We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Madame Curie Bioscience Database [Internet]. Austin (TX): Landes Bioscience; 2000-.

Bookshelf ID: NBK6372

LIM Domain and Its Binding to Target Proteins

Algirdas Velyvis and Jun Qin.

LIM domain is a unique double-zinc finger motif found in a variety of proteins such as homeodomain transcription factors, kinases, and adaptors. The LIM-containing proteins are involved in diverse biological processes including cytoskeleton organization, cell lineage specification and organ development. Dysfunctions of LIM domains induce pathological effects including muscle detachment, embryonic lethality, and oncogenesis. Acting as a protein-protein interaction motif, the LIM domain has a conserved scaffold but highly variable mode in recognizing diverse target proteins. This chapter describes the structure and function of LIM domain proteins and discusses the molecular basis by which the domain mediates protein-protein interactions.

Introduction

LIM domains were discovered as a novel double zinc finger sequence motif, C-X₂-C-X_17-19-H-X₂-C-X2-C-X2-C-X_15-19-C, with conserved distribution of cysteine and histidine residues in lin-11, Isl-1 and mec-3 (hence - LIM) gene products.¹^,² LIM domains are found in proteins from varied branches of eukaryotes: plants, animals, fungi and mycetozoa (D. discoideum) and have been classified as A, B, C or D types based on their sequence similarity.³^,⁴ Several domains fall outside of these classes due to pronounced sequence divergence.

LIM domains have been also classified into three groups, 1, 2 and 3. LIM domains from classes A and B are most frequently found fused to other functional domains such as kinase domain. But nuclear LIM-only (LMO) proteins, which contain two LIM domains only, are also important members of this group, called group 1.⁵ Class C domains are called group 2. This group_s proteins often contain two copies of LIM domain per protein molecule where the domains are more similar to each other than those from classes A and B. There are no other domains in association with the C-class LIM domains. It was suggested that class C-containing proteins arose via internal duplication of LIM domains.⁴ Class D serves as a sorting bin for LIM domains lacking homology to other classes and with little similarity among themselves. Given that sequence similarity of LIM domains in classes A - C predicts rather well the global architecture of the entire protein, perhaps it is not surprising that the motley collection of LIM domains in class D appears in dissimilar proteins. These are called group 3 and harbor from one to five LIM domains, either with or without additional functional domains or motifs.

Diverse Functions of LIM-Containing Proteins

Proteins containing LIM domains are involved in a variety of biological processes including regulation of gene transcription, cytoskeleton organization, cell lineage specification and organ development. LIM-containing proteins can have significantly different functions even in the same groups. Three commonly known subclasses of group 1 LIM-containing proteins are: (i) LIM-homeodomain (LIM-HD) transcription factors, which participate in activation of transcription and development of the nervous system.⁶ (ii) LMO proteins that are thought to act as molecular adapters involved in development and oncogenesis.⁷ The LMO-1 and LMO-2 were first discovered at translocation breakpoints of T-cell leukemia patients.⁸^,⁹ The LMO family has been now extended to LMO-3 and LMO-4.¹⁰ (iii) LIM kinases, which are involved in cytoskeleton establishment and regulation. LIM kinases (Limk1 and Limk2) are known to phosphorylate and thereby inhibit cofilin.¹¹^,¹² Cofilin functions by depolymerizing F-actin, hence activation of LIM-kinases by the upstream small GTPase dependent PAK1 or ROCK kinases leads to accumulation of actin filaments.¹³

A functionally well-characterized subclass of group 2 LIM proteins is cysteine-rich proteins (CRP1 - CRP3), which are prominent in myogenesis and muscle structure. Expression of all three is upregulated at distinct stages of embryo development and myogenesis.¹³ Antisense disruption of CRP3 blocks terminal differentiation in myoblasts.¹⁴ Interestingly, in CRP3 knock-out the development appears to be normal, but the arrangement of striated muscle myofibrils is disrupted, arguing for a structural or structureregulating role of CRP3. A structural function for all three CRPs in fibroblast cells is implied by robust intracellular staining at actin filaments¹⁵ and in focal adhesions (at least for CRP1).¹⁶ Given that these proteins bind to α-actinin (a well known actin bundle cross-linking protein) and to zyxin (a focal adhesion marker protein), such intracellular localization is expected.¹³ CRP3 is also reported to bind to βI-spectrin in muscle cells and colocalizes with spectrin at the plasma membrane where it overlies Z- and M- lines of striated muscle.¹⁷ The composition and function of these regions of muscle cells is rather similar to focal adhesions, therefore CRP proteins have similar structural roles in different cell types.

The class D LIM domains that belong to group 3 are exemplified by zyxin (3 LIM domains), paxillin (4 LIM domains), and PINCH (5 LIM domains) - three adaptor proteins that are classical markers of focal adhesion (FA) plaques.¹⁸^,¹⁹ All are thought to act as adaptor proteins that orchestrate assembly of FAs. These proteins regulate cell shape change and spreading via distinct LIM-mediated protein-protein interactions.

