Write a comment

PREreview of “Engineering ERα degraders with pleiotropic ubiquitin ligase ligands maximizes therapeutic efficacy by co-opting distinct effector ligases”

by Mathéo Lodé, Elodie Lafont, Xavier Guillory of the Proteostasis and Cancer Team, INSERM U1242.

Summary of the article

In this article by Shemorry et al., the authors evaluate a series of bifunctional compounds they called chemical inducers of degradation (CIDEs) designed to degrade Estrogen Receptor-alpha (ERα) by linking an ERα inhibitor to different ubiquitin ligase ligands. The approach chosen by the authors to investigate a panel of E3 ligase ligands (instead of the usual VH-032 and IMiDs derivatives) is not often chosen as it represents a significant amount of synthetic work. Here, this effort was rewarded by the discovery of multiple potent CIDEs and that one compound in particular, their pan-IAP/Erα CIDE, induced a distinct TNFα-dependent cell death phenotype, offering a new dual therapeutic approach.

General comments

The main text of this manuscript by Shemorry et al. is clear, concise and reads easily. Extensive experimental work support nicely the main findings and allows the reader to follow along. The findings of this study reveal for the first time a bifunctional degrader that, on top of inducing the degradation of the protein of interest, can co-opt additional discrete functions of cellular effector proteins (such as E3 ligases) through its E3 ligand. This paves the way to the rational design of ligands capable of inducing/exploiting multiple effector proteins and cellular responses at once for maximum therapeutic efficacy, not just the degradation of the protein of interest. This work also highlights therefore the need for novel actionable E3 ligases and associated ligands, for now still limited to a handful of molecules.

We believe this manuscript could be improved on some specific points detailed below.

Specific comments

All western blot panels: MW ladder should be shown.

Figure 2 and corresponding Figure S2

The authors claim a correlation between K_D and D_max (Fig 2C) and a more modest correlation between average t_(1/2 )and D_max (Supp 2C). Regarding K_D and D_max, there might be a correlation in the XIAP/ERa serie, but the other would need more example to really assert a trend. For the t_(1/2 )and D_max, correlation is truly not obvious. A quantification of this correlation would help defend this affirmation.

Figure 3

The legend does not disclose how many replicates were used to estimate confluence, nor what doses of compounds were used. These details should be added (the same comment applies to other figures in this manuscript).

The Incucyte graphs (here and in other figures, but in particular Fig 3D) are overall very crowded with overlapping curves, making it very difficult to interpret the data. We encourage the authors to either find a way to improve the readability on one graph, or to split the results and make two graphs side-by-side.

 3B – There is a clear difference in cell counts between DMSO and Untreated (untreated cells look quite sparse, are they even alive?). Yet, the differences are assessed compared to DMSO. Could the authors clarify this choice and comment on the effects of DMSO on MCF7 cell proliferation?

 3C and 3D - It looks like GDC-0152 is the only compound where the PROTAC activity is highly impacted by washout. Compounds where the confluence is almost not impacted by washout could indicate a really stable mode of action, even hours after washout. It would be nice to investigate such stability. For instance, Fulvestrant, that is acting by retaining ERα as a monomer, is still active 4 days after washout: the authors may wish to comment on this aspect.

Figure 4

4A: The number of replicates, quantification of the gels and statistical analysis where appropriate should be provided.

Figure 5

5C: We observed that the condition GDC-0152*+PBMCs were not included in this experiment, whereas it was in the previous 2 panels. Additionally, the moment at which the cells were treated and at which dose are not indicated.

Figure 6

6C - In this in vivo model, the Enbrel effect looks really minor, which contrasts with the results obtained in figure 5. Could the authors comment on that? Is there a possible role for additional death receptor/ligand pairs, or other modes of death induction, here? Additionally, a comparison to fulvestrant or another ERalpha degrader such as ARV-471 would be interesting here to really compare the therapeutic potential of this novel dual mode of action compared to more established approaches.

6D - Regarding individual tumors, Enbrel works only on a few individuals and for 20 days. It would be of interest to characterize inter-tumoral differences that could explain such divergent Enbrel effects.

Also, tumors under pan-IAP/ERα-CIDE + Enbrel seem to rapidly grow at different time points. This raise the question; would a longer longitudinal study show less or more of an effect of Enbrel?

Competing Interests

The authors declare that they have no competing interests.

Competing interests

The authors declare that they have no competing interests.