Hormonal Signaling in Prostatic Hyperplasia and Neoplasia*

The PROSTATE is the site of two common diseases in aging men, benign prostatic hyperplasia (BPH) and prostate cancer (CaP). Although mortality rates due to BPH are less than 0.5 per 100,000 population, this disease affects the quality of life and health care costs exceed 4 billion dollars per year. If a man in the United States lives to be 80 yr of age, there is a 25% chance that he will require surgical treatment for BPH. In 1999, CaP will be diagnosed in 179,000 American men and it will cause 37,000 deaths. This represents the second most common cause of cancer deaths in men. Histological BPH increases with age in all ethnic groups. By age 80, it is present in more than 80% of men. In contrast, mean prostate weight at autopsy or prostate volume in community-based studies increases after age 50 or 60. Comparing 70–79-yr-old men, 93% of men in Olmstead County, Minnesota, but only 37% of Japanese men had volumes exceeding 20 mm3 (1). This suggests that there may be genetic differences among ethnic groups. However, greater prostate volumes in urban-dwelling Chinese than in rural-dwelling Chinese suggest that environmental factors contribute to prostate volume, and twin studies in Utah indicate that both genetic and environmental factors contribute to prostate volume. Population-based studies of prostate enlargement in Africans and African-Americans are lacking. Most recent data using National Hospital Discharge Data Survey indicate that the percentage of African-American men undergoing prostatectomy for prostate enlargement is similar to the percentage of whites when adjusted for age (2). This suggests that the prevalence of prostate enlargement in African-American men is similar to that in white American men. In 1999, CaP will account for 28.7% of the total number of new, nonskin cancers, but only 12.7% of cancer-related deaths for American males. These numbers imply that the large majority of affected men will die with this disease, rather than from it. This is, in part, due to the age at the time of diagnosis, to the low degree of malignancy of many CaP, and to the effectiveness of treatment. The epidemiology of CaP is complicated by the fact that 15–30% of men over the age of 50 have histological evidence of CaP at autopsy (latent CaP) (3). The factors responsible for causing a latent CaP to become a clinical CaP are poorly understood. The prevalence and mortality rates from CaP vary among different ethnic groups. They are higher in African-Americans (224 and 55/100,000 persons/yr), intermediate among whites (150 and 24/100,000), and lower among Orientals living in Asia (82 and 10/100,000) (4). This is in sharp contrast with the prevalence of latent CaP, which is roughly the same worldwide (5). Thus, some ethnic differences in the prevalence of prostate enlargement and significant ethnic differences in the incidence of clinical CaP seem to provide opportunities for learning more about the pathogenesis of these important diseases. Potential differences in sex steroid levels have been the focus for many studies. The development, growth, and differentiation of the normal prostate are regulated by androgens, acting through the androgen receptor (AR). The concepts that a normal androgen-signaling pathway is required for the development of a normal prostate and for the development of BPH and CaP come from the following observations: 1) abnormal androgen biosynthesis and inactivating mutations of the AR and 5α-reductase type II genes are associated with absent or rudimentary development of the prostate in 46-XY individuals; 2) there is anecdotal evidence that BPH and CaP do not develop in men who underwent castration around the time of puberty; 3) androgen can induce experimental prostate enlargement or cancer in susceptible animals. Although other hormones, such as insulin-like growth factor (IGF)-I, vitamin D, KGF, estradiol (E2), and PRL may affect the prostate, this review focuses mainly on the androgen-signaling pathway. Our aim is to critically assess the most important hormonal mechanisms related to the androgen signaling pathway that may be involved in the etiology and in the response to treatment of these two common disorders. It would be very convenient if one could correlate serum levels of sex steroids with risks of developing BPH or CaP. Although there are many studies, their significance is limited by several factors: 1) to establish linkage between development of a disease and certain steroid hormone levels, studies need to be prospective and should last until the disease of interest has developed; 2) circulating hormone levels vary significantly throughout the day and over the years. It is difficult to determine to what extent the level of a hormone measured once reflects a lifetime index of hormone status; and 3) the levels of circulating steroid hormones do not necessarily reflect their intraprostatic concentrations. Studies have been performed to see whether there is a correlation between sex steroid levels at various ages and the prevalence of prostatic diseases in the various ethnic groups. It seems that African-American fetuses are exposed to somewhat higher levels of testosterone (T) than white fetuses. Additionally, young adult African-American men have higher circulating levels of T than age-matched Caucasian-American men do. However, levels in young Japanese men are similar to Caucasians. Ethnic differences in male hair distribution suggest differences in serum androgen levels or in androgen effects on androgen-sensitive hair follicles and, perhaps, the prostate. Dihydrotestosterone (DHT) is the principle androgen for both. Hair follicles are known to have high levels of 5α reductase type 1 activity, and the prostate is known to have high