Overall summary of the article and its findings:

The authors accurately describe the current state of knowledge when it comes to PCa therapy. Blocking AR signalling has been a mainstay for PCa therapy since 1941. The results of surgical or chemical castration are only temporary and lead to CRPC emergence, therefore second-generation anti-androgens, like ENZ, have been developed. Nevertheless, PCa cells adapt revealing a gap in knowledge

Kregel et al 2016 is a comprehensive analysis of the mechanisms underlying treatment-driven enzalutamide resistance using 4 prostate cancer cell lines as a model. The key findings include: 1) Cells treated with ENZ for over 6 months exhibit different growth rates than the parental cell lines; 2) The ENZ-R cells injected into mice form tumors and metastasize at a higher rate; 3) ENZ-R cells undergo complex expression changes, concerning both AR-regulated and AR-independent genes.

Overall strengths of the article and what impact it might have in the prostate cancer field:

The authors present a compelling story that sheds light on the mechanisms of acquired enzalutamide resistance. This is a pertinent issue as 1) enzalutamide is the first approved anti-androgen indicated for mCRPC and 2) patients eventually progress on ENZ despite promising initial results. AR mutations conferring ENZ-resistance have been previously described in the literature, but only a small fraction of the patients presents them. Understanding various avenues of acquired ENZ resistance is of great clinical value and can potentially redefine the standard of care for this lethal disease.

Major points in the article which need clarification, refinement, reanalysis, rewrites and/or additional information and suggestions for what could be done to improve the article:

Overall, the story is clear and well-founded but there are some gaps that when addressed, would improve the quality of this paper.

1) On page 26261, paragraph 2, the authors hypothesized that ENZ-R cells might be also resistant to other drugs used in PCa therapy and attempted to test this hypothesis (Figure 1C). However, the only tested drug docetaxel, is a microtubule stabilization inhibitor. To support the claim that ENZ-R doesn’t lead to multi-drug resistance, I would suggest testing a drug that targets AR directly (another anti-androgen like bicalutamide) and indirectly (any of the AR cofactor inhibitors, or AR degraders).

2) On page 26261, paragraph 2 – the authors make a claim that “In all four cell lines, EnzR cells continued to be sensitive to docetaxel-induced cell death (Figure 1C)”. Figure 1C does not include any statistical analysis to support the claim. In fact, taking a closer look at the VCAP-ENZ-R cells indicates they are more resistant to docetaxel (0.5nM, 1nM, 5nM) when compared to the corresponding treatments of the parental cell line. Adding a few more docetaxel concentrations would allow to model IC50 of this drug and definitely determine drug-sensitivity differences. I believe this should be further investigated and if the statistical analysis reveals that indeed more VCAP-R cells survive the docetaxel treatment, it should be thoroughly discussed.

3) On page 26268, the very last paragraph, the authors indicate: “our analysis did not prioritize pathways of the negatively regulated AR-target genes GR and Sox2, which we have previously identified as mediating resistance to AR antagonists”. I might be misinterpreting the intention behind the word 'prioritization' however, this study could benefit from an alternative to IPA approach to investigate the nature of this issue. I would recommend selecting a few of the genes in question for a qPCR analysis and present the results using the approach from Figure 4.

Minor points like figures/tables not being mentioned in the text, a missing reference, typos, and other inconsistencies.

Overall, the data is presented appropriately. Most figures are clear with informative titles and labels. The text in the results adds to the data creating a coherent story. There are a few issues that could be addressed to improve the paper:

Title: The title is informative and relevant to what the focus of this study is but wordy. Could be shortened to: “Acquired resistance to enzalutamide in castration-resistant prostate cancer.” OR more accurately reflect the results of the study: “Pleiotropic mechanisms contribute to enzalutamide resistance acquisition in prostate cancer cells".

Figure 1A – missing scale bar in the images

Figure 1B – text extensively discusses % changes in the PI uptake but the y axis of the figure shows fractions instead. I would recommend harmonizing.

Figure 1B and statistical analyses – it looks like the authors attempt to conduct a comparative analysis between 3 different conditions using a t-test (the cross). Typically the t-test is reserved for comparisons of only 2 conditions and ANOVA is more suitable for 3+. Please consult a biostatistician to ensure the correct statistical test was chosen.

Figure 3B – the authors claim that “AR-V7 decreased in the CWR-R1-EnzR cells” but when looking at the corresponding WB and loading controls, I’m not certain this is the case. I would suggest repeating this extraction and WB to see if the result is reproducible. Also, please include non-cropped WB images in the supplemental material.

Figure 3C – the figure legend indicates that "Each lane represents 100,000 cells worth of lysate". Without knowing the ratios at which these extracts were diluted/loaded on the gel it is difficult to make a comparative analysis. For instance, the whole cell extract band intensity for parental LAPC-4 is definitely not a sum of nuclear + cytoplasmic band intensities. What are the dilution ratios between each fraction?

Figure 5D – I’m not an expert in IPA analysis but this looks like a very interesting visualization technique for big data analyses. The text could benefit from a thorough description of this method, especially considering the grey arrows in the figure are not annotated but vary in thickness. I believe the readers could benefit from knowing what does it indicate.

Figure 6B – the figure legend reads: “Overall numbers of genes Non-AR-Associated genes identified in each cell line.” This is unclear, possibly the word ‘genes’ is used too many times by accident.

Figure 6D – The figure legend contains proper annotation of what the red and blue arrows are, but what are the yellow and grey ones?

Materials and methods: a) In paragraph 1 the authors describe the procedure for their growth assays but forgot to mention what kind serum was used. Depending on the assay the cells might be cultured in FBS or CSS (charcoal stripped serum with hormones removed). This detail is important for proper result interpretation.

b) The authors need to be commended for maintaining in culture the parental cell lines while generating ENZ-R cells. Cancer cells are unstable and prolonged culturing can alter the cells irrespectively of ENZ treatment. In paragraph 2, could you please indicate if the parental lines were vehicle (presumably DMSO) treated during this time.

Overall, this looks like an excellent study. I appreciate the authors candor regarding the limitations and setting the ground for future studies. I'm looking forward to reading subsequent publications from this lab.