J. Biochem. 116, 636-642 (1994)

Molecular Cloning of a Chicken Lung cDNA Encoding a Novel with N-Terminal Two LIM/Double Finger Motifs1

Kazumasa Ohashi,* Jiro Toshima,* Katsunori Tajinda,* Toshikazu Nakamura,' and

Kensaku Mizuno*•2

*Department of Biology , Faculty of Science, Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812; and •õ Biomedical Research Center, Osaka University Medical School, Suita, Osaka 565

Received for publication, April 25, 1994

Using the eDNA fragment of chicken c-sea receptor as a probe, we isolated from a chicken lung cDNA library overlapping cDNA clones encoding a novel , which we termed LIM-kinase (LIMK). The predicted polypeptide of 642 amino acid residues contains remarkable structural features, composed of the N-terminal two tandem- ly arrayed LIM/double zinc finger motifs and the C-terminal unusual protein kinase domain. To our knowledge, a protein kinase containing the LIM motif in the molecule has not heretofore been described. The protein kinase domain of LIMK shares highly conserved residues with the known protein , but LIMK is unique in that it contains the sequence DLNSHN in subdomain VIB and a short, highly basic insert sequence, which may function as a signal for nuclear localization, between subdomain VII and VIII in the protein kinase domain. Northern blot analysis revealed that the single species of LIMK mRNA of 3.8kb is expressed predominantly in the lung, and faintly in the kidney, liver, brain, spleen, gizzard, and intestine. As the LIM motif is thought to be involved in protein-protein interactions by binding to another LIM motif, and is often present in the homeodomain- containing involved in cell fate determination and in the oncogenic nuclear proteins (rhombotins), it is likely that LIMK is involved in developmental or oncogenic processes through interactions with these LIM-containing proteins.

Key words: LIMK, LIM-kinase, LIM motif, protein kinase, zinc finger.

Protein kinases play an important role in the regulation of ture (5, 6). Recently, an increasing number of protein diverse cell functions, including cell growth, differentia kinase genes have been isolated using the chain tion, and other responses to external stimuli. A large reaction (PCR) or low-stringency screening (7, 8). Some of number of protein kinases have been heretofore identified, these newly identified protein kinases contain a domain and they are classified into several subfamilies, based on structure not related to any known members of the protein their structures and functions (1-3). Protein kinases in a kinase family. These unclassified protein kinases seem to subfamily contain, in addition to a similar sequence in the be regulated in a different manner and to have specific protein kinase catalytic domain, similar extracatalytic functions in intracellular pathways not domain structures. Since the extracatalytic domain is previously established. In view of the central role of protein usually involved in the regulation of kinase catalytic kinases in diverse cell activities, further searches for a activity, each member of a protein kinase subfamily has a novel class of protein kinases are important to better similar mode of regulation. For example, the members of understand signal-transducing mechanisms. protein kinase C subfamily have a similar extracatalytic In the present study, we isolated a chicken cDNA coding domain composed of a diacylglycerol-binding domain and a for a novel protein kinase; it has an unusual sequence in the calcium ion-binding domain, and the kinase activity of protein kinase domain as well as a unique extracatalytic protein kinase Cs is modulated by binding of appropriate domain. The encoded protein, termed LIM-kinase (LIMK), effectors to these regulatory domains (4). Similarly, Src contains two tandem repeats of LIM/double zinc finger type protein tyrosine kinases and calmodulin-dependent motif at the N-terminus and a protein kinase consensus protein kinases have their own mode of regulation, which sequence at the C-terminus. The LIM/double zinc finger depends upon a characteristic non-catalytic domain struc motif, named after three homeodomain-containing pro teins, Lin-11, Isl-1, and Mec-3 (9-11), is a putative zinc- 1 This study was supported in part by a Grant-in-Aid for Scientific binding motif with a conserved placement of cysteine and Research from the Ministry of Education, Science and Culture of histidine residues, (CX2CX16-19HX2C)X2(CX2CX16-20- Japan. The nucleotide sequence of chick LIMK reported in this paper CX2-3C/H/D) (12). It can be defined as a paired zinc finger has been submitted to the DDBJ, EMBL, and GenBank databases separated by a two-amino-acid linker (12). Whereas most with the accession number D26310. 2 To whom correspondence should be addressed. zinc-finger motifs function by binding to specific DNA sequences (13), the LIM motif is suggested to be involved Abbreviations: CRP, cysteine-rich protein; LIMK, LIM motif-con taining protein kinase; PCR, polymerase chain reaction. in protein-protein interactions (14, 15). LIMK in the

