Download Bachelor thesis Jocelyn De Wever

Document related concepts
Transcript
Bachelor thesis
Mutation screening of the exonuclease domain of POLD1
and POLE genes in patients with multiple adenomas and
colorectal cancer
Institute of Human Genetics at the Medical University
Graz
Univ.-ass. Mag. Dr.rer.nat. Ellen Heitzer
Jocelyn De Wever
Department GEZ-LA
Bachelor Biomedical Laboratory technology
Specialization Pharmaceutical and biomedical laboratory
technology 2013-2014
“The laws of genetics apply even if you refuse to learn them.”
- Allison Plowden
The study of inheritable Colorectal Cancer by genetic mutation detection of known predisposition
genes and exonuclease domain mutations in POLD1 and POLE.
Acknowledgement
First of all, I want to thank Doctor Ellen Heitzer who has helped, advised and encouraged me
throughout the course of this thesis. She reviewed my manuscripts thoroughly and without
her support the thesis would not have become what it is now. Even though she had a busy
agenda she always made time if necessary.
I also want to thank the colleagues, Verena Rupp, Ricarda Graf, Sumitra Mohan, Katharina
Woisetschläger etcetera, who have showed me some techniques and to who I always could
ask any question or feedback. Also I want to thank the colleagues from diagnostic who were
always supportive and always were happy to give tips during my work. All these people have
shown me how a real working environment runs and how important collaborations and
communications can be.
Of course I also want to thank my parents, who supported me with whole their heart and
effort to make this Erasmus possible.
And last but not least, my friend Joren Remue who was not here in person but also showed
from a distant to be supportive.
Organigram
Function
Name
Head, laboratory manager
Univ.-Prof. Dr. med. Michael Speicher
Deputy head
Ao. Univ.-Prof. DDr. Erwin Petek
Deputy laboratory manager
Ao. Univ. Prof. Dr. Klaus Wagner
Quality manager
Michael Andrée
Deputy quality manager
Ao. Univ-Prof. DDr. Erwin Petek
Routine
diagnostics
management
Prenatal/postnatal group
group
executive
Ao. Univ-Prof. Dr. med. Peter Kroisel
Dr. med. Werner Emberger
Tumor cytogenetics
Dr. med. Werner Emberger
Dr. Kristina Orendi
Evaluator
Ao. Univ.-Prof. Dr. med. Peter Kroisel
Dr. med. Werner Emberger
FISH pre-/postnatal
Ao. Prof. DDR. Erwin Petek
Assoz. Prof. Dr. led. Jochen Geigl
Array CGH
Ao. Univ Prof. Dr. Klaus Wagner
Univ.-Prof. Dr. med. Michael Speicher
Approved signatories of the GCG
All reports form moleculargenetic diagnostics
Ao. Univ. Prof. Dr. Klaus Wzgner
Priv. Doz. Mag. Dr. Thomas Schwarzbraun
Dr. Ellen Heitzer
Ao. Univ. Prof. DDr. Erwin Petek
Approved signatories of the GCG
Mag.rer.nat. Dr. scient.med. Lisa
Ziegenfuss
Ao. Univ.-Prof. Dr. med. Peter Kroisel
All reports from cytogenetic diagnostics
Dr. med. Werner Amberger
Assoz. Prof. Dr. med. Jochen Geigl
Univ-Prof. Dr. med. Michael Speicher
Group leaders research
Assoz. Prof Dr. med. Jochen Geigl
Ao. Univ. Prof. DDr. Erwin Petek
Univ-Prof. Dr. med. Michael Speicher
Assoz. Prof. Dr. Christian Windpassinger
Patient care
Ao. Univ.-Prof Dr. med. Peter Kroisel
Univ-Prof. Dr. med. Michael Speicher
Secretary of institute - head Prof. Speicher
Ida Tieber
Secretary for outpatients - head Prof. Speicher
Susanne Berger, Stefanie Schalk,
Elisabeth Fuchs, Elisabeth Schuller
Marita Fuchs, Katharina Ledinek
Karina Lichtenegger,
Safety representative
Ao. Univ.-Prof. Dr. med. Peter Kroisel
Priv. Doz. Mag. Dr. Thomas Schwarzbraun
Hygiene specialis
Assoz. Prof. Dr. med. Jochen Geigl
EDV - representative
Ao. Univ Prof. Dr. Klaus Wagner
Deputy EDV - representative
Michael Andrée
Institute speaker - Head Prof. Speicher
Mag. Maria Langer-Winter
Ofner-
Table of Content
1. Abstract .................................................................................................................................. 1
2. Objective ................................................................................................................................ 3
3. List of abbreviations .............................................................................................................. 4
4. Background and status of research....................................................................................... 7
4.1 Hallmarks of Cancer ............................................................................................................ 7
4.1.1 Somatic versus germline mutations ........................................................................................7
4.1.2 Oncogenes ...............................................................................................................................9
4.1.3 Tumor suppressor genes ...................................................................................................... 10
4.1.4 DNA polymerases ................................................................................................................. 11
4.1.4.1 DNA replication............................................................................................................. 12
4.1.4.2 DNA replication fidelity ................................................................................................. 13
4.1.4.3 Exonuclease domain mutations (EDM) ........................................................................ 14
4.1.5 Caretaker genes.................................................................................................................... 15
4.1.5.1 Mismatch repair genes (MMR) ..................................................................................... 15
4.1.5.2 Nucleotide excision repair (NER) & Base excision repair (BER) .................................... 15
4.1.5.3 Other stability genes ..................................................................................................... 16
4.1.5.4 Mutations in stability genes ......................................................................................... 16
4.2 Colorectal Cancer ...............................................................................................................18
4.2.1 Sporadic Versus familial cases.............................................................................................. 19
4.2.2 Penetrance of hereditary colorectal cancer syndromes ....................................................... 20
4.2.3 Genetic pathways of CRC...................................................................................................... 20
4.2.4 Familial adenomatous polyposis (FAP)................................................................................. 21
4.2.4.1 Adenomatous polyposis coli (APC)................................................................................ 21
4.2.4.2 Affected pathways ........................................................................................................ 22
4.2.4.3 Attenuated FAP (AFAP) & MYH associated polyposis (MAP)........................................ 22
4.2.5 Lynch Syndrome (LS) ............................................................................................................. 23
4.2.5.1 Pathogenesis of LS ........................................................................................................ 23
4.2.5.2 Mutation spectra in LS .................................................................................................. 24
4.2.5.3 Tumorigenesis and presence of MSI in LS ..................................................................... 25
4.2.6 Hamartomatous polyposis syndromes ................................................................................. 25
4.2.6.1 Peutz-Jeghers syndromes (PJS) ..................................................................................... 25
4.2.6.2 Juvenile polyposis syndromes (JPS)............................................................................... 25
4.2.6.3 PTEN hamartomatous tumor syndromes (PHTS).......................................................... 26
4.2.7 Polymerase proofreading associated polyposis (PPAP) ....................................................... 26
4.2.8 Short summary of genes associated with a high susceptibility of CRC ................................ 28