levels of 5α reductase type 2 activity. Lookingbill et al. (6) evaluated 184 Caucasian and Chinese subjects. Mean chest hair scores using a scale of 0–4 were 3.0 vs. 0.8 (P < 0.0001), Caucasian vs. Chinese. Serum levels of total and bioavailable T did not differ, but serum levels of precursor androgens, dehydroepiandrosterone sulfate, and androstenedione were higher in Caucasians. Levels of 5α-reduced androgen products, AAG and androsterone glucuronide, were higher in Caucasians. A follow-up study sought to determine whether the lower 5α-reduced products in Chinese were caused by a reduction in precursor steroids or decreased 5α-reductase activity (7). They found no evidence of reduced 5α-reductase activity, but there was significant reduction in T production rates, plasma T, and sex hormone binding globulin (SHBG) levels in Chinese men living in Beijing compared with Chinese men living in Pennsylvania. No differences in T levels or metabolism were noted between Chinese and Caucasians living in Pennsylvania. These studies indicate differences in sex steroid levels among Chinese and Caucasians are due more to environmental factors than genetic factors. It is likely that differences in diets contribute to the differences because a diet higher in polyunsaturated fat intake and fiber is known to decrease serum total and free T levels. If the differences in hair patterns in Orientals and Caucasians are not due to differences in circulating levels of androgens or to differences in 5α-reductase activity, it would seem that they must be related to differences in the AR or to postreceptor signaling. Differences in AR or in postreceptor signaling also could contribute to the ethnic differences in the incidence of BPH and CaP. Another approach is to compare sex hormone levels in men who develop prostate enlargement or CaP with men from the same ethnic group who do not develop these diseases. Many investigators have evaluated sex steroid levels at the time when prostate enlargement was diagnosed and/or at the time of surgery for lower urinary tract symptoms and prostate enlargement. It should be noted that prostate size correlates poorly with symptoms and urine flow rates; however, men with prostate enlargement are three to four times more likely to undergo surgical intervention when followed for several years (8). Although some studies have noted differences in one or more hormonal parameters, consistent hormonal differences have not been observed within an ethnic group at the time when prostate enlargement was diagnosed. The relationship between plasma hormone levels obtained before development of clinical symptoms and subsequent surgical treatment for BPH has been evaluated in 320 men participating in the Physicians’ Health Study (9). An additional 320 men who did not undergo prostatectomy served as controls. The plasma specimens were obtained up to 9 years before surgery. Using multivariate analyses, they found a strong trend for increasing risk across quintiles of E2 (Q5 vs. Q1 OR = 3.56). This excess risk was confined to men with lower androgen levels. Plasma T was not correlated with subsequent prostate surgery. Sex steroid hormone levels and CaP have been correlated in two prospective studies. Serum levels for T, DHT, and AAG in 59 men who developed CaP between 1973 and 1994 were compared with those in 180 patients who were CaP-free in 1994 (10). No association was detected when the initial levels of these hormones were individually correlated with CaP risk. In another study, levels in 222 men who provided plasma in 1982 and developed CaP by 1992 were compared with those in 390 controls (11). When T and SHBG levels were adjusted simultaneously, CaP risk correlated positively with levels of T and negatively with levels of SHBG. Additionally, serum levels of DHT and AAG did not seem to be correlated to CaP risk. Epidemiological observations suggest the following factors influence the development of CaP: diet, genetic susceptibility, sexually transmitted agents, environmental carcinogens and endocrine function. There is good evidence for inherited and genetic components (12); however, the fact that Japanese and Chinese males migrating to the United States experience a sharp increase in CaP prevalence in the first and second generation, indicates the importance of environmental factors. There is a strong correlation between per capita fat intake and CaP incidence and mortality; however, saturated fat is thought to account for only 10–15% of the differences in CaP incidence. Perhaps, total calorie intake rather than fat intake, per se, should be considered. Restriction of energy intake in mice inhibits prostate tumor growth by inhibiting tumor angiogenesis. The model proposed by these authors predicts that total energy intake rather than fat may be a determinant for CaP growth. Many components of the androgen-signaling pathway have been investigated to identify potential mechanisms responsible for the development of macronodular BPH and CaP. Since the 1930s, it has been recognized that T is the principal circulating androgen secreted by the testis. However, DHT is the predominant androgen bound to the AR in the nuclei of target cells in the prostate. In the prostate, T is converted to DHT by the enzyme steroid 5α-reductase type II (SRD5A2). The effects of androgen are mediated by AR, a member of the steroid-thyroid-retinoid superfamily of nuclear receptors. Nonligand-bound AR is thought to be in the cytoplasm of target cells complexed to a number of ancillary protein members of the heat shock protein family. Binding of hormone is followed by dissociation from heat shock proteins, phosphorylation, dimerization, and binding to specific DNA sequences within or adjacent to androgen-responsive genes. The ligand-activated receptor regulates