Using a RNAi-mediated screen of the phosphatome, we have identified several phosphatases that were essential for cell survival after HU challenge (). As with other similar screening approaches, potential candidates may not be identified due to various reasons, including incomplete depletion and functional redundancy. Caveats are also required for the interpretations of the positive candidates. Although not all the candidates have been characterized in detail, it is likely that not all the RNAi effects were specific. In fact, the mRNA of some of the candidates were not effectively depleted by the shRNAs (). Nevertheless, our data indicated the ability of the assay to identify physiological-relevant candidates. Depletion of the classic checkpoint component, CHK1, was used as a positive control (Fig. S1
). Moreover, one of the positive clones, dual specificity phosphatase 11 (DUSP11), has already been implicated in the G2
DNA damage checkpoint 
. It was found that U2OS cells that expressed DUSP11 shRNA retained histone H3Ser10
phosphorylation after irradiation.
Our data indicated that downregulation of SHP2 with either siRNA or shRNA, both with transient transfection and in stable cell lines, enhanced HU-mediated cell death ( and ). SHP2 was also required for cell survival after challenge with other DNA stress-inducing agents, including CIS () and IR (). Unlike HU, CIS could generate inter-and intra-strand crosslinks, causing single and double-strand breaks in addition to replication blockage. This may explain the relatively high level of cell death associated with CIS. Consistent with these results, we found that SHP2 was activated (as indicated by SHP2Tyr542
phosphorylation) upon treatment with HU (), CIS (), and IR (). In agreement with our results, Kuo et al
. recently demonstrated using an activity-based probe that SHP2 is activated by CIS 
. The specificity of the effects on SHP2 was demonstrated with a number of approaches, including rescue experiments using shSHP2-resistant SHP2 () as well as using a small inhibitor of SHP2 ().
SHP2 has mainly been implicated in cell growth regulation. Indeed, depletion of SHP2 extended the cell cycle by ~2 h in HeLa cells (). We found that the cell cycle delay was probably caused by a reduction of cyclin E and an extension of G1 phase. However, the cell cycle defects were not responsible for the increase in cell death after CIS treatment in SHP2-depleted cells ().
We found that the increase in cell death after SHP2 downregulation was at least in part due to defective checkpoint mechanisms. The activation of CHK1 after HU (), CIS (), and IR () treatment was reduced after SHP2 depletion. Consistent with the weakening of the checkpoint, the reduction of histone H3Ser10
phosphorylation after DNA damage was prevented in SHP2-depleted cells ( and ). Indeed, SHP2-depleted cells prematurely entered mitosis after IR treatment (). Another indication of the weakening of the checkpoint after SHP2 depletion was the impairment of DNA repair ( and Fig. S8
). The premature entry into mitosis of HU- or CIS-treated S phase cells probably resulted directly in apoptosis ( and ). In accordance with our findings, the IR-mediated G2
arrest response was found to be diminished in MEFs lacking functional Shp2 
. It should be noted, however, SHP2 probably only plays a peripheral role in the checkpoint. Depletion of the classic checkpoint component CHK1 induced significantly more checkpoint defects than SHP2 ().
As SHP2 is implicated in the regulation of several proteins that can potentially regulate apoptosis (including AKT, ERK, and JNK), it is perhaps not surprising that depletion of SHP2 affected cell survival in our studies. However, previously published evidence of SHP2’s effects on apoptosis is contentious. Consistent with our findings, Shp2-deficient mouse embryonic fibroblasts (MEFs) are hypersensitive to etoposide- or proteasome inhibitor MG115-induced apoptosis 
. Likewise, inactivation of Shp2 sensitizes MEFs to epigallocatechin-3-gallate (green tea polyphenol)-mediated apoptosis through p53- and p73-dependent mechanisms 
. Murine hemopoietic cell line Ba/F3 overexpressing catalytically inactive Shp2 are also more susceptible to cell death 
. However, contrary to our findings, Shp2-deficient MEFs were found to be less sensitive to CIS-induced apoptosis 
. Shp2 has also been reported to negatively regulate hematopoietic cell survival by dephosphorylation of Stat5 
. It should be noted that the studies using Shp2-deficent cells involves the use of immortalized MEFs with a deletion within exon 3 of Shp2, which results in a truncation that lacks the NH2
-terminal SH2 domain (N-SH2), but retains catalytic activity 
. The effects of the Shp2 truncation in MEFs may differ from the depletion of the entire SHP2 in this study. In accordance with this, Shp2−/−
mice died at peri-implantation, much earlier than Shp2exon 3−/−
mice (midgestation) 
. In agreement with our results, RNAi-mediated knockdown of Shp2 in murine myoblast C2C12 also reduces cell survival 
In conclusion, SHP2 is both activated upon several DNA stress signals and is required for maintaining the checkpoint. The depletion of SHP2 resulted in defects in DNA damage-induced CHK1 activation, cell cycle arrest, and DNA repair, thereby promoting apoptosis.