Proc. Natd. Acad. Sci. USAVol. 91, pp. 6388-6392, July 1994Biochemistry

Sak, a murine protein-serine/threonine kinase that is related to theDrosophila polo kinase and involved in cell proliferationCAROL FODE*t, BENNY MOTRO**, SHIDA YOUSEFI*, MIKE HEFFERNAN*t, AND JAMES W. DENNIS*t*Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, ON Canada M5G 1X5; and tDepartment of Medical Genetics,University of Toronto, Toronto, ON Canada M5S 1A8

Communicated by Henry Friesen, March 4, 1994

ABSTRACT We have isolated murine cDNAs encodingtwo isoforms of a putative protein-serine/threonine kinase,designated Sak-a and Sak-b, which differ in their noncatalyticC-terminal ends. The kinase domain of Sak is related to thecatalyic domains of the Drosophila polo, Saccharomyces cere-vuide CDC5, and murine Snk and Pik kinases, a family ofproteins for which a role in controlling cell proliferation hasbeen established (polo, CDC5) or implicated (Snk, Pik). North-ern and in situ RNA analyses of Sak gene expression in mouseembryos and adult tissues revealed that expression was asso-ciated with mitotic and meiotic cell division. In addition, duringembryogenesis, Sak expression was prominent in the respira-tory and olfactory mucosa. The pattern of Sak expression andits sequence homology with thepolo gene family suggest that theSak kinase may play a role in cell proliferation. In support ofthis, cell growth was suppressed by expression of a Sak-aantisesefragment in CHO cells.

The Drosophila gene polo and the Saccharomyces cerevisiaegene CDC5 encode protein-serine/threonine kinases whichare highly related in sequence (47% identity in the catalyticdomain) and in function (1, 2). Phenotypic analysis ofmutantalleles ofpolo and CDC5 indicates that both are required forthe proper formation and function of the spindle duringmitotic and meiotic cell divisions (1, 3-5). Two relatedmurine genes, Snk (6) and Plk (7), have been cloned and mayalso play a role in regulating cell proliferation as shown by cellcycle-dependent changes in their transcription levels and acorrelation between the tissue-specific pattern ofPik expres-sion and active cell division. We report here the cloning of arelated murine gene, Sak, which encodes two isoforms of aputative serine/threonine kinase.§ Determination of Sak ex-pression during embryogenesis and in adult tissues revealeda correlation with mitotic activity. Furthermore, expressionof an antisense Sak-a transcript inhibited the proliferation ofCHO cells.

MATERIALS AND METHODScDNA Cloning and Molecular Analysis. A cDNA library

was prepared in pCDM8 (Invitrogen) using poly(A)+ RNAfrom D33W25, a murine lymphoid tumor cell line (8). CHOPcells, a subline of CHO cells which expresses the polyomalarge tumor C() antigen, were transiently transfected with thecDNA library by a modified DEAE-dextran procedure (9).After 72 hr, transfected cells were placed in selection mediumcontaining the cytotoxic lectin wheat germ agglutinin (WGA,50 pg/ml), and 3 weeks later plasmid DNA was recoveredfrom the WGA-resistant clones by the Hirt procedure (10).Radiolabeled probes were prepared by random priming usingthe T7 Quick Prime kit (Pharmacia), and the dideoxynucle-otide DNA sequencing method was done according to theinstructions accompanying the Sequenase kit (United States

Biochemical). cDNA sequences were compared with theGenBank/EMBL databases (GenBank release 71) by theFASTA program and amino acid alignments were done by theMALIGNED program, both from the University of WisconsinGenetics Computer Group.In Situ RNA Analysis. In situ RNA hybridization was

carried out on 8- to 10-pm cryostat sections (11). Adjacentsections were probed with Sak-a and Sak-b antisense RNAprobes. Control probes included a Sak-a sense probe, theneuron-specific marker SCG10, c-kit, and SA; the latter threeprobes gave specific nonoverlapping patterns (data notshown). The 1.5-kb fragment 17S, unique to the Sak-atranscript, was subcloned into pBluescript SK (Stratagene)and antisense probes were generated by T7 RNA polymerase(probe 1, Fig. 1). To generate a Sak-b-specific probe, the 3'EcoRI-Xho I fiagment (1.5 kb) was transcribed with T3 RNApolymerase (probe 3, Fig. 1). Post-hybridization washingsincluded treatment with RNase A (50 pg/ml) at 370C or 42DCfor 30 min and two stringent washes of 20 min each at 600Cin 0.lx standard saline citrate (15 mM NaCl/1.5 mM sodiumcitrate, pH 7). The slides were dipped in Kodak NTB-2emulsion, exposed for 4-6 days, developed, and stained withtoluidine blue.

