Structure-function relationships in dystrophin and ... - Semantic Scholar

Muscular Dystrophy and Allied Disorders: Research Strategies

Structure-function relationships in dystrophin and utrophin S. J. Winder MRC Centre, Hills Road, Cambridge CB2 2QH, U.K. 497

Introduction T h e integrity of the animal cell membrane is believed to be maintained in part by the large (approx. 400 kDa) cytoskeletal proteins dystrophin and utrophin. Both proteins are capable of forming a link, probably flexible and potentially extensible/compressible, between the actin cytoskeleton and the cell membrane itself. Utrophin is expressed in all cell types, whereas dystrophin expression is restricted to muscle and tissues of neuronal origin. Mutations in these genes, utrophin on human chromosome 6 and dystrophin on the X chromosome, would be expected to have severe consequences for the cell. Mutations in the dystrophin gene are known, giving rise to the Duchenne and Becker muscular dystrophy phenotypes (DMD and BMD respectively). Mutations that lead to the functional absence of dystrophin protein - promoter defects, premature stop codons and other missense mutations generally present as the more common and severe DMD phenotype. Deletions, truncations and substitutions that maintain the reading frame of the DNA and allow the expression of altered dystrophin protein generally lead to the milder but less common BMD phenotype. These rarer mutations are therefore much more informative in ascribing functional importance to the dystrophin protein, as disease severity can be correlated with the position of any mutations in the protein. Despite the correspondingly large size of the utrophin gene, no mutations in utrophin have yet been found, leading to the suggestion that the protein is essential for life. An understanding of the functions of the various regions of the dystrophin protein and its close relative, utrophin, will provide useful information regarding the potential use of utrophin in gene therapy approaches in DMD and BMD. Dystrophin and utrophin share considerable sequence and presumed functional similarity; Abbreviations used: BMD, Becker muscular dystrophy; DMD, Duchenne muscular dystrophy; ABS, actin binding site; DAP, dystrophin-associated protein. Present address: Institute of Cell and Molecular Biology, University of Edinburgh, Michael Swann Building, King’s Buildings, Mayfield Road, Edinburgh EH9 3JR, Scotland, U.K.

each possesses an N-terminal F-actin-binding domain, a series of triple-helical coiled-coil repeats making up approx. 75% of each protein and a C-terminal region comprising several domains involved in protein-protein interactions and possibly signalling functions. In-frame deletions in the central triple-helical coiled-coil regions, by far the most common, present with the mildest BMD phenotypes, whereas mutations in the N-terminal region present with severe BMD and mutations in the C-terminal region lead to DMD. On this basis it can be seen that the regions in dystrophin that form the connections to the actin cytoskeleton and to the membrane are the most important, with the central (spacer/flexible) region to some extent being redundant. The present paper will therefore concentrate on recent structural and functional analysis of the N- and C-terminal domains, with details of the central coiled-coil region presented elsewhere [ 1,2].

N-terminal actin-binding domains T h e low concentrations of both dystrophin and utrophin in cells has made the biochemical characterization of these proteins difficult. Although small quantities of both proteins have been isolated, there has generally been insufficient material for thorough biochemical analyses. Most workers have therefore chosen to overexpress dystrophin and utrophin domains in bacteria in order to generate sufficient protein for analysis. We have expressed the putative actin-binding domains from both dystrophin and utrophin in Escherichia coli and characterized their interactions with actin [3]. Expressed dystrophin and utrophin actin-binding domains have similar affinities for skeletal muscle F-actin and non-muscle F-actin, although the affinity for non-muscle actin is approx. fourfold higher in each case. Earlier experiments with a utrophin actin-binding domain (utrophin residues 1-261, UTR261) containing a conservative substitution (Thr to Ser) within the actin binding site (ABS) 2 [3], bound F-actin with 19 pM affinity and a stoichiometry of 2 mol of utrophin/mol of F-actin [3]. Subsequent analysis has shown that a nonfusion utrophin actin-binding domain with the wild-type sequence (UTR261+) also bound to

I996

Biochemical Society Transactions

498

F-actin with 19 pM affinity but with a stoichiomol Of utrophin/mol Of F-actin (see metry Of Figure l), in keeping with the observed stoichiometries for the related dystrophin and cx-actinin actin-binding domains '1 * It is possible therefore that the COIlServatiVe substitution Of a threonine for serine within ABS2 has a profound effect on the stoichiometry of binding without affecting affinity' this unlikely. Thus, despite the (approx. twofold) difference in affinity for F-actin, the actinbinding domains of dystrophin and utrophin can be considered to be functionally equivalent, in as much as they simply bind to F-actin and do not cap, cross-link or sever actin filaments or affect actin polymerization kinetics. T h e affinities of dystrophin and utrophin actin-binding domains for F-actin are shown in Table 1. One difference between the actin-binding domains of dystrophin and utrnphin is in their interaction with calmodulin. Whereas calmodulin binds to both expressed actin-binding domains in the presence of calcium but not in its absence [6,7], only the binding of utrophin to F-actin is affected by cal-

