Response to Lehrer

To the Editor: We thank Dr Lehrer for his thoughtful correspondence and The Cancer Genome Atlas (TCGA) based analysis of CDK12, MED1, and ERBB2 expression.1 Dr Lehrer’s findings reinforce the close association between CDK12 and MED1 across breast cancer subtypes and extend discussion on the context-dependent CDK12 dysregulation in HER2-positive and ER-positive/HER2-negative disease. As noted, CDK12 and MED1 show strong co-expression even independent of ERBB2 copy number. This supports our mechanistic evidence that CDK12 functionally cooperates with MED1 and ER to sustain transcription in advanced ER-positive tumors. Although transcriptomic correlations highlight potential co-regulation, our study demonstrated that CDK12 directly influences ER chromatin recruitment and transcriptional output, and this is critical to understanding the functional hierarchy between CDK12, MED1, and ER signaling. We agree that the proposed dual model of genomic co-amplification in HER2-positive cancers and transcriptional co-activation in ER-positive/HER2-negative cancers offers a useful framework. Clinical series of HER2-positive breast cancers have similarly shown recurrent ERBB2/CDK12 co-amplification and linked it to therapeutic outcome,2,3 which supports the genomic dosage context described by Dr Lehrer. Moreover, targeting CDK12 in HER2-amplified models reduces PI3K/AKT signaling and restores anti-HER2 sensitivity,4,5 also confirming functional cooperation between CDK12 and HER2 pathways. Beyond receptor-positive disease, pharmacological studies in triple-negative breast cancer have demonstrated that dual CDK12/CDK13 inhibition provokes intronic polyadenylation, suppresses DNA-damage-repair gene expression, and induces a “BRCAness” phenotype that sensitizes tumors to PARP inhibitors and platinum agents.6,7 These observations indicate that CDK12 can regulate distinct transcriptional programs depending on molecular context, including genomic co-amplification in HER2-positive disease, CDK12/MED1-driven co-activation in luminal tumors, and potential DNA-repair gene control in basal-like disease (Figure 1). In practice, distinguishing between these contexts will be essential when designing therapeutic strategies that target the CDK12 complex, because these mechanisms may not be mutually exclusive. Distinct contexts of CDK12 dysregulation in breast cancer. (A) In HER2-positive tumors, CDK12 and ERBB2 are co-amplified on chromosome 17q12, leading to increased mRNA expression primarily reflecting genomic dosage. (B) In ER-positive/HER2-negative tumors, CDK12 phosphorylates and activates MED1, facilitating estrogen receptor (ER) chromatin recruitment and transcriptional activation of ER-responsive genes. (C) In triple-negative breast cancer, loss or pharmacological inhibition of CDK12/CDK13 promotes intronic polyadenylation and reduced DNA-damage-repair (DDR) gene expression, resulting in a BRCAness phenotype and enhanced sensitivity to PARP inhibitors and DNA-damaging agents. Figure created with BioRender.com. The funder had no role in the design of the study; the collection, analysis, or interpretation of the data; or the writing of the manuscript and decision to submit it for publication. Damir Varešlija (Conceptualization, Funding acquisition, Supervision, Writing—original draft, Writing—review & editing), Daniela Ottaviani (Conceptualization, Formal analysis, Funding acquisition, Supervision, Writing—original draft, Writing—review & editing), and Leonie Young (Conceptualization, Funding acquisition, Investigation, Supervision, Writing—original draft, Writing—review & editing) The authors are supported by funding from Enterprise Ireland (IP/2019/0796; L.Y., D.V.); Breast Cancer Ireland (18239A01, L.Y.); Research Ireland (19/FFP/6443, L.Y.), (20/FFP-P/8597, D.V.), (23/SPP/11783, L.Y., D.V.), Science Foundation Ireland (Research Ireland)-Irish Research Council SFI-IRC Pathway grant (22/PATH-S/10866, D.O.); Breast Cancer Now (2021JulyPCC1460, L.Y., D.V.), (2019AugSF1310 with the generous support of Walk the Walk), and EU Intereg BOTs (EAPA_0048/2022). The authors declare no competing interests. No new data was analyzed as part of this article.