Sense and Antisense Sak Expression Vectors. The hygromy-cin-resistance gene driven by the j-actin promoter wasexcised from the pSP-3-hygroA vector (gift from CeciliaMoens and Dr. J. Rossant, Mount Sinai Hospital) byEcoRV/Pvu II digestion and cloned into the Sca I site of pCDM8vectors containing either no insert (control) or sense andantisense Sak cDNA fragments under the control of thecytomegalovirus promoter. For the colony-formation assay,30 pg ofeach plasmid was linearized by Cla I, Hindu, or SacI and electroporated into 5 x 106 CHO cells. Colonies wereselected in a minimal essential medium with 10%o fetal bovineserum and hygromycin B at 400 pg/ml for 10 days. Colonieswere stained in 0.06% methylene blue and 1.25% glutaralde-hyde in phosphate-buffered saline for 1 hr and then counted.

RESULTSIsolati of Murine Sak cDNAs. To clone genes regulating

sialylation, a murine lymphoid cDNA library was transfectedintoCHOP cells, and 3 days later the sialic acid-binding lectinWGA was added to the culture medium. From this screen,four WGA-resistant CHOP clones retained episomal plasmidcarrying an identical 1.5-kb cDNA insert (clone 17S, Fig.1A), corresponding to a partial fragment of the Sak-a gene(see below) in an antisense orientation. However, the rela-tionship between the undersialylated phenotype and Sakexpression remains to be- determined, since transient trans-fections of CHOP cells with an expression vector containing

Abbreviations: WGA, wheat germ agglutinun; dpc, day(s) postcon-ception.tPresent address: Department of Life Sciences, Bar-Ilan University,Ramat-Gan 52-900, Israel.§The sequences reported in this paper have been deposited in theGenBank data base (accession nos. L29479 and L29480).

6388

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

1, 2

020

Biochemistry: Fode et al.

Kinase Domain

ATG TAA

Sak-a I I

AATAAA AATAAA

PROBE 1(17S)PROBE 2

TAA..........

* -"..... -. ,,''.: PROBE 3

ATGSak-b

500 BP

B

1 1 GTCCT -ATGGGCGCCAAGCCT GCTGGCTT C TCG>GACGCTSCCGTCGGGAGGGGACT GCGAGAAGGC CGAGC CC CGGGCG CCGG CGG C T C^GGGACATGGCGG CG T GCAT C1 U A A

2 21 GGAAG>GATCGAGCT TTA.T T-GAT CTAC TCGG TAAAGGAT CAT T TGCTGGTG TCTACAGAGC TGAGT CCA T ACACwAC T GGT w GAA T - GCAAT- ZAAAA '

6 G E R I E D F K V G N L L G K G S F A G YV R A E S I H T , A K M331 GATAGATAAGAAAGCCATGTACAAAGCTGGAT GGTACAGAGAGTCCAAT GAGGTGAALTA CA T TGCCAGT TGuAACAC CC CT CTGT ,- -TGGAG C - C A TAA TT AC-43 . O K K A M Y K A G M V Q R V Q N E V K ' H C O L K H P S V -I N t :

4 41 T TGAAAT. ACAT TAT GTCTACC TGGT ATTGGAT GTGCCACATGGAGAATGAACAGATATC TGAAGAACAGAAT GAAGC CT T C T CAAAAGGGAG CTAGGCA C79 E D N N Y V Y L V L E C H N G E M N R Y L K N R M K 0 F S R A R

5S 1 T TCATGCACCAG T TATCACAGGATGTT ATATCT TCAT TCTCATGOCATAT TGCACCGGGACCTCACACT CTCTAACATCT TAC TTACGCGGAA TATGAACATAAAAA -

116 F M H Q I I T G M L Y L H S H G I L H R O L T L S N I L N N : K6 61 TGCTrA CT TTGG CTAGCA4XCAG TTGAATATGCCACATGAAAGCAC TATACAC TCTGT GGGACTCC TAA TT ATAT T TCACCAGAAA T - GAACo-C5 AAG TG&CA CA T153 A D F G L A T O L N P H E K H Y T L C G T P N Y r S P F ' A. D S A -

7 71 GACTTGAATCTGATAT TTGGT CAT TG&GCT GTA TG TC TTATACGT TACT TA T TGGAAGACCACC TT T TGACA CTGACACAGT CAAGAA CA4A - TGAA CAMAG TAG T C CT189 L E S D I W S L G C S Y T L L I G R P P F 0 T D T V K N K V V

881 GCAGATTATGATGCCAGCCT TTTTGTCACGAGAGGCCCAGGACCT TATCCACCAGT TAC TTCGTAGAAACCC TGCAGATCGG TTAAG-TC-G.JTC T TC TGT GT TGGAOCA226 A Y E M P A F L S R E A C 3 L I H Q L L R R N P A R L S S S V L 0 "

