Abstract
Background
The regeneration of peripheral nerve is associated with a change in the alternative splicing of the fibronectin primary gene transcript to re-express embryonic isoforms containing a binding site for α4β1 integrins that promote neurite outgrowth. Here we use PC12 cells to examine the role of the interaction between paxillin and the α4 integrin cytoplasmic domain in neurite outgrowth.
Results
Expression of α4 with mutations in the paxillin-binding domain reduced neurite outgrowth on recombinant embryonic fibronectin fragments relative to wild type α4. Over-expression of paxillin promoted neurite outgrowth while a mutant isoform lacking the LD4 domain implicated in the regulation of ARF and Rac GTPases was less effective. Optimal α4-mediated migration in leucocytes requires spatial regulation of α4 phosphorylation at Ser988, a post-translational modification that blocks paxillin binding to the integrin cytoplasmic domain. In keeping with this α4(S988D), which mimics phosphorylated α4, did not promote neurite outgrowth. However, α4 was not phosphorylated in the PC12 cells, and a non-phosphorylatable α4(S988A) mutant promoted neurite outgrowth indistinguishably from the wild type integrin.
Conclusion
We establish the importance of the α4 integrin-paxillin interaction in a model of axonal regeneration and highlight differing dependence on phosphorylation of α4 for extension of neuronal growth cones and migration of non-neural cells.
Background
Regrowth of the peripheral nervous system (PNS) after injuries such as axotomy is associated with changes in the expression of extracellular matrix (ECM) molecules in the environment traversed by the regenerating growth cones [1]. These changes may promote regeneration in at least two different ways. First, by generating increased levels of molecules such as laminin (Ln) and fibronectin (Fn) that can promote neurite outgrowth in cell culture and are therefore likely to promote regeneration in vivo [2,3]. Second, by altering the pattern of alternative splicing so as to switch from isoforms of ECM proteins expressed in adult nerve to those expressed earlier during a time of axon growth in development. An example of the second mechanism is provided by Fn, whose primary gene transcript is alternatively spliced in three regions [4-9]. Two of these include or exclude single Fn type III repeats (EIIIA and EIIIB) while the third partially or completely excludes a region (V or IIICS) that contains a cell binding sequence Leu-Asp-Val (LDV) recognised by the integrin α4β1 [10,11]. While these regions are excluded from adult Fns in a cell and tissue specific pattern [12-14], we have previously found that sciatic nerve injury and regeneration in the adult rat is associated with a change in the pattern of alternative splicing and the re-expression of embryonic Fns that include all these regions [15,16].
Recombinant Fns that include the entire V region promote neurite outgrowth from dorsal root ganglion (DRG) neurons or PC12 cells (a cell line model of PNS neurones) more effectively than those containing only the RGD sequence present in the 10th type III repeat present in all Fns [17]. As the α4 integrin subunit is expressed on regenerating growth cones in vivo, we have therefore proposed that the re-expression of embryonic Fns following injury contributes to regeneration of the adult PNS by enhancing α4β1 integrin signalling [17]. The identification of α4 integrins as part of the regenerative mechanism in the PNS is of particular interest as these integrins play an important role in cell migration in other cell types [18,19]. The adaptor protein, paxillin, which binds directly to the α4 cytoplasmic domain, is an important downstream signalling molecule in this process [20]. Transgenic knock-in mice expressing only α4 integrin with a Y991A mutation within the paxillin binding site that blocks the interaction show impaired leucocyte recruitment during inflammation [21]. A physiological modification that blocks binding of paxillin is generated by phosphorylation of the α4 integrin cytoplasmic domain, detected in certain T cell lines using an antiserum raised against a synthetic peptide containing phosphoserine at Ser988 [22]. Phosphorylation is localised to the leading edge of migrating CHO cells [23], and the reversibility of paxillin binding enabled by phosphorylation and dephosphorylation of α4 may be important for migration, as both S988A (that inhibits phosphorylation) and S988D (that mimics phosphorylation) mutations inhibit migration in cell culture [24]. We have shown previously that paxillin is also associated specifically with α4 in DRG neurons and PC12 cells [17]. This suggests the hypothesis that paxillin also makes an essential contribution to α4-mediated growth cone motility and PNS regeneration by virtue of the association with the integrin cytoplasmic domain. Here we test this hypothesis by examining the requirement for the α4-paxillin interaction in neurite outgrowth and find that the interaction is required; however, in contrast to non-neuronal cell migration, α4 phosphorylation is dispensable for neurite outgrowth.
Results
To examine α4-mediated neurite outgrowth we used PC12 cells as model neurons. These cells do not express α4 integrin and transfection with α4 enables them to manifest neurite outgrowth on Fn isoforms that contain the α4 integrin binding site [17]. To assess the requirement for paxillin binding to the α4 cytoplasmic domain for this neurite outgrowth, we first expressed α4 subunits into which alanine substitutions had been inserted into either or both of two amino acids essential for paxillin binding; Glu983 and Tyr991 [25]. As controls we expressed either full-length α4 or α4 lacking the cytoplasmic domain sequences C-terminal to the GFFKR motif (Fig 1A). To demonstrate the expected effects of these mutations on paxillin binding, we immunoprecipitated lysates of the PC12 cells with antibodies against the extracellular domain of α4 and then western blotted the protein complexes pulled down with anti-paxillin antibodies. As shown in Fig 1B, the α4 subunits were all expressed in the PC12 cells but bound paxillin to differing degrees; while full-length α4 complexed with high levels of paxillin, α4 with either single mutation bound much lower levels and α4 with the double mutation showed virtually no binding. As expected, the α4 subunit lacking the majority of the cytoplasmic domain did not bind paxillin, illustrating the specificity of the assay.
