Characterization of a Monoclonal Antibody Panel Shows That the Myotonic Dystrophy Protein Kinase, DMPK, Is Expressed Almost Exclusively in Muscle and Heart

Characterization of a monoclonal antibody panel shows that the myotonic dystrophy protein kinase, DMPK, is expressed almost exclusively in muscle and heart

L.T. Lam, Y.C.N. Pham, Nguyen thi Man and G.E. Morris+

MRIC Biochemistry Group, PP18, North East Wales Institute, Mold Road, Wrexham LL11 2AW, UK

Received 11 May 2000; Accepted 28 June 2000 DDBJ/EMBL/GenBank accession no. AF250871

Myotonic dystrophy (DM) is a multisystemic disorder ness), brain (mental retardation), eyes (cataracts), testis caused by an inherited CTG repeat expansion which (atrophy) and heart (conduction defects) (10). It is important, affects three genes encoding the DM protein kinase therefore, to establish the normal tissue distribution of all three (DMPK), a homeobox protein Six5 and a protein proteins produced from the DM locus. containing WD repeats. Using a panel of 16 mono- The human DMPK cDNA has 15 exons and predicts a ∼ clonal antibodies against several different DMPK protein of 70 kDa, although there may be variation in the size epitopes we detected DMPK, as a single protein of and sequence of the first exon (1–3,11,12). DMPK is a serine/ threonine protein kinase in which the catalytic domain ∼80 kDa, only in skeletal muscle, cardiac muscle and, (∼43 kDa) is followed by a coiled-coil domain (∼12 kDa) and to a lesser extent, smooth muscle. Many earlier a hydrophobic C-terminal domain. Alternative splicing has reports of DMPK with different sizes and tissue distri- been observed in a short VSGGG sequence between domains butions appear to be due to antibody cross-reactions and in the C-terminal domain (13). Characterization of DMPK with more abundant proteins. One such antibody, protein using antibodies has proven to be very difficult and is MANDM1, was used to isolate two related protein still a source of considerable confusion. This may be largely kinases, MRCKα and β, from a human brain cDNA due to antibody cross-reaction, to which DMPK antibodies library and the shared epitope was located at the seem to be particularly vulnerable. Anti-peptide antisera against catalytic site of DMPK using a phage-displayed the catalytic domain have recognized proteins of 52–55 kDa in random peptide library. The peptide library also iden- skeletal muscle (14,15), 42 and 60 kDa in brain and heart tified an epitope shared between DMPK and a 55 kDa (14,15) or 53 kDa in heart and skeletal muscle (16). Several antisera prepared by Timchenko et al. (17) did recognize a muscle-specific protein. The results suggest that protein of the expected size (72 kDa) in fibroblasts. A mono- effects of the repeat expansion on the DMPK gene clonal antibody produced using recombinant catalytic and coil may be responsible for muscle and heart features of domains as immunogen recognized a 64 kDa protein in muscle DM, whereas clinical changes in other tissues may be and a 79 kDa protein in brain (18). However, variable results due to effects on the other two genes. have also been obtained with antibodies against the C-terminal domain, which is not known to share sequences with any other protein. Polyclonal antisera raised against the whole C-terminal INTRODUCTION domain recognized proteins of 71–74 and 80–85 kDa in skel- Myotonic dystrophy (DM) is caused by an unstable CTG etal and cardiac muscle (19–21). However, a similar study repeat sequence in the 3′-untranslated region of a protein produced antiserum recognizing proteins of 85 and 54 kDa kinase (DMPK) gene on chromosome 19 (1–3). The CTG repeat (22) and antisera against C-terminal peptides recognized proteins increases from 5–30 in the normal population to 50–1000+ in of 70–72 and 50–55 kDa in skeletal muscle (23,24). Finally, a DM patients and the increase is correlated with disease 45 kDa protein was detected in brain using C-terminal antibodies severity (4). Nuclear retention of DMPK transcripts with (21). expanded CTG repeats has been demonstrated (5) and The correct approach to this problem is to prepare a panel of decreased DMPK poly(A)+ mRNA levels have been reported monoclonal antibodies (mAbs) which recognize several in DM tissues (6,7). The expanded CTG repeat at the DM locus different epitopes on DMPK. Although mAbs, like antisera, is also within the 5′ end of the Six5 gene predicted to encode a often display cross-reactions with other proteins, these