androgen receptor mrna expression in xenopus laevis cns: sexual dimorphism and regulation in...

Androgen Receptor mRNA Expression in Xenopus laevis CNS: Sexual Dimorphism and Regulation in Laryngeal Motor Nucleus

Julio Perez, * Michael A. Cohen, a n d Darcy B. Kelley

Department of Biology, Columbia University, 903 Fairchild, New York, New York 10027

SUMMARY

Using Northern analysis, in situ hybridization, and nuclease protection assays, the expression and regulation of androgen receptor messenger RNA (AR mRNA) was examined in the C N S of juvenile Xenopus luevis. Only one of the AR mRNA isoforms expressed in X. Iuevis is transcribed in the C N S as shown by Northern blot analysis. Nuclease protection assays demonstrate that the expression of AR mRNA is higher in the brain stem than in the telencephalon and diencephalon. Although expression of AR mRNA is widespread throughout the CNS, cells of cranial nerve nucleus IX-X (N. IX-X) and spinal cord display the highest in situ hybridization signals in their cytoplasm. Double labeling using horseradish peroxidase and digoxigenin labeled AR probes reveals that

laryngeal and anterior spinal cord motor neurons express AR mRNA. More cells express AR mRNA in N. IX-X of males than of females. The number of AR expressing cells in N. IX-X decreases following gonadectomy in both sexes, and dihydrotestosterone (DHT) treatment for 1 month reverses this effect. Increased expression of AR mRNA in the brain of DHT treated animals is also apparent in nuclease protection assays. Sex differences in number of AR expressing cells and hormone regulation of AR mRNA expression in motor nuclei may influence neuromuscular systems devoted to sexually differentiated behaviors. Keywords: androgen receptor, mRNA expression, motor neurons, sexual dimorphism.

k 1YY6 John %‘ile> &Sons, tnc.

INTRODUCTION

Steroid hormones regulate neuroendocrine functions and behaviors related to reproduction. Among the most dramatic effects of steroids are alterations in developmental programs leading to sex differences in the CNS. In many cases, hormone sensitive neurons have been implicated in the control of sex-specific reproductive behaviors. For example, nuclei of the preoptic area and hypothalamus of mammals have been shown to be sexually dimorphic and are involved in reproductive con- duct (Tobet et al., 1986; Paredesand Baum, 1995).

Received February 7, 1996; accepted March 25, 1996 Journal of Neurobiology, Vol. 30, No. 4, pp. 556-568 ( 1996) 0 1996 John Wiley Sr Sons, Inc. CCC 0022-3034/96/040556- I3

* To whom correspondence should be addressed (e-mail: [email protected]).

556

In canaries, sexually dimorphic regions of the telencephalon undergo seasonal modifications in size related to variations in bird song (Nottebohm, 1980). In the frog Xenopus furvis, the IX-X nerve differs between sexes in the number of axons that innervate the larynx or vocal organ (Kelley and Dennison, 1990). The larynx itself varies in muscle fiber type and size due to an androgen-driven pro- gram of masculinization; sex differences in this neuromuscular system result in sex-specific pat- terns of vocalization (Kelley and Pfaff, 1976; Han- nigan and Kelley, 1986; Watson and Kelley, 1992; Watson et al., 1993). We examined the possibility that nuclei of the brain involved in differentiated behaviors are distinct in males and females in re- gard to their sensitivity to hormones. Thus, we analyzed in the cranial nerve nucleus (N.) IX-X whether the expression of androgen receptor messenger RNA (AR mRNA) is sexually dimorphic.

Androgens interact with an intracellular recep-

CNS Androgen Receptor Regulution 557

tor that induces gene transcription (Berger and Watson, 1989). In the rat CNS, AR mRNA and protein are widely expressed in regions including the telencephalon, neurons within the cranial nerve nuclei, and motor neurons of the spinal cord (Simerly et al., 1990; Osada et al., 1993; A1 Shamma and Arnold, 1995 ). In X . luevis, two isoforms of AR are expressed in the larynx (Fischer et al., 1993, 1995). In the CNS, diencephalic, mesencephalic, and medullary nuclei, including cells within the N. IX-X and anterior spinal cord, accu- mulate testosterone and dihydrotestosterone (DHT; Kelley et al., 1975; Kelley, 1980, 1981; Er- ulkar et al., 1981; Gorlick and Kelley, 1986). An unresolved issue within this system is the identification of cell types that express AR. Anterior spinal cord and N. IX-X include both motor neurons and interneurons (Erulkar et al., 1981; Watson and Kelley, 1989). We sought to determine specifically whether motor neurons in these nuclei express AR mRNA and could thus directly participate in androgen-evoked responses.

Adult male X . luevis have more axons in N. IX- X than do females (Kelley and Dennison, 1990). The larger number of axons observed in male tad- poles requires androgen exposure (Robertson et al., 1994). In juvenile females, testis implantation increases the number of laryngeal nerve axons to male levels (Watson et al., 1993). Although the mechanism by which changes in axon number is achieved is still unclear, these results suggest that motor neurons in N. IX-X of juveniles are sensitive to androgens. Because androgen regulates the expression of its own receptor in the rat brain and prostate and the Xmupzrs larynx (Menard and Harlan, 1993; Takeda et al., 1991: Fischer et al., 1995), we explored the effect of endocrine manipulations on AR mRNA expression in several CNS regions, including N. IX-X. Our results suggest that AR expression in N. IX-X of juveniles is upregulated by androgens, a process that could fa- cilitate responses induced by androgens in motor neurons.

