Calreticulin (CRT) and calnexin (CNX) are members of a family of endoplasmic reticulum (ER) chaperones that fold newly synthesized polypeptides. Aside from their role as foldases in the ER, our laboratory has shown that all members of this family of proteins modulate Ca²⁺ oscillations. In Xenopus oocytes and other cells, stimulation by G-protein and tyrosine coupled receptors results in Ca²⁺ release from the Inositol 1,4,5 trisphosphate receptor (IP₃R) located in the ER. Following release, Ca²⁺ is re-sequestered into the ER by Ca²⁺ ATPases of the SERCA family. CRT and CNX overexpression inhibit Ca²⁺ oscillations when co-expressed with SERCA2b or when oocytes are treated with pyruvate malate to induce oscillations. By domain deletion mutagenesis of CRT we have determined that the N and P domains are necessary for the inhibition of Ca²⁺ oscillations. The mechanism of inhibition may involve a lectin-like interaction since mutagenesis of a lumenal asparagine to alanine in SERCA2b destroys the inhibitory effect. Coexpression of SERCA2a (which lacks the luminal asparagine) with either CRT or CNX does not inhibit Ca²⁺ oscillations, consistent with the notion that a lectin interaction may be involved. Unlike CRT, which is entirely lumenal, CNX has a cytosolic domain that is phosphorylated by multiple kinases. Mutagenesis of two PKC/PDK residues in CNX indicated that S562 supports phosphorylation. Expression of SERCA2b with a mutated CNX in S562 prevents the inhibition of Ca²⁺ oscillations suggesting that this residue serves as a phosphorylatable regulatory switch controlling the interaction of CNX with SERCA2b. Indeed, immunoprecipitations with a CNX specific antibody of oocytes treated with or without IP₃ and preloaded with [γ-³²P]-ATP demonstrated that S562 is phosphorylated at rest and dephosphorylated in response to IP₃. Phosphorylation-mediated control of the interaction of CNX with SERCA2b is of significance since it suggests a bi-directional mode of communication between the Ca²⁺ signaling system and the folding machinery in the ER to maintain Ca²⁺ homeostasis in the organelle. The maintenance of Ca²⁺ homeostasis in the ER is then essential for protein folding.

Introduction

In the past six years the goal of our laboratory has been to understand the role of ER chaperone proteins in regulating Ca²⁺ signaling. Our studies have primarily focused on Ca²⁺ re-uptake into the ER lumen by the family of sarco-endoplasmic reticulum Ca²⁺ ATPases (SERCAs). Prior to our work, SERCAs were largely perceived as housekeeping enzymes, with very little dynamic regulation. In addition, ER chaperone proteins were largely thought to be involved only in nascent protein folding. We discovered that at least two ER chaperone proteins, CRT and CNX, can modulate Ca²⁺ ATPase activity. This regulation occurs with ER resident mature proteins such as SERCA2b and is consistent with the need to maintain ER Ca²⁺ concentrations at levels that are optimal for protein folding. These findings underscore the necessity to expand our classical views of Ca²⁺ signaling and ER chaperones. In this chapter, we review the evidence that has lead to this new model of ER Ca²⁺ signaling.

Our initial studies arose from the observations that increments in the state of Ca²⁺ store refilling correlated with the sensitivity of the inositol 1,4,5 trisphosphate receptor (IP₃R) to release Ca²⁺.¹ These reports suggested that the luminal Ca²⁺ concentration could regulate Ca²⁺ release. To investigate this problem, we chose to manipulate ER luminal Ca²⁺ concentrations by overexpressing calreticulin (CRT), a prototypical Ca²⁺ storage protein2 that is expressed at millimolar concentrations in the ER.^3,4 Our hypothesis was simply that higher Ca²⁺ levels in the lumen of the ER would increase Ca²⁺ oscillations via the IP₃R (to increase release) or via SERCA (to increase uptake). As presented below, we successfully determined that IP₃ induced Ca²⁺ release could in fact be regulated by altering the expression levels of CRT. In addition, overexpression of CNX, a closely related, but membrane anchored family member of CRT also regulated Ca²⁺ release. However, our approach also led to several unexpected findings that changed our understanding of the underlying physiology. Namely, that high capacity Ca²⁺ storage in CRT was not responsible for the effects observed, as discussed later. Before we discuss these data, we briefly review our experimental system.