4.3 Genetic testing ...................................................................................................................28
4.3.1 Routine mutation screening of CRC ...................................................................................... 29
4.3.1.1 Germline detection of CRC ............................................................................................ 29
4.3.1.2 Genetic testing for FAP ................................................................................................. 29
4.3.1.3 Genetic testing for LS .................................................................................................... 30
4.3.2. The search of more predisposition genes ............................................................................ 32
5. Experimental Research ........................................................................................................ 33
5.1 Material and Method .........................................................................................................33
5.1.1 Patient samples .................................................................................................................... 33
5.1.2 DNA quantification ............................................................................................................... 33
5.1.2.1 Reagents ....................................................................................................................... 33
5.1.2.2 Protocol......................................................................................................................... 34
5.1.3 Polymerase chain reaction (PCR).......................................................................................... 35
5.1.3.1 Reagents ....................................................................................................................... 36
5.1.3.2 Primer design ................................................................................................................ 36
5.1.3.3 Protocol PCR reaction ................................................................................................... 37
5.1.4 Gel electrophoresis and Agilent 2100 Bioanalyzer ............................................................... 38
5.1.4.1 Reagents ....................................................................................................................... 38
5.1.4.2 Protocol......................................................................................................................... 39
5.1.5 Sanger sequencing................................................................................................................ 39
5.1.5.1 Reagents ....................................................................................................................... 41
5.1.5.2 Protocol......................................................................................................................... 41
5.1.6 Statistics ............................................................................................................................... 43
5.2 Results ...............................................................................................................................44
5.2.1 DNA quantification ............................................................................................................... 44
5.2.2 PCR quality check.................................................................................................................. 45
5.2.2.1 Gel electrophoresis ....................................................................................................... 45
5.2.2.2 Agilent 2100 Bioanalyzer .............................................................................................. 46
5.2.3 Sanger sequencing................................................................................................................ 47
5.2.4 Patient characteristics .......................................................................................................... 47
5.2.5 Mutation detection of patient samples ................................................................................ 47
5.2.5.1 overview mutations found in patients .......................................................................... 48
5.2.5.2 putative mutations ...................................................................................................... 51
5.2.5.3 Single nucleotide polymorphisms (SNP) ....................................................................... 54
5.3 Discussion ..........................................................................................................................55
5.3.1 Technical considerations ...................................................................................................... 55
5.3.2 Interpretation of mutational screening results .................................................................... 56
6. Conclusion ............................................................................................................................ 64
7. References............................................................................................................................ 65
8. Appendix ................................................................................................................................. I
8.1 PCR quality check results from Agilent 2100 Bioanalyzer ....................................................... I
8.2 Sequencing results human genomic female DNA .................................................................. V
8.2.1 Sequencing results of POLD1 exons ........................................................................................ V
8.2.2 Sequencing results of POLE exons ......................................................................................... VI
8.3 Amino acid conservation .................................................................................................. VII
8.4 Detailed information of identified SNPs ............................................................................ VIII
8.4.1 SNP in POLD1 intron 9 ......................................................................................................... VIII
8.4.2 SNP in POLD1 intron 10 ....................................................................................................... VIII
8.4.3 SNP in POLD1 intron 13 ......................................................................................................... IX
8.4.4 SNP in POLE intron 9 .............................................................................................................. IX
8.4.4 SNP in POLE intron 12 ............................................................................................................ IX
8.4.4 SNP in POLE intron 13 ............................................................................................................. X
1. Abstract
The occurrence of cancer depends on the accumulation of mutations in genes caused by
external factors or by inheritance. In the last decades genetic research has grown to identify
these genetic lesions and their consequences. New screening methods led to early detection
of cancer, which increases the survival rates. Mutational screening of predisposition genes is
routinely performed in patients with early onset of cancer or a clear family history. One of the
best known inheritable cancers is colorectal cancer (CRC), which is the third most common
cancer worldwide. To date ten predisposition genes are screened for mutations for molecular
genetic diagnosis of heritable CRC syndromes. However, disease causing mutations are only
found in about 30% of patients suggesting that there may be other genes responsible for
hereditary colon cancer that have not been identified yet. Recently Palles et al. identified
“exonuclease domain mutations” (EDMs) in two genes POLD1 and POLE, encoding for
replicative polymerases, that predispose to "polymerase proofreading-associated polyposis"
(PPAP). PPAP is a disease characterised by multiple colorectal adenomas and carcinomas. The
mutational screening of these genes is not yet part of the routine molecular diagnostic of the
common CRC-associated tumor syndromes in Austria.
