Novel Repeat Elements Direct Rat Proenkephalin Transcription during Spermatogenesis*

(Received for publication, August 19, 1996, and in revised form, November, 20, 1996)

Feng Liu , John Tokeson Dagger , Stephan P. Persengiev , Karl Ebert § and Daniel L. Kilpatrick

From the Neurobiology Group, Worcester Foundation for Biomedical Research, Shrewsbury, Massachusetts 01545 and the § Department of Anatomy and Cell Biology, Tufts University School of Veterinary Medicine, North Grafton, Massachusetts 01536

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES


ABSTRACT

The developmental program controlling sperm formation occurs in multiple stages that sequentially involve mitosis, meiosis, and spermiogenesis. The transcriptional mechanisms regulating these distinct phases are poorly understood. In particular, while a required role for the germ cell transcription factor cyclic AMP response element modulator-tau during spermiogenesis has recently been demonstrated, the transcriptional mechanisms leading to early haploid cell formation are unknown. The rat and mouse proenkephalin genes are selectively expressed from an alternate, germ cell-specific promoter in meiotic and early haploid cells. In this study, the minimal rat proenkephalin germ line promoter was localized to a 116-bp region encompassing the transcriptional start site region. Further, a proximal 51-bp sequence located in the 5'-flanking region is absolutely required for germ line promoter activity. This 51 bp sequence corresponds to a previously characterized binding element (GCP1) that forms cell-specific complexes with rat spermatogenic cell nuclear factors distinct from cyclic AMP response element binding proteins. Further, GCP1 contains novel direct repeat sequences required for factor binding and transgene expression in spermatogenic cells. These repeat elements are highly similar to sequences within the active regions of other male germ line promoters expressed during meiosis. GCP1 may therefore contain transcriptional elements that participate more generally during meiosis in the differentiation of spermatocytes and early haploid spermatids.


INTRODUCTION

Spermatogenesis consists of a well defined developmental program involving a series of proliferative and differentiative stages. These events can be divided into three phases: 1) mitosis involving spermatogonial cell types, 2) meiotic division and differentiation of spermatocytes into haploid spermatids, and 3) spermiogenesis in which spermatids morphologically differentiate into spermatozoa. Successful integration of these phases requires the sequential expression of different subsets of genes and involves, in part, stage-dependent transcription from spermatogenic cell-specific promoters. Developmental regulation of distinct and sometimes cell-specific transcription factors appears to be a critical mechanism controlling this spermatogenic program. For example, a number of proliferation-associated transcriptional regulators, including c-fos, c-jun, and c-myc, are expressed exclusively in spermatogonial cell types (1, 2). Similarly, germ cell-specific forms of CREB1 and CREM, as well as novel mRNAs for Sp1, are expressed during meiotic and/or postmeiotic stages (3-5).

Recent studies have begun to shed light on the role of transcriptional regulators during specific phases of sperm development. CREMtau , a spermatogenic cell-specific CREM isoform, functions as a transcriptional activator of promoters containing cAMP response elements (CREs) (6). Expression of CREMtau protein is reportedly restricted to postmeiotic (spermatid) stages, suggesting a specific role during spermiogenesis (4). Consistent with this, it has been shown to activate transcription from a number of haploid-specific promoters that contain CREs, including those for RT7, protamine-1, and transition protein-1 (6-8). In addition, CREM-deficient mice are infertile due to the failure to complete spermiogenesis beyond the round spermatid stage, and they lack testicular expression of several haploid-specific, CRE-dependent genes (9, 10). Clearly, while CREMtau has a critical role in regulating spermiogenesis via CRE-containing promoters, additional transcription pathways are required for appropriate sperm maturation. For example, several germ cell promoters are active during meiosis alone or in both late meiotic and early postmeiotic germ cells, including those for histones H1 and H2B, lactate dehydrogenase c, Pdha-2, proacrosin, and pgk-2 (11-16). Further, formation of pachytene spermatocytes and early (round) spermatids is not interrupted in CREM-deficient mice, and proacrosin expression persists in these mutants (9, 10). Thus, CREM-independent transcriptional mechanisms clearly are important for meiosis and haploid cell formation.

The proenkephalin gene, which codes for enkephalin-containing opioid peptides, is expressed in a stage-dependent manner during rat and mouse spermatogenesis (17, 18). Opioid peptides have been implicated in paracrine interactions between germ cells and Sertoli cells that may be important for maintenance of spermatogenesis (19, 20). Proenkephalin expression occurs initially at low levels in early meiotic stages and then increases markedly in late pachytene spermatocytes and round spermatids and declines thereafter. These cells use an independent, alternate promoter within the single copy proenkephalin gene to synthesize a novel 1700-nt germ cell-specific mRNA (21). Previous studies have demonstrated that the rat proenkephalin germ line promoter is contained within a 500-bp TATA-less sequence located within the first somatic intron (22). While the upstream somatic promoter contains functional CRE elements (23), the germ line proenkephalin promoter lacks such consensus sequences. Since this promoter is highly active in both pachytene spermatocytes and round spermatids, its characterization should provide valuable insight into CRE-independent transcriptional mechanisms governing formation of late meiotic and early haploid cells. Here we report the identification of novel cis-acting repeat elements required for expression of the rat germ line proenkephalin promoter using a transgenic mouse model. The existence of germ cell-specific nuclear factors interacting with these elements is also demonstrated.


