Multiple Transcripts Encoding Heme Oxygenase-2 in Rat ... - CiteSeerX

BIOLOGY OF REPRODUCTION 53, 1330-1338 (1995)

Multiple Transcripts Encoding Heme Oxygenase-2 in Rat Testis: Developmental and Cell-Specific Regulation of Transcripts and Protein1 William K. McCoubrey, Jr., Benay Eke, 3 and Mahin D. Maines 2 University of Rochester School of Medicine Departments of Biophysics and EnvironmentalMedicine, Rochester, New York 14642 ABSTRACT We report for the first time that heme oxygenase-2 (HO-2) expression is regulated by developmental and cell type-specific factors in the testis, and we describe the presence of three unique sizes of HO-2 transcripts in the testis. HO-2, together with HO-1 (HSP32), catalyzes oxidative cleavage of the heme molecule to biliverdin, carbon monoxide, and iron; HO-2 is the major isozyme of the testis. Northern blot analysis was used to demonstrate the presence of five transcripts for HO-2 in rat testis mRNA; they range from -1.3 to -2.1 kb in length with a predominant 1.45-kb message;three of the transcripts, -1.45 kb, -1.7 kb, and -2.1 kb, are unique to testis. The two other transcripts of -1.3 and -1.9 kb are common to every tissue examined, including the testis. Analysis of three distinct cDNAs isolated from rat libraries in phage lambda indicates that all are identical from -37, relative to translation initiation through the coding region to the first of two poly(A) signals previously identified inthe HO-2 gene (McCoubrey and Maines, 1994). Upstream of -37, the 5' untranslated sequences of the isolates differ in both length and sequence. Comparison with the genomic sequence suggests that the multiple transcripts arise by splicing of alternative first exons as well as use of alternate poly(A) signals. Northern hybridization with probes specific for the unique portion of each cDNA are consistent with this interpretation. Further, unlike HO-1, HO-2 messages are developmentally regulated; only -1.3- and -1.9-kb transcripts were detected, at minute levels, inthe testis RNA of 7-day-old rats. A pronounced increase intotal message level was observed by Day 28 postpartum, although the level had not reached the marked amplification seen in the adult testis. Further, the transcript patterns differed when Day 28 and adult testis were compared to Day 7 testis. The very predominant -1.45-kb band and the -1.7- and -2.1-kb bands were absent from Day 7 testis. Heme oxygenase activity and HO-2 protein levels, as assessed by Western blot, reflect the increases at the RNA level. Interestingly, although abundant HO-2 mRNA can be detected by in situ hybridization in spermatogonia, spermatocytes, and spermatids, HO-2 protein was detected, by immunocytochemistry, only in spermatids. These observations demonstrate tissue and cell specificity of HO-2 gene expression and suggest that in the testis, HO-2 expression is regulated at the transcriptional and translational levels.

INTRODUCTION The microsomal heme oxygenases, HO-1 (HSP32) and HO-2 [1], act in concert with NADPH-cytochrome P450 reductase to catalyze the oxidative cleavage of the a-meso carbon bridge of heme b (Fe-protoporphyrin IX) at the initial, rate-limiting step of heme catabolism (reviewed in [2] and [3]). In mammalian and certain fish species, one product of this reaction, biliverdin IXa, is subsequently converted to bilirubin through the action of the dual nucleotide-dependent cytosolic enzyme, biliverdin reductase [4]. The two HO isozymes are the products of separate genes [5] and differ in both regulation and tissue distribution. HO-2 is the predominant form in the brain and testis, two organs that display impressive heme oxygenase activity [2]; however, unlike expression in the brain [6], HO-2 expression in the testis is translationally regulated [7]. All products of heme degradation activity-biliverdin, CO, and Fe-are known to display biological activity. Biliverdin and bilirubin are potent antioxidants [8, 9], and Fe Accepted July 13, 1995. Received May 17. 1995. 1This study was supported by NIH grants R037 ES04391, ES01247, and the Burroughs Wellcome fund. 2 Correspondence: University of Rochester School of Medicine, Department of Biophysics, 601 Elmwood Avenue, Rochester, NY 14642. FAX: (716) 275-6007. 3 Current address: University of Ankara, Ankara. Turkey.

released by heme oxygenase activity is recognized as a gene regulator for the Fe storage protein, ferritin [10]. Fe also regulates transferrin and NO synthase gene expression [11, 12]. CO is a regulator of cGMP production in the brain and cardiovascular system [13, 14] and hence may function as a cellular signaling molecule [6, 15-17]. It is noteworthy that iron plays an important role in the development of sperm cells [18]; cleavage of the heme molecule by HO isozymes represents the only known enzymatic means of releasing Fe from the molecule. Furthermore, germ cells are exquisitely sensitive to lipid peroxidation, and, indeed, oxygen-free radical species are conjectured to be involved in human sperm dysfunction [19-21]. It is well documented that there are quantitative and qualitative differences in a variety of genes expressed in the testis and other organs [22]. The diversity may include genes that are solely expressed in the testis, genes that give rise to high levels of a particular transcript(s) in the testis, or genes that display stage-specific expression [23, 24]. The present study was undertaken to explore whether there are differences between HO-2 transcripts in the testis and other tissues and whether stage-specific transcripts can be detected in the developing testis. We show that testis RNA contains HO-2 homologous transcripts not detected in other tissues and that the expression of these transcripts is developmentally regulated. We also show that HO-2 protein ex1330

