|
|
||||||||
Endocrinology and Metabolism |
1 Department of Endocrinology, VU University Medical Centre, Amsterdam, The Netherlands.
2 Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, Toronto, Ontario, Canada.
3 Department of Laboratory Medicine and Pathobiology, University of Toronto, Ontario, Canada.
4 Department of Biochemistry and Molecular Biology, University of Athens, Athens, Greece.
5 Geestelijke Gezondheidszorg (GGZ) Delfland, Institute of Mental Health, Delft, The Netherlands.
aAddress correspondence to this author at: Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, 600 University Ave., Toronto, ON, M5G 1X5 Canada. Fax 416-586-8628; e-mail ediamandis{at}mtsinai.on.ca.
| Abstract |
|---|
|
|
|---|
Methods: Thirty-five male-to-female transsexuals receiving cyproterone acetate and estrogens (orally or transdermally) were included in this study. Serum and urine samples were collected before initiation of therapy and 4 and 12 months post therapy. ELISAs were used to measure multiple kallikreins in serum and urine.
Results: After antiandrogen and estrogen therapy, serum testosterone concentrations decreased dramatically, as did serum and urinary concentrations of human glandular kallikrein (hK2) and prostate-specific antigen (PSA; hK3). Statistically significant but relatively small changes in serum and urinary concentrations of many other kallikreins were also seen. Kallikreins in serum and urine were correlated before and after treatment.
Conclusions: The concentrations of hK2 and hK3, but not of any other kallikreins, decrease dramatically after combined antiandrogen and estrogen treatment in male-to-female transsexuals. The smaller responses of the other kallikreins presumably reflect their expression in multiple tissues.
| Introduction |
|---|
|
|
|---|
hK3, better known as PSA, is a valuable tumor marker of prostate cancer (7). The potential significance of the other kallikreins in tumor biology has been reviewed(1)(4)(5)(6)(8). Accumulating evidence indicates that, in addition to hK2 and hK3, many other members of the human kallikrein gene family are also implicated in carcinogenesis, in particular, in endocrine-dependent malignancies. Recent reports suggest a possible role for kallikreins in controlling vital elements of tumor biology, such as apoptosis, angiogenesis, and tumor metastasis, by cleavage of critical substrates such as growth factors, hormones, receptors, or extracellular matrix proteins(5)(9)(10). Furthermore, many kallikreins are known to be up- or down-regulated by steroid hormones in breast and prostate cancer cell lines and may represent important new tumor markers for the diagnosis, monitoring, and treatment of endocrine-dependent malignancies(4)(5)(9)(11)(12).
Studies on the regulation of the kallikrein genes have been performed mainly in vitro in prostate, endometrial, and breast cancer cell lines (13)(14)(15)(16). Most genes appear to be up- or down-regulated by estrogens, androgens, and progestins but with different potencies. One notable observation is the tissue-specific pattern of regulation of some genes (e.g., the prostate-specific regulation of KLK3 and KLK2) and the different patterns of hormonal regulation in different tissues; e.g., KLK4 is up-regulated by androgens in prostate cancer cell lines(17) and by estrogens in endometrial cancer cell lines(16). The control of kallikrein gene transcription may involve integration of many transcription factors and pathways, which provide diversity in regulation mechanisms and create opportunities for cell- and tissue-specific interactions(14).
Sex steroid hormones impact on the initiation and/or progression of endocrine-dependent malignancies (18). As outlined above, most kallikreins appear to be under sex steroid hormonal regulation. Taken together, kallikreins may belong to the targets by which hormones affect the initiation or progression of such tumors(19). Therefore, a better understanding of the hormonal regulation of kallikreins could increase our insights into tumor biology. To date, mainly cancer cell lines have been studied. The role of sex steroids on the expression of kallikreins has been studied in vivo only for hK2 and hK3(20)(21)(22). These studies, mainly performed in female-to-male and male-to-female transsexuals receiving cross-sex hormones, have shown that hK2 and hK3 are up-regulated in serum and urine by testosterone(21) and down-regulated by antiandrogens and estrogens, respectively(20). Recently, specific ELISAs for most other kallikreins have been developed. This allowed us to study the hormonal regulation of other kallikreins in vivo.