gS 1 T CC TTTCA TGTCACGAAA TCCSTTCACCAAAGAGTAAAACGTAGGGAC TGTAGAGGCGT CAATGGATAGTGGGCA TGCTACAC TT TCCACAACAA T TACAGCCT CTT CT G263 P F S R N P P K S |K 0 V T E 0 S M 0 S G JHA L S T - - A S S GMI0 GTACCAGT TTGAGTGGCAGCCTACTTGACAGAAGCTT T TGGT TOGTCACCACTTCCAATA ATTAC TGTAT TTCAAAAATAAAAA T TCAAG TGACT TT TC T TCA299 T S L S G S L L R L L V G O P L P N K I T V F Q K N K N S S F S SZ1211 GGAGATGGAAGTAATTT'TTGTACTCAATOGOGAASATCCAGAACAAGAAGCTAATAGTAGGGGACGGGGGAGAGTGAT TGAAGATGCAGAAGAGAGGCCGCAT TCTCGA TA336 G 0 G S N F C T Q W G N P E Q E A N S R G R G R V I E D A E R S R1321 CCTGCGCAGAGCTCATTCCTCTGATAGAGCCAGCCCCTCTAATCAGTCTCGAGCAAAAACATACTCAGTAGAACGTTGTCACTCAGTAGAAATGC TT'OAAAGCCTAGAA373 . R R A H S S D R A S P S N Q S R A K T Y S V H S K P R

1 431 GATCACTGGATGAAAATCA.ACACAGTTCCAATCATCAT TlGTCTAGGAAAAACTCGTT T TCCATTTGCAGACCAGACACCT CAGA TGC;AAA TOGT ACAGCAG .GGC;TTTGuG409 S L E N Q H S S N H H C G K T P F P F A 0 0 - P 0 M E f W F G

1541 AATCTGCAAATiAAT&CTCATTTAGGAGAAACTAATGAGCACCACACCGTTAGCCCAAACAGAGATTTCCAGGACTATCCAGA TTGCAGGA OACGTACGAAACGCT-G446 N N A H L G E T N E H H T S P N R F D Y P L C R A

1 65 1 GA CTGACACGAGAGCCAwGCAAGAAT GCT GAT AC TT C TGCCAA TG T TCAT&CT GTAAAGCAGCT∾TGCCAT GAAATACA TGAGTGCACA TC ACCATAAGCC T5AI:G TCLA483 T 0 T R A S K N AD s A N V H A V K L S A M K Y S A Pi K VSN

1 761 TGCCACAGGAGCCGGGCCTACATCCT CAT TCTGAACAAAGCAAGAATAGAAGTATGGAG~TCGACACTGGGTTACCAGAAACCTACCT TAAGAAG TA T-ACAT CTCCuT-G519 P Q E P G L H P H S E Q S K N R S M E S LG Y K P T L R S S P

1 871 ATTGCTCACAGAT TAAGCAAT CAGACAGAAAACCAAAAAGGCTGTGGTGAGCAT CCTTt3AT TCAiAGAGGAOl:;TGTGTGGAGCT T C TGAAGAGT&T5CliTC .GAAGG556 I A H R L K P I R a K T K K A V S 1 L S E V C V E L Q F A S r

1 981 ATATGTGAAAGAAGT;CT TCAGATATCGAG3TGATOGt;ACTATGATCACTGTTTATTACCCGAACGATGGAAGAGGCT TTCC TCT TGCTGACAGACC TCCCT TGCCTACTG593 Y V K E V L I S S D G T b I T V Y Y P N G R G F P L A P P 3 - v - JZ2091 ACAACA TCAGTAGGTACA.GCT T TGACAA TCTACCAGAAAAA TACTGGCGGAAAT AlTCAG TAT GCT TCCA&A T TCA TTCAGCT AG TAAGA T C-AAAA C- C CCAAAA T CA C629 I S R Y S F D N L P E K Y W R K Y Y A S R I _ V R S K K

2 201 TAT TTTACAAGATATGCTAAATTATTT TGATGGAAAATTGT CCTGGTGCTGAT TT CGAAGT TTGGT T TTATGATGGA&GCAAAATACA TAAAAC7GAAAAT TTAAT TCA666 Y F T R Y A K C I L E N S P G A D F E V W F Y G A K H E2311I CATAAT TGAGAAAACAGGGATATCT TATALATT TAAAAAATGAAAATGAAGT TACCAGCCTGAAAGAGGAAGTAAAAGTATATATGGACCATG 4TAA TGAGGGTCACCG TA703 I E K T I S Y N L K N E N E V T S L K E E V K V Y M D rj A N - " R2421 T TTGCTTGTCACTGGAATCTGTAAT CT CTGAGGAGGAAAAGAGAAGCAGGGGTTCT TCAT TCT TCCCTATAATCGTAGGAAGAAAACCGTGTAA TATAUT T CAGCTAAA739 C L S L E S V I S E E E K R S R G S S F F P I I V G 8R K P G 4 T S S P K|

2 531 GCC T TATCAGCTCCTCCT GT GGCC CAAGC TGCTGT AAGGGAGAGCAGGCG TCAGCAAGCAGACT GAGCG TGAAT AG TGCCGC TT TC C CCA CACAG TCC CCAGGAC TCAG776 A L S A P P V D P S C C K G E Q A S A S |RL S V N S A A F P T Q1 S P G L S

