Research articles

By Dr. Sumaira N Syed
Corresponding Author Dr. Sumaira N Syed
General Surgery, Bedford Hospital, 72midland road - United Kingdom MK40 1QH
Submitting Author Dr. Sumaira N Syed

Breast Cancer, Biomarkers, Metabolomics

Syed SN. Biomarkers of Breast Cancer Cell Lines A; Pilot Study on Human Breast Cancer Metabolomics. WebmedCentral BREAST 2013;4(3):WMC004092
doi: 10.9754/journal.wmc.2013.004092

This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC-BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Submitted on: 05 Mar 2013 10:04:10 AM GMT
Published on: 05 Mar 2013 12:32:39 PM GMT


Metabolism of a cancer cell is significantly different from that of a normal cell. Therefore the metabolites of a breast cancer cell are also different from the metabolites of a normal breast epithelial cell and identification of the altered metabolites in body fluids may help us identify the presense of cancer in the body. Metabolites produced by breast cancer cells can  serve as potential biomarkers of breast cancer.

In this study the breast cancer cell lines were cultured and the metabolites released into the culture media were obtained into three extracts. Liquid chromatography/mass spectrometry were then used to separate and analyze these metabolites. Data sets produced were aligned by mass lynx software and subsequently subjected to multivariate statistical analysis. Metabolites  present in higher amounts in breast cancer cell lines were identified by comparison with the known masses in databases like METLIN. The fold rise of metabolites in breast cancer cell lines compared to the non tumourigenic cell line was used to identify the potential biomarkers of  breast cancer cell lines. It was found that there are more than eighty metabolites which can be regarded as biomarkers of cancer cell lines and there are six metabolites which can be regarded as specific potential biomarkers of breast cancer cell lines.


Breast cancer is the most common cancer among women in the U.K(1). Every year more than 40000  women in the UK develop breast cancer and nearly 10000 die because of this disease(2). To detect breast cancer at such a stage where  medical intervention can alter mortality and morbidity  statistics of this condition, we need to identify some measurable substance in the body that is  specifically associated with breast cancer. Such a substance if identified would serve as a marker of breast cancer. Its presense or rise in the normal value (if present under normal  conditions) would mean breast cancer is present in the body. Such substances often called  biomarkers of cancer are usually products of metabolism of tumour cells(3). Although it has been known for quite a long time now that tumour cell metabolism is different from the metabolism of  normal cell, the altered metabolites of breast cancer cells have not yet been utilized as  metabolic biomarkers for breast cancer screening. It would be extremely useful to identify the  biomarkers of breast cancer in body fluids as they would provide non biopsy tests which would  be highly sensitive and specific for this condition. Identification of the altered metabolites  originating from breast cancer cell lines that differ significantly from the metabolites of non tumourigenic breast epithelial cell line was the focus of this study.

Aims & Objectives

The overall aim was to identify classes of metabolites that are associated with the development  of breast tumour phenotype in the cell lines. The specific aims were

a) To develop sample extraction methods in order to profile a wide range of metabolites in cultured breast epithelial cell lines.
b) To use liquid chromatography/mass spectrometry and multivariate statistical methods to determine how metabolite profiles differ between normal and tumourigenic breast cell lines.

Materials And Methods

Cell Lines: Three breast cell lines were used; MCF-7, an early stage breast cell line obtained  from Cell Line Services (CLS), Germany.

MDA-MB-231, an invasive human breast cancer cell line obtained from Cell Line Services (CLS) ,Germany.

MCF-10A, a non tumourigenic breast epithelial cell line obtained from American Type Culture Collection, USA used for comparison.

Cell Culture Media: The breast cell lines were cultured in the DMEM/F12 medium (Invitrogen Cat  no 21331020). It was supplemented with fetal bovine serum, pencillin, streptomycin, glut amine, epidermal growth factor, hydrocortisone, cholera toxin and bovine insulin. After being purchased the breast cell lines were maintained in an incubator at 37° C with 5% CO2 & splited 2-3      times/week. The culture  media with supplements was put in the six well culture plates. 3ml of  a cell line (having 2.0×10>cells/ml) was placed in the walls of  these culture plates. Thus we   had many replicates of each cell line. These culture plates were kept in the incubator overnight and thus the three cell lines got cultured under similar conditions. Metabolites produced by  the cell lines during culture were expected to be present in the culture media. Therefore only  the supernatant from the culture media was taken in tubes that were labeled as the replicate  numbers of the cell lines. Some tubes were kept as controls and these had media that was not in  contact with cell cultures. All the tubes were stored at -80°C.