Zinc Coordination in LIM Domains

The presence of a conserved pattern of cysteine and histidine residues in LIM domains suggested that LIM domains likely contain structural Zn²⁺ ions and this is indeed so. From the current data, the consensus pattern appears to be C - X₂ - C - X_16-23 - H/C - X₂ - C/H - X₂ - C - X₂ - C - X_15-30 - C - X_1-3 - C/H/D (/indicates alternative amino acid residue). Metal coordination by LIM domains was extensively studied on two members of the family, CRP1 and CRIP proteins. These proteins contain two and one copies of LIM domains respectively, with metal coordination consensus residues CCHCCCCC, which all belong to class C. It was established that these LIM domains, expressed in heterologous bacterial hosts or isolated from natural sources, contain approximately 2 equivalents of zinc per equivalent of protein.²⁰ Bound zinc was also demonstrated for native preparations of zyxin.¹⁶ Metal ions can be stripped from LIM domains by denaturation with low pH or guanidinium hydrochloride. A number of other metals- Cd²⁺, Co²⁺, Cu(I),- can be reconstituted into LIM domains upon refolding, again at two ions per domain as established by atomic absorption spectroscopy¹⁶ and by titration.²¹ UV spectroscopy shows bands consistent with formation of S-metal bonds in the complex.²¹ A strong CoS₄ band dominates the Co²⁺ spectrum. Interestingly, the two metal sites appear to have different affinities for metal ions. This observation was craftily exploited by preparing a LIM domain with the high affinity site occupied by Zn²⁺ and the second site by Co²⁺.²¹ This preparation yielded a spectrum consistent with CoS₃N type complex. 1D ¹¹³Cd NMR spectra of Cd²⁺-substituted LIM domains provided further information on the coordination. Chemical shifts of two signals per LIM domain observed in Cd²⁺ NMR are consistent with Cd-S₄ coordination.²¹ The chemical shift of one of the signals, however, is in the region consistent with both S₄ and S₃N coordination. The possibility for S₃N coordination was strongly supported by observation of J-coupling between Cd²⁺ and hydrogen atoms in the approximately 7 parts per million region, which can be explained only by the presence of Cd2+ coordinated to a histidine imidazole ring.²¹ However, none of the above studies determined which of the 7 conserved cysteines form the S₄ and the S₃N site respectively. Based on a generally valid assumption that individual C-X₂-C/H units coordinate the same metal ion, Beckerle and coworkers used mutagenesis to address this question.²² Results of Co²⁺ UV-visible and ¹¹³Cd NMR spectroscopy on the mutant proteins indicated that the N-terminal 4 residues in the protein sequence (CCHC) form the S₃N site, while the S₄ site is composed of the 4 C-terminal cysteines. On a side note, among the mutants tested was a cysteine to aspartic acid substituted domain, in which the C-terminal CCCC zinc finger was transformed into CCCD. Aspartic acid was predicted to be a zinc binding residue in many LIM domains among LIM-HD proteins, but generally it is not common as a Zn²⁺-coordinating group. The cysteine to aspartic acid mutant folded and ¹¹³Cd NMR and Co UV-visible spectra are consistent with conversion of the S₄ to the (apparently perfectly functional) S₃O Zn²⁺-coordination sphere. Finally, the arrangements of metal coordinating residues in the LIM domains were derived from NMR structure determination.²³^,²⁴

LIM Domain As Protein Interaction Motif

Currently, the only function ascribed to LIM domains is protein-protein recognition. Two recent reviews¹⁰^,¹³ present an impressive list of about 30 known protein-protein interaction pairs for LIM domains. Given that the LIM domain count is in the hundreds and the consensus opinion in the field that LIM domains function by recognizing target proteins, it is clear that this is but the tip of the iceberg. The following sections present some well-characterized examples of LIM domain-mediated protein-protein interactions.

LIM-HD and LMO Proteins

Well-known partners for LIM domains, discovered simultaneously for LIM-HD²⁵ and LMO²⁶ proteins, are proteins from the Ldb (LIM domain binding) family. For LIM-HD proteins, in nearly all cases protein-protein interactions depend on LIM domains.⁶ It is the LIM-Ldb interaction mediated formation of different heterodimers of LIM-HD proteins that is thought to determine distinct identities for motor neurons during development.²⁷ Ldb1 specifically binds to nuclear group 1 LIM proteins: LMO1, LMO2, Lmx1, Lhx1, Isl1, Mec3, but not LIM kinase, paxillin, zyxin, enigma or CRP1.²⁶ Deletion mutagenesis was used to determine the LIM-interacting part of Ldb1. A 39-residue stretch within Ldb1 (residues 300-338 of mouse protein) was identified that is sufficient to bind LIM domains.²⁸^,²⁹ This region is termed “LID” _ LIM interaction domain. The LID region contains both highly hydrophobic and hydrophilic stretches of residues and encompasses a 34 residue sequence where human and C. elegans proteins share 23 identical residues (additionally, 7 residues are conservatively substituted), which implies strong evolutionary pressure to preserve the LIM - LID interaction. It is not clear so far what features of LIM domains are responsible for recognition of LID: in some instances it is LIM1 (first LIM), while in others the LIM2 (second LIM) domain of the LIM-HD protein is implicated. Also, some conflicting data exist for X. laevis Xlim-1 where both LIM1 and LIM2 appear to bind to Ldb1.²⁹ LID of Ldb1 appears to be functionally independent from the Ldb1 dimerization domain,²⁸ and monomeric behaviour of LID LIM (of LMO2 or LMO4 proteins) fusion proteins suggests 1:1 binding stoichiometry.³⁰ The importance of LIM domain-mediated interactions for LIM-HD proteins is underscored by the finding that Y116C mutation in a human LIM-HD protein (Lhx3) causes a genetic disease, combined pituitary hormone deficiency.³¹ Residues equivalent to Tyr116 play a central role in hydrophobic cores of known LIM domain structures, hence this mutation is predicted to cause unfolding and loss of function of the LIM domain. Indeed, mutant Lhx3 behaves consistently with such prediction.³² While binding to Ldb by LIM-HD and LMO proteins is important for the activation of transcription, a recently discovered RLIM protein seems to have a negative regulatory role³³ by binding to Lhx3 via its basic region of about 100 residues.³⁴

