- Email: [email protected]
The bovine protamine 2 gene: Evidence for alternative splicing
Biochimica et BiophysicaActa, 1132(1992) 133-139 © 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4781/92/$05.00
133
BBAEXP 92411
The bovine protamine 2 gene: evidence for alternative splicing Hannelore Kremling, Notker Reinhart, Manfred Schl6sser and Wolfgang Engel Institut fiir Humangenetikder Uniuersitiit, G6ttingen (Germany) (Received 20 march 1992)
Key words: Protamine 2; Alternative splicing; Nuclear protein Protamine 2 (PRM2) is a low molecular weight arginine-rich protein which is present in haploid spermatogenic cells of human and mouse. Although the bull PRM2 gene is translated and transcribed at low levels, the protein could not be detected. The gene was isolated from a cosmid library and was found to consist of two exons (298 and 50 bp, respectively) interrupted by an intron of 142 bp. As compared to the PRM2 genes of man, mouse and rat the bovine gene lacks a highly conserved sequence coding for the amino acids RLHRIH. Furthermore, primer extension experiments on bull PRM2 mRNA and sequencing of junction fragments revealed alternative splicing of mRNA resulting in two putative isoforms of the protein. The most abundant transcript is spliced at the conserved splice donor site found in exon 1 at position 236 giving rise to an in-frame deletion of 63 bp as compared to the cDNA sequence (Maier et al. (1990) Nucleic Acids Res. 18, 1249-1254). The less abundant longer mRNA was not detectable by radioactive primer extension. The corresponding cDNA was obtained by performing PCR with reverse transcribed bull testis RNA or with a spermatid specific cDNA library. Alternative splicing should result in an addition of 21 nonpolar amino acids in the derived polypeptide and an altered protein conformation and function.
Introduction During spermiogenesis in mammals transition proteins and protamines replace histones forming a compact chromatin structure in spermatozoa [2,3]. Whereas protamine 1 (PRM1) could be detected in all mammal species during elongation of spermatids, protamine 2 was found only in spermatogenic cells of human, mouse, hamster, stallion and monkey [4-8]. However, in the genomes of all species studied to date that lack protamine 2 in their spermatozoa, the PRM2 gene has been detected and was found to be transcribed [9-11]. For boar and bull, Maier et al. [1] demonstrated transcription and translation on low levels of PRM2 in testis, cDNA sequence analysis revealed various mutational events which should alter the primary structure of the protein. In boar a deletion of eight amino acids has removed the sequence motif R L H R I H from the amino-terminus of the molecule, while in bull numerous point mutations have accumulated neutral and hydrophobic amino acids in the motif which reduce the affinity of the protamine 2 to DNA. It was suggested that due to these mutational events the mature PRM2 is not functional and therefore absent from the sperm nucleus.
Correspondence to: W. Engel, Institut fiir Humangenetik, Gosslerstr. 12D, W-3400 G6ttingen, Germany.
We have isolated genomic clones for bull protamine 2 from a genomic library and compared the nucleotide sequence with those of human, mouse and rat as well as with the cDNA described by Maier et al. [1]. The mutational events in the sequence coding for the motif R L H R I H could be verified at the genomic level. At the level of mRNA, evidence for alternative splicing which results in two isoforms of the protein could be obtained.
Materials and Methods
Isolation and sequence analyses of genomic clones for bull protamine 2 A bovine cosmid library constructed in pcos 2 E M B L [12] was screened by replicating clones onto nitrocellulose filters [13] and hybridized with 32p multiprimed bull protamine 2 cDNA [1]. Hybridization and washing conditions were the same as described earlier [14]. The isolated clones were subcloned into pUC 18 and sequenced by the dideoxynucleotide-chain-termination method [15]. The D N A sequence was compared with the ' D N A Star' computer program 'Align' [16]. Primer extension and Northern blot analyses For primer extension an oligonucleotide primer, complementary to nucleotides 12 to 33 (B1, Fig. lb), was 5' end labeled with T4 polynucleotide kinase (Biolabs) and [y-aEp]ATP (3000 Ci/mmol, Amersham) and
134 Northern blot analysis was performed with 30 /xg bull testis total RNA and 15 /xg poly(A) + RNA, respectively. Hybridization was carried out with the 32p. labeled oligonucleotide probes B8 (complementary to nucleotides 11 to 58) and B9 (complementary to nucleotides 244 to 291). Hybridization conditions were the same as described earlier [18].
incubated with 10 ~g bull testis poly(A) + RNA or 30 fig total testis RNA at 85°C for 15 min, and then hybridized at 42°C for 2 h. The extension reaction was performed with reverse transcriptase (Superscript, BRL) as described earlier [17]. Primer extension, using primer B2 (Fig. lb) which is complementary to nucleotides 475 to 498, was performed as described above.
a
cDNA P
H
3'
5'
Gene
'
8 I
P I
E I
B I
8 |
5'
8m
A
I
I
i
p
ip
", '~,H
8
'r---1
l
~ 1
i_.J I
b
-730
:3'
E2
E1
-700
#
IOObp
-~)70
CTGCCTACCACGCTCT~CCCC, GCGGCi'GCTATGGTT~'GATTCTC~-,---,--'CATGC~~G~C~~ -640 -61o G~Ca~;GGCACA',-,-:-I-,~;GGGG^GGt,~G CGTGTG~ C C ~ G T C ~ G ~ , G A G G G C T ~ C , -55o -s2o -460
-.~se C A T A ~ ~ ~ -~9o
-430
-~00
~V,~CCI~C*C.~AC~XGCAT~,.AGAT~,aCCC'~..CaC*CaT~CCCC~TCC~CCCO~AGA~C~C~~ -370
-340
-.~10
GGCAGCAGGGAAGCTT~AGGAACCCC(~AGGA~AGGGGCA~GGTGTGA~&TCC~GACAAGGAG~YlT~~ -280
-250
-220
CAGAGG~TFLGTcCCT~TTT~GAGG~ATGGGGGAC~CGTTTTACC~GGTGAGGA~TGGCTGC~C~CGC~CTGTG~~G~ -19o
-16o
•
•
GGGGCC~.AGG'X~T,G G A C 2 I ~ T G G ' X ~ " ~ C C G -lO0
~
•
-J,3o ,
w
•
CC21~GTCACAG'I~,GGGGGCCCC~IT~'ATATA~GTCCCCC C~ ~ C ~ A C -70 -4o
AGACCC~ACCAACACT~TTAACAGCA~GAGCAGGGG~AGCCTCAGc~CCTcCCCG~CCCCCCAA~CCAGCTC~C~TCTCAGcCC~AGT I .......... -10 •
CGCCCA~CCGTGACCC~TCCATCACG :---::-i~--z-I
....
>
B3
..........
~.---->
+1 ~ T G GT(: ~
.......................
..-.8
< . . . . . . . . . . . . . B1 . . . . . . . . . . . . m ~ ~ ~ ~ R~T ~ ~ ~ ~ o
o
......
.-:= . . . . . . .
4u ~
•
•
~
~
"r-
iCJ~G
Net Val Arg Cy$ His Val Ly$ Ser Pro Thr G1u $er Pro Pro G1y Gin .80 ~oo
Gin GZy Ser Gly Gin Gin Gly G1u Thr Glu His Pro Asp Gin Ala Arg G1u Leu Arg Pro G1u Asp Ile 120
140
~60
].80
Pro VaJ Tyr GIF Arg Thr His Arg G1y Arg Tyr His Tyr Arg His Arg Set His Thr Arg Arg Arg Pro ~6
........
.6 .......
~-::::::J.---.4
.......
- ......