Xenopus Oocytes As an Expression System

Xenopus oocytes were used throughout these studies. A complete description of this system is available.⁵ Briefly, we used IP₃-induced Ca²⁺ wave activity to assay both Ca²⁺ release and Ca²⁺ uptake. The rising phase of individual waves reflects the activity of the IP₃Rs and the decay phase reflects the uptake processes contributed mainly by Ca²⁺ ATPases of the SERCA family. Mitochondria can also contribute to the decay phase in the presence of energizing respiratory chain substrates.⁶ Cytosolic Ca²⁺ is imaged with Ca²⁺ indicator dyes and confocal microscopy. The advantage of this approach is that all experimental measurements can be carried out in vivo. The disadvantage is that contributions to the Ca²⁺ signal from Ca²⁺ release vs Ca²⁺ uptake cannot be precisely distinguished. Consequently, the functional Ca²⁺ oscillations assay is complemented by molecular and biochemical techniques.

Calreticulin and Calnexin Have an Inhibitory Effect on Ca²⁺ Oscillations

Overexpression of CRT⁷ or CNX⁸ in oocytes reduces the number of oocytes displaying IP3-mediated Ca²⁺ oscillations. In the remaining oocytes that exhibit repetitive activity, the amplitude and frequency of Ca²⁺ oscillations was significantly lower.^7,9 These observations were consistent with an effect of CRT or CNX on the endogenous SERCA pump to inhibit uptake or on the IP₃R to favor Ca²⁺ release. We decided to test the “SERCA” hypothesis first and this is what we have focused in this chapter. We had previously reported that overexpression of SERCA pumps leads to an increase in the frequency of Ca²⁺ oscillations.¹⁰ To determine whether the inhibitory effect of CRT and CNX on Ca²⁺ oscillations was due to inhibition of pump activity, we co-expressed CRT^7,9 and CNX⁸ with SERCA2b, the ubiquitous Ca²⁺ ATPase of the ER. In this high background of Ca²⁺ oscillations, CRT still had a strong inhibitory effect. There was a significant decrease in the number of oocytes displaying Ca²⁺ oscillations, and in those oocytes that had repetitive Ca²⁺ release, there was a decrease in period between waves and a corresponding decrease in the decay time of individual oscillations (measured as the 50% decrease in T_1/2).^8,9 Energization of mitochondria by respiratory substrates (pyruvate-malate, PM) also modulates cytosolic Ca²⁺ by eliciting robust, low-frequency synchronized Ca²⁺ oscillations.⁶ In yet another control experiement, we demonstrated that overexpression of either CRT⁷ or CNX ⁸ in oocytes treated with PM also reduces the number of oocytes that have Ca²⁺ oscillations and lowered the amplitude and frequency of oscillations in those oocytes that still had repetitive activity. We concluded from these experiments that the most parsimonious explanation of the data was that Ca²⁺ oscillations were decreased by an inhibition of endogenous SERCA2b.^7,8 Furthermore, since both CRT and CNX have only the conserved P-domain in common, inhibition of Ca²⁺ oscillations was likely being mediated by a lectin or chaperone interaction with the SERCA substrate as discussed below.

The observed inhibitory effects of CRT and CNX may be due to a physical interaction with the pump. Both chaperones contribute to the folding and oligomerization of glycoproteins in the ER.¹¹¹⁴ The lectin domain was thought to be in the central portion of the P-domain.¹⁵ However more recently it is the N domain that is considered to contain the lectin binding domain.¹⁶ In CRT, this domain has a low affinity Ca²⁺ binding site and is responsible for binding to the substrate during protein folding. Recent evidence suggests that the P-domain is flexible and is responsible for binding to the thioreductase ERp57, which together with CRT and CNX forms inter or intra disulfide bonds with the substrate.¹⁶¹⁹ The Ca²⁺ storage domain is in the C-domain of CRT.² We performed domain deletion mutagenesis of CRT to determine whether the inhibition of Ca²⁺ oscillations was due to Ca²⁺ storage or lectin interaction with SERCA2b, which has a consensus site for glycosylation on the 11^th transmembrane segment. Mutant ΔC contains the N+P domains, but lacks the C domain. Similarly, the ΔP mutant has the N+C domain and lacks the P-domain, while the ΔPC mutant only carries the N-domain. All mutants were made with a KDEL ER retention signal and we confirmed their localization to the ER by confocal immunofluorescence.⁷ We found that inhibition of Ca²⁺ oscillations in CRT overexpressing oocytes requires the N+P domain.⁷ Our initial bias had been that overexpression of CRT would increase the amount of releasable Ca²⁺ from the ER by working as a Ca²⁺ storage protein. However, deletion of the C-domain did not remove the inhibitory action of CRT on Ca²⁺ oscillations, clearing demonstrating that the Ca²⁺ storage properties of CRT were not responsible for our observed effects. Rather, we thought that the inhibitory activity of CRT on Ca²⁺ oscillations may be mediated by a lectin-like interaction between CRT and the pump since the N+P domains are required. This model is also consistent with the inhibitory effect of CNX on Ca²⁺ oscillations as discussed further below.

