Address correspondence and reprint requests to Prof. Edon Melloni, University of Genoa, DIMES, Biochemistry Section, Viale Benedetto XV, 1-16132 Genoa. E-mail: firstname.lastname@example.org
In resting human neuronal cells, nitric oxide synthase (nNOS) is present in its native 160 kDa form in a quiescent state predominantly co-localized on the plasma membrane, via its PDZ (Psd-95/Discs-large/Zona Occludens) domain, with NMDA receptor (NMDA-R) and in tight association with heat shock protein 90 (HSP90). Following exposure of the cells to Ca2+-ionophore or to NMDA, nNOS undergoes proteolytic removal of the PDZ domain being converted into a fully active 130 kDa form. The newly generated nNO synthase form dissociates from NMDA-R and extensively diffuses into the cytosol in direct correlation with NO production. Intracellular redistribution and activation of nNOS are completely prevented in cells preloaded with calpain inhibitor-1, indicating that these processes are triggered by a concomitant activation of calpain. The role of calpain has been confirmed by immunoprecipitation experiments revealing that also μ-calpain is specifically recruited into the NMDA-R-nNOS-HSP90 complex following calcium loading. Thus, the formation of clusters containing HSP90, μ-calpain, nNOS and NMDA-R represents the limiting step for the operation of the mechanism that links an efficient synthesis of NO to a local increase in Ca2+ influx.
Although these multiprotein complexes in subregions of different cell types have been identified, the relationship between their formation and function remains still to be completely defined (d’Anglemont de Tassigny et al. 2009). Moreover, it has not been clarified if the activation of the synthase occurs in its still associated form or if, similarly to other NOS isoforms, activation occurs in the soluble cell fraction (Arundine et al. 2003).
The neuronal synthase form contains an additional calpain site of cleavage located in the N-terminal region which causes the removal of the PDZ domain without affecting the catalytic site of the enzyme (Araujo and Carvalho 2005; Averna et al. 2007).
Furthermore, degradation and inactivation of nNOS by calpain have been shown to occur in excitotoxic conditions accompanied by progressive neurodegeneration and cell death (Araujo and Carvalho 2005; Cao et al. 2005; Volbracht et al. 2005). As a paradox, in acute neurodegenerative disorders, characterized by a profound alteration in Ca2+ homeostasis, down-regulation of nNOS has been proposed as a neuroprotective mechanism that prevents overproduction of NO highly toxic for the cell. However, in spite of all these information, it still remains to be proved if proteolytic degradation of nNOS is of physiological relevance in the regulation of its activity.
In this respect, we have recently shown that in in vivo conditions, in which activation of nNOS is concomitant to that of calpain, the synthase is protected from inactivation by interaction with HSP90 (Averna et al. 2007, 2008a,b). This protective mechanism was due to the formation of a nNOS-HSP90 binary complex and to the recruitment in the complex of calpain in the presence of Ca2+. In this form both nNOS and HSP90 become resistant to degradation and inactivation by the co-associated protease (Averna et al. 2007).
Interestingly, it was also observed that in the heterotrimeric complex nNOS retained full catalytic activity although having undergone a calpain-mediated proteolytic conversion into a fully active 130 kDa form resulting from the removal of the PDZ domain (Averna et al. 2007). These findings explained how activation of calpain, concomitantly occurring to that of nNOS, did not affect NO production. Also in the case of eNOS the synthase was preserved from digestion by the concomitant activated calpain through the formation of an eNOS-HSP90-calpain heterocomplex (Averna et al. 2008a).
Altogether these findings indicated a new function of HSP90 and proved that the levels of this chaperone were directly related to susceptibility of NOS to proteolytic degradation.
The present study was designed to explore in more details the mechanism leading to intracellular activation of nNOS. We found that upon intracellular Ca2+ increase translocation and activation of calpain at the membrane level induced the removal of the PDZ domain from the 160 kDa native NOS producing a cytosolic fully active 130 kDa form in direct correlation with NO production. In cells preloaded with a synthetic calpain inhibitor all these events were completely prevented.
