Roderick HL et al. (2000), Cytosolic phosphorylation of calnexin controls ...

XB-ART-10868

J Cell Biol 2000 Jun 12;1496:1235-48. doi: 10.1083/jcb.149.6.1235.

Show Gene links Show Anatomy links

Cytosolic phosphorylation of calnexin controls intracellular Ca(2+) oscillations via an interaction with SERCA2b.

Roderick HL, Lechleiter JD, Camacho P.

???displayArticle.abstract???
Calreticulin (CRT) and calnexin (CLNX) are lectin chaperones that participate in protein folding in the endoplasmic reticulum (ER). CRT is a soluble ER lumenal protein, whereas CLNX is a transmembrane protein with a cytosolic domain that contains two consensus motifs for protein kinase (PK) C/proline- directed kinase (PDK) phosphorylation. Using confocal Ca(2+) imaging in Xenopus oocytes, we report here that coexpression of CLNX with sarco endoplasmic reticulum calcium ATPase (SERCA) 2b results in inhibition of intracellular Ca(2+) oscillations, suggesting a functional inhibition of the pump. By site-directed mutagenesis, we demonstrate that this interaction is regulated by a COOH-terminal serine residue (S562) in CLNX. Furthermore, inositol 1,4,5-trisphosphate- mediated Ca(2+) release results in a dephosphorylation of this residue. We also demonstrate by coimmunoprecipitation that CLNX physically interacts with the COOH terminus of SERCA2b and that after dephosphorylation treatment, this interaction is significantly reduced. Together, our results suggest that CRT is uniquely regulated by ER lumenal conditions, whereas CLNX is, in addition, regulated by the phosphorylation status of its cytosolic domain. The S562 residue in CLNX acts as a molecular switch that regulates the interaction of the chaperone with SERCA2b, thereby affecting Ca(2+) signaling and controlling Ca(2+)-sensitive chaperone functions in the ER.

???displayArticle.pubmedLink??? 10851021
???displayArticle.pmcLink??? PMC2175122
???displayArticle.link??? J Cell Biol
???displayArticle.grants??? [+]

Species referenced: Xenopus
Genes referenced: calr canx

???attribute.lit??? ???displayArticles.show???