I
I :::::::::::::::::::::
Tyr Arg Arg Arg Arg Arg Arg Ala Cys Arg His Arg Arg Arg Arg Arg GZY Ala Ala GJ~ Pro Pro CYs --> . . . . . . . . . . . . .
.,
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
:1
• 330
300
Ala Pro I1e Pro GIv Thr Pro Gln Ala Ser Aro Gln GIv Set G •
•
360 •
•
•
390 •
o
•
420 •
cccaccctetcagacagccccctcatcc{gtgcccccaccccacccc9gcagcccccc~tccccacca~gttccctgg~tggccccgt~ac 450
470
< . . . . . . . . . . . B2 .............
tggaact{tttccaaaag GC T ~ OG~ ~IG & T G & G G A ~ & G G ~ & G G *G~ TGO &GA A ~ C J ~ C T C T ~ GCTTC ly Cys Arg Arg Her Arg Arg Arg Arg Arg Arg Cy$ Arg Arg Gin Leu *** --I . ~2o pso
CCTG~GCCCACCAG~%AGAGTCAC~TGCCAGGAA~CCCCTCGCAAGACCATACC~CATCCCC Fig. 1. (a) Diagram of bovine protamine 2 c D N A and gene. The coding portions of the m R N A are boxed, shadings indicate the alternative spliced area. Restriction sites: S = Styl, P ~ P s t I , E = E c o R I , B = B a m H I , Sm = S m a l , A = A c c l , H = H i n d I I l . (b) Nucleotide sequence of the bovine protamine 2 gene, including 5' and 3' flanking regions. The amino acid sequence is shown below the nucleotide sequence. The transcription start point is marked by a star at residue G at position - 1 2 5 , C A A T - and TATA-boxes are underlined. The alternative spliced sequence is double underlined. Primers used for primer extension, P C R and Northern blot analysis are shown above the sequence (B1-B9).
135
PCR-analysis of protamine 2 gene of different bovine breeds
Results
DNA was prepared from lymphocytes of six bovine breeds (Galloway, Holstein Friesian, Swiss Brown, Murnau Werdenfels cattle, Pustertaler Sprinzen, Simmental). Protamine 2 sequences were amplified by PCR according to Maier et al. [1] with the following modifications: first PCR round was performed for 30 cycles with 50 pmol primer B3 (positions - 8 to 13, Fig. lb) and 50 pmol primer B4 (complementary to positions 215 to 235). Cycling conditions were 1 min at 92°C, 45 s at 40°C and 1 min at 63°C. The second round was performed with 1 /xl of the first PCR product, 50 pmol primer B5 (positions 48 to 62) and 50 pmol primer B6 (complementary to positions 200 to 220). Cycling conditions were the same as for the first round except the annealing temperature was increased to 50°C. Single stranded PCR procucts were obtained by asymmetric PCR using 1 pmol primer B5 (Fig. lb), which contains the M13 reverse primer sequence at the 5' end, and 50 pmol primer B6 (Fig. lb). Automatic sequencing was performed with M13 reverse dye primer on an ABI model 373A sequencer according to the manufacturer's protocol (Applied Biosystems, USA).
Isolation and characterization of the bull protamine 2 gene
Isolation of protamine 2 cDNAs from a bull spermatid cDNA library and synthesis of cDNAs from bull testis RNA 1 /zl (104 pfu) of a bull spermatid cDNA library constructed in hgtll was incubated with 100/zl digestion buffer (100 mM Tris, pH 8, 10 mM EDTA, 100 mM NaCI, 0.1% SDS, 50 mM DTT, 0.5 ~g//zl proteinase K) over night at 37°C. After purification of the DNA, PCR was performed with 100 ng protamine specific primer B1 and 100 ng hgt11 specific primer. PCR conditions were 1 min at 94°C for denaturation, 2 min at 37°C for annealing and 3 min at 72°C for elongation with a time extension of 3 s. For the second round protamine 2 specific primers, B3 and B4 (complementary to positions 475 to 497) were used. Cycling conditions were 1 min at 94°C, 1.5 min at 55°C, 2 min at 72°C for 30 cycles. A third PCR round was performed with 1/zl of the PCR product from the second round, 50 pmol primer B7 (239 to 258) and 50 pmol primer B2 (complementary to 475 to 498). The PCR conditions were the same as for the second round except that the annealing temperature was raised to 60°C. To synthesize protamine 2 cDNA from bull testis RNA, 30/xg total testis RNA were incubated with 200 ng primer B2 and 200 U reverse transcriptase (Superscript, BRL). After alkaline hydrolysis, PCR was performed with the primers B2/B5 for the first round and B2/B7 for the second round. Sequencing of all PCR products were carried out on an ABI sequencer (Applied Biosystems, USA).
In a previous report we described the isolation of a cDNA clone for bull protamine 2 [1]. Using this cDNA clone as a hybridization probe, a bull genomic library was screened. A total of 4.105 cosmid clones were screened and one positive genomic clone was selected. Restriction fragments corresponding to the cDNA hybridizing regions of this genomic clone were subcloned and sequenced (Fig. la). Comparison of the nucleotide sequence with the cDNA revealed that the gene is composed of two exons. The nucleotide sequence of the bull protamine gene encompassing the two exons, plus 746 bp of the upstream and 67 bp of the downstream flanking sequences is shown in Fig. lb. The nucleotide sequence differs from that of the cDNA in three areas. At positions 178 to 183 six nucleotides coding for two arginines were present in the genomic DNA but absent in the cDNA. A change of a G to a C (position 184) results in a proline instead of an alanine. At positions 225 to 227 three additional nucleotides were identified which create an arginine and also change a serine to an arginine. In exon 2 of the genomic sequence an A (position 476) is present instead of a G in the cDNA, which results in an amino acid change from glycine to arginine. Some other nucleotide changes were identified in the 5' and 3' untranslated regions. The discrepancies between the cDNA and the genomic sequence can be explained by errors in the cDNA synthesis during construction of the library and by mistakes in sequence analysis. To confirm the sequence of the genomic clones the regions containing these discrepancies were amplified from genomic DNA by PCR and subsequently sequenced.
Characterization of the 5' flanking region of the bull protamine 2 gene To map the transcription initiation site of the bull protamine 2 gene, primer extension analysis was performed. The primer B1 (Fig. lb) was hybridized to testis RNA, and extended using reverse transcriptase. Two primer extension products, a major and a minor one, 159 and 157 nucleotides long, respectively, were obtained (Fig. 2a). The major 159 nucleotide fragment indicates that the 5' untranslated region is 125 bp in length. The transcription initiation site can be assigned to the G at position - 1 2 5 (Fig. lb). The minor fragment indicates either the presence of a second transcription initiation site, or incomplete RNA. A TATA-box sequence (TATATA) is located 22 bp upstream of the transcription initiation site (position -147). A CAAT-box can be found at position -191, 65 bp upstream of the transcription initiation site.
136
a GATC
1
2
b 3 GA
T C
Fig. 2. (a) Determination of the transcription start site of protamine 2. Primer extension analysis was performed with primer B1 and poly(A) + RNA (1) and total RNA (2) from bull testis. (b) Primer extension using primer B2 and total testis RNA. The length of the product indicates the abundance of the short transcript (3). Lanes G, A, T, C represent dideoxy reactions.