Inhibition of Ca²⁺ Oscillations Is Mediated by the COOH Terminus of SERCA2b

The SERCA2 gene generates two alternative spliced products that are expressed in a tissue and developmental specific manner.²⁰²⁴ SERCA2a, the heart isoform²⁴ is shorter having only the prototypical 10 transmembrane (TM) segments. SERCA2b is ubiquitously expressed ²⁴ and deviates structurally from other Ca²⁺ ATPases by having an 11^th TM segment. The COOH terminus of SERCA terminates in the ER lumen with a glycosylation consensus signal at asparagine N1036 ²⁵. To test the hypothesis that glycosylation was required for pump inhibition, we first performed experiments in which SERCA2a or SERCA2b were co-expressed with CRT.⁹ Our prediction was that CRT would interact with the COOH terminus of SERCA2b causing inhibition of Ca²⁺ oscillations whereas oscillations should not be affected by co-expression of the CRT with SERCA2a, which only has 10 TM segments. Indeed this was the case. Ca²⁺ oscillations were similar in all respects regardless of whether the oocytes overexpressed either SERCA2a alone or SERCA2a + CRT.⁹ On the other hand Ca²⁺ oscillations were inhibited when SERCA2b was expressed with CRT.^7,9 This critical finding suggested that a direct interaction between CRT and the COOH terminus of SERCA2b was responsible for the luminal modulation of the pump. To test this hypothesis further, we created a site directed mutant of SERCA2b (SERCA2b-N1036A) in which the asparagine was mutated to alanine. Two groups of oocytes were overexpressed, those coexpressing SERCA2b-N1036A + ΔC mutant and oocytes expressing SERCA2b-N1036A alone.⁹ We found that N1036 was absolutely required for the inhibitory effect of CRT. These results, together with the findings from the another group of investigators demonstrate that progressive deletion of the COOH terminus of SERCA2b converts SERCA2b to a SERCA2a^22,23 suggest that the COOH terminus of SERCA2b is critical in determining the differences between the two SERCA2 Ca²⁺ ATPases. Furthermore the findings suggest that an interaction of CRT and CNX at the COOH terminus of SERCA2b may require N1036 glycosylation.

Interaction of CNX with the COOH Terminus of SERCA2b

CNX behaves similarly to CRT in the control of Ca²⁺ oscillations.⁸ Since it has a similar P-like domain in the lumen, but no Ca²⁺ binding domain there, the inhibition of Ca²⁺ oscillations is consistent with a glycosylation mediated effect. To test this hypothesis, we determined first whether SERCA2b was indeed glycosylated. We generated constructs for expression of the TM9 to TM11 of SERCA2b (SERCA2b/TM9-11) as well as the equivalent mutant with the 1036A mutation. In addition, we generated a SERCA2a/TM9-10 mutant. Correct polytopic insertion in the ER membrane of similar constructs was previously demonstrated.²⁵ We performed in vitro translations in rabbit reticulosite lysate in the presence of canine pancreatic microsomes. All constructs ran at the predicted molecular mass (SERCA2b and SERCA2b-N1036A/TM9-11, ˜13.2 kD; and SERCA2a ˜7.2 kD). When run on SDS-PAGE, we observed no migrational difference between SERCAb/TM9-11 and SERCA2b-N1036/ TM9-11, despite the fact that a positive control, S. cerevisiae α factor displayed full glycosylation at three sites.⁸ This indicated that N1036 was not glycosylated. To corroborate this finding, we treated In vitro translated products SERCAb/TM9-11 and SERCA2b-N1036/TM9-11 with endoglycosidase H (endo H). Addition of EndoH did not alter the mobility shift of SERCAb/ TM9-11 and SERCA2b-N1036/TM9-11 despite the fact that the enzyme caused a complete downward shift of positive control S. cerevisiae α factor, which is glycosylated in three asparagines.⁸ This finding may not yet rule out glycosylation at this N1036 and more sophisticated analysis is required to conclusively demonstrate glycosylation. Further more rigorous analysis needs to be completed to demonstrate whether SERCA2b is glycosylated at N1036. Irrespective of the state of glycosylation of N1036, it is still possible that CNX engages the SERCA2b substrate via a peptide-peptide interaction. Indeed, both CRT and CNX are known to bind to substrates via peptide-peptide interactions as traditional chaperones do.^26,27 To determine if an interaction existed between CNX and the COOH tail of SERCA2b, endogenous CNX from the microsomes were immunoprecipitated by a CNX-specific antibody from the in vitro translated products of SERCA2a/TM9-10 or SERCA2b/TM9-11. Co-immunoprecipitated proteins were subsequently detected by fluorography. We demonstrated an interaction of CNX with SERCA2b, a very reduced interaction with SERCA2b-N1036A and as expected, no interaction with SERCA2a.⁸ Furthermore, the controls behaved as expected, i.e., S. cerevisiae α factor was shown to interact with CNX and the negative control (no mRNA supplemented in the in vitro translation reaction) did not show an interaction. Together, these results suggest that there is a specific interaction involving CNX and SERCA 2b but not with SERCA2a. Further, this interaction is localized at the COOH tail of SERCA2b. These observations combined with our imaging data suggest that the differences between SERCA2a and SERCA2b are due to an interaction of CNX with the COOH terminus of SERCA2b and strengthen our view that the inhibition of Ca²⁺ oscillations is dependent on this interaction.

