ELAPOR1 induces the classical/progenitor subtype and contributes to reduced disease aggressiveness through metabolic reprogramming in pancreatic cancer

Pancreatic ductal adenocarcinoma (PDAC) is a heterogeneous disease with distinct molecular subtypes described as classical/progenitor and basal‐like/squamous PDAC. We hypothesized that integrative transcriptome and metabolome approaches can identify candidate genes whose inactivation contributes to the development of the aggressive basal‐like/squamous subtype. Using our integrated approach, we identified endosome–lysosome associated apoptosis and autophagy regulator 1 (ELAPOR1/KIAA1324) as a candidate tumor suppressor in both our NCI‐UMD‐German cohort and additional validation cohorts. Diminished ELAPOR1 expression was linked to high histological grade, advanced disease stage, the basal‐like/squamous subtype, and reduced patient survival in PDAC. In vitro experiments demonstrated that ELAPOR1 transgene expression not only inhibited the migration and invasion of PDAC cells but also induced gene expression characteristics associated with the classical/progenitor subtype. Metabolome analysis of patient tumors and PDAC cells revealed a metabolic program associated with both upregulated ELAPOR1 and the classical/progenitor subtype, encompassing upregulated lipogenesis and downregulated amino acid metabolism. 1‐Methylnicotinamide, a known oncometabolite derived from S‐adenosylmethionine, was inversely associated with ELAPOR1 expression and promoted migration and invasion of PDAC cells in vitro. Taken together, our data suggest that enhanced ELAPOR1 expression promotes transcriptome and metabolome characteristics that are indicative of the classical/progenitor subtype, whereas its reduction associates with basal‐like/squamous tumors with increased disease aggressiveness in PDAC patients. These findings position ELAPOR1 as a promising candidate for diagnostic and therapeutic targeting in PDAC.


What's new?
Pancreatic ductal adenocarcinoma (PDAC) has two distinct subtypes: classical/progenitor and basal-like/squamous, with the latter being more aggressive and exhibiting worse outcomes.
Here, the authors probed the transcriptome and metabolome and identified endosomelysosome associated apoptosis and autophagy regulator 1 (ELAPOR1) as a candidate tumor suppressor associated with the latter subtype.Increased expression of ELAPOR1 contributed to metabolic characteristics associated with the classical/progenitor subtype, reduced cell invasion, and better survival, while decreased ELAPOR1 correlated with the development of the basallike/squamous subtype.ELAPOR1 could be a useful target in diagnosing and treating PDAC.

| INTRODUCTION
Pancreatic cancer is a highly lethal cancer with a 5-year survival of only 12%. 1 Pancreatic ductal adenocarcinoma (PDAC) is the most common type of pancreatic cancer, accounting for more than 95% of all malignancies in the pancreas. 24][5] Bailey et al. proposed a classification system for PDAC based on four subtypes: pancreatic progenitor, aberrantly differentiated endocrine exocrine (ADEX), squamous, and immunogenic. 3th pancreatic progenitor and ADEX subtypes upregulate transcriptional networks associated with pancreatic development/differentiation of exocrine and neuroendocrine lineages. 3The squamous subtype is associated with the most aggressive disease and exhibits distinct characteristics, including hypoxia, squamous differentiation, and MYC activation. 3A more comprehensive subtype analysis recently proposed that PDAC can be classified into two major subtypes: "classical/progenitor" and "basal-like/squamous." 6 Beyond this distinction, the classical/ progenitor subtype possesses more of an endodermal-pancreatic identity that is lost in the basal-like/squamous subtype. 7[12] Moreover, the molecular subtypes of PDAC exhibit distinctively different metabolic profiles.[15][16] However, the metabolic adaptations and their potential interaction with tumor gene-expression profiles have not been well studied in PDAC.Many of the previous studies were based on a small number of patient samples and cell lines.Using large PDAC patient cohorts together with human PDAC cell lines, we recently described SERPINB3 as an oncogenic driver of aggressive basal-like/squamous subtype of PDAC. 17In the present study, we aimed to examine the interactive role of potential tumor suppressor genes and metabolic reprogramming in the development of molecular subtypes in PDAC.
Our findings showed that downregulation of endosome-lysosome associated apoptosis and autophagy regulator 1 (ELAPOR1, also known as KIAA1324/EIG121) contributes to the metabolic adaptation in the basal-like/squamous subtype in patients with PDAC.

