A mouse thymidylate synthase pseudogene derived from an aberrantly processed RNA molecule

363

Gme. 82 (1989) 363-370 Elsevier GENE

03154

A mouse thymidylate synthase pseudogene derived from an aberrantly processed RNA molecule (Recombin~t

DNA;

polyadenyiation;

mRNA;

alternative

splicing;

intron;

sequence

evolution)

Dawei Li and Lee F. Johnson of Biochemistry

Departments Received

by M. Belfort:

Revised:

7 May 1989

Accepted:

and Molecular Genetics, The Ohio State University, Columbus, OH 43210 (U.S.A.)

3 April 1989

8 May 1989

__._ SIJMMARY

A DNA fragment containing a mouse-thymidylate-synthase(TS) processed pseudogene was cloned and analyzed. Comparison with the sequences of the mouse TS-encoding gene (ts) and cDNA revealed that the pseudogene started at one of the normal 5’ termini of TS mRNA, ended with a poly(A) tail, and was flanked by 16-nucleotide (nt) direct repeats. The region corresponding to the open reading frame was 97.3 y0 identicai to that of the. cDNA. Two unusual features were observed. First, the poly(A) tail of the pseudogene was located 2 kb downstream from the normal location. Second, the final 10 nt of intron 5 were retained in the ‘coding region’ of the pseudogene. Therefore, it appears that the pseudogene was derived from a nonfictional TS ‘mRNA’ that was aberrantly spliced and polyadenylated. Analysis of the sequence of intron 5 of the ts gene revealed the presence of an alternative 3’ splice site 10 nt upstream from the normal splice site. Sl-nuclease protection assays showed that about 10 r/; of TS mRNA isolated from mouse cells was spliced at the alternative site. ._.---.

_.-. .._. ._

We have been studying the structure and expression of the gene for thymidylate synthase (TS; EC

Correspondence

to: Dr. L.F. Johnson,

try, The Ohio State University, OH 43210 (U.S.A.) Abbreviations: 10h years;

Department

Tel. (614)292-9479;

Fax (614)292-1538.

bp, base pair(s); kb, kilobase nt, nucleotide(s);

ORF,

open

citrate

pH 7.6; TS, thymidyIate

reading

frame;

ing TS; rsp, transcription

(1378-I 1 IY~R9:503.50

0

of Biochemis-

484 West 12th Ave., Columbus,

oiigo,

or 1000 bp; myr,

oligodeoxyribonucleotide;

SSC, 0.15 M NaC1/0.015 synthase;

start point(s).

19R9 Elsevier

Science

M Na,.

ts, gene (DNA) encod-

Publishers

R.V.

2.1.1.45) in cultured mouse 3T6 fibroblasts. We have isolated a ~uorodeoxyuridine-resistant derivative (LU3-7) that overproduces TS (Rossana et al., 1982) and its mRNA (Geyer and Johnson, 1984) by a factor of about 50. Southern-blot analyses revealed that all of the restriction fragments that contain exons of the normal ts gene (Deng et al., 1986) are amplified about 50-fold in the LU3-7 cells (Jenh et al., 1985b). However, one restriction fragment (or more, depending on the enzyme used) detected by the ts cDNA probe is present at the same level in LU3-7 and 37’6 cells. This ‘unampli~ed ts sequence’ may represent another functional ts gene, a different

(Biomedical Division)

364

gene with a related sequence or a ts pseudogene. To distinguish between these possibilities, we have

(b) Sequence analysis

cloned

Appropriate were subcloned

and analyzed

the unampli~ed

ts sequence.

