|
|
|
|
|
|
|
|
|||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | |||||
First published online 15 May 2007
doi: 10.1242/jcs.03459
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
1 Ludwig Institute for Cancer Research, Hospital Alemão Oswaldo Cruz, São Paulo, Brazil
2 Departamento de Bioquímica, Instituto de Química, Universidade de São Paulo, São Paulo, Brazil
3 Centro de Tratamento e Pesquisa Hospital do Câncer, São Paulo, Brazil
4 Departamento de Patologia Básica, Universidade Federal do Paraná, Curitiba, Brazil
5 Departamento de Biologia Celular, Universidade Federal do Paraná, Curitiba, Brazil
6 INFAR, Universidade Federal de São Paulo, São Paulo, Brazil
* Author for correspondence (e-mail: vmartins{at}ludwig.org.br)
Accepted 11 April 2007
 |
Summary |
|---|
 Top Summary IntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
v
3 activity. Our findings indicate that PrPC plays an important role in axonal growth, and this function may be rescued in PrPC-knockout animals by integrin compensatory mechanisms.
Key words: Dorsal root ganglia, Extracellular matrix, Cellular prion protein, Vitronectin, Axon growth, Integrins
|
Introduction |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
). It has the ability to bind copper (Brown et al., 1997), to protect against oxidative stress (Brown and Besinger, 1998) and has been shown to influence cell signaling mechanisms, neuronal survival and differentiation (Chen, S. et al., 2003; Chiarini et al., 2002; Lopes et al., 2005; Mouillet-Richard et al., 2000). PrPC has also been shown to bind the laminin receptor (Gauczynski et al., 2001; Hundt et al., 2001), to promote neuritogenesis through its interaction with NCAM (Santuccione et al., 2005; Schmitt-Ulms et al., 2001) and to induce neurite maintenance and neuronal differentiation through binding to the extracellular matrix (ECM) protein laminin (Ln) (Graner et al., 2000a; Graner et al., 2000b).
ECM proteins are known to regulate neuronal differentiation and axonal regeneration (Turney and Bridgman, 2005; Easley et al., 2006; Tom et al., 2004). Vitronectin (Vn) is expressed during development on embryonic day 10 (E10) in mice, mainly in the central nervous system (Seiffert et al., 1995), and has been shown to support proliferation and differentiation of cultured neurons (Martinez-Morales et al., 1995). In dorsal root ganglia (DRG) neurons, Vn-mediated axonal growth can be inhibited by anti-Vn antibody (Isahara and Yamamoto, 1995). Vn can also induce motor neuron differentiation, and anti-Vn antibodies reduce the number of motor neurons generated in chicken embryos (Martinez-Morales et al., 1997; Pons and Marti, 2000).
The classical ECM receptors, integrins, can bind several molecules and have been associated with ECM biological functions. The integrin recognition sequence Arg-Gly-Asp (RGD) is present in numerous ECM proteins including collagen, Vn and Fibronectin (Fn). RGD peptide is biologically active and able to substitute for ECM proteins in a variety of situations (Hynes, 2002; Pierschbacher and Ruoslahti, 1984).
Given that PrPC binds Ln, we hypothesized that PrPC might act as a broad ECM ligand and tested whether PrPC binds the ECM proteins Vn, Fn and type IV collagen. We also sought to characterize the cellular events associated with any such binding. We assessed the role of PrPC-ECM interaction in DRG axon outgrowth using primary cultures obtained from two different PrPC-null mouse strains (ZrchI and Npu) and their respective wild-type controls. The role of integrins in wild-type and PrPC-null DRG axonal growth was also addressed and integrin activity was evaluated using specific antibodies and in functional assays using RGD peptide.
|
Results |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
), and also interacts with Vn (note that 10 µg of Ln represents 10 times fewer moles than the other proteins, owing to its higher molecular mass of 
900 kDa). Conversely, PrPC did not associate with Fn or type IV collagen. Binding assays demonstrated that His6-PrPC binding to Vn is dose dependent and saturable (Fig. 1b), with a Kd of 12 nM. His6-PrPC refolded in the presence of copper (Zanata et al., 2002b) presented a similar binding to Vn (data not shown). It is important to note that the Vn used here was highly pure (see supplementary material Fig. S1a), and recombinant PrPC analyzed by circular dichroism spectra showed an -helix structure (Cordeiro et al., 2004a; Cordeiro et al., 2004b).
|
PrPC interaction with 125I-Vn could be blocked dose-dependently by competition with increasing concentrations of unlabeled Vn, as well as by other PrPC ligands, such as stress-inducible protein 1 (STI1) (Zanata et al., 2002b) and Ln (Graner et al., 2000a; Graner et al., 2000b), but not by BSA (Fig. 1c). Thus, only specific PrPC ligands can disrupt its binding interactions. The peptide representing the specific PrPC binding site in the STI1 molecule (STI1 peptide) (Chiarini et al., 2002; Zanata et al., 2002b), but not that from Ln (Ln 
-1 peptide) (Graner et al., 2000a), competed for the PrPC-Vn interaction (Fig. 1d). These data suggest that STI1 and Vn share a binding site in the PrPC molecule, whereas Ln must interact with another PrPC domain. Other Vn ligands such as type I collagen, type IV collagen and heparin were also tested for their ability to compete with the PrPC-Vn interaction (supplementary material Fig. S2b). Although they do not bind Vn in the same domain as PrPC (supplementary material Fig. S2a), they are able to disturb the PrPC-Vn interaction, probably by steric hindrance.