MATERIALS AND METHODS

DNA Constructs

Previous transgenic studies (22) used a rat proenkephalin chloramphenicol acetyltransferase (CAT) fusion construct encompassing the germ cell proenkephalin start site region as well as adjacent 3'- and 5'-sequences (RPKCAT0.5). Derivatives of this transgene were generated as follows (see Fig. 1). RPKCAT0.5Delta MSP and RPKCAT0.5Delta BSM were prepared by isolation of HindIII-MspI and HindIII-BsmI fragments, respectively, from RPKCAT0.5 and subcloning into the promoterless RPKCAT vector using HindIII and SmaI sites. RPKCAT0.4 was made by deletion of the first 100 bp from RPKCAT0.5 using TfiI and XhoI and re-ligation of blunt ends.


Fig. 1. Expression patterns of different RPKCAT transgenes. Promoter regions contained within different transgene constructs are shown as solid bars above the corresponding rat proenkephalin promoter regions. The black boxes in the GCP1mut construct indicate the locations of mutated repeat elements. The initiation sites for the somatic (upstream) and germ cell (downstream) transcripts are indicated by arrows. The locations of restriction sites used in generating different constructs and of the GCP1 element are also shown. Exon 1S is the first exon for the somatic transcript. Exon 1T is the first exon for the germ cell-specific transcript. Expression data for the CAT reporter gene in transgenic mice are presented in the table as the numbers of independent lines that specifically contained CAT transcripts in the testis and germ cells, but not in somatic tissues, and that were initiated from the appropriate start sites (plus column). The minus column shows the number of transgenic lines that did not express CAT either in testis or in somatic tissues. The asterisk indicates aberrant CAT expression in both testis and somatic tissues in one transgenic line for the RPKCAT0.4 construct.
[View Larger Version of this Image (32K GIF file)]


Constructs RPKCAT0.5-3'Del, RPKCAT0.4-3'Del, RPKCAT5'Del, RPKCAT0.45, and RPKCAT0.45-3'Del were all generated using the polymerase chain reaction (PCR). Vent DNA polymerase (New England Biolabs, Inc., Beverly, MA) was used in 20 cycle reactions. The DNA sequences of 5'-primers were as follows: 0.5-3'Del, GTACTACCATCGATGCATGCGATCGTCCAACTTTCT; 0.4-3'Del, CAGTACCGCTCGAGAATCTTGCCCTAAGCAACCGAC; 0.45 and 0.45-3'Del, ACTACCATCGATGCGCCTGCTCCCGAGCCG; and 0.5-5'Del, GTACTACCATCGATCAGGAAGACAGGATGCCCCA. The 3'-primer used for each RPKCAT-3'Del construct was TCCTGTCCTTCCGAGGGCGC, while that used for constructs 0.5-5' Del and 0.45 was GCTGGTACCAGATCTGAGCTC. Restriction sites within the RPKCAT vector used for subcloning of PCR products were as follows: 0.5-3'Del, 0.5-5' Del, 0.45 and 0.45-3' Del for ClaI/SmaI; and 0.4-3' Del for XhoI/SmaI.

The transgene GCP1mut was prepared by annealing sense and antisense oligonucleotides containing GCP1 sequences that extended to the third repeat element at its 3'-end where a TfiI site is located and that contained mutations in the second and third repeats (shown in lower case): CGATGCGCCTGCTCCCGAGCCGCGAAtaCagtTGTTCGtaCagt (sense) and actGtaCGAACAactGtaTTCGCGGCTGGGAGCAGGGCGCAT (antisense). This annealed fragment contained a cohesive PstI overhang at its 5'-end and a blunt 3'-end. A DNA fragment containing the remaining 3'-sequences for RPKCATO.45 was isolated as a TfiI-XmaI fragment, with the TfiI site blunt-ended using Klenow. These two DNA fragments were then ligated to a PstI/XmaI fragment prepared from pRPKCAT0.5 containing CAT, SV40, and vector sequences.

All constructs were confirmed by DNA sequencing before use in transgenic experiments. Transgenes containing promoter, CAT, and SV40 3' sequences were released by restriction digestion of plasmid DNAs and gel purified prior to injection. Numbering of sequences within the germ line promoter are based on assignment of the most upstream initiation site as +1.

Generation of Transgenic Mice

Transgenic mice were made by microinjection of the purified transgenes into the male pronucleus of fertilized mouse eggs. To determine transgene integration, tail DNA from founder mice was analyzed by PCR using primers to CAT sequences (ACGTTTCAGTTTGCTCATGG and AGCTAAGGAAGCTAAAATGG) (24). TSHbeta primers (TCCTCAAAGATGCTCATTAG and GTAACTCACTCATGCAAAGT) were used as a positive internal control.

Positive founders were bred to obtain male transgenic offspring prior to analysis. Tissues were removed from males for each positive line following sexual maturation. A sufficient number of independent transgenic lines (typically 4-8) were generated for each construct to ensure clear interpretation of results. In particular, all non-functional constructs in this study yielded no positive lines, while expressing transgenes had an expression rate of typically 67% or greater. This success rate for functionally active transgenes is consistent with previous transgenic studies (22, 25).