Solid Phase Extraction (SPE)

These method was used for concentrating metabolites from the media in  the tubes. 0.5ml methanol had been added to the tubes. SPE method involved loading the sample  solutionon to SPE phase, wash away undesired components and then washing off the desired analytes with another solvent into a collection tube. For all extractions we made an internal  standard mixture of stable isotopes. In 100µl ethanol we added 1ong each of d4 estradiol, d4  estradiol sulfate and d9progesterone. The stock solution of 1µg/ml concentration was kept at  -20°C. The tubes were taken out from the 80°C freezer and thawed. 5ml of media from the  replicates were taken each time. Then the tubes were vortexed and centrifuged. The top 2ml of  the supernatent were placed in labelled glass tubes. 2ml of 2% formic acid in water was added  to buffer them. 100µl of internal standard was added and tubes were vortexed. The cartridges  used for SPE were strata X-AW 60mg/3ml manufac tured by Phenomenex. The sorbent lot number used  was S308-19. For conditioning the cartridges 3ml ethyl acetate, 3ml methanol and 3ml 2% formic  acid in water was used. Then the sample was applied to SPE. 3ml deionised pure water was used  for washing each tube and the tubes were then dried for 15 minutes. One set of tubes was put under SPE, eluted with 3ml ethyl acetate and labelled ETAC n. Second set of tubes was put under SPE eluted with MEOH and labelled MEOHn. Third set of tubes was put under SPE eluted with ammonium hydroxide in methanol and labelled AMMn. N indicated the replicate no.of the cellline. The samples were stored at -20C overnight. ETAC extract of  the media was expected to contain  neutral or lipophilic metabolites (eg- steroids, nucleosides, fatty acids, phospholipids) MEOH extract of the media was expected to contain more polar neutral molecules, prostaglandins and phospholipids. AMM fraction was expected to contain conjugated anionic 3metabolites including  organic acids. Ultra performance liquid chromatography electrospray ionisation time of flight mass spectroscopy [UPLC-ESI-T OF MS] Chromatographic separation was performed using a Waters ACQUITY UPLC™ system (Waters Corp., Milford, USA), equipped with a binary solvent delivery  system and an auto-sampler. A Waters 100mm × 2.1 mm ACQUITY C18  1.7 µm column was used to separate the endogenous metabolites. The mobile phase consisted of SOLVENT (A) 0.2% formic acid in water and 5% acetonitrile in water(B) 100% acetonitrile and 0.2% formic acid in water. The following gradient program was used for the MS analysis in positive mode: 0-15.0 min from 0.0 to 100% B then held in 100% B for 10 mins. In negative ESI mode the same gradient program was used. Analytes were detected with a Micromass (Waters,Manchester, UK) TO F-MS system with an ESI source operated in either negative or positive mode. Capillary voltage was set at 2.60kV in positive mode and at between -2.60V and -2.75V in negative mode. Argon was used as collision gas at TOF penning pressures of 274.83 × 10^-7 to 5×10^-7 mbar. Collision energy was set at 10 eV  to avoid fragmentation of the analytes. Sulfadimethoxine (5pg/µl in methanol/water, 1:1, v/v, plus, in positive mode only, 0.1% formic acid) was used as internal lock mass infused at 40µl/min via a lockspray interface (baffling frequency;0.2 s^-1) to ensure accurate mass measurement. The internal lockmass m/z ratios were 311.0814 and 309.0658 in positive and  negative mode respectively. Source temperature was 1000C and desolvation temperature  was 3000C. The nebulising and desolvation nitrogen flows were maintained at 100 and 400 l/h respectively. The mass spectrometer was calibrated with sodium iodide and the spectra were  collected in full scan mode from 100 to 1000 m/z. Full scan mass spectra of the range of  metabolites were recorded in positive and negative modes in order to select the most abundant m/z ion. Thetheoretical parent ions for each metabolite were calculated from the atomic mass of  the most abundant isotope of each element by using the Molecular Weight Calculator software ( Mass Lynx 4.1 Software). The ETAC &  MEOH extracts of samples were analysed in +ESI & AMM  extracts were analyzed in– ESI  mode. Mass Lynx Software: Mass Lynx Software was used for MS analyses. Extraction of the spectral peaks from the raw data and then chromatogram alignment were carried out automatically by using Marker lynx v 4.1 softwarepackage [waters corporation, Milford, MA, USA] The parameters used for detecting the spectral peaks were optimised to minimize noise level of the detected signal. The parameters used were; mass accuracy of the acquired data[mass tolerance]:0.05 Da ; width of an average peak at 5% height: 5s; baseline  noise between the peaks [peak to peak baseline noise]:100; number of masses per RT submitted to  the collection algorithm:50 minimum intensity allowed for a spectral peak to be defined as a marker: 1% of the base peak intensity [BPI]. The BPI chromatograms were used to calculate the presense of internal standard in this study (Figure 1)

The mass lynx software provided base peak intensity  (BPI) chromatograms of the samples. The following figure illustrates BPI chromatograms of the media(M)samples analyzed in +ESI mode.