MULTIPLE HO-2 TRANSCRIPTS IN TESTIS

pression is linked to germ cell maturation, suggesting regulation by cell type-specific factors. Furthermore, sequence and Northern blot analysis of several HO-2 cDNAs suggest that some of the HO-2 transcripts arise by splicing of alternative first exons. MATERIALS AND METHODS Materials Oligo(dT)-cellulose, proteinase K, salmon testis DNA, Triton X-100, yeast tRNA, dextran sulfate, 4-chloronapthol, and paraformaldehyde were obtained from Sigma Chemical Company (St. Louis, MO). [a- 32P]dCTP, [a- 35S]dATP, and [y32 P]ATP was supplied by Amersham Corporation (Indianapolis, IN) or DuPont/NEN (Boston, MA). Fluorescein isothiocyanate-conjugated goat anti-rabbit IgG (FITC-GAR), goat anti-rabbit IgG (GAR), and horseradish peroxidaseconjugated goat anti-rabbit IgG (HRP-GAR) were obtained from Organon Teknika-Cappel Corporation (Westchester, PA). The Histostain-SP Streptavidin-biotin system for peroxidase stain of testicular tissue was obtained from Zymed Laboratories (San Francisco, CA). Nytran filters with a 0.2Atm pore size were from Schleicher & Scheull (Keene, NH). Biodyne-A filters for plaque lifts were from Pall BioSupport (Glencove, NY). Immobilon membrane for Western analysis was obtained from Millipore Corp. (Bedford, MA). All chemicals were of the highest purity commercially available. All oligonucleotides used during the course of this study were obtained in HPLC-purified form from Research Genetics (Huntsville, AL) or Midland Certified Reagent Co. (Midland, TX). U.S. Biochemical Corporation (Cleveland, OH) was one source of Taq DNA polymerase and supplied Sequenase Ver. 2.0 and all DNA modifying and restriction enzymes used in these studies. Amplitaq polymerase was also obtained from Perkin-Elmer (Norwalk, CT). Digoxigenin-11dUTP, anti-digoxigenin antibody conjugated to alkaline phosphatase, nitrotetrazolium blue, and 5-bromo-4-chloro3-indolyl phosphate were purchased from BoehringerMannheim (Indianapolis, IN). Animals Sprague-Dawley rats used as a tissue source in these studies were purchased from Harlan Industries (Madison, WI). All animal use procedures were in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the local Animal Care Committee. Rat testis HO-2 was purified and used to raise polyclonal antiserum in New Zealand white rabbits as described previously [1, 25]. Library Screening and Sequence Analysis A rat testis -1.3-kb HO-2 cDNA clone previously referred to as isolate 18B [261, now referred to as rHO-2, was

1331

labeled with [a32P]dCTP via the random primers labeling method [27] and then used as a hybridization probe for screening of rat kidney, brain, and testis cDNA libraries in Xgtll (Clontech Laboratories, Palo Alto, CA). Approximately 5 X 105 plaques on Escherichia coli strain Y1090 were transferred to Biodyne-A filters from each library and probed as described previously [26]. Positive candidates were purified to homogeneity. The inserts ranged in size from -0.5 to -1.7 kb. Three isolates containing the intact coding region, as assessed by polymerase chain reaction (PCR) analysis, were subcloned into pBS + (Stratagene, San Diego, CA) and sequenced by the double-stranded method of Chen and Seeburg [28], with use of M13 universal primers, oligonucleotide primers complementary to previously determined rat HO-2 cDNA sequences [26], and primers specific to the cDNA as required. Probes For Northern blot analysis, the HO-1 hybridization probe was a 569-bp HO-1 fragment corresponding to nucleotides (nt) + 86 to + 654 as reported by Shibahara et al. [29], which was generated via an adaptation of the PCR as previously described [30] and subsequently subcloned as an EcoRI to Pst I fragment in vector pBS +. The HO-2 hybridization probe was the full-length (1300 bp) HO-2 cDNA purified by this laboratory from a rat testis DNA library [26]. Additional probes for unique portions of untranslated sequences with the exception of that for rHO-2-1 were generated via PCR using the appropriate cDNA as template as indicated in figure legends. Products were subcloned into the pCRII vector (Invitrogen, San Diego, CA) and sequenced to confirm their identity. All double-stranded DNA probes used in this study for Northern analysis, including the mouse a-actin cDNA probe [31], were labeled according to the manufacturer's instructions with [a- 32 P]dCTP by the random priming method (Random Primers DNA Labeling System; U.S. Biochemical, Cleveland, OH) and were further purified by spin chromatography [32]. The oligonucleotide probe for rHO2-1 (5'CTCAGCCGCTCTGTAGGTCAGTC-3') was labeled by use of T4 polynucleotide kinase and [y- 32 P]ATP and was also purified by spin column chromatography. RNA Preparationand Northern Blot Analysis Total RNA was prepared from rat tissue by the method of Chirgwin et al. [33], and polyadenylated RNA was isolated by oligo(dT) cellulose chromatography [34]. Formaldehydedenatured RNA was fractionated on a 2.2 M formaldehyde, 1.2% (w/v) agarose gel and transferred to Nytran. For extended resolution electrophoresis, fractionation was carried out in an 18-cm gel at 2 v/cm for 7 h rather than the conventional 9-cm gel run for 3 h. Prehybridization, hybridization of the appropriate 32P-labeled cDNA, and posthybridization treatment of filters were performed essentially as

1332

MCCOUBREY JR. ET AL.