RNase Protection

Total RNA from tissues or isolated cells was prepared by the method of Chirgwin et al. (26). RNase protection was performed as described previously (25). Briefly, radiolabeled antisense RNA probes were generated from linearized plasmids using [alpha -32P]UTP and either T7 or T3 RNA polymerase. The RNA probes were annealed to total RNA in a final volume of 30 µl of hybridization buffer and then digested with 40 µg/ml RNase A and 2 µg/ml RNase T1. Samples were treated with proteinase K (20 mg/ml) and then extracted, precipitated, and analyzed on 5% polyacrylamide sequencing gels. Riboprobes vectors for CAT (pCAT) sequences and the rat proenkephalin germ line start site region (pAva) have been previously described (25). The vector p0.45-3'DelCAT was used for start site analysis of the RPKCAT-3'Del constructs. It was prepared by subcloning a XhoI/EcoRI fragment from RPKCAT0.45-3'Del into the same sites within pBS-SK(-) (Stratagene, La Jolla, CA).

Preparation of Germ Cells and Nuclear Extracts

Spermatogenic cells enriched in pachytene spermatocytes and spermatids were prepared by enzymatic digestion of adult transgenic mouse testes essentially as described by Bellve et al. (27). For nuclear extracts, enriched spermatogenic cells were isolated from adult rat testes and homogenized in the presence of a protease inhibitor mixture (5 µg/ml each aprotonin, pepstatin A, and leupeptin, 2 mM benzamidine, and 0.5 mM phenylmethylsulfonyl fluoride) (22). Nuclei were isolated on sucrose gradients, and nuclear proteins were extracted as described previously (22, 28) and stored at -70 C.

Gel Mobility Shift Assays

Gel-shift assays were performed according to previous protocols (22) with some modifications. Briefly, 0.25-1 ng of 32P-labeled DNA (oligonucleotide or genomic) was incubated on ice for 20 min with 2 µg of nuclear extract in a total reaction buffer volume of 14 µl. The reaction buffer contained 10 mM HEPES (pH 7.9), 30 mM NaCl, 10% (v/v) glycerol, 0.5 mM dithiothreitol, 6 mM MgCl2, 6 mM spermidine, 0.5 µg of sonicated salmon sperm DNA, and 0.5 µg of poly(dA-dT). DNA-protein complexes were resolved on 4% non-denaturing polyacrylamide gels using low ionic strength buffer containing 12.5 mM Tris borate (pH 8.0) and 0.25 mM EDTA. Competition assays were performed by addition of excess amounts of unlabeled oligonucleotide into the reaction mixture.

Double-stranded oligonucleotides and genomic restriction fragments containing 5' overhangs were labeled with [alpha -32P]dCTP by fill-in reactions with Klenow. Genomic fragments were prepared by restriction digestion of pRPKCAT0.5 (22). DNA sequences of oligonucleotide probes were as follows: GCP1, GCGCCTGCTCCCGAGCCGCGAACTCCAGTGTTCGCTCCAGAATCTTGCCCT; and GCP1-mut123, GCGCCTGtaCagtAGCCGCGAAtaCagtTGTTCGtacagtAATCTTGCCCT. GCP1mut1, GCP1mut2, GCP1mut3, and GCP1mut23 oligonucleotides contained the same mutated bases of the indicated repeat elements as in GCP1mut123. TATA-mut sequences were used as nonspecific competitor DNA: GGGGGGGGAGAAAAGGGGT. Additional competitor DNAs used were: CRE, AGAGATTGCCTGACGTCAGAGAGCTAG; CREmut, AGAGATTGCCTGTGGTCAGAGAGCTAG; and Gli, GAA AGATTGTCCCTGCTGGTCCTGCTCCACGACCCACCCGGCAAGGTT.

Autoradiographs of competition experiments using wild-type and mutated GCP1 sequences were quantified using densitometric scanning (PDI, Huntington Station, NY).

Sequence Analysis

A search for transcription factor binding elements within rat proenkephalin promoter sequences was performed using the Tfsites program (Genetics Computer Group, Madison, WI).


RESULTS

Functional Mapping of Downstream Sequences within the Testis-specific Proenkephalin Promoter

Previous work demonstrated that a fusion gene construct (RPKCAT0.5) containing 500 bp of the rat proenkephalin gene linked to the reporter gene CAT was sufficient for cell- and stage-specific expression of the germ line proenkephalin promoter (22). RPKCAT0.5 contains the transcriptional start site region for the rat promoter as well as neighboring 5'- and 3'-sequences (Fig. 1). To further map cis-elements that specify spermatogenic cell transcription, a series of proenkephalin-CAT constructs containing different portions of this 500 bp promoter sequence were generated and tested in transgenic mice. Fig. 1 summarizes the results for a series of nine different constructs in comparison with RPKCAT0.5. Constructs RPKCAT0.5Delta MSP (-118 to +312), RPKCAT0.5Delta Bsm (-118 to +168), RPKCATO.8 (which also contains the upstream somatic promoter), and RPKCAT0.5-3'Del (-118 to +62) comprise sequential deletions of the 3' promoter region. All four constructs actively expressed CAT transcripts in testes of transgenic mice but not in somatic tissues (kidney, liver, heart, or cerebellum) as determined by RNase protection analysis of total RNA using an antisense riboprobe to CAT sequences. In each case, a 250-nt CAT protection fragment was selectively detected in testis RNA (Fig. 2; data not shown). To confirm that the transgenes were expressed in the male germ line, RNase protection was performed using RNA from enriched preparations of pachytene spermatocytes and round spermatids, stages that express rat proenkephalin in high amounts. CAT transcripts were abundantly detected in these germ cells for all four transgenes (Fig. 3A and data not shown). The level of transgene expression in germ cells was equivalent to that for the original RPKCAT0.5 construct, as shown for RPKCAT0.5-3'Del (Fig. 3A).


