Academic Year 2015 - 2016

LIPID METABOLISM IN COLORECTAL CANCER: ALTERATIONS, THERAPEUTIC OPPORTUNITIES, AND OUTLOOK ON

MOLECULAR DIAGNOSTICS

Stijn DE KEUKELEIRE

Promotor: Prof. Dr. Lennart Martens

Dissertation presented in the 2nd Master year in the programme of

Master of Medicine in Medicine

FACULTY OF MEDICINE AND

HEALTH SCIENCES

My gratitude goes out to my girlfriend and family, whose support and personal interest have

been a major encouragement in writing this review. I would also like to thank my supervisor,

Prof. Lennart Martens, as for each time we had an appointment, he inspired me with his

knowledge and personal interest regarding this topic.

This thesis is dedicated to my mother: her fight against injustice goes on. May my work

inspire others to investigate and develop possibilities for battling this cruel illness.

1 ABSTRACT ............................................................................................................................................... 1

2 INTRODUCTION ....................................................................................................................................... 2

3 METHOD ................................................................................................................................................. 4

4 RESULTS .................................................................................................................................................. 5

4.1 COLORECTAL CANCER ............................................................................................................................... 5 4.1.1 Epidemiology ................................................................................................................................... 5 4.1.2 Prevention and screening ................................................................................................................ 5 4.1.3 Classification method ...................................................................................................................... 6 4.1.4 Treatment ........................................................................................................................................ 6

4.1.4.1 Non-metastatic colorectal cancer ................................................................................................................ 6 4.1.4.2 Metastatic colorectal cancer ........................................................................................................................ 6

4.2 METABOLOMICS IN COLORECTAL CANCER ................................................................................................ 7 4.2.1 Current and upcoming ‘omics’ approaches for biomarkers ........................................................... 7 4.2.2 Lipidomics and its analytical technology ........................................................................................ 8

4.3 LIPID METABOLISM AND COLORECTAL CANCER:........................................................................................ 9 4.3.1 Free Fatty Acids ............................................................................................................................ 10

4.3.1.1 Pathway .................................................................................................................................................... 10 4.3.1.2 Known alterations in colorectal cancer ..................................................................................................... 13 4.3.1.3 Alterations in free fatty acid synthesis pathways and distribution ............................................................ 14

4.3.1.3.1 Enzymes and other proteins ........................................................................................................... 14 4.3.1.3.2 Fatty acid distribution..................................................................................................................... 16

4.3.1.4 Alterations in eicosanoid metabolism and distribution ............................................................................. 18 4.3.1.4.1 COX-pathway ................................................................................................................................ 18 4.3.1.4.2 LOX-pathway ................................................................................................................................. 19 4.3.1.4.3 CYT P450 epoxygenase pathway ................................................................................................... 20

4.3.2 Glycerolipids .................................................................................................................................. 21 4.3.2.1 Pathway ............................................................................................................................................... 21 4.3.2.2 Glycerolipids and colorectal cancer ..................................................................................................... 25

4.3.2.2.1 Alterations in the glycerophosphate pathway ............................................................................... 25 4.3.2.2.2 Glycerophospholipids ..................................................................................................................... 26 4.3.2.2.3 Phospholipases ............................................................................................................................... 27 4.3.2.2.4 Triglycerids ..................................................................................................................................... 28

4.3.3 Sphingolipids ................................................................................................................................. 29 4.3.3.1 Pathway ............................................................................................................................................... 29 4.3.3.2 Sphingolipids in colorectal cancer ....................................................................................................... 31

4.3.4 Sterols ............................................................................................................................................ 35 4.3.4.1 Pathway ............................................................................................................................................... 35 4.3.4.2 Alterations in colorectal cancer ........................................................................................................... 36

4.3.5 Lipoproteins and lipid droplets ...................................................................................................... 39 4.4 AN INTERPRETATION OF PREVIOUSLY OBTAINED EXPERIMENTAL DATA .................................................. 41

5 DISCUSSION .......................................................................................................................................... 42

6 SUMMARY IN DUTCH ............................................................................................................................ 44

7 REFERENCES .......................................................................................................................................... 45

8 APPENDIX ............................................................................................................................................. 55

1

1 Abstract

Colorectal cancer (CRC) is one of the most prevalent and deadliest cancers worldwide, which

has prompted scientists to search of better screening methods and adequate therapies for this

disease. Lipidomics, the study of lipid profiles in cells, tissues and other organisms, is an

under-investigated research field, but has already showed promising results in many types of

cancers, including CRC. This literature review tries to comprehensively resume all alterations

in lipid metabolism that are linked with CRC. There are four major lipid groups to be

discussed: free fatty acids, glycerolipids (including triglycerides and

glycerophosphospholipids), sphingolipids and sterol lipids. The changing distribution of lipid

bodies and lipoproteins in CRC will also be mentioned. Firstly, free fatty acids in tumour

samples showed up-regulation of de novo synthesis-related enzymes (for instance ACLY and

FASN) and significant alterations in fatty acid distribution, shifting towards more desaturated,

pro-inflammatory ω-6 fatty acids. Regarding glycerolipids, triglycerides were mainly down-

regulated and phosphatidylcholine was more augmented in CRC tissue samples, while

lysophosphatidylcholines was shown up-regulated in plasma of CRC patients. Ceramide is the

major component in sphingolipid metabolism and due to its anti-proliferative, pro-apoptotic

properties CRC cells try to evade pathways that lead to the generation of this metabolite.

Cholesterol de novo synthesis and uptake from lipoproteins was enhanced in these neoplastic

cells, which was also highlighted by their higher expression of LDL and HDL cellular

receptors. Lastly, intracellular lipid droplet augmentation has already been associated with

malignancy and indeed was present in colon cancer cell lines. These findings indicate that

lipidomics could provide many opportunities in targeted therapy and prognostic or screening

biomarkers in CRC, although more intensive research is necessary if lipid metabolism is to

be integrated in our therapeutic and screening policy, not only for CRC but also other types of

cancers.

2

2 Introduction

Colorectal cancer still remains a major health topic: last year’s data show that the yearly

incidence amounts to almost 2 million people with more than 600 000 deaths throughout the

globe. These data vary in different countries, depending on age, sex, hereditary factors and

risk-factors concerning lifestyle-issues such as alcohol consumption, smoking, red meat

intake, obesity, diabetes and inflammatory bowel diseases (IBD). (1)

The current treatment strategy and prognosis are mainly determined by the clinicopathological

staging system TNM. Based on the Dukes staging system, it provides information concerning

the local tumor growth, lymph nodal invasion and metastasis of the colorectal neoplasm. (2)

Today, surgery still remains the main curative option for non-metastatic colorectal cancer,

with adjuvant or neo-adjuvant therapy (using radiotherapy, chemotherapy or radio-

chemotherapy) conserved for higher stages or high-risk patients. (1)

Metabolomics is a new research field, attempting to profile intra- and extracellular

metabolites (sugars, amino acids, nucleotides and lipids) to acquire new biomarkers for

diagnosis and prognosis, and to gain better insight in the fundamental steps involved in

carcinogenesis. Metabolomics has shown promising results in CRC staging and metastasis

detection and some minor results were achieved when pharmaco-metabolomics were used in

cancer treatment. (3) Lipidomics is a branch of metabolomics that examines the lipid

pathways and networks by quantitatively defining lipid profiles in organelles, cells and

organisms and has taken some huge steps forward regarding applications and further

analytical technologies. Mass spectrometry remains the cornerstone for investigating these

lipidomes. (4)

Lipids are complex but unique molecules with an irreplaceable role in cell structure

maintenance, energy provision, as signaling. (5) According to the classification by Fahy et al.-

, these lipid molecules are commonly divided into eight main groups: free fatty acids,

sphingolipids, glycerolipids, phosphoglycerolipids, sterols, prenols, saccharolipids and

polyketides. (6) The first five main groups contain bioactive (signaling) lipid molecules such

as fatty acids, diacylglycerol, ceramide, sphingosine, lysophosphatic acid that interfere in the

regulation or (de)activation of different signaling lipid pathways.

3

Prenol lipids, saccharolipids and polyketides however are mainly of bacterial, fungal and

plant origin and will therefore not be discussed in this thesis. (5)

There is substantial evidence that neoplasm cells show alterations in their lipid metabolism.

(7) These alterations can cause changes in cell membrane structures, energy homeostasis and

cell signaling with defects in gene expression, protein distribution and cell functioning,

including the processes of apoptosis, autophagy, necrosis, proliferation, differentiation,

growth, drug resistance and chemotherapy response. (5) (7)

By comprehending this phenomenon of lipid alteration, we could obtain vital information

about the pathogenesis in CRC, create new biomarkers for screening or diagnosing CRC in

earlier stages, and suggest new molecular targets for anticancer therapy. (8)

In this review, we will describe current used approaches for lipid profiling and detection of

lipid deformities in CRC with an eye to biomarker discovery and potential treatment options.

Most importantly, we will give a brief description of the metabolic pathways of each lipid

subgroup and its function in the colorectal tissue, and the abnormalities in these pathways that

could lead to the development of CRC.

4

3 Method

The related articles were acquired via the PubMed database by querying results with different

MeSH-terms. The most useful MeSH-terms were ‘colorectal neoplasms’, ‘neoplasm staging’

and ‘lipids/metabolism’.

The review by Huang et al. contains a division of those lipids in different categories: fatty

acids, sterols, sphingolipids, glycerolipids and glycerophospholipids (5) and these terms were

subsequently used for further investigation in combination with the above noted MeSH terms.

This method allowed enough reviews to be retrieved to obtain a sufficient background on

these topics.

However, a more profound analysis was necessary, so the investigation was expanded by

utilizing Google Scholar and Google as additional search engines. When a sufficient amount

of relevant articles were acquired, the exploration was continued by using specialized terms

in these search engines to deepen the investigation on certain topics such as ‘epoxygenases’,

‘lipoxygenases’, ‘cyclooxygenases’, ‘eicosonaids’, ‘glycerophosphate acyltransferase’,

‘sphingomyeline’, and ‘ceramide’ all coupled with ‘colorectal cancer’. Lastly, the snowball

effect method was used based on certain articles of interest, in which the references included

in these articles led to additional articles of interest. The results were filtered by restricting the

retrieved articles to those written in English.

5

4 Results

4.1 Colorectal cancer

4.1.1 Epidemiology

CRC is one of the most diagnosed cancers and causes the majority of cancer-related deaths in

the Western world, with a global incidence of two million people, more than half a million

deaths per year and a cumulative lifetime risk of developing CRC of approximately 5%.(9)

In recent years, more and more risk factors have been discovered including age (with a peak

incidence at 70 years), male sex, family history, obesity, diabetes, IBD, alcohol consumption,

tobacco use, red or processed meat consumption, low fruit and vegetable intake and even

intestinal bacterial infections (e.g. Helicobacter pylorum). (1) A high variability in incidence

exists between countries, though these incidence rates are highest in Western European

countries, Australia and New Zealand. This suggests that a Western life style may be a

significant risk factor in CRC development. It is important to note that the increasing

prevalence can also be due to improved screening procedures, which allow detection of more

asymptomatic patients, combined with the beneficial progression in treatment, which expands

the patients’ survival.(10)

4.1.2 Prevention and screening

Knowing that CRC has such a large impact on global health, prevention is an essential part of

this medical topic. Primary prevention is the most effective strategy in avoiding CRC and can

be maintained by reducing the adjustable risk factors and promoting a healthy lifestyle. (11)

Primary prevention by intake of certain drugs such as NSAID or hormone therapies however,

has important side effects and still lacks fundamental evidence.

Secondary prevention comprises early detection of CRC, which can be obtained by screening

with faecal occult blood tests, colonoscopy or sigmoidoscopy and even virtual colonoscopy

by using CT imaging. This latter technique has been proposed as a screening technique,

though exposion to radiation remains an important consequence, while the cost-effectiveness

of this procedure is doubted, especially for primary prevention.

Options for tertiary prevention of surviving CRC patients have not been fully explored yet.

Physical exercises, cessation of smoking and long term intake of aspirin or NSAIDs have

been recommended, but more investigation remains necessary. (10)

6

4.1.3 Classification method

Officially, there are two classification methods that are utilized for staging CRC: Dukes

stadium and the TNM system. Biopsy samples and imaging techniques

(ultrasonography,(PET-) CT, MRI) are necessary for determining which stage belongs to a

specific patient. Because the Dukes staging classification received the connotation of being

out-dated and confusing, tumors are expected to be categorized according to the latest version

of the UICC TNM-system since 2003. (2)(12) Table 1 (see appendix) provides an overview of

the Dukes and TNM classification systems, and describes the different stages according to the

findings of imaging techniques and histopathological investigation. Categorizing the stage of

the patient’s CRC is important for determining the prognosis and type of therapy this patient

should receive. (11)

4.1.4 Treatment

4.1.4.1 Non-metastatic colorectal cancer

As mentioned above, surgery remains the imperative curative option in non-metastatic CRC.

Colon surgery is performed by removing the tumor and its corresponding lymph nodes with a

clear resection margin. Treatment of rectal cancer is executed by total mesorectal excision.

These procedures were initially performed using open surgery, but laparoscopy has proven to

be equally efficient with a lower mortality, morbidity and shorter hospitalization, although it

is technically more difficult for the surgeon.

Locoregional recurrence is a risk factor after surgery and is associated with a poorer

prognosis. Due to the anatomical complexity of the pelvis, rectal cancer has a higher risk of

recurrence compared with colon cancer. Avoiding recurrence can be accomplished by

resection using wide margins and, if necessary, combined with pre-operative and/or post-

operative chemotherapy, radiotherapy or chemo-radiotherapy. Adjuvant chemotherapy can

also be advised when there is an increased risk for distant metastasis after resection. The

decision wheter surgery with or without (neo-) adjuvant therapy should be applied depends on

the CRC stage, together with the patient’s own preference. (1) (8) (11)

4.1.4.2 Metastatic colorectal cancer

The treatment policy for CRC patients with distant metastasis varies from case to case.

Surgery of liver or lung metastases can be offered if these metastases seem resectable, but

when irresectable, palliative chemotherapy should be the standard. Alongside traditional

7

chemotherapeutics, more and more new drugs can be utilized as a form of personalized

therapy. Examples are Bevacizumab and Aflibercept, targeting VEGF (vascular endothelial

growth factor) or Cetixumab and Panitumumab, targeting EPGF (epidermal growth factor).

(1)(11)

4.2 Metabolomics in colorectal cancer

4.2.1 Current and upcoming ‘omics’ approaches for biomarkers

Over the past years, genomics and proteomics have been the center of investigation for

carcinogenesis. This has provided an abundance of information that led to the development of

new insights, new therapeutic targets and strategies, and nucleotide- or protein-based

biomarkers. (13) Some of the most prevalently used and investigated biomarkers are

described in this section.

