| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
1 Institute for Mind and Biology, Departments of
2 Psychology
3 Medicine
4 School of Social Service Administration, University of Chicago, Illinois.
Address correspondence to Martha K. McClintock, Department of Psychology, University of Chicago, 5730 Woodlawn Ave., Chicago, IL 60637. E-mail: mkm1{at}uchicago.edu
 |
Abstract |
|---|
|
NEED FOR SPECIFYING MULTIPLE ENVIRONMENTS INTERACTING WITH GENES |
|---|
; Yu et al., 2002). How does the amount of information in the genome compare with that needed to account for human behavior? The human brain has 1010 neurons interconnected by 1015 synapses, with 10 synaptic events per synapse per second. Thus, during the half hour spent reading this article, a person's brain will experience at least 1.8 x 1020 synaptic events. The difference in the amount of information between genome and brain activity, let alone brain structure and the rest of the body, is on the order of 1016. In miles, this is the diameter of a galaxy.
Obviously, the genome can only create a living being in interaction with its environment—the only other source of information in gene–environment interactions. A common misconception is to view this interaction as hierarchical—to view genes as the fundamental building blocks that interact with the nuclear environment to create cells, which interact to create structures or physiologic systems upward to individual psychology and social behaviors. But this view ignores the fact that a genome is not created de novo—It evolves through selection at the level of the individual, in dynamic interaction with its social and physical environment, enabling survival and production of offspring that also survive to reproduce. This creates a selection pressure for genes that can or cannot be expressed in response to the demands of the environment. Therefore, the interaction between genes and environment is reciprocal—including not only genetic production of traits, but environmental regulation of genes throughout the life span. Indeed, only 25% of the variability in the life span of humans or baboons is attributed to genetic factors (Martin, Mahaney, Bronikowski, Dee Carey, Dyke, & Comuzzie, 2002; Skytthe et al., 2003).
Both the development of traits and the regulation of gene expression involve interactions not only with "the" environment but with many environments at multiple levels of analysis, ranging from the ecosystem, social systems, and the individual's immediate experiences, as well as environments within the individual set by physiologic systems, cell–cell interactions, structures within the cell's cytoplasm and the nucleus. These interactions also take place over multiple time spans: life in the moment (milliseconds to hours), life span development and aging (decades), cultural and ecologic systems (centuries), and evolutionary time. These multiple levels and time frames are all nested within one another, all potentially operating simultaneously. Thus, in principle, one cannot reduce the understanding of behavior or physical traits to genetic information. Nor can we know a priori which levels and time frames will be both necessary and sufficient for explaining the emergence of genetic traits.
One effective strategy for identifying environments that regulate gene expression is to begin with a genetic trait for which specific genes are known. Then we can ask: What is the full range of environments that affect the expression of those genes? To illustrate these general principles, we present a hypothesis based on a synthesis of our collective research, which spans levels of analysis from ethnic populations to mechanisms for cell death and gene regulation. The trait in question is mammary cancer.
Mammary cancer develops after multiple environmental events, or "hits," to the genome as well as inherited mutations. This involves a variety of tumor suppressor genes such as BRCA1 and BRCA2 as well as cooperative oncogenes such as HER-2/neu and MYC (Futreal et al., 1994; Lancaster et al., 1996; Miki, Katagiri, Kasumi, & Yoshimoto, 1996; Schmitt & Reis-Filho, 2003). Second, cells dividing unchecked by normal programmed cell death, termed "apoptosis," are at greater risk for mutations producing cancer (Bai et al., 2001; Brash & Ponten, 1998; Soung et al., 2004; Sun, Li, & Sun, 1999). What environments might regulate these genetic mechanisms, and could they range from the cellular to the social?
|
DISPARITY OF BREAST CANCER MORTALITY BETWEEN BLACK AND WHITE WOMEN |
|---|
; English, Cleveland, & Barber, 2002; Hankey, Miller, Curtis, & Kosary, 1994; Shiao, Chen, Lehmann, Wu, & Correa, 1997). The difference in age dynamics is particularly striking, with young Black women experiencing a sharp increase between 30 and 44 years of age and then a relatively low rate thereafter that is relatively independent of age. In sharp contrast, breast cancer incidence in White women increases exponentially with age, with the greatest age-dependent frequency occurring after menopause. Black women are also more likely than Whites to have massive nonmalignant fibroadenomas, particularly in puberty (El-Tamer, Song, & Wait, 1999).