The focus of this thesis was a mutational screening using Sanger sequencing of exons
encoding the exonuclease domain – which is responsible for proofreading – of the POLD1
and POLE genes, respectively. In total, 59 patients, with a high penetrance CRC phenotype,
who underwent routine diagnostics for the most common high penetrance genes, but where
no causative mutations have been found, were included in this study. A total of two EDMs
were identified in the analysed patient cohort, one in POLD1 and one in POLE. The POLD1
mutation was a deletion of three nucleotides leading to a deletion of two amino acids and an
insertion of a new amino acid. The mutation has not been reported before. The POLE
mutation was a missense mutation that was listed in dbSNP, but without any allele
frequency. In both cases it remains to be elucidated whether the mutations are indeed
disease causing or rather neutral variants. Although both patients had an early onset of
cancer, they did not show clear family history of cancer. Therefore, segregation analysis of
the identified mutation will most likely not be informative. Taken together, further research
is needed to understand pathogenesis and functional impact of these mutations.
Nevertheless, inclusion of POLD1 and POLE in routine diagnostics for hereditary cancer
syndrome is worthy of consideration.
1
Samenvatting
Het voorkomen van kanker hangt af van het voorkomen van mutaties in genen veroorzaakt
door externe factoren of door erfelijkheid. In de laatste decennia is het belang van genetisch
onderzoek voor de identificatie van genetische alteraties en hun consequenties meer en meer
gestegen. Nieuwe screeningmethodes zorgden voor een vroege detectie van kanker,
waardoor de overlevingskansen van patiënten stijgen. Hierbij hoort het routineus screenen,
van patiënten met een vroegtijdige kanker of met een duidelijke familiale kankergeschiedenis
op mutaties in de predispositie genen. Colorectale kanker is één van de best gekende erfelijke
kankers, waarbij het de derde meest voorkomende kanker is wereldwijd. Vandaag de dag
worden, voor moleculaire genetische diagnose van erfelijke colorectale kanker syndromen,
tien predispositie genen gescreend op mutaties. Er worden echter slechts in 30% van deze
patiënten ziekmakende mutaties gevonden. Dit suggereert dat er mogelijks andere genen
betrokken zijn bij erfelijke colonkanker syndromen die nog niet zijn gedefinieerd. Recent
identificeerde Palles et al “exonuclease domein mutaties” (EDMs) in twee genen POLD1 en
POLE, coderend voor replicatieve polymerases, dewelke aanleg vertonen tot “polymerase
proofreading-geassocieerde polyposis” (PPAP). PPAP is een aandoening gekenmerkt door
multipele colorectale adenomen en carcinomen. De mutationele screening van deze genen
maken nog geen deel uit van de routineuze moleculaire diagnostiek voor veel voorkomende
colorectale kanker-geassocieerde tumor syndromen in Oostenrijk.
De focus van deze thesis ligt op het mutationeel screenen via Sanger sequencing van de
exonen, coderend voor de exonuclease domeinen van POLD1 en POLE genen. Deze
exonuclease domeinen zijn verantwoordelijk voor de proofreading capaciteit van DNA
polymerase δ en DNA polymerase ε. In totaal werden 59 patiënten met een hoog penetrant
CRC fenotype, welke na routineuze moleculaire diagnostiek van de veel voorkomende hoge
penetrante predispositie genen geen oorzakelijke mutaties vertoonden, geïncludeerd in deze
studie. In totaal werden twee EDMs, één in POLD1 en één in POLE, gedefinieerd en
geanalyseerd in deze populatie. De POLD1 mutatie hield een deletie van 3 nucleotiden in, die
resulteerde in een deletie van twee aminozuren en een insertie van één nieuw aminozuur.
Deze mutatie was na verder onderzoek nog niet eerder gedefinieerd. De POLE mutatie was
een missense mutatie, dewelke eerder werd vermeld in dbSNP, maar zonder enige data van
allel frequenties. Beide EDMs dienen verder onderzocht te worden om na te gaan of deze
inderdaad ziekmakend zijn of eerder neutrale varianten. Alhoewel beide betrokken patiënten
een vroegtijdige colorectale kanker vertoonden, hadden ze beiden geen duidelijke familiale
kankerhistorie. Segregatie analyse van de geïdentificeerde mutaties zal dus hoogst
waarschijnlijk niet informatief zijn. Al bij al, is verder onderzoek noodzakelijk voor het
identificeren van de pathogenese en functionele impact van deze mutaties. Desondanks dient
de inclusie van POLD1 en POLE genen in routineuze diagnostiek voor erfelijke kanker
syndromen worden overwogen.
2
2. Objective
Cancer is one of the leading causes of death worldwide accounting for 8.2 million deaths in
2012. It is predicted that the number of deaths will increase to 13.1 million in 2030 (WHO,
2008 & WHO, 2014). Cancer genetics is becoming an evolving field of research, where genetic
testing of high-penetrance tumor susceptibility genes helped identifying individuals early with
a high risk of cancer development. This can lead to an effective prevention and surveillance of
cancer (Pasche, 2010).
