FASEB J. Pierce now sold as Thermo Scientific
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.06-7644com.
 (The FASEB Journal. 2007;21:1383-1392.)
© 2007 FASEB
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
fj.06-7644comv1
21/7/1383    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mrosek, M.
Right arrow Articles by Mayans, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mrosek, M.
Right arrow Articles by Mayans, O.

Molecular determinants for the recruitment of the ubiquitin-ligase MuRF-1 onto M-line titin

Michael Mrosek*, Dietmar Labeit{ddagger}, Stephanie Witt{ddagger}, Heiko Heerklotz{dagger}, Eleonore von Castelmur*, Siegfried Labeit{ddagger} and Olga Mayans*,1

* Division of Structural Biology and

{dagger} Chemical Biophysics, Biozentrum, University of Basel, Klingelbergstr 70, Basel, Switzerland;

{ddagger} Institut für Anästhesiologie und Operative Intensivmedizin, Universitätsklinikum Mannheim, Mannheim, Germany

1Correspondence: Division of Structural Biology Biozentrum, University of Basel, Klingelbergstr. 70, CH-4056 Basel, Switzerland. E-mail: olga.mayans{at}unibas.ch


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Titin forms an intrasarcomeric filament system in vertebrate striated muscle, which has elastic and signaling properties and is thereby central to mechanotransduction. Near its C-terminus and directly preceding a kinase domain, titin contains a conserved pattern of Ig and FnIII modules (IgA168-IgA169-FnIIIA170, hereby A168-A170) that recruits the E3 ubiquitin-ligase MuRF-1 to the filament. This interaction is thought to regulate myofibril turnover and the trophic state of muscle. We have elucidated the crystal structure of A168-A170, characterized MuRF-1 variants by circular dichroism (CD) and SEC-MALS, and studied the interaction of both components by isothermal calorimetry, SPOTS blots, and pull-down assays. This has led to the identification of the molecular determinants of the binding. A168-A170 shows an extended, rigid architecture, which is characterized by a shallow surface groove that spans its full length and a distinct loop protrusion in its middle point. In MuRF-1, a C-terminal helical domain is sufficient to bind A168-A170 with high affinity. This helical region predictably docks into the surface groove of A168-A170. Furthermore, pull-down assays demonstrate that the loop protrusion in A168-A170 is a key mediator of MuRF-1 recognition. Our findings indicate that this region of titin could serve as a target to attempt therapeutic inhibition of MuRF-1-mediated muscle turnover, where binding of small molecules to its distinctive structural features could block MuRF-1 access.—Mrosek, M., Labeit, D., Witt, S., Heerklotz, H., von Castelmur, E., Labeit, S., Mayans, O. Molecular determinants for the recruitment of the ubiquitin-ligase MuRF-1 onto M-line titin


Key Words: elastic filament titin • muscle atrophy • X-ray crystallography • binding studies


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE STRIATED MUSCLE TISSUE from vertebrates undergoes constant remodeling in adaptation to its mechanical load (reviewed in 1). Adaptive remodeling influences muscle tissue architecture, fiber-type composition, enzymatic apparatus, and innervation. Accordingly, numerous signaling pathways have been identified recently, which are controlled by myofibrillar stretch-strain. A prime candidate for the mediation of mechanical signaling in myocytes is the giant protein titin, the only sarcomeric component that forms a continuous filament system spanning full myofibrils, which can measure their tension and contains signaling domains (reviewed in 2).

Titin is composed of up to 33000 aminoacids, spanning in situ ~2 µm, from the Z-disc to the central M-line of the sarcomere (2) . It folds into a total of 195 immunoglobulin (Ig) and 132 fibronectin-III (FnIII) domains, which together amount to over 90% of its mass. To date, structures of polydomain tandems of titin are limited to the Ig-doublet Z1Z2 from the N terminus of the filament, which has been elucidated in its free (3) and telethonin-bound states (4) . Titin contains one single catalytic domain, a Ser/Thr kinase (TK), located near its C-terminus at the sarcomeric M-line. TK has been implicated in myofibrillogenesis (5) as well as in control of gene expression and protein turnover in mature muscle (6) . The deformation of the M-line during sarcomere stretch has lead to suggest that TK could become activated via mechanically-induced conformational changes (7) , thereby exerting its action on vicinal titin-associated proteins (6) . The latter are thought to shuttle to the nucleus, where they interact further with regulators of gene expression (6 , 8) . According to this hypothesis, titin mediates a net translation of mechanical strain into chemical signaling.