We found that the sequence has many of the properties typical of a processed pseudogene (Vanin, 1985; Wilde,

1985).

fragments from the insert of TX-4 into Bluescribe and Bluescript plas-

mids (Stratagene) chain-termination sequencing parison

and sequenced using the dideoxy method (Sanger et al., 1977). The

strategy

is shown

of the sequence

in Fig. 1. The com-

of the cloned fragment with

those of the ts cDNA (Perryman et al., 1986) and the flanking regions of the ts gene (Deng et al., 1986) is EXPERIMENTAL

AND DISCUSSION

shown

in Fig. 2, A and B. Similarity

unamplified (a) Isolation and analysis of the pseudogene Southern-blot analysis of DNA isolated from LU3-7 (overproducing) and 3T6 (parental) cells revealed that the only unamplified ts sequence that was detected at the stringency of hybridization used in our analyses (1 x SSC, 65”C), was contained within a single 9.5kb XbaI restriction fragment (data not shown). The unamplified zs sequence was cloned using standard procedures (Maniatis et al., 1982). Briefly, high M,. DNA was isolated from 3T6 cells and digested with XbaI. DNA fragments ranging from 7-19 kb were isolated by sucrose-~adient sedimentation, ligated to theXba1 site of phage a Charon 35 (Loenen and Blattner, 1983) and packaged into phage particles. E. co/i LE392 were infected with the phage and a library of lo6 plaques was obtained. The library was screened with a full length l.s cDNA probe, pMTS-3 (Geyer and Johnson, 1984). DNA isolated from positive phages was further analyzed by restriction mapping and Southern-blot analysis. One clone, TX-4, corresponded to the unamplified 9.5-kb IS sequence. To be certain that the cloned fragment was not rearranged and that it corresponded to the unamplitied ts sequence rather than a portion of the ts gene, the fragment was analyzed by Southern-blot analysis using multiple restriction enzymes. The patterns were compared with those for DNA isolated from 3T6 and LU3-7 cells. For each set of restriction enzymes used in these analyses (XbaI + BamHI or EcoRI or HindIII), all of the restriction fragments from LU3-‘7 cells that contained unamplified ts sequences corresponded to fragments present in the TX-4 insert, and no other unamplified fragments were detected (not shown). The restriction map of the cloned insert is shown in Fig. 1.

between

the

ts sequence and the ts genomic sequence

was first observed 88 bp upstream from the AUG start codon. This corresponds to the tsp for TS mRNA moiecules with the longest 5’-untranslated region (Deng et al., 1986). Comparison of the sequence downstream from this point with that of ts cDNA revealed a high degree of similarity across the entire coding region. Thus it appears that the unamplified ts sequence corresponds to a processed ts pseudogene. The pseudogene has suffered a number of mutations (single nt changes as well as small insertions and deletions) across the coding region,

Fig. I. Restriction

maps and sequencing

tion maps of the pseudogene rs gene (including

strategies.

The restric-

and the 3’ end of the normal mouse

the last two exons and the 3’ flanking

region)

are shown. A plus symbol indicates

the sites where the sequences

diverge. The thick lines correspond

to coding regions. The arrows

indicate

of the sequences

the length

determined. Ap,ApaI; HindIII;

and direction

The following restriction B,BamHI;

Hf, Hid;

Ba, WI;

sites are indicated:

that were A,AccI;

Bs, BsmI: C, ClaI; E, EcoRI; H,

P, PsII; S, SacI; Sp, SphI; St, StuI; X, XbaI.

365

which precludes the possibility that the sequence could encode a functional TS enzyme.

adenylated

at the stop codon, while another

polyadenylated

at additional

15 ?, are

sites less than 2.50 nt

Other features that are commonly found in a processed pseudogene in&de a poly(A) tail at the 3’ end and direct repeats of chromosomal DNA

downstream

immediately adjacent to the 5’ end and the poly(A) tail of the pseudogene (Vanin, 1985; Wilde, 1985).

ts cDNA probe, but were longer than 2 kb. These iarge molecules represented only a few 9; of the total