See Illustration 1

Figure 1: The BPI chromatograms of these media samples show similar features as expected since none of these samples have any breast cells. Some differences can however be noted. It was  expected that the chromatograms within a class (eg MCF10A) in any extract (eg methanol) should be similar. Infact the BPI chromatograms were very similar to each other as expected although some differences were also there. The differences could either be due to biological variability in the samples or due to the variability in extraction of the samples. The variation in extraction was revealed by the difference in the content of internal standards (calculated from their BPI chromatograms) in the samples of the same class.

Multivariate Analyses Of The Metabolomic Data: Data was the nexported to SIMCA-P software [Umetrics UK ltd, Winkfield,Windsor Berkshire,UK] for analysis. Before the multivariate analysis, dataa recentred pare to scaled and log transformed in order to optimise data and  limit skewness. Each dataset comprised an SPE fraction of the control and samples. Initially, principal component analysis [PCA] was performed to obtain an overview of the data and to identify outliers. Data was then subjected to projections using PLS-DA to find class– separating differences [variables] in pairwise comparisons of the treatments. Finally, OPLS-DA was  performed with the data to filter the information that was only due to class separation. Cross validation CV, default parameters) was used to determine the significant components of the models and thus minimise overfitting. The performance of the models was then described by the explained variation [R2X for PCA and OPLS-DA and R2Y for PLS-DA and OPLS-DA] and predictive  ability [Q2] parameters of the models { WORK FLOW illustrated in Figure 2}

The following figures represent the  models obtained by the SIMCA software after the mass lynx  data analysis. In the models the three classes indicate the three cell lines. The blue dots  represent the MDA-MB-231 cell line replicates, the red dots represent the MCF-7 cell line  replicates and MCF 10A cell line replicates are represented by black dots. The model  characteristics are given in the tables following the figures.

See Illustration 2

Figure 1a: shows the PCA model (labelled M21 in the  study) of MEOH extracts of the three cell  line replicates & it can be seen that classes are not grouped together

See Illustration 3

Figure 2b: shows three dimensional view of the same M21 model. The characteristics of this PCA- model M21 are shown in the table below.

See Illustration 4

Figure 2a: shows the PLS-DA model(M23) of the earlier shown PCA model, here class separation is  much better

See Illustration 5

Figure 2b: shows the three dimensional view of the above figure The class separation is  evident. This  PLS-DA  model  M23  had  the  following  characteristics:

See Illustration 6

Figure 3: Indicates the OPLS model of the same data set from which the preceding PCA & PLS-DA  models were obtained. The cell line MCF 10A (black) replicates & cell line MCF-7 (red) replicates are distinctly separate.

See illustration 7

Figure 4: Shows the S-plot obtained from the OPLS model in the previous figure. The lower end  of the S-plot has the masses of metabolites which are significantly produced in higher amounts  by the cancer cell line MCF-7 replicates as compared to the MCF-10A non tumourigenic cell line  replicates.

Six S-plots were obtained during this study. Three could be used to compare the amount of metabolites produced differentially by MCF-7 cell line replicates and MCF-10A cell line replicates in the three different extracts.

The other three S-plots were used to compare the amount of metabolites produced differentially  by MDA-MB-231 cell line replicates and MCF-10A cell line replicates in the three different  extracts.

After noting the metabolite masses at the extreme ends of the S-plots, comparisons could be  made between their presense in the breast cancer cell line sample replicates against the  control i.e MCF 10A cell line replicates.

Compounds that were totally absent in the normal cell line but present in significant amount in  the cancer cell line were selected as potential biomarkers

Their identity could be established by comparing them with the masses of known compounds in the known data bases.


After noting the metabolites (m/z i.e. mass/charge ratios) at the extreme ends of the S-plots, comparisons could be made between theirpresense in the breast cancer cell line sample replicates against the control i-eMCF 10A cell line replicates. The metabolites were identified from their  m/z values by comparison with the known compounds in data bases like human metabolome database  and metlin. The information obtained by the comparison has been summarized in the form of table1. The first column shows the retention time (RT) of the metabolite. The second column shows the mass charge ratio of the metabolite (m/z). The third column shows the range in         which the metabolite is produced by the breast cancer cell line (MCF 7 or MDM-MB-231) the  fourth column shows the range in which the metabolite is produced by the breast cell line  MCF10A. The fifth column shows the fold rise (^)    

The sixth column indicates the mass of the closest matched known substance found in the  databases. The last column gives the common names of the known substances of similar masses. The columns up headings MCF 7, MCF10A & MDM-MB-231 within the table indicate the cell extract  in  which the metabolite was found as given in the table. There are also subheadings AMM, MEOH  & ETAC which indicate the extract in which the following metabolites were found. The matches  from the human metabolome are given first under the heading common name of known compound. The matches from the metlin database are written below in the same column shaded in blue. The  metabolites analyzed in Amm extract were searched for in [M-H] mode & the metabolites in ETAC &  MEOH extract were searched in [M+H] mode. {Table 1} The table shows the identified metabolites belong to all classes of compounds which means that metabolism of carbohydrates, fats, proteins and all other classes of compounds is altered in a tumourigenic cell.

See Illustration 8

This means that a metabolite found in the ETAC extract having mass/charge [m/z] ratio of  332. 326 with a retention time [Rt] 9.29 is produced by some MDM-MB-231 cancer cells upto a  value of 100 & by some upto a value of 5; therefore range used is 5-100. But the same metabolite is produced by the normal cell MCF 10A upto a value of 0-4. So I infer that this is a metabolite that is normally produced by a breast epithelial cell and its production can rise upto 25 fold  in breast cancer {100/4=25}. The fold rise has been calculated by the ratio as illustrated. The  fold rise has been rounded to nearest non-decimal value in all rows of the column.

Thus it is clear from the table that there are more than eighty metabolites whose levels are  significantly higher in breast cancer cell line extracts as compared to non cancer cell lines.


The compounds that were found to be differentially produced in two cell groups in this study  are numerous. But only those metabolites can be potential biomarkers of breast cancer which are  markedly different as indicated by their fold rise in the tables. e.g. Substances with a fold  rise 2 are produced in double amount by tumour cell lines than by non tumour cell lines. But  this may happen in  inflammatory diseases also. So I concentrated on substances whose fold rise  was much higher.

The following information about the metaboites identified as potential biomarkers was obtained  from the human metabolome database. The chemical formulas and the mass weight differences are  also given in the website but I have not noted them here because there are numerous possible  adduct ions and I only had to list the nearest matches in the table that I made & only seven  columns could be incorporated in the table.

There are more than eighty metabolites identified in the table which seem to be potential  biomarkers of cancer cell lines in general as the metabolism is altered in every cancer. However I noted six metabolites which seem to be potential specific breast cancer cell line biomarkers.

PE(14:1(9Z)/14:1(9Z)) a phosphatidylethanolamine (PE) was found to be upto 14 fold more in the  ETAC extract of the MCF 7 breast cancer cell extracts than in non cancer MCF 10A cell line extracts in our experiment. Previous studies using NMR spectro scopy(25) have shown high  amounts of PE in breast cancer cells. Our experiment confirms that and we also find that the  amount produced by breast cancer cell can be upto 14 times more than that produced by non  cancer breast epithelial cell. Although glycerophosphoethanolamines can have many different com binations of fatty acids, but PE(14:1(9Z)/14:1(9Z)), in particular, consists of two chains of myristoleic acid at the C-1 and C-2 posi tions. The myristoleic acid moieties are derived from milk fats. Milk is produced only in the breast tissue. So I think that this metabolite can be a  specific  breast  cancer  biomarker. Its  increased  production  also  means  that  phospholipid degradation is much more in breast cancer cells than in normal breast epithelial cell. This was  illustrated further by the increased amounts of other phospholipids found in the present study.

Another glycerophosphoethanolamine in the ETAC extract of MCF-7 breast cancer cell lines was   PE(22:6(4Z,7Z,10Z,13Z,16Z,19Z )/14:1(9Z)), which consists of one chain of docosahexaenoic acid at the C-1 position and one chain of myristoleic acid at the C-2 position. The myristoleic acid moiety is derived from milk fats which makes this also a specific breast cancer marker. Its  levels were double in the breast cancer cell line extracts. PE(15:0/18:4(6Z, 9Z, 12Z,15Z)) a phosphatidylethanolamine(PE) consisting of one chain of pentadecanoic acid at the C-1 position and one chain of stearidonic acid at the C-2 position. The pentadecanoic acid moiety is derive ed from milk fat. Thus this is also specific to the breast tissue. It was found in the ETAC  extract of MCF 7 CA cells. It was found to be upto 37 fold more in CA cell extracts. PE(18:4(6Z,9Z, 12 Z,15Z)/15:0) consists of one chain of stearidonic acid at the C-1 position and one chain of pentadecanoic acid at the C-2 position. The pentadecanoic acid moiety is derived from milk fat making this PE specific to breast tissue. It was found to be three fold more in  breast cancer ETAC extracts of MCF 7 cell line.

DG(14:1(9Z)/20:5(5Z,8Z,11Z,14Z,17Z)/0:0) a diacylglycerol which usually can have many different combinations of fatty acids but DG(14:1(9Z)/20:5(5Z,8Z,11Z,14Z,17Z)/0:0), in particular, consis ts of one chain of myristoleic acid at the C-1 position and one chain of eicosapentaenoic acid at the C-2 position. The myristoleic acid moiety is derived from milk fats making it a specific  breast tissue metabolite. It was found elevated in MCF-7 breast cancer cell line ammonium  extracts as compared to the MCF10A non tumour igenic cell line.

