Accepted Manuscript
Understanding the role of androgens and placental AR variants: Insight into steroid- dependent fetal-placental growth and development
A.S. Meakin, V.L. Clifton
PII: S0143-4004(18)31172-X DOI: https://doi.org/10.1016/j.placenta.2019.03.006 Reference: YPLAC 3954
To appear in: Placenta
Received Date: 8 November 2018 Revised Date: 7 March 2019 Accepted Date: 14 March 2019
Please cite this article as: Meakin AS, Clifton VL, Understanding the role of androgens and placental AR variants: Insight into steroid-dependent fetal-placental growth and development, Placenta (2019), doi: https://doi.org/10.1016/j.placenta.2019.03.006.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT 1 Review: Understanding the role of androgens and placental AR variants:
2 insight into steroid-dependent fetal-placental growth and development
3
4 A.S Meakin 1 & V.L Clifton 1
5 1 Mater Medical Research Institute, University of Queensland, Brisbane, Australia
6
7 Corresponding Author: Professor Vicki L Clifton
8 Mater Medical Research Institute,
9 Level 4, Translational Research Institute
10 37 Kent St, Woolloongabba, Qld. 4101 11 Brisbane, Australia MANUSCRIPT 12 [email protected]
13 Phone 617 3443 7640
14
15 Word count: 3200
16 Key words: androgens, androgen receptor, placenta, growth
17 Short title: PlacentalACCEPTED androgen signalling
1
ACCEPTED MANUSCRIPT
18 Abstract
19 The role of steroids throughout pregnancy and their effect on placental physiology is
20 well established, especially for estrogens, progestogens, and glucocorticoids;
21 however, the role of androgens – particularly within the context of placental
22 physiology – remains largely unexplored. Androgens are often defined as the male
23 sex-steroids and are fundamental for the defeminisation and masculinisation of male
24 fetuses. Therefore, the placenta may adapt to these steroids in a sex-specific
25 manner, with males being more receptive to changes in these steroids
26 concentrations, when compared with females; however, their involvement in female
27 intrauterine development has been investigated in several studies and may suggest
28 females have a level of responsiveness to these steroids. While the former may be
29 true, studies have reported sex-specific differences in the expression and activity of 30 factors involved in androgen biosynthesis andMANUSCRIPT bioavailability, with males consistently 31 demonstrating greater degrees of altered protein and gene expression when
32 compared with females. Understanding the placental androgen axis is essential as
33 many pregnancy comorbidities are associated with elevated concentrations of
34 androgens and perturbed intrauterine development or growth. Indeed, it appears that
35 specific pathophysiologies of pregnancy can modulate the activity of key factors
36 involved in the placental androgen axis and this may contribute to the etiology of
37 sex-specific developmental outcomes from certain pregnancy complications. This 38 review will provideACCEPTED insight into what is currently known regarding androgen signalling 39 and the human placenta, and how this complex system may regulate sex-specific
40 growth and developmental outcomes in normal and adverse pregnancies.
2
ACCEPTED MANUSCRIPT
41 Androgens and pregnancy
42 Throughout gestation, the levels of maternal circulating androgens including
43 testosterone, dihydrotestosterone (DHT), dehydroepiandrosterone (DHEA), and
44 androstenedione increase with concentrations increasing three-fold by the third
45 trimester when compared to levels of non-pregnant women [1]. There are sex
46 differences in the concentrations of androgens with cord blood testosterone
47 concentrations being higher in males than females in normal pregnancies [2, 3].
48 Studies further demonstrate that pregnancy comorbidities such as polycystic ovarian
49 syndrome (PCOS), preeclampsia and gestational diabetes mellitus (GDM) are
50 associated with an increase in the concentration of these sex-steroids [4-8]. The
51 synthesis of androgens is dependent on the bioavailability of pregnenolone, an
52 androgen precursor which is metabolised from cholesterol (Figure 1) in the
53 mitochondria of select cell types, one of which is the placental trophoblast [9-11]. MANUSCRIPT 54 Numerous proteins involved in androgen biosynthesis and signalling have been
55 identified in the human placenta, and some studies would suggest that the
56 expression and activity of these proteins may be altered by fetal-placental sex [4, 12-
57 14]. Under normal conditions, placental synthesised androgens are converted to
58 estrogens by placental aromatase (CYP19A1) (Figure 1). There are limited studies
59 which have examined how placental sex or the presence of a pregnancy
60 complication affects this process; however, one study examining the sex-specific
61 dysregulation ofACCEPTED androgen biosynthesis from placentae of preeclamptic pregnancies
62 demonstrated decreased expression of CYP19A1 protein and mRNA in males, when
63 compared to females [4]. Interestingly, female placentae of preeclamptic
64 pregnancies had increased expression of CYP19A1 compared to female placentae
65 from an uncomplicated pregnancy. Although the activity of aromatase was not
3
ACCEPTED MANUSCRIPT 66 measured, these findings, in conjunction with a reported increase in androgen
67 concentrations in women with preeclampsia and a male fetus suggest that in the
68 presence of a pregnancy complication, males may prioritise androgen bioavailability
69 by suppressing aromatase expression. Under similar adverse conditions, females
70 prioritise estrogen bioavailability by upregulating the expression of aromatase.