636 J. Biochem. Cloning of a LIM Motif-Containing Protein Kinase 637

present study is a unique protein in that it contains in the molecule both a protein kinase domain and two LIM RESULTS AND DISCUSSION domains. We describe here the complete structure of LIMK and expression patterns of its mRNA. The potential role of The initial objective of our study was to isolate the cDNA LIMK in cellular functions is discussed. clone of c-sea, the cellular homologue of the avian erythrob lastosis virus oncogene v-sea. The cDNA from chicken lung

EXPERIMENTAL PROCEDURES RNA was subjected to PCR amplification, using primers corresponding to the reported sequence of v-sea (18), and RNA Preparation-Total RNA was isolated from chicken the amplified 0.6-kb cDNA fragment of chicken c-sea was tissues by the acid guanidinium thiocyanate-phenol used as a probe to screen a chicken lung cDNA library. chloroform method (16). Poly(A)+ RNA was purified by Screening of 8 x 105 phages produced approximately 200 two cycles of Oligotex dT-30 (Rosch, Tokyo) adsorption, positive clones, of which 6 clones with a relatively high according to the manufacturer's instructions. intensity were plaque-purified and subcloned. Based on the Isolation of Chicken LIMK cDNAs•\Total RNA from sequences at 5•L- and 3•L-termini, 4 isolated clones were chicken lung (2ƒÊg) was subjected to reverse-transcription found to code for chicken c-sea receptor tyrosine kinase using random hexamer primers and (22). The other 2 clones, chS-5 of a 3.8-kb insert and chS-6 (BRL). A chicken c-sea cDNA fragment was PCR-amplifi of 3.3kb, had novel overlapping sequences. Nucleotide ed, as described (17), using a sense primer SEAlS (5•L- sequence analysis of the insert of chS-5 clone revealed that ATAGGATCCAGCTGCTGGAGGAGGTGAAGGAC- the 3,806-bp sequence contained a single long open reading AT-3•L; containing BamHI site at the 5•L end) and an frame predicted to encode a 642-amino-acid polypeptide antisense primer SEA2AS (5•L-ATAGAATTCCACGTTCTTTTTCCACGTCTTTTG (Fig. 1), which we named LIM-kinase (LIMK) for LIM ACTTGGTAGTGAATTTT- 3•L; containing EcoRI site at the motif-containing protein kinase (see below). The size of the 5•L end) corresponding to the v-sea sequence (805-830) and insert of chS-5 (3.8kb) coincides with the size of mRNA, (1398-1422), respectively (18). An oligo(dT)-primed measured by Northern analysis (see below), hence, the X -ZAP II cDNA library (8 x 105 independent plaques) clone probably contains a practically full length cDNA. The prepared from chicken lung poly(A)+ RNA was hybridized initiation codon was assigned to the first in-frame ATG at with the 32P-labeled 0.6-kb chicken c-sea cDNA fragment, nucleotides 83-85, which reside in the Kozak consensus in a solution containing 43% formamide, 5 x SSPE (1 x sequence (23). The stop codon is followed by a 1.8-kb SSPE: 0.15M NaCl, 10mM sodium phosphate, 1mM 3•L-noncoding sequence, including poly(A) tail at the 3•L- EDTA, pH 7.4), 5 x Denhardt's solution (1 X Denhardt's terminus. There is an ATTTA sequence at nucleotides solution: 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% 3750-3754, which is known to be the recognition signal for bovine serum albumin), and 50 ,u g/ml of salmon sperm mRNA instability (24); thus it is likely to be the primary DNA at 42•Ž for 24h. The filter was washed twice at room response gene regulated in a manner similar to other temperature and once at 42•Ž in a mixture of 2 x SSPE and short-lived mRNAs (24). 0.5% SDS. From 200 positive clones, 6 were selected, The predicted amino acid sequence of LIMK has promi plaque-purified, and subcloned into pBluescript (Strata- nent structural features, composed of a protein kinase gene). One of these clones (chS-5) with a 3.8-kb insert was domain in the C-terminal half and two repeats of a cysteine- subjected to sequence analysis. rich LIM/double zinc finger motif at the extreme N- DNA Sequencing-The complete sequence was deter- terminus (Fig. 2). The interposing approximately 200- mined on both strands by the dideoxy termination method amino-acid region flanked by the LIM domain and the (19), using Taq polymerase and a dye primer cycle se protein kinase domain has no similarity to any known quencing kit with 370A DNA sequencer (Applied Biosys sequence motifs. Hydropathy plot analysis (25) showed no tems). Overlapping fragments were generated by se long stretch of hydrophobic residues that could serve as a signal sequence for protein secretion or a transmembrane quential exonuclease III digestion (20) or by priming with sequence-specific oligonucleotides. Sequences were ana domain (data not shown), hence LIMK seems to function lyzed and compared using the DNASIS DNA sequence intracellularly. The protein kinase domain of LIMK shares highly program (Hitachi). Northern Hybridization-Poly(A)+ RNA was denatured conserved amino acid elements with the known protein with formaldehyde and electrophoresed on a 1% agarose/ kinases. As shown in Fig. 3A, the conserved amino acid 0.7% formaldehyde gel (21). were transferred to a residues, which are thought to be essential for kinase Biodyne B nylon membrane filter (Pall BioSupport) and catalytic activity (2, 3), including the ATP-, hybridized, under the conditions described above for GXGXXG (residues 338-343) and a lysine at the position screening a cDNA library, with the 32P-labeled 1.5-kb 360, and the short motifs DFG (residues 469-471) and APE 5•L-terminal EcoRI fragment of chicken LIMK cDNA (chS- (residues 514-516) in the subdomain VII and VIII, respec 5). Filters were also hybridized with the 32P-labeled cDNA tively, are present in LIMK. However, the remainder of of rat glyceraldehyde-3-phosphate dehydrogenase (GAP- the sequence of the protein kinase domain of LIMK is DH) as a control for constant loading. Filters were washed significantly different from those of other members of the in 2 x SSPE and 0.5% SDS at 65°C, and densitometrically protein kinase family (about 25-30% identity compared analyzed using a BAS2000 Bio-Imaging Analyzer (Fuji with other protein kinases). The short region of amino acid Film, Tokyo). sequence in subdomain VIB is usually used to predict the substrate specificity of the protein kinase (2, 3). In this region, LIMK contains the sequence DLNSHN (residues 451-456), which does not fit the consensus sequence charac-