Characterization of the PrPC and Vn interacting domains
Of the twenty peptides from mouse PrPC covering the whole protein sequence used to compete for PrPC-Vn binding, two of them, corresponding to PrPC residues 103-122 and 113-132, effectively blocked (by 80%) PrPC-Vn binding (Fig. 2a). This implies that the amino acid sequence shared by both peptides may represent the putative binding site for Vn in the PrPC molecule. The binding and competition assays were always performed with a freshly prepared peptide solution to avoid possible neurotoxic aggregates (Chiarini et al., 2002; Ettaiche et al., 2000). The lack of PrPC-Vn interaction blockade by peptide 93-112, which can also form aggregates owing to the partial presence of the hydrophobic domain, provides further evidence that the interaction blockade was not the result of peptide aggregation.
|
Binding assays performed using four PrPC molecules presenting small deletions on the putative Vn binding site, revealed that the PrPC deletion mutants 
105-112, 113-119, and 105-128 did not bind Vn, whereas mutants 51-90 and 120-125 exhibited binding capacity similar to that of the wild-type molecule (Fig. 2b). These data corroborate that the region comprising a.a. 105-119 of PrPC includes the binding site for Vn. As this domain is inserted into the N-terminal random coil of the PrPC molecule (Riek et al., 1996), these deletions are not expected to disturb PrPC secondary or tertiary structure. Since all the experiments herein were conducted with E. coli recombinant His6-PrPC, we were unable to evaluate possible roles of PrPC sugar residues in this interaction. However, as the binding site was mapped to a region lying outside the glycosylation site, it is unlikely that sugar residues on PrPC are essential for its interaction with Vn.
|
), which states that peptides presenting opposite hydropathy profiles can bind one another, we used HYDROLOG software to search for Vn sequences whose hydropathy profile was more than 70% complementary to the PrPC 105-119 peptide. Two human Vn domains presented this profile, one from residues 262-275 (Fig. 2c) and another from residues 309-322 (Fig. 2d). We performed competition assays using both peptides and, as controls, two peptides randomly chosen from within the Vn sequence (Vn161-174 and Vn289-302) along with a peptide containing the RGD sequence (VnRGD) (Ruoslahti and Pierschbacher, 1987). The peptides Vn309-322Hu (human Vn) and Vn307-320Mo (the equivalent peptide for mouse Vn) competed for PrPC-Vn interaction, thus identifying this domain as the putative binding site for PrPC (Fig. 2e).
We also performed binding assays using the 125I-Vn307-320Mo peptide and wild-type His6-PrPC or PrPC deletion mutants 105-128, 105-119 and 113-119. Vn307-320Mo peptide readily bound wild-type PrPC, whereas very low binding was observed with all of the deletion mutants (Fig. 2f). Using overlay assays we demonstrated that peptide Vn307-320Mo directly binds the PrPC peptide corresponding to a.a. 103-122, but not to a.a. 43-62 (Fig. 2f inset). These data are consistent with the presence of an interaction between the PrPC domain 105-119 and the Vn domain 307-320 (mouse Vn).
Supplementary material Fig. S2a shows a schematic representation of PrPC and Vn, along with their main ligands and respective binding sites (Lee et al., 2003; Schvartz et al., 1999). Although there is moderate sequence diversity in Vn across species, the PrPC binding domain is generally well-conserved (see supplementary material Fig. S3a). Chicken Vn, which has the greatest divergence relative to human Vn, bound PrPC in the same manner as human Vn (see supplementary material Fig. S3b). Furthermore, peptides from the PrPC-binding domain derived from chicken and mouse Vn sequences were also able to inhibit the PrPC interaction with human Vn (see supplementary material Fig. S3c), suggesting that the PrPC-Vn interaction is evolutionarily conserved.
PrPC and Vn expression in DRG from embryonic mice
Since Vn and PrPC have been detected early in development (Miele et al., 2003; Seiffert et al., 1995), we performed immunohistochemistry assays in E12.5 mice embryos. Immunohistochemistry assays were performed in ZrchI Prnp+/+ (Fig. 3a-l) and ZrchI Prnp0/0 (Fig. 3m-o) E12.5 mice embryos using anti-PrPC (Fig. 3a-c,g-i,m,n) and anti-Vn antibodies (Fig. 3d-f,j-l,o) in sagittal sections (Fig. 3a-f,m-o) and coronal sections (Fig. 3g-l). The antibodies used have been extensively tested in western blotting assays, and show specificity for PrPC (Zanata et al., 2002b) or Vn (data not shown). Control experiments with mouse or rabbit serum produced no immunolabeling (Fig. 3 insets in a and j, respectively).
|
), we observed strong PrPC immunoreactivity in the developing brain, spinal cord (Fig. 3g) and DRG (Fig. 3a,g). In DRG, although axons and forming axonal fibers were heavily labeled, nuclei were not labeled and neuronal bodies showed only weak labeling (Fig. 3b,c,h,i). PrPC expression could be detected only at E10, in the nervous system and in the heart (data not shown). As development proceeds, those organs still show high levels of immunoreactivity. Immunolabeling in other organs such as kidney, lungs and muscles could only be seen later (E18) in development (data not shown). Sections from Prnp0/0 mouse embryos did not bind the anti-PrPC antibody, proving the specificity of the immunohistochemistry reaction (Fig. 3m,n). Vn immunolabeling was more intense than PrPC, but presented a similar pattern in brain, spinal cord (Fig. 3j) and DRG (Fig. 3d,j). In the DRGs, expression was predominantly observed in growing axons and nerves (Fig. 3e,f,k,l). Vn starts to be expressed at E8, through the whole embryo. Similarly to PrPC, at E10 it presents high expression in the heart and nervous system. As development proceeds, those organs still present high levels of immunoreactivity, whereas immunolabeling can also be seen in other organs such as kidney, lung and muscles. Expression in the liver is very abundant (data not shown), since this organ secretes Vn to the blood. In PrPC-null mice, the pattern of Vn expression in DRG (Fig. 3o) as well as in other tissues (data not shown) was the same as that in wild-type mice. Immunofluorescent confocal images confirmed that PrPC and Vn strongly co-localized in DRG cells (Fig. 4a,b) and in growing nerves (Fig. 4c).