MATERIALS AND METHODS

Animals and Hormone Treatments

Juvenile frogs (stage PM 1,3 months after metamorpho- sis is complete; Tobias et al., I99 I ) were used in the present study. Animals were separated into male and female groups that were either left intact or gonadectomized. Within the gonadectomized group, frogs were either im-

planted with an empty Silastic tube (Dow Corning, VST 030065) or with a tube containing 5 mg of 5a-17P-ol-3- one androstan (DHT, Sigma). An additional group of gonadectomized females received a SilasticTM tube containing 5 mg of I7@-estradiol (Sigma). One month after gonadectomy and Silastic tube implantation, animals were intracardially perfused with 5 m L of 0.6% NaCl followed by 10 m L of 4% paraformaldehyde in I X PBS (2.6 mMKCI, 1.4 m M K H 2 P 0 4 , 136 mMNaC1,8 m M Na2HP04, pH 7.2). The brain and spinal cord (CNS) were removed and incubated for 2 h in fixative followed by 20% sucrose in 1 X PBS for 4-12 h. Twenty-micron horizontal sections of the brain were cut in a cryostat, thaw mounted onto Superfrost plus microscope slides (Fisher), and stored at -70°C until use.

Northern Blotting

For Northern blot hybridization, an 843-base pair (bp) fragment of the AR, extending from the first zinc finger of the DNA binding domain into the ligand binding domain, was used [Fig. I (A) ; He et al., 19901. This fragment was polymerase chain reaction (PCR) amplified from PM I and PM3 laryngeal musclecDNA usingprim- ers containing the SP6 (sense) and T7 (antisense) pro- moters and then gel purified. The fragment was labeled with [ 32P] -dCTP using the random primer extension reaction. Total RNA was isolated from the brain and spinal cord of 10 PM 1 males or 10 PM 1 females by the RNAzol method (Tel-Test, Inc.) and denatured using glyoxal. Twenty micrograms of total RNA was electrophoreti- cally separated on a 1.2% agarose gel and blotted onto Genescreen membranes ( DuPont) by standard methods. Hybridization was carried out overnight a t 42°C under high stringency conditions. Blots were washed in 2X standard saline citrate (SSC), 0. I % SDS at room temper- ature for 1 h and 0. I X SSC, 0.1 ’% SDS at 65°C for I h.

Nuclease Protection Assays

A 495-bp fragment was PCR amplified using the primers Sp6-5GGTATTCCACAGCTTGA3’ and T7-5’GTC- GACAGACCTTCCAC3’ and the 843-bp AR cDNA described previously as the template [Fig. 1 ( A ) ] . The PCR product was inserted into a pCR-I1 vector and transfected into INVaF‘ cells (Invitrogen). We se- quenced this cloned PCR fragment; it encodes a portion of the ligand binding domain of the X . luevis a A R ( H e et al., 1990). After purification, this DNA was cut with Xho I and used to generate 3’P-antisense RNA by in vitro transcription (specific activity 4.7 x 10’ cpm/Fg). Brains were obtained from gonadectomized and DHT- treated gonadectomized males and females. Total RNA was isolated using Trizol (Lifetime Tech.) from two regions: the anterior brain (telencephalon, diencephalon, and mesencephalon) and the region of the brain stem containing N. IX-X [Fig. 2( A)]. RNA concentration was measured by spectrophotometry and compared in

558 Pkrez, Cohen, and Kelley

A HYPERVARIABLE DNA-BINDING LOAN0 BINDING

B

Figure 1 (A) AR cDNA and probes used for Northern blots, in situ hybridization, and nuclease protection assays. Probes are either complementary to 843 nucleo- tides (n t ) of the DNA and ligand binding domains (Northern) or to the first 495 nt of the ligand binding domain of the AR (nuclease and in situ, assays). The probe for nuclease protection assays contained, in addition, 19 nt of the Sp6 promoter and 43 nt of the pCR-I1 vector. ( B ) Northern blot analysis demonstrates that only a single transcript of approximately 9.6 kb is expressed in the CNS of PM 1 X . luevis males and females.

1% agarose gels to confirm the quality and concentration of RNA. To determine regulation of AR mRNA by DHT, 5 pg of total RNA from the experimental groups were assayed and the relative amounts of transcript compared. Gel purified probe, 2 X lo5 cpm, was hybridized to total RNA for 10 rnin in hybridization buffer and treated with 1:250 of RNAse A/T mix (HybSpeed RPA kit, Ambion). Hybridization of the probe to tRNA ( S O p g ) was used as negative control. Samples were analyzed on a 7 M urea/4% polyacrylamide gel, and the mobility of the probe and protected fragment was compared to 32P-5’ end labeled RNA molecular weight markers ( Promega). After film exposure (X-OMAT, Kodak), gels were reexposed to a phosphor image screen and the image analyzed (Molecular Dynamics Image). The optical density of the background (signals from portions of the gel running more rapidly than the protected fragment) was subtracted from values for the protected fragment. Data are presented as optical density units.

Horseradish Peroxidase (HRP) Histochemistry

One group of intact animals was prepared for HRP histochemistry. The larynx was accessed through an incision in the side of the body under the arm and theflexor car17i radiulis muscle was approached by an incision in the skin of the forearm. Frogs received HRP type VI (Sigma, approximately 1 mg/animal) inserted as crys- tals into the respective muscles. Three days later, animals were anesthetized, perfused, and brains treated as before. HRP detection was carried out as described previously (Llewellyn-Smith eta]., 1992). Briefly, cryostat sections were preincubated in a solution containing 0.006% tetra- methylbenzidine (Sigma) in I X PBS for 20 rnin and de- veloped with 0.005% H202. Sections were next incubated in a solution containing 1 mg/mL 3-3’-diamino- benzidine (DAB) and 16 mg/mL CoCI2. Finally, a brown precipitate was formed when incubated in 0.05% H202. Sections were then processed for AR in situ hybridization according to the procedure described below.