Proteins associated to TK include MURF-1, MURF-2 (as shown in situ and in vitro; 8, 9), and nbr1 and p62 (in vitro data; 6). MuRF-1 binds to titin domains IgA168-IgA169-FnIIIA170 (hereby A168-A170), located just prior to TK (Fig. 1 A) (9) . MuRF-1 can heteromerize with MuRF-2, allegedly through coiled-coil segments (9) . In vitro, MuRF-2 binds p62, and the latter nbr1, which interacts with TK (6) . However, with the exception of MuRF-1, it is not currently known to what extent theseproteins could form stable assemblies on the titin filament. MuRFs are muscle-specific proteins with a RING-finger/B-box/coiled-coil tripartite fold known as TRIM (Fig. 1B ) (10) . Although three MuRF isoforms (MuRF-1, -2, -3; encoded by different genes; 9) have been identified to date, only MuRF-1 has been shown to interact directly with titin. MuRF-1 is an E3 ubiquitin-ligase that becomes up-regulated during muscle loss and has been proposed as a molecular mediator of muscle atrophy (11) . Muscle atrophy is characterized by an accelerated protein degradation via the ubiquitin proteasome system and it is an associated condition to numerous diseases, including cancer, diabetes and AIDS, as well as physiological states such as long-term immobilization and microgravity. A role of MuRF-1 in muscle turnover has been confirmed by a knockout mouse model that showed increased resistance to muscle atrophy (11) , which suggests the potential pharmacological relevance of this protein. Yet, the cellular function of MuRF-1 might be complex to modulate, since it interacts with a broad range of cellular factors, including the ubiquitin-like modifier SUMO-3 (12) , isopeptidase T-3, and the SUMO E2 ligase Ubc9 (8) , which thereby might exert their influence on the titin cytoskeleton. Furthermore, nuclear forms of MuRF-1 interact with the transcription regulator GMEB-1 (8) . Taken together, current functional data on MuRF-1 suggest its involvement in turnover and muscle gene expression, possibly forming part of mechanotransduction pathways in control of the titin composition of the sarcomere which, in turn, determines mechanical stiffness. Thus, in order to gain an insight into the structure of M-line titin and its recruitment of MuRF-1, we have elucidated the crystal structure of A168-A170 and interpreted it in the context of its binding to MuRF-1. Our findings aid the understanding of the molecular determinants of the interaction and point to a therapeutic interest of this region of titin.


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		Figure 1. Domain composition of M-line titin and MuRF-1. A) M-line titin. Ig and FnIII domains are represented by filled boxes, TK by a globe. The construct under study, A168-A170, is colored; (B) MuRF-1 domain composition. The extent of ring finger, MuRF-specific (MFC) and B-box domains is as reported (9) . The helical fraction, including an expected coiled-coil motif, has been predicted from sequence data and is in orange. The sequence previously suggested as primary titin binding site (26) is boxed in blue. The C-terminal acidic tail (AT) is in red.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cloning
For crystallographic studies, A168-A170 from human titin (residues 24431–24731, EMBL X90568) was cloned into the pETM-11 vector (EMBL collection) at restriction sites NcoI and MIuI. pETM-11 is a variant of pET-24d (Novagen, Madison, WI, USA), which includes an N-terminal His6-tag and a TEV (tobacco etch virus) protease cleavage site prior to the inserted gene. Pull-down assays used A168-A170 cloned in pETM-44 (EMBL collection), which provided the N-terminal fusion of maltose binding protein (MBP) to the titin fragment. A mutated version of this construct, where the 102KTLE105 motif had been replaced by AAAA, was obtained using the QuikChange protocol (Stratagen, La Jolla, CA, USA).

Fragments from MuRF-1 (Swiss-Prot Q969Q1), MuRF-1(166–341) and MuRF-1(166–315), were inserted in pETM-11 via NcoI and KpnI restriction sites. Both constructs correspond to a C280S mutated variant designed to prevent unspecific aggregation of the samples due to oxidation during storage.

Protein production
Overexpression of A168-A170 for structural analysis was carried out in E. coli BL21(DE3) Rosetta (Novagen). Cultures were grown at 37°C up to an OD600 of 0.6 in Luria Bertani medium supplemented with 25 µg/ml kanamycin and 34 µg/ml chloramphenicol. Expression was induced by addition of isopropyl-ß-D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM. Cultures were further grown at 25°C for ~18 additional hours. Cells were harvested by centrifugation at 2800 g and 4°C. Bacterial pellet was resuspended in 50 mM Tris-HCl, pH 8.0; 150 mM NaCl; containing a protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany) and DNase I. Lysis was by addition of lysozyme and sonication. The homogenate was clarified by centrifugation at 15000 g and 4°C. The supernatant was applied to a Ni2+-chelating HisTrap column (GE Healthcare) equilibrated in lysis buffer and eluted using 250 mM imidazole. The eluent was dialyzed against 50 mM Tris-HCl, pH 8.0; 150 mM NaCl; and 2 mM DTT in the presence of TEV protease. Given that A168-A170 still interacted with the Ni2+-chelating resin after tag removal, the digested mixture was further purified by gel filtration on a Superdex 200 Hiload 16/60PG column (GE Healthcare) equilibrated in dialysis buffer. The samples were finally dialyzed against 25mM MOPS, pH 8.0; 150 mM KCl; 1 mM DTT, 1 mM EGTA; and stored at 4°C for further use.

MuRF-1 fragments were prepared as above, but purification of the protease digest used subtractive affinity chromatography prior to gel filtration.

Structure elucidation
Crystals grew in hanging drops at 20°C from solutions containing 2.1 M NaCl; 0.1 M MES, pH 5.5; and 0.6 M Li2SO4, mixed in a 1:1 ratio with a 12 mg/ml protein solution. Crystals were in the habit of long rods, hexagonal in cross section with approximate dimensions of 600 x 100 x 100 µm3 and reached their final size over a 2 month period.

