GENOMICS

13,613-621

(1992)

Structure and Expression of the Human ~58~‘~-’ Protein Kinase Chromosomal Gene PETER

G. EIms,*~t

JILL

M. LAHTI,* ANDVINCENT

J. Km*st+’

*Department of Tumor Cell Biology, St. Jude Children’s Research Hospital, Memphis, Tennessee 38105; t Laboratory of Medical Genetics, University of Alabama, Birmingham, Alabama 35294; and *Department of Biochemistry, University of Tennessee School of Medicine, Memphis, Tennessee 38105 Received

February

20, 1992;

March

26, 1992

normal cell division (Bunnell et al., 1990a,b). When this protein kinase is slightly overexpressed in fibroblasts, the cells are sequestered at late telophase for an extended period of time (Bunnell et al., 1990a). Those cells that eventually exit this temporary block exhibit an increased frequency of abnormal cytokinesis. Conversely, when the expression of the ~58”‘~~’protein kinase is diminished, cells replicate their DNA at an accelerated rate (Bunnell et al., 1990b). Of all the cell cycle genes examined thus far, the ~58”‘~~’protein kinase is the only one whose abnormal expression leads to a noticeable change in cell cycle progression and in cell phenotype. Abnormal cellular appearance and growth in response to altered ~58”‘~~’ gene expression suggest that this kinase may function to negatively regulate some aspect of normal cell cycle progression. Additionally, this gene has been mapped to a region of human chromosome l,lp36, that is frequently deleted in numerous tumors (neuroblastoma, malignant melanoma, endocrine neoplasia, etc.) and is believed to harbor a possible tumor suppressor gene (Eipers et al., 1991; Bishop, 1991). Furthermore, the structure of the predicted protein encoded by ~58”“~’ is 46% identical to the master mitotic cell cycle protein kinase, ~34”~“~ (B unnell et al., 1990a). The events that regulate the normal cell cycle in eukaryotic cells have been revealed by a combination of genetic, biochemical, and cellular studies. The onset of mitosis is controlled by the cooperative interaction of at least four distinct proteins that act as subunits in a combinatorial fashion, ~34”~“‘, cyclin B, pl3’““, and cdc25 (for reviews see Murray and Kirschner, 1989; Doree, 1990; Nurse, 1990; Pines and Hunter, 1990; Maller, 1991). p34”d”2 is the catalytic subunit of this complex, containing all of the requisite functional domains necessary for protein kinase activity (Hanks et al., 1988), while cyclin B, p13”“c’, and cdc25 function as regulatory subunits of the complex (Dunphy et al., 1988; Gautier et al., 1988, 1991; Arion et al., 1988; Lohka et al., 1988; Draetta et al., 1989; Ducommun et al., 1991; Galaktionov and Beach, 1991; Kumagai and Dunphy, 1991; Strausfield et al., 1991; Dunphy and Kumagai, 1991). All of these various components of the mitotic promoting factor (MPF) are members of larger gene fami-

A cDNA corresponding to a 58-kDa cell division control-related protein kinase, p58”“-‘, has previously been isolated, sequenced, and assigned to human chromosome 1~36. Aberrant expression of this protein kinase negatively regulates normal cellular growth. The protein contains a central domain of 299 amino P58c1k-1 acids that is 46% identical to human ~34’~“, the master mitotic protein kinase. Deletion of 1~36 has been correlated to numerous tumors, and this chromosome region has been suggested to harbor a putative tumor suppressor gene on the basis of the growth characteristics of these tumors. In this report we detail the complete structure of the p58c’kS’ chromosomal gene, including its putative promoter region, transcriptional start sites, exonic sequences, and intron/exon boundary sequences. The gene is 10 kb in size and contains 12 exons and 11 introns. Interestingly, the rather large 2.0-kb 3’ untranslated region is interrupted by an intron that separates a region containing numerous AUUUA destabilization motifs from the coding region. Furthermore, we detail the expression of this gene in normal human tissues as well as several human tumor cell samples and lines. The origin of multiple human transcripts from the same chromosomal gene, and the possible differential stability of these various transcripts, is discussed with regard to the transcriptional and post-transcriptional regulation of this gene. This is the first report of the chromosomal gene structure of a member of the ~34’~‘~ supergene family. o Issz Academic Press,

revised

Inc.

INTRODUCTION

The ~58”‘~~’protein kinase gene has been isolated and its expression and activity appear to be closely linked to Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under Accession Nos. M88553, M88554, M88555, M88557, M88559, M88560, M88561, M88562, M88563, and M88565. ’ To whom correspondence should be addressed at the Department of Tumor Cell Biology, St. Jude Children’s Research Hospital Memphis, TN 38105. 613

All

Copyright 0 1992 rights of reproduction

0888-7543/92 $5.00 by Academic Press, Inc. in any form reserved.