transcription of androgen-dependent genes. The AR contains two microsatellites consisting of poly-Q and poly-G repeats in exon 1. The length of these two repeats is highly polymorphic in the general population. Investigators agree that an expanded poly-Q tract is associated with reduced transcriptional activity (13). This modification is associated with normal 3H-DHT binding ability. This initial observation was followed by studies in which differences in the size of the poly-Q repeat were correlated to the transcriptional activity of AR. Some (14), but not all authors (15), have concluded that there is an inverse correlation between poly-Q size and transcriptional activity. The poly-Q repeats exert an inhibitory effect on transcription, either directly (14), or indirectly by affecting AR mRNA stability (16). Investigators have proposed several hypotheses to explain the development of pathological prostatic enlargement. Sex steroids are invoked directly or indirectly in each. Berry et al. observed that pathological prostate enlargement is present only when there is histological BPH. Assuming that this observation is correct, one might consider what causes micronodular BPH and what is responsible for the development of macronodular BPH. The pathogenesis of each may be different. In this respect, micronodular BPH is like latent CaP and prostate enlargement or macronodular BPH is like clinical CaP. It is useful to note that macronodular disease primarily involves the stroma. The increase in prostate volume seen with BPH is caused by cellular hyperplasia and reduced apoptosis. Stereological analysis indicates that the relative volume of stroma in macronodular BPH is increased 33% over that in the normal prostate. These findings have been confirmed by quantitative morphometry in men with symptomatic BPH (17). Apoptosis is reduced and the life span of BPH stromal cells is increased. Using estimates of superoxide dismutase activity, the average life span of BPH stromal cells has been estimated to be more than 30 yr (18). A recent study found the mean proliferation index in the epithelium (0.142 ± 0.097) and stroma (0.121 ± 0.082) to be similar (19). The apoptotic rate in epithelium (0.172 ± 0.156) was similar to the proliferative but no apoptotic cells were detected in the stroma. These investigators that stromal growth in BPH is due to proliferation in the of androgen is required for prostate development, it has been proposed that increased prostatic of androgen or increased androgen activity causes macronodular BPH. BPH has increased 5α-reductase activity however, the mRNA for both in BPH were decreased when compared with levels in normal from and normal men. studies that BPH higher of DHT, but et al. that differences were related to metabolism of DHT in before the time of DHT levels in epithelium from normal and BPH decreased with increasing age of the however, stromal DHT in both normal and BPH was not affected by T levels did not with increasing The is to have higher of T, DHT, and AAG than intermediate and Thus, evidence indicates that macronodular enlargement of the prostate is not caused by an increase in DHT, but DHT may AR AR levels, and and AR transcriptional activity have been in BPH higher and lower levels of AR mRNA have in BPH than in normal prostatic Using binding AR levels in BPH have been to be or increased. that AR is present in both BPH cells and stromal of AR are higher in the Using DNA from men who had lower urinary tract symptoms and an prostate by and who underwent prostate surgery and DNA from controls risk of surgery decreased with AR Comparing repeat length of or less to or the for prostate surgery was with a of to Thus, there is evidence for increased AR in the BPH and for poly-Q repeats in patients who undergo surgery for BPH. for the of is mainly on observations that E2 the effects of androgen in BPH in by increasing AR levels and, perhaps, by other mechanisms and the fact that the serum to T increases with age in men. of T to E2 or of androstenedione to also could contribute to higher prostate of mRNA can be detected by in most BPH but not levels of E2 and increase with age, in a very significant increase in the in the stroma receptor and mRNA have been primarily in the stroma. The recent of a second in the prostate analysis of It is of interest that prostatic hyperplasia is observed in aging mice with the did not prostate Thus, data are but they do not a strong for in the development and of macronodular BPH. of growth factors and growth factors are to be the factors that directly enlargement of the prostate. Many of these growth factors are at regulated by sex of growth factor production and has been compared with growth potential of prostatic stroma. studies of the pathogenesis of BPH have on growth factors such as KGF, and and their and increase with aging in the or of BPH. they may not be a for stromal but there is evidence that this pathway inhibits apoptosis. this time it is not whether the increases in these growth factors are or by the androgen-signaling pathway. The suggests that hyperplasia from an increase in the number of cells of at an age and after prostate size from normal adult to androgen or may affect the number of and this could determine the number of potential and potential prostate proliferation and of the An increased a decreased rate of or a of the two would increase prostate such would in a enlargement of the prostate. BPH is to the and primarily of stromal it is not these observations are to the development of BPH. studies suggest that can be caused by decreased of and protein of could cause hyperplasia and/or In there is evidence that the androgen-signaling pathway to the development of BPH. Differences in androgen signaling and signaling normal prostate development