For X-ray data collection, crystals were frozen at 100 K in mother liquor solutions supplemented with 20% [v/v] glycerol as cryoprotectant. X-ray data processing was in XDS (13) , and data statistics and crystal parameters are given in Table 1 . Crystals contained two copies of A168-A170 per asymmetric unit, which were related by an imperfect two-fold axis approximately colocalized with the crystallographic z-axis. Since attempts at phasing by molecular replacement were unsuccessful, experimental phases were obtained by SIR combined with a 3{lambda} MAD experiment on a crystal derivatized with [Ta6Br12]2+ by soaking for 30 min at saturation. Derivative data statistics are given in Table 3 . Anomalous scatterers (2 sites) were located using CNS (14) and phases calculated and refined in SHARP (15) (Table 3) . Density modification used solvent flattening, histogram matching, and NCS domain averaging in DM (16) . Manual model building was in O (17) . Model refinement applied overall anisotropic B-factor scaling, bulk solvent correction, and grouped B-factor minimization as implemented in CNS (14) . NCS restraints were applied across individual domains throughout refinement and moderately loosened in the final cycle of refinement (resulting rmsd values were 0.33 Å for C{alpha} atoms across copies of domain IgA168, 0.25 Å for IgA169, and 0.24 Å for FnA170). For refinement, experimental data were divided into a working and a test set using FREERFLAG (16) . Solvent molecules were built using the water-pick routine in CNS and validated visually in electron density maps. The final model includes all protein residues of A168-A170, at exception of the 12 C-terminal residues that were disordered in both NCS copies. Two N-terminal residues in positions –1 and –2, remnants of the TEV cleavage site, could be observed and are part of the final model.


View this table:
[in this window]
[in a new window]

  		Table 1. Native X-ray data and model refinement statistics


View this table:
[in this window]
[in a new window]

  		Table 2. Domain arrangement in poly-domain fragments from titin


View this table:
[in this window]
[in a new window]

  		Table 3. [Ta6Br12]2+ X-ray data and phasing statistics

Model coordinates and experimental X-ray data have been deposited at the Protein Data Bank with accession code 2NZI.

Circular dichroism
MuRF-1(166–315) was extensively dialyzed against 10 mM sodium tetraborate, pH 8.5; 20 mM NaCl; and concentrated up to 0.1 mg/ml as determined by BCA (Pierce, Rockford, IL, USA) assay. Data were acquired at 25°C on a 62A DS circular dichroism spectropolarimeter (AVIV) equipped with a temperature-controlled quartz cell of 1 mm optical path. The spectrum was recorded at a scan speed of 50 nm/min with a response time of 5 s and averaged over 8 scans. Protein conformation was calculated by fitting the experimental data with a combination of reference spectra for known secondary structure elements (18) .

Size exclusion chromatography combined with multiangle light scattering (SEC-MALS)
The oligomeric state of MuRF-1(166–315) in solution was determined via SEC-MALS measurements performed on an ÄKTA basic system (GE Healthcare) connected to a triangle static light-scattering detector and a differential refractometer (miniDAWN Tristar and Optilab, respectively). A Superdex 200 HR 10/300 GL column (GE Healthcare) was used in 50 mM Tris, pH 8.0, and 100 mM NaCl at a flow rate of 0.7 ml/min. Sample volumes of 100 µl were injected at a concentration of 2 mg/ml. A specific refractive index increment (dn/dc) of 0.185 ml/g was used. The data were recorded and processed using ASTRA software (Wyatt Technology, Santa Barbara, CA, USA). To determine the detector delay volumes and normalization coefficients for the light scattering detector, a BSA sample (Pierce) was used as reference.

Isothermal titration calorimetry
Data were recorded with a VP-ITC calorimeter (Microcal LLC, Northhampton, MA, USA) at 25°C running Origin v7 software. All samples were previously dialyzed against 50 mM Tris-HCl, pH 8.0, and 20 mM NaCl. Titrations consisted of 5–7 µl injections of A168-A170 concentrated to 1.4 mg/ml into MuRF-1(166–341) or MuRF-1(166–315) solutions at 1 mg/ml and 0.7 mg/ml, respectively (as determined by BCA assay; 19). The time interval between injections was 200–300 s. The data were corrected for the heat of dilution by subtraction of the small constant heat obtained at the end of the titration, where no further binding occurs.

Binding assays
Titin residues mediating MuRF-1 binding were revealed by using a SPOTS blot membrane (JPT, Berlin, Germany) that displayed A168-A170 as a series of 32 overlapping peptides of 20-aa length. Peptides were acetylated at their N terminus to enhance stability. The membrane was initially washed with ethanol and abundantly with TBS, followed by an overnight blocking step in TBS supplemented with 5% (w/v) milk powder. For binding, MuRF-1(166–315) was added at a final concentration of 50 µg/ml and reacted overnight. Unbound proteins were removed by washes of TBS supplemented with 0.05% (v/v) Tween (Sigma, St. Louis, MO, USA). Bound fragments were reacted for 3 h with specific antibodies against MuRF-1 (rabbit polyclonal, 0.5 µg/ml). Bound antibodies were then detected using the Elite Kit (Vectastain) employing biotinylated anti-rabbit IgG (PK6101), and the detection kit LumiLight Plus (Roche Diagnostics) used as indicated by the manufacturer.

Pull-down assays used MBP-A168–170 fusion products, both wild-type and as a KTLE to AAAA mutated variant, which had been immobilized onto amylose resin (New England Biolabs, Beverly, MA, USA). Purified MuRF-1(166–315) was flowed through the column and incubated for 1 h. MBP in isolation was tested as control. Mixtures were washed with 20 column volumes of buffer solutions containing 20 mM Tris-HCl, pH 7.4; 200 mM NaCl; 1 mM EDTA; and 10 mM ß-mercaptoethanol. Elution used 10 mM maltose.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Crystal structure of A168-A170
The structure of A168-A170 from human titin has been elucidated at 2.9 Å resolution using X-ray crystallography. Data statistics and model parameters are given in Table 1 . A168-A170 comprises two Ig domains (IgA168-IgA169) and one FnIII (FnA170). The molecule adopts an extended conformation with an end-to-end distance of ~115 Å, where the two N-terminal Ig lie almost perfectly coaxial but the FnIII domain is bent away from the molecular axis (Fig. 2 A). Domain arrangements are described in Table 2 . The crystal form used in this study contains two A168-A170 copies in its asymmetric unit. These are essentially identical in conformation (rmsd 0.68 Å for 288 C{alpha} atoms of the whole molecule, calculated with SPDBV; 20), suggesting that this fragment of titin has a well-defined long-range order, and a high degree of stiffness.