Zyxin

A number of binding partners are known for zyxin: CRP proteins, ³⁵^,¹⁶ α-actinin,³⁶ proto-oncogene Vav,³⁷ p130^Cas,³⁸ and members of the Ena/VASP family of proteins.³⁹ LIM1 of zyxin is necessary and sufficient for binding to CRP, as demonstrated by blot overlay and GST pull-down assays.⁴⁰ Interestingly, although two zinc fingers in LIM domains form a single domain, individual fingers appear to have some functional independence.²²^,⁴¹ This is also implied by the study that the N-terminal and not the C-terminal zinc finger of zyxin LIM1 is responsible for CRP1 recognition.⁴² The C-terminal zinc finger in LIM1 of zyxin could be swapped with the C-terminal zinc finger from another LIM domain (but not deleted!) with no apparent effect on binding to CRP1. In CRP1, LIM1 and LIM2 appear to cooperate in binding to zyxin, so that deletion of either reduces the extent of binding more than twenty-fold in blot overlay assay.⁴³ The entire LIM₁ is required, but the C-terminal zinc finger in LIM2 is disposable. The linker between the two LIM domains is inert. Taken together these studies present an interesting picture_LIM domains in CRP1 are linked by a long disordered segment and structurally seem to be entirely independent.⁴⁴ Zyxin was also suggested to bind to actin filaments via α-actinin and play a regulatory rather than a structural role in FAs. Ample biochemical data support this notion. Interaction with Ena/VASP proteins is important in targeting zyxin to lamellipodia,⁴⁵ although under normal cellular situations the vast majority of zyxin resides in focal adhesions.⁴⁶ Ena/VASP proteins are known to bind profilin—an actin polymerizing protein. In an important development, it was shown that zyxin-VASP complex is able to initiate actin polymerization.⁴⁷

Paxillin

Paxillin is one of the central proteins in FAs acting as a scaffold for a myriad of binding partners.⁴⁸ The LIM domain-containing C-terminal half of paxillin is known to bind to protein tyrosine phosphatase – PEST⁴⁹^,⁵⁰ and this binding is postulated to bring PTP-PEST into the vicinity of its target, phosphorylated p130^Cas.⁵¹ PTP-PEST is required for disassembly of FAs possibly by dephosphorylating p130^Cas and paxillin⁵² and therefore modulating their FA targeting regions. The LIM domain region is responsible for targeting paxillin to Fas,⁵³ via a yet unknown binding event. It was recently shown that Paxillin LIM domains bind to α- and γ-tubulin as well,⁵⁴ which led to a proposal that paxillin is responsible for the interplay between actin filaments and microtubules. A fragment containing both the C-terminal LIM3 and LIM4 of paxillin is competent to bind to PTP-PEST, while LIM3 or LIM4 in isolation are not.⁵¹ Deletions of LIM3 or LIM4 from an otherwise intact paxillin or its fragments abolish binding, as does disruption of the N- or C-terminal zinc fingers within LIM3 and the N-terminal finger in LIM4 with point mutagenesis. Comprehensive deletion of PTP-PEST showed that a part of the protein that centers on the 2^nd (of five total) proline-rich region of PTP-PEST is the binding epitope. The structure of this PTP-PEST peptide is not known, but the 50% content of proline likely precludes formation of a globular fold. It is interesting that a close homologue of paxillin, Hic-5 protein, also binds to PTP-PEST.⁵⁵ Broadly the same region of the phosphatase, rich in prolines, is involved. Here, in contrast to paxillin, LIM2, LIM3 and LIM4 are involved in binding, with LIM3 being absolutely necessary for detectable binding, while the presence of LIM2 and LIM4 clearly augments the extent of binding.⁵⁵ Given that the LIM domains of paxillin and His-5 share 68% sequence similarity,⁵¹ it is not obvious whether the protein interface is the same for both members of the family, or only the net result, formation of the complex with PTP-PEST, is conserved.