may be seems to be required for pathological prostate and it is likely that effects on cells are mediated by growth factors. The growth factors by both and mechanisms to proliferation and apoptosis. may contribute to this both and stromal cells contribute to prostate the increase in stromal cells in most with macronodular BPH. Population-based for the of mutations in various ethnic have correlated CaP risk with the prevalence of certain two mutations of this have been and the prevalence of which activity by 33% from is more common among men prevalence of CaP is the than among African-Americans and Caucasians prevalence of CaP is the which an enzyme with a higher for T is more common among men with CaP. Thus, these authors suggest that in the can affect activity, intraprostatic levels of DHT and risk of developing CaP. AR and AR is in all of CaP prostatic and disease both before and after androgen CaP are AR and, the the AR In of the of AR in of CaP, investigators have to determine whether mutations of this correlate with the development of CaP. have been detected only in of of CaP In in et of no AR was Thus, a prevalence of AR mutations in the AR mutations as a common in the pathogenesis of CaP. et al. proposed that inactivating mutations of AR, which are in latent CaP in Japanese men and absent in American men, would the of latent CaP a in this ethnic group AR in the size of the poly-Q repeat and transcriptional activity could in increased AR activity. When the number of repeats was compared with the incidence of CaP in the general population, there was an association between an increased risk of developing the disease and a poly-Q In some studies have also detected an association between a poly-Q repeat and the of disease, high histological and age of there is also a correlation between the number of and the increased higher mortality and more of CaP in the African-American population. A correlation between a poly-G repeat and CaP risk has also been by but not other However, because AR with a reduced poly-G tract has reduced transcriptional activity, it is not such receptor may to CaP. In the data are only for a of the androgen-signaling pathway in the pathogenesis of CaP. They identify a relationship between the prevalence of mutations and of CaP in various ethnic groups. However, it is not whether this has clinical They indicate that AR mutations do not a significant in the pathogenesis of CaP. they identify a correlation between in the size of the poly-Q repeat of the AR with CaP However, at the transcriptional of AR with various repeats have been performed only in It is not known if they have significant effect in This very important will be by the development of AR with various length of the are only for or symptomatic BPH. macronodular BPH is that decrease DHT levels T and DHT levels or androgen effects prostate volume only by symptoms and increase urine flow only by an average of less than 2 of prostatic DHT levels can of the disease however, prostatic DHT not provide good for to BPH. The for the limited effectiveness of that DHT and/or T are poorly understood. Although AR are present in the and stromal cells of the prostate, only the to undergo in patients with a 5α-reductase Thus, 5α-reductase may be more useful in rather than enlargement of the prostate. with CaP have been with androgen the observation that of androgen by castration of the disease in as many as 80% of the a response of CaP cells the to in the of a very low of of men with die of CaP. Thus, the of androgen in CaP is one of the most in this of to explain the of androgen have on the various components of the androgen-signaling pathway. In of the of AR in most of CaP, many of these hypotheses are on of this 1) of the AR which would tumor growth at very low of the 2) AR mutations associated with a a a receptor protein that is by other than inactivating mutations of the AR, CaP cells with a and in the size of the repeat of AR that may increase response to with and of with the to growth and 3) increased intraprostatic of DHT to of of AR by signaling mediated by with various of androgen such as and of after androgen Most of these hypotheses are by significant However, at it is not whether some of such as of AR, of or growth, are AR or AR mutations have been detected in a of patients with CaP In have detected mutations in specimens of prostate cancer et of with these AR have the of some of and mechanisms of growth. Thus, on these one could that AR mutations might growth and/or response to treatment. it to be using in that these mutations an in the of CaP. Serum sex steroids within the normal do not correlate with development of prostatic enlargement and however, BPH and CaP do not in males who develop androgen before age 20 if they do not androgen Although most investigators that androgens the development of macronodular BPH and CaP, the evidence indicates that most of these patients have normal levels of androgens acting a normal androgen-signaling pathway. The importance of environmental factors is by differences in serum hormone levels between Chinese living in and differences in prostate size between and urban-dwelling differences in prostate size among and the increased prevalence of CaP in second and men. The of an abnormal androgen-signaling pathway is more likely when CaP androgen the causes of androgen or mutations of the AR and growth by seem most for are than for It is that a large percentage of the male will a 5α-reductase to prostate enlargement. However, if the were to that caused a significantly lower incidence of both prostate enlargement and CaP, many men would prostate enlargement or CaP is it is that will be A approach is to the apoptotic pathway of the have the effectiveness of this approach in et for The is to develop an all of the