We showed previously

number

that the structure

of the pre-

dominant mouse TS mRNA is highly unusual in that it lacks a 3’-untranslated region; the poly(A) tail immediately

follows

the

UAA

translational

stop

codon (Perryman et al., 1986). The stop codon in the fs genomic sequence is UAG and is not followed by a poiy(f%) sequence (fenh et al., 1986). We were surprised to find that the fs pseudogene lacked a poly(A) sequence at the position corresponding to the poly(A) tail ofTS mRNA. Fig. 2A shows that the sequence of the pseudogene was very similar to that of the normal mouse gene for at least a few hundred nt downstream from the UAG stop codon. Xdditional physical mapping studies and preliminary sequence analyses (not shown) indicated that the pseudogene was similar to the normal fs gene for about 2 kb downstream From the coding region, then diverged. To define more precisely the point of divergence, the sequence of an Ew Rl-ACCI fragment in the 3’ flanking region of the pseudogene (Fig. 1) was determined and compared with the corresponding sequence from the 3’ end of the IS gene. Fig. 2B shows that there is a high degree of similarity between the sequences of the pseudogene and gene up to but not beyond an A-rich region in the pseudogene. This region consists of a stretch of 16 A residues followed by 13 repeats of GAAA. Eleven nt upstream from the poty(A) stretch is the sequence ATTAAC, which is similar to the upstream polyadenylation consensus signals AATAAA or ATTAAA (Birnstiel et al.. 1985; Jenh et al., 1986). The sequence imtnediatel~ downstream from the A-rich region is AAGACAGCCACCTGAA, which differs at only two positions from the sequence AAGAAAGCCACCTGAG that is immediately upstream from the 5’ terminus of the pseudogene. Thus, the processed pseudogene has a poly(A) tail at its 3’ terminus and is flanked by direct repeats of genomic DNA, as expected for a processed pseudogene. Earlier S 1 nuclease protection studies showed that about 80”,,, of the TS mRNA molecules are poly-

from the stop codon (Deng et al., 1986).

Northern-blot anafyses reveaied the existence of additional RNA molecules that hybridized with the

spond

of TS mRNA to TS

adenylated at minor signals (Geyer and 1985a). Perhaps

molecules

mRNA

moiecuies

and may correthat

are

poly-

downstream polyadenylation Johnson, 1984; Jenh et al.,

the processed

ts pseudogene

corre-

sponds to one ofthese minor mRNA species. Similar observations have been made previously for other processed pseudogenes (Scarpulla and Wu, 1983 ; Lee et al., 1983). Another interesting p~~ssibiiity is that at the time of formation of the pseudogene. the primary site for poiyadcnylation of TS mRNA was at the site observed for the pseudogene. Subsequent mutations in the ts gene led to the creation (or the increase in efficiency) of the polyadenylation signal that is used at the present time. Comparison of the genomic and pseudogene sequences immediately downstream from the stop codon revealed several nt substitutions as well as a difference in the length of the o&go(T) stretch. We have recently found that some of these dowrlstream sequences, including the oligo(T) stretch, are components of the poIyadenylati~~n signal formouseTS mRNA (C.J. Harendza and L.F.J., in preparation). (ci Sequence divergence The overall degree of sequence similarity, when comparing the coding region of the gene with the corresponding region of the pseudogene (ignoring insertions and delet~onsj, is 97.3”,. There are seven differences at the first position, four at the second position and 14 at the third position of the codon in the 921-nt ORF. Similar analyses revealed 95.5’5;, identity in the 5’-noncoding region and 94.4”” in the 3’-untranslated region. It is not surprising that the noncoding regions have diverged to a greater extent than the coding regions since there are few selective pressures on the former, whereas there are severe restraints on changes in the highly conserved ~Perryi~lan et al., 1986) coding region of the functional IS gene, especially at the tirst and second posi-

366

AATAGCAAGTTGAATATATATATCT

GAAAGAAAGAAAGAAAGAAAGAAAGACAGC_CACCTGAATGCTGGGTGTGATGACGCACGCCTrTGATCCCA

CCTCTTGGCATGTGCCTCTTTTTAGACTGATATGRTTCCGAGATCTGGATGCCCGTCT?AGACTGTCTGGC

P

553

GCACTTGGGAGGCAGAGACAGGCGGATTTCTGAGTTCGAGGCCAGCCTGGTCTAC

CCTCATAGCAAACCTTATTCCCGATTCAGACCAAAACCACAATATGTACTTAGAATGC

P

624 (SSlllI)

