Regulation of Cancerous inhibitor of PP 2 A ( CIP 2 A ) by small molecule inhibitor for c-Jun NH-Terminal Kinases ( JNKs ) , SP 600125 , in Human Fibrosarcoma ( HT 1080 ) cells

Protein phosphatase 2A inhibition is one of the pre-requisites for Background: human cell transformation. Previously, we have identified an endogenous inhibitor of PP2A, CIP2A (Cancerous Inhibitor of Protein Phosphatase 2A) in human fibrosarcoma cells (HT1080) using tandem affinity purification. CIP2A over expression has been demonstrated in almost every tumour type studied so far. However, our understanding on the mechanisms regulating CIP2A expression in human cancers, especially in sarcomas, is still emerging. Human fibrosarcoma (HT1080) cells were treated with small Methods: molecule inhibitors against the three major signalling pathways, namely p38, MEK and JNK pathways to identify the pathway regulating CIP2A expression in the sarcoma cells. This was followed by verification of the results using small interfering RNAs (siRNA) for the kinases. In line with previous observations, small molecule inhibitor for MEK Results: pathway (PD98059) decreased CIP2A mRNA and protein expression. Interestingly, small molecule inhibitor for the JNK pathway, SP600125 decreased mRNA and protein levels of CIP2A oncoprotein with negligible effect of SB203580 (p38 kinase) inhibitor on CIP2A expression in HT1080 cells. However, siRNAs specific to either JNK1 or JNK2 kinases did not result in decrease in CIP2A expression. Contrarily, two different CIP2A siRNAs, which were used as positive controls, decreased JNK2 expression in HT1080 cells. Although it is well established that SP600125 inhibits JNK Conclusion: kinases, it has also been shown to inhibit a spectra of other kinases. SP600125 inhibits CIP2A protein expression both in time and concentration dependent manner. However, depletion of both JNK1 and JNK2 kinases using specific siRNAs fails to decrease CIP2A protein expression levels, thereby indicating the need to verify the results obtained by treatment with small molecular inhibitors of kinases by independent approaches like two different target specific siRNAs. Finally, fortuitously we identify JNK2 as a CIP2A downstream target in HT1080 cells. 2

. Experimentally, it was demonstrated that activation of RAS and telomerase (TERT), along with inactivation of the tumour suppressor proteins P53 and Retinoblastoma protein (RB) can immortalize a variety of human cell types, which can subsequently transform to a tumourigenic state in response to inhibition of protein phosphatase 2A (PP2A) 1,2 .Various independent studies have shown that inhibition of PP2A activity is a pre-requisite for human cell transformation 1,[3][4][5] .Therefore, understanding the mechanisms by which PP2A is inhibited in cancer cells is vital for developing new anti-cancer therapies.
Although several transcription factors like MYC 8 , ETS1 17 , E2F1 9 and ATF2 19 have been identified as positive regulators of CIP2A in various carcinomas, factors influencing CIP2A expression in non-hematopoietic mesenchymal cells or sarcomas, are yet to be discovered.Notably, CIP2A amplification has been observed in soft-tissue sarcomas 21 .In addition, since CIP2A was identified using HT1080 (human fibrosarcoma cell line) cell extracts 6 this cell line was selected to dissect the mechanisms for high CIP2A expression in sarcomas.Since p38, ERK and JNK signalling pathways are commonly perturbed in cancers, we assessed the role of these pathways in CIP2A expression in HT1080 cells.To this end, respective small molecule kinase inhibitors, namely SB203580 (p38 pathway inhibitor), PD98059 (MEK pathway inhibitor) and SP600125 (JNK pathway inhibitor) were used to inhibit signalling through these pathways in HT1080 cells.

Bioinformatics
Cancer cell line encyclopaedia (http://www.broadinstitute.org/software/cprg/?q=node/11) was used to get the expression levels of JNK1 and JNK2 in HT1080 cells 22 .This is a resource which provides analysis and visualization of DNA copy number, mRNA expression, mutation data and more, for 1000 cancer cell lines.

Statistics
Student T-test was used obtain the statistical significance value using Graph Pad software.