| PDAC cohorts
We utilized gene expression datasets from the "Bailey" cohort (GSE36924), 3 the "Moffitt" cohort (GSE71729), 4 and our NCI-UMD-German cohort (GSE183795) 18 to identify genes involved in the development of PDAC molecular subtypes.Partek Genomics Suite 7.0 (Partek Inc., Chesterfield, MO) was used to compare the basallike/squamous subtype with the classical/progenitor subtype.For the integrative transcriptome and metabolome analyses, we interrogated RNA sequencing data from patient PDAC tumors in the NCI-UMD-German cohort (GSE224564). 17

| Cell lines and culture condition
Human pancreatic cancer cell lines were purchased from American Type Culture Collection (ATCC), Rockville, Maryland.All cell lines were authenticated using short tandem repeat (STR) profiling within the last 3 years.

| ELAPOR1 overexpression after lentiviral infection
The ELAPOR1 construct (EX-L2758-Lv122) and the corresponding empty vector control (EX-NEG-Lv122) were purchased from Genecopoeia (Rockville, MD).To establish stable cell lines overexpressing ELAPOR1, PDAC cells (Panc 10.05, SU.86.86) were infected with lentiviral particles produced by transfecting 293 T cells with the lentiviral expression vectors and the Lenti-Pac™ HIV Expression Packaging system from Genecopoeia.Stable clones of the PDAC cells were obtained by selection with 4 μg/mL puromycin (Thermo Fisher Scientific).

| RNA sequencing
RNA sequencing data were obtained by performing quadruplicate RNA sequencing using total RNA isolated from human PDAC cell line (Panc 10.05 +/À ELAPOR1).The PDAC cells were cultured in RPMI 1640 medium, GlutaMax™, supplemented with 10% FBS and 1% penicillin-streptomycin for 72 h before RNA extraction.Libraries were prepared by the Sequencing Facility at NCI-Leidos using the TruSeq Stranded mRNA Kit (Illumina, San Diego, CA) and sequenced paired-end on NextSeq (Illumina) with 2 Â 101 bp read lengths, as described earlier. 17Briefly, a total of approximately 32-60 million paired-end reads were generated with a base call quality of ≥Q30.
Sequence reads in fastq format were then aligned to the human reference genome hg38 using STAR and RSEM to obtain gene expression as transcript per million with FPKM mapped reads.Differential expression analysis was performed using DESeq2.Ingenuity pathway analysis (IPA, QIAGEN, Venlo, Netherlands) and Gene Set Enrichment Analysis (GSEA) were used for the enrichment analysis in archived pathways and datasets.The RNA sequencing data for human PDAC cell lines were deposited in the NCBI's GEO database under accession number GSE243879.The sequencing coverage and quality statistics for each sample are summarized in Table S1.

| Quantitative real-time PCR for ELAPOR1
The High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) was used to create first-strand cDNA from total RNA.Quantitative real-time PCR (qPCR) assays were then performed using Taqman probes (Thermo Fisher Scientific): ELAPOR1 (Hs00331399_m1) and GAPDH (Hs99999905_m1).

| Metabolic profiling and data analysis of PDAC
Metabolic profiling of the tumor samples was conducted by Metabolon Inc. (Morrisville, NC), using their standard protocol as described earlier. 19,20Metabolon's untargeted metabolic platform uses two separate ultra-high performance liquid chromatography/tandem mass spectrometry (UHPLC/MS/MS) injections and one gas chromatography/mass spectrometry injection for each sample, to measure all metabolites.The resulting dataset was normalized across samples before data delivery, using a Metabolon protocol.The normalized relative abundance levels for each metabolite were used for further data analysis (Table S2).The same protocol was used to perform metabolic profiling of cultured cells.Here, we performed quadruplicate metabolic profiling of human PDAC cell line (Panc 10.05 +/À ELAPOR1).
After incubation, cells that had migrated or invaded through the membrane were fixed with 100% methanol (Thermo Fisher Scientific) and stained with Crystal violet solution (MilliporeSigma, Burlington, MA).
The cells were then counted.To examine the effect of 1-methylnicotinamide (MilliporeSigma), 10 mM of the compound (final concentration) was added into the lower chamber.
After incubation, cells were washed with PBS, trypsinized, and further washed with PBS.PDAC cells with ELAPOR1 transgene expression and PDAC cells with the empty vector control (vectors containing GFP-tag) were subjected to flow cytometry.Cells underwent viability gating, singularity gating to isolate single cells, and GFP-positivity gating (GFP_A cutoff of >5 Â 10 3 ).Subsequently, the median values of forward scatter (FSC_A) and side scatter (SSC_A) were measured to assess cell size and complexity.Normalization was carried out using PDAC cells with the empty vector control as a reference.Flow cytometry analysis was conducted using the ID7000™ Spectral Cell Analyzer (Sony Group Corporation, Tokyo, Japan).The gating process and data analysis were performed on this analyzer at a core facility, involving trained personnel with expertise.