(AccI)

ACTTVCSAAGACAITAACCTTIGCCTCCA~ ACTTATAAGACATTAACCTTTGCClCCAATTTrTTTACATGGCTCTGTGGTGTAiTATTAGTAAACCACGG ;l@@@@{d

AAAAAAAhAAAAAAGAAAGAAAGMAGAAAGAAAGAAAGAAA

TGCTAATATCTGAAACTTCTGCCTTTCTAGCCTTAArArA~T TGCTAATATCTGAAACTTCTGCCTTTCTAGCCTTAATATATGCTATGTCCTGCCCTGTTGTTAAGACTGCT

TGITTTGTTTTGTTTTGTTTTGTTTT~ATCTCTGGCCAAITCATCA~TG~T~CAAGTGTCCAAGTGACTTT TTATCTCTGGCCAATTCATCAATGTTCCAAGTGTCCAAGTGACTTT

TC~AGG~GATTCAATGCACTTA~GTTT~TrTTGTTTTGTTTTGTTTTGTTTTGTTTTGTTT~GTTTTGTTT TCTAGG~~fiTTCAATGCACTTAGGTTTTTTTTGTTT GTTT GUT GTTTTGTTTT

GTGCCAAT

P 482

411

P

P 311

C?CCCCCTT:CTTTlTATATTCAGGAGTGCTCTAAAGATGAGTGGkAACACAACGATTTACTA~TAGAATA CTCCCCCTTTCTTTTTATATTCAGGAGTGCTCTAAA CACAACGATTTACTAATAGAATA

P ?lfJ

CCCAA

GAGTCCTCCCAGGr~CCTCTGCCTGCCAAGCAAACTTCACTG~~~TGT~TGTAGCATT AACTCTGCCTCtiTGAGTGCTCCCAGCCACCTCTGCCTGCCAAGCAAACTTCACTGTAGTGTtiTGTAGCATT

P 143

GTGCCAATAATAA:AGCAAGTTGAATa:AiAIAT;jTdTATATATATATATATATATATATATATAACCCAA

AA~GTGGGGATRGGGTATGAGTATA~~GGTTATGAGAAC~C~AAAGTTCCAGG~GGTGAGTTTC~~T~ AACGTGGGGATAGGGTATGAGTATACAGGTTATGAGAACACTAAAGTTCCAGGAGGTGAGTTTCTGTtiAAA

(EcoRI) GAATTCAATCTTAAAAGGCACTTATAATAAGATAATACTAAAAG~GAGCAC~~GGA~CCATACACCAGACT SAATTCAATCTTGAAAGGCACTTATAATAAIACACGTGGAGCCATACACCAGACT

72

p

:

P

B 2. Analysis

of the IS pseudogene.

is also included

(Panel B) The sequence

Plus symbols indicate

..r

*

r......

1

to the junction

2

of exons 5 and 6 are compared

positions

of sequence

identity.

the template of the pseudogene.

splice acceptor

to exon sequences, Two potential

correspond

by asterisks.

letters

1 was used for

to intron sequences. Splice site

letters correspond

Splice site 2 is used for normal TS mRNA. Colons indicate

sites are indicated

whereas lower-case

at the with the sequence

of the cDNA and pseudogene

of the final 26 nt of intron 5 to the first 9 nt of exon 6 of the mouse ts gene. Capital

region corresponding

splicing of intron 5. The sequences

of an polyadenylation

,;,v,i ii~troi? '9 . ..t(.tttt?ti.ct~;t~.dy~ciitttttdg(‘Tiil;(;TCAi...'il!)n D; ** **

;"!j"i:i,i!j~:!;

of a possible upstream

the position

of the two sequences.