71 Likewise, a study conducted by Maliqueo et al. [14] reported reduced aromatase
72 activity in placentae from women diagnosed with PCOS, when compared to controls.
73 This study, however, did not account for placental sex: rather, the study reported
74 lower androstenedione and higher estriol concentrations in the cord blood of female
75 newborns of women with PCOS when compared to control, however the exact
76 source of these steroids’ biosynthesis remains unclear. Taken together, the two
77 studies would indicate there are sex-specific alterations of androgen metabolism in
78 the presence of a maternal complication, with an increase in placental aromatase 79 activity in females and a decrease in males MANUSCRIPT resulting in a rise in estrogen or 80 androgen concentrations, respectively.
81 Androgen precursors that are not converted to estrogens are metabolised into the
82 potent androgens testosterone and DHT. Protein and mRNA expression of
83 cytochrome P450 17A1 (CYP17A1), an enzyme involved in androgen metabolism,
84 has been detected in primary human trophoblast cells and the human trophoblast
85 cell line JEG-3 [12]. These cells were able to generate testosterone de novo , 86 however placentalACCEPTED sex was not accounted for, which leaves questions surrounding 87 placental sex and testosterone synthesis unanswered. Previous findings from our
88 group have demonstrated male placentae at term have increased expression of 5 α-
89 reductase (SRD5A) compared to females [13]. This enzyme is involved in reducing
90 testosterone to DHT (Figure 1), a potent androgen derivative with a higher binding
4
ACCEPTED MANUSCRIPT 91 affinity for the androgen receptor (AR) than testosterone [15] and suggests that not
92 only is the placenta a source of androgens, but that this organ may contribute to
93 increased activity of the androgen axis in a sex-specific manner.
94 Current data suggests there are concentration differences in androgens between the
95 sexes in pregnancy and they are essential for sex determination, sex differences in
96 fetal growth and possibly important in conferring sex-specific differences of the fetal-
97 placental response to pregnancy complications or perturbations. Therefore,
98 understanding i) the role androgens have in male and female intrauterine
99 development ii) the role the placenta has in mediating these processes and iii) how
100 pregnancy complications and perturbations of the placental androgen axis affects
101 critical periods of growth and development are of particular interest.
102 Androgens: role in fetal growth and development
103 Androgens are essential for defeminisation MANUSCRIPT (Müllerian ducts regression) and 104 masculinisation (differentiation of Wollfian ducts to the epididymis, vasa deferentia,
105 and the seminal vesicles) of the male fetus (for a review, see Rey et al . [16]).
106 Interestingly, perturbations to androgen signalling via various mutations in the AR
107 gene results in 46,XY karyotyped neonates presenting with female-like phenotypes
108 [17, 18]. These 46,XY neonates, who are clinically diagnosed as having partial or
109 complete androgen insensitivity syndrome (AIS), were reported to have significantly 110 reduced birthweightACCEPTED adjusted for gestational age when compared to a control male 111 population [18]. Studies using global ARKO mouse models show similar findings,
112 with males displaying infertility and a hyperfeminised phenotype absent of male
113 reproductive organs, except testes [19]. It appears androgens and their receptor are
5
ACCEPTED MANUSCRIPT 114 fundamental for appropriate male growth and development in utero and any
115 perturbation to this signalling pathway results in adverse outcomes.
116 Recent studies have demonstrated an involvement for androgens in female
117 intrauterine development. Inhibition of androgen action via the antiandrogen
118 flutamide resulted in perturbed folliculogenesis [20]. Knapczyk-Stwora et al . [20]
119 reported altered expression of key targets involved in this process including IGF-1,
120 IGF-1R, c-KIT and KL. A similar study using global and ovarian granulosa cell-
121 specific AR knockout mouse models demonstrated, both in vitro and in vivo, that
122 androgens at a physiological concentration attenuate follicular atresia via miR-125b
123 and enhance follicle-stimulating hormone (FSH) receptor expression, which further
124 increases FSH-mediated follicle growth and development [21].
125 Although androgens are important for female development, excessive levels present 126 risks for female fetuses and the mother. PlacentalMANUSCRIPT aromatase insufficiency has been 127 shown to result in virilisation of female fetuses, due to the excessive levels of
128 androgens in both maternal and fetal circulations [22-24]. Further, intrauterine
129 exposure to high androgen levels is thought to be a contributing factor to the
130 development of diseases later in life such as PCOS, metabolic disorders, and
131 cardiovascular disease (CVD) [25-27]. The placenta is thought to mitigate excessive
132 intrauterine androgen exposure of the female fetus through aromatase activity [28].
133 Certain pregnancy comorbidities, however, may detrimentally alter the expression
134 and activity ofACCEPTED androgen-specific placental enzymes and adversely affect female
135 development. For this reason, understanding how pregnancy comorbidities affect
136 key enzymes involved in androgen metabolism and biosynthesis needs further
137 investigation. It appears that the dynamic state of androgen biosynthesis and
138 bioavailability is fundamental for the appropriate development of female fetuses, and
6
ACCEPTED MANUSCRIPT 139 any perturbations to these processes can have adverse developmental outcomes
140 and contribute to disease risks in adulthood.