Vn binds PrPC in vivo
To verify whether Vn could bind PrPC in the cell surface, we labeled Vn with Alexa Fluor 568 (see supplementary material Fig. S1b). Dissociated DRG cells were treated with Alexa 568-Vn and subjected to PrPC fluorescent immunocytochemistry (Fig. 5a upper panel). We observed the co-localization of Vn and PrPC at the cell surface. Images of live SN-56 cells (Blusztajn et al., 1992; Hammond et al., 1990) transfected with green fluorescent protein (GFP)-PrPC and treated with Alexa 568-Vn showed co-localization of PrPC with Vn at the cell surface (Fig. 5a bottom panel).
|
60 kDa [the expected molecular mass for GFP-PrPC (Lee et al., 2001)] when anti-GFP (lane 3) or anti-PrPC antibodies (lane 6) were applied (Fig. 5b). When the same procedure was conducted with extracts from non-transfected cells (lanes 1 and 4) or with those from cells transfected with GFP alone (lanes 2 and 5) no binding to the Vn-Sepharose was observed. These results indicate that PrPC, but not GFP alone, binds to Vn. The levels of endogenous PrPC in HEK293 cells were too low for their association with Vn to be detected.
PrPC-Vn interaction mediates DRG axonal growth
We investigated the possible role in axonal growth of the interaction between PrPC and Vn, which are expressed in elongating DRG and medulla axons and are implicated in neuronal differentiation (Graner et al., 2000a; Graner et al., 2000b; Martinez-Morales et al., 1995; Martinez-Morales et al., 1997; Pons et al., 2001; Sales et al., 2002). Cultured DRG explants from E12.5 mice expressing PrPC (ZrchI Prnp+/+) (Fig. 6a inset) in the presence of Vn for 36 hours produced axonal growth (Fig. 6a). Vn peptide, Vn307-320Mo, corresponding to the PrPC-binding site of the mouse Vn molecule elicited the same effect as that triggered by the whole molecule (Fig. 6b), whereas Vn peptide Vn161-174 (Fig. 6d) had no effect, even at concentrations 20-fold higher than peptide Vn307-320Mo.
|
ZrchI Prnp0/0 mouse (Bueler et al., 1992) DRG neurons cultured on Vn presented neurite growth rates (Fig. 6g) similar to that of wild-type DRG neurons (Fig. 6a). Nevertheless, in opposition to what has been demonstrated for wild-type DRG, Vn peptide Vn307-320Mo (Fig. 6f), which mimics the PrPC binding site, was unable to induce axon growth in Prnp0/0 DRG neurons. Vn peptide Vn161-174 also had no effect on Prnp0/0 DRG neurons (Fig. 6e). Treatment of Vn-stimulated DRG with rabbit anti-PrPC antibody blocked axonal growth in cultures from wild-type animals but not in those from PrPC-null mice (Fig. 6h). Data representing the average axon length from each treatment are summarized in Fig. 6h.
To rule out the possibility of spurious results because of genetic background, we conducted our key experiments in both ZrchI (Bueler et al., 1992) and Npu (Manson et al., 1994) PrPC-null mice. DRG from Npu Prnp–/– mice and their wild-type controls plated for 24 hours in the presence of Vn presented similar axonal growth (Fig. 6i). On the other hand, when plated with Vn307-320Mo, only Npu Prnp+/+ DRG grew axons and no effect was observed with either knockout or wild-type cells when an irrelevant Vn peptide (Vn161-174) was used (Fig. 6i). Inclusion of an antibody against PrPC peptide 106-126 (Chiarini et al., 2002) (Fig. 6i) completely blocked axonal growth in cultures from wild-type animals, whereas non-immune purified IgG had no effect (Fig. 6i). We also carried out dissociated DRG cell cultures in the presence of Vn to measure the percentage of cells with neurites, and observed that the positive responsiveness to Vn was the same for ZrchI Prnp+/+ and Prnp0/0 neurons (Fig. 6j).
The experiments using anti-PrPC antibodies or the Vn307-320 peptide demonstrated that axonal growth can be supported specifically by the PrPC-Vn interaction. Nonetheless, the whole Vn molecule induces the same axonal outgrowth pattern in wild-type and PrPC-null neurons. The similarity between the effect of Vn in wild-type and PrPC-null DRG suggests that there may be another Vn receptor that can compensate for the PrPC deficiency.
In fact, flow cytometry analysis shows that Alexa 568-Vn is able to bind equally to the surface of PrPC-null and wild-type cells (see supplementary material Fig. S4a). Indicating that in the absence of PrPC other Vn receptors are present at the cell surface.
Integrin participation in Vn-induced axonal growth is enhanced in the absence of PrPC
The obvious targets for the putative compensatory mechanism suggested by the above data are integrins, the classical Vn receptors that interact with this molecule through the Vn-RGD peptide. We performed functional assays using the RGD peptide, which has an advantage over antibodies in that it can simultaneously trigger or halt several integrin dimers (Isahara and Yamamoto, 1995; Monier-Gavelle and Duband, 1997). The RGD peptide can be used for inhibition or stimulation of the neuritogenesis depending on its presentation. In solution, the RGD peptide does not support cell adhesion and thus can be used to perform competition assays (Pierschbacher and Ruoslahti, 1987). Conversely, when coupled to BSA, the peptide adheres to the coverslip and supports cell adhesion (Danilov and Juliano, 1989).