614

EIPERS,

LAHTI,

lies in vertebrates, suggesting that regulation of mitosis, as well as other points of the cell cycle (i.e., the restriction point controlling entry into S-phase), is regulated in a complex manner by these reiterated genes. Establishing a link between altered cell cycle function and cellular transformation could provide valuable insight into the mechanisms of uncontrolled cell growth. One of the more likely targets for transformation is the cyclin regulatory subunit, since its function is essential to the proper regulation of the catalytic protein kinase, and possibly the catalytic tyrosine phosphatase, subunits. Furthermore, the kinetics of G, cyclin mRNA and protein induction in response to extracellular signals suggest that these molecules may provide an advantageous target for cellular transforming events when they are expressed at high levels and escape normal cellular regulation (Hunter and Pines, 1991). Thus far, both cyclin A and cyclin D have been implicated in tumors where these genes have been disrupted either by viral integration, cyclin A, or by chromosomal translocation, cyclin Dl (Wang et al., 1990; Motokura et al., 1991). In addition, the protein kinase catalytic subunits, either by their direct or by their indirect interaction with substrates, may also prove to be targets of cellular transforming events. Several proto-oncogene products, including ~60”.““, cabl, pRb, and ~53, have been found to be potential substrates of ~34”~“~ and/or p33”&‘, a closely related protein kinase that has been found in association with cyclin A during early S-phase (Hunter, 1991; Wang et al., 1991; Faha et al., 1992; Ewen et al,, 1992; Cao et al., 1992; Devoto et al., 1992). These observations may indirectly link the ~34’~“, and/or p33cdk2, protein kinase activities to cellular transforming events. Additional members of the p34”d’2 supergene family may be more directly implicated in transforming events, in part due to their more specialized function in somatic cells. In this study, we have examined the normal structure and expression of the p5E?’ chromosomal gene that maps to human chromosome 1~36. The gene spans 10 kb of genomic DNA and contains 12 exons and 11 introns. The ~58”~-’ gene has one major transcription start site and several minor start sites as mapped by rapid amplification of cDNA ends (RACE) and Sl nuclease protection. The gene also contains multiple polyadenylation signal sites that could potentially function in uiuo in two different exons that encode the 3’ untranslated region of the gene. Furthermore, the major portion, but not all, of this large 3’ untranslated region, which contains numerous DNA sequence motifs implicated in mRNA stability (AUUUA), is contained in a separate exon. This pattern of structural organization may have implications for regulation of p58c’k-’ gene expression. These data are supported further by differential hybridization studies utilizing probes from this exon on Northern blot analysis of human mRNAs. Finally, the p58”k-’ gene is expressed ubiquitously in adult human tissues, as was previously shown for its murine counterpart (Kidd et al., 1991). These studies should provide important structural and regulatory information that may be utilized in the analy-

AND

KIDD

sis of the ~58~‘~~’ structural gene and mRNA in human tumors that are deleted in the 1~36 region, as well as provide important information regarding the regulation of a potentially important cell cycle-related gene product. MATERIALS

AND

METHODS

Isolation and characterization of the p58”“~’ chromosomal gene. Two different human cosmid genomic libraries, one constructed from leukocyte DNA of a normal human male and cloned into the vector pCV107 (Kidd, unpublished) and a second commercially available cosmid library of human placental DNA cloned into the vector pWE15 (Stratagene), were screened (5 X lo5 colonies) with the coding region of the human ~58 cDNA (Bunnell et al., 1990a). Three different cosmids, one from the human male pCV107 library (hp58cos2A) and two from the human placental library (hp58cos9 and hp58cos15), were isolated. One of these cosmids, hp58cos2A, was chosen for further analysis because it contained the entire ~58 chromosomal gene region and was previously used to localize the gene (Eipers et al., 1991). A minimal restriction endonuclease map of the ~58”‘~-’ gene region was established for the cosmid clones by differential hybridization of the cosmid DNA with various fragments of the human ~58”~-’ cDNA. DNA sequence analysis was performed as previously described (Bunnell et al., 1990a; Kidd et al., 1991) using synthetic oligonucleotides (Oligos Etc.) that corresponded to various regions of the human cDNA. These oligonucleotides were 18 nucleotides in length and were spaced at 150. to 250.bp intervals along both strands of the DNA. In most instances, gene fragments were subcloned into the pKS or pSK phagemid vectors and double-stranded DNA sequence analysis was performed. In regions of extensive secondary structure, ITP or deazaGTP was substituted in the DNA sequence reaction and/or new sets of synthetic oligonucleotides were constructed for additional sequence analysis. RACE cDNA analysis. Construction of a cDNA pool containing the 5’ end of the human ~58”~-’ message was carried out according to the method of Frohman and Martin (1989), with slight modification. First-strand cDNA synthesis was carried out using a gene-specific oligonucleotide (GSP-RT 5’-GAGGATGGTGTTGATCTCCCTCAG-3’) and 5 pg of total cellular RNA from the human pre-B ALL cell line 207. To synthesize cDNA, RNA was heated at 65°C for 3 min, cooled on ice, and then added to a reaction mix containing 20 mM of each dNTP, 1.5 pM reverse transcriptase (GSP-RT) primer, 10 units RNasin, 0.01 M DTT, 1X BRL (Bethesda Research Laboratories) RT buffer, and 200 units MuMLV RT (BRL). This mix was then incubated at 37°C for 90 min. The resulting cDNA pool was then purified over a Centricon 100 filtration unit (Amicon) before being tailed with terminal deoxynucleotidyltransferase, TdT (BRL), according to the manufacturer’s directions. The pool of poly(A)+ tailed cDNA was then subjected to two rounds of polymerase chain reaction (PCR) amplification using two nested gene-specific primers corresponding to the 3’ end of exon 2 of the gene (GSP-15’-GAAGCCCTCCTTCTCCTT-3’ and GSP-2 5’-CAGCCGCTTTAGAGCCAC-3’) and two primers corresponding to an adapter and a complimentary primer containing a poly(dT) stretch and several restriction endonuclease sites (adapter 5’- GACTCGAGTCGACATCGTTTTTTTTTTTTTTTTT3’ and primer 5’.GACTCGAGTCGACATCG-3’), which allowed anchored extension of the cDNA product in the 5’ direction of the mRNA. The resulting 860-bp PCR product was directionally subcloned into the plasmid pTZ-19R by utilizing the PstI site inside the gene-specific primers and the Sal1 site contained in the adapter. Sl n&ease mapping. A 3.0-kb BglII restriction fragment from hp58cos2A containing the 5’ untranslated region (UTR) was dephosphorylated and end labeled with [y-32P]ATP for use as a probe in Sl nuclease analysis as described by Weaver and Weissman (1979). The hybridization mix containing 20 ~1 of hybridization buffer, 50 pg of total cellular RNA from 207 cells and 100,000 dpm of probe was heated to 95°C for 3 min and then incubated overnight at 52°C. The hybridization reaction was then incubated with 200 ~1 of Sl buffer and