FIG. 1. Northern blot analysis of HO-2 transcripts in rat tissues. Poly(A)' RNA was isolated from rat thymus (lane 1), spleen (lane 2), brain (lane 3), testis (lane 4), kidney (lane 5), and liver (lane 6) and subjected to Northern blot hybridization analysis. Each lane was loaded with 4 pg RNA from the appropriate source. Upper panel: filter probed with full-length (1.3 kb) HO-2 cDNA [261. Lower panel: same filter probed with actin. Cytoplasmic 2.1 factin message is indicated. 1.6-kb a-actin message can also be seen intestis RNA.

described before [30]. For the oligonucleotide probe, the filter was prehybridized for 1 h at 42°C in 6-strength SSC (single-strength SSC is 0.15 M NaC1, 15 mM sodium citrate), 5-strength Denhardt's solution (single-strength Denhardt's solution is 0.02% BSA, 0.02% Ficoll, and 0.02% polyvinylpyrrolidone), 0.05% sodium pyrophosphate, 100 plg/ml denatured salmon testis DNA, and 0.5% SDS. The blot was then hybridized for 18-20 h at 42 0C in 5-strength salinesodium phosphate-ethylenediaminetetraacetic acid, 5strength Denhardt's solution, 0.05% sodium pyrophosphate, 0.5% NP40, and 100 jtg/ml denatured salmon testis DNA containing -1 X 106 cpm/l of 32 -P end-labeled [351 oligonucleotide probe. After hybridization, the filter was washed in 6-strength SSC, 0.05% sodium pyrophosphate twice at room temperature for 15 min and twice at 52°C for 15 min. All blots were exposed to Kodak XAR-5 (Eastman Kodak, Rochester, NY) film at -80°C with intensifying screen. Northern blots were quantified densitometrically with an Ultroscan XI, densitometer (LKB). Immunohistochemisty and In Situ Hybridization of Testis HO-2 Rats were killed by decapitation, testes were removed and fixed in 4% (v/v) paraformaldehyde 16 h at 4°C, and paraffin-embedded and 5-im-thick sections were obtained.

Immunostaining was carried out as detailed previously [36] with FITC-GAR used according to manufacturer's instructions. The pattern of immunofluorescence was obtained in testis from at least 3 individual rats. Furthermore, direct comparison of sections was performed by photographing the specimens under identical conditions of lighting and time exposure. The specificity of immunochemical localization by antibodies was established by preabsorption of HO-2 antiserum with an excess of rat testis HO-2 protein. Incubation of additional testis sections with the preabsorbed antiserum abolished the staining that was observed under immunohistochemistry conditions. For in situ hybridization, rat testes were fixed and paraffin-embedded as described for immunocytochemistry of HO-2 protein. From each specimen, 5-tm-thick sections were cut onto Superfrost Plus (Fisher Scientific, Pittsburgh, PA) slides, and tissues were processed essentially as described before [6]. Immunochemical detection of digoxigenin-labeled HO-2 DNA:mRNA hybrids was carried out using a probe corresponding to nt -32 -+ nt + 504 of the HO-2 sequence. Complexes were visualized by incubation of slides in development buffer containing 0.61 mM nitrotetrazolium blue, 0.52 mM 5-bromo-4-chloro-3-indolyl phosphate, and 0.024% (w/v) levamisole for 5 h in the dark. HO-2 mRNA patterns were obtained in testis of 3 individual rats. In addition to the HO-2 antisense probe, the appropriate HO-2 sense probe (- control) was also tested. Furthermore, a control was run in which no probe was added to the tissue in order to confirm the absence of endogenous alkaline phosphatase activity in the testis sections. Measurement of Microsomal Enzyme Activities A microsomal fraction from testis was prepared and used for determination of heme oxygenase activity as described before, in an assay system supplemented with purified NADPH cytochrome P450 reductase [37]. Protein content was determined by the method of Lowry et al. [381. Duplicate Day 7 samples each consisted of pooled testes from seven animals whereas testes from four individuals were separately processed and analyzed for Day 28 and adult rats. Heme oxygenase activity is expressed as units per milligram of protein and is reported as mean value ± standard deviation. One unit is defined as 1 nmol bilirubin/h. Western Blot Analysis Rat testis microsomes, 150 rtg/lane, were subjected to SDS-PAGE [391 under denaturing conditions, transferred to Millipore Immobilon PVDT membrane, and subjected to Western Blot analysis as described before [5]. The HO-2 protein was visualized by use of monospecific polyclonal rabbit antibody against rat testis HO-2 [251. Purified HO-2 protein was used as the standard.