As shown in Fig. 7a, the VnRGD peptide at a concentration of 8 µM inhibited Vn-stimulated axonal growth of ZrchI Prnp0/0 DRG neurons. The impairment reached poly-L-lysine growth levels (about 50% of that observed in Vn). Vn-stimulated axonal growth in Prnp+/+ neurons was impaired only when the VnRGD peptide reached a concentration of 12 µM. Additionally, we measured the axonal growth induced by VnRGD-BSA and observed that ZrchI Prnp0/0 DRG are more responsive than Prnp+/+ DRG (Fig. 7b). Prnp0/0 DRG neurons extended axons in 0.5 nmol of adsorbed peptide, whereas Prnp+/+ DRG neurons did not exhibit any axonal growth, even in the presence of a 20-fold higher concentration of the peptide. Thus, Prnp0/0 DRG cells were more responsive than the Prnp+/+ cells to the RGD peptide. To further confirm these results, we plated Npu Prnp–/– and Prnp+/+ DRG over adsorbed RGD-BSA. We observed that Npu Prnp–/– DRG neurons were also more responsive to RGD-BSA than Npu Prnp+/+ DRG neurons (Fig. 7c). Thus, we can be confident that the above findings were not due to any spurious effect present only in ZrchI animals. These data demonstrate that PrPC-ablated DRG neurons have a greater dependence upon integrins for axonal outgrowth than do their respective wild-type cells.
|
To test whether the activation state of Vn-binding integrins was altered, we measured the level of active v3 integrin in adhering ZrchI Prnp0/0 DRG cells through an immunofluorescence assay with the ligand-mimetic antibody WOW-1 (Pampori et al., 1999). ZrchI Prnp0/0 DRG neurons showed a 30% higher level of v3 activation than Prnp+/+ DRG neurons (Fig. 7d). We tested whether Npu Prnp–/– neurons presented a similar increase in integrin activity using the commercially available anti-ligand-induced binding sites (LIBS) antibody AP5 (Faccio et al., 2002). Dissociated Npu Prnp–/– DRG neurons showed greater 3 activity than Prnp+/+ cells (Fig. 7e). Accordingly, ZrchI Prnp0/0 primary mouse embryonic fibroblasts (PMEFs) also showed 30% more integrin v subunit expression than wild-type PMEFs (see supplementary material Fig. S4b). Together, these findings indicate that a higher integrin activity, particularly that from v3 integrin, is present in DRG after ablation of PrPC.
|
Discussion |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
; Graner et al., 2000b) and glycosaminoglycans (Gonzalez-Iglesias et al., 2002) have been of particular interest. These previous findings suggested that PrPC might act as a wide-range cell surface receptor, capable of binding numerous ECM proteins. However, the present findings indicate that PrPC interaction with ECM proteins is not a broad-spectrum phenomenon, but rather that PrPC specifically binds Vn with high affinity, and does not readily bind Fn or type IV collagen.
We proposed that PrPC participates in a multi-protein complex, the biological effects of which would be dependent both on the affinity and accessibility of each member of the complex (Martins and Brentani, 2002). The present competition experiments demonstrated that PrPC-Vn interaction in vitro has a higher affinity, Kd 10–8 M, than that between PrPC and STI1, Kd 10–7 M (Zanata et al., 2002b). Furthermore, the Vn binding domain in PrPC in residues 105-119 (Fig. 2a) overlaps with the STI1-binding domain (Zanata et al., 2002b), indicating that PrPC interactions with Vn or STI1 are mutually exclusive, with the first being more favorable and conditioning the latter to local protein availability and levels. The ability of PrPC-STI1 to provide neuroprotection against programmed cell death (Chiarini et al., 2002) and neuritogenesis in hippocampal neurons (Lopes et al., 2005), whereas PrPC-Vn engagement promotes axonal growth in DRG, provide prime examples of how cell fate can be directly influenced by the molecular environment of the cell.
PrPC domain 105-128, which contains the Vn and STI1 binding sites, is highly conserved across species (Gabriel et al., 1992), and is indeed identical in mice and humans. This domain also includes proteolytic sites (Jimenez-Huete et al., 1998) and hereditary prion disease-associated mutations (Mastrianni and Roos, 2000), indicating that this region and probably its interaction with Vn and STI1 have an important role in vivo. For example, it was recently reported that this domain is linked to the toxicity of PrPC accumulated in the cytosol through binding to Bcl-2 (Rambold et al., 2006).
Although PrPC-Ln and PrPC-Vn interactions present a similar Kd, (Graner et al., 2000a), we demonstrated that lower concentrations of Vn than Ln are needed to disrupt the PrPC-Vn interaction. Since the Ln-binding site is located in residues 173-183 of PrPC (Coitinho et al., 2006), the Ln 1 peptide (the PrPC-binding site in Ln) does not disrupt the PrPC-Vn complex (Fig. 1d). Thus, it is plausible that in vitro, the Ln molecule may interact with PrPC, and thereby disturb PrPC-Vn binding by steric hindrance. On the other hand, steric hindrance may not occur in vivo because of the ECM organization or to the presence of biologically active proteolytic fragments from the Ln molecule (Chen, Z. et al., 2003). Indeed, the signals triggered by each of these PrPC ligands may have cooperative roles in some biological events. In developing cerebellar granule cells, the presence of Ln induces proliferation, whereas during migration these cells find Vn and differentiate (Pons et al., 2001). Therefore, we believe that PrPC has pleiotropic functions that depend upon its cellular expression as well as the greater contextual cellular milieu.
We exploited the complementary hydropathy principle (Brentani, 1988) to map a.a. 309-322 as the PrPC-binding site in the human Vn molecule. To date, more than 40 protein-protein interactions have been shown to comply with this principle (Heal et al., 2002), including a Vn and fibrinogen interaction with the integrin IIb3 (Gartner et al., 1991) and the PrPC-STI1 interaction (Martins et al., 1997; Zanata et al., 2002b). This region (residues 309-322) of human Vn has never been shown to bind other proteins (Schvartz et al., 1999). According to the three-dimensional theoretical model of Vn (Xu et al., 2001), this mainly hydrophobic peptide is partially buried. Conversely, it should be considered that the threading algorithm (Xu and Xu, 2000) used to create this model makes use of an energy function that penalizes the exposure of hydrophobic side chains, whereas it is known that protein-binding sites are generally hydrophobic (Tsai et al., 1997). Since it was not considered that this particular region could be a binding site, protein interactions were not taken into consideration in these calculations (Xu et al., 2001).
