Pathways regulating the trafficking and turnover of pannexin1 protein and the role of the C-terminal domain - PubMed (original) (raw)

Pathways regulating the trafficking and turnover of pannexin1 protein and the role of the C-terminal domain

Ruchi Gehi et al. J Biol Chem. 2011.

Abstract

Pannexin1 (Panx1) is an integral membrane protein comprised of three species as follows: an unglycosylated core-Gly0, a high mannose-Gly1, and a complex glycosylated Gly2 species. Although Panx1 channels mediate several cellular responses, the domain regulating its oligomerization and cell surface trafficking and the mechanisms governing its internalization and degradation have not been identified. This study characterizes the role of the Panx1 C-tail domain by truncating the polypeptide at residue 307 and expressing the mutant in BICR-M1R(k) and HEK-293T cells. Enzymatic digestion and immunolabeling assays revealed that the Panx1(T307)-RFP was glycosylated primarily to the high mannose species consistent with its retention in the endoplasmic reticulum. Co-expression of Panx1(T307)-RFP with Panx1 followed by co-immunoprecipitation assays revealed that the mutant and Panx1 could interact, whereas biotinylation assays showed that this interaction inhibited Panx1 from maturing into the Gly2 species and reaching the cell surface. Additional inhibitor studies indicated that the degradation of the mutant was via proteasomes, whereas Panx1 was degraded by lysosomes. Analysis of the pathways important in Panx1 internalization revealed partial co-distribution of Panx1 with many molecular constituents of the endocytic machinery that include clathrin, AP2, dynamin II, caveolin-1, and caveolin-2. However, co-immunoprecipitation assays together with the disruption of lipid rafts by methyl-β-cyclodextrin suggest that Panx1 does not engage this endocytic machinery. Furthermore, dominant-negative and pharmacological studies revealed that Panx1 internalization was dynamin II-independent. Collectively, these results indicate that the oligomerization and trafficking of Panx1 are regulated by the C-terminal domain, whereas internalization of long lived Panx1 channels occurs in a manner that is distinct from classical endocytic pathways.

PubMed Disclaimer

Figures

FIGURE 1.

Truncated Panx1 fails to reach the cell surface. When expressed in BICR-M1Rk cells, Panx1T307-RFP was localized to intracellular compartments (A, red), and Panx1-RFP was found at the cell surface (B, red). Immunolabeling with anti-RFP antibody detected both Panx1-RFP and Panx1T307-RFP (A and B, green). Furthermore, protein lysates from HEK-293T WT cells and cells overexpressing Panx1-GFP, Panx1-RFP, or Panx1T307-RFP were immunoblotted with either anti-RFP antibody (C) or anti-Panx1 antibody (D). Panx1T307-RFP was detected primarily as a doublet of ∼50 kDa (C), and Panx1-RFP (C and D) was resolved similar to Panx1-GFP (D) in a multiple banding profile above ∼75 kDa. GAPDH was used as a protein loading control. Protein lysates of HEK-293T cells labeled with (+) or without (−) biotin were precipitated with NeutrAvidin beads prior to immunoblotting for RFP and Panx1. Cell surface biotinylation of Panx1-RFP revealed that all glycosylated species of Panx1 trafficked to the cell surface in the NeutrAvidin pulled samples (F), and no clear cell surface expression of Panx1T307-RFP was detected (E). GAPDH was used as a control to detect any internalization of biotin during the labeling procedure.

FIGURE 2.

Panx1T307-RFP is primarily retained in the ER and is glycosylated to a high mannose species. Cells ectopically expressing Panx1T307-RFP (red) or Panx1-RFP (red) were immunolabeled with antibodies against PDI, calnexin, or GM130. The distribution of Panx1T307-RFP primarily overlapped with the resident proteins of the endoplasmic reticulum-PDI and calnexin (A and B, merge) but not the perinucleus-localized Golgi marker GM130 (C, green). The bulk of Panx1-RFP trafficked to the cell surface with no obvious co-localization with PDI (A, merge) or calnexin (B, merge). Digestion of cell lysates containing Panx1-RFP or Panx1T307-RFP with _N_-glycosidase F (D and E) and Endo H (F and G) revealed that although Panx1-RFP can exist as a core (Gly0), high mannose (Gly1), and complex glycoprotein species (Gly2), Panx1T307-RFP primarily consisted of core (Gly0) and high mannose (Gly1) species. GAPDH was used as a protein loading control.

FIGURE 3.

Panx1T307-RFP reduces the cell surface localization of Panx1. HEK-293T cells transfected with Panx1 alone (A) or with Panx1-RFP (B, red) and Panx1T307-RFP (C, red) were immunolabeled for Panx1 (green). When expressed alone (A) or with Panx1-RFP (B), Panx1 reached the cell surface and also co-localized with Panx1-RFP (B, merge). The intracellular localization pattern of Panx1T307-RFP (red) was evident in the presence of Panx1 (green) (C). Nuclei were counterstained with Hoechst (blue). Bars, 10 μm. Protein lysates of HEK-293T cells labeled with (+) or without (−) biotin were precipitated with NeutrAvidin beads prior to immunoblotting for RFP and Panx1. Cell surface biotinylation failed to detect Panx1T307-RFP even when co-expressed with Panx1 (D); however, Panx1T307-RFP expression reduced the overall cell surface level of Panx1 (E). The NeutrAvidin fraction contained both Panx1-RFP and Panx1 when they were co-expressed (E). Total protein lysates labeled with (+) or without (−) biotin were also assessed for the expression levels of Panx1, Panx1-RFP, and Panx1T307-RFP. GAPDH was used to detect any internalization of biotin during the labeling procedure.

FIGURE 4.

Truncated Panx1 has limited interaction with full-length Panx1. Immunoprecipitation of Panx1 was performed on cell lysates from WT HEK-293T cells and cells ectopically expressing either Panx1 alone or with Panx1-RFP and Panx1T307-RFP. A robust interaction of the Gly1 and Gly2 species of Panx1-RFP with Panx1 was revealed when immunoblotted for RFP (A) and Panx1 (B). Panx1T307-RFP also co-immunoprecipitated with Panx1; however, it was mainly the Gly1 species that was detected (A). Co-immunoprecipitation of Panx1 with Panx1T307-RFP revealed reduced interaction with the Gly2 species of Panx1 (B). Expression levels of all Panx1 variants were also assessed by immunolabeling protein lysates for RFP or Panx1. IP, immunoprecipitation; IB, immunoblotting.

FIGURE 5.

Panx1T307-RFP is primarily targeted for proteasome-mediated degradation. Panx1T307-RFP- and Panx1-RFP-expressing HEK-293T cells were exposed to either proteasomal (lactacystin) or lysosomal (chloroquine) inhibitors for 20 h prior to immunoblotting for RFP or Panx1. Expression of both Gly0 and Gly1 species of Panx1T307-RFP increased with lactacystin exposure, with no detectable change upon chloroquine treatment (A). Lactacystin-treated cells expressing Panx1-RFP (B) revealed a slight reduction in the expression of the Gly2 species, and chloroquine treatment resulted in the accumulation of the Gly1 and Gly2 species. When co-expressed together with Panx1, chloroquine treatment increased the Gly1 species in cells expressing Panx1T307-RFP (C).

FIGURE 6.

Panx1 is destined for lysosome-mediated degradation. HEK-293T cells expressing Panx1 together with Panx1T307-RFP (A) or BICR-M1Rk cells expressing Panx1 alone (B) were exposed to proteasomal (lactacystin) and lysosomal (chloroquine) inhibitors prior to immunoblotting for Panx1. Co-expression of Panx1 with Panx1T307-RFP clearly reduced the Gly2 form of Panx1 (A) when compared with Panx1 expressed alone (B). Panx1-expressing cells were also exposed to BFA and chloroquine alone or together for 20 h prior to immunoblotting (B) and immunolabeling (C) for Panx1. Chloroquine treatment caused a sustained expression of the Gly2 species of Panx1 with an accumulation of the Gly1 form (B), which was marked by an increased intracellular distribution of Panx1 (C, red). BFA-induced reduction of the Panx1 Gly2 species (B) was correlated by the increased loss of the cell surface population of Panx1 with a subsequent accumulation in the intracellular compartment (C, red). GAPDH was used as a protein loading control. Nuclei were counterstained with Hoechst. Bar, 10 μm.

FIGURE 7.

Panx1 co-distributes but does not co-immunoprecipitate with clathrin or AP2. Panx1 expressing BICR-M1Rk cells were immunolabeled for clathrin heavy chain (A) or anti-AP2 (B and C). Alexa Fluor488-conjugated Panx1 antibody detected Panx1 (green) with clathrin (red) at the cell surface (A, merge). Double immunofluorescent labeling revealed the co-distribution of Panx1 (green) with AP2 (red) at the cell surface before and after a 20-h BFA exposure (C). Nuclei were stained with Hoechst. Bar, 10 μm. Immunoprecipitates of Panx1 from WT and Panx1-expressing cells were immunoblotted for Panx1, clathrin, and AP2. Panx1 (∼43–50 kDa) was successfully pulled down from Panx1-expressing cells; however, clathrin (∼192 kDa) or AP2 (∼105–110 kDa) did not co-immunoprecipitate with Panx1 (D). IB, immunoblotting; IP, immunoprecipitation.

FIGURE 8.

Cell surface Panx1 is localized to a distinct compartment and does not co-immunoprecipitate with Cav-1 or Cav-1. BICR-M1Rk cells endogenously expressing Cav-1 and Cav-2 and stably expressing Panx1 were fixed with formaldehyde or acetone/methanol prior to immunolabeling. Panx1 (green) was found to co-distribute with intracellular and plasma membrane-localized Cav-1 and Cav-2 (red) (A–D). Immunoprecipitates of Panx1 (E and G) from WT and Panx1-expressing cells were immunoblotted for Panx1, Cav-1, and Cav-2. Panx1 was detected only in the immunoprecipitates of Panx1 (E and G) and not in the immunoprecipitates of Cav-1 (F) or Cav-2 (H). Conversely, Cav-1 (F) and Cav-2 (H) were found in their respective immunoprecipitates and not in the Panx1 immunoprecipitates (E and G). Treatment of Panx1-expressing cells up to 24 h of MβC did not alter the expression levels of the Panx1 glycosylation species (I); the highly phosphorylated form of Cx43 was reduced within 1 h of MβC exposure (I). The cell surface distribution of Panx1 (red) remained relatively uniform during the 24 h of MβC treatment (J). Nuclei were counterstained with Hoechst. Bar, 10 μm. GAPDH was used as a protein loading control. IB, immunoblot; IP, immunoprecipitation.

FIGURE 9.

Panx1 internalization is not dynamin II-dependent. Cells stably expressing Panx1 were engineered to express WT Dyn II or the K44A Dyn II mutant. In untreated cells, WT and K44A Dyn II partially co-distributed with Panx1 primarily at the cell surface (A and C). BFA treatment induced similar intracellular and cell surface distribution of Panx1 in the presence of WT (B) or K44A DynII (D). Bar, 10 μm. Nuclei were counterstained with Hoechst. Protein lysates of untreated and BFA-treated cells expressing Panx1 alone or in combination with WT and K44A DynII were subjected to immunoblotting using anti-Panx1 and anti-DynII antibodies. A significant reduction of the Panx1 Gly2 species and accumulation of the Gly1 species was observed when cells were exposed to BFA (E and F). The overall Panx1 expression level remained unaltered with the co-expression of WT and K44A DynII (E and F). Transient transfection of WT and K44A DynII resulted in their expression at approximately equal levels (E). In contrast to the WT DynII, temperature-sensitive uptake of transferrin-Alexa Fluor555 in live cells was greatly reduced in the presence of K44A DynII (G). Immunoprecipitates of Panx1 from WT and Panx1-expressing cells were probed with anti-Panx1 and anti-DynII antibodies. Panx1 but not DynII was detected in the immunoprecipitates of Panx1 (H). GAPDH was used as a protein loading control. For statistical analysis, ratios between glycosylation species of Panx1 (Gly0, Gly1, and Gly2) and GAPDH loading control were taken, and a one-way analysis of variance was performed followed by Tukey's test (where a is significantly different from b). The y axis reflects arbitrary numbers. Bar, 10 μm.

References

1. Panchin Y., Kelmanson I., Matz M., Lukyanov K., Usman N., Lukyanov S. (2000) Curr. Biol. 10, R473–R474 - PubMed
1. Bao L., Locovei S., Dahl G. (2004) FEBS Lett. 572, 65–68 - PubMed
1. Sridharan M., Adderley S. P., Bowles E. A., Egan T. M., Stephenson A. H., Ellsworth M. L., Sprague R. S. (2010) Am. J. Physiol. Heart Circ. Physiol. 299, H1146–H1152 - PMC - PubMed
1. Woehrle T., Yip L., Elkhal A., Sumi Y., Chen Y., Yao Y., Insel P. A., Junger W. G. (2010) Blood 116, 3475–3484 - PMC - PubMed
1. Locovei S., Wang J., Dahl G. (2006) FEBS Lett. 580, 239–244 - PubMed

Pathways regulating the trafficking and turnover of pannexin1 protein and the role of the C-terminal domain - PubMed (original) (raw)