Protein phosphatase 1 regulatory inhibitor subunit 14C promotes triple‐negative breast cancer progression via sustaining inactive glycogen synthase kinase 3 beta

  1. Yunting Jian
  2. Lingzhi Kong
  3. Hongyi Xu
  4. Yawei Shi
  5. Xinjian Huang
  6. Wenjing Zhong
  7. Shumei Huang
  8. Yue Li
  9. Dongni Shi
  10. Yunyun Xiao
  11. Muwen Yang
  12. Siqi Li
  13. Xiangfu Chen
  14. Ying Ouyang
  15. Yameng Hu
  16. Xin Chen
  17. Libing Song
  18. Runyi Ye
  19. Weidong Wei

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    The manuscript presents data that high expression of Protein Phosphatase 1 (PP1) inhibitor in triple-negative breast cancer contributes to poor outcomes by downregulation of an important kinase, GSK3β. The study clearly demonstrates that changes in PPP1R14C expression alter the behaviour of the studied cancer cells and mouse models and proposes a mechanism linking PP1 inhibitor to GSK3β. If this mechanism were substantiated, this would enhance our understanding of the pathophysiology of this important disease and might suggest new treatment options.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Triple‐negative breast cancer (TNBC) is fast‐growing and highly metastatic with the poorest prognosis among the breast cancer subtypes. Inactivation of glycogen synthase kinase 3 beta (GSK3β) plays a vital role in the aggressiveness of TNBC; however, the underlying mechanism for sustained GSK3β inhibition remains largely unknown. Here, we find that protein phosphatase 1 regulatory inhibitor subunit 14C (PPP1R14C) is upregulated in TNBC and relevant to poor prognosis in patients. Overexpression of PPP1R14C facilitates cell proliferation and the aggressive phenotype of TNBC cells, whereas the depletion of PPP1R14C elicits opposite effects. Moreover, PPP1R14C is phosphorylated and activated by protein kinase C iota (PRKCI) at Thr73. p‐PPP1R14C then represses Ser/Thr protein phosphatase type 1 (PP1) to retain GSK3β phosphorylation at high levels. Furthermore, p‐PPP1R14C recruits E3 ligase, TRIM25, toward the ubiquitylation and degradation of non‐phosphorylated GSK3β. Importantly, the blockade of PPP1R14C phosphorylation inhibits xenograft tumorigenesis and lung metastasis of TNBC cells. These findings provide a novel mechanism for sustained GSK3β inactivation in TNBC and suggest that PPP1R14C might be a potential therapeutic target.

Article activity feed

  1. eLife

    Evaluation Summary:

    The manuscript presents data that high expression of Protein Phosphatase 1 (PP1) inhibitor in triple-negative breast cancer contributes to poor outcomes by downregulation of an important kinase, GSK3β. The study clearly demonstrates that changes in PPP1R14C expression alter the behaviour of the studied cancer cells and mouse models and proposes a mechanism linking PP1 inhibitor to GSK3β. If this mechanism were substantiated, this would enhance our understanding of the pathophysiology of this important disease and might suggest new treatment options.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  2. eLife

    Reviewer #1 (Public Review):

    This manuscript represents a potentially important study that attempts to elucidate the mechanisms through which the glycogen synthase kinase 3β (GSK3β) is involved in the progression of triple negative breast cancer (TNBC). Starting with an analysis of the TCGA and GEO databases the authors identify the serine/threonine phosphatase, PPP1R14C, as significantly upregulated in TNBC - because of this and the known ability of PPP1R14C to regulate the ability of PP1 to dephosphorylate GSK3β the focus of the manuscript is on the relationship of these molecules in TNBC. The significance of the role of PPP1R14C in TNBC was demonstrated by the negative correlation of PPP1R14C expression levels and the outcomes of TNBC patients; the higher the expression of PPP1R14C in TNBC, the poorer the outcome. These data certainly support the importance of investigating the mechanistic link between PPP1R14C, GSK3β and TNBC. Mechanistically, the authors demonstrate using gain-of- and loss-of-function approaches that PPP1R14C promotes and antagonizes breast cancer cell proliferation, mobility and anchorage-independent growth, respectively. Consistent with this, xenograft studies showed that increased PPP1R14C expression promoted whereas knockdown reduced tumor growth. These data are generally convincing on face value but unfortunately are limited in the robustness of the derived conclusions given the issues that surround interfering with the expression of regulatory PP1 subunits (see below). The authors show that PPP1R14C appears to specifically regulate GSK3β phosphorylation at Ser9 through modulating PP1's ability to dephosphorylate this site which subsequently controlled TRIM-23-mediated GSK3β stability. Finally, a claimed PP1 activator was used to provide further mechanistic insight into the actions of PP1 on GSK3β phosphorylation and stability. These results showed that the inhibitor and PP1 silencing led to reduced tumor volume. Collectively, this is an interesting manuscript, but it falls short in terms of strong mechanistic insight, largely due to limitations in the strength of evidence provided from many of the experiments.

