| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Human Genetics, University of Michigan Medical School, Ann Arbor, Michigan 48109-0618
Address all correspondence and requests for reprints to: Diane M. Robins, Department of Human Genetics, University of Michigan Medical School, Ann Arbor, Michigan 48109-0618. E-mail: drobins{at}umich.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mouse sex-limited protein (Slp) gene is also dimorphic in expression and allows genetic analysis based on the rich variety of cis- and trans-acting alleles existing in inbred mouse strains. Slp arose as a duplication of the complement component gene C4, with which it retains greater than 95% identity (9, 10). Unlike C4, Slp is normally expressed only in mature males, suggesting androgen dependence, and has diverged in function in the immune system (11, 12, 13). The androgen dependence of Slp occurs by the direct action of androgen receptor (AR) on the gene in kidney, but in liver androgen acts indirectly by controlling GH release from the pituitary (14, 15). Direct molecular measurements, and the existence of cis-acting Slp alleles with altered regulation, demonstrate that Slp’s induction is transcriptional (16, 17). Trans-acting mutations further clarify tissue-specific distinctions in Slp regulation. For example, AR-deficient mice (Ar-, or Tfm, testicular feminization) do not express Slp but can be made to do so in liver by pulsatile GH administration (14).
Evidence that repression as well as induction is required for sexually dimorphic Slp expression is found in mice homozygous for rsl (regulator of sex limitation) alleles. In these mice, Slp expression initiates in females, as well as in males, at puberty (18). Inbred strains of rsl mice (e.g. FM and PL/J) appear otherwise normal, and the loss of sex limitation of Slp is recessive to Rsl alleles. Measurement of Slp serum levels in congenic rsl strains indicates that Slp expression is increased in rsl males as well, by an amount similar to that observed in females (19). Because rsl is recessive and its effect is to increase expression of its target gene, it is likely that the Rsl gene product functions, at some level, to repress Slp synthesis. This repression appears to operate in both sexes but is overcome in males by hormonal induction at puberty.
As a prelude to our positional cloning of Rsl (20), we characterized the breadth of Rsl’s effects and how its mechanism of action interfaced with hormonal regulation. We found that Rsl affects Slp in liver but not kidney, suggesting tissue specificity of Rsl regulation. Further, female expression of other male-specific liver genes occurs in rsl mice, including mouse urinary proteins (MUPs) and members of the P450 family. Rsl regulation is independent of the hormonal induction of its target genes because homozygosity for rsl allows expression of Slp in mice lacking GH or androgen signaling (due to hypophysectomy or genetic deficiency, respectively). Further, Rsl action occurs before hormonal induction. In sum, this exploration reveals a novel repression conferred by Rsl that operates in both sexes on male-specific liver genes. While not itself hormone dependent, Rsl’s action serves to accentuate gender differences in hepatic gene expression in mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All animal experimentation was performed in accord with the NIH Guidelines for Care and Use of Experimental Animals. Hypophysectomy was performed in the University of Michigan Laboratory for Animal Medicine, using protocols approved by the University Committee on Use and Care of Animals. Following surgery, survivors received 5% sucrose in their drinking water until they were killed 2–3 wk later. For parabiosis, Rsl and rslPL females were surgically conjoined at 6 wk of age using a modified Bunster-Meyer procedure (21). Following recovery, mice were fed ad libitum and housed for 3 wk. At the time the rats were killed, 200 µl of blood were collected from each heart and DNA isolated using the QIAmp Blood kit (QIAGEN, Valencia, CA). Thirty cycles of PCR using 32P-labeled primers for D13Mit253 were performed, and products were separated on an 8% sequencing gel and visualized by autoradiography.
Ribonuclease (RNase) protection assays
RNA was prepared as previously described (15) using the RNasol method (22). In cases where organs were pooled from several animals, multiple independent pools were examined; in most cases, individual animals were assessed, with at least three examined per point. Ten micrograms of total RNA were assayed per liver sample. To adjust for lower Slp expression in kidney where Slp is only expressed in the proximal convoluted tubules, compared with broader as well as more abundant expression in liver, 30 rather than 10 µg of RNA were assayed and autoradiogram exposures were for 3 d rather than 16 h. Synthesis of the Slp riboprobe and RNase protection were performed as described (23). Actin riboprobes were transcribed from the pTRI-B-actin plasmid (Ambion, Inc., Austin, TX) using the T7 promoter.