Research on “colorectal cancer” (CRC) is leading the way to understand cancer genetics and
the underlying mechanisms to a predisposition to CRC (de la Chapelle B., 2004). To date ten
predisposition genes are routinely screened for mutations to identify hereditary forms of
CRCs. However, only in about 30% of the patients with a phenotype of multiple, large
colorectal adenomas, early-onset CRC or “endometrial cancer” (EC), disease causing
mutations have been identified. In 70% of the patients, no germline mutations can be
identified despite a high a priori probability that the cancer may be attributed to a highpenetrance germline mutation. This does not necessarily mean that these patients and their
relatives do not harbour a predisposition in their genome that is responsible for their
increased risk of cancer development. It could be that the mutated gene is not known or has
not been identified yet. Therefore, there is a high need for identifying new predisposition
genes to increase the percentage of detection and to decrease the development of these
common CRC-associated tumor syndromes.
Recently, Palles et al identified germline “exonuclease domain mutations” (EDMs) in two
genes that encode for replicative polymerase genes, i.e. POLD1 and POLE predisposing to
"polymerase proofreading associated polyposis" (PPAP). PPAP is a disease characterised by
multiple colorectal adenomas and carcinomas (Palles et al., 2013 & Heitzer E. & Tomlinson I.,
2014). Mutational screening of these genes is not yet part of the routine molecular diagnostic
of the common CRC-associated tumor syndromes in Austria. However, at the Institute of
Human Genetics there is a cohort of patients with CRC available that already underwent
routine diagnostic for the most common high penetrance genes, i.e. APC or the mismatch
repair genes, but where no causative mutation has been found. In response to the findings
from Palles et al we aimed to screen the exonuclease domain of the POLD1 and POLE genes in
this cohort of patients. This in order to confirm that mutations in these genes are indeed
causative for the predisposition to CRC and to evaluate whether inclusion of these genes into
routine diagnostics might contribute to a higher detection rate of predisposing mutations.
3
3. List of abbreviations
AB:
Applied Biosystems
aCRC:
Advanced CRC
APC:
Adenomatous polyposis coli
BER:
Base-excision repair
BLM:
Bloom syndrome
BMP:
Bone morphogenetic protein
BMPR1A:
Bone morphogenetic protein receptor, type IA
Bp:
Base pairs
BR:
Broad Range
BRAF:
B- rapidly accelerated fibrosarcoma
BRCA 1:
Breast cancer 1
BRCA 2:
Breast cancer 2
BRRS:
Bannayan-Riley-Ruvalcaba syndrome
CE:
Capillary electrophorese
CCND1:
Cyclin-D1
CHEK2:
Checkpoint kinase 2
CIMP:
CpG island methylator phenotype
CIN:
Chromosomal instability
c-myc:
Cellular myelocyto-matosis
CRC:
Colorectal cancer
DGGE:
Denaturing gradient gel electrophoresis
DMSO:
Ditmethylsulfoxide
Df:
Degree of freedoms
dNTP:
Deoxyribonucleoside triphosphates
ddNTP:
2’, 3’ – dideoxyribonucleoside triphosphate
Ds:
Double stranded
EC:
Endometrial cancer
4
EDM:
Exonuclease domain mutation
Exo:
Exonuclease deficient
FAP:
Familial adenomatous polyposis
FWD:
Forward primer
GI:
Gastrointestinal
GREM1:
Gremlin 1, DAN family BMP antagonist
HNPCC:
Hereditary nonpolyposis colorectal cancer
HRT:
Hormone-replacement therapy
HS:
High sensitivity
H0:
Null hypothesis
H1:
Alternative hypothesis
Ig:
Immunoglobulin
IHC:
Immunohistochemistry
KRAS:
Kirsten rat sarcoma
LOH:
loss of heterozygosity
LS:
Lynch syndrome
MAF:
Minor allele frequency
MAP:
MYH associated polyposis
MLH1:
Mut L homologue 1
MMR:
Mismatch repair
MSH2:
Human mut S homologue 2
MSH6:
Human mut S homologue 6
MSI:
Microsatellite instability
MUTYH or MYH:
Mut y homologue,
ND:
NanoDrop
NER:
Nucleotide-excision repair
NSAID:
Non-steroidal inflammatory drugs
NTC:
Negative control
5
PC:
Positive control
PCNA:
Proliferating cell nuclear antigen
PCR:
Polymerase chain reaction
PHTS:
PTEN hamartoma tumor syndrome (PHTS),
PJS:
Peutz-Jeghers syndrome
PMS2:
Postmitotic segregation, a Mut L homologue
POLD1:
DNA polymerase Delta 1
POLE:
DNA polymerase ε
PPAP:
Polymerase proofreading associated polyposis
PTEN:
Phosphatase and tensin homolog
PTT:
Protein truncation testing
Ras:
Rat sarcoma
Rcf:
Rotational centrifuge force
REV:
Reverse primer
Rpm:
Rounds per minute
SMAD4 /DPC4:
Deleted in pancreatic carcinoma in locus 4
SNP:
Short nucleotide polymorphism
Ss:
Single stranded
SSCP:
Single stranded conformational polymorphism
STK11-LKB1:
Serine threonine kinase
TBE:
Tris-Borat-EDTA
Tcf:
T cell factor
TCGA:
The Cancer Genome Atlas
TGF-β:
Transforming growth factor β
TP53:
Tumor Protein 53
UV:
Ultraviolet
UV:
Unknown Variant
WHO:
World health organisation
6
4.
Background and status of research
4.1
Hallmarks of cancer
Cancer is one of the most common diseases in humans and one of the leading causes of death
worldwide, accounting for 7.6 million deaths (13% of all deaths) in 2008 and 8.2 million deaths
in 2012. It is predicted that the number of deaths will increase to 13.1 million in 2030 (Cancer,
2013 & WHO, 2008 & WHO, 2014).
Cancer, also called the disease of
abnormal gene function, is caused by
genetic alterations of one or more genes
in one single cell resulting in an abnormal
growth of the cell (Figure 1) (Cancer, 2013
& Grady W.M. & Carethers J.M., 2008 &
Hereditary and cancer, 2013). The tumor
entity is determined by the type of cells
that are involved in the transformation Figure 1. Cancer is a results of an abnormal growth and
proliferation of one cell (Waters J.D., 2010).
resulting in neoplasms (Davis, 2011).
When cells divide uncontrollably they form a mass of tissue called neoplasms. Neoplasms can
be benign, pre-malignant (carcinoma in situ) or lead to the formation of malignant tumors
(cancer). Carcinoma in situ do not invade neighbouring tissue or metastasize and grow slowly
compared to malignant tumors that are capable of spreading by invasion, also called
metastasis (What is Carcinoma in situ, 2014 & What is a tumor, 2012 & What is Cancer? What
Causes Cancer? 2013). Tumors are furthermore able to form their own blood vessel, by a
process called angiogenesis, facilitating the provision of nutrients.
Metastasis is caused by the movement of tumor cells through the body by the lymph and
blood system, metastases account for 90% of the human cancer deaths (Mehlen P. & Puisieux
A., 2006 & Fidler I.J.,2003). Therefore, cancer research is very import to understand the
pathogenesis of cancer at systemic, cellular and molecular levels, to improve the survival
chance of individuals (Fidler I.J., 2003).
4.1.1 Somatic versus germline mutations
Cancer is a genetic disease, so the occurrence of a cancer is driven by epigenetic and genetic
abnormalities (Papou E.P. & Ahuja N., 2010 & de la Chapelle A., 2004). The epigenetic
abnormalities are caused by the environment and the lifestyle of an individual including
tobacco use, overweight and hormonal factors (Peto J., 2001). This may explain the rising
incidence of cancer with age, caused by the accumulation of mutations (Cancer, 2013 &
7
Knudson A.G., 2001). Most cancer-causing mutations are somatic, i.e. mutations occurring in
one cell of a specific tissue during a person’s lifetime which are not inherited.
However, 5-10% of all cancers are inherited, caused by predisposing mutations that affect the
germline and are inherited from a parent (Hereditary and cancer, 2013 & de la Chapelle A.,
2004 & Garber J.E & Offit K., 2004). Such inherited genetic mutations increase a person’s risk
of developing cancer owing to the fact that every cell of the body is affected from a
predisposing mutation. These mutations mainly occur in tumor suppressor genes and include
point mutations, deletions or insertions in one allele of a gene. Therefore hereditary cancer
syndromes are inherited in an autosomal manner (Vogelstein B. & Kinzler W.K., 2004 & de la
Chapelle A., 2004 & Garber J.E. & Offit K., 2005 & Hereditary and cancer, 2013). However,
according to Knudson’s “two-hit hypothesis”, for tumor initiation, also the unaffected allele
needs to be inactivated. Therefore, heterozygosity alone is not a sufficient condition for the
development of cancer, proven by the fact that penetrance is incomplete for cancer mutations
in the germline (Knudson A.G., 2001 & Knudson A.G., 1996). To date numerous genes have
been identified that predispose to cancer and are routinely analysed in the clinical diagnostics
(Peto J., 2001 & de la Chapelle A., 2004).
Cell division is a physiological process and under physiological conditions proliferation and
programmed cell death is balanced. This balance is regulated by numerous genes, whereby
alterations in these genes are associated with cancer (Vogelstein B. & Kinzler W.K., 2004 &
Malumbres M. & Barbacid M., 2001). The concerning genes are classified in two types i.e.
oncogenes and tumor suppressor genes, whereas the late include gatekeeper, caretaker and
landscaper genes. When one or more of these genes, involved in cellular responses to
oncogenic stimuli such as DNA repair or apoptosis, are altered, the cell cycle regulation can be
hampered. Further accumulation of genetic changes can occur and cause tumorigenesis
(Vogelstein B. & Kinzler W.K., 2004 & What Is Cancer? What Causes Cancer?, 2013 &
Fearnhead et al., 2002 & Michor et al., 2004). Other genes mutated in cancer include those
that inactivate the apoptotic pathways, induce genomic instability and promote angiogenesis
(Malumbres M. & Barbacid M., 2001).
Tumor suppressor genes constitute the largest group of cloned hereditary cancer genes
(Knudson A.G., 2002). Even though, it has been noted that the most common forms of
hereditary cancer predisposition, leading to breast and colon cancers, are caused by inherited
mutations of stability genes rather than tumor suppressor genes or oncogenes (Vogelstein B.
& Kinzler W.K., 2004). Although, recently mutations in polymerases have been identified that
can have a function in tumorigenesis (Heitzer E. & Tomlinson I., 2014 & Palles et al., 2013 &
Church et al., 2013).
8
4.1.2 Oncogenes
Oncogenes are responsible for cell proliferation, growth, differentiation, the control of the cell
cycle and apoptosis. These genes encode for grow factors, grow factor receptors, signal
transducers, apoptosis regulators and transcription factors (Vogelstein B. & Kinzler W.K., 2004
& Papou E.P. & Ahuja N., 2010).
Mutations in oncogenes are typically dominant and include chromosomal translocations,
intragenic mutations or gene amplifications resulting in a constitutively active pathway (Papou
E.P. & Ahuja N., 2010 & Vogelstein B. & Kinzler W.K., 2004). The normally regulated genes are
called proto-oncogenes and are only called oncogenes after an activating mutation of the
genes. Owing to dominance only one allele needs to be activated to cause uncontrolled
growth and differentiation of the cell, similar to a stuck accelerator that keeps a car going on
(Papou E.P. & Ahuja N., 2010 & Vogelstein et al, 2010).
The conversion from proto-oncogene
to oncogene can occur in two forms,
a quantitative form and a qualitative
form, respectively (Figure 2).
Quantitative forms of oncogene
activation include the amplification or
translocation of the gene to an active
site of a chromosome, thereby
bringing the genes under the control
of a strong promoter. For example, c- Figure 2. Conversion from proto-oncogene to oncogene occurs in 2
MYC can be translocated to forms, the qualitative form and the quantitative form, respectively.
Causing the occurrence of a fusion protein or aberrant expression of
“immunoglobulin” (Ig) loci on normal protein, resulting in a high expression of the gene (Hoffbrand,
chromosome 2 or 22 resulting in an A.V. & Moss, P.A.H., 2011).
upregulated expression. In contrast qualitative changes of oncogenes include point mutations
or production of novel proteins out of chimeric genes, leading to constitutionally activated
proteins (Pappou E.P. & Ahuja N., 2010 & Chan V.T.W. & McGee J.O'D, 1987).
The first discovered proto-oncogene family RAS is activated by qualitative changes, i.e. point
mutations in distinct hot spot regions that are found in approximately 50% of all colon cancers
(Papou E.P. & Ahuja N., 2010 & Fearnhead et al., 2002). The RAS genes encode for G-proteins
in the plasma membrane that initiate a protein kinase cascade for the activation of nuclear
transcription factors that alters the gene expression by either inducing or repressing it.
Mutations in the RAS genes lead to an independent activation of the downstream signalling
transduction system, causing stimulated proliferation, the regulation of multiple growth and
survival signals and inhibition of apoptosis (Papou E.P. & Ahuja N., 2010). Another example is
9
mutations in the BRAF gene, encoding for a serine/threonine kinase located immediately
downstream of RAS (Papou E.P. & Ahuja N., 2010 & Vogelstein B. & Kinzler W.K., 2004). Both
BRAF and KRAS participate in the same proliferation-supporting signalling pathways, whereby
KRAS or BRAF mutations occur mutually exclusive in CRCs (Lynch et al., 2009). The BRAF
mutations occur in approximately 15% of CRCs and have been linked to non-Lynch Syndrome
associated tumor and sporadic microsatellite unstable CRCs, whereas KRAS mutations are
frequently observed in “Lynch syndrome” (LS) associated tumor CRCs (Lynch et al., 2009 &
Kohlmann, W. & Gruber, S.B., 2012).
4.1.3 Tumor suppressor genes
Tumor suppressor genes are mainly involved in the control of cell division. Mutant tumor
suppressor alleles are recessive, therefore, unlike oncogenes, both alleles have to be
inactivated to drive tumorigenesis. When only one allele for the gene is damaged, the second
can still produce the correct protein. It is analogous to a dysfunctional brake in an automobile.
If one brake is damaged, the other one can maintain its function. However, if the second brake
is also broke, the car is out of control (Vogelstein B. & Kinzler W.K., 2004 & Mach, M., 2013).
Tumor suppressor gene and oncogene mutations operate similarly, they drive the neoplastic
process by increasing tumor cell number through the stimulation of cell birth or by the
inhibition of cell death or cell-cycle arrest (Vogelstein B. & Kinzler W.K., 2004). The important
difference between tumor suppressor genes and oncogenes is that oncogenes results from
activation of proto-oncogenes, while tumor suppressor genes cause cancer when they are
inactivated (Hereditary and cancer, 2013).
In hereditary cancer syndromes, the individuals carry a germline mutation, that is linked to
cancer susceptibility, in one of the alleles in every single cell. Therefore, these individuals are
at high risk for developing cancer during their lifetime (Vogelstein B. & Kinzler W.K., 2004 &
Mach, M., 2013). Taken together, for the development of sporadic cancers both copies of a
tumor suppressor gene need to be inactivated somatically. Whereas in hereditary cancer
syndrome one mutated copy was already inherited and the other copy is inactivated by a
second somatic hit. This was already stated in 1971 by Alfred Knudson by the “second-hit”
hypothesis (Figure 3). The difference between hereditary and non-hereditary cases is the
timing of the first hit which is prezygotic or postzygotic, respectively (Knudson A.G, 1996 &
Knudson A.G. Jr., 1971). The first hit (somatic or inherited) leads to a heterozygous state for
the respective mutation. Inactivating mutations include missense mutations, nonsense
mutations or small or large insertions and deletions. The second hit involves usually another
mutation or aberrant promoter methylation. If the second copy of the gene is lost by deletion
a condition called “loss of heterozygosity” (LOH) occurs. Although most tumor suppressor
genes are recessive, there are some tumor suppressor genes, where inactivation of one copy
is sufficient to contribute to tumor progression. This phenomenon is called haploinsufficiency
10
(Figure 3) (Balmain et al, 2003 & Santorosa M. & Ashworth A., 2004 & Knudson A.G., 1996 &
Knudson A.G Jr., 1971).
Figure 3. (a,b) The classical Knudson “two-hit” model, involving a mutational event leading to gene inactivation, after which
LOH results in functional inactivation of both alleles. The first mutation can occur somatically (a) or can be inherited
through the germline predisposing highly to tumor development (b). (c,d) Haploinsufficient tumor suppressor genes does
not have to lose the two functional copies to increase the cancer risk. The loss of a single copy gene may occur by mutation,
deletion or silencing (c) or may be inherited through the germline (d) (Balmain et al, 2003).
Germline mutations that predispose to colorectal cancer are frequently found in several
tumor suppressor genes including STK11 (STK11-LKB1), that is associated with “Peutz-Jeghers
syndrome” (PJS), SMAD4, which is a key intracellular messenger in the “transforming growth
factor β” (TGF-β) pathway, and APC gene, that is associated with “familial adenomatous
polyposis” (FAP) (Su et al., 1999 & Vogelstein B. & Kinzler W.K., 2004 & Schwarte-Waldhoff et
al, 2000).
4.1.4 DNA polymerases
DNA polymerases are enzymes that are responsible for
synthesizing DNA and are essential for replication, DNA repair
and genetic recombination (Heitzer E. & Tomlinson I., 2014 &
Lange et al., 2011). They have structural domains regularly
similar to a “right hand”, in which DNA lies in the “palm” and
is surrounded by “fingers” and “thumb” domains as shown
for A-family DNA polymerase in Figure 4 (Lange et al., 2011).
The palm has a function in catalysis of the phosphoryl transfer
reaction, whereas the fingers have a function in interacting
with the incoming nucleoside triphosphates as well as the
template base to which it is paired. The thumb plays a role in
positioning the DNA (Steitz T.A., 1999).
Figure 4. A family DNA polymerase
structure (Lange et al., 2011).
11
The main function of these proteins is the duplication of the genome and the protection of
cells against the effects of DNA damage caused by many sources like UV, water-catalysed
reactions, ionizing radiation, etc. To date 15 different genes for DNA polymerases have been
encoded in the mammalian genome. They are specialized for different functions, including
DNA replication, DNA repair or the tolerance of DNA damage (Lange et al., 2011). DNA
polymerases can be divided in different DNA polymerase families based on amino acid
sequence comparisons and crystal structure analysis. Representative crystal structures are
known for enzymes in four of these families, including polymerase family A, B, X and Y.
Polymerase B family is the most important polymerase family and is involved in DNA
replication (Steitz T.A., 1999 & McCulloch S.D. & Kunkel T.A., 2008 & Lange et al., 2011).
4.1.4.1 DNA replication
The most important function of polymerases is the DNA replication, which is a highly complex
process, requiring the combined activity of dozens of proteins including 3 members of the Bfamily polymerases, polymerases α, δ and ε (Figure 5) (McCulloch S.D. & Kunkel T.A., 2008 &
Heitzer E. & Tomlinson I., 2014).
The four subunit polymerase α – primase complex
synthesizes RNA-DNA hybrid primers on the
leading and lagging strand at the replication forks,
thereby initiating the DNA replication. The
synthesis of the leading strand starts when a
polymerase binds and extends the primers in a
continuous way for as long as the polymerase is
able to stay bound. On the other hand the
replication of the lagging strand, which is a
discontinuous mode of synthesis, occurs in
patches of approximately 250 “base pairs” (bp)
called the Okazaki fragments, each of which must
be initiated by polymerase α – primase activity.
(McCulloch S.D. & Kunkel T.A., 2008).
Figure 5. Cartoon model of a eukaryotic replication
fork, necessary for DNA replication by complex
process of the combined activity of dozens of
proteins, whereof polymerases α, δ and ε (McCulloch
S.D. & Kunkel T.A., 2008).
The elongation of these primers on the leading and
lagging strand differs and is carried out by different
polymerases. Polymerases ε and δ are responsible for the leading- and lagging strand
synthesis, respectively (Yoshida et al., 2011 & McCulloch S.D. & Kunkel T.A., 2008 & Lange et
al., 2011 & Church et al., 2013). However, it is not fully elucidated if this is always the case, as
Pavlov proposed a model where polymerase ε starts replicating the leading strand, but may
dissociates later whereby polymerase δ takes over to complete the replication. This
hypothesis is endorsed by the higher mutation rate in polymerase δ exonuclease deficient
12
strains compared to polymerase ε exonuclease deficient strains (Heitzer E. & Tomlinson I.,
2014 & Pavlov et al., 2006).
Polymerases δ and ε are the principal replicases of the eukaryotic replication fork, by which
the major components are encoded by POLE and POLD1, respectively (Church et al., 2013).
They perform the bulk of replication with very high fidelity, which is ensured by a WatsonCrick base pairing and their possession of a catalytic subunit that includes 3’ exonuclease
domain to excise faults. This is known as proofreading which ensures an error rate of
approximately 10-5 to 10-7 substitutions per base, which is the lowest of all characterized DNA
polymerases. While most polymerases only rely on Watson-Crick base pairing for their fidelity
and lack intrinsic error-checking activity, like polymerase α (McCulloch S.D. & Kunkel T.A.,
2008 & Lange et al., 2011 & Heitzer E. & Tomlinson I., 2014). In addition, polymerases δ and ε
are thought to be involved in several repair pathways such as “mismatch repair” (MMR) ,
“nucleotide excision repair” (NER) and repair of “double strand” (ds) breaks, which are
important in maintaining DNA fidelity (Heitzer E. & Tomlinson I., 2014).
4.1.4.2 DNA replication fidelity
The replicative polymerases are part of the high fidelity replication and are highly accurate by
generating less than 1 mutation for every 10 000 base pairs copied, due to a combination of
accurate incorporation of nucleotides into the nascent DNA strand and by removal of
mispaired nucleotides by the exonuclease activity, also known as proofreading. The major
polymerases have evolved mechanisms to strongly favor correct over incorrect
“deoxyribonucleoside triphosphates” (dNTPs) incorporation. In addition, several DNA
polymerases, like polymerase δ and ε, have an associated 3' to 5' exonuclease activity that
catalyzes the hydrolysis of non-complementary nucleotides at the 3’-terminus. The faithful
replication of DNA is important to maintain genomic stability and to prevent mutagenesis and
tumor development (Church et al., 2013 & Yoshida et al., 2011 & Heitzer E. & Tomlinson I.,
2014 & McCulloch S.D. & Kunkel T.A., 2008).