cross- homeobox protein (8) and a third gene, DMWD (gene 59, or reactions tend to be very specific for each epitope. Any protein DMR-N9 in the mouse), which lies 5′ to DMPK,isalso on western blots which contains most, if not all, of the epitopes affected by CTG expansions (9). Effects on expression of non- is likely to be authentic DMPK, whereas proteins sharing only DMPK proteins could therefore be responsible for some, if not one epitope with DMPK are likely cross-reactions. Our initial all, of the clinical features of DM. DM affects many different results with a panel of 12 mAbs against the catalytic or coil tissues, including muscle (myotonia and progressive weak- domains suggested that DMPK was expressed as 80 and

+To whom correspondence should be addressed. Tel: +44 1978 293214; Fax: +44 1978 290008; Email: [email protected] 2168 Human Molecular Genetics, 2000, Vol. 9, No. 14

Table 1. Sixteen mAbs against DMPK catalytic or coil domains mAb Clone no. Species specificitya Domain recognizedb MANDM1 6G8 hu, rb, mo Catalytic MANDM2 10A7 hu Catalytic MANDM3 8B5 hu Coil MANDM4 9H9 hu Catalytic MANDM5 5G12 hu Catalytic MANDM6 7E4B7 hu Catalytic MANDM7 (55 kDa) 12G5 hu, rb Catalytic MANDM8 (55 kDa) 7E7 hu, rb Catalytic MANDM9 3D10 hu Coil MANDM10 2H12 hu Coil MANDM11 6B6 hu Coil MANDM12 7E4G2 hu Coil MANDM13 12A4 hu Coil MANDM14 7E6 hu Coil MANDM15 5C9 hu Coil MANDM16 9A4 hu Coil ahu, human; rb, rabbit; mo, mouse. bDomain recognition was determined using recombinant DMPK fragments encoded by exons 2–8 (catalytic) or exons 9–12 (coil) (25).

72 kDa proteins, the 80 kDa protein being present mainly in skeletal and cardiac muscle whereas the 72 kDa protein is widely expressed (25). We now show, however, that only the Figure 1. Western blot detection of DMPK and cross-reacting proteins in 80 kDa protein is authentic DMPK and that it is detectable only (a) skeletal muscle, (b) lung and (c) cerebral cortex, using a panel of 15 mAbs. in skeletal muscle, in cardiac muscle and, at lower levels, in DMPK is found only in skeletal muscle (a). Human tissue extracts were prepared by homogenization and boiling in 1% SDS buffer before separation as a smooth muscle. Cross-reacting proteins of 190 and 72 kDa strip on 3–12.5% polyacrylamide gradient gels as described previously. The recognized by mAb MANDM1 appear to be isoforms of western blot on nitrocellulose was cut into vertical strips, each of which was myotonic dystrophy-related cdc42-binding kinase (MRCK) incubated with a different mAb (MANDM mAb numbers at top of blot; C, no- (26–28). mAb control; MANDM12 was omitted). The strips were reassembled in the presence of chemiluminescent substrate for exposure to X-ray film. Mr, bioti- nylated markers (catalase, 60 kDa; phosphorylase, 97 kDa); MRCK, myotonic RESULTS dystrophy-related cdc42-binding kinase; CRP, cross-reacting protein. A panel of 12 mAbs, prepared using a DMPK fragment containing the catalytic and coil domains, has been described alternatively spliced isoforms is that no mRNA splicing that previously (25). Seven mAbs recognized the catalytic domain removes the coil domain has been observed, so all nine mAbs and five recognized the coil domain. A further four mAbs, against the coil domain should recognize all known splicing MANDM13–16, were prepared using recombinant coil plus isoforms (13). C-terminal domains and all four of these mapped to the coil Figure 1b and c shows that the 80 kDa DMPK band is not domain (Table 1). detectable in brain or lung, since many mAbs detect nothing at Figure 1a shows that the only protein recognized by nearly all. Some cross-reacting proteins are very prominent, espe- all mAbs on a western blot of human skeletal muscle is a single cially in brain. mAb MANDM1 is the only one to detect a band ∼ band of 80 kDa. Since these mAbs recognize several different close to the Mr of DMPK in lung and brain, but Figure 2a epitopes and they all recognize the DMPK recombinant fusion shows that this is a CRP of slightly lower Mr. The upper band protein (data not shown), this 80 kDa band is authentic DMPK. of authentic DMPK is detected only in cardiac and skeletal Many of the mAbs also recognize other proteins of higher or muscle whereas CRP is also present in lung, brain and three ∼ lower Mr, notably a 55 kDa protein, proteins of 200 kDa and different human skin fibroblast cell lines (F1–F3). Figure 2b a protein that migrates slightly faster than DMPK at ∼72 kDa shows that the upper band is also detectable by MANDM1 at [cross-reacting protein (CRP)]. Since these proteins are recog- very low levels in fetal stomach (smooth muscle), but this nized by only one or two of 16 mAbs, they must be cross- could not be detected by the whole panel of mAbs (data not reacting proteins with some sequence or structural similarity shown), either because the other mAbs are weaker or because to DMPK. The argument against the lower Mr bands being the band is not authentic DMPK. DMPK was also detected by Human Molecular Genetics, 2000, Vol. 9, No. 14 2169