MULTIPLE HO-2 TRANSCRIPTS IN TESTIS RESULTS The relative distribution of HO-2 message in several adult rat organs was examined by Northern blotting with use of a coding region probe for rat HO-2. As shown in Figure 1 in all organs but the testis, two discrete HO-2 homologous transcripts were detected, although their abundance differed. Specifically, the spleen (lane 2) had the lowest level of HO-2 transcripts, while the brain (lane 3) and testis (lane 4), as predicted from their exceedingly high levels of heme oxygenase activity [2], had the highest levels. Furthermore, there were marked differences in the proportions of the two transcripts across tissues. For instance, in the liver (lane 6), the larger message was nearly undetectable, whereas in organs such as the kidney (lane 5) and brain (lane 3), both transcripts were present in nearly equal abundance. Previously, we had reported the presence of two transcripts of HO-2 in rat brain and kidney [40]. In the testis (lane 4), under the same conditions, HO-2 transcripts did not resolve. However, the overall pattern of hybridization for this tissue appeared to differ from that in other organs. To resolve the HO-2 transcript signal in rat testis, Northern hybridization was carried out with the same probe on control testis from adult rats, as well as testis from Day 7 and Day 28 rats, but under conditions of electrophoresis allowing optimal separation of RNA in the 1-2.5-kb range. Equal amounts of poly(A)+ RNA was loaded for all preparations. As shown in Figure 2a, in the adult, two discrete bands of -2.1 and -1.7 kb could be resolved, while the bulk of the HO-2 transcript appeared to lie between -1.45 and -1.3 kb. HO-2 transcripts were observed to show differential expression in the developing testis. In the Day 7 and Day 28 organs, the overall message abundance was significantly lower than that seen in the adult. Additionally, the distribution of transcripts was also seen to differ in different stages of development. In Day 7 animals, barely detectable transcripts were observed; these transcripts measured - 1.3 and -1.9 kb in length. In Day 28 testis, additional populations were detected. Although they did not resolve under the conditions used, there were clearly transcripts that were not present in Day 7 rats. These included two

FIG. 2. Developmental expression of HO-2 in testis. a) Duplicate 4-pg aliquots of + poly(A) RNA from testes of Day 7 and Day 28 postpartum rats and adults were subjected to extended resolution electrophoresis and Northern blot analysis. Sizes of hybridizing bands were determined by linear regression based on RNA markers run in same gel and stained with ethidium bromide. b) Blot probed with actin. c) Rat testis microsomes from Day 7 and adult animals were analyzed by SDS-PAGE and stained for presence of HO-2 immunoreactive proteins. Duplicate lanes were loaded with 150 ugof protein. Every other lane was loaded with excess microsomes in order to visualize HO-2 in Day 7 testes; some bleeding into empty lanes has occurred for adult sample. d) Microsomal fractions from testes were analyzed for heme oxygenase activity. Activity is shown in units/mg protein; 1 unit catalyzes formation of 1 nmol bilirubin/h. Results represent duplicate pools of testes from seven animals for Day 7 animals and 4 testes for Day 28 and adult animals. Error bars represent standard deviation.

1333

more slowly migrating bands of about -2.1 and -1.45 kb. An extremely prominent signal at - 1.45 kb was clearly developmentally regulated in that it was the most prevalent form in the Day 28 testis and became so prominent in the adult testis that it overwhelmed the resolution of other forms. This level of prominence explains both the impos-

1334

MCCOUBREY JR. ET AL.

FIG. 3. Comparative migration pattern of HO-2 transcripts on extended resolution gel. Northern blot analysis of multiple transcripts of HO-2 inrat tissue poly(A) + RNA was carried out as described for Figure 2 with use of 4 Ig poly(A)' from brain and prostate, and 2 ig poly(A)t RNA from testis. Upper panel: Blot probed with fulllength HO-2 probe. Lower panel: Blot probed with actin.

sibility of detecting the -1.7-kb transcript on conventional gels and the original report on the presence of only one HO-2 transcript in the testis [26]. Probing the same blot with actin (Fig. 2b) confirmed that the increases in message levels for HO-2 were real and not the result of underloading of the Day 7 and Day 28 samples. The increases in overall message level with age were reflected in both HO-2 protein levels as assessed by Western blot (Fig. 2c) and heme oxygenase activity in testis microsomes (Fig.2d). In the Day 7 testis, HO-2 protein was barely detectable; thus accurate quantification of the molecular weight was not possible; however, the molecular weights of the Day 28 and adult testis proteins were essentially identical. The migration pattern of HO-2 purified from adult testis did not differ from that of the brain when examined by Western blotting [41]. To examine whether heme oxygenase activity is reflective of developmental changes in HO-1, the level of this transcript was assessed. Essentially, the level of transcript was not changed in developing tissue (data not shown). To assess whether the same heterogeneity of transcripts occurs in other tissues, Northern analysis was carried out under the same conditions comparing brain (an organ with high levels of two transcripts) and prostate (another organ that is part of the reproductive system) with testis RNA. As shown in Figure 3, the -1.45-, -1.7-, and -2.1-kb populations visible in adult testis, but not seen in Day 7 testis, were not detected in either the brain or prostate. In fact, analysis of other tissues including thymus, kidney, heart, spleen, and liver revealed only the -1.3- and -1.9-kb transcripts (data not shown). This figure also demonstrates that, because of the abundance of the -1.45-kb message in the testis, it was not possible to