This conservative proteolytic conversion of nNOS was shown to occur only in the presence of HSP90 which preserved the 130 kDa nNOS form from further calpain digestion. Altogether, these findings are indicating that, in neuronal cells, the limiting step of nNOS activation is represented by dissociation and translocation of the active synthase form into the cytosol in concomitance with NO production.
Leupeptin, aprotinin, calmodulin, 4,5-diamino-fluorescein diacetate (DAF-2DA), Ca2+-ionophore A23187, N-nitro-l-arginine methyl ester (L-NAME) and calpain inhibitor-1 were purchased from Sigma-Aldrich Milan, Italy. 4-(2-aminoethyl) benzenesulfonylfluoride was obtained from Calbiochem, Canada, USA. ECL® Detection System, NADPH, FAD, FMN, tetrahydrobiopterin, L-[14C]arginine (25nCi; specific activity 308 Ci·mol−1) were obtained from GE Healthcare (Milan, Italy). Dowex 50W8 Na+ form resin was obtained from Bio-Rad (Segrate, Milan, Italy). NMDA was purchased from Tocris Biosciences (Bristol, UK). Human erythrocyte calpain was purified and quantified as indicated previously (Michetti et al. 1996). Rat brain HSP90 was purified as reported previously (Averna et al. 2007).
Monoclonal mouse IgG1 nNOS and HSP90 antibodies were purchased from BD Transduction Laboratories, Milan, Italy; rabbit immunoaffinity purified IgG anti-NR1 pan was obtained from Cell Signaling Technology, Inc.3 TraskLane Danvers, MA, USA; rabbit affinity isolated antibody anti-μ-calpain (domain IV) and monoclonal mouse IgG1 m-calpain (domain III/IV) clone 107-82 were purchased from Sigma-Aldrich.
SK-N-BE human neuronal cells were kindly provided by Claudia Cantoni (Department of Experimental Medicine (DIMES), Section of General Pathology, University of Genoa, Viale Benedetto XV, 1, 16132 Genoa, Italy) and maintained in continuous culture at 37°C (5% CO2) with RPMI 1640 growth medium containing 10% fetal calf serum, 10 U/mL penicillin, 100 μg/mL streptomycin and 2 mMl-glutamine.
Immunofluorescence confocal microscopy
SK-N-BE cells grown on glass slides (8 × 104 cells) were fixed and permeabilized by the Triton/paraformaldehyde method, as described in (De Tullio et al. 1999). Cells were treated with 2.5 μg/mL nNOS, or HSP90 antibodies diluted in phosphate-buffered saline (PBS) solution containing 5% (v/v) fetal calf serum. After incubation for 3 h at 25°C, cells were washed three times with PBS and treated with 4 μg/mL chicken anti-mouse Alexa fluor 488 conjugate (Molecular Probes, Eugene, OR, USA) secondary antibody for 1 h. Images were collected by a Bio-Rad MRC1024 confocal microscopy (Bio-Rad), using a 60× Plan Apo objective with numerical aperture 1.4.
Immunoprecipitation and immunoblot
SK-N-BE cells (6 × 106 cells) were lysed by three cycles of freezing and thawing in 400 μL of ice-cold 50 mM sodium borate, 1 mM EDTA, 10 μg/mL aprotinin, 20 μg/mL leupeptin, 10 μg/mL 4-(2-aminoethyl) benzenesulfonylfluoride, pH 7.5 (lysis buffer) and protein quantification was performed using the Lowry method. Cell lysates were then centrifuged at 60 000 g for 15 min and the membranes were solubilized in 400 μL of 50 mM sodium borate, 0.1 mM EDTA, pH 7.5 containing 0.1% Triton X-100 and then 60 μg of HSP90 purified from rat brain (Averna et al. 2007) were added. The mixture, treated as described elsewhere in this paper has been pre-cleared with protein G-Sepharose and then incubated in the presence of 2 μg of anti-HSP90 mAb at 4°C, over night. Protein G-Sepharose was then added to each sample and incubated for an additional 1 h. The immunocomplexes were washed three times with lysis buffer without enzyme inhibitors, heated in sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer for 5 min and submitted to 8% SDS-PAGE (Leammli 1970). Proteins were then transferred by electroblotting onto a nitrocellulose membrane and saturated with PBS, pH 7.5, containing 5% skim milk powder. The blots were probed with specific antibodies, followed by a peroxidase-conjugated secondary antibody and then developed with an ECL detection system as previously described (Palejwala and Goldsmith 1992). The immunoreactive material was detected with a Bio-Rad Chemi Doc XRS apparatus and quantified using the Quantity One 4.6.1 software (Bio-Rad).