Micro-satellite Instability (MSI), a short repetitive DNA nucleotide sequence involved in

DNA repair after replication, is currently one of the most used biomarkers. The MSI status is

divided in 3 different categories: MSI high (>30%), MSI low (10-30%) or MSS

(microsatellite stable). It still remains unclear if the MSI status can be interpreted as a positive

or negative factor for the prognosis. Studies have showed that the MSI status is associated

with some clinicopathological variables and patients with a high MSI status have a higher

general five year survival rate. (8) (14)

CIMP or hypermethylation of the CpG promoter island, is a well-known transcriptional

silencer of DNA repair genes and tumor suppressor genes. It is often correlated with MSI and

CIN and CRCs are therefore increasingly divided according to their CIMP status: CIMP-high,

CIMP-low and CIMP-negative. (8)

CIN or Chromosomal instability is a characteristic found in almost 60-80% of all CRCs. It

depicts the chromosome alterations, either structural or numerical, in CRC cells and is

associated with poor prognosis and moderately differentiated types of cancer cells. (8)(12)

KRAS is a proto-oncogen that, when mutated, continuously activates the MAPK pathway

and PI3K/Akt, both of which promote carcinogenesis by cellular proliferation. (8) It is present

in 30-50% of all CRCs but still hasn’t been considered as an essential predictive biomarker.

(14)

8

BRAF is an inhibitor of the RAS/MAPK intracellular signaling pathway, encoding a

serine/threonine kinase. Mutations of this gene are correlated with the early stages of

colorectal cancer development. Moreover, KRAS and BRAF mutations have both been

associated with poor response on EPGF-inhibitors. (12) (14)

VEGF, vascular endothelial growth factor, is a pro-angiogenic factor in the development of

CRC and has been proposed as a molecular biomarker for lymph node metastasis. (8)

LOH or Loss Of Heterozygosity, especially located in chromosome 18q, is able to inactivate

tumor-suppressor genes and has a certain prognostic value, though this potential biomarker

needs further investigation. (8)

Interestingly there is also a metabolomics based biomarker under development: a prediction

model using the metabolites 2-hydroxy butyrate, aspartic acid, kynurenine and cystamine has

already been tested and proposed as an early detection method for CRC stage 1 and 2. (15)

4.2.2 Lipidomics and its analytical technology

In recent years, lipidomics has been introduced as a new chapter in metabolomics,

representing the analysis of lipid metabolism in organelles, cells, tissues and organisms, in

health and disease. (6) (13) Lipids embody essential roles in human physiological and

pathological processes such as apoptosis, proliferation and carcinogenesis, and although they

are extremely complex, more and more scientists are investing time in the development of

new techniques for separation and identification of these lipid molecules. (16)

The first step in lipid analysis comprises extraction of lipid molecules. This can be

accomplished by using a mixture of methanol, chloroform and water (the Blight and Dyer-

method). Many more techniques for lipid extraction from samples are available but will not be

further discussed here. (16)

Due to its high reproducibility, selectivity and sensitivity for detecting different lipid

components in body fluids or tissue samples, mass spectrometry (MS) is the most widespread

standard procedure for identification of lipid molecules. MS can be combined with separation

techniques such as capillary electrophoresis (CE), gas chromatography (GC) or in most cases,

high performance liquid chromatography (HP-LC). (3)(6)

9

HP-LC separates the different lipid components according to their chemical or physical

properties by introducing the lipids/liquid to a high pressurized monolithic silica tube or

column, called the stationary phase. A compound solvent, dubbed the mobile phase, is then

pushed over the column at high pressure, and this compound solvent is differentially mixed

from hydrophilic and hydrophobic primary solvents. Lipids first adhere to the column, but

dissolve into the compound solvent mixture and thus elute from the column when the

hydrophobicity is sufficiently high. This system thus effectively allows the separation of

lipids with different physiochemical properties over time.

Once the separated components have left the column, these lipids are in the MS instrument.

Initially, ionization of the lipids is performed, usually via electrospay ionization (ESI), which

introduces a charge on the lipid molecules, creating lipid ions. Next, separation according to

the mass-over-charge ratio (m/z) is performed by bringing the ionized molecules into a mass

analyzer, typically the rapidly oscillating magnetic field of a quadrupole for lipids. The m/z-

separated ionized lipid molecules then impact an electron multiplier that serves as a detector,

allowing a spectrum to be created of the number of ion impacts per m/z.. (17)

4.3 Lipid metabolism and colorectal cancer:

Lipids are apolar, hydrophobic molecules with an essential role in the thermal isolation of the

human body and the emulsification of food in the intestines using bile acids. Most importantly

for this thesis, they are also an alternative energy supply for cells, provide structural integrity

for cellular membranes, and are essential signaling components, be it as second messengers or

as hormones.

Cellular lipid metabolism contains complex processes and includes uptake, synthesis,

degradation and transport. Every lipid can be regulated by different pathways in different

tissues or cells, and activated or inhibited by physiological, pathological or therapeutic

circumstances.

Bioactive lipids transduct signals to regulate lipid metabolism using different pathways:

o G-protein coupled receptors

o Tyrosine kinases

o Integrin signaling

o Ion channel signaling

o pH changes

o oxidative stress

10

o mechanical stress

Cancer cells gain defects or alterations in their lipid metabolism, causing changes different

cellular processes such as proliferation, differentiation, motility and growth. (5) This could be

induced by the over-expression of essential lipid enzymes, altering the cell’s functions and

enhancing lipid biosynthesis. Excessive lipogenic enzyme function is correlated with

enhanced transcription of lipogenic genes, for instance by the transcription-regulator SREBP

(sterol regulating element-binding protein) by growth factors or by growth signaling

pathways. Epidermal growth factor, keratinocyte growth factor, erBb-receptors and Her2/neu

receptor tyrosine kinase have already been identified, while the PI3K/Akt pathway is

activated in multiple types of cancer and plays a central role in the expression of lipogenic

enzymes. (18)

According to the Lipid Classification and Nomenclature Committee, lipids are divided in

eight main groups: free fatty acids, glycerolipids, glycerophospholipids, sphingolipids, sterol

lipids, prenol lipids, saccharolipids and polyketides. (5) However, in this thesis, a slight

modification in this classification method will be applied. Because mono- ,di- and

triacylglyceroles as well as glycerophospholipids all originate from a glycerol-backbone and

undergo a common biosynthetic pathway (until phosphatic acid). ‘Glycerolipids’ and

‘glycerophospholipids’ will therefore be unified as one term in this thesis: glycerolipids.

In addition, lipoproteins could also be included as unique lipid entities. (19) These different

lipid classes will be discussed in more detail in the subsequent sections, with a specific focus

on their pathway, and any alterations thereof in CRC.

4.3.1 Free Fatty Acids

4.3.1.1 Pathway

The FA biosynthesis usually begins with the provision of acetyl-groups by citrate, generated

in the Krebs-cycle. Subsequently, citrate is converted into acetyl-CoA and oxaleacetate by

ATP citrate-lyase (ACLY). Acetyl-CoA carboxylase (ACC) is the rate-limiting step of fatty

acid-biosynthesis, converting acetyl-CoA into malonyl-CoA. Acetyl and malonyl groups are

attached together by fatty acid synthase (FASN). Eventually this generates the substrate

palmitic acid (16 C) due to repeated condensations of acetyl-groups. Desaturation is an

optional step in the fatty acid biosynthesis, taking place in the endoplasmatic reticulum and

executed by SCD (stearoyl-CoA desaturase). This introduces a double bond in the saturated

11

fatty acids, making it a mono-unsaturated fatty acid. Palmitic acid is the main saturated fatty

acid and can be converted into other saturated or unsaturated fatty acids such as stearic acids,

oleic acid and more. (20)

It is important to notice that the pathway described above represents the production of non-

essential fatty acids only. The essential fatty acids cannot be produced by humans and other

mammals, and can only be obtained through the diet. (21)

The eicosanoids are derived from 20-carbon essential fatty acids: prostanoids, leukotrienes

and epoxygenases are produced from ω-6 arachidonic acid (AA), while resolvins and

protectins are obtained from ω-3 PUFA (poly-unsaturated fatty acids), including

eicosapentaenoic (EPA) and docosahexaenoic acid (DHA). (22)

AA is an essential fatty acid and a key metabolite for the production of eicosanoids via the

COX, LOX and CYT P450 (EPOX) pathways. The grand majority of eicosanoids acquire

their AA from membrane glycerophospholipids by activation of PLA2, which releases the

substrate at the sn2-position of the glycerol-backbone. Another fraction of the AA is obtained

via linoleic acid (LA): firstly, γ-linoleic-CoA (GLA) can be obtained by diet or by the slow

reaction mechanism of LA-CoA with a ∆6-desaturase. When going through the microsomal

elongation pathway, GLA-CoA is exchanged for dihomo-γ-linoleoyl-CoA (DGLA) which is

converted into AA by a ∆5-desaturase. (23)

AA can eventually be metabolized through three different pathways for generating

eicosanoids: cyclooxygenase (COX)-pathway providing prostanoids, lipoxygenases,

leukotrienes and P450 epoxygenase supplying hydroxyeicosatetraenoic acid (HETE),

epoxyeicosatetraenoic acid (ETE) and hydroperoxyeicosatetraenoic (HPETE). (22) An

overview of the eicosanoid pathways and its products can be seen in the figure below.

12

Fig. 1: overview of arachidonic acid metabolism

13

As mentioned above, a small portion of the eicosanoids are derived from the ω-3 PUFA, EPA

and DHA, obtained by the α-linolenic acid-metabolism or by diet, and may as well go through

the three oxidative pathways in the same way as arachidonic acid. This creates variant series

of (benign) prostaglandins, leukotriens, resolvins, protectins and other bioactive lipid

mediators as depicted in the figure below. The red marked compounds represent moderately

pro-inflammatory products while the green marked compounds represent anti-inflammatory

and anti-tumorigenic products. (24)

Fig.2: ω-3 fatty acids metabolism

4.3.1.2 Known alterations in colorectal cancer

When manufacturing new structural components, healthy cells commonly rely on dietary

circulating fatty acids or exogenous intake of carbons, which can be converted into FA’s by

liver or adipose tissue. With the exception of some metabolically active tissues such as fetal

lung tissue, hepatocytes, adipocytes, lactating breasts and the cycling endometrium, de novo

lipogenesis-pathway is reduced in most healthy cells. (18) (25)

Cancer cells require sufficient molecular building blocks as a response for their enhanced

proliferation and growth capabilities. (20) In these cases, ‘de novo’ synthesis of FA’s is

essential for building lipid membranes and signaling molecules, although some tumors are

able to capture FA’s from their environment (e.g. using FABP4). Enhanced influx of carbons,

introduced as glucose or glutamine, supports the anabolic pathways for building these new

fatty acids. Consequently, blocking FA de novo synthesis by limiting its supplies, enhancing

14

FA degradation, pushing FA’s into storage, and/or downsizing FA release from storage could

offer new therapeutic strategies in cancer therapy. (18) (26) (27)

4.3.1.3 Alterations in free fatty acid synthesis pathways and distribution

4.3.1.3.1 Enzymes and other proteins

The simplest way to obstruct synthesis is the down-regulation of the conversion of

carbohydrates to citrate in the Krebs cycle, blocking the mitochondrial citrate carrier (CIC or

protein SLC25A1) or inhibiting the enzymes that take part in the enhanced de novo

lipogenesis in neoplastic cells. (20) The mitochondrial citrate carrier is an essential transporter

in de novo FA synthesis: citrate exits the mitochondria and is cleaved into oxaloacetate and

two acetyl-CoA molecules. Some studies reported that these citrate carriers are up-regulated

in colon cancer. (28)

ACLY is a homotetrameric cytosolic enzyme that converts mitochondrial citrate into acetyl-

CoA, and thus may provide essential substrates for the FA synthesis pathway and the

mevalonate-pathway. A significant increase in ACLY expression and activity has been shown

in colonic cancer cells. (25) Another study noted that ACLY was up-regulated in CRC

compared to normal tissue, as well in chemo-resistant CRC cells compared to chemo-naive

variants. In addition, activation of the PI3K/Akt-pathway, a signal-transduction pathway that

promotes cell survival and growth as a response to extracellular factors, is correlated with

ACLY activity: enhanced glucose uptake and metabolism leads to upregulated PI3/Akt which

phosphorylates and triggers ACLY, increasing FA synthesis and being partially responsible

for carcinogenesis. (29) (30)

These findings suggest that inhibition of ALCY may be introduced as a new therapy for

targeting cancer cells. RNAi (RNA interference), or pharmacological inhibitors could be

functional for down-regulating ALCY, resulting in a growth arrest of those tumor cells. (25)

Experiments with the citrate analogues (+) and (-) 2,2-difluorocitrate, on rat livers resulted in

a reduction of ACLY-activity. This outcome was also reached with (-) hydroxycitrate in

HepG2cells (a cell-line of polar human hepatocytes), although a restricted membrane

transport and necessity of high concentrations of this citrate analogue were profound

limitations. Radicicol, an antifungal macrolide, also inhibits the enzyme in rat liver cells,

whereas SB-204990 restricts proliferation in lung cancer cells. Also, combining the ALCY

15

inhibitors with a statin seems to have profound anti-tumorigenic effects. These findings are

still somewhat controversial and further investigation is desirable. (25)

As in many cancers, fatty acid synthase (FASN) is over-expressed in CRC, with the

metastatic variant taking account for the highest expression and correlating with a poor

prognosis. The theory behind this relationship remains unclear. (31) FASN knockdown

neutralizes CRC cells by restraining the cells’ energy metabolism and with it, the mTOR-

pathway, a serine/threonine kinase member of the PI3K pathway that enhances expression of

different lipogenic enzymes . (32)

It seems that the clinical stage of CRC is significantly correlated with the serum FASN of the

patient: concentrations were higher in stage 3 and 4 than in stage 1 and 2. (33) Another study

suggested that the FASN serum level is associated with the TNM-stage: patients with a tumor

extent T1 and T2 without lymph node metastasis nor distant metastasis have a significantly

lower FASN serum than patients with a T3 or T4 tumor extent and lymph node metastasis or

distant metastasis. Furthermore, the disease-free interval and the five-year overall survival

rate was smaller for patients with an elevated serum FASN level. This suggests that FASN

could be a potential biomarker, informing us about the clinical state of the tumor and patient

prognosis. (34)

FASN could also be a potential anti-angiogenetic target: knockdown of this enzyme decreases

the microvascular density and inhibits release of VEGF-A. This stabilization in secretion of

anti- and pro-angiogenetic factors, could inhibit the tumor vascularisation and by

consequence, reduce the probability of metastasis. (31) Inhibitors of FASN, cerulenin, orlistat

and C75, induce apoptosis in several cancer cell lines, including colorectal cancer, and could

be a new effective treatment strategy. (35)

Acetyl-CoA carboxyl (ACC) has two subtypes in the human body, both having different

metabolic functions: ACCα and ACCβ. (20) TOFA(5-tetradecyloxy-2-furoic acid), an

allosteric inhibitor of ACCα and a cytotoxic agent, disturbs de novo lipogenesis, induces

apoptosis in cells, and thus may be of use as a therapeutic in colon cancer. (36)

Stearoyl COA-desaturase (SCD) has a clear role in the lipid metabolism and growth pathways

of cancer cells: by increasing lipogenesis through inhibition of FA oxidation and alteration of

several intracellular pathways, activation of Akt-pathway and deactivation of AMPK-

16

pathways, it creates a benevolent environment for cell survival and proliferation. Inhibition is

known to cause apoptosis in many cancer types, including colonic tissue, and could be a

promising new therapeutic target in cancer interventions. (20) Although a higher expression

and activity of SCD1 were found in some colonic carcinoma tissues, there is still insufficient

evidence regarding this enzyme in CRC. (37) One study measured the SCD1-activity by

calculating the 16:1 n-7/16:0 ratio, which showed a lower mean activity in colorectal cancer

than in normal mucosa. The SCD2 activity on the other hand, measured by the 18:1 n-9/18:0

ratio, showed an augmentation of this enzyme in CRC tissue. (38)

Despite the proof that blockage of lipogenic enzymes could reduce tumor growth in colorectal

cancer cells in vitro, more research has to be performed to examine if this could be a long

term strategy in CRC treatment in vivo, especially because more evidence has been revealed

that an alternative mitochondrial pathway for de novo lipogenesis in mammalian cells exists.