Vol. 116, No. 3, 1994 638 K. Ohashi et al.

Fig. 1. Nucleotide and deduced amino acid sequence of chicken LIM-kinase (LIMK) cDNA. The nucleotides and amino acids are numbered on the left and right. An asterisk denotes the termination codon. The N-terminal two repeats of the LIM domain and a C-terminal protein kinase domain are underlined. The LIM consensus residues are circled. The basic insert in the kinase domain is boxed. The putative signal for mRNA instability is doubly underlined.

J. Biochem. Cloning of a LIM Motif-Containing Protein Kinase 639

Fig. 2. Schematic representation of the predicted structure of the LIMK protein. Four putative zinc finger configurations are illustrated in such a way that the conserved cysteine, histidine, and aspartic acid residues coordinate a zinc ion. The amino acid residue numbers are indicated.

Fig. 3. Alignments of the protein kinase domain (A) and the LIM motifs (B) of chicken LIMK (chLIMK) with the related sequences of other proteins. (A) The first amino acid positions of the respective sequences are shown on the left of the align ment. The highly conserved amino acid resi dues throughout the all protein kinases (3) are indicated by asterisks. The conserved subdomains in protein kinases (3) are indicat ed by Roman numerals on the bottom of the alignment. The amino acids identical with LIMK are shaded. Gaps (-) are introduced to increase similarity. (B) The first residue number is shown on the left. The LIM consen sus residues are shaded. The positions of the highly conserved residues within the LIM motifs are shown on the bottom of the align ment with hydrophobic residues indicated by asterisks. Sequence data sources are as fol lows: mouse activin receptor (mACTR) (39); human protein kinase C-ƒ¿ (hPKC-ƒ¿) (40); chicken c-Sea (chSEA) (22); Lin-11 (9); human rhombotin-1 (Ttg-1) (30); CRP2 (31).