Figure 2
View larger version (33K):
[in this window]
[in a new window]

  		Figure 2. Crystal structure of A168-A170. A) Ribbon representation. The insertion loop in IgA169 is displayed in cyan. The inset shows the molecular conformation schematically; (B) Surface representation in two related views, where green emphasizes the local curvature. A ridge is present in the concave side of the molecule. The distinct loop of IgA169 is marked; (C) Docking of an {alpha}-helical peptide in coiled-coil conformation onto the surface of A168-A170 to satisfy shape complementarity. The helical peptide shown derives from lamin A coil 2B (PDB code 1X8Y).

Domains IgA168 and IgA169 belong to the I(intermediate)-set of Ig folds (21) . They share 20% seq. id. and a high structural similarity (rmsd 0.99 Å for 82 matching C{alpha} atoms, SPDBV). Both belong to the "N-conserved" type of Ig from titin (22) , characterized by a conserved N-terminal loop cluster comprising proline residues prior to ß-strand A, a PxP motif in the BC loop, and an extended FG ß-hairpin hosting an NxxG sequence (Fig. 3 C). This Ig type predominates in the Z-disc, the skeletal I-band and the A-band fractions of titin, but not in its constitutive cardiac I-band. Domains IgA168 and FnA170 show no significant deviation from other equivalent modules across titin. However, IgA169 shows a unique 9-residue loop protrusion between ß-strands A and A’ with sequence PKTLEGMGA, which adopts in part an {alpha}-helical conformation (Figs. 2A and Fig. 5 A). According to structure and sequence data on titin domains, a loop insertion at this position is not detectable in any other Ig of titin.


Figure 3
View larger version (61K):
[in this window]
[in a new window]

  		Figure 3. Domain interfaces. A) IgA168-IgA169 and (B) IgA169-FnA170 interfaces. Domain color code as in Fig. 2 A. Hydrophobic contributions are shown in green. Selected van der Waals surfaces are displayed. The C{alpha} atoms of residues in the NxxG motif of the FG ß-turns of both IgA169 and FnA170 are shown as spheres. Hydrogen bonds are indicated by dotted lines; C) Structure-based sequence alignment of titin Ig and FnIII of known structure (PDB accession codes are given). ß-Strand composition is shown in yellow; residues at Ig-Ig and Ig-FnIII interfaces are highlighted in red and cyan, respectively. Features characteristic of "N-conserved" Ig (A168, A169, I1, Z1, Z2 and M5) are shown in blue. I91 was formerly I27. Strand nomenclature for Ig as in (21) and for FnIII as in (31) . The distinct loop of IgA169 is shown in gray; D) A-band of titin. Ig and FnIII domains are shown as filled boxes, where Ig domains flanked by FnIII are in brown, FnIII following an N-terminal Ig in gray, other FnIII in white and Ig followed by another C-terminal Ig in blue. FnA71, the only other FnIII of titin with known structure (1BPV), is in cyan. The two residues in variable positions of the NxxG ß-turn of FnIII domains are displayed. Domain super-repeats are indicated. The pattern of alternating charges is displayed, where red and blue indicate negative and positive charges, respectively.


Figure 5
View larger version (35K):
[in this window]
[in a new window]

  		Figure 5. Binding of MuRF-1 to A168-A170. A) SPOTS blots of peptide fragments from A168-A170 assayed vs. MuRF-1(166–315). Data corresponding to the insertion motif in IgA169 are boxed in red; B) Mapping of interacting SPOTS peptides (orange) onto the molecular surface of A168-A170. They cluster at the concave face (left), while only one peptide (peptide 18) can be mapped to the convex side of the molecule (right), which otherwise remains unmatched. The distinct loop insertion of IgA169 is marked; C) Binding of MuRF-1 to A169-A170 monitored by ITC. The experimental pattern of injections of MuRF-1(166–341) is shown (upper panel). Heats of binding for titrations of A168-A170 into MuRF-1(166–341) (filled circles) and MuRF-1(166–315) (open triangles) recorded in 50 mM Tris-HCl, pH 8.0; 20 mM NaCl at 25°C is displayed (lower panel). Both curves show the same Kd and {Delta}H and differ only slightly in the active fraction of MuRF-1; D) Superimposition of Ig domains from titin for which a structural model is available, namely IgA169 (cyan), IgA168 (green), I1 (magenta; PDB accession code 1G1C), M5 (yellow; 1NCT), Z1 (salmon; 2A38) and Z2 (gray, 2A38). The unique loop insertion of IgA169 is clearly identifiable. The C{alpha} atoms of its component residues are displayed as solid spheres and the sequence given. I91 (formerly I27), which does not belong to this Ig subgroup and has shorter loop regions, is excluded from the superimposition to ease visualization. As the other Ig, I91 does not comprise loop insertions in its N-terminal ß-strand region. The KTLE motif altered by site-site directed mutagenesis is in red; E) SDS-PAGE of a pull-down assay where MBP-A168-A170 was immobilized in affinity resin in the presence of MuRF-1(166–315). Lane 1 shows isolated MBP as control; lane 2 reveals the binding of MuRF-1 to wild-type A168-A170; lane 3 employed tetra-mutated A168-A170, where the KTLE motif had been disrupted, with the resulting loss of MuRF-1 binding.

