Lnc‐TPT1‐AS1/CBP/ATIC Axis Mediated Purine Metabolism Activation Promotes Breast Cancer Progression

ABSTRACT

The purine biosynthetic pathway was recently identified to play a crucial role in breast cancer progression. However, little was known about the regulatory mechanisms of long non‐coding RNA in breast cancer purine metabolism. In this study, we discovered that LncRNA TPT1‐AS1 (TPT1‐AS1) was downregulated in breast cancer tissues. Its introduction in breast cancer cells markedly suppressed tumor growth and metastasis in xenograft tumor models. Mass spectrometric analysis suggested that the purine biosynthetic pathway was activated in TPT1‐AS1‐knockdown MCF‐7 cells. Inosine monophosphate (IMP), the product of de novo purine biosynthesis, was significantly upregulated. Mechanistically, we found that TPT1‐AS1 could physically interact with CBP (CREB‐binding protein), which consequently led to the loss of H3K27Ac in the promoter area of ATIC, the key enzyme of IMP synthesis. This process could block breast cancer purine metabolism and inhibit breast cancer progression. In conclusion, our findings illustrate the role of non‐coding RNAs in breast cancer purine metabolism reprogramming and present a potential candidate for breast cancer therapy.

Overexpression of TPT1‐AS1 could suppress the malignant phenotype and purine synthesis activation of breast cancer through competitive interaction with CBP, subsequently reducing the promoter H3K27ac signal and transcription of ATIC. Therefore, targeting the TPT1‐AS1/CBP/ATIC axis may provide a useful strategy for the treatment of patients with breast cancer.

Abbreviations

AMP

adenosine monophosphate

ATIC

aminoimidazole carboxamide ribonucleotide transformylase inosine monophosphate cyclohydrolase

CBP

CREB‐binding protein

ceRNA

competing endogenous RNA

ChIP

chromatin immunoprecipitation

FISH

fluorescent in situ hybridization

GMP

guanosine monophosphate

HAT

histone acetyltransferase

HRP

horseradish peroxidase conjugated

IF

immunofluorescence

IHC

immunohistochemical

IMP

inosine monophosphate

ISH

in situ hybridization

LncRNAs

long non‐coding RNAs

MS

mass spectrometry

PVDF

polyvinylidene fluoride

RIP

RNA‐binding protein immunoprecipitation

TPT1‐AS1

LncRNA TPT1‐AS1

WGCNA

weighted gene co‐expression network analysis

1. Introduction

As the most commonly diagnosed cancer worldwide, breast cancer poses a significant threat to women's health [1, 2]. During the past two decades, the incidence of breast cancer has been rising, with 2.3 million women diagnosed with breast cancer globally per year [3]. Despite the emerging diagnosis and treatment strategies, the prognosis for patients with relapse and metastasis is dismal. Therefore, it is crucial to find an alternative to address this situation, which could lead to new opportunities for treating breast cancer.

Disturbed purine metabolism is closely related to progression as well as metastasis in tumors. Studies have demonstrated that protein‐dependent purine biosynthesis notably inhibited therapy response in prostate cancer [4]. Meanwhile, the aberrant expression of purine biosynthesis‐related enzymes can also promote stem characteristics of tumor cells and predict a poor prognosis in glioma [5]. However, whether the role of purine metabolism in breast cancer is in line with the findings above has not been well defined.

Long non‐coding RNAs (LncRNAs) are a category of poorly conserved endogenous RNAs longer than 200 nucleotides that do not encode functional proteins but regulate gene expression [6]. Our previous studies, along with other accumulating evidence, indicated that LncRNAs were able to modulate critical cellular functions and had the potential to be promising diagnostic, predictive biomarkers, and therapeutic targets in breast cancer [7, 8, 9]. Mechanistically, the involvement of LncRNAs in tumorigenesis through multiple pathways has been revealed, such as miRNA sponging, epigenetic modification, and transcription regulation [10, 11, 12]. We have demonstrated that lncRNA TPT1‐AS1 was one of the pivotal regulators of LncRNAs in colorectal cancer [13]. However, the precise role and regulatory mechanisms of TPT1‐AS1 in breast cancer, especially metabolism‐related, remain to be further explored.

Purine nucleotides are essential for the biosynthesis of DNA and RNA. Cells are able to harness two principal pathways to produce purine nucleotides, namely the de novo and salvage pathways. In malignant diseases, increased proliferation often leads to altered purine metabolism [14]. Many enzymes involved in purine biosynthesis are overexpressed in breast cancer, including glycyl‐tRNA synthetase (GARS) [15], Glycinamide ribonucleotide transformylase (GART) and 5‐aminoimidazole‐4‐carboxamide ribonucleotide formyltransferase (ATIC) etc. [16]. The dysfunction of purine metabolism‐related enzymes and metabolites can further affect other biological processes within the malignancies, such as immune evasion and treatment resistance [17, 18]. The interaction of purine‐related pathways and cancer development was reviewed in great detail by Singh et al. [19]. ATIC, which catalyzes the final two steps of de novo purine biosynthesis, is a pivotal regulator in cancer progression [20]. Although less studied in breast cancer, clinical data have demonstrated that ATIC was overexpressed in breast cancer tissues and associated with a poor prognosis [21]. Mechanically, ATIC was proven to participate in multiple tumor‐promoting biological processes, including proliferation and migration. In lung adenocarcinoma, ATIC enhances proliferation and migration by regulating the expression level of Myc [22]. In hepatocellular cancer, ATIC promotes malignant phenotypes via the AKT/FOXO3 axis and the AMPK‐mTOR‐S6 K1 signaling [23, 24]. ATIC can also affect the treatment responses of malignancies. The polymorphism of the ATIC gene is related to drug resistance in pediatric osteosarcoma and chronic leukemia [25, 26]. Moreover, the upregulation of ATIC is associated with endocrine therapy‐insensitivity in breast cancer [27]. Based on the findings above, ATIC has the potential to be a promising anti‐cancer drug target. Brooks et al. have already found that blocking ATIC function led to tumor‐suppression in preclinical breast cancer models [28].

