YTHDF1 amplification is correlated with worse outcome and lower immune cell infiltrations in breast cancer

Abstract

OBJECTIVE:

N6-methyladenosine (m6A) is a common RNA modification on eukaryotic mRNA and some of the m6A regulatory proteins play a crucial role in breast cancer. However, the copy number variations for m6A regulatory proteins and their role in clinicopathological characteristics and survival in breast cancer remain unclear.

METHODS:

In this study, we screened the m6A related genes alterations in breast cancer by analyzing the Molecular Taxonomy of Breast Cancer International Consortium and The Cancer Genome Atlas database, and further analyzed the clinical prognostic value of YTHDF1 amplification.

RESULTS:

The YTH domain family (YTHDF3 and YTHDF1) amplification exhibited higher alteration rates among 10 m6A regulatory genes. YTHDF1 and YTHDF3 amplification resulted in higher mRNA expression ( $P<$ 0.0001). Protein expression of YTHDF1 and YTHDF3 were higher in breast cancer ( $P<$ 0.0001). YTHDF1 amplification presented a high correlation with worse clinicopathological characteristics and overall survival in patients with breast cancer. Cox regression analysis showed that YTHDF1 amplification was an independent risk factor for 10-year overall survival in breast cancer (Hazard ratio: 1.663; 95% confidence interval: 1.298–2.131; $P<$ 0.001). Gene set enrichment analysis revealed that the downstream target of YTHDF1 may be related to MYC signaling regulation and T cell differentiation. Moreover, YTHDF1 amplification and high expression resulted in lower immune cell infiltration. YTHDF1 knockdown retrained proliferation, migration and invasion in breast cancer cells in vitro.

CONCLUSIONS:

We found significant worse clinical characteristics and lower immune infiltrates in patients with YTHDF1 amplification. The findings indicate that YTHDF1 amplification may be a potential target for the treatment of breast cancer.

Keywords

Breast cancer survival N6-methyladenosine YTHDF1 amplification

Abbreviations

m6A	N6-methyladenosine
TCGA	The Cancer Genome Atlas
METABRIC	Molecular Taxonomy of Breast Cancer International Consortium
CNV	copy number variation
OS	overall survival
K-M	Kaplan-Meier

RFS	relapse free survival
DMFS	distance metastasis free survival
mRNA	messenger RNA
DC	dendritic cell
GSEA	gene set enrichment analysis
WT	wild type
MUT	mutant
HR	hazards ratio
CI	confidence intervals
ER	estrogen receptor
PR	progesteron receptor

1. Introduction

Breast cancer is the most common type of cancer with the second highest mortality rate worldwide, affecting a large number of women. In 2017, $>$ 25,000 women were newly diagnosed with breast cancer in the United States of America [1]. According to Cancer Stat Facts of the National Cancer Institute, 12.8% of women will suffer from breast cancer during their lifetime. The high mortality rate is mainly due to recurrence following the development of resistance to chemotherapy and distant metastasis. Although considerable efforts have been made in clinical trials in terms of offering new therapy options to improve clinical outcomes [2, 3, 4, 5], the molecular mechanism of advanced breast cancer remains to be elucidated. Although gene expression patterns in breast cancer have been comprehensively studied [6, 7, 8], post-transcriptional regulation remains to be investigated. A more comprehensive understanding of correlation between cancer survival and post-transcriptional gene regulation may explain cancer initiation and progression.

RNA modification has attracted broad interest due to its potential to regulate genes at the post-transcriptional level. N6-methyladenosine (m6A) is the most prevalent type of RNA modification on eukaryotic mRNA [9]. The important roles of m6A in mRNA translation efficiency [10], regulation of gene expression [11, 12], and other biological processes, thereby causing diseases in humans, were recently reported [13]. The regulatory proteins of m6A include “writer” and “eraser” acting as methyltransferase and demethylase, respectively [13, 14], and “reader” which recognizes m6A-modified mRNA [14]. Several m6A “writer”, “eraser” and “reader” proteins have been reported [12, 15, 16, 17, 18]. The members of the YTH domain family 1–3 (YTHDF1, YTHDF2, YTHDF3) are important m6A cytoplasmic readers. YTHDF1 and YTHDF3 promote the translation of m6A-methylated mRNA by recruiting translation initiation factors [17]. Although the members of the YTH domain family are closely related to pathogenesis, there is limited knowledge regarding their role in the correlations with breast cancer.

In the present study, we sought to characterize the copy number variations (CNVs) of m6A RNA regulatory genes in cohorts of patients with breast cancer, and further evaluate the association between YTHDF1 and YTHDF3 amplification with clinicopathologic characterization and survival. Furthermore, we also analyzed the associated molecular pathway and underlying mechanism in patients with YTHDF1 amplification.

2. Material and methods

2.1 Cohort data collection

We extracted the clinical, somatic CNV data and mRNA expression data from the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) and The Cancer Genome Atlas (TCGA) data from cBioPortal (http://www.cbioportal.org) [6, 19, 20, 21].

2.2 Kaplan-Meier plotter

To analyze the prognostic value of YTHDF1 and YTHDF3, survival analysis was performed on KM plotter (http://kmplot.com) [22]. The Kaplan-Meier plotter includes breast cancer patients ( $n=$ 6,234) from GEO and TCGA, which is capable of evaluating the effect of genes on survival. Patient samples were split into high expression and low expression groups according to upper quartile. Hazard ratio and 95% confidence interval and log rank $p$ value were also analyzed.

2.3 Gene set enrichment analysis (GSEA)

GSEA was performed with the TCGA data set using the GSEA v4.0.1 software (Broad Institute, UT, USA) [23, 24]. Patients were classified into an YTHDF1 amplification group and wild-type group. Enrichment of gene sets for YTHDF1 amplification and non-altered groups was computed with the form of “h” versus “l” and the permutation number of 1,000. Hallmark and immune signature gene data sets were downloaded through the GSEA software. Gene data sets were considered to be significantly enriched with a false discovery rate $<$ 0.25 and P-value $<$ 0.05.

2.4 UALCAN and clinical proteomic tumor analysis consortium (CPTAC)

We observed YTHDF1 and YTHDF3 protein expression profiles from clinical proteomic tumor analysis consortium (CPTAC) on UALCAN (http://ualcan.path. uab.edu) [25]. The Log2 spectrum count ratio value from CPTAC is firstly normalized in each sample configuration file, and then normalized between each sample. The Z value represents the standard deviation with comparison of breast cancer and normal tissue. YTHDF1 expression data from 24 types of cancer in TCGA were analyzed on UALCAN [26].

2.5 Immune cell infiltration analysis

Tumor Immune Estimation Resource (TIMER) is a public web server for the analysis of immune cell infiltration across all major types of cancer [27, 28]. Six types of immune cells were included (i.e., B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils, and dendritic cells). We used the somatic copy number alteration (SCNA) module and expression level to investigate the extend of infiltration of these six immune cells. The SCNAs were classified according to GISTIC 2.0 (Genomic Identification of Significant Targets in Cancer) [29]. The comparison of immune cell infiltration between two groups was performed using a two-sided Wilcoxon rank-sum test.

Table 1
Association between YTHDF1 amplification status and clinicopathological characteristics

	YTHDF1 AMP		YTHDF1 WT
	No. (% in AMP)		No. (% in WT)		$P$ value
Pathologic type
Breast invasive ductal carcinoma	129	(92.1)	1531	(79.0)	$<$ 0.001
Breast mixed ductal and lobular	5	(3.6)	213	(11.0)
Carcinoma
Breast invasive lobular carcinoma	5	(3.6)	167	(8.6)
Breast invasive mixed mucinous carcinoma	0	(0.0)	25	(1.3)
Metaplastic breast cancer	1	(0.7)	1	(0.1)
Age (years)	59.4 $\pm$ 13.7		60.7 $\pm$ 13.1		0.235
Tumor size	28.5 $\pm$ 18.9		26.3 $\pm$ 15.5		0.127
Node status
Neg (pN0)	55	(39.9)	998	(53.5)	0.002
Pos (1–3)	49	(35.5)	582	(31.2)
Pos ( $\geqslant$ 4)	34	(24.6)	286	(15.3)
Estrogen receptor
Pos	108	(74.0)	1509	(75.7)	0.644
Neg	38	(26.0)	485	(24.3)
Progesteron receptor
Pos	51	(38.6)	989	(53.5)	0.001
Neg	81	(61.4)	859	(46.5)
HER2 status
Pos	24	(18.2)	223	(12.1)	0.040
Neg	108	(81.8)	1625	(87.9)
Subtype
HER2+/HR+	11	(8.3)	102	(5.5)	0.239
HER2+/HR-	13	(9.8)	121	(6.5)
HER2-/HR+	88	(66.7)	1325	(71.7)
HER2-/HR-	20	(15.2)	300	(16.2)
Pam50 subtype
Basal	11	(8.4)	198	(10.7)	$<$ 0.001
LumA	25	(19.1)	675	(36.6)
LumB	53	(40.5)	422	(22.9)
Claudin-low	13	(9.9)	205	(11.1)
HER2	23	(17.6)	201	(10.9)
Normal	6	(4.6)	142	(7.7)
TP53
MUT	73	(49.7)	673	(33.2)	$<$ 0.001
WT	74	(50.3)	1353	(66.8)
PIK3CA
MUT	50	(34.0)	865	(42.7)	0.040
WT	97	(66.0)	1161	(57.3)
MYC
AMP	80	(54.4)	474	(23.4)	$<$ 0.001
WT	67	(45.6)	1552	(76.6)

HR, hormone receptor; WT, wild type; MUT, mutant; Pos, possitive; Neg, negative; AMP, amplification.

2.6 Cell culture and transfection

MDA-MB-231 and MCF-7 cells were obtained from the Cell Bank of Type Culture Collection (Chinese Academy of Sciences, Shanghai, China). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco, Grand Island, USA) containing 10% fetal bovine serum (Gibco) at 37 ${}^{\circ}$ C in a humidified atmosphere with 5% CO ${}_{2}$ . To generate RBM22-shRNA vectors, psi-LVRU6GP vector was digested by BamH I and EcoR I and following cloned with the indicated sequences according to the manufacturer’s instructions. hYTHDF1 shRNA-1 F: GATCCGTTCGTTACATCAGAAGGATATCAAGAGTATCCTTCTGATGTAACG AACTTTTTTG; shRNA-1 R: AATTCAAAAAAGTTCGTTACATCAGAAGGATACTCTTGATATCCTTCTGATGTAACGAACG; hYTHDF1 shRNA-2 F: GATCGTTCGTTACATCAGAAGGATATCAAGAGTATCCTTTTCCTGATGTAACGAACTTTTTTG; shRNA-2 R: AATTAAAAAAGTTCGTTACATCAGAAGGATACTCTTGATATCCTTCTGATGTAACGAACG; Cells were transfected with plasmids by Lipofectamine 2000 Transfection Reagent according to the manufacturer’s protocol. Lipofectamine 2000 Transfection Reagent was from Invitrogen (Invitrogen, CA, USA). Cell Counting Kit was obtained from DONGJI (DONGJI, Kyushu, Japan). All RNA samples from MDA-MB-231 and MCF-7 cells were extracted by conducting RNeasy Mini Kit (QIAGEN, Germany) according to the manufacturer’s instructions. mRNA reverse transcription was completed by using a PrimeScriptTM RT reagent Kit (Takara Biotechnology, Dalian, China), and qRT-PCR was conducted by SYBR Premix Ex TaqTM II (Takara). Primers specific for YTHDF1 and GAPDH were synthesized by Sangon Biotech. The primer sequences are listed as follows: YTHDF1 F: 5’-TCAGGCTGGAGAATAACGA-3’; R: 5’-GGTTGTGTGCTTGTAGGAACT-3’; GAPDH F: 5’-GCACCGTCAAGGCTGAGAAC-3’; R: 5’-TGGTGAAGACGCCAGTGGA-3’.

For each sample, mRNA expression was normalized to GAPDH. Relative fold changes were calculated utilizing the 2- $\Delta\Delta$ Ct method.

2.7 Wound healing assay

To examine the migratory ability, MDA-MB-231 and MCF-7 cells were incubated in 6-wells overnight. Then cells were transfected with psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids. Three straight scratches across each well were created in the cell layer by a pipette tip after transfected with indicated plasmids. Then cells were cultured with the medium containing 2% FBS for 48 h. Wound closure was observed under a microscope after 48 h of incubation. The migratory activity was assessed by measuring the distance of migration into the scratches in each well and the migration rate was calculated as the ratio of the relative migration distance in the shRNA group to the relative migration distance in the vector group.

Figure 1.

Characterization of m6A regulatory gene alteration in breast cancer patients. (A) Percentages of CNVs of ten m6A regulators and three oncogenes in breast cancer patients based on METABRIC cohort. (B) OncoPrint of ten m6A and three oncogenes regulators in breast cancer patients based on METABRIC cohort. The data were obtained and visualized by cBioPortal database [19, 20] using the METABRIC Breast Invasive Carcinoma dataset with modifications.

2.8 Transwell assay

MDA-MB-231 and MCF-7 cells were incubated in 6-well plates and allowed to attach overnight. MDA-MB-231 and MCF-7 cells were transfected with psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids by Lipofectamine 2000 Transfection Reagent according to the manufacturer’s protocol. Then cells were digested and suspended in starvation medium. Cells were counted by 0.4% Trypan Blue, 1 $\times$ 10 ${}^{5}$ surviving cells in 0.2 ml of starvation medium were then seeded in the inserts. In migration assay, 0.2 ml of starvation medium with 1 $\times$ 10 ${}^{5}$ surviving cells were added to the top transwell chambers, and 0.7 ml of complete medium were filled in the lower chambers. In invasion assay, chambers were pre-coated with matrigel. The medium in the top chambers was removed after 2 h. Then 0.2 ml of starvation medium with 1 $\times$ 10 ${}^{5}$ surviving cells were added to the top transwell chambers, and 0.7 ml of complete medium were filled in the lower chambers. After 24 $\sim$ 48 h of cultivation, cells under the lower chamber were washed with PBS and stained with 4% paraformaldehyde and 0.1% crystal violet. Five random fields were imaged from insert, and cells were counted by an inverted light microscope at 5 $\times$ magnification.

Table 2
Association between YTHDF3 amplification status and clinicopathological characteristics

	YTHDF3 AMP		YTHDF3 WT
	No.		No.		$P$ value
Pathologic type
Breast invasive ductal carcinoma	242	(84.6)	1418	(79.2)	0.244
Breast mixed ductal and lobular carcinoma	26	(9.1)	192	(10.7)
Breast invasive lobular carcinoma	16	(5.6)	156	(8.7)
Breast invasive mixed mucinous carcinoma	2	(0.7)	23	(1.3)
Metaplastic breast cancer	0	(0.0)	2	(0.1)
Age (years)	61.9 $\pm$ 12.5		60.4 $\pm$ 13.2		0.054
Tumor size	26.2 $\pm$ 12.6		26.5 $\pm$ 16.2		0.791
Node status
Neg (pN0)	120	(43.5)	933	(54.0)	0.003
Pos (1-3)	98	(35.5)	533	(30.8)
Pos (>4)	58	(21.0)	262	(15.2)
Estrogen receptor
Pos	214	(72.8)	1403	(75.6)	0.234
Neg	80	(27.2)	453	(24.4)
Progesteron receptor
Pos	125	(46.3)	915	(53.5)	0.027
Neg	145	(53.7)	795	(46.5)
HER2 status
Pos	45	(16.7)	202	(11.8)	0.040
Neg	225	(83.3)	1508	(88.2)
Subtype
HER2+/HR+	20	(7.4)	93	(5.4)	0.126
HER2+/HR-	25	(9.3)	109	(6.4)
HER2-/HR+	179	(66.3)	1234	(72.2)
HER2-/HR-	46	(17.0)	274	(16.0)
Pam50 subtype
Basal	11	(8.4)	198	(10.7)	$<$ 0.001
LumA	25	(19.1)	675	(36.6)
LumB	53	(40.5)	422	(22.9)
Claudin-low	13	(9.9)	205	(11.1)
HER2	23	(17.6)	201	(10.9)
Normal	6	(4.6)	142	(7.7)
TP53
MUT	111	(37.6)	635	(33.8)	0.200
WT	184	(62.4)	1243	(66.2)
PIK3CA
MUT	108	(36.6)	807	(43.0)	0.040
WT	187	(63.4)	1071	(57.0)
MYC
AMP	73	(24.7)	481	(25.6)	0.751
WT	222	(75.3)	1397	(74.4)

HR, hormone receptor; WT, wild type; MUT, mutant; Pos, positive; Neg, negative; AMP, amplification.

2.9 Statistical analysis

Pearson’s chi-squared test, Fisher’s exact test, $t$ -test, or Kruskal-Wallis test were selected to examine differences among variables. Continuous variables were compared using Student’s $t$ -test or the Mann-Whitney U test. Categorical data were compared using the chi-squared test. The one-way analysis of variance (ANOVA) was used for multiple-group comparisons. $P$ -values were two-tailed and $P$ -values $<$ 0.05 denoted statistically significant differences. Survival analysis was conducted using the Kaplan-Meier method, and two groups were compared through the log-rank test. Cox proportional hazards regression was conducted for patient survival. Statistical analyses were conducted using the SPSS version 20.0 (IBM Corp., Armonk, NY, USA), GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA) software and R statistical software (version 3.5.2, http://www.R-project.org).

3. Results

3.1 CNVs of m6A regulatory genes in patients with breast cancer

Firstly, we targeted 10 m6A regulatory genes, namely YTHDF3, YTHDF1, FTO, YTHDC1, ALKBH5, METTL3, WTAP, METTL14, YTHDF2, and YTHDC2. Most frequently altered genes, including PIK3CA, TP53, and MYC, were also analyzed for comparison with the above 10 genes.

The CNVs of m6A regulatory genes in 2,173 patients of the METABRIC cohort are described in Fig. 1A. Among them, the m6A reader genes YTHDF3 and YTHDF1 ranked the highest in alteration rate (14% and 7%, respectively). In contrast, we found only a small proportion of gene alterations among the other eight genes. An oncoprint spectrum of 10 m6A regulatory genes and three genes with frequent alterations from cBioPortal [19, 20] is shown in Fig. 1B, in which individual samples and genes are represented as columns and rows, respectively. Obviously, amplification accounted for the majority of gene alterations. In addition, deep deletion in CNVs was observed in a relatively small proportion of patients. Alterations of m6A regulatory genes in breast cancer were partially overlapped with those of the three frequently altered genes (Fig. 1B and Tables 1 and 2).

3.2 YTHDF1 and YTHDF3 amplification were associated with higher expression

Next, the effects of YTHDF1 and YTHDF3 amplification on gene expression were analyzed. As expected, the mRNA levels of YTHDF1 and YTHDF3 were significantly higher in the amplification group in both YTHDF1 and YTHDF3 groups (Figs 2A and B). Moreover, the protein expression of YTHDF1 and YTHDF3 were higher than normal tissue (Figs 3A and B). Analysis of the TCGA pan-cancer cohorts [26] showed that YTHDF1 also presented high expression in breast cancer tumors across 24 types of cancer (Supplementary Fig. 1).

Figure 2.

Correlation between mRNA expression levels according to gene amplification. (A) YTHDF1 expression levels comparison between YTHDF1 amplification YTHDF1 wild type. (B) YTHDF3 expression levels comparison between YTHDF3 amplification and YTHDF3 wild type. WT, wild type. ${}^{****}P<$ 0.0001.

Figure 3.

Comparison between protein expression of YTHDF1 (A) and YTHDF3 (B) between breast cancer and normal tissue. ${}^{****}P<$ 0.0001.

3.3 Amplification of m6A regulatory genes was associated with clinicopathological features in patients with breast cancer

YTHDF1 and YTHDF3 exhibited a higher incidence of amplifications in patients with breast cancer. Therefore, we assessed the relationship between the YTHDF1 or YTHDF3 amplification status and clinicopathological characteristics of patients with breast cancer using the METABRIC data set. We observed a higher proportion of lymph node metastasis, progesterone receptor negative rate, human epidermal growth factor receptor 2 (HER2) positive rate, and PAM50 subtype in both the YTHDF1 (Table 1) and YTHDF3 (Table 2) amplification groups.

PIK3CA, TP53, MYC are the most frequently altered genes and play important roles in breast cancer. Hence, we also assessed the relationship between YTHDF1, YTHDF3, and those three frequently mutated genes. As expected, YTHDF1 amplification had a high correlation with PIK3CA mutation, TP53 mutation, and MYC amplification (Table 1 and Fig. 1B), while YTHDF3 amplification was only correlated with PIK3CA mutation (Table 2).

3.4 Association between YTHDF1 and YTHDF3 amplification and survival of patients with breast cancer

We further estimated the effect of CNV on overall survival (OS), especially YTHDF1 and YTHDF3 amplification, to evaluate the prognostic value of m6A regulatory gene CNVs.

Table 3
Univariate and multivariate Cox proportional hazard analysis for 10-year overall survival

	Univariate			Multivariate
	HR	95% CI	$P$ value	HR	95% CI	$P$ value
Node status	1.718	1.566–1.885	$<$ 0.001	1.549	1.408–1.705	$<$ 0.001
Tumor size	1.016	1.013–1.019	$<$ 0.001
ER status	0.620	0.530–0.725	$<$ 0.001	0.526	0.440–0.629	$<$ 0.001
PR status	0.612	0.531–0.707	$<$ 0.001
HER2 status	1.747	1.445–2.112	$<$ 0.001	1.517	1.234–1.866	$<$ 0.001
Age	1.024	1.018–1.030	$<$ 0.001	1.033	1.026–1.039	$<$ 0.001
YTHDF1	1.811	1.419–2.313	$<$ 0.001	1.663	1.298–2.131	$<$ 0.001

HR, hazards ratio; CI, confidence intervals; ER, Estrogen receptor; PR, Progesteron receptor.

Figure 4.

Kaplan-Meier estimates of overall survival (OS) are shown according to YTHDF1 and YTHDF3 amplification status. (A) Kaplan-Meier (K-M) curves of OS according to YTHDF1 amplification status in METABRIC cohort. (B) K-M curves of OS according to YTHDF3 amplification status. OS, overall survival; K-M, Kaplan-Meier.

Subsequently, we focused on the prognostic effect of YTHDF1 and YTHDF3 on OS in the METABRIC breast cancer cohort. Patients with YTHDF1 amplification had significantly lower OS those with YTHDF1 non-amplification ( $P<$ 0.001) (Fig. 4A). We also observe lower OS with YTHDF3 amplification when compared to the non-amplification groups ( $P=$ 0.016) (Fig. 4B). Given that the YTHDF1 and YTHDF3 were related to OS, Cox regression was performed to confirm their prognostic effect. Univariate and multivariate Cox regression analyses are shown in Table 3. YTHDF1 amplification, lymph node status, HER2 status, estrogen receptor status, and age, were independent risk factors for 10-year OS in patients with breast cancer. Among these factors, YTHDF1 amplification had the highest hazard ratio (1.663; 95% confidence interval: 1.298–2.131; $P<$ 0.001).

Figure 5.

Kaplan-Meier estimates of survival are shown according to YTHDF1 and YTHDF3 mRNA expression. (A) Kaplan-Meier curves of overall survival (OS) are shown according to YTHDF1 expression. (B) Kaplan-Meier curves of OS are shown according to YTHDF3 expression. (C) Kaplan-Meier curves of relapse free survival (RFS) are shown according to YTHDF1 expression. (D) Kaplan-Meier curves of RFS are shown according to YTHDF3 expression. (E) Kaplan-Meier curves of distance metastasis free survival (DMFS) are shown according to YTHDF1 expression. (F) Kaplan-Meier curves of DMFS are shown according to YTHDF3 expression. OS, overall survival; K-M, Kaplan-Meier; RFS, relapse free survival; DMFS, distance metastasis free survival.

Figure 6.

GSEA results of YTHDF1 amplification status. Gene set enrichment plots of G2M checkpoint (A), E2F targets (B), MYC targets (C) and DNA repair (D) correlated to YTHDF1 amplification in breast cancer of TCGA data set.

3.5 Correlation between the expression of m6A regulatory genes and survival in patients with breast cancer

YTHDF1 and YTHDF3 amplification resulted in higher mRNA expression. Thus, we further confirmed the correlation between YTHDF1 and YTHDF3 expression and survival. After analyzing on KM plotter [22], we observed that the YTHDF1 high expression group (upper quarter) was associated with a significant lower relapse free survival (RFS) ( $P=$ 0.026, Fig. 5C) and distance free survival (DMFS) ( $P=$ 0.00075, Fig. 5E) versus the YTHDF1 low expression group. However, no statistically significant lower OS was identified in YTHDF1 high expression group ( $P=$ 0.16, Fig. 5A). Moreover, patients with YTHDF3 high expression had lower RFS ( $P=$ 0.0018, Fig. 5D). For OS and DMFS, no difference was observed between YTHDF3 high/low expression patients (Fig. 5B and F).

3.6 Enrichment analysis of YTHDF1 amplification

On the basis of the role of YTHDF1 in m6A methylation and the above results, we further evaluated the enriched gene sets in samples with and without amplification of YTHDF1. The GSEA analysis revealed that YTHDF1 amplification significantly enriched biological pathways, including the G2/M checkpoint (Fig. 6A), cell cycle-related targets of E2F transcription (Fig. 6B), MYC regulation (Fig. 6C), and DNA repair regulation (Fig. 6D). This finding indicated the important function of YTHDF1 in cell cycle regulation, in line with previous studies [17, 30]. We also performed GSEA with immunologic signatures gene sets, observing a high enrichment in the polarization of T-cell differentiation (Supplementary Fig. 2). The regulation of downstream genes by YTHDF1 remains under investigation.

3.7 Effect of YTHDF1 amplification on the infiltration of immune cells in breast cancer

We applied the DNA SCNA in TIMER server (cistrome.shinyapps.io/timer) to illustrate the infiltration of immune cells in breast cancer patients with YTHDF1 amplification [27, 28]. Six tumor infiltration immune cells, including dendritic cells, macrophages, neutrophils, CD4+ T cells, CD8+ T cells, and B cells, were analyzed in the TCGA breast cancer cohort with genomic data. CD4+ T cells, macrophages, dendritic cells, and B cells were diminished in patients with YTHDF1 amplification versus diploid/normal patients (Fig. 7A). As expected, immune infiltrates in YTHDF1 high expression were lower than those in YTHDF1 low expression group (Fig. 7B). Given the GSEA result of enrichment in T-cell differentiation, these data showed the potential function of YTHDF1 in the regulation of immune cells by affecting the modification of m6A.

Figure 7.

Tumor-infiltrating immune cell comparison according to YTHDF1 status. (A) Abundance of tumor-infiltrating immune cell for YTHDF1 copy number status in breast cancer in The Cancer Genome Atlas (TCGA). (B) Abundance of tumor-infiltrating immune cell for YTHDF1 expression status in breast cancer in TCGA. The abundance of six tumor-infiltrating immune cell types (B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells) in TCGA are shown. Box plots are presented to show the distributions of each tumor-infiltrating immune cell subset for selected group. $P$ -value Significant Codes: 0 $\leqslant$ *** $<$ 0.001 $\leqslant$ ** $<$ 0.01 $\leqslant$ *s $<$ 0.05 $\leqslant$ . $<$ 0.1.

Figure 8.

Effects of YTHDF1 down-regulation on MDA-MB-231 and MCF-7 cell proliferation, migration and invasion. (A) YTHDF1 mRNA expressions were detected in MDA-MB-231 and MCF-7 cells after transfected with psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids. (B) CCK8 assays showed the cell proliferation of MDA-MB-231 and MCF-7 cells after transfected with psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids. The survival cells were detected on day 0, 1, 2, 3, 5. (C) Wound healing assays revealed the migration ability of psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids transfected MDA-MB-231 and MCF-7 cells. Upper: Representative micrographs of wound healing assay. Below: The quantification data for migration rates are shown in histograms. (D) Transwell assays showed the migration ability and invasion ability of MDA-MB-231 and MCF-7 cells after transfected with psi-LVRU6GP vector, shRNA-1 and shRNA-2 plasmids. Left: Representative micrographs of transwell migration assays. Right: Representative micrographs of matrigel invasion assays. The quantification data for migration and invasion numbers are shown in histograms. All data are the means $\pm$ SEM of triplicates from three independent experiments. * $P<$ 0.05, ** $P<$ 0.01, *** $P<$ 0.001 vs. Vector.

3.8 Knockdown of YTHDF1 inhibited breast cancer cell proliferation, migration and invasion

To explore the loss of function of YTHDF1 in breast cancer cell, we transfected two breast cancer cell lines, MDA-MB-231 and MCF-7, with two independent shRNAs targeted YTHDF1 mRNA. YTHDF1 mRNA expression were confirmed down-regulated in shRNA-1 and shRNA-2 groups in both MDA-MB-231 and MCF-7 cells (Fig. 8A). After transfected with shRNA-1 and shRNA-2, MDA-MB-231 and MCF-7 cells showed lower proliferation in CCK8 assays (Fig. 8B). The wound healing assays indicated the migration ability were inhibited by shRNA-1 and shRNA-2 in both MDA-MB-231 and MCF-7 cells (Fig. 8C). The transwell assays revealed the migration and invasion were restrained by knockdown of YTHDF1 (Fig. 8D).

4. Discussion

Large-scale, next-generation sequencing allows us to understand the biogenesis of breast cancer in a convenient manner to determine the mutation landscape of breast cancer. Thus, copy number alterations (e.g., ERBB2 amplification [2], MYC amplification [31], FGFR1 amplification [32], and CCND1 amplification [33]) offer great potential for the development of gene-targeted therapy. Molecular profiling has revealed that breast cancer is heterogeneous, and an increasing number of CNVs in driver genes and targetable genes have been identified [7]. In a recent study of clear cell renal cell carcinoma (ccRCC), YTHDF2 showed a high alteration rate (55.11%), followed by METTL3 (30.11%) [18]. Although changes in the expression of m6A regulatory genes have been investigated, CNVs in breast cancer remain unclear. For this reason, we performed the first investigation of CNVs of m6A regulatory genes in patients with breast cancer.

The frequency of CNVs among the 10 m6A regulatory genes examined was relatively lower than those reported in ccRCC [18] and acute myeloid leukemia (AML) [34]. More importantly, other CNVs types (e.g., deep deletion, shallow deletion, and copy number gain) were more likely to occur in ccRCC and AML rather than the predominant amplification observed in our cohorts. Notably, the frequency of YTHDF1 (22.73%) in ccRCC was also higher than that recorded in our METABRIC results (14%). Compare to the low frequency of CNVs in breast cancer in TCGA, the frequency of CNVs in METABRIC was higher. Hence, we performed a clinicopathologic correlation analysis with METABRIC. Moreover, it is also worth noting that the 10 m6A regulatory genes exhibited a low concurrent alteration rate, except the YTHDF1 and YTHDF3 overlap. This result indicated that CNVs in regulatory genes may function independently.

In addition to the high frequency of CNVs, the functions of YTHDF1 and YTHDF3 show a great potential in the pathogenesis of breast cancer. YTH domain family genes interplay with hundreds of mutual targets. In the cytoplasm, m6A-binding protein YTHDF1 recruits m6A-modified mRNA to accelerate the recruitment of translation initiation factors and ribosomes, thereby increasing the efficiency of translation [10]. Shi et al. described that YTHDF3 interplays with YTHDF1 and modulates the translation of target cellular mRNA in HeLa cells [17]. Thus, we estimated the effect of YTHDF1 and YTHDF3 in the survival of patients with breast cancer. Since there are few events of amplification and shorter follow-up period in TCGA cohort, we finally used METABRIC to perform survival and Cox regression analysis. According to our results, YTHDF1 and YTHDF3 were significantly related to OS. Obviously, YTHDF1 exerted a more profound effect on OS, considering that patients with YTHDF1 amplification were associated with a poor outcome. The prognostic effects were also confirmed by Cox regression analysis. Similar to our results, CNVs in ccRCC [18] and AML [34] also predicted worse survival. Because recurrence data was unknown in METATRIC, RFS and DMFS were not compared. In addiction, with the use of KM-plotter, we confirmed the effect of YTHDF1 and YTHDF3 expression on survival including OS, RFS and DMFS with a larger cohort.

YTHDF3, the third member of the YTH family, shares $>$ 65% protein sequence identity with YTHDF1. In our results, the correlation with clinicopathologic factors and prognostic value of YTHDF3 was not as important as that observed for YTHDF1. Of note, YTHDF3 may collaborates with YTHDF1. However, YTHDF3 was partially amplified concurrently with YTHDF1; this was inconsistent with results showing that other m6A regulator CNVs had no concurrence. Moreover, they may not be simultaneously amplified, but function in cooperation. Indeed, we also evaluated the effects of other eight m6A regulatory genes. Unlike previous reports of ccRCC and AML, we did not observe significance in the CNVs of those genes. The low frequency of CNVs in these genes renders them inappropriate for inclusion in Kaplan-Meier survival analysis.

Amplification of the YTHDF1 and YTHDF3 m6A regulatory genes caused an increase in mRNA expression. The effect of YTHDF1 and YTHDF3 amplification in prognosis may be closely related to high mRNA expression; however, such evidence has not been previously reported. The mRNA expression in m6A “writer” genes (WTAP, METTL3 and METTL14) was lower in breast cancer tumors than in normal tissues [35]. In addition, the m6A “eraser” gene FTO showed high expression in breast cancer tumors, causing BNIP3 m6A demethylation and subsequent progression of breast cancer [36]. m6A “reader” genes, including YTHDF1 and YTHDF3, were not well investigated in the above two reports. Herein, we reported for the first time that YTHDF1 and YTHDF3 may be novel targets of m6A regulation in breast cancer.

A more recent study revealed that antigen-specific CD8+ T cell antitumor immune response was elevated in YTHDF1-knockout mice [37]. Lysosomal proteases encoding m6A-modified mRNA binds with YTHDF1 in dendritic cells, and subsequently suppresses the cross-priming of CD8+ T cells and cross-presentation of tumor antigens. More interestingly, the effect of anti-PD-L1 therapy in YTHDF1-knockout mice is elevated, showing the potential of YTHDF1 as a target for checkpoint inhibition therapy. We found that macrophages, dendritic cells, and B cells were reduced in patients with YTHDF1 amplification, prompting that YTHDF1 may regulate the infiltration of immune cells. They also confirmed that colon cancer patients with low YTHDF1 expression present lower CD8+ T cell infiltration; however, this result was not confirmed in breast cancer. In the current study, we found that CD4+ T cells, macrophages, dendritic cells, and B cells exhibited less infiltration in patients with YTHDF1 amplification. Nevertheless, the mechanism through which YTHDF1 influences the tumor microenvironment by affecting m6A modification in breast cancer remains unclear. Moreover, highly enriched gene sets in polarization of T-cell differentiation were found in the YTHDF1 amplification group, indicating that YTHDF1 may regulate T-cell differentiation through m6A modification. Hence, YTHDF1 appears to be a novel target for immune checkpoint therapy.

Of note, $>$ 50% (80/147) of patients with YTHDF1 amplification concurrently had MYC amplification, and the GSEA results showed enrichment in the subgroup of genes regulated by MYC. These findings imply that YTHDF1 may be involved in the MYC signaling pathway. METTL14 regulates MYC mRNA through m6A modification, enhancing the stability and translation of MYC in the pathogenesis of AML [38]. FTO also promotes the stability and translation of MYC by blocking the YTHDF2-mediated RNA decay in MYC mRNA [39]. Thus far, there are no previous studies providing evidence on the function of YTHDF1 in the MYC signaling pathway in breast cancer. A study investigating AML reported that CNVs in m6A regulatory genes show a close correlation with TP53 mutation [34]. In the present study, we also found a high concurrence in YTHDF1 amplification and TP53 mutation. We previously reported that ALKBH5, another m6A “eraser” gene, regulates m6A demethylation in NANOG mRNA, enhancing breast cancer stem cells in the hypoxia inducible factor-dependent hypoxic tumor microenvironment [40]. Observing the in vitro study, breast cancer cell proliferation and migration were restrained by knockdown of YTHDF1. But the oncogenic effect of YTHDF1 and the mechanism through which YTHDF1 functions downstream in the pathogenesis and progression of breast cancer require further investigation.

In conclusion, the present analysis illustrated the genetic alterations of m6A regulatory genes in breast cancer. The findings revealed a significant correlation between clinical characteristics and YTHDF1 amplification. YTHDF1 amplification confers a poorer outcome and is an independent risk factor for 10-year OS in patients with breast cancer. Moreover, YTHDF1 amplification resulted in less infiltration of immune cells. The potential regulator/downstream targets of YTHDF1 in the pathogenesis and progression of breast cancer may be the MYC signaling pathway and T-cell differentiation. YTHDF1 down-regulation inhibits proliferation, migration and invasion in breast cancer cells in vitro. However, the oncogenic effect of YTHDF1 still requires further mechanistic studies.

Author contributions

For every author, his or her contribution to the manuscript needs to be provided using the following categories:

Conception: Ning Liao; Cheukfai Li; Chuanzhao Zhang.

Interpretation or analysis of data: Cheukfai Li; Chuanzhao Zhang.

Preparation of the manuscript: Cheukfai Li; Chuanzhao Zhang.

Revision for important intellectual content: Guochun Zhang; Bo Chen; Xuerui Li; Kai Li; Chongyang Ren; Lingzhu Wen; Ning Liao.

Supervision: Ning Liao.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-203103.

sj-docx-1-cbm-10.3233_CBM-203103.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-203103.docx

Footnotes

Acknowledgments

This study was supported by funding from Guangdong Basic and Applied Basic Research Foundation (2021A1515111003), Research Funds from Guangzhou Municipal Science and Technology Project (202102020 195), Project of Administration of Traditional Chinese Medicine of Guangdong Province of China (20221002) and Project of Doctor scientific research from Guangdong Provincial People’s Hospital (2020bq12).

References

Waks

A.G.

and Winer

E.P.

, Breast cancer treatment: A review, JAMA 321 (2019), 288–300.

D.Y.

and Bang

Y.J.

, HER2-targeted therapies – a role beyond breast cancer, Nat Rev Clin Oncol (2019).

Voorwerk

Slagter

Horlings

H.M.

Sikorska

van de Vijver

K.K.

de Maaker

Nederlof

Kluin

R.J.C.

Warren

Ong

Wiersma

T.G.

Russell

N.S.

Lalezari

Schouten

P.C.

Bakker

N.A.M.

Ketelaars

S.L.C.

Peters

Lange

C.A.H.

van Werkhoven

van Tinteren

Mandjes

I.A.M.

Kemper

Onderwater

Chalabi

Wilgenhof

Haanen

Salgado

de Visser

K.E.

Sonke

G.S.

Wessels

L.F.A.

Linn

S.C.

Schumacher

T.N.

Blank

C.U.

and Kok

, Immune induction strategies in metastatic triple-negative breast cancer to enhance the sensitivity to PD-1 blockade: the TONIC trial, Nat Med 25 (2019), 920–928.

Schmid

Adams

Rugo

H.S.

Schneeweiss

Barrios

C.H.

Iwata

Dieras

Hegg

S.A.

Shaw Wright

Henschel

Molinero

Chui

S.Y.

Funke

Husain

Winer

E.P.

Loi

Emens

L.A.

and Investigators

I.M.T.

, Atezolizumab and Nab-paclitaxel in advanced triple-negative breast cancer, N Engl J Med 379 (2018), 2108–2121.

Turner

N.C.

Slamon

D.J.

Bondarenko

S.A.

Masuda

Colleoni

DeMichele

Loi

Verma

Iwata

Harbeck

Loibl

Andre

Puyana Theall

Huang

Giorgetti

Huang Bartlett

and Cristofanilli

, Overall survival with palbociclib and fulvestrant in advanced breast cancer, N Engl J Med 379 (2018), 1926–1936.

Cancer Genome Atlas

, Comprehensive molecular portraits of human breast tumours, Nature 490 (2012), 61–70.

Kalimutho

Nones

Srihari

Duijf

P.H.G.

Waddell

and Khanna

K.K.

, Patterns of Genomic Instability in Breast Cancer, Trends Pharmacol Sci 40 (2019), 198–211.

Chin

DeVries

Fridlyand

Spellman

P.T.

Roydasgupta

Kuo

W.L.

Lapuk

Neve

R.M.

Qian

Ryder

Chen

Feiler

Tokuyasu

Kingsley

Dairkee

Meng

Chew

Pinkel

Jain

Ljung

B.M.

Esserman

Albertson

D.G.

Waldman

F.M.

and Gray

J.W.

, Genomic and transcriptional aberrations linked to breast cancer pathophysiologies, Cancer Cell 10 (2006), 529–41.

Desrosiers

Friderici

and Rottman

, Identification of methylated nucleosides in messenger RNA from Novikoff hepatoma cells, Proc Natl Acad Sci USA 71 (1974), 3971–5.

10.

Wang

Zhao

B.S.

Roundtree

I.A.

Han

Weng

Chen

Shi

and He

, N(6)-methyladenosine Modulates Messenger RNA Translation Efficiency, Cell 161 (2015), 1388–99.

11.

Frye

Harada

B.T.

Behm

and He

, RNA modifications modulate gene expression during development, Science 361 (2018), 1346–1349.

12.

Roundtree

I.A.

Evans

M.E.

Pan

and He

, Dynamic RNA modifications in gene expression regulation, Cell 169 (2017), 1187–1200.

13.

Barbieri

Tzelepis

Pandolfini

Shi

Millan-Zambrano

Robson

S.C.

Aspris

Migliori

Bannister

A.J.

Han

De Braekeleer

Ponstingl

Hendrick

Vakoc

C.R.

Vassiliou

G.S.

and Kouzarides

, Promoter-bound METTL3 maintains myeloid leukaemia by m(6)A-dependent translation control, Nature 552 (2017), 126–131.

14.

Shi

Wei

and He

, Where, When, and How: Context-Dependent Functions of RNA Methylation Writers, Readers, and Erasers, Mol Cell 74 (2019), 640–650.

15.

Liu

Dai

Zheng

Parisien

and Pan

, N(6)-methyladenosine-dependent RNA structural switches regulate RNA-protein interactions, Nature 518 (2015), 560–4.

16.

Zhou

Wan

Gao

Zhang

Jaffrey

S.R.

and Qian

S.B.

, Dynamic m(6)A mRNA methylation directs translational control of heat shock response, Nature 526 (2015), 591–4.

17.

Shi

Wang

Zhao

B.S.

Hsu

P.J.

Liu

and He

, YTHDF3 facilitates translation and decay of N(6)-methyladenosine-modified RNA, Cell Res 27 (2017), 315–328.

18.

Qian

J.Y.

Gao

Sun

Cao

M.D.

Shi

Xia

T.S.

Zhou

W.B.

Wang

Ding

and Wei

J.F.

, KIAA1429 acts as an oncogenic factor in breast cancer by regulating CDK1 in an N6-methyladenosine-independent manner, Oncogene 38 (2019), 6123–6141.

19.

Cerami

Gao

Dogrusoz

Gross

B.E.

Sumer

S.O.

Aksoy

B.A.

Jacobsen

Byrne

C.J.

Heuer

M.L.

Larsson

Antipin

Reva

Goldberg

A.P.

Sander

and Schultz

, The cBio cancer genomics portal: An open platform for exploring multidimensional cancer genomics data, Cancer Discov 2 (2012), 401–4.

20.

Gao

Aksoy

B.A.

Dogrusoz

Dresdner

Gross

Sumer

S.O.

Sun

Jacobsen

Sinha

Larsson

Cerami

Sander

and Schultz

, Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal, Sci Signal 6 (2013), pl1.

21.

Pereira

Chin

S.-F.

Rueda

O.M.

Vollan

H.-K.M.

Provenzano

Bardwell

H.A.

Pugh

Jones

Russell

Sammut

S.-J.

Tsui

D.W.Y.

Liu

Dawson

S.-J.

Abraham

Northen

Peden

J.F.

Mukherjee

Turashvili

Green

A.R.

McKinney

Oloumi

Shah

Rosenfeld

Murphy

Bentley

D.R.

Ellis

I.O.

Purushotham

Pinder

S.E.

Børresen-Dale

A.-L.

Earl

H.M.

Pharoah

P.D.

Ross

M.T.

Aparicio

and Caldas

, The somatic mutation profiles of 2433, breast cancers refine their genomic and transcriptomic landscapes, Nature Communications 7 (2016).

22.

Gyorffy

Lanczky

Eklund

A.C.

Denkert

Budczies

and Szallasi

, An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients, Breast Cancer Res Treat 123 (2010), 725–31.

23.

Subramanian

Tamayo

Mootha

V.K.

Mukherjee

Ebert

B.L.

Gillette

M.A.

Paulovich

Pomeroy

S.L.

Golub

T.R.

Lander

E.S.

and Mesirov

J.P.

, Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles, Proc Natl Acad Sci USA 102 (2005), 15545–50.

24.

Huang da

Sherman

B.T.

and Lempicki

R.A.

, Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists, Nucleic Acids Res 37 (2009), 1–13.

25.

Chen

Chandrashekar

D.S.

Varambally

and Creighton

C.J.

, Pan-cancer molecular subtypes revealed by mass-spectrometry-based proteomic characterization of more than 500 human cancers, Nat Commun 10 (2019), 5679.

26.

Chandrashekar

D.S.

Bashel

Balasubramanya

S.A.H.

Creighton

C.J.

Ponce-Rodriguez

Chakravarthi

and Varambally

, UALCAN: A portal for facilitating tumor subgroup gene expression and survival analyses, Neoplasia 19 (2017), 649–658.

27.

Fan

Wang

Traugh

Chen

Liu

J.S.

and Liu

X.S.

, TIMER: A web server for comprehensive analysis of tumor-infiltrating immune cells, Cancer Res 77 (2017), e108–e110.

28.

Severson

Pignon

J.C.

Zhao

Novak

Jiang

Shen

Aster

J.C.

Rodig

Signoretti

Liu

J.S.

and Liu

X.S.

, Comprehensive analyses of tumor immunity: implications for cancer immunotherapy, Genome Biol 17 (2016), 174.

29.

Mermel

C.H.

Schumacher

S.E.

Hill

Meyerson

M.L.

Beroukhim

and Getz

, GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers, Genome Biol 12 (2011), R41.

30.

Hsu

P.J.

Zhu

Guo

Shi

Liu

Shi

Wang

Cheng

Luo

Dai

Liu

Guo

Sha

Shen

and He

, Ythdc2 is an N6-methyladenosine binding protein that regulates mammalian spermatogenesis, Cell Research 27 (2017), 1115–1127.

31.

Zhang

Wang

Chen

Guo

Cao

Ren

Wen

Jia

Mok

Chen

Wei

Lin

Zhang

Hou

Han-Zhang

Liu

Balch

C.M.

Meric-Bernstam

and Liao

, Characterization of frequently mutated cancer genes in Chinese breast tumors: a comparison of Chinese and TCGA cohorts, Ann Transl Med 7 (2019), 179.

32.

Hortobagyi

G.N.

Chen

Piccart

Rugo

H.S.

Burris

H.A.

, 3rd Pritchard

K.I.

Campone

Noguchi

Perez

A.T.

Deleu

Shtivelband

Masuda

Dakhil

Anderson

Robinson

D.M.

Garg

McDonald

E.R.

, 3rd Bitter

Huang

Taran

Bachelot

Lebrun

Lebwohl

and Baselga

, Correlative analysis of genetic alterations and everolimus benefit in hormone receptor-positive, human epidermal growthfactor receptor 2-negative advanced breast cancer: Results from BOLERO-2, J Clin Oncol 34 (2016), 419–26.

33.

Gong

Litchfield

L.M.

Webster

Chio

L.C.

Wong

S.S.

Stewart

T.R.

Dowless

Dempsey

Zeng

Torres

Boehnke

Mur

Marugan

Baquero

Bray

S.M.

Wulur

I.H.

Chu

Qian

H.R.

Iversen

P.W.

Merzoug

F.F.

X.S.

Reinhard

De Dios

Caldwell

C.W.

Lallena

M.J.

Beckmann

R.P.

and Buchanan

S.G.

, Genomic aberrations that activate D-type cyclins are associated with enhanced sensitivity to the CDK4 and CDK6 inhibitor abemaciclib, Cancer Cell 32 (2017), 761–776 e6.

34.

Kwok

C.T.

Marshall

A.D.

Rasko

J.E.

and Wong

J.J.

, Genetic alterations of m(6)A regulators predict poorer survival in acute myeloid leukemia, J Hematol Oncol 10 (2017), 39.

35.

Ning

Liu

and Zhang

, Changes of N6-methyladenosine modulators promote breast cancer progression, BMC Cancer 19 (2019).

36.

Niu

Lin

Wan

Chen

Liang

Sun

Wang

Xiong

X.F.

Wei

and Wan

, RNA N6-methyladenosine demethylase FTO promotes breast tumor progression through inhibiting BNIP3, Mol Cancer 18 (2019), 46.

37.

Han

Liu

Chen

Dong

Liu

Chang

Huang

Liu

Wang

Dougherty

Bissonnette

M.B.

Shen

Weichselbaum

R.R.

M.M.

and He

, Anti-tumour immunity controlled through mRNA m(6)A methylation and YTHDF1 in dendritic cells, Nature 566 (2019), 270–274.

38.

Weng

Huang

Qin

Zhao

B.S.

Dong

Shi

Skibbe

Shen

Sheng

Wang

Wunderlich

Zhang

Dore

L.C.

Deng

Ferchen

Sun

Jiang

Marcucci

Mulloy

J.C.

Yang

Qian

Wei

and Chen

, METTL14 Inhibits Hematopoietic Stem/Progenitor Differentiation and Promotes Leukemogenesis via mRNA m(6)A Modification, Cell Stem Cell 22 (2018), 191–205 e9.

39.

Dong

Nachtergaele

Wunderlich

Qing

Deng

Wang

Weng

Skibbe

Dai

Zou

Weng

Huang

Ferchen

Qin

Zhang

Sasaki

A.T.

Plas

D.R.

Bradner

J.E.

Wei

Marcucci

Jiang

Mulloy

J.C.

Jin

and Chen

, R-2HG Exhibits Anti-tumor Activity by Targeting FTO/m(6)A/MYC/CEBPA Signaling, Cell 172 (2018), 90–105 e23.

40.

Zhang

Samanta

Bullen

J.W.

Zhang

Chen

and Semenza

G.L.

, Hypoxia induces the breast cancer stem cell phenotype by HIF-dependent and ALKBH5-mediated m6A-demethylation of NANOG mRNA, Proceedings of the National Academy of Sciences 113 (2016), E2047–E2056.

YTHDF1 amplification is correlated with worse outcome and lower immune cell infiltrations in breast cancer

Abstract

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

Abbreviations

1. Introduction

2. Material and methods

2.1 Cohort data collection

2.2 Kaplan-Meier plotter

2.3 Gene set enrichment analysis (GSEA)

2.4 UALCAN and clinical proteomic tumor analysis consortium (CPTAC)

2.5 Immune cell infiltration analysis

Table 1 Association between YTHDF1 amplification status and clinicopathological characteristics

2.7 Wound healing assay

Table 2 Association between YTHDF3 amplification status and clinicopathological characteristics

3. Results

3.1 CNVs of m6A regulatory genes in patients with breast cancer

3.2 YTHDF1 and YTHDF3 amplification were associated with higher expression

3.4 Association between YTHDF1 and YTHDF3 amplification and survival of patients with breast cancer

Table 3 Univariate and multivariate Cox proportional hazard analysis for 10-year overall survival

3.6 Enrichment analysis of YTHDF1 amplification

3.7 Effect of YTHDF1 amplification on the infiltration of immune cells in breast cancer

4. Discussion

Author contributions

Supplementary data

sj-docx-1-cbm-10.3233_CBM-203103.docx - Supplemental material

Footnotes

Acknowledgments

References

Table 1
Association between YTHDF1 amplification status and clinicopathological characteristics

Table 2
Association between YTHDF3 amplification status and clinicopathological characteristics

Table 3
Univariate and multivariate Cox proportional hazard analysis for 10-year overall survival