To examine the effects of these mutations on α4-mediated neurite outgrowth, we then grew these PC12 cells on recombinant Fn fragments containing the V120 region for 24 hrs. The extent of neurite outgrowth correlated with the degree of paxillin binding to the α4 cytoplasmic domain (Fig 2A). Either single mutation reduced neurite outgrowth while the double mutation resulted in a complete loss of α4-mediated neurite outgrowth, with the degree of outgrowth now being identical to the low level seen with wild type PC12 cells or with PC12 cells expressing the α4 subunit lacking the majority of the cytoplasmic domain. To ask whether this effect of perturbing paxillin interactions was specific for α4-mediated neurite outgrowth, we repeated the experiments with cells expressing wild type α4 or the double mutation on laminin, collagen or Fn V120. Laminin and collagen bind to different integrins on the cell surface (α1β1, α3β1, α6β1 and α7β1 for laminin, α1β1 and α2β1 for collagen [26]) and promote neurite outgrowth in PC12 cells [27,28]. Expression of α4, either wild-type or mutant, had no effect on this neurite outgrowth while, as before, only the cells expressing wild-type α4 showed enhanced neurite outgrowth on Fn V120 (Fig 2B) confirming that the loss of the paxillin association with α4 only perturbs α4-mediated neurite outgrowth.
To show directly that paxillin promotes neurite outgrowth in response to Fn V120 substrates, we over-expressed wild-type or mutant paxillins in PC12 cells also expressing wild-type α4. The structure of wild-type paxillin is shown in Fig 3A, and the two different paxillin mutants used in the study are shown in Fig. 3B. One lacks the LD4 domain that promotes neurite outgrowth in PC12 cells grown on collagen substrates [28]. The other has Tyr to Phe substitutions at Tyr31, Tyr118 and Tyr182. This prevents the phosphorylation of Tyr31 and Tyr118 implicated in cell motility [29] but has no effect on neurite outgrowth on collagen [28]. Western blot experiments confirmed that tyrosine phosphorylation was not observed in the triple mutant paxillin, whilst the wild type and LD4 deletion mutant forms were recognised by the anti-phosphotyrosine antibody (Fig 3C). To determine the relative expression levels of expressed and endogenous paxillin, we performed western blots using antibodies against paxillin (that will recognise both the endogenous and expressed molecules) and against the FLAG-tag present only in the expressed form of paxillin. As shown in Fig 3D, there were greater levels of the expressed paxillins than the endogenous paxillin, and the levels of the mutant paxillins were comparable with those seen with the wild-type construct (Fig 3D). To confirm that the expressed paxillins associated with α4, we western blotted using anti-paxillin antibodies cell lysates immunoprecipitated first with anti-α4 and then with the anti FLAG-tag antibody that detects only the expressed paxillin. All the expressed paxillins associated with α4 (Fig 3E). Notably the LD4 deletion mutant associated with the integrin, even though the critical residues for α4 binding have been shown to be present in an Ala176 to Asp275 fragment containing the LD3 and LD4 domains [30].
Next, we examined neurite outgrowth in PC12 cells co-expressing wild-type α4 and the different paxillins. In cells expressing wild-type paxillin, we found enhanced neurite outgrowth on Fn V120 compared with cells expressing α4 alone (Fig 4). In keeping with the previous studies [28], no significant difference in the extent of neurite outgrowth on Fn V120 substrates was seen between PC12 cells expressing wild type and the triple tyrosine mutant paxillin, while the LD4 domain mutant supported neurite outgrowth to a significantly reduced extent (Fig 4). These over-expression studies therefore confirm a role for paxillin in promoting neurite outgrowth on Fn V120 substrates. Taken together with the studies above showing that association of paxillin with the integrin cytoplasmic domain is required for outgrowth, they show that the recruitment of paxillin by α4 integrin represents an important mechanism for neurite outgrowth on these embryonic fibronectins.