Enigma Protein

A relatively well-characterized system is Enigma protein that contains three LIM domains at its C terminus. Yeast two-hybrid screening revealed that Enigma binds to the Insulin receptor (InsR) internalization motif.⁵⁶ Point mutants in two internalization-driving tetrapeptides within this segment somewhat diminished (weaker peptide) and thoroughly blocked (stronger peptide) interaction with full-length Enigma and LIM3 construct. The interaction is specific since the internalization segments of IGF1, EGF, sequence-optimized EGF, Transferrin and LDL receptors failed to bind. However, the paired sequence-optimized EGF receptor motif (the tandem repeat of the internalization motif of EGF that has been optimized to display high activity in the internalization assay) did bind, albeit to a lesser extent. The relevance of the last observation is not clear, since LIM2 in Mec-3, a nuclear LIM-HD protein which has no known function in receptor endocytosis also binds to the paired sequence-optimized EGFR motif. Furthermore, this paired sequence-optimized motif binds equally well to LIM2, LIM3 and the C-terminally truncated LIM3 of Enigma.⁵⁶ Yeast two-hybrid screening was also instrumental in discovering that the LIM domain fragment of Enigma also binds to Ret Receptor tyrosine kinase.⁵⁷ Mutating the most C-terminal tyrosine at position 1062 to phenylalanine or deleting twenty-three C-terminal residues (including Tyr1062) from Ret abolished the interaction. In fact, a 61 C-terminal residue segment of Ret is perfectly capable of binding to enigma⁵⁸ in a GST pull-down assay. Their association is not phosphorylation-dependent, and is specific for Enigma, as Zyxin and CRP1 did not bind. Enigma-Ret binding was confirmed by the observation that coinjection of DNA for the LIM domain-containing part of Enigma with the Ret construct into fibroblasts abolished the mitogenic effect of Ret in a dominant-negative style.⁵⁷ Unlike InsR, Ret peptide is specifically recognized by Enigma LIM2,⁵⁸ while the LIM domains from LIM-HD proteins, zyxin, CRP1 or paxillin do not bind Ret or InsR peptides. Which residues are important for Ret - LIM2 recognition was further assessed with a competition assay.⁵⁸ Wild type (NKLY1062) and mutant (AKLA and NKLF) peptides within an otherwise intact 20-residue segment of Ret Tyr1062 region were used as competitors for Ret in GST-pull down. While AKLA peptide and InsR LIM3-recognizing peptide failed to inhibit the interaction, NKLY and NKLF peptides were inhibitors. The capacity of NKLF peptide to act as an inhibitor, however, contradicts previous data⁵⁷ that the Y1062F mutation disrupts binding. Also, the GST pull-down titration experiment showed that lower amounts of Y1062F mutant Ret are retained by GST-LIM₂. The apparent discrepancy may arise from the fact that the amount of observed binding in GST-pull down assay depends on times and volumes of the binding and washing steps, in addition to the affinity of interaction. Also, 20-residue peptides may be suboptimal for binding, since high concentrations were required to document competition.⁵⁸ A random peptide library was screened to identify sequence requirements for peptides recognized by Enigma LIM3.⁵⁷ A screen with fixed tyrosine which is assigned at position 0, and is prominent in both internalization-targeting peptides in InsR and recognized by LIM3, identified strong preference for proline at positions _1 and +2. A second screen with a fixed PXXP feature yielded a consensus GP-Hyd1-GP-Hyd2-Y/F-A, where Hyd1 is Met, Phe, Tyr or Ile and Hyd2 is Ile, Met or Val. This compares nicely with the sequence naturally present in InsR: GPLGPLYA. Moreover, GPLY is present, which is the strong InsR internalization-driving motif whose mutation blocks binding to LIM3.⁵⁶

PINCH

PINCH (Particularly Interesting New Cys-His) protein was discovered in an immunological screen for senescent erythrocyte antigens.⁵⁹ Nearly the entire protein sequence of PINCH is contained in the LIM domains, with very short interdomain linker peptides, and a C-terminal extension with numerous positive charges. A yeast two-hybrid screen of the human lung library revealed that PINCH interacts with the bait that consisted of the N-terminal ankyrin repeat domain of integrin-linked kinase (ILK).⁶⁰ Paxillin or zyxin LIM domain fragments failed to bind to ILK. MBP-PINCH is able to pull down ILK from cell lysates and ILK is retained on an anti-PINCH immunoglobulin column. Binding was confirmed in vitro with an ELISA assay between recombinant MBP-PINCH and GST-ILK. This establishes that even though binding in the cells might be mediated by bridging proteins, it is likely a direct association. Deletion mutagenesis with a subsequent yeast two-hybrid assay established that LIM1 of PINCH is necessary and sufficient for binding. LIM1 was also competent to bind to ILK in pull-down and ELISA⁶¹^,⁶⁰ experiments. The ankyrin repeat domain of ILK is solely responsible for binding to PINCH.⁶⁰^,⁶¹ The LIM4 domain of PINCH was found to interact with Nck2 protein that is also an adaptor.⁶² Deletion mutagenesis revealed that the third SH3 domain of Nck2 is responsible for binding to LIM4,⁶⁰ however, the affinity of the interaction was found to be very weak.²⁴

In summary, unlike short peptide-recognizing SH3 or SH2 domains, LIM domains bind to a wide variety of partners, which make it difficult to pinpoint their binding specificities. Hence, despite the success in the case of Enigma LIM3, combinatorial peptide synthesis and modification techniques are of limited value for research in the field of LIM domains. Structure determination, which would define conformations of specificity-defining groups at domain surfaces and would directly reveal the nature of the interfaces between the molecules in the complex, seems to be the most robust way to uncover the biochemical basis for recognition of targets by LIM domains.