SP600125 inhibits CIP2A mRNA and protein expression in a time and concentration dependent manner in HT1080 cells
To determine the oncogenic signalling pathways that maybe involved in regulating CIP2A expression in human fibrosarcoma, HT1080 cells were treated with small molecule inhibitors for the p38 (SB203580; 20 µM), JNK (SP600125; 10 µM) and ERK (PD98058; 20 µM) signalling pathways.As previously observed in gastric cancer cells 17 , PD98058 reduced CIP2A mRNA and protein expression in HT1080 cells (Figure 1A and B).Importantly, while SB203580 showed negligible effect, SP600125 potently inhibited CIP2A mRNA and protein expression in HT1080 (Figure 1A and B) cells.Furthermore, inhibition of CIP2A protein expression by SP600125 was observed to be both time and concentration dependent (Figure 1C and D).
Neither JNK1 nor JNK2 regulate CIP2A protein expression in HT1080 SP600125-mediated inhibition of CIP2A expression suggested the involvement of the c-Jun N-Terminal Kinases (JNKs) in regulation of CIP2A expression in HT1080 cells.Since the JNK3 isoform is well known to be neural-specific 23 , specific validated siRNAs against JNK1 and JNK2 isoforms were transfected into HT1080 cells and CIP2A expression estimated.Although, both JNK1 and JNK2 siRNAs reduced their target proteins (JNK1 and JNK2 respectively), there was no change in CIP2A protein expression with either of them individually (Figure 2A   (Figure 2C).Since the JNK2 isoform is expressed more than 10-fold higher than the JNK1 isoform in the HT1080 22 (Figure 2D) cell line we transfected two validated and specific siRNAs against JNK2.In addition, we also transfected the HT1080 cells with two different CIP2A siRNA as positive controls.In line with our previous observation (Figure 2B and C), the two different siRNAs for JNK2 knocked out JNK2 expression (Figure 2E), while CIP2A expression remained unaltered.Surprisingly, two different CIP2A siRNA, which worked well as positive controls, efficiently decreased JNK2 expression (Figure 2E).

Discussion
Even though small molecule inhibitors are an emerging therapeutic option against cancers, the specificity issues limit their potential to be used in clinics.They have been extensively used to study various cell signalling pathways.In particular SP600125 has been used to study the effect of c-Jun N-Terminal Kinases (JNKs) in various processes [24][25][26] .Our results suggest that even though we were able to see decrease in CIP2A expression in HT1080 cells on treatment with SP600125 (Figure 1), at doses used previously to inhibit c-Jun NH 2 -terminal kinase (JNKs) activity 27,28 , we were not able to validate the findings using two different siRNAs specific to the JNK2 kinase (Figure 2E).Interestingly, a decade ago a previous study emphatically demonstrated the effect of SP600125 on different kinases 29 .The study revealed that SP600125, although a JNK inhibitor, could inhibit the activity of several other kinases 29 .In fact, SP600125 was demonstrated to inhibit kinases like CK1, DYRK1 and AMPK even more effectively than JNK kinase itself 29 (Supplementary Table 1).
Surprisingly, two different CIP2A siRNAs used as positive controls, decreased JNK2 expression levels in HT1080.This has also been observed in a separate study in an epithelial origin cell line, HeLa 20 .
Interestingly, JNK2 has been shown to regulate CIP2A expression via ATF2 transcription factor in mouse embryo fibroblasts (MEFs) 19 in the pre-transformed stage.Since we observe the vice versa in fully transformed HT1080 cells, it can suggest that there may be a molecular switch between JNK2 and CIP2A which may have a possible role in the RAS-transformation of mesenchymal cells.Nevertheless, the functional consequence of CIP2A-mediated JNK2 expression in mesenchymal cells would require further exploration.
Altogether, our study highlights the need for the validation of results obtained by small molecule treatments with independent approaches like two or more target specific siRNAs, shRNAs or use of inducible systems like RNAi or Tamoxifen/Tetracycline-induced overexpression systems 8 .