2 641 T CCTTCCACTGTGACAGT TG.AGGACT TGGCCACACAGCGAC TGCCACAGGAACAGGCGT C TCT TCAAGT CT TCCTAAAT C TGCA CAGC T T-TGAAA TCTGT T TT T G TGA2 751 AAAA T&GTTGGT TGGGCTACACAGC TAACTAGCGGAGCT GTGTGGGTTCAGT T T uTGATGGGTCACAGT TGGTTG T CCAGGCAGGAG T A CT TCTCATCGAG T TACACA T t.A849 N V G W A r Q L T S G A V W V O F N D G S O L V V Q A G v 5 SS T S

Z861 CCAGATGGTCAGACAACTAGGTATGGAAAATGAAAAAT TACCTGAATACATCAAACAGAAAT TACAGTGT CT TTCTT CCATCCT T C TGAT- -T 'TCTAAT CCAAGTCrrsa6 P D G Q T T R Y G E N E K L P E Y I K a K L 5 C --S S , 4,. FShP-

2971t TAAT TTYTCAGTAT TTAAGTCTCGAUGT CTATATT TAAT AAATGACTT TTrTGGCTGGCT'TT'AAGTAAGTGA TT T TTTAAATTTrAC T ''.TAC T +CAGAAAGCCT Ty AT923 N F Q*

3081 T TAAACAGAiT TT TAATATACACAATAAAAATATkATAAGAAAACAATAAAAT T TCAGTTACCTALATATAGTGGTCATAAGiGCTAGGACA /C-AAT7~T TGCT CCAAGCA!T3191 GTAATCCTTCAAAGlT TTGTGCTCCTATGTTTGTAT TGAACTAAGT TGTGTATGGCTTGTTTGT TT TTGT TAT T TTCT TTACTAATAAGACAT TGAGAATCACGGACAAAA3301 CATAGTTTTCA.ATTT lTTGAATGTGTAAATAATGTATTATAAGCAATATGTAAATGTGTATATT T TATAT TTAT TTT TATAG;CACT TGTGT ^TGATAAGAT T TC-GCAAA -

3 4 11 ACAT TT T ATAAAA TAAACACAGTGC:T AAGiTTtT TC C TT 3 44 7

660

89

880

990

3c

43C

409

445

-492

870

980

592

2090829

565

'9C39

'53u

2640812

549

4 30

0960

885

308C92531 9C330034 0

c145 4 AGGTAT TCACCCACfuAAGCATGT CAAT GT T TTAACT TCATTAAACACCAAACAGCCAATAGT TAAGGAT CTT TTGAAAGACCGT AAA TGACTGAGCAGTA TAAGGA417 R Y S P T K S N V N V L T S L N T K P . v K D K D- ' - - v e K D

1564 TAATCTTTTAAACTTATTGAACAAT0TTGATCGCTAA 1600453 N L L N L L N K F 0 R 465

this antisense fragment did not result in a high frequency ofWGA-resistant colonies (data not shown).

Additional cDNA clones were isolated from the lymphoidcDNA library by using the 17S partial cDNA fragment as aprobe (probe 1, Fig. 1A), and subsequently full-length cloneswere obtained by screening the library with a more-5' fiag-ment (probe 2, Fig. 1A). Sequence analysis of the isolatedclones identified two transcriptional units, Sak-a and Sak-b(Fig. 1). The Sak-a and Sak-b cDNAs most likely representalternatively spliced forms of the gene, given that the se-quences diverge after an AG dinucleotide (nt 1456 and 1457),a sequence frequently found 5' to splice donor sites (12). A205-nt 5' untranslated region precedes the predicted start site

FIG. 1. Schematic of the SakcDNAs and nucleotide sequence.(A) The open reading frames (blackboxes) ofthe Sak-a and Sak-b tran-scriptional units are flanked byATG start and TAA stop codons.The point of sequence divergenceis marked in Sak-b by the stripedbox. The position of the kinasedomain is indicated. Probe 1 cor-responds to the insert 17S and isspecific to Sak-a. Probe 2 was usedto screen the library for full-lengthcDNAclones. Probe 3 is specific tothe 3' untranslated region of theSak-b transcript. (B) Nucleotidesequence of the Sak-a transcrip-tional unit and the deduced aminoacid sequence. PEST amino acidsequences are boxed, with the mid-dle basic residue breaking the se-quence in two PEST regions. Con-sensus polyadenylylation signalsare underlined. (C) Nucleotide andpredicted amino acid sequences ofthe coding region of Sak-b, begin-ning from the AG dinucleotide thatmarks the point of sequence diver-gence from Sak-a.

of translation, with the AACAIQG sequence including theinitiation site matching closely the Kozak consensus forinitiator methionine codons (13). The Sak-a transcriptionalunit encodes a 925-aa protein of 103 kDa, and Sak-b encodesa protein of 464 aa and 53 kDa. The first 416 aa of bothproteins are identical; the predicted Sak-a and Sak-b proteinshave different C-terminal regions of 509 aa (Fig. 1B) and 48aa (Fig. 1C), respectively.The 205-nt 5' untranslated region of Sak is 71% G+C, and

computer folding of this region suggests that several ener-getically favorable secondary structures are possible (i.e., 8G= -89 kcal), which may reduce translational efficiency (14).In Sak-a clones, 3' untranslated regions of52 nt or 464 ntwere