HUMAN

1~36

p58c’k-’

615

GENE

A. p58 protein 1

76

374

436

COOH

NH2 13

71100

142

178

218

255

II

I

294

355

405

B. Human p58 gene Pvull Pstl Hindlll BamHl EcoRl

I I I ,

I

I

I

I I ,I

I II

I

I I

ATG I UII 1

23

I

I

Ill

I

IIUI

4

557

s

9 10

I

TGA I 11

AATAAA I I

12

rkbl

FIG. 1. Schematic diagram and comparison of p58’lk-’ protein and gene structure. (A) Schematic diagram of the human ~58”~” protein and the position of introns relative to both the overall protein sequence and the structural subdomains of the protein. The solid box represents the catalytic domain, and the stippled box the unique putative amino-terminal regulatory domain of the protein, the open box the ~34”~” -related carboxyl-terminal domain. The positions of the amino acids that define these various domains of the protein are indicated above the protein. gene relative to amino acid positions in the protein. (B) Restriction map and Below the schematic are the positions of introns in the ~58”“.’ schematic diagram of the human 1~36 p58’ik-’ gene. Open boxes represent either 5’ or 3’ untranslated regions of the mRNA and solid boxes represent the coding regions of the protein. The relative positions of the ATG initiation codon, the TGA termination codon, and the final AATAAA polyadenylation signal are shown above the gene; numbers corresponding to the exonic regions of the gene are shown below. Positions of the indicated restriction endonuclease sites are shown above the gene. 200 U of SI nuclease at 37°C for 30 min. Mock incubations were performed in an identical manner, with the exception that tRNA was included in place of the 207 cellular RNA. Digestion products were then EtOH precipitated, dried, and resuspended in 4 ~1 of formamide loading dye and electrophoresed on a 6% acrylamide sequencing gel. The gel was exposed either overnight or for 2 days at -80°C to Kodak XAR film. Northern blot analysis. Five micrograms of poly(A)+ RNA from various human tumor patient material (chronic lymphocytic leukemia, CLL, or acute myelogenous leukemia, AML) or established B cell lines (kindly provided by Dr. Peter Burrows) was subjected to Northern blot analysis as previously described (Bunnell et al., 1990a; Kidd et al., 1991). Probes for the Northern blot analyses were derived from either the entire coding region for the p58”‘+’ cDNA contained in a 1.5kb EcoRI fragment (Probe A) or a 422.bp PuuII-EcoRI fragment (Probe B) created from the 3’untranslated region. The 5’untranslated region probe was derived from either the previously isolated murine products created in this study. Various P58”i+’ cDNA or the RACE adult human tissue poly(A)+ RNAs, 2 ag, were obtained from Clonetech and analyzed in a similar manner. A human y-actin cDNA control was used to rehybridize the blots.