Figure l Binding of UTR26 I

+ to F-actin

Binding curve for bacterially expressed utrophin residues 1-26 I , wild-type sequence (UTR261 +), in the presence of 10 p M Factin (mean of two separate experiments). F-actin sedimentation assays were performed as previously described and quantified by densitometric analysis of SDS/polyacrylamide gels [3], Inset: parabound t o F-actin with an meters from curve fitting. UTR261 apparent stoichiometry of I , I (ml) and affinity of I 9 f2 p M (m2). MO is a function of x ; Chisq, chi squared; R is the regression fit of the data to the calculated line; NA. not applicable.

+

1.2 A

c

1 1 rL - 0.8 t

-Ef

1

\

8 + r 8

n

y - m l +(MO/(MO+mZ))

0.6

I

Value1

Error

0.4

0.2

p:

0

50

100 150 200 250 300 350 400

[UTR261+] pM

Volume 24

Table I

Affinities of expressed dystrophin and utrophin actinbinding domains for skeletal and non-muscle actins The affinities of the dystrophin actin-binding domain, residues 1-246 (DMD246), and utrophin actin.blnding domain, residues 1-261 ( ~ ~ ~ and 2 6 ~1 ~ ~ +), 2 for 6 1skeletal and non-muscle (platelet) F-actin were determined in saturation sedimentation Factin-binding assays as described previously [3]. UTR261 is the cleaved product of a glutathione S-transferase fusion construct containing a threonine for serine substitution at residue I 13: UTR261 is a non-fusion product of wild-type sequence. Afinities are in pM fS.E.M.; ND, not determined.

+

Actin-binding construct

Skeletal-muscle actin

Non-muscle actin

DMD246 UTR26 I UTR261

44*

I 3 k2t 5+ I t ND

+

19+3t 19+2

*From Way et al. [5]. +From Winder et al. [3].

cium/calmodulin [7]. In sedimentation F-actinbinding assays, half-maximum displacement of 20 pM UTR261 from 10 pM F-actin occurred at 50 pM calcium/calmodulin [7], whereas in the absence of calcium UTR261 was not displaced from F-actin even at a 10-fold molar excess of calmodulin. T h e dystrophin actin-binding domain, however, was not displaced in a calciumdependent manner [6,7]. These findings have important implications for the organization and regulation of dystrophin and utrophin within the cytoskeleton of muscle cells. T h e finding that dystrophin binds to non-muscle actin with approx. fourfold higher affinity than to skeletal muscle actin suggests that dystrophin actually binds to a cortical network of non-muscle actin within the muscle cell rather than directly to the contractile apparatus, as has previously been suggested. T h e anchoring of dystrophin to a continuous network of non-muscle actin overlying the contractile apparatus is infinitely more appealing and also fits with immunofluorescence data showing a continuous network of dystrophin staining beneath the muscle cell membrane [&lo]. Furthermore, if dystrophin was anchored to the contractile machinery, then gaps in the network would be apparent in relaxed muscles corresponding to the bare zone of the A-band where there are no thin (actin) filaments; this situation does not arise.


