PREreview of Identification of Glucosidase II that regulates STIM1 activation dynamics

Pre-review of “Identification of Glucosidase II that regulates STIM1 activation dynamics”

by Ketsia Bakambamba, Elodie Lafont and Eric Chevet of the Proteostasis and Cancer team INSERM U1242

Summary of the article

In this article by Du et al, the authors identify glucosidase II (GANAB and PRKCSH) as partners of STIM1. They then explore the impact of STIM1 affinity for Ca2+ and store-operated Ca2+ entry and propose glucosidase II as a new modulator of STIM1 activation.

General comments

The results obtained by the authors are interesting, yet we think that the study could benefit from additional experiments, reanalysis of existing ones and text refinement (see suggestions in specific points below).

Specific comments

Figure 1 and corresponding Figure S1:

1A. How can STIM1 be biotinylated in its cytosolic part?

1B. How long is the Tg treatment?

1C, D It is not clear whether this is obtained upon Tg stimulation (which may have introduced a bias).

Figure 1 and S1 in general: Comparison of all 4 conditions for the proteomic results (+/- Tg for STIM1 and STIM2) results would be informative here. Given the topology of the constructs used, it is a bit surprising that actin filament organization-related proteins are enriched here. What is the author’s take on this result?

Figure 2 and corresponding Figure S2:

The authors should specify that this is live microscopy.

Why not using Tg here?

What are the STIM1-associated punctae generally representing in this study: misfolded clumps or PM-ER contact sites?

Is the relocalization simply due to protein aggregation?

A negative control would be nice here: would a random transmembrane glycoprotein or a soluble cargo also colocalize with STIM1 in the different conditions used?

Why not test UGGT1 as well in Figure 2?

The experiments performed here should also be done on endogenous STIM1, PRKCSH, GANAB.

Providing a scheme depicting STIM1 domains and mutations used would be useful here.

2E-F: For the STIM1 cytosolic domain deleted protein (1-442): how does it work for forming punctae: is it really associated to PM? How would ORAI1 and STIM1 interact in this context? Is this form of STIM1 localized at th ER?

Figure S2A: STIM1 209-685 is already in punctae: how can this be explained?

S2B-D: why is the cell aspect so different? Is STIM1-C227W improperly folded?

Using a marker of ER, and staining ORAI1 for PM contact sites, would also be meaningful in these figures.

Figure 3

Western blot for comparing overexpression levels for the different constructs used in the paper would be nice.

3B: is there a signal peptide in the MRH-mScarlet construct? How does it get localized to the ER?

The punctae in this figure have different aspect than the others depicted before, in particular for the 1-442 STIM1-expressing cells at rest: how can it be explained? More generally: are the different mutants of STIM1 used in the study already affecting the morphology of the ER?

Figure 4 and corresponding Figures S3 and S4

STIM1: why use a new mutant here: what about the previous ones?

mNG△N5-SOAR1L: it is not clear from the text what exactly is the construct used here and how this experiment is done.

4C: displaying these different RNA levels on the same graph is confusing.

4D: Molecular weight ladder lacking.

4F and 4G: the differences are very slight: do the authors think this is biologically relevant?

Glucosidase inhibitor (eg castanospermine, deoxynojirimycin) could have been used in this study to assess if this activity impacts STIM1 localization and Ca2+ binding.

4E: UGGT1 should be investigated in the previous figures as well (impact on localization of STIM1?), not just in Figure 4, on should be checked in Figure 5 as well.

S4: STIM1-2NQ: It looks like there a different pattern when overexpressing GANAB or PRKCSH. Could the authors comment on that?

Evaluating the status of glycosylation and folding of STIM1 could also be done by checking its retention time on calnexin throughout the study.

Figure 5

It is not clear when/how the mean of SOCE is measured in this Figure.

5C: why is there a kinetic difference between shGANAB and shPRKCSH when they are supposed to affect the same enzymatic complex?

Are PRKCSH and other interactors calcium binding proteins? If so, their function could be affected by calcium depletion per se.

Choices of colors in the figures are not color-blind friendly.

Competing interests

The authors declare that they have no competing interests.

Competing interests

The authors declare that they have no competing interests.