In Situ Hybridization

The 495-bp PCR product described previously was in vitro transcribed using the corresponding RNA polymerase (Sp6, sense; T7, antisense) in the presence of digoxigenin-dUTP (dig-UTP) according to the manufacturer’s protocol ( Boehringer Mannheim). Residual DNA was digested with DNAse. Sense and antisense probes were purified by chromatography through Nensorb-20 columns ( DuPont) and stored at -70°C until hybridization. The probe concentration was assessed on dot blots by comparison to dig-UTP labeled RNA of known concentration (Boehringer Mannheim). Sections of the brain and spinal cord were postfixed for 20 rnin at room tem- perature in 4% paraformaldehyde. The slides were washed twice in 1 X PBS for 5 rnin each, incubated in proteinase K ( 10 pg/mL) in 0.1 MTris-HC1, pH 7.5, 10 mMEDTA for 30 min, followed by a 10-min incubation in 0.25% acetic acid anhydride in 0.1 Mtriethanolamine. Sections were dehydrated in a graded series of ethanol solutions, placed in chloroform for 5 rnin and 100 and 95% ethanol for 1 rnin each, and finally air dried. The AR probe was diluted in hybridization solution (50% formamide, 0.6 M NaCI, 0.06 11.1 sodium citrate, pH 7 , 1 X Denhardt’s solution, 0.5 mg/mL yeast tRNA, 0.2 p g / mL salmon sperm DNA, 10% dextran sulfate) to a final concentration of 2 pg/pL and applied to sections. Slides were covered with parafilm and incubated in a humid chamber at 50°C overnight. The following day, sections were washed with 2 X SSC (0.3 A4 NaCl, 0.03 M Na citrate, pH 7) and 1 X SSC for 1 h each, followed by 0.5 X SSC for 1 h at 50°C and RNAse A ( 1 0 pg/mL) in RNAse buffer ( 10 m M Tris, pH 8.0, 1 mhf EDTA, 0.5 MNaCl) for 30 rnin at 37°C. Probe concentrations of 10, 2, and 0.1 pg/pL in hybridization solution were assayed to determine the saturating concentration at - 12 h

CNSAndsogen Receptos Regulation 559

Figure 2 ( A ) Diagram of Xenopus luevis brain. Two regions were surveyed for AR mRNA expression by nuclease protection assays: the anterior brain (AB) and brain stem region containing N. IX-X (IX-X). (B) Sensitivity of the nuclease protection assay. Signals due to the protected fragment increase linearly with increasing amounts of total RNA (2.5, 5, and 10 pg) in the assay. (C) Nuclease protection assay comparing expression of AR mRNA in IX-X and AB. Optical density analysis reveals that expression in the brain stem is -58% higher than in the anterior brain. Insert: Total RNA from the brain stem and anterior brain analyzed in 1% agarose gel. (B, C ) MW lane: RNA molecular weight markers; arrowheads, 562 and 363 nt. P lane: Undigested AR probe. tRNA lane: As an internal control, the AR probe was hybridized to tRNA. Occasionally, a band of undigested probe was seen in the tRNA lane and in the 2.5 and 10 wg lanes.

(overnight) of hybridization [Fig. 3( A-C)]. To control for specificity, hybridization signals with a 843 nucleo- tide ( n t ) dig-labeled probe (similar t o that used for Northern blot) were compared to signals with the 495 nt probe; similar results were obtained (data not shown).

In addition, one group of sections was incubated with similar concentrations of dig-labeled sense riboprobe [Fig. 3(D)]. To control for nonspecific signals another set of sections was preincubated with 20 pg/mL RNAse A for 30 min at 37°C [Fig. 3(E) ] . The immunohisto-

Figure 3 Nonisotopic hybridization signals in N. IX-X in control experiments. The tissue was hybridized with (A) 10 pg/pL, (B) 2 pg/pL, and (C) 0.1 pg/pL of probe in the hybridization solution. No hybridization signals were found in sections hybridized with ( D ) sense probe or pretreated with RNAse A. Scale bar ( E ) = 50 pm.

560 Pirez, Cohen, and Kelltiy

chemical procedure for recognizing dig-labled probes was as previously described (Pkrez and Hoyer, 1995). Nuclei labeled with the AR probe were identified according to Simpson et al. ( 1986).

Cell Quantification and Statistical Analysis

Cells expressing AR mRNA were recognizable by blue staining in the cytoplasm. The number of AR expressing cells in N. IX-X was determined in consecutive horizontal sections at 250X using a light microscope. Only cells with a visible nucleus were counted. Statistical comparison of the experimental groups was performed using analysis of variance (ANOVA) with the post hoc New- man-Keuls test.

RESULTS

Northern Blot Analysis

Because two forms of the AR mRNA have been reported in the Xenopzis larynx, we first determined if these mRNA forms are expressed in the CNS. A probe that recognizes the two isoforms of the X. laevis AR, ARcv and ARP, was used to identify transcripts expressed in the CNS [Fig. 1 ( A ) ] . The ARcv, which is a 9.6 kb mRNA, is the only transcript expressed in both male and female CNS at PM1 [Fig. 1 (B) ] .

Expression of AR mRNA in X . Iaevis Brain

Nuclease protection assay produced a single protected fragment, demonstrating the specificity of the probe used for this assay and in situ hybridization [Figs. I ( A ) , 2( B,C)] . To determine the sensi-

tivity of the nuclease protection assay, increasing concentrations (2.5, 5, and 10 pg) of total RNA from the anterior brain were hybridized to the AR probe [Fig. 2 (B) ] . The optical density values ofthe respective protected fragments were 18,800, 34,83 1, and 78,550 units and are thus linearly related to RNA content in the assay. In the first assay in the females, the optical density value of the protected fragment representing AR mRNA was 72% higher in the brain stem than in the anterior brain (optical density values in the brain stem = 5 1,267 vs. anterior brain = 14,353). This result was con- firmed in a second assay [Fig. 2(C)] ; the optical densities of the protected fragment were anterior brain 25,195 and brain stem 45,881 (45% higher). These results suggest that levels of AR mRNA are higher in the brain stem than in the anterior brain.

Localization of AR mRNA Expression in X. laevis CNS

AR probe concentrations of 10, 2, and 0.1 pg/pL hybridization solution were compared to determine the optimum concentration for in sitii hybridization [Fig. 3( A-C)]. A concentration of 2 pg/pL hybridization solution was used for quantification of the number of AR mRNA expressing cells because this concentration saturated the hybridization reaction [Fig. 3 ( B ) ] without the increase in background observed at the higher concentration [ Fig. 3 (A)] . No hybridization signal was observed in sections incubated with dig-labeled sense probe [Fig. 3 ( D ) ] or preincubated with RNAseA [Fig. 3(E)].