Domain interfaces
The domains of A168-A170 form a tight tandem. They are connected through one-residue linkers in extended conformation that effectively result in the continuation of structural elements from one domain into the next. The extended arrangement of the molecule is characterized by domain interfaces with unusually small buried areas (Table 2) and engages a minimal number of residues in each domain. Both Ig-Ig and Ig-FnIII interfaces are void of specific interactions and only involve small hydrophobic clusters.

Domains IgA168-IgA169 interact through hydrophobic groups that surround the linker strand dorsally and ventrally (Fig. 3A ). Given their coaxial orientation, the interacting residues originate from the linker region (V92, aliphatic portion of E93, V94) and the C-terminus of ß-strand A’ (aliphatic portion of R13, Y14) in IgA168 and the conserved elements of this Ig type in IgA169, namely the FG turn (F180) and the PxP motif of the BC loop (P128, K127) (Fig. 3C ). A comparison of IgA168-IgA169 and Z1Z2 (the only other Ig doublet of titin structurally characterized to date) (3 , 4) reveals certain similarities. Also Z1Z2 exhibit an extended conformation, lack specific interdomain contacts, and include a small hydrophobic cluster between modules. As that of IgA168-IgA169, the cluster of Z1Z2 comprises residues C-terminal to ß-strand A’, the BC loop and linker residues (Fig. 3C ). Despite, Z1Z2 displays a different domain orientation (Table 2) and has a longer, three-residue linker that allows certain modular dynamics (3) . It can then be concluded that the mere presence of elements characteristic of the "N-conserved" Ig type does not determine domain orientations in these doublets. Since the identified interface residues are not conserved in Ig across titin (23) , it cannot be predicted at this point how closely the observed conformations represent those of other tandems. Yet, the overall structural principles of Ig arraying revealed by these studies might be generic to the titin filament as indicated by the fact that all poly-Ig arrays of titin characterized to date by small angle X-ray scattering have revealed comparable extended arrangements independently of their domain or linker composition (22) .

The interface of domains IgA169-FnA170 exhibits a hydrophobic component as that of IgA168-IgA169 (Fig. 3b ). Similarly, contacts involve ß-strand A’ (H111) in the initial domain IgA169 and the NxxG motif of ß-turn FG in FnA170 (K272, F273) (Fig. 3C ). However, in contrast to Ig, the FG turns in FnIII of titin are highly conserved in sequence, as revealed by the alignment of 123 FnIII sequences of A-band titin (CLUSTALW; 23). This analysis showed that those FnIII domains following an Ig within the 11-domain super-repeats of the central C-zone region of A-band titin, supposedly part of the myosin-based thick filament, consistently host a conserved residue pair in the variable positions of their NxxG motif (Fig. 3D ). These invariably include a charged residue followed by an aromatic group, most often tyrosine or phenylalanine. Since these residues are part of the modular interface, Ig-FnIII pairs from the C-zone of titin can be expected to exhibit a similar interdomain conformation as that of A169-A170. Strikingly, the charges within this motif are distinctly distributed along the C-zone repeats, where every first FnIII hosts a negative charge and the rest a positive group (Fig. 3D ). Such segregation is not observed in the 7-domain super-repeats of the preceding D-zone (Fig. 3D ). In the crystal structure, the lateral amino group of K272 is not involved in interactions, suggesting its availability for binding to other sarcomeric proteins. Myosin-binding protein C (MyBP-C) binds titin by interacting with the first Ig domain of every super-repeat in the C-zone, so that titin possibly aids thereby the regular polymerization of myosin filaments (25) . Given that the FG ß-turn of FnA170 is located toward the intermodular space, its charge might be speculated to contribute to MyBP-C docking. Interestingly, the FG turns of FnIII domains at other positions within repeats of the D- and C-zones also exhibit certain conservation of their sequence motifs (Fig. 3D ), suggesting that these might be generically important for FnIII-FnIII arrangements and/or interactions to other A-band components.

MuRF-1 binding
The surface of A168-A170 reveals a shallow groove at the concave side of the molecule that spans its full length (Fig. 2B ). This could be expected to accommodate the binding counterpart of MuRF-1. Based on SPOTS blots of MuRF-1 peptides, the latter has been proposed to bind titin through a sequence just prior to its acidic C-terminal tail (Fig. 1B ) (26) . According to our own analysis (Jpred software; 27), the C-terminal half of MuRF-1 (residues 166–315, which include the proposed binding sequence) has high tendency to {alpha}-helical formation, with at least its middle section predictably following a coiled-coil association (COILS; 28) (Fig. 1B ). This is supported by CD data on a correspondingly truncated MuRF-1 variant, MuRF-1(166–315), whose spectrum exhibits a maximum at 190 and minima at 208 and 222 nm indicative of a predominant {alpha}-helical composition (Fig. 4 A). An estimation of the secondary structure content of this sample based on these data suggested an approximate {alpha}-helical content of 70% and 30% random coil (fitting parameter R2=99.8%). This is in excellent agreement with results from both manual examination and docking prediction software (PatchDock; 29), which indicated that the dimensions and geometry of the surface groove in A168-A170 can well accommodate an {alpha}-helix along its length. In particular, a best docking is obtained if the helix exhibits a long-range bending of its axis as that of components of a dimeric coiled-coil. For example, as seen in human lamin A coil 2B (30) used as a prototypic structural template to probe the surface of A168-A170 in this study (Fig. 2C ). To explore further the interaction area between titin and MuRF-1, we investigated their binding by SPOTS blots of peptide fragments of A168-A170 assayed against the central helical fraction of MuRF-1, MuRF-1(166–315). The results confirm that all three modules in A168-A170 are involved in the interaction (Fig. 5A ). Furthermore, the mapping of the selected peptides onto the crystal structure of A168-A170 showed that, with the exception of one peptide, all interacting fragments colocalized at the concave face of the molecule (Fig. 5B ). Thus, it could be concluded that this is the primary MuRF-1 binding interface.