In this study, we examined the effect of hormone treatment on serum and urine concentrations of hK2 to hK8, hK10, hK11, hK13, and hK14, obtained from male-to-female transsexuals who were receiving combined estrogen and antiandrogen therapy.
| Materials and Methods |
|---|
|
|
|---|
hormone measurements
Serum measurements were performed for 17ß-estradiol (RIA; Sorin Biomedica), testosterone (Coat-A-Count RIA; Diagnostic Products Corporation), luteinizing hormone (Amerlite immunometric luminescence assay; Amersham Pharmacia Biotech), follicle-stimulating hormone (immunometric luminescence assay; Amerlite), and prolactin (IRMA; Biosource Technologies, Inc.).
kallikrein assays
For all kallikrein measurements in serum and urine, we used ELISA-type immunofluorometric procedures developed in-house. Most of these methods have been previously described and validated (Table 1
). The cross-reactivity of these ELISA assays against all other kallikreins was negligible (23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33).
|
statistical analysis
For data analysis, patients were subdivided into groups based on different types of treatment. Because the distributions of kallikrein concentrations were not gaussian, we used the nonparametric Wilcoxon signed-rank test to determine the differences between kallikrein concentrations before and after treatment. Correlations between kallikreins before and after treatment were assessed by the Spearman correlation coefficient. These tests treated each kallikrein concentration as a continuous variable. For all analyses, a P value <0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
|
Serum kallikrein concentrations before and after administration of antiandrogens plus estrogens, specified for the type of estrogen used, oral or transdermal, are presented in Table 3
. We found no difference between 4 months and 12 months of hormone administration (data not shown). The combined administration of antiandrogens and estrogens had a profound impact on serum concentrations of some, but not all, kallikreins. The effects of oral and transdermal estrogens were sometimes discrepant. Serum hK2 and hK3 concentrations were dramatically suppressed by both types of treatment. Serum hK6 and hK10 concentrations were also suppressed by both treatments but to a much lesser degree. Antiandrogens plus oral estrogens suppressed serum hK5 and hK11 concentrations, but these kallikreins were not suppressed by antiandrogens plus transdermal estrogens. Conversely, antiandrogens plus transdermal estrogens suppressed serum hK4 and hK7 concentrations, but these kallikreins were not suppressed by antiandrogens plus oral estrogens. No suppression by antiandrogens and either type of estrogen was noted for serum hK8, hK13, and hK14.
|
In general, the most consistent and dramatic changes in serum kallikrein concentrations after estrogen and antiandrogen treatment were seen only with hK2 and hK3.
The correlations between serum concentrations of kallikreins before hormone administration are shown in Table 1 of the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol52/issue7/. We found significant correlations between serum concentrations of hK2 and hK3 (positive) or hK10 (negative), between hK4 and hK13 or hK14 (both positive), between hK5 and hK6 (positive), and between hK7 and hK8, hK13, or hK14 (all positive). The strongest correlations (positive) were between hK2 and hK3 and between hK7 and hK8.
The correlations between serum kallikreins after 1 year of administration of 100 mg cyproterone acetate plus oral ethinyl estradiol are presented in Table 2 of the online Data Supplement. After 1 year of treatment, serum concentrations of hK2, hK4, and hK10 no longer showed any correlations with any of the other kallikreins. Serum hK3 no longer showed a correlation with hK2 but still showed positive correlations with hK6 and hK11. Serum hK5 was positively correlated with hK6, as before hormone treatment. Serum hK6 was correlated with hK13. Serum hK7 was positively correlated, as before hormone administration, with hK8, hK13, and hK14.
The correlations between serum kallikreins after 1 year of administration of 100 mg of cyproterone acetate plus transdermal 17ß-estradiol are shown in Table 3 of the online Data Supplement. Similar to the changes induced by cyproterone acetate plus oral ethinyl estradiol, serum concentrations of hK2, hK4, and hK10 no longer showed correlations with any of the other kallikreins. This was also the case for hK5, hK11, and hK13. Before hormone administration, serum hK6 correlated with hK5, but this correlation was lost; a negative correlation with hK13 became apparent, which was also observed with oral estrogens but in the opposite direction. Serum hK7 continued to correlate with hK8, but its correlations with hK13 and hK14, observed both before treatment and with oral estrogens, were lost after 1 year of treatment with cyproterone acetate plus transdermal 17ß-estradiol. Serum hK8 showed a positive correlation with hK14, not present before hormone administration, which not observed with oral estrogens.