Structures of LIM Domains

LIM2 of cCRP1 is the first LIM domain for which the 3D structure has been determined.⁶³ The structure revealed two contiguous zinc fingers. In each zinc finger, two of the Zn²⁺-coordinating cysteine residues are located at the turns connecting individual β strands in β hairpins. The NOE patterns in these turns closely resemble patterns observed in the cysteine-containing turns (also metal-coordinating) in P. furiosus rubredoxin (Rd) and hence these turns have been called “rubredoxin knuckles” (fig. 1). Moreover, these turns contain a unique feature—H^N-S hydrogen bonds from backbone amide protons to metal chelating sulfur atoms, as demonstrated for Rd itself. Heteronuclear Spin-Echo difference (HSED) spectrum on the Cd²⁺-substituted LIM2 sample showed clearly the existence of partially covalent H^N-S H-bonds. 1H-¹¹³Cd HMQC experiment on Cd²⁺-substituted LIM₂ unequivocally established metal coordination in LIM₂. The two zinc fingers in cCRP1 LIM2 stack together (fig. 2) to form a single domain with several bulky hydrophobic residues forming a small core. The core is dominated by stacking of the aromatic rings of Trp138 and Phe143 (fig. 3) that reside in the very C terminus of the N-terminal zinc finger and the inter-finger dipeptide, respectively. The importance of these residues is reflected in their conservation: among currently known LIM domain sequences, the position equivalent to equivalent to Phe143 is occupied by phenylalanine (80.4%), leucine (10.8%), or tyrosine, valine, tryptophan or isoleucine (6.9%) (total 98.1%). It is quite possible that the preference for aromatic residues at these two positions is due to strengthening of the hydrophobic cluster by additional aromatic π - π interaction. Other hydrophobic residues (also highly conserved) pile on the sides of these two: Val131 from the N-terminal zinc finger and Leu152, Leu157, Ieu164 from the C-terminal finger. Interestingly, Perez-Alvarado et al⁶³ also discovered a rather good superposition between the C-terminal zinc finger of LIM2 and the DNA-binding zinc fingers in the glucocorticoid receptor and in the GATA-1 transcription factor (backbone rmsds 2.0 and 1.7 Å, respectively), which suggested a possible a role for LIM domains as DNA-binding molecules. However, so far this hypothesis is not supported by experimental evidence.

Figure 1

Rubredoxin (Rd) turns. (a) N-terminal turn in P. furiosus Rd (PDB ID: 1brf). (b) N-terminal turn in qCRP2 LIM2 (PDB ID 1CTL). Spheres – Fe³⁺ (a) and Zn²⁺ (b). H^N-S H-bonds are denoted by lines.

Figure 2

Ribbon representation of cCRP1 LIM2, with secondary structure elements labeled and Zn²⁺ ions in spheres.

Figure 3

Hydrophobic core in cCRP2 LIM2. Backbone is shown as a tube. The hydrophobic residues in the core are labeled.

Since the report of cCRP LIM2 structure, numerous structures of other LIM domains have been determined by NMR spectroscopy. These include all three groups of LIM domains: (i) LIM1 domains of LMO2 and LMO4,⁶⁴ which belong to group 1. (ii) LIM domains in group 2 family including LIM2 of cCRP1 as mentioned above, CRIP protein (one LIM domain),⁶⁵ full-length cCRP1 (LIM1 and LIM2),43 LIM1 and LIM2 of qCRP2.⁶⁶^,⁶⁷ In addition, the structure of an isolated N-terminal zinc finger of LIM domain in Lasp-1 has been determined.⁴¹ (iii) LIM1 and LIM4 domains of PINCH, which belong to group 3.²³^,²⁴ The main structural features of these LIM domains were found to be the same as cCRP LIM2: four short β hairpins with a C-terminal α helix; two “rubredoxin knuckles”; small hydrophobic core contributed by both zinc fingers. Although the global fold is similar among all known LIM domain structures, there are some differences, notably the relative orientation of the two zinc fingers varies about the long axis of the domain (fig. 4). Different orientations between the fingers in LIM domains may be important for protein-protein recognition. Each zinc finger appears to be an independent folding unit despite the interactions between the two fingers. This is evidenced by the determination of the structure of isolated N-terminal CCHC half of the LIM domain of Lasp-1 protein.⁴¹ Clearly, four zinc binding residues are the key structural determinants, which orchestrate the fold of this fragment. Other residues in Lasp-1 also adopt conformations close to those observed in the N-terminal halves of full-length LIM domains.

Figure 4

Overlay of CRIP LIM domain (red, PDB ID 1IML) with cCRP1 LIM2 (blue), illustrating the relative twist of the C-terminal Znfinger in CRIP. Backbone residues 1-23 of CRIP were best fit superimposed onto corresponding residues of LIM2. (a) Overall view; (more...)

A key issue about the LIM domain is its structural basis in recognizing target proteins. Recent chemical shift mapping studies on PINCH LIM domains have provided important insights into this issue:²³^,²⁴ For PINCH LIM1, it was shown that C-terminal hydrophobic patch and several adjoining residues are most perturbed upon binding to ankyrin repepat (ANK) domain of ILK (fig. 5a) whereas residues from N-terminal basic region of PINCH LIM4 appear to be responsible for binding to SH3 domain (fig. 5). The structure of LMO-Ldb1 complex revealed a rather extensive binding interface involving both zinc finger 1 and zinc finger 2.⁶⁴ Hence, LIM domains do not appear to have a single, well defined binding epitope, but rather present on their surfaces clusters of different functional groups, individually tailored for diverse binding partners. Such diverse recognition mode has also been observed in PTB and SH2 domains.⁶⁸^,⁶⁹ Clearly, more structures of LIM domains in complex with target proteins are necessary for a thorough understanding of this fascinating protein interaction motif.

Figure 5

Comparison of PINCH LIM1 and LIM4 binding modes to target proteins: (a) the major binding surface (shaded area with labeled residues) of LIM1 (PDB ID 1G47) to ILK ANK domain; (b) the major binding surface (shaded area with labeled residues)

Acknowledgments

The work was supported by NIH grants to J.Q. (HL58758, GM62823).