not shown). Figure 3 shows that MRCKα is highly conserved between human and rat at both the mRNA and protein levels. Since the sequence of the protein kinase PK428 is identical to that of MRCKα, it may be a splicing isoform containing the complete catalytic domain (27). MANDM1 is therefore likely to recognize PK428 also. PK428 is reported to migrate on SDS–PAGE with an Mr of 65 kDa, so CRP in Figures 1 and 2 may be this protein kinase. It is possible, however, that PK428 is not detectably expressed as protein in any of the tissues tested here and that CRP is some other cross-reacting protein. Detailed epitope mapping was attempted by using phage- displayed random peptide libraries. A library of bacteriophage fd-tet expressing 15mer peptides was selected with mixtures of anti-DMPK mAbs as described previously (30) and selected clones were tested for reaction with individual mAbs in the mixture. The expressed peptide sequence was identified by DNA sequencing and matched with the known DMPK sequence (Fig. 4). MANDM2 and MANDM4 both recognized the sequence FDLVxDG near the end of the catalytic domain [only 5 amino acids away from the alternatively spliced VSGGG sequence (13)], whereas the two 55 kDa mAbs, MANDM7 and MANDM8, recognized the nearby sequence Figure 2. Western blot identification of DMPK and CRPs in human tissues and skin fibroblast cultures using mAb MANDM1. DMPK is detected only in heart (T)PD(F)E(G). Both sequences are significantly different in and skeletal muscle. The method followed was as in that in Figure 1 except that MRCKα and −β, explaining why these four mAbs do not 7% polyacrylamide was used to increase resolution and samples were cross-react with those proteins. GenBank was searched for loaded in slots. (a) F1 and F3, human skin fibroblast primary cell lines at TPDFEG (using Advanced Blastp with ‘word length = 2’ and low passage number; F2, human skin fibroblast line that has been passaged many times; Mu, human thigh muscle; Ht, human cardiac muscle (ventricle); Br, ‘expectation = 10 000’), in the hope of identifying the 55 kDa human brain (cortex); Lu, human lung. (b) FeS, human fetal stomach. FeS and protein, and two human proteins were found which contained Ht are on a different blot from the other three lanes. The 190 kDa MRCK is PDFEG. One was sarcoplasmic reticulum calcium ATPase and clearly present in muscle and lung [barely visible in (a), but clearly also present the second was the P100 subunit of nuclear factor NF-κB in Fig. 1a and b], but at lower levels than in brain and heart. Variation in band (Fig. 4), but neither of these are 55 kDa proteins. MANDM1 intensity between blots is partly due to different film exposure times, necessary appears to recognize the sequence RxIKxxxI in the middle of to optimize DMPK and its separation from CRP, though high Mr bands may have transferred less efficiently during blotting in (a). the catalytic domain and the RxIK sequence is conserved in MRCKα and −β, although not in Rho kinase, consistent with the observed cross-reactions of this mAb. The MANDM2/4 epitope sequence is altered in mouse DMPK (FDVVxDR the same panel of mAbs in fetal skeletal muscle, but not in fetal instead of FDLVxDG; SwissProt accession no. AAC60666) brain cortex, lung or liver (data not shown). Overall, the data whereas the MANDM1 epitope sequence is identical in mouse, suggest that DMPK is expressed mainly in skeletal and cardiac consistent with the species specificities of these mAbs. muscle. Lower levels may be present in smooth muscle, but DMPK levels were below the limits of detection in any other cell or tissue tested. DMPK mRNA has been detected by RT–PCR in DISCUSSION non-muscle cells (7,29) so the possibility that DMPK protein is Our studies on the distribution of DMPK in human tissues have also produced at very low levels in non-muscle cells cannot be shown that DMPK is produced in substantial amounts only in ruled