No cells in telencephalon, diencephalon, or mesencephalon showed high levels of hybridization to

Figure 4 Hybridization signals in X . luevis CNS. Horizontal sections of the brain are shown from (A) rostra1 to (F) caudal. In each figure, up is anterior and down is posterior. Cells labeled with the AR probe display a blue cytoplasmic signal. Brain stem nuclei were identified according to Simpson et al. ( 1986). (A) Weak hybridization signals were found in the stnatum (St), preoptic area (Pa), and thalamus (Th) . ( B ) Hybridization signals were observed in the vestibular (Ve) and lateral line (NLL) nuclei; the AR probe did not hybridize to granule cells of the cerebellum (Cer). ( C ) hybridization signals in cells of the nucleus isthmus (Is) and dorsal tegmental nucleus ofthe medulla (Dt). No hybridization signals with the AR probe were found in the central grey (Cg). (D) Hybridization signals in the trigeminal ( V ) and facial (VI I ) motor nuclei. (E ) In situ hybridization signals in the motor nucleus of cranial nerves IX-X (IX-X) and reticular formation (Re). (F) Motor neuron columns in the anterior spinal cord display high AR hybridization signals. ( G ) Cells in the postotic or acoustic ganglion were strongly labeled with AR probe. ( H ) Section at the level of the spinal cord: no hybridization signals were apparent in sections hybridized with sense probe. Original magnifications in (A-F) and (H)areasin(A).Scalebar= (A) 100pm,(G)25pm.

562 Pbrez, Cohen, und Kt.1lr.v

the AR probe. Cells with light or moderate label in the cytoplasm following hybridization with the AR probe were found in the dorsolateral striatum of the telencephalon [Fig. 4(A)], thalamus [Fig. 4(A)], magnocellular nucleus of the preoptic area [Fig. 4( A)], anterior hypothalamus (not shown), and mesencephalic nucleus of the trigeminal nerve ( not shown). In contrast, cells in the anterior preoptic nucleus, anterior thalamus, and optic tectum were not labeled with the AR probe (not shown).

Identification of brain stem nuclei follows the atlas and terminology of Simpson et al. ( 1986). In the brain stem, cells in the nucleus of the lateral line nerve [Fig. 4( B)] the vestibular nucleus [Fig. 4(B)], nucleus isthmus [Fig. 4(C)], and motor nuclei of cranial nerves V and VII [Fig. 4( D ) ] displayed moderate hybridization signals. Cells of the interpeduncular nucleus were also moderately labeled (not shown). A subpopulation of cells within the dorsal tegmental area of the anterior medulla [Fig. 4( C) ] displayed moderate to high hybridization. No hybridization was observed in the granular layer of the cerebellum or within the central grey [Fig. 4(B,C)].

The strongest hybridization signals to the AR probe in the CNS were found in cells of N. IX-X [Fig. 4( E)] and the motor column of the spinal cord [Fig. 4( F ) ] . Cells of the reticular formation displayed low to moderate hybridization signals [Fig. 4( E)] . In the peripheral nervous system, cells of the postotic ganglia displayed a strong in sitzi hybridization signal with the AR probe [Fig. 4 ( G) ] . No hybridization signal was seen when the sense probe was used [Fig. 4( H ) ] .

HRP-labeled neurons in the N. 1X-X [Fig. 5(A)] and spinal cord at the level of the second and third spinal nerves [Fig. 5 (B)] express AR mRNA. In N. IX-X, we also found HRP-labeled motor neurons with no hybridization to the AR probe and HRP negative cells that express AR mRNA [Fig. 5 (A)]. Double-labeled cells were observed in N. IX-X of both sexes. The number of HRP-labeled cells was smaller than the number of N. IX-X motor neurons expected for PMI (Robertson et al., 1994), suggesting that the HRP injection did not label all N. IX-X motor neurons.

Hormonal Regulation of Number of AR mRNA Expressing Cells in N. IX-X

Gonadectomy and DHT treatment both had a marked effect on the number of AR expressing cells in N. IX-X. An increase in AR mRNA signals within N. IX-X cells was observed in DHT treated

gonadectomized males and females as compared to untreated gonadectomized animals (Fig. 6 ) . A two-way ANOVA comparing intact, gonadectomized, and DHT treated males and females revealed significant effects of treatment (F = 17.93, d f = 2 , p < 0 . 0 0 0 1 ) a n d s e x ( F = 11.35,df= l , p < 0.0 1 ). This analysis was followed by a one-way ANOVA with these six groups of animals ( F = 9.55, df = 5, p < 0.000 1 ) and a Newman-Keuls post hoc analysis for comparisons. In intact animals, the number of AR mRNA positive cells in N. IX-X of males was significantly higher than in females ( Newman-Keuls, p < 0.05; Fig. 7 ) . In contrast, the number of AR expressing cells was not significantly different between sexes in either untreated or DHT treated gonadectomized groups (Fig. 7). One month after gonadectomy, the number of AR expressing cells in N. IX-X was significantly reduced in males, but not in females, compared to intact animals (Newman-Keuls, p < 0.05). Gonadectomized males and females treated with DHT had significantly more AR mRNA positive cells than the respective untreated gonadectomized animals (for both comparisons, Newman-Keuls, p < 0.0 1 ). The effect of estradiol treatment was independently analyzed in females using a one-way ANOVA ( F = 10.7, df = 3 , p < 0.008). The number of AR expressing cells in gonadectomized females treated with estradiol was significantly smaller than in DHT treated gonadectomized females ( Newman-Keuls, p < 0.05) but did not significantly differ from intact or gonadectomized females ( p > 0.05; Fig. 7 ) .