ProGlyRsyl... . ..G';CCTGCM; CCAGSTGAT... . . . . ..i.i;TrIl;Cdr,jcatttt:~lrCCASljT';AT... . . .. . . . .. . . . . .. ..

, , ,f,l”i~~~i’;l~

the position

correspond

at the 5’ and 3’ ends

Symbol Y indicates

the points of divergence

are the direct repeats

the two nt that flank an intron. Symbol @ indicates

<;I)?Jfi

Fig. 3. Aberrant

signal.

nt substitution.

denote

of the pseudogene.

for reference.

site to the Bsml site. The numbers

sequences

with that ofthe normal gene from the f&RI to the normal gene. The underlined Asterisks

regions of the gene (Deng et al.,

begins at the A of the AUG start codon. The

from the EcoRI site to theAcc1 site in the 3’ flanking region is compared

of the enzyme

ofthe pseudogene

sequence

deduced

sequence

at the UAG stop codon) flanking

of the cDNA

1986). Numbering

or 5’ and 3’ (starting

(P) from

et al., 1986)

of the pseudogene

with that ofthe cDNA (Perryman

(Panel A) The sequence

the Wind111 site to the Sac1 site is compared


368

.

. EXON

6...........

5' ttCCtgtgag~CatttttagCCAGGTGATTTTGTCC~CACTTTGGG

3' GENE

3’

5' PROBE

AABGZ.CACTCTGT -TC(IGTCCACT-CAOOTGTGTGZG.ACCC*

III -A-

III -N-

tions of the codon. It has been estimated that the rate of nt substitutions for neutral mutations is approx. 0.57” per myr (Li et al., 1985). Assuming that the 3’-untranslated regions of both the gene and the pseudogene are under no selective pressure, these regions should diverge about

1 y0 every myr. There-

fore, it appears that the pseudogene about 5.6 myr ago.

was formed

(d) The pseudogene contains sequences at the 3’ end of intron 5 Another hallmark of processed pseudogenes is the absence of intron sequences, in line with the idea that they are derived from mature mRNA molecules (Vanin, 1985; Wilde, 1985). Analysis of the ‘coding region’ of the processed zs pseudogene (Fig. 2A) revealed that in every case but one, the intron sequences were precisely removed. However, we noted that there was a IO-nt insertion between the nt representing the 3’ end of exon 5 and the 5’ end of exon 6. Surprisingly, these inserted nt were identical to the final 10 nt of intron 5 of the mouse ts gene (Fig. 3). Further analysis of the sequence of intron 5 revealed that a pyrimidine stretch followed by an AG are immediately upstream from the 10 nt insertion. This sequence is in reasonable agreement with the consensus sequence for a 3’ splice site (Mount, 1982). These observations strongly suggest that the 10 nt insertion is the result of an error in splicing of the RNA molecule that served as the template for the processed pseudogene.

I

‘;

(e) Alternative splicing of mouse TS mRNA Fig. 4. Sl

nuclease

(A)‘mRNA

protection

ducing cell line) and from a ts had been transiently minigene

(pI3d4)

nuclease

protection

hybridization quantities

assays.

hamster

transfected that

lacks assays

conditions

V79 cell line (lane 3) that

intron

5 (Deng

were

performed

of these mRNAs,

purified

Hybridization

using

aqueous

as well as with tRNA (lane 2). The

to the final 20 nt of intron and

et al., 1989). Sl

(Ausubel et al., 1987) with the indicated

6. The probe was 5’ end-labeled kinase

poly-

for two days with a mouse ts

probe (top of the figure) was a synthetic mentary

Cytoplasmic

was isolated from LLJ3-7 cells (lane 4) (TS-overpro-

with ‘*PO, using polynucleotide

by polyacrylamide

temperature

oligo that was comple-

5 and the first 26 nt of exon gel electrophoresis.