A PKC Phosphorylation Site in CNX Regulates Inhibition of Ca²⁺ Oscillations

CNX is a Type I, single pass transmembrane segment protein with multiple phosphorylation sites in the cytosol. These sites fit the consensus sequences for protein kinase C (PKC/PDK) and casein kinase 2 (CK2).^28,29 The PKC sites provided the attractive possibility that activation of the IP₃R signaling system could directly regulate their phosphorylation by activating a Ca²⁺ sensitive PKC. The PKC/PDK sites would then serve as a regulatory switch in the control of Ca²⁺ oscillations or in the control of glycoprotein binding to substrates in the ER. Our strategy to test this hypothesis was to mutate the phosphorylation sites and overexpress the constructs in oocytes. Two PKC/PDK sites are present in CNX, S562 and S485. Mutations to alanine were made singly and at both sites. When co-expressed with SERCA2b, we found that the COOH terminal mutant (CNX-S562) did not inhibit Ca²⁺ oscillations. This result indicated that phosphorylation controlled the interaction of CNX with the pump and supported our model that phosphorylation of this site behaved as a regulatory switch controlling Ca²⁺ oscillations.⁸ To corroborate this point, a cytosolic peptide spanning the S562 site was generated (CNXcyt) to compete with the endogenous kinase controlling the inhibition of Ca²⁺ oscillations. When this peptide was co-expressed with SERCA2b and CNX, it prevented the inhibition of Ca²⁺ oscillations, suggesting that it was efficiently acting as a pseudosubstrate for the endogenous kinase that otherwise was responsible for phosphorylating the PKC/PDK sites.⁸ The identity of this kinase appears to be PDK.²⁹

We determined the state of phosphorylation of CNX at rest and under conditions of IP₃-mediated mobilization of Ca²⁺. CNX + SERCA2b or CNX-S562A mutant + SERCA2b, or CNX + SERCA2b + CNXcyt. were overexpressed in oocytes. We prelabelled oocytes with [γ³²P]ATP and injected them with either water (control) or IP₃ (300 nM final) to mobilize Ca²⁺ from intracellular stores. CNX was then immunoprecipitated with a specific antibody and found to be phosphorylated at rest and dephosphorylated by IP₃ Ca²⁺ release.8 Importantly, immunoprecipitation of CNX from oocytes overexpressing CNX-S562A mutant + SERCA2b showed that CNX was minimally phosphorylated at rest and this level of phosphorylation was not changed by mutagenesis, suggesting that S562 supports phosphorylation and it is dephosphorylated by a Ca²⁺ dependent phosphatase.⁸ Ongoing work in our laboratory has indicated that the identity of this phosphatase is calcineurin (CN), which was initially recognized as Ca²⁺ dependent by Klee and co-workers.³⁰ These data suggest that CN dephosphorylates CNX and that this dephosphorylation controls the dissociation of CNX with SERCA2b.