RESULTS

AND

DISCUSSION

p58”lkm’Gene Sequence and Organization Three overlapping cosmid clones corresponding to the human p5SCrkV’ gene located on human chromosome 1~36 were isolated from two different human cosmid libraries. One of these clones, hpMcos2A, has been previously reported and was used to map the gene to human chromosome 1~36 (Eipers et al., 1991). The remaining two cosmids overlap this isolate, but both extend further in the 3’ direction from the gene (data not shown). Therefore, hp58cos2A was chosen for further analysis of the chromosomal gene structure. This cosmid contained approxi-

mately 36 kb of DNA, with approximately 14 kb of DNA extending 5’ of the p5E~“‘~-’gene and 12 kb of DNA extending 3’ of the gene. The structure of the ~58”‘~~’chromosomal gene is shown in Fig. 1B. The gene spans 10 kb of DNA and contains 12 exons and 11 introns. The DNA sequence of the putative promoter region, all exons, all intron/exon boundaries, and the entire 3’ untranslated region are shown in Fig. 2. This DNA sequence confirmed the previous open reading frame (ORF) predicted by the human fetal liver cDNA isolate (Bunnell et aZ., 1990a), with the exception of several nucleotide changes in the amino-terminal region of the protein (exon 2), which were all due to sequencing errors, and a single nucleotide transition in the catalytic domain of the kinase (exon 7), which is apparently due to allelic variance. These changes resulted in changes in the predicted amino acid sequence of the protein for amino acid residues 1-41, 46, and 242 (Fig. 2). Both strands of the genomic DNA were routinely sequenced for the entire gene, confirming these changes. When the human cDNA isolate was resequenced in these same regions, the alterations observed in the amino-terminal region (exon 2) of the genomic DNA were confirmed. This region of the cDNA contains homopurine and GCrich sequences that apparently resulted in sequence compression in the original gels and misinterpretation of the sequence in this region. The single nucleotide change in exon 7 was not found in the original human cDNA isolate and presumably represents allelic variance in the genomic DNA. Similarly, a single nucleotide change due to allelic variance in exon 9 of the gene, which does not result in any change of the predicted protein sequence, was found in the gene but not the

EIPERS,

LAHTI,

AND

KIDD

FIG. 2.

Nucleotide sequence of the human 1~36 chromosomal gene. The complete nucleotide sequence of all exons and 1.0 kb of DNA immediately 5’of the ATG initiation codon are shown, as well as partial or complete sequences of the intronic regions. The predicted ORF of the previously isolated human cDNA (top sequence) is compared directly to the predicted ORF of the gene (bottom sequence). Differences in nucleotide sequences are indicated by * and differences in the amino acid sequence are underlined. Gaps or insertions in the nucleotide sequence are indicated by -. Amino acid positions are indicated by numbers below the protein sequence. The position of the major transcriptional start site is indicated by the large arrow, the secondary site by the smaller arrow, and the minor start sites by arrowheads. Exonic sequence is indicated by capital letters and intronic or flanking sequence by lowercase letters. An additional ATG initiation codon in the 5’ untranslated region is underlined once, and in-frame termination codons following this codon are underlined twice. The locations of the BglII restriction sites used for the Sl analysis are boxed in the 5’untranslated region and in exon 5. Potential polyadenylation splice signal sites in exons 11 and 12 are underlined. Interruptions in intronic sequence are indicated by /.

cDNA isolate. Several nucleotide changes were found scattered throughout the 3’ untranslated region and are indicated appropriately (Fig. 2). These differences are most likely due to original sequencing errors. The first 32 bp of the original human ~58~‘~-’ cDNA, corresponding to a portion of the 5’ untranslated region, could not be found in the genomic DNA in a similar location or within 1 kb of the initiation codon of the ORF (Fig. 2). This sequence may represent a cloning artifact in the original cDNA based on further analyses presented below, but it may also represent an infrequently utilized splicing variant involving a distal exonic region not found in this cosmid. All of the intron/exon boundaries of the gene were flanked by consensus splice donor and splice acceptor sequences (Table 1). The structure of the human p58c’k-’ chromosomal gene demonstrates that the gene most likely arose by gene duplication and divergence from the ancestral ~34’~”

gene followed by gene fusion, which was originally suggested based on a comparison of the genes in the ~34”~“~ supergene family (Kidd, 1992). The unique amino-terminal region of ~58~‘~~’ is encoded by two distinct exons (Fig. 1A). Similarly, the unique carboxyl-terminal region of the protein is encoded by two distinct exons (Fig. 1A). The catalytic domain of the ~58”‘~~’ protein kinase is encoded by seven separate exons that are approximately loo-120 bp in size each. It will be of interest to compare the organization of the ~34”~“~ chromosomal gene to that of p58”lk-’ since the two genes are 68% homologous to one another in this region (Bunnell et al., 1990a). Finally, the 3’ untranslated region of this gene is interrupted by an intron (Figs. 1B and 2). To our knowledge, this is the first report of an intron in the 3’ untranslated region of a mammalian gene. Functional poly(A) splice signal sites can be found in appropriate locations of both exons 11 and 12 (Fig. 2) (Nussinov, 1986).

a-991 IC~YYC~~YYY~

FIG.