141 It is unclear how androgens affect intrauterine growth outcomes. Human population
142 studies have demonstrated small for gestational age (SGA) male neonates have
143 significantly reduced levels of testosterone when compared to appropriate for
144 gestational age (AGA) neonates [29]. Similarly, Carlsen et al . [30] demonstrated
145 negative associations between testosterone levels at 17 and 33 weeks’ gestation
146 and birthweight (g) in female neonates. In contrast, Voegtline et al . [31] reported
147 reduced male birthweight, but increased postnatal weight gain in pregnancies with
148 high maternal plasma testosterone concentrations when compared with low
149 testosterone. The same study reported increased female birthweight in the high
150 testosterone group compared to the low testosterone group, but no change in 151 postnatal weight gain. These discrepancies MANUSCRIPTbetween studies are likely products of 152 study design, which highlights the necessity for additional research investigating
153 associations between androgen levels and birthweight outcomes, especially when
154 considering a specific pregnancy comorbidity.
155 Sex differences in human fetal growth in pregnancy complications may be due to
156 altered androgen levels. Our group has previously reported sex-specific birthweight
157 outcomes from pregnancies complicated by mild preeclampsia [32]. Specifically,
158 female neonates from preeclamptic pregnancies had significantly reduced
159 birthweight centileACCEPTED (BWC) when compared to neonates from healthy control groups.
160 In contrast, male BWC remained unaltered between normotensive and pre-eclamptic
161 groups. Given previous studies report increased testosterone concentrations in
162 pregnancies complicated by pre-eclampsia [5, 33], it is possible that androgens have
7
ACCEPTED MANUSCRIPT 163 a role in regulating the differential intrauterine growth response between the sexes.
164 However, further studies are required to substantiate this possibility.
165 In vivo animal studies have demonstrated that high levels of androgens during
166 gestation can lead to IUGR-like outcomes, reduced fetal viability, and reduced litter
167 size [34-38]. Cleys et al . [35] reported that the average birthweight of female
168 offspring of testosterone propionate (TP) treated ewe was significantly reduced
169 compared to controls, and that no difference in male birthweights were observed
170 between treatment groups. These findings are in agreement with previous research
171 conducted by Beckett et al . [34], which identified consistently lower birthweights of
172 female offspring born to aromatisable (testosterone) and non-aromatisable (DHT)
173 androgen treated ewes. In contrast, Sathishkumar et al . [38] reported increased
174 maternal testosterone concentrations in a rodent model suppressed fetal growth 175 indirectly by affecting placental amino acidMANUSCRIPT delivery to the fetus. Prenatal TP 176 exposure significantly reduced slc38a4 mRNA and protein expression in both male
177 and female rat placentae and was associated with reduced placental amino acid
178 transport capacity with a greater effect observed in males. Based on both human
179 and animal data, it is postulated that sex differences in the growth effects of
180 androgens may be the result of changes in androgen signalling, particularly in the
181 placenta where nutrient transport and growth mechanisms impact fetal growth.
182 Androgen signalling ACCEPTED 183 Androgen signalling primarily occurs through the AR, a steroid receptor belonging to
184 the nuclear receptor 3-ketosteroid group C (NR3C) subfamily. The AR gene has
185 been localised to Xq11-12, contains eight exons and codes for a 919-amino acid
186 protein (Figure 2). Exon 1 of the AR gene encodes for the N-terminal domain (NTD):
8
ACCEPTED MANUSCRIPT 187 this domain contains an activation function 1 (AF1) region, which interacts with
188 coregulatory proteins to enhance transcriptional regulation of AR-target genes. The
189 second and third exons encode two distinct zinc-fingers (collectively referred to as
190 the DNA-binding domain) required for interactions with androgen response elements
191 (ARE) located within promoter regions of AR target genes. The remaining exons of
192 the AR gene (4-8) encode a short flexible hinge region (which contains the nuclear
193 localisation signal) and an alpha-helical sandwich, composed of 11 alpha-helices,
194 which forms the AR C-terminal domain (CTD). This CTD is comprised of the ligand
195 binding domain (LBD) and the AF2 co-regulator binding interface [39]. In its ligand-
196 free state, the AR protein is localised in the cytoplasm where it is bound by a number
197 of chaperone proteins including Hsp90 and immunophilins [40]. Upon interaction
198 between androgen ligands and the LBD, a conformational change of the AR protein
199 exposes the nuclear localisation signal in the hinge region, thereby allowing direct 200 interactions with importin-α which facilitates MANUSCRIPT subsequent nuclear translocation [41]. 201 Within the nucleus, dimerization of ligand-bound AR units at promoter elements
202 results in the transcriptional regulation of AR target genes, leading to either activated
203 or suppressed expression, depending on the co-regulatory proteins with which this
204 AR-dimer is interacting [42].
205 There is a plethora of reported co-regulatory proteins that modulate transcriptional
206 activity of AR-target genes. Prostate cancer research has demonstrated ligand 207 dependent co-repressorACCEPTED (LCoR), forkhead box protein (FOX)A1, and FOXO1 208 modulate androgen-dependent and –independent AR activity [43, 44]. It is unclear
209 whether these protein-protein interactions are present in the human placenta and if a
210 pregnancy complication affects their expression. Recent studies identified AR
211 complexes containing histone lysine demethylases (KDMs) in ovine trophoblast cells
9
ACCEPTED MANUSCRIPT 212 [35]. This complex was shown to co-localise and bind to the ARE in the AR target
213 genes VEGF and IGF-1. KDMs demethylate histone lysine residues increasing the
214 accessibility of ARE on methylated AR target genes, which can enhance cell
215 proliferation and growth [45]. Cleys et al . [35] observed that ewes treated with TP
216 had significantly increased IGF-1 and VEGF mRNA and protein expression, when
217 compared to controls. Furthermore, the same study identified KDM1A and KDM4D
218 localisation in syncytiotrophoblast and villous stromal cells of first trimester human
219 placentae. These findings suggest that AR-KDM complexes may modulate the
220 transcriptional regulation of androgen-mediated target genes in human placental
221 tissue cells, thereby contributing to growth stimulation by IGF-1 or VEGF-dependent
222 mechanisms.