A

Proc. Natl. Acad. Sci. USA 91 (1994) 6389

-

Dow

nloa

ded

by g

uest

on

Mar

ch 1

1, 2

020

Proc. Nati. Acad. Sci. USA 91 (1994)

identified, containing a consensus polyadenylylation signal(AATAAA) 22 nt and 20 nt upstream of a poly(A) tract,respectively (Fig. 1B). This region is A+T-rich and containsthree ATTTA sequences, features which have been shown todecrease the intracellular half-lives of transcripts (15). Thesequence ofthe 3' untranslated region of the Sak-b transcrip-tional unit is ":2 kb long and lacks a poly(A) consensus signaland tail, suggesting that the corresponding cDNA clones maynot represent a full-length transcript (sequence not shown).The Sak Protein Is Related to the polo Subfamily of Protein-

Serine/Threonine Kinases. The N-terminal region common toboth Sak-a and Sak-b proteins has significant homology to thepolo subfamily of serine/threonine kinases, including theDrosophila polo kinase (1), the murine Snk protein (6), theCDC5 kinase of S. cerevisiae (2), and the murine Plk protein(7) (42%, 41%, 39%6, and 37% identity, respectively). Inaddition, these proteins have a common structural organiza-tion, with the kinase domain located in the N terminus,followed by a C-terminal domain ofunknown function. Giventhat Sak is most closely related to the polo subfamily ofkinases and is the third murine member ofthis group, we havenamed it Sak, for Snk/Plk-akin kinase.An alignment of the kinase domains of Sak and the polo

subfamily members is shown in Fig. 2. The original four polosubfamily members contain the amino acid sequence GXG-GFAXC within subdomain I, with both the alanine andcysteine residues being uncommon in these positions. TheSak kinase contains the alanine residue of this motif, but thecysteine is replaced with valine, a residue more frequentlyfound at this position in other serine/threonine kinases. TheSak kinase domain also diverges significantly from othermembers of the polo subfamily in subdomain VI-B. Withinthis subdomain, the threonine residue in the Sak sequenceDLTLSN is in a position normally occupied by a lysine, notonly in proteins within the polo subfamily, but in the majorityof serine/threonine kinases. The serine residue in this se-quence is replaced by a glycine in all other members-of thepolo subfamily. Sak also has the sequence GTPNYISPE insubdomain VIII, which matches closely with the GTXXYX-APE consensus sequence characteristic of serine/threoninekinases (16).The C-terminal domains of Sak-a and Sak-b showed no

significant similarities to each other or to any sequencesfound in the databank. In addition, a 30-aa homology domainidentified in the C terminus of the polo, CDC5, PLk, and Snk

IINNL~~~~~~~~~~~K~~~S~~~~ ~ ~ ~ N~~~~~II~;Q 14III:11[KKC F T(IrIll.L PKi[

i N:K 1)YHRGHFLFL31 NrK

. K VII' PDYJRGQLF Kt, C: D F.K-n*

HH'NYIEY LKT ~ I~i!IVF

OF 1Y KH RS 1A iT KfL(DINP K

1

lF fNl N TVY I V

IV

.L.IVjI; UK T!;I~L K,-31;VP L L V;I I QNVMEL R!~EF R PI IS VLINV fi SI L CVM I.

,; f-NY141 I Y IEL- BkCPi;j K ![ Y IA , I LIVL0J1E-.[.F8HY K Y Y3 CSI. ;F 3K W'Tr .1NVXYr3l\ G!PLRO 2 ;rJ,.Is .3 C!*|tT !t' }.; r,-;F tF

V

;VK z i.| ' 0'''uPJ17 ;..AA3 l N LAMt

T' ArLK!~-)TY I Y.- F';.'IIR 1Tm SVLii.itl..,..SLVlTM]R3ir

Y. I rP: !.' T v 'K l FI:KH I t s.~~~~~~I F

FiG. 2. Amino acid alignment of the kinase domains of Sak,Drosophila polo, murine Pik, mudne Snk, and yeast CDC5. Numbersat right refer to amino acid position. Amino acids common to all fivekinases are higghted in black, those common to 4 are in dark gray,and those common to 3 are in light gray. The protein kinasesubdomains I-XI, as defined by Hanks (16), are indicated.

kinases is absent from both of the predicted Sak proteins (7).Three "PEST" regions, each containing two sequences intandem in Sak-a and one in Sak-b, were identified in theC-terminal region of the proteins (Fig. 2). PEST sequencesare rich in proline, serine, threonine, aspartate, and gluta-mate residues and are flanked on either side by a basic aminoacid. They are found in a number of proteins with shortintracellular half-lives and appear to contribute directly tomessage instability (17).