Determination of NO production with DAF-2DA
SK-N-BE cells (2 × 104 cells) grown on a 96-well microplate were incubated at 37°C for 30 min in 200 μL of oxygenated physiological salt solution having the following composition: 10 mM HEPES, 0.14 M NaCl, 5 mM KCl, 5 mM glucose, 1 mM MgCl2, pH 7.4 (HEPES buffer) containing 100 μM l-arginine and 5 μM DAF-2DA. This compound is non-fluorescent but reacts with NO in the presence of oxygen to form the highly fluorescent derivative DAF-2 triazole (Kojima et al. 1998; Nagatsubo et al. 1998), whose intensity is proportional to NO levels. The cells were then washed twice with HEPES buffer and after the addition of 1 μM calcium ionophore A23187 in 200 μL of HEPES buffer containing 2 mM CaCl2, in the absence or presence of 1 mM L-NAME, the fluorescence (Excitation wavelength 485 nm; Emission wavelength 535 nm) was continuously measured using the top reading mode in the fluorescence multilabel reader LB 940 Mithras (Berthold Technologies, Germany).
Assay of intracellular calpain activity
SK-N-BE cells (2 × 104 cells) grown on a 96-well microplate were incubated at 37°C for 20 min in 100 μL of oxygenated HEPES buffer containing 50 μM t-Boc-Leu-Met-CMAC fluorogenic calpain substrate, 10 μM glycine and 2 mM CaCl2. Cells were then washed twice with HEPES buffer to remove excess substrate and after addition of 250 μM or 500 μM NMDA in 100 μL of HEPES buffer containing 10 μM glycine and 2 mM CaCl2, the fluorescence emission was continuously monitored for 180 min with a Mithras LB 940 plate reader (Berthold Technologies). The excitation/emission wavelengths were 355/485 nm, respectively.
Detection of nNOS from isolated neuronal cell membranes and assay of NOS activity
SK-N-BE cells (6 × 106 cells) were collected and lysed by three cycles of freezing and thawing in 400 μL of ice-cold lysis buffer. Cell lysates were centrifuged at 60 000 g for 15 min at 4°C and the particulate material was washed three times with 50 mM sodium borate buffer pH 7.5 containing 0.1 mM EDTA. After centrifugation at 100 000 g for 5 min membranes were incubated (100 μL) in different conditions for 10 min at 37°C. Aliquots of the incubations (50 μL) were centrifuged at 60 000 g for 15 min and the clear supernatants and the membranes of each sample were separately collected and heated in SDS-PAGE loading buffer for 5 min and submitted to 8% SDS-PAGE. Proteins were then transferred by electroblotting onto a nitrocellulose membrane and nNOS was detected as described above.
Alternatively, after the incubation, the samples were directly utilized to assay NOS activity detecting the production of citrulline from l-[14C] arginine as previously described (Averna et al. 2007).
Association of nNOS to NMDA receptor in human neuroblastoma SK-N-BE cells
Confocal microscopy inspection revealed that nNOS was preferentially localized in unstimulated growing human neuroblastoma SK-N-BE cells on the inner surface of the plasma membranes (Fig. 1a). Following cell lysis, nNOS protein was present almost exclusively in the membrane being undetectable in the cytosolic fraction (Fig. 1b). Since nNOS is not an integral membrane protein, efforts were made to identify the synthase partners mediating this association. For this purpose, we have taken advantage of the known high capacity (Averna et al. 2007, 2008a) of nNOS to associate with HSP90 and accordingly the interaction of nNOS with membrane proteins was explored by co-immunoprecipitation experiments using an immobilized HSP90 antibody. As shown in Fig. 1c the immunoprecipitates were found to contain, in addition to nNOS also the NMDA receptor (Fig. 1c). These data are in agreement with previous findings indicating the presence, at the membrane level, of a NMDA-R-PSD95-nNOS complex in which the N-terminal nNOS PDZ domain provided the anchoring region (Kim and Sheng 2004).