Alternatiue splicing of the bull protamine 2 The genes for protamine 2 have been isolated and sequenced from human [17], mouse [19], rat [10], boar [20] and bull. All of these genes contain one single intron, located towards the COOH-terminal end. A comparison of the exon-intron structure revealed the splice donor sequence C G C A G A G / g t to be conserved in the genes of human, mouse, rat and boar (Fig. 3). This conserved sequence can also be identified in the bull gene, but comparison of the gene with the cDNA described by Maier et al. [1] revealed, that the coding region of exon 1 is not spliced at this position. The
BULL
coding sequence continues with GT, and 63 bp downstream of the conserved donor sequence, the intron starts with the nucleotides gt. The alignment of the nucleotide sequence of exon 1 of bull and boar protamine 2 [20] revealed a similarity of about 84%, but this similarity is only apparent in the sequence from the translation start codon ATG to the sequence motif CGCAGAG (pos. 229 to 235, Fig. lb). The remaining 63 bp of exon 1 of the bull gene do not show any homology. Alignments of exon 1 of human, mouse and rat show the same result. To prove whether there is a second transcript of bull protamine 2, which is spliced at the conserved sequence CGCAGAG, we used PCR to amplify protamine 2 cDNA from a bull spermatid library. The cDNA library was identical to the one which Maier et al. [1] had used to isolate their cDNA clone. Using the pair of primers B2/B5 (Fig. lb) a PCR product of approx. 250 bp was obtained (Fig. 4). Sequence analysis and comparison of the PCR product with the genomic sequence unambiguously locates the splice donor site at the conserved sequence CGCAGAG (Fig. lb, pos. 229 to 235). This result indicates that the bull spermatid cDNA library contains two variants of protamine 2, which differ in their splice donor site. Although the primer pair B2/B5 should produce two cDNA variants, a short and a large one, only the short one was obtained. The detection of the large cDNA variant was achieved by using the primer pair B2/B7 (Fig. lb) which amplifies a fragment of 130 bp in length representing only the large cDNA variant. The PCR product contains the splice donor site 63 bp downstream of the conserved sequence CGCAGAG as shown by sequencing and comparison to the genomic sequence. To investigate whether one of the two cDNA variants is due to errors in the cDNA synthesis during the construction of the spermatid library, the same PCR analyses were
CGCAGAG/GT GGCTCAG/gt :::::::::::::::::::::::::::::::::::::::::
CGCAGAG/gt
ag/GCTGC
ag/GCTGC
BOAR
CGCAGAG/gt
ag/GCTGC
HUMAN
CGCAGAG/gt
ag/GCTGC
HOUSE
CGCAGAG/gt
ag/GCTGC
RAT
Fig. 3. Diagram of the protamine 2 exon-intron structures of bull, boar, human, mouse and rat. The black boxes represent the coding regions of the mRNA. Shadings illustrate the alternative spliced area in bull protamine 2. The nucleotides of the exon intron boundaries are shown, capitals indicate the nucleotides of the exons, small letters indicate the nucleotides of the introns.
137
1
2
3
4
5
W
4
--250 --130
B5--> B7-->
<--B2 <--B2
Fig. 4. Ethidium bromide stained 1.5% agarose gel containing PCR products of the region surrounding the alternativesplice sites amplified from cDNA prepared from testis RNA (2,3) and isolated from a spermatid library (4,5). (1) Size marker (kb-ladder, BRL), (2,4) PCR products amplifiedwith primers B2 and B5, representinga fragment of the short transcript. (3,5) PCR products amplifiedwith primers B2 and B7, representinga fragmentof the large transcript. The drawing gives the protamine 2 cDNA with the positions of the two different splice acceptor sites (235 and 298, respectively,see Fig. lb) and the positions of the primers used for PCR analysis. The shaded area indicates the alternativespliced region.
repeated on bull testis total RNA. Identical PCR products from both experiments were obtained (Fig. 4). Southern blot analysis with bovine genomic DNA digested with several restriction enzymes strongly suggests that bull protamine 2 is a single copy gene (data not shown). These results clearly demonstrate that bull protamine 2 is alternatively spliced. The short transcript consists of 235 bp of exon 1 and 50 bp of exon 2, which corresponds to the intron-exon splicing observed in other mammals (human, mouse, rat, boar) [10,17,19,20]. The larger transcript contains 63 additional bases at the 3' end of exon 1, exon 2 remains unchanged. The sequence a g / G C T G C at the splice acceptor site is conserved in all five mammmals (Fig. 3). All splice junction sequences conform to the G T / A G rule [21]. The intron splits the codon for glycine (Fig. lb, pos. 298) at the first nucleotide position in both m R N A variants. To determine which of the two cDNAs represents the most abundant transcript in the bull testis, Northern blot analysis was performed. Total and poly(A) + RNA were hybridized with the oligonucleotides B8 and B9 (Fig. lb). With the oligonucleotide B8, which should hybridize with both transcripts, a hybridization signal at approx. 900 bp was obtained. Whereas the oligonucleotide B9, which is specific for the large transcript, did not produce a hybridization signal (data not shown).
A further indication that the short transcript is the most abundant in bull testis was obtained from primer extension analysis. As shown before (Figs. lb and 2a) the transcription initiation site of bull protamine 2 gene is located 125 bp upstream of ATG. Using primer B2 (exon 2, Fig. lb), a major extension product of 416 bases was obtained (Fig. 2b) which corresponds with the length of the short transcript. Even after 4 days of exposure no signal of the large transcript was detected. Therefore, the short transcript represents the most abundant variant of the two protamine 2 transcripts. The large one is present at a very low level and can be detected only by the very sensitive method of PCR. We estimate that the relative abundance of the large and the short transcript is less than 1 : 100.
Conservation of bull protamine 2 among bovine breeds Maier et al. [1] described the cDNA sequence of bull protamine 2. They demonstrated a derived amino acid motif R L H R I H which is highly conserved in mouse and human but diverges in bull. Although sequence analysis of genomic clones revealed two additional codons for arginine (pos. 178-183) and the change of an alanine to proline (pos. 184-186) the derived amino acid motif R L H R I H is not present in bull protamine 2. To confirm the nucleotide sequence in this region and to examine whether the underlying mutation is present in the protamine 2 genes of different bovine breeds (Gallaway, Holstein Friesian, Murnau Werdenfels cattle, Pustertaler Sprinzen, Swiss Brown, Simmental), PCR and direct sequence analysis were performed. It could be observed, that all examined bovine breeds do not contain the nucleotide sequence coding for the amino acid motif RLHRIH. Discussion
How to explain the lack of protamine 2 in bull spermiogenesis Protamine 2 has been found in spermatozoa of mouse and human but not in bull, boar and rat [4,7,22]. The respective transcripts were identified in these species but they were less abundant than in mouse and human [1,9,23]. The comparison of the exon-intron structure revealed no differences between species that possess protamine 2 and those that have only a reduced amount of m R N A and lack the protein. The coding sequences are split by one single intron and the proportions of exon 1 and 2 are comparable between the species (Fig. 3). Therefore, no information is gained about the differences in protamine 2 gene expression between mouse, human, bull, boar and rat, respectively, when the exon-intron structure is compared. The protamine genes of human, mouse and bull are closely linked in the genome [14,17,24] and might be controlled by similar trans-acting factors. To identify