2-Continued

618

EIPERS,

TABLE Nucleotide

1

Sequence of Intron/Exon in Human ~58

5’ UT region-Exon acttttccagTTCCA-Exon tccttcccagGGGCT-Exon ttcttcacagATGAA-Exon tgccctgcagGAGAT-Exon tccctcccagGGGAG-Exon gtcgttgcagGTGGG-Exon -_ggcgtcttagGAATA-Exon t&tttcaaGACCT-Exon ctccccccagGTTCC-Exon tcgtccccagGGTGA-Exon actagactagCTAAA-Exon

LAHTI,

Junctions

l-GGTGgtaagaacct ZWZTGCAgtcaggccct 33AACAGgtgggtgggg 44TTAGAgtgagcatgg 5-GCCAGgtacagccgc 66TGAAGgtgagccccc 7-CCAAGgtcctgggat %TCAAG&gg&tgg 99AACAAatganccaat -I_ lo-AGCTGgtgaggggcc ll-TCCTTgttgtggagg 12-AATAAA

-

Mapping the Transcriptional Start Sites of the p58’lkm1Gene The transcriptional start sites of the ~58”‘~~’gene were determined by two distinct methods. First, the entire 5’ untranslated region of the gene was cloned, and the putative transcriptional start site mapped, by RACE using total RNA from the human pre-B ALL cell line 207. After extension from the anchored oligonucleotide corresponding to the previously determined cDNA sequence, 20 separate RACE isolates were selected and sequenced. In all 20 isolates the cDNA clones initiated at an identical position, the G nucleotide located at position -430 of the ~58”‘~~’ gene sequence (Fig. 2). Second, the putative start site was mapped by Sl nuclease protection analysis of total RNA from the same cell line. A 3-kb BgZII fragment from the gene was isolated from a larger 6.5kb BamHI fragment containing the complete ORF. Digestion of the larger BumHI fragment with BglII resulted in two separate ends that were subsequently labeled with 32P,both of which would protect specific regions of the mRNA molecule. One of the BglII sites was located in exon 5 of the gene (Fig. 2) and would protect a 34-nucleotide region of the mRNA, while the other BglII site was located in the 5’ untranslated region (Fig. 2) and would protect a 47-nucleotide region of the mRNA based on the initiation site predicted by RACE. Thus, the exon 5 restriction site provided an internal control for the Sl mapping studies. Appropriate undigested and digested probe controls were also included in the experiment. Examination of the Sl nuclease digestion products by gel electrophoresis revealed the expected 34-nucleotide band and a slightly smaller than expected 40-nucleotide band in approximately equimolar ratio (Fig. 3). One possible explanation for the discrepancy in the expected size of the larger band is the extensive homopurine/homopyrimidine DNA sequence in this region of the 5’ untranslated region (Fig. 2). Unusual secondary structure of the DNA in this region may have created structures that, when hybridized as an RNA/DNA duplex, were digested anomalously by the Sl nuclease. However, based on the unequivocal analyses of numerous distinct RACE clones

AND

KIDD

and their absolute conservation with the genomic sequence, the major transcriptional start site is most likely located at nucleotide -420. In addition, the Sl nuclease data support the location of the major transcriptional start site within 5-8 nucleotides of this region. A much less abundant secondary transcriptional start site was revealed by a longer exposure of the Sl gel (Fig. 3). This site maps to a G nucleotide at position -537 of the genomic sequence (Fig. 2). Based on the relative abundance of the protected product, this alternative transcriptional start site is utilized at approximately 2-5% of the frequency of the major start site, which would account for its absence in the RACE clones that were analyzed previously. Finally, consistent with many TATA-less promoter regions, the ~58”~~’ gene promoter also contains several very minor Sl nuclease protected fragments that may represent infrequent transcriptional start sites (Figs. 2 and 3). Numerous potential probe-

t=

TF

---major

si tart

-exon

5

b FIG. 3. Sl nuclease protection studies. Autoradiographs of Sl nuclease protection analysis of total RNA from 207 cells are shown with the sizes of molecular weight markers shown on the left. The position of undigested probe is shown. The lanes flanking either side of the center lane contain mock incubations with tRNA, while the center lane contains RNA from the 207 cell line. The left lane contains labeled molecular weight markers. The panel on the left is a 2-day exposure to Kodak XAR-5 film at -8O”C, while the panel on the right is an 8-h exposure of the same gel. The positions of the various protected fragments, including the protected portion of exon 5 and the major start site, are indicated by arrows to the right of the panels.