223 There is limited information about AR signalling in the human placenta and how 224 interactions with co-regulatory proteins modulateMANUSCRIPT AR function. Identification and 225 characterisation of these interactions within the placenta may provide insight into the
226 molecular mechanisms underpinning placental androgen signalling. Furthermore,
227 how pregnancy comorbidities affect AR interactions with co-regulator proteins would
228 assist in understanding sex-specific fetal-placental growth and developmental
229 outcomes. While AR-coregulatory factors can modulate androgen-mediated
230 signalling, another mechanism by which this pathway may be altered is through
231 differential expression of AR transcript and protein variants. ACCEPTED 232 AR variants
233 At present, there are at least 20 different AR transcript variants reported in the
234 literature [46-52], which gives rise to numerous protein variants with molecular
235 weights (MW) ranging from 45-120kDa. Canonical genomic androgen signalling is
10
ACCEPTED MANUSCRIPT 236 mediated via the full-length AR protein (AR-FL). In vitro and in vivo studies have
237 shown AR-FL necessitates growth in tumour and non-tumour tissue via regulation of
238 downstream signalling pathways. Conversely, selective silencing of AR-FL has been
239 reported to suppress tumour growth and proliferation. Hickey et al . [53] reported in
240 vitro knockdown of AR-FL significantly reduced the proliferation of MDA-MB-453
241 cells. Similar findings have been reported in in vitro and in vivo models of other
242 cancers including prostate [54] and bladder [55, 56]. Testosterone stimulation of
243 muscle-derived cells significantly increased IGF-1 mRNA expression and this
244 response was abrogated by the AR-antagonist flutamide [57]. Very few studies have
245 examined the expression of AR-FL in the human placenta and whether placental sex
246 or pregnancy complications alter its expression. Horie et al. [58] reported the
247 presence of AR in the human placental trophoblast[59]. More recently preeclamptic
248 pregnancies were reported to be associated with increased AR-FL protein and 249 mRNA expression in the placenta [4], regardless MANUSCRIPT of fetal sex. 250 Post-transcriptional modifications of AR can result in C- and N-terminally truncated
251 variants. Alternative splicing of AR exons 1, 2, and cryptic exon (CE) 3 generate AR-
252 V7. This C-terminally truncated, ligand-independent variant has been reported to
253 regulate transcription of AR-FL target genes [53, 60, 61], as well as a unique subset
254 of inflammatory genes, some of which include CXCL10, CCL5, IL8, and MMP13. AR-
255 V7 overexpression in the breast cancer cell line MDA-MB-453 resulted in a gene 256 expression profileACCEPTED that shared no overlap with gene expression profiles in the 257 prostate cancer cell line LNCaP or MDA-MB-453 overexpressing AR-FL. These
258 findings highlight the unique transcriptional regulatory role AR-V7 has and suggests
259 this C-terminally truncated variant regulates specific mechanisms distinct from AR-
260 FL. Expression of AR-V7 has primarily been reported in prostate and breast cancer
11
ACCEPTED MANUSCRIPT 261 tissue and cell lines, but it has also been found in non-tumour prostate tissue cells,
262 suggesting a role in androgen signalling in normal tissue [60].
263 Similar to AR-V7, AR-V1 is produced through alternative splicing of AR exons 1, 2,
264 and CE1. AR-V1 lacks the C-terminal LBD and has been shown to inhibit AR-V7 by
265 sequestering it in the cytoplasm. Co-overexpression of AR-V1 and AR-V7
266 suppressed the transcriptional activity of the AR-V7 specific gene, UBEC2C,
267 demonstrating the inhibitory action of AR-V1 when expressed with AR-V7 [62]. In
268 contrast, AR-FL dimerization with AR-V1 in LNCaP cells resulted in an increase in
269 prostate specific antigen (PSA) mRNA expression, suggesting that the AR-V1/FL
270 complex enhances the transactivational capacity of AR-FL in the presence or
271 absence of androgens. Additional C-terminally truncated variants have also been
272 reported (for a review, refer to Dehm & Tindall [51]), however the functions of these
273 variants remain largely unknown. MANUSCRIPT 274 AR-45 is a N-terminally truncated AR isoform, which is generated through alternative
275 splicing of exon 1b, an intronically localised exon. AR-45 contains a unique seven
276 amino acid sequence in place of the highly conserved NTD [63]. Studies assessing
277 the function of AR-45 have demonstrated it suppresses cellular growth via a
278 postulated AR-45/FL non-responsive heterodimer; however, the same study reported
279 that in the androgen-independent cell line, PC-3, AR-45 co-expressed with TIF2
280 upregulated MMTV and PSA promoter activity in the presence of DHT or 281 androstenedioneACCEPTED [50]. Although AR-45 lacks the AF1, co-regulatory factors binding 282 to the LBD-localised AF2 are able to enhance transcription of target genes [39, 64,
283 65]. Moreover, in the absence of AR-FL, AR-45 may function as a membrane-bound
284 AR in neuronal SNpc and N27 cells mediating androgen-dependent non-genomic
285 signalling [66]. To our knowledge, no study has reported multiple AR protein variants
12
ACCEPTED MANUSCRIPT 286 in the human placenta or in placentae of animal models. Given their capacity to
287 modulate androgen-dependent and -independent signalling, assessing the
288 expression and function of AR variants has been a focus within our group. Of
289 particular interest is the differential expression of placental AR variants between the
290 sexes and in the presence of pregnancy complications, and how may these variant
291 alter the placenta’s response to androgens.