As outlined in the introduction, the binding of paxillin to the α4 integrin is inhibited by phosphorylation of Ser988. The temporal and spatial regulation of paxillin binding enabled by phosphorylation and dephosphorylation may have an important physiological role in cell motility, as both S988A (that inhibits phosphorylation) and S988D (that mimics phosphorylation) mutations inhibit migration in cell culture [24]. In the final part of this study, therefore, we examined the role of phosphorylation of Ser988 in neurite outgrowth on Fn V120 substrates. We transfected PC12 cells with either wild type α4 or α4 with S988A or S988D mutations. As an additional control, we used a control S990A mutation that does not alter paxillin binding (Fig 5A). All four α4 constructs were equally expressed in PC12 cells (Fig 5B). Wild type α4 was not phosphorylated in PC12 cells as judged by immuno blotting with a phospho-α4 specific antibody whereas it was heavily phosphorylated in Jurkat cells (a T cell line) (Fig. 5B). In contrast, the α4(S988D) was labelled by this antibody in PC12 cells suggesting that this mutation is also phospho-mimetic with respect to recognition by this antibody. We confirmed the effect of these mutations on the binding of paxillin to the α4 cytoplasmic domain by immunoprecipitating α4 from lysates of cells expressing each of the four different α4 subunits and detecting paxillin by immunoblotting. As expected, no associated paxillin was seen with the phospho-mimicking S988D mutation, whilst all others bound paxillin as expected (Fig 6A). Lastly, we examined neurite outgrowth in PC12 cells expressing each of the four α4 subunits on Fn V120. As before, wild-type α4 promoted neurite outgrowth and this was not affected by the S988A or S990A mutations. In contrast, however, the S988D mutation that blocked paxillin binding showed no α4-mediated neurite outgrowth, as the levels of outgrowth were identical to those of PC12 cells not expressing α4 integrin (Fig 6B). We conclude that mutations that mimic phosphorylation of Ser988 will prevent paxillin binding in PC12 cells, and therefore inhibit the neurite outgrowth promoted by α4 integrin on Fn V120 substrates. However, the lack of detectable phosphorylation of wild-type α4 in the PC12 cells and the efficacy of α4(S988A) in promoting neurite outgrowth suggests that the reversibility of paxillin binding enabled by phosphorylation of Ser988 is not necessary for neurite outgrowth.
Discussion
We have shown that the association of the adaptor protein, paxillin, with the α4 integrin cytoplasmic domain is required for α4-mediated neurite outgrowth in PC12 cells on embryonic Fns that are expressed during nerve regeneration. In the absence of this association, the enhanced neurite outgrowth on Fn V120 is lost. Conversely, overexpression of paxillin enhances α4-dependent neurite outgrowth on this substrate. As alternative splicing of the Fn primary gene transcript following PNS injury increases the number of Fns containing V120 in the regenerating nerve, and α4 is expressed on regenerating growth cones in vivo [16,17], we suggest that the α4-paxillin interaction may play a key role in PNS regeneration. The upregulation of different ECM molecules during PNS regeneration suggests that other integrins will also contribute to regeneration. Two candidates are α5β1, that is also expressed on regenerating growth cones in vivo [2], and α7β1, a laminin receptor expressed on a subset of DRG neurons and also shown to be required for normal regeneration of motor neurones in the facial nerve [31,32]. However, as α4 and the closely related α9 are the only integrin α subunits reported to bind paxillin directly [20,33], it is likely that these integrins will activate different and parallel downstream pathways so providing an enhanced regenerative response triggered by the change in Fn splicing.
The demonstration of a functional requirement for paxillin binding to a specific integrin to promote neurite outgrowth extends previous work on signalling within growth cones. Paxillin is localised within neuronal growth cones within dynamic focal adhesions that assemble at the leading edge, enlarge and finally disassemble while remaining in place as the growth cone advances [34]. It becomes tyrosine phosphorylated on neurite outgrowth [35,36]. Tyrosine phosphorylation of paxillin at Tyr31 and Tyr118 provide a binding site for the adaptor protein Crk [37,38]. Overexpression of Crk promotes neurite outgrowth in PC12 cells [39], and also promotes cell motility in a bladder cancer cell line [29]. Surprisingly, therefore, but in agreement with others [28], we find no evidence that inhibiting tyrosine phosphorylation by expression of a paxillin with Tyr to Phe mutations in these residues inhibits the ability of over-expressed paxillin to enhance neurite outgrowth. Two other regions of paxillin have been implicated in neurite outgrowth signalling. First, Ser 85 (Ser 83 in the rat) that is phosphorylated by p38 MAPK. This phosphorylation promotes outgrowth that can be inhibited by a Ser to Ala mutation in this position [40]. Second, the LD4 domain within the leucine-rich domains, that provides a binding site for a number of other signalling proteins [41,42] and deletion of which inhibits neurite outgrowth in PC12 cells growing on collagen substrates [28]. Our results show markedly less neurite outgrowth following expression of the LD4 deletion mutant than full-length paxillin, suggesting an important role of the LD4 domain in α4 integrin-mediated neurite formation.
At least three possible mechanisms by which the LD4 domain might promote neurite outgrowth have been described. First, by the recruitment of the focal adhesion kinases FAK and Pyk2, both of which interact directly with the LD4 domain [41], are co-immunoprecipitated with paxillin in PC12 cells [43] and are activated by interaction with the α4 cytoplasmic domain in T cells [44]. Activation of these kinases occurs during growth factor-induced neurite outgrowth in PC12 cells, and inhibition of their activity by expression of a C-terminal construct (PRNK and FRNK, which act as dominant negative isoforms) blocks this outgrowth [28]. Second, by recruitment of the Arf-GAP PKL. This in turn recruits the p21-associated kinase (PAK1) via a linking interaction with PIX [45], and PAK1 localisation to the membrane in PC12 cells has been shown to promote neurite outgrowth [46], potentially via the activation of cdc42 [42]. Third, by the regulation of the Rho GTPase Rac, as demonstrated by the prolonged levels of Rac activation, loss of directional migration and randomly-orientated lamellopodia formation in CHO cells expressing an LD4 deletion mutant of paxillin [47]. As expression of a constitutively active form of Rac in PC12 cells promotes cell spreading but not neurite outgrowth, whilst dominant negative Rac inhibits neurite outgrowth [46], precise regulation of Rac activity is clearly essential for neurite formation by PC12 cells.
