This paper uses the lab’s well known and useful RASless MEFs to show that elevated expression of KSR1 (and KSR2) can compensate for the absence of RAS to promote Raf and ERK activation and colony formation. The authors contend that KSR1 expression shifts the killing curve of cells expressing RasG12C to the right and call it resistance. While the cells appear to be less sensitive, it is not clear that the response can accurately be called resistance. The cells expressing KSR1 are less sensitive after so short a treatment. Resistance, in the context used has clinical implications that may require further examination to prove. While the decreased sensitivity may reflect the likelihood of drug resistance, the authors may wish to temper their conclusions.

Using deletion constructs, the authors show that the amino terminus of KSR1 including the CA1 region is required, and argue, based on mutations intended to disable the ATP binding of the KSR1 pseudokinase domain, that ATP binding is also necessary. The authors claim this affects heterodimerization, which appears to be based on the loss of function due to the mutations in KSR1 and the fact that previously published crystal structures of the KSR pseudokinase domain and the Raf1 kinase. Not co-precipitation, proximity ligation assays, or bioID experiments are provided to demonstrate that interaction of Raf with the pseudokinase domain of KSR1 is disrupted by disruption of ATP binding.

Specific comments

1. For all plots of colony formation shown as ratios of test/control, the authors should indicate the number of control colonies so the reader can calculate the efficiency and effectiveness of the experiment.
2. Fig. 1C. What is the amount of exogenous KSR1 relative to endogenous KSR1 expressed in RASless MEFs? Kortum et al. (MCB 2004) (ref. 9) showed that there is an optimal level of KSR1 to maximal Ras-dependent signaling. Is this true for Ras-independent signaling in RASless MEFs as well? If not, what does this say about Ras-independent activation of Raf by KSR1?
3. Fig. 2C. This experiment shows that CA1 region is important for colony formation. It would be valuable to test a construct with the CA1 region only, or with the KSR1 sequences amino terminal to the pseudokinase domain to show that rescue of colony formation requires both the CA1 domain and the pseudokinase domain and not that, for example, that the KSR1pseudokinase domain is inhibitory to the CA1 region. Zhou et al. JMB 2002) implicated the C1 domain (CA3 region) in membrane localization. Did the authors consider testing this?
4. Fig. 2D. Is the lack of an effect of the deltaCA1 construct due to relatively low level of expression?
5. Fig. 3B. There appears to be no loading control, and less total KSR1 in the F701A lane. It would be preferable to show the specificity of the 32P-ATP interaction with KSR1 by competing with unlabeled ATP.
6. Fig. 3C. KSR1 D683A, 700A also failed to induce proliferation- how does this result indicate that ATP binding is essential for KSR1. The authors should clarify.
7. Fig. 6A-6B. Perhaps the wording should be "KSR1 expression decreased the responsiveness to Sotorasib treatment" Treatment for 72 h- does not define resistance. This comment also pertains to the discussion section.
8. Fig. 6A. The authors should show the standard deviation and indicate of the differences is are statistically significant.
9. Fig. 6B. Do MIA PACA-2 cells not have endogenous KSR1 expression?
10. Fig. 6 and Discussion. If elevated KSR1 does induce resistance to Ras inhibition, does drug-induced resistance to Ras inhibition increase in the expression of endogenous KSR1?

The authors did a fine job of responding to the reviewers. My one remaining and very minor comment is that the authors claim that the KSR1F701A mutant was less active at promoting ERK activation- it is not evident to me from Figure 6.