clearly resolve the -1.3-kb message in this tissue even when lower amounts of RNA were loaded. HO-2 is the product of a single-copy gene; therefore, the multiple transcripts could arise from alternative splicing, or from polymorphism in 5' or 3' untranslated sequences. In order to assess these possibilities, we sought to identify additional HO-2 cDNAs from brain, kidney, and testis libraries in phage k. One additional full-length cDNA was obtained from each library. These are shown schematically in Figure 4, upper panel, along with the previously reported [261 testis cDNA, designated here as rHO-2. Two of the clones, one from kidney and another from the brain library, had identical 5' untranslated regions (UTR) except for the presence of an additional 4 nucleotides on the 5' end of the latter and are presented as a single drawing-we refer to this as rHO-2-1. The new isolate from the testis library is designated rHO-2-2. All of the isolates were found to be identical in sequence from nt -37, relative to translation initiation, to nt + 1092, the site of polyadenylation of two of the clones, rHO-2 and rHO-2-2. The kidney and brain cDNAs utilized the second polyadenylation signal previously identified in the HO-2 gene and, therefore, have an additional 559 nt in the 3' UTR; the sequence of this region was identical to that reported for the gene [40]. Upstream of nt - 37, the cDNAs differed in both length and sequence; specifically, the first 141 nt of the rHO-2 5' UTR are replaced by unique sequences of 23 nt in rHO-2-1 and 522 nt in rHO2-2 (Fig. 4, lower panel). To examine expression of the various messages represented by these cDNAs in testis, poly(A) + RNA from testis was subjected to Northern blot analysis with use of probes from the unique 5' UTR of each cDNA as well as the region between the two polyadenylation signals. The results, as summarized in Table 1, indicate that all three 5' sequences are utilized in testis, producing two bands that differ in length by approximately the length of sequence between the two signals, with the exception of those hybridizing with rHO-2-2. Although it was not clearly resolved in the adult testis when a full-length probe was used, the rHO2-1 probe revealed the presence of the 1.9-kb transcript seen in Day 7 and Day 28 testis RNA. Differences in transcript size of less than the anticipated 559 bp, such as that seen with rHO-2-2, may reflect differing degrees of polyadenylation or the presence of closely related UTR not yet identified as cDNAs. Consistent with the latter possibility, the rHO-2-2 probe detected an additional band of 0.9 kb (data not shown) that was not detected by the fulllength probe, suggesting that similar UTR may be used on messages other than HO-2. As differences in HO-2 message levels were observed in the developing testis, we wished to determine whether there was also stage-specific expression in developing germ cells. In situ hybridization with a probe from the common 5' region of the HO-2 cDNA, nt - 32 + 504, revealed abundant levels of message in spermatogonia, spermatocytes,

MULTIPLE HO-2 TRANSCRIPTS IN TESTIS -37 +1

-178

I

I

rHO-2

1335

+945 I

I

C

I

+1092 -1 I

-59

+1650

- I

rHO-2-1

I

I

L

-559 rHO-2-2

I i_

I

I

I I

a1 I

rHO-2

-178

ATAAGTCTGG CTATCTCT CCAGGCACGG AGTGTTGGGG TGTGCCATCT CCCCCTFGG GACCACCTTG GGTTGCCCTC TTAGCAAAGT TGGCCTTACC AAGGAGCAGT CATCTTGGAT TATATAATTT GAATGAGCCA AGGACCGAAG TGAGGGCAGC ACAAACAACT CAGCAA -38

rHO-2-1

-59

GCTGACCTAC AGAGCGGCTG AG -38

rHO-2-2

-559

TTCTGGGATC CCGTTCCA TCAGGATCAG AATGTCCTTT GAAACCCTGA ACGGGAGGAT AGCCTATGTG GACACCGAGA

CTCCCGTGTC GTGCTTGTGT GTTCTGCTGC GCGCTAAGAG GTTGFAAA CCCAGAACAG ACGACTTCTC CAGGTGGGAA

TCATGCCTGG GGGTGACTCC CTACTGCTTA CCTGGTCTTG CTGGAAACTA CAGCACTGAG TGCATAAAAG GTGCCTCCGA

TCACGGAGAG GGGCTGCTGC ATCACTGCCC CCTCCACATC GCCTCClTm ATTCCCTGCC AGGTGCTGTG TGTTGATACT GATGGAGCAG CAGCCCCTCT CCCCTGCTTA CAGATCCCTC GTTACGGTGT CCTAGGCAGT CT -38

TCCATCTAGT GGTCTCATCC CCATCTCTCT GGCCCmCG CCGTGACCAG TGAACAGCCC GACGCACGGC

TTCCAGTTGA TCAGCATAGC GGGGCAAATA GCTTCTCTCT GGCATGAGAC AGCAACAGCA CCTCAGGCTG

FIG. 4. Identification of multiple cDNAs for HO-2. Three full-length cDNAs were cloned from Xgt 11libraries and sequenced as described in Materials and Methods. Upper panel: diagrammatic representation of two additional clones and previously reported clone rHO-2. Narrow lines represent sequences unique to each cDNA; boxes indicate common sequences. Coding region is shown as filled box. Polyadenylation signals previously identified in genomic DNA [401 are indicated by arrow heads over clones; nucleotide numbers, relative to ATG start codon with the A residue as + 1,are indicated above. Lower panel: nucleotide sequences of unique 5' regions of cDNA clones.

