 1996Go; Rodger et al., 1997Go) increased angiogenesis after luteal rescue is not apparent. Interspecies differences may account for these findings. However, it would not explain the different results of this study compared with results by Rodger et al. in the human corpus luteum (Rodger et al., 1997Go). Measurements by Rodger et al. were performed per unit area of luteal tissue and they did not address changes in the expansion of granulosa lutein cells during luteal rescue, which could serve to spread proliferating endothelial cells further apart and give a result not representative of the true rescued environment. The marked impact of lutein cell size on the proportion per unit area of other tissue compartments can be demonstrated by applying the conversion factors calculated in this study to the figures given in Rodger's study. After this adjustment a comparable increase in proliferation during luteal rescue (data not shown) is found as estimated in this study by calculation of Ki-67-positive nuclei per total number of nuclei.

Rodger et al., detecting endothelial cells by von Willebrand factor immunocytochemistry, further demonstrated an increase in endothelial cell number from the early to mid-luteal phase which remained high in the late corpus luteum and after luteal rescue (Rodger et al., 1997Go). However, a concomitant elevation of cell proliferation after rescue was not demonstrated. Again, the different results as compared with our study may be due to the failure of quantifying endothelial cell number per area without taking account of the tissue expansion and shrinkage throughout the cycle and after luteal rescue. In this study, after HCG-induced luteal rescue a significant increase in endothelial cell content of the corpus luteum was demonstrated. This, taken with the fact that endothelial cell proliferation increased at this stage, suggests that rescue of the human corpus luteum is associated with an increased angiogenic activity.

The molecular mechanisms that regulate the angiogenic process in the human corpus luteum are not fully elucidated, but clearly involve the expression of angiogenic factors such as VEGF and angiopoietins (Wulff et al., 2000Go). In this study, intense VEGF protein expression was observed in lutein cells adjacent to the antrum of the developing corpus luteum, giving support for the endothelial cell mitogenicity of VEGF, in the antral invasion of blood vessels forming the mature mid-luteal corpus luteum. High expression in the mid-luteal corpus luteum suggests a role for VEGF in the ongoing angiogenic process, and in the survival and maintenance of the extensive mid-luteal vasculature. In non-human primates, immunoneutralization of VEGF at this time results in a decline in angiogenesis, an increase in apoptosis and a decrease in luteal function (Dickson et al., 2001Go). It could also reflect non-angiogenic functions of VEGF, such as the regulation of vascular permeability. Area of immunostaining for VEGF was significantly increased in the rescued corpus luteum. This finding is consistent with increased levels of VEGF mRNA in the human corpus luteum after HCG treatment (Wulff et al., 2000Go), indicating that in the human the molecular environment would be conducive for intense angiogenic activity in the rescued corpus luteum.

The corpus luteum of pregnancy survives and functions for several months, requiring a stable vasculature with prolonged endothelial cell survival, and previous studies in the human (Rodger et al., 1997Go), rhesus monkey (Christenson and Stouffer, 1996Go) and sheep (Jablonka-Shariff et al., 1993Go) suggested that early pregnancy may be associated with an increase in vessel stability. Stabilization of the newly formed vasculature is achieved by recruitment of periendothelial support cells such as pericytes. Recruitment of pericytes has been documented soon after induction of bovine luteal angiogenesis and the percentage of vessels with associated pericytes increases as the luteal phase progresses (Goede et al., 1998Go). Coverage of blood vessels with differentiated pericytes can be regarded as a parameter of the maturity of a developing neovasculature. In this study, an increase in pericyte area was observed from the early to mid-luteal phase, indicating that after the initial angiogenic burst during luteal formation, newly formed vessels are stabilized in the functional corpus luteum of the mid-luteal phase. In addition, it was demonstrated that rescue of the human corpus luteum is associated with high pericyte coverage as observed by a further increase in area of {alpha}-smooth muscle actin immunostaining. Furthermore, it was shown that the percentage of non-endothelial cell proliferation (presumably pericytes) was highest during luteal rescue. It has been shown that <5% of lutein cells proliferate during luteal rescue (Rodger et al., 1997Go). Thus, it is suggested that the proliferating non-endothelial cells are most likely to be pericytes, indicating that pericytes are not only recruited but also proliferate as they coat the vasculature. The increase in pericyte coverage, together with increased VEGF protein and the continued presence of small capillaries with no associated pericytes, indicates that during luteal rescue both vessel growth and stabilization are occurring. It is suggested that this second angiogenic wave is restricted to the endothelium of small capillaries. These capillaries are not terminally differentiated by contact with pericytes recruited in the mid-luteal phase, and remain growth factor responsive. Pericyte recruitment also takes place during luteal rescue to stabilize the newly formed vessels. The extension of the vasculature in the corpus luteum of early pregnancy is probably needed to provide for the increased requirement of hormone precursors, nutrients and oxygen for increased progesterone synthesis and to deliver secreted progesterone into the periphery. Pericyte recruitment and vessel stabilization extends the lifespan of the endothelium which is required for the corpus luteum to function for the first three months of pregnancy.