292 Placental androgen signalling in pregnancies complicated
293 by asthma
294 Previous work from our team has identified that male fetuses grow normally and
295 female fetuses reduce growth in the presence of maternal asthma [67]. We propose
296 that this differential response is mediated by sex-specific alterations in placental
297 function with male placentae inducing a state of glucocorticoid resistance to avoid 298 the anti-proliferative effects of increased concent MANUSCRIPTrations of glucocorticoids. Previous 299 studies have indicated that increased fetal cortisol concentrations are positively
300 correlated with testosterone levels in cord blood [3]. It is possible, therefore, that
301 males continue to grow in the presence of maternal asthma or under stress
302 conditions that result in a rise in cortisol because of androgen induced anabolic
303 mechanisms in the placenta. We hypothesise that the growth difference between
304 males and females is conferred by sex-specific differences in placental AR variant 305 expression. ACCEPTED 306 Preliminary data from our group has demonstrated sex-specific expression of
307 placental AR variants which vary in the presence or absence of maternal asthma.
308 Male placentae from asthmatic pregnancies had a significant reduction in the nuclear
309 expression of three known AR isoforms including AR-FL, AR-V7 and AR-V1 but an
13
ACCEPTED MANUSCRIPT 310 increase in the nuclear expression of AR-45 [68]. There were no changes observed
311 in female placentae which may suggest alternative pathways are prioritised in
312 females that contribute to reduced growth in the presence of maternal asthma.
313 Interestingly, increased AR-45 expression in male placentae from asthmatic
314 pregnancies did not appear to inhibit mRNA expression of androgen-mediated target
315 genes including IGF-1, IGF-1R , IGFBP-3, IGFBP-5 or VEGF ; in fact, the data
316 demonstrated that the presence of maternal asthma increased expression of these
317 growth targets, and correlation data – as well as preliminary in vitro studies – would
318 suggest AR-45 may be involved in IGF-1R and IGFBP-5 transcription (unpublished
319 data). Our preliminary findings justify further investigations to explore the
320 mechanistic function of placental AR variants, which may provide evidence
321 supporting a crucial role of AR protein variants in placental pathophysiology.
322 Conclusion MANUSCRIPT 323 It is evident that the placental androgen action is mediated by a complex network of
324 pathways that remain to be fully understood. The involvement of androgens in male
325 and female intrauterine growth and development has been documented, and it is
326 clear that pregnancy comorbidities can perturb placental androgen signalling and
327 give rise to fetal morbidities (or, in extreme cases, mortality) and may contribute to
328 the development of diseases later in life. Our group’s findings suggest that canonical
329 androgen signalling is not the established mechanism of androgen action in the
330 human placenta,ACCEPTED which necessitates further investigations of the role of AR variants
331 in the human placental function. Understanding the complexity of placental androgen
332 signalling and its role in sex-specific growth and developmental outcomes, especially
333 in circumstances where a specific pathophysiology of pregnancy exists, is critical.