This striking health disparity likely arises from the reciprocal interplay of culture and biology that defines ethnicity and race. It is well established that human populations from a specific geographic origin can have lower genetic variance for some traits than do humans as a whole. Indigenous tribes of the island of Taiwan are genetically homogeneous within a tribe but diversified among them. Moreover, they have little genetic relationship to the Han of Mainland China (Lin et al., 2000). Likewise, the variant of a gene that increases the risk of prostate cancer (CYP3A4*1B) is found in 76% of Africans (N = 391), 58% of African Americans (N = 261), 7% of U.S. Caucasians (N = 573), and 0% of Asians (N = 210) (Rebbeck, Jaffee, Walker, Wein, & Malkowicz, 1998; Zeigler-Johnson et al., 2002).
The disparity in mammary cancer may involve the BRCA1 gene. BRCA1-associated breast cancers occur at an earlier average age (44 years) than sporadic breast cancers (Ford, Easton, Bishop, Narod, & Goldgar, 1994), and it has been observed that Blacks have a greater breast cancer incidence between 30 and 49 years than Whites (Adebamowo & Adekunle, 1999; Eley, 1994). Additionally, compared with noncarriers, BRCA1-associated breast cancers are characterized by higher than expected frequencies of medullary or atypical medullary carcinoma, poor differentiation (high tumor grade), aneuploidy, high S-phase fraction, and hormone receptor negativity (Phillips et al., 1999; Shiao et al., 1997). These aggressive histologic features are also characteristic of breast cancer in Black women. These similarities suggest that alterations in BRCA1 or related pathways might contribute to breast cancer in young Black women, although limited data are available from this population to evaluate the possibility.
Inherited germ-line mutations in BRCA1 may reflect a genetic founder effect and create disparities in breast cancer among young Black women in the United States and Nigeria. Indeed, unique mutations in BRCA1 (1832del5, 5296del4, and 3883insA) are found in Black but not White families (Gao, Neeuhausen, Cummings, Luce, & Olopade, 1997; Gao et al., 2000). These, as well as other genes with less penetrance, may be shared among all Black women of the African diaspora, particularly those originating in West Africa. Clinically, this is plausible, and genetic tests of tumors of Nigerian women are underway. Supporting this hypothesis is the clinical observation that Black and White women in Nigeria have a similar disparity in breast cancer and disease. Among young women, Blacks are most likely to develop breast cancer. The average age at diagnosis is 43 years, whereas Whites of European ancestry do not develop breast cancer until 60 years of age on average (Adebamowo & Adekunle, 1999; Ihekwaba, 1992). Again, Black pubertal girls are more likely to develop rapidly growing fibroadenomas (Schneider, Laubenberger, Kommoss, Madjar, Grone, & Langer, 1997).
However, inherited genetic mutations can be only one part of the disparity in breast cancer. Overall, only 5–10% of breast cancers have germ-line mutations as their origin (Kelsey, 1989; King, 1992; Skolnick & Cannon-Albright, 1992). Moreover, not all women with germ-line mutations of BRCA1 develop breast cancer. Finally, there is no racial disparity in the frequency of germ-line (i.e., inherited) mutations (Gao et al., 1997, 2000). Environmental and physiologic factors regulate their penetrance. In these cases, psychosocial factors may increase the probability of somatic gene alterations, leading to breast cancers in women without a family history.