(39)

The CPT (carnitine palmitoyl transferase) 1-transporter, transferring the long FA chains into

the mitochondria where β-oxidation takes place, seems to be significantly decreased in

colorectal and breast carcinoma. (40) CPT-1 participates during the metabolic transformation

(solid) cancer cells undergo when faced with metabolic stress. Continuous tumor growth

deprives cancer cells from their oxygen and nutrient (glucose) supply, forcing them to search

for an alternative, anaerobic form of energy. Apparently, cancer cells up-regulate their

CPT1C-gene, leading to promoted FFA-oxidation and generation of ATP as energy source.

Indeed, experiments on colorectal cancer xenografts showed a decrease in tumor growth when

CPT1C was depleted. (41)

4.3.1.3.2 Fatty acid distribution

FA distribution of healthy and CRC tissues illustrated that there is a definite correlation

between the colorectal cancer stage and FA profiles. Both samples contained non-essential

palmitic acid, stearic acid, oleic acid, and the essential linolic acid as the most profuse FA.

(38)

CRC tissue showed a shift in FA distribution with a higher rate of saturated FA. While a 50%

increase of stearic acid was noted , the amount of myristic and palmytic acid remained

relatively stable in CRC. On the contrary, the mono-unsaturated FA palmitoleic and oleic acid

were respectively 50 and 20% lower in CRC than in normal mucosa. (38)

17

The ratio in ω-6/ ω-3 PUFA was increased in CRC with the total ω-3 PUFA being decreased

compared to normal samples.(21) Dihomo-γ linolic acid (DGLA) and AA showed an

augmentation of respectively 34% and 38%. (38) One study proved that linolic acid levels

were downsized: this finding combined with the elevated values of AA could indicate that

there is a higher turnover into this pro-inflammatory product. (21) This hypothesis was

contradicted by another study, where linolic acid levels were slightly elevated together with

arachidonic acid. Furthermore, AA/LA, AA/DGLA, and AA/ -total ω-6 PUFA ratios were

higher, thus indicating that desaturase-enzymes may be up-regulated in CRC.(38)

The concentrations of EPA and DHA, described as percentages of total FA concentration,

were respectively 88% and 37% decreased in CRC samples compared to healthy samples.

These two products are anti-inflammatory and decrease the levels of colorectal biomarkers.

(38)

The relationship between FA distribution and clinicopathological stages seems to remain

unclear. When comparing the saturated FA and mono unsaturated fatty acids (MUFA), there

were no significant differences between colorectal tissues with Duke B and C stages. Only

palmitoyl acid and 18:1 ω-9 were almost half the amount in the C-stadium than in the B-

stadium. Concerning the PUFA, the ratio ω-6/ ω-3 was strongly increased in the Dukes C

stadium compared with the B stadium. AA levels in stage C were six times higher than stage

B, but the linoleic acid level was reduced by almost 66% in stage C compared with stage B.

Another study observed no differences in SFA, MUFA and PUFA between the

clinicopathological stages (I, II and III-IV) or lymph node metastasis.(21).

Ω-3 fatty acids (EPA and DHA) suppress AA-derived eicosanoid production by replacing the

membrane bound-AA by ω-3 PUFA, competing with desaturases and elongases and inhibiting

formation of LA to AA. Altogether this makes AA less available and consequently,

suppresses AA-derived eicosanoid production. (42) These ω-3 PUFA replace pro-

inflammatory substrates in the COX-2-pathway while EPA acts as the preferred substrate for

the LOX-pathway, blocking pro-angiogenetic and inflammatory prostaglandins or

leukotrienes. (28)(42) In addition, EPA and DHA seem to have a protective effect against

CRC by enhancing the caspase-dependent apoptosis pathway through down-regulation of

two regulatory elements, FLIP and XIAP. Another encouraging finding is that EPA and DHA

do not trigger apoptosis in healthy colorectal cells, which could make these potential and

effective adjuvants with other chemotherapeutic agents. (44)

18

These results suggest that the ω-6 PUFA, AA in particular, promote carcinogenesis in

colorectal tissue and that ω-3 PUFA, DHA and EPA, attempt to restrain it.

4.3.1.4 Alterations in eicosanoid metabolism and distribution

4.3.1.4.1 COX-pathway

There are 3 variants of COX-enzymes: COX-1, COX-2 and COX-3. COX-1 is a house-keeper

enzyme of gastric mucosa, renal bloodflow and platelet activation that remains stable under

physiological and pathological circumstances. COX-2 has an unmistakable role in

inflammation and other pathophysiological processes. COX-3 is a variant of COX-2 with a

currently unclear function.

It is well well-known that COX-2 expression is augmented in CRC: 50% of adenoma and

85% of adenocarcinoma had elevated values. (45)

The prostaglandin PGE2 seems to be a key factor in colorectal tumorigenesis, promoting cell

growth, angiogenesis (via VEGF) and tumor survival. (31) PGE2 was strongly up-regulated in

metastatic tumors, thus enhancing invasion and urine samples with elevated PGE2 values

gave an increased risk in colorectal cancer development. (22) (46) It is the most dominant

prostaglandin present in CRC tissue. (45) Lung, head and neck and breast cancer also had an

elevated quantity of PGE2 and were all associated with poor prognosis. (28,29) PGE2

prostaglandins are degraded by 15-PGDH, which if knocked out in murine subject animals,

induced a 7.6 times increase in risk for developing colon tumor with a doubling of PGE2

values. Transgenic mice deficient in prostaglandin target receptors EP (1-4), display a

significant reduction in colon cancer incidence. (22)

Prostacycline- or PGI2-values appear to be decreased in CRC tissue, meaning that this

prostaglandin may not be essential in colorectal carcinogenesis, though a study showed that

addition of PGI2 analogues had a protective effect against metastasis. (22) It may have a role

in tumor progression due to the fact that it activates PPARδ, thus accelerating intestinal tumor

growth in mice subjects. Further investigation will hopefully clarify the specific role of PGI2.

(45)

Although prostanoid receptors DP2 gradually stop being expressed during the adenoma-

carcinoma sequence, PGD2s function still remains unclear. Inhibition of CRC cell

proliferation has been observed when treated with PGD2, but its metabolites (PGJ2 and

15dPGJ2) induced cancer cell proliferation and survival. (22) One proposed hypothesis is that

15dPGJ2 may inhibit tumor cell growth through binding of the PPARγ-receptor. (47) Another

19

theory is that over-expression of PGD-synthase could lead to a metabolic shift in which less

PGE2 is produced and PGH2 is predominantly converted into PGD2, suppressing tumor

growth. More investigation about PGD2 and its effects in CRC and other cancer cells seems

necesarry. (45)

Prostaglandin F2α has no known role in carcinogenesis of CRC and will not be further

discussed. (48)

TxA2 is derived from the COX-1 and COX-2 pathway and contributes to tumor growth and

angiogenesis promotion. (45) Targeting COX-1 with aspirin proved to have an anti-

carcinogenic effect. Furthermore, a direct addition of TxA2 after deletion of tromboxane

synthase erased cell arrest in CRC and promoted proliferation. The tromboxane synthase

seems to gain some expression in CRC tissue compared to normal mucosa, with TxB2

metabolites and their urinary excretion being significantly raised in these patients. (22) Also,

a TxA2 synthase-inhibitor was shown to be able to block liver metastasis in CRC. (45)

NSAIDs have proven their effects in blocking inflammation, but could also offer anti-

carcinogenic effects thus reducing the risk of colon cancer and other prevalent solid tumors

(breast, prostate, lung) via COX-2 inhibition. (47) A quantity of studies confirmed that intake

of aspirin or other NSAID on a regular basis during a 10-15years period gives a relative risk

reduction of 40-50% for the development of colorectal adenoma. (45) As mentioned before,

the question thus arises if NSAIDs or aspirins could be part of the colorectal cancer

prevention strategy, albeit in combination with proper exercise, healthy diet and avoidance of

other carcinogenic risk factors. (49)

4.3.1.4.2 LOX-pathway

CRC patients often gain enhanced activity of 5-LOX and 12-LOX. With 5-LOX being over-

expressed in human tubular adenoma, villous adenoma and colorectal adenocarcinoma, it may

be associated with chronic inflammation and carcinogenesis. 12-LOX over-expression could

be more of an incentive for metastasis, while also participating in angiogenesis and cancer

cell proliferation. (42) Eicosanoid profiling illustrated that not only PGE2 and AA, but also

12-HETE was significantly altered in CRC tissue compared to normal mucosa, though

without any correlation with Dukes stadium for this metabolite. (48) Even though 15-LOX-2

didn’t show any significant alterations in expression, a 125-patient prospective study found

down-regulated 13-S-HODE (a 15-LOX-1 product from linoleic acid) concentrations when

going through the mucosa-adenoma-carcinoma sequence of the colon. (22) Indeed, 13-HODE

20

showed significantly augmented values in normal colorectal mucosa compared to colorectal

adenoma. (50) 13-S-HODE has a tumor-suppressive role, inducing cell cycle arrest in cancer

cells and restoring apoptosis via PPAR-δ activation. (51) Other LOX products effects are also

induced via the superficial G-protein cell receptors or activation of the PPAR family. (22)

Both COX-2 and LOX-5 have pro-carcinogenic effects and blocking these pathways could

enhance a shift of free AA to another, producing more pro-tumorigenic metabolites. Dual

blockade of both enzymes could provide as a new chemo-preventive strategy: it has already

been found safe and effective as treatment in osteoporosis. (42)

An inverse relation between 15-LOX-1 and COX-2 has also been noted: this was confirmed

in a study where metabolism of LA shifted towards the COX-2 and away from the 15-LOX-1

pathway when progressing through the adenoma-carcinoma sequence: while 96% of low-

grade adenomas expressed the 15-LOX-1 gene, this was only the case for 43% of carcinoma-

in-adenoma lesion subjects. Regarding COX-2, 2% of low-grade adenoma expressed the

gene, whereas this was 71% for the carcinoma-in-adenoma and even 92% of the advanced

carcinomas. (52) Furthermore, products that up-regulate 15-LOX-1 and down-regulate COX-

2, such as sulforaphane or Honokiol, inhibit intestinal polyp formation and gastric

carcinogenesis, respectively. Development of new chemotherapeutic agents and using these

metabolites as molecular biomarkers could become valuable in colorectal cancer screening,

prevention or treatment. (46)

4.3.1.4.3 CYT P450 epoxygenase pathway

The explicit role of this pathway in cancer development has not been cleared up yet, though

CYP2J2 and other CYP enzymes seem to be over-expressed in a variety of cancer cell lines,

including colon cancer.(42) It is supposed that that this pathway metabolizes free AA,

inhibiting ceramide production and apoptosis (see 4.3.3.2), but also gaining 14-,15-EET,

inhibitors of apoptosis via the PI3K-Akt pathway. (53) Though there is little known about the

role of EETs in CRC progression, patients with these types of solid tumors have elevated

levels of these compounds in blood and urine. Treatment of endothelial cells with 14-,15-EET

induced tumor growth and metastasis promotion. On the other hand, inhibitors of

epoxygenases or EET-antagonists were capable of containing tumor development in some

cancer types in rat subjects (for example glioblastoma), thus prolonging the animals’ survival.

(54) Whether these findings could be similar for CRC in humans needs to be further

investigated.

21

As mentioned, DHA is in competition with AA for being processed along the epoxygenase

pathway. This generates epoxydocosapentaenoic acids (EDPs), mediators with anti-

angiogenic and anti-tumorigenic effects. In mice, administration of 19,20-EDP blocked

VEGF- and FGF2 –induced angiogenesis, restrained primary tumor growth by 50-90% and

gave a 70% reduction of lung metastasis foci and weight. The question arises if DHA or other

ω-3 PUFA can be administered to CRC patients with high epoxygenase expression profiles to

reduce EET levels and increase EDP levels in patients, thus contributing to a better prognosis.

(24)

4.3.2 Glycerolipids

4.3.2.1 Pathway

The triglycerides (TG) and glycerophospholipids are glycerolipids that share a common

biosynthetic pathway: the glycerophosphate pathway. (5) Both products can be obtained via

the intermediate phosphatidic acid (PA), though all products have a completely differing FA

composition. (55)

TG are the first type of glycerolipids, consisting of a glycerol backbone and three esterified

FA’s, all different in length and saturation level. TG can be stored in lipoproteins (e.g. LDL,

HDL chylomicrons…) as well as in lipid droplets, present within every cell-type, but most

commonly found in adipocytes. (55)(56)

Glycerophospholipids are also built of a glycerol backbone with two esterified FA at the sn1-

and sn2-position but with a polar head group attached to position sn3. These are the cell

membrane’s most abundant building blocks, for which the composition varies from each type

of cell, organelle, inner or outer membrane. Each composition in turn shows dissimilarities in

cellular functions such as vesicular transport, membrane viscosity and signal transduction.

Due to their FA remodeling system (see further), glycerophospholipids are an extremely

important source of lipid mediators such as lysophospholipids and saturated or unsaturated

FA’s, which can be converted into eicosanoids and other lipids. (57)

The glycerophospholipids are classified according to the structure of their polar head groups.