Tissue-Specific Expression of Sak. Northern analysis ofadult tissues showed Sak-a expression to be most abundantin the testis compared with all other tissues examined (Fig. 3).Transcript levels in the testis were comparatively low at day8 (i.e., before meiosis) and increased with age in a mannerthat reflected the increase in meiotic activity occurring atthese time points. Interestingly, a 4.5-kb transcript wasdetected at day 8, whereas a 4-kb transcript was predominantin 3-week and adult testis (Fig. 3). Transcripts of 4.5 and 4.0kb were also detected at lower levels in the spleen, thymus,ovary, and 13.5-dpc embryonic RNA and not in the heart,liver, kidney, or brain. Northern blots probed with sequencesunique to the 3' untranslated region of Sak-b (probe 3, Fig. 1)identified transcripts of 3.5 and 4 kb that were expressed atseveralfold lower levels than Sak-a but with the same tissue-specific patterns, including the stage-specific change in ex-pression level and transcript size seen in the testis uponmaturation (data not shown).

Spatial LocI ation of Sak Transcripts by in Situ Hybrid-ization. To analyze Sak expression during embryonic devel-opment, 35S-labeled antisense probes specific to the Sak-a(probe 1) or Sak-b (probe 3, Fig. 1) transcripts were hybrid-ized in situ to sections of embryos at various gestationalstages. The expression patterns of Sak-a and Sak-b weresimilar, but the relative amount of transcripts was muchhigher for Sak-a.At the 7.5-dpc primitive-streak stage, Sak transcripts were

evenly distributed in both embryonic and extraembryonictissues, and much lower levels were detected within thematernal decidua (Fig. 4B). With the onset of organogenesis,Sak expression became more restricted in a manner thatreflected the regionalization of proliferating zones. For ex-ample, Sak expression in the central nervous system wasconfined to the ventricular zones of the brain and spinal cordat all stages ofdevelopment ofthese structures [e.g., 11.5 dpc(Fig. 4D) and 17.5 dpc (Fig. 4F)]. During embryogenesis,expression of Sak was detected within many organs duringtheir proliferative stages, including the skin, liver, thymus,small intestine, and the cortical layer of the kidney (Fig. 4 D

Fio. 3. Northern analysis of Sak-a expression in murine tissues.Total RNA (10 pa) from mouse tissues was separated by agarose gelelectrophoresis, transferred to a nylon membrane, and probed witheitheraSak-a [probe 1(17S), Fig. 1] ora 3-actinprobe. The rigtmsttwo lanes of testis RNA from 8-day and adult mice more cearlydemonstrate the change in Sak-a transcript size (4.5kb to 4.0 kb) thatoccurs following pubertal maturation. The embryo RNA was ob-tained at 13.5 days postconception (dpc).

6390 Biochemistry: Fode et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

1, 2

020

Proc. Nati. Acad. Sci. USA 91 (1994) 6391

A *~~~d

Mai

.I>ss

,

oldie

IN .,

FIG. 4. In situ RNA localiza-tion of Sak-a expression during

SCmurine embryonic development.(A and B) Bright- and darkfleld 2

m-photomicrographs of sagittal sec-tions through a7.5-dpc embryo em-bedded in the uterus, hybridized C

with a Sak-a probe. (C and D)Bright- and darkfield photomicro-graphs of sagittal sections throughan 11.5-dpc embryo hybridized .witha Sak-aprobe.(E,K, MandF,L, N) Bright- and darkflield photo-micrographs of sagittal sections "F Mthrough a 17.5-dpc embryo hybrid-ized with a Sak-a probe. Arrow-heads in K mark the border be-tween the olfactory and respiratory dVepithelia. (G-J) Brightfield photo-micrograph(G) ofa coronal section E __through a 13.5-dpc embryo withadjacent sections probed withSak-a (H), SCG10 (I), and Sak-asense control (J). 4v, Fourth ven-tricle; ac, amniotc cavity; aq, aq-ueduct; bi, basal layer; de, de- mmcidua; ec, ectoplacental cone; es,esophagus; ex, exocoelom; gu, gut;he, heart; Jo, Jacobson's organ; ki,kidney; la, larynx; li, liver; lu, lung;lv, lateral ventricle; mp, maxillarypalate; nc, nasal cavity; np, nasopharynx; ns, nasal septum; ob, olfactory bulb; oe, olfactory epithelium; ol, olfactory pit; op, oropharynx; re,respiratory epitheium; sc, spinal cord; st, stomach; tb, turbinate bone; th, thymus; to, tongue; tr, trachea; ur, urogenital ridge.

and F). In the adult small intestine, Sak expression wasrestricted to the region at the base of the crypts, where celldivision was occurring, and could not be detected in theepithelium lining the villi (Fig. 5F).

Interestingly, Sak transcripts were particularly prominentduring embryogenesis in the nasal cavity, within both theolfactory and the nasal mucosa (Fig. 4D and F). To examineexpression in the olfactory epithelium more closely, consec-utive sections cut through the head of a 13.5-dpc embryowere hybridized with Sak (Fig. 4H) and SCG10 (Fig. 41)probes. In transverse sections through the head of 13.5-dpcembryos, the Sak probe hybridized to a region basal to theolfactory neuron layer identified by the neuron-specificSCG10 probe (18), suggesting that Sak is expressed in thebasal layer ofthe olfactory epithelium and not in the epithelialor neuronal layers. In these sections, Sakexpression was alsodetected in the vomeronasal organ, which contains an acces-sory olfactory epithelium resembling the olfactory mucosa.Control sections hybridized with a Sak sense probe did notshow expression in these regions, confirming the specificityof this expression pattern (Fig. 4J ). Expression in the basalcell layer of the olfactory epithelium persisted until at least17.5 dpc (Fig. 4F and L). In addition, at 17.5 dpc, high levelsof Sak transcripts were detected in the epithelium lining thenonneuronal portion of the nasal cavity (Fig. 4 F and L), anexpression pattern which continued into the similarly struc-tured epithelial layer of the upper respiratory tract, bronchi,and large bronchioles (Fig. 4 F and N).