	Figure 1. Overexpression of CLNX in Xenopus oocytes inhibits repetitive Ca2+ waves. (a) Confocal Ca2+ imaging in a control oocyte overexpressing rat SERCA2b shows high frequency Ca2+ oscillations after an injection of IP3 (∼300 nM final). Inhibition of Ca2+ oscillations is also shown for a representative oocyte coexpressing CLNX + SERCA2b. In this and subsequent figures, the top image contains a spatio–temporal stack of Ca2+ wave activity followed to the right by a representative confocal image of Ca2+ wave activity at the indicated timepoint. Bar, 50 μm. Also in this and subsequent figures, the trace under each image represents the change in fluorescence from resting levels (ΔF/F) shown as a function of time. The white square represents the 5 × 5 pixel area used in determining ΔF/F. (b) The percentage of oocytes exhibiting repetitive Ca2+ wave activity was decreased in oocytes coexpressing CLNX + SERCA2b (n = 63, 30%), relative to oocytes overexpressing SERCA 2b alone (n = 52, 88%) (left). The asterisk indicates statistical significance in comparison to oocytes overexpressing SERCA2b alone, P < 0.005 (Chi-square test). In this figure and subsequent histograms, the color convention used is as follows: black for SERCA 2b and light gray for CLNX + SERCA2b overexpressing oocytes. Detailed analysis of Ca2+ waves was performed only on those oocytes that exhibited repetitive Ca2+ waves. Compared with oocytes overexpressing SERCA 2b alone, overexpression of CLNX + SERCA2b increased the period between waves (middle) and the decay time of individual waves (right). The asterisk denotes a statistical significant difference P < 0.05 (t test) in comparison to control oocytes overexpressing SERCA2b alone (see also Table ).
	Figure 2. Overexpression of CLNX in Xenopus oocytes inhibits repetitive Ca2+ waves induced by mitochondrial energization. (a) Confocal Ca2+ imaging of a control oocyte (IP3 ∼300 nM final) energized previously with pyruvate malate (10 mM final). In this representative oocyte, a Ca2+ tide is followed by low-frequency Ca2+ oscillations. Ca2+ oscillations are inhibited in a representative oocyte overexpressing CLNX. Bar, 100 μm. (b) The percentage of oocytes exhibiting repetitive Ca2+ wave activity was decreased in oocytes overexpressing CLNX (n = 18, 71%), relative to control oocytes (n = 25, 84%) (left). The asterisk indicates statistical significance in comparison to oocytes overexpressing SERCA2b alone, P < 0.05 (Chi-square test). Detailed analysis of Ca2+ waves was performed only on those oocytes that exhibited repetitive Ca2+ activity. Compared with control oocytes, overexpression of CLNX increased the period between waves (middle) and the decay time of individual waves (right). The asterisk denotes a statistical significant difference P < 0.05 (t test) in comparison to control oocytes (see also Table ).
	Figure 3. CLNX does not inhibit Ca2+ oscillations induced by overexpression of the SERCA2b-N1036A mutant. Confocal Ca2+ imaging in an oocyte overexpressing SERCA2b-N1036A after an injection of IP3 (∼300 nM final) shows high-frequency Ca2+ oscillations (top). These oscillations are not inhibited in oocytes coexpressing CLNX + SERCA2b-N1036A (bottom). Bar, 50 μm.
	Figure 4. CLNX coimmunoprecipitates with the COOH terminus of SERCA2b. In vitro translations of synthetic mRNAs were performed in rabbit reticulocyte lysate supplemented with canine pancreatic microsomes and L-[35S]methionine. Samples were subjected to 15% SDS-PAGE and detected by fluorography. (a) Membrane fractions isolated from in vitro translation reactions were loaded as follows: S. cerevisiae α factor (lane 1), negative control without RNA (lane 2), SERCA2a/TM9-10 (lane 3), SERCA2b/TM9-11 (lane 4), and SERCA2b-N1036A/TM9-11 (lane 5). (b) Endo H treatment demonstrates that the COOH terminus of SERCA2b is not glycosylated. Samples were loaded as follows: negative control without mRNA (lane 1), paired samples were loaded ± endo H as follows: S. cerevisiae α factor (second set of lanes), SERCA2b/TM9-11 (third set of lanes), SERCA2b-N1036A/TM9-11 (fourth set of lanes), and SERCA2a/TM9-10 (fifth set of lanes). (c) Coimmunoprecipitations of CLNX with SERCA COOH terminus constructs demonstrates an interaction of CLNX with SERCA2b, a reduced interaction with the SERCA2b-N1036A mutant, but no interaction with SERCA2a. Lanes were loaded as follows: negative control without RNA in the translation (lane 1), S. cerevisiae α factor (lane 2), SERCA2a/TM11-10 (lane 3), SERCA2b/TM9-11 (lane 4), and SERCA2b-N1036A/TM9-11 (lane 5).
	Figure 5. Mutagenesis of consensus sites for phosphorylation by PKC/PDK in CLNX. The ER lumenal, single TM, and cytosolic domain boundaries for CLNX are indicated. Partial sequence of the cytosolic domain is shown in single letter amino acid code. PKC phosphorylation consensus sites are indicated in bold and underline. Specific phosphorylated residues are shown in larger font and depicted as filled circles. Sequence mutations generated are indicated in amino acid code and depicted by clear circles. The full sequence and position relative to wild-type CLNX is shown for the CLNXcyt cytosolic peptide. To the right of each construct, the corresponding symbol is given and is used in subsequent figures.