Directional cell migration is associated with high levels of Rac activity at the leading edge of the cell [48]. A key question for future studies will therefore be the role of paxillin and α4 integrin in the spatial control of Rac activity within the growth cone. The phosphorylated α4 integrins found in T cell lines will not bind paxillin [22], and are localised to the leading edge of migrating CHO cells, while non-phosphorylated integrins are found in the lateral edges [23]. As the α4-paxillin interaction inhibits Rac activity and lamellopodia formation, the extension of cellular protrusions driven by Rac and required for movement are therefore restricted to the leading edge of the cell and directionality of migration is achieved [23,49]. A mechanism by which the paxillin-integrin interaction reduces Rac activity is provided by recent work showing that recruitment of the Arf-GAP GIT1 inactivates Arf activity, which in turn results in decreased Rac activation [49]. Integrin phosphorylation therefore plays a key role in directional migration in some non-neural cells by providing a mechanism for the spatial regulation of paxillin dissociation from the α4 cytoplasmic domain.
Conclusion
Peripheral nerve repair is associated with a re-expression of embryonic isoforms of fibronectin that contain a binding site for the α4β1 integrin. This integrin is expressed on regenerating growth cones and promotes neurite outgrowth on embryonic fibronectins in cell culture. Using PC12 neuronal cells as a model system, we have shown here that the interaction of the adaptor protein paxillin with the integrin cytoplasmic domain is required for this outgrowth. We have also found that, in contrast to studies using T cells, there is no biochemical evidence for phosphorylation of α4 integrins in PC12 cells, suggesting that the dissociation of paxillin from the integrin mediated by this phosphorylation is either not required for growth cone motility or is achieved through different means. This points to potentially interesting differences between the mechanisms of two distinct forms of cell movement; directional migration in isolated cells and growth cone motility at the end of a cell process or axon.
Methods
Reagents
The following antibodies were used: mouse anti-human α4, clone HP2/1 (Chemicon); hamster anti-rat β1, clone HA2/5 (Pharmingen), mouse anti-paxillin, clone 349 (Transduction Laboratories), mouse anti-FLAG M2 (Sigma), anti-phosphotyrosine antibody PY20 (Transduction Laboratories). The rabbit phospho-specific anti-α1 antibody was generated against a synthetic peptide containing phospho serine at Ser988 as previously described [22]. Recombinant Fn fragments encompassing repeats III-8-15 and containing all three possible combinations of the V region (and therefore the α4β1 integrin binding site) were prepared as described previously [17].
Integrin constructs
Construction of pLXIN (Clontech) retroviral vectors encoding α4 integrin chimeras was previously described [17]. α4 cytoplasmic domain mutations were introduced in the full length α4 chain by site-directed mutagenesis using the QuickChange kit (Stratagene). Alanine substitutions were introduced at Y991 and E983 using the following primers: 5'-C AGA AGA GAC AGT TGG AGT GCT ATC AAC AGT AAA AGC AAT G-3' (sense) and the corresponding antisense primer for substitution of Y991 (Y991A), and 5'-C AAA TCT ATC CTA CAA GAA GCA AAC AGA AGA GAC AGT TGG-3' (sense) with the matching antisense for substitution of E983 (E983A). Double mutants were generated by successive rounds of mutagenesis, first E983, followed by Y991 substitution. Alanine and aspartic acid substitutions were introduced at S988 and S990 using the following sense primers: S988A, 5'-C CTA CAA GAA GAA AAC AGA AGA GAC GCT TGG AGT TAT ATC AAC AG-3'; S988D, 5'-C CTA CAA GAA GAA AAC AGA AGA GAC GAT TGG AGT TAT ATC AAC AG-3'; S990A, 5'-GAA AAC AGA GAC AGT TGG GCT TAT ATC AAC AGT AAA AGC-3'. Base substitution at the sites of mutation was confirmed by DNA sequencing. The nomenclature used for the different constructs is as follows: α4Y991A, α4E983A, α4EY, α4S988A, α4S988D, α4S990A.
Paxillin constructs
The cDNA sequence encoding wild-type chicken paxillin, and paxillin with three Y to F substitutions at Y31, Y118 and Y 182, were a gift from I Dikic (Uppsala, Sweden). PCR was used to generate a Xba I site replacing the ATG of paxillin at the 5' end, and a Xho I site at the 3'end. A Not I-Xba I linker fragment including a Xho I site and a FLAG tag, DYKDDDDK, was ligated at the 5' end of the paxillin sequence. The sequence encoding the FLAG tag in frame with paxillin was excised using Xho I and cloned into the pMSCVpuro retroviral vector (Clontech). Paxillin lacking the LD4 domain was generated by two sequential rounds of site-directed mutagenesis, using the following deletion primers: 5'-GCT TCT TCA GCT ACA CGA GAA GAG CTG ATG GCG TCC CTC-3' and 5'-GAG GGA CGC CAT CTC TTC TCG TGT AGC TGA AGA AGC-3'. All cDNAs were sequenced.