and spermatids as well as some Leydig cells (Fig. 5a). The authenticity of this staining of cDNA:mRNA hybrids was confirmed by the near loss of signal when the sense orientation of the same probe was used for hybridization (Fig. 5b). Interestingly, immunocytochemistry demonstrated the presence of HO-2 immunoreactive protein primarily associated with the spermatids of the seminiferous tubules, while spermatogonia lacked detectable levels of HO-2 protein (Fig. 5c). The pattern seen could be abolished by preadsorption of the primary antibody with excess HO-2 protein. DISCUSSION The process of spermatogenesis in mammals involves a unique and continuous mitotic and meiotic division of stem cells, culminating in the production of mature spermatids. The incredibly complex nature of spermatogenesis and tes-

ticular differentiation extends to the developmental changes that testes undergo in the process of animal maturation and includes developmentally regulated expression of genes at the transcriptional and translational levels. Presently, we report that HO-2 is among that class of genes that shows qualitative and quantitative differential expression of transcript and protein in the testis. We have detected three sizes of transcripts present only in the maturing and adult testis that were not detected in other organs examined or in the neonatal testis. Aside from the highly abundant -1.45-kb transcript, we detected two other transcripts of -1.7 kb and -2.1 kb that are also specifically present in the maturing and adult testis. Considering that the -1.45-kb transcript is by far the most abundant transcript in the maturing and adult testis, it is evident that the appearance of this transcript is under developmental

TABLE 1. Northern blot analysis of duplicate 4- and 8-pg samples of testis poly (A)+ RNA subjected to electrophoresis with HO-2 probe.* Full length 2.1 1.7 1.45 1.3

rHO-2 (-165 to -39)

rHO-2-1 (-60 to -38)

rHO-2-2 (- 225 to -39)

3' UTR (+1271 to +1482

2.1

1.9 1.6

2.1 1.5

2.1 1.7

1.45

*Sizes of hybridizing transcripts are indicated, determined by linear regression based on the mobility of ethidium bromide stained RNA standards.

1336

MCCOUBREY JR. ET AL.

1337

MULTIPLE HO-2 TRANSCRIPTS IN TESTIS

control. The same kind of control can be surmised for the lower-abundance -1.7- and -2.1-kb transcripts. Translation of HO-2 protein is, on the other hand, clearly linked to maturation of germ cells, as HO-2 protein appears associated only with the mature postmeiotic germ cells. At this time it is not clear which transcript is translated to the protein detected in the mature germ cells, but considering the fact that the -1.45-kb message level is by far the most prominent of the three transcripts, it would seem reasonable to suggest that the HO-2 protein expressed in the testis is to a large extent a translation product of this transcript. The significance of the use of other 5' and 3' UTR that generate the -1.3-kb and -1.9-kb transcripts remains to be elucidated. We have previously shown, with -1.3-kb and -1.9kb transcripts in the brain, that there may also be differences in the translational efficiency of the transcripts [30]. A recent report [42] describes similar observations in mouse male germ cells in which use of alternative promoters of the gene encoding superoxide dismutase leads to formation of multiple transcripts that are apparently translated with different efficiencies. Moreover, we have shown that in transformed cells, the ability to utilize the second poly(A) signal is apparently lost [40]. Accordingly, it is quite plausible that the transcripts unique to the testis have differential translational efficiency and are controlled by a factor(s) that is specific to different stages of germ cell differentiation. However, differential message stability [43] and targeting of message to different subcellular locations [44, 45] could be among the factors that signify generation of other transcripts. The facts that the -1.3-kb and -1.9-kb transcripts are the only two transcripts in somatic cells and that they are the only sizes of transcripts present in Day 7 testis suggest that these transcripts are constitutive and common to all organs and that the others are unique to germ cells. HO-2 is a single-copy gene [40]; therefore, the five HO2 transcripts of different sizes can result from processes such as differential initiation of transcription, alternative use of polyadenylation sites, differing degrees of polyadenylation or differential splicing, and stage-specific exon utilization. The identification of three different 5' untranslated sequences of different length argues for the use of different initiation points in generation of different transcripts. This, plus the presence of two poly(A) signals on HO-2 cDNA [40], would allow generation of transcripts of different sizes. In fact, Northern blot hybridization with probes specific to the 5' untranslated sequences and the nucleotide sequence between the two poly(A) signals indicate that the differently FIG. 5. Detection of HO-2 message and protein in adult testis. a)Testis from adult rats was sectioned and subjected to in situ hybridization with digoxigenin-labeled antisense probe corresponding to nt - 32 - nt +504 of HO-2 cDNA as described in Materials and Methods. b) In situ hybridization with digoxigenin-labeled sense probe for HO-2. c) Adult testis sections were examined by immunocytochemistry with rabbit anti-rat HO-2 antiserum and FITC-conjugated goat anti-rabbit secondary antiserum.

sized transcripts arise from combined use of different 5' untranslated sequences along with alternate usage of the two poly(A) signals. The occurrence of organ-specific exon utilization, although it cannot be ruled out, is not consistent with the similarity in size of the protein, visualized by Western blot analysis of data for the HO-2 protein in the testis and brain of adult rats [41]. Also, as shown here, HO-2 displays a similar mobility when proteins from Day 28 and adult testis are compared. In addition, the coding regions of rHO-2-1 and rHO-2-2, the newly described HO-2 transcripts, are identical to that of rHO-2, with divergence in nucleotide sequence only in the 5' and 3' untranslated domains. Because the previously reported HO-2 gene structure [40] indicates the presence of a splice site between - 36 and - 37, the multiple transcripts of HO-2 probably arise by combined use of alternative first exons and alternate polyadenylation sites. The presence of germ cell-specific proteins is known, and there are suggestions as to the biological importance of such proteins, including "paracrine" intratesticular regulation of functions and fertilization [46, 47]. Although HO-2 is not a germ cell-specific protein, the fact that it is more abundant in the testis than other organs leads us to suggest an important function of the protein in the organ aside from its ability to regulate heme levels. We suggest that both its cell maturation- and organ maturation-dependent expression have relevance to testicular functions and fertility. Considering the integral role of iron in testicular function, it is reasonable to suppose that the metal, released by heme oxygenase isozymes, may be reutilized in testis not only as an essential element for various oxidation-reduction processes but also to modulate iron-responsive gene expression such as that of transferrin, ferritin, and nitric oxide synthase [1012] in the organ. Also, the extreme sensitivity of sperm function to oxygen radicals, together with the high-level ability of spermatozoa to generate reactive oxygen species [19, 48] due to the prevalence of membrane polyunsaturated lipids, creates a demand for an extensive arsenal of defense mechanisms. This, in part, could be provided by the ability of the mature spermatids to generate the antioxidants biliverdin and bilirubin [8]. In this context, it is noteworthy that bilirubin is a lipid soluble compound and concentrates in the membrane lipid bilayer, and also that sperm plasma lipid is most vulnerable to lipid peroxidation; lipid peroxidation is believed to be largely responsible for defective sperm function. Also, the localization of nitric oxide synthase in the various segments of the rat testis, including blood vessels [49], may suggest a role for Fe released by heme oxygenase activity in the regulation of the synthase.