Vol. 116, No. 3, 1994 640 K. Ohashi et al.

subdomains (2, 3), the basic insert may function as the nuclear localization signal for LIMK. Another striking structural feature of LIMK is the presence of two tandem repeats of the LIM/double zinc finger motif at the extreme N-terminus of the molecule. Figure 3B shows an alignment of the two LIM sequences of LIMK with those of other LIM motif-containing proteins, Lin-11, rhombotin-1, and cysteine-rich protein 2 (CRP2) (9, 30, 31). The cysteine, histidine and aspartic acid residues of the LIM consensus motif exactly align. Both LIM sequences in LIMK also share hydrophobic residues (asterisks in Fig. 3B) conserved within the LIM proteins. The extracatalytic domain of the protein kinases C family contains two repeats of cysteine-rich zinc finger-like motifs, as the diacylglycerol-binding domain (4), but there is no homology between the zinc finger motif in protein kinase C and the LIM/double zinc finger motif in LIMK. Since the protein kinase containing the LIM motif in the molecule has not heretofore been described, it is of much interest to study the functional role of LIMK in intracel lular signaling pathways. The LIM motif has been identified in various homeo domain-containing proteins, such as Caenorhabditis ele gans Lin-11 (9), Mec-3 (10), murine Isl-1 (11), Lmx-1 Fig. 4. Northern blot analysis of LIMK mRNA expression in (14), Xenopus Xlim-1 (32), and DrosophilaApterous (33), chicken tissues. Poly(A)+ RNA (2 Jig) was electrophoresed, trans involved in transcriptional regulation and determination of ferred, and hybridized with 32P-labeled fragment of chicken LIMK cDNA, as described inn "EXPERIMENTAL PROCEDURES." Posi cell fate. For example, Lin-11 and Mec-3 are the proteins tions of LIMK mRNA (3.8kb) are indicated by an arrow. The lower required for asymmetric divisions of vulval precursor cells panel shows the expression of glyceraldehyde phosphate dehy and differentiation of touch receptor neurons, respectively, drogenase (GAPDH), as a control. in the nematode C. elegans (9, 10), and Isl-1 is a rat insulin gene enhancer-binding protein (11). LIM motifs are also present in proteins lacking a homeodomain or any obvious teristic to either serine/threonine-specific kinases (DLKX functional domain, such as rhombotins (30, 34), cysteine- XN) or to tyrosine-specific kinases (DLAARN or DLRA- rich intestinal protein (CRIP) (35), cysteine-rich protein AN); hence we could not predict from this sequence (CRP) (36), cysteine-rich protein 2 (CRP2) (31), and zyxin whether LIMK has specificity to serine/threonine or tyro (15). While the cellular functions of these homeodomain- sine residues. deficient LIM-proteins are unknown, rhombotins are sug Phylogenetic analysis by comparing kinase domains of gested to be oncogenicnuclear proteins (37, 38). While the LIMK and other protein kinases (26) revealed that LIMK function of LIM motif in LIM-containing proteins is not is related to the activin receptor and Daf-1 (data not well understood, it was proposed to be involved in protein- shown). However, the relationships between LIMK and protein interactions (14, 15). The finding that zyxin is these kinases are not obvious, because the bootstrap associated with CRP provides strong evidence for the probability of the branching point (27) is relatively low involvement of LIM motifs in protein-protein interactions (64%). The observations that receptor and non-receptor through association between LIM motifs (15). Given that type tyrosine kinases form an obvious cluster on the tree the LIM motifs in LIMK are involved in association with (bootstrap probability; 97%) and are clearly separated other LIM-containing proteins through LIM-LIM interac from LIMK (data not shown), mean that LIMK probably tions, such interaction presumably serves to modulate the belongs to a serine/threonine kinase family. However, it is kinase activity of LIMK or to attract the target protein of still possible that LIMK has tyrosine kinase activity, as do LIMK. Since most LIM-containing proteins, including some members of serine/threonine kinase family, such as LIM-homeodomain proteins and rhombotins, are nuclear MAP kinase and activin receptor phosphorylate tyrosine proteins involved in cell-fate determination or in oncogene residues as well as serine/threonine residues (28, 29). sis, it is probable that LIMK participates in cell differentia Kinase activity of the LIMK protein will need to be tion or oncogenic processes through interaction with these measured. proteins. In this respect, the presence of the potential It is noted that the protein kinase domain of LIMK nuclear targeting signal within the kinase insert of LIMK contains the short, highly basic insert sequence KKRTLR may support the notion that LIMK functions in the nucleus. KSDRKKR (residues 491-503) between the subdomain On the other hand, LIMK may interact with the cytoskele VII and VIII (Fig. 3A). Since such an insert is absent in tal LIM-containing protein, zyxin, isolated from cell adhe other protein kinases, it may play a specific role in LIMK sion plaques (15). Since the mechanism of signal transduc function. The highly basic character of this insert suggests tion mediated through cell adhesion plaques remains that it may function as a nuclear targeting signal. As the unknown, LIMK might be a component in the process. kinase insert region is thought to exist in a loop structure Determination of the subcellular localization of LIMK is a protruding from the globular folding of the conserved prerequisite to understand the cellular function of LIMK;