Vn is composed of multiple domains known to bind distinct proteins (see supplementary material Fig. S2a). The first 44 amino acids comprise a somatomedin-like domain, which harbors plasminogen activator inhibitor-1. This somatomedin-like domain is followed by the RGD peptide domain (residues 53-64), which is responsible for integrin binding, and an acidic stretch involved in binding of the thrombin-antithrombin III complex. Two binding sites for collagen have been described, one adjacent to the RGD site and other adjacent to the heparin-binding site. The Vn is composed of six hemopexin domains (a.a. 132-459). The C-terminal part of the molecule harbors a plasminogen (a.a. 332-348), a heparin (a.a. 348-346) and a glycosaminoglycan-binding site (a.a. 348-361). A subregion of this domain (a.a. 348-370) has also been implicated in plasminogen activator inhibitor-1 binding (Preissner, 1991; Schvartz et al., 1999). Thus although the PrPC-binding site in Vn has not been assigned to bind other proteins, it is close to the heparin-binding domain. Heparin is efficient in competing PrPC-Vn interaction (see supplementary material Fig. S2a), since it is able to bind both PrPC (Pan et al., 2002) and Vn (Francois et al., 1999). Thus, as proposed above, PrPC pleiotropic functions are dependent on the greater contextual cellular milieu.
There are many Vn receptors found on the cell surface, such as integrins IIb3, v1, v3 and v5 (Felding-Habermann and Cheresh, 1993). Each one of these binds to the Vn RGD domain with a different affinity and induces specific signals within cells (Takagi et al., 2002). Our demonstration herein that RGD peptide did not compete for the PrPC-Vn interaction indicates that Vn binds PrPC at a domain that is distinct from the integrin-binding domain.
Since both PrPC and integrins participate in axonal growth, it was surprising to observe a complete inhibition of Vn-stimulated axonal growth in the presence of anti-PrPC antibody. This finding suggests that PrPC occupies a part of the macromolecular complex such that its inactivation by antibodies may disturb the whole complex. The ability of Vn peptide containing the PrPC-binding site to reproduce the biological effects of the whole Vn in wild-type but not in PrPC-null DRG, further suggests that PrPC is an exclusive ligand for the Vn 307-320 domain. The fact that Vn307-320Mo (Fig. 7) and VnRGD (Danilov and Juliano, 1989) are both able to substitute for the whole Vn molecule in the axonal growth phenomenon is consistent with the hypothesis that PrPC and integrins act through the same signal transduction pathway. Notably, the data showing that VnRGD is able to inhibit axonal growth in wild-type neurons highlights the importance of the integrins in this process.
In spite of the important functions for PrPC described over the past few years (Chiarini et al., 2002; Lopes et al., 2005; Mouillet-Richard et al., 2000; Rambold et al., 2006; Steele et al., 2006), PrPC-null mice have only minor defects (Bueler et al., 1992). One explanation for this near-normal phenotype is that PrPC ablation may be compensated by proteins with redundant functions (Bueler et al., 1992). This seems to be the case here, because the whole Vn molecule induced DRG axonal growth in both wild-type and PrPC-null neurons. On the other hand, PrPC-null (both ZrchI and Npu) DRG axonal growth is more sensitive to RGD than the wild-type, indicating that the cellular signaling involved in this phenomenon is more dependent upon integrins in neurons from the knockout animals. Additionally, we observed 30% more activated v3 in DRG from PrPC-null mice than in their wild-type counterparts. This increase in activated integrins may represent a compensatory mechanism developed by the PrPC-null animals, where the use of proteins already involved in this specific phenotype can prevent malformation of the nervous system.
Numerous examples of compensatory mechanisms have been reported (Kitami and Nadeau, 2002; Schwarz et al., 2002) and promiscuous cell signaling transduction pathways are targets for molecular redundancy (Xian et al., 2001). Indeed, PrPC-null mice have been reported to have hyper-activation of extracellular signal-related kinase (ERK) (Chiarini et al., 2002; Lopes et al., 2005; Brown et al., 2002). PrPC (Zanata et al., 2002b) and integrins (Roberts et al., 2003) are upstream MAPK effectors, which means that at least in this situation, an integrin could substitute for PrPC signaling.
GPI-anchored proteins, such as PrPC, can perform signaling roles, integrating the ECM with the cytoskeleton. For example the GPI-anchored raft-associated protein urokinase type plasminogen activator receptor (uPAR) performs complex signaling involved in cell adhesion, proliferation and migration in response to several ligands including Vn (Blasi and Carmeliet, 2002). To modulate internal cell signaling, GPI-anchored proteins must interact with transmembrane adaptors, such as integrins, G-protein-coupled receptors and caveolins (Blasi and Carmeliet, 2002). Despite the involvement of uPAR in cell signaling and diverse biological functions, uPAR-null mice have an apparently normal phenotype (Bugge et al., 1995). Thus, a GPI-anchored protein can be associated with critical regulatory cell tasks and the lack of a severe phenotype in knockout mice cannot demonstrate that the protein is not normally involved in important phenotypes.