out. heart and skeletal muscle. This conclusion agrees with in situ In order to identify some of the cross-reacting proteins, a hybridization studies that showed DMPK mRNA in heart, human brain cDNA library in the λZapexpressionvectorwas skeletal muscle and, at lower levels, smooth muscle (31). screened with a mixture of mAbs which displayed strong non- Several early northern blot studies also showed that DMPK DMPK bands on western blots of brain (Fig. 1b). Apart from mRNA was present in both heart and skeletal muscle, but diffi- several clones that reacted with all anti-DMPK mAbs (and cult to detect in other tissues (brain, lung, liver, etc.) (2,13,32). presumably contained DMPK sequences), two clones were This suggests that DMPK defects may be involved in the obtained that reacted with MANDM1 only. The first of these cardiac and skeletal muscle problems in DM patients. This contained the human MRCKβ sequence in the correct reading view is supported by the mild myopathy (31,33) and cardiac frame and identical to the published sequence (28; GenBank conduction defects (34) observed in DMPK knockout mice. β accession no. NM_006035). MRCK has an Mr of 190 kDa, Several studies of DMPK protein using antibodies or of very close to the 200 kDa band recognized by MANDM1 in DMPK mRNA by RT–PCR, however, have suggested that brain and other tissues (Fig. 1). The second clone contained DMPK is a widely distributed protein, found in many tissues MRCKα catalytic domain sequence (Fig. 3a), identified by and cultured cell lines. In an earlier study (25) we interpreted a comparison with the published rat sequence (26). Recombinant 72–80 kDa doublet as DMPK, though it is now clear that only full-length rat MRCKα wasusedonawesternblottoconfirm the upper 80 kDa band is authentic DMPK. Studies on rabbit binding of MANDM1 only and no other DMPK mAbs (data tissues could only be done with MANDM1, since other mAbs 2170 Human Molecular Genetics, 2000, Vol. 9, No. 14

Figure 3. The DNA sequence from a phage clone selected by MANDM1 matches rat MRCKα and includes human PK428. (a) The human PK428 sequence (lower case; GenBank accession no. U59305) is shown above the newly determined human MRCKα sequence (upper case; GenBank accession no. AF250871). Differ- ences in the rat MRCKα sequence (GenBank accession no. AF021935) are shown below. The rat sequence is numbered according to GenBank in which the coding sequence begins at nucleotide 530. The phage clone insert ended at the 3′ end shown, so no more of the sequence could be determined. (b) Translation to amino acid sequences. The rat sequence is numbered at E584 according to the SwissProt database. If the N-terminal sequence of human MRCKα is identical to human PK428, then this would also end at E584 (since PK428 ends at I496). are human specific and this led to confusion with CRP, which muscle tissues (data not shown). An alternative possibility is is abundant in non-muscle tissues (Fig. 2). This reinforces our that DMPK protein is more stable in muscle tissues and accu- assertion that only a panel of mAbs against different DMPK mulates to higher levels. The presence of even small amounts epitopes can ensure that authentic DMPK is being studied. The of DMPK in non-muscle tissues would suggest that it has some present work has been done with human tissues only, enabling function there, but the much higher levels in heart and muscle us to use mAbs against at least four different catalytic domain suggest a special function in these tissues. epitopes and at least four different coil domain epitopes (based Cross-reaction of mAb MANDM1 with MRCKα and −β has on epitope mapping and western blot cross-reactions). RT–PCR been confirmed unequivocally by its selection of these cDNAs detection of DMPK mRNA in non-muscle tissues may be from a λ human brain cDNA expression library and its binding explained by the greater sensitivity of this technique compared to recombinant MRCKα. The molecular basis for this cross- with western blotting. However, prolonged exposure of the X-ray reaction is explained by MANDM1 selection of peptides film for western blots still did not detect 80 kDa DMPK in non- containing RxIK from a phage-displayed random peptide Human Molecular Genetics, 2000, Vol. 9, No. 14 2171