| Statistical analysis
Statistical analyses were performed using GraphPad Prizm 9 (Graph-Pad Software, La Jolla, CA).To determine the difference in overall survival between patient groups, we used the Kaplan-Meier method and the log-rank test.We assessed group differences using unpaired twotailed Student's t-tests (for two groups) or ANOVA (for three or more groups).The results are presented as mean ± SD, and statistical significance was defined as a P-value <.05.

| ELAPOR1 is downregulated in the basal-like/ squamous PDAC and associates with advanced stage and poor patient survival
We aimed to identify candidate tumor suppressor genes involved in the development of the basal-like/squamous subtype of PDAC.Initially, we analyzed transcriptome data from two PDAC cohorts (Bailey cohort 3 and Moffitt cohort 4 ).This analysis revealed 399 genes displaying differential expression in the basal-like/squamous compared to the classical/progenitor subtype (cutoffs: fold change >1.5 or <À1.5, P < .05),with 268 of these genes being downregulated in the basal-like/squamous subtype (Figure 1A).We further analyzed the genes based on their association with patient survival (hazard ratio <0.666 using COX regression analysis, P < .05) in our NCI-UMD-German cohort, yielding a narrowed list of 60 genes (Table S3).
Among them, IAPP and ELAPOR1 were ranked as the first and the second most downregulated genes in tumor tissues compared to nontumor tissues, and ELAPOR1 was more significantly downregulated in the basal-like/squamous subtype in both the Bailey and Moffitt cohort datasets (Table S3).Previous studies have shown that the basal-like/squamous subtype is enriched for TP53 mutations, 3,6 while the KRAS mutational status can also impact subtype differentiation. 4,23,24PDAC cases with either a TP53 or a KRAS mutation exhibited lower expression of ELAPOR1 (Figure S1A).ELAPOR1 is an estrogen-regulated gene 25 and plays a role in lysosomal and autophagy processes. 26,279][30] Based on our aforementioned findings and the published literature, we selected ELAPOR1 for further investigations and hypothesized that ELAPOR1 inhibits PDAC progression through metabolic reprogramming.Downregulation of ELAPOR1 transcript levels associated with decreased patient survival, and ELAPOR1 transcripts were downregulated in tumor tissues F I G U R E 1 Transcriptome analysis identifies ELAPOR1 as a candidate tumor suppressor in basal-like/squamous PDAC tumors and as a prognostic marker of patient survival.(A) Approach to identify potential suppressor genes of the basal-like/squamous subtype.Candidate genes are identified by analyzing two cohorts (Bailey cohort 3 ; GSE36924, Moffitt cohort 4 ; GSE71729), followed by a survival analysis in the NCI-UMD-German cohort. 18Progenitor; n = 30, Squamous; n = 25, Classical; n = 103, Basal-like; n = 42.(B-E) ELAPOR1 transcript levels in PDAC tumors compared to adjacent noncancerous tissues.ELAPOR1 is found to be downregulated in tumors in both the NCI-UMD-German cohort (mRNA via qPCR) and the validation cohort (Moffitt cohort 4 ; GSE71729; mRNA).The lower panels show Kaplan-Meier plots and log-rank test results, depicting the association between decreased ELAPOR1 and decreased PDAC patient survival in the NCI-UMD-German and the validation cohort (TCGA cohort).For this survival analysis, patients in the upper 75% quartile and lower 25% quartile of ELAPOR1 expression were compared.Nontumor compared to nontumor tissues in both the NCI-UMD-German cohort and the validation cohorts (Moffitt cohort and TCGA cohort) (Figure 1B-E).Moreover, the downregulation of ELAPOR1 was associated with advanced stage (III/IV) and high-grade PDAC (poorly or undifferentiated PDAC histology) (Figure 1F,G), consistent with a previous study indicating an association between the basal-like/ squamous subtype and high-grade PDAC. 5 IHC revealed prevalent ELAPOR1 protein detection in the cytoplasm of acinar cells and endocrine cells, with lower expression in tumor cells (Figure S1B,C).ELA-POR1 protein downregulation was observed in epithelial cells exhibiting acinar-to-ductal metaplasia (ADM), suggesting that ELA-POR1 might be involved in maintaining acinar cell differentiation.
While qPCR results suggested a correlation between ELAPOR1 levels and tumor histology (Figure 1G), IHC did not reveal a clear association (data not shown).ELAPOR1 is also known as a downstream target of MIST1/BHLHA15, 27 which has the function of a transcription factor in the classical/progenitor subtype. 3Consistently, we found a positive correlation between MIST1 and ELAPOR1 transcript levels in the classical/progenitor subtype, but not in the basal-like/squamous subtype (Figure S1D), suggesting that the functional interaction between MIST1 and ELAPOR1 may be specific to the classical/progenitor subtype.Together, these findings support the hypothesis that ELAPOR1 has a tumor inhibitory function in PDAC, where its loss is associated with poor patient survival and a basal-like/squamous differentiation.

| Upregulation of ELAPOR1 downregulates cellular movement in PDAC
We next investigated whether ELAPOR1 expression may decrease basal-like/squamous subtype characteristics and reduce disease aggressiveness.We previously defined the molecular subtypes among 175 patients in our NCI-UMD-German cohort 17 and showed that pathways related to cellular movement are more activated in the basal-like/squamous than the classical/progenitor subtype.The basallike/squamous subtype exhibited a downregulation of ELAPOR1 expression in comparison to the other subtypes (Figure S2A).Pathway enrichment analysis using the Ingenuity Pathways Analysis (IPA) revealed that ELAPOR1-high PDAC tumors enriched the pathways, including cellular movement and cellular development (Figure 2A).
To confirm these observations from patient tumors, we evaluated ELAPOR1 expression in human PDAC cell lines.ELAPOR1 expression was low to undetectable in all the PDAC cell lines (Figure S2B).Panc 10.05 and SU.86.86 cells were then selected to establish cell lines with ELAPOR1 transgene expression.ELAPOR1 expression was confirmed at the mRNA and the protein levels in these cell lines (Figure S2C,D).Pathway enrichment analysis using IPA showed that the human PDAC cells with ELAPOR1 transgene expression had an enrichment pattern in pathways coherent with ELAPOR1-high PDAC tumors of the NCI-UMD-German cohort, showing again an enrichment of genes in the cellular movement pathway (Figure 2B).To further investigate ELAPOR1 functions, we performed additional in vitro experiments using these ELAPOR1-overexpressing PDAC cells.CCK-8/WST-8 assay indicated that ELAPOR1 did not affect proliferation (data not shown).However, ELAPOR1 expression significantly inhibited the migration and invasion of PDAC cells (Figure 2C,D).Furthermore, IPA and gene set enrichment analysis (GSEA) in PDAC tumors revealed that ELAPOR1 may upregulate neuroendocrine differentiation and downregulate HIF1A/hypoxia signaling (Figures 3A,B and   S3A), which are known characteristics of the classical/progenitor subtype. 3We conducted cell morphology examinations to study the in vitro effects of ELAPOR1 expression.Flow cytometry analysis revealed that ELAPOR1 expression led to an increase in both cell size and complexity (Figure S3B).These findings suggest a potential enhancement of differentiation in vitro, and the observed increase in cell size may be attributed to the retention of secretory properties.
Collectively, our findings suggest that ELAPOR1 plays a role in suppressing the characteristics of the basal-like/squamous subtype of PDAC.

| Upregulation of lipid metabolism and downregulation of amino acid metabolism in ELAPOR1-high PDAC
Our aim was to investigate the impact of ELAPOR1 on metabolism in PDAC.Applying GSEA to the transcriptome data revealed that ELA-POR1 expression associated with increased lipogenesis but downregulated glycolysis and polyamine metabolism, a downstream pathway of amino acid metabolism (Figures 3C and S3C).
Subsequently, we delved into the effect of ELAPOR1 on metabolism in PDAC, using both transcriptome and metabolome datasets from a total of 50 patients in our NCI-UMD-German cohort (Table S2).In this analysis, we identified 70 significantly upregulated and 64 downregulated metabolites in ELAPOR1-high tumors compared to ELAPOR1-low tumors (Figure 4A and Table S4).Pathway enrichment analysis using MetaboAnalyst 5.0 (https://www.metaboanalyst.ca) 21with these differential metabolites as input pointed to the upregulation of lipid metabolism and the downregulation of amino acid and carbohydrate metabolism in ELAPOR1-high tumors (Figure 4B,C).ELAPOR1 transcript expression levels in the unclassified and the classical/progenitor subtypes were rather similar (Figure S2A).Therefore, we hypothesized that ELAPOR1 is not involved in maintaining differences between these two subtypes.
Consequently, we opted to treat the classical/progenitor subtype and the unclassified subtype as a combined group, referred to as the "unified classical/progenitor subtype."We then compared the metabolic profile of the "unified classical/progenitor subtype" with that of the basal-like/squamous subtype.Exactly 98 metabolites were significantly increased and 155 were decreased in the unified classical/ progenitor subtype when compared to the basal-like/squamous subtype (Figure 4D and Table S5).Pathway enrichment analysis using MetaboAnalyst 5.0 unveiled the upregulation of lipid metabolism and the downregulation of amino acid and carbohydrate metabolism in the unified classical/progenitor subtype (Figure 4E,F).Furthermore, the enrichment analysis using IPA with the input of upregulated metabolites for both the ELAPOR1-high tumors and the unified classical/progenitor subtype tumors indicated the common activation of neuronal pathways (Figure S4A,B).
To continue our investigation into the role of these metabolites, we performed a global metabolome analysis in human PDAC cells.
Here, we observed that 64 metabolites were significantly upregulated and 85 were downregulated in ELAPOR1 transgene-overexpressing Panc 10.05 cells when compared to vector control cells (Figure 4G and Table S6).Pathway enrichment analysis using MetaboAnalyst 5.0 revealed that ELAPOR1-overexpressing Panc 10.05 cells upregulated lipogenesis, while downregulating amino acid and carbohydrate metabolism (Figure 4H,I).Previous studies from our group have shown that amino acid metabolism promotes the aggressiveness of PDAC, 17 while lipogenesis inhibits disease progression. 20We observed similar metabolic adaptation profiles in basal-like/squamous PDAC and ELA-POR1-low PDAC cells, which may potentially contribute to the disease progression.found a strong correlation between MNA and methionine levels in the PDAC tumors (P < .0001,r = .7411,Figure S4D).

| 1-Methylnicotinamide accelerates migration and invasion in PDAC cells
S-adenosylmethionine originates from methionine.2][33] Hence, we hypothesized that the downregulation of amino acid metabolism in ELAPOR1-high tumors may result in a decrease of MNA, leading to decreased disease aggressiveness in PDAC patients.PDAC cells (Figure S5A,B).These findings suggest that ELAPOR1-induced metabolic alterations may have a potential role in reducing disease aggressiveness.However, the oncometabolite MNA, on its own, could not revert ELAPOR1-induced phenotypes in transgene-expressing PDAC cells.

| DISCUSSION
In this study, we observed that downregulation of ELAPOR1 associates with the basal-like/squamous subtype and an unfavorable patient prognosis in PDAC.Upregulation of ELAPOR1 induced changes to cancer cell metabolism and inhibited migration and invasion in PDAC cells.Moreover, ELAPOR1 expression associated with neuroendocrine differentiation, a characteristic of the classical/progenitor subtype in which ELAPOR1 expression is maintained.
ELAPOR1, also known as KIAA1324 or estrogen-induced gene 121 (EIG121), encodes a transmembrane protein that localizes to various cellular membranes, including the plasma membrane, lysosome, endosome, and Golgi apparatus. 26,34ELAPOR1 regulates secretory granule differentiation 34 and autophagy under stress conditions. 26,35,368][39] In pancreatic β cells, ELAPOR1 inhibits insulin hormone and insulin-like growth factor 1 signaling by decreasing the number of cell-surface receptors and blunting the response to insulin. 402][43] ELAPOR1 also suppresses the progression of gastric cancer by inhibiting the oncoprotein GRP78. 29However, the function of ELAPOR1 in breast cancer remains controversial. 26,35,36A link between ELAPOR1 and metabolism has been reported 44 ; however, the role of ELAPOR1 in PDAC has not been described.
Cancer cells increase glycolysis, and lipid and amino acid synthesis through the induction of the pentose phosphate pathway. 8,94][15][16][17] Amino acids are vital metabolites for PDAC progression, 13,45,46 while lipids may have more complex effects. 20,46,47We previously reported that fatty acids exert growth-inhibitory effects in PDAC. 20Consistently, the alterations to the metabolism induced by ELAPOR1 may promote the classical/progenitor subtype because of the downregulation of carbohydrate and amino acid metabolism and the upregulation of lipogenesis.However, its loss may redirect differentiation into the basallike/squamous subtype with an increase in glycolysis and a decrease in lipogenesis.
As a metabolite of interest, MNA was one of the most strikingly downregulated metabolites in ELAPOR1-overexpressing Panc 10.05 cells.2][33] Other investigations characterized MNA as an immune-regulatory metabolite in human ovarian cancer, demonstrating its role in inhibiting interferongamma signaling in immune cells and contributing to tumor progression. 48Whereas an assessment of this signaling pathway in immune cells was not within the scope of the present study, we will investigate the impact of MNA on immune cell function in future research within the context of PDAC.Other metabolites, such as NAA, NAAG, guanidinoacetate, DHA, and ornithine were also highly upregulated in ELAPOR1-overexpressing Panc 10.05 cells.Moreover, the function of most of these metabolites has not been described in PDAC.These and other observations should be followed up in future studies.
Molecular subtyping in breast cancer has undergone extensive research and has proven to be clinically valuable. 49

Exactly 1 .
0 Â 10 4 of PDAC cells were seeded on 8-well Nunc™ Lab-Tek™ II CC2™ Chamber Slides (Thermo Fisher Scientific).Cells were washed with PBS and fixed in 4% paraformaldehyde in PBS at room temperature for 10 min, rinsed twice and permeabilized with 0.2% Triton X-100 in PBS for 20 min.Cells were rinsed again with PBS and blocked in Animal-Free Blocker ® and Diluent, R.T.U.(Vector Laboratories, Inc., Newark, CA; SP-5035-100) for 2 h at room temperature before overnight incubation with rabbit polyclonal anti-ELAPOR1 antibody (Thermo Fisher Scientific, PA5-72691, 1:150).Cells were then washed three times with PBS for 15 min and incubated with anti-rabbit IgG (H + L) highly cross-adsorbed Alexa Fluor 594 secondary antibody (Thermo Fisher Scientific, A-21207, 1:2000) for 1 h in the dark and at room temperature.Cells were then washed for 15 min, cultured with DAPI (Thermo Fisher Scientific, 62248, 1:1000) for 10 min, and mounted using ProLong™ Glass Antifade Mountant (Thermo Fisher Scientific).All samples were photographed at identical exposure times.

F
Figure 5A).NAAG and NAA are metabolites with neurotransmitter function, supporting our findings from the transcriptome analysis that