In summary, the characterization of PrPC as a ligand for Vn and its involvement in axonal growth allowed us to demonstrate the relevance of PrPC in the development of the peripheral nervous system (PNS). Additionally, compensatory mechanisms occurring during embryogenesis are turned on when PrPC is ablated. Thus, at least for this event, our findings indicate that redundancy for PrPC interactions resides within the integrin pathway. The importance of PrPC in the PNS has been increasingly recognized because PNS is a target for PrPC conversion to PrPSc, and thus is critically involved in the prion neuroinvasion (Glatzel et al., 2004). Providing further support for PNS involvement in prion neuroinvasion, are the observations that PrPC is retrogradely transported in peripheral nerves (Moya et al., 2004) and that prion accumulation occurs in DRG and in autonomic ganglia isolated from patients with Creutzfeldt-Jakob and Gerstmann-Sträussler-Scheinker (Haik et al., 2003; Ishida et al., 2005; Lee et al., 2005). The present findings together with these previous observations indicate that the physiological functions of PrPC in peripheral nerves warrant further investigation.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
; Yatohgo et al., 1988) and mouse His6-PrPC was cloned (Zahn et al., 1997) and expressed (Zanata et al., 2002b). Type IV collagen and albumin are from Sigma. Four PrPC deletion mutants were constructed using wild-type cDNA by sequential PCR amplification (Ausubel et al., 1993); cloned into pRSET A vector (Invitrogen) and expressed and purified as the wild-type His6-PrPC. Mutant His6-PrPC molecules are soluble, sensitive to proteinase K and posses nearly the same -helix and -sheet content as the wild-type protein. The following internal primers were used for sequential PCR amplification: 51-900R-TACCCCCTCCTGGGTAACGGTTGCCTCC; F-GAGGGATCCAAAAAGCGGCCAAAG; 105-112R-CCAGCTGCCGCAGCCCCTGGTTGGCTGG; F-CCCAGCAAACCAGGGGCTGCGGCAGCTGG; 113-119R-CCCCATTACTGCCACATGCTTGAGGTTG; F-AAGCATGTGGCAGTAGTGGGGGGCCTT; 120-125R-CAGCATGTAGCCTGCCCCAGCTGCCGC; F-GCAGCTGGGGCAGGCTACATGCGGGAGC; 105-128F-GGAACAAGCCCAGCAAACCACTGGGGAGCGCCATGACGG;R-GTCCATGGCGCTCCCCAGTGGTTTGCTGGGCTTTGTTCC. The following external primers were used for all mutants: F-AGAGAATTCTCAGCTGGATCTTCTCCCGTC; R-GAGGGATCCAAAAAGCGGCCAAAG.
Peptides
Twenty peptides covering the whole mouse PrPC (23-231) (Zanata et al., 2002b) were used. STI1 peptide (230-ELGNDAYKKKDFDKAL-245) (Zanata et al., 2002b), laminin -1 peptide (1575-RNIAEIIKDI-1584) (Graner et al., 2000a) and six Vn peptides: VnRGD KPQVTRGDVFTMPE; Vn161-174 AEEELCSGKPFDAF; Vn262-275 AHSYSGRERVYFFK; Vn289-302 SQEECEGSSLSAVF; Vn309-322Hu QRDSWEDIFELLFW and Vn307-320Mo QRDSWENIFELLFW were synthesized by Neosystem (Strasbourg, France). The subscript numbers indicate amino acid position in the Vn molecule; the first five peptides follow the human (Hu) Vn sequence and the last one follows the mouse (Mo) Vn sequence.
Antibodies
Anti-PrPC used for western blotting, immunohistochemistry and immunofluorescence reactions is a polyclonal antibody obtained by His6-PrPC immunization in PrPC-null mice. Rabbit polyclonal antibodies obtained by His6-PrPC (Bethyl) or PrPC peptide 106-126 (Neosystem) immunization were used in DRG axon growth assays (Chiarini et al., 2002; Zanata et al., 2002b).
The WOW-1 antibody, kindly provided by Prof. Sanford Shattil (University of California), recognizes the active form of v3; the heavy chain hypervariable region of an antibody against activated IIb3 was replaced with a single v integrin-binding domain (from an adenovirus RGD-rich protein involved in its internalization mediated by v) (Felding-Habermann et al., 2001; Pampori et al., 1999). The AP5 antibody (GTI Diagnostics) is an anti-LIBS antibody that recognizes the 3 N-terminus and is regulated by cation binding at a site distinct from the LIBS (Honda et al., 1995). At normal extracellular Ca2+ levels, AP5 binds to 3 only when the integrin is in the `activated' conformation (Faccio et al., 2002). Anti-GFP was from Becton Dickinson and anti-Vn was obtained by rabbit immunization with purified Vn.
Overlay assay
The indicated amounts of Ln, Fn, Vn, type IV collagen and BSA were adsorbed onto nitrocellulose membranes. Blocking was performed in 5% milk in TBST (TBS 0.05% pH 7.4 Tween-20) for 2 hours at room temperature and membranes were then washed. His6-PrPC (7 µg) was labeled with 0.5 mCi Na125I (Amersham Biosciences) using one iodobead (Pierce). The labeled protein was incubated with the membrane for 16 hours at 4°C. After washing, an X-ray film (Hyperfilm, Amersham Biosciences) was exposed to the membrane.
3 µg His6-PrPC or PrPC peptides 43-62 or 103-122 was adsorbed onto nitrocellulose membranes. After blocking, Vn307-320Mo was labeled with biotin using an EZ-Link Sulfo-NHS-LC-Biotinylation Kit (Pierce) and incubated with the membrane for 16 hours at 4°C, followed by Streptavidin-HRP (Sigma) for 1 hour at room temperature.
Binding assays
Binding experiments were conducted as previously described (Graner et al., 2000a; Martins et al., 1997). Briefly, His6-PrPC or His6-PrPC deletion mutants (2 µg) were adsorbed in polystyrene wells and blocked with 2% BSA. Wells were incubated for 3 hours at 37°C with Na125I-labeled Vn or Vn307-320Mo coupled to BSA. The wells were washed, and radioactivity was measured in a gamma counter (Mini gamma counter, LKB-Wallac). Data were analyzed using the Scatchard Method (Scatchard, 1949).
Competition assays
Unlabeled PrPC peptides (32 µM) were pre-incubated with 38 nM of 125I-Vn for 2 hours at room temperature. Peptides and 125I-Vn were added to the His6-PrPC coated wells and incubated for 3 hours at 37°C. The wells were washed and the radioactivity was measured. Vn, STI1, Ln whole proteins or STI1 and Ln peptides were pre-incubated with coated His6-PrPC for 2 hours at room temperature followed by 3 hours at 37°C with 125I-Vn. After washing, the radioactivity was measured.