In sections through the testis the highest levels ofexpressionwere in the meiotically dividing spermatocytes, and muchlower levels were seen in the spermatogonia, the immediatepostmeiotic round spermatids, and mature spermatozoa (Fig.SD). No expression was detected in the epithelial Sertoli cells

or stromal Leydig cells. Examination of gonadotropin-superovulated ovaries at various stages of the cycle revealedhigh levels of Sak expression in oocytes during all phases ofgrowth. Lower levels of Sak transcripts were also detectedwithin the granulosa cells after the initiation of folliculargrowth by pregnant-mare serum gonadotropins, with expres-sion levels correlating with the proliferative state.Er n of a Sak-a Anene F ment S Cell

Prolifertn. To determine whether Sak expression was re-quired for cell growth, we used an antisense approach toexamine the effect ofdecreasing expression of this gene on theability of CHO cells to form colonies. Colony formation hasbeen used to assess the ability of cyclin D1 (19), as well as thep53 (20) and retinoblastoma (21) tumor-suppressor genes, tosuppress cell growth. Growth suppression was measured bytransfecting CHO cells with a vector expressing both theantisense Sak-a fragment (17S) and the hygromycin B-resis-tance gene. Sak-a antisense expression decreased the efficiencyofcolony formation in the presence ofhygromycin B by a factorof 20 compared with cells transfected with vector containingonly the hygromycin-resistance gene, whereas expression ofsense Sak-a or Sak-b had little or no effect on colony formation(Table 1). In addition, digesting the antisense constructbetweenthe cytomegalovirus promoter and the 17S insert with HidIll(Table 1) or Sac I (data not shown) restored colony formationto normal levels,- demonstrating that antisense expression isresponsible for the inhibition of cell growth.

DISCUSSIONIn this report, we describe the cloning of Sak, a murineprotein kinase, during a screen designed to isolate cDNAscapable ofconferring resistance to the cytotoxic lectin WGA.The 1.5-kb cDNA originally isolated corresponded to the 3'

Biochemistry: Fode et aL

Dow

nloa

ded

by g

uest

on

Mar

ch 1

1, 2

020

Proc. Natl. Acad. Sci. USA 91 (1994)

N~~~~~~~"e

x~

FIG. 5. In situRNA localization ofSak expression in the adult gutand gonads. All sections were hybridized with a Sak-a-specificprobe. (A and B) Bright- and darkfield photomicrographs of sectionsthrough the ovary of a superovulated female induced by pregnant-mare serum gonadotropins. gr, Granulosa cell; oo, oocyte. (C andD)Bright- and darkfield photomicrographs of sections through an adulttestis. Le, Leydig cell; sg, spermatogonia; sp, spermatocyte; st,spermatids. (E and F) Bright- and darkfield photomicrographs ofsections through the adult gut. cr, Crypt cells; il, intestinal lumen;sm, smooth muscle layer; vi, villi.

translated region of the Sak-a transcriptional unit in anantisense orientation. Although expression of this antisensefragment failed to directly confer WGA resistance, its ex-pression did suppress growth of CHO cells, suggesting thatthe Sak-a protein is required for cell proliferation. It ispossible that Sak-a antisense expression initially had a cy-tostatic effect that protected the cells from WGA toxicity andallowed time for somatic mutations affecting sialylation lev-els to occur during the 3-week selection period. Interestingly,in a subsequent experiment, plasmids expressing c-src an-tisense were isolated from two independent WGA-resistantcolonies. Reduction ofc-src expression levels using antisenseconstructs has been shown to suppress cell growth of rat andmouse fibroblasts, suggesting that this cloning system mayhave some ability to select for genes that regulate cellproliferation (22).

Table 1. Suppression of colony formation by Sak-a antisense% hygromycin B-

Construct Digestion resistant colonies n

No Sakinsert Cla I 100 3Sak-a antisense Cla I 4.2 + 1.6 3Sak-a antisense HindIII 87 1Sak-a sense Cla I 85 1Sak-b sense Cla I 82 1

CHO cells were transfected with pCDM8 vectors containingcDNA inserts under the control ofthe cytomegalovirus promoter andthe hygromycin-resistance (hyg) cassette. Cla I linearized the vector,and HindIII cut the vector between the cytomegalovirus promoterand the Sak antisense cDNA to prevent expression of this fragment.Transfected cells were selected in hygromycmi B for 10 days and thenenumerated. Results are expressed as a percentage of colony for-mation in cultures transfected with the construct with no Sak insertwhich produced 318, 239, and 159 colonies per 5 x 105 transfectedcells in n = 3 independent experiments.