14
ACCEPTED MANUSCRIPT 334
335 References
336 1. Makieva, S., P.T. Saunders, and J.E. Norman, Androgens in pregnancy: roles in parturition. 337 Hum Reprod Update, 2014. 20 (4): p. 542-59. 338 2. Maccoby, E.E., et al., Concentrations of sex hormones in umbilical-cord blood: their relation 339 to sex and birth order of infants. Child Dev, 1979. 50 (3): p. 632-42. 340 3. Gitau, R., et al., Fetal plasma testosterone correlates positively with cortisol. Arch Dis Child 341 Fetal Neonatal Ed, 2005. 90 (2): p. F166-9. 342 4. Sathishkumar, K., et al., Fetal sex-related dysregulation in testosterone production and their 343 receptor expression in the human placenta with preeclampsia. J Perinatol, 2012. 32 (5): p. 344 328-35. 345 5. Salamalekis, E., et al., Androgen levels in the third trimester of pregnancy in patients with 346 preeclampsia. Eur J Obstet Gynecol Reprod Biol, 2006. 126 (1): p. 16-9. 347 6. Falbo, A., et al., Changes in androgens and insulin sensitivity indexes throughout pregnancy 348 in women with polycystic ovary syndrome (PCOS): relationships with adverse outcomes. 349 Journal of Ovarian Research, 2010. 3(1): p. 23. 350 7. Sir-Petermann, T., et al., Maternal serum androgens in pregnant women with polycystic 351 ovarian syndrome: possible implications in prenatal androgenization. Hum Reprod, 2002. 352 17 (10): p. 2573-9. 353 8. Morisset, A.S., et al., Androgens in the maternal and fetal circulation: association with insulin 354 resistance. J Matern Fetal Neonatal Med, 2013. 26 (5): p. 513-9. 355 9. Tuckey, R.C., Progesterone synthesis by the human placenta. Placenta, 2005. 26 (4): p. 273- 356 281. 357 10. Miller, W.L., Steroid hormone synthesis MANUSCRIPT in mitochondria. Molecular and Cellular 358 Endocrinology, 2013. 379 (1): p. 62-73. 359 11. Bose, H.S., et al., N-218 MLN64, a protein with StAR-like steroidogenic activity, is folded and 360 cleaved similarly to StAR. Biochemistry, 2000. 39 (38): p. 11722-31. 361 12. Escobar, J.C., et al., The human placenta expresses CYP17 and generates androgens de novo. 362 J Clin Endocrinol Metab, 2011. 96 (5): p. 1385-92. 363 13. Vu, T.T., et al., Changes in human placental 5alpha-reductase isoenzyme expression with 364 advancing gestation: effects of fetal sex and glucocorticoid exposure. Reprod Fertil Dev, 365 2009. 21 (4): p. 599-607. 366 14. Maliqueo, M., et al., Placental steroidogenesis in pregnant women with polycystic ovary 367 syndrome. Eur J Obstet Gynecol Reprod Biol, 2013. 166 (2): p. 151-5. 368 15. Oettel, M., Testosterone metabolism, dose-response relationships and receptor 369 polymorphisms: selected pharmacological/toxicological considerations on benefits versus 370 risks of testosterone therapy in men. Aging Male, 2003. 6(4): p. 230-56. 371 16. Rey, R.A. and R.P. Grinspon, Normal male sexual differentiation and aetiology of disorders of 372 sex development. Best Pract Res Clin Endocrinol Metab, 2011. 25 (2): p. 221-38. 373 17. Lek, N., etACCEPTED al., Low frequency of androgen receptor gene mutations in 46 XY DSD, and fetal 374 growth restriction. Arch Dis Child, 2014. 99 (4): p. 358-61. 375 18. de Zegher, F., et al., Androgens and fetal growth. Horm Res, 1998. 50 (4): p. 243-4. 376 19. Zhou, X., Roles of androgen receptor in male and female reproduction: lessons from global 377 and cell-specific androgen receptor knockout (ARKO) mice. J Androl, 2010. 31 (3): p. 235-43. 378 20. Knapczyk-Stwora, K., et al., Effect of flutamide on folliculogenesis in the fetal porcine ovary-- 379 regulation by Kit ligand/c-Kit and IGF1/IGF1R systems. Anim Reprod Sci, 2013. 142 (3-4): p. 380 160-7.
15
ACCEPTED MANUSCRIPT 381 21. Sen, A., et al., Androgens regulate ovarian follicular development by increasing follicle 382 stimulating hormone receptor and microRNA-125b expression. Proc Natl Acad Sci U S A, 383 2014. 111 (8): p. 3008-13. 384 22. Shozu, M., et al., A new cause of female pseudohermaphroditism: placental aromatase 385 deficiency. J Clin Endocrinol Metab, 1991. 72 (3): p. 560-6. 386 23. Stalberg, K., et al., Aromatase Deficiency in Fetal Virilization. Journal of Pediatric Urology, 387 2010. 6: p. S85. 388 24. Jones, M.E.E., et al., Recognizing rare disorders: aromatase deficiency. Nature Clinical 389 Practice Endocrinology &Amp; Metabolism, 2007. 3: p. 414. 390 25. Dumesic, D.A., D.H. Abbott, and V. Padmanabhan, Polycystic ovary syndrome and its 391 developmental origins. Rev Endocr Metab Disord, 2007. 8(2): p. 127-41. 392 26. Padmanabhan, V. and A. Veiga-Lopez, Animal models of the polycystic ovary syndrome 393 phenotype. Steroids, 2013. 78 (8): p. 734-40. 394 27. Vyas, A.K., et al., Prenatal programming: adverse cardiac programming by gestational 395 testosterone excess. Sci Rep, 2016. 6: p. 28335. 396 28. Stocco, C., Tissue physiology and pathology of aromatase. Steroids, 2012. 77 (1-2): p. 27-35. 397 29. Cicognani, A., et al., Low birth weight for gestational age and subsequent male gonadal 398 function. J Pediatr, 2002. 