Regulation of AR mRNA Expression in Anterior Brain

The effect of DHT on AR mRNA content in the an- tenor brain was analyzed by nuclease protection assay. In a first set of experiments 5 pg of total RNA from eight to nine animals per group was analyzed (Fig. 8). Optical density values for the male groups were: gonadectomized = 25,152 and gonadectomized-DHT treated = 73,461, and for the female groups were: gonadectomized = 16,174 and gonadectomized-DHT treated = 19,052. Thus nuclease protection assays revealed that AR mRNA levels in the anterior brain were 66% higher in DHT-treated gonadectomized males and 15% higher in DHT-treated gonadectomized females. To confirm these results, total RNA was extracted from a second set ofanimals ( 12- 18 animals per group) and analyzed as above. Optical density values for the male groups were gonadectomized = 19,427, gonadectomized-DHT

CNS Androgen Rrwptor Regulation 563

Figure 5 ( A ) Motor neurons in N. IX-X and (B) motor neurons that project to the arm were labeled with H R P (brown reaction) and AR probe (blue reaction). Some N. IX-X motor neurons express AR mRNA (arrowheads) whereas others d o not (arrows). Scale bar = ( A ) 30 pm, ( B ) 50 pm.

treated = 24,53 1 (20.8% increase), and for the fe- DISCUSSION male groups were gonadectomized = 14,353, gonadectomized-DHT treated = 26,186 (45% higher). We conclude that DHT treatment increases AR mRNA

Motor Neurons in N. IX-X Express AR mRNA

levels in the anterior brain of both gonadectomized males and females.

Target cells whose functions are directly influenced by steroid hormones require the expression of spe-

Figure 6 AR mRNA expressing cells of N. IX-X in (A, C) gonadectomized and (B, D) gonadectomized/DHT treated (A, B) males and (C, D) females. IX-X, nucleus IX-X. An increase in hybridization signal was observed in cells of gonadectomized/ DHT treated groups as compared to untreated/gonadectomized animals. Scale bar = 100 pm.

564 Pkrez, Cohm, und K d k y

60C

In

-1 50C

z $ 2

$ 5

II) In Y 400

L

2 5 30C =

200

3 m = 100 I z

0

I P C O O’ 1 I r p < O 0 1 1

I N T A C T

MALES 0 FEMALES

1 p<O 05

GX GX-DH1

T R E A T M E N T S

I G X ~ E 2

Figure 7 Androgen receptor mRNA expressing cells in N. IX-X of males (striped bars) and females (white bars). Mean values +_ S.E.M. are presented; number of animals analyzed are given with each bar. Intact females have significantly fewer AR mRNA expressing cells than males ( Newmann-Keuls, p < 0.05 ). Gonadectomized (GX) males have significantly fewer AR mRNA expressing cells than intact males ( Newman-Keuls, p < 0.05). Gonadectomized and D H T treated (GX-DHT) males and females have significantly more AR mRNA expressing cells than untreated gonadectomized animals (Newman-Keuls, p < 0.01 ). Gonadectomized females treated with estradiol (GX-E2) have significantly fewer AR mRNA expressing cells than DHT treated females ( Newman-Keuls, p < 0.05).

cific receptors. Once steroid receptors are ligand ac- tivated, they act as transcription factors controlling the expression of hormone-regulated genes. Two neuromuscular systems in X . laevis have been shown to be regulated by androgens: one that con- trols reproductive vocalizations and another involved in the control of clasping during amplexus (reviewed in Kelley, 1996). The muscles in these systems express high levels of AR as demonstrated by binding assays, immunocytochemistry, and in situ hybridization for AR mRNA (Erulkar et al., 1981; Fischeret al., 1993, Dorlochteret al., 1994). AR binding was also demonstrated in motor nuclei ofthe CNS ofX. laevis (Kelley et al., 1975; Kelley, 1981; Erulkar et al., 1981; Gorlick and Kelley, 1986; Brennan and Henderson, 1995 ).

While it seemed very likely that some, if not all, of the cells that express AR in motor nuclei of laryngeal and arm nerves were motor neurons (Kelley, 1980; Erulkar et al., 1981), this identification had not been made directly. The specifica- tion of cell types expressing AR in mRNA in N.

IX-X and anterior spinal cord was approached in the present study by analyzing in situ hybridization signals in cells back labeled from laryngeal and arm muscles with HRP. Results of the present study in- dicate that motor neurons in the medulla that project to laryngeal muscle and motor neurons in anterior spinal cord that project to theflexor carpi vadi- ulis muscle of the arm express AR mRNA. In addition, A R niRNA was not detected in a subpopulation of motor neurons. Thus, in intact males and females, a population of laryngeal motor neurons may not express AR mRNA. N. IX-X con- tains about twice as many cells as there are laryngeal axons (Kelley and Dennison, 1990), and the only motor neurons in the nucleus are those pro- jecting to the larynx (Simpson et al., 1986). Back- fills with DiI demonstrated the presence of interneurons in N. IX-X that are sexually dimorphic in dendritic extent (Kelley et al., 1988; Watson and Kelley, 1989). Because not all motor neurons were HRP filled in the present study (i.e., the number of HRP-labeled cells was smaller than the number of laryngeal motor neurons expected at PM 1 ; Robert- son et al., 1994) and because we do not have spe-

Figure 8 Nuclease protection analysis of gonadectomized ( G X ) and G X / D H T treated ( D H T ) males and females. Total RNA ( 5 p g ) from the anterior brain ofthe experimental groups was hybridized to the AR probe and digested with RNAse. G X / D H T treated animals have relatively more AR mRNA than gonadectomized animals. MW lane: RNA molecular weight markers are in- dicated with arrowheads, 562 and 363 nt (below). Plane: Undigested AR probe. tRNA lane: No protected fragment was found after hybridization of the probe to tRNA. Insert: Total RNA from the respective experimental groups was compared in a 1 % agarose gel.

CNS Androgen Receplor Regiddon 565

cific markers for interneurons, we cannot yet determine whether interneurons in N. IX-X express AR mRNA. However, we can conclude that in intact juvenile male and females the population of laryngeal motor neurons is mixed with respect to AR mRNA expression.