ACKNOWLEDGMENTS We thank Xiao Dan Zhao for performing in situ and immunocytochemical studies, and Cindy Burke and Justin Thornton for preparation of the manuscript.

MCCOUBREY JR. ET AL.

1338 REFERENCES

1. Maines MD, Trakshel GM, Kutty RK. Characterization of two constitutive forms of heme oxygenase: only one molecular species of the enzyme is inducible. J Biol Chem 1986: 261:411-419. 2. Maines MD. Heme oxygenase: function, multiplicity, regulatory mechanisms, and clinical applications. FASEB J 1988; 2:2557-2568. 3. Maines MD. Heme Oxygenase: Clinical Applications and Functions. Boca Raton, FL:CRC Press; 1992. 4. Kutty RK, Maines MD. Purification and characterization of biliverdin reductase from the rat liver. J Biol Chem 1981: 256:3956-3962. 5. Cruse 1, Maines MD. Evidence suggesting that the two forms of heme oxygenase are products of different genes. J Biol Chem 1988; 263:3348-3353 6. Ewing JF, Maines MD. In situ hybridization and immunohistochemical localization of heme oxygenase-2 mRNA and protein in normal rat brain: differential distribution of isozyme 1 and 2. Mol Cell Neurosci 1992; 3:559-570. 7 Ewing JF. Maines MD. Distribution of constitutive (HO-2) and heat inducible heme oxygenase (HO-1) isozymes in rat testes: HO-2 displays stage-specific expression in spermatocytes. Endocrinology 1995; 136:2294-2302. 8. Stocker R, Yamamoto Y,McDonagh AF. Glazer AN, Ames BN. Bilimbin is an antioxidant of possible physiological importance. Science 1987; 235:1045-1047. 9. Neuzil J, Stocker R. Free and albumin bound bilirubin are efficient co-antioxidants for u-tocopherol, inhibiting plasma and low density lipoprotein lipid peroxidation. J Biol Chem 1994; 269:16717-16719. 10. Vile GF, Tyrell RM. Oxidative stress resulting from UJVA irradiation of human skin fibro blasts leads to a heme oxygenase-dependent increase in ferritin. J Biol Chem 1993 268:14678-14681. 11. HuggenvikJ, Sylvester SR,Griswold MD. Control of transferrin mRNA synthesis in Sertoli cells. Ann NY Acad Sci 1984; 438:1-7. 12. Weiss G, Werner Felmayer G, Werner ER. Griinewald K, Wachter H., Hentz MW Iron regulates nitric oxide synthase activity by controlling nuclear transcription. J Exp Med 1994; 180:969-976. 13. Maines MD. Carbon monoxide: an emerging regulator of cGMP in the brain. Mol Cell Neurosci 1993;4:389-397. 14. Ewing JF, Raju VS, Maines MD. Induction of heart heme oxygenase-1 (HSP32) by hyperthermia: possible role in stress-mediated elevation of cyclic 3':5'-guanosine monophosphate. J Pharmacol Exp Ther 1994; 271:408-414. 15. Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. Carbon monoxide: a putative neural messenger. Science 1993; 259:381-383. 16 Zhou M,Small SA,Kandel ER, Hawkins RD. Nitric oxide and carbon monoxide produce activity-dependent long term synaptic enhancement in hippocampus. Science 1993 260:1946-1950. 17. Prabhakar NR, Dinerman JL, Agani FH, Snyder SH. Carlxbon monoxide: a role in carotid body chemoreception. Neurobiology 1995 92:1994-1997. 18. Griswold MD. Protein secretion of Sertoli cells. Int Rev Cytol 1988; 110:133-156. 19. Alvarez JG, Touchstone JC, Blasco 1, Storey BT. Spontaneous lipid peroxidation and production of hydrogen peroxide and superoxide in human spermatozoa. J Androl 1987: 8:338-348. 20. Aitken RJ,Buckingham DW, Harkiss D. Use of a xanthine oxidase free radical generating system to investigate the cytotoxic effects of reactive oxygen species on human sper matozoa.J Reprod Fertil 1993; 97:441-450. 21. Griveau JF, Dumont E, Renard P, Callegari JP, Le Lannou D. Reactive oxygen species, lipid peroxidation and enzymatic defence systems in human spermatozoa. J Reprod Fertil 1995 103:17-26. 22. Wolgemuth DJ, Watrin F. List of cloned mouse genes with unique expression patterns during spermatogenesis. Mammal Gen 1991; 1:283-288. 23. Sarge KD, Park-Sarge O-K, Kirby JD, Mayo KE, Morimoto RI. Expression of heat shock factor 2 in mouse testis: potential role as a regulator of heat-shock protein gene expres sion during spermatogenesis. Biol Reprod 1994 50:1334-1343. 24. Walter L, Dirks B, Rothermel E, Heyens M, Szpirer C, Levan G, Gunther E. A novel

25.