138 elements possibly responsible for the reduced expression of the bull protamine 2 gene, the 5' flanking region of bull protamine 2 was compared with those of bull protamine 1 [25], mouse and human protamine 2 [17,19]. In the two protamine genes of the mouse as well as in the genes of human several common sequence motifs arranged in clusters have been found [17,19]. The 5' flanking regions of bull protamine 1 and 2 genes lack those common sequence motifs. Therefore, it can be suggested that the reduced expression of the bull protamine 2 gene is due to the lack of cis-acting elements in the 5' region. The results of in vitro transcription and gel retardation assays with mouse protamine 2 gene indicate that the region 140 bp to 23 bp upstream of the transcription start point is sufficient for maximal transcription. In this region three motifs (E,B, #8) have been found which are in common with mouse protamine 1 [26,27]. The motif #8 which was proposed to be a modified SPl-binding site, is also present in the bull protamine 2 but the nucleotide sequence of the motifs E and B are altered. Thus, it can be suggested that these changes in the 5' untranslated sequence have a negative effect on the transcription rate of the protamine 2 gene. However, the comparison between the 5' untranslated region of mouse and rat protamine 2 let us suggest that also other motifs than E, B and #8 are included in protamine 2 gene transcription. Protamine 2 of the rat was found to be transcribed with only 30% of the efficiency of mouse protamine 2 [23], although the three motifs are identical in mouse and rat. Another important difference between the gene for bull protamine 2 and the genes for protamines of other species are observed in the nucleotide sequence of the polyadenylation signal. The consensus sequence AAUAAA is changed to AAUAAG. This mutation could be responsible for the altered transcription rate of bull protamine 2 gene, as well. An identical mutation in the polyadenylation signal of the human a2globin gene resulted in a reduction of the transcription rate to 5% [28]. The reason, why bull protamine 2 is poorly transcribed in contrast to mouse protamine 2 or bull protamine 1 can not be answered by sequence comparisons of the 5' flanking regions. Protein binding studies and in vitro mutagenesis experiments should help to obtain further information about the expression of bull protamine 2. An additional cause for the absence of protamine 2 in bull spermiogenesis could be found at the translational level. Kwon and Hecht [29] demonstrated by gel retardation assays that two sequence motifs in the 3' untranslated region of mouse protamine 2 form specific RNA-protein complexes. These motifs (Y,H) are conserved in mammalian protamine and transition protein genes for which translational control is known. The 3' untranslated region of bull protamine 2 cDNA
[1] was found to contain only one of these motifs (H). If both elements are necessary for translational control, the lack of element Y could be responsible for a disturbed mRNA translation in vivo. However, Maier et al. [1] have shown that bull protamine 2 mRNA can be translated in vitro. The lack of protamine 2 in bull spermatozoa could be due to a disturbed protein function. Maier et al. [1] compared the cDNA sequence of bull and boar protamine 2 with that of mouse and human. A mutation was found in bull and boar which result in the absence of an amino acid motif RLHRIH. We have shown that this mutation is not only present at the cDNA level but also at the genomic level, including the genomic DNA of six bovine breeds. However, this motif is conserved in human and mouse in whose spermatozoa protamine 2 is present. Therefore, the assumption of Maier et al. [1] that the motif RLHRIH is of functional importance for DNA-protein interaction is supported.
Alternative splicing in the bull protamine 2 transcript Bull protamine 2 mRNA exists in two different variants in the bull testis (Fig. 4). The large transcript, 63 bp longer at the 3' end of exon 1, is present at a very low level, only detectable by PCR. Maier et al. [1] have isolated this less abundant cDNA variant from a bull spermatid library. To exclude the possibility that this large cDNA is an artifact of the library, PCR analysis was performed on total testis RNA. Both transcripts were detected and verified by sequencing. These results confirm that both transcripts are the products of alternative splicing. The most abundant short form has a splice donor site which is comparable with those of protamine 2 transcripts of other species. The two mRNA variants of the bull protamine 2 differ in 63 bp at the 3' end of exon 1, but these additional nucleotides do not result in a reading frame shift. However, the additional amino acids could lead to an altered protein conformation and function. Protamines are rich in arginines which are clustered. Due to the insertion of the 21 neutral and acidic amino acids a region in the bull protamine 2 protein which might be important for DNA-protein interaction is altered. It remains unclear if both transcripts for bull protamine 2 are translated and have different binding activities to DNA.
Acknowledgements We thank Prof. Dr. W. Holtz for providing the blood of the bovine breeds, Stephanie Bunkowski for technical assistance, Derek Murphy and Frauke Rininsland for reading the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (En 84/19-2).
139 References 1 Maier, W.M., Nussbaum, G., Domenjoud, L., Klemm, U. and Engel, W. (1990) Nucleic Acids Res. 18, 1249-1254. 2 Grimes, S.R., Meistrich, M.L., Platz, R.D. and Hnilica, L.S. (1977) Exp. Cell Res. 110, 31-39. 3 Hecht, N.B. (1989) Molecular Biology of Chromosomal Function (Adolph, K.W., ed.), pp. 396-420, Springer, New York. 4 McKay, D.J., Renaux, B.S. and Dixon, G.H. (1986) Eur. J. Biochem. 156, 5-8. 5 Belaiche, D., Loir, M., Kruggle, W. and Sautiere, P. (1987) Biochim. Biophys. Acta 913, 145-149. 6 Corzett, M., Blacher, R., Mazrimas, J.A. and Balhorn, R. (1987) J. Cell Biol. 105, 152a. 7 Bellv~, A.R., McKay, D.J., Renaux, B.S. and Dixon, G.H. (1988) Biochemistry 27, 2890-2897. 8 Balhorn, R. (1989) Molecular Biology of Chromosomal Function (Adolph, K.W., ed.), pp. 366-395, Springer, New York. 9 Bower, P., Yelick, P.C. and Hecht, N.B. (1987) Biol. Reprod. 37, 479-488. 10 Tanhauser, S.M. and Hecht, N.B. (1989) Nucleic Acids Res. 17, 4395. 11 Johnson, P.A., Yelick, P.C., Liem, H. and Hecht, N.B. (1988) Gamete Res. 19, 169-175. 12 Plenz, G., Prokop, C.M., Kratzberg, T. and Geldermann, H. (1988) J. Anim. Breed. Genet. 105, 389-399. 13 Hanahan, D. and Meselson, M. (1980) Gene 10, 63-67. 14 Reinhart, N., Kremling, H., Luerssen, H., Adham, I.M. and Engel, W. (1991) Biol. Chem. Hoppe-Seyler 372, 431-436. 15 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.
16 Wilbur, W.J. and Lipman, D.J. (1982) Proc. Natl. Acad. Sci. USA 80, 726-730. 17 Domenjoud, L., Nussbaum, G., Adham, I.M., Greeske, G. and Engel, W. (1990) Genomics 8, 127-133. 18 Kremling, H., Luerssen, H., Adham, I.M., Klemm, U., Tsaousidou, S. (1989) Differentiation 40, 184-190. 19 Johnson, P.A., Pechorn, J.J., Yelick, P.C., Palmiter, R.D. and Hecht, N.B. (1988) Biochim. Biophys. Acta 950, 45-53. 20 Keime, S., Heidland, K., Kumm, S., Schl6sser, M., Hroch, N., Holtz, W. and Engel, W. (1992) Biol. Chem. Hoppe-Seyler, in press. 21 Mount, S.M. (1982) Nucleic Acids Res. 10, 459-472. 22 Balhorn, R., Weston, S., Thomas, C. and Wyrobek, A.J. (1984) Exp. Cell Res. 150, 298-308. 23 Bunick, D., Balhorn, R., Stanker, L.H. and Hecht, N.B. (1990) Exp. Cell Res. 188, 147-152. 24 Hecht, N.B., Kleene, K.C., Yelick, T.C., Johnson, P.A., Pravtcheva, D.D. and Ruddle, F.H. (1986) Som. Cell Mol. Gen. 12, 191-196. 25 Krawetz, S.A., Connor, W. and Dixon, G.H. (1988) J. Biol. Chem. 263, 321-326. 26 Bunick, D., Johnson, P.A., Johnson, T.R. and Hecht, N.B. (1990) Proc. Natl. Acad. Sci USA 87, 891-895. 27 Johnson, P.A., Bunick, D. and Hecht, N.B. (1991) Biol. Reprod. 44, 127-134. 28 Higgs, D.R., Goodbourn, S.E.Y., Lamb, J., Clegg, J.B., Weatherall, D.J. and Proudfoot, N.J. (1983) Nature 306, 398-400. 29 Kwon, Y.K. and Hecht, N.B. (1991) Proc. Natl. Acad. Sci. USA 88, 3584-3588.