Northern blotting
RNA was denatured by heating in formaldehyde-formamide, fractionated through agarose gels containing 6% formaldehyde and transferred to Zetaprobe-GT membranes (Bio-Rad Laboratories, Inc., Hercules, CA). Filters were hybridized to a random-primed probe (Cyp 2d-9, Cyp 2a-4, RPL19) under standard conditions. For MUP detection, gene-specific oligonucleotides were end-labeled with 32P and hybridized in (5x sodium chloride/sodium citrate; 20 mM Na2HPO4; pH 7.2; 7% sodium dodecyl sulfate; 1x Denhardt’s) at 45 C and washed in 1x sodium chloride/sodium citrate at 45 C (24).
Probes for Northerns were as follows; MUP gene-specific oligonucleotide probes were synthesized by the University of Michigan DNA Core Facility. MUPs I and II: 5' GAGCACTCTTCATCTCTTACAGT; MUP III: 5' TCGCAGTCATTTCGGTGC (25). Cyp 2d-9, Cyp 2a-4, and RPL19 were synthesized by RT-PCR from B10.D2 liver mRNA using the following primers: Cyp 2d-9-5' AAGAATACCATAGACTCCAGA (forward) and 5' GTGGTCCGTGACCTGTTTGG (reverse); Cyp 2a-4-5' CTCTCTCATGAAGATCAGCC (forward) and 5' ATTGTGTTCCACTTTCTTGG (reverse): RPL19-5' CTGAAGGTCAAAGGGAATGTG (forward) and 5' GGACAGAGTCTTGATGATCTC (reverse). Identity of PCR products was confirmed by restriction enzyme mapping. Products were gel purified and subcloned into pGemTA (Promega Corp., Madison, WI) and a SacII-SalI fragment used as probe.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) (23). In Fig. 1, liver mRNA levels in B10.D2 mice, where Slp expression is limited to males, was compared with expression in the two congenic strains, B10.D2.PL-rsl and B10.D2.FM-rsl (referred to as rslPL and rslFM, respectively). Both these strains express Slp in females but at different levels, which is likely due to distinct rsl alleles originating from the different strains. Quantitation by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) analysis of multiple independent experiments showed that rslPL female expression ranged from 45–60% of B10.D2 male levels. The rslFM female level was substantially less than the rslPL female level, but readily detected compared with Rsl females. PhosphorImager analysis also corroborated data from serum that Slp expression increases about 2-fold in rslPL compared with Rsl males. The increase in rsl gene expression in males is more evident for other targets (see below). In contrast to the marked effect on Slp, there was no effect on C4 (Fig. 1A), actin or RPL19 (see below), indicating target gene selectivity of rsl action.
|
). Females bearing either rsl allele did not show Slp expression greater than the few percent of male expression that also occurs in Rsl female kidney (23). Therefore, the rsl relief of sex limitation of Slp expression occurs in liver, but not in kidney, revealing a cell-specific control additional to distinctions in hormonal induction in these tissues.
The rsl pathway regulates MUPs as well as Slp, and male- but not female-specific P450s
The dramatic effect of the rsl alleles on Slp expression led us to question whether genes under similar hormonal control were also affected. We first examined members of the MUP (mouse urinary protein) gene family, which are thought to function as pheromone carriers upon excretion (26). Some MUPs are abundantly expressed in male liver, where, like Slp, their androgen dependence is GH-mediated (27). We probed Northern blots with oligonucleotides that distinguish members of this highly homologous family; one probe recognizes two male-specific RNAs (MUP I and II) and the other is specific for MUP III (24, 25). A single Northern filter, containing 10 µg total RNA per lane, except for 2 µg polyA+ RNA in the two leftmost lanes, was probed first for RPL19, a ribosomal protein mRNA, to allow normalization of samples, and then sequentially with the MUP oligonucleotides (Fig. 2). In the B10.D2 Rsl mice, MUPs I and II are 4-fold, and MUP III 12-fold, more abundant in males than females by PhosphorImager analysis. The rsl alleles in the congenic B10.D2 background increased MUP expression in females, and also clearly in males, as for Slp. Interestingly, there were differential effects on the different MUP genes. Comparing female to male expression within each strain, the rslFM allele appeared to more greatly affect MUP III than MUP I and II, whereas the rslPL allele more strongly effected MUPs I and II. These differences could be due to biological variation in MUP gene expression or to allelic variation in these strains.
|
-hydroxylase (Cyp 2d-9) mRNA was present at a few-fold higher levels in male than female B10.D2 mice, using a probe derived by RT-PCR from the cDNA sequence (29) hybridized to the same Northern filter as previously (Fig. 2). In the rsl mice, both males and females showed increased levels of Cyp 2d-9 RNA (Fig. 2), suggesting this gene is also an rsl target. In contrast, the female-prominent Cyp 2a-4 gene, encoding 15-hydroxylase, was unaffected by rsl. Therefore, Rsl targets appear to include male-specific but not female-specific hepatic genes.