References

1.: Karlsson O, Thor S, Norberg T. et al. Insulin gene enhancer binding protein Isl-1 is a member of a novel class of proteins containing both a homeo- and a Cys-His domain. Nature. 1990;344:879–882. [PubMed: 1691825]
2.: Freyd G, Kim SK, Horvitz HR. Novel cysteine-rich motif and homeodomain in the product of the Caenorhabditis elegans cell lineage gene lin-11. Nature. 1990;344:876–879. [PubMed: 1970421]
3.: Taira M, Evrard JL, Steinmetz A. et al. Classification of LIM proteins. Trends Genet. 1995;11:431–432. [PubMed: 8578598]
4.: Dawid IB, Toyama R, Taira M. LIM domain proteins. CR Acad Sci III. 1995;318:295–306. [PubMed: 7788499]
5.: Dawid IB, Breen JJ, Toyama R. LIM domains: Multiple roles as adapters and functional modifiers in protein interactions. Trends Genet. 1998;14:156–162. [PubMed: 9594664]
6.: Hobert O, Westphal H. Functions of LIM-homeobox genes. Trends Genet. 2000;16:75–83. [PubMed: 10652534]
7.: Rabbitts T.H. LMO T-cell translocation oncogenes typify genes activated by chromosomal translocations that alter transcription and developmental processes. Genes Dev. 1998;12(17):2651–7. [PubMed: 9732263]
8.: Larson RC, Osada H, Larson TA. et al. The oncogenic LIM protein Rbtn2 causes thymic developmental aberrations that precede malignancy in transgenic mice. Oncogene. 1995;11(5):853–62. [PubMed: 7545805]
9.: Larson RC, Lavenir I, Larson TA. et al. Protein dimerization between Lmo2 (Rbtn2) and Tal1 alters thymocyte development and potentiates T cell tumorigenesis in transgenic mice. EMBO J. 1996;15(5):1021–7. [PMC free article: PMC449997] [PubMed: 8605871]
10.: Bach I. The LIM domain: Regulation by association. Mech. Dev. 2000;91:5–17. [PubMed: 10704826]
11.: Arber S, Barbayannis FA, Hanser H. et al. Regulation of actin dynamics through phosphorylation of cofilin by LIM-kinase. Nature. 1998;39:805–809. [PubMed: 9655397]
12.: Yang N, Higuchi O, Ohashi K. et al. Cofilin phosphorylation by LIM-kinase 1 and its role in Rac-mediated actin reorganization. Nature. 1998;393:809–812. [PubMed: 9655398]
13.: Khurana T, Khurana B, Noegel AA. LIM proteins: Association with the actin cytoskeleton. Protoplasma. 2002;219:1–12. [PubMed: 11926060]
14.: Arber S, Halder G, Caroni P. Muscle LIM protein, a novel essential regulator of myogenesis, promotes myogenic differentiation. Cell. 1994;79:221–231. [PubMed: 7954791]
15.: Louis HA, Pino JD, Schmeichel KL. et al. Comparison of three members of the cysteine-rich protein family reveals functional conservation and divergent patterns of gene expression. J Biol Chem. 1997;272:27484–27491. [PubMed: 9341203]
16.: Sadler I, Crawford AW, Michelsen JW. et al. Zyxin and cCRP: Two interactive LIM domain proteins associated with the cytoskeleton. J Cell Biol. 1992;119:1573–1587. [PMC free article: PMC2289750] [PubMed: 1469049]
17.: Flick MJ, Konieczny SF. The muscle regulatory and structural protein MLP is a cytoskeletal binding partner of betaI-spectrin. J Cell Sci. 2000;113(Pt 9):1553–1564. [PubMed: 10751147]
18.: Wu C, Dedhar S. Integrin-linked kinase (ILK) and its interactors: A new paradigm for the coupling of extracellular matrix to actin cytoskeleton and signaling complexes. J Cell Biol. 2001;155:505–510. [PMC free article: PMC2198863] [PubMed: 11696562]
19.: Wang Y, Gilmore TD. Zyxin and paxillin proteins: Focal adhesion plaque LIM domain proteins go nuclear. Biochim Biophys Acta. 2003;1593(2-3):115–20. [PubMed: 12581855]
20.: Michelsen JW, Schmeichel KL, Beckerle MC. et al. The LIM motif defines a specific zinc-binding protein domain. Proc Natl Acad Sci USA. 1993;90:4404–4408. [PMC free article: PMC46519] [PubMed: 8506279]
21.: Kosa JL, Michelsen JW, Louis HA. et al. Common metal ion coordination in LIM domain proteins. Biochemistry. 1994;33:468–477. [PubMed: 8286377]
22.: Michelsen JW, Sewell AK, Louis HA. et al. Mutational analysis of the metal sites in an LIM domain. J Biol Chem. 1994;269:11108–11113. [PubMed: 8157637]
23.: Velyvis A, Yang Y, Wu C. et al. Solution structure of the focal adhesion adaptor PINCH LIM1 domain and characterization of its interaction with the integrin-linked kinase ankyrin repeat domain. J Biol Chem. 2001;276:4932–4939. [PubMed: 11078733]
24.: Velyvis A, Vaynberg J, Yang Y. et al. Structural and functional insights into PINCH LIM4 domain-mediated integrin signaling. Nat Struct Biol. 2003;10:558–564. [PubMed: 12794636]
25.: Agulnick AD, Taira M, Breen JJ. et al. Interactions of the LIM-domain-binding factor Ldb1 with LIM homeodomain proteins. Nature. 1996;384:270–272. [PubMed: 8918878]