133
BBAEXP 92411
The bovine protamine 2 gene: evidence for alternative splicing Hannelore Kremling, Notker Reinhart, Manfred Schl6sser and Wolfgang Engel Institut fiir Humangenetikder Uniuersitiit, G6ttingen (Germany) (Received 20 march 1992)
Key words: Protamine 2; Alternative splicing; Nuclear protein Protamine 2 (PRM2) is a low molecular weight arginine-rich protein which is present in haploid spermatogenic cells of human and mouse. Although the bull PRM2 gene is translated and transcribed at low levels, the protein could not be detected. The gene was isolated from a cosmid library and was found to consist of two exons (298 and 50 bp, respectively) interrupted by an intron of 142 bp. As compared to the PRM2 genes of man, mouse and rat the bovine gene lacks a highly conserved sequence coding for the amino acids RLHRIH. Furthermore, primer extension experiments on bull PRM2 mRNA and sequencing of junction fragments revealed alternative splicing of mRNA resulting in two putative isoforms of the protein. The most abundant transcript is spliced at the conserved splice donor site found in exon 1 at position 236 giving rise to an in-frame deletion of 63 bp as compared to the cDNA sequence (Maier et al. (1990) Nucleic Acids Res. 18, 1249-1254). The less abundant longer mRNA was not detectable by radioactive primer extension. The corresponding cDNA was obtained by performing PCR with reverse transcribed bull testis RNA or with a spermatid specific cDNA library. Alternative splicing should result in an addition of 21 nonpolar amino acids in the derived polypeptide and an altered protein conformation and function.
Introduction During spermiogenesis in mammals transition proteins and protamines replace histones forming a compact chromatin structure in spermatozoa [2,3]. Whereas protamine 1 (PRM1) could be detected in all mammal species during elongation of spermatids, protamine 2 was found only in spermatogenic cells of human, mouse, hamster, stallion and monkey [4-8]. However, in the genomes of all species studied to date that lack protamine 2 in their spermatozoa, the PRM2 gene has been detected and was found to be transcribed [9-11]. For boar and bull, Maier et al. [1] demonstrated transcription and translation on low levels of PRM2 in testis, cDNA sequence analysis revealed various mutational events which should alter the primary structure of the protein. In boar a deletion of eight amino acids has removed the sequence motif R L H R I H from the amino-terminus of the molecule, while in bull numerous point mutations have accumulated neutral and hydrophobic amino acids in the motif which reduce the affinity of the protamine 2 to DNA. It was suggested that due to these mutational events the mature PRM2 is not functional and therefore absent from the sperm nucleus.
Correspondence to: W. Engel, Institut fiir Humangenetik, Gosslerstr. 12D, W-3400 G6ttingen, Germany.
We have isolated genomic clones for bull protamine 2 from a genomic library and compared the nucleotide sequence with those of human, mouse and rat as well as with the cDNA described by Maier et al. [1]. The mutational events in the sequence coding for the motif R L H R I H could be verified at the genomic level. At the level of mRNA, evidence for alternative splicing which results in two isoforms of the protein could be obtained.
Materials and Methods
Isolation and sequence analyses of genomic clones for bull protamine 2 A bovine cosmid library constructed in pcos 2 E M B L [12] was screened by replicating clones onto nitrocellulose filters [13] and hybridized with 32p multiprimed bull protamine 2 cDNA [1]. Hybridization and washing conditions were the same as described earlier [14]. The isolated clones were subcloned into pUC 18 and sequenced by the dideoxynucleotide-chain-termination method [15]. The D N A sequence was compared with the ' D N A Star' computer program 'Align' [16]. Primer extension and Northern blot analyses For primer extension an oligonucleotide primer, complementary to nucleotides 12 to 33 (B1, Fig. lb), was 5' end labeled with T4 polynucleotide kinase (Biolabs) and [y-aEp]ATP (3000 Ci/mmol, Amersham) and
134 Northern blot analysis was performed with 30 /xg bull testis total RNA and 15 /xg poly(A) + RNA, respectively. Hybridization was carried out with the 32p. labeled oligonucleotide probes B8 (complementary to nucleotides 11 to 58) and B9 (complementary to nucleotides 244 to 291). Hybridization conditions were the same as described earlier [18].
incubated with 10 ~g bull testis poly(A) + RNA or 30 fig total testis RNA at 85°C for 15 min, and then hybridized at 42°C for 2 h. The extension reaction was performed with reverse transcriptase (Superscript, BRL) as described earlier [17]. Primer extension, using primer B2 (Fig. lb) which is complementary to nucleotides 475 to 498, was performed as described above.
a
cDNA P
H
3'
5'
Gene
'
8 I
P I
E I
B I
8 |
5'
8m
A
I
I
i
p
ip
", '~,H
8
'r---1
l
~ 1
i_.J I
b
-730
:3'
E2
E1
-700
#
IOObp
-~)70
CTGCCTACCACGCTCT~CCCC, GCGGCi'GCTATGGTT~'GATTCTC~-,---,--'CATGC~~G~C~~ -640 -61o G~Ca~;GGCACA',-,-:-I-,~;GGGG^GGt,~G CGTGTG~ C C ~ G T C ~ G ~ , G A G G G C T ~ C , -55o -s2o -460
-.~se C A T A ~ ~ ~ -~9o
-430
-~00
~V,~CCI~C*C.~AC~XGCAT~,.AGAT~,aCCC'~..CaC*CaT~CCCC~TCC~CCCO~AGA~C~C~~ -370
-340
-.~10
GGCAGCAGGGAAGCTT~AGGAACCCC(~AGGA~AGGGGCA~GGTGTGA~&TCC~GACAAGGAG~YlT~~ -280
-250
-220
CAGAGG~TFLGTcCCT~TTT~GAGG~ATGGGGGAC~CGTTTTACC~GGTGAGGA~TGGCTGC~C~CGC~CTGTG~~G~ -19o
-16o
•
•
GGGGCC~.AGG'X~T,G G A C 2 I ~ T G G ' X ~ " ~ C C G -lO0
~
•
-J,3o ,
w
•
CC21~GTCACAG'I~,GGGGGCCCC~IT~'ATATA~GTCCCCC C~ ~ C ~ A C -70 -4o
AGACCC~ACCAACACT~TTAACAGCA~GAGCAGGGG~AGCCTCAGc~CCTcCCCG~CCCCCCAA~CCAGCTC~C~TCTCAGcCC~AGT I .......... -10 •
CGCCCA~CCGTGACCC~TCCATCACG :---::-i~--z-I
....
>
B3
..........
~.---->
+1 ~ T G GT(: ~
.......................
..-.8
< . . . . . . . . . . . . . B1 . . . . . . . . . . . . m ~ ~ ~ ~ R~T ~ ~ ~ ~ o
o
......
.-:= . . . . . . .
4u ~
•
•
~
~
"r-
iCJ~G
Net Val Arg Cy$ His Val Ly$ Ser Pro Thr G1u $er Pro Pro G1y Gin .80 ~oo
Gin GZy Ser Gly Gin Gin Gly G1u Thr Glu His Pro Asp Gin Ala Arg G1u Leu Arg Pro G1u Asp Ile 120
140
~60
].80
Pro VaJ Tyr GIF Arg Thr His Arg G1y Arg Tyr His Tyr Arg His Arg Set His Thr Arg Arg Arg Pro ~6
........
.6 .......
~-::::::J.---.4
.......
- ......
I
I :::::::::::::::::::::
Tyr Arg Arg Arg Arg Arg Arg Ala Cys Arg His Arg Arg Arg Arg Arg GZY Ala Ala GJ~ Pro Pro CYs --> . . . . . . . . . . . . .