141 (3): p. 376-9. 399 30. Carlsen, S.M., G. Jacobsen, and P. Romundstad, Maternal testosterone levels during 400 pregnancy are associated with offspring size at birth. Eur J Endocrinol, 2006. 155 (2): p. 365- 401 70. 402 31. Voegtline, K.M., et al., Sex-specific associations of maternal prenatal testosterone levels with 403 birth weight and weight gain in infancy. J Dev Orig Health Dis, 2013. 4(4): p. 280-4. 404 32. Stark, M.J., V.L. Clifton, and I.M. Wright, Neonates born to mothers with preeclampsia exhibit 405 sex-specific alterations in microvascular function. Pediatr Res, 2009. 65 (3): p. 292-5. 406 33. Acromite, M.T., et al., Androgens in preeclampsia. Am J Obstet Gynecol, 1999. 180 (1 Pt 1): p. 407 60-3. 408 34. Beckett, E.M., et al., Developmental programing:MANUSCRIPT impact of testosterone on placental 409 differentiation. Reproduction, 2014. 148 (2): p. 199-209. 410 35. Cleys, E.R., et al., Androgen receptor and histone lysine demethylases in ovine placenta. PLoS 411 One, 2015. 10 (2): p. e0117472. 412 36. Blesson, C.S., et al., Prenatal testosterone exposure induces hypertension in adult females via 413 androgen receptor-dependent protein kinase Cdelta-mediated mechanism. Hypertension, 414 2015. 65 (3): p. 683-690. 415 37. Gopalakrishnan, K., et al., Elevated testosterone reduces uterine blood flow, spiral artery 416 elongation and placental oxygenation in pregnant rats. Hypertension, 2016. 67 (3): p. 630-9. 417 38. Sathishkumar, K., et al., Prenatal testosterone-induced fetal growth restriction is associated 418 with down-regulation of rat placental amino acid transport. Reprod Biol Endocrinol, 2011. 9: 419 p. 110. 420 39. He, B., et al., Dependence of selective gene activation on the androgen receptor NH2- and 421 COOH-terminal interaction. J Biol Chem, 2002. 277 (28): p. 25631-9. 422 40. Pratt, W.B. and D.O. Toft, Steroid receptor interactions with heat shock protein and 423 immunophilinACCEPTED chaperones. Endocr Rev, 1997. 18 (3): p. 306-60. 424 41. Cutress, M.L., et al., Structural basis for the nuclear import of the human androgen receptor. 425 J Cell Sci, 2008. 121 (Pt 7): p. 957-68. 426 42. Heinlein, C.A. and C. Chang, Androgen receptor (AR) coregulators: an overview. Endocr Rev, 427 2002. 23 (2): p. 175-200. 428 43. Asim, M., et al., Ligand-dependent corepressor acts as a novel androgen receptor 429 corepressor, inhibits prostate cancer growth, and is functionally inactivated by the Src 430 protein kinase. J Biol Chem, 2011. 286 (43): p. 37108-17.
16
ACCEPTED MANUSCRIPT 431 44. Zhao, Y., D.J. Tindall, and H. Huang, Modulation of Androgen Receptor by FOXA1 and FOXO1 432 Factors in Prostate Cancer. Int J Biol Sci, 2014. 10 (6): p. 614-9. 433 45. Metzger, E., et al., LSD1 demethylates repressive histone marks to promote androgen- 434 receptor-dependent transcription. Nature, 2005. 437 (7057): p. 436-9. 435 46. Li, Y., et al., AR intragenic deletions linked to androgen receptor splice variant expression and 436 activity in models of prostate cancer progression. Oncogene, 2012. 31 (45): p. 4759-67. 437 47. Nyquist, M.D., et al., TALEN-engineered AR gene rearrangements reveal endocrine 438 uncoupling of androgen receptor in prostate cancer. Proc Natl Acad Sci U S A, 2013. 110 (43): 439 p. 17492-7. 440 48. Lu, C. and J. Luo, Decoding the androgen receptor splice variants. Transl Androl Urol, 2013. 441 2(3): p. 178-186. 442 49. Hu, D.G., et al., Identification of androgen receptor splice variant transcripts in breast cancer 443 cell lines and human tissues. Horm Cancer, 2014. 5(2): p. 61-71. 444 50. Ahrens-Fath, I., et al., Androgen receptor function is modulated by the tissue-specific AR45 445 variant. FEBS Journal, 2004. 272 (1): p. 74-84. 446 51. Dehm, S.M. and D.J. Tindall, Alternatively spliced androgen receptor variants. Endocr Relat 447 Cancer, 2011. 18 (5): p. R183-96. 448 52. Stope, M.B., et al., Shortened isoforms of the androgen receptor are regulated by the 449 cytoprotective heat-shock protein HSPB1 and the tumor-suppressive microRNA miR-1 in 450 prostate cancer cells. Anticancer Res, 2013. 33 (11): p. 4921-6. 451 53. Hickey, T., et al., Expression of androgen receptor splice variants in clinical breast cancers. 452 Oncotarget, 2015. 7: p. 17. 453 54. Snoek, R., et al., In vivo knockdown of the androgen receptor results in growth inhibition and 454 regression of well-established, castration-resistant prostate tumors. Clin Cancer Res, 2009. 455 15 (1): p. 39-47. 456 55. Li, P., J. Chen, and H. Miyamoto, Androgen Receptor Signaling in Bladder Cancer. Cancers 457 (Basel), 2017. 9(2). 458 56. Miyamoto, H., et al., Promotion of bladder cancerMANUSCRIPT development and progression by androgen 459 receptor signals. J Natl Cancer Inst, 2007. 99 (7): p. 558-68. 460 57. Sculthorpe, N., et al., Androgens affect myogenesis in vitro and increase local IGF-1 461 expression. Med Sci Sports Exerc, 2012. 