In conclusion, in the human corpus luteum a cyclic dynamic process of vascular development and regression takes place in the normal menstrual cycle. The calculation of a conversion factor prevents changes in lutein cell volume affecting quantification of the vascular cells in question. The early luteal phase is associated with intense angiogenesis followed by formation of a stable vasculature during the mid-luteal phase, while late luteal phase regression is associated with degradation of capillaries and stable vessels. Moreover, it has now been established that simulating early pregnancy in the human corpus luteum by HCG treatment leads to increased expression of angiogenic factors and is associated with both increased angiogenesis and vessel stabilization.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
This work was supported in part by a grant (to C.W.) from the Deutsche Forschungsgemeinschaft


    Notes
 
3 To whom correspondence should be addressed. E-mail: Christine-wulff{at}onlinehome.de Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Christenson, L.K. and Stouffer, R.L. (1996) Proliferation of microvascular endothelial cells in the primate corpus luteum during the menstrual cycle and simulated early pregnancy. Endocrinology, 137, 367–374.[Abstract]

Dickson, S.E. and Fraser, H.M. (2000) Inhibition of early luteal angiogenesis by gonadotropin-releasing hormone antagonist treatment in the primate. J. Clin. Endocrinol. Metab., 85, 2339–2344.[Abstract/Free Full Text]

Dickson, S.E., Bicknell, R. and Fraser, H.M. (2001) Mid-luteal angiogenesis and function in the primate is dependent on vascular endothelial growth factor. J. Endocrinol., 168, 409–416.[Abstract/Free Full Text]

Duncan, W.C., McNeilly, A.S., Fraser, H.M. et al. (1996) Luteinizing hormone receptor in the human corpus luteum: lack of down-regulation during maternal recognition of pregnancy. Hum. Reprod., 11, 2291–2297.[Abstract]

Duncan, W.C., McNeilly, A.S. and Illingworth, P.J. (1998) The effect of luteal `rescue' on the expression and localization of matrix metalloproteinases and their tissue inhibitors in the human corpus luteum. J. Clin. Endocrino. Metab., 83, 2470–2478.[Abstract/Free Full Text]

Duncan, W.C., Cowen, G.M. and Illingworth, P.J. (1999) Steroidogenic enzyme expression in human corpora lutea in the presence and absence of exogenous human chorionic gonadotrophin (HCG). Mol. Hum. Reprod., 5, 291–298.[Abstract/Free Full Text]

Gaytan, F., Morales, C., Garcia-Pardo, L. et al. (1998) Macrophages, cell proliferation, and cell death in the human menstrual corpus luteum. Biol. Reprod., 59, 417–425.[Abstract/Free Full Text]

Gaytan, F., Morales, C., García-Pardo, L. et al. (1999) A quantitative study of changes in the human corpus luteum microvasulature during the menstrual cycle. Biol. Reprod., 60, 914–919.[Abstract/Free Full Text]

Goede, V., Schmidt, T., Kimmina, S. et al. (1998) Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis. Lab. Invest., 78, 1385–1394.[ISI][Medline]

Illingworth, P.J., Reddi, K., Smith, K. et al. (1990) Pharmacologic `rescue' of the corpus luteum results in increased inhibin production. Clin. Endocrinol., 33, 323–332.[ISI][Medline]

Jablonka-Shariff, A., Grazul-Bilska, A.T., Redmer, D.A. et al. (1993) Growth and cellular proliferation of ovine corpora lutea throughout the estrous cycle. Endocrinology, 133, 1871–1879.[Abstract]

Leung, D., Cachianes, G., Kuang, W. et al. (1989) Vascular endothelial growth factor is a secreted angiogenic mitogen. Science, 246, 1306–1309.[ISI][Medline]

Li, T.C., Rogers, A.W. and Dockery, P. (1988) A new method of histologic dating of human endometrium in the luteal phase. Fertil. Steril., 50, 52–60.[ISI][Medline]

Rodger, F.E., Young, F.M., Fraser, H.M. et al. (1997) Endothelial cell proliferation follows the mid-cycle luteinizing hormone surge, but not human chorionic gonadotrophin rescue, in the human corpus luteum. Hum. Reprod., 12, 1723–1729.[Abstract]

Wulff, C., Wilson, H., Largue, P. et al. (2000) Angiogenesis in the human corpus luteum: localization and changes in angiopoietins, Tie-2 and vascular endothelial growth factor messenger ribonucleic acid. J. Clin. Endocrinol. Metab., 85, 4302–4309.[Abstract/Free Full Text]

Young, F.M., Rodger, F.E., Illingworth, P.J. et al. (2000) Cell proliferation and vascular morphology in the marmoset corpus luteum. Hum. Reprod., 15, 557–566.[Abstract/Free Full Text]

Zheng, J., Fricke, P.M., Reynolds, L.P. et al. (1994) Evaluation of growth, cell proliferation, and cell death in bovine corpora lutea throughout the estrous cycle. Biol. Reprod., 51, 623–632.[Abstract]

Submitted on May 31, 2001; accepted on September 3, 2001.