was 37°C. RNA-DNA

hybrids

were

digested

with 400 units of Sl nuclease

analyzed

on an 8”/, polyacrylamide

indicates

the labeled nt and two short vertical lines indicate a site

of Sl digestion.

and the products

sequencing

were

gel. The asterisk

In the first lane the probe was subjected

To determine if alternatively spliced TS mRNA could be detected in mouse cells, Sl nuclease prowere performed. Cytoplasmic tection assays poly(A) + mRNA was isolated from LU3-7 cells and probed with a synthetic oligo that corresponded to the 3’ end of intron 5 and the first 26 nt of exon 6 of the normal mouse ts gene. As a control, mRNA was also isolated from ts- hamster V79 cells that had been transiently transfected with a mouse ts minigene (pI3d4). In this minigene, the regions of the ts gene

to the

A + G cleavage

reactions

signals corresponding spliced TS mRNAs, on the left margin.

of Maxam

to the normally

and Gilbert

(1977). The

(N) and alternatively

as well as undigested

probe,

(A)

are indicated

369

corresponding 5 through Portions

to exons 1 through 3, as well as exons

7, were replaced

by mouse

ts cDNA.

of intron 3 and all of in&on 4 were retained

to ensure high-level expression of the transfected minigene (Deng et al., 1989). Since the minigene lacked intron 5, mRNA derived from the minigene cannot be spliced at the alternate site. Properly S 1-resistant

spliced

TS mRNA

should

result in an

fragment that is 28 nt in length, whereas

RNA that was spliced

at the upstream

site would

result in a 38-m S l-resistant fragment. The protected fragment is 2 nt longer than the position of the in&on-exon

boundary

REFERENCES

since the last 2 nt of exon 5, as

well as the Iast 2 nt of the in&on, are AG (Fig. 3). Fig. 4 shows that, when the mRNA from the transfected cells was analyzed, all of the Sl-resistant DNA was within 1 or 2 nt of the size expected for normally spliced TS mRNA. No Sl-resistant fragments were observed when mRNA from untransfected b cells was analyzed (not shown). However, when mRNA from LU3-7 mouse fibroblasts was analyzed, an additional set of fragments was observed that corresponded to aberrantly spliced TS mRNA. Densitometric analysis of the autoradiogram indicated that the upstream site is used about ten times less frequently than the normal site. The 10-nt insertion would lead to a change in the reading frame of the mRNA and polypeptide chain termination six codons downstream from the start of the insertion (Fig. 31. Since a normal C-terminal region of the enzyme is essential for TS catalytic activity (Au11 et al., 1974), it is clear that the aberrantly spliced RNA molecule is incapable of encoding a functional TS enzyme. The biological significance (if any) of this alternatively spliced RNA and the protein it encodes, remains to be determined.

A&, J.L., Loebie, R.R. and Dunlap, dependent

inactivation

Ausubel,

F.M., Brent, R., Kingston,

J.A.,

Seidman,

Molecular

J.G.

J. Biol.

D.D., Smith,

K.: Current

Protocols

1.Wiley-Interscience,

in

New York,

4.6.7.

M.L., Bussiinger,

mination

R.E., Moore,

and Struhl,

Biology, Suppl.

1987, chapter Birnstiel,

M. and Strub, K.: Transcription

and 3’ end processing:

ter-

the end is in sight! Cell 41

(I985) 349-359. Deng, T., Li, D., Jenh, C.-W. and Johnson, gene for mouse and multipie (1986)

thymidylate

L.F.: Structure

synthase:

transcriptional

start

locations

of the

of introns

sites. J. Biol. Chem.

261

16~-~6~~5.

Deng, T., Li, Y. and Johnson, expression

is stimulated

L.F.: Thymidylate

synthase

by some (but not all)introns.

gene

Nucleic

Acids Res. 17 (1989) 645-658. Geyer,

P.K. and Johnson,

L.F.:

Molecular

of DNA

complementary

messenger

RNA. J. Biol. C&em. 259 (1384) 7X%-7211.

