YB1 Is a Major Contributor to Health Disparities in Triple Negative Breast Cancer

Abstract

Triple negative breast cancer (TNBC) is the most aggressive amongst all breast cancer (BC) subtypes. While TNBC tumors represent less than 20% of all BC subtypes, they are responsible for the most BC-related deaths. More significantly, when considering TNBC incidence across all racial/ethnic groups, TNBC accounts for less than 20% of all BCs. However, in non-Hispanic black women, the incidence rate of TNBC is more than 40%, which may be a contributing factor to the higher BC-related death rate in this population. These disparities remain strong even after accounting for differences in socioeconomic status, healthcare access, and lifestyle factors. Increased evidence now points to biological mechanisms that are intrinsic to the tumor that contribute to disparate TNBC disease burdens. Here, we show that YB1, a multifunction gene, plays a major role in the TNBC disparities between African American (AA) and Caucasian American (CA) women. We show in three independent TNBC tumors cohorts, that YB1 is significantly highly expressed in AA TNBC tumors when compared to CAs, and that increased levels of YB1 correlate with poor survival of AA patients with TNBC. We used a combination of genetic manipulation of YB1 and chemotherapy treatment, both in vitro and in animal models of TNBC to show that YB1 oncogenic activity is more enhanced in TNBC cell lines of AA origin, by increasing their tumorigenic and aggressive behaviors, trough the activation of cancer stem cell phenotype and resistance to chemotherapeutic treatments.

Keywords: TNBC; WAVE3; YB1; cancer disparities; cancer stem cells; chemoresistance.

Figure 1

(A) Quantification of Yb1 mRNA expression levels based on log2 RSEM values batch normalized from Illumina HiSeq RNASeq data by breast cancer subtype in the breast cancer BRCA patient data from TCGA PanCancer Atlas. Expression levels of YB1 are significantly higher (*** p < 0.0001, Wilcoxon) in the basal (TNBC) subtype when compared to other BC subtypes. (B) Survival analysis based on YB1 expression levels (median values) in the basal subtype of the TCGA BRCA PanCancer Atlas cohort. YB1 activation in TNBCs is associated with poorer overall survival (p = 0.049). (C) KM plot correlating survival of 4929 BC patients with YB1 mRNA expression levels. High YB1 expression levels correlate with poor survival probability in BC patients (p < 1.9 × 10⁻⁸). (D) KM plot correlating the survival of BC patients with YB1 protein expression levels. High YB1 expression levels correlate with poor survival probability in BC patients (p = 0.047). (E) KM plot correlating the survival of TNBC patients with YB1 mRNA expression levels. High YB1 expression levels correlate with poor survival probability in TNBC patients (p = 0.03).

Figure 2

(A) YB1 mRNA transcript quantification in AA and CA TNBC cell lines was determined by qt-RT-PCR and data was plotted as the fold change in HCC1806. (B) Average YB1 mRNA transcripts in AA and CA TNBC cell lines was plotted as the fold change in AA cell lines. YB1 transcript expression levels are significantly higher (* p < 0.001, Student ’t test) in AA TNBC cell lines when compared to their CA counterparts. (C) Comparison of YB1 expression levels, levels based on log2 RSEM values batch normalized from Illumina HiSeq RNASeq data, between AA and CA BC patients from the TCGA BRCA PanCancer Atlas cohort. YB1 expression levels are significantly higher (* p < 0.01, Wilcoxon) in AA when compared to CA BC tumors. (D) Comparison of YB1 expression levels between CA and AA TNBC from the MetroHealth TNBC cohort, as determined by qt-RT-PCR. Data was plotted as the fold change to AA tumors. YB1 expression levels are significantly higher (* p < 0.01, Student ’t test) in AA when compared to CA BC tumors.