AR mRNA Widely Expressed in X. laevis CNS

I n situ hybridization signals and nuclease protection assays demonstrated that there is wide expression of AR mRNA in the CNS. Levels of transcript are higher in the brain stem than in anterior regions ofthe brain. The in situ hybridization signals found in the cytoplasm of cells in the motor column of the spinal cord and in cells of motor nuclei of the brain stem were comparatively higher than signals in cells of other nuclei of the CNS. Previous to this study, the distribution of AR binding in the X . luevis CNS was known from steroid autoradiographic examination using radioactive testosterone and DHT (Kelley et al., 1975; Kelley, 1980, 1981; Gorlick and Kelley, 1986). Binding of radioac- tively labeled hormone is more restricted than AR mRNA expression. Cells of the striatum, posterior preoptic area, motor nucleus V and VII, nucleus isthmi, and interpeduncular nucleus do not concentrate testosterone or DHT but do express AR mRNA. On the other hand, AR mRNA expression and ligand binding correlate in the ventral thalamus, the dorsal tegmental nucleus, N. IX-X, reticular formation, and spinal cord. Both methods re- veal the absence of AR expression in the anterior preoptic nucleus, anterior thalamus, granular layer of the cerebellum, and within the central grey. AR mRNA expression in the postotic (acoustic) ganglion agrees with observations of DHT binding (D. Kelley, unpubl. observ.). Androgens may contrib- ute to auditory sensitivity (Kelley, 1980).

A broad distribution of AR mRNA was also described in the rat brain, including cortex and cerebellum that do not concentrate significant amounts of radioactive androgens (Simerly et al., 1990; Osada et al., 1993). The probe used in this study was directed against the ligand-binding domain of the receptor. This region is highly specific for each steroid receptor, and it is thus unlikely that the pat- tern of hybridization is the result of cross hybridization to other receptors. Further, specificity of the probe is demonstrated in the nuclease protection assay as a single band of the appropriate molecular weight.

Differences between autoradiography and in

situ hybridization might be due to differences in the sensitivity of the assays or to unrelated ratios of AR mRNA expression to translation. For example, some neurons may not express sufficient AR mRNA for efficient translation. We attempted to address this question using the PG-2 1 antibody, a polyclonal antisera raised against the first 2 1 amino acids of the rat and human AR at the amino termi- nus (Prins et al., 199 1 ). Cell nuclei of HRP-labeled motor neurons were immunoreactive to the antibody. However, in Western analysis of proteins obtained from X . luevis spinal cord, the antibody rec- ognized two protein bands: one at approximately 100 kDa, probably corresponding to the AR, and another at 170 kDa, of unknown identity. Because the PG-2 1 antibody recognizes multiple bands, analysis of AR mRNA and AR protein expression in the CNS of X . luevis awaits development of more specific reagents.

One AR transcript is present in juvenile X . laevis CNS. This result contrasts with the two isoforms expressed in the rat brain (Burgess and Handa, 1993). We previously described two AR mRNA isoforms(9.6 and 8 kb; Fischeret al., 1993) expressed in the larynx of juvenile X . laevis; inter- estingly, the shorter transcript is related to androgen-evoked induction of laryngeal cell proliferation (Fischer et al., 1995). We do not yet know whether this shorter transcript is expressed earlier in the development of the CNS when cell proliferation is prevalent (Gorlick and Kelley, 1987).

As in binding experiments, in situ hybridization and nuclease protection assays suggest that motor neuron nuclei are the major sites of AR expression in the X . luevis CNS. The wide distribution of AR mRNA suggests that cells in different regions of the CNS have the potential to express the protein and that androgens might participate in a variety of CNS functions.

Sexual Dimorphism in AR mRNA Expression in N. IX-X

The number of AR mRNA expressing cells in N. IX-X is higher in males than in females. It is possi- ble that this characteristic may also occur in sexually dimorphic nuclei of other vertebrates (for a re- view see Arnold and Gorski, 1984). At least two mechanisms may explain sex differences in the number of AR mRNA expressing cells in the CNS. First, cell division and/or death of this population may differ in the sexes. Second, the expression of AR mRNA may be under the control of hormones whose circulating levels differ in the sexes. These

explanations are not mutually exclusive, and the second may serve as a mechanism for the first. While no sex difference in cell proliferation is apparent in N. IX-X of X . lacvis, an augmented decrease in the number of axons in the laryngeal nerve of females during tadpole stages leads to a sexually dimorphic axonal content in juveniles and adults (Gorlick and Kelley, 1987; Kelley and Den- nison. 1990). During tadpole stages, axonal decreases can be prevented by androgens (Robertson et al., 1994). Superficially, these results appear to parallel the control of motor neuron numbers in the SNB nucleus of the rat spinal cord, where androgen in males prevents ontogenetic cell death (Nordeen et al., 1985). Thus, androgen attenua- tion of cell death in N. IX-X is a mechanism that may explain differences in the number of cells. It would be of interest to determine whether AR expressing motor neurons are preferentially main- tained by hormones during naturally occurring cell death.

Hormonal Regulation of AR mRNA Expression in N. IX-X

In these juveniles, endocrine manipulations had powerful effects on the number of AR mRNA expressing cells in N. IX-X. Decreases in AR mRNA expressing cells were found in gonadectomized animals, and DHT treatment reversed this effect. The increase in AR mRNA expressing cells was hormone specific because estradiol, a metabolite oftes- tosterone, did not have a significant effect. The effect ofgonadectomy, or androgen removal, is also marked in these animals because they have fewer AR mRNA expressing cells than intact animals. At the moment we cannot determine if the augmented number of AR mRNA expressing cells in males represents more expression per cell or more cells in the nucleus due to reduced cell death. Because cells in N. IX-X are extremely sensitive to circulating androgen levels, the sexual dimorphism in the number of AR mRNA expressing cells in N. IX-X at PM I could be due simply to sex differences in circulating levels of androgens. However, at PM 1 there are no sex differences in DHT or testosterone levels; levels in males are not significantly higher than levels in females until PM4, some 6 months later (Kang et al., I995 ). Regardless, the expression of the receptor is susceptible to variations in circulating levels of androgens in juvenile animals.

