In our experience, the detection of PERK phosphorylation (or IRE1 phosphorylation) as markers of ER stress is fraught with great difficulties. Except in rare cells that are highly enriched in their ER content, the anti-phosphoPERK antisera we have tested are unable to reliably detect the protein in straight immunoblots. To detect PERK activation we resort to the laborious procedure of immunoprecipitation of PERK from detergent lysates followed by immunoblot. This procedure is technically difficult and consumes large amount of sample and antiserum (see Heather Harding’s protocol on detecting P-PERK by sequential immunoprecipitation and blotting). Our stocks of antiserum to PERK have been depleted over the years and we are unable to share them. Fortunately, Cell Signaling Technology has come out with a monoclonal antibody (#3179) that detects phosph-PERK by direct immunoblot of lysates of stressed cells (they used AR42J cells in the example posted on the web), but other labs, noteably that of Umut Ozcan (at Children's Hospital in Boston), have succesfully used this antiserum to detect phosphorylated PERK in straight lysates from cells and tissues.

It is our opinion that in many cases use of downstream protein markers for ER stress such as CHOP, BiP and ATF4 is a better course of action. An alternative worth considering is to follow XBP-1 splicing by an RT PCR assay, for which you can download a protocol here. On occasion there will be a specific question related to PERK activation that can not be answered by the surrogate markers, however in most cases detecting PERK activation adds little and consumes a lot of time and resources.

We are unable to meet the demand for antisera to ER stress markers. However, Affinity Bioreagents has recently commercialized the 9C8 anti-CHOP MoAb that we had originally developed. They distribute IgG derived from ascites produced with this hybridoma. We have tested their product and found it suitable for immunoblots. While we have not tested it for IP, our own preparations of this MoAb have worked well in IP and ICC in the past.

Shenolikar and colleagues had published a paper (Brush et al, Mol Cell Biol. 2003 February; 23(4): 1292–1303) in which they report on the successful application of the SCBT anti-GADD34 H-193 serum to stain endogenous GADD34 in NIH 3T3 cells. This is a remarkable achievement and suggests that the antiserum is very good and detects the mouse protein. We have not tried to reproduce this observation using SCBT anti-GADD34 H-193

The Santa Cruz rabbit polyclonal antiserum to XBP-1 (sc-7160) is good too, it picks up both the human and the mouse. (Please note that our paper on XBP-1 (Calfon, et al., 2002, Nature; 415:92) contains an unfortunate error in that we used the rabbit polyclonal to XBP-1 (SC-7160) for our successful blots and not the murine monoclonal to XBP-1 (SC-8015) as stated erroneously in the paper. We were unable to detect the endogenous XBP-1 with the SC-8015 monoclonal antiserum)

Detecting phosphorylated eIF2a is not easy. We have enjoyed some success with the BioSource affinity purified rabbit antiserum (44-728G), but perfomance of the antiserum varies by lot.

Over-expression of a malfolding protein as a means to induce ER stress:

In most circumstances, we have been unable to detect the phosphorylation of PERK or IRE1 in response to the over-expression of an ER client protein. We suspect that the reason for this is the slowish rate at which ER stress ensues (compared to the more rapid induction of ER stress in cells treated with toxins that massively and synchronously perturb ER function). Therefore, we would not recommend going after P-PERK and P-IRE1 as markers of ER stress. Rather we would look at the downstream markers such as CHOP and XBP-1 protein accumulation first.