C-terminal domains T h e increasing collection of motifs and domains that constitute the C-terminal regions of dystrophin and utrophin are predominantly involved in the association of these proteins with components of a group of intracellular, transmembrane and extracellular proteins collectively known as the dystrophin-associated proteins (DAPs). It is via these DAPs that dystrophin and utrophin are anchored at the cell membrane and are effectively linked to the extracellular matrix. T h e most recent additions to this complex region are the WW and ZZ domains. Protein sequence alignments of these two domains from dystrophin, utrophin and other proteins are shown in Figure 2. T h e WW domain [13,14], so called because of its delimiting and conserved tryptophan residues (single-letter code W), forms the first part of the C-terminal region of dystrophin and utrophin and effectively marks the boundary between the end of the triple-helical coiled-coil domain and the start of the C-terminal domains. Although the WW domains in dystrophin and utrophin have not as yet been directly implicated in any protein-protein interaction, this is believed to be their function [15,16]. T h e WW domain of the Yes-associated protein (YAP) has recently been shown to associate with prolinerich sequences in two putative ligands, WBP-1 and WBP-2 [15]. T h e involvement of a consensus XPPXY motif for recognition by the WW domain is reminiscent of the consensus PXXP motif for recognition by SH3 domains. Although clearly distinct, these similarities have drawn obvious speculation about a role for the WW domain in targeting of proteins to the plasma membrane or to the cytoskeleton or in mediating the targeting of proteins to their substrates, all of which would fit with a role for WW in dystrophin and utrophin in mediating interactions with the DAPs (see below). T h e ZZ domain [17], so called by analogy with the previously described WW domain and because of its potential for binding two zinc ions, is a putative zinc finger located within the cysteine-rich region of dystrophin, utrophin, the related 87-kDa acetylcholine receptor-associated protein and other unrelated proteins (Figure 2) [17]. T h e putative zinc finger has the form Cys)4-Cy~-)4~-Cys-&-Cys, with the zinc ion co-ordinated to the four cysteine residues, as shown schematically in Figure 2(C). T h e related 87-kDa acetycholine receptor-associated protein, which also has a ZZ domain but with six cysteine

residues, may co-ordinate two zinc ions, as shown (Figure 2C). Although a specific function for the ZZ domain has yet to be found, it is assumed that the putative zinc finger is involved in protein-protein interactions, in a manner analogous to the cytoskeletal zinc finger-containing protein, zyxin [ 181, and not protein-nucleic acid interactions. Zyxin contains three copies of the LIM (C. efeguns Lin-11, rat Isl-1, C. Efeguns Mec-3) zinc finger domain, which are involved in the association of zyxin with ct-actinin and cysteine-rich protein. T h e ZZ domain is equivalent to (although clearly distinct from) one half of a single LIM domain, with a complete LIM Figure 2

WW and ZZ domains in dystrophin and utrophin (A) WW domains of dystrophin and utrophin. Amino acid sequences of the WW domains of dystrophin and utrophin and the consensus, taken from [13], with residues matching the consensus shown in bold type. (6)ZZ domains of dystrophin, utrophin and other proteins. Amino acid sequences of the Z Z domains of dystrophin. utrophin and other sequences and the consensus, derived from [ 171, with cysteine residues potentially able to coordinate to zinc shown in bold type; h represents hydrophobic residues. Abbreviations: 87kDa, 87-kDa acetylcholine receptorassociated protein; Drn ref(Z)P, Drosophilo melonogoster ref(2)P gene product [ I I]; Sc ADA2, Socchorornyces cereviseoe ADA2 gene product [12]. (C) Schematic representation of the ZZ domains of dystrophin and utrophin (left) with one zinc ion (Zn) co-ordinated to four cysteine (C) residues and the ZZ domains of the 87-kDa acetylcholine receptor-associated protein, Dm ref(2)P and Sc ADA2 with two zinc ions co-ordinated to the six cysteines and each other. Numbers in the loops represent numbers of amino acids. R Dystrcphin T S V Q G P N L R A I S P N K V P Y Y I M E E T Q T T C N D H P K M T E L Y 3 8 9 IltTSVQLPNQRSISHNKVPYYIIPBQZQTQTTCNDHPKMTELF 28M

. .LP .GNE.. ....G. . Y Y .PB .T .T T . Y . .P . . .. . .

Ccnsensus

a 5 p t . e

gLm e

reff2)p

scAap2

AKCNICKECPIIGFRYRSLKHFNYDICQSCFF AKCNICKECPIVGFRYRSLKHFNYDVCQSCFF V E C SYCRCESMMGFRYRCQQCHNYQLCQNCFW VECDGCGLAPLIGFRYKCVQCSNYDLCQKCES F HCDVCSADCTNRVRVSCAICPEYDLCVPCVS h.C. .C..

.. h . .

3342 3093 278 154 36

.RYRC. . C DYDLC. . C . .