In this study, we identified 14 aberrantly expressed LncRNAs in breast cancer tissues, including lncRNA TPT1‐AS1, by bioinformatic analyses. Functional studies showed that TPT1‐AS1 inhibited the proliferation, invasion, and stemness properties of breast cancer cells. Mass spectrometric analysis showed that the purine biosynthetic pathway was activated in TPT1‐AS1‐knockdown MCF‐7 cells. In terms of direct mechanisms, we discovered that TPT1‐AS1 could physically interact with CBP (CREB‐binding protein), which consequently led to the loss of H3K27Ac in the promoter area of ATIC, the key enzyme of IMP synthesis. The downregulation of ATIC subsequently impeded purine metabolism and cancer progression in breast cancer. Our findings elucidate the function of LncRNA TPT1‐AS1 in breast cancer purine metabolism reprogramming and suggest TPT1‐AS1 as a potential candidate for breast cancer therapy [22].

2. Materials and Methods

2.1. Cell Lines and Breast Cancer Specimens

All breast cancer cell lines were obtained from The Chinese Type Culture Collection, Chinese Academic of Science. Fresh breast cancer tissues and normal breast tissues were collected from 197 patients who underwent surgery in the Department of General Surgery of the Second Affiliated Hospital of Harbin Medical University between May 2013 and August 2020. The correlated clinical information of the patients was collected and analyzed. The stage of breast cancer was evaluated according to the TNM classification proposed by the AJCC cancer staging manual. The samples used for RNA isolation were immediately frozen at −80°C after surgical resection; the samples used for tissue chips were formalin‐fixed. The study was approved by the ethics and scientific committee of Harbin Medical University.

2.2. Identification of Hub LncRNAs in Breast Cancer

WGCNA was employed to identify co‐expression modules of LncRNAs related to breast cancer (BC) obtained from LncRBase (http://bicresources.jcbose.ac.in/zhumur/lncrbase/index_download.html; Appendix S1). The LncRNAs in highly correlated modules (blue and red modules) were subjected to enrichment analysis using the LncSEA online tool (http://bio.liclab.net/LncSEA/Analysis.php) to identify significantly enriched biological functions and pathways [29]. LncSEA is a database containing large amounts of LncRNA‐based resources, including an online enrichment analysis platform. Briefly, the names of LncRNAs belonging to the red and blue modules were submitted and annotated by the LncSEA enrichment platform to produce a list of significant pathways with respective P values. Using the Cytohubba plugin in Cytoscape software, we identified the hub LncRNAs and constructed the interaction network of the top 20 LncRNAs. We then selected the differentially expressed genes (DEGs) by comparing the expression distribution of hub genes in BC tissues and normal tissues (P‐value < 0.05 was considered statistically significant).

2.3. RNA Isolation and qRT‐PCR

RNA isolation and real‐time quantitative PCR were conducted as described in detail in our previous studies [8, 13]. The primers used for qPCR in this study are listed in Table S1. Relative RNA expression was calculated by the 2^−△△CT formula.

2.4. Pharmaceuticals

C646, the histone acetyltransferase (HAT) inhibitor, was purchased from MCE (Medchem Express, USA). Appropriate numbers of the indicated cells were seeded into the 6‐well plates with the media containing 20 μM of C646 or not. The cells were cultured for 12 h before being harvested, and the total mRNA was isolated.

2.5. siRNAs Synthesis, Vector Construction, and Transfection

The synthetic oligonucleotides used for silencing TPT1‐AS1 (sh‐TPT1‐AS1#1, sh‐TPT1‐AS1#2) and oligonucleotides for the overexpression of TPT1‐AS1 (Lv‐TPT1‐AS1) were designed and synthesized by GenePharma Company (Shanghai, China). Negative control sh‐NC or Lv‐NC was also obtained from GenePharma Company. ATIC shRNAs were designed and synthesized by Ribobio Company (Guangzhou, China); CBP siRNAs were purchased from Oligobio Company (Beijing, China). Transient transfection was performed using Lipofectamine 2000 reagent (Invitrogen, USA) according to the manufacturer's instructions. The detailed shRNA sequences used in this study are listed in Table S1, and the detailed methods were depicted in Appendix S1.

2.6. Proliferation, Migration and Invasion Assays

To assess the effects of LncRNA knockdown on cell proliferation, migration, and invasion, soft agar assay, 3D morphogenesis matrigel culture, and Transwell assays were performed. The experiment details were described in Appendix S1, and all the experiments were performed in triplicate.