26.: Jurata LW, Kenny DA, Gill GN. Nuclear LIM interactor, a rhombotin and LIM homeodomain interacting protein, is expressed early in neuronal development. Proc Natl Acad Sci USA. 1996;93:11693–11698. [PMC free article: PMC38120] [PubMed: 8876198]
27.: Jurata LW, Pfaff SL, Gill GN. The nuclear LIM domain interactor NLI mediates homo- and heterodimerization of LIM domain transcription factors. J Biol Chem. 1998;273:3152–3157. [PubMed: 9452425]
28.: Jurata LW, Gill GN. Functional analysis of the nuclear LIM domain interactor NLI. Mol Cell Biol. 1997;17:5688–5698. [PMC free article: PMC232417] [PubMed: 9315627]
29.: Breen JJ, Agulnick AD, Westphal H. et al. Interactions between LIM domains and the LIM domain-binding protein Ldb1. J Biol Chem. 1998;273:4712–4717. [PubMed: 9468533]
30.: Deane JE, Sum E, Mackay JP. et al. Design, production and characterization of FLIN2 and FLIN4: The engineering of intramolecular ldb1: LMO complexes. Protein Eng. 2001;14:493–499. [PubMed: 11522923]
31.: Netchine I, Sobrier ML, Krude H. et al. Mutations in LHX3 result in a new syndrome revealed by combined pituitary hormone deficiency. Nat Genet. 2000;25:182–186. [PubMed: 10835633]
32.: Howard PW, Maurer RA. A Point Mutation in the LIM Domain of Lhx3 Reduces Activation of the Glycoprotein Hormone alpha -Subunit Promoter. J Biol Chem. 2001;276:19020–19026. [PubMed: 11279219]
33.: Bach I, Rodriguez-Esteban C, Carriere C. et al. RLIM inhibits functional activity of LIM homeodomain transcription factors via recruitment of the histone deacetylase complex. Nat Genet. 1999;22:394–399. [PubMed: 10431247]
34.: Ostendorff HP, Peirano RI, Peters MA. et al. Ubiquitination-dependent cofactor exchange on LIM homeodomain transcription factors. Nature. 2002;416:99–103. [PubMed: 11882901]
35.: Crawford AW, Pino JD, Beckerle MC. Biochemical and molecular characterization of the chicken cysteine-rich protein, a developmentally regulated LIM-domain protein that is associated with the actin cytoskeleton. J Cell Biol. 1994;124:117–127. [PMC free article: PMC2119904] [PubMed: 8294495]
36.: Crawford AW, Michelsen JW, Beckerle MC. An interaction between zyxin and alpha-actinin. J Cell Biol. 1992;116:1381–1393. [PMC free article: PMC2289370] [PubMed: 1541635]
37.: Hobert O, Schilling JW, Beckerle MC. et al. SH3 domain-dependent interaction of the proto-oncogene product Vav with the focal contact protein zyxin. Oncogene. 1996;12:1577–1581. [PubMed: 8622875]
38.: Yi J, Kloeker S, Jensen CC. et al. Members of the Zyxin family of LIM proteins interact with members of the p130Cas family of signal transducers. J Biol Chem. 2002;277:9580–9589. [PubMed: 11782456]
39.: Drees B, Friederich E, Fradelizi J. et al. Characterization of the interaction between zyxin and members of the Ena/ vasodilator-stimulated phosphoprotein family of proteins. J Biol Chem. 2000;275:22503–22511. [PubMed: 10801818]
40.: Schmeichel KL, Beckerle MC. The LIM domain is a modular protein-binding interface. Cell. 1994;79:211–219. [PubMed: 7954790]
41.: Hammarstrom A, Berndt KD, Sillard R. et al. Solution structure of a naturally-occurring zinc-peptide complex demonstrates that the N-terminal zinc-binding module of the Lasp-1 LIM domain is an independent folding unit. Biochem. 1996;35:12723–12732. [PubMed: 8841116]
42.: Schmeichel KL, Beckerle MC. Molecular dissection of a LIM domain. Mol Biol Cell. 1997;8:219–230. [PMC free article: PMC276075] [PubMed: 9190203]
43.: Schmeichel KL, Beckerle MC. LIM domains of cysteine-rich protein 1 (CRP1) are essential for its zyxin-binding function. Biochem J. 1998;331(Pt 3):885–892. [PMC free article: PMC1219431] [PubMed: 9560318]
44.: Yao X, Perez-Alvarado GC, Louis HA. et al. Solution structure of the chicken cysteine-rich protein, CRP1, a double-LIM protein implicated in muscle differentiation. Biochem. 1999;38:5701–5713. [PubMed: 10231520]
45.: Nix DA, Fradelizi J, Bockholt S. et al. Targeting of zyxin to sites of actin membrane interaction and to the nucleus. J Biol Chem. 2001;276:34759–34767. [PubMed: 11395501]
46.: Beckerle MC. Zyxin: Zinc fingers at sites of cell adhesion. Bioessays. 1997;19:949–957. [PubMed: 9394617]
47.: Fradelizi J, Noireaux V, Plastino J. et al. ActA and human zyxin harbour Arp2/3-independent actin-polymerization activity. Nat Cell Biol. 2001;3:699–707. [PubMed: 11483954]
48.: Turner CE. Paxillin and focal adhesion signaling. Nat Cell Biol. 2000;2:E231–E236. [PubMed: 11146675]