4
Upregulation of lipid metabolism and downregulation of amino acid metabolism in ELAPOR1-high PDAC.Analysis of the metabolome in PDAC tumors (NCI-UMD-German cohort) and Panc 10.05 human PDAC cells, contrasting ELAPOR1-high (n = 25) versus ELAPOR1-low (n = 25) PDAC tumors, classical/progenitor + unclassified tumors (n = 35) versus basal-like/squamous (n = 15) tumors, and Panc 10.05 cells with ELAPOR1 transgene expression (ELAPOR1-high) versus vector control (ELAPOR1-low).(A) Volcano plot of the metabolites with differential abundance in ELAPOR1-high versus ELAPOR1-low tumors.Seventy metabolites are significantly upregulated and 64 are downregulated in ELAPOR1-high tumors (P < .05).The dotted line indicates Àlog 10 (P = .05).(B,C) Pathway enrichment analysis with differential metabolites using MetaboAnalyst 5.0, indicating that lipogenesis-related pathways are upregulated (red arrows) and amino acid (blue arrows) and carbohydrate metabolism (orange arrow) are downregulated in ELAPOR1-high tumors.(D) Volcano plot of the metabolites with differential abundance comparing the "unified classical/progenitor subtype" (classical/progenitor subtype + unclassified subtype) with the basal-like/ squamous PDAC.Ninety-eight metabolites are significantly upregulated and 155 are downregulated in the unified classical/progenitor subtype when compared to the basal-like/squamous subtype.The dotted line indicates Àlog 10 (P = .05)thresholds.(E,F) Pathway enrichment analysis using MetaboAnalyst 5.0, indicating that the unified classical/progenitor subtype shows upregulation of lipogenesis (red arrows) and downregulation of amino acid (blue arrows) and carbohydrate metabolism (orange arrow).(G) Volcano plot of the metabolites with differential abundance in ELAPOR1-overexpressing versus vector control Panc 10.05 cells.Sixty-four metabolites are significantly upregulated and 85 are downregulated in ELAPOR1-overexpressing Panc 10.05 cells.The dotted line indicates Àlog 10 (P = .05).(H,I) Pathway enrichment analysis with differential metabolites using MetaboAnalyst 5.0, indicating that ELAPOR1-overexpressing Panc 10.05 cells exhibit upregulated lipogenesis (red arrows) and downregulated amino acid (blue arrows) and carbohydrate metabolism (orange arrow).
Breast cancer is commonly classified into four distinct molecular subtypes, with one subtype exhibiting basal-like characteristics.Diagnosis of the molecular subtype in each patient's breast cancer is typically determined using IHC, enabling the selection of appropriate therapeutic agents based on the diagnosis.Conversely, molecular subtyping in PDAC is still in its exploratory phase and PDAC molecular subtypes are currently being investigated using whole exome sequencing methods such as RNA sequencing.This study identified ELAPOR1 as a potential diagnostic marker for the classical/progenitor PDAC subtype, associated with favorable survival.If the molecular subtype of PDAC can be readily diagnosed using techniques such as qPCR or IHC, it may have significant clinical implications.The utilization of estrogen receptor in the molecular subtyping of breast cancer,49 where estrogen receptor-positive cases exhibit favorable survival, aligns with the identification of ELAPOR1, an estrogen-induced gene, as a marker for the classical/progenitor subtype that is associated with favorable survival.This observation strengthens the potential utility of ELAPOR1 as a reliable marker in PDAC molecular subtyping.This study has several limitations.One limitation arises from the absence of loss-of-function studies.Given that all the PDAC cell lines that we utilized lacked endogenous ELAPOR1 expression, the focus was on gain-of-function experiments.Future investigations should incorporate loss-of-function and rescue experiments to complement gain-of-function experiments with ELAPOR1 to provide a more comprehensive understanding of ELAPOR1's role in PDAC.Furthermore, we explored the association between ELAPOR1 protein expression and patient survival/histology using IHC.However, no significant difference by ELAPOR1 expression status was observed in the IHC.Given that we had fewer samples for IHC than for qPCR, the limited sample size may have impacted the findings.Another limitation is related to our experiments with MNA supplementation.MNA failed to enhance migration and invasion in PDAC cells with ELAPOR1 transgene expression but promoted these processes in PDAC cells with the control vector (FigureS5A,B).This discrepancy may arise from ELAPOR1-induced metabolic reprogramming that may have affected the MNA signaling pathway.Taken together, our data suggest that persistence of ELAPOR1 expression promotes transcriptome and metabolome characteristics indicative of the classical/progenitor subtype, whereas its loss associates with basal-like/squamous tumors and increased aggressiveness of PDAC.ELAPOR1 expression induced a distinct metabolic signature characterized by upregulation of lipogenesis and downregulation of amino acid metabolism, commonly observed in the classical/ progenitor PDAC subtype.Upregulation of ELAPOR1 also associated with improved PDAC survival and exerted inhibitory effects on migration and invasion in PDAC cells, indicating its potential tumor inhibitory function and may be further explored for designing novel interventional approaches in PDAC patients.