(58) Phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS),

phosphatidylinositol (PI), phosphatidylglycerol (PG) and diphosphatidylglycerol (DPG) or

22

cardiolipin are the major classes. These phospholipids can be further subclassified according

to the type of chemical bond with the FA at the sn1-position: diacyl, alkylacyl and

alkenylacyl. (55) (56) The heterogenity of glycerophospholipids is further increased due to

the different FA-groups at the sn1 and sn2-position: while saturated FA’s are more common

on the sn1-position, unsaturated FA’s are mainly present on the sn2-position. (55)

In tissues other than adipocytes, synthesis of PA commences with the glycerol-backbone,

which is activated by phosphorylation at position C3. The glycerol-3-phosphate is

incorporated with a FA-CoA via the Glycerol-3-Phosphate acyltransferase (GPAT),

converting it into lysophosphatic acid (LPA). This is followed by 1-acylglycerol-3-phosphate

acyltransferase (AGPAT), creating phosphatic acid (PA), the key intermediate in triglyceride

and glycerophospholipid synthesis. This metabolite can also be produced via a different

pathway that is mainly used in adipocytes, which are lacking the glycerol-kinase enzyme. (56)

Consequently, 1-alkyl LPA is gained through use of dihydroxyaceton-phosphate (DHAP), a

metabolite produced during glycolysis, by acylation with DHAP-acyltransferase (DHAPAT)

and reduction by DHAP oxido-reductase. Eventually, AGPAT becomes involved in the

process and leads to the formation of 1-alkyl PA. (55) Lastly, PA can also be obtained

through phosphorylation of DAG, using the DAG-kinase enzyme. (58) The mitochondrial

GPAT is more likely to utilize saturated FA’s, while the microsomal GPAT generally uses

unsaturated FA’s. (59) There also seems to be a correlation between the type of mitochondrial

GPAT and AGPAT and the selection of fatty acids as substrates, although the mechanism

behind this remains uncertain. (60)

When PA follows the TAG pathway, PA-phosphatase (PAP) removes the phosphate group,

gaining 1,2-diacylglycerol (1,2-DAG). Eventually, the last fatty acid group is esterified by

DAG acyltransferase (DGAT), recovering TAG. (60) Though this pathway is commonly

present in liver and adipose tissue, intestines use the mono-acyl glycerol pathway in gaining

TAG, where absorbed dietary MAG can be sequentially converted into TAG by MGAT and

DGAT. (55)

23

Fig. 3: Overview of glycerolipid metabolism; adapted from (58)

Homeostasis of glycerophospholipids is extremely complex and the mechanisms concerning

de novo synthesis, degradation, fatty acid remodeling and interorganelle transport are not yet

fully understood. (58) Briefly summarized, glycerophospholipids can be gained via two

different pathways. Initially, after dephosphorilation of PA by PAP, DAG is converted into

PC and PE: this is the CDP-choline and CDP-ethanolamine pathway, also referred to as the

Kennedy pathway. The CDP-DAG pathway on the other hand, produces PI, PS and PG

through the reaction of PA with inositol, serine and glycerol respectively, while being aided

by CDP-DAG-synthase. (5)(55)(61) Furthermore, glycerophospholipids are able to convert

themselves into another type of phospholipid: PG can be irreversibly converted into

cardiolipin, while PS can alter into PE and vice versa. (61) A more detailed description of the

synthesis of glycerophospholipids is comprehensively described in a review about

glycerophospholipid homeostasis in mammalian cells. (58)

Newly synthesized glycerophospholipids have an altered fatty acid composition compared to

mature phospholipids. (61) All glycerophospholipids are submitted to fatty acid remodeling,

also called the Land’s cycle, thus creating a deacylation-reacylation reaction between

phospholipids and lysophospholipids. Deacylation is catalyzed by PLA1 or PLA2, depending

on whether it takes place at the sn1 or sn2 positions respectively. Lysophospholipid acyl

transferase (LPAT) substitutes the original FA by another FA on the glycerol backbone and

24

are divided in two protein families according to their substrate specificity: AGPAT (also

mentioned in the glycerophosphate pathway) and membrane bound O-acyltransferase

(MBOAT). (55)

Fig. 4: Fatty acid remodeling in glycerophospholipids; adapted from (55)

Degradation is another important step in the glycerophospholipid homeostasis process and it

is maintained via different lysosomal and non-lysosomal phospholipases. Alongside PLA1

and PLA2, PLC and PLD may also participate in the glycerophospholipid homeostasis. How

the coordination between biosynthesis and degradation of glycerophospholipids is realized,

still remains unclear. (58)

Fig. 5: Action points of phospholipases on the glycerol backbone; adapted

from (5)

25

4.3.2.2 Glycerolipids and colorectal cancer

4.3.2.2.1 Alterations in the glycerophosphate pathway

When starting with the first step in the glycerophosphate pathway, an enhanced uptake of

glycerol in HCT-15 colon cancer cells was observed. (62) Glycerol, an important metabolite

in gluconeogenesis, oxidation and lipogenesis , is transported via a specific Na+-dependent

carrier. This glycerol transporter may be a new target for drug development by reducing the

rate of the glycerophospholipid pathway, though more research is necessary on this subject.

(62) (63)

While GPAT seems to deliver no contribution to the carcinogenesis process, some evidence

indicated that AGPAT is over-expressed in several human cancers. AGPAT2 inhibition

induces growth arrest, necrosis or apoptosis and blocks RAS/RAF/Erk and PI3kinase/Akt

pathways. In addition, the expression of AGPAT11/LPCAT2 and AGPAT9/LPCAT1 were

found to be up-regulated in CRC tissue in comparison with normal mucosa. (64) This could

imply that there is a particular role for PA, being involved in cancer cell progression by

amplifying the Ras signal and activating the pro-survival MAPK and PI3K/Akt pathways and

thus the survival of tumors. (61)

LPA is an essential intermediate, influencing many pathological processes such as fibrosis,

inflammation, asthma, atherosclerosis and cancer. LPA is produced after Lyso-PLC reacts

with Lyspophospholipase D or when PA is deacylated by PLA1 or PLA2. The effects of LPA

are maintained by LPA receptors, which are coupled with specific G-proteins for regulating

downstream receptors. (65) Although LPA is a mitogen capable of enhancing proliferation in

CRC via interference with the APC/β-catenine pathway, this specific pathway is mutated in

the majority of tumor cells. There are hypotheses that LPA compensates this by over-

expression of Krüppel-like factor, a transcription factor that enhances intestinal crypt cell

proliferation. (66) Either way, the proliferation effects of LPA in an affected colon cancer cell

line depends on the type of LPA receptor. For instance, proliferation of DLD1 colon cancer

cells is promoted by LPA1, while HCT116 cells require LPA2 and LPA3. Moreover,

activation of LPA2 would deliver anti-apoptotic signals to colon cancer cells. (67) LPA2 also

showed elevated levels in CRC, while LPA1 receptors were down-regulated in colorectal

adenocarcinoma compared with normal colonic mucosa. The fact that these receptors are

aberrantly expressed in CRC and with LPA demonstrated to have proliferating, anti-apoptotic

26

effects could mean these play a certain role in the development of this disease. (68) Lastly,

cyclic Phosphatidic acid (cPA) is a structural analogue of LPA but instead, it has anti-

proliferative effects on DLD1 colon cancer cells. It inhibits cyclin D expression, blocks the

PI3K pathway and may have some potential as a therapeutic inhibitor of CRC progression.

(69)

4.3.2.2.2 Glycerophospholipids

The fatty acid remodeling enzymes, LPCAT, are able catalyze alterations in the colorectal

lipid profile, consequently contributing to cell malignancy. (67) Phosphatidyl choline (PC) is

an important structural element in cell membranes and plays a key role in cell cycle

regulation, proliferation and apoptosis by providing lipid second messengers. (27) Augmented

values of PC (16/0:16/1) were observed in CRC samples with more advanced stages.

Moreover, the ratio of PC (16/0:16/1)/LPC (16/0:16/1) was increased in colon cancer cell

membranes, implying an up-regulated activity of LPCAT4. Yet it still remains unclear if other

factors also play a role in this PC increase. (70)

Distribution of choline containing phospholipids in plasma were compared between healthy

subjects, individuals with adenomatous polyps and CRC patients. A detailed description of

these data can be read in Li, S. et al.

Shortly summarized, LPC plasma levels were gradually decreased through the colorectal

adenoma-carcinoma sequence, with the lowest concentrations present in healthy individuals.

Instead of colonoscopies, different species of LPC could be utilized as new detection

biomarkers for CRC. These could even be combined with other lipid metabolites that are

typically decreased in patient plasma such as sphingomyeline or sphingophosphocholine. (71)

PI molecules are best known for their roles in the PI3K-Akt/mTOR pathways: PI3Ks are

important kinases in gaining inactive PIP, PIP2 and active PIP3, which act as second

messengers and transduct signals from trans-membrane receptors to the cytosol. Since these

interfere in different cellular processes, failing of this pathway may result in continuous

activation with up-regulated angiogenesis, proliferation and cell survival in tumors due to

enhanced lipid metabolism and protein synthesis. Deletion of or mutations in the PTEN tumor

suppressor gene, which codes for a phosphatase that degenerates PIP3 was observed in CRC

tissues. Methods for inhibition of PI3K or conversion of active PIP3 to PIP2 or PIP, could be

valuable as new therapies in many different cancers, including CRC. Recent years, different

27

small molecule PI3K inhibitors have already been tested in clinical trials, with promising

results, but more research still remains necessary. (72) (73) (74)

4.3.2.2.3 Phospholipases

Findings suggest that PLA’s may be important regulators of cell growth.

Glycerophospholipids are known to be catabolized by PLA’s into FFA and lysolipids. Using

PC as example, these metabolites has potential proliferative effects on the one hand, but can

also re-enhance PC synthesis on the other hand by integrating these catabolized products

again in the CDP-choline pathway. This mechanism could be similar for the catabolism of

other types of glycerophospholipids. (27)

PS-PLA1 is a phospholipase that acts specifically on PS (phosphatidylserine) to produce

lysophosphatidylserine (LPS). Expression of PS-PLA1 was significantly higher in CRC

samples with increased tumor size, deeper invasion and presence of hematogenous metastasis.

Higher PS-PLA1 levels were also correlated with a decrease in disease-free survival. (75)

PLA2 is essential for releasing FFA at the sn2 position of glycerophospholipids. This includes

AA, a metabolite of the COX-1 and, more important, COX-2 pathways. As described earlier,

this is the source of pro-tumorigenic prostaglandins, for example PGE2. Defects in PLA2

could therefore be a protective factor in colorectal tumorigenesis. Cytoplasmic PLA2 was

already showed to be augmented in almost 50% of all types of colon cancer cell lines and

indeed, COX-2 expression was significantly correlated with cPLA2 expression. cPLA2

expression could enhance colon cancer development, although possibly in a less determining

way in comparison to COX-2, which is over-expressed in 80% of all colon cancer cell lines.

(76) In addition, it seems that (groupIIA) PLA2 positive cancer cells are correlated with a

lower disease free survival and patients live significantly shorter than those with PLA2

negative tumors. Furthermore, PLA2 is most abundantly expressed in stage II colorectal

cancer and the colorectal tumor’s position showed some correlation with the PLA2

expression: almost 70% of right-sided tumors were completely negative for PLA2 while this

was the case for only 42% of the left sided tumors. (77)

Platelet activating factor (PAF) does not only play a role in coagulation and immunology, but

is also suspected to take part in the carcinogenesis of colorectal and other tissue. PAF is

generated from a membrane-bound glycerophosphocholine by PLA2, gaining Lyso-PAF

28

which is followed by acetylation of the Lyso-group. Lastly, PAF can be catabolised by

acetylhydrolase (AHA). Remarkably, a worsening prognosis correlates with the amount of

catabolic products and enzymes of this pathway that are present in CRC tissue: T1-T4N0M0

had up-regulated activity of PLA2 and AHA with elevated tissue levels of LysoPAF and PAF,

TxN1M0 CRC showed the same profile but lacks elevation of PAF, while distant metastasis

(TxNxM1) CRC only showed an augmentation of AHA levels. (78) Furthermore, plasma

levels of sPLA2 and AHA were significantly, though modestly, increased in patients with

CRC. (79) These results imply that PAF de novo synthesis is increased in CRC, maybe

causing amplification of angiogenesis by VEGF production. Much still remains unclear

however, and the question thus arises if these findings are relevant. In particular,

information about this topic has apparently not evolved since 2003, which may indicate that

further investigation has not been successful, or that PAF has no interesting prospective in

CRC. (78)(79)

PLCδ1, a variant of phospolipase C, had a significantly reduced expression in CRC cell lines.

The data indicated that this enzyme inhibits tumor development, motility and invasiveness.

(80)

Phospholipase D degrades PC into PA and choline, while being aided by cofactor PIP2.

phospholipase D has a dual function: it provides structural integrity to (intra)cellular

membranes but also has a signaling function via PA and protein-protein interactions with

GTPases, kinases and phosphatases. Over-expression of phospholipase D has been found in a

variety of tumor types including CRC and is directly associated with angiogenesis, tumor

survival, cell migration and metastasis. Small molecules that are capable of inhibiting PLD

could be a promising new strategy in many types of cancers. (81)

4.3.2.2.4 Triglycerids

Serum and tissue concentrations of triglycerides have also been intensively investigated in

healthy and CRC patients. When serum and tissue lipid levels were compared between normal

and malignant colorectal tissue, tissue triglycerides were significantly down-regulated in

cancerous tissue but no significant correlation was noted for serum triglycerides. The clinical

TNM stages of colorectal cancer did show a correlation for both tissue as serum triglycerides,

where a progression in reduced triglycerides levels was noted in stage 3 and 4. Triglyceride

levels were lower in all patients with lymph node metastasis than those without lymph node

29

metastasis.(82) Furthermore, a dose-dependent positive association between adenoma

occurrence and triglyceride levels, with the distal colonic adenoma having the highest

significance , suggests a location dependent mechanism of lipid metabolism in colon cancer

development. (82) (83)

4.3.3 Sphingolipids

4.3.3.1 Pathway

Sphingolipids are composed of a sphingoid base backbone, most commonly sphingosine,

sphinganine or dihydrosphingosine, connected to a FFA via an amide-bond and a head group

(hydrogen, choline, serine, ethanolamine…) via an oxygen-bond. Depending on the type of

head group, the sphingolipids are divided in three main groups: sphingomyeline (SM),

ceramide and glycosphingolipids. (84)

Fig. 6: overview of sphingolipid metabolism

Ceramide is the central molecule in the sphingolipid metabolism. De novo synthesis

commences with palmitoyl-CoA and serine, followed by the intermediate products

sphinganine and dihydroceramide and eventually, synthesis of ceramide. (85) Ceramide can

be converted into ceramide-1-P, SM, glycerosphingolipids and sphingosine. A phosphate-

30

group can be added to this latter product, creating sphingosine-1-phosphate (S1P). Up to this

point, all mentioned steps can also be reversed for generating ceramide. Lastly, S1P-lyase

irreversibly divides S1P into ethanolamine-phosphate and palmitaldehyde.