Many health behaviors, diets, and environmental factors modulate breast cancer risk, both in humans and in animal models. These include, among others: obesity, lack of exercise, high dietary fat, and exposure to carcinogens (Bernstein, 2002; Kim, 2002; Salih & Fentiman, 2001). In addition, there are disparate availability and access to health care systems, which can increase mortality from breast cancer (Coughlin, Thompson, Hall, Logan, & Uhler, 2002). Many of these factors, however, have been insufficient to explain the extreme health disparity in breast cancer between Blacks and Whites, particularly in young women (Jatoi, Becher, & Leake, 2003; Marie Swanson, Haslam, & Azzouz, 2003). Moreover, cancer is a multistep process, and there are undoubtedly a host of ways that the environment and behavior can increase the series of mutations necessary for the initiation and growth of mammary cancer.
In this article, we identify three environments at different levels of analysis: social isolation, hypervigilance (constant alertness for potential danger), and ovarian function throughout the life span. Each has a demonstrated disparity between Blacks and Whites, and each, in animal models, increases risk factors for mammary cancer. We hypothesize that it is the interaction of each of these environments that increases mammary cancer risk. The interaction changes cellular and genetic functions, such as preventing programmed cell death in breast tissue, and creates somatic alterations in tumor-regulating genes, such as hypermethylation of BRCA1.
|
PSYCHOSOCIAL PREDISEASE PATHWAYS AND RISK FACTORS |
|---|
; House, Robbins, & Metzner, 1982) demonstrated that social isolation and felt loneliness (i.e., perceived and self-reported) were correlated with high rates of all-cause mortality. The question now becomes: What are the myriad routes by which the social environment "gets under the skin" and exacerbates disease (Harrington, 2000; Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002)? Cacioppo and colleagues (Cacioppo et al., 2002a, 2002b; Hawkley, Burleson, Berntson, & Cacioppo, 2003) have identified some possible routes. Felt loneliness is associated with increased total peripheral resistance (i.e., activation of the sympathetic nervous system), disrupted sleep, and altered neuroendocrine function. Importantly, neither the frequency of actually being alone nor health behaviors differed between the lonely and nonlonely groups, providing strong evidence that it is felt loneliness, rather than being alone, that is the important component of a predisease pathway for cardiovascular disease.
There is emerging evidence that Blacks, as a group, are socially isolated and lonely. In Heat Wave: A Social Autopsy of Disaster in Chicago, Klinenberg (2002) determined that elderly poor Blacks were the group most likely to die during a recent Chicago heat wave. Hispanics with the same socioeconomic profiles were much less likely to die, because their social structures provided relationships with people who looked after them, whereas poor Blacks were the most isolated. Also, Patterson (1997) in his book titled The Ordeal of Integration: Progress and Resentment in America's "Racial" Crisis, while discussing marriage patterns, suggests that Blacks might be the "loneliest people in the world."
Other psychological environments may also provide a route "under the skin." We anticipate that high rates of crime and dilapidation of housing in some Black neighborhoods on the South Side of Chicago contribute to health disparities in a number of ways. First, violent crime decreases social trust and increases the expectation that others will take advantage of you (Wilkinson, 1999). And, those who do not trust others are more likely to live alone. Crime likewise contributes to a threatening environment and the need for hypervigilance for potential threats, especially if one lives alone in unsafe housing (Sampson, Raudenbush, & Earls, 1997). Black women might also be more hypervigilant, by virtue of the disproportionate burden of economic uncertainty that they share (Steele & Sherman, 1999; Wilson, 1996). Living as a minority itself can also entail a daily accumulation of stressors, loneliness, and social isolation (Williams, Neighbors, & Jackson, 2003).
Finally, the problems of an extended family environment fall disproportionately on many middle-aged women, who are at risk for breast cancer, such as the increasing number of Black women who are faced with parenting their grandchildren (Waite, 2000). For many women, this may be exacerbated in the household context, with its associated daily hassles and demands. In contrast, social supports are certainly available in Black communities, particularly attending church, which is associated with lower mortality from cancer (Hummer, Rogers, Nam, & Ellison, 1999). As Waite and Hughes argue (Waite, 2000), the key factor in determining hypervigilance and social isolation may well not be the actual burdens, threats, and stressors present in the environment, but rather the fact that they outweigh the available supports.