Fig. 2. 3' deletion analysis of the germ line promoter by RNase protection. Total RNA (50 µg) prepared from various tissues was examined by RNase protection for different RPKCAT transgenic lines harboring different 3' deletions of the proenkephalin germ line promoter (see Fig. 1). The riboprobe used was pCAT, which contains 250 nt of CAT coding sequence. Transgene designations are given above the bars and transgenic line numbers and tissues are indicated above each lane. T, testis; K, kidney; H, heart; L, liver. CAT sense strand, which contains 270 nt of complementary CAT and plasmid sequence, was used as a positive control. Riboprobe digested in the presence of yeast RNA served as a negative control (probe). MspI-digested fragments of pBR322 were used as size markers (M), and their positions are indicated in base pairs in the margins. Arrows indicate the 250 nt protection products.
[View Larger Version of this Image (49K GIF file)]



Fig. 3. Expression of the RPKCAT0.5-3'Del transgene in spermatogenic cells. A, protection analysis of CAT transcripts in total RNA from spermatogenic cells. Fifty µg of total RNA was used except for the first lane of line 1 for the 0.5-3'Del transgene (0.5-3'Del-1) (100 µg). RNA from purified germ cells of RPKCAT0.5 transgenic mice (0.5(+)) was used as a positive control while kidney RNA (0.5-3'Del-1K(-)) served as a negative control. Probe, pCAT riboprobe digested alone; M, size markers. B, analysis of transcription start sites for the RPKCAT0.5-3'Del construct. RNase protection was performed with the p0.45-3'DELCAT riboprobe using total RNA from transgenic mouse testes and kidney (50 µg of each). Total RNA from non-transgenic mouse testis (MT) was used as a negative control. Numbers above lanes refer to specific transgenic lines. K, kidney; T, testis. The major protected bands have lengths of approximately 300, 309, 320, and 334 nt and are shown by arrows. This result agrees well with the sizes predicted for transcripts initiated from the 3'-Del transgenes, as shown in the schematic in panel C. C, the region encompassed by the p0.45-3'DELCAT riboprobe is shown together with the four observed protection fragments generated from the initiation region (In) below it.
[View Larger Version of this Image (37K GIF file)]


Transcriptional initiation of the 1.7-kilobase proenkephalin mRNAs in rat and mouse testis occurs at multiple sites spanning a 35-bp region within the germ line promoter. RNase protection using probes encompassing this region was performed to determine whether transcription of the four transgenes initiated from these same sites. Analysis of testicular RNA from RPKCAT0.5-3'Del mice using the p0.45-3'DELCAT riboprobe generated multiple protection fragments of approximately 300, 309, 320, and 334 nt that agree well with the predicted sizes for products initiated from the germ line start sites (Fig. 3, B and C). Appropriate initiation from the other 3' deleted transgenes (RPKCAT0.5Delta MSP, Delta BSM, and RPKCAT0.8) was also confirmed by RNase protection (data not shown). Based on these results, sequences downstream of position +62 are dispensable for appropriate stage- and cell-specific initiation from the rat proenkephalin promoter in male germ cells.

Analysis of 5'-Flanking Sequences

In contrast to the above findings, deletion of either 128 (-118 to +10) (RPKCAT5'-Del) or 105 bp (-118 to -14) (RPKCAT0.4) of the 5'-flanking region resulted in complete loss of detectable CAT transcripts in testes of transgenic mice (Figs. 1 and 4). For all but one transgenic line, CAT expression was also absent in somatic tissues of mice harboring these transgenes. One RPKCAT0.4 line exhibited expression in multiple somatic tissues as well as testis (data not shown). This appears to reflect the fortuitous insertion of the transgene adjacent to or within a ubiquitously active transcription unit. These results demonstrated the existence of one or more functional cis-elements within the 5'-flanking sequence extending from -118 to -14.


Fig. 4. 5'-deletion analysis in transgenic mice. Total RNA (50 µg) from different tissues of mice containing the RPKCAT-5'Del and 0.4 constructs was submitted to RNase protection using the pCAT riboprobe as described in the legend to Fig. 2. Number designations above each lane refer to the specific transgenic line examined. T, testis; K, kidney.
[View Larger Version of this Image (21K GIF file)]


Previous experiments detected multiple sites within the 500 bp proenkephalin germ line promoter that were bound by nuclear proteins from spermatogenic cells (22). One of these sites, termed GCP1 (-54 to -4), was contained within the functional 105 bp 5'-flanking sequence, immediately upstream of the initiation region (Fig. 1). To determine whether binding within this region was functionally important, another transgene (RPKCAT0.45) was tested in which the 51-bp GCP1 sequence was selectively added back to the RPKCAT0.4 transgene. Readdition of these sequences restored testis- specific transgene expression (Fig. 5A). The level of RPKCAT0.45 expression was also comparable with that found for the other functional transgenes (Fig. 5A). Further analysis also confirmed that the transgene was expressed in spermatogenic cells (Fig. 5B). To determine whether transcription initiated from the appropriate start sites, RNase protection was performed on transgenic testis RNA using the pAva riboprobe (Fig. 5C). Multiple protection fragments identical in size to those generated from rat testis RNA were observed. Thus, sequences from -54 to -4 contain regulatory elements that are critical for stage-dependent, spermatogenic cell-specific expression of the rat proenkephalin gene.