Jenh,

C-H.,

Geyer,

thymidylate

to mouse

cloning

sequences

P.K.

synthase

in an overproducing

and

thymidylate

Johnson,

mRNA content

L.F.:

synthase Control

of

and gene transcription

mouse cell line. Mol. Cell. Biol. 5 (1985a)

2521-2532. Jenh,

C.-H.,

Geyer,

Thymidylate (1985b) Jenh,

P.K.,

synthase

uridine-resistant

Baskin, gene

mouse

F.

and

Johnson,

arnp~i~~ation

ceit

lines.

L.F.:

in Auorodeuxy-

Mol.

Pharmacol.

28

80-85,

C-H.,

Deng, T., Li, D., DeWille,

Mouse

thymidylate

untranslated

synthase

region.

J. and Johnson,

messenger

Proc. Natl. Acad.

RNA

L.F.:

lacks

a 3’

Sci. USA 83 (1986)

8482-8486. Lee, M.G.-S., tionary

Lewis, S.A., Wilde, CD. and Cowan,

history

[?-tubulin

of a muttigene

gene and three

N.J.: Evofu-

family: an expressed

processed

human

pseudogenes.

Cell 33

(1983) 477-487. Li, W.-H.,

Luo,

sequences. Genetics. Loenen,

C.-C.

Pienum,

Maxam,

C-I.:

Evolution

of DNA

Evolutionary

1985, pp. 1-94.

F.R.: Lambda

Charon

vectors

for DNA cloning in recombi-

hosts. Gene 26 (1983) 171-179. E.F. and Sambrook,

Manual.

Harbor,

J.: Molecular

Cold Spring Harbor

Cloning.

I.aboratory,

Cold

NY, 1982.

A.M. and Gilbert,

DNA. Mount,

New York,

and 35) adapted

T., Fritsch,

A Labbratory

Wu,

R.J. (Ed.), Molecular

and Blattner,

nation-deficient

Spring

and

In Macfntyre,

W.A.M.

Maniatis,

We thank Drs. Paul Fuerst, Scott Falkenthal and Elio Vanin for helpful discussions. These studies were supported by grant GM29356 from the National Institute for General Medical Sciences and by grant CA16058 from the National Cancer Institute.

synthetase.

Chcm. 249 (1974) 1167-l 172.

(Ch32, 33,34,

ACKNOWLEDGEMENTS

R.B.: The carbuxypeptidase-

of thymidylate

Proc. Nat]. Acad.

S.M.: A catalog

W.: A new method

for sequencing

Sci. USA 74 (1977) 560-564.

of splice junction

sequences.

Nucleic

Acids Res. 10 (1982) 459-472. Perryman, Johnson, synthase enzyme.

S.M.,

Rossana,

L.F.: Sequence reveals

striking

G.,

Deng,

of a cDNA similarity

T., Vanin,

E.F.

and

For mouse thymidyfate with

Mol. Biol. Evoi. 3 (1985) 3 I3-321.

the prokaryotic

370

Rossana, thetase

C., Rao, L.C. and Johnson, overproduction

mouse fibroblasts. Sanger, F., Nicklen, chain-terminating

in

L.F.: Thymidylate

Mol. Cell. Biol. 2 (1982) S. and Cot&on, inhibitors.

syn-

5-fluorodeoxyuridine-resistant

Vanin, E.F.: Processed with

Proc. Natl. Acad. Sci. USA 74

(1977) 5463-5467. Scarpulla,

R.C. and Wu, R.: Nonallehc

of the cyto-

family of the rat may arise through

Wilde, CD.:

dif-

RNAs. Cell 32 (1983) 473-482. pseudogenes:

tion. Annu. Rev. Genet. 323-352.

members

c multigene

ferent messenger

1118-I125.

A.R.: DNA sequencing

chrome

Pseudogenes.

characteristics

and evolu-

19 (1985) 253-272. CRC Crit. Rev. Biochem.

19 (1985)