Hormonal regulation of AR expression differs in different tissues and species. Our data are consis- tent with decreases in AR immunoreactivity in the

rat brain following long-term gonadectomy and recovery of AR expression due to steroid treatments (Menard and Harlan, 1993; Sar et al., 1990). Sim- ilarly, in cultured insect cells infected with a bacu- lovirus containing an AR cDNA, upregulation of AR expression can be induced by short exposure to testosterone (Kallio et al., 1994). However, in the rat brain Burgess and Handa (1993) observed upregulation of AR mRNA expression by long- term gonadectomy and downregulation following short-term treatments with DHT. In the PMl or PM2 X . laevis larynx, androgen treatment down- regulates AR mRNA expression; the kidney, however, displays increases in AR mRNA content (Fischer et al., 1995).

We can conclude that the expression and regulation of AR mRNA are highly specific to tissue, developmental stage, and hormone treatment. The functional consequences of this regulation are as yet unclear. However, because androgen treatment in juvenile animals causes marked increases in the number of axons in the IX-X nerve, it is likely that this effect requires an increase in AR expression in motor neurons, as opposed to a decrease, to achieve hormone-evoked induction of transcription related to these complex structural changes.

As in other species including humans, AR mRNA expression in the CNS o fX. luevis is higher in the motor nuclei and spinal cord than other brain regions. There is a sexual dimorphism in the number of AR mRNA expressing cells in N. IX- X. The transcription of the receptor can be altered in this nucleus either by absence of endogenous ag- onist or hormone replacement. Our data suggest that motor neurons in this nucleus are direct targets for androgen, and that regulation of AR mRNA expression could play an important role in the involvement of these cells in sexually differentiated behaviors.

We thank Dr. Gail Prins for gift of the PG-21 antibody and Dr. Martha Tobias for helpful comments on the manuscript. This work was supported by NS19949 and by a fellowship from the Ministry of Education and Science, Spain (EX95 50691 740).

REFERENCES

AL SHAMMA, H. A. and ARNOLD, A. P. ( 1995). Impor- tance of target innervation in recovery from axotomy- induced loss of androgen receptor in rat perineal motorneurons. J. Nriivobiol. 28:34 1-353.

ARNOLD, A. P. and GORSKY, R. A. ( 1984). Gonadal ste-

roid induction of structural sex differences in the central nervous system. Annzi. Rev. Ncirrosci. 7:4 13-442.

BERGER, F. G. and WATSON, G. ( 1989). Androgen-regulated gene expression. Anntr. Rcv. Physiol. 515 1-65.

BRENNAN, C. and HENDERSON, L. P. (1995). Androgen regulation of neuromuscular junction structure and function in a sexually dimorphic muscle of the frog. Xenopirs luevis. J . Nezrrohiol. 27: 172- 188.

BURGESS, L. H. and HANDA, R. J. (1993). Hormonal regulation of androgen receptor mRNA in the brain and anterior pituitary gland of the male rat. Mol. Bruin

DORLOCHTER, M., ASTROW, S. H., and HERRERA, A. A. ( 1994). Effects of testosterone on a sexually dimorphic frog muscle: repeated in vivo observations and androgen receptor distribution. J. Nrurobiol. 25:897- 916.

ERIJLKAR, S., KELLEY, D. B., J U R M A N , M., ZLMLAN, F.. SCtiNEIDER, G., and KRIEGE.R, N. (1981). The modulation of the neural control of the clasp reflex in male Xmopi i s Iuevis by androgens. Proc. Nut/. Acud.

FISCHER, L. M., CAI-Z, D., and KTLLEY, D. B. (1993). An androgen receptor mRNA isoform associated with hormone-induced cell proliferation. Proc. Null. Acad. Sci. USA 9053254-8258.

FISCHER, L. M., CATZ, D., and KELLEY, D. B. ( 1995). Androgen-directed development of the Xenopiis Iucvis larynx: control of androgen receptor expression and tissue differentiation. DLJV. Biol. 170: 1 15- 126.

GORLIC'K, D. L. and KEI.LEY, D. B. ( 1986). The ontog- eny of androgen receptors in the CNS of Xrnopus lucvis frogs. Dev. Bruin Rex 26: 193-200.

GmLicti, D. L. and KELLEY, D. B. ( 1987). Neurogen- esis in the vocalization pathway of Xcw)pzi.s 1uwi.s J . Comp. N w d . 257:6 14-627.

HANNIGAN, P. and KELLEY, D. B. (1986). Androgen- induced alterations in vocalizations of female Xcwopiis Iuevis: modifiability and constraints. J. Comp. Phys- iol. 158517-527.

HE, W. W., FISCHER, L. M., SUN, S., BILHARTZ, D. L., ZHU, X., YOUNG, C. Y.-F., KELLEY, D. B., and TIN- DALL, D. J. ( 1990). Molecular cloning of androgen receptors from divergent species with a polymerase chain reaction technique: complete cDNA sequence of the mouse androgen receptor and isolation of androgen receptor cDNA probes from dog, guinea pig and clawed frog. Biochcm. Biophj!y. Res. Commun.

KALLIO, P. J., PALVIMO, J . J., MELITO, M., and JANNE, 0. A. ( 1994). Analysis ofandrogen receptor-DNA in- teractions with receptor proteins produced in insect cells. J. Biol. Chem. 269: 1 15 14- 1 1522.

U N G , L., MARIN, M., and KELLEY, D. (1995). Andro- gen biosynthesis and secretion in developing Xenopus Iuevis. Gen. Comp. Endocrinol. 100:293-307.

KELLEY, D. B. ( 1980). Auditory and vocal nuclei in the

RCS. 19:3 1-38.

Sci. USA 78:5876-5880.