other kinases, but PDFEG is highly conserved in sarcoplasmic reticulum calcium ATPases and is also found in human NF-κB, suggesting two potential sources of cross-reaction. However, both of these proteins are larger than 55 kDa, so the identity of the 55 kDa protein remains unknown, although we have shown that it is adult specific, as well as skeletal muscle specific, being absent from 16 day fetal muscle (data not shown). Some workers have tried to avoid cross-reaction by using the coil and/or C-terminal domains as immunogens (19–24), since these regions have little sequence homology with known proteins. However, various protein bands were also detected by such antibodies and many mAbs against the coil domain in Figure 1 also show cross-reacting bands. We suggest that DMPK migrates on SDS–PAGE at ∼80 kDa, since it migrates between 97 and 60 kDa Mr markers and rather closer to the former. The increase over the predicted size of 69 kDa may be due to post-translational modification, possibly involving a glycosaminoglycan moiety and the unique VSGGG sequence in DMPK (13). Mr should not be used as a criterion for identi- Figure 4. Epitope mapping using phage-displayed peptide libraries. In (a), fication of authentic DMPK since its migration in relation to epitopes recognized by mAbs MANDM7/8 and MANDM2/4 are almost adja- Mr markers may be affected by modification and electro- cent at the C-terminal end of the DMPK catalytic domain, whereas (b)shows phoresis conditions. In particular, our results do not exclude the MANDM1 epitope in the middle of the catalytic domain. 15mer peptide all ‘DMPK’ proteins previously reported as 70–74 kDa nor sequences are shown above the DMPK sequence with matching amino acids in do they mean that all ‘DMPK’ protein bands described as bold in the peptides and underlined in the DMPK sequence. The numbers of separate phage clones sequenced are shown in parentheses. Corresponding rat 80–85 kDa are authentic. The demonstration of several different and human MRCK sequences are shown below the DMPK sequence with DMPK epitopes should be the only acceptable criterion. amino acids which match the epitope underlined. The last amino acid in each It would clearly be valuable to use the mAbs that are mono- sequence is numbered as in the SwissProt database and DMPK numbering is specific on western blots (e.g. MANDM5) to localize DMPK based on the current SwissProt reference sequence NP_004400. in heart and skeletal muscle sections by immunofluorescence microscopy. We showed earlier that most of the mAbs stain intercalated discs in the heart, though rather faintly, but none of library, a sequence which is shared by MRCKα and −β.This them show any specific localization in skeletal muscle (25). sequence is also shared by PK428, which is a likely splicing This suggests that the mAbs only recognize denatured DMPK α on blots, but not native DMPK on frozen tissue sections. For isoform of MRCK and a candidate for the lower Mr CRP band on western blots (Fig. 2), but is not shared by other example, in native DMPK the ‘catalytic loop’ sequence recog- related kinases, such as p160 Rho kinase (35–37). Although nized by MANDM1 may be either inaccessible or unable to the three-dimensional structure of DMPK has not been deter- adopt the flexible structure which MANDM1 seems to recog- mined, comparison with human cAMP-dependent protein nize in free peptides and denatured DMPK. We have observed kinase A (Brookhaven pdb accession no. 1YDT) shows that with another globular enzyme, creatine kinase, that mAbs tend to recognize either the native or the denatured form, but rarely the MANDM1 epitope includes part of the conserved ‘catalytic both (39). Attempts to reveal DMPK by fixation or SDS treat- loop’ which accepts protons from phosphorylatable Ser/Thr ment of muscle sections (40), however, have not yet been residues(HRDLKPENinPKAandHRDIKPDNinDMPK) successful and new ‘native DMPK-specific’ mAbs may be (38). Two mAbs which recognize only DMPK in human skel- required for immunolocalization. etal muscle, MANDM2 and MANDM4, selected a peptide containing FDLVxDG from the phage library and this sequence is not shared by the MRCKs. MANDM2 and MATERIALS AND METHODS MANDM4 differ in their exact epitope specificity, since MANDM2 cross-reacts with several proteins in human brain Production of mAbs (Fig. 1b); this could be explained if MANDM4 tolerated less A panel of 12 mAbs against the catalytic and coil domains of deviation from the epitope sequence than MANDM2. Identi- DMPK has been described previously (25). A DMPK coil/ fication of the MANDM7/8 epitope is of particular interest C-terminal cDNA fragment, produced as previously described