Fig. 5. Expression of the RPKCAT0.45 transgene in testis and spermatogenic cells. A, RNase protection analysis of transgenic mouse tissues. Lanes are labeled as in Fig. 2, with numbers above lanes referring to the individual transgenic lines tested. T, testis; K, kidney; C, cerebellum. The arrow indicates the specific 250-nt CAT protection fragment. Testis RNA for line 58-1 was run in two amounts (50 and 20 µg); all other samples were 50 µg. B, protection analysis of spermatogenic cells from RPKCAT0.45 transgenic mice. Samples were analyzed as described in the legend to Fig. 3A. K(-), transgenic mouse kidney RNA; 0.5(+), total RNA from RPKCAT0.5 germ cells; 0.45-1, germ cell RNA from RPKCAT0.45 line 1. The arrow indicates the 250-nt CAT transcript. C, transcriptional start sites were determined using the pAva riboprobe and 50 µg of total RNA from transgenic mouse testis. The specific protection products are shown by the bracket. Rat testis (RT) and non-transgenic mouse testis (MT) served as positive and negative controls, respectively. The transgenic lines tested are given by numbers above each lane. T, testis; K, kidney.
[View Larger Version of this Image (72K GIF file)]


To define more accurately the minimal promoter sequences required for the proenkephalin germ line promoter, the construct RPKCAT0.45-3'Del was studied (Fig. 1). This transgene was equally active in directing testis-specific expression (Fig. 6A) that also localized to germ cells and initiated from the appropriate sites (data not shown), thus further defining the minimal promoter region to a 116-bp sequence (-54 to +62). In addition to GCP1, this minimal promoter encompasses the initiation region and a previously defined (22) downstream protein interaction site (GCP2) (Fig. 6B).


Fig. 6. Analysis of the RPKCAT0.45-3'Del transgene. A, RNase protection analysis was performed on tissues from different transgenic lines as in Fig. 2. The individual transgenic lines tested are indicated by the numbers above each lane. T, testis; K, kidney; C, cerebellum. Larger bands above the specific 250-nt CAT protected fragments (arrow) are products of incomplete digestion by RNase. B, schematic of the 116-bp minimal rat proenkephalin germ line promoter. Two regions that bind spermatogenic cell nuclear factors, GCP1 (1) and GCP2 (2), are shown together with the transcriptional initiation region (In). The minimal promoter within the RPKCAT0.5 sequence is indicated above by the bar. Exon 1T, exon 1 for the rat proenkephalin germ cell transcripts.
[View Larger Version of this Image (30K GIF file)]


Novel Spermatogenic Cell Nuclear Factors Specifically Interact with GCP1

The preceding findings indicated that nuclear protein interactions with GCP1 sequences are critical for activity of the proenkephalin germ line promoter during spermatogenesis. Sequence analysis did not reveal any consensus binding sites for known transcription factors within this region. However, the GCP1 sequence does contain three direct repeats (CTCCA/CG) (Fig. 7A) that could function to regulate transcription. The possible involvement of these repeated sequences in factor binding was, therefore, examined using gel-shift analysis. GCP1 sequences form a major and additional minor complexes of higher mobility with rat testis and germ cell nuclear extracts that are distinct from those formed with liver nuclear proteins (Fig. 7B). The relative abundance of the higher mobility complexes varied in different experiments (Fig. 7, C and D), and they may reflect probe heterogeneity and/or partial proteolysis in some experiments. Competition experiments demonstrated that binding of germ cell factors to GCP1 was DNA sequence-specific (Fig. 7C). This analysis also revealed that GCP1 sequences were not recognized by CRE-binding proteins present in spermatogenic cell nuclear extracts (Fig. 7C). Further, mutation of the three direct repeats abolished the binding of germ cell factors (Fig. 7D), suggesting the direct involvement of these elements in GCP1/protein interactions.


Fig. 7. Mobility shift assays using the GCP1 oligonucleotide probe. A, nucleotide sequences of wild-type (W.T.) and mutated (Mut123) GCP1 oligonucleotides. The repeat regions (R1, R2, and R3) and mutated sequences are underlined. B, gel-shift analysis of GCP1 binding to rat germ cell (G) and liver (L) nuclear extracts. Arrows indicate complexes detected in liver and germ cells. C, 10- and 25-fold excess of unlabeled GCP1, CRE, CRE-Mut, and nonspecific TATA-mut (NS) oligos were used as cold competitors of binding to rat germ cell extracts. C, control; P, probe alone. D, competition using a 10-fold excess of either wild-type GCP1, GCP1mut23 (M23), or GCPmut123 (M123) with rat germ cell extracts. C, control.
[View Larger Version of this Image (61K GIF file)]


Functional Analysis of GCP1

Additional studies of the repeat sequences were performed to determine their relative contributions to factor binding by GCP1. Mutation of the first repeat alone had a negligible effect on this interaction (Fig. 8). However, mutation of either the second or third repeat resulted in partial loss of binding activity (50-65% decrease) and mutation of both resulted in essentially complete disruption of the GCP1/protein complex (Fig. 8; see also Fig. 7D). Mutation of repeat 1 together with either the second or third repeat did not further reduce GCP1 binding activity, nor did deletion of the 5'-end of GCP1 including the first repeat (Fig. 8). Thus, the second and third repeats, which are identical in sequence and differ from the first repeat in the fifth position (A instead of C), are required for maximal binding of GCP1 to germ cell nuclear factors. It is possible that the first repeat is inactive because of this single base difference in the fifth position or because of its positioning within GCP1, or both.


