Hsp40 Couples with the CSPα Chaperone Complex upon Induction of the Heat Shock Response

In response to a conditioning stress, the expression of a set of molecular chaperones called heat shock proteins is increased. In neurons, stress-induced and constitutively expressed molecular chaperones protect against damage induced by ischemia and neurodegenerative diseases, however the molecular basis of this protection is not known. Here we have investigated the crosstalk between stress-induced chaperones and cysteine string protein (CSPα). CSPα is a constitutively expressed synaptic vesicle protein bearing a J domain and a cysteine rich “string” region that has been implicated in the long term functional integrity of synaptic transmission and the defense against neurodegeneration. We have shown previously that the CSPα chaperone complex increases isoproterenol-mediated signaling by stimulating GDP/GTP exchange of Gαs. In this report we demonstrate that in response to heat shock or treatment with the Hsp90 inhibitor geldanamycin, the J protein Hsp40 becomes a major component of the CSPα complex. Association of Hsp40 with CSPα decreases CSPα-CSPα dimerization and enhances the CSPα-induced increase in steady state GTP hydrolysis of Gαs. This newly identified CSPα-Hsp40 association reveals a previously undescribed coupling of J proteins. In view of the crucial importance of stress-induced chaperones in the protection against cell death, our data attribute a role for Hsp40 crosstalk with CSPα in neuroprotection.


Introduction
In response to a range of stressful stimuli including hyperthermia and ischemia, an ancient evolutionarily conserved cellular program called the heat shock response is activated and the expression of several chaperones is induced to enhance cell survival to subsequent insults. The heat shock response also involves the translocation of several chaperones [1,2]. Although the mechanistic basis of chaperone cytoprotection is not yet understood, the chaperone anti-apoptotic activity is thought to be due to the ability of chaperones to rid the cell of misfolded proteins. In addition to the stress-induced chaperones, many molecular chaperones are expressed constitutively and are widely held to have basic and indispensable functions in maintaining protein conformation. Constitutive and stress-induced chaperones maintain an important balance in the cell between protein refolding and protein elimination, however, while transient upregulation of molecular chaperones is critical for cell survival, chronic up-regulation of chaperones (eg Hsp70) correlates with poor prognosis in some types of cancer [3]. Neurons are thought to be especially vulnerable to the risk of protein misfolding due to the conformational flexibility required of the cellular machinery underlying synaptic transmission. Furthermore, post-mitotic cells, such as neurons, cannot dilute aggregated proteins through cell division, leading to the potential accumulation of misfolded proteins. Huntington's, Alzheimer's, Parkinson's and Prion diseases are caused by defects in protein folding, underlining the biological importance of the problem of aberrant protein folding in neurons. In experimental models, molecular chaperones, either constitutive or stress-induced, are inhibitors of neurodegeneration (reviewed: [4][5][6]). Pharmacological modulators that directly regulate chaperone and stress-induced chaperone activity have been identified, emphasizing the potential of the heat shock response as a pharmaceutical target (Reviewed: [7]).
Cysteine string protein (CSPa) is a synaptic vesicle protein bearing a signature J domain and a cysteine rich string region that is implicated in the defense against neurodegeneration. The cytosolic proteins Hsc70 (heat shock cognate protein of 70 kDa), SGT (small glutamine rich tetratricopeptide repeat domain protein) and HIP (Hsc70 interacting protein) form a complex with CSPa which is tethered to the synaptic vesicle. Assembly of the chaperone components is thought to allow for localized activation of Hsc70, a cytosolic ATPase that couples energy from ATP hydrolysis to conformational work on target proteins [8][9][10][11]. The assembly of the CSPa chaperone complex is nucleotide sensitive, emphasizing the dynamic nature of the complex [8,10,12]. Why a specialized synaptic vesicle chaperone system evolved remains a mystery.