It is important to notice that SM is commonly obtained by release from the cell membrane due

to sphingomyelinase (SMase). (86)

Glycosphingolipids are an extremely complex subgroup and it falls beyond the reach of this

thesis to characterize these. Briefly summarized, ceramide can be converted into

galactosylceramide or glucosylceramide, which in their turn can be elongated with other

saccharides such as mannose, fucose, galactose, N-acetylgalactosamine and N-

acetylneuraminic acid. The complexity of this process is illustrated in the figure below, which

represents the synthesis of glycosphingolipids in the brain. (87)

31

Fig.7: Overview of glycosphingolipid synthesis; adapted from (87)

4.3.3.2 Sphingolipids in colorectal cancer

Sphingolipids are abundant in the gut system, with twice the amount located in the small

intestine compared to the colon. (84) Not only are these compounds crucial structural

components in cell membranes, but these also provide important bioactive metabolites for

intracellular signaling and regulation of countless cellular processes including apoptosis,

growth, differentiation, proliferation, angiogenesis, cell adhesion, migration, inflammation

and lymphocyte traffic. (43)

Even though ceramide, sphingosine and S1P are considered the most important bioactive

sphingolipids, other metabolites also have their role. (89) An important remark is that

sphingolipids are quite restricted in traveling to other cellular compartments. Therefore,

effects of sphingolipids will largely take place within the compartment where they are

normally settled. (90)

As sphingolipids encompass many cellular functions, it is not improbable that alterations in

their metabolism could trigger processes that promote colorectal carcinogenesis or that grant

them resistance against certain (chemo)therapies. (90) Discrepancies in CRC concerning the

metabolism of every sphingolipid subgroup will be described in the next paragraphs with

some suggestions for therapy or screening opportunities. (86)

Regarding the SMnases, there are three subtypes of this enzyme: neutral SMase, alkaline

SMase and acidic SMase. A 1997 study showed distinct alterations in activity or presence of

these subtypes in CRC patients, who were ordered according to their Dukes stadium (A,B or

C). Alkaline SMase and neutral SMase had a decrease of respectively 75% and 50% in CRC

tissue compared with normal colorectal mucosa. Both enzymes were significantly lower in

stage B and C than in stage A. Acid SMase on the other hand, had no significant activities in

32

CRC. An important issue is that this study contained only 18 subjects, so these results should

be interpreted cautiously. (91)

Introducing 1,2-dimethylhydrazine, a DNA-methylating carcinogenic agent, to rat colon

mucosa gave a significant SM increase and SMase decrease, just before the stage when

colonic adenoma developed. (86) (88) This strongly suggested that a reduction in hydrolysis

of SM prior to the malignancy process takes place. (9)(88) In addition, Dillehay et al.

discovered that a SM-rich diet introduced to mice, who were again treated with 1,2-

dimethylhydrazine, restricted formation of aberrant crypts (pre-stadium lesions of colon

cancer). However, colon cancer incidence was not significantly reduced in these mice. (92)

Plasma levels of SM were also decreased in patients with adenomatosa polyposis (AP)

compared with healthy subjects and CRC patients, meaning that this could be a useful

biomarker for detecting AP. (71) Altogether, these results point out that SM could play its

part in the early stages of colorectal cancer development.(86) (93) (94)

Ceramide is a bioactive molecule that inhibits cell proliferation and activates apoptosis by

modification of different molecules and pathways (AKT, Bcl-2, PKCα…) (88) It can undergo

glycolysation, hydrolysis or phosphorylation, which determines whether this sphingolipid

metabolite will obtain apoptotic or mitogenic capacities. Additionally, an abundant quantity

of ceramide is generated when the colorectal cancer cells are stressed by radiotherapy,

chemotherapy, hypoxia or nutrient deprivation which activates apoptotic pathways. (95)

Investigation of primary and metastatic colon cancer in human subjects, pointed out that

ceramide levels were less than half the amount found in normal colon mucosa. (93)

Ceramidase, which also is divided in a neutral, acid and alkaline subtype, is an important

regulator of sphingosine and S1P. However, the ceramidase activity did not appear altered in

CRC patients, implying that low ceramide levels are predominantly caused by reduced SMase

activity. (91) This theory however, still requires more evidence because other factors may be

responsible for the changes in SM and ceramide levels such as alterations in de novo

sphingolipid synthesis, ceramide glycolysation and other processes. (86)

Here again, administration of ceramide analogues, C2 and C6, or inhibitors of ceramidase, D-

MAPP and B13, induced cell death in human CRC cell lines (SW403). In particular B13 had

astounding effects when introduced to mice, who had their liver injected with human colon

33

cancer cells . 70% of the animals with the SW403 and 100% of the animals with the Lovo cell

line remained tumor-free after treatment with B13. (93)

Furthermore, treatment of CRC with chemotherapy combined with an adjuvant that blocked

de novo ceramide synthesis (Fumonisin B1) or GlcCer production (1-phenyl- 2-

palmitoylamino-3-morpholino-1-propanol or PPMP) was observed. Fumonisin B1 as adjuvant

significantly decreased the rate of CRC cell apoptosis and increased the GlcCer levels in

comparison with chemotherapy alone. PPMP administration on the other hand doubled the

ceramide concentrations and increased the cell death ratio by 88%. (96) All of these

molecules that interfere with the ceramide metabolism by augmenting ceramide

concentrations, may prove useful as adjuvant chemotherapeutic agents against colorectal

cancer. (97)

Breakdown of ceramide by ceramidase offers sphingosine, a sphingoid base. Literature

claimed this molecule is quantitatively restricted in comparison to its phosphorylated variant,

S1P. Sphingosine has been put forward as a signaling molecule with pro-apoptotic and tumor-

suppressive properties by regulating Protein Kinase C-isomorfs and acidic nuclear

phospoproteins. (98) When sphingosine and sphinganine were added to a human colon cancer

cell line, this induced cell arrest at G2/M and augmented the rate of apoptosis in the

neoplasmic cells. This effect was similar for C2-ceramide, a short chain ceramide analogue,

but didn’t emerge for C2-dihydroceramide, a short chain dihydroceramide analogue,

indicating that the effects have something to do with the 4,5-trans double bond in ceramide,

which is absent in dihydroceramide. (99)

S1P is more likely to be a pro-carcinogenic factor, with mitogenic and pro-angiogenic

properties, inhibiting apoptosis and promoting maturation of colorectal neoplasm cells via

S1P-receptors. (86) It neutralizes the pro-apoptotic effects of ceramide and sphingosine when

added to tumor cells under stress reactions. (98) S1P also appears to be the missing

connection between colorectal inflammation and CRC development: by continuously

activating NF-kB and STAT-3, it provides proliferative and survival advantages to colorectal

cells. Moreover, STAT-3 again activates the S1P-R, creating a vicious circle in the

promotion of cell clonal expansion. (100) New therapeutic strategies with specific anti-S1P

antibodies already appear quite promising in different tumor lineages, including CRC. (101)

34

Deficiency of sphingosine-kinase-1 (SphK1) in mice inhibits colon polyp formation and

furthermore , is generally over-expressed in human colon cancer. (100)

Observation of SphK in colorectal tissue showed no expression in normal mucosa, negative or

moderately positive expression in adenoma, but positive expression in 89% of

adenocarcinoma and significantly higher expression in all metastatic adenocarcinoma. There

was, however, no correlation between SphK-expression and tumor invasion. (89)

Furthermore, S1P-lyase and Sphingosine-Posphatase were under-expressed in human colon

cancer cell lines, thus blocking S1P catabolism and apoptosis. The fact that S1P-lyase was

under-expressed could even suggest that sphingolipids obtained by diet, are not metabolized

in colon cancer. (102)

The S1P-pathway and the COX-2 pathway have been shown to share a connection: SphK

shutdown decreased COX-2 and PGE2 production, while S1P enhances COX-2 expression

and PGE2 production. (88) Then again, ceramide kinase and ceramide-1-P are necessary for

translocation and activation of cPLA2, which provides substrates (arachidonic acid) for the

COX-2 inflammatory and pro-carcinogenic pathway. Both these sphingolipid metabolites

significantly augment PGE2-values via two different but completely coordinated and

synergistic pathways than when given separately and in the same quantity. (103) Interestingly,

tissue samples of colorectal adenocarcinoma and their metastatic variants with an elevated

COX-2 expression were also positive for SphK, while a positive expression of SphK was not

always linked to positive COX-2 expression. (89)

As previously mentioned, glycospingolipids are extremely diverse, although changes in the

glycolysation pattern of glycolipids is a familiar phenomenon in a large quantity of cancers,

including CRC. (104) As an illustration: sialidation is often over-expressed in colon cancer

cells. Indeed, human Neu3 activity is elevated in neoplastic colons, blocking programmed cell

death and even modulating cell differentiation. (105) These findings are partially confirmed

by the augmentation of lactosylceramide (2 to 5 times higher than in normal mucosa) and a

sialidase product in colon cancer cells . (106) Another study revealed that glycosphingolipids

in CRC obtained an increased fucosylation, a decreased glycan acetylation, sulfation and

ganglioside disialylation. (104) Furthermore, the higher the Dukes stage or metastatic

potential of the patient’s CRC, the higher the sulfogalactosylceramide levels. (107) Lastly,

chemosensitizers combined with GCS-inhibitors induced cytotoxicity and therefore, apoptosis

in colon cancer. (97) Sphingolipid glycolysation could lead us to new insights on colorectal

35

carcinogenesis, new strategies for therapy or new biomarkers for screening. However, a real

pattern in abberant glycolysation has not actually been determined yet, because these differ in

every type of cancer and make use of different glycan-groups. (104)

4.3.4 Sterols

4.3.4.1 Pathway Cholesterol is an extremely important molecule in the human body: it is an essential

component in cell membranes, in which the majority is non-esterified cholesterol, and

intracellular membranes, where the greater part consists of esterified cholesterol. Furthermore,

it is a precursor of bile acids, Vitamin D and steroid hormones. (108)

Fig. 8: chemical structure of cholesterol

Cholesterol is a polycyclic structure containing a total of 27 carbon atoms. The hydroxyl-

group on C3 can be replaced by an ester-bond. Both cholesterol forms are transported in

lipoprotein-particles (chylomicrons, VLDL, LDL, IDL and HDL) due to its lack of solubility

in water or plasma. Cholesterol can be obtained through diet, but is mainly de novo

synthesized. Cholesterol biosynthesis is a complex process : aided by the mevalonate

pathway, it commences with condensation of acetyl-CoA and acetoacetyl-CoA, producing

HMG-CoA. Eventually, this product is reduced to mevalonate by HMGCoA-reductase, the

rate-limiting step of cholesterol biosynthesis. (7) Afterwards, two kinase reactions

sequentially take place, directly followed by a decarboxylation, forming

isopentenylpyrophosphate (IPP). This latter product can be converted into

dimethylallylpyrophosphate (DMPP) by isomerisation. Condensation of IPP and DMPP

produces geranylpyrophosphate (GPP), which sequentially undergoes a condensation reaction

with IPP, gaining farnesylpyrophosphate (FPP). Both these condensations are sustained by

36

geranylpyrophosphate-synthase (GPPs) and farnesylpyrophosphate-synthase (FPPs). Squalene

is synthesized in the next step, derived from FPP. With help of the squalene-epoxidase and

lanosterol-synthase, it is converted into lanosterol. From here on, 19 different reactions aided

by nine different enzymes, are necessary to gain the final product, cholesterol. (108)

In addition, cholesterol can be esterified with a long FFA by acyl-coenzyme cholesterol

acyltransferase (ACAT), gaining cholesteryl esters. This is the main form in which cholesterol

can be stored in cells, and transported through the blood via lipoproteins.

Fig.9: Overview of cholesterol biosynthesis

4.3.4.2 Alterations in colorectal cancer

Cholesterol and cholesterol precursors are known to be important metabolites in

carcinogenesis: cancer cells require higher concentrations during their development and

further growth. Targeting this pathway could supply new therapeutic methods against CRC,

although changes in this pathway may also contribute to chemotherapy resistance. (109)

Cholesterol transport also showed some discrepancies in malignant cells, but this will be

described in the next section on lipoproteins.

37

Oxysterols are catabolic products of cholesterol, produced by cytP450-enzymes or by

reactions with reactive oxygen and nitrogen. (110) They are able to connect with LXR-

receptors and activate different cholesterol efflux pump and other regulatory proteins.

Oxysterols are also involved in LDL-receptor degradation. (109) Some oxysterols are capable

of initiating colorectal tumorigenesis by augmenting ROS/RNS activity, thus activating pro-

carcinogenic proteins like COX-2 and even supporting tumor progression by promoting

migration of neoplastic cells. (110) While it can therefore be stated that oxysterols have a

protective influence during carcinogenesis by inhibiting cholesterol abundance in cells, the

fact that some oxysterols have pro-inflammatory and pro-oxidative capacities (thus being pro-

carcinogenic), contests the former theory. (109) (110)

The cellular mechanism of cholesteryl ester (CE) homeostasis remains uncertain, but findings

indicate that many cancer cell lines are submitted to higher CE synthesis and increasing

uptake of CE delivered by LDL and HDL.(111) In breast cancer, CE cell accumulation has

already been associated with a malignant phenotype and proliferative effects, but the effects

on CRC specifically have not yet been investigated. (112)

Regarding the mevalonate pathway, an elevation of HMGCoA-reductase, FPP synthase and

farnesyltransferase have been noticed in different types of cancer, especially in colorectal

cancer.

HMGCoA-reductase is the rate limiting step in cholesterol biosynthesis, meaning this

enzyme has a major impact on pathway progression. (113) Its product, mevalonate acid, is

known to rapidly initiate DNA-replication, enhancing control loss of the synthesis pathway

and contributing to malignant transformation. (114) Although most CRC samples have

increased levels of surface LDL-receptors, HMGCoA-reducate activity in CRC was more

significantly elevated in LDL-R negative colorectal tumors, as absence of LDL-R in

malignant cells imply they are solely dependent on endogenous synthesis of cholesterol

instead on environmental uptake. (115) Quite contradictory with the former findings, absence

of LDL-R in colorectal tumors had a negative impact on patient survival, but expression of

HMGCoA-reductase is more up-regulated in TNM stage I and II than in stage III and IV, thus

giving patients a better prognostic value. (116) (117)

It also seems that HMGCoA-activity is correlated with the tumor localization in the colon:

tumors on the left side of the colon had three times the enzyme activity compared with rectal

tumors. (115) As the greatest risk in colon cancer is metastasis and in rectal cancer,

38

recurrence, these arguments feed the theory that different tumor localizations are associated

with different molecular mechanisms in carcinogenesis of which the cholesterol pathway has

its own unique function. (113)

Besides elevated mevalonate levels, many studies indicate that isoprenoid compounds FPP

and GPP are elevated in malignancy, activating oncogenes and promoting tumorigenesis.

(114)(115) Indeed, FFPs activity was enhanced in CRC tissue compared to normal mucosa.

Furthermore, FPP is the essential substrate for farnesylation, a protein isoprenylation process.

These protein alterations provide them with unique capacities considering proliferation,

differentiation and cell survival, properties that could be altered in neoplasmic cells. An up-

regulation of farnesyl transferase was noted in CRC hence confirming this theory. (114)

Inhibitors of HMGCoA-reductases, statins, are widely used for lowering cholesterol serum

levels as a part of cardiovascular risk prevention but might prove themselves worthy as

chemo-preventive agents in colorectal cancer. More and more evidence suggests that statin

intake may also interfere in tumor proliferation, growth and metastasis. (118) Besides down-

regulation of cholesterol, it reduces-PP and geranylgeranyl-PP concentrations, inhibiting

prenylation of proteins and hereby restricting essential physiological functions in tumor

development. (119) Although the role of simvastatin remains somewhat unclear: it showed

promotion of apoptosis while reducing human cancer cell proliferation in vitro. (115) (117)

Another study showed no effects in cancer cell proliferation or apoptosis, but administration

inhibited cell migration and thus metastasis probability. However,

geranylgeranylpyrophosphate or mevalonate addition reduced statins inhibitory effects on cell

migration. (120) Conclusively, a meta-analysis reported a small but significant protective

effect of statins in CRC by decreasing invasiveness, metastatic properties, and even

chemosensitizing the neoplastic cells for other chemotherapeutics. (119)

Biphosphonates are commonly used for the treatment of osteoporosis, but since these are

capable of targeting the FPPs in the mevalonate pathway, they could be valuable in blocking

cholesterol synthesis, thus restricting further CRC development.

