Referee 1's Review:

Rosado etal. describe a multiparametric live cell screening approach to identify selective inhibitors of Akt signaling. The system is based on the ability of constitutive active mutants of Akt (myrAkt) and STAT5 (STAT5A1*6) to rescue Ba/F3 cells from factor (IL3) dependence via distinct pathways. Perturbing myrAkt or STAT5A1*6 signaling in the respective cell lines leads to Ba/F3 cell death in the absence of IL3. Cocultures of myrAkt and STAT5A1*6 expressing Ba/F3 cell lines are screened in a high throughput live cell imaging format to reveal agents that specifically target the myrAkt-dependent (or STAT5A1*6) pathway, distinguishing these from compounds that affect neither, both or unrelated pathways. Cells are identified by DNA (DRAQ5) staining, segmentation algorithm and enumerated by a unique fluorescent marker (CFG or YFG expression).

IL3-dependent Ba/F3 cells are a well-studied model system and as such a good choice for this approach. The main advantage forwarded in the paper is the use of cocultures, facilitating HTS and specificity calculations via cell ratios. This is a useful concept and the authors provide evidence to support their claims. Methods are adequately described and the paper is generally well written. The central proof-of-principle compound screening data show good reproducibility and there is a clear distinction between selective and non-selective compounds, the major goal of the system.

This paper will be of particular interest to investigators pursuing high throughput drug screens using high content imaging analysis.

Some issues warrant discussion and should be addressed in a revision:

It would be helpful if the authors could describe more in depth the difference between how myrAkt and STAT5A1*6 generate IL3-independence. This is important to interpret the results of the screening.

Figure 3. The individual panels are not properly labeled. “i” and “j” are missing and are not mentioned in the figure legend. In “k”, the graph symbols are not defined and not described in the figure legend. It is unclear in the legend the difference between cell survival (Alamar Blue) and viability (trypan blue exclusion).

Notably the compound screening is conducted at 12 hours but no data is presented as to the cell line viability at this time point. What is the extent of apoptosis at 12 hours? The time course data shown suggest that the surviving mrAkt and STAT5A1*6 cell lines achieve an apparent steady-state at 1-2 days. Is there any cell division following IL3-withdrawal or are these cells arrested? Why not treat the cells at this timepoint?

Figure 5. The compound selectivity is determined by the % ratio between the number of CFP/GFP expressing cells in the well. This calculation does not discriminate between toxicity and no effect at all. Hence it would be desirable to show data for an additional viability parameter based on the DRAQ5 staining (eg total # number normal nuclei and or pyknotic nuclei) to reveal toxic compounds as indicated in the Discussion.

The authors should explain more clearly why the PI3 kinase inhibitor LY294002 has no effect on the CFG/YFG ratio as the “PI3K/Akt” pathway is mentioned in the text.
It is confusing that while the PI3 kinase inhibitor LY294002 has no effect in this assay, apparently inhibiting PDK1 does; PDK1 is also activated via its PH-domain interaction with the PI3K product PIP3. It is expected that PDK1 phosphorylation of myrAkt on Thr308 is required for Akt kinase activity; the authors show that myrAkt is phosphorylated in Figure 2f. The authors should suggest an explanation for this observation.

The authors refer to the staurosporine derivative UCN01 as a PDK1 inhibitor. UCN01 is however relatively non-specific and will affect several kinases at the concentrations used in the screen. Staurosporine also inhibits PDK1 in addition to several other kinases. Hence it is interesting that UCN01 and staurosporine are resolved by this assay and this should be discussed based on the known kinase-inhibition profiles of these compounds (see reference #13: Komander et al., 2003).

The Akt inhibitor X compound displays a dose response in the assay: is the IC50 commensurate with published values?

Minor points:

I would suggest changing the wording in the abstract to reflect that the assay measures differential cell survival using a multiparametric live cell imaging assay in lieu of “fluorescence emitted by these paired cell lines…”

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.