Such social and psychological disparities between Blacks and Whites are not unique to Americans or even to the contrast between Blacks and Whites. In Ibadan, Oyo State, Nigeria, psychosocial disparities among Nigerian women appear to be as marked as those between some Black and White women in America. Ibadan has been a major city of the Yoruba people for hundreds of years, and women of families who have lived there for generations enjoy considerable family and financial supports, as do the highly educated élite. In sharp contrast, other women are caught in social conflicts caused by dramatic rapid modernization. This is particularly true for women transplanted from rural society to the city in response to a rapidly globalizing economy, creating conflicts between urban and rural practices and values (Arason, Barkardottir, & Egisson, 1993). It is among these women that social isolation and vigilance are likely to be highest. Rural women who move to the city to find better markets for their small-scale commerce (e.g., Ijesa osomaalo, or cloth traders) often lose the social supports typical of their rural community (van't Veer et al., 2002). This is particularly so for women of non-Yoruba tribal groups such as Igbo or Housa.
Such disparities in social support have been documented in South Africa; urban Blacks experience lower social support than their rural counterparts (Mboya, 2000). In rural towns as well, commercialization of the economy has changed the traditional system of economic supports flowing along family lines (Guyer, 1990). Even when members of the same family lineage move to Ibadan, few live in the same urban compounds or housing complexes (Arason et al., 1993), eroding economic and social supports enjoyed by extended family living together in a rural town. In becoming urbanized, women may also lose the highly valued communal quality of rural life (Gugler & Flanagan, 1978). Finally, Yoruba women traditionally expect to be financially independent, yet the urban economy amplifies the disparity of women's earning capacity (Zeitlin, Megawangi, Kramer, Colletta, Babatunde, & Garnabm, 1995). Women have more difficulty than do men finding jobs that offset the high cost of urban living. Often, jobs are not close to their homes and childcare support is scarce (van't Veer et al., 2002). Thus, we hypothesize that it is these Nigerian women in particular, not all Black Nigerian women, that are at risk for early aggressive breast cancer.
|
DISRUPTION OF DNA METHYLATION IN SPORADIC BREAST CANCER |
|---|
; Futreal et al., 1994; Hosking et al., 1995; Lancaster et al., 1996; Merajver et al., 1995; Miki et al., 1996; Takahashi et al., 1996; Teng et al., 1996). Despite the lack of heritable mutations, several lines of evidence implicate malfunction of BRCA1 and BRCA2 genes in sporadic tumors. Such malfunction can be acquired during a woman's lifetime through a variety of mutagenic environmental events, particularly random mutations during failure of apoptosis (cell death), as is discussed later. More interestingly, malignancy is associated with hypermethylation of tumor-suppressor genes, which "silences" it or its promoter, and hypomethylation of tumor-enhancing genes.
Disruption of the normal DNA methylation patterns is an established common hallmark of human cancer cells. In a healthy cell, the DNA methylation patterns are conserved through cell divisions, allowing the expression of the particular set of cellular genes necessary for that cell type and blocking the expression of exogenously inserted sequences (Esteller et al., 2001). Cancer cells often exhibit the dual phenomena of global hypomethylation accompanied by hypermethylation of several small CpG islands, an area, typically near promoters, with a high frequency of the C-G sequence, connected by a phosphor-diester bond (Jones & Laird, 1999). The aberrant methylation of the CpG island located in the 5'-promoter region of several tumor suppressor genes such as hMLH1, VHL, CDH1, p16INK4a, and APC shuts down the expression of these contiguous genes.
Even though it is not mutated, BRCA1 is likely involved in sporadic tumors. Its high frequency of loss of heterozygosity in sporadic tumors (Cropp et al., 1994; Foulkes, Black, Stamp, Solomon, & Trowsdale, 1993; Lindblom, Skoog, Andersen, Rotstein, Nordenskjold, & Larsson, 1993), decreased activity in sporadic tumors once they become invasive (Thompson, Jensen, Obermiller, Page, & Holt, 1995), increased proliferation of mammary epithelium with antisense oligonucleotides to BRCA1 (Rao, Shao, Ahmad, & Reddy, 1996), and the decreased tumorigenicity of cell lines derived from sporadic tumors with the introduction of a normal copy of the BRCA1 gene (Holt et al., 1996) all indicate involvement in sporadic breast cancer. For BRCA2, the loss of heterozygosity observed in 30–40% of sporadic primary breast cancers suggests it, too, is involved in sporadic cases (Cleton-Jansen et al., 1995).