Cell culture
COS7, NIH 3T3 and GP+86 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS), penicillin/streptomycin, and 2 mM glutamine (Sigma). Rat pheochromocytoma (PC12) cells were grown on Poly-D-Lysine-coated tissue culture flasks in DMEM supplemented with 10% horse serum (HS), 5% FCS, penicillin/streptomycin, and 2 mM glutamine. For priming with NGF, PC12 cells were passaged onto collagen-coated (calf-skin type I collagen; Sigma) dishes at a density of 104 cells/cm2 and cultured for 4–5 days in DMEM complemented with 1% ITS+™ premix (BD Biosciences) and 50 ng/ml 2.5S NGF (Serotec). For neurite outgrowth assays, NGF-primed cells were passaged by gentle trituration and cultured on the desired substrate in the same defined medium.
Transfection and selection of stable clones
The constructs were transfected into the GP+86 packaging cell line, using FuGENE 6 as transfection reagent (Roche). Cells were maintained in DMEM supplemented with 10% FCS, penicillin/streptomycin, 2 mM glutamine and 1 mg/ml G418 (Life Technologies) for pLXIN construct or 2 μg/ml puromycin for pMSCV constructs until clones appeared. The GP+86 cells were then replated and when at 60–70% confluence, were washed and medium without G418 or puromycin added. The conditioned medium was collected after 24 h, filtered (0.45 μm pore size), and either used immediately or aliquoted, frozen and kept at -70°C. Virus titer was determined using NIH 3T3 cells and equalized. PC12 cells were transfected using polybrene, and selected and maintained with 1 mg/ml G418 or 2 μg/ml puromycin.
Neurite outgrowth assay
All neurite outgrowth assays were performed in 4-well tissue culture plates. For preparation of the substrates, ligands (recombinant Fn fragments, type-1 collagen (Sigma) or laminin-1 (Sigma)) were diluted into PBS, and deposited as 30 μl drops in the centre of the wells. Where recombinant Fn fragments were used, the dishes were first coated with 10 μg/ml anti-IgG1 Fc for 3 h at RT, washed 2 × with PBS and then post-adsorbed with the various fragments overnight at 4°C. The wells were washed 3 × with PBS, and PC12 cells were deposited in 30 μl drops on the substrates at a density of 5 × 102/cm2. After 1 h, when the cells were attached, the medium was added, to a final volume of 200 μl. Measures were taken after 12 h, and 24 h (separate sets of dishes). Dishes were viewed under a Zeiss phase microscope and random fields were recorded using a digital camera and analyzed using the Openlab software (Improvision). The length of neurites was determined using the Openlab software. Determinations were made on at least 5 separate experiments, and the longest neurite of at least 100 PC12 cells/well was measured. Untransfected and mock-transfected cells were used as negative controls. Assays were conducted on both experimental and control cell lines in parallel, and the results normalized to the control values. Each assay was repeated 5 times with three independently generated cell lines, and the mean determined. Statistical analysis was performed using Students t test.
Immunoprecipitations and Western blotting
Cell surface molecules were labelled with 0.1 mg/ml NHS-LC-Biotin (Pierce) in PBS at 37°C for 30 min. Cells were washed three times with cell wash buffer (50 mM Tris-HCl pH 7.5, 0.15 M NaCl, 1 mM CaCl2, 1 mM MgCl2), scraped, and lysed on ice for 30 min in extraction buffer: cell wash buffer plus 1% Triton X-100, 0.05% Tween 20, 2 mM PMSF, 1 μg/ml pepstatin A, 2 μg/ml aprotinin, 5 μg/ml leupeptin, 2 mM sodium vanadate, 4 mM sodium pyrophospate, and 2 mM sodium fluoride. After clarification by centrifuging at 14000 rpm for 20 min at 4°C, the cell lysates were incubated with 30 μl protein A-sepharose (Pharmacia). The lysates were pre-cleared by two sequential 2 h incubations (to remove proteins showing non-specific interactions with protein-A) and immunoprecipitations were then carried out overnight at 4°C. Where hamster or mouse monoclonal antibodies were used, rabbit anti-hamster or rabbit anti-mouse antisera (Nordic Immunological Laboratories) were also added to the tubes (1 in 250). All antibodies were used at a dilution of 1:250. The beads were washed 5 × with immunoprecipitation buffer (identical to cell wash buffer except for 0.5 M NaCl and 1% NP40) and the precipitated polypeptides were extracted with SDS sample buffer. Precipitated cell surface biotin-labelled molecules were separated by SDS-PAGE under non-reducing conditions and detected with streptavidin-peroxidase followed by ECL (Amersham). Co-precipitated paxillin was detected by immuno-blotting of reduced immunoprecipitates with anti-paxillin antibody, followed by anti-mouse-peroxidase conjugated antibody and ECL.
Authors' contributions
MV performed all the experiments in this study, with UBF contributing to those analysing the effects of paxillin over-expression. JH generated and validated the phospho-specific anti-α4 integrin antibody. Cff-C and MHG conceived the study, based on previous work in their respective laboratories. Cff-C wrote the first draft of the manuscript, with subsequent contributions from all authors who also approved the final manuscript.
