2.2. Canine mammary gland tumor (CMT) primary cell line establishment

The primary CMT cell line was established directly from digested CMT tissues obtained from the Veterinary Medical Teaching Hospital of Seoul National University using TrypLE Express solution containing the Rho kinase inhibitor Y-27632, DNase I, and Collagenase/Dispase. The digested cells were filtered and equilibrated in PBS containing 1X Mammary Epithelial Growth Supplement (MEGS). Primary cells were first cultured in advanced DMEM/F12 supplemented with 2 mM L-glutamine, penicillin (100 U/mL)–streptomycin (100 μg/mL), and 1X MEGS. Subsequent cell cultures were grown in DMEM with the same supplementation as mentioned above, except without additional L-glutamine and replacing MEGS with 10 % fetal bovine serum (FBS). All cells were grown in a humidified incubator at 37 °C with 5 % CO₂ and were tested regularly for mycoplasma contamination.

2.3. Cell culture

Human breast cancer cell lines and breast epithelial cell line MCF-10A were obtained from the Korean Cell Line Bank and cultured as previously reported [21]. Briefly, normal epithelial MCF-10A cells were cultured in MEBM Basal Medium (CC-3151; Lonza, Basel, Switzerland) or MEGM supplemented medium (CC4136; Lonza, Basel, Switzerland). SKBR3 cells were cultured in RPMI 1640 medium supplemented with L-glutamine (SH30027.01, Hyclone, UT, USA). MCF-7 cells were cultured in DMEM/high glucose supplemented with L-glutamine and sodium pyruvate (SH30243.01; Hyclone, UT, USA). CHMp canine mammary gland adenocarcinoma cell line was purchased from N. Sasaki lab [22] and grown in RPMI 1640 medium (HyClone, SH30027) containing 10 % FBS; Gibco 1600044) and 50 μg/mL gentamicin (Sigma‒Aldrich, G1272). All cell lines were cultured at 37 °C in a 5 % CO₂ atmosphere. All cells were cultured in a humid incubator at 37 °C with 5 % CO₂, and their mycoplasma contamination was verified to be absent.

2.4. Expression and human ChIP-seq data set

Online datasets were retrieved from Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) and The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/). This study utilized three public cohorts, GSE65194 and GSE119810, which encompass microarray expression profiles of normal and cancerous human breast and canine mammary tissues, respectively. Additionally, GSE85158 provided the H3K4me3 ChIP-sequencing data for human breast cancer cell lines. Student's t-tests were used to detect significant differences between two or more groups.

2.5. siRNA transfection

A total of 30 pmol each of ALYREF siRNA (Genolution, Seoul, South Korea) and SN-1012 (Bioneer, Daejeon, Republic of Korea) as negative controls were transfected with Lipofectamine RNAiMAX transfection reagent (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. The cells were harvested 48 and 72 h after incubation for further analysis. The siRNA sequences are listed in Supplementary Table 1.

2.6. RNA extraction and quantitative RT-PCR

RNA extraction and quantitative RT-qPCR (qRT-PCR) were performed as previously reported [21]. Briefly, total RNAs were extracted using the TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. Total RNA (2 μg) was reverse transcribed into cDNA using the CellScript All‐in‐One 5 × First Strand cDNA Synthesis Master Mix (CellSafe, Yong-In, Republic of Korea) according to the manufacturer's instructions. The primers used for RT-qPCR are listed in Supplementary Table 1. The samples were run in triplicate and the expression of each target gene mRNA level was analyzed using the 2-ΔΔCT method and normalized to GAPDH expression in humans and dogs.

2.7. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis

GO and KEGG pathway enrichment analyses were performed using Database for Annotation, Visualization, and Integrated Discovery (DAVID) bioinformatics resources to predict the gene function of identified differentially histone-modified genes. The promoter region was defined as 2 kb upstream and downstream of the TSS.

2.8. Cell growth assay

Cell viability was determined using a Cell Counting Kit-8 (CCK-8; Dojindo, Tokyo, Japan) as previously described [23]. Briefly, cells were seeded into 96-well plates at 5000 cells for CHMp and 7500 cells (MCF7 and SKBR3) per well. A volume of 10 μL of CCK-8 reagent was added to each well. Plates were incubated for 3 h at 37 °C, and the absorbance value (OD) of each well was measured at 450 nm using a spectrophotometer (Epoch Microplate spectrophotometer, BioTek Instruments, Winooski, VT, USA) according to the manufacturer's instructions.