49.: Cote JF, Turner CE, Tremblay ML. Intact LIM 3 and LIM 4 domains of paxillin are required for the association to a novel polyproline region (Pro 2) of protein-tyrosine phosphatase-PEST. J Biol Chem. 1999;274:20550–20560. [PubMed: 10400685]
50.: Shen Y, Schneider G, Cloutier JF. et al. Direct association of protein-tyrosine phosphatase PTP-PEST with paxillin. J Biol Chem. 1998;273:6474–6481. [PubMed: 9497381]
51.: Cote JF, Turner CE, Tremblay ML. Intact LIM 3 and LIM 4 domains of paxillin are required for the association to a novel polyproline region (Pro 2) of protein-tyrosine phosphatase-PEST. J Biol Chem. 1999;274:20550–20560. [PubMed: 10400685]
52.: Shen Y, Lyons P, Cooley M. et al. The noncatalytic domain of protein-tyrosine phosphatase-PEST targets paxillin for dephosphorylation in vivo. J Biol Chem. 2000;275:1405–1413. [PubMed: 10625692]
53.: Brown MC, Perrotta JA, Turner CE. Identification of LIM3 as the principal determinant of paxillin focal adhesion localization and characterization of a novel motif on paxillin directing vinculin and focal adhesion kinase binding. J Cell Biol. 1996;135:1109–1123. [PMC free article: PMC2133378] [PubMed: 8922390]
54.: Herreros L, Rodriguez-Fernandez JL, Brown MC. et al. Paxillin localizes to the lymphocyte microtubule organizing center and associates with the microtubule cytoskeleton. J Biol Chem. 2000;275:26436–26440. [PubMed: 10840040]
55.: Nishiya N, Iwabuchi Y, Shibanuma M. et al. Hic-5, a Paxillin Homologue, Binds to the Protein-tyrosine Phosphatase PEST (PTP-PEST) through Its LIM 3_in. J Biol Chem. 1999;274:9847–9853. [PubMed: 10092676]
56.: Wu RY, Gill GN. LIM domain recognition of a tyrosine-containing tight turn. J Biol Chem. 1994;269:25085–25090. [PubMed: 7929196]
57.: Durick K, Wu RY, Gill GN. et al. Mitogenic Signaling by Ret/ptc2 Requires Association with Enigma via a LIM Domain. J Biol Chem. 1996;271:12691–12694. [PubMed: 8662982]
58.: Wu RY, Durick K, Songyang Z. et al. Specificity of LIM Domain Interactions with Receptor Tyrosine Kinases. J Biol Chem. 1996;271:15934–15941. [PubMed: 8663233]
59.: Rearden A. A new LIM protein containing an autoepitope homologous to “senescent cell antigen” Biochem Biophys Res Commun. 1994;201:1124–1131. [PubMed: 7517666]
60.: Tu Y, Li F, Goicoechea S. et al. The LIM-only protein PINCH directly interacts with integrin-linked kinase and is recruited to integrin-rich sites in spreading cells. Mol Cell Biol. 1999;19:2425–2434. [PMC free article: PMC84035] [PubMed: 10022929]
61.: Li F, Zhang Y, Wu C. Integrin-linked kinase is localized to cell-matrix focal adhesions but not cell-cell adhesion sites and the focal adhesion localization of integrin-linked kinase is regulated by the PINCH-binding ANK repeats. J Cell Sci. 1999;112(Pt 24):4589–4599. [PubMed: 10574708]
62.: Tu Y, Li F, Wu C. Nck-2, a novel Src homology2/3-containing adaptor protein that interacts with the LIM-only protein PINCH and components of growth factor receptor kinase-signaling pathways. Mol Biol Cell. 1998;9:3367–3382. [PMC free article: PMC25640] [PubMed: 9843575]
63.: Perez-Alvarado GC, Miles C, Michelsen JW. et al. Structure of the carboxy-terminal LIM domain from the cysteine rich protein CRP. Nat Struct Biol. 1994;1:388–398. [PubMed: 7664053]
64.: Deane JE, Mackay JP, Kwan AH. et al. Structural basis for the recognition of ldb1 by the N-terminal LIM domains of LMO2 and LMO4. EMBO J. 2003;22(9):2224–33. [PMC free article: PMC156068] [PubMed: 12727888]
65.: Perez-Alvarado GC, Kosa JL, Louis HA. et al. Structure of the cysteine-rich intestinal protein, CRIP. J Mol Biol. 1996;257:153–174. [PubMed: 8632452]
66.: Konrat R, Weiskirchen R, Krautler B. et al. Solution structure of the carboxyl-terminal LIM domain from quail cysteine-rich protein CRP2. J Biol Chem. 1997;272:12001–12007. [PubMed: 9115265]
67.: Kontaxis G, Konrat R, Krautler B. et al. Structure and intramodular dynamics of the amino-terminal LIM domain from quail cysteineand glycine-rich protein CRP2. Biochem. 1998;37:7127–7134. [PubMed: 9585524]
68.: Li SC, Zwahlen C, Vincent SJ. et al. Structure of a Numb PTB domain-peptide complex suggests a basis for diverse binding specificity. Nat Struct Biol. 1998;5:1075–83. [PubMed: 9846878]
69.: Mallis RJ, Brazin KN, Fulton DB. et al. Structural characterization of a proline-driven conformational switch within the Itk SH2 domain. Nat Struct Biol. 2002;9:900–5. [PubMed: 12402030]