since these mAbs recognize a 55 kDa skeletal muscle-specific (25), was expressed as a thioredoxin fusion from pET32a and protein and react only weakly, if at all, with authentic DMPK purified by His-tag affinity chromatography according to the (Fig. 1a) (25). Both mAbs recognize recombinant DMPK supplier’s instructions (Novagen, Madison, WI). For immuni- produced in bacteria, however, which suggests that post- zation, it was precipitated with 50% ethanol and redissolved in translational modification may prevent recognition of DMPK 6 M urea in phosphate-buffered saline. A further four mAbs from human muscle. The MANDM7/8 epitope lies within 20 (MANDM13–16) were produced by immunization of BALB/c amino acids of the alternatively spliced VSGGG sequence mice and fusion of spleen cells with Sp2/0 myeloma cells as which is thought to influence post-translational modification described elsewhere (41). Hybridoma culture supernatants (13). This linear epitope sequence (TPDFEG) is absent from were tested by ELISA (41) and cell lines were established by 2172 Human Molecular Genetics, 2000, Vol. 9, No. 14 two rounds of cloning by limiting dilution. All four mAbs were 5. Taneja, K.L., McCurrach, M., Schalling, M., Housman, D. and Singer, mapped to the coil domain using DMPK subfragments as R.H. (1995) Foci of trinucleotide repeat transcripts in nuclei of myotonic dystrophy cells and tissues. J. Cell Biol., 128, 995–1002. previously described (25). 6. Wang, J.Z., Pegoraro, E., Menegazzo, E., Gennarelli, M., Hoop, R.C., Angelini, C. and Hoffmann, E.P. (1995) Myotonic dystrophy—evidence for SDS–PAGE and western blotting a possible dominant negative RNA mutation. Hum. Mol. Genet., 4, 599–606. 7. Hamshere, M.G., Newman, E.E., Alwazzan, M., Athwal, B.S. and Brook, SDS–PAGE and western blotting were carried out essentially J.D. (1997) Transcriptional abnormality in myotonic dystrophy affects as described elsewhere (42). Antibody-reacting bands were DMPK but not neighboring genes. Proc. Natl Acad. Sci. USA, 94, 7394–7399. visualized following development with a biotin–avidin detec- 8.Boucher,C.A.,King,S.K.,Carey,N.,Krahe,R.,Winchester,C.L., Rahman, S., Creavin, T., Meghji, P., Bailey, M.E.S., Chartier, F.L. et al. tion system for mouse immunoglobulin (Vectastain kit; Vector (1995) A novel homeodomain encoding gene is associated with a large Laboratories, Burlingame, CA) followed by a chemilumines- CpG island interrupted by the myotonic dystrophy unstable (CTG)n cent detection system (SuperSignal; Pierce, Rockford, IL). repeat. Hum. Mol. Genet., 4, 1919–1925. Recombinant rat MRCKα was a gift from T. Leung and L. Lim 9. Jansen, G., Bachner, D., Coerwinkel, M., Wormskamp, N., Hameister, H. and Wieringa, B. (1995) Structural organization and developmental (University of Singapore, Singapore). expression pattern of the mouse WD repeat gene DMR N9 immediately upstream of the myotonic dystrophy locus. Hum. Mol. Genet., 4, 843–852. Screening cDNA libraries 10. Harper, P.S. (1989) Myotonic Dystrophy, 2nd edn. W.B. Saunders, London, UK. A human brain cDNA library in the λZAP II vector 11. Fu, Y.H., Friedman, D.L., Richards, S., Pearlman, J.A., Gibbs, R.A., Pizzuti, A., (Stratagene, La Jolla, CA) was plated to produce plaques in a Ashizawa, T., Perryman, M.B., Scarlato, G., Fenwick, R.G. and Caskey, C.T. lawn of Escherichia coli, as described in the supplier’s (1993) Decreased expression of myotonin protein kinase messenger RNA and protein in adult form of myotonic dystrophy. Science, 260, 235–238. protocols. Lifts on nitrocellulose were screened with mAb 12. Mahadevan, M.S., Amemiya, C., Jansen, G., Sabourin, L., Baird, S., MANDM1 as described above for western blotting, except that Neville,C.E.,Wormskamp,N.,Segers,B.,Batzer,M.,Lamerdin,J.et al. diaminobenzidine was used as substrate. Positive plaques were (1993) Structure and genomic sequence of the myotonic dystrophy (DM amplified and cloned to homogeneity before recovery of the kinase) gene. Hum. Mol. Genet., 2, 299–304. cDNA inserts in pBluescript plasmids, which were then 13. Groenen, P.J.T.A., Wansink, D.G., Coerwinkel, M., van den Broek., W., Jansen, G. and Wieringa, B. (2000) Constitutive and regulated modes of subjected to dye-based automated DNA sequencing using splicing produce six major myotonic dystrophy protein kinase (DMPK) plasmid primers (Cambridge Bioscience, Cambridge, UK). isoforms with distinct properties. Hum. Mol. Genet., 9, 605–616. Novel human MRCKα sequence was submitted to GenBank 14. Brewster, B.S., Jeal, S. and Strong, P.N. (1993) Identification of a protein (accession no. AF250871). product of the myotonic dystrophy gene using peptide specific antibodies. Biochem. Biophys. Res. Commun., 194, 1256–1260. 