2.9. Cell migration assay

Cell migration was assessed as previously described [24]. The assay was performed using SPL Transwell inserts (#37224; SPL Life Sciences, Seoul, Korea) in 24-well plates. For all cells, approximately 5 × 104 cells in 200 μL of DMEM/high glucose medium (Hyclone, Logan, UT, USA) supplemented with 10 % US origin FBS (Hyclone, Logan, UT, USA) were placed in the chamber, and 500 μL of the same medium was added in the chamber. The plates were incubated at 37 °C in 5 % CO₂. After washing, cells were fixed and stained with crystal violet. Cells were counted under a microscope using ToupView software.

2.10. Colony formation assay

Five hundred cells were seeded 24 h after transfection in a 6-well plate with the appropriate media. After incubating the cells for 10–14 days at 37 °C in 5 % CO₂, they were fixed and stained with 2.5 % crystal violet and the number of colonies was counted.

2.11. Genome-wide ChIP-seq and ChIP-PCR

ChIP assays were performed on canine mammary tissue and human breast cancer cell lines to evaluate H3K4me3 (Abcam, ab8580, USA) and H3K27me3 (Abcam, ab6002, USA) enrichment, according to a previous study. Briefly, the samples were fixed with 1 % formaldehyde in a rocker at room temperature for 10 min to cross-link histones to DNA, and the reaction was stopped by adding glycine to a final concentration of 125 mM. The DNA was then sheared by sonication to generate 300–600 bp DNA fragments. Immunoprecipitation was performed with the H3K4me3 antibody at a concentration of 5 μg per 25 μg chromatin, and the same amount of normal IgG (AC-005, Abclonal, China) was used as control. The precipitated DNA was detected by RT-PCR using specific primers (Supplementary Table S1). Across all generated in-house ChIP-seq datasets, the CanFam3.1 genome was used as a reference genome assembly. The ChIP-seq data were processed as in our previous study [25].

2.12. Protein extraction and western blot analysis

Protein extraction and western blotting were performed as described previously [21]. Immunoblotting was performed and antibodies specific for ALYREF (1:1000, A4298, Abclonal, China), BAX (1:500, a12009, Abclonal, China), and B-Actin (1:1000, sc-47778, Santa Cruz Biotechnology, Santa Cruz, USA). Images were quantified using ImageJ and relative optical densities were calculated by adjusting to B-actin loading control.

3.2. Gene ontology (GO) and KEGG pathway analysis of the gene set differentially marked by H3K4me3

To investigate the H3K4me3-marked genes that were aberrantly enriched in tumors, GO functional analysis was performed using DAVID. The results revealed diverse cancer-associated terms. A total of 329 genes with a more than 1.5-fold highly enriched H3K4me3 marks in tumors compared with adjacent normal tissues were categorized into 41 subcategories in biological processes (BP), 23 subcategories in molecular function (MF), and 28 subcategories in the cellular compartment (CC) (data not shown). The top 10 GO terms in the BP categories are listed according to the p-value in Fig. 2a and b. We observed processes associated with carcinogenesis, such as cell cycle regulation (GO:0051726) organic substance metabolic processes (GO:0071704), and upregulated GO. These GO terms were upregulated in human breast cancer [26,27]. The terms of H3K4me3 downregulated genes were highly enriched in the nervous system (GO:007399)-related terms (Fig. 2b).

Open in a separate window

Fig. 2

Gene ontology (GO) and KEGG pathway analysis.

(a–c) Enrichment analysis of gene ontology (GO) and KEGG pathway using differentially histone-marked genes in CMT. The top 10 significant GO_BP terms and the top 5 KEGG terms are depicted. (a) Gene Ontology (GO) analysis of genes exhibiting a 1.5-fold increase and (b) decrease in H3K4me3, and (c) KEGG analysis of genes with increased H3K4me3.

KEGG pathway analysis revealed that genes with high levels of H3K4me3 were predominantly enriched in metabolism-related pathways, including those for drugs, nucleotides, and purines. Among the 39 genes associated with these pathways, we identified several well-known breast cancer-related genes, including SQLE, SRD5A1, and UAP1. Analysis of human cohorts and Kaplan–Meier (KM) plots indicated that the upregulation of SQLE, SRD5A1, and UAP1 was associated with poor prognosis and reduced survival in patients with breast cancer [[28], [29], [30]]. To validate the impact of H3K4me3 histone modifications on metabolism-related gene expression in cancer cells, their expression levels were assessed. The expression of all three genes was upregulated in CMT and human breast cancer cell lines compared to normal cells (Supplementary Figs. 1a–c). A web-based Kaplan–Meier plotter (https://kmplot.com/analysis/index.php?p=service) was used to draw the KM plots. The KM plots revealed that SQLE, SRD5A1, and UAP11 upregulation led to poor prognosis and survival in patients with breast cancer.

In summary, we found that the upregulation of metabolism-related genes SRD5A1, UAP1, and SQLE by higher H3K4me3 marks in CMT and their higher expression levels were also demonstrated in both canine and human breast cancers (Supplementary Figs. 1a–c). Our findings suggest that histone modifications could potentially influence expression changes in genes associated with metabolism.

3.3. H3K4me3 modification of ALYREF promoter changes in canine mammary tumor

The top 20 genes in the H3K4me3 highly enriched group are listed in Table 2 based on their respective p-values. The KM plot (Supplementary Figs. 2a–h) revealed that recurrence-free survival (RFS) was significantly less favorable in patients with high KIAA0895, SQLE, TONSL, COMMD5, DOLK, CCDC137, ALYREF, and P3H4 (SC65) expression (Table 2 and Supplementary Figs. 2a–h). We conducted a literature search on KIAA0895, SQLE [28,31,32], TONSL [33,34], COMMD5 [35,36], DOLK, CCDC137 [37], ALYREF [17,19,20], and P3H4(SC65) [38,39] to identify gene signatures associated with breast cancer. Despite having the lowest p-value in the KM plot, we found that research on ALYREF in hormone receptor-positive breast cancer is lacking, except for TNBC [20]. ALYREF, implicated in mRNA expression and RNA splicing, has shown elevated expression in numerous cancer patients based on data from TCGA. Despite the relevance of ALYREF in various types of cancer, including its potential involvement in tumorigenesis, its role in hormone receptor-positive breast cancer remains unclear. Consequently, we conducted subsequent experiments utilizing ALYREF.

We hypothesized that ALYREF might be involved in the malignant transformation of breast cancer in dogs and humans. To test this hypothesis, we initially conducted a comparative analysis of the ALYREF gene structure in humans, dogs, and mice (Fig. 3a). DNA sequence homology and the NCBI database were used to construct a phylogenetic tree based on the amino acid sequences of ALYREF isolated from humans, canines, and mice (Fig. 3b and c). The genetic architectures of humans and dogs exhibited a greater degree of similarity than those of humans and mice. Moreover, DNA and amino acid sequences displayed a high level of homology between humans and dogs.

Open in a separate window

Open in a separate window

Fig. 3

Comparison of ALYREF orthologous regions of the human, dog, and mouse genome and analysis of ALYREF gene expression in cancer and normal specimens from Gene Expression Omnibus (GEO) dataset.

(a) Schematic structures showing ALYREF genes of humans, dogs, and mice. (b) Query cover and homology of ALYREF genes in DNA sequences between human and dog, and between human and mouse, respectively. (c) Query cover and homology in amino acid sequence between human and dog, and between human and mouse, respectively. (d) IGV browser screenshots of H3K4me3 ChIP-seq analysis in 3 pairs of CMTs. Each peak of H3K4me3 reveals respective specimens. The top three tracks reveal H3K4me3 signals in cancer tissues, others show signals in normal tissues. (e) Quantification of ALYREF expression in CHMp, a pair of primary cell lines. ALYREF expression was higher in CMT cell lines compared with normal cell lines. Data are presented as mean ± SEM (n = 3). The Student's t-test was used to analyze the data, and statistical analysis was performed. p < 0.05, p < 0.01, p < 0.001. (f–g) Quantification of ALYREF expression in CMT tissues (n = 158) and healthy controls (n = 64) using the GEO dataset (GSE119810). (f) Expression of the ALYREF gene was relatively higher in CMT tissues compared with normal tissues. (g) Expression of the ALYREF gene was relatively upregulated in CMT tissues based on tumor grade. The x-axis represents sample numbers and tumor grades, and the y-axis represents the FPKM value. (h) IGV browser screenshots depict H3K4me3 ChIP-seq peaks in breast cancer cell lines, utilizing the GEO dataset (GSE85185). The top three tracks reveal SKBR, MCF7, and MDA-MB-231 (red), and the bottom track signals represent a normal cell line, MCF 10A (blue). (i) Quantification of ALYREF expression in breast cancer cell lines. ALYREF expression was higher in breast cancer cell lines compared with normal cell lines. Data are presented as mean ± SEM (n = 3). The Student's t-test was used to analyze the data, and statistical analysis was performed. p < 0.05, p < 0.01, p < 0.001. (j–k) Quantification of ALYREF expression in human breast cancer tissues (n = 115) and healthy control (n = 11) using the GEO dataset (GSE65194). (j) Expression of the ALYREF gene was relatively higher in human breast cancer tissues compared with normal tissues. (k) Expression of the ALYREF gene was relatively upregulated in breast cancer tissues based on tumor grade. The x-axis represents sample numbers and tumor subtypes, and the y-axis represents relative expression.

To clarify the relationship between histone modification and ALYREF expression, we analyzed RNA-seq data from CMT tissues, a pair of primary CMT cell lines, and a CHMp cell line using RT-qPCR. Based on the ChIP-seq data obtained in our lab, we observed peaks of H3K4me3 at chr9: 397, 497-398, 880, located 710bp upstream from the TSS of ALYREF. These peaks were highly enriched in a fraction of CMT tissues compared to those in normal tissues (Fig. 3d). Additionally, RNA expression levels were higher in primary cancer and CHMp cell lines than in primary normal cells (Fig. 3e). Our previous RNA-seq study performed in CMT tissues, had shown that ALYREF expression was higher in cancer tissues than in normal tissues. Moreover, as the tumor grade of the cancer tissue increased, ALYREF expression levels tended to increase (Fig. 3f and g). Similar to the findings in canines, a comparable pattern was observed in humans, wherein breast cancer cell lines (SKBR3 and MCF7) exhibited elevated levels of H3K4me3and ALYREF expression (Fig. 3h and i) compared to normal cell lines. Moreover, ALYREF expression was significantly higher in the cancer tissues than in the normal tissues (Fig. 3j). As the tumor grade increased in cancer tissues, there was a concurrent increase in ALYREF expression (Fig. 3k). The correlation between ALYREF expression and cancer grade indicates the involvement of ALYREF in the malignant transformation of non-TNBC breast cancers.

3.4. ALYREF affects cell viability, colony formation and apoptosis

To investigate the function of ALYREF in breast cancer cells, ALYREF was depleted in three independent cell lines: CHMp, SKBR3, and MCF7. The depletion efficiency of ALYREF in cell lines was estimated by RT-qPCR and Western blot analyses (Fig. 4a–e; since the ALYREF antibody did not guarantee cross-reactivity in dogs, Western blot data for canines are not presented). The CCK-8 assay revealed that cell viability was significantly reduced in all three ALYREF-depleted cell lines (Fig. 4f–h). To demonstrate tumor formation in vitro, we performed a colony formation assay. Fewer colonies were formed in ALYREF-depleted cell lines than in scrambled siRNA control cells (Fig. 4i–k), suggesting that ALYREF affects tumor cell proliferation and colony formation.

Open in a separate window

Open in a separate window

Fig. 4

Cellular consequences of loss of ALYREF function in CMT and human breast cancer cells.

(a–c) Quantification of depletion efficiency of ALYREF in CHMp and different breast cancer cell lines after treatment with specific siRNAs against ALYREF or control siRNA 72 h after transfection on mRNA level via RT-qPCR using GAPDH as a reference gene. n = 3, ±SD. p < 0.001. (d–e) Verification of siRNA-mediated depletion efficiency of ALYREF in (d) SKBR3 and (e) MCF7 cells on protein level analyzed by Western blot using β-Actin as loading control. (f–h) Cellular growth assay in breast cancer cells over 72 h under control conditions (control siRNA; gray, untreated parental control; blue) or after siRNA-mediated depletion of ALYREF (red). n = 3, ±SD. p < 0.001, p < 0.01 (f) Depletion of ALYREF inhibited proliferation of the CMT cell line. Depletion of ALYREF inhibited the proliferation of (g) SKBR3 and (h) MCF7. (i–k) Colony formation assay in (i) CHMp, (j) SKBR3, and (k) MCF7. Below the bar graphs show representative stained colony formation figures with 3 wells each under control conditions or after siRNA-mediated ALYREF depletion. n = 3, ±SD. p < 0.001. (l–n) Flow-cytometry-based AnnexinV staining of the CMT cell line, SKBR3, and MCF7 72 h after silencing by control siRNA or siALYREF. Statistical significance between cells was assessed using Student's t-test. n = 3, ±SD. p < 0.05, p < 0.01, and p < 0.001.

After identifying ALYREF as an important factor in tumor cell viability and colony formation, we investigated the effect of ALYREF depletion on the migration of breast cancer and CMT cell lines. Despite ALYREF depletion, there were no significant differences in tumor cell migration and cell cycle distribution compared to control cells (Supplementary Figs. 3a–f).

Annexin V-fluorescence-activated cell sorting (FACS) was performed to determine the role of ALYREF in apoptosis. ALYREF depletion increased apoptosis in all three breast cancer cell lines (Fig. 4l–n).

We thank Dr. Kang-Hoon Lee, GeneOne Life Science, Inc., Seoul, Republic of Korea, for insightful discussions, Keun Hong Son for providing part of the bioinformatics analysis, and M. Aldonza for helping in establishing the CMT primary cell line. We would like to thank Wan Hee Kim, College of Veterinary Medicine, Seoul National University, for providing us with well-controlled clinical specimens.