J. Biochem. Cloning of a LIM Motif-Containing Protein Kinase 641

this study using a specific antibody for LIMK is ongoing in 12. Wang, X., Lee, G., Liebhaber, S.A., and Cooke, N.E. (1992) our laboratory. Human cysteine-rich protein: A member of the LIM/double- The tissue distribution of LIMK mRNA expression was finger family displaying coordinate serum induction with c-myc. examined by Northern blot analysis of poly(A)+ RNAs J. Biol. Chem. 267, 9176-9184 from various chicken tissues, using as a probe the 1.5-kb 13. Harrison, S.C. (1991) A structural taxonomy of DNA-binding EcoRI fragment of chicken LIMK cDNA. The single band of domains. Nature 353, 715-719 14. German, M.S., Wang, J., Chadwick, R.B., and Rutter, W.J. 3.8kb was detected in the highest level in RNA from lung , (1992) Synergistic activation of the insulin gene by a LIM-homeo and there were weaker signals with RNAs from the kidney, domain protein and a basic helix-loop-helix protein: Building a brain, liver, spleen, gizzard, and large intestine (Fig. 4). functional insulin minienhancer complex. Genes Deu. 6, 2165- The high level of expression of LIMK in lung suggests a 2176 specific role of LIMK in this organ. 15. Sadler, I., Crawford, A.W., Michelsen, J.W., and Beckerle, M.C. Recently we isolated from rat brain and human hepatoma (1992) Zyxin and cCRP: Two interactive LIM domain proteins HepG2 cell cDNA libraries cDNA clones encoding LIMK associated with the cytoskeleton. J. Cell Biol. 119, 1573-1587 16. Chomczynski, P. and Sacchi, N. (1987) Single-step method of and LIMK-related proteins (to be published). The deduced RNA isolation by acid guanidinium thiocyanate-phenol-chloro sequences of the encoded proteins have an overall structure form extraction. Anal. Biochem. 162, 156-159 similar to that of chicken LIMK described in the present 17. Ohashi, K., Mizuno, K., Kuma, K., Miyata, T., and Nakamura, T. study. They have common domain structures composed of (1994) Cloning of the cDNA for a novel receptor tyrosine kinase, the N-terminal two LIM domains, the intervening domain, Sky, predominantly expressed in brain. Oncogene 9, 699-705 and the C-terminal protein kinase domain. The characteris 18. Smith, D.R., Vogt, P.K., and Hayman, M.J. (1989) The v-sea tic sequence motifs in the protein kinase domain of chicken oncogene of avian erythroblastosis retrovirus S13: Another member of the protein-tyrosine kinase gene family. Proc. Natl. LIMK, such as the DLNSHN motif in the subdomain VIB Acad. Sci. USA 86, 5291-5295 and the basic kinase insert between the subdomains VII and 19. Sanger, F., Nicklen, S., and Coulson, A.R. (1977) DNA sequenc VIII, are also conserved in rat and human LIMK.related ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. proteins. These observations suggest that these LIM- USA 74,5463-5467 kinases form a subfamily in the protein kinase family and 20. Henikoff, S. (1987) Unidirectional digestion with exonuclease III that they are likely to have specialized functions in the in DNA sequence analysis. Methods Enzymol. 155, 156-165 21. Sambrook, J., Fristch, E.F., and Maniatis, T. (1989) Molecular previously unknown signaling pathways. Cloning. A Laboratory Manual 2nd ed. Cold Spring Harbor Laboratory Press, New York We thank Dr. Kei-ichi Kuma and Dr. Takashi Miyata of Kyoto 22. Huff, J.L., Jelinek, M.A., Borgman, C.A., Lansing, T.J., and University for phylogenetic analysis, and M. Ohara for valuable Parsons, J.T. (1993) The protooncogene c-sea encodes a trans- comments. membrane protein-tyrosine kinase related to the Met/hepatocyte growth factor/scatter factor receptor. Proc. Natl. Acad. Sci. USA REFERENCES 90,6140--6144 23. Kozak, M. (1989) The scanning model for translation: An update. 