Rsl regulation is not dependent on GH secretory pattern or amount
Characterization thus far suggested that Rsl intersects a regulatory pathway common to male-specific liver genes. A possible basis for the rsl phenotype was acquisition by these mice of an altered pattern of GH release. In particular, it was conceivable that rsl females secreted GH with a pulse pattern that induced the male-specific hepatic target genes. However, the male-specific GH pulse also represses expression of female-specific liver genes in males (8), and, as shown in Fig. 2, expression of female-specific Cyp 2a-4 was not reduced in female rsl mice. Further, GH pulsatile secretion in males accounts for their greater postpuberal size than females, as shown by partial dwarfing of male STAT5b null mice that cannot transduce the GH pulse signal (8). In contrast, there was no effect on growth in either sex of rsl mice: Rsl and rslPL mice were identical in sex-specific rate of growth and size attained (data not shown).
To test not only the secretory pattern but any involvement of GH in the rsl effect, mice were hypophysectomized and expression of the Slp target gene examined after 2 and 3 wk, by which time Slp is undetectable in wild-type hypophysectomized mice (14, 15). Male mice were used, in part to allow loss of Slp expression in kidney, where rsl had no impact (Fig. 1), to serve as an internal control for efficacy of pituitary ablation. In addition, reduction of GH was monitored after surgery by decreased MUP protein in urine assayed by SDS-PAGE (not shown), and after they were killed, by diminished IGF-I mRNA in liver (see Fig. 3). As shown in Fig. 3, Slp expression determined by RNase protection was absent 2 wk after hypophysectomy in Rsl mice, but in rslPL mice was sustained at about 20% of the intact male level (by PhosphorImager analysis), even 3 wk post surgery. Slp mRNA levels in kidney were depleted in all hypophysectomized animals used for analysis. Reduced levels of IGF-I mRNA were used to corroborate successful pituitary ablation because GH enhances IGF-I gene expression (30). In Rsl mice, IGF-I mRNA was decreased 2- to 3-fold 2 wk after surgery; in rsl mice, the IGF-I decline was substantially more dramatic. This suggested that in the absence of a pituitary factor, IGF-I was sustained to a greater extent in Rsl than rsl mice. Regardless of the relative difference in IGF-I level, reduction of IGF-I mRNA correlated with hypophysectomy. Hypophysectomy did not affect C4 or actin levels in either tissue. Therefore, the mechanism by which rsl alleles lead to Slp expression is not dependent on pulsatile GH secretion, and in fact, has no stringent requirement for any pituitary factor.
|
). Among the parental and backcross second generation mice, Slp was expressed in all males and in females homozygous for rsl, as expected. Slp mRNA was also present at about 50% of the B10.D2 Rsl male level in livers of Tfm/Y mice homozygous for rsl, in contrast to no expression in Tfm/Y mice that have at least one Rsl allele. As expected, Slp was not expressed in kidney of Tfm/Y mice, whether homozygous or heterozygous for rsl (not shown). This confirms at the molecular level that the rsl component of Slp expression is independent of functional AR.
|
). Whereas 2 wk suffice to eliminate Slp by hypophysectomy or castration in males, or to induce Slp in females treated with androgen, there was no change in Slp levels following parabiosis. Circulatory mixing was qualitatively demonstrated by the presence of cells of both genotypes in heart blood, using microsatellite markers that distinguish Rsl and rsl alleles. Whereas critical concentrations of a putative factor might be difficult to attain following parabiosis, particularly if such a factor is short-lived, this result nevertheless supports the idea that Rsl acts autonomously in the liver and is not dependent on long-lived circulatory factors readily transmissable in serum.