c

Dystrophin Utrophln

87kDa ref(2)P ADA2

I996

499

Biochemical Society Transactions

500

domain forming a double zinc finger, each coordinating one zinc ion. T h e target sequence for the 22 domain has not been identified and awaits further direct experimentation to identify a putative ligand and its mode of interaction. As discussed below, however, one good candidate already exists. From gel overlay assays of /J-dystroglycan (a 43 kDa transmembrane glycoprotein DAP) with various glutathione S-transferase fusion constructs spanning parts of the dystrophin C-terminus, Ozawa and co-workers [ 19-21] identified a region within the C-terminus of dystrophin involved in binding to /J-dystroglycan. T h e delineated region encompasses both the WW and 22 domains, and thus both domains may be required for the interaction of dystrophin (and utrophin) with P-dystroglycan. A recently described mutation in dystrophin that affects the 22 domain led to the severe DMD phenotype [22], circumstantial evidence that this region of dystrophin is important for the interaction with the DMs. T h e mutation caused an in-frame deletion of exons 68 and 69 of dystrophin by exon skipping, resulting in the deletion of the 22 domain. One other piece of evidence for the involvement of the 22-domain putative zinc finger in binding to P-dystroglycan is that the acetylcholine receptor-associated protein rapsyn, which also associates with B-dystroglycan and in an apparently mutually exclusive way with respect to utrophin [23], also contains a zinc finger (although not a 22 type finger). T h e zinc finger in rapsyn is essential for rapsyn function [24], the clustering of acetylcholine receptors, and is mediated via P-dystroglycan [24]. It is therefore possible that the zinc finger in rapsyn and the putative 22-domain zinc finger in dystrophin (utrophin) both associate with the same region of P-dystroglycan.

Conclusions It is clear that dystrophin and utrophin form a link between the actin cystoskeleton and the cell membrane and via the membrane-anchored dystrophin-associated proteins through to the extracellular matrix. Disruptions in this link, especially in the case of dystrophin, result in membrane damage and necrosis. Genetic analysis has revealed the relative importance of the N- and C-terminal domains in this link with a less important role played by the central coiledcoil region; however, structural and functional

Volume 24

analyses of dystrophin and utrophin are beginning to reveal their fundamental roles in maintaining the integrity of the cell membrane.

I am grateful to Dr. J. Kendrick-Jones, MRC Laboratory of Molecular Biology, Cambridge, where this work was carried out, for his enthusiasm, encouragement and guidance throughout the course of these studies. I am also grateful to the Muscular Dystrophy Group of Great Britain and Northern Ireland for financial support. 1 Winder, S. J., Gibson, T. J. and Kendrick-Jones, J. (1996) Biochem. SOC.Trans. 24,280s 2 Winder, S. J., Gibson, T. J. and Kendrick-Jones, J. (1995) FEBS Lett. 369, 27-33 3 Winder, S. J., Hemmings, L., Maciver, S. K., Bolton, S. J., Tinsley, J. M., Davies, K. E., Critchley, D. R. and Kendrick-Jones, J. (1995) J. Cell Sci. 108, 63-71 4 Way, M., Pope, B. and Weeds, A. G. (1992) J. Cell Biol. 119, 835-842 5 Way, M., Pope, B., Cross, R. A., Kendrick-Jones,J. and Weeds, A. G. (1992) FEBS Lett. 301,243-245 6 Bonet-Kerrache, A., Fabbrizio, E. and Mornet, D. (1994) FEBS Lett. 355, 49-53 7 Winder, S. J. and Kendrick-Jones, J. (1995) FEBS Lett. 357, 125-128 8 Masuda, T., Fujimaki, N., Ozawa, E. and Ishikawa, H. (1992) J. Cell Biol. 119, 543-548 9 Porter, G. A., Dmytrenko, G. M., Winkelmann, J. C. and Bloch, R. J. (1992) J. Cell Biol. 117, 997- 1005 10 Straub, V., Bittner, R. E., Leger, J. J. and Voit, T. (1992) J. Cell Biol. 119, 1183-1191 1 1 Dezelee, S., Bras, F., Contamine, D., LopezFerber, M., Segretain, D. and Teninges, D. (1989) EMBO J. 8, 3437-3446 12 Berger, S. L., Pina, B., Silverman, N., Marcus, G. A., Agapite, J., Regier, J. L., Triezenberg, S. J. and Guarente, L. (1992) Cell 70, 251-265 13 Bork, P. and Sudol, M. (1994) Trends Biochem. Sci. 19, 531-533 14 Andre, B. and Springael, J. Y. (1994) Biochem. Biophys. Res. Commun. 205, 1201-1205 15 Chen, H. I. and Sudol, M. (1995) Proc. Natl. Acad. Sci. U.S.A. 92, 7819-7823 16 Sudol, M., Bork, P., Einbond, A., Kastury, K., Druck, T., Negrini, M., Huebner, K. and Lehman, D. (1995) J. Biol. Chem. 270, 14733-14741 17 Ponting, C. P., Blake, D. J., Davies, K. E., Kendrick-Jones, J. and Winder, S. J. (1996) Trends Biochem. Sci. 21, 11-13 18 Schmeichel, K. L. and Beckerle, M. C. (1994) Cell 79,211-219 19 Suzuki, A., Yoshida, M., Yamamoto, H. and Ozawa, E. (1992) FEBS Lett. 308, 154-160 20 Suzuki, A., Yoshida, M., Hayashi, K., Mizuno, Y.,