Figure 4
View larger version (15K):
[in this window]
[in a new window]

  		Figure 4. Biophysical characterization of the titin-binding domain of MuRF-1. A) CD spectrum of MuRF-1(166–315) exhibiting features characteristic of a sample with a predominantly {alpha}-helical content; B) SEC-MALS measurements of MuRF-1(166–315). Average molecular mass per vol unit (gray) and normalized refractive index (continuous line) are shown. The theoretical mass of a dimer of this construct is indicated by a dashed horizontal line.

Further, we have established the interaction between A168-A170 and the helical fraction of MuRF-1 in solution and quantified it using isothermal calorimetry (ITC). This study used two N-terminally truncated MuRF-1 constructs, MuRF-1(166–341) and MuRF-1(166–315) (Fig. 1B ), where the latter lacks in addition the acidic C-terminal tail. Binding monitored by ITC showed that both MuRF-1 variants interacted strongly with A168-A170, yielding heat release values of 14 and 28 kcal/mol and Kd affinities of 35 and 37 nM, respectively (Fig. 5C ). Since Kd values are similar for both constructs, it can be concluded that the acidic tail of MuRF-1 does not influence binding but that this is determined by its helical fraction. It was also found that saturation occurs at a molar ratio of ~0.1, which indicates that ~10 MuRF-1 molecules bind one A168-A170. Such high ratio, however, is unusual and points to either a high oligomerization state of MuRF-1 variants or possible technical problems in the estimation of this parameter by ITC. To gain a further insight into this issue, we investigated the association state of the helical fraction of MuRF-1 in isolation using size-exclusion chromatography combined with multiangle static light scattering (SEC-MALS). This technique yields an accurate determination of the molecular mass (MM) without being influenced by molecular shape or hydrodynamic parameters. SEC-MALS measurements of MuRF-1(166–315) yielded a MM of 32.3 kDa (protein concentration estimated from refractive index), which is in excellent agreement with the calculated MM of 34.7 kDa for a dimer of this sample (Fig. 4B ). This supports predictions based on sequence data that suggest a self-association of MuRF-1 via a dimeric coiled-coil segment (COILS software; 28) (Fig. 1B ). Since A168-A170 is monomeric according to structural and chromatographic data, it could be expected that the basic stoichiometry of the titin/MuRF-1 complex is 1:2 (or a multiple of this ratio). Unfortunately, we have not been able to investigate directly the composition of titin/MuRF-1 assemblies by other biophysical techniques during this study. Mixing and storage of complexed samples at concentrations and in conditions compatible with biophysical analysis led to precipitation that rendered the preparations unsuitable for study. Thus, at this time, we cannot discard the possibility that the ratio measured by ITC might represent a more complex scenario in which, on binding, conformational changes take place in the individual components that result in a higher level of assembly. However, the differing ITC value is more likely to be of technical origin. Since ITC curves were consistent over the several repetitions of the recordings and for both constructs, MuRF-1(166–341) and MuRF-1(166–315), the measured stoichiometry cannot be due to random or variable effects, such as unspecific aggregation of MuRF-1 samples. The latter, for which no experimental evidence exists, would have caused drifts in the saturation point of the different ITC measurements depending on sample batch, storage times and conditions. Instead, ITC data most probably reflect a consistent error in estimating the concentration of MuRF-1 fragments in this study. The MuRF-1 helical domain has a low content of aromatic residues and calculation of its concentration required of the BCA colorimetric method, whose response to different proteins can vary widely. Moderate, cumulative errors in quantifying the amounts of MuRF-1 variants and/or A168-A170 could well explain the deviation of recorded ITC values (~0.1) from the expected (~0.5) titin/MuRF-1 ratio based on the oligomeric state of the individual components.

Structural and biochemical data here presented indicate that MuRF-1 and titin interact through an extensive contact area. However, in A168-A170 the unique loop insertion of IgA169 is located at the midpoint of its surface groove (Fig. 2B, C ) and it could be expected to play a central role in MuRF-1 binding or function. Effectively, SPOTS blots confirmed that this motif is involved in MuRF-1 interaction (Fig. 5A ). To further assess the relevance of this feature in the binding, we constructed a tetra-mutated variant of A168-A170, where residues 102-KTLE-105 had been exchanged for 102AAAA105, and tested its interaction to MuRF-1(166–315) using pull-down assays (Fig. 5D ). Results confirmed that alteration of this unique motif abolishes the interaction with MuRF-1 and, therefore, that this is central to its molecular recognition by A168-A170.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The muscle filament titin, despite its colossal dimensions, has a simple and repetitive architecture consisting of linear tandems of Ig and FnIII modules. The structure of A168-A170 reveals the molecular details of Ig-Ig and Ig-FnIII interfaces, where the latter are likely to be generic to Ig-FnIII domain pairs along titin. A168-A170 has a pronouncedly extended conformation and exhibits a well-defined long-range order with limited conformational freedom. This seems to result from extremely short linker sequences and the presence of small hydrophobic clusters that restrict the modular orientations. Specific contacts are not found across any of these domain pairs. Hydrophobic interactions include groups from similar structural elements in Ig and FnIII modules, with residues around the linker region and the FG ß-turn being involved in both cases (Fig. 3C ). The observed Ig-Ig contacts are poorly conserved (Fig. 3C ). In agreement, domain orientations in IgA168-IgA169 and Z1Z2 differ somewhat (Table 2) . On the contrary, sequence conservation at the Ig-FnIII interface indicates that this modular conformation might be common to equivalent pairs from the A-band. It is yet to be established whether the conservation of interface groups in this case, in particular the FG ß-turn of FnIII domains, might play an additional role in interactions to other sarcomeric proteins—such as MyBP-C (25) .