The kallikrein concentrations in urine before and after 1 year of treatment with antiandrogens plus estrogens (oral or transdermal) are presented in Table 4
. No difference in effects was found between 4 months and 12 months of hormone administration (data not shown). The combined administration of antiandrogens plus estrogens had a profound impact on urine concentrations of some, but not all, kallikreins. There was a dramatic decrease in urine concentrations of hK2 and hK3 with both treatment regimens. The decrease in hK4 was much smaller and reached significance only with treatment with antiandrogens plus transdermal 17ß-estradiol. For the urine concentrations of hK5hK8, hK10, and hK13, we observed a trend toward increased concentrations, but statistical significance was only achieved for hK6, hK10, and hK13 on administration of antiandrogens plus oral ethinyl estradiol. No significant effects were observed on urine concentrations of hK5, hK7, hK8, hK11, and hK14.
|
The correlations between urinary kallikrein concentrations before hormone administration are shown in Table 4 of the online Data Supplement. We found significant correlations between urine concentrations of hK2 and hK3, hK10, and hK11 (all positive); between hK3 and hK4 (negative); between hK10 and hK11 and hK13 (all positive); between hK4 and hK6 and hK10 (both negative) or hK14 (positive); between hK5 and hK6, hK7, hK8, hK11, and hK13 (all positive); between hK6 and hK7, hK8, hK10, hK11, and hK13 (all positive); between hK7 and hK8 and hK13 (both positive); between hK8 and hK10 and hK13 (both positive); between hK10 and hK11 and hK13 (both positive); and between hK11 with hK13 (positive). The correlations between urinary kallikrein concentrations after 1 year of treatment with 100 mg of cyproterone acetate plus oral ethinyl estradiol are shown in Table 5 of the online Data Supplement, and the correlations after 1 year of treatment with 100 mg of cyproterone acetate plus transdermal 17ß-estradiol are shown in Table 6 of the online Data Supplement. In Tables 79 of the online Data Supplement, we present the correlations between urinary and serum kallikrein concentrations before and after the 2 types of treatment. The statistically significant correlations between serum and urine kallikreins were rare and weak.
| Discussion |
|---|
|
|
|---|
|
The "classic" hormone-dependent expression model proposes that androgen activation of the androgen receptor leads to HR binding to androgen response elements (AREs), thus leading to transcriptional activation of PSA and hK2. The genes that encode PSA and hK2 contain AREs in the proximal promoter region (100 to 400 bp) and an ARE-enhancer element located
3000 to 4000 bp upstream from the transcriptional start site. Therefore, similarities between kallikrein expression and hormone stimulation between these genes can account for their sensitivity to androgen stimulation (41)(42)(43)(44).
Norgestrel, a synthetic progestin, has also been shown to activate PSA and hK2 expression in BT-474 and T-47D breast cancer cell lines along with dihydrotestosterone (DHT) (34). This may explain why hirsute women continue to display high PSA with oral contraceptive treatments, as oral contraceptives usually contain a synthetic progestin ingredient(45)(46).
Several other kallikreins also show single-hormonedependent sensitivity. hK15 and hK1 are androgen sensitive in some tissue culture cells lines; however, their expression cannot be easily explained because the genes apparently lack AREs in their proximal or upstream promoter regions (17)(40). The same can be said for estradiol-specific expression of hK5, hK6, and hK8, in several breast cancer cell lines, but predictive estrogen response elements have not been found in upstream promoter regions(26)(37).
The hormonal regulation is further complicated by evidence that multiple kallikreins can be activated by several different hormones. These kallikreins include hK10, -11, -13, and -14. In BT-474 cells, these kallikreins are under the regulation of estradiol, whereas in T-47D cells, they are activated by DHT stimulation (37)(39). In MCF-7 breast cancer cell lines, these 4 kallikreins are stimulated by estradiol, DHT, and Norgestrel treatment (our unpublished data). Promoter-deletion constructs for hK10 failed to identify any proximal promoter region that are required for hormonal regulation in different cell lines, and scanning of upstream sequences did not identify any predictive HREs for any of these hormones(37). Although apparently lacking HREs, these kallikreins are sensitive to anti-hormone receptor treatment with Nilutamide and ICI 182,780(47)(48) androgen and estrogen inhibitors, respectively, thus also making it difficult to associate their coordinated expression to a single-hormonedependent trans-acting factor. Therefore, close examination of Table 5
shows how the kallikrein expression profile for one cell line can differ from another cell line when treated with the same steroid hormone.
Many kallikreins often have a localized tissue expression profile that is associated with different hormone-dependent endocrine organs. Therefore, hormone treatments that cause increases in kallikrein expression in serum and other biological fluids are most often associated with cancer. Studies have correlated HR status of cancer patients to the prognostic value of kallikreins. In turn, this has predictive value in selecting patients who are likely to respond to therapy. However, only prostate cancer, which is associated with increased PSA, is treated with antiandrogenic drugs because of its androgen-dependent regulation.