Fig. 8. Role of individual repeat elements in germ cell factor binding to GCP1. Gel-shift assays were performed using rat germ cell nuclear extracts and the GCP1 probe in the presence and absence of varying concentrations of competitor DNAs. Mutations of the GCP1 repeat elements (R1, R2, or R3) are indicated by black boxes (open boxes signify wild type sequences). The amount of the specific GCP1 complex (shown by arrow in Fig. 7D) in each case was quantified by densitometry, and the binding activity of each competitor relative to GCP1 is indicated as follows: ++++, 90-100%; ++, 35-50%; -, <10%.
[View Larger Version of this Image (14K GIF file)]


To examine the functional significance of these in vitro protein interactions with GCP1 sequences, an additional transgene, containing mutations of the second and third repeats identical to those used in gel-shift analysis, was constructed and tested (GCP1mut; Fig. 1). Mutation of these two sites resulted in complete loss of transgene expression in mouse testis and did not produce abberant expression in somatic tissues (Fig. 9). These findings indicate that the interactions between spermatogenic cell nuclear proteins and the second and third repeat elements within GCP1 are critical for proenkephalin germ line promoter activity.


Fig. 9. Functional analysis of the GCP1 repeat elements. RNase protection analysis of tissues from GCP1mut transgenic mice was performed as described in Fig. 2. The arrows indicate the predicted position for the 250 nt CAT protection fragment. Total RNA from testes of RPKCAT0.45 transgenic mice was used as a positive control.
[View Larger Version of this Image (46K GIF file)]



DISCUSSION

Spermatogenic cells are characterized by the production of a large number of unique transcript forms often encoding ubiquitously expressed proteins (30). Some of these mRNAs encode germ cell-specific isoforms that may perform specialized functions during spermatogenesis (30). In other cases, novel untranslated sequences are present that may have specific regulatory roles, as in the translational control of testis-specific superoxide dismutase-1 mRNA involving its 5'-untranslated region (31). Transcription from an alternate spermatogenic cell-specific promoter is often involved in generation of these unique transcripts (32-34), as for rat and mouse proenkephalin. Alternate promoters expressed in somatic cells frequently function selectively during cell development and differentiation (35). It is therefore possible that some germ cell-specific transcripts reflect the necessity for an alternate promoter that provides appropriate cell-specific and developmentally regulated expression in the male germ line.

Previous work showed that the rat proenkephalin germ line promoter functions independently of the upstream somatic promoter and that a 500-bp region could mediate cell- and stage-specific expression (22). In this study, we have demonstrated that a 116-bp region encompassing the start site region is sufficient for appropriate transcription of the proenkephalin promoter in spermatocytes and spermatids and that a 51-bp 5'-flanking sequence (GCP1) is absolutely required for its activity. These data are consistent with the presence of one or more cell-specific transactivators in these spermatogenic stages that interact with specific elements in GCP1. At present, there is no evidence for the involvement of transcriptional repression in somatic cells since deletion analyses did not result in ectopic transgene expression in somatic tissues. However, a role for silencer elements located elsewhere within the minimal promoter region cannot be ruled out. This is relevant since repression has been implicated in spermatogenic cell transcription of the c-mos, lactate dehydrogenase c, Tctex-1, and histone H2b promoters (12, 36-38).

Recent studies have begun to characterize transcriptional regulation during spermatogenesis using transgenic, cell transfection and in vitro transcription approaches (7, 11, 15, 22, 32, 39-44). While numerous germ cell-specific promoters have been analyzed, much remains to be learned regarding the specific transcriptional mechanisms mediating stage-dependent gene expression in the male germ line. A critical role for CREM proteins and CRE-dependent promoter regulation in mediating spermiogenesis has been established (9, 10). However, additional mechanisms appear necessary for formation of meiotic and early postmeiotic germ cells. In fact, several testis-specific DNA binding proteins apparently unrelated to CREMs have been implicated in promoter regulation in these spermatogenic stages. For example, novel ets-domain-like proteins appear to regulate the pgk-2 promoter in mouse germ cells (39), and a testis-specific protein recognizes a novel palindromic regulatory element involved in testis-specific transcription of the mouse lactate dehydrogenase c gene (37). Both of these promoters function during spermatocyte and spermatid stages. In addition, testis-specific factors bind to regulatory elements (TE1 and TE2) unrelated to CREs within the spermatocyte-specific histone H1t promoter (45). However, the identities and functional roles for these and other putative germ cell-specific trans-factors, such as Tet-1 (46), are generally unknown.