A168-A170 selectively recruits MuRF-1 to the M-line region of titin. The specificity of the interaction appears achieved through two idiosyncratic features of this fragment of titin: i) the topography of a shallow groove spanning the length of its surface and defined by the long-range domain arrangement of this distinct tandem; and ii) the unique loop insertion of IgA169. ITC data show that the titin/MuRF-1 interaction is of high affinity, where the {alpha}-helical region of MuRF-1 (residues 166–315) is sufficient for binding. Based on the oligomeric state of individual components, we propose the stoichiometry of the titin/MuRF-1 complex to be 1:2. SPOTS blots support the view that the groove in A168-A170 constitutes the binding interface to MuRF-1 as predicted from the crystal structure. The groove could host up to 70% of the maximal hypothetical coiled-coil length achievable by MuRF-1, if its full helical region was to form such motif. The fact that the interaction involves more than one modular unit in titin suggests that this filament uses long-range steric determinants, provided by its interdomain arrangements, to achieve specificity in the recruitment of other proteins to distinct locations of its otherwise repetitive frame. Finally, we have identified a 102KTLE105 insertion motif in IgA169 as primary molecular determinant of MuRF-1 interaction by pull-down assays on a A168-A170 mutated variant. So that even if the interacting surface in the MuRF-1/titin complex appears extensive, the binding could be modulated by targeting this key motif. The high affinity of the interaction and the fact that this is primarily governed by a defined sequence motif suggest that the A168-A170 region of titin could serve as a potential therapeutic target against muscle atrophy, where binding of small molecules to its distinctive structural features could block the access of MuRF-1.


   ACKNOWLEDGMENTS
 
Our gratitude goes to B. Sheppard and D. Nurizzo (ESRF) for their support during X-ray data collection and to A. Tsamaloukas (Biozentrum) for help with ITC. M.M. was supported by SNF grant (3100A0–100852) to OM. This project was further financed by DFG (La668/7–2 and La1969/1–1).