The studies described here have focused on the “SERCA” hypothesis. Based on our data, CRT (unpublished data) and CNX⁸ are associated with SERCA2b and inhibit its activity when Ca²⁺ stores are full (i.e., under resting conditions). In this state, the pump is sufficiently active to maintain the ER lumen at full Ca²⁺ capacity. This environment is optimal for protein folding given the requirement of ER chaperone activity for Ca²⁺.³¹ When the IP₃ signaling pathway is activated several rapid changes occur. First, IP₃-mediated Ca²⁺ release depletes Ca²⁺ from the ER with a corresponding mirror image increase in the cytosol. These cytosolic increases cause activation of the Ca²⁺ dependent phosphatase CN, which de-phosphorylates CNX. This results in the dissociation of CNX from SERCA2b, removing pump inhibition. The return to maximum pumping activity rapidly refills the ER lumen and minimizes the potential risk of impaired protein folding during cytosolic Ca²⁺ signaling. In our view, the role of CRT/CNX regulation of SERCA2b activity is to minimize the duration of ER Ca²⁺ depletion (see Fig. 1).

Figure 1

Model of CRT/CNX regulation of ER Ca²⁺ signaling. Adapted from Roderick et al 2001.

References

1.

Missiaen L, Taylor CW, Berridge MJ. Luminal Ca²⁺ promoting spontaneous Ca²⁺ release from inositol trisphosphate-sensitive stores in rat hepatocytes. J Physiol. 1992;455:623–40. [PMC free article: PMC1175662] [PubMed: 1484365]

2.

Baksh S, Michalak M. Expression of calreticulin in Escherichia coli and identification of its Ca²⁺binding domains. J Biol Chem. 1991;266:21458–65. [PubMed: 1939178]

3.

Fliegel L, Burns K, MacLennan DH. et al. Molecular cloning of the high affinity calcium-binding protein (calreticulin) of skeletal muscle sarcoplasmic reticulum. J Biol Chem. 1989;264(36):21522–8. [PubMed: 2600080]

4.

Milner RE, Baksh S, Shemanko C. et al. Calreticulin, and not calsequestrin, is the major calcium binding protein of smooth muscle sarcoplasmic reticulum and liver endoplasmic reticulum. J Biol Chem. 1991;266(11):7155–65. [PubMed: 2016321]

5.

Camacho P, Lechleiter JD. Xenpus oocytes as a tool in calcium signaling researchIn: Putney J, ed.Calcium Signaling Boca Raton: CRC Press,2000157–81.

6.

Jouaville LS, Ichas F, Holmuhamedov EL. et al. Synchronization of calcium waves by mitochondrial substrates in Xenopus laevis oocytes. Nature. 1995;377(6548):438–41. [PubMed: 7566122]

7.

Camacho P, Lechleiter JD. Calreticulin inhibits repetitive intracellular Ca²⁺ waves. Cell. 1995;82(5):765–71. [PubMed: 7671304]

8.

Roderick HL, Lechleiter JD, Camacho P. Cytosolic phosphorylation of calnexin controls intracellular Ca(²⁺) oscillations via an interaction with SERCA2b. J Cell Biol. 2000;149(6):1235–48. [PMC free article: PMC2175122] [PubMed: 10851021]

9.

John LM, Lechleiter JD, Camacho P. Differential modulation of SERCA2 isoforms by calreticulin. J Cell Biol. 1998;142(4):963–73. [PMC free article: PMC2132884] [PubMed: 9722609]

10.

Camacho P, Lechleiter JD. Increased frequency of calcium waves in Xenopus laevis oocytes that express a calcium-ATPase. Science. 1993;260(5105):226–9. [PubMed: 8385800]

11.

Bergeron JJ, Brenner MB, Thomas DY. et al. Calnexin: a membrane-bound chaperone of the endoplasmic reticulum. Trends Biochem Sci. 1994;19(3):124–8. [PubMed: 8203019]

12.

Ellgaard L, Helenius A. ER quality control: towards an understanding at the molecular level. Curr Opin Cell Biol. 2001;13(4):431–7. [PubMed: 11454449]

13.

Jakob CA, Chevet E, Thomas DY. et al. Lectins of the ER quality control machinery. Results Probl Cell Differ. 2001;33:1–17. [PubMed: 11190669]

14.

Helenius A, Trombetta ES, Hebert DN. et al. Calnexin, Calreticulin, and the folding of glycoproteins. Trends Cell Biol. 1997;7:193–200. [PubMed: 17708944]

15.