Open Peer Review
Current Referee Status: Inhibition of the protein phosphatase PP2A contributes to the acquisition of a cancerous phenotype in cells.One mechanism by which the inhibition of PP2A can be achieved is through the up-regulation of endogenous inhibitors of PP2A such as CIP2A.However, the signaling pathways regulating CIP2A expression in cancer cells remain poorly defined.In this article, Anchit Khanna has studied the role of the different mitogen-activated protein (MAP) kinase cascades in regulating the expression of CIP2A in the HT1080 human fibrosarcoma cell line.Treatment of HT1080 cells with the pharmacological inhibitor SP600125 led to a drop in the mRNA and protein expression of CIP2A in HT1080 cells.SP600125 has often been used to implicate the protein kinases of the JNK family in the cellular process under study, but as pointed out by the author, this compound has many off-target effects.Anchit Khanna therefore attempted to validate the results using siRNA mediated knock-down of JNK1 and JNK2 expression in HT1080 cells.Unfortunately, CIP2A expression was normal in the cells in which JNK1 and/or JNK2 were depleted using siRNA oligos.The author concluded that the effects of SP600125 on CIP2A expression were not due to JNK1 or JNK2 inhibition, and stated the importance of validating experimental results using independent methods.
The data as presented does not confirm or refute a role for the JNKs in regulating CIP2A expression in HT1080 cells, as alternative explanations for the data remain possible.For instance, the knock-down of JNK1 or JNK2 may not be sufficiently effective, leaving residual JNK1/2 catalytic activity in cells which drives the expression of CIP2A.Indeed, the data presented in figure 2A show that JNK1 expression was suppressed by only 50% and the author failed to show the level of JNK1 expression in the experiment depicted in figure 2C.In addition to blotting for the respective kinase, it would have also been informative in both the inhibitor and RNAi studies to measure the phosphorylation of a substrate such as c-Jun to verify that these treatments indeed blocked JNK function in HT1080 cells.
It is also good practice to use structurally-unrelated protein kinase inhibitors to reduce the chance of off-target effects contributing to the phenotype under study.Since SP600125 entered the market, much improved JNK inhibitors have been developed such as JNK-in-8 (Zhang .2012 Chem Biol).It would et al be worthwhile to test the effect of such compounds on CIP2A expression, which could help establish whether a role for JNKs exist in this pathway.
The author is correct in stating that one must validate results using different approaches.I was therefore very surprised to read that the author concluded that CIP2A is involved in regulating the expression of JNK2 in HT1080 using a single approach.RNAi technology, like other methodologies, is fraught with off-target problems.It is essential to validate the results of these experiments by re-expressing the protein using an RNAi-resistant cDNA of CIP2A.This control experiment would strongly support the author's F1000Research off-target problems.It is essential to validate the results of these experiments by re-expressing the protein using an RNAi-resistant cDNA of CIP2A.This control experiment would strongly support the author's conclusion that CIP2A controls JNK2 expression in HT1080 cells.I acknowledge that these experiments are challenging and the author may be able to provide supporting evidence using alternative approaches to the one used in this article.
Although the experiments are well performed, the study is incomplete and does not allow one to make a strong case for or against involvement of the JNKs in regulating CIP2A expression.Based on this assessment, I have scored this paper 'Approved with Reservations'.
Additional Minor Points: 1.The author omitted to indicate the statistical significance of the data in Figure 1A.2. In a single author paper, one should not use the terms 'we' and 'us' rather 'I' and 'me.'I have read this submission.I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
No competing interests were disclosed.

Competing Interests:
Author Response 11 Dec 2013 , University of New South Wales, Australia Anchit Khanna ''The data as presented does not confirm or refute a role for the JNKs in regulating CIP2A expression in HT1080 cells, as alternative explanations for the data remain possible.For instance, the knock-down of JNK1 or JNK2 may not be sufficiently effective, leaving residual JNK1/2 catalytic activity in cells which drives the expression of CIP2A."I thank Dr. Clark for his prompt and valuable comments on the manuscript.However, I would like to highlight Figures 2B, C and E wherein almost 100% knockdown of JNK2 kinase was achieved.In fact, two different siRNAs for JNK2 isoform, which is the isoform that is majorly expressed in HT1080 (Fig. 2D) cells, completely abolished JNK2 expression leaving little room for the existence of residual JNK2 activity in these cells for maintaining CIP2A levels.Additionally, 10 fold higher expression of JNK2 isoform in comparison to the JNK1 isoform in Ht1080 cells (Fig 2D), suggests it to be more relevant functionally.