The existence of the heat shock response raises several questions. How do stress-induced chaperones protect synaptic transmission and prevent neurodegeneration? Is there crosstalk between stress-induced chaperones and constituitively expressed chaperones or do these chaperone machines have separate nonoverlapping cellular tasks? As a first step toward testing the hypothesis that stress-induced chaperones are coupled to the CSPa chaperone system, we evaluated the components of the CSPa complex during transient expression of stress chaperones. CSPa expression levels do not increase in response to heat shock or geldanamycin treatment in neural cell lines. The previously reported 70 kDa CSPa dimer [13,14] was reduced following heat shock. Suprisingly our findings demonstrate that the stress-induced J protein Hsp40 becomes a major component of the CSPa multimeric complex after heat shock. Following heat shock, Hsp40 expression is increased and Hsp40 localizes to the plasma membrane. Geldanamycin, like heat shock, triggers the assembly of Hsp40 with the CSPa multimeric complex. In vitro, the stimulatory effect of CSPa on the steady-state hydrolysis of GTP by Ga s is enhanced in the presence of Hsp40. Furthermore, transient transfection of CSPa or induction of the heat shock response increased isoproterenol-stimulated phosphorylation of synapsin.
Our results suggest that the transient assembly of Hsp40 with the CSPa complex is important in the maintenance of synaptic function in the face of environmental stress, and emphasize the complexity and functional elegance of the J protein chaperone machines.

CSPa expression is not altered in response to heat shock
To begin to test the role of CSPa in the heat shock response, the expression of CSPa and Hsp70 was examined in CAD mouse neuroblastoma cells before and after 40 min of conditioning heat shock at 42uC. CSPa is constitutively expressed in brain and exocrine/endocrine secretory tissues, however, its expression has been reported to be influenced by antidepressants [15][16][17], amphetamines [18] and diabetes [19]. The cellular mechanisms that underlie changes in CSPa expression are currently unknown. The cDNA clone for rat CSPa (594 bp open reading frame) contains 181 nucleotides of 59 untranslated region, 1.2 kb of 39 untranslated region and encodes a 35 kDa protein with extensive lipid modification [13]. Figure 1 shows that CSPa levels are not increased by heat shock. As expected, Hsp70 and Hsp40 were induced by heat shock and their induction was blocked by pretreatment with quercetin, an inhibitor of the heat shock response. Quercetin did not alter CSPa expression.
We then evaluated the time course of expression of select chaperones in response to heat shock. Following a conditioning stress, Hsp70 is rapidly expressed in CAD, LAN1, and PC12 cell lines, consistent with Hsp70 induction in neurons [1]. Figure 2 demonstrates that Hsp70 protein expression is clearly increased ,3 hours after heat shock. Although Hsp70 is detected in LAN1 cells prior to the conditioning stress, the levels of Hsp70 expression were still observed to increase in response to heat shock. The time course of Hsp70 expression showed slower onset in differentiated compared to undifferentiated CAD cells, however robust Hsp70 expression was observed at ,6 hours in differentiated CAD cells. Figure 2 shows that Hsp25/27 expression also increased in CAD, LAN1, PC12 cell lines ,5 hours after a conditioning heat shock. Although Hsp25/27 was expressed in unstressed LAN1 and PC12 cells, further increases were observed following a conditioning heat shock. Hsp40 levels increased ,3 hours after a conditioning heat shock in CAD and LAN1 cells, however no increase was seen in PC12 cells which already had high control levels of Hsp40 preheat shock. In contrast to Hsp70, Hsp40 and Hsp25/27, the expression of actin was not altered in response to heat shock demonstrating the specificity of the cellular response to a conditioning stress. Following heat shock CSPa levels are not altered in CAD cells (Figure 1), however in PC12 and LAN1 cells reduced CSPa levels were observed ( Figure 2). These data are consistent with those found in vivo where the severity of the stress required to trigger the heat shock response is reported to vary among neural populations [6].