HUMAN

1~36

~58”‘~-’

GENE

619

tween the two mRNA species mentioned above. This is also consistent with the Sl nuclease mapping and RACE results, both of which predicted a single major transcriptional initiation site, and, therefore, precludes the usage of multiple major initiation sites. To test this hypothesis, several different human poly(A)+ RNAs were probed with either a coding region probe corresponding to exons l-11 or a 3’ untranslated region probe corresponding to the 3’ end of exon 12 (Fig. 4A). When RNAs were probed with the coding region probe, two p58C’k-’mRNAs, 1.71.8 and 3.5-3.7 kb in size, were observed (Fig. 4B). The larger transcript represents mRNAs at 3.5 and 3.7 kb that were not resolved under these conditions because these two transcripts are observed in most tissues. Iden- 3.7 tically sized transcripts were also seen when this same blot was hybridized to a 5’ untranslated region probe derived from the RACE clones (data not shown). Conversely, when the probe corresponding to the 3’untranscl.8 lated region from exon 12 was used on the same blot, only the larger 3.7-kb mRNA species was observed (Fig. 4B). The levels of the various mRNAs varied considerkb ably in the RNA derived from in uiuo tumor cells (espe1234 56 cially the smaller 1.7- to 1.8-kb transcript), whereas the FIG. 4. Northern blot analysis of various human poly(A)+ RNAs larger transcripts were more abundant in the established using distinct probes for exons l-11 or 12. (A) Schematic diagram of human cell lines (Fig. 4B), which is consistent with prethe human p58c’k-’ cDNA showing selected restriction endonuclease vious observations (Bunnell et al., 1990a,b; Kidd, et al., sites, the position of intron 11 in the 3’ untranslated region, the location of potential polyadenylation signal sites, and the location of the 1991; Kraft et al., 1992). The potential significance, if two probes derived from the cDNA and used in the analysis. (B) any, of the abundant smaller transcript in certain tumor Northern blots of l-5 /lg of the following poly(A)+ RNAs; lane 1, a p/y cell samples is unknown. IgG CLL-1 patient; lane 2, a l/y IgG CLL-2 patient; lane 3, LBW 25 We have also previously shown that the expression of B cell line; lane 4, AML patient; lane 5, Nalm 6 pre-B ALL cell line; the ~58~‘~~’gene in normal adult tissues from the mouse lane 6, 207 pre-B ALL cell line. The probe used for each blot is indicated above each panel. The sizes of the major transcripts are indiis ubiquitous (Kidd et al., 1991). To determine whether cated to the right. the gene is also ubiquitously expressed in normal adult human tissues, a similar Northern blot analysis was perbinding sites for SPl, AP-1, GRE, HSE, Octl, ATF, and formed (Fig. 5). Two equivalently expressed transcripts, 3.7 and 3.5 kb in size, were observed in normal adult NFKB transcriptional factors were also observed in this putative promoter sequence. However, further analysis heart, brain, placenta, lung, liver, skeletal muscle, kidof this putative promoter region will be required before ney, and pancreas (Fig. 5). Similarly sized transcripts were also observed in the human U937 cell line (Kraft et specific regions can be definitively identified as funcal., 1992) and have been observed in several of the pre-B tional components in uiuo. cell lines used in Fig. 4 when the gel is resolved to a higher degree by electrophoresis (data not shown). Northern Blot Analysis of Selected Tumor Cell RNAs Therefore, there is no bona fide difference between the and Normal Adult Tissue RNAs sizes of the ~58~‘~~’mRNA transcripts seen in human The previous analyses demonstrated that the p58C1k-’ tumor cell lines and those seen in normal human tissues. chromosomal gene potentially encoded several mRNA Based on the human ~58”‘~~’ cDNA and genomic setranscripts from the same locus and that differences in quence, it appears most likely that these transcripts these transcripts might be regulated by differential utiliarise from the differential utilization of polyadenylation zation of polyadenylation splice signal sites in exons 11 splice signal sites located 184 bp apart in exon 12 of the and 12. We have previously shown that human, mouse, gene (Fig. 2). Hybridization experiments utilizing a maand Chinese hamster cell lines and tissues produce as jor portion of the 5’ untranslated region from the RACE many as three differently sized ~58”‘~~~mRNAs in varyclones (nucleotides -420 to -180 of the genomic seing amounts (Bunnell et al., 1990a,b; Kidd et al., 1991; quence) revealed that both of the mRNAs hybridized Kraft et al., 1992). It is quite possible that at least two of with this probe (data not shown), which in combination these mRNAs, 3.7 and 1.8 kb in size, are derived by utiliwith the Sl nuclease protection results suggest that both zation of the polyadenylation signal sites in exons 11 and of the transcripts were derived from the same gene and 12. If this occurred, the mRNAs would differ in size by that they were transcriptionally initiated at similar exactly 1969 bp based on the gene sequence (Fig. 2), sites. Both polyadenylation signal sites in exon 12 are which is equivalent to the observed size difference be- perfect consensus signal sequences, and both sites are

620

EIPERS,

LAHTI,

-1.8 kb actin

AND

KIDD

mainder of the 3’ untranslated region by an intronic sequence. This is the first observation of such an intron in a mammalian 3’ untranslated region, and it is possible that this particular structural gene organizational motif is of some regulatory importance. Inclusion or exclusion of these 3’ untranslated regions in different ~58”‘~~’transcripts was observed in certain human tumor cell samples and lines, supporting this notion. Finally, it will be of interest to compare the structural organization and promoter regions of the various ~34”~“’ supergene family members to determine how these genes evolved, how they are differentially expressed during the cell cycle, and the potential for abnormal expression and/or function of these genes in human tumors ACKNOWLEDGMENTS