Hagiwara, Y. and Ozawa, E. (1994) Eur. J. Biochem. 220,283-292 21 Suzuki, A., Yoshida, M. and Ozawa, E. (1995) J. Cell Biol. 128, 373-381 22 Tuffery, S., Lenk, U., Roberts, R. G., Coubes, C., Demaille, J. and Claustres, M. (1995) Hum. Mutat. 6, 126-135 23 Apel, E. D., Roberds, S. L., Campbell, K. P. and

Merlie, J. P. (1995) Neuron 15, 115-126 24 Scotland, P. B., Colledge, M., Melinkova, I., Dai, 2. and Froehner, S. C. (1993) J. Cell Biol. 123, 7 19-728 50 I

Received 13 November 1995

Costameric distribution of fl-dystroglycan (43 kDa dystrophin-associated glycoprotein) in normal and dystrophin-deficient human skeletal muscle R. Herrmann*, L. V. 6. Andenont and T. Voit*$ *Department of Paediatrics, University of Essen, Hufelandstrasse 55, Essen, Germany, and +University School of Neuroscience and Muscular Dystrophy Group Laboratories, Newcastle upon Tyne, U.K.

Introduction Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy are caused by different mutations of the dystrophin gene. Its gene product, the 430 kDa protein dystrophin [l], is tightly associated with the skeletal muscle sarcolemma [2-41. Dystrophin is predicted to be a 125 to 145 nm rod-shaped protein [5-71, which may be organized as dimer, and shares significant sequence homologies with other cytoskeletal proteins such as spectrin and a-actinin [8-111. T h e precise function of dystrophin is still not known, but two actin binding sites at the Nterminal portion and the tight connection at the C-terminus to a group of membrane proteins suggest a mechanical role within the submembrane cytoskeleton connecting the plasma membrane to the sarcomeres [ 121. This dystrophin-associated glycoprotein (DAG) complex can be subdivided into three groups. At the cytoplasmic face of the sarcolemma is a triplet of 59 kDa dystrophin-associated proteins, called syntrophins, which bind dystrophin at the C-terminus [13-151. A group of transmembrane proteins, the sarcoglycan complex, is composed of a 50 kDa DAG, adhalin or a-sarcoglycan; a 43 kDa DAG, /3-sarcoglycan; a 35 kDa DAG, y-sarcoglycan; and possibly a 25 kDa dystrophin-associated protein [ 16,171. Primary and secondary deficiency of adhalin has been found in severe, childAbbreviations used: DMD, Duchenne muscular dystrophy; DAG, dystrophin-associated protein; WB, Western blot; IF, immunofluorescence; FITC, fluorescein isothiocyanate; T R , Texas Red. +To whom correspondence should be addressed.

hood, autosomal recessive muscular dystrophy (SCARMD, LGMD2D or LGMDZC) [ 18-25]. T h e expression of the sarcoglycan complex is restricted to cardiac and skeletal muscle [26]. T h e main connection between the contractile apparatus and the extracellular matrix may be formed by the dystroglycan complex. This complex is composed of 156 kDa DAG (a-dystroglycan), which is located at the extracellular site of the sarcolemma and which binds laminin-2 (merosin), a major component of the extracellular matrix [27], and 43 kDa DAG (P-dystroglycan), a transmembrane glycoprotein that binds on one side to a-dystroglycan and on the cytoplasmic side of the sarcolemma to the cysteine-rich domain of dystrophin [27,28]. As such, dystrophin may act as one link of a lattice connecting the extracellular matrix to the sarcomeres. These two proteins of the dystroglycan complex are expressed ubiquitously in several tissues and are transcribed from a single gene [29,30]. It has been suggested that the dystroglycan complex not only has a mechanical role in muscle but may also be involved in signal transduction via interaction with the adapter protein Grb2 [31-331. Protein studies have shown that a lack of or vast reduction in dystrophin in DMD muscle leads to a secondary loss of DAG expression at the muscle sarcolemma [34]. This loss of DAG is supposed to be a pathogenic factor ultimately leading to muscle necrosis by further disruption of the connection of sarcomeres to the extracellular matrix [ 121. However, DMD patients with large in-frame deletions of the dystrophin gene and preserved expression of the dystroglycan-binding dystrophin domain are not less affected clinically than those lacking dystro-

I996