.,
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
:1
• 330
300
Ala Pro I1e Pro GIv Thr Pro Gln Ala Ser Aro Gln GIv Set G •
•
360 •
•
•
390 •
o
•
420 •
cccaccctetcagacagccccctcatcc{gtgcccccaccccacccc9gcagcccccc~tccccacca~gttccctgg~tggccccgt~ac 450
470
< . . . . . . . . . . . B2 .............
tggaact{tttccaaaag GC T ~ OG~ ~IG & T G & G G A ~ & G G ~ & G G *G~ TGO &GA A ~ C J ~ C T C T ~ GCTTC ly Cys Arg Arg Her Arg Arg Arg Arg Arg Arg Cy$ Arg Arg Gin Leu *** --I . ~2o pso
CCTG~GCCCACCAG~%AGAGTCAC~TGCCAGGAA~CCCCTCGCAAGACCATACC~CATCCCC Fig. 1. (a) Diagram of bovine protamine 2 c D N A and gene. The coding portions of the m R N A are boxed, shadings indicate the alternative spliced area. Restriction sites: S = Styl, P ~ P s t I , E = E c o R I , B = B a m H I , Sm = S m a l , A = A c c l , H = H i n d I I l . (b) Nucleotide sequence of the bovine protamine 2 gene, including 5' and 3' flanking regions. The amino acid sequence is shown below the nucleotide sequence. The transcription start point is marked by a star at residue G at position - 1 2 5 , C A A T - and TATA-boxes are underlined. The alternative spliced sequence is double underlined. Primers used for primer extension, P C R and Northern blot analysis are shown above the sequence (B1-B9).
135
PCR-analysis of protamine 2 gene of different bovine breeds
Results
DNA was prepared from lymphocytes of six bovine breeds (Galloway, Holstein Friesian, Swiss Brown, Murnau Werdenfels cattle, Pustertaler Sprinzen, Simmental). Protamine 2 sequences were amplified by PCR according to Maier et al. [1] with the following modifications: first PCR round was performed for 30 cycles with 50 pmol primer B3 (positions - 8 to 13, Fig. lb) and 50 pmol primer B4 (complementary to positions 215 to 235). Cycling conditions were 1 min at 92°C, 45 s at 40°C and 1 min at 63°C. The second round was performed with 1 /xl of the first PCR product, 50 pmol primer B5 (positions 48 to 62) and 50 pmol primer B6 (complementary to positions 200 to 220). Cycling conditions were the same as for the first round except the annealing temperature was increased to 50°C. Single stranded PCR procucts were obtained by asymmetric PCR using 1 pmol primer B5 (Fig. lb), which contains the M13 reverse primer sequence at the 5' end, and 50 pmol primer B6 (Fig. lb). Automatic sequencing was performed with M13 reverse dye primer on an ABI model 373A sequencer according to the manufacturer's protocol (Applied Biosystems, USA).
Isolation and characterization of the bull protamine 2 gene
Isolation of protamine 2 cDNAs from a bull spermatid cDNA library and synthesis of cDNAs from bull testis RNA 1 /zl (104 pfu) of a bull spermatid cDNA library constructed in hgtll was incubated with 100/zl digestion buffer (100 mM Tris, pH 8, 10 mM EDTA, 100 mM NaCI, 0.1% SDS, 50 mM DTT, 0.5 ~g//zl proteinase K) over night at 37°C. After purification of the DNA, PCR was performed with 100 ng protamine specific primer B1 and 100 ng hgt11 specific primer. PCR conditions were 1 min at 94°C for denaturation, 2 min at 37°C for annealing and 3 min at 72°C for elongation with a time extension of 3 s. For the second round protamine 2 specific primers, B3 and B4 (complementary to positions 475 to 497) were used. Cycling conditions were 1 min at 94°C, 1.5 min at 55°C, 2 min at 72°C for 30 cycles. A third PCR round was performed with 1/zl of the PCR product from the second round, 50 pmol primer B7 (239 to 258) and 50 pmol primer B2 (complementary to 475 to 498). The PCR conditions were the same as for the second round except that the annealing temperature was raised to 60°C. To synthesize protamine 2 cDNA from bull testis RNA, 30/xg total testis RNA were incubated with 200 ng primer B2 and 200 U reverse transcriptase (Superscript, BRL). After alkaline hydrolysis, PCR was performed with the primers B2/B5 for the first round and B2/B7 for the second round. Sequencing of all PCR products were carried out on an ABI sequencer (Applied Biosystems, USA).
In a previous report we described the isolation of a cDNA clone for bull protamine 2 [1]. Using this cDNA clone as a hybridization probe, a bull genomic library was screened. A total of 4.105 cosmid clones were screened and one positive genomic clone was selected. Restriction fragments corresponding to the cDNA hybridizing regions of this genomic clone were subcloned and sequenced (Fig. la). Comparison of the nucleotide sequence with the cDNA revealed that the gene is composed of two exons. The nucleotide sequence of the bull protamine gene encompassing the two exons, plus 746 bp of the upstream and 67 bp of the downstream flanking sequences is shown in Fig. lb. The nucleotide sequence differs from that of the cDNA in three areas. At positions 178 to 183 six nucleotides coding for two arginines were present in the genomic DNA but absent in the cDNA. A change of a G to a C (position 184) results in a proline instead of an alanine. At positions 225 to 227 three additional nucleotides were identified which create an arginine and also change a serine to an arginine. In exon 2 of the genomic sequence an A (position 476) is present instead of a G in the cDNA, which results in an amino acid change from glycine to arginine. Some other nucleotide changes were identified in the 5' and 3' untranslated regions. The discrepancies between the cDNA and the genomic sequence can be explained by errors in the cDNA synthesis during construction of the library and by mistakes in sequence analysis. To confirm the sequence of the genomic clones the regions containing these discrepancies were amplified from genomic DNA by PCR and subsequently sequenced.
Characterization of the 5' flanking region of the bull protamine 2 gene To map the transcription initiation site of the bull protamine 2 gene, primer extension analysis was performed. The primer B1 (Fig. lb) was hybridized to testis RNA, and extended using reverse transcriptase. Two primer extension products, a major and a minor one, 159 and 157 nucleotides long, respectively, were obtained (Fig. 2a). The major 159 nucleotide fragment indicates that the 5' untranslated region is 125 bp in length. The transcription initiation site can be assigned to the G at position - 1 2 5 (Fig. lb). The minor fragment indicates either the presence of a second transcription initiation site, or incomplete RNA. A TATA-box sequence (TATATA) is located 22 bp upstream of the transcription initiation site (position -147). A CAAT-box can be found at position -191, 65 bp upstream of the transcription initiation site.
136
a GATC
1
2
b 3 GA
T C
Fig. 2. (a) Determination of the transcription start site of protamine 2. Primer extension analysis was performed with primer B1 and poly(A) + RNA (1) and total RNA (2) from bull testis. (b) Primer extension using primer B2 and total testis RNA. The length of the product indicates the abundance of the short transcript (3). Lanes G, A, T, C represent dideoxy reactions.