171~697-704.

frog brain concentrate sex hormones. Science 207: 553-555.

KELLEY, D. B. ( 198 1 ). Locations of androgen-concen- trating cells in the brain of Xrnopzts lucvis: autoradiography with 3H-dihydrotestosterone. J . Comp. Neurol. 199:22 1-23 1.

KELLEY, D. B. ( 1996). Sexual differentiation in Xenopiis luevis. In: The Biology ofXenopiis. R. C . Tinsley and H. R. Kobel, Eds. Oxford University Press, Oxford,

KELLEY, D. B. and DENNISON, J. ( 1990). The vocal motor neurons of Xcwopus Iuevis: Development of sex differences in axon number. J. Neurobiol. 21:869- 882.

KELLEY, D., FENSTEMAKER, S., HANNIGAN, P., and SHIH, S. (1988). The sexually dimorphic laryngeal motor neurons ofXenopirs luevis: a quantitative Golgi study. J . Neiirohiol. 19:413-429.

KELLEY, D. B., MORKELL, J. I., and PFAFF, D. W. ( 1975). Autoradiographic localization of hormone-concentrat- ing cells in the brain of an amphibian, Xenopus luevis. 1. Testosterone. J. Comp. Nclurol. 16447-62.

KELLEY, D. B. and PFAFF, D. W. (1976). Hormone effects on male sex behavior in adult South African clawed frogs, Xenoprrs lucvis. Horm. Behuv. 7: I 59- 182.

LLFWELLYN-SMITH, I. J., PILOWSKY, P., and MINSON, J . B. ( 1992). Retrograde tracers for light and electron microscopy. In: lCvppc~rimcntu1 Nc~urounutoiny, .4 Pruc- ticulApprouch. IRL Press, Oxford, U. K., pp. 31-58.

MENARD, C. S. and HARI-AN, R. E. (1993). Up-regulation of androgen receptor immunoreactivity in the rat brain by androgenic-anabolic steroids. Bruin R m 622:

NORDEEN, E. J., NORDEEN, K. W., SENGELAUB, D. R., and ARNOLD, A. P. ( 1985). Androgens prevent nor- mally occurring cell death in a sexually dimorphic spinal nucleus. Science 229:67 1-673.

NOTTEBOHM, F. ( 1980). Testosterone triggers growth of brain vocal control nuclei in adult female canaries. Bruin Res. 189:429-436.

KATO, J. ( 1993). Detection and levels of androgen receptor messenger ribonucleic acid in the rat brain by means of reverse transcription-polymerase chain reaction. Endocr. J . 40:439-446.

PAREDES, R. G. and BAIJM, M. J. ( 1995). Altered sexual partner preference in male ferrets given excitotoxic le- sions of the preoptic area/anterior hypothalamus. J . Nt.urosci. 1566 19-6630.

P ~ R E Z , J. and HOYER, D. (1995). Co-expression ofso- matostatin SSTR-3 and SSTR-4 receptor messenger RNAs in the rat brain. Neuroscience 64~24 1-253.

PRINS, G. S., BIRCH, L., and GREENE, G. L. ( 199 I ). An- drogen receptor localization in different cell types of the adult rat prostate. Endocrinology 129:3 187-3 199.

ROBERTSON, J. C., WATSON, J . T., and KELLEY, D. B. ( 1994). Androgen directs sexual differentiation of la-

U. K., pp. 143-176.

226-236.

OSADA, T., HIRATA, s., HIRAI, M., HAGIHARA, K., and

568 P&z, Cohen, and Kel lq

ryngeal innervation in developing Xenopus laevis. J . Neurobiol. 25: 1625- 1636.

SAR, M., LUBAHN, D. B., FRENCH, F. S., and WILSON, E. M. ( 1990). Immunohistochemical localization of the androgen receptor in rat and human tissues. Endo- crinology 127:3180-3186.

SIMERLY, R. B., CHANG, c., MURAMATSU, M., and SWANSON, L. W. ( 1990). Distribution of androgen and estrogen receptor mRNA-containing cells in the rat brain: an in situ hybridization study. J. Comp. Neurol. 29476-95.

SIMPSON, H. B., TOBIAS, M. L., and KELLEY, D. B. ( 1986). Origin and identification of fibers in the cranial nerve IX-X complex of Xrnvpus Iuevis; Lucifer Yellow backfills in vitro. J. C'omp. Neurol. 244430- 444.

TAKEDA, H., NAKAMOTO, T., KOKONTIS, J., CHODAK, G., and CHANG, C. ( 1991 ). Autoregulation of androgen receptor expression in rodent prostate: immunohistochemical and in situ hybridization analysis. Bio- chem. Biophys. Rex Commun. 177:488-496.

TOBET, S. A., ZAHNISER, D. J., and BAUM, M. J. ( 1986). Sexual dimorphism in the preoptic/anterior hypotha- lamic area of ferrets: effect of adult exposure to sex steroids. Bruin Res. 364:249-257.

TOBIAS, M. L., MARIN, M. L., and KELLEY, D. B. ( I99 1 ). Development of functional sex differences in the larynx ofXenopus 1uevi.s. Dev. Bid. 147:25 1-259.

WATSON, J. and KELLEY, D. B. ( 1989). Development of sex differences in dendritic length in Xrnupus luevis laryngeal motorneurons. Soc. Neurosci. ,4bstr. 15579.

WATSON, J. and KELLEY, D. B. (1992). Testicular masculinization of vocal behavior in juvenile female Xen- opus Iuevis reveals sensitive periods for song duration, rate and frequency spectra. J . Comp. Physiol. 171:

WATSON, J. T., ROBERTSON, J., SACHDEV, U., and KEL- LEY, D. B. ( 1993). Laryngeal muscle and motor neuron plasticity in Xenopus 1uevi.s; testicular masculinization of a developing neuromuscular system. J. N m robiol. 24:1615-1625.

343-350.

androgen receptor mrna expression in xenopus laevis cns: sexual dimorphism and regulation in...

Documents