15. Koga, R., Nakao, Y., Kurano, Y., Tsukahara, T., Nakamura, A., Ishiura, Phage-displayed peptide library S., Nonaka, I. and Arahata, K. (1994) Decreased myotonin protein kinase in the skeletal and cardiac muscles in myotonic dystrophy. Biochem. Bio- A library of random 15mer peptides in the fuse5 filamentous phys. Res. Commun., 202, 577–585. phage was generously provided by George P. Smith (University 16. Van der Ven, P.F.M., Jansen, G., Vankuppevelt, T.H.M.S.M., Perryman, of Missouri, Columbia, MO) and was subjected to reiterative M.B.,Lupa,M.,Dunne,P.W.,Terlaak,H.J.,Jap,P.H.K.,Veerkamp,J.H., selection with mAb mixtures as previously described (30). After Epstein, H.F. and Wieringa, B. (1993) Myotonic dystrophy kinase is a component of neuromuscular junctions. Hum. Mol. Genet., 2, 1889–1894. cloning to homogeneity, peptide insert sequences recognized by 17. Timchenko, L., Nastainczyk, W., Schneider, T., Patel, B., Hofmann, F. individual mAbs were determined by dideoxy DNA sequencing and Caskey, C.T. (1995) Full-length myotonin protein kinase (72 kDa) using Sequenase (Amersham International, Little Chalfont, UK). displays serine kinase activity. Proc. Natl Acad. Sci. USA, 92, 5366–5370. 18.Dunne,P.W.,Ma,L.,Casey,D.L.,Harati,Y.andEpstein,H.F.(1996) Localisation of myotonic dystrophy protein kinase in skeletal muscle and ACKNOWLEDGEMENTS its alteration with disease. Cell Motil. Cytoskeleton, 33, 52–63. 19.Maeda,M.,Taft,C.S.,Bush,E.W.,Holder,E.,Bailey,W.M.,Neville,H., We thank George P. Smith (University of Missouri) for the Perryman, M.B. and Bies, R.D. (1995) Identification, tissue-specific phage-display library, Thomas Leung and Louis Lim expression, and subcellular localisation of the 80 kda and 71 kda forms of α myotonic dystrophy kinase protein. J. Biol. Chem., 270, 20246–20249. (University of Singapore) for recombinant MRCK and the 20. Waring, J.D., Haq, R., Tamai, K., Sabourin, L.A., Ikeda, J.E. and Korneluk, Muscular Dystrophy Campaign for research grants. R.G. (1996) Investigation of myotonic dystrophy kinase isoform translocation and membrane association. J. Biol. Chem., 271, 15187–15193. 21. Whiting, E.J., Waring, J.D., Tama, K., Somerville, M.J., Hincke, M., REFERENCES Staines, W.A., Ikeda, J.E. and Korneluk, R.G. (1995) Characterization of myotonic dystrophy kinase (DMPK) protein in human and rodent muscle 1.Fu,Y.H.,Pizzuti,A.,Fenwick,R.G.,King,J.,Rajnarayan,S.,Dunne, and central nervous tissue. Hum. Mol. Genet., 4, 1063–1072. P.W.,Dubel,J.,Nasser,G.A.,Ashizawa,T.,Dejong,P.et al. (1992) An 22. Salvatori, S., Biral, D., Furlan, S. and Marin, O. (1997) Evidence for local- unstable triplet repeat in a gene related to myotonic muscular dystrophy. isation of the myotonic dystrophy protein kinase to the terminal cisternae Science, 255, 1256–1258. of the sarcoplasmic reticulum. J. Muscle Res. Cell Motil., 18, 429–440. 2. Brook, J.D., McCurrach, M.E., Harley, H.G., Buckler, A.J., Church, D., 23. Salvatori, S., Biral, D., Furlan, S. and Marin, O. (1994) Identification and Aburatani, H., Hunter, K., Stanton, V.P., Thirion, J.P., Hudson, T. et al. localisation of the myotonic dystrophy gene product in skeletal and car- (1992) Molecular basis of myotonic dystrophy—expansion of a trinucleo- diac muscles. Biochem. Biophys. Res. Commun., 203, 1365–1370. tide (CTG) repeat at the 3′ end of a transcript encoding a protein kinase 24. Shimokawa, M., Ishiura, S., Kameda, N., Yamamoto, M., Sasagawa, N., family member. Cell, 68, 799–808. Saitoh, N., Sorimachi, H., Ueda, H., Ohno, S., Suzuki, K. and Kobayashi, 3. Mahadevan, M., Tsilfildis, C., Sabouri, L., Shutler, G., Amemiya, C., T. (1997) Novel isoform of myotonin protein kinase gene product of Jansen,G.,Neville,C.,Narang,M.,Barcelo,J.,O’Hoy,K.et al. (1992) myotonic dystrophy is localized in the sarcoplasmic reticulum of skeletal Myotonic dystrophy mutation: an unstable CTG repeat in the 3′ untrans- muscle. Am.J.Pathol., 150, 1285–1295. lated region of the gene. Science, 255, 1253–1255. 25. Pham, Y.N.C., Nguyen thi Man, Le, L. and Morris, G.E. (1998) Localization 4. Wieringa, B. (1994) Myotonic dystrophy reviewed: back to the future. of myotonic dystrophy protein kinase in human and rabbit tissues using a new Hum. Mol. Genet., 3,1–7. panel of monoclonal antibodies. Hum. Mol. Genet., 7, 1957–1965. Human Molecular Genetics, 2000, Vol. 9, No. 14 2173