	Figure 6. Inhibition of repetitive Ca2+ waves is abrogated by a single amino acid mutation in serine residue 562 of CLNX. (a) Changes in fluorescence from resting levels (ΔF/F) are shown for oocytes as labeled. To the right of each trace, a single confocal image of Ca2+ wave activity at the indicated time is shown for each representative oocyte. (b) The percentage of oocytes exhibiting repetitive Ca2+ wave activity in oocytes coexpressing CLNX-S562A + SERCA2b is increased from that of the wild-type CLNX, and reaches the levels of SERCA2b alone control oocytes (depicted as a dashed line). This percentage was not significantly different in oocytes overexpressing the proximal PKC mutant CLNX-S485A + SERCA2b or the double mutant CLNX-S485A/S562A + SERCA2b when compared with oocytes overexpressing wild-type CLNX + SERCA2b. The asterisk denotes a statistical significant difference (P < 0.005, Chi-square) for the comparison of CLNX-S562A mutant to wild-type CLNX.
	Figure 7. Dominant-negative expression of a competing peptide spanning serine residue 562 of CLNX reverses the inhibition of Ca2+ oscillations by the wild-type protein. (a) ΔF/F is shown for a representative control oocyte coexpressing CLNX + SERCA2b (upper panel) and for an oocyte overexpressing CLNXcyt + CLNX + SERCA2b (lower panel). To the right of each trace, a corresponding single confocal image of Ca2+ wave activity is shown. (b) The percentage of oocytes exhibiting repetitive Ca2+ wave in oocytes coexpressing CLNX + SERCA2b + CLNXcyt is increased when compared with oocytes in which the cytosolic peptide was omitted (CLNX + SERCA2b). The asterisk indicates statistical significance of P < 0.005 (Chi-square) for a comparison between oocytes overexpressing CLNXcyt + CLNX + SERCA2b and control oocytes in which the cytosolic peptide was omitted.
	Figure 8. CLNX is dephosphorylated after IP3 injection. (a) Immunoprecipitations of [γ-32P]ATP-labeled CLNX and CLNX-S562A mutant from oocytes injected with IP3 (+) or noninjected (−) control oocytes run on SDS-PAGE (upper panel). Paired samples were loaded ± IP3 for oocytes overexpressing the following: SERCA2b alone (first set of lanes), CLNX + SERCA2b (second set of lanes), CLNX-S562A + SERCA2b (third set of lanes), and CLNXcyt + CLNX + SERCA2b (fourth set of lanes). Densitometric analysis of the phosphorylated protein bands was normalized to the intensity of the wild-type CLNX (−IP3) and plotted in arbitrary units (lower panel). (b) Immunodetection of CLNX detected by Western blotting in oocytes overexpressing the following: CLNX and CLNX mutants with SERCA2b (upper panel). The gel was loaded as follows: SERCA2b (lane 1), CLNX + SERCA2b (lane 2), CLNX-S485A/S562A + SERCA2b (lane 3), CLNX-S485A + SERCA2b (lane 4), and CLNX-S562A + SERCA2b (lane 5). Immunodetection of SERCA2b in oocytes coexpressing SERCA2b with CLNX and CLNX mutants (lower panel). The lanes correspond to protein fractions from oocytes that were injected with mRNAs encoding the following: SERCA2b (lane 1), CLNX + SERCA2b (lane 2), CLNX-S562A + SERCA2b (lane 3), CLNX-S485A + SERCA2b (lane 4), and CLNX-S485A/S562A + SERCA2b (lane 5).
	Figure 9. Dephosphorylation of CLNX reduces its interaction with SERCA2b. In vitro translations of synthetic mRNAs were performed in rabbit reticulocyte lysate supplemented with canine pancreatic microsomes and L-[35S]methionine. Isolated microsomes were treated with alkaline phosphatase as described in Materials and Methods, the samples were subjected to 15% SDS-PAGE, and detected by fluorography. A positive glycosylation control of S. cerevisiae α factor was also translated. Paired samples ± alkaline phosphatase treatment that were coimmunoprecipitated with CLNX were loaded in the following order: negative control without RNA in the translation (lane 1), S. cerevisiae α factor (second set of lanes), SERCA2b/TM9-11 (third set of lanes), and SERCA2b-N1036A/TM9-11 (fourth set of lanes).
	Figure 10. Model depicting the functional consequences of cytosolic phosphorylation of CLNX. (1) Under resting conditions, CLNX is phosphorylated on S562 located in the cytosolic domain of CLNX as shown. In this state, we suggest that CLNX is free to interact with the COOH terminus of SERCA2b. CLNX may also interact with the ribosome. The Ca2+ ATPase is inhibited, the Ca2+ stores are expected to be full (dark gray) and in an optimal condition for protein folding. (2) Mobilization of the Ca2+ stores by IP3 (light gray) results in a Ca2+-dependent dephosphorylation of S562 in CLNX. In the dephosphorylated state, inhibition of repetitive Ca2+ waves by CLNX is not observed, suggesting a loss of interaction with the pump resulting in Ca2+ store refilling. This phosphorylation switch implies that cytosolic Ca2+ regulates interactions of CLNX with lumenal proteins (e.g., SERCA2b), resulting in control of Ca2+ uptake and also modulation of protein folding in the ER lumen. In contrast, a bi-directional arrow is shown to indicate that the interaction of CRT with targets is determined by lumenal conditions. (See Discussion for further details).

References [+] :

Baksh, Expression of calreticulin in Escherichia coli and identification of its Ca2+ binding domains. 1991, Pubmed