The proenkephalin germ line promoter also becomes highly active in pachytene spermatocytes, and the 51-bp GCP1 regulatory region does not contain CRE-like sequences and is not responsive to CREMtau or CREB in transient co-transfection assays (data not shown). In addition, proenkephalin expression is not reduced in testes of CREM-deficient mice.2 Therefore, the proenkephalin gene is also regulated by CREM-independent mechanisms during spermatogenesis. Consistent with this, germ cell-specific nuclear proteins distinct from CREMtau and other CRE-binding proteins specifically bind to this region. Factor binding involves a novel sequence (CTCCAG) repeated twice within GCP1, and more importantly, mutation of these two elements abolishes proenkephalin germ line promoter activity. Interestingly, this repeat resembles sequences within the lactate dehydrogenase c palindromic element that are required for germ cell-specific transcription (CTCCTG) (37). In both promoters, these sequences are located just upstream of the start sites, and in both cases, germ cell proteins bind to DNA segments containing these elements. Similarly, the testis-specific element TE2 found within the rat histone H1t promoter also contains sequences resembling the GCP1 repeats (CCCCAG) (45). It is therefore possible that sequences related to the repeat elements within GCP1 may regulate multiple germ cell-specific promoters expressed during late meiosis. Repeated elements are often required for proper DNA binding or enhanced activation by trans-acting factors. For example, DNA binding by thyroid hormone requires appropriately spaced direct repeats (47), and Sp1 exhibits synergistic promoter activation when multiple binding sites are present (48).

Functional analysis of other sequences within the 116-bp promoter region is an important goal of future analyses. For example, initiation regions are often critical for activation of TATA-less promoters (49) although sequences resembling typical initiator (Inr) elements are not found within the start site region of the proenkephalin germ line promoter. In addition, another region is present within the defined minimal promoter (GCP2) that binds germ cell-specific nuclear factors and contains a palindromic GC-box (see Fig. 6B). Finally, characterization of the GCP1 binding proteins, their developmental expression, and potential interactions with other promoters active during late meiosis are clearly important to pursue.


FOOTNOTES

*   This work was supported by U. S. Public Health Service Grant DK36468 and National Research Service Award F32 DK09006. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    Present address: Dept. of Life Sciences, Virginia State University, Petersburg, VA 23806.
   To whom correspondence should be addressed. Tel.: 508-842-8921, ext. 134; Fax: 508-842-9632; E-mail: kilpatrick{at}sci.wfbr.edu.
1    The abbreviations used are: CREB, cyclic AMP response element binding protein; CRE, cyclic AMP response element; CREMtau , cyclic AMP response element modulator-tau ; CAT, chloramphenicol acetyltransferase; nt, nucleotide(s); bp, base pair(s).
2    J. Blendy and G. Shütz, unpublished observations.

Acknowledgments

We thank Donna Marshall for excellent technical assistance and Dr. Priti Raval for contributions to initial aspects in this work. We also thank Cathy Warren for invaluable help in preparing this manuscript.