1. Hunter, T. (1987) A thousand and one protein kinases. Cell 50, J. Cell Biol. 108, 229-241 823-829 24. Shaw, G. and Kamen, R. (1986) A conserved AU sequence from 2. Hanks, S.K., Quinn, A.M., and Hunter, T. (1988) The protein the 3•L untranslated region of CM-CSF mRNA mediates selective kinase family: Conserved features and deduced phylogenyof the mRNA degradation. Cell 46, 659-667 catalytic domains. Science 241, 42-52 25. Kyte, J. and Doolittle, R.F. (1982) A simple method for display 3. Hanks, S.K. and Quinn, A.M. (1991) Protein kinase catalytic ing the hydropathic character of a protein. J. Mol. Biol. 157, 105- domain sequence database: Identificationof conserved features of 132 primary structure and classificationof family members. Methods 26. Saitou, N. and Nei, M. (1987) The neighbor-joining method: A Enzymol. 200, 38-63 new method for reconstructing phylogenetic trees. Mol. Biol. 4. Nishizuka, Y. (1992) Intracellular signaling by hydrolysis of Evol. 4, 406-425 phospholipids and activation of protein kinase C. Science 258, 27. Felsenstein, J. (1985) Confidence limits on phylogenies: An 607-614 approach using the bootstrap. Evolution 39, 783-791 5. Koch, C.A., Anderson, D., Moran, M.F., Ellis, C., and Pawson, T. 28. Seger, R., Ahn, N.G., Boulton, T.G., Yancopoulos, G.D., Panayo (1991) SH2 and SH3 domains: Elements that controlinteractions tatos, N., Radziejewska, E., Ericsson, L., Bratlien, R.L., Cobb, M. of cytoplasmic signaling proteins. Science 252, 668-674 H., and Krebs, E.G. (1991) Microtubule-associated protein 2 6. Hanson, P.I. and Schulman, H. (1992) Neuronal Cat+/calmodu kinases, ERK1 and ERK2, undergo autophosphorylation on both lin-dependent protein kinases. Annu. Rev. Biochem.61, 559-601 tyrosine and threonine residues: Implication for their mechanism 7. Wilks, A.F. (1989) Two putative protein-tyrosine kinases iden of activation. Proc. Natl. Acad. Sci. USA 88, 6142-6146 tified by application of the polymerase chain reaction. Proc. Natl. 29. Nakamura, T., Sugino, K., Kurosawa, N., Sawai, M., Takio, K., Acad. Sci. USA 86, 1603-1607 Eto, Y., Iwashita, S., Muramatsu, M., Titani, K., and Sugino, H. 8. Partanen, J., Makela, T.P., Alitalo, R., Lehvaslaiho, H., and (1992) Isolation and characterization of activin receptor from Alitolo, K. (1990) Putative tyrosine kinases expressed in K-562 mouse embryonal carcinoma cells: Identification of its serine/ human leukemia cells. Proc. Natl. Acad. Sci. USA 87, 8913- threonine/tyrosine protein kinase activity. J. Biol. Chem. 267, 8917 18924-18928 9. Freyd, G., Kim, S.K., and Horvitz, H.R. (1990) Novel cysteine- 30. McGuire, E.A., Hockett, R.D., Pollock, K.M., Barthholdi, M.F., rich motif and homeodomainin the product of the Caenorhabditis O'Brien, S.J., and Korsmeyer, S.J. (1989) The t(11;14)(pl5q11) elegans cell lineage gene lin-11. Nature 344, 876-879 in a T-cell acute lymphoblastic leukemia cell line activates 10. Way, J.C. and Chalfie, M. (1988) mec-3, a homeobox-containing multiple transcripts, including Ttg-1, a gene encoding a potential gene that specifiesdifferentiation of the touch receptor neurons in zinc finger protein. Mol. Cell. Biol. 9, 2124-2132 C. elegans. Cell 54, 5-16 31. Okano, I., Yamamoto, T., Kaji, A., Kimura, T., Mizuno, K., and 11. Karlsson, 0., Thor, S., Norberg, T., Ohlsson, H., and Edlund, T. Nakamura, T. (1993) Cloning of CRP2, a novel member of the (1990) Insulin gene enhancer binding protein Isl-1 is a member of cysteine-rich with two repeats of an unusual LIM/ a novel class of proteins containing both a homeo- and a Cys-His double zinc-finger motif. FEES Lett. 333, 51-55 domain. Nature 344, 879-882 32. Taira, M., Jamrich, M., Good, P.J., and Dawid, LB. (1992) The