Promoter hypermethylation has recently been suggested as mechanism for BRCA1 inactivation in sporadic breast and ovarian cancers ranging in frequency from 11% to 20% (Biano, Hussey, & Dobrovic, 1999; Catteau, Harris, Xu, & Solomon, 1999; Dobrovic & Simpfendorfer, 1997). In contrast, BRCA2 seems to undergo hypomethylation and is frequently overexpressed in tumor cells (Thyaga Rajan & Felten, 2002), hence our focus on BRCA1 promoter methylation as a candidate for environmental regulation of breast cancer. The challenge, then, is to establish physiologic mechanisms through which psychosocial environments such as social isolation and hypervigilance can alter methylation of specific cancer genes and their promoters or increase the probability of spontaneous mutations through hyperproliferation of cells.
|
A RODENT MODEL FOR ENVIRONMENTAL REGULATION OF BREAST CANCER GENETICS |
|---|
). Thus, the genetic, hormonal, and immunologic systems for coping with disease co-evolved in the physiologic environment associated with social living. When the female rats' social system is disrupted by isolation and loss of supportive social interactions, they develop mammary carcinomas and hyperplasia at four times the rate of their same aged counterparts living in a group (Figure 1). This occurs at 14 months of age, equivalent to 35–45 years of age in humans (Minino & Smith, 2001) and similar to the peak of early onset of premenopausal breast cancer observed in Black women (English et al., 2002). By 17 months of age, the mammary cancer burden in isolated females is 16 times that of those living in groups (Hermes & McClintock, submitted). In contrast, group-living female rats do not develop a high rate of mammary tumors until late in their life span, at 23 months (see Figure 1), which is equivalent to 60–70 years of age in humans, similar to the pattern observed in postmenopausal White women.
|
). Neophobic males and socially isolated females both have prolonged secretion of glucocorticoids in response to the stressors of everyday life (Cavigelli & McClintock, 2003; G. Hermes & M. K. McClintock, submitted). In this rodent model, the disparity in early fatal mammary tumors cannot be genetic, because the animals are randomly assigned to their respective social conditions, balanced for family of origin. Moreover, obesity and exercise levels do not explain the disparity. In addition, they all live in the same laboratory room where they can see, smell, and hear each other, with the same food, light, temperature, and animal care protocols. Thus, in the animal model, it is the psychosocial environment that is regulating the cellular and genetic processes producing tumors.
Our rat model focuses on tumor development in the context of natural physiology, particularly the dynamics of undisrupted ovarian function throughout the life span and responses of the adrenal axis to mild stressors during a protected life in a laboratory colony. Thus, it extends the scope and generality of in vivo models of cancer, beyond artificially manipulating hormonal states through chronic implants of different hormones. Here we have the opportunity to consider the dynamics of natural variation in ovarian and adrenal states as mechanisms of mammary development and tumorigenesis.
|
PUBERTY AND ADULT OVARIAN FUNCTION |
|---|
). Strikingly, there is a much smaller disparity in age at menarche, maturation of ovarian function, between Black and White girls (Cameron, Grieve, Kruger, & Leschner, 1993; Herman-Giddens et al., 1997; Lee, Guo, & Kulin, 2001; Morrison, Barton, Biro, Sprecher, Falkner, & Obarzanek, 1994; Mul, Fredriks, van Buren, Oostdijk, Verloove-Vanhorick, & Wit, 2001). Therefore, a crucial pubertal risk factor may not be early puberty per se, but instead prolonged breast development prior to the onset of menarche and afterward during cyclic exposure to ovarian steroids.