, 2013 et al "It is also good practice to use structurally-unrelated protein kinase inhibitors to reduce the chance of off-target effects contributing to the phenotype under study"
I thank Dr. Clark for sharing his expertise in the field and totally agree that structurally-unrelated kinase inhibitors have decreased off target effects and their use would definitely further confirm the negative findings.However, the main aim of this study is to highlight the un-specificity of SP600125 compound which has been used extensively as specific JNK inhibitor.In addition, using two different siRNAs (to overcome off-target effects due to one siRNA) that knock down JNK2 expression almost completely (Fig 2E ) without changing the levels of CIP2A significantly strongly suggest that JNK2 does not regulate CIP2A expression in HT1080 cells.
"The author is correct in stating that one must validate results using different approaches.I was therefore very surprised to read that the author concluded that CIP2A is involved in regulating the expression of JNK2 in HT1080 using a single approach.RNAi technology, like other methodologies, is fraught with off-target problems.It is essential to validate the results of these experiments by re-expressing the protein using an RNAi-resistant cDNA of CIP2A.This control experiment would strongly support the author's conclusion that CIP2A controls JNK2 expression in HT1080 cells.I acknowledge that these experiments are challenging and the author may be able to provide supporting evidence using alternative approaches to the one used in this article."I thank Dr. Clark for highlighting the issue of siRNA specificity and totally agree that there are off-target effects that have to be ruled out with this approach.It is for this reason most journals and the general scientific community requests the use of two independent siRNAs for verifying the effects due to depletion of the target gene.Accordingly, as shown in Fig 2E, there have been two independent siRNAs against both JNK2 and CIP2A used which not only almost completely deplete their target proteins but also rule out the element of un-specificity which usually arises by the use of a single siRNA against the target protein.
Additionally, an independent study ( ) using microarray approach Niemelä 2012 et al., identified JNK2 as downstream target of CIP2A in HeLa cells, thereby lending further confidence to these findings.Moreover, as mentioned in the manuscript this finding surfaced just by chance as CIP2A siRNAs were being used as positive controls in the experiment.Additionally, I agree and have now further highlighted in the manuscript as well that CIP2A-mediated regulation of JNK2 would need further exploration.Finally, I would like to emphasize that the main aim of this study is to highlight that SP600125, concentrations at which it has been demonstrated to inhibit JNK activity, may also inhibit other kinases as highlighted by an elegant study previously (kinases that it inhibits have been listed in supplementary table 1).Interestingly, we recently identified CHK1 kinase (which is almost as sensitive as JNK to SP600125) as a positive regulator of CIP2A expression (Khanna et  ).Therefore, to explore the possible role of CHK1 and other SP600125 sensitive , 2013 al. kinases in regulating CIP2A expression, we did a pilot screen and transfected HT1080 cells with 2 different siRNAs against CHK1 (Kinase almost as sensitive to SP600125 as JNK according to previously published study) and with siRNAs against CDK2 and DYRK1 (two other kinases that are more sensitive to SP600125 than JNK2).As shown in supplementary figure 1B, two different siRNAs against CHK1 decreased CIP2A expression as observed previously, while CDK2 and DYRK1 didn't change CIP2A expression.This

Figure 1 .
Figure 1.SP600125 positively regulates CIP2A expression in HT1080 cells.A. qRT-PCR showing the effect of small molecule inhibitors against the p38 (SB203580), MEK-ERK (PD98059) and JNK (SP600125) pathways on CIP2A mRNA expression (12h timepoint; Shown is mean values ± S.D., of representative results from three independent experiments.Student T-test was used obtain the statistical significance value) B. Western blot showing the effect of small molecule inhibitors against the p38 (SB203580), MEK-ERK (PD98059) and JNK (SP600125) pathways on CIP2A protein expression (24h timepoint).Shown is the representative picture of two independent experiments.C and D. Effect of small molecule inhibitors against JNK (SP600125) pathway on CIP2A protein expression in concentration (C) and time (D) dependent manner.Shown is the representative picture of two independent experiments.

Figure 2 .
Figure 2. JNK1 and JNK2 do not regulate CIP2A expression in HT1080 cells.A, B, and C. Western blots showing the effect of JNK1 (A), JNK2 (B) and combination of both JNK1 and JNK2 (C) siRNAs on CIP2A protein expression, 72h post-transfection.Shown is a representative result from two independent experiments.D. mRNA expression of JNK1 and JNK2 in HT1080 cell line from the cancer cell encyclopaedia study.E. Western blot showing the effect of two different siRNAs specific for JNK2 and CIP2A proteins and their protein expression levels 72h post-transfection.The numbers below the blot are the quantified values for CIP2A and JNK2 protein levels normalized to Actin protein levels, relative to the levels in Scrambled (control) transfected cells.Shown is a representative result of two independent experiments.
Unit, University of Dundee, Dundee, UK "...the data presented in figure2Ashow that JNK1 expression was suppressed by only 50% and the author failed to show the level of JNK1 expression in the experiment depicted in " figure2C.I thank Dr. Clark for this relevant observation.I have now attached the JNK1 levels from the experiment shown in Fig 2C as supplementary figure 1A with quantifications.As shown in the figure, despite complete knockdown of JNK1 in these cells, CIP2A levels remain almost unaltered, thereby re-emphasizing the fact that JNK1 does not regulate CIP2A expression in HT1080 cells.Complete knockdown of JNK1 also rules out the possibility of residual JNK1 activity driving CIP2A expression and the use of c-Jun as marker of JNK function in this experiment.With reference to SP600125, the concentrations used in Figure 1 have been demonstrated previously in several studies to inhibit JNK activity (p-Jun levels).Few examples of the studies that have used similar concentrations of SP600125 to inhibit JNK activity are (used 20uM of SP600125), (5uM of SP600125 Yue 2011 et al., Duan 2011 et al.,

Introduction
It has been recently established that regardless of phenotypic variability between different cancer types, perturbation of a limited number of genetic elements is sufficient to induce transformation in different human cell types