Limits...
Secretion of protein disulphide isomerase AGR2 confers tumorigenic properties

View Article: PubMed Central - PubMed

ABSTRACT

The extracellular matrix (ECM) plays an instrumental role in determining the spatial orientation of epithelial polarity and the formation of lumens in glandular tissues during morphogenesis. Here, we show that the Endoplasmic Reticulum (ER)-resident protein anterior gradient-2 (AGR2), a soluble protein-disulfide isomerase involved in ER protein folding and quality control, is secreted and interacts with the ECM. Extracellular AGR2 (eAGR2) is a microenvironmental regulator of epithelial tissue architecture, which plays a role in the preneoplastic phenotype and contributes to epithelial tumorigenicity. Indeed, eAGR2, is secreted as a functionally active protein independently of its thioredoxin-like domain (CXXS) and of its ER-retention domain (KTEL), and is sufficient, by itself, to promote the acquisition of invasive and metastatic features. Therefore, we conclude that eAGR2 plays an extracellular role independent of its ER function and we elucidate this gain-of-function as a novel and unexpected critical ECM microenvironmental pro-oncogenic regulator of epithelial morphogenesis and tumorigenesis.

Doi:: http://dx.doi.org/10.7554/eLife.13887.001

No MeSH data available.


AGR2 expression does not alter the secretory pathway and AGR2 secreted is not O-GlcNacylated.(A) Proteins in the intracellular lysate in H23 organoids silenced (Sh-AGR2) or not (Sh-ctl) for AGR2 were subjected to immunoblotting with anti-AGR2 antibody. GAPDH is shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (B) Proteins in the intracellular lysate in HBEC infected with either empty vector (EV) or AGR2, as analyzed by immunoblot with anti-AGR2 antibody. GAPDH concentrations are shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (C) Extracellular AGR2 from medium organoid lysates of HBEC organoids infected with either empty vector (EV) or AGR2 was immunoprecipitated with anti-AGR2 antibody and immunoblotted with anti-AGR2 or anti-O-GlcNAc antibodies. Intracellular organoid lysates were also probed with anti-AGR2 and anti-O-GlcNAc antibodies. One representative experiment (n = 3) is shown.DOI:http://dx.doi.org/10.7554/eLife.13887.009
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4940162&req=5

fig5s1: AGR2 expression does not alter the secretory pathway and AGR2 secreted is not O-GlcNacylated.(A) Proteins in the intracellular lysate in H23 organoids silenced (Sh-AGR2) or not (Sh-ctl) for AGR2 were subjected to immunoblotting with anti-AGR2 antibody. GAPDH is shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (B) Proteins in the intracellular lysate in HBEC infected with either empty vector (EV) or AGR2, as analyzed by immunoblot with anti-AGR2 antibody. GAPDH concentrations are shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (C) Extracellular AGR2 from medium organoid lysates of HBEC organoids infected with either empty vector (EV) or AGR2 was immunoprecipitated with anti-AGR2 antibody and immunoblotted with anti-AGR2 or anti-O-GlcNAc antibodies. Intracellular organoid lysates were also probed with anti-AGR2 and anti-O-GlcNAc antibodies. One representative experiment (n = 3) is shown.DOI:http://dx.doi.org/10.7554/eLife.13887.009