|
), using two to three mice per point, in each of two independent experiments. Slp mRNA was undetectable in Rsl females in these assays, as expected (not shown). Slp became evident at 5 wk of age in Rsl males, accumulating rapidly thereafter. There was no Slp mRNA at 2 wk in either sex of rsl mice, but by 3 wk of age, before weaning, Slp expression was apparent in both sexes, and attained maximal levels by 5 wk. When Slp mRNA was normalized to C4 levels, which do not vary with age, sex, or Rsl genotype, the rate of accumulation appeared similar in the two strains, but initiated at least 2 wk earlier in the rsl mice. This suggests that the repression normally enforced by Rsl, and revealed by Slp expression due to absence of repression in rsl mice, begins to act at around 3 wk, before the GH induction that occurs in males later in puberty. Thus Rsl is temporally, as well as mechanistically, independent of the hormonal induction of the male-specific liver genes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The pathway in which Rsl acts can be grossly described from the present analysis, and is confirmed by our recent identification of the gene product as a KRAB zinc finger protein (20). Firstly, that pathway is directly involved in gene regulation. Noted as a trans-acting locus affecting Slp, Rsl affects the expression of a set of genes that, like Slp, are induced in male liver at puberty. In male kidney, expression of Slp is directly induced by androgen and its receptor, but in liver, Slp induction by androgen is indirect, exerted via gonadal control of pulsatile GH secretion from the pituitary (summarized in Fig. 7A; Refs. 14, 15). As shown in this study, homozygous rsl alleles lead to high levels of Slp mRNA in females, in liver but not in kidney. Similarly higher expression occurs in males as well as in females, for the spectrum of male-specific liver genes. Because rsl increases gene expression of the targets, and because there is no such increase in heterozygous Rsl/rsl mice, the wild-type Rsl likely functions, at some level, in repression of male-specific liver genes. The alternative, that rsl encodes an activator absent in Rsl mice, is unlikely because of the lack of effect in heterozygotes.
|
). Slp induction at puberty occurs by increased mRNA synthesis, as shown in nuclear run-on assays where Slp initiation in Rsl females is 1/50th the rate in males, paralleling steady-state RNA levels (17). In addition, some alleles of Slp are constitutively expressed due to recombination placing C4 regulatory sequences before Slp coding regions (16, 33). This implies that sequences required for male expression reside in the 5' flank, as corroborated by identification of an upstream enhancer responsive to androgen and binding sites for STAT5b, the transducer of the pulsatile GH signal (17, 34, 35). Thus, posttranscriptional effects, including message stabilization and translational control, are unlikely to account for expression in rsl mice. Further, a posttranscriptional mechanism for rsl action should produce detectable expression in Rsl/rsl mice, but this is not seen. Therefore, evidence suggests that Rsl dictates transcriptional repression of the set of genes that, in mice, are male specific in liver. Some promoter/enhancer elements appear to be common to male-specific hepatic genes, including some of the targets examined here (e.g. Refs. 17, 36, 37), and may be candidates for an Rsl "silencer" site. A third intriguing feature of the Rsl pathway is that, although apparently critical for sexually dimorphic gene expression, Rsl itself does not appear dependent on either pituitary or androgen regulation. Although GH and AR are not required for Rsl regulation, there still may be some hormonal control involved in the effect. This could explain lower Slp levels sustained in rsl mice that are hypophysectomized compared with those that are Tfm/Y. Alternatively, strain differences in additional loci that modify Slp expression could influence this quantitative difference. In addition, the decrease in IGF-I expression following hypophysectomy is much greater in rsl than Rsl mice, suggesting a possible positive effect of Rsl on IGF-I level, but only in the presence of GH. As a test for unidentified hormones that might be required for Rsl regulation, we performed parabiosis to determine whether such factors could be exchanged through the circulation. Parabiosis is a classic means to distinguish whether a genetic defect resides in a hormone or its receptor, and provided a crucial foundation for the discovery of leptin (31, 38). The lack of alteration of Slp expression following parabiosis does not rule out any hormonal control, but is consistent with Rsl functioning in a liver-intrinsic manner.
A paradox is why target gene expression is not seen before puberty in rsl mice, where absence of repression seems sufficient to account for expression in adults. This suggests that Rsl itself may be regulated during maturation in some manner that is not dependent on the same factors that induce the target genes. This is supported by initiation of Slp expression very early in puberty in rsl mice, compared with relatively late in puberty in Rsl mice. Thus, the Rsl repression mechanism begins to operate well before hormonal induction of the target genes, and continues to operate simultaneously with induction in males. It may be that a general and global repression silences the target genes before puberty, and only once the general repression is removed can either specific induction or repression be observed.