FIG. 5. Northern blot analysis of poly(A)+ RNAs from various adult human tissues. Two micrograms of poly(A)+ RNA from adult human tissues, indicated below the respective lanes, was analyzed with either the p58c’k-’ coding region cDNA probe (top) or a human y-actin cDNA probe (bottom). The sizes of the various transcripts are indicated to the right of the panel. The ~58 panel was exposed to Kodak XAR-5 film at -80°C for 10 h; the actin panel was exposed at -80°C for 2 h.

preceded by similar AUUUA destabilization motifs (Caput et al., 1986; Shaw and Kamen, 1986). Upon prolonged exposure, a smaller 1.8-kb transcript was also seen (data not shown), but the level of expression was much lower than that of the two larger transcripts. This is consistent with the imperfect consensus polyadenylation signal sequence found in exon 11 versus the two perfect consensus sites in exon 12, and may indicate biased utilization of the sites found in exon 12, while the site found in exon 11 may be more specialized in function. This point is currently being investigated further with regard to potential differences in gene regulation and the stability of the various mRNA transcripts. Concluding Remarks In this report we have presented data detailing the structure and expression of the human p5Bcik-’ chromosomal gene, a protein kinase that appears to negatively regulate cell cycle progression in vertebrate cells. Structural analyses have revealed that this gene may be complexly regulated by utilization of numerous polyadenylation splice signal sites located in two distinct exons. Interestingly, a major portion of the rather large 3’ untranslated region containing all of the potential AUUUA mRNA destabilization motifs (Caput et al., 1986; Shaw and Kamen, 1986) is separated from the re-

We thank Dr. Linda Shapiro and Mr. Bart Jones for help and advice regarding the RACE analysis and Sl nuclease mapping, respectively, and Dr. Patrick Umeda (UAB) for confirmation of the hp58”“.i cDNA sequence. We also thank the other members of the Kidd laboratory for helpful comments and suggestions regarding the manuscript. This work was supported by grants from the NIH (GM 44088), the American Cancer Society (CD-389B), and the American Lebanese Syrian Associated Charities of St. Jude Children’s Research Hospital. Peter G. Eipers was supported in part by a training grant from NIH (MCJ000905-27-0).

REFERENCES Arion, D., Meijer, L., Brizuela, L., and Beach, D. (1988). cdc2 is a component of the metaphase-specific histone Hl kinase: Evidence for identity with MPF. Cell 55: 371-378. Bishop, J. M. (1991). Molecular themes in oncogenesis. Cell 64: 235248. Bunnell, B. A., Heath, L. S., Adams, D. E., Lahti, J. M., and Kidd, V. J. (1990a). Elevated expression of a 58 kDa protein kinase leads to changes in the CHO cell cycle. Proc. N&l. Acad. Sci. USA 87: 7467-7471. Bunnell, B. A., Fillmore, H., Gregory, P., and Kidd, V. J. (1990b). A dominant negative mutation in two proteins created by ectopic expression of an AU-rich 3’ untranslated region. Somatic Cell Mol. Genet. 16: 151-162. Caput, D., Beutler, B., Hartog, K., Thayer, R., Brown-Shimer, S., and Cerami, A. (1986). Identification of a common nucleotide sequence in the 3’ untranslated region of mRNA molecules specifying inflammatory mediators. Proc. Natl. Acad. Sci. USA 83: 1670-1674. Cao, L., Faha, B., Dembski, M., Tsai, L-H., Harlow, E., and Dyson, N. (1992). Independent binding of the retinonblastoma protein and ~107 to the transcription factor E2F. Nature 355: 176-179. Devoto, S. H., Mudryj, M., Pines, J., Hunter, T., and Nevins, J. R. (1992). A cyclin A-protein kinase complex possesses sequence-speof the E2F-cyclin cific DNA binding activity: p33 cdks is a component A complex. Cell 68: 167-176. Doree, M. (1990). Control of M-phase by maturation promoting factor. Curr. Opin. Cell Biol. 2: 269-273. Draetta, G., Piwnica-Worms, H., Morrison, D., Druker, B., Roberts, T., and Beach, D. (1989). cdc2 protein kinase is complexed with both cyclin A and B: Evidence for proteolytic inactivation of MPF. Cell 56: 829-836. Ducommun, B., Brambilla, P., and Draetta, G. (1991). Mutations at sites involved in sucl binding inactivate Cdc2. Mol. Cell. Biol. 11: 6177-6184. Dunphy, W. G., Brizuela, L., Beach, D., and Newport, J. (1988). The

HUMAN

Xenopus cdc2 protein is a component of MPF, a cytoplasmic tor of mitosis. Cell 54: 423-431. Dunphy, W. G., and Kumagai, A. (1991). The cdc25 protein an intrinsic phosphatase activity. Cell 67: 1899196.