    The major confounding issue of this manuscript rests on the interpretation of the overexpression and silencing experiments of PPP1R14C. Due to the actions of PPP1R14C as a regulatory subunit of PP1, disrupting its expression either through upregulation or downregulation shifts the stoichiometry of binding with the multitude of other PP1-interacting proteins. Therefore, one cannot conclude that the effects on tumor formation, proliferation, migration, anchorage independent growth etc as a result of either overexperessing/silencing PPP1R14C is due solely to its actions rather than a secondary consequence of disrupted PP1 complex formation with its other interacting proteins.

    In relation to the abovementioned point, it was very curious as to why either overexpressing or silencing PPP1R14C was so selective to the effects on GSK3β. The authors report a lack of effect on other PP1 substrates. The quality of the immunoblots of these other substrates (e.g. Myc) was insufficient.

    To provide a mechanism for the effects of PPP1R14C in regulating GSK3β phosphorylation at Ser9 and subsequently ubiquitination by TRIM25, the authors perform a series of co-immunoprecipitation experiments. These experiments show that while TRIM25 does form a complex with PPP1R14C/GSK3β but they do not provide proof that the E3-ubiquitin ligase event is in fact mediated by TRIM25.

    The pharmacological use of the C2 ceramide provides an appealing orthogonal, experimental manipulation, but it is hard to see evidence that this is a specific activator of PP1. In that regard it is even harder to understand how it appears to recapitulate the effects of PPP1R14C silencing.

  3. eLife

    Reviewer #2 (Public Review):

    This manuscript aims to provide novel insights into the molecular mechanisms contributing to the progression of triple-negative breast cancer (TNBC), the most lethal type of breast cancer. It starts from the observation that PPP1R14C (also known as KEPI), an established inhibitor of protein phosphatase-1, is upregulated in TNBC. The authors subsequently show that PPP1R14C upregulation correlates with a poor prognosis in patients and tumor progression. They also provide data indicating that PPP1R14C enhances the phosphorylation of an inhibitory site (Ser9) of protein kinase GSK3beta and targets GSK3beta for proteolytic degradation via the E3 ubiquitin ligase TRIM25. Co-immunoprecipitation data suggest that PPP1R14C forms two ternary complexes, one with GSK3beta and PP1, and another one with GSK3beta and TRIM25, and it is suggested that these complexes mediate the observed effects of PPP1R14C overexpression on GSK3beta inhibition and degradation.

    The topic is important and interesting but the provided data do not justify the major conclusions made. No compelling evidence is provided for a direct interaction between PPP1R14C and GSK3beta or TRIM25. Also, PPP1R14C is only an inhibitor of PP1 when phosphorylated on a specific residue but such phosphorylation is not demonstrated. PP1 knockdown or PPP1R14C overexpression experiments are not sufficiently conclusive because they affect many distinct PP1 holoenzymes. The manuscript also suffers from a lack of essential control- and validation experiments.

  4. eLife

    Reviewer #3 (Public Review):

    The manuscript presents data that high expression of Protein Phosphatase 1 inhibitor in triple-negative breast cancer contributes to the poor outcome by downregulation of an important kinase, GSK3β. If substantiated, this would enhance our understanding of the pathophysiology of this important disease and might suggest new treatment options. Indeed, changes in PPP1R14C expression alter the behaviour of TNBC in cells and in mouse models, but the mechanistic links to GSK3 are not robustly established.