Hsp70 and Hsp25/27 expression is variable among neural cell lines
Next we directly compared non-heat shocked (control) CAD, differentiated CAD, PC12, differentiated PC12, LAN1 and HEK cells for CSPa expression. Figure 3A shows that CSPa was present in all cell lines except HEK cells. The cell lines were then evaluated for basal (control) levels of heat shock chaperones including Hsp70, Hsp25/27 (heat shock protein 25/27) and Hsp40. Hsp70 was not detectable in non-heat shocked (control) CAD, differentiated CAD or PC12 cells, but was present at high levels in LAN1 and HEK cells and detectable in differentiated PC12 cells. Notably, some tumor cells have been reported to constitutively express high levels of the anti-apoptotic chaperone Hsp70. Furthermore, high Hsp70 levels have been reported to correlate with poor prognosis in some types of cancer [3,7]. In contrast, Hsp25/Hsp27 was absent from CAD cells but detectable in HEK cells and abundant in LAN1 and PC12 cells. Hsp25 and Hsp27 are homologous proteins. Hsp25 is present in rat and mouse, while Hsp27 is present in humans. Anti-Hsp25 antibody does not cross-react with Hsp27 and vice versa. Actin is shown as a loading control (Figure 3, panel 5). The basal expression of select chaperones in unstressed rat brain is shown for comparison in Figure 3B. Hsp40 and Hsc70 are abundant in both the cytosolic (S) and membrane particulate (M) fractions of rat brain while the stress-inducible Hsp70 was not detected. Hsp25 was detected in the cytosolic fraction. Taken together, these observations indicate that while CSPa was present in all neural cell lines as expected, the background expression of stress-induced chaperones varied extensively between the cell lines.

The CSPa dimer is reduced by heat shock and increased by quercetin treatment
Although CSPa is present in neural cell lines, its expression is ,10 fold lower (1061.7, n = 3, data not shown) than that found in rat brain homogenates. To conduct a more detailed analysis of the CSPa chaperone complex, CSPa was examined in CAD mouse neuroblastoma cells transiently transfected with CSPa in order to bring CSPa levels up to those found in adult rat brain. After transfection a 70 kDa CSPa immunoreactive band was observed ( Figure 4). 70 kDa CSPa dimers have previously been reported in rat brain [13,20], rat hippocampus [12], rat pancreas [13], PC12 cells transiently expressing CSPa [14,21], and HEK293 cells transiently expressing CSPa [22]. The cellular role of the CSPa dimer is not known. Figure 4 shows that in CAD cells transfected with CSPa the 70 kDa CSPa dimer is stable, SDS-resistant and maintained after incubation in sample buffer at 80uC for 10 min. Figure 4 clearly demonstrates that following 40 minutes of conditioning heat shock at 42uC, there is a decline in the CSPa dimer detected. In contrast, quercetin increases the CSPa-CSPa complex in control and heat shocked cells ( Figure 4). The upper panel in Figure 4 is an overexposure of the 70 kDa CSPa dimer demonstrating its presence in control but not heat shocked CAD cells. In transfected ( Figure 4) but not untransfected ( Figure 1) CAD cells, quercetin was observed to increase expression levels of the CSPa monomer. Taken together, these data show that the extremely stable CSPa dimer is regulated by quercetin as well as by a conditioning heat shock.