Acknowledgments
Acknowledgements
This work was supported by a grant from Action Medical Research (Horsham, UK) to MV and Cff-C, a Wellcome Trust Studentship to UBF, Grant AR27214 from the National Institutes of Health to MHG and an Arthritis Investigator award from the Arthritis Foundation to JH. We are very grateful to Dr Ivan Dikic for the gift of paxillin constructs.
Contributor Information
Mariette Vogelezang, Email: mariette@MIT.EDU.
Ulrike B Forster, Email: buf20@cam.ac.uk.
Jaewon Han, Email: jahan@ucsd.edu.
Mark H Ginsberg, Email: mhginsberg@ucsd.edu.
Charles ffrench-Constant, Email: cfc@mole.bio.cam.ac.uk.
References
- Martini R. Expression and functional roles of neural cell surface molecules and extracellular matrix components during development and regeneration of peripheral nerves. J Neurocytol. 1994;23:1–28. doi: 10.1007/BF01189813. [DOI] [PubMed] [Google Scholar]
- Lefcort F, Venstrom K, McDonald JA, Reichardt LF. Regulation of expression of fibronectin and its receptor, alpha 5 beta 1, during development and regeneration of peripheral nerve. Development. 1992;116:767–782. doi: 10.1242/dev.116.3.767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wallquist W, Patarroyo M, Thams S, Carlstedt T, Stark B, Cullheim S, Hammarberg H. Laminin chains in rat and human peripheral nerve: distribution and regulation during development and after axonal injury. J Comp Neurol. 2002;454:284–293. doi: 10.1002/cne.10434. [DOI] [PubMed] [Google Scholar]
- Schwarzbauer JE, Tamkun JW, Lemischka IR, Hynes RO. Three different fibronectin mRNAs arise by alternative splicing within the coding region. Cell. 1983;35:421–431. doi: 10.1016/0092-8674(83)90175-7. [DOI] [PubMed] [Google Scholar]
- Schwarzbauer JE, Patel RS, Fonda D, Hynes RO. Multiple sites of alternative splicing of the rat fibronectin gene transcript. Embo J. 1987;6:2573–2580. doi: 10.1002/j.1460-2075.1987.tb02547.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tamkun JW, Schwarzbauer JE, Hynes RO. A single rat fibronectin gene generates three different mRNAs by alternative splicing of a complex exon. Proc Natl Acad Sci U S A. 1984;81:5140–5144. doi: 10.1073/pnas.81.16.5140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kornblihtt AR, Umezawa K, Vibe-Pedersen K, Baralle FE. Primary structure of human fibronectin: differential splicing may generate at least 10 polypeptides from a single gene. Embo J. 1985;4:1755–1759. doi: 10.1002/j.1460-2075.1985.tb03847.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gutman A, Kornblihtt AR. Identification of a third region of cell-specific alternative splicing in human fibronectin mRNA. Proc Natl Acad Sci U S A. 1987;84:7179–7182. doi: 10.1073/pnas.84.20.7179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zardi L, Carnemolla B, Siri A, Petersen TE, Paolella G, Sebastio G, Baralle FE. Transformed human cells produce a new fibronectin isoform by preferential alternative splicing of a previously unobserved exon. Embo J. 1987;6:2337–2342. doi: 10.1002/j.1460-2075.1987.tb02509.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wayner EA, Garcia-Pardo A, Humphries MJ, McDonald JA, Carter WG. Identification and characterization of the T lymphocyte adhesion receptor for an alternative cell attachment domain (CS-1) in plasma fibronectin. J Cell Biol. 1989;109:1321–1330. doi: 10.1083/jcb.109.3.1321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guan JL, Hynes RO. Lymphoid cells recognize an alternatively spliced segment of fibronectin via the integrin receptor alpha 4 beta 1. Cell. 1990;60:53–61. doi: 10.1016/0092-8674(90)90715-Q. [DOI] [PubMed] [Google Scholar]
- ffrench-Constant C. Alternative splicing of fibronectin--many different proteins but few different functions. Exp Cell Res. 1995;221:261–271. doi: 10.1006/excr.1995.1374. [DOI] [PubMed] [Google Scholar]
- Peters JH, Hynes RO. Fibronectin isoform distribution in the mouse. I. The alternatively spliced EIIIB, EIIIA, and V segments show widespread codistribution in the developing mouse embryo. Cell Adhes Commun. 1996;4:103–125. doi: 10.3109/15419069609010766. [DOI] [PubMed] [Google Scholar]
- Peters JH, Chen GE, Hynes RO. Fibronectin isoform distribution in the mouse. II. Differential distribution of the alternatively spliced EIIIB, EIIIA, and V segments in the adult mouse. Cell Adhes Commun. 1996;4:127–148. doi: 10.3109/15419069609010767. [DOI] [PubMed] [Google Scholar]
- Mathews GA, ffrench-Constant C. Embryonic fibronectins are up-regulated following peripheral nerve injury in rats. J Neurobiol. 1995;26:171–188. doi: 10.1002/neu.480260203. [DOI] [PubMed] [Google Scholar]