During the interval between breast development and menarche, the hormonal milieu differs markedly from that once regular ovarian cycles are established. Sex hormone binding globulin is variable (Bedecarras, Gryngarten, Ayuso, Escobar, Bergada, & Campo, 1998), growth hormone is higher than it is after menarche, with higher basal and nighttime levels and more high pulses (Neville, McFadden, & Forsyth, 2002; Wennink, Delemarre-van de Waal, Schoemaker, Blaauw, & van den Braken, 1991), and estradiol release is pulsatile, rather than continuous, particularly at night (Mitamura, Yano, Suzuki, Ito, Makita, & Okuna, 2000). Each of these hormone dynamics affects mammary development and the development of fibroadenomas in Black adolescents (Naidu, Thomson, & Nirmul, 1989) as well as risk for mammary cancer. It is intriguing that the Black African populations in which these pubertal disparities do not occur are also those with relatively high social supports and networks: that is, rural, not urban, girls in South Africa (Cameron & Wright, 1990; Cameron et al., 1993) and upper middle class urban schoolgirls in Nigeria (Fakeye & Fagbule, 1990).
If exposure to ovarian steroids after menarche is an environmental risk factor, it is likely that the temporal dynamics of spontaneous ovarian cycles are critical. It is well established that it is the dynamic change in hormones that regulates other endocrine systems (Larsen, Kronenberg, Melmed, & Polonsky, 2002). A rapid increase in estradiol, not its level, triggers the luteinizing hormone surge. It is the rise in corticosterone, not its level, that regulates negative feedback on the hypothalamic–pituitary–adrenal axis.
In both humans and rats, social interactions, mediated by pheromones, change ovarian function. Specifically, they regulate the timing of the preovulatory surge of luteinizing hormone, thereby affecting the duration of unopposed estrogen, prolactin pulses, and corticosterone levels (McClintock, 1978, 1983, 1984; Stern & McClintock, 1998). Social isolation disrupts this social regulation of ovarian function. It also changes independent neuroendocrine mechanisms underlying reproductive behaviors (Gans & McClintock, 1993; Gans, Stamper, Butler, & McClintock, 1995) and undoubtedly alters the hormonal milieu of mammary tissue during puberty and adulthood.
There is a paradox, however, later in the life span when tumors are developing, social isolation accelerates reproductive aging, and isolated females at the age of tumorigenesis are no longer having spontaneous ovarian cycles (Figure 2) (G. Hermes & M. K. McClintock, submitted; LeFevre & McClintock, 1988, 1991). They have reached reproductive senescence and are producing only low tonic levels of estrogen. The strong role of estrogen receptors in some mammary cancers predicts the opposite pattern in rats as well as humans. This points to yet another hormonal environment created by social isolation that may accelerate tumor growth late in the life span.
|
|
STRESS HORMONES AND INHIBITION OF PROGRAMMED CELL DEATH |
|---|
). Thus, it appears to accelerate aging of the hypothalamic–pituitary–adrenal axis and hippocampus (Sapolsky, Krey, & McEwen, 1986). The failure of socially isolated rats to rapidly return to a baseline level of corticosterone is postulated to result in chronic GR activation, thereby inhibiting apoptosis of the ductal epithelium and increasing risk for spontaneous mutations causing mammary cancer.
The identification of antiapoptotic signaling mechanisms has provided an experimental framework for understanding how premalignant and malignant cells can escape environmental signals that would normally initiate programmed cell death. The critical importance of identifying survival pathways relevant to specific cell types is highlighted by studies that have associated activation of individual antiapoptotic pathways with treatment failure in some tumor cell types and not others (Schmitt & Lowe, 1999).
In mammary epithelial cells, Conzen and colleagues (Moran, Gray, Mikosz, & Conzen, 2000) recently defined a novel GR-mediated survival mechanism that is induced by prolonged exposure to physiologic concentrations of glucocorticoid and inhibited by GR-specific antagonists (antiglucocorticoids). GR antagonists have been shown previously to interfere with glucocorticoid function by a two-step process: competitive inhibition of glucocorticoid receptor binding and competition of the antagonist-bound receptor with that of the glucocorticoid-bound receptor on DNA response elements within target gene promoters (Wagner et al., 1999). Because active antiglucocorticoid compounds inhibit the transactivation (and possibly the repression) of GR-specific target genes and also block GR-mediated survival signaling, we hypothesize that the mechanism through which the GR induces survival requires the induction and/or repression of cell type–specific "survival genes" (Mikosz, Brickley, Sharkey, Moran, & Conzen, 2001).