Dissociation from membranes and activation of nNOS in SK-N-BE cells loaded with Ca2+
To reproduce the physiological conditions required for nNOS activation, neuroblastoma SK-N-BE cells were exposed to 1 μM Ca2+-ionophore A23187 for different periods of time. Following 1 or 2 h of incubation (Fig. 2a), nNOS was found to dissociate from the membranes and to become largely diffused into the cytosol. These changes in cellular distribution were accompanied by the conversion of the native 160 kDa form into low molecular mass species, the most represented one having a molecular mass of 130 kDa generated by the removal of the N-terminal PDZ domain (Fig. 2b), a process promoted by calpain (Averna et al. 2008b).
Immunoprecipitation of HSP90 from the cytosolic fraction of Ca2+-loaded cells revealed that the 130 kDa nNOS form was still associated only to the chaperone whereas, as expected, the NMDA receptor, revealed with anti-NR1 pan antibody, was recovered on the membranes without any appreciable loss of protein (Fig. 2c). The same results were obtained using anti NMDAR2B antibody (data not shown). Thus, the association of HSP90 to nNOS occurring on the membrane surface was not abolished following the cellular redistribution of the synthase.
To establish how these events were related to the activation of the synthase, the rate of NO production in cells was measured immediately after Ca2+ loading or alternatively after 1 or 2 h of incubation.
As shown in Fig. 2d, NO production became detectable approximately 40 min after addition of the Ca2+-ionophore, whereas in cells pre-exposed to the ionophore for 1 or 2 h, NO production occurred without any delay. In all these conditions, the rate of NO production was directly related to the amount of active 130 kDa form generated (Fig. 2d). The results obtained in these experiments suggested that, both degradation and activation of nNOS involved calpain activity. To verify this hypothesis, prior to the exposure to Ca2+-ionophore, cells were preloaded with the synthetic calpain inhibitor c.i.-1. In these conditions, both NO production and the appearance of the 130 kDa nNOS form were completely prevented, thus confirming the role of the Ca2+-dependent protease in the translocation and activation of the synthase.
Dissociation from membranes and activation of nNOS in SK-N-BE cells exposed to NMDA
Since it is generally proposed that NMDA-R stimulation triggers also nNOS activation, we exposed SK-N-BE cells to increasing concentrations of NMDA and measured NO production, conversion of native nNOS into the 130 kDa active form as well as intracellular calpain activity. As shown in Fig. 3a, in the presence of 250 μM NMDA, an appreciable amount of NO production became detectable only after 3 h, whereas in the presence of 500 μM NMDA, NO production became detectable after 1 h and resulted to be 2.5-fold higher after 3 h. In NMDA-stimulated cells, NOS activation evaluated by NO production was found to be concomitant with the appearance of the 130 kDa NOS form, suggesting that also in these conditions, a parallel activation of calpain occurred. This was directly confirmed by measurement of the intracellular protease activity (Fig. 3b) which increased concomitantly to the production of NO. These findings were suggesting that an increase in [Ca2+]i promoted translocation and activation of μ-calpain in proximity of the NMDA-R-nNOS-HSP90 protein clusters resulting in the limited degradation of the synthase.
Protection of nNOS by HSP90 during its release from membranes in SK-N-BE cells loaded with Ca2+
The essential role of HSP90 in the overall activation process of nNOS and the presence of HSP90-nNOS complexes in resting or in Ca2+-stimulated neuronal cells suggested to explore the availability of the chaperone in the previously described experimental conditions. The results obtained (Fig. 4) indicated that the bulk of HSP90 was present in the cytosolic fraction and that it remained unaffected following 2 h of exposure to Ca2+-ionophore.