- Vogelezang MG, Scherer SS, Fawcett JW, ffrench-Constant C. Regulation of fibronectin alternative splicing during peripheral nerve repair. J Neurosci Res. 1999;56:323–333. doi: 10.1002/(SICI)1097-4547(19990515)56:4<323::AID-JNR1>3.0.CO;2-6. [DOI] [PubMed] [Google Scholar]
- Vogelezang MG, Liu Z, Relvas JB, Raivich G, Scherer SS, ffrench-Constant C. Alpha4 integrin is expressed during peripheral nerve regeneration and enhances neurite outgrowth. J Neurosci. 2001;21:6732–6744. doi: 10.1523/JNEUROSCI.21-17-06732.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chan BM, Kassner PD, Schiro JA, Byers HR, Kupper TS, Hemler ME. Distinct cellular functions mediated by different VLA integrin alpha subunit cytoplasmic domains. Cell. 1992;68:1051–1060. doi: 10.1016/0092-8674(92)90077-P. [DOI] [PubMed] [Google Scholar]
- Kassner PD, Alon R, Springer TA, Hemler ME. Specialized functional properties of the integrin alpha 4 cytoplasmic domain. Mol Biol Cell. 1995;6:661–674. doi: 10.1091/mbc.6.6.661. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu S, Thomas SM, Woodside DG, Rose DM, Kiosses WB, Pfaff M, Ginsberg MH. Binding of paxillin to alpha4 integrins modifies integrin-dependent biological responses. Nature. 1999;402:676–681. doi: 10.1038/45264. [DOI] [PubMed] [Google Scholar]
- Feral CC, Rose DM, Han J, Fox N, Silverman GJ, Kaushansky K, Ginsberg MH. Blocking the alpha 4 integrin-paxillin interaction selectively impairs mononuclear leukocyte recruitment to an inflammatory site. J Clin Invest. 2006;116:715–723. doi: 10.1172/JCI26091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Han J, Liu S, Rose DM, Schlaepfer DD, McDonald H, Ginsberg MH. Phosphorylation of the integrin alpha 4 cytoplasmic domain regulates paxillin binding. J Biol Chem. 2001;276:40903–40909. doi: 10.1074/jbc.M102665200. [DOI] [PubMed] [Google Scholar]
- Goldfinger LE, Han J, Kiosses WB, Howe AK, Ginsberg MH. Spatial restriction of alpha4 integrin phosphorylation regulates lamellipodial stability and alpha4beta1-dependent cell migration. J Cell Biol. 2003;162:731–741. doi: 10.1083/jcb.200304031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Han J, Rose DM, Woodside DG, Goldfinger LE, Ginsberg MH. Integrin alpha 4 beta 1-dependent T cell migration requires both phosphorylation and dephosphorylation of the alpha 4 cytoplasmic domain to regulate the reversible binding of paxillin. J Biol Chem. 2003;278:34845–34853. doi: 10.1074/jbc.M304691200. [DOI] [PubMed] [Google Scholar]
- Liu S, Ginsberg MH. Paxillin binding to a conserved sequence motif in the alpha 4 integrin cytoplasmic domain. J Biol Chem. 2000;275:22736–22742. doi: 10.1074/jbc.M000388200. [DOI] [PubMed] [Google Scholar]
- Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell. 1992;69:11–25. doi: 10.1016/0092-8674(92)90115-S. [DOI] [PubMed] [Google Scholar]
- Tomaselli KJ, Hall DE, Flier LA, Gehlsen KR, Turner DC, Carbonetto S, Reichardt LF. A neuronal cell line (PC12) expresses two beta 1-class integrins-alpha 1 beta 1 and alpha 3 beta 1-that recognize different neurite outgrowth-promoting domains in laminin. Neuron. 1990;5:651–662. doi: 10.1016/0896-6273(90)90219-6. [DOI] [PubMed] [Google Scholar]
- Ivankovic-Dikic I, Gronroos E, Blaukat A, Barth BU, Dikic I. Pyk2 and FAK regulate neurite outgrowth induced by growth factors and integrins. Nat Cell Biol. 2000;2:574–581. doi: 10.1038/35023515. [DOI] [PubMed] [Google Scholar]
- Petit V, Boyer B, Lentz D, Turner CE, Thiery JP, Valles AM. Phosphorylation of tyrosine residues 31 and 118 on paxillin regulates cell migration through an association with CRK in NBT-II cells. J Cell Biol. 2000;148:957–970. doi: 10.1083/jcb.148.5.957. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu S, Kiosses WB, Rose DM, Slepak M, Salgia R, Griffin JD, Turner CE, Schwartz MA, Ginsberg MH. A fragment of paxillin binds the alpha 4 integrin cytoplasmic domain (tail) and selectively inhibits alpha 4-mediated cell migration. J Biol Chem. 2002;277:20887–20894. doi: 10.1074/jbc.M110928200. [DOI] [PubMed] [Google Scholar]
- Werner A, Willem M, Jones LL, Kreutzberg GW, Mayer U, Raivich G. Impaired axonal regeneration in alpha7 integrin-deficient mice. J Neurosci. 2000;20:1822–1830. doi: 10.1523/JNEUROSCI.20-05-01822.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gardiner NJ, Fernyhough P, Tomlinson DR, Mayer U, von der Mark H, Streuli CH. Alpha7 integrin mediates neurite outgrowth of distinct populations of adult sensory neurons. Mol Cell Neurosci. 2005;28:229–240. doi: 10.1016/j.mcn.2004.08.017. [DOI] [PubMed] [Google Scholar]