DG(20:5(5Z,8Z,11Z,14Z,17Z)/14:1(9Z)/0:0) a glyceride consists of one chain of eicosapentaenoic acid at the C-1 position and one chain of myristoleic acid at the C-2 position. The myristoleic acid moiety is derived from milk fat making this a specific metabolite of breast tissue. It was  found in the ammonium extract of MDA-MB-231 breast cancer cell lines. The amount produced was  upto five fold more in the breast cancer cell lines as compared to the MCF 10A non cancer cell  lines.

LysoPE(0:0/20:3(11Z,14Z,17Z)) a lysophospholipid(LPL) that was found to be produced upto 14  fold more by breast cancer cells. LPLs are breakdown products of phosphatidylethanol amine found in all the cells. But the amount produced is much smaller. That indicates that it might be  useful as a general tumour marker but not as a specific breast cancer marker.

LysoPC(20:2(11Z,14Z)) another lysophospholipid (LyP) is found in normal conditions in the blood  plasma. An enzyme lecithin: cholesterol acyltransferase (LCAT) secreted from the liver is  invoved in determining its plasma level under normal conditions. In this study it was found  that a breast cancer cell produces it upto 7 fold more of this metabolite than a normal breast  epithelial cell. This I feel can also be a suitable marker as it is measured in the plasma at  present also and no new tests need to be devized. Although it is not specific to breast cancer, but if we do know the normal range in plasma, we can evaluate its levels in breast cancer to  see how useful its measurement can be.

Sphingomyelin SM(d17:1/24:1(15Z)) or SM(d17:1/24:1(15Z)) and Sphingomyelin (d18:0/14:0) or SM(d18:0/14:0) are both sphingolipids found in cell membranes especially in the myelin sheath  of nerve cells. Sphingomelins besides having a ceramide core (sphingosine bonded to a fatty acid via an amide linkage) additionally contain either phosphocholine or phosphoethanolamine. In  breast cancer cell extract not only these two sphingomyelins but also their component molecules  were found in abundance which again reflects the increased metabolism (20 fold more in our exp eriment) of sphingolipids in breast cancer cells. Ceramide was found to be produced upto 100  fold more by some cancer cells in contrast to non cancer cells. Whether this rise is associated  only with breast cancer cells or it occurs in other cancers also needs further studies.

TG(20:5(5Z,8Z,11Z,14Z,17Z)/18:3(9Z,12Z,15Z)/22:6(4Z,7Z,10Z,13Z,16Z,19Z))[iso6] is a monodocosahexaenoic acid triglyceride. It was found to be present 2-20 times more in breast  cancer extracts as compared to normal cells.

Another metabolite found to rise upto thirty fold more was Farnesol. It is an intermediate in the  isoprenoid/cholesterol bio synthetic pathway & plays a role in controlling the degradation of 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase. Studies have shown that it can  activate the farnesoid receptor (FXR), a nuclear receptor that forms a functional heterodimer. The exogenous farnesol effects various physiological processes like inhibition of phosphatidylcholine  biosynthesis induction of apoptosis, inhibition of cell cycle progression and actin cytoskeletal disorganization. The elevated farnesol found in breast cancer extracts in our study thus validate these  findings as the cancer cell division is uncontrolled apoptosis is absent and phospholipid  turnover is high. However I cant be sure if its level is so high in breast cancer only (thirty fold rise) or in other cancers also. But it can be again a general tumour marker.

A substance which is usually found in the nerve tissue was found to be particularly raised in  breast cancer extracts in our study. This was Gamma Glutamylglutamic acid. It was found to be  20 fold more in cancer as compared to non cancer cell extracts. It is made of two glutamate  molecules. Normally glutamate plays a role in synaptic plasticity. In brain injury it accumulates out side the cells which causes calcium ions to enter the cells via NMDA receptor channels. That causes neuronal damage and cell death. Excessively high intracellular Ca2+ is  believed to bring about the fatal changes by damaging mitochondria and inducing pro-apoptotic  genes. I think that this could be a mechanism of the normal breast epithelial cell death under  the effect of metabolites produced by breast cancer cells nearby.

S-(PGJ2)-glutathione a glutathione conjugate of prostaglandin J2 was found to be two  to five fold more in cancer cell extracts. It is one of the PGD2 dehydration product 9-deoxy- Δ9-PGD2 (also called prostaglandin J2). It has known cytotoxic activity and its elevated levels indicate that  glutathione conjugation is increased in breast cancer cells. It is two to five  fold increased in breast cancer cell extracts but is absent/negligible in normal cells. So it  may be a potential marker of this tumour.

Similarly 2-N-[5-(4-bromophenyl)-1,3,4-thiadiazol-2-yl]-1-N-(3,4- difluoro phenyl) pyrrolidine-1,2-dicarboxamide is present in cancer cell extracts but absent in normal  cell extracts. This can be a potential biomarker specific for breast cancer.

3 alpha,7alpha-Dihydroxy-5beta-cholestan-26-al is an intermediate in bile acid biosynthesis, specifically in the synthesis of chenodeoxyglycocholate and litho cholate. It was found to be  20 fold more in breast cancer cell extracts than in non cancerous cells. Bile acids are  believed to regulate all key enzymes involved in cholesterol homeostasis. Bile acids have  potent membrane disrupting potential. These facts indicate that in breast cancer cells there  are possibly enzymes which direct bile acid synthesis although normally this only occurs in  liver. If we consider the other way round it could be that there are enzymes which degrade  cholesterol and lipids to compounds which are similar to bile synthesis intermediates.

27-Norcholestanehexol is a bile alcohol present in minute amounts in the bile and urine in  healthy subjects. Bile alcohols are end products for cholesterol elimination. Presense of  elevated amounts of this bile alcohol in breast cancer cell extracts indicates the increased  metabolism of cholesterol in cancer cells.

Another metabolite found to be 20 fold more in breast cancer cell extracts. This is either a,24R,25-Trihydroxyvitamin D3 or 7alpha,26-Dihydroxy-4-cholesten-3-one. A precursor of chenodeoxycholic acid, 7 alpha, 26-dihydroxy-4-cholesten-3-one, found in elevated amounts again  establishes the presense of bile compounds in breast tissue. As regards alpha 24R, 25-trihydroxy vitamin D3 it is known that prostate cells can produce 1a,25-dihydroxyvitamin D3 (1a,25(OH)2D3) from 25-hydroxyvitamin D3 (25(OH)D3) to regulate their own growth. Whether the same can occur  in breast cancer cells needs further studies.

Demethylphylloquinone a form of vitamin K was found to be raised 5 fold to 70 fold in breast  cancer cell extracts. This is therefore a potential marker for breast cancer.

Mesaconic acid and itaconic acid were found to be five fold more in cancer cells than normal  cells. Increased amounts of these metabolites establishes the fact that anaerobic carbohydrate  metabolism is carried out excessively in breast cancer cells. Deuteroporphyrin IX is a  non-natural dicarboxylic porphyrin usually described as a fecal porphyrin in patients with endemic chronicar senic poisoning. Deuteroporphyrin IX was found more than thirty fold more in  cancer cells than in non cancer cells. Protoporphyrinogen IX was a similar metabolite found to  be 4 fold more from cancer cells. Gamma-delta-Dioxovaleric acid is produced by enzymes acting  in porphyrin metablosm. Its levels in breast cancer extracts were four fold more than in non  cancer cell extracts. Thus we can conclude that even porphyrin metabolism products are produced  in increased amounts by breast cancer cells.

Nicotinate D-ribonucleoside was found to be 2-6 fold more in breast cancer cell extracts than in non cancer cell extracts. This validates the finding in previous studies that ribonucleoside  levels are altered in breast cancer cells as compared to non cancer breast epithelial cells. Normally it is involved in the nicotinate and nicotinamide metabolism pathways. Its altered  levels indicate that energy production pathways are also altered as expected in tumour.

Uridine 2',3'-cyclic phosphate is a cyclic nucleotide. Cyclic phosphates are commonly found at the 3' end of mRNAs and other small RNAs. Uridine 2',3'-cyclic phosphate is a substrate for the enzyme 2',3'-cyclic nucleotide-3'-phosphodiesterase (CNPase, EC which hydrolyses it to Uridine 2'-phosphate. CNPase is a unique RNase in that it only cleaves nucleoside 2',3'-cyclic phosphates and not the RNA internucleotide linkage, like other RNases such as RNase A and RNase T1.

1,7-dimethylxanthine (paraxanthine) is a metabolite belonging to Super Class- Nucleosides and Nucleoside conjugates. It is the preferential path of caffeine metabolism in humans. It was  found to be produced more than 20 fold by breast cancer cells than by non cancer cells  indicating a massive alteration of nucleoside metabolism in cancer cells (33).

Cervonyl carnitine is an acylcarnitine. It was found to be three fold more in breast cancer  cell extracts than in non cancer cell extracts. It could well be a marker of tumour.

4-Nitrocatechol a by-product of the hydroxylation of 4-Nitrophenol by the human cytochrome P450 (CYP) 2E1 is a useful metabolic marker for the presence of functional cytochrome P450 2E1 in microsomes  of  the  cells.

Enterostatin VPGPR (Val-Pro-Gly-Pro-Arg) is a pentapeptide that is released from procolipase during fat digestion. Enterostatin levels are elevated in the plasma of obese women but were  found to be raised in breast cancer cell lines also.

There are also metabolites which are significantly higher in the normal breast epithelial cells  and almost produced negligibly by the breast cancer cell lines. These are located in the S-plot  on the end opposite (i.e. the higher extreme) to the end where the cancer cell line marker  metabolites are located (i.e.lower  extreme).

Discussion and Conclusion

Historical background:

The field of cancer metabolomics is relatively new. The few studies done on breast cancer  metabolism so far haverevealed high phosphocholine levels in breast cancer cells as compared to  normal breast cells. 31 P spectroscopy showed more phosphomonoesters in  breast cancer cells  than in normal cells and the phosphocholine and phosphoethanolamine were found to contribute to  the high PME sig nal in breast cancer cells in vivo in a multinuclear Nuclearmagnetic resonance  spectroscopy study (4). In a proteomics study, proteome analysis of different breast cancer  cell lines identified differentially expressed proteins. O16 /O18 peptide labelling was done so  as tocompare peptides in one sample [labelled with O18] with peptides in another sample [label led with O16] by mass spectrometry. Heirarchial clustering showed that various proteins were  differentially expressed in the cancer cell lines (5). Another study focussed on the lipid and carbohydrate metabolites of breast cancer cells. Lipid biosynthesis in tumour cells was     found to be altered as compared to normal cells.Levels of enzymes of lipid biosynthesis pathway  like fattyacid synthase & 2,4-dienoyl coenzyme. A reductase were also found to be different    as compared to normal cells. The altered carbohydrate metabolism in tumour cells was found to be characterized by increased glucose uptake and elevated glycolysis. Expression of lactate dehydrogenase and otherglycolytic control enzymes was also found to be altered (6) Studies on  nucleic acid metabolites in breast cancer cells showed that there occur DNA/RNA modifications  and there is an elevation in the amount of excreted modified nucleosides in breast cancer cells  as compared to normal cells. 26 of 36 metabolites identified with breast cancer cells, were      found to be modified ribonucleosides. Ribonucleosides and those compounds which have cis-diol structure were detected. Marked differences were found in 5-methyluridine, N2,N2,7-trimethylgua nosine,N6-methyl-N6-threonylcarbamoyl adenosine and 3-(3-amino carbox ypropyl)-uridine. 1-ribosyl-4-carboxamido-5-amino imidazole andS-adenosyl methionine occurredonly in supernatants of MCF-7 cells(7). Systematic methods were proposed for identifying metabolic markers inurine  samples of breast cancer patients and comparison was done between 50 breast cancer patients and 50 normal persons in one study. Nine metabolic pathways were found to be altered in the breast cancer patients. Four metabolic substances (Homovanillate, 4-hydroxyphenylacetate, 5-hydroxy indoleacetate and urea) were identified to be significantly different in cancer & normal subjects (8). In another study urine from breast cancer patients was analysed for metabolites by using multivariate methods and five potential urinary markers forbreast cancer could be identified with high accuracy(9).

Present study:

There are more than eighty identified metabolites in Table 1 that seem to be potential bio markers of cancer cell lines. Among these are six potential specific breast cancer cell line biomarkers. What makes them specific is their origin from the milk fats as can be inferred from the presense of pentadecanoic acid moiety or myristoleic acid moiety. Milk is produced only by the breast tissue the human body. Thus anything derived from milk fat can be regarded as specific for breast. The  following are the metabolites which can be regarded as the six potential specific biomarkers of breast cancer celllines

1. PE(14:1(9Z)/14:1(9Z)
2. PE(22:6(4Z,7Z,10Z,13Z,16Z,19Z)/14:1(9Z
3. PE(15:0/18:4(6Z,9Z,12Z,15Z))
4. PE(18:4(6Z,9Z,12Z,15Z)/15:0)
5. DG(14:1(9Z)/20:5(5Z,8Z,11Z,14Z,17Z)/0:0)
6. DG(20:5(5Z,8Z,11Z,14Z,17Z)/14:1(9Z)/0;0)

PE(14:1(9Z)/14:1(9Z)) (having two chains of myristoleic acid at the C-1 and C-2 positions) was fourteen fold more, PE (22:6(4Z,7Z,10Z,13Z,16Z,19Z)/14:1(9Z)) (having one chain of docosahexaenoic acid at the C-1 position and one chain of myristoleic acid at the C-2 position) was double PE (15:0/18:4(6Z,9Z,12Z,15Z)) (having one chain of pentadecanoic acid at the C-1 position and one chain of stearidonic acid at the C-2 position) was thirty seven fold  more PE (18:4(6Z,9Z,12Z,15Z)/15:0)) (having one chain of stearidonic acid at the C-1 position and one chain of pentadecanoic acid at the C-2 position) was three fold more in the breast cancer cellline MCF-7 than by non cancer cellline MCF-10A.

DG(14:1(9Z)/20:5(5Z,8Z,11Z,14Z,17Z)/0:0) containing myristoleicacid moiety derived from milk  fats was found to be 5 fold more in the AMM extracts of MDA-MB-231 CA cells

Suggested future work:

The findings of this study must be validated further by tandem massspectrometry. It should be  established whether the identified markers are general tumour markers or not and whether the six identified metabolites are the ideal potential specific biomarkers of breast cancer cell lines. It is important to verify whether the metabolites found elevated in this study involving breast cancer cell lines are also elevated in all patients who suffer from breastcancer. If so then tests should be devised to measure these biomarkers in body fluids of patients suffering from breast cancer and find out how the levels fluctuate at different stages of breast cancer so as to use these biomarker levels in body fluids for early detection of breast cancer. That  would help to initiate treatment early and thus prevent mortality due to this condition.