26. 27. 28. 29 30. 31. 32. 33 34. 35.

36

37 38. 39. 40. 41. 42.

43.

44.

4i. 46.

47. 48.

49.

conserved gene of the rat that is developmentally regulated in the testis. Mammal Gen 1994; 5:216-221. Trakshel GM, Kutty RK, Maines MD. Purification and characterization of the major constitutive form of testicular heme oxygenase: the non-inducible isoform.J Biol Chem 1986: 261:11131-11137. Rotenberg MO, Maines MD. Isolation, characterization, and expression in Eschechia coli of a cDNA encoding rat heme oxygenase 2. J Biol Chem 1990; 265:7501-7506. Feinberg AP, Vogelstein B. A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1983: 132:6-13. Chen EY,Seeburg PH. Supercoil sequencing: a fast and simple method for sequencing plasmid DNA. DNA (New York) 1985 4:165-170. Shibaharra S, Mueller RM. Taguchi H, Yoshida T. Cloning and expression of cDNA for rat heme oxygenase. Proc Natl Acad Sci USA 1985; 240:7865-7869. Sun Y,Rotenberg MO, Maines MD. Developmental expression of heme oxygenase isozymes in rat brain: two HO-2 mRNAs are detected. J Biol Chem 1990: 265:8212-8217. Minty AJ, Caravatti M, Robert B. Cohen A, Dauber 1', Weydert A, Gros F, Buckingham ME. Mouse actin messenger RNA's. J Biol Chem 1981: 256:1008-114. SambrookJ, Fritsch EF,Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY:Cold Spring Harbor Laboratory Press; 1989. Chirgwin JM, Przybyla AE. MacDonald RJ,Rutter WJ. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979; 18:5294-5299. Ausubel FM, Brent R, Kingston RE, Moore Dl), Smith JA, Seidman J, Struhl K. Current Protocols in Molecular Biology. New York, NY:John Wiley and Sons: 1987. Mischke D, Pardue ML.Organization and expression of alpha-tubulin genes in Drosophila melanogaster One member of the alpha-tuulin multigene family is transcribed in both oogenesis and later embryonic development. J Mol Biol 1982:156:449-466. Ewing JF, Maines MD. Rapid induction of hemeoxygenase-I mRNA and protein by hy perthermia in rat brain: heme oxygenase-2 is not a heat-shock protein. Proc Natl Acad Sci USA1991; 88:5364-5368 Maines MD, Chung A, Kutty RK. Inhibition of testicular heme oxygenase activity by cadmium: a novel cellular response. J Biol Chem 1982: 257:14116-14121. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ.Protein measurement with folin phenol reagent. J Biol Chem 1951;193:256-275. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-685. McCoubrey WK JR, Maines MD. The structure, organization and differential expression of the gene encoding rat heme oxygenase-2. Gene 1994; 139:155-161. Mark JA, Maines Ml). Tin-protoporphyrin-mediated disruption in vivo of heme oxygenase-2 protein integrity and activity in rat brain. Pediatr Res 1992; 32:324-329. Gu W. Morales C, Hecht NB. In male germ cells, copper-zinc superoxide dismutase utilizes alternate promoters that produce multiple transcripts with different translation potential. J Biol Chem 1995 270:236-243. Sachs AB. Messenger RNA degradation in eukaryotes. Cell 1993;74:413-421. Kislauskis EH, Li Z, Singer RH, Taneja KL.Isoform specific 3' untranslated sequences sort alpha-cardiac and beta-cytoplasmic actin messenger RNAs to different cytoplasmic compartments.J CellBiol 1993; 123:165-172. Kislauskis EH, Zhu X, Singer RH. Sequences responsible for intracellular localization of beta-actin messenger RNA also affect cell phenotype. J Cell Biol 1994 127:441-451. O'Hara B, Donovan DM, Lindberg I, Brannock MT, Ricker DD, Moffatt CA, Klaunberg BA, Schindler C, Chang TSK, Nelson RJ,Uhl GR. Proenkephalin transgenic mice: a short promoter confers high testis expression and reduced fertility. Mol Reprod l)ev 1994: 38:275-284. Xiong Y, Hales DB. Immune-endocrine interactions in the mouse testis: cytokine-mle diated inhibition of Leydigcell steroidogenesis. EndocrJ 1994: 2:223-228. Aitken RJ, ClarksenJS. Cellular basis of defective sperm function and its association with the genesis of reactive oxygen species by human spermatozoa. J Reprod Fertil 1987: 81:459-469. Burnett AL, Ricker Dl), Chamness SL,Maguire MP, CroneJK, Bredt DS, Snyder SH, Chang TSK.Localization of nitric oxide synthase in the reproductive organs of the male rat. Biol Reprod 1995 52:1-7.