Received for publication October 31, 2006. Accepted for publication December 6, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Bassel-Duby, R., Olson, E. N. (2006) Signaling pathways in skeletal muscle remodeling. Annu. Rev. Biochem. 75,19-37[CrossRef][Medline]
  2. Granzier, H. L., Labeit, S. (2004) The giant protein titin: a major player in myocardial mechanics, signaling, and disease. Circ. Res. 94,284-295[Abstract/Free Full Text]
  3. Marino, M., Zou, P., Svergun, D., Garcia, P., Edlich, C., Simon, B., Wilmanns, M., Muhle-Goll, C., Mayans, O. (2006) The Ig-doublet Z1Z2: a model system for the hybrid analysis of conformational dynamics in Ig tandems from titin. Structure 14,1437-1447[Medline]
  4. Zou, P., Pinotsis, N., Lange, S., Song, Y. H., Popov, A., Mavridis, I., Mayans, O., Gautel, M., Wilmanns, M. (2006) Palindromic assembly of the giant muscle protein titin in the sarcomeric Z-disk. Nature 439,229-233[CrossRef][Medline]
  5. Mayans, O., van der Ven, P. F., Wilm, M., Mues, A., Young, P., Furst, D. O., Wilmanns, M., Gautel, M. (1998) Structural basis for activation of the titin kinase domain during myofibrillogenesis. Nature 395,863-869[CrossRef][Medline]
  6. Lange, S., Xiang, F., Yakovenko, A., Vihola, A., Hackman, P., Rostkova, E., Kristensen, J., Brandmeier, B., Franzen, G., Hedberg, B., et al (2005) The kinase domain of titin controls muscle gene expression and protein turnover. Science 308,1599-1603[Abstract/Free Full Text]
  7. Grater, F., Shen, J., Jiang, H., Gautel, M., Grubmuller, H. (2005) Mechanically induced titin kinase activation studied by force-probe molecular dynamics simulations. Biophys. J. 88,790-804[CrossRef][Medline]
  8. McElhinny, A. S., Kakinuma, K., Sorimachi, H., Labeit, S., Gregorio, C. C. (2002) Muscle-specific RING finger-1 interacts with titin to regulate sarcomeric M-line and thick filament structure and may have nuclear functions via its interaction with glucocorticoid modulatory element binding protein-1. J. Cell Biol. 157,125-136[Abstract/Free Full Text]
  9. Centner, T., Yano, J., Kimura, E., McElhinny, A. S., Pelin, K., Witt, C. C., Bang, M. L., Trombitas, K., Granzier, H., Gregorio, C.C., et al (2001) Identification of muscle specific ring finger proteins as potential regulators of the titin kinase domain. J. Mol. Biol. 306,717-726[CrossRef][Medline]
  10. Meroni, G., Diez-Roux, G. (2005) TRIM/RBCC, a novel class of single protein RING finger E3 ubiquitin ligases. Bioessays 27,1147-1157[CrossRef][Medline]
  11. Bodine, S.C., Latres, E., Baumhueter, S., Lai, V.K., Nunez, L., Clarke, B.A., Poueymirou, W.T., Panaro, F.J., Na, E., Dharmarajan, K., et al (2001) Identification of ubiquitin ligases required for skeletal muscle atrophy. Science 294,1704-1708[Abstract/Free Full Text]
  12. Dai, K. S., Liew, C. C. (2001) A novel human striated muscle RING zinc finger protein, SMRZ, interacts with SMT3b via its RING domain. J. Biol. Chem. 276,23992-23999[Abstract/Free Full Text]
  13. Kabsch, W. (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J. Appl. Cryst. 26,795-800[CrossRef]
  14. Brunger, A.T., et al (1998) Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. Allogr. D54,905-921[CrossRef]
  15. de la Fortelle, E., Bricogne, G. (1997) Maximum-likelihood heavy-atom parameter refinement for multiple isomorphous replacement and multiwavelength anomalous diffraction methods. Methods Enzymol. 276,472-494
  16. . CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr. Allogr. D50,760-763
  17. Jones, T.A., Zou, J.Y., Cowan, S. W., Kjeldgaard, M. (1991) Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. Allogr. A47,110-119[CrossRef]
  18. Yang, J. T., Wu, C. S., Martinez, H. M. (1986) Calculation of protein conformation from circular dichroism. Methods Enzymol 130,208-269[Medline]
  19. Brown, R. E. (1989) Protein measurement using bicinchoninic acid: elimination of interfering substances. Anal. Biochem. 180,136[CrossRef][Medline]
  20. Guex, N., Peitsch, M. C. (1997) SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis 18,2714-2723[CrossRef][Medline]
  21. Harpaz, Y., Chothia, C. (1994) Many of the immunoglobulin superfamily domains in cell adhesion molecules and surface receptors belong to a new structural set which is close to that containing variable domains. J. Mol. Biol. 238,528-539[CrossRef][Medline]
  22. Marino, M., Svergun, D. I., Kreplak, L., Konarev, P. V., Maco, B., Labeit, D., Mayans, O. (2005) Poly-Ig tandems from I-band titin share extended domain arrangements irrespective of the distinct features of their modular constituents. J. Muscle. Res. Cell. Motil. 26,355-365[CrossRef][Medline]
  23. Witt, C. C., Olivieri, N., Centner, T., Kolmerer, B., Millevoi, S., Morell, J., Labeit, D., Labeit, S., Jockusch, H., Pastore, A. (1998) A survey of the primary structure and the interspecies conservation of I-band titin’s elastic elements in vertebrates. J. Struct. Biol. 122,206-215[CrossRef][Medline]
  24. Higgins, D., Thompson, J., Gibson, T., Thompson, J. D., Higgins, D. G., Gibson, T. J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22,4673-4680[Abstract/Free Full Text]
  25. Freiburg, A., Gautel, M. (1996) A molecular map of the interactions between titin and myosin-binding protein C. Implications for sarcomeric assembly in familial hypertrophic cardiomyopathy. Eur. J. Biochem. 235,317-323[Medline]
  26. Witt, S. H., Granzier, H., Witt, C. C., Labeit, S. (2005) MURF-1 and MURF-2 target a specific subset of myofibrillar proteins redundantly: towards understanding MURF-dependent muscle ubiquitination. J. Mol. Biol. 350,713-722[CrossRef][Medline]
  27. Cuff, J. A., Clamp, M. E., Siddiqui, A. S., Finlay, M., Barton, G. J. (1998) Jpred: a consensus secondary structure prediction server. Bioinformatics 14,892-893[Abstract/Free Full Text]
  28. Lupas, A., Van Dyke, M., Stock, J. (1991) Predicting coiled coils from protein sequences. Science 252,1162-1164[CrossRef][Medline]
  29. Schneidman-Duhovny, D., Inbar, Y., Nussinov, R., Wolfson, H. J. (2005) PatchDock and SymmDock: servers for rigid and symmetric docking. Nucleic Acids Res. 33,W363-W367[Abstract/Free Full Text]
  30. Strelkov, S. V., Schumacher, J., Burkhard, P., Aebi, U., Herrmann, H. (2004) Crystal structure of the human lamin A coil 2B dimer: implications for the head-to-tail association of nuclear lamins. J. Mol. Biol. 343,1067-1080[CrossRef][Medline]
  31. Leahy, D. J., Hendrickson, W. A., Aukhil, I., Erickson, H. P. (1992) Structure of a fibronectin type III domain from tenascin phased by MAD analysis of the selenomethionyl protein. Science 258,987-991[Abstract/Free Full Text]
  32. Laskowski, R. A., MacArthur, M. W., Moss, D. S., Thornton, J. M. (1993) PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Cryst. 26,283-291[CrossRef]
  33. Lee, B., Richards, F. M. (1971) The interpretation of protein structures: Estimation of static accessibility. J. Mol. Biol. 55,379-400[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
fj.06-7644comv1
21/7/1383    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mrosek, M.
Right arrow Articles by Mayans, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mrosek, M.
Right arrow Articles by Mayans, O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS