Neuronatin is a modifier of estrogen receptor-positive breast cancer incidence and outcome

Abstract

Purpose: Understanding the molecular mediators of breast cancer survival is critical for accurate disease prognosis and improving therapies. Here, we identified Neuronatin (NNAT) as a novel antiproliferative modifier of estrogen receptor-alpha (ER+) breast cancer.

Experimental design: Genomic regions harboring breast cancer modifiers were identified by congenic mapping in a rat model of carcinogen-induced mammary cancer. Tumors from susceptible and resistant congenics were analyzed by RNAseq to identify candidate genes. Candidates were prioritized by correlation with outcome, using a consensus of three breast cancer patient cohorts. NNAT was transgenically expressed in ER+ breast cancer lines (T47D and ZR75), followed by transcriptomic and phenotypic characterization.

Results: We identified a region on rat chromosome 3 (142-178 Mb) that modified mammary tumor incidence. RNAseq of the mammary tumors narrowed the candidate list to three differentially expressed genes: NNAT, SLC35C2, and FAM210B. NNAT mRNA and protein also correlated with survival in human breast cancer patients. Quantitative immunohistochemistry of NNAT protein revealed an inverse correlation with survival in a univariate analysis of patients with invasive ER+ breast cancer (training cohort: n = 444, HR = 0.62, p = 0.031; validation cohort: n = 430, HR = 0.48, p = 0.004). NNAT also held up as an independent predictor of survival after multivariable adjustment (HR = 0.64, p = 0.038). NNAT significantly reduced proliferation and migration of ER+ breast cancer cells, which coincided with altered expression of multiple related pathways.

Conclusions: Collectively, these data implicate NNAT as a novel mediator of cell proliferation and migration, which correlates with decreased tumorigenic potential and prolonged patient survival.

Keywords: Breast Cancer; Cell cycle; Estrogen; NNAT; Prognosis.

Fig. 1

a Congenic mapping localized a 36 Mb region on RNO3 (142–178 Mb) that modifies mammary tumor incidence. Schematic representation of SS-3^BN congenic strains that were generated by introgressing segments of BN chromosome 3 (black) into the genetic background of the parental SS strain (white) by marker-assisted breeding. Thin black bars represent confidence intervals, which are chromosomal regions that could be BN or SS. Following exposure to DMBA carcinogen, mammary tumor incidence and latency were recorded weekly for 15 weeks. b Tumor latency represented as the percentage of rats with mammary tumors at each time point. Closed circles = SS, open circles = line A, closed squares = SS-3^BN, open squares = line B. c Six genes that were differentially upregulated in expression within the RNO3 region (142–178 Mb) was detected by RNAseq analysis performed on tumors from rat strains with the SS-derived alleles within the region [SS (n = 5) and line A (n = 6)] compared with those that have the BN-derived alleles [SS-3^BN (n = 3) and line B (n = 6)]. (D) 4 genes that were differentially downregulated in expression within the RNO3 region (142–178 Mb) was detected by RNAseq analysis performed on tumors from rat strains with the SS-derived alleles within the region [SS (n = 5) and line A (n = 6)] compared with those that have the BN-derived alleles [SS-3^BN (n = 3) and line B (n = 6)]. *p < 0.05, **p < 0.01, ***p < 0.001, as determined by Fisher’s Exact Test (incidence), Log-Rank Test (latency), and FDR-corrected Student t test (RNAseq)

Fig. 2

NNAT immunostaining in ER+ breast cancer and Kaplan–Meier curves for progression-free survival by dichotomized NNAT protein expression. Immunofluorescent imaging of NNAT protein expression (red signal) that is colocalized to pan-CK+ tumor cells (green signal) at low resolution (a–c; × 12.9) and high resolution (d, e; × 60). Note the perinuclear localization of the NNAT within the cytoplasm of pan-CK+ malignant tumor cells (d). Scale bar represents 100 μm (a–c) and 20 μm (d, e). Median cancer cell immunofluorescence signal for NNAT protein was computed for each tumor in the training cohort and an objective data-driven cutpoint dichotomized tumors into Low NNAT (black line) and High NNAT (red line) expression which was used to obtain the Kaplan–Meier plot (f). The same cutpoint was then applied to generate a Kaplan–Meier plot for the Validation cohort (g). Censored cases are represented by hash marks on the plot lines

Fig. 3

Single-cell RNAseq analysis of NNAT expression in the mouse mammary gland. a Single-cell RNAseq data were downloaded from the Tabula Muris consortium to identify the organism-wide distribution of NNAT + cells in the mouse. b Single-cell RNAseq data from the mouse cell atlas revealed a subclustering of NNAT expression in mammary gland cells undergoing involution. c A Spearman’s correlation of genes expressed with NNAT showed a significant enrichment of pathways involved in cell proliferation (blue dots) and motility (red dots)

Fig. 4

RNAseq analysis of ER+ tumor cells following transgenic overexpression of NNAT or GFP in T47D and ZR75 cells (n = 3 per group). a Transgenic overexpression of NNAT was confirmed by Western blot. b Gene network analysis revealed significant enrichment of the ESR1 (ERα) and E2F1 pathways. c, d Heatmaps of the enriched gene networks that re-associated with ESR1 (c) and E2F1 (d). e Bar graphs plots of the differential expression (FDR < 0.05) of key cell cycle mediators that are implicated in the ESR1 and E2F1 pathways. Data are presented as the mean fold expressions ± SEM (n = 3 per group). f Western blotting of CDKN1A, phosphorylated ERK1/2, total ERK1/2, and β-actin (n = 3 per group). g Proliferation of ER+ tumor cells following transgenic overexpression of NNAT or GFP in T47D and ZR75 cells. Data are presented as mean cell counts per well ± SEM (n = 18 per group). ***p < 0.001, as determined by Student t test

Fig. 5

NNAT overexpression inhibits the migration of ER+ breast cancer cell lines. a Representative image of T47D GFP cells on the bottom surface. b Representative image of T47D NNAT cells on the bottom surface. (C) Representative image of ZR75 GFP cells on the bottom surface. d Representative image of ZR75 NNAT cells on the bottom surface. e Average number of T47D GFP and T47D NNAT cells from 4 independent fields from 3 independent experiments. f Average number of ZR75 GFP and ZR75 NNAT cells from 4 independent fields from 3 independent experiments. g Gene network analysis revealed significant enrichment of the cell migration pathway. Date shown are individual chamber averages and the means of the data from at least 3 independent experiments (n = 15 wells/group). ***p < 0.001, as determined by Student t test

Fig. 6

Schematic of cell cycle regulators that were regulated by NNAT expression during the G₁-S phases of the cell cycle. The gene expression changes caused by transgenic overexpression of NNAT are denoted by red (increased mRNA expression), blue (decreased mRNA expression), or black (no change). Interactions that activate/increase expression, inhibit/decrease expression, or are required to pass the G₁ restriction point are indicated within the figure