Mentions: To further characterize the extent of AGR2 secretion, a ratiometric comparison between intracellular and extracellular AGR2 proteins was performed (Figure 5C–D). AGR2 protein was immunoprecipitated from both intracellular and extracellular organoid extracts and analyzed by Western blotting (Figure 5C). In adenocarcinoma organoids (H23), we found ~20% of eAGR2 and ~80% of iAGR2 (Figure 5D), while in non-tumor organoids (HBEC-EV), AGR2 protein was exclusively intracellular (Figure 5D). In contrast, AGR2 overexpression in non-tumor organoids (HBEC-AGR2) forced the secretion of eAGR2 (~30% in the extracellular medium) (Figure 5D). Next, to test if the distribution of AGR2 could modify other components of the secretory pathway, a ratiometric comparison of both BiP and CANX was performed in tumor organoids (Sh-ctl) as compared to tumor organoids silenced for AGR2 (Sh-AGR2) (Figure 5E and Figure 5—figure supplement 1A). Results showed that of AGR2 depletion had no impact on BiP and CANX expression. Similarly, we found no difference in the expression of BiP and CANX between non-tumor organoids overexpressing AGR2 (HBEC-AGR2) or not (HBEC-EV) (Figure 5F and Figure 5-—figure supplement 1B) indicating that the secretory pathway was not altered upon changes in AGR2 expression. To further correlate the amount of AGR2 produced by tumor organoids to the capacity of cancer cells to secrete eAGR2, we transiently blocked protein synthesis using a cycloheximide (CHX) pulse-chase approach and evaluated the amounts of iAGR2 and eAGR2 in tumor organoids. After 8 hr of CHX treatment, eAGR2 was practically undetectable (Figure 5G–H), whereas ~60% of iAGR2 was still present. These results demonstrate that CHX treatment significantly reduces AGR2 protein synthesis which impacts on AGR2 protein secretion. Indeed, extracellular AGR2 rapidly disappeared from the culture medium through yet unknown mechanisms and became undetectable within 8 hr of CHX treatment, due to the attenuated iAGR2 synthesis and the subsequent reduction of the secreted pool. Although we have not yet identified the mechanisms by which eAGR2 disappears from the medium, several hypothesis could be postulated concerning its extracellular fate. As such the secreted protein could be i) trapped in the ECM and become unavailable as a free protein in the medium; ii) aggregated and therefore remain in the unsoluble fraction before biochemical analysis; iii) degraded by extracellular proteases released by the cell or even iv) internalized by the cells and degraded through an endolysosomal pathway. Thus, the precise mechanisms by which eAGR2 disappears remain to be further investigated.


Secretion of protein disulphide isomerase AGR2 confers tumorigenic properties
AGR2 expression does not alter the secretory pathway and AGR2 secreted is not O-GlcNacylated.(A) Proteins in the intracellular lysate in H23 organoids silenced (Sh-AGR2) or not (Sh-ctl) for AGR2 were subjected to immunoblotting with anti-AGR2 antibody. GAPDH is shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (B) Proteins in the intracellular lysate in HBEC infected with either empty vector (EV) or AGR2, as analyzed by immunoblot with anti-AGR2 antibody. GAPDH concentrations are shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (C) Extracellular AGR2 from medium organoid lysates of HBEC organoids infected with either empty vector (EV) or AGR2 was immunoprecipitated with anti-AGR2 antibody and immunoblotted with anti-AGR2 or anti-O-GlcNAc antibodies. Intracellular organoid lysates were also probed with anti-AGR2 and anti-O-GlcNAc antibodies. One representative experiment (n = 3) is shown.DOI:http://dx.doi.org/10.7554/eLife.13887.009
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4940162&req=5