- Young BA, Taooka Y, Liu S, Askins KJ, Yokosaki Y, Thomas SM, Sheppard D. The cytoplasmic domain of the integrin alpha9 subunit requires the adaptor protein paxillin to inhibit cell spreading but promotes cell migration in a paxillin-independent manner. Mol Biol Cell. 2001;12:3214–3225. doi: 10.1091/mbc.12.10.3214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Woo S, Gomez TM. Rac1 and RhoA promote neurite outgrowth through formation and stabilization of growth cone point contacts. J Neurosci. 2006;26:1418–1428. doi: 10.1523/JNEUROSCI.4209-05.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khan MA, Okumura N, Okada M, Kobayashi S, Nakagawa H. Nerve growth factor stimulates tyrosine phosphorylation of paxillin in PC12h cells. FEBS Lett. 1995;362:201–204. doi: 10.1016/0014-5793(95)00250-D. [DOI] [PubMed] [Google Scholar]
- Leventhal PS, Feldman EL. Tyrosine phosphorylation and enhanced expression of paxillin during neuronal differentiation in vitro. J Biol Chem. 1996;271:5957–5960. doi: 10.1074/jbc.271.11.5957. [DOI] [PubMed] [Google Scholar]
- Schaller MD, Parsons JT. pp125FAK-dependent tyrosine phosphorylation of paxillin creates a high-affinity binding site for Crk. Mol Cell Biol. 1995;15:2635–2645. doi: 10.1128/mcb.15.5.2635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Birge RB, Fajardo JE, Reichman C, Shoelson SE, Songyang Z, Cantley LC, Hanafusa H. Identification and characterization of a high-affinity interaction between v-Crk and tyrosine-phosphorylated paxillin in CT10-transformed fibroblasts. Mol Cell Biol. 1993;13:4648–4656. doi: 10.1128/mcb.13.8.4648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hempstead BL, Birge RB, Fajardo JE, Glassman R, Mahadeo D, Kraemer R, Hanafusa H. Expression of the v-crk oncogene product in PC12 cells results in rapid differentiation by both nerve growth factor- and epidermal growth factor-dependent pathways. Mol Cell Biol. 1994;14:1964–1971. doi: 10.1128/mcb.14.3.1964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang C, Borchers CH, Schaller MD, Jacobson K. Phosphorylation of paxillin by p38MAPK is involved in the neurite extension of PC-12 cells. J Cell Biol. 2004;164:593–602. doi: 10.1083/jcb.200307081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Turner CE. Paxillin and focal adhesion signalling. Nat Cell Biol. 2000;2:E231–6. doi: 10.1038/35046659. [DOI] [PubMed] [Google Scholar]
- Brown MC, Turner CE. Paxillin: adapting to change. Physiol Rev. 2004;84:1315–1339. doi: 10.1152/physrev.00002.2004. [DOI] [PubMed] [Google Scholar]
- Park SY, Avraham H, Avraham S. Characterization of the tyrosine kinases RAFTK/Pyk2 and FAK in nerve growth factor-induced neuronal differentiation. J Biol Chem. 2000;275:19768–19777. doi: 10.1074/jbc.M909932199. [DOI] [PubMed] [Google Scholar]
- Rose DM, Liu S, Woodside DG, Han J, Schlaepfer DD, Ginsberg MH. Paxillin binding to the alpha 4 integrin subunit stimulates LFA-1 (integrin alpha L beta 2)-dependent T cell migration by augmenting the activation of focal adhesion kinase/proline-rich tyrosine kinase-2. J Immunol. 2003;170:5912–5918. doi: 10.4049/jimmunol.170.12.5912. [DOI] [PubMed] [Google Scholar]
- Turner CE, Brown MC, Perrotta JA, Riedy MC, Nikolopoulos SN, McDonald AR, Bagrodia S, Thomas S, Leventhal PS. Paxillin LD4 motif binds PAK and PIX through a novel 95-kD ankyrin repeat, ARF-GAP protein: A role in cytoskeletal remodeling. J Cell Biol. 1999;145:851–863. doi: 10.1083/jcb.145.4.851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Daniels RH, Hall PS, Bokoch GM. Membrane targeting of p21-activated kinase 1 (PAK1) induces neurite outgrowth from PC12 cells. Embo J. 1998;17:754–764. doi: 10.1093/emboj/17.3.754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- West KA, Zhang H, Brown MC, Nikolopoulos SN, Riedy MC, Horwitz AF, Turner CE. The LD4 motif of paxillin regulates cell spreading and motility through an interaction with paxillin kinase linker (PKL) J Cell Biol. 2001;154:161–176. doi: 10.1083/jcb.200101039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kraynov VS, Chamberlain C, Bokoch GM, Schwartz MA, Slabaugh S, Hahn KM. Localized Rac activation dynamics visualized in living cells. Science. 2000;290:333–337. doi: 10.1126/science.290.5490.333. [DOI] [PubMed] [Google Scholar]
- Nishiya N, Kiosses WB, Han J, Ginsberg MH. An alpha4 integrin-paxillin-Arf-GAP complex restricts Rac activation to the leading edge of migrating cells. Nat Cell Biol. 2005;7:343–352. doi: 10.1038/ncb1234. [DOI] [PubMed] [Google Scholar]