Hsp40 is a major component of the CSPa complex after a conditioning heat shock Two possible scenarios could mediate the anti-apoptotic activity of stress chaperones. It is possible that (1) stress inducible chaperones have evolved to perform the same function as constitutively expressed chaperones or that (2) they carry out specialized functions to specifically cope with physiological stress. In order to distinguish between these two possibilities, we examined the association between CSPa and heat shock proteins. Glutathione-S-transferase (GST) fusion proteins consisting of CSPa 1-112 or CSPa  (full length) were coupled to glutathione agarose beads and used in an in vitro binding assay. The beads were incubated with cell homogenate, washed, and bound proteins were eluted and evaluated by Western blot. We have previously shown that CSPa 1-112 contains two binding sites for G proteins and has guanine nucleotide exchange (GEF) activity for Ga s [11]. Recombinant full length CSPa 1-198 also contains both binding sites for G proteins, but only has GEF activity for Ga s in the presence of Hsc70 and SGT (small glutamine-rich tetratricopeptide repeat domain protein). Specifically, we have shown that CSPa regulates heterotrimeric GTP binding proteins (G proteins) by preferentially targeting the inactive GDP-bound form of Ga s and promoting GDP/GTP exchange which increases cAMP levels and downstream phosphorylation. In vitro, Hsc70 and SGT trigger a switch in CSPa from an inactive GEF to an active GEF. Therefore we examined both CSPa  (requires Hsc70/SGT activation) and CSPa 1-112 (active GEF) for their possible associations with stress-induced chaperones.
Chaperones homologous to those found in the CSPa complex are induced following a preconditioning heat shock (eg. Hsp70 (70 kDa heat shock protein) and Hsp40 (40 kDa heat shock protein)). Hsc70 and Hsp70 both associate with the J domain of CSPa [8,9,20]. Despite the robust induction of Hsp70 expression by heat shock and its significant homology (85%) with Hsc70 [23], a CSPa/Hsc70 complex was favored over a CSPa/Hsp70 complex in CAD cell homogenates ( Figure 5). After a conditioning heat shock, Hsp70 was induced and associated weakly with the GST-CSPa 1-112 and CSPa  complex. These data confirm the relatively weaker CSPa association and ATPase activation of Hsp70 compared to Hsc70 utilizing yeast two hybrid and ATPase assay techniques [9,20]. The ATPase Hsp90 weakly associated with the CSPa complex. Like Hsc70, Hsp70 association was increased in the presence of ATP. In contrast, Hsp25 was not found to associate with the CSPa complex either before or after heat shock.
If Hsp40 and CSPa have evolved to perform separate functions, then it is possible that induction of Hsp40 will reduce the assembly of CSPa with Hsc70. Given the homology between the J domains of CSPa and Hsp40, one might expect elevated levels of Hsp40 to disrupt the CSPa/Hsc70 complex and favor a Hsp40/Hsc70 complex by competing with CSPa for association with Hsc70 and Hsp70. To our surprise, however, Hsp40 was found to associate robustly with both GST-CSPa 1-112 and CSPa 1-198 after heat shock. Figure 5 shows that Hsp40 association with CSPa 1-198 was nucleotide independent, however association of Hsp40 with CSPa 1-112 was greater in the presence of either ATP or GDP. These data indicate that the components of the CSPa chaperone complex are altered in response to a conditioning stress to include the cytosolic stress-induced J protein Hsp40.
While these data are consistent with a direct interaction between Hsp40 and CSPa, it does not permit us to rule out the possibility that CSPa/Hsp40 interact indirectly. To investigate this possibility, we examined the ability of immobilized CSPa fusion proteins to interact with soluble Hsp40. As shown in Figure 6A, immobilized GST-CSPa  and GST-CSPa 1-112 were able to bind purified, soluble Hsp40 in a pull down assay, indicating that there is indeed a direct interaction between CSPa and Hsp40.