(114) A meta-analysis concluded that oral administration of biphosphonates clearly reduces

the risk of CRC, depending on the dose and duration of drug administration. (121)

39

Total cholesterol (TCH) and free cholesterol (FCH) were measured in patients with CRC and

compared with healthy patients or benign colorectal lesions. A drop in serum TCH was

observed while serum FCH was significantly elevated. Furthermore, only serum TCH

concentrations were significantly and progressively reduced in CRC stage III and IV. The

fact that neoplasm cells have higher need of cholesterol, could explain the drop in cholesterol

serum levels due to higher uptake, while the low levels of tissue cholesterol may be caused

by the cells limited cholesterol synthesis and transport capacities, while trying to keep up with

the enhanced cholesterol need. This hypothesis remains vague and the mechanism behind this

theory hasn’t been elucidated yet. The only conclusion that can be made is that the

physiological homeostasis of cholesterol is diminished in CRC, which could offer new

therapeutic targets or predictive biomarkers. One final remark is that these alterations in lipid

levels could have been present before the patients illness, as they are independent risk factors

in the development of CRC. (82)

Besides alterations in biosynthesis, cholesterol homeostasis can also be deregulated by

deficiencies in the cholesterol efflux pumps of cells. An example is the ABCA1 efflux

function in mitochondria: inhibited ABCA1 expression in colon cancer cells increases the

mitochondrial cholesterol concentrations and reduces the release of cell death promoting

molecules. Consequently, these tumor cells are able to have prolonged survival rates. (122)

Although cholesterol is the essential precursor molecule of Vitamin D, bile acids and sterols,

alterations in their metabolism and/or the association with CRC falls beyond the reach of this

thesis.

4.3.5 Lipoproteins and lipid droplets

Lipoproteins mainly comprise molecules that are capable of transporting lipids such as

cholesterol, cholesteryl esters, TG and vitamins, to all cells. The tend to occur in a structure

that is composed of a phospholipid monolayer stabilized by neutral apolipoprotein, with a

central hydrophobic core within. Lipoproteins are classified into four classes according to

their progressing density: chylomicrons, VLDL, IDL, LDL, and HDL. Additionally, the

hydrophobic core is an ideal environment for the transportation of lipophilic agents such as

certain chemotherapeutic agents, as these remain stabilized within this core and are not

directly removed from the bloodstream, offering an elevated half-life for these compounds.

(118)

40

Many types of cancers over-express HDL- and LDL-receptors on their cell surface, with

serum HDL showing decreased values and augmented quantities of tissue HDL in cancer

patients. Indeed, CRC patients had reduced serum HDL and LDL values while tissue HDL

was increased in CRC. These values were even more highlighted when progressing through

the TNM stages due to the fact that higher tumor grade is correlated with higher demand for

cholesterol, and thus enhanced cholesterol transport. (82) (118) As Apolipoproteins form the

fundamental basis in the synthesis of lipoproteins, over-expression of these structures could

be playing a role in tumor development. Apo-A1 and Apo-B may have an altered expression

that could be associated with the HDL and LDL shifts seen in CRC. (82) Apolipoprotein E1

gene polymorphisms have already been correlated with higher CRC risk. (123)

Lipid droplets or lipid bodies are small, dynamic organelles, consisting of a lipid monolayer

and capable of stocking all cytosolic lipids in human cells. Besides lipids, these organelles

also contain a significant amount of proteins that are involved in lipid metabolism, but also

other processes that could enhance carcinogenesis. Well known examples are PI3K, MAPK,

cPLA2, COX, and LOX-enzymes, of which the latter three enzymes claim essential roles in

the pro-inflammatory eicosanoid pathway while using AA as substrate. (124) Lipid bodies

were significantly increased in colon carcinoma cell lines in comparison to normal mucosa

when investigated using Raman microscopy. (125)

Also, expression of COX-2 was fortified in lipid droplets of colon cancer cells, which was

correlated with enhanced production of PGE2. As this is a pro-inflammatory and pro-

tumorigenic factor, as described above, lipid droplets may contribute to the development of

CRC and could be suggested as a new therapeutic target, while the amount of lipid droplets in

CRC cells could be brought forward as a prognostic biomarker. (124)

41

4.4 An interpretation of previously obtained experimental data

Dr J.Foster performed data analysis for his PhD dissertation at Cambridge University, UK of

comparative lipidomic profiling data of CRC samples with normal mucosa samples. His

analysis not only aimed for identifying alterations of general lipid subclasses, but also of

individual lipids that were organized according to their specific acyl bond.

Generally, tumor samples showed a significant increase (P = 0.0025) in the total sum of lipid

concentrations when compared to normal samples. This phenomenon repeated itself as the

total amount of lipids corresponding to a specific subclass was significantly altered in CRC

samples (P < 0.05) in all except three subclasses; only LPC, SM and PA did not show

significant alteration. Furthermore, the concentrations of lipid substrates within these

subclasses were increased in all tumor samples compared to normal mucosa, but subclasses

LPA, monoacylglycerols and TG were decreased.

Glycerolipids TG, PC and PE were altered most frequently in CRC and were mainly down-

regulated. These tumor samples showed a clear preference for long chain desaturated PS

(phosphatidylserine) species, while short chain saturated PS species were moderately down-

regulated. The total amount of all PE species concentrations was absolutely increased in

tumor tissues, with evidence that some reactions shifted towards the production of PE, by

using substrates PC, PS and DAG (diacylglycerol). LPC was also strongly altered in tumor

tissue in comparison with normal mucosa.

Other important findings were the shifts that manifested when lipid profiles were compared

between tumor and normal mucosa samples, tumors with different stages and different tumor

sizes. Ceramide 16:0 levels were increased in tumor size 1 samples compared to other

samples, while ceramide 18:0 levels were decreased. Additionally, PE species of patients who

received radiotherapy prior to tissue sample collection were evaluated: this highlighted that

PE 36:1 and PE 36:2 were reduced, while PE 38:4 was augmented in pre-radio samples in

comparison with all other samples. Lastly, PI and PS values of tumor size 2 samples were put

side by side with those of tumor size 3 samples. While PI 34:1 PI 36:1 and PI 36:2 were

elevated, PI 38:4 was down-regulated in tumor size 3 samples compared to tumor size 2.

Furthermore, tumor size 3 shifted towards lower PS 44:7 and PS 44:8 values. (126)

42

5 Discussion

Alterations of lipid metabolism in CRC cover an enormous quantity of information due to its

extreme complexity. Lipidomics remain a somewhat under-investigated area and the findings

that have been published so far are likely to reveal only the tip of the iceberg.

A lot of opportunities for therapy and screening methods lay within this territory, of which

some have already been mentioned above.

Regarding the FFA, ACLY and FASN have been intensively examined enzymes.

Experiments with ACLY and FASN-inhibiting medicines already showed promising results,

while serum FASN values could provide predictive and prognostic information. The shift in

FFA distribution in CRC tissue compared to healthy colorectal cells is another interesting

finding. What mechanisms drive the cancer cells in producing more ‘aggressive’ ω-6 FFA,

AA in particular, in comparison with more ‘protective’ ω-3 FFA, EPA and DHA? Although a

lot of attention has been devoted to the production of pro-inflammatory eicosanoids (e.g.

PGE2, TxA2, 12-HETE) and their role in colorectal carcinogenesis, decreasing the amount of

substrates used for these pathways may be of exceptional value for inhibiting further

development of CRC.

Glycerolipids, and mainly glycerophospholipids, have an unmistakable role in CRC

pathogenesis. In literature as well as in Foster’s thesis, TG levels were demonstrated to be

altered in different CRC stages, although it is still questionable what clinical benefits could

be acquired from these findings in CRC. LPA, PC and PLA2 have also been intensively

examined throughout the years and showed more potential for future use. PI and moreover,

PIP3, is part of a key regulator in many intracellular pathways that augment pro-tumorigenic

pathways. Suppressing PI3K/Akt/mTOR pathways have already been tested with

pharmaceuticals in different cancer types. This offered promising results at the beginning of

the trial, but chronic treatment forced the cancer cells to adapt and use other growth-

stimulating pathways, making these medicines unable to fully exterminate the cancer cells.

(127) There could therefore be room for different approaches in targeting these pathways in

CRC, for instance by inhibiting synthesis of PI-molecules.

Ceramide is likely to be at the epicenter of spingolipid metabolism alterations in CRC cells.

All results indicate that cancer cells try to avoid pathways that favor ceramide production,

43

thereby evading its apoptotic and anti-proliferative capacities. Developing methods for

enhancing ceramide production, or inhibit pathways that cut off ceramide supplies could be of

therapeutic value.

In cholesterol metabolism, all evidence points out that de novo synthesis is up-regulated in

CRC, together with enhanced uptake of cholesterol provided by lipoproteins LDL or HDL.

Not only enhanced cholesterol synthesis, but also the increased isoprenylation activity from

its intermediates FPP and GPP could be interesting targets. For decades, cholesterol

metabolism has been a major topic in cardiovascular diseases while data are showing some

significance in malignancy processes. This reinforces speculations that pharmaceuticals that

lower cholesterol uptake or that inhibit de novo synthesis could also be of use in the therapy

or prevention of CRC.

I conclude that there is plenty of information that suggests an essential role for lipid

metabolism in CRC pathogenesis. This may lead to new opportunities in treatment and

prevention of CRC, granting the patient a better prognosis due to improved and more

personalized medicines and earlier detection of the tumor, preferably during curable stages.

Nowadays, CRC is still classified according to the TNM or TNM clinical staging system, an

instrument that is primarily focused on surgical treatment of CRC. As every CRC covers a

specific, heterogeneous metabolic profile, the question rises if lipidomics (and other ‘omics’)

approaches could become the new standard in adequately categorizing CRC on a molecular

basis. This molecular classification could offer patients a personalized therapy-schedule,

depending on the type of molecular defects their colorectal tumor acquired. Still, this thesis

confirms that lipidomics is standing in its infancy, as we know so much and yet so little about

the normal biological processes in mammals that interfere with lipids, let alone in

pathological processes such as CRC. The majority of the literature that I have investigated

mainly focused on the alterations of lipid molecules in CRC regarding absolute concentrations

within a lipid sub-class. A handful of articles, including Foster’s thesis, were the only ones

who attempted to examine lipid metabolism alterations in CRC by observing each individual

lipid from each lipid sub-class on their quantity, distribution and composition in comparison

to normal mucosa. Indeed, this is an expensive and time consuming method of investigation,

but this may be the only way to fully understand the mechanisms behind CRC pathogenesis.

A long road still lays ahead of us if we want to efficiently progress within this specific

territory as much work remains to be done.

44

6 Summary in Dutch

Colorectale darmkanker blijft één van de meest prevalente, incidentrijkste en dodelijkste

maligniteiten ter wereld. Wetenschappers testen steeds nieuwe technieken te uit om

efficiëntere screeningsmethoden en therapieën te ontwikkelen . Lipidomics is onderdeel van

het netwerk van de metabolomics en bestudeert de distributie van lipiden in menselijke cellen,

weefsels, maar ook andere organismen. Lipidomics blijft vooralsnog een weinig onderzocht

terrein, doch heeft al veelbelovende resultaten geboekt in verscheidene soorten maligniteiten.

In deze literatuurstudie werd getracht om zoveel mogelijk informatie samen te vatten die tot

op heden gekend is inzake het verband tussen lipidenmetabolisme en colorectale darmkanker.

Naast de lipoproteïnen en de intracellulaire lipidendruppels, zullen de meest significante

veranderingen binnen de grootste lipidengroepen worden besproken. Dit betreft de vrije

vetzuren, glycerolipiden (met name de triglyceriden en de fosfoglycerolipiden),

sphingolipiden en sterolen. Er werd alvast gestegen activiteit waargenomen van enzymes die

de synthese van vrije vetzuren reguleren, evenals significante veranderingen in de

samenstelling van vrije vetzuren in colorectale tumor specimens. Er werd alsook een gedaalde

hoeveelheid triglyceriden en phosphaticylcholine glycerophospholipiden genoteerd in deze

specimens. Gezien ceramide de centrale molecule is binnen het sphingolipidenmetabolisme

en beschikt over pro-apoptosis en anti-proliferatieve eigenschappen, pogen colorectale

maligne cellen de productie van dit specifiek metaboliet te vermijden. Een gestegen de novo

synthese van cholesterol werd waargenomen, evenals een toegenomen opname van

cholesterol vanuit de omgeving via lipoproteïnen, wat duidelijk kon aangetoond worden

gezien de gestegen cellulaire expressie van HDL- en LDL-receptoren. Tot slot is maligniteit

al enige tijd geassocieerd aan een verhoogde aanwezigheid van intracellulaire lipidendruppels

en dit fenomeen werd inderdaad bevestigd in colorectale darmkanker specimens. Desondanks

de bovenvermelde bevindingen die aangeven dat er weldegelijk een associatie bestaat tussen

colorectale darmkanker en lipidenmetabolisme en er voldoende opportuniteiten bestaan om

nieuwe therapieën en biomerkers te ontwikkelen, staat dit gebied nog steeds in haar

kinderschoenen. Heel wat onderzoek zal nog moeten verricht worden willen we het

lipidenmetabolisme integreren in het huidige screenings-en therapiebeleid, niet alleen voor

colorectale kanker maar ook voor andere soorten maligniteiten.

45

7 References

1. Brenner H, Kloor M, Pox CP. Colorectal cancer. Lancet [Internet]. Elsevier Ltd; 2014

Apr 26 [cited 2014 Jul 10];383(9927):1490–502.

2. Cserni G. Nodal staging of colorectal carcinomas and sentinel nodes. J Clin Pathol

[Internet]. 2003 May 1;56(5):327–35.

3. Vermeersch KA, Styczynski MP. Applications of metabolomics in cancer research. J

Carcinog [Internet]. Medknow Publications and Media Pvt. Ltd.; 2013 Jan 1 [cited

2015 May 22];12(1):9.

4. Lipidomics: Analysis of the Lipid Composition of Cells and Subcellular Organelles by

Electrospray Ionization Mass Spectrometry [Internet]. [cited 2016 Mar 26].

5. Huang C, Freter C. Lipid Metabolism, Apoptosis and Cancer Therapy. Int J Mol Sci

[Internet]. 2015;16:924–49.

6. Li M, Zhou Z, Nie H, Bai Y, Liu H. Recent advances of chromatography and mass

spectrometry in lipidomics. Anal Bioanal Chem. 2011;399(1):243–9.

7. Santos CR, Schulze A. Lipid metabolism in cancer. FEBS J [Internet]. 2012 Aug [cited

2014 Jul 11];279(15):2610–23.

8. Sideris M, Papagrigoriadis S, Hill D. Molecular Biomarkers and Classification Models

in the Evaluation of the Prognosis of Colorectal Cancer. 2014;2068:2061–8.

9. García-Barros M, Coant N, Truman J-P, Snider AJ, Hannun Y a. Sphingolipids in

colon cancer. Biochim Biophys Acta [Internet]. Elsevier B.V.; 2014 May [cited 2014

Nov 4];1841(5):773–82.

10. Chawla N, Butler EN, Lund J, Warren JL, Harlan LC, Yabroff KR. Patterns of

colorectal cancer care in Europe, Australia, and New Zealand. J Natl Cancer Inst -

Monogr. 2013;(46):36–61.

11. Binefa G, Rodríguez-Moranta F, Teule A, Medina-Hayas M. Colorectal cancer: From

prevention to personalized medicine. World J Gastroenterol [Internet].