1. id=575 (office for national  statistics, 26 August  2010)
3. Early Detection of Second Breast Cancer Can Almost Double Survival-2009-03-17T08:00:00-04:00Crystal Phend
4. Degani.H,Katz-Brull.R,Margalit.R Choline metabolism in breast cancer 2H,13C,and 31P-NMRstudiesofcells and  tumoursMAGMA-Magnetic Resonance  materials inbiology,physics and  medicine6[1998]44-52
5. Patwardhan.A.J,Strittmatter.E.F,Smith.D.R Quantitative proteome analysis of breast cancer     cell lines using 18O labelling and an accurate masss and time tag strategy Proteomics [2006]2903-2915
6. Lynn M Knowles1 and Jeffrey W Smith;Genome-wide changes accompanying knockdown of fatty acid synth ase  in breast cancerJ.Biomed Biotechnol [2006]
7. Bullinger.D,Fehm.T,Laufer.S, Kammerer.B Metabolic  signature  of  breast  cancer  cell  line MCF 7;profiling  of  modified nucleosides  via LC-IT  MS  coupling BMC Biochemistry [2007]
8. Nam.H, Chul .C.B, Kim.Y, Lee.K,  Lee.D Combining tissue transcriptomics and urine metabolomics for breast cancer biomarker identificationBioinformatics25(23)-[2009]3151-3157
9.Kim.Y52 Multivariate Classification of Urine Metabolome Profiles  for  breast cancer  diagnosis CIKM [2009]
11. Oakman et al;Uncovering the metabolomic fingerprint of breast cancer Int J Biochem Cell Biol. 2010 May 10 ( 2010
12. Jeane Silva*, Somsankar Dasgupta*, Guanghu Wang*, Kannan Krishnamurthy*, EdmondRitter  and Erhard Bieberich1 Lipids isolated from bone induce the migration of human breast cancer cells
Journal of Lipid Research, Vol.47, 724-733, April 2006 (
13. Jennifer A. Cuthbert;Mevalonates, Ras and  Breast cancer  accessed from (

Source(s) of Funding


Competing Interests



This article has been downloaded from WebmedCentral. With our unique author driven post publication peer review, contents posted on this web portal do not undergo any prepublication peer or editorial review. It is completely the responsibility of the authors to ensure not only scientific and ethical standards of the manuscript but also its grammatical accuracy. Authors must ensure that they obtain all the necessary permissions before submitting any information that requires obtaining a consent or approval from a third party. Authors should also ensure not to submit any information which they do not have the copyright of or of which they have transferred the copyrights to a third party.
Contents on WebmedCentral are purely for biomedical researchers and scientists. They are not meant to cater to the needs of an individual patient. The web portal or any content(s) therein is neither designed to support, nor replace, the relationship that exists between a patient/site visitor and his/her physician. Your use of the WebmedCentral site and its contents is entirely at your own risk. We do not take any responsibility for any harm that you may suffer or inflict on a third person by following the contents of this website.

4 reviews posted so far

Biomarkers of Breast Cell Lines
Posted by Dr. William J Maloney on 12 Jun 2014 05:56:01 PM GMT Reviewed by Interested Peers

Thanks a lot for reviewing ... View more
Responded by Dr. Sumaira N Syed on 15 Mar 2013 11:24:45 AM GMT

1 comment posted so far

Specific Biomarkers for Breast Cancer Posted by Dr. Ekaterina V Moiseeva on 06 Mar 2013 10:50:24 AM GMT

Thanks for the comment I would be glad if you review the paper... View more
Responded by Dr. Sumaira N Syed on 14 Mar 2013 07:45:10 AM GMT

Please use this functionality to flag objectionable, inappropriate, inaccurate, and offensive content to WebmedCentral Team and the authors.


Author Comments
0 comments posted so far


What is article Popularity?

Article popularity is calculated by considering the scores: age of the article
Popularity = (P - 1) / (T + 2)^1.5
P : points is the sum of individual scores, which includes article Views, Downloads, Reviews, Comments and their weightage

Scores   Weightage
Views Points X 1
Download Points X 2
Comment Points X 5
Review Points X 10
Points= sum(Views Points + Download Points + Comment Points + Review Points)
T : time since submission in hours.
P is subtracted by 1 to negate submitter's vote.
Age factor is (time since submission in hours plus two) to the power of 1.5.factor.

How Article Quality Works?

For each article Authors/Readers, Reviewers and WMC Editors can review/rate the articles. These ratings are used to determine Feedback Scores.

In most cases, article receive ratings in the range of 0 to 10. We calculate average of all the ratings and consider it as article quality.

Quality=Average(Authors/Readers Ratings + Reviewers Ratings + WMC Editor Ratings)