Alternatiue splicing of the bull protamine 2 The genes for protamine 2 have been isolated and sequenced from human [17], mouse [19], rat [10], boar [20] and bull. All of these genes contain one single intron, located towards the COOH-terminal end. A comparison of the exon-intron structure revealed the splice donor sequence C G C A G A G / g t to be conserved in the genes of human, mouse, rat and boar (Fig. 3). This conserved sequence can also be identified in the bull gene, but comparison of the gene with the cDNA described by Maier et al. [1] revealed, that the coding region of exon 1 is not spliced at this position. The
BULL
coding sequence continues with GT, and 63 bp downstream of the conserved donor sequence, the intron starts with the nucleotides gt. The alignment of the nucleotide sequence of exon 1 of bull and boar protamine 2 [20] revealed a similarity of about 84%, but this similarity is only apparent in the sequence from the translation start codon ATG to the sequence motif CGCAGAG (pos. 229 to 235, Fig. lb). The remaining 63 bp of exon 1 of the bull gene do not show any homology. Alignments of exon 1 of human, mouse and rat show the same result. To prove whether there is a second transcript of bull protamine 2, which is spliced at the conserved sequence CGCAGAG, we used PCR to amplify protamine 2 cDNA from a bull spermatid library. The cDNA library was identical to the one which Maier et al. [1] had used to isolate their cDNA clone. Using the pair of primers B2/B5 (Fig. lb) a PCR product of approx. 250 bp was obtained (Fig. 4). Sequence analysis and comparison of the PCR product with the genomic sequence unambiguously locates the splice donor site at the conserved sequence CGCAGAG (Fig. lb, pos. 229 to 235). This result indicates that the bull spermatid cDNA library contains two variants of protamine 2, which differ in their splice donor site. Although the primer pair B2/B5 should produce two cDNA variants, a short and a large one, only the short one was obtained. The detection of the large cDNA variant was achieved by using the primer pair B2/B7 (Fig. lb) which amplifies a fragment of 130 bp in length representing only the large cDNA variant. The PCR product contains the splice donor site 63 bp downstream of the conserved sequence CGCAGAG as shown by sequencing and comparison to the genomic sequence. To investigate whether one of the two cDNA variants is due to errors in the cDNA synthesis during the construction of the spermatid library, the same PCR analyses were
CGCAGAG/GT GGCTCAG/gt :::::::::::::::::::::::::::::::::::::::::
CGCAGAG/gt
ag/GCTGC
ag/GCTGC
BOAR
CGCAGAG/gt
ag/GCTGC
HUMAN
CGCAGAG/gt
ag/GCTGC
HOUSE
CGCAGAG/gt
ag/GCTGC
RAT
Fig. 3. Diagram of the protamine 2 exon-intron structures of bull, boar, human, mouse and rat. The black boxes represent the coding regions of the mRNA. Shadings illustrate the alternative spliced area in bull protamine 2. The nucleotides of the exon intron boundaries are shown, capitals indicate the nucleotides of the exons, small letters indicate the nucleotides of the introns.
137
1
2
3
4
5
W
4
--250 --130
B5--> B7-->
<--B2 <--B2
Fig. 4. Ethidium bromide stained 1.5% agarose gel containing PCR products of the region surrounding the alternativesplice sites amplified from cDNA prepared from testis RNA (2,3) and isolated from a spermatid library (4,5). (1) Size marker (kb-ladder, BRL), (2,4) PCR products amplifiedwith primers B2 and B5, representinga fragment of the short transcript. (3,5) PCR products amplifiedwith primers B2 and B7, representinga fragmentof the large transcript. The drawing gives the protamine 2 cDNA with the positions of the two different splice acceptor sites (235 and 298, respectively,see Fig. lb) and the positions of the primers used for PCR analysis. The shaded area indicates the alternativespliced region.
repeated on bull testis total RNA. Identical PCR products from both experiments were obtained (Fig. 4). Southern blot analysis with bovine genomic DNA digested with several restriction enzymes strongly suggests that bull protamine 2 is a single copy gene (data not shown). These results clearly demonstrate that bull protamine 2 is alternatively spliced. The short transcript consists of 235 bp of exon 1 and 50 bp of exon 2, which corresponds to the intron-exon splicing observed in other mammals (human, mouse, rat, boar) [10,17,19,20]. The larger transcript contains 63 additional bases at the 3' end of exon 1, exon 2 remains unchanged. The sequence a g / G C T G C at the splice acceptor site is conserved in all five mammmals (Fig. 3). All splice junction sequences conform to the G T / A G rule [21]. The intron splits the codon for glycine (Fig. lb, pos. 298) at the first nucleotide position in both m R N A variants. To determine which of the two cDNAs represents the most abundant transcript in the bull testis, Northern blot analysis was performed. Total and poly(A) + RNA were hybridized with the oligonucleotides B8 and B9 (Fig. lb). With the oligonucleotide B8, which should hybridize with both transcripts, a hybridization signal at approx. 900 bp was obtained. Whereas the oligonucleotide B9, which is specific for the large transcript, did not produce a hybridization signal (data not shown).
A further indication that the short transcript is the most abundant in bull testis was obtained from primer extension analysis. As shown before (Figs. lb and 2a) the transcription initiation site of bull protamine 2 gene is located 125 bp upstream of ATG. Using primer B2 (exon 2, Fig. lb), a major extension product of 416 bases was obtained (Fig. 2b) which corresponds with the length of the short transcript. Even after 4 days of exposure no signal of the large transcript was detected. Therefore, the short transcript represents the most abundant variant of the two protamine 2 transcripts. The large one is present at a very low level and can be detected only by the very sensitive method of PCR. We estimate that the relative abundance of the large and the short transcript is less than 1 : 100.
Conservation of bull protamine 2 among bovine breeds Maier et al. [1] described the cDNA sequence of bull protamine 2. They demonstrated a derived amino acid motif R L H R I H which is highly conserved in mouse and human but diverges in bull. Although sequence analysis of genomic clones revealed two additional codons for arginine (pos. 178-183) and the change of an alanine to proline (pos. 184-186) the derived amino acid motif R L H R I H is not present in bull protamine 2. To confirm the nucleotide sequence in this region and to examine whether the underlying mutation is present in the protamine 2 genes of different bovine breeds (Gallaway, Holstein Friesian, Murnau Werdenfels cattle, Pustertaler Sprinzen, Swiss Brown, Simmental), PCR and direct sequence analysis were performed. It could be observed, that all examined bovine breeds do not contain the nucleotide sequence coding for the amino acid motif RLHRIH. Discussion
How to explain the lack of protamine 2 in bull spermiogenesis Protamine 2 has been found in spermatozoa of mouse and human but not in bull, boar and rat [4,7,22]. The respective transcripts were identified in these species but they were less abundant than in mouse and human [1,9,23]. The comparison of the exon-intron structure revealed no differences between species that possess protamine 2 and those that have only a reduced amount of m R N A and lack the protein. The coding sequences are split by one single intron and the proportions of exon 1 and 2 are comparable between the species (Fig. 3). Therefore, no information is gained about the differences in protamine 2 gene expression between mouse, human, bull, boar and rat, respectively, when the exon-intron structure is compared. The protamine genes of human, mouse and bull are closely linked in the genome [14,17,24] and might be controlled by similar trans-acting factors. To identify
138 elements possibly responsible for the reduced expression of the bull protamine 2 gene, the 5' flanking region of bull protamine 2 was compared with those of bull protamine 1 [25], mouse and human protamine 2 [17,19]. In the two protamine genes of the mouse as well as in the genes of human several common sequence motifs arranged in clusters have been found [17,19]. The 5' flanking regions of bull protamine 1 and 2 genes lack those common sequence motifs. Therefore, it can be suggested that the reduced expression of the bull protamine 2 gene is due to the lack of cis-acting elements in the 5' region. The results of in vitro transcription and gel retardation assays with mouse protamine 2 gene indicate that the region 140 bp to 23 bp upstream of the transcription start point is sufficient for maximal transcription. In this region three motifs (E,B, #8) have been found which are in common with mouse protamine 1 [26,27]. The motif #8 which was proposed to be a modified SPl-binding site, is also present in the bull protamine 2 but the nucleotide sequence of the motifs E and B are altered. Thus, it can be suggested that these changes in the 5' untranslated sequence have a negative effect on the transcription rate of the protamine 2 gene. However, the comparison between the 5' untranslated region of mouse and rat protamine 2 let us suggest that also other motifs than E, B and #8 are included in protamine 2 gene transcription. Protamine 2 of the rat was found to be transcribed with only 30% of the efficiency of mouse protamine 2 [23], although the three motifs are identical in mouse and rat. Another important difference between the gene for bull protamine 2 and the genes for protamines of other species are observed in the nucleotide sequence of the polyadenylation signal. The consensus sequence AAUAAA is changed to AAUAAG. This mutation could be responsible for the altered transcription rate of bull protamine 2 gene, as well. An identical mutation in the polyadenylation signal of the human a2globin gene resulted in a reduction of the transcription rate to 5% [28]. The reason, why bull protamine 2 is poorly transcribed in contrast to mouse protamine 2 or bull protamine 1 can not be answered by sequence comparisons of the 5' flanking regions. Protein binding studies and in vitro mutagenesis experiments should help to obtain further information about the expression of bull protamine 2. An additional cause for the absence of protamine 2 in bull spermiogenesis could be found at the translational level. Kwon and Hecht [29] demonstrated by gel retardation assays that two sequence motifs in the 3' untranslated region of mouse protamine 2 form specific RNA-protein complexes. These motifs (Y,H) are conserved in mammalian protamine and transition protein genes for which translational control is known. The 3' untranslated region of bull protamine 2 cDNA
[1] was found to contain only one of these motifs (H). If both elements are necessary for translational control, the lack of element Y could be responsible for a disturbed mRNA translation in vivo. However, Maier et al. [1] have shown that bull protamine 2 mRNA can be translated in vitro. The lack of protamine 2 in bull spermatozoa could be due to a disturbed protein function. Maier et al. [1] compared the cDNA sequence of bull and boar protamine 2 with that of mouse and human. A mutation was found in bull and boar which result in the absence of an amino acid motif RLHRIH. We have shown that this mutation is not only present at the cDNA level but also at the genomic level, including the genomic DNA of six bovine breeds. However, this motif is conserved in human and mouse in whose spermatozoa protamine 2 is present. Therefore, the assumption of Maier et al. [1] that the motif RLHRIH is of functional importance for DNA-protein interaction is supported.