Figure 3

(A) Western Blot analysis of protein lysates from parental (Ctrl) AA (black font) and CA (red font) TNBC cell lines and their YB1-KO derivatives, that were probed with the indicated antibodies. β-Actin was used a loading control. (B,C) Colony formation assay. The numbers under the bands represent the fold change of the Western Blot signal normalized to that of the parental cell line, as determined by densitometry analyses from one representative of 3 replicate blots. (B) Representative images of colony formation of parental and YB1-KO AA (black font) and CA (red font) TNBC cell lines. (C) Quantification of the number of colonies in each plate. Data is plotted as the percentage change from the control cells. Data shown are representative of 3 replicates (***, p < 0.0001, Student ’t test).

Figure 4

(A) Incidence (top) and latency (bottom) of tumor growth of prenatal 4T1 or its YB1-KO derivative cells that were inoculated in the mammary fat pads of female Balb/C mice (n = 5 mice per group). (B) Quantification of tumor volume of tumors generated from the inoculation of control 4T1 cells or their YB1-KO derivatives into the mammary fat pads of female Balb/C mice. (C, top panel) Quantification of lung metastasis nodules from the animal experiment described in (B). The bottom panel shows representative lungs from each mouse group. (D) Quantification of the tumor volume of tumors generated from the inoculation of control MDA-MB-231 (grey graph), MDA-MB-468 (black graph) cells or their respective YB1-KO derivatives (yellow and red graphs, respectively) into the mammary fat pads of female NSG mice (n = 5 mice per group). (E,F) Quantification of lung metastasis nodules from the MDA-MB-468 (E) and the MDA-MB-231 animal experiment described in (D) (**, p < 0.001, Student ’t-test).

Figure 5

(A–D) Comparison of the expression levels of Oct4 (A), Nanog (B), Sox2 (C), and WAVE3 (D) between the BC subtypes in the TCGA BRCA PanCancer Atlas cohort. (**, p < 0.001; ***, p < 0.0001, Wilcoxon). (E–F) Comparison of the expression levels of Oct4 (E), Nanog (F), Sox2 (G), and WAVE3 (H) between AA and CA BC patients in the TCGA BRCA PanCancer Atlas cohort. (*, p < 0.01; **, p < 0.001; ***, p < 0.0001, Wilcoxon). In all cases, mRNA expression levels were based on log2 RSEM values batch normalized from Illumina HiSeq RNASeq data. (I–L) Comparison of the expression levels of Oct4 (I), Nanog (J), Sox2 (K), and WAVE3 (L) between AA and CA BC patients in the MetroHealth cohort, as determined by qt-RT-PCR. Data were plotted as fold change to AA tumors. (* p < 0.01, **, p < 0.001; ***, p < 0.0001, Wilcoxon).

Figure 6

(A–E) Comparisons of the IC50 values of cisplatin (A,B), doxorubicin (C,D) and docetaxel (E,F) between AA (black bars) and CA (red bars) TNBC cell lines. The IC50 values were derived from the CCLE database. Average IC50 in the AA and CA TNBC cell lines was plotted as fold change to AA cell lines for cisplatin (B), doxorubicin (D) and docetaxel (F), (*, p < 0.01, Fisher’s Exact). (G) Gene set enrichment analysis of the TCGA BRCA PanCancer Atlas dataset shows that the gene signature associated with docetaxel resistance is enriched in the BC tumors that express high levels of YB1. (H) Western Blots with the indicated antibodies of protein lysates from parental MDA-MB-MB-468 cells or those that were treated with doxorubicin for 48 h (transient treatment), or MDA-MB-468 cells that were generated to be permanently resistant (chronic) to doxorubicin (468-Dox-res). The numbers under the bands represent the fold change of the Western Blot signal normalized to that of the parental cell line. (I) Western Blots with the indicated antibodies of protein lysates from parental MDA-MB-468, their cisplatin resistant (Cis-R) or doxorubicin (Dox-R) derivatives. β-Actin was used as a loading control. The numbers under the bands represent the fold change of the Western Blot signal normalized to that of the parental cell line. (J) Quantification of the tumor volume of tumors generated from the inoculation of parental MDA-MB-468 cells (black graph), or their cisplatin resistant (Cis-R) derivatives into the mammary fat pads of female NSG mice (n = 5 mice per group). Treatment with cisplatin started 21 days post inoculation. (***, p < 0.0001, Student ’t test).