44 (4): p. 610-5. 462 58. Horie, K., et al., Immunohistochemical localization of androgen receptor in the human 463 endometrium, decidua, placenta and pathological conditions of the endometrium. Hum 464 Reprod, 1992. 7(10): p. 1461-6. 465 59. Hsu, T., et al., Expression of Androgen Receptor in Human Placentas From Normal and 466 Preeclamptic Pregnancies. Taiwan J Obstet Gynecol, 2009. 48 (3): p. 5. 467 60. Hu, R., et al., Ligand-independent androgen receptor variants derived from splicing of cryptic 468 exons signify hormone-refractory prostate cancer. Cancer Res, 2009. 69 (1): p. 16-22. 469 61. Tepper, C.G., et al., Characterization of a novel androgen receptor mutation in a relapsed 470 CWR22 prostate cancer xenograft and cell line. Cancer Res, 2002. 62 (22): p. 6606-14. 471 62. Zhan, Y., et al., Interplay between Cytoplasmic and Nuclear Androgen Receptor Splice 472 Variants Mediates Castration Resistance. Mol Cancer Res, 2017. 15 (1): p. 59-68. 473 63. Weiss, B.,ACCEPTED H. Faus, and B. Haendler, Phylogenetic conservation of the androgen receptor 474 AR45 variant form in placental mammals. Gene, 2007. 399 (2): p. 105-11. 475 64. He, B., et al., Structural Basis for Androgen Receptor Interdomain and Coactivator 476 Interactions Suggests a Transition in Nuclear Receptor Activation Function Dominance. 477 Molecular Cell, 2004. 16 (3): p. 425-438. 478 65. Bevan, C.L., et al., The AF1 and AF2 Domains of the Androgen Receptor Interact with Distinct 479 Regions of SRC1. Mol Cell Biol, 1999. 19 (12): p. 8383-92. 480 66. Garza-Contreras, J., et al., Presence of Androgen Receptor Variant in Neuronal Lipid Rafts. 481 eNeuro, 2017. 4(4).
17
ACCEPTED MANUSCRIPT 482 67. Murphy, V.E., et al., Maternal asthma is associated with reduced female fetal growth. Am J 483 Respir Crit Care Med, 2003. 168 (11): p. 1317-23. 484 68. Meakin, A., Z. Saif, and V. Clifton, Identification of placental androgen receptor isoforms and 485 their relationship to fetal growth in pregnancies complicated by asthma. Placenta, 2017. 57 : 486 p. 306. 487
MANUSCRIPT
ACCEPTED
18
ACCEPTED MANUSCRIPT Figure 1 – Androgens are synthesised via a cascade of enzymatic reactions.
The first reaction in the process of androgen synthesis involves cleavage of the cholesterol side-chain via the cholesterol side-chain cleavage enzyme (CYP11A1).
The resulting pregnenolone is either converted to progesterone or 17 α-
hydroxypregnenolone via 3β-hydroxystyeroid dehydrogenase (HSD3B) or
cytochrome P450 17A1 (CYP17A1), respectively. Both progesterone and 17 α-
hydroxypregnenolone are then converted to 17 α-hydroxyprogesterone via CYP17A1 or HSD3B, respectively, however 17 α-hydroxypregnenolone can be directly
metabolised to Dehydroepiandrosterone (DHEA) via CYP17A1. Androstenedione is
synthesised via HSD3B’s oxidation of DHEA or via hydroxylation of 17 α-
hydroxyprogesterone: this androgen derivative is then reduced to 5 α-
androstanedione via 5 α-reductase (SRD5A). DHEA, androstenedione, and 5 α-
androstanedione can be converted to androstenediol, testosterone, and dihydrotestosterone, respectively, via 17 β -hydroxysteroidMANUSCRIPT dehydrogenase type 3 (HSD17B3) or 17 β-hydroxysteroid dehydrogenase type 5 (AKR1C3). Testosterone itself can also be synthesised via the oxidation of androstenediol, and this well- known androgen can then be reduced to dihydrotestosterone via SRD5A. The androgens testosterone and 5 α-Androstanedione can further be metabolised to the estrogens estradiol and estrone, respectively, via aromatase (CYP19A1).
Figure 2 – AR gene and protein schematic. The AR gene is localised to Xq11-12, contains eightACCEPTED exons, and codes for a 919-amino acid protein. The AR protein contains three major functional domains required for appropriate protein function.
Exon 1 of the AR gene encodes for the N-terminal domain (NTD). The NTD contains two primary transactivation domains (transactivation uni-1 (Tau-1) and -2 (Tau-2)) within the activation function region 1 (AF1). Exon 2 and 3 of the AR gene encodes ACCEPTED MANUSCRIPT the DNA-binding domain (DBD) which interacts to the androgen response element
(ARE) within the promoter region of target genes. This interaction is facilitated by two zinc fingers. Exon 4 of the AR gene encodes for a hinge region which contains the nuclear localisation signal that facilitates nuclear translocation. Exons 5-8 encode the ligand binding domain (LBD) which interacts with androgen ligands and also contains the AF2, which allows weak interactions with co-regulatory complexes.
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT
MANUSCRIPT
ACCEPTED ACCEPTED MANUSCRIPT Highlights
Male and female fetuses have differing growth trajectories in complicated pregnancies
Androgens may allow male fetuses to continue to grow under adverse conditions
Androgen receptor variants may mediate the differential growth response between males and females
MANUSCRIPT
ACCEPTED