In humans, opportunities to monitor the effects of profound changes in hormonal milieus on biological variables are limited. Treatment of transsexuals with cross-sex hormones provides such an opportunity. The patients are usually rather young and not affected by concurrent illnesses; therefore, the alterations in biological variables can be reliably ascribed to changes in hormonal milieu after administration of cross-sex hormones.
The patients studied in this report were male-to-female transsexuals receiving a combination of the antiandrogen cyproterone acetate with an estrogen preparation, either oral ethinyl estradiol or transdermal 17ß-estradiol. This combination suppresses plasma testosterone concentrations to values below the detection limit of the assay (Table 2
). With this treatment, at the same time, a potent estrogen stimulus is introduced. The results of this study must therefore be viewed in the light of 2 significant alterations in hormonal milieu: the elimination of androgens and the introduction of a potent estrogen stimulus. The endocrine milieu introduced by the combination of cyproterone acetate and transdermal 17ß-estradiol is similar to that in women in the late follicular phase of the cycle, whereas the administration of oral ethinyl estradiol constitutes a much more potent estrogenic stimulus.
The above hormonal intervention produced strong and significant effects on various members of the kallikrein gene family. As previously described, the most pronounced effects were the decreases in serum and urinary hK2 and hK3 after administration of antiandrogen with either oral and transdermal estrogens (20). In addition, serum hK6 and hK10 decreased to a lesser, although statistically significantly, degree with both estrogen regimens. The decreases in hK2 and hK3 were observed earlier in our laboratory in a similar group of patients(20). In view of the report that hK2 and hK3 are up-regulated by androgens, this effect could probably be ascribed to androgen deprivation in the persons studied rather than to an estrogenic stimulus.
Although it has been reported that estrogens up-regulate many human kallikreins (Table 5
), this was not the case in our results. If anything, we observed modest decreases.
There were discrepancies between the effects on kallikreins of antiandrogens plus oral ethinyl estradiol on the one hand and antiandrogens plus transdermal 17ß-estradiol on the other. Serum concentrations of hK5 and hK11 were suppressed by administration of antiandrogens plus oral ethinyl estradiol, whereas they were not significantly suppressed by antiandrogens plus transdermal 17ß-estradiol. Conversely, serum concentrations of hK4 and hK7 were suppressed by administration of antiandrogens plus transdermal 17ß-estradiol but not by antiandrogens plus oral ethinyl estradiol. An intriguing finding is that the type of estrogen (oral or transdermal) had different impacts on some members of the kallikrein gene family, but the explanation for these differences between the estrogen regimen is not obvious. However, a parallel may be drawn with other studies that found discrepancies between the effects of oral ethinyl estradiol vs transdermal 17ß-estradiol. The "first-pass" of oral estrogens through the liver has been invoked to explain the increases in clotting factors (49), C-reactive protein(50), insulin-like growth factor(51)(52), and triglycerides and the suppression of plasma tissue-type plasminogen activator(50). Oral, but not transdermal, administration of estrogens lowers tissue-type plasminogen activator concentrations in humans without affecting endothelial synthesis. These changes are lower with the transdermal route of estradiol administration. Furthermore, oral ethinyl estradiol is more potent than 17ß-estradiol(53)(54).
In summary, the administration of antiandrogens plus estrogens to adult males had an impact on some, but not all, serum kallikrein concentrations. The design of the study does not allow us to determine whether these effects could be ascribed to the resulting profound androgen deprivation, to the induction of a potent estrogen stimulus, or to both. Administration of antiandrogens combined with oral ethinyl estradiol or transdermal 17ß-estradiol had similar effects on serum concentrations of several kallikreins, but surprisingly, some others were affected differently, which is potentially explained by the hepatic effect of oral ethinyl estradiol, which does not occur with transdermal 17ß-estradiol. The most clear effects of treatment with an antiandrogen plus estradiol were seen with hK2 and hK3 in both serum and urine (dramatic down-regulation). This is very likely attributable to the well-known up-regulation of these enzymes by androgens in both cancer cell lines and in vivo. Regarding the other kallikreins, 2 factors may complicate our interpretations: (a) most kallikreins other than hK2 and hK3 are not clearly regulated by the classic hormone/HR/HRE system but by more complex mechanisms; and (b) our data may reflect the highly specific expression of hK2 and hK3 in the prostate and the diverse expression of most other kallikreins in many different tissues.
| Footnotes |
|---|
2 Nonstandard abbreviations: hK1, pancreatic/renal kallikrein protein; hK2, glandular kallikrein; hK3 (PSA), prostate-specific antigen; hK4hK15, human kallikreins 415; HR, hormone receptor; HRE, hormone response element; ARE, androgen response element; and DHT, dihydroxytestosterone. ![]()
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |