Decreased serum levels of glycerol-3- phosphate dehydrogenase 1 and monoacylglycerol lipase act as diagnostic biomarkers for breast cancer

Abstract

BACKGROUND:

Breast cancer (BC) is one of the most life-threatening cancer types among women. Despite major developments in medical sciences and technologies, the incidence and mortality rates of BC cases are still increasing. One of the reasons for this increase is the absence of an easy to perform early-diagnostic tool. Although there are defined BC biomarkers routinely used for diagnostic and prognostic purposes, none of these biomarkers is useful for early diagnosis. Therefore, early diagnosis of BC remains an important challenge and there is a great need for the early-diagnostic biomarker(s).

OBJECTIVE:

In this study, we aimed to evaluate the diagnostic and prognostic values of glycerol-3-phosphate dehydrogenase (GPD1) and monoacylglycerol lipase (MAGL) proteins as non-invasive serum biomarkers.

METHODS:

GPD1 and MAGL serum levels were determined by ELISA for BC patients ( $n=$ 100) from five different subtypes, and healthy controls ( $n=$ 20), and a comparative analysis was performed to determine statistically significant expression differences among the groups.

RESULTS:

The results provided evidence that GPD1 acted as a diagnostic biomarker in distinguishing triple-negative breast cancer (TNBC) patients from other subtypes, and MAGL acted as a diagnostic biomarker in distinguishing healthy individuals from BC patients.

CONCLUSION:

GPD1 and MAGL might be proposed as non-invasive diagnostic biomarkers for BC with high sensitivity and specificity.

Keywords

Breast cancer monoacylglycerol lipase glycerol-3-phosphate dehydrogenase diagnostic biomarkers ELISA

1. Introduction

Breast cancer (BC) is a worrisome disease for women around the world [1, 2]. According to the World Health Organization (WHO), 2.1 million women each year suffer from this debilitating disorder [3]. In 2018, WHO estimated that 627.000 women died from BC which accounted for 15% of all cancer deaths among women [3]. To improve BC outcomes and survival rates, better diagnostic and prognostic tools are needed. Those tools should help to screen large women populations and allow early detection of BC. Protein-based biomarkers circulating in serum are among the tools used for the screening of various diseases including BC. Unfortunately, so far, none of the known protein-based serum biomarkers allows the early detection of BC with high specificity and sensitivity [4, 5, 6, 7]. Therefore, the ongoing quest for the discovery of such biomarkers continues.

In a recent comparative proteomics study, several differentially regulated proteins were identified in BC tissue samples [8]. Bioinformatics analysis of the differentially regulated proteins underlined the importance of glycerol-3-phosphate dehydrogenase (GPD1, Uniprot# P21695) and monoacylglycerol lipase (MAGL, Uniprot# Q99685) which play essential roles in triacylglycerol and lipid metabolisms. Our previous findings provided evidence that GPD1 and MAGL were the two tissue-based diagnostic and prognostic biomarker candidates for BC and also held the potential to differentiate the triple-negative breast cancer (TNBC) subtype from the other BC subtypes. Although our findings, overall, are very promising, they may only be useful for monitoring the changes at the tissue level. Therefore, their use in clinics is limited and less practical. In this study, we aimed to carry out a follow-up study to evaluate the possibility of using GPD1 and MAGL as serum-based biomarkers for the diagnosis/prognosis of BC. For this purpose, GPD1 and MAGL serum levels were measured by ELISA, and the diagnostic and prognostic values of these two proteins as non-invasive serum biomarkers were evaluated.

2. Materials and methods

2.1 Ethical approval

Ethical approval for this study was obtained from the Ethics Committee of Kocaeli University with approval ID of KU-GOKAEK-2018/49. Informed consent was obtained from each patient.

2.2 Patient characteristics

Patients who were diagnosed with invasive or infiltrating ductal carcinoma at the General Surgery Department of Kocaeli University Medical School and underwent lumpectomy or mastectomy between 2015 and 2017 were evaluated for the study. The patients who received neoadjuvant treatment such as endocrine therapy and/or chemotherapy were excluded from the study group. A total of 100 female BC patients aged between 30–60 years old were selected and included in the study.

2.3 Procurement and storage of the sera

Five mL of blood was collected from each BC patient ( $n=$ 100) before surgery, and healthy volunteers ( $n=$ 20). Within 30 minutes after the collection of blood samples, serum was separated via centrifugation at 2000 xg for 10 minutes. All serum samples were then stored at $-$ 80 ${}^{\circ}$ C after snap-frozen in liquid nitrogen for long-term storage and were not subjected to freeze-thaw cycle.

2.4 Molecular subtyping of tumour samples

Molecular subtyping of tumour samples was performed at the Pathology Department of Kocaeli University Medical School based on the expression levels of ER, PR, HER2, and Ki-67 proteins. The BC patients were grouped into five different subtypes, namely Luminal A (ER $+$ and/or PR $+$ , HER2-, Ki-67 $<$ 15%); Luminal B (ER $+$ and/or PR $+$ , HER2-, Ki-67 $\geqslant$ 15%); Luminal B-like HER2 $+$ (ER $+$ and/or PR $+$ , HER2 $+$ , Ki-67 $\geqslant$ 15%); TNBC (ER-, PR-, HER2-), and HER2-enriched (ER-, PR-, HER2 $+$ ) [5, 9]. Overall, Luminal A (LumA), Luminal B (LumB), Luminal B-like HER2 $+$ (LumB-like), TNBC, and HER2-enriched (HER2-enr.) groups included 34, 33, 14, 11, and 8 patients, respectively. Detailed information on clinical parameters of BC patients was reported in our previous work [8].

2.5 Preparation of serum pools

Serum pools for each tumour subtype were created using equal volumes of serum samples from each patient of the same subtype. The control serum pool was similarly created using the serum samples of healthy volunteers ( $n=$ 20).

Table 1
Serum GPD1 and MAGL levels in the BC subtypes and the control group (mean $\pm$ standard deviation)

Subtype	LumA	LumB	LumB-like HER2 $+$	TNBC	HER2-enriched	Control
GPD1 level (ng/mL)	6.04 $\pm$ 0.38	4.72 $\pm$ 0.32	2.95 $\pm$ 0.22	1.77 $\pm$ 1.11	8.41 $\pm$ 0.75	5.36 $\pm$ 0.75
MAGL level (ng/mL)	6.41 $\pm$ 0.34	4.19 $\pm$ 0.35	3.30 $\pm$ 0.68	3.03 $\pm$ 0.69	2.57 $\pm$ 0.43	10.31 $\pm$ 1.22

2.6 Determination of serum GPD1 and MAGL levels by ELISA

Serum GPD1 and MAGL levels were determined by following the instructions of commercially available ELISA kits (MyBioSource, cat. no. MBS9344172 and MyBioSource, cat. no. MBS167556, respectively). In brief, the positions for the standards, samples, and the blank were assigned to the wells of the ELISA plates and added to each corresponding assigned well. HRP-conjugated reagents were added to each well and the plates were incubated at 37 ${}^{\circ}$ C for 60 minutes. After washing all wells, substrate solutions were added to each well and incubated at 37 ${}^{\circ}$ C in the dark for 10 minutes. At the end of the incubation, the reaction was stopped by adding the stop solution to each well, and optical density was measured at 450 nm within 15 minutes. GPD1 and MAGL concentrations were calculated using the optical density of known standards.

2.7 Statistical analysis

The statistical analyses were carried out using R (Version 3.5.3). All variables were expressed as mean $\pm$ standard deviation. The Mann-Whitney U test was applied to determine the statistical significance of the differences among the groups. Spearman’s rank correlation test was used to evaluate the pairwise correlations between groups. $P$ -values less than 0.05 were considered to be statistically significant. An outlier analysis was performed as needed, and data with at least 3 standard deviations from the sample mean were considered as outliers and excluded from the sample set.

2.8 Evaluation of the diagnostic values of the biomarkers

The diagnostic values of GPD1 and MAGL were determined through sensitivity and specificity metrics, and the Receiver Operating Characteristic (ROC) curve. The optimal cut-off level for biomarker quantification was calculated by the Youden index.

2.9 Evaluation of the prognostic values of the biomarkers

The prognostic values of GPD1 and MAGL were determined by using BC RNA-Seq dataset retrieved from The Cancer Genome Atlas (TGCA) and survival analysis via the SurvExpress validation tool [10]. In SurvExpress, the samples were partitioned into low- and high-risk groups according to their prognostic index. The prognostic capabilities of the potential biomarkers were characterized through Kaplan-Meier plots using the overall survival times of patients, the log-rank test, and the hazard ratio (HR).

3. Results

3.1 Serum GPD1 levels discriminated TNBC patients from the other BC subtypes and the control

Serum GPD1 levels for each BC subtype and the control were initially analyzed in serum pools by ELISA. Serum GPD1 levels for the TNBC, LumB-like HER2 $+$ , and HER2-enriched subtypes displayed statistically significant differences from the control pool (Fig. 1A). The TNBC subtype, especially, displayed significantly lower serum GPD1 levels when compared to other subtypes as well as the control. The differences in GPD1 levels between TNBC, other subtypes, and the control were statistically significant (HER2-enriched ( $p=4.73\times 10^{-5}$ ), LumA ( $p=5.46\times 10^{-6}$ ), LumB ( $p=8.34\times 10^{-5}$ ), LumB-like HER2 $+$ ( $p=$ 0.0206), and the control ( $p=1.10\times 10^{-3}$ ). Similar to the TNBC samples, serum GPD1 levels in LumB-like HER2 $+$ samples were also lower than GPD1 levels in LumA ( $p=8.59\times 10^{-4}$ ), LumB ( $p=2.47\times 10^{-3}$ ), HER2-enriched ( $p=3.70\times 10^{-3}$ ), and the control ( $p=$ 0.023). On the other hand, serum GPD1 level was significantly higher in the HER2-enriched group when compared to all other groups ( $p<$ 0.01).

Figure 1.

Serum expression levels of GPD1 and MAGL determined by ELISA (A) Comparative analysis of the GPD1 serum expression levels among the BC subtypes and control group (B) Comparative analysis of the MAGL serum expression levels among the BC subtypes and control group.

To further investigate serum GPD1 levels in the individual TNBC serum samples ( $n=$ 11), each sample was subjected to ELISA analysis. The results verified our previous finding and demonstrated that the serum GPD1 levels were significantly lower in TNBC patients (1.77 $\pm$ 1.11 ng/mL) than the healthy controls (5.36 $\pm$ 0.75 ng/mL) and the other BC subtypes (Table 1). Three samples (P3, P5, and P11) that displayed consistently high serum GPD1 levels ( $>$ 3 standard deviations away from the sample mean) were considered as outliers and thus excluded from the analysis. However, the high variability of GPD1 values in TNBC samples, as the variation (1.11) is still comparable to the mean (1.77) even though we excluded three samples.

3.2 Serum MAGL levels discriminated healthy individuals from all other BC subtypes

Serum MAGL levels for each BC subtype and the control were determined in serum pools by ELISA. The serum MAGL levels in all BC subtypes were significantly lower than in the control pool (Fig. 1B; Table 1). The lowest MAGL level was detected in the HER2-enriched serum pool while the control pool had the highest MAGL level. Like GPD1, serum MAGL levels in each TNBC sample were determined for cross-comparisons. Serum MAGL levels of two individuals (P2 and P6) were similar to the level detected in the control group, whereas all the other nine individuals had lower MAGL levels than the control pool. It was possible to state that control samples had much higher MAGL levels than all BC subtypes. The difference in serum MAGL levels between each BC subtype and the control were statistically significant (TNBC ( $p=$ 0.024), LumA ( $p=$ 0.0245), LumB ( $p=$ 0.0087), LumB-like HER2 $+$ ( $p=$ 0.0027), and HER2-enriched ( $p=$ 0.0041)) (Fig. 1B).

Figure 2.

GPD1 and MAGL mRNA expression levels were determined using the BC RNA-Seq dataset retrieved from The Cancer Genome Atlas (A) Comparative analysis of the GPD1 mRNA expression levels among the BC subtypes and control group (B) Comparative analysis of the MAGL mRNA expression levels among the BC subtypes and control group.

3.3 Serum GPD1 levels were more predictive in discriminating TNBC patients than GPD1 mRNA levels

Using the mRNA expression data retrieved from TCGA (BRCA dataset with 155 TNBC, 74 HER2, 507 LumA, 191 LumB, and 34 control samples), the GPD1 mRNA expression levels of each subtype were evaluated (Fig. 2A). A profile similar to the serum GPD1 levels was observed in GPD1 at mRNA expression levels for BC patients. However, when pairwise comparisons were performed between the subtypes using the mRNA levels, the differences were significant for only TNBC vs. LumA ( $p=6.22\times 10^{-24}$ ) and TNBC vs. control ( $p=3.81\times 10^{-14}$ ). There were no statistically significant differences among other subtypes except LumA vs. LumB ( $p=1.50\times 10^{-25}$ ) and LumB vs. control ( $p=4.79\times 10^{-14}$ ).

3.4 Serum MAGL levels rather than mRNA levels were more predictive in discriminating TNBC patients

Using the mRNA expression data retrieved from TCGA, the mRNA expression levels of MAGL in each subtype were also evaluated (Fig. 2B). mRNA expressions were the lowest in the HER2-enriched subtype and highest in the control group. But the differences were not statistically significant.

3.5 Decreased serum GPD1 level may be used as a potential diagnostic biomarker for TNBC

To assess the value of serum GPD1 levels as a diagnostic biomarker in differentiating the TNBC subtype from the other subtypes, a ROC curve was created and the area under the curve (AUC) metric was determined (Fig. 3A). The ROC curve indicated an AUC that corresponded to test the accuracy of 0.948 (95% CI: 0.928–0.968, $p<$ 10 ${}^{-3}$ ). The optimal cut-off value was determined by the Youden index as 3.18 ng/mL with corresponding sensitivity and specificity values of 95.0% and 77.5%, respectively (Fig. 3B). This finding implied that a patient with a GPD1 level of less than 3.18 ng/mL could be diagnosed as TNBC with 95% sensitivity and 77.5% specificity.

Figure 3.

Evaluation of the diagnostic value of GPD1 as a biomarker in distinguishing TNBC patients from other BC subtypes (A) To determine the diagnostic value of GPD1, the Receiver Operating Characteristic (ROC) curve, and the area under the curve (AUC) metric was employed. The ROC analysis showed that sensitivity and specificity were 95.0% and 77.5%, respectively and the AUC was 0.948 (95% CI: 0.928–0.968, $p<10^{-3}$ ). (B) The optimal cut-off level for GPD1 was calculated by the Youden index. The optimal cut-off value was shown to be 3.18 ng/mL.

Figure 4.

Evaluation of the diagnostic value of MAGL as a biomarker in distinguishing healthy subjects from BC patients from any subtypes (A) To determine the diagnostic value of MAGL, the Receiver Operating Characteristic (ROC) curve and the area under the curve (AUC) metric was employed. The ROC analysis showed that the sensitivity and specificity were 100% and 98.5%, respectively and AUC was 0.980 (95% CI: 0.968–0.992, $p<10^{-3}$ ). (B) The optimal cut-off level for GPD1 quantification was calculated by the Youden index. The optimal cut-off value was shown to be 7.23 ng/mL.

Figure 5.

Evaluation of the prognostic value of GPD1 and MAGL. Patients’ outcomes were analysed through Kaplan-Meier survival analysis using mRNA expression data retrieved from TCGA. (A) Box blot graph representing low- and high-risk groups for GPD1 in BC patients. Using the mean expression level of GPD1 mRNA, the samples were partitioned into two risk groups. (B) Box blot graph representing low- and high-risk groups for MAGL in BC patients. Using the mean expression level of MAGL mRNA, the samples were partitioned into two risk groups. (C) Kaplan-Meier survival curves of BC patients for GPD1. The group with higher GPD1 gene expression levels received a poor prognosis compared to the low-risk group with lower GPD1 levels (log-rank $p=$ 0.065 and HR $=$ 1.38). (D) Kaplan-Meier survival curves of BC patients for MAGL. The group with higher MAGL gene expression levels received a poor prognosis compared to the low-risk group with lower MAGL levels (log-rank $p=2.2\times 10^{-4}$ and HR $=$ 1.93).

3.6 Decreased serum MAGL level may be used as a potential diagnostic biomarker for BC

To assess the value of the serum MAGL levels as a diagnostic biomarker in differentiating healthy individuals from the BC patients, a ROC curve was created and the AUC was analysed (Fig. 4A). The ROC analysis indicated an AUC that corresponded to test the accuracy of 0.980 (95% CI: 0.968–0.992, $p<$ 10 ${}^{-3}$ ). The optimal cut-off value was determined by the Youden index as 7.23 ng/mL with corresponding sensitivity and specificity values of 100.0% and 98.5%, respectively (Fig. 4B). This finding implied that a person with a serum MAGL level higher than 7.23 ng/mL can be considered as healthy with 100% sensitivity and 98.5% specificity.

3.7 Elevated GPD1 and MAGL mRNA levels were correlated with the poor prognosis of BC

To further explore the prognostic performances of GPD1 and MAGL in BC, the outcomes of the patients were analysed through the Kaplan-Meier survival estimator (Fig. 5). GPD1 and MAGL mRNA data retrieved from TCGA was used to partition the samples into two risk groups (Fig. 5A and B). The group with higher GPD1 gene expression levels received a poor prognosis compared to the low-risk group with lower GPD1 levels (log-rank $p=$ 0.065 and HR $=$ 1.38) (Fig. 5C). The group with higher MAGL gene expression levels received a poor prognosis compared to the low-risk group with lower MAGL levels (log-rank $p=2.2\times 10^{-4}$ and HR $=$ 1.93) (Fig. 5D).

4. Discussion

BC is a disease with serious consequences and may result in death. The only way to escape from this disorder is to diagnose it at an early stage so that necessary precautions can be taken to prevent its spread. However, early detection of BC is not an easy task since it requires a physical examination of the person. In most cases, highly sophisticated instrumentation like the mammogram and MRI is also required. In cases where a suspicious lump is detected in the breast, then a biopsy sample is to be taken for a definite diagnosis. Unfortunately, the presently used screening methods for the diagnosis of BC are cumbersome and not practical [11, 12]. A more practical way of screening BC is to use serum biomarkers. However, the currently available serum biomarkers are not that helpful for the diagnosis of BC since they have low specificity [13]. For example, one of the most commonly used biomarkers, CEA, has a sensitivity of 95% with 18% specificity [14]. Similarly, CA 15-3, the soluble form of MUC-1 protein and widely used serum marker for BC, has a sensitivity of 95% with 32% specificity [14].

It has been long known that cancer cells can reprogram their central carbon metabolism and maintain an elevated level of glycolysis to meet their high energy needs [15, 16]. In addition to the changes in carbon metabolism, another aspect of metabolic reprogramming in cancer cells is aberrant lipogenesis [17]. The dysregulation of core enzymes in lipid metabolism and their contribution to cancer cell migration, invasion, and metastasis has been well established in the literature [18, 19]. The genes involved in cellular fatty acid uptake and de novo lipogenesis are amplified in metastatic tumours and their expressions have been associated with the aggressive properties of malignant cancers [20, 21]. In our previous work, we identified two differentially regulated proteins, namely GPD1 and MAGL, and established an association between lipid metabolism and BC. These two proteins were down-regulated in tumour tissues of BC subtypes in comparison to the control group, and the lowest tissue expression was in the TNBC subtype.

GPD1 is a member of the NAD ${}^{+}$ -dependent glycerol-3-phosphate dehydrogenase family and catalyses the reversible conversion of dihydroxyacetone phosphate to glycerol-3-phosphate which is converted to glycerol or used for the synthesis of triglycerides and phospholipids [22, 23]. This conversion is the key step in the integration of carbohydrate, lipid, and glycerol metabolism [24]. Based on the Human Protein Atlas, GPD1 expression can be detected in most malignant tumours including renal and hepatocellular carcinomas, and it is moderately expressed in BC tumour tissues compared to other cancer types [25].

MAGL is an enzyme responsible for the breakdown of monoacylglycerols to free fatty acids (FFA) and is considered to be the key enzyme in the regulation of the FFAs network [26]. Its expression was stated to be regulated in numerous aggressive tumours such as colorectal cancer, neuroblastoma, and nasopharyngeal carcinoma [27, 28, 29]. It is reported in the literature that cancer cells use de novo fatty acid synthesis rather than hydrolysis of neutral lipid stores [21, 30, 31, 32, 33]. However, there is also a compelling study stating that cancer cells use MAGL-dependent hydrolysis of monoacylglycerols as the major source of FFAs in aggressive malignant cells [34]. The current knowledge about the function of MAGL is not sufficient to explain this contradiction.

In this study, a comparative analysis of the serum GPD1 and MAGL levels was performed among the BC subtypes. The results indicated significantly low serum concentrations of both proteins in the TNBC subtype compared with the other BC subtypes. These results were consistent with our previous tissue-based mass spectrometry results [8]. GPD1 serum expression levels less than 3.18 ng/mL could be used to diagnose TNBC with a sensitivity of 95% and a specificity of 77.5%. The ELISA results suggested that serum GPD1 levels in BC patients had a higher diagnostic value than those used in clinics today. These results are also consistent with transcriptome data from TCGA. However, monitoring serum GPD1 levels was more informative than mRNA levels in distinguishing TNBC patients. The prognostic value of GPD1 in BC was determined via Kaplan-Meier survival analysis using mRNA expression data. Patients with higher GPD1 gene expression levels had a worse prognosis than the low-risk group with lower GPD1 levels. Although GPD1 had a high value as a diagnostic biomarker for BC, its prognostic value was not the same.

Comparative analysis of serum MAGL levels among BC subtypes and healthy subjects revealed significantly higher MAGL levels in controls. When serum levels were compared with the MAGL mRNA levels, serum MAGL levels were found to be more informative in distinguishing healthy subjects from BC patients. Analysis of the ROC revealed that subjects with serum MAGL levels higher than $>$ 7.23 ng/mL might be used to identify healthy subjects with a sensitivity of 100% and a specificity of 98.5%. Overall, the data showed that MAGL may have a potential value as a diagnostic biomarker for BC.

5. Conclusion

The quest to find a biomarker for early detection of BC continues. The biomarkers in use today are either not specific or do not allow early detection of BC. The classical radiological scans and the other clinical approaches do not appear to be powerful and convenient enough to allow early detection. In the late few years, novel biomarker candidates have emerged but they also suffered from sensitivity and specificity issues [35, 36]. Gene/protein signatures are proposed as biomarker panels but they stay short due to their high cost and availability. The work presented here analysed the serum expression profiles of GPD1 and MAGL proteins in BC subtypes compared to healthy controls. Statistical analysis of the ELISA results provided strong evidence that GPD1 acts as a diagnostic biomarker in distinguishing TNBC patients from other subtypes (sensitivity of 95% and specificity of 77.5%). Also, MAGL displayed value as a diagnostic biomarker in distinguishing healthy individuals from BC patients (sensitivity of 100% and specificity of 98.5%). The experimental findings directed us to evaluate the prognostic values of these two proteins at the transcriptome level by retrieving data from the TCGA database. The statistical analysis of transcriptomics data failed to support the evidence that GPD1 and MAGL hold prognostic values. However, our experimental data suggested that GPD1 and MAGL might be used as non-invasive biomarkers with high sensitivity and specificity for the diagnosis of BC.

Footnotes

Acknowledgments

This study was carried out in the Medical Biology Proteomics laboratory at Kocaeli University Medical School and supported by Kocaeli University, Scientific Research Unit under the grant number 2018/123. The funding body had no role in the design or execution of the study.

Author contributions

Conception: Murat Kasap, Kubra Karaosmanoglu yoneten.

Interpretation or analysis of data: Kubra Karaosmanoglu Yoneten, Kazim Yalcin Arga, Murat Kasap.

Preparation of the manuscript: Kubra Karaosmanoglu Yoneten, Murat Kasap.

Revision for important intellectual content: Gurler Akpinar, Kazim Yalcin Arga.

Supervision: Murat Kasap, Nihat Zafer Utkan.

References

Jemal

Siegel

and Ward

, Cancer statistics, 2010, CA: A Cancer Journal for Clinicians 60 (2010). doi: 10.3322/caac.20073.

DeSantis

C.E.

Fedewa

S.A.

Goding Sauer

Kramer

J.L.

Smith

R.A.

and Jemal

, Breast cancer statistics, 2015: Convergence of incidence rates between black and white women, CA: A Cancer Journal for Clinicians 66 (2016). doi: 10.3322/caac.21320.

WHO, https://www.who.int/cancer/prevention/diagnosis- screening/breast-cancer/en/, Accessed October, 2020.

Misek

D.E.

and Kim

E.H.

, Protein biomarkers for the early detection of breast cancer, International Journal of Proteomics 2011 (2011). doi: 10.1155/2011/343582.

Goldhirsch

Winer

E.P.

Coates

A.S.

Gelber

R.D.

Piccart-Gebhart

Thürlimann

Senn

H.J.

Albain

K.S.

André

Bergh

Bonnefoi

Bretel-Morales

Burstein

Cardoso

Castiglione-Gertsch

Coates

A.S.

Colleoni

Costa

Curigliano

Davidson

N.E.

di Leo

Ejlertsen

Forbes

J.F.

Gelber

R.D.

Gnant

Goldhirsch

Goodwin

Goss

P.E.

Harris

J.R.

Hayes

D.F.

Hudis

C.A.

Ingle

J.N.

Jassem

Jiang

Karlsson

Loibl

Morrow

Namer

Osborne

C.K.

Partridge

A.H.

Penault-Llorca

Perou

C.M.

Piccart-Gebhart

M.J.

Pritchard

K.I.

Rutgers

E.J.T.

Sedlmayer

Semiglazov

Shao

Z.M.

Smith

Thürlimann

Toi

Tutt

Untch

Viale

Watanabe

Wilcken

Winer

E.P.

and Wood

W.C.

, Personalizing the treatment of women with early breast cancer: Highlights of the st gallen international expert consensus on the primary therapy of early breast Cancer 2013, Annals of Oncology 24 (2013), 2206–2223. doi: 10.1093/annonc/mdt303.

Kazarian

Blyuss

Metodieva

Gentry-Maharaj

Ryan

Kiseleva

E.M.

Prytomanova

O.M.

Jacobs

I.J.

Widschwendter

Menon

and Timms

J.F.

, Testing breast cancer serum biomarkers for early detection and prognosis in pre-diagnosis samples, British Journal of Cancer 116 (2017), 501–508. doi: 10.1038/bjc.2016.433.

Loke

S.Y.

and Lee

A.S.G.

, The future of blood-based biomarkers for the early detection of breast cancer, European Journal of Cancer 92 (2018). doi: 10.1016/j.ejca.2017.12.025.

Yoneten

K.K.

Kasap

Akpinar

Gunes

Gurel

and Utkan

N.Z.

, Comparative proteome analysis of breast cancer tissues highlights the importance of glycerol-3phosphate dehydrogenase 1 and monoacylglycerol lipase in breast cancer metabolism, Cancer Genomics and Proteomics 16 (2019), 377–397. doi: 10.21873/cgp.20143.

Cho

, Molecular subtypes and imaging phenotypes of breast cancer, Ultrasonography 35 (2016), 281–288. doi: 10.14366/usg.16030.

10.

Aguirre-Gamboa

Gomez-Rueda

Martínez-Ledesma

Martínez-Torteya

Chacolla-Huaringa

Rodriguez-Barrientos

Tamez-Peña

J.G.

and Treviño

, SurvExpress: An online biomarker validation tool and database for cancer gene expression data using survival analysis, PLoS ONE 8 (2013). doi: 10.1371/journal.pone.0074250.

11.

Hellquist

B.N.

Czene

Hjälm

Nyström

and Jonsson

, Effectiveness of population-based service screening with mammography for women ages 40 to 49 years with a high or low risk of breast cancer: Socioeconomic status, parity, and age at birth of first child, Cancer 121 (2015). doi: 10.1002/cncr.29011.

12.

Onega

Goldman

L.E.

Walker

R.L.

Miglioretti

D.L.

Buist

D.S.

Taplin

Geller

B.M.

Hill

D.A.

and Smith-Bindman

, Facility mammography volume in relation to breast cancer screening outcomes, Journal of Medical Screening 23 (2016). doi: 10.1177/0969141315595254.

13.

Bates

S.E.

, Clinical applications of serum tumor markers, Annals of Internal Medicine 115 (1991). doi: 10.7326/0003-4819-115-8-623.

14.

Zaleski

Kobilay

Schroeder

Debald

Semaan

Hettwer

Uhlig

Kuhn

Hartmann

and Holdenrieder

, Improved sensitivity for detection of breast cancer by combination of miR-34a and tumor markers CA 15-3 or CEA, 2018, www.oncotarget.com.

15.

Zhou

Liotta

L.A.

and Petricoin

E.F.

, Cancer metabolism: What we can learn from proteomic analysis by mass spectrometry HHS public access, 2012.

16.

Liberti

M.V.

and Locasale

J.W.

, The Warburg effect: How does it benefit cancer cells? Trends in Biochemical Sciences 41 (2016), 211–218. doi: 10.1016/j.tibs.2015.12.001.

17.

Luo

Cheng

Tan

Tang

Yang

and Cao

, Emerging roles of lipid metabolism in cancer metastasis, Molecular Cancer 16 (2017). doi: 10.1186/s12943-017-0646-3.

18.

Santos

C.R.

and Schulze

, Lipid metabolism in cancer, FEBS Journal 279 (2012), 2610–2623. doi: 10.1111/j.1742-4658.2012.08644.x.

19.

Ray

and Roy

S.S.

, Aberrant lipid metabolism in cancer cells – the role of oncolipid-activated signaling, FEBS Journal 285 (2018), 432–443. doi: 10.1111/febs.14281.

20.

Kim

H.-Y.

Lee

K.-M.

Kim

S.-H.

Kwon

Y.-J.

Chun

Y.-J.

and Choi

H.-K.

, Comparative metabolic and lipidomic profiling of human breast cancer cells with different metastatic potentials, 2016. www.impactjournals.com/oncotarget.

21.

Menendez

J.A.

and Lupu

, Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis, Nature Reviews Cancer 7 (2007). doi: 10.1038/nrc2222.

22.

Hopkinson

D.A.

Peters

and Harris

, Rare electrophoretic variants of glycerol-3-phosphate dehydrogenase: evidence for two structural gene loci (GPD1 and GPD2), Annals of Human Genetics 37 (1974). doi: 10.1111/j.1469-1809.1974.tb01852.x.

23.

Menaya

González-Manchón

Parrilla

and Ayuso

M.S.

, Molecular cloning, sequencing and expression of a cDNA encoding a human liver NAD-dependent α-glycerol-3-phosphate dehydrogenase, Biochimica et Biophysica Acta (BBA) – Gene Structure and Expression 1262 (1995). doi: 10.1016/0167-4781(95)00069-S.

24.

Reshef

Olswang

Cassuto

Blum

Croniger

C.M.

Kalhan

S.C.

Tilghman

S.M.

and Hanson

R.W.

, Glyceroneogenesis and the triglyceride/fatty acid cycle, Journal of Biological Chemistry 278 (2003), 30413–30416. doi: 10.1074/jbc.R300017200.

25.

The Human Protein Atlas, http://v15.proteinatlas.org/ENSG00000167588-GPD1/cancer, 2020.

26.

Zhang

Liu

Lian

Liao

Chen

Qin

Wang

Jiang

Wang

and Gong

, Monoacylglycerol lipase: A novel potential therapeutic target and prognostic indicator for hepatocellular carcinoma, Scientific Reports 6 (2016). doi: 10.1038/srep35784.

27.

Zhang

Seviour

E.G.

Tao

Liu

Ling

Chen

and Wang

, Monoacylglycerol lipase (MAGL) knockdown inhibits tumor cells growth in colorectal cancer, Cancer Letters 307 (2011). doi: 10.1016/j.canlet.2011.03.007.

28.

Matuszak

Hamtiaux

Baldeyroux

Muccioli

G.G.

Poupaert

J.H.

Lansiaux

and Lambert

D.M.

, Dual inhibition of MAGL and type II topoisomerase by N-phenylmaleimides as a potential strategy to reduce neuroblastoma cell growth, European Journal of Pharmaceutical Sciences 45 (2012). doi: 10.1016/j.ejps.2011.11.011.

29.

W.-R.

Lian

Y.-F.

Peng

L.-X.

Lei

J.-J.

Deng

C.-C.

Feng

Q.-S.

Chen

L.-Z.

Bei

J.-X.

and Zeng

Y.-X.

, Monoacylglycerol lipase promotes metastases in nasopharyngeal carcinoma, 2014. www.ijcep.com/.

30.

Brusselmans

Timmermans

van de Sande

van Veldhoven

P.P.

Guan

Shechter

Claessens

Verhoeven

and Swinnen

J.V.

, Squalene synthase, a determinant of raft-associated cholesterol and modulator of cancer cell proliferation, Journal of Biological Chemistry 282 (2007), 18777–18785. doi: 10.1074/jbc.M611763200.

31.

Sebti

S.M.

, Protein farnesylation: Implications for normal physiology, malignant transformation, and cancer therapy, Cancer Cell 7 (2005), 297–300. doi: 10.1016/j.ccr.2005.04.005.

32.

Kuemmerle

N.B.

Rysman

Lombardo

P.S.

Flanagan

A.J.

Lipe

B.C.

Wells

W.A.

Pettus

J.R.

Froehlich

H.M.

Memoli

V.A.

Morganelli

P.M.

Swinnen

J.V.

Timmerman

L.A.

Chaychi

Fricano

C.J.

Eisenberg

B.L.

Coleman

W.B.

and Kinlaw

W.B.

, Lipoprotein lipase links dietary fat to solid tumor cell proliferation, Molecular Cancer Therapeutics 10 (2011), 427–436. doi: 10.1158/1535-7163.MCT-10-0802.

33.

Zaidi

Lupien

Kuemmerle

N.B.

Kinlaw

W.B.

Swinnen

J.V.

and Smans

, Lipogenesis and lipolysis: The pathways exploited by the cancer cells to acquire fatty acids, Progress in Lipid Research 52 (2013), 585–589. doi: 10.1016/j.plipres.2013.08.005.

34.

Nomura

D.K.

Long

J.Z.

Niessen

Hoover

H.S.

S.W.

and Cravatt

B.F.

, Monoacylglycerol lipase regulates a fatty acid network that promotes cancer pathogenesis, Cell 140 (2010), 49–61. doi: 10.1016/j.cell.2009.11.027.

35.

Jia

Chen

Wang

Jayasinghe

Luo

Wei

Wang

Xiong

Chen

Wang

Zhao

and Zhou

, Exosome: emerging biomarker in breast cancer, 2017. www.impactjournals.com/oncotarget/.

36.

Weigel

M.T.

and Dowsett

, Current and emerging biomarkers in breast cancer: Prognosis and prediction, Endocrine-Related Cancer 17 (2010). doi: 10.1677/ERC-10-0136.

Decreased serum levels of glycerol-3- phosphate dehydrogenase 1 and monoacylglycerol lipase act as diagnostic biomarkers for breast cancer

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Ethical approval

2.2 Patient characteristics

2.3 Procurement and storage of the sera

2.4 Molecular subtyping of tumour samples

2.5 Preparation of serum pools

Table 1 Serum GPD1 and MAGL levels in the BC subtypes and the control group (mean ± standard deviation)

2.7 Statistical analysis

2.8 Evaluation of the diagnostic values of the biomarkers

2.9 Evaluation of the prognostic values of the biomarkers

3. Results

3.1 Serum GPD1 levels discriminated TNBC patients from the other BC subtypes and the control

3.4 Serum MAGL levels rather than mRNA levels were more predictive in discriminating TNBC patients

3.5 Decreased serum GPD1 level may be used as a potential diagnostic biomarker for TNBC

3.7 Elevated GPD1 and MAGL mRNA levels were correlated with the poor prognosis of BC

4. Discussion

5. Conclusion

Footnotes

Acknowledgments

Author contributions

References

Table 1
Serum GPD1 and MAGL levels in the BC subtypes and the control group (mean $\pm$ standard deviation)