Cell transfection and pull-down assay
HEK293 cells were transfected with pEGFP-C1 (Clontech) or pEGFP-STI1 (Zanata et al., 2002b) by calcium phosphate co-precipitation as previously described (Puchel et al., 1995). After 48 hours in culture, transfected cells were lysed in NP40 (0.5% in PBS). Cell extracts were pre-cleared by treatment with 30 µl of inactive CNBr-Sepharose (reactive groups previously blocked) for 1 hour at 4°C.
Vn was covalently coupled to CNBr-Sepharose 4B (Amersham) according to the manufacturer's instructions, and 30 µl Vn-CNBr-Sepharose was incubated in pre-cleared cell extracts for 16 hours at 4°C. After washing with 0.5% NP-40 in PBS, bound proteins were eluted with Laemmli buffer and analyzed by western blotting using mouse anti-PrPC (1:1000) or anti-GFP (1:3000) followed by anti-mouse HRP.
Alexa Fluor 568 Vn labeling and co-localization assay
Vn labeling was performed using an Alexa Fluor 568 labeling kit (Molecular Probes). SN-56 cells, a mouse cholinergic septal neuronal cell line (Blusztajn et al., 1992; Hammond et al., 1990), were transfected with GFP-PrPC using lipofectamine (Invitrogen) as described previously (Lee et al., 2001). Transfected and differentiated (1 mM cAMP for 1 day) SN-56 cells, were treated with 4 µg Alexa 568-Vn for 1 hour at 4°C. After several washes with PBS, images of live cells were acquired using a Bio-Rad Radiance 2100 laser-scanning confocal system coupled to a Nikon microscope (TE2000-U). Dissociated DRG cells were incubated with 4 µg labelled Vn for 1 hour at 4°C, fixed, submitted to immunofluorescence using mouse anti-PrPC followed by anti-mouse Alexa Fluor 488, and images were acquired as for SN-56 cells.
Animals
The `Principles of laboratory animal care' (NIH publication 85-23, 1996) were strictly followed in all experiments. ZrchI Prnp0/0 were provided by Dr Charles Weissmann (Scripps Florida, FL) (Bueler et al., 1992). ZrchI Prnp+/+ mice were generated by crossing F1 descendants from 129/SV and C57BL/6J matings. Npu Prnp–/– (Manson et al., 1994) were provided by Bruce Chesebro and Richard Race (Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, MT). These animals were backcrossed to C57BL/10 mice for at least eight generations. Heterozygous animals were mated and homozygous F1 descendants from the same litter were crossed to generate Npu Prnp–/– or Npu Prnp+/+ embryos. All of the adult animals used to generate Npu Prnp–/– and Npu Prnp+/+ embryos were genotyped by PCR (Manson et al., 1994).
DRG explants
DRG from E12.5 mice (Bueler et al., 1992) were dissected and cultured in poly-L-lysine coated glass coverslip in neurobasal medium (Invitrogen) with 2 mM glutamine, 100 IU penicillin, 100 µg/ml streptomicyn, B-27 (Invitrogen) and 50 ng/ml nerve growth factor (Sigma). In Fig. 7b,d, DRGs were plated in VnRGD-BSA adsorbed coverslips. Treatments were performed immediately after plating and ganglia were cultured for 24 (Fig. 6i,j, Fig. 7a-c) or 36 hours (Fig. 6a-h), fixed in 4% paraformaldehyde/0.12 M sucrose and stained with haematoxylin. The neurite length was measured as the distance from the edge of the DRG to the tip of the three longest neurites, and the mean value was used as the neurite length for each DRG (Zanata et al., 2002a). At least 12 ganglia from three independent experiments were considered for each individual data point. Cultures were dissociated by enzymatic digestion of the dissected ganglia for 30 minutes with 1% trypsin in neurobasal medium. After mechanical dissociation, 5x104 cells per 13 mm2 well were plated in the presence of Vn for 6 hours. Cells were fixed and stained as described above and the percentage of cells presenting a neurite longer than one cell body was calculated.
Immunohistochemistry
DRG explants grown in the presence of Vn were fixed and incubated for 4 hours at 4°C with a mouse polyclonal antibody anti-PrPC, 1:250 (Chiarini et al., 2002) followed by Alexa Fluor 568 anti-mouse IgG (Molecular Probes, Eugene, OR), for 40 minutes at room temperature. Ganglia were viewed with an Olympus IX70 microscope equipped with epifluorescence.
Formalin-fixed E12.5 mice embryos were embedded in paraffin and sections submitted to immunohistochemistry as previously described (Lopes et al., 2005) with mouse polyclonal anti-PrPC antibody (1:1000) (Chiarini et al., 2002) or with rabbit polyclonal anti-Vn antibody (1:250).
Confocal immunofluorescence
E12.5 mouse embryos were immediately frozen and 3-µm-thick cryostat sections submitted to immunofluorescence as previously described (Lopes et al., 2005) with a mouse polyclonal anti-PrPC antibody (1:250) (Chiarini et al., 2002) and rabbit anti-Vn serum (1:100).
v3 activity assays
Dissociated DRG cells were plated on poly-L-lysine-coated coverslips and incubated for 24 hours, fixed, and blocked as described above. Immunofluorescence reaction with WOW-1 antibody (1:4) (Pampori et al., 1999) was undertaken for 16 hours at room temperature followed by Alexa Fluor 568 anti-mouse IgG (1:3000) (Molecular Probes). Immunofluorescence with AP5 antibody (50 µg/ml) was performed in PBS with 3 mM Ca2+ for 16 hours at room temperature followed by Alexa Fluor 568 anti-mouse IgG (1:3000). Images were acquired with an Olympus IX70 microscope equipped with epifluorescence. To acquire the images, the digital camera (Olympus DP70) exposure was set so that no fluorescence could be observed in cells incubated with only secondary antibody. At least five fields of each coverslip were imaged and the fluorescence of each cell was measured with the Image-Pro Plus 4.1 (Media Cybernetics). At least 100 cells per coverslip were considered.