2014;20(22):6786–808.

12. Malesci A, Laghi L. Novel prognostic biomarkers in colorectal cancer. Dig Dis.

2012;30:296–303.

13. Gassler N, Klaus C, Kaemmerer E, Reinartz A. Modifier-concept of colorectal

carcinogenesis: Lipidomics as a technical tool in pathway analysis. World J

Gastroenterol. 2010;16(15):1820–7.

14. Coppedè F, Lopomo A, Spisni R, Migliore L. Genetic and epigenetic biomarkers for

diagnosis, prognosis and treatment of colorectal cancer. World J Gastroenterol

46

[Internet]. 2014;20(4):943–56.

15. Nishiumi S, Kobayashi T, Ikeda A, Yoshie T, Kibi M, Izumi Y, et al. A novel serum

metabolomics-based diagnostic approach for colorectal cancer. 2012;7(7):1–10.

16. Li L, Han J, Wang Z, Liu J, Wei J, Xiong S, et al. Mass spectrometry methodology in

lipid analysis. Int J Mol Sci. 2014;15(6):10492–507.

17. Schwudke D, Schuhmann K, Herzog R, Bornstein SR, Shevchenko a. Shotgun

Lipidomics on High Resolution Mass Spectrometers. Cold Spring Harb Perspect Biol.

2011;3(9):a004614–a004614.

18. Swinnen J V, Brusselmans K, Verhoeven G. Increased lipogenesis in cancer cells: new

players, novel targets. Curr Opin Clin Nutr Metab Care. 2006;9(4):358–65.

19. Balog CI a, Stavenhagen K, Fung WLJ, Koeleman C a, McDonnell L a, Verhoeven a,

et al. N-glycosylation of colorectal cancer tissues: A liquid chromatography and mass

spectrometry-based investigation. Mol Cell Proteomics [Internet]. 2012 Sep [cited

2014 Nov 11];11(9):571–85.

20. Currie E, Schulze A, Zechner R, Walther TC, Farese R V. Cellular fatty acid

metabolism and cancer. Cell Metab [Internet]. Elsevier Inc.; 2013;18(2):153–61.

21. Yang K. Expression profile of polyunsaturated fatty acids in colorectal cancer. World J

Gastroenterol [Internet]. 2015;21(8):2405.

22. Cathcart M-C, Lysaght J, Pidgeon GP. Eicosanoid signalling pathways in the

development and progression of colorectal cancer: novel approaches for

prevention/intervention. Cancer Metastasis Rev [Internet]. 2011 Dec [cited 2014 Nov

11];30(3-4):363–85.

23. Eicosanoid Synthesis and Metabolism: Prostaglandins, Thromboxanes, Leukotrienes,

Lipoxins [Internet]. [cited 2015 Nov 15].

24. Zhang G, Panigrahy D, Mahakian LM, Yang J, Liu J-Y, Stephen Lee KS, et al. Epoxy

metabolites of docosahexaenoic acid (DHA) inhibit angiogenesis, tumor growth, and

metastasis. Proc Natl Acad Sci U S A [Internet]. 2013;110:6530–5.

25. Zaidi N, Swinnen J V., Smans K. ATP-Citrate Lyase: A Key Player in Cancer

Metabolism. Cancer Res [Internet]. 2012;72(15):3709–14.

26. Medes G, Weinhouse S, Thomas A. Metabolism of Neoplastia Tissue. IV. A Study of

Lipid Synthesis in Neoplastia Tissue Slices in Vitro. Cancer Res [Internet].

1953;13(2):27–9.

27. Ridgway ND. The role of phosphatidylcholine and choline metabolites to cell

proliferation and survival. Crit Rev Biochem Mol Biol [Internet]. 2013;48(1):20–38.

28. Gutiérrez-Aguilar M, Baines CP. Physiological and pathological roles of mitochondrial

SLC25 carriers. Biochem J [Internet]. 2013;454(3):371–86.

47

29. Zhou Y, Bollu LR, Tozzi F, Ye X, Bhattacharya R, Gao G, et al. ATP Citrate Lyase

Mediates Resistance of Colorectal Cancer Cells to SN38. Mol Cancer Ther [Internet].

2013;12(12):2782–91.

30. Hatzivassiliou G, Zhao F, Bauer DE, Andreadis C, Shaw AN, Dhanak D, et al. ATP

citrate lyase inhibition can suppress tumor cell growth. Cancer Cell [Internet].

2005;8(October):311–21.

31. Zaytseva YY, Elliott V a., Rychahou P, Mustain WC, Kim JT, Valentino J, et al.

Cancer cell-associated fatty acid synthase activates endothelial cells and promotes

angiogenesis in colorectal cancer. Carcinogenesis [Internet]. 2014;35(6):1341–51.

32. Chang L, Wu P, Senthilkumar R, Tian X, Liu H, Shen X, et al. Loss of fatty acid

synthase suppresses the malignant phenotype of colorectal cancer cells by down-

regulating energy metabolism and mTOR signaling pathway. J Cancer Res Clin Oncol

[Internet]. Springer Berlin Heidelberg; 2015;

33. Notarnicola M, Tutino V, Calvani M, Lorusso D, Guerra V, Caruso MG. Serum Levels

of Fatty Acid Synthase in Colorectal Cancer Patients Are Associated with Tumor

Stage. J Gastrointest Cancer [Internet]. 2012;43:508–11.

34. Long Q, Yi Y, Qiu J, Xu C, Huang P. Fatty acid synthase (FASN) levels in serum of

colorectal cancer patients: correlation with clinical outcomes. Tumor Biol [Internet].

2014;35:3855–9.

35. Flavin R, Peluso S, Nguyen P, Loda M. Fatty acid synthase as a potential therapeutic

target in cancer. Futur Oncol [Internet]. 2010;6(4):551–62.

36. Wang C, Xu C, Sun M, Luo D, Liao D, Cao D. Acetyl-CoA carboxylase-α inhibitor

TOFA induces human cancer cell apoptosis. Biochem Biophys Res Commun

[Internet]. 2009;385(3):302–6.

37. Igal R a. Stearoyl-CoA desaturase-1: a novel key player in the mechanisms of cell

proliferation, programmed cell death and transformation to cancer. 2010;31(9):1509–

15.

38. Zhang J, Zhang L, Ye X, Chen L, Zhang L, Gao Y, et al. Characteristics of fatty acid

distribution is associated with colorectal cancer prognosis. Prostaglandins Leukot

Essent Fatty Acids [Internet]. Elsevier; 2013 May [cited 2014 Nov 11];88(5):355–60.

39. Zhan Y, Ginanni N, Tota MR, Wu M, Bays NW, Richon VM, et al. Control of Cell

Growth and Survival by Enzymes of the Fatty Acid Synthesis Pathway in HCT-116

Colon Cancer Cells. Clin Cancer Res [Internet]. 2008;14(18):5735–42.

40. Mazzarelli P, Pucci S, Bonanno E, Sesti F, Calvani M, Giusto Spagnoli L. Carnitine

palmitoyltransferase I in human carcinomas: A novel role in histone deacetylation?

Cancer Biol Ther [Internet]. 2007;6(October 2015):1606–13.

41. Zaugg K, Yao Y, Reilly PT, Kannan K, Kiarash R, Mason J, et al. Carnitine

48

palmitoyltransferase 1C promotes cell survival and tumor growth under conditions of

metabolic stress. 2011;1041–51.

42. Janakiram NB, Mohammed A, Rao C V. Role of lipoxins, resolvins, and other

bioactive lipids in colon and pancreatic cancer. Cancer Metastasis Rev [Internet]. 2011

Dec [cited 2014 Nov 11];30(3-4):507–23.

43. Cockbain a. J, Toogood GJ, Hull M a. Omega-3 polyunsaturated fatty acids for the

treatment and prevention of colorectal cancer. Gut. 2012;61(1):135–49.

44. Giros A, Grzybowski M, Sohn VR, Pons E, Fernandez-Morales J, Xicola RM, et al.

Regulation of colorectal cancer cell apoptosis by the n-3 polyunsaturated fatty acids

Docosahexaenoic and Eicosapentaenoic. Cancer Prev Res (Phila) [Internet].

2009;2(8):732–42.

45. Wang D, DuBois RN. Pro-inflammatory prostaglandins and progression of colorectal

cancer. 2009;267(2):197–203.

46. Zuo X, Shureiqi I. Eicosanoid profiling in colon cancer: emergence of a pattern.

Prostaglandins Other Lipid Mediat [Internet]. Elsevier Inc.; 2013 [cited 2014 Nov

11];104-105:139–43.

47. Wang D, Dubois RN. Eicosanoids and cancer. 2011;10(3):181–93.

48. Mal M, Koh PK, Cheah PY, Chan ECY. Ultra-pressure liquid chromatography/tandem

mass spectrometry targeted profiling of arachidonic acid and eicosanoids in human

colorectal cancer. Rapid Commun Mass Spectrom [Internet]. 2011 Mar 30 [cited 2014

Nov 11];25(6):755–64.

49. Bertagnolli MM. The potential of non-steroidal anti-inflammatory drugs (NSAIDs) for

colorectal cancer prevention. J Surg Oncol [Internet]. 2003 Nov [cited 2014 Nov

11];84(3):113–9.

50. Dennis MK, Field AS, Burai R, Ramesh C, Whitney K, Bologa CG, et al. NIH Public

Access. 2012;127(24):358–66.

51. Shureiqi I, Jiang W, Zuo X, Wu Y, Stimmel JB, Leesnitzer LM, et al. The 15-

lipoxygenase-1 product 13-S-hydroxyoctadecadienoic acid down-regulates PPAR-delta

to induce apoptosis in colorectal cancer cells. Proc Natl Acad Sci U S A.

2003;100:9968–73.

52. Yuri M, Sasahira T, Nakai K, Ishimaru S, Ohmori H, Kuniyasu H. Reversal of

expression of 15-lipoxygenase-1 to cyclooxygenase-2 is associated with development

of colonic cancer. Histopathology [Internet]. 2007 Oct [cited 2014 Nov 11];51(4):520–

7.

53. Chen JK, Capdevila J, Harris RC. Cytochrome p450 epoxygenase metabolism of

arachidonic acid inhibits apoptosis. Mol Cell Biol [Internet]. 2001;21(18):6322–31.

49

54. Wang D, DuBois RN. Epoxyeicosatrienoic acids: A double-edged sword in

cardiovascular diseases and cancer. J Clin Invest. 2012;122(1):19–22.

55. Yamashita A, Hayashi Y, Nemoto-Sasaki Y, Ito M, Oka S, Tanikawa T, et al.

Acyltransferases and transacylases that determine the fatty acid composition of

glycerolipids and the metabolism of bioactive lipid mediators in mammalian cells and

model organisms. Prog Lipid Res [Internet]. Elsevier Ltd; 2014;53(1):18–81.

56. Fatty Acid, Triglyceride, Phospholipid Synthesis and Metabolism [Internet]. [cited

2015 Nov 29].

57. Hishikawa D, Hashidate T, Shimizu T, Shindou H. Diversity and function of

membrane glycerophospholipids generated by the remodeling pathway in mammalian

cells. J Lipid Res [Internet]. 2014;55:799–807.

58. Hermansson M, Hokynar K, Somerharju P. Mechanisms of glycerophospholipid

homeostasis in mammalian cells. Prog Lipid Res [Internet]. Elsevier Ltd;

2011;50(3):240–57.

59. Athenstaedt K, Daum G. Phosphatidic acid, a key intermediate in lipid metabolism. Eur

J Biochem. 1999;266(August):1–16.

60. Duncan RE, Jaworski K. Triacylglycerol metabolism in adipose tissue. 2009;2(2):229–

37.

61. Yamashita A, Hayashi Y, Matsumoto N, Nemoto-Sasaki Y, Oka S, Tanikawa T, et al.

Glycerophosphate/Acylglycerophosphate Acyltransferases. Biology (Basel) [Internet].

2014;3:801–30.

62. Fujimoto N, Inoue K, Hayashi Y, Yuasa H. Glycerol uptake in HCT-15 human colon

cancer cell line by Na(+)-dependent carrier-mediated transport. Biol Pharm Bull

[Internet]. 2006;29(1):150–4.

63. Fujimoto N, Inoue K, Ohgusu Y, Hayashi Y, Yuasa H. Enhanced uptake of glycerol by

butyrate treatment in HCT-15 human colon cancer cell line. Drug

MetabPharmacokinet. 2007;22(3):195–8.

64. Agarwal AK, Garg A. Enzymatic activity of the human 1-acylglycerol-3-phosphate-O-

acyltransferase isoform 11: upregulated in breast and cervical cancers. J Lipid Res

[Internet]. 2010;51(8):2143–52.

65. Angelica MD, Fong Y. Colorectal cancer cells – proliferation, survival and invasion by

lysophosphatidic acid. October. 2008;141(4):520–9.

66. Zhang H, Bialkowska A, Rusovici R, Chanchevalap S, Shim H, Katz JP, et al.

Lysophosphatidic acid facilitates proliferation of colon cancer cells via induction of

Kr??ppel-like factor 5. J Biol Chem. 2007;282(21):15541–9.

67. Mansilla F, da Costa K-A, Wang S, Kruhøffer M, Lewin TM, Ørntoft TF, et al.

50

Lysophosphatidylcholine acyltransferase 1 (LPCAT1) overexpression in human

colorectal cancer. 2008;42(1):157–62.

68. Shida D, Watanabe T, Aoki J, Hama K, Kitayama J, Sonoda H, et al. Aberrant

expression of lysophosphatidic acid (LPA) receptors in human colorectal cancer. Lab

Invest. 2004;84(10):1352–62.

69. Tsukahara T, Haniu H, Matsuda Y. Cyclic phosphatidic acid induces G0/G1 arrest,

inhibits AKT phosphorylation, and downregulates cyclin D1 expression in colorectal

cancer cells. Cell Mol Biol Lett [Internet]. 2015;20(1):1–10.

70. Kurabe N, Hayasaka T, Ogawa M, Masaki N, Ide Y, Waki M, et al. Accumulated

phosphatidylcholine (16:0/16:1) in human colorectal cancer; possible involvement of

LPCAT4. Cancer Sci. 2013;104(10):1295–302.

71. Li S, Guo B, Song J, Deng X, Cong Y, Li P, et al. Plasma choline-containing

phospholipids: Potential biomarkers for colorectal cancer progression. Metabolomics.

2013;9(1):202–12.

72. Akinleye A, Avvaru P, Furqan M, Song Y, Liu D. Phosphtidylinositol 3-kinase (PI3K)

Inhibitors As Cancer Therapeutics. 2013;1–17.

73. Chong ML, Loh M, Thakkar B, Pang B, Iacopetta B, Soong R. Phosphatidylinositol-3-

kinase pathway aberrations in gastric and colorectal cancer: Meta-analysis, co-

occurrence and ethnic variation. Int J Cancer. 2014;134(5):1232–8.

74. Yu M, Grady WM. Therapeutic targeting of the phosphatidylinositol 3-kinase signaling

pathway: novel targeted therapies and advances in the treatment of colorectal cancer.

Ther Adv Gastroenterol [Internet]. 2012;5(5):319–37.