To further understand the structural requirements for assembly of Hsp40 with the CSPa complex, Hsp40 deletion and point mutants were constructed and the regions of Hsp40 required for its binding to CSPa were determined. In each assay, an equal amount of fusion protein was immobilized to sepharose beads as confirmed by Ponceau S staining. The presence of Hsp40 was analyzed by Western blotting. Hsp40 and CSPa belong to a large and diverse protein family [24]. Each member has a conserved J domain that functions to stimulate Hsc70/Hsp70 ATPase (eg CSPa stimulates Hsc70 ATPase [8]). There is no functional one to one correspondence between members of the J protein family and members of the Hsp70 family. J domains are a ,70 amino acid region of homology comprised of four a helices with a highly conserved tripeptide of histidine, proline and aspartic acid (HPD motif) located between helices II and III. The structures of the J domain of CSPa (mouse) and Hsp40 (human) have been determined from nuclear magnetic resonance studies [25]. Figure  S1 shows the comparison of the amino acid sequence between rat CSPa and Hsp40. The J domain (magenta) and cysteine string region (red) of CSPa and the J domain and the DnaJ C-terminal domain (cyan) of Hsp40 are indicated. CSPa and Hsp40 show only 16% identity over their entire amino acid sequences, but 51% identity within the J domains. Secondary structure predictions for helix 1 show a weak ''b sheet'' (the probability is 0.4 sheet compared to 0.3 helix; 1 is high, 0 is low) for position 5-8 a.a. of Hsp40_Rat and 17-20 a.a. of CSPa_Rat, while predictions are stronger for helices 2, 3 and 4 (probability is 1, 1 and 0.7 respectively). Mutation of the highly conserved HPD tripeptide of Hsp40 (Hsp40 HPDRAAA ) did not abolish binding to CSPa, indicating that CSPa/Hsp40 association is not dependent on this conserved motif. Furthermore, Hsp40 1-80 as well as Hsp40 1-80HPDRAAA were not found to associate with CSPa  or CSPa 1-112 ( Figure 6B&C). These data therefore define Hsp40 residues 81-340 as important for binding to CSPa.
To understand better the nature of the CSPa/Hsp40 association, we evaluated the distribution of CSPa and Hsp40 in CAD cells transiently expressing myc-CSPa (Figure 7). In control cells (C), CSPa was observed to be concentrated at the cell-cell contact sites (indicated with an arrow) and, in contrast, Hsp40 was broadly expressed. Following heat shock (H.S.) we found CSPa and Hsp40 to be present primarily in the plasma membrane however reduced at the cell-cell contacts, indicating that Hsp40 as well as CSPa was relocalized. These observations are consistent with our biochemical data (Figures 5&6) demonstrating a CSPa:Hsp40 association following heat shock. Panel 4 clearly demonstrates that quercetin blocked the heat shock general redistribution of Hsp40 to the plasma membrane. Intracellular localization of CSPa and Hsp40 was increased in cells treated with quercetin (Q) as well as quercetin followed by heat shock (Q+HS). To our surprise, Hsp40 but not CSPa localized to cell-cell contacts following the quercetin/heat shock. Plasma membrane colocalization of CSPa and Hsp40 in single differentiated cells is shown in yellow in the right hand panel (panel 5). Taken together, these biochemical and histochemical data demonstrate that following heat shock, cellular levels of Hsp40 are increased and both Hsp40 and CSPa undergo redistribution. These data indicate that changes in the expression levels of Hsp40 as well as changes in the cellular localization of Hsp40 and CSPa facilitate the assembly of Hsp40 with the CSPa chaperone complex.

Hsp40 is a major component of the CSPa complex after Geldanamycin treatment
Next we evaluated a series of agents to see if they altered either CSPa levels or the components of the CSPa complex ( Figure  S2A). The Hsp90 ATPase inhibitors geldanamycin, 17-AAG and novobiocin robustly induced the expression of Hsp70 in PC12 cells. Lithium ions have been reported to enhance CSPa expression [26] and we have previously shown that CSPa increased the response to the b2 adrenergic agonist isoproterenol [11]. No Hsp70 was detected in PC12 cells treated with isoproterenol, LiCl or quercetin. In contrast to Hsp70 and Hsp25, the expression of actin was not altered in response to geldanamycin or 17AAG demonstrating the specificity of the cellular response to the Hsp90 inhibitors. Following geldanamycin treatment of PC12 cells, Hsp70 expression is detected as early as 5 hours and continues to increase ( Figure S2B). To our surprise, geldanamycin did not trigger the expression of Hsp70 but did induce Hsp40 expression without increasing CSPa expression in CAD mouse neuroblastoma cells ( Figures S2C&D). The molecular events that underlie the difference between geldanamycin-induced and heat shock-induced Hsp70 expression in CAD cells remains to be established, however, it provided us the opportunity to investigate the CSPa/Hsp40 association in the absence of Hsp70.