Furthermore, to better characterize the role of calpain in the release as well as in the activation of nNOS, isolated SK-N-BE cell membranes were exposed to various cofactors, all required for activation of the synthase (Fig. 5). In the presence of Ca2+-CAM with or without HSP90 no release of the synthase from the membrane surface could be observed. Following addition of calpain to these incubation mixtures, in the absence of HSP90, the 160 kDa native synthase band almost completely disappeared. However, if the chaperone was also present native nNOS was converted in the 130 kDa form which was then released from the membranes.
These data are consistent with previous observations (Averna et al. 2007) indicating that in the presence of HSP90, digestion of native nNOS by calpain involved only the removal of the PDZ domain without any loss of activity.
Selective association of μ-calpain to NMDA-R multiprotein complexes
To better relate nNOS digestion with the production of NO as it occurs in intact cells (see Figs 2 and 3), the catalytic efficiency of native synthase was measured in its soluble isolated form or when still associated to membranes. As shown in Table 1, both isolated soluble forms of native 160 kDa and of digested 130 kDa nNOS expressed full catalytic efficiency in the absence or in the presence of HSP90. However, when bound to membranes the native nNOS showed very low catalytic efficiency even in the presence of HSP90. These results together with those described in the preceding sections, indicated an essential role of calpain in nNOS activation and were suggesting that in the course of intracellular Ca2+ increase the protease had to be translocated in proximity of the NMDA-R-nNOS-HSP90 protein clusters. To verify such hypothesis, Ca2+-stimulated cells were preloaded with the synthetic calpain inhibitor in order to allow translocation, but not activation of the protease. We have observed that, in these conditions, Ca2+ stimulation did not promote conversion of the native nNOS in the 130 kDa form, but both μ- and m-calpain isozymes, became now detectable in association with membranes (Fig. 6a).
Table 1. Relationship between nNOS forms and catalytic activity of the synthase
nNOS form (kDa)
Activity (counts × per min/sample)
nNOS forms were exposed to saturating amounts of cofactors and substrate for 30 min at 37°C and the amount of labeled citrulline was determined as described in Experimental procedures and in (Averna et al. 2007). Where indicated HSP90 purified from rat brain (Averna et al. 2007) was added to the mixtures. Native nNOS 160 kDa was isolated from rat brain as previously described (Averna et al. 2007), membrane bound native nNOS 160 kDa and nNOS 130 kDa were produced as described in Experimental procedures and in legend to Fig. 5.
Isolated native (160)
Isolated native (160)
Membrane bound native (160)
Membrane bound native (160)
Furthermore, immunoprecipitation with the anti-HSP90 antibody (Fig. 6b) revealed that μ-calpain co-immunoprecipitated together with NMDA-R and nNOS. This process resulted to be highly selective since m-calpain, although being translocated to the membrane fraction, was not detectable in the NMDA-R clusters. Comparing the amount of total μ-calpain translocated to the membrane fraction with that detected in the immunocomplex, it can be roughly estimated that 15–25% of total μ-calpain was directly associated to the functional protein clusters.
The intracellular localization of the Ca2+-dependent nNOS seems to be functionally correlated to the requirements for its activation. Several lines of evidence have in fact indicated that in neuronal cells the synthase resides in co-localization via PDZ anchoring domains interactions in a NMDA-R-PSD95-nNOS complex, a situation that can efficiently couple local variations in Ca2+ influx to the Ca2+-dependent activation of the synthase (Hajimohammadreza et al. 1997;Wu et al. 2005). Although many reports (Kornau et al. 1995; Sattler et al. 1999; Aarts et al. 2002;Kim and Sheng 2004; Rameau et al. 2007) seem to agree in that this location is a necessary prerequisite for nNOS activation, it has not yet been completely established whether in resting cells the enzyme pre-exists in this localization or if it is present in the cytosol and becomes associated only in response to a specific cell stimuli (d’Anglemont de Tassigny et al. 2009). It must however be considered that some of these conclusions have been derived from experiments performed in non-neuronal cells by over-expressing nNOS (Arundine et al. 2003) or alternatively a protein formed by nNOS PDZ domain coupled to the yellow fluorescent protein molecule (Ohnishi et al. 2008).
