Expression of therapy-induced senescence markers in breast cancer samples upon incomplete response to neoadjuvant chemotherapy

Abstract

Senescence is a cell stress response induced by replicative, oxidative, oncogenic, and genotoxic stresses. Tumor cells undergo senescence in response to several cancer therapeutics in vitro (Therapy-Induced Senescence, TIS), including agents utilized as neoadjuvant chemotherapy (NAC) in the treatment of invasive breast cancer. TIS has been proposed to contribute to adverse therapy outcomes including relapse. However, there is limited evidence on the induction of senescence in response to NAC in clinical cancer and its contribution to disease outcomes. In this work, the expression of three senescence-associated markers (p21CIP1, H3K9Me3 (histone H3 lysine 9 trimethylation), and Lamin B1) was investigated in breast cancer samples that developed partial or incomplete pathological response to NAC (n=37). Accordingly, 40.54% of all samples showed marker expression consistent with a senescence-like phenotype, while the remainders were either negative or inconclusive for senescence (2.70 and 56.8%, respectively). Moreover, analysis of core-needle biopsies revealed minimal changes in p21CIP1 and H3K9Me3, but significant changes in Lamin B1 expression levels following NAC, highlighting a more predictive role of Lamin B1 in senescence detection. However, our analysis did not establish an association between TIS and cancer relapse as only three patients (8.1%) with a senescence-like profile developed short-term recurrent disease. Our analysis indicates that identification of TIS in tumor samples requires large-scale transcriptomic and protein marker analyses and extended clinical follow-up. Better understanding of in vivo senescence should elucidate its contribution to therapy outcomes and pave the way for the utilization of senolytic approaches as potential adjuvant cancer therapy.

Keywords: H3K9Me3; Lamin B1; Neoadjuvant Chemotherapy; Senescence; breast cancers; p21CIP1.

Figure 1. Expression levels of p21^CIP1, H3K9Me3, and Lamin B1 in breast cancer samples following NAC

(A) The percentage of samples with positive or negative expression of each senescence-associated biomarker in samples of patients that only received NAC and developed partial or incomplete pathological response to therapy (n=37). (B) Table demonstrates correlational expression of the three tested senescence-associated biomarkers. Upper panel shows direct comparison of Lamin B1 expression with both p21^CIP1 and H3K9Me3 separately, while lower panel shows direct comparison of p21^CIP1 expression with H3K9Me3 expression. P-values are included as calculated using Fisher’s exact test.

Figure 2. Expression levels of p21^CIP1, H3K9Me3, and Lamin B1 in breast cancer samples before and after NAC

Pre-NAC staining was performed on core needle biopsy samples of 11 patients that received NAC only and developed partial or incomplete pathological response (n=11). Representative bright-field microscopic images of the immunohistochemical expression levels of p21^CIP1, H3K9Me3, and Lamin B1 in breast cancer sample of patient A28, showing insignificant reduction in p21^CIP1 and H3K9Me3 expression levels, but significant decrease in Lamin B1 levels following NAC. All images were taken using bright-field microscopy (Olympus BX 25, Olympus, Tokyo, Japan) under 40× objective lens.

Figure 3. Determination of evidence on the induction senescence-like phenotype based on p21^CIP1, H3K9Me3, and Lamin B1 staining

Determining senescence induction was based on the evaluation of all three tested biomarkers combined. Only samples that show positive expression for H3K9Me3 and p21^CIP1 and negative expression of Lamin B1 were considered positive for senescence. Consequently, samples that were negative for H3K9Me3 and p21^CIP1 and positive expression of Lamin B1 were considered negative for senescence. Lastly, all other expression possibilities were considered as inconclusive for senescence. Upper panel shows representative images for patient A34 whose samples were positive for p21^CIP1, H3K9Me3 and negative for Lamin B1, and hence, considered positive for senescence-like phenotype. Conversely, middle panel shows representative images for patient A8 whose samples were negative for p21^CIP1, H3K9Me3 and positive for Lamin B1, and hence, considered negative for senescence-like phenotype. Other possibilities of expression, as shown in representative images in lower panel (patient A36) were considered inconclusive for senescence-like phenotype. The percentage of samples positive for senescence-like phenotype was 40.54% while the remainder were either negative or inconclusive (2.70 and 56.8%, respectively) (n=37). All images were taken using bright-field microscopy (Olympus BX 25, Olympus, Tokyo, Japan) under 40× objective lens.