We then examined the CSPa complex in geldanamycin treated CAD cells. CSPa 1-112 or CSPa 1-198 fusion proteins were coupled to glutathione agarose beads and used in an in vitro binding assay. Following geldanamycin treatment, Hsp40-CSPa 1-112 and Hsp40-CSPa 1-198 complexes were abundant in the presence of ATP (Figure 8), similiar to that found after heat shock ( Figure 5). Again, Hsp90 was observed to associate weakly with the CSPa complex. Thus the CSPa chaperone complex is dynamic, undergoing changes in components in the post-stress response and in response to Hsp90 inhibitors. Taken together, our results indicate that as the cellular levels of Hsp40 rise in response to a conditioning stress or after treatment with Hsp90 inhibitors, the CSPa complex becomes an Hsp40-CSPa complex.

Hsp40 promotes CSPa's GEF activity in vitro and synapsin phosphorylation in CAD neuroblastoma cells
Next we evaluated Hsp40's effect on the steady state hydrolysis of GTP by Ga s . We have previously demonstrated that the stimulation of GTPase activity by Ga s by CSPa requires Hsc70 and SGT [11]. Figure 9A shows that Hsp40 enhanced the CSPa-Hsc70-SGT-stimulated increase in GTP hydrolysis by Ga s . GTP hydrolysis by Ga s was not altered in the presence of Hsp40 alone, indicating that Hsp40 does not effect the hydrolysis of GTP by Ga s alone. Furthermore, Hsp40 does not hydrolyze GTP. To gain further insight into the transient association of Hsp40 with the CSPa chaperone complex, we evaluated the activation of cellular mechanisms downstream of G proteins after heat shock. Phosphorylation of synapsin in CAD cells serves as an independent readout of cellular signaling through Ga s . Figure 9B shows that isoproterenol-induced phosphorylation of synapsin in CAD cells was increased after heat shock (3.7 fold) or transfection with myc-tagged CSPa (4.2 fold). This is consistent with previous work showing that expression of CSPa in HEK cells increased isoproterenol-stimulated cAMP levels [11] as well as G protein inhibition of N-type calcium channels [12]. Moreover, we have recently reported that the expression of CSPa in CAD cells increases isoproterenol-mediated phosphorylation of the transcription factor CREB [27]. These results show that the heat shock response modulates signaling through Ga s pathways.

Discussion
In this study, we demonstrate for the first time that the stressinduced J protein, Hsp40, physically couples with the constitutive CSPa chaperone complex. CSPa is a synaptic vesicle J protein that is essential for periods of extended neurotransmission and is implicated in the defense against neurodegeneration [28][29][30]. CSPa is constitutively expressed on synaptic vesicles and is thought to tether Hsc70 for conformational work at the synaptic vesicle site. Hsp40 is a cytosolic J protein that is rapidly and transiently induced in response to stress, including, but not exclusive to, heat shock and geldanamycin treatment. Stressinduced expression of several chaperones, including Hsp40, is widely thought to protect cells from the deleterious effects of subsequent stress. Interference with the heat shock response would be expected to have enormous cellular consequences and reduce cell survival. It is fully anticipated that transient expression of stress-induced chaperones in vivo occurs routinely in neurons to cope with the stresses rendered by various insults. Intriguingly, some neurons have a particularly higher threshold for the induction of the heat shock response compared with other neurons [6]. In this study, we provide evidence that Hsp40 specifically and directly associates with the constitutive CSPa synaptic vesicle chaperone complex following either a conditioning heat shock or geldanamycin treatment. Following heat shock, Hsp40 expression is increased and both CSPa and Hsp40 undergo redistribution. The association of Hsp40 with CSPa correlates with reduced CSPa dimerization, translocation of Hsp40 to the plasma membrane and enhanced CSPa-mediated augmentation of steady state GTP hydrolysis. Association with CSPa is mediated via C terminal binding sites of Hsp40 and does not involve the J domain. Since neurotransmitter release relies on complex interactions between multiple cellular components, we speculate that the crosstalk between CSPa chaperone machinery and the stress-induced Hsp40 is important in maintaining functionally competent synapses.