fig5s1: AGR2 expression does not alter the secretory pathway and AGR2 secreted is not O-GlcNacylated.(A) Proteins in the intracellular lysate in H23 organoids silenced (Sh-AGR2) or not (Sh-ctl) for AGR2 were subjected to immunoblotting with anti-AGR2 antibody. GAPDH is shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (B) Proteins in the intracellular lysate in HBEC infected with either empty vector (EV) or AGR2, as analyzed by immunoblot with anti-AGR2 antibody. GAPDH concentrations are shown as loading control. The transmembrane protein Calnexin (CANX) and the KDEL containing protein BiP are shown as control ER-resident proteins in intracellular lysates. One representative experiment (n = 3) is shown. (C) Extracellular AGR2 from medium organoid lysates of HBEC organoids infected with either empty vector (EV) or AGR2 was immunoprecipitated with anti-AGR2 antibody and immunoblotted with anti-AGR2 or anti-O-GlcNAc antibodies. Intracellular organoid lysates were also probed with anti-AGR2 and anti-O-GlcNAc antibodies. One representative experiment (n = 3) is shown.DOI:http://dx.doi.org/10.7554/eLife.13887.009
Mentions: To further characterize the extent of AGR2 secretion, a ratiometric comparison between intracellular and extracellular AGR2 proteins was performed (Figure 5C–D). AGR2 protein was immunoprecipitated from both intracellular and extracellular organoid extracts and analyzed by Western blotting (Figure 5C). In adenocarcinoma organoids (H23), we found ~20% of eAGR2 and ~80% of iAGR2 (Figure 5D), while in non-tumor organoids (HBEC-EV), AGR2 protein was exclusively intracellular (Figure 5D). In contrast, AGR2 overexpression in non-tumor organoids (HBEC-AGR2) forced the secretion of eAGR2 (~30% in the extracellular medium) (Figure 5D). Next, to test if the distribution of AGR2 could modify other components of the secretory pathway, a ratiometric comparison of both BiP and CANX was performed in tumor organoids (Sh-ctl) as compared to tumor organoids silenced for AGR2 (Sh-AGR2) (Figure 5E and Figure 5—figure supplement 1A). Results showed that of AGR2 depletion had no impact on BiP and CANX expression. Similarly, we found no difference in the expression of BiP and CANX between non-tumor organoids overexpressing AGR2 (HBEC-AGR2) or not (HBEC-EV) (Figure 5F and Figure 5-—figure supplement 1B) indicating that the secretory pathway was not altered upon changes in AGR2 expression. To further correlate the amount of AGR2 produced by tumor organoids to the capacity of cancer cells to secrete eAGR2, we transiently blocked protein synthesis using a cycloheximide (CHX) pulse-chase approach and evaluated the amounts of iAGR2 and eAGR2 in tumor organoids. After 8 hr of CHX treatment, eAGR2 was practically undetectable (Figure 5G–H), whereas ~60% of iAGR2 was still present. These results demonstrate that CHX treatment significantly reduces AGR2 protein synthesis which impacts on AGR2 protein secretion. Indeed, extracellular AGR2 rapidly disappeared from the culture medium through yet unknown mechanisms and became undetectable within 8 hr of CHX treatment, due to the attenuated iAGR2 synthesis and the subsequent reduction of the secreted pool. Although we have not yet identified the mechanisms by which eAGR2 disappears from the medium, several hypothesis could be postulated concerning its extracellular fate. As such the secreted protein could be i) trapped in the ECM and become unavailable as a free protein in the medium; ii) aggregated and therefore remain in the unsoluble fraction before biochemical analysis; iii) degraded by extracellular proteases released by the cell or even iv) internalized by the cells and degraded through an endolysosomal pathway. Thus, the precise mechanisms by which eAGR2 disappears remain to be further investigated.

View Article: PubMed Central - PubMed

ABSTRACT

The extracellular matrix (ECM) plays an instrumental role in determining the spatial orientation of epithelial polarity and the formation of lumens in glandular tissues during morphogenesis. Here, we show that the Endoplasmic Reticulum (ER)-resident protein anterior gradient-2 (AGR2), a soluble protein-disulfide isomerase involved in ER protein folding and quality control, is secreted and interacts with the ECM. Extracellular AGR2 (eAGR2) is a microenvironmental regulator of epithelial tissue architecture, which plays a role in the preneoplastic phenotype and contributes to epithelial tumorigenicity. Indeed, eAGR2, is secreted as a functionally active protein independently of its thioredoxin-like domain (CXXS) and of its ER-retention domain (KTEL), and is sufficient, by itself, to promote the acquisition of invasive and metastatic features. Therefore, we conclude that eAGR2 plays an extracellular role independent of its ER function and we elucidate this gain-of-function as a novel and unexpected critical ECM microenvironmental pro-oncogenic regulator of epithelial morphogenesis and tumorigenesis.

Doi:: http://dx.doi.org/10.7554/eLife.13887.001

No MeSH data available.