In the present study, we confirmed that in unstimulated human neuroblastoma SK-N-BE cells nNOS is present in a quiescent state predominantly co-localized, via its PDZ domain, with NMDA-R and in tight association with HSP90.
However, the most significant finding is the demonstration that the mechanism of nNOS activation is promoted by a calpain-mediated limited proteolysis through conversion of native 160 kDa nNOS into a fully active 130 kDa which dissociates from the membranes and freely diffuses into the cytosol retaining its association with HSP90. The formation of the 130 kDa synthase is directly correlated to NO production, whereas the presence of HSP90 provides, as previously shown (Averna et al. 2008b), protection of the catalytic moiety of the synthase from complete degradation by calpain concomitantly activated in condition of increase in [Ca2+]i.
In cells preloaded with synthetic calpain inhibitor all these sequential events are completely abrogated. Furthermore, of particular relevance was the observation that also the activation of nNOS induced by cell stimulation with NMDA involves an identical calpain-mediated post-translational modification of the native synthase accompanied by changes in its intracellular localization.
In reconstructed systems we demonstrated that the isolated native 160 kDa nNOS, but not the membrane associated form, catalyzed the production of NO (Averna et al. 2007). The purified 130 kDa form of the partially digested synthase expressed a catalytic activity comparable to that of isolated native nNOS and resulted to be the only enzyme form released from membranes. Thus, interaction of nNOS to the NMDA-R-PSD-95 complex appears to be a prerequisite for its activation which takes place by the additional association of μ-calpain promoting the dissociation of the synthase into a fully active cytosolic form. The present findings provide an additional evidence in support of the concept that calpain-mediated proteolysis can act as a regulatory rather than a destructive mechanism; as indicated also by previous observations demonstrating that specific neuronal functions are dependent on a selective calpain-mediated cleavage of a number of enzyme proteins (Patel et al. 1994; Hong et al. 1995; Wu et al. 2004, 2007; Goudenege et al. 2005).
The interplay between a regulatory or a destructive role of calpain-mediated proteolysis, which can operate either in physiological neuronal functions or in the course of neurodegenerative disorders, can be determined by several factors among which the extent of [Ca2+]i increase, the level of HSP90 expression as well as the intracellular distribution of calpastatin are those of highest significance (De Tullio et al. 1999).
An additional consideration emerging from the findings herewith reported is that nNOS activation appears to be similar in many aspects to that of eNOS. Both enzymes are, in a quiescent state, localized on the plasma membrane and become active following dissociation and translocation within the cell cytosol (Alderton et al. 2001; Kone et al. 2003). In the course of activation of both synthase isoforms, HSP90 exerts an essential role in protecting the active synthase core from digestion by calpain which is concomitantly activated. However, whereas in the case of eNOS, binding to HSP90 promotes at the same time dissociation from caveolae and protection from degradation (Averna et al. 2008b), in the case of nNOS, in the presence of HSP90 calpain digestion is limited to the removal of the PDZ domain, a step required for the dissociation of nNOS in an active form fully protected by the chaperone.
Finally, it is interesting to report that recent observations (Kim et al. 2007) have indicated that, in cultured motor neuron cells transfected with human mutant superoxide dismutase type 1 (SOD1), elevation in [Ca2+]i produces NO-dependent aggregation of SOD1. Aggregation was suppressed by preloading cells with calpain or nNOS inhibitors; but this aggregation was not attenuated if cells were treated with an exogenous NO source. On the basis of the present findings, these observations can now be considered as an indirect evidence of the crucial role of calpain in intracellular nNOS activation and NO production.
Our observations, including the protective role of HSP90, are summarized in the following model:
Proposed model for the role of calpain on the activation of nNOS in human neuronal cells. () In resting cells, NMDA receptor and PDZ-binding proteins (PDZ-BP) recruits the nNOS-HSP90 complex. In this form the synthase is inactive. () Following an increase in [Ca2+]i, calpain is translocated on the cluster, through association to HSP90, and nNOS remains inactive. () The synthase diffuses out of the membranes in an active form through the removal of its PDZ domain, catalyzed by calpain.
This work was supported in part by grants from MIUR and PRIN projects, and from the University of Genoa.