Whether stress-induced chaperones typically perform independent cellular tasks or are coupled to constitutive chaperone machines, such as CSPa, is a current biological question. The notion that Hsp40 contributes to the folding activity of the CSPa  [30][31][32]. The Drosophila CSPa null mutants exhibit temperature sensitive paralysis and die as larvae or within days of adulthood [30]. Deletion of CSPa in mice causes blindness followed by progressive motor and sensorial impairment and neurodegeneration with no survival beyond 4 months [29,33,34]. Thus, these reports indicate that CSPa is important for the long term functional integrity of the synaptic machinery. It is possible, given the data presented here that the prevention of presynaptic neurodegeneration by the synaptic vesicle CSPa chaperone complex involves a transient interaction with the highly conserved stress induced protein Hsp40.
In conclusion, our combined findings suggest a model in which heat shock alters the composition of the CSPa complex to include Hsp40. Understanding the composition of the CSPa chaperone complex, either basal or following a conditioning stress, is crucial to understanding the physiological role(s) of CSPa. Data presented here show that in vitro, Hsp40 increased the CSPa-induced increase in steady state GTP hydrolysis of Ga s . Furthermore, in CAD cells, induction of the heat shock response decreased CSPa dimerization and increased isoproterenol-stimulated phosphorylation of synapsin. In view of the crucial importance of stressinduced chaperones in protection against cell death, our data attribute a key role to the association of Hsp40 with the CSPa chaperone complex in neuroprotection.

Reagents and Chemicals
Anti-CSPa polyclonal was prepared as described previously [13]. Anti-Hsp70 mouse monoclonal, anti-Hsp40 rabbit polyclonal, anti-Hsp90 rat monoclonal, anti-Hsp25 rabbit polyclonal, anti-Hsp27 mouse monoclonal were from Assay Designs. Antiphosphosynapsin and anti-synapsin rabbit polyclonal antibodies were from Cell Signaling Technology. Anti-c-myc mouse monoclonal was from Clonetech. Anti-actin mouse monoclonal, anti-Hsp70/Hsc70 mouse monoclonal, quercetin and isoproterenol were from Sigma. Geldanamycin was from Calbiochem. 17AAG was from Invitrogen. Novobiocin was from EMO Biosciences Inc.
CAD mouse neuroblastoma cells were seeded into 6 well plates and grown in DMEM/F12 medium supplemented with 10% fetal bovine serum and 1% Penicillin/streptomycin. For differentiation cells were grown in Opti-MEM for 60 hrs. LAN1 human neuroblastoma cells were seeded into 6 well plates and grown in RPMI medium supplemented with 10% fetal bovine serum and 1% Penicillin/streptomycin. PC12 cells were obtained from ATCC. PC12 cells were grown in Dulbecco's modified Eagle's Medium supplemented with 10% heat-inactivated horse serum and 5% fetal calf serum. For differentiation PC12 cells were treated with 50 ng/ml mouse NGF (R&D systems) for 9 days. Human embryonic kidney tsa-201 (HEK) cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Cells were lysed in 40 mM Tris (pH 7.4), 150 mM NaCl, 2 mM EDTA, 1 mM EGTA, 1 mM Na 3 VO 4 , 0.1% SDS, 1% Tx100, 0.5 mM PMSF and protease inhibitor (Sigma) end-over-end at 4uC for 1 hr. Lysates were centrifuged at 150006g for 5 min at 4uC and the supernatant was collected. Protein concentration was determined using a Bradford reagent (BioRad).