The transcriptional repression dictated by Rsl operates on male-specific liver genes in both sexes, but is overridden by hormonal induction in males. In rsl mice, repression is relieved and target gene expression increases during puberty in both sexes, even without hormonal induction (as in female, Tfm, or hypophysectomized mice) and to an even greater extent in males where induction and lack of repression are additive. This demonstrates that Rsl normally functions to enforce and refine sexually dimorphic gene expression. Given the evolutionary conservation of genomes, particularly for regulatory networks, it is likely that Rsl has human counterparts, which may be involved in repression of a similar set of target genes. In humans, the P450s exhibit genetic and regulatory variation that may enter into many aspects of steroid and drug metabolism. Although sexually dimorphic expression of particular P450s is less well studied in humans than in rodents, there are numerous examples of drugs that are metabolized differently in men and women, as well as other aspects of liver pathology that differ between the sexes (39, 40, 41, 42). Rsl may shed light on these syndromes and also may be more broadly informative to liver function because it is the rodent target genes, and not Rsl itself, that is gender specific. These studies establish a foundation for future studies of Rsl, which may provide unique insights into basic mechanisms of tissue- and gene-specific transcriptional repression.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: AR, Androgen receptor; C4, fourth component of complement; MUP, mouse urinary protein; RNase, ribonuclease; rsl, regulator of sex limitation; Slp, sex-limited protein.
Received December 23, 2002.
Accepted for publication January 28, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase (C-P-45016) in mice. J Biol Chem 264:2920–2927This article has been cited by other articles:
 |
|
 |
|
  M. G. Holloway, G. D. Miles, A. A. Dombkowski, and D. J. Waxman Liver-Specific Hepatocyte Nuclear Factor-4{alpha} Deficiency: Greater Impact on Gene Expression in Male than in Female Mouse Liver Mol. Endocrinol., May 1, 2008; 22(5): 1274 - 1286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Gilbert Sex, Salt, and Senescence: Sorting Out Mechanisms of the Developmental Origins of Hypertension Hypertension, April 1, 2008; 51(4): 997 - 999. [Full Text] [PDF]
|
|
|
|
 |
|
 K. H. Clodfelter, G. D. Miles, V. Wauthier, M. G. Holloway, X. Zhang, P. Hodor, W. J. Ray, and D. J. Waxman Role of STAT5a in regulation of sex-specific gene expression in female but not male mouse liver revealed by microarray analysis Physiol Genomics, September 11, 2007; 31(1): 63 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Waxman and C. O'Connor Growth Hormone Regulation of Sex-Dependent Liver Gene Expression Mol. Endocrinol., November 1, 2006; 20(11): 2613 - 2629. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Yang, E. E. Schadt, S. Wang, H. Wang, A. P. Arnold, L. Ingram-Drake, T. A. Drake, and A. J. Lusis Tissue-specific expression and regulation of sexually dimorphic genes in mice Genome Res., August 1, 2006; 16(8): 995 - 1004. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Clodfelter, M. G. Holloway, P. Hodor, S.-H. Park, W. J. Ray, and D. J. Waxman Sex-Dependent Liver Gene Expression Is Extensive and Largely Dependent upon Signal Transducer and Activator of Transcription 5b (STAT5b): STAT5b-Dependent Activation of Male Genes and Repression of Female Genes Revealed by Microarray Analysis Mol. Endocrinol., June 1, 2006; 20(6): 1333 - 1351. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Yu, N. Dadgar, M. Albertelli, A. Scheller, R. L. Albin, D. M. Robins, and A. P. Lieberman Abnormalities of Germ Cell Maturation and Sertoli Cell Cytoskeleton in Androgen Receptor 113 CAG Knock-In Mice Reveal Toxic Effects of the Mutant Protein Am. J. Pathol., January 1, 2006; 168(1): 195 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Wiwi, M. Gupte, and D. J. Waxman Sexually Dimorphic P450 Gene Expression in Liver-Specific Hepatocyte Nuclear Factor 4{alpha}-Deficient Mice Mol. Endocrinol., August 1, 2004; 18(8): 1975 - 1987. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Waxman and J. L. Celenza Sexual dimorphism of hepatic gene expression: novel biological role of KRAB zinc finger repressors revealed Genes & Dev., November 1, 2003; 17(21): 2607 - 2613. [Full Text] [PDF]
|
|
|
|
|
|
C. J. Krebs, L. K. Larkins, R. Price, K. M. Tullis, R. D. Miller, and D. M. Robins Regulator of sex-limitation (Rsl) encodes a pair of KRAB zinc-finger genes that control sexually dimorphic liver gene expression Genes & Dev., November 1, 2003; 17(21): 2664 - 2674. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