REFERENCES

  1. Propst, F., Rosenberg, M. P., and Vande Woude, G. F. (1988) Trends Genet. 4, 183-187 [CrossRef][Medline] [Order article via Infotrieve]
  2. Torry, D. S. (1992) Am. J. Reprod. Immunol. 27, 167-170 [Medline] [Order article via Infotrieve]
  3. Waeber, G., Meyer, T. E., LeSieur, M., Hermann, H. L., Gerard, N., and Habener, J. F. (1991) Mol. Endocrinol. 5, 1418-1430 [Abstract]
  4. Delmas, V., van der Hoorn, F., Mellstrom, B., Jegou, B., and Sassone-Corsi, P. (1993) Mol. Endocrinol. 7, 1502-1514 [Abstract]
  5. Persengiev, S. P., Raval, P. J., Rabinovitch, S., Millette, C. F., and Kilpatrick, D. L. (1996) Endocrinology 137, 638-646 [Abstract]
  6. Delmas, V., and Sassone-Corsi, P. (1994) Mol. Cell. Endocrinol. 100, 121-124 [CrossRef][Medline] [Order article via Infotrieve]
  7. Kistler, M. K., Sassone-Corsi, P., and Kistler, W. S. (1994) Biol. Reprod. 51, 1322-1329 [Abstract]
  8. Zambrowicz, B. P., and Palmiter, R. D. (1994) Biol. Reprod. 50, 65-72 [Abstract]
  9. Nantel, F., Monaco, L., Foulkes, N. S., Masquiller, D., LeMeur, M., Henriksen, K., Dierich, A., Parvinen, M., and Sasson-Corsi, P. (1996) Nature 380, 159-162 [CrossRef][Medline] [Order article via Infotrieve]
  10. Blendy, J. A., Kaestner, K. H., Weinbauer, G. F., Nieschlag, E., and Schutz, G. (1996) Nature 380, 162-165 [CrossRef][Medline] [Order article via Infotrieve]
  11. van Wert, J. M., Wolfe, S. A., and Grimes, S. R. (1995) Biochemistry 34, 8733-8743 [Medline] [Order article via Infotrieve]
  12. Lim, K., and Chae, C.-B. (1992) J. Biol. Chem. 267, 15271-15273 [Abstract/Free Full Text]
  13. Thomas, K., Del Mazo, J., Eversole, P., Bellve, A., Hiraoka, Y., Li, S. L., and Simon, M. (1990) Development 109, 483-493 [Abstract]
  14. Iannello, R. C., Kola, I., and Dahl, H.-H. M. (1993) J. Biol. Chem. 268, 22581-22590 [Abstract/Free Full Text]
  15. Nayernia, K., Nieter, S., Kremling, H., Oberwinkler, H., and Engel, W. (1994) J. Biol. Chem. 269, 32181-32186 [Abstract/Free Full Text]
  16. Robinson, M. O., McCarrey, J. R., and Simon, M. I. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 8437-8441 [Abstract]
  17. Kilpatrick, D. L., and Millette, C. F. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 5015-5081 [Abstract]
  18. Kilpatrick, D. L., Borland, K., and Jin, D. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 5695-5699 [Abstract]
  19. Fabbri, A., Tsai-Morris, C. H., Luna, S., Fraioli, F., and Dufau, M. L. (1985) Endocrinology 117, 2544-2546 [Abstract]
  20. Morris, P. L., Vale, W. W., and Bardin, C. W. (1987) Biochem. Biophys. Res. Commun. 148, 1513-1519 [Medline] [Order article via Infotrieve]
  21. Kilpatrick, D. L., Zinn, S. A., Fitzgerald, M., Higuchi, H., Sabol, S. L., and Meyerhardt, J. (1990) Mol. Cell. Biol. 10, 3717-3726 [Medline] [Order article via Infotrieve]
  22. Galcheva-Gargova, Z., Tokeson, J. P., Karagyosov, L. K., Ebert, K. M., and Kilpatrick, D. L. (1993) Mol. Endocrinol. 7, 979-991 [Abstract]
  23. Hyman, S. E., Comb, M., Lin, Y.-S., Pearlberg, J., Green, M. R., and Goodman, H. M. (1988) Mol. Cell. Biol. 8, 4225-4233 [Medline] [Order article via Infotrieve]
  24. Walter, C. A., Nasr-Schirf, D., and Luna, V. J. (1989) BioTechniques 7, 1065-1070 [Medline] [Order article via Infotrieve]
  25. Zinn, S. A., Ebert, K. M., Mehta, N. D., Joshi, J., and Kilpatrick, D. L. (1991) J. Biol. Chem. 266, 23850-23855 [Abstract/Free Full Text]
  26. Chirgwin, J. M., Przbyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979) Biochemistry 18, 5294-5299 [Medline] [Order article via Infotrieve]
  27. Bellve, A. R., Millette, C. F., Bhatnagar, Y. M., and O'Brien, D. A. (1977) J. Histochem. Cytochem
  28. Cereghini, S., Raymondjean, M., Carranca, A. G., Herbomel, P., and Yaniv, M. (1987) Cell 50, 627-638 [Medline] [Order article via Infotrieve]
  29. Williams, M., Brys, A., Weiner, A. M., and Maizels, N. (1992) Nucleic Acids Res. 20, 4935-4936 [Medline] [Order article via Infotrieve]
  30. Wolgemuth, D., and Watrin, F. (1991) Mamm. Genome 1, 283-288 [Medline] [Order article via Infotrieve]
  31. Gu, W., and Hecht, N. B. (1996) Mol. Cell. Biol. 16, 4535-4543 [Abstract]
  32. Sun, Z., and Means, A. R. (1995) J. Biol. Chem. 270, 20962-20967 [Abstract/Free Full Text]
  33. Teruya, J. H., Kutsunai, S. Y., Spear, D. H., Edwards, P., and Clarke, C. F. (1990) Mol. Cell. Biol. 10, 2315-2326 [Medline] [Order article via Infotrieve]
  34. Howard, T. E., Shai, S.-Y., Langford, K. G., and Martin, B. M. (1990) Mol. Cell. Biol. 10, 4294-4302 [Medline] [Order article via Infotrieve]
  35. Schibler, U., and Sierra, F. (1987) Trends Genet. 21, 237-257
  36. Zinkel, S. S., Pal, S. K., Szeberenyi, J., and Cooper, G. M. (1992) Mol. Cell. Biol. 12, 2029-2036 [Abstract]
  37. Zhou, W., and Goldberg, E. (1996) Biol. Reprod. 54, 84-90 [Abstract]
  38. O'Neill, M. J., and Artzt, K. (1995) Development 121, 561-568 [Abstract/Free Full Text]
  39. Goto, M., Masamune, Y., and Nakanishi, Y. (1993) Nucleic Acids Res. 21, 209-214 [Abstract]
  40. van der Hoorn, F. A., and Tarnasky, H. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 703-707 [Abstract]
  41. Langford, K. G., Shai, S.-Y., Howard, T. E., Kovac, M. J., Overbeek, P. A., and Bernstein, K. E. (1991) J. Biol. Chem. 266, 15559-15562 [Abstract/Free Full Text]
  42. Zhou, W., Xu, J., and Goldberg, E. (1994) Biol. Reprod. 51, 425-432 [Abstract]
  43. Bunick, D., Johnson, P. A., Johnson, T. R., and Hecht, N. B. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 891-895 [Abstract]
  44. Zambrowicz, B. P., Harendza, C. J., Zimmermann, J. W., Brinster, R. L., and Palmiter, R. D. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 5071-5075 [Abstract]
  45. Wolfe, S. A., van Wert, J. M., and Grimes, S. R. (1995) Biochemistry 34, 12461-12469 [Medline] [Order article via Infotrieve]
  46. Tamura, T., Makino, Y., Mikoshiba, K., and Muramatsu, M. (1992) J. Biol. Chem. 267, 4327-4332 [Abstract/Free Full Text]
  47. Glass, C. K. (1994) Endocr. Rev. 15, 391-407 [Medline] [Order article via Infotrieve]
  48. Pascal, E., and Tjian, R. (1991) Genes & Dev. 5, 1646-1656 [Abstract]
  49. Smale, S., and Baltimore, D. (1989) Cell 57, 103-113 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.