2.7. In Vivo Assay

To evaluate the tumor formation ability of the engineered cells, xenograft mouse models were established. 5‐week‐old female SCID mice were purchased from Vital River (Beijing, China) and housed under specific pathogen‐free conditions at 24°C, 50% humidity, with a 12 h light/dark cycle. Briefly, orthotopic xenografts were established using breast cancer cell lines after transfection and controls (details in Appendix S1). All mice were imaged by the Xenogen IVIS Spectrum Imaging System (Caliper Life Sciences, USA).

2.8. Immunofluorescence (IF) and RNA Fluorescent In Situ Hybridization (FISH)

For IF, the 3D‐cultured spheroids in Matrigel were fixed, followed by pre‐blocking and incubating with primary antibodies. Subsequently, they were incubated with corresponding fluorescent‐labeled secondary antibodies and observed using confocal microscopy. To detect TPT1‐AS1 expression in breast cancer cells, RNA FISH was performed with lncRNA FISH Probe Mix and Fluorescence in situ Hybridization Kit (RIBO Bio, China) according to the manufacturer's instruction (details were described in Appendix S1).

2.9. Dual‐Luciferase Reporter Assay

Briefly, sequences containing the ATIC promoter region were cloned into restriction sites of the pGL3‐basic vector to generate the pGL3‐ATIC reporter construct. The luciferase activity was measured using the Dual‐Luciferase Reporter Assay System (E1910, Promega) (details in Appendix S1).

2.10. In Situ Hybridization (ISH)

In situ hybridization was performed on the tissue microarrays (details in Appendix S1).

2.11. Metabolites Analysis

Cell extracts were analyzed by LC/MS/MS to measure the relative levels of intracellular metabolites, as described previously [9] (see Appendix S1 for details).

2.12. RNA Pulldown Assay and Western Blot

For the RNA pulldown assay, full‐length TPT1‐AS1 and antisense TPT1‐AS1 RNA were biotin‐labeled, transfected, and isolated. RNA oligomers were incubated with protein lysates from MCF‐7 cells afterwards. Then streptavidin agarose beads were added to the reaction mix, followed by standard western blot analysis (details see Appendix S1).

2.13. RNA‐Binding Protein Immunoprecipitation Assay

RNA‐binding protein immunoprecipitation (RIP) assay was performed using the EZ‐Magna RIP RNA‐Binding Protein Immunoprecipitation Kit (Millipore, USA) according to the manufacturer's instructions (details in Appendix S1). The primers specific to TPT1‐AS1 immunoprecipitate are listed below: 5′‐GCCACCACTCCCAGATCTTC‐3′ (forward) and 5′‐TGCTTGGGATTTACTGGAGGAC‐3′ (reverse).

2.14. Chromatin Immunoprecipitation (ChIP) Assay

ChIP assay was performed using the EZ ChIP Kit (Millipore) following the manufacturer's protocol (see Appendix S1). The primers used for detecting the ATIC promoter were listed below: 5′‐ACTGTCATCATTCCTCTCAT CCAG‐3′(forward) and 5′‐TTTAAGTGAGGGACTCCAGAGG‐3′ (reverse).

2.15. Statistical Analysis

Independent experiments were performed in triplicate, and the data were processed and plotted with GraphPad Prism 7.0 software. Detailed methods of statistical analysis are documented in Appendix S1.

3. Results

3.1. Hub LncRNAs Related to BC Were Identified by WGCNA

We collected the lncRNA gene set (3264 LncRNAs) and gene expression profiles (1068 BC samples) from the LncRBase database and TCGA, respectively. Next, we performed WGCNA analysis, which identified clusters of LncRNAs that were highly correlated with each other among patients. The LncRNAs were allocated to modules of different colors, with 16 modules identified in total (Gray 1303 genes, yellow 187, tan 54, cyan 34, green‐yellow 55, turquoise 441, blue 230, red 148, pink 99, purple 59, black 146, magenta 80, brown 199, salmon 44, cyan 34, midnight‐blue 32) at a soft threshold of 2. The association between each module was evaluated according to Spearman's correlation coefficient, and the result was visualized with a heatmap (Figure 1A and Figure S1a–c). We observed that LncRNAs in the red and blue modules had better clustering. Thus, enrichment analysis was performed separately for these two modules by LncSEA [29]. The results indicated that the LncRNAs involved in the blue module were significantly relevant to proliferation, metastasis, and invasion (Figure 1B). In contrast, the red module showed a limited correlation with cancer. Therefore, a total of 230 LncRNAs in the blue module were selected for further analysis. To further narrow the range of the 230 genes, we loaded the node and edge data representing the interactions between LncRNAs in the blue module calculated by WGCNA into Cytoscape software for further analysis. Using the Cytohubba plug‐in, 20 LncRNAs with the most connections in the network were selected as hub LncRNAs (Figure 1C). Next, we compared the expression distribution of the 20 hub LncRNAs with paracancerous tissues in TCGA‐BRCA and found that 14 were aberrantly expressed (Figure 1D). Interestingly, TPT1‐AS1, which was previously studied in our team, was markedly downregulated in breast cancer tissues.

FIGURE 1.