1~36

regulacontains

Eipers, P. G., Barnoski, B. L., Han, J., Carroll, A. J., and Kidd, V. J. (1991). Localization of the expressed human p58 protein kinase chromosomal gene to chromosome 1~36 and a highly related sequence to chromosome 15. Genomics 11: 621-629. Ewen, M. E., Faha, B., Harlow, E., and Livingston, D. M. (1992). Interaction of ~107 with cyclin A independent of complex formation with viral oncoproteins. Science 255: 85-87. Faha, B., Ewen, M. E., Tsai, L-H., Livingston, D. M., and Harlow, E. (1992). Interaction between human cyclin A and adenovirus ElAassociated ~107 protein. Science 255: 87-90. Frohman, M. A., and Martin, G. R. (1989). Rapid amplification of cDNA ends using nested primers. Technique 1: 165-170. Galationov, K., and Beach, D. (1991). Specific activation of cdc25 tyrosine phosphatases by B-type cyclins: Evidence for multiple roles of mitotic cyclins. Cell 67: 1181-1194. Gautier, J., Norbury, C., Lohka, M., Nurse, P., and Maller, J. (1988). Purified maturation-promoting factor contains the product of a Xenopus homolog of the fission yeast cell cycle control gene cdc2+.

Cell54:433-439. Gautier, J., Solomon, M. J., Booher, R. N., Bazan, J. F., and Kirschner, M. W. (1991). cdc25 is a specific tyrosine phosphatase that directly activates ~34”~“. Cell 67: 197-211. Hanks, S. K., Quinn, A. M., and Hunter, T. (1988). The protein kinase family: Conserved features and deduced phylogeny of the catalytic domains. Science 241: 42-52. Hunter, T. (1991). Cooperation between oncogenes. Cell 64: 249-270. Hunter, T., and Pines, J. (1991). Cyclins and cancer. Cell 66: 1071-

1074. Kidd, V. J. (1992). Cell division control-related protein kinases: Putative origins and functions. Mol. Carcinog. 5: 95-101. Kidd, V. J., Luo, W., Xiang, J., Tu, F., Easton, J., McCune, S., and Snead, M. S. (1991). Regulated expression of a cell division controlrelated protein kinase during development. Cell Growth Differ. 2: 85-93. Kraft, A. S., Wang, S. S., Xiang, J., Pouncey, L., and Kidd, V. J. (1992). Regulation of the p58 cell division control-related protein

p58”“-’

621

GENE

kinase during differentiation

phorbol of U937

12-myristate 13-acetate-induced cells. Oncogene 7: 501-506.

terminal

Kumagai, A., and Dunphy, W. G. (1991). The cdc25 protein controls tyrosine dephosphorylation of the cdc2 protein in a cell-free system. Cell 64:903-914. Lohka, M. J., Hayes, M. K., and Maller, J. L. (1988). Purification of maturation-promoting factor, an intracellular regulator of early mitotic events. hoc. Natl. Acad. Sci. USA 85: 3009-3013. Maller, J. L. (1991). Mitotic control. Curr. Opin. Cell Biol. 3: 269-275. Motokura, T., Bloom, T., Kim, H. G., Juppner, H., Ruderman, J. V., Kronenberg, H. M., and Arnold, A. (1991). A novel cyclin encoded by a bcl-l-linked candidate oncogene. Nature 350: 512-515. Murray, A. W., and Kirschner, M. W. (1989). Dominoes and clocks: The union of two views of the cell cycle. Science 246: 614-621. Nurse, P. (1990). mitosis. Nature

Universal control 344: 503-508.

Nussinov, R. (1986). Sequence signals cient formation of mRNA 3’ termini.

mechanism

regulating

which may be required Nucleic Acids Res.

onset

of

for effi-

14: 3557-

3572. Pines, J., and Hunter, associated protein ture 346: 760-763.

T. (1990). Human cyclin p60 and behaves differently

A is adenovirus from cyclin

ElAB. Nu-

Shaw, G., and Kamen, R. (1986). A conserved AU sequence from the 3’ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46: 659-667. Strausfield, U., Labbe, J. C., Fesquet, D., Cavadore, J. C., Picard, A., Sadhu, K., Russell, P., and Doree, M. (1991). Dephosphorylation and activation of a ~34”~“’ cyclin-B complex in vitro by human cdc25 protein. Nature 351: 242-245. Wang, J., Chenivesse, X., Henglein, B., andBrechot, C. (1990). titis B virus integration in a cyclin A gene in a hepatocellular noma. Nature 343: 555-557.

Hepacarci-

Wang, H-G. H., Draetta, G., and Moran, E. (1991). ElA induces phosphorylation of the retinoblastoma protein independently of direct physical association between the ElA and retinoblastoma products. Mol. Cell. Biol. 11: 4253-4265. Weaver, R. F., and Weissman, C. (1979). Mapping of RNA by a modification of the Berk-Sharp procedure: The 5’ termini of 15s P-globin mRNA precursor and mature 10s fl-globin mRNA have identical map coordinates. Nucleic. Acids Res. 7: 1175-1193.