    Fig 1-2 identified the PPP1R14C as upregulated in TNBC and with a significant correlation with worse outcome. Fig 3 and 4 show in vitro and in vivo effects of changes in PP1R14C consistent with increased proliferation, migration and metastasis in vivo. These studies look very solid and appear to identify a role for this phosphatase regulator in TNBC.

    The weaker part of the manuscript is the mechanistic link to GSK3 regulation. Over-expression and knockdown of PPP1R14C have effects on GSK3β phosphorylation and downstream targets, but the direct connection is unclear and made challenging by a number of complex experimental issues.

    The big questions -
    1. Is GSK3 directly ubiquitylated by TRIM25 on K183? I don't think the data are strong here, for reasons elaborated on below.

    2. Is GSK3 really the important target of PPP1R14C/PP1 complex? The biological data are correlative and the direct experiment, does GSK3β (S9A/K183R) rescue PPP1R14C over-expression, would need to be done. But since I suspect K183R is kinase-dead, this may fail.

    3. The studies with C2 are confounded by the broad effects (including on PP2A) of treating cells with ceramide. Calling C2 a specific PP1 activator is I think unwarranted.

    Specific comments:
    Why is there a band in Fig 5D lane 2, the Flag-PPP1R14C lane, in the absence of Flag-PPP1R14C?

    Why in Fig 5E, F, G are there two bands in the pGSK3bS9 blot?
    The authors would need to show the total GSK3 coming down here too, and the total GSK3 present in Fig 5H as well.

    I have trouble understanding the result in Fig 5H. According to this, global PP1 phosphatase activity increases 3 fold when PPP1R14C is knocked down. First, there is no method noted for this assay. How do we know this is specific to PP1? Second, PPP1R14C is only one of many PP1 interactors. How can its knockdown change cellular PP1 activity 3-fold? I note the knockout mouse for PPP1R14C had a 15% increase in thalamus PP1 activity (see fig 3, https://doi.org/10.1016/j.neuroscience.2009.10.007). This experiment needs much more in the way of controls.

    Fig 6 evaluates the role of PPP1R14C in GSK3 protein stability. There is a fundamental weakness here - How do the authors know the ubiquitylated smear in the various Fig 6 assays is GSK3 versus a ubiquitylated protein that interacts with active GSK3? GSK3 phosphorylation directs many proteins (famously β-catenin and Myc) for ubiquitylation and degradation, so the co-IP of ubiquitylated proteins with GSK3 is to be expected if the IP stringency is not very very high. This is consistent with inactive pSER9 GSK3 not bringing down ubiquitylated proteins. An IP after for example boiling in SDS to break up large complexes would be needed to test if GSK3 itself, rather than associated substrates, is directly ubiquitylated.

    Is TRIM25 specific for GSK3? It's identified by mass spectrometry. However, when I plug TRIM25 into the CRAPome database (https://reprint-apms.org) I find it comes down in 136/716 (19%) of all MS IP studies, making it a very common contaminant in IP. Thus the bar is high to show this is specific. Here the interaction is validated with over-expression of various truncation mutants.

    Line 235: "K183 of GSK3β has been recognized as the ubiquitylation site". First, what is the reference for this statement? I found one paper (https://doi.org/10.1074/jbc.M116.771667 that claims this residue is important for FBXO17 K48 modification, not the K63 linkage associated with TRIM25). In the crystal structure of GSK3β, that K183 appears to coordinate the phosphates of ATP, so the effect of the K183R mutation may be to make the kinase inactive, which would confound their results. So an important experiment is, does K183R retain wildtype kinase activity? Or is it inactive, and so act like the phosphorylated S9 GSK3?

    The reference for ceramide as a PP1 activator is not a primary reference, it is to a paper in the Journal of Endodontics, which uses it. It would be important to cite primary literature for this usage of C2. I note that many papers cite C2 ceramide as a PP2A activator. It is unclear what the rationale is for using it as a specific PP1 activator?

  5. Version published to 10.1002/ctm2.725
  6. Version published to 10.1101/2021.05.09.443305v1 on bioRxiv