Vol. 116, No. 3, 1994 642 K. Ohashi et al.

LIM domain-containing homeo box gene Xlim-1 is expressed Cooke, N.E. (1990) Characterization of a human cDNA encoding specifically in the organizer region of Xenopus gastrula embryos. a widely expressed and highly conserved cysteine-rich protein Genes Dec. 6, 356-366 with an unusual zinc-finger motif. Nucleic Acids Res. 18, 3871- 33. Cohen, B., McGuffin, M.E., Pfeifle, C., Segal, D., and Cohen, 3879 S.M. (1992) apterous, a gene required for imaginal disc develop 37. McGuire, E.A., Davis, A.R., and Korsmeyer, S.J. (1991) T-cell ment in Drosophila encodes a member of the LIM family of translocation gene 1 (Ttg-1) encodes a nuclear protein normally developmental regulatory proteins. Genes Dec. 6, 715-729 expressed in neural lineage cells. Blood 77, 599-606 34. Boehm, T., Greenberg, J.M., Buluwela, L., Lavenir, I., Forster, 38. McGuire, E.A., Rintoul, C.E., and Korsmeyer, S.J. (1992) A., and Rabbitts, T.H. (1990) An unusual structure of a putative Thymic overexpression of Ttg-1 in transgenic mice results in T cell oncogene which allows production of similar proteins from T-cell acute lymphoblastic leukemia/lymphoma. Mol. Cell. Biol. distinct mRNAs. EMBO J. 9, 857-868 12,4186-4196 35. Birkenmeier, E.H. and Gordon, J.I. (1986) Developmental 39. Mathews, L.S. and Vale, W.W. (1991) Expression cloning of an regulation of a gene that encodes a cysteine-rich intestinal protein activin receptor, a predicted transmembrane serine kinase. Cell and maps near the murine immunoglobulin heavy chain locus. 65,973-982 Proc. Natl. Acad. Sci. USA 83, 2516-2520 40. Finkenzeller, G., Marme, D., and Hug, H. (1990) Sequence of 36. Liebhaber, S.A., Emery, J.G., Urbanek, M., Wang, X., and human protein kinase Ca. Nucleic Acids Res. 18, 2183

J. Biochem.