Glucocorticoids are well known for their anti-inflammatory and immunosuppressive properties as well as for their essential role in embryonic development. The majority of glucocorticoids' properties are thought to be a consequence of the ability of the activated GR to act as a transcription factor, either through a direct DNA binding–dependent mechanism or through cross-talk and often interference with other transcription factors such as AP1, STAT5, and nuclear factor-
B (Dumont et al., 1998). In addition, so-called "nongenomic" effects may play a role in the rapid effects of glucocorticoids on cell signaling (Borski, 2000).
Although the GR is expressed ubiquitously in normal human mammary epithelium as well as breast cancers, the role of the GR in breast cancer biology and mammary gland development has, until recently, received relatively little attention. Indeed, 100 articles have been published on the role of GRs in mammary tumor biology in contrast to >11,000 on estrogen receptors (2004 PubMed search with key terms of "breast cancer" and "glucocorticoid receptor" or "estrogen receptor").
In vitro, a physiologic concentration of hydrocortisone (10–6 M) has long been added to the mixture of survival factors required for successful epithelial cell growth in serum-free conditions; furthermore, the importance of glucocorticoids for optimal plating efficiency of mammary epithelial cells has suggested a possible role in cell survival (Hammond, Ham, & Stampfer, 1984). In vivo, systemic glucocorticoid treatment has been observed to prevent mammary gland involution and concomitant mammary epithelial cell apoptosis in the glands of lactating mice weaning their young (Lund et al., 1996). Moreover, a GR DNA binding mutant knock-in (GRdim) has recently been shown to have defects in mammary gland development, suggesting a direct role for GR transcriptional activation in mammary epithelial cell proliferation and possibly apoptosis (Reichardt et al., 2001).
What is the intracellular pathway transducing the glucocorticoid hormonal environment, triggered by psychosocial factors, into the cellular environment regulating cell death? One such important downstream effector of glucocorticoids' antiapoptotic function, in both mammary epithelial cells and breast cancer cell lines subjected to growth factor deprivation–induced apoptosis, is serum- and glucocorticoid-inducible kinase (SGK-1) (Mikosz et al., 2001). SGK-1 is a novel serine–threonine kinase that was identified as a transcriptional target of glucocorticoid in a rat mammary tumor cell line (Webster, Goya, Ge, Maiyar, & Firestone, 1993) and was initially studied as a potential mammary cell cycle regulatory protein (Buse, Tran, Luther, Phu, Aponte, & Firestone, 1999) in rat tumor cells. Recently, two additional isoforms of SGK have been identified: SGK-2 and SGK-3. However, neither of these isoforms appears to be regulated transcriptionally (Casamayor, Torrance, Kobayashi, Thorner, & Alessi, 1999).
Both GR activation and SGK-1 expression can also inhibit apoptosis in human breast cancer cell lines subjected to paclitaxel- or doxorubicin-induced apoptosis at clinically relevant concentrations (Wu, Chaudhuri, Brickley, Pang, Karison, & Conzen, 2004) (Figure 3). The administration of RU486, a GR blocker, fully reinstates cell death in response to these chemotherapeutic agents, demonstrating the central role of GRs in regulating apoptosis (see Figure 3B).
|
|
IMPLICATIONS |
|---|
|
Acknowledgments |
|---|
|
References |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. B. Warnecke, A. Oh, N. Breen, S. Gehlert, E. Paskett, K. L. Tucker, N. Lurie, T. Rebbeck, J. Goodwin, J. Flack, et al. Approaching Health Disparities From a Population Perspective: The National Institutes of Health Centers for Population Health and Health Disparities Am J Public Health, September 1, 2008; 98(9): 1608 - 1615. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
TOP
Abstract
Need for Specifying Multiple...