A detailed study of Sak expression by Northern and in situRNA hybridization analyses has shown that Sak expressioncorrelates with mitotic and meiotic activity in embryonic andadult tissues. For example, Sak transcripts in the embryoniccentral nervous system are confined to the ventricular zones,where neuroblasts are actively dividing, and are not detectedin the intermediate or marginal zones, where the postmitoticneurons are located. In adult mice, Sak is expressed in tissueswith a mitotic component, including hematopoietic tissues,the stem cells of the intestinal crypt, the granulosa cells ofdeveloping ovarian follicles, and the matrix, or growth re-gion, of the hair follicle (data not shown). In situ RNAanalysis revealed high levels of Sak transcripts in meioticspermatocytes and oocytes.

In addition to mitotically dividing cells, the respiratory andolfactory mucosa represent major sites of Sak expressionduring embryogenesis. Sak is expressed in the basal region ofthe olfactory epithelium, consisting mainly of progenitorstem cells for the olfactory neuronal lineage. In the respira-tory system, high levels of Sak expression are seen in themucus-secreting and ciliated epithelium of the nasal cavityand larger airways, with levels decreasing as the respiratorytract branches. A common feature of ciliated and dividingcells is the motile nature of the microtubular system. Defectsin the organization of the microtubular system during celldivision are seen in polo' homozygotes and CDC5tS mutants,suggesting that this family of kinases, including Sak, mayregulate microtubule-mediated processes.

Part of this work was carried out in the laboratory of Dr. AlanBernstein. We thank him for his support and encouragement as wellas his critical reading of the manuscript. We thank Mike Demetriouand Francois Guillemot for critically reading the manuscript andZofia Krzyzek for secretarial assistance. C.F. was supported by aNational Science and Engineering Research Council (Canada) stu-dent award. B.M. was a fellow of the Leukemia Research Fund.J.W.D. is a Senior Research Scientist of the National CancerInstitute of Canada. This work was supported by grants from theNational Cancer Institute of Canada and the Medical ResearchCouncil of Canada to J.W.D. and Alan Bernstein.

1. Unmazres, S., Moreira, A., Tavares, A., Girdham, C., Spruce,B. A., Gonzales, C., Karess, R. E., Glover, D. M. & Sunkel, C. E.(1991) Genes Dev. 5, 2153-2165.

2. Kitada, K., Johnson, A. L., Johnston, L. H. & Sugino, A. (1993)Mol. Cell. Biol. 13, 4445-4457.

3. Fenton, B. & Glover, D. M. (1993) Nature (London) 363, 637-640.4. Sharon, G. & Simchen, G. (1990) Genetics 125, 475-485.5. Schild, D. & Byers, B. (1980) Genetics 96, 859-876.6. Simmons, D. L., Neel, B. G., Stevens, R., Evett, G. & Erikson,

R. L. (1992) Mol. Cell. Biol. 12, 4164-4169.7. Clay, F. J., McEwen, S. J., Bertoncello, I., Wilks, A. F. & Dunn,

A. R. (1993) Proc. Nati. Acad. Sci. USA 90, 4882-4886.8. Dennis, J. W. (1986) Cancer Res. 46, 4594-4600.9. Heffernan, M. & Dennis, J. W. (1991) NucleicAcids Res. 19, 85-92.

10. Hirt, B. (1967) J. Mol. Biol. 26, 365-369.11. Peter, M., Nakagawa, J., Doree, M., Labbe, J. C. & Nigg, E. A.

(1990) Cell 60, 791-801.12. Padgett, R. A., Grabowski, P. J., Konarska, M. M., Seiler, S. &

Sharp, P. A. (1986) Annu. Rev. Biochem. 55, 1119-1150.13. Kozak, M. (1986) Cell 44, 283-292.14. Kozak, M. (1992) Annu. Rev. Cell Biol. 8, 197-225.15. Shaw, G. & Kamen, R. (1986) Cell 46, 659-667.16. Hanks, S. K. & Quinn, A. M. (1991) Methods Enzymol. 200,38-63.17. Rogers, S., Wells, R. & Rechsteiner, M. (1986) Science 234,

364-368.18. Stein, R., Mori, N., Matthews, K., Lo, L.-C. & Anderson, D. J.

(1988) Neuron 1, 463-476.19. Quelle, D. E., Ashmun, R. A., Shurtleff, S. A., Kato, J.-Y., Bar-

Sagi, D., Roussel, M. F. & Sherr, C. J. (1993) Genes Dev. 7,1559-1571.

20. Baker, S. J., Markowitz, S., Fearon, E. R., Willson, J. K. V. &Vogelstein, B. (1993) Science 249, 912-915.

21. Qii, X.-Q., Chittenden, T., Livingston, D. M. & Kaelin, W. G., Jr.(1992) Genes Dev. 6, 953-964.

22. Talmage, D. A., Riney, C. & Benjamin, T. L. (1991) Cell GrowthDifferen. 2, 51-58.

6392 Biochemistry: Fode et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

1, 2

020