On 2017 Jun 19, Stefano Casola commented:

Authors acknowledge the insightful comments. We would like to draw the attention of Dr Woodgett to the following observations:<br> • The antibody used to detect GSK3-beta Serine-9 phosphorylation does not recognize phosphorylated forms of GSK3-alpha (https://www.cellsignal.com/products/primary-antibodies/phospho-gsk-3-beta-ser9-d3a4-rabbit-mab/9322). Thus, whereas our data indicate that loss of the BCR in MYC-driven lymphoma cells leads to a reduction in GSK3-beta Ser-9 phosphorylation, it remains to be investigated whether GSK3-alpha is similarly affected in these cells.

• GSK3-beta knock-down experiments were performed using six independent shRNAs (referred to in the Methods section of the paper). Data obtained with the two most effective hairpins are shown in Extended Data Figure 5c, d. Importantly, the shRNAs were selected for their ability to target GSK3-beta but sparing GSK3-alpha. Despite only partial GSK3-beta knock-down, lymphoma cells losing BCR expression resisted substantially better to their BCR+ counterparts in competition assays, with the most effective hairpin (shRNA# 2) causing a complete block of their counter selection (Extended Data Figure 5e). These results closely mirror those obtained studying BCR+/BCR- lymphoma cell competitions treated with the GSK3 inhibitor CHIR99021 (Figures 3d and Extended Data Figure 5a).

Therefore, whereas we cannot exclude a contribution of GSK3-alpha, our data indicate that modest changes in GSK3-beta expression/phosphorylation are sufficient to critically affect BCR-controlled fitness of MYC-driven lymphoma cells.

On 2017 Jun 15, Jim Woodgett commented:

This is an interesting paper but from what I can see, the evidence for these effects being dependent on GSK3beta (rather than a combination of GSK-3beta plus GSK-3alpha) is limited to a partial (maximally 35%) knockdown by siRNA in extended data figure 5 - where a marginal effect was observed (partial knockdown of GSK3alpha may have given a similar result). The pharmacological inhibitor used, CHIR99021, has NO significant selectivity for GSK3beta over GSK3alpha (the authors do refer to it being in GSK3 inhibitor in two places). In every example where phosphorylation of Serine 9 of GSK3beta has been examined along with phosphorylation of Serine 21 of GSK3alpha, they are phosphorylated in parallel. There are no kinases that target these sites selectively (not true of a more C-terminal site, Ser389, targeted on GSK3beta by p38 MAPK). Why is this important? Because throughout the paper, there are over 50 mentions of GSK3beta, including the title, yet there was no measurement of GSK3alpha phosphorylation or knockdown.

On 2017 Jun 15, Jim Woodgett commented:

This is an interesting paper but from what I can see, the evidence for these effects being dependent on GSK3beta (rather than a combination of GSK-3beta plus GSK-3alpha) is limited to a partial (maximally 35%) knockdown by siRNA in extended data figure 5 - where a marginal effect was observed (partial knockdown of GSK3alpha may have given a similar result). The pharmacological inhibitor used, CHIR99021, has NO significant selectivity for GSK3beta over GSK3alpha (the authors do refer to it being in GSK3 inhibitor in two places). In every example where phosphorylation of Serine 9 of GSK3beta has been examined along with phosphorylation of Serine 21 of GSK3alpha, they are phosphorylated in parallel. There are no kinases that target these sites selectively (not true of a more C-terminal site, Ser389, targeted on GSK3beta by p38 MAPK). Why is this important? Because throughout the paper, there are over 50 mentions of GSK3beta, including the title, yet there was no measurement of GSK3alpha phosphorylation or knockdown.

On 2017 Jun 19, Stefano Casola commented:

Authors acknowledge the insightful comments. We would like to draw the attention of Dr Woodgett to the following observations:<br> • The antibody used to detect GSK3-beta Serine-9 phosphorylation does not recognize phosphorylated forms of GSK3-alpha (https://www.cellsignal.com/products/primary-antibodies/phospho-gsk-3-beta-ser9-d3a4-rabbit-mab/9322). Thus, whereas our data indicate that loss of the BCR in MYC-driven lymphoma cells leads to a reduction in GSK3-beta Ser-9 phosphorylation, it remains to be investigated whether GSK3-alpha is similarly affected in these cells.

• GSK3-beta knock-down experiments were performed using six independent shRNAs (referred to in the Methods section of the paper). Data obtained with the two most effective hairpins are shown in Extended Data Figure 5c, d. Importantly, the shRNAs were selected for their ability to target GSK3-beta but sparing GSK3-alpha. Despite only partial GSK3-beta knock-down, lymphoma cells losing BCR expression resisted substantially better to their BCR+ counterparts in competition assays, with the most effective hairpin (shRNA# 2) causing a complete block of their counter selection (Extended Data Figure 5e). These results closely mirror those obtained studying BCR+/BCR- lymphoma cell competitions treated with the GSK3 inhibitor CHIR99021 (Figures 3d and Extended Data Figure 5a).

Therefore, whereas we cannot exclude a contribution of GSK3-alpha, our data indicate that modest changes in GSK3-beta expression/phosphorylation are sufficient to critically affect BCR-controlled fitness of MYC-driven lymphoma cells.