Alternative splicing in the bull protamine 2 transcript Bull protamine 2 mRNA exists in two different variants in the bull testis (Fig. 4). The large transcript, 63 bp longer at the 3' end of exon 1, is present at a very low level, only detectable by PCR. Maier et al. [1] have isolated this less abundant cDNA variant from a bull spermatid library. To exclude the possibility that this large cDNA is an artifact of the library, PCR analysis was performed on total testis RNA. Both transcripts were detected and verified by sequencing. These results confirm that both transcripts are the products of alternative splicing. The most abundant short form has a splice donor site which is comparable with those of protamine 2 transcripts of other species. The two mRNA variants of the bull protamine 2 differ in 63 bp at the 3' end of exon 1, but these additional nucleotides do not result in a reading frame shift. However, the additional amino acids could lead to an altered protein conformation and function. Protamines are rich in arginines which are clustered. Due to the insertion of the 21 neutral and acidic amino acids a region in the bull protamine 2 protein which might be important for DNA-protein interaction is altered. It remains unclear if both transcripts for bull protamine 2 are translated and have different binding activities to DNA.
Acknowledgements We thank Prof. Dr. W. Holtz for providing the blood of the bovine breeds, Stephanie Bunkowski for technical assistance, Derek Murphy and Frauke Rininsland for reading the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (En 84/19-2).
139 References 1 Maier, W.M., Nussbaum, G., Domenjoud, L., Klemm, U. and Engel, W. (1990) Nucleic Acids Res. 18, 1249-1254. 2 Grimes, S.R., Meistrich, M.L., Platz, R.D. and Hnilica, L.S. (1977) Exp. Cell Res. 110, 31-39. 3 Hecht, N.B. (1989) Molecular Biology of Chromosomal Function (Adolph, K.W., ed.), pp. 396-420, Springer, New York. 4 McKay, D.J., Renaux, B.S. and Dixon, G.H. (1986) Eur. J. Biochem. 156, 5-8. 5 Belaiche, D., Loir, M., Kruggle, W. and Sautiere, P. (1987) Biochim. Biophys. Acta 913, 145-149. 6 Corzett, M., Blacher, R., Mazrimas, J.A. and Balhorn, R. (1987) J. Cell Biol. 105, 152a. 7 Bellv~, A.R., McKay, D.J., Renaux, B.S. and Dixon, G.H. (1988) Biochemistry 27, 2890-2897. 8 Balhorn, R. (1989) Molecular Biology of Chromosomal Function (Adolph, K.W., ed.), pp. 366-395, Springer, New York. 9 Bower, P., Yelick, P.C. and Hecht, N.B. (1987) Biol. Reprod. 37, 479-488. 10 Tanhauser, S.M. and Hecht, N.B. (1989) Nucleic Acids Res. 17, 4395. 11 Johnson, P.A., Yelick, P.C., Liem, H. and Hecht, N.B. (1988) Gamete Res. 19, 169-175. 12 Plenz, G., Prokop, C.M., Kratzberg, T. and Geldermann, H. (1988) J. Anim. Breed. Genet. 105, 389-399. 13 Hanahan, D. and Meselson, M. (1980) Gene 10, 63-67. 14 Reinhart, N., Kremling, H., Luerssen, H., Adham, I.M. and Engel, W. (1991) Biol. Chem. Hoppe-Seyler 372, 431-436. 15 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.
16 Wilbur, W.J. and Lipman, D.J. (1982) Proc. Natl. Acad. Sci. USA 80, 726-730. 17 Domenjoud, L., Nussbaum, G., Adham, I.M., Greeske, G. and Engel, W. (1990) Genomics 8, 127-133. 18 Kremling, H., Luerssen, H., Adham, I.M., Klemm, U., Tsaousidou, S. (1989) Differentiation 40, 184-190. 19 Johnson, P.A., Pechorn, J.J., Yelick, P.C., Palmiter, R.D. and Hecht, N.B. (1988) Biochim. Biophys. Acta 950, 45-53. 20 Keime, S., Heidland, K., Kumm, S., Schl6sser, M., Hroch, N., Holtz, W. and Engel, W. (1992) Biol. Chem. Hoppe-Seyler, in press. 21 Mount, S.M. (1982) Nucleic Acids Res. 10, 459-472. 22 Balhorn, R., Weston, S., Thomas, C. and Wyrobek, A.J. (1984) Exp. Cell Res. 150, 298-308. 23 Bunick, D., Balhorn, R., Stanker, L.H. and Hecht, N.B. (1990) Exp. Cell Res. 188, 147-152. 24 Hecht, N.B., Kleene, K.C., Yelick, T.C., Johnson, P.A., Pravtcheva, D.D. and Ruddle, F.H. (1986) Som. Cell Mol. Gen. 12, 191-196. 25 Krawetz, S.A., Connor, W. and Dixon, G.H. (1988) J. Biol. Chem. 263, 321-326. 26 Bunick, D., Johnson, P.A., Johnson, T.R. and Hecht, N.B. (1990) Proc. Natl. Acad. Sci USA 87, 891-895. 27 Johnson, P.A., Bunick, D. and Hecht, N.B. (1991) Biol. Reprod. 44, 127-134. 28 Higgs, D.R., Goodbourn, S.E.Y., Lamb, J., Clegg, J.B., Weatherall, D.J. and Proudfoot, N.J. (1983) Nature 306, 398-400. 29 Kwon, Y.K. and Hecht, N.B. (1991) Proc. Natl. Acad. Sci. USA 88, 3584-3588.