26. Leung, T., Chen, X.Q., Tan, I., Manser, E. and Lim, L. (1998) Myotonic 35. Leung, T., Chen, X.Q., Manser, E. and Lim, L. (1996) The p160 RHOA dystrophy kinase-related Cdc42-binding kinase acts as a Cdc42 effector in binding kinase ROK-alpha is a member of a kinase family and is involved promoting cytoskeletal reorganization. Mol. Cell. Biol., 18, 130–140. in the reorganization of the cytoskeleton. Mol. Cell. Biol., 16, 5313–5327. 27. Zhao, Y., Kidd, V. and Kraft, A.S. (1997) Cloning and chromosomal loca- 36. Ishizaki, T., Maekawa, M., Fujisawa, K., Okawa, K., Iwamatsu, A., Fujita, tion of a novel member of the myotonic dystrophy family of protein A., Watanabe, N., Saito, Y., Kakizuka, A., Morii, N. and Narumiya, S. kinases. J. Biol. Chem., 272, 10013–10020. (1996) The small GTP binding protein RHO binds to and activates a 160 28. Moncrieff, C.L., Bailey, M.E.S., Morrison, N. and Johnson, K.J. (1999) kda Ser/Thr protein kinase homologous to myotonic dystrophy kinase. EMBO J., 15, 1885–1893. Cloning and chromosomal localization of human Cdc42-binding protein 37. Matsui, T., Amano, M., Yamamoto, T., Chihara, K., Nakafuku, M., Ito, kinase beta. Genomics, 57, 297–300. M.,Nakano,T.,Okawa,K.,Iwamatsu,A.andKaibuchi,K.(1996)Rho- 29. Jansen, G., Bartolomei, M., Kalscheuer, V., Merkx, G., Wormskamp, N., associated kinase, a novel serine/threonine kinase, as a putative target for Mariman, E., Smeets, D., Ropers, H.H. and Wieringa, B. (1993) No small GTP binding protein Rho. EMBO J., 15, 2208–2216. imprinting involved in the expression of DM-kinase mRNAs in mouse 38. Smith, C.M., Shindyalov, I.N., Veretnik, S., Gribskov, M., Taylor, S.S., and human tissues. Hum. Mol. Genet., 2, 1221–1227. Ten Eyck, L.F. and Bourne, P.E. (1997) The protein kinase resource. 30. Pereboev, A. and Morris, G.E. (1996) Reiterative screening of phage- Trends Biochem. Sci., 22, 444–446. display peptide libraries with antibodies. In Morris, G.E. (ed.), Methods in 39. Nguyen thi Man, Cartwright, A.J., Andrews, K. and Morris, G.E. (1989) Molecular Biology, Vol. 66, Epitope Mapping Protocols. Humana Press, Treatment of human muscle creatine kinase with glutaraldehyde prefer- Totowa, NJ, pp. 195–206. entially increases the immunogenicity of the native conformation and 31. Jansen, G., Groenen, P.J.T.A., Bachner, D., Jap, P.H.K., Coerwinkel, M., permits production of high-affinity monoclonal antibodies which recog- Oerlemans, F., Vandenbroek, W., Gohlsch, B., Pette, D., Plomp, J.J. et al. nize two distinct surface epitopes. J. Immunol. Methods, 125, 251–259. (1996) Abnormal myotonic dystrophy protein kinase levels produce only 40. Nguyen thi Man, Ginjaar, H.B., van Ommen, G.J.B. and Morris, G.E. (1992) mild myopathy in mice. Nature Genet., 13, 316–324. Monoclonal antibodies for dystrophin analysis: epitope mapping and 32. Sabourin, L.A., Mahadevan, M.S., Narang, M., Lee, D.S., Surh, L.C. and improved binding to SDS-treated muscle sections. Biochem. J., 288, 663–668. Korneluk, R.G. (1993) Effect of the myotonic dystrophy (DM) mutation 41. Nguyen thi Man and Morris, G.E. (1996) Production of panels of monoclonal antibodies by the hybridoma method. In Morris, G.E. (ed.), Methods on mRNA levels of the DM gene. Nature Genet., 4, 233–238. in Molecular Biology, Vol. 66, Epitope Mapping Protocols. Humana Press, 33. Reddy, S., Smith, D.B., Rich, M.M., Leferovich, J.M., Reilly, P., Davis, Totowa, NJ, pp. 377–389. B.M., Tran, K., Rayburn, H., Bronson, R., Cros, D. et al. (1996) Mice 42. Nguyen thi Man, Ellis, J.M., Love, D.R., Davies, K.E., Gatter, K.C., lacking the myotonic dystrophy protein kinase develop a late onset pro- Dickson, G. and Morris, G.E. (1991) Localisation of the DMDL gene- gressive myopathy. Nature Genet., 13, 325–335. encoded dystrophin-related protein using a panel of nineteen monoclonal 34. Berul, C.I., Maguire, C.T., Aronovitz, M.J., Greenwood, J., Miller, C., antibodies: presence at neuromuscular junctions, in the sarcolemma of Gehrmann, J., Housman, D., Mendelsohn, M.E. and Reddy, S. (1999) dystrophin skeletal muscle, in vascular and other smooth muscles, and in DMPK dosage alterations result in atrioventricular conduction abnormal- proliferating brain cell lines. J. Cell Biol., 115, 1695–1700. ities in a mouse myotonic dystrophy model. J. Clin. Invest., 103, R1–R7. 2174 Human Molecular Genetics, 2000, Vol. 9, No. 14