Statistical analysis
The mean values of at least three independent datasets are shown in the figures; the error bars represent s.d. Fit to a normal distribution was evaluated with the Kolmogorov-Smirnov test. The homogeneity of variances was assessed using Levene's test. The comparison of means for two independent samples was performed using Student's t-test or Mann-Whitney's test. When comparing more than two groups, an ANOVA or Kruskal Wallis test was used and a Tukey-HSD test was used for multiple comparisons.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. D., Smith, J. A. and Struhl, K. (1993). Constructing recombinant DNA molecules by the polymerase chain reaction. In Current Protocols in Molecular Biology (1st edn), p. 3.17.1. New York: Wiley Interscience.
Blasi, F. and Carmeliet, P. (2002). uPAR: a versatile signalling orchestrator. Nat. Rev. Mol. Cell Biol. 3, 932-943.[CrossRef][Medline]
Blusztajn, J. K., Venturini, A., Jackson, D. A., Lee, H. J. and Wainer, B. H. (1992). Acetylcholine synthesis and release is enhanced by dibutyryl cyclic AMP in a neuronal cell line derived from mouse septum. J. Neurosci. 12, 793-799.[Abstract]
Brentani, R. R. (1988). Biological implications of complementary hydropathy of amino acids. J. Theor. Biol. 135, 495-499.[CrossRef][Medline]
Brown, D. R. and Besinger, A. (1998). Prion protein expression and superoxide dismutase activity. Biochem. J. 334, 423-429.[Medline]
Brown, D. R., Qin, K., Herms, J. W., Madlung, A., Manson, J., Strome, R., Fraser, P. E., Kruck, T., von Bohlen, A., Schulz-Schaeffer, W. et al. (1997). The cellular prion protein binds copper in vivo. Nature 390, 684-687.[Medline]
Brown, D. R., Nicholas, R. S. and Canevari, L. (2002). Lack of prion protein expression results in a neuronal phenotype sensitive to stress. J. Neurosci. Res. 67, 211-224.[CrossRef][Medline]
Bueler, H., Fischer, M., Lang, Y., Bluethmann, H., Lipp, H. P., DeArmond, S. J., Prusiner, S. B., Aguet, M. and Weissmann, C. (1992). Normal development and behaviour of mice lacking the neuronal cell-surface PrP protein. Nature 356, 577-582.[CrossRef][Medline]
Bugge, T. H., Suh, T. T., Flick, M. J., Daugherty, C. C., Romer, J., Solberg, H., Ellis, V., Dano, K. and Degen, J. L. (1995). The receptor for urokinase-type plasminogen activator is not essential for mouse development or fertility. J. Biol. Chem. 270, 16886-16894.
Chen, S., Mange, A., Dong, L., Lehmann, S. and Schachner, M. (2003). Prion protein as trans-interacting partner for neurons is involved in neurite outgrowth and neuronal survival. Mol. Cell. Neurosci. 22, 227-233.[CrossRef][Medline]
Chen, Z. L., Indyk, J. A. and Strickland, S. (2003). The hippocampal laminin matrix is dynamic and critical for neuronal survival. Mol. Biol. Cell 14, 2665-2676.
Chiarini, L. B., Freitas, A. R., Zanata, S. M., Brentani, R. R., Martins, V. R. and Linden, R. (2002). Cellular prion protein transduces neuroprotective signals. EMBO J. 21, 3317-3326.[CrossRef][Medline]
Coitinho, A. S., Freitas, A. R., Lopes, M. H., Hajj, G. N., Roesler, R., Walz, R., Rossato, J. I., Cammarota, M., Izquierdo, I., Martins, V. R. et al. (2006). The interaction between prion protein and laminin modulates memory consolidation. Eur. J. Neurosci. 24, 3255-3264.[CrossRef][Medline]
Cordeiro, Y., Kraineva, J., Ravindra, R., Lima, L. M., Gomes, M. P., Foguel, D., Winter, R. and Silva, J. L. (2004a). Hydration and packing effects on prion folding and beta-sheet conversion. High pressure spectroscopy and pressure perturbation calorimetry studies. J. Biol. Chem. 279, 32354-32359.
Cordeiro, Y., Lima, L. M., Gomes, M. P., Foguel, D. and Silva, J. L. (2004b). Modulation of prion protein oligomerization, aggregation, and beta-sheet conversion by 4,4'-dianilino-1,1'-binaphthyl-5,5'-sulfonate (bis-ANS). J. Biol. Chem. 279, 5346-5352.
Danilov, Y. N. and Juliano, R. L. (1989). (Arg-Gly-Asp)n-albumin conjugates as a model substratum for integrin-mediated cell adhesion. Exp. Cell Res. 182, 186-196.[CrossRef][Medline]
Easley, C. A., Faison, M. O., Kirsch, T. L., Lee, J. A., Seward, M. E. and Tombes, R. M. (2006). Laminin activates CaMK-II to stabilize nascent embryonic axons. Brain Res. 1092, 59-68.[CrossRef][Medline]
Engvall, E. and Ruoslahti, E. (1977). Binding of soluble form of fibroblast surface protein, fibronectin, to collagen. Int. J. Cancer 20, 1-5.[Medline]
Ettaiche, M., Pichot, R., Vincent, J. P. and Chabry, J. (2000). In vivo cytotoxicity of the prion protein fragment 106-126. J. Biol. Chem. 275, 36487-36490.
Faccio, R., Grano, M., Colucci, S., Villa, A., Giannelli, G., Quaranta, V. and Zallone, A. (2002). Localization and possible role of two different alpha v beta 3 integrin conformations in resting and resorbing osteoclasts. J. Cell Sci. 115, 2919-2929.