Transient transfection of CAD cells
CAD cells were washed in PBS and transiently transfected with 0.5 ug myc-tagged rat CSPa 1-198 DNA using Lipofectamine-2000 (Invitrogen) in Opti-MEM, and maintained in culture for 24 hrs prior to heat shock or drug treatment.

Cell lines and lysate preparation
Whole rat brains were homogenized in 20 mM Tris-HCl buffer (pH 7.4), 2 mM MgSO 4 , 1 mM PMSF and EDTA-free inhibitor cocktail as previously described [51]. The homogenate was centrifuged at 100,0006g for 1 h at 4uC. The resultant soluble fraction was removed and designated the soluble cytosolic fraction (S). The remaining pellet was solubilized in homogenizing buffer containing 1% (w/v) n-dodecyl-b-D-maltoside (Calbiochem) for 60 min at 4uC. Following centrifugation at 100,0006g for 1 hr at 4uC, the resulting supernatant constituted the detergent-solubilized membrane particulate fraction (P). All procedures were carried out in strict accordance with a protocol approved by the University of Calgary Animal Care Committee.

Immunoblotting
Proteins were transferred from polyacrylamide gels to nitrocellulose (0.45 mm) in 20 mM Tris, 150 mM glycine, and 12% methanol. Membranes were blocked with 4% milk solution (prepared in PBS with 0.1% Tween 20) and incubated with primary antibody for 2 hours at room temperature or overnight at 4uC. The membranes were washed in blocking solution and incubated with horseradish peroxidase-coupled secondary anti-body. The signal was developed using West Pico Pierce reagent (Pierce Biotechnology, Inc.) and exposed to Kodak film.
Sequence analysis CSPa from Rattus norvegicus (accession number: NP_077075.1) and Hsp40 from Rattus norvegicus (accession number: EDL92267.1) were evaluated for homology between J domains. Alignments of sequences were obtained using CLUSTAL-W with default settings in place and EMBOSS pairwise global alignment using an implementation of the Needleman-Wunsch algorithm [52]. Domains were identified with InterProScan and secondary structure prediction was carried out using PredictProtein [53].
Immunofluorescence CAD cells were grown on coverslips coated with glycerol in DMEM/F12 or Opti-MEM for differentiation. Cells were washed in PBS, fixed in 2% paraformaldehyde for 10 min, permeabilized in ice cold methanol for 10 min and rinsed with PBS. Cells were blocked in a 3% BSA, 0.05% Tween 20 solution in PBS for 30 min. Incubations of cells with primary antibodies were carried out sequentially overnight at 4uC and 1 hr at room temperature. Following incubation with primary antibody, cells were washed 36with PBS and incubated with either goat anti-mouse conjugated to Alexa 488 or sheep anti-rabbit conjugated to Cy3 secondary antibody in the blocking solution for 1 hour at room temperature. Following secondary antibody incubation, cells were washed 36with PBS, mounted onto glass slides with DABCO (Sigma) and photographed with a Leica confocal microscope. The fluorophores Alexa 488 and Cy3 were excited at 470 nm and 535 nm respectively, and images were collected at 525 nm and 610 nm.

GTPase assay
Steady-state GTPase reactions were performed at 25uC in the presence of Ga s (0.3 mM), Gb 1 c 1 (0.6 mM), and the absence or presence of 0.3 mM Hsp40, CSPa, Hsc70 and SGT [54]. Proteins were mixed with 10 mM [c-32 P]GTP (1mCi) in a final volume of 50 ml of 20 mM Tris-HCl (pH 8.0) buffer containing 130 mM NaCl and 10 mM MgSO 4 and the reaction was started by the addition of Ga s . Aliquots (20 ml) were withdrawn after 30 min and transferred to 100 ml of 7% (v/v) perchloric acid. Nucleotides were precipitated with 700 ml of 10% (w/v) charcoal suspension in phosphate-buffered saline and free [ 32 P i ] was measured with liquid scintillation counting. Results were fit with linear regression.