Michalak M, Milner RE, Burns K. et al. Calreticulin. Biochem J. 1992;285(Pt 3):681–92. [PMC free article: PMC1132847] [PubMed: 1497605]

16.

Schrag JD, Bergeron JJ, Li Y. et al. The Structure of calnexin, an ER chaperone involved in quality control of protein folding. Mol Cell. 2001;8(3):633–44. [PubMed: 11583625]

17.

Ellgaard L, Riek R, Herrmann T. et al. NMR structure of the calreticulin P-domain. Proc Natl Acad Sci USA. 2001;98(6):3133–8. [PMC free article: PMC30619] [PubMed: 11248044]

18.

Ellgaard L, Riek R, Braun D. et al. Three-dimensional structure topology of the calreticulin P-domain based on NMR assignment. FEBS Letters. 2001;488(1-2):69–73. [PubMed: 11163798]

19.

Frickel EM, Riek R, Jelesarov I. et al. TROSY-NMR reveals interaction between ERp57 and the tip of the calreticulin P-domain. Proc Natl Acad Sci USA. 2002;99(4):1954–9. [PMC free article: PMC122301] [PubMed: 11842220]

20.

Gunteski-Hamblin AM, Greeb J, Shull GE. A novel Ca²⁺ pump expressed in brain, kidney, and stomach is encoded by an alternative transcript of the slow-twitch muscle sarcoplasmic reticulum Ca-ATPase gene. Identification of cDNAs encoding Ca²⁺ and other cation-transporting ATPases using an oligonucleotide probe derived from the ATP-binding site. J Biol Chem. 1988;263(29):15032–40. [PubMed: 2844797]

21.

Lytton J, Westlin M, Burk SE. et al. Functional comparisons between isoforms of the sarcoplasmic or endoplasmic reticulum family of calcium pumps. J Biol Chem. 1992;267(20):14483–9. [PubMed: 1385815]

22.

Verboomen H, Wuytack F, De Smedt H. et al. Functional difference between SERCA2a and SERCA2b Ca²⁺ pumps and their modulation by phospholamban. Biochem J. 1992;286(Pt 2):591–5. [PMC free article: PMC1132938] [PubMed: 1326945]

23.

Verboomen H, Wuytack F, Van den Bosch L. et al. The functional importance of the extreme C-terminal tail in the gene 2 organellar Ca(2+)-transport ATPase (SERCA2a/b). Biochem J. 1994;303(Pt 3):979–84. [PMC free article: PMC1137642] [PubMed: 7980471]

24.

Wu KD, Lee WS, Wey J. et al. Localization and quantification of endoplasmic reticulum Ca(2+)-ATPase isoform transcripts. Am J Physiol. 1995;269(3Pt 1):C775–84. [PubMed: 7573409]

25.

Bayle D, Weeks D, Sachs G. The membrane topology of the rat sarcoplasmic and endoplasmic reticulum calcium ATPases by in vitro translation scanning. J Bio Chem. 1995;270(43):25678–84. [PubMed: 7592746]

26.

Ihara Y, Cohen-Doyle MF, Saito Y. et al. Calnexin discriminates between protein conformational states and functions as a molecular chaperone in vitro. Mol Cell. 1999;4(3):331–41. [PubMed: 10518214]

27.

Saito Y, Ihara Y, Leach MR. et al. Calreticulin functions in vitro as a molecular chaperone for both glycosylated and non-glycosylated proteins. EMBO J. 1999;18(23):6718–29. [PMC free article: PMC1171734] [PubMed: 10581245]

28.

Tjoelker LW, Seyfried CE, Eddy RL Jr. et al. Human, mouse, and rat calnexin cDNA cloning: identification of potential calcium binding motifs and gene localization to human chromosome 5. Biochemistry. 1994;33(11):3229–36. [PubMed: 8136357]

29.

Wong HN, Ward MA, Bell AW. et al. Conserved in vivo phosphorylation of calnexin at casein kinase II sites as well as a protein kinase C/proline-directed kinase site. J Biol Chem. 1998;273(27):17227–35. [PubMed: 9642293]

30.

Klee CB, Ren H, Wang X. Regulation of the calmodulin-stimulated protein phosphatase, calcineurin. J Biol Chem. 1998;273(22):13367–70. [PubMed: 9593662]

31.

Corbett EF, Oikawa K, Francois P. et al. Ca²⁺ regulation of interactions between endoplasmic reticulum chaperones. J Biol Chem. 1999;274(10):6203–11. [PubMed: 10037706]