75. Iida Y, Sunami E, Yamashita H, Hiyoshi M. Expression in Colorectal Cancer

Correlates with Tumor Invasion and Hematogenous Metastasis. 2015;1464:1459–64.

76. Panel V, Boëlle P-Y, Ayala-Sanmartin J, Jouniaux A-M, Hamelin R, Masliah J, et al.

Cytoplasmic phospholipase A2 expression in human colon adenocarcinoma is

correlated with cyclooxygenase-2 expression and contributes to prostaglandin E2

production. Cancer Lett [Internet]. 2006 Nov 18 [cited 2014 Nov 11];243(2):255–63.

77. Buhmeida A, Bendardaf R, Hilska M, Laine J, Collan Y, Laato M, et al. PLA2 (group

IIA phospholipase A2) as a prognostic determinant in stage II colorectal carcinoma.

Ann Oncol [Internet]. 2009 Jul 1 [cited 2014 Nov 11];20(7):1230–5.

78. Denizot Y, Gainant A, Guglielmi L, Bouvier S, Cubertafond P, Mathonnet M. Tissue

concentrations of platelet-activating factor in colorectal carcinoma: inverse

relationships with Dukes’ stage of patients. Oncogene [Internet]. 2003 Oct 16 [cited

2014 Nov 11];22(46):7222–4.

79. Denizot Y, Truffinet V, Bouvier S, Gainant A, Cubertafond P, Mathonnet M. Elevated

plasma phospholipase A2 and platelet-activating factor acetylhydrolase activity in

51

colorectal cancer. Mediators Inflamm [Internet]. 2004;13(1):53–4.

80. Satow R, Hirano T, Batori R, Nakamura T, Murayama Y, Fukami K. Phospholipase

Cδ1 induces E-cadherin expression and suppresses malignancy in colorectal cancer

cells. Proc Natl Acad Sci U S A [Internet]. 2014;111(37):13505–10.

81. Gomez-Cambronero J. Phosphatidic acid, phospholipase D and tumorigenesis. Adv

Biol Regul [Internet]. The Author; 2014;54(1):197–206.

82. Zhang X, Zhao X-W, Liu D-B, Han C-Z, Du L-L, Jing J-X, et al. Lipid levels in serum

and cancerous tissues of colorectal cancer patients. World J Gastroenterol [Internet].

2014;20(26):8646–52.

83. Otani T, Iwasaki M, Ikeda S, Kozu T, Saito H, Mutoh M, et al. Serum triglycerides and

colorectal adenoma in a case-control study among cancer screening examinees (Japan).

Cancer Causes Control [Internet]. 2006 Dec [cited 2014 Nov 11];17(10):1245–52.

84. Kurek K, Bart BA, Piotrowska DM, B PW, Ma A, Chabowska B, et al. Metabolism ,

Physiological Role , and Clinical Implications of Sphingolipids in Gastrointestinal

Tract. 2013;2013.

85. Gault CR, Obeid LM, Hannun Y a. An overview of sphingolipid metabolism: From

synthesis to breakdown. Adv Exp Med Biol. 2010;688:1–23.

86. Furuya H, Shimizu Y, Kawamori T. Sphingolipids in cancer. Cancer Metastasis Rev

[Internet]. 2011 Dec [cited 2014 Nov 11];30(3-4):567–76.

87. Schnaar RL, Suzuki A, Stanley P. Glycosphingolipids [Internet]. Cold Spring Harbor

Laboratory Press; 2009 [cited 2015 Nov 15].

88. Duan R-D, Nilsson A. Metabolism of sphingolipids in the gut and its relation to

inflammation and cancer development. Prog Lipid Res [Internet]. Elsevier Ltd; 2009

Jan [cited 2014 Nov 6];48(1):62–72.

89. Kawamori T, Kaneshiro T, Okumura M, Maalouf S, Uflacker A, Bielawski J, et al.

Role for sphingosine kinase 1 in colon carcinogenesis. FASEB J [Internet].

2009;23(2):405–14.

90. Modrak DE. Sphingolipid targets in cancer therapy. Mol Cancer Ther [Internet].

2006;5(2):200–8.

91. Hertervig E, Nilsson A, Nyberg L, Duan RD. Alkaline sphingomyelinase activity is

decreased in human colorectal carcinoma. Cancer. 1997;79:448–53.

92. Dillehay DL, Webb SK, Andalfred ES, Merrill H. Biochemical and Molecular Roles of

Nutrients Dietary Sphingomyelin Inhibits Colon Cancer in CF1 Mice. J Nutr.

1994;(November 1993):615–20.

93. Selzner M, Bielawska a, Morse M a, Rüdiger H a, Sindram D, Hannun Y a, et al.

Induction of apoptotic cell death and prevention of tumor growth by ceramide

52

analogues in metastatic human colon cancer. Cancer Res [Internet]. 2001;61(3):1233–

40.

94. Dahiya R, Dudeja PK, Brasitus TA. Premalignant alterations in the glycosphingolipid

composition of colonic epithelial cells of rats treated with 1,2-dimethylhydrazine.

Cancer Res [Internet]. 1987;47(4):1031–5.

95. Sambasivarao S V. Tamoxifen magnifies therapeutic impact of ceramide in human

colorectal cancer cells independent of p53. 2013;18(9):1199–216.

96. Litvak DA, Bilchik AJ, Cabot MC. Modulators of ceramide metabolism sensitize

colorectal cancer cells to chemotherapy: a novel treatment strategy. J Gastrointest Surg

[Internet]. 2003;7(1):140–8.

97. Huang W-C, Chen C-L, Lin Y-S, Lin C-F. Apoptotic Sphingolipid Ceramide in Cancer

Therapy. J Lipids [Internet]. 2011;2011:1–15.

98. Don A, Lim X, Couttas T. Re-Configuration of Sphingolipid Metabolism by

Oncogenic Transformation. Biomolecules [Internet]. 2014;4(1):315–53.

99. AHN EH, SCHROEDER JJ. Induction of Apoptosis by Sphingosine, Sphinganine, and

C2-Ceramide in Human Colon Cancer Cells, but not by C2-Dihydroceramide.

Anticancer Res [Internet]. 2010;30(7):2881–4.

100. Pyne NJ, Pyne S. Sphingosine 1-Phosphate Is a Missing Link between Chronic

Inflammation and Colon Cancer. Cancer Cell [Internet]. Elsevier; 2013;23(1):5–7.

101. Visentin B, Vekich J a., Sibbald BJ, Cavalli AL, Moreno KM, Matteo RG, et al.

Validation of an anti-sphingosine-1-phosphate antibody as a potential therapeutic in

reducing growth, invasion, and angiogenesis in multiple tumor lineages. Cancer Cell.

2006;9(3):225–38.

102. Oskouian B, Sooriyakumaran P, Borowsky AD, Crans A, Dillard-Telm L, Tam YY, et

al. Sphingosine-1-phosphate lyase potentiates apoptosis via p53- and p38-dependent

pathways and is down-regulated in colon cancer. Proc Natl Acad Sci U S A.

2006;103(46):17384–9.

103. Pettus BJ, Kitatani K, Chalfant CE, Taha TA, Kawamori T, Bielawski J, et al. The

Coordination of Prostaglandin E2 Production by Sphingosine-1-phosphate and

Ceramide-1-phosphate. Mol Pharmacol. 2005;68(2):330–5.

104. Holst S, Stavenhagen K, Balog CI a, Koeleman C a M, McDonnell LM, Mayboroda O

a, et al. Investigations on aberrant glycosylation of glycosphingolipids in colorectal

cancer tissues using liquid chromatography and matrix-assisted laser desorption time-

of-flight mass spectrometry (MALDI-TOF-MS). Mol Cell Proteomics [Internet]. 2013

Nov [cited 2014 Nov 11];12(11):3081–93.

105. Kakugawa Y, Wada T, Yamaguchi K, Yamanami H, Ouchi K, Sato I, et al. Up-

regulation of plasma membrane-associated ganglioside sialidase (Neu3) in human

53

colon cancer and its involvement in apoptosis suppression. Proc Natl Acad Sci U S A

[Internet]. 2002;99(16):10718–23.

106. Siddiqui B, Whitehead J, Kim Y. Glycosphingolipids in human colonic

adenocarcinoma. J Biol Chem [Internet]. 1978;253(7):2168–75.

107. Morichika H, Hamanaka Y, Tai T, Ishizuka I. Sulfatides as a predictive factor of lymph

node metastasis in patients with colorectal adenocarcinoma. Cancer [Internet].

1996;78(1):43–7.

108. Cholesterol: Synthesis, Metabolism, Regulation [Internet]. [cited 2015 Nov 21].

109. Gorin A, Gabitova L, Astsaturov I. Regulation of cholesterol biosynthesis and cancer

signaling. Curr Opin Pharmacol [Internet]. Elsevier Ltd; 2012 Dec [cited 2014 Nov

11];12(6):710–6.

110. Jusakul A, Yongvanit P, Loilome W, Namwat N, Kuver R. Mechanisms of oxysterol-

induced carcinogenesis. Lipids Health Dis [Internet]. BioMed Central Ltd; 2011 Jan

[cited 2014 Dec 2];10(1):44.

111. Tosi MR, Tugnoli V. Cholesteryl esters in malignancy. Clin Chim Acta. 2005;359(1-

2):27–45.

112. de Gonzalo-Calvo D, López-Vilaró L, Nasarre L, Perez-Olabarria M, Vázquez T,

Escuin D, et al. Intratumor cholesteryl ester accumulation is associated with human

breast cancer proliferation and aggressive potential: a molecular and

clinicopathological study. BMC Cancer [Internet]. 2015 Jan [cited 2016 Mar

22];15:460.

113. Caruso MG, Notarnicola M. Biochemical changes of mevalonate pathway in human

colorectal cancer. Anticancer Res [Internet]. 2005;25(5):3393–7.

114. Messa C, Cavallini A, Bifulco M, Tecce MF, Eletto D, Di Leo A, et al. Higher farnesyl

diphosphate synthase activity in human colorectal cancer inhibition of cellular

apoptosis. Oncology [Internet]. 2004;67(5-6):351–8.

115. Notarnicola M, Messa C, Pricci M, Guerra V, DF A, Montemurro S, et al. Up-

regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity in left-sided

human colon cancer. Anticancer Res [Internet]. 2004;24(6):3837–42.

116. Caruso MG, Osella AR, Notarnicola M, Berloco P, Leo S, Bonfiglio C, et al.

Prognostic value of low density lipoprotein receptor expression in colorectal

carcinoma. Oncol Rep [Internet]. Spandidos Publications; 1998 Jul 1 [cited 2015 Nov

27];5(4):927–57.

117. Bengtsson E, Nerjovaj P, Wangefjord S, Nodin B, Eberhard J, Uhlen M, et al. HMG-

CoA reductase expression in primary colorectal cancer correlates with favourable

clinicopathological characteristics and an improved clinical outcome. Diagn Pathol

[Internet]. 2014;9:78.

54

118. Cruz PMR, Mo H, McConathy WJ, Sabnis N, Lacko AG. The role of cholesterol

metabolism and cholesterol transport in carcinogenesis: A review of scientific findings,

relevant to future cancer therapeutics. Front Pharmacol. 2013;4 SEP(September):1–7.

119. Bardou M, Barkun A, Martel M. Effect of statin therapy on colorectal cancer. Gut.

2010;59(11):1572–85.

120. Al-Haidari A a, Syk I, Thorlacius H. HMG-CoA reductase regulates CCL17-induced

colon cancer cell migration via geranylgeranylation and RhoA activation. Biochem

Biophys Res Commun [Internet]. Elsevier Inc.; 2014;446(1):68–72.

121. Thosani N, Thosani SN, Kumar S, Nugent Z, Jimenez C, Singh H, et al. Reduced Risk

of Colorectal Cancer With Use of Oral Bisphosphonates: A Systematic Review and

Meta-Analysis. J Clin Oncol [Internet]. 2013;31(5):623–30.

122. Smith B, Land H. Anti-cancer activity of the cholesterol exporter ABCA1 gene.

2012;2(3):580–90.

123. Watson MA, Gay L, Stebbings WSL, Speakman CTM, Bingham SA, Loktionov A.

Apolipoprotein E gene polymorphism and colorectal cancer: gender-specific

modulation of risk and prognosis. Clin Sci (Lond) [Internet]. Portland Press Limited;

2003 May 1 [cited 2016 Mar 14];104(5):537–45.

124. Accioly MT, Pacheco P, Maya-Monteiro CM, Carrossini N, Robbs BK, Oliveira SS, et

al. Lipid bodies are reservoirs of cyclooxygenase-2 and sites of prostaglandin-E2

synthesis in colon cancer cells. Cancer Res. 2008;68(6):1732–40.

125. Scalfi-Happ C, Udart M, Hauser C, Rück A. Investigation of lipid bodies in a colon

carcinoma cell line by confocal Raman microscopy. Med Laser Appl [Internet].

Australasian Society of Cardiac and Thoracic Surgeons and The Cardiac Society of

Australia and New Zealand; 2011;26(4):152–7.

126. Foster JM. Transomics : Integrating core ‘ omics ’ concepts. Cambridge: University of

Cambridge; 2012.

127. Porta C, Paglino C, Mosca A. Targeting PI3K/Akt/mTOR Signaling in Cancer. Front

Oncol [Internet]. 2014 Jan [cited 2016 Mar 11];4:64.

55

8 Appendix

Dukes classificiation

Stage A: invading submucosa

Stage B1: invading into muscularis propria

Stage B2: invading through muscularis propria

Stage B3: equals T4N0M0

Stage C1: as B1 but with lymph node invasion

Stage C2: as B2 but with lymph node invasion

Stage C3: as B3 but with lymph node invasion

Stage D: distant metastasis

TNM classification

T=tumor

T0= no primary tumor present

Tis: tumor in situ (located intra-epithelial or invading lamina propria)

T1: tumor invading submucosa

T2: tumor invading muscularis propria

T3: tumor invading subserosa or non-peritoneal pericolic or perirectal

structures.

T4a: tumor invading visceral peritoneal surface

T4b: tumor invading or adherent to surrounding organs

Tx: primary tumor cannot be identified

N: lymph node invasion

N0: no lymph node metastasis

N1a: metastasis in one regional lymph node

N1b: metastasis in 2-3 regional lymph nodes

N2a: metastasis in 4-6 regional lymph nodes

N2b: metastasis in 7 or more regional lymph nodes

Nx: lymph node invasion cannot be identified

M= distant metastasis

M0: no distant metastasis

56

M1a: distant metastasis to organ or other site

M1b: distant metastasis to multiple organs, sites or peritoneum

Mx: distant metastasis cannot be identified

TNM disease stages

Stage 1: T1-T2N0M0

Stage 2A: T3, N0, M0

Stage 2B: T4a, N0, M0

Stage 2C: T4b, N0, M0

Stage 3A: T1-T2, N1, M0 and T1, N2a, M0

Stage 3B: T1-T2, N2b, M0; T2-T3, N2a, M0 and T3-T4a, N1, M0

Stage 3C: T3-T4a, N2b, M0 and T4b, N1-N2, M0 and T4a, N2a, M0

Stage 4A: M1a with any T and any N

Stage 4B: M1b with any T and any N