Identifying co‐expressed LncRNAs related to BC by WGCNA. (A) Heat map of the Topological Overlap Matrix in WGCNA analysis. (B) Functional enrichment of genes in the blue module. (C) The hub genes in the blue module by Cytohubba. (D) The expression distribution of hub genes in tumor tissues and normal tissues. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3.2. TPT1‐AS1 Is Downregulated in Breast Cancer Tissues, Which Predicts Poor Prognosis

We identified TPT1‐AS1 as a significantly downregulated hub lncRNA in breast cancer tissues using bioinformatic methods. Moreover, data from TCGA and GTEx databases (normal breast tissues) revealed similar results (Figure S2a). To further investigate the association between TPT1‐AS1 expression and breast cancer, we measured TPT1‐AS1 expression using a cohort of 102 patients. As shown in Figure 2A, TPT1‐AS1 was downregulated in breast cancer tissues compared with the adjacent normal tissues. In parallel with the aforementioned, the level of TPT1‐AS1 was decreased in 83 of the 102 (81.4%) breast cancer tissues (Figure 2B). Moreover, lower expression levels of TPT1‐AS1 showed unfavorable relapse‐free survival in a cohort including 1016 breast cancer patients (Figure S2b). To clarify the relevance between TPT1‐AS1 expression and clinicopathologic features in breast cancer patients, we divided the samples into two groups according to the median TPT1‐AS1 expression. We found that TPT1‐AS1 expression was negatively correlated with tumor differentiation grade, lymph node metastasis and TNM stage (Figure 2C,D, Table 1 and Figure S2c). We also performed in situ hybridization (ISH) staining on a breast cancer tissue array comprised of 197 clinical samples. Our data uncovered a lower expression of TPT1‐AS1 in cancer samples compared with normal tissues (Figure 2F); moreover, the percentage of cells expressing TPT1‐AS1 was negatively correlated with tumor differentiation grade, TNM stage, and lymph node metastasis (Figure S2d–f). We then investigated TPT1‐AS1 expression in human breast cancer cell lines (MDA‐MB‐231, MDA‐MB‐468, T47D, MDA‐MB‐453, MCF‐7) and the normal breast epithelial cell line (MCF‐10A). As expected, TPT1‐AS1 expression was higher in low‐metastatic cell lines (Figure 2E). Taken together, these results showed that TPT1‐AS1 was downregulated in breast cancer tissues and associated with poor prognosis.

FIGURE 2.

The downregulated expression of TPT1‐AS1 in breast cancer tissues correlates with poor prognosis. (A, B) TPT1‐AS1 expression in 102 pairs of cancerous and adjacent normal breast tissues detected by real‐time PCR. (C) The correlation between TPT1‐AS1 expression and tumor differentiation grade. (D) The correlation between the relative expression of TPT1‐AS1 and lymph node metastasis. (E) qRT‐PCR was used to analyze the level of TPT1‐AS1 in the normal breast epithelial cell line MCF10A and breast cancer cell lines (MDA‐MB‐231, MDA‐MB‐468, T47D, MDA‐MB‐453 and MCF‐7) with different metastatic potentials. (F) Analysis of the ISH staining of TPT1‐AS1 in breast cancer tissues. *p < 0.05, **p < 0.01, ***p < 0.001.

TABLE 1.

Relationship between TPT1‐AS1 expression and clinicopathologic features of BC patients (n = 102).

Variable	Relative TPT1‐AS1 expression		p
	Low (n = 51)	High (n = 51)
Age
< 50	27	21	NS
> 50	24	30
Histological differentiation
Well	10	17	< 0.05
Moderate	13	22
Poor	28	12
Tumor size
< 2 cm	17	25	NS
2–5 cm	26	21
> 5 cm	8	5
Lymph node metastasis
Yes	32	15	< 0.05
No	19	36
Tumor stage
I	10	23	< 0.05
II	19	20
III	22	8
Molecular subtype
Luminal like	26	30	NS
Her‐2	12	14
Triple negative	13	7

Note: BC patients were divided into TPT1‐AS1 high group and low group according to the analysis of qRT‐PCR detection. Differences among variables were evaluated by χ ² or Fisher's exact χ ² test.

Abbreviation: NS, not significant between different groups.

3.3. TPT1‐AS1 Inhibits Breast Cancer Cell Proliferation, Invasion, and Stemness Features In Vitro

Having noted the connection between TPT1‐AS1 higher expression and a better survival rate in breast cancer patients, we set out to functionally confirm the effects of TPT1‐AS1 on breast cancer cells. To this end, we knocked down TPT1‐AS1 in MCF‐7 and T47D cells with two shot hairpin RNAs (sh‐TPT1‐AS1#1, sh‐TPT1‐AS1#2) targeting different sites within TPT1‐AS1. And sh‐TPT1‐AS1#2 had better knockdown efficiency, so it was chosen for all subsequent experiments (Figure S2g). In addition to TPT1‐AS1 knockdown, MDA‐MB‐468 cells with lower intrinsic TPT1‐AS1 were stably transfected with the lentiviral vector LV‐TPT1‐AS1 to achieve overexpression (Figure S2h). Soft agar assays implied that silencing TPT1‐AS1 in MCF‐7 cells significantly inhibited the proliferation ability of breast cancer cells, while the overexpression of TPT1‐AS1 in MDA‐MB‐468 cells showed an opposite result (Figure 3A). We then investigated the function of TPT1‐AS1 in the migratory and invasive abilities of breast cancer cells. The results from transwell assays demonstrated that TPT1‐AS1 overexpression significantly inhibited cell migration and invasion in MDA‐MB‐468 cells. Conversely, TPT1‐AS1 depletion markedly enhanced the migrative and invasive ability of MCF‐7 cells (Figure 3B). In addition, we analyzed the sphere‐forming ability of the indicated cells. As expected, MDA‐MB‐468 cells with TPT1‐AS1 transfection generated more spheres that were larger in size than those derived from control cells. Moreover, MCF‐7 cells with TPT1‐AS1 depletion generated fewer spheres, which were also smaller compared with control cells (Figure 3C). We also performed the functional assays in T47D and MDA‐MB‐231 cells and found similar results (Figure S3a–c). As further verified by immunofluorescence assays, the levels of Oct4 and SOX2, two typical stemness markers, were markedly decreased in TPT1‐AS1‐overexpressing spheres (Figure 3D). The western blot assay also confirmed that OCT4 and SOX2 were downregulated in TPT1‐AS1‐overexpressing‐cells (Figures S3d and S5b). Together, our findings indicated that TPT1‐AS1 inhibited the proliferation, invasion, and stemness features of breast cancer cells in vitro.

FIGURE 3.

TPT1‐AS1 inhibits proliferation, invasion, and stemness features of breast cancer cells. (A) Soft agar assays indicated the proliferation ability of breast cancer cells with TPT1‐AS1 silencing and increasing (Scale bar: 100 μm). (B) Transwell assays were used to detect the migratory and invasive abilities of indicated cells(Scale bar: 10 μm). (C) The sphere‐forming ability of the indicated cells with TPT1‐AS1 silencing and increasing was examined by a 3D‐cultured mammosphere model (Scale bar: 100 μm). (D) IF staining of Sox2 and Oct4 in the indicated cell spheres was shown (Scale bar: 50 μm). *p < 0.05, **p < 0.01, ***p < 0.001.

3.4. TPT1‐AS1 Overexpression Significantly Reduces the Growth and Metastasis of Breast Cancer In Vivo

To further assess the biological function of TPT1‐AS1 in vivo, MDA‐MB‐468 and MCF‐7 cells were injected into the mammary fat pads of SCID mice, respectively, to build orthotopic mice models. In contrast with the control group, TPT1‐AS1‐transfected MDA‐MB‐468 cells developed smaller subcutaneous tumors (Figure 4Ai). Meanwhile, TPT1‐AS1 knockdown in MCF‐7 cells reversed such effects (Figure 4Aii), which further validated the crucial role of TPT1‐AS1 in facilitating the proliferation ability of breast cancer cells in vivo. To further confirm the influence of TPT1‐AS1 in tumor metastasis, MDA‐MB‐468 cells stably transfected with LV‐NC and LV‐TPT1‐AS1 were injected into the tail vein of SCID mice to establish lung metastasis models. The results showed that metastatic nodules at the lung surface in mice bearing TPT1‐AS1‐overexpressed cells were markedly decreased compared with the LV‐NC group (Figure 4Bi). The results were further verified by H&E staining (Figure 4Bii), indicating that increased TPT1‐AS1 expression impaired the lung metastasis of breast cancer cells. To conclude, the observations consistently confirmed the pivotal role of TPT1‐AS1 in inhibiting the growth and metastasis of breast cancer both in vitro and in vivo.

FIGURE 4.

TPT1‐AS1 inhibits breast cancer cell proliferation and metastasis in vivo. (Ai) and (Aii) Representative images of the SCID mice injected with indicated engineered MDA‐MB‐468, MCF‐7, and control cells (five mice in each group) and the tumor volumes were measured. (Bi) Lung metastatic mouse models established by tail vein injection with indicated cells were used to explore the metastatic ability, and representative images of H&E staining of metastatic foci in the lungs (Scale bar upper: 5 mm; Scale bar lower: 250 μm). (Bii) Metastasis nodules were counted and analyzed. *p < 0.05, **p < 0.01, ***p < 0.001.

3.5. TPT1‐AS1 Loss Activates the Purine Synthesis Pathway by Upregulating the Expression of ATIC

Altered metabolic activities drive tumorigenesis and metastasis in various malignancies. Therefore, we utilized mass spectrometry (MS)‐based metabolic profiling of MCF‐7 cells transfected with sh‐control or sh‐TPT1‐AS1 to delineate whether TPT1‐AS1 contributes to reprogramming cancer‐related metabolic pathways. Intriguingly, in contrast to the sh‐NC group, the intracellular pools of purine intermediates such as adenosine monophosphate (AMP), guanosine monophosphate (GMP) and especially inosine monophosphate (IMP), were dramatically increased in TPT1‐AS1 silencing MCF‐7 cells (Figure 5A), indicating that silencing TPT1‐AS1 might facilitate the purine synthesis pathway. It was well established that ATIC, the final enzyme in the purine de novo synthesis pathway, is able to transfer a formyl group to AICAR to produce the end product IMP. Data from TCGA and GTEx databases indicated that ATIC is aberrantly expressed in breast cancer tissues versus normal tissues (Figure 5B), and the negative correlation between ATIC and TPT1‐AS1 was also confirmed (Figure 5C). Next, we set out to detect whether the critical rate‐limiting enzymes in the purine synthesis pathway were increased in TPT1‐AS1‐knockdown MCF‐7 cells. As shown in Figure 5D, the mRNA level of ATIC was markedly decreased among the enzymes in TPT1‐AS1‐knockdown MCF‐7 cells. Western blot assays also confirmed that the protein level of ATIC was reduced in MCF‐7 and T47D cells transfected with TPT1‐AS1 (Figure 5E, Figure S5a). Furthermore, we performed restoration assays to analyze the effect of sh‐ATIC in TPT1‐AS1‐knockdown MCF‐7 and T47D cells. As shown in Figure 5F, the enhanced invasion and migration induced by TPT1‐AS1 silencing were reversed by the knockdown of ATIC in MCF‐7 or T47D cells. To further confirm the role of TPT1‐AS1 in purine synthesis, we used liquid chromatography (LC)–MS/MS to gauge the relative flux of 13C‐glycine. Not surprisingly, silencing TPT1‐AS1 in MCF‐7 cells increased the amounts of 13C‐glycine intermediates including IMP, AMP, and GMP. At the same time, the introduction of sh‐ATIC reversed the enhanced amounts of the 13C‐glycine intermediates (Figure 5G). In contrast, overexpression of ATIC restored the decreased 13C‐glycine intermediates in TPT1‐AS1‐transfected MDA‐MB‐468 cells (Figure 5H). The results indicated that TPT1‐AS1 loss might promote breast cancer cell purine synthesis and malignant phenotypes by regulating the expression of ATIC.

FIGURE 5.

TPT1‐AS1 inhibits purine synthesis of breast cancer cells through ATIC. (A) Heat map showed the expression of multiple metabolites measured by mass spectrometry (MS) in MCF‐7 cells with or without sh‐TPT1‐AS1 transfection. (B) TCGA and GTEx data was analyzed to compare the expression of ATIC in breast cancer or corresponding normal tissues. (C) TCGA and GTEx data was obtained to detect the relationship between ATIC and TPT1‐AS1 expression in breast cancer. (D) qRT–PCR assays were used to measure the mRNA level of purine synthesis‐related enzymes in TPT1‐AS1‐knockdown MCF‐7 cells. (E) Western blot assays were used to detect the expression of ATIC in MCF‐7 and T47D cells transfected with sh‐TPT1‐AS1 or sh‐NC. (F) Invasion and migration of MCF‐7 and T47D cells transfected with or without sh‐TPT1‐AS1 and sh‐ATIC were assessed by transwell assays (Scale bar: 20 mm). (G) ¹³C‐labeled intermediates of purine synthesis were examined in TPT1‐AS1‐knockdown MCF‐7 cells transfected with or without sh‐ATIC. (H) ¹³C‐labeled intermediates of purine synthesis were examined in TPT1‐AS1‐transfected MDA‐MB‐468 cells with or without ATIC overexpression. *p < 0.05, **p < 0.01, ***p < 0.001.

3.6. TPT1‐AS1 Regulates ATIC Expression Through Histone Acetylation at the Transcriptional Level

To determine the mechanism underlying the downregulation of ATIC by TPT1‐AS1, we performed an RNA FISH experiment to determine the intracellular localization of TPT1‐AS1. The result showed that TPT1‐AS1 was mainly located in the nucleus (Figure 6A), indicating that TPT1‐AS1 might function in the nucleus and regulate gene expression at the transcriptional level. To further confirm the hypothesis, we performed bioinformatics analysis and found abundant H3K27ac signals in the promoter region of ATIC (Figure 6B), suggesting that ATIC might be transcriptionally regulated by chromatin histone acetylation. We, therefore, probed whether knocking down TPT1‐AS1 expression would promote ATIC mRNA synthesis transcriptionally. Luciferase reporter assays showed that depletion of TPT1‐AS1 enhanced the activity of the ATIC reporter signal, indicating that TPT1‐AS1 could regulate ATIC expression at the transcriptional level (Figure 6C,D). We next treated TPT1‐AS1‐knockdown MCF‐7 and T47D cells with C646, a histone acetyltransferase inhibitor, and found that the ATIC mRNA level markedly decreased in response to the addition of C646 (Figure 6E,F). Western blot assays also proved that TPT1‐AS1 depleting could upregulate the expression of Ac‐H3 (Figures S4a and S5d). It was established that CREB‐binding protein (CBP) was the most critical enzyme that could modify histone acetylation at the transcriptional level, so we doubted if CBP participated in regulating ATIC expression. To this end, we used qRT‐PCR and western blot assays to detect the effect of CBP knockdown in multiple breast cancer cell lines. Consistent with our assumption, decreasing the expression of CBP using siRNAs significantly reduced the mRNA and protein levels of ATIC (Figure 6G,H). Meanwhile, the protein level of Ac‐histone 3 also decreased in CBP‐knockdown breast cancer cells (Figure 6H, Figure S5c). ChIP assays were also performed to confirm the association between CBP and the ATIC promoter region (Figure 7C, Figures S3e and S4b). To conclude, TPT1‐AS1 knockdown upregulated ATIC mRNA expression transcriptionally through histone acetylation.

FIGURE 6.

TPT1‐AS1 regulates ATIC expression through histone acetylation at transcriptional level. (A) TPT1‐AS1 intracellular localization was detected in MCF‐7 cells by RNA FISH assay. The DAPI‐stained nuclei are shown in blue. (Scale bar: 10 μm). (B) The H3K27ac signal of ATIC promoter. (C, D) Luciferase reporter assay was performed using vector‐containing sequences of the ATIC promoter region in indicated cells. (E, F) qRT‐PCR assays were used to detect the ATIC expression after C646 treatment in MCF‐7 and T47D cells. (G, H) qRT‐PCR and western blot analysis of ATIC expression after CBP knockdown in breast cancer cells. *p < 0.05, **p < 0.01, ***p < 0.001, # no significance.

FIGURE 7.

TPT1‐AS1 inhibits ATIC expression through binding with CBP and further inhibits breast cancer cell proliferation. (A) qRT‐PCR was used to detect the expression of CBP in MCF‐7 and T47D cells transfected with sh‐TPT1‐AS1 or sh‐NC. (B) ChIP‐qPCR assays were performed to show the enrichment of the ATIC promoter region that was binding with CBP or H3K27Ac in MCF‐7 cells transfected with sh‐TPT1‐AS1 or sh‐NC. (C) RIP experiments showed that the CBP antibody could interact with TPT1‐AS1 (top and middle rows) and ChIP‐PCR assay indicated that TPT1‐AS1 bonds with the ATIC promoter in MCF‐7 cells (bottom row). (D) RNA pull‐down assays demonstrated that TPT1‐AS1 could retrieve CBP in MCF‐7 cells as detected by western blot. FT represents the flow‐through protein after the RNA‐protein binding reaction. Input indicates the total cell protein used. (E, F) Luciferase reporter assays were performed to show the influence of TPT1‐AS1 on ATIC transcription after CBP knockdown. (G) Soft agar assays were used to explore the proliferative ability of the indicated cells (Scale bar: 100 μm). *p < 0.05, **p < 0.01, ***p < 0.001, # no significance.

3.7. TPT1‐AS1 Interacts With CBP and Reduces the Association Between CBP and AITC Promoter

To test if TPT1‐AS1 regulates ATIC mRNA expression through CBP‐mediated histone acetylation, we first performed western blot assays to detect CBP expression in TPT1‐AS1‐knockdown MCF‐7 cells. The results showed that TPT1‐AS1 failed to regulate CBP expression (Figure 7A). Next, we doubted whether TPT1‐AS1 influenced the association between CBP and the ATIC promoter. As mentioned, TPT1‐AS1 is mainly located in the nucleus and regulates ATIC expression at the transcriptional level. Thus, ChIP assays were performed using CBP and H3K27Ac antibodies. We found that the depletion of TPT1‐AS1 markedly increased the enrichment of CBP and H3K27ac acetylation at the promoter region of ATIC (Figure 7B), which made us speculate that TPT1‐AS1 might bind to CBP and disrupt the interaction between CBP and the ATIC promoter. To this end, RIP assays were carried out to detect the interaction between TPT1‐AS1 and CBP. We observed a marked enrichment of TPT1‐AS1 pulled down by the CBP antibody compared with the IgG control antibody (Figure 7C, Figure S3e). RNA‐protein pull‐down assays further confirmed that TPT1‐AS1 distinctively retrieved CBP from MCF‐7 total protein extracts but not antisense TPT1‐AS1 (Figure 7D). Next, we used functional assays to explore if the effect of TPT1‐AS1 on breast cancer cells was mediated by CBP. A luciferase reporter assay showed that the depletion of CBP attenuated the ATIC expression at the transcriptional level, regardless of the TPT1‐AS1 knockdown status (Figure 7E,F). Soft agar assays implied that after silencing endogenous CBP, TPT1‐AS1 failed to alter the proliferative and invasive ability of breast cancer cells (Figure 7G). Therefore, we concluded that TPT1‐AS1 interacted with CBP and subsequently decreased the enrichment of CBP and histone acetylation at the promoter region of ATIC, finally inducing the transcriptional inhibition of ATIC (Figure 8).

FIGURE 8.

Schematic diagram of the mechanism that TPT1‐AS1 interacts with CBP and decreases the enrichment of CBP and histone acetylation at the promoter of ATIC.

4. Discussion

LncRNAs are pivotal parts that regulate the process of tumorigenesis and thus are associated with the prognosis of breast cancer. In this study, we screened BC‐related LncRNAs and identified 20 key LncRNAs. We chose TPT1‐AS1 for further exploration and discovered that TPT1‐AS1 was downregulated in breast cancer tissues and highly metastatic cell lines. Importantly, we confirmed that TPT1‐AS1 influenced IMP levels by regulating H3K27Ac in the promoter region of ATIC. Moreover, we identified that TPT1‐AS1 could physically interact with CBP and consequently led to the loss of H3K27Ac.

Consistent with our findings, two recent studies reported that TPT1‐AS1 was downregulated in breast cancer and acted as a tumor suppressor [30, 31]. Hu et al. analyzed the expression of TPT1‐AS1 in TCGA and GSE databases and performed in vitro experiments to prove that low expression of TPT1‐AS1 could predict a poor prognosis and promote the proliferation and invasion of breast cancer cells [30]. Another team found a similar expression pattern of TPT1‐AS1 in breast cancer samples and claimed that TPT1‐AS1 enhanced the sensitivity of breast cancer cell to paclitaxel [31]. However, the aforementioned studies mainly focused on the competing endogenous RNA (ceRNA) mechanism and the upstream regulation of TPT1‐AS1, while our study elucidated a more detailed and precise relationship between TPT1‐AS1 and clinical pathological features of breast cancer. Moreover, our work demonstrated a previously unknown function of TPT1‐AS1, namely the interaction of histone modification enzymes and the effect on metabolism reprogramming in cancer. Since drugs that target purine metabolism have been standard regimens in cancer [17], our study aims to provide novel knowledge on the alteration of purine biosynthesis in breast cancer, hoping to fuel the development of anti‐purine metabolism‐related drugs in breast cancer. We performed qRT‐PCR and ISH staining to detect the expression of TPT1‐AS1 and drew the conclusion that TPT1‐AS1 expression was negatively correlated with tumor differentiation grade and lymph node metastasis. Moreover, both in vitro and in vivo assays were conducted to prove the suppressive role of TPT1‐AS1 in breast cancer proliferation, invasion, metastasis, and stemness property. Furthermore, we conducted mass spectrometry (MS) analysis to detect the altered metabolic activities of TPT1‐AS1‐depleting MCF‐7 cells and found that TPT1‐AS1 could inhibit the activity of the purine synthesis pathway. This finding delineated the role of TPT1‐AS1 in reprogramming cancer‐related metabolic pathways for the first time and elaborated a more detailed understanding of the function of TPT1‐AS1.

Aminoimidazole carboxamide ribonucleotide transformylase/inosine monophosphate cyclohydrolase (ATIC) is a 64‐kDa bifunctional enzyme that catalyzes the last two steps of the de novo purine biosynthesis pathway [20]. As a master rate‐limiting enzyme in purine metabolism, the crucial role of ATIC in cancer development has been extensively described in liver cancer [24]. A recent study reported that ATIC was closely related to tamoxifen resistance in breast cancer and the elevated expression of ATIC was associated with a poor prognosis [27]. Additionally, we also found that ATIC was transcriptionally activated in triple‐negative breast cancer under high glucose induction (data unpublished). Thus, strategies targeting ATIC to regulate purine synthesis might be an ideal way for breast cancer therapeutics. Although the crosstalk of post‐translational modifications of ATIC and how ATIC acted in its function through the active motif were well described [32, 33], the transcriptional regulation of ATIC remained unclear. In this study, we revealed that TPT1‐AS1 transcriptionally inhibited ATIC by interacting with CBP and subsequently attenuating the H3K27Ac signal in the promoter region of ATIC. This finding elucidated a specific regulatory network of TPT1‐AS1, CBP, and ATIC. Intriguingly, a recent study suggested that p300/CBP epigenetically regulated the expression of several metabolic enzymes through modulation of histone acetylation in HCC, including ATIC [34]. This finding strengthened our hypothesis to a large extent. Nevertheless, histone acetylation was one of the histone modifications in the promoter region of genes; H3K4me and H4R3me were also detected in the database. Thus, we could not exclude the possibility that factors other than the TPT1‐AS1/CBP axis might regulate ATIC transcriptionally. Further study will undoubtedly enrich the recognition of the regulatory mechanisms of ATIC in breast cancer. Of note, the two published articles focusing on the regulatory network of TPT1‐AS1 both regarded microRNAs as downstream targets. Unlike the complementary base pairing model, we suggest that TPT1‐AS1 could competitively bind with RNA‐binding proteins and decrease the H3K27Ac levels in the ATIC promoter region. A number of eRNAs were reported to interact with CBP and regulate the expression of target genes through histone acetylation [35], which further strengthens our findings. To conclude, the TPT1‐AS1/CBP/ATIC axis provides new insights into the relationship between long non‐coding RNAs and purine metabolism of breast cancer for the first time.

Author Contributions

Yiyun Zhang: investigation. Hanyu Zhang: investigation. Mingcui Li: writing – original draft, writing – review and editing. Yanling Li: investigation. Zhuo‐Ran Wang: writing – review and editing. Weilun Cheng: formal analysis. Yansong Liu: formal analysis. Zhengbo Fang: formal analysis. Ang Zheng: project administration. Jingxuan Wang: conceptualization, writing – original draft. Fei Ma: conceptualization, funding acquisition.

Disclosure

Our research was performed in strict accordance with the Guide for the Administration of Affairs Concerning Experimental Animals. All the experimental procedures were conducted under the approval of the Animal Experimental Ethics Committee of Harbin Medical University.

Ethics Statement

This study was reviewed and approved by the Ethical Committees of Harbin Medical University and performed in accordance with the Declaration of Helsinki.

Consent

Written informed consent was obtained from each patient.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1. Doc. S1 Supplementary methods and materials.

Figure S1. Identify co‐expressed lncRNAs related to BC by WGCNA.

Figure S2. Expression levels of TPT1‐AS1 associated with patients’ characteristics.

Figure S3. TPT1‐AS1 influences proliferation, migration, invasion, stemness and interacts with CBP.

Figure S4. WB and ChIP results related to figure 6.

Figure S5. Quantification of WB assays.

Table S1. Primers and siRNA sequences.

Data Availability Statement

All data and materials in this study are available upon request.