However, sometimes the polymerases make errors that escape the proofreading activity.
These errors made by polymerases α, δ and ε are often misincorporations of dNTPs resulting
in base substitution errors, or insertions and deletions of one or more nucleotides during DNA
synthesis. These errors resulting from strand misalignments cause one or more unpaired bases
which occurs either in the primer strand, leading to insertions, or in the template strand,
leading to deletions. Usually the MMR apparatus, also called post-replication surveillance,
monitors and excises these DNA errors and correctly re-synthesize the DNA. The combination
of these correcting mechanisms results in an in vivo mutation rate that is estimated to be
lower than 10-9, i.e. less than one error every billion base pairs copied (Yoshida et al., 2011 &
McCulloch S.D. & Kunkel T.A., 2008 & Lange et al., 2011).
13
4.1.4.3 Exonuclease domain mutations (EDM)
As mentioned before polymerases δ and ε, heterotetramers in higher eukaryotes, contain
proofreading capacity which is important to maintain replication fidelity. Polymerases δ and ε
both compromise a catalytic subunit, POLD1 and POLE respectively. The accessory subunits
POLE 2/3/4 and POLD 2/3/4 are involved in regulating synthesis and in binding cofactors such
as “Proliferating Cell Nuclear Antigen” (PCNA). Both genes are highly expressed and show
great levels of evolutionary conservation. For an overview of these two genes see Table 1.
They differ in their length but are homologous, i.e. 23% identical and 37% similar, over their
exonuclease domain (residues 268-471 of POLE and 304-517 of POLD1) (Palles et al.,2013 &
Heitzer E. & Tomlinson I., 2014 & Lange et al.,2011 & Briggs S. & Tomlinson I., 2013).
Table 1. Overview polymerases δ and ε (Lange et al., 2011).
Recently mice carrying artificial alleles with substitutions at essential amino acids residues in
the proofreading domains of polymerases δ and ε have been reported, whereby tumor growth
in various organs of the animals arose (Yoshida et al., 2011 & McCulloch S.D. & Kunkel T.A.,
2008).
Mice with homozygous germline pole and/or pold1 mutations at the exonuclease active site
had distinct, but overlapping tissue-specific tumor phenotypes. The poleexo/exo mice died
prematurely of intestinal adenomas and adenocarcinomas, whereas the pold1exo/exo mice died
at an age of 8 months from thymic lymphomas or developed skin tumors, lung
adenocarcinomas or teratomas (Lange et al., 2011 & Heitzer E. & Tomlinson I., 2014). The
pole-mutant animals showed predominantly nodal lymphomas and histiocytic sarcomas,
whereas pold1-mutants showed thymic lymphomas and skin papillomas/sarcomas. Both types
of mice had intestinal adenomas and lung tumors, whereby the intestinal adenomas were
more common in pole-mutants and the lung tumors more in pold1-mutants (Heitzer E. &
Tomlinson I., 2014).
Table 2. Phenotypes of DNA polymerases δ and ε knockout in mice (Lange et al., 2011).
It has been shown that mutations in the exonuclease domains, called “exonuclease domain
mutations” (EDMs), lead to selective inactivation of polymerases δ and ε causing 10-100 times
higher mutations rates than the wild-types strains increasing the frequency of mutagenesis
14
(Yoshida et al., 2011 & Lange et al., 2011). These EDMs have been identified in POLD1,
encoding for the p124 subunit of polymerase δ, and in POLE, encoding for p262 catalytic
subunit of polymerase ε (McCulloch S.D. & Kunkel T.A., 2008 & Lange et al., 2011).
POLE and POLD1 are thought not to act as classical tumor suppressor genes, proven by the
unlikeliness of enzyme loss-of-function mutations to be pathogenic, since proofreading can
only fail after successful polymerisation. Also the fact that only a minority of tumors with POLE
and POLD1 EDMs show LOH or other inactivating mutations that could act as “second hits”
like in the classical tumor suppressor model of Knudson. Other data suggest that the mutator
phenotype and increased frequency of tumor formation is only indicated when POLE
mutations are homozygous. Another explanation might be that pathogenic EDMs are
selectively haploinsufficient and secondary somatic mutations in MSH2 and MSH6 mutations
contribute to tumorigenesis. The consequences of polymerase EDMs are not yet clear and
further analysis will be needed to fully understand the contribution of these mutations to
tumorigenesis (Heitzer E. & Tomlinson I., 2014).
4.1.5
Caretaker genes
The stability genes, also known as caretaker genes, keep genetic alterations to a minimum.
These genes include MMR, the NER and the BER genes responsible for repairing subtle
mistakes made during normal DNA replication or induced by exposure to mutagens
(Vogelstein B. & Kinzler W.K., 2004). These DNA repair mechanisms are generally involved in
the removal of damaged or incorrect incorporated bases, by which a DNA polymerase is
required to resynthesize DNA by using the undamaged strand as a template (Lange et al.,
2011).
4.1.5.1 Miss match repair (MMR)
The MMR genes include proteins that recognize and repair mismatches made by polymerases
and missed by the proofreading mechanism. They are responsible for maintaining the
replicative fidelity of DNA. MMR removes mismatched bases by excising a segment of DNA
between the mismatch and the nearby nick, recognized by mismatch recognition proteins.
The remaining gap in the DNA is filled by polymerase δ (Lange et al., 2011 & de la Chapelle, A.,
2004 & Grady W.M. & Carethers J.M., 2008). Approximately 30% of the genes in the human
genome are encoding for proteins that regulates DNA fidelity achieved through the combined
action of accurate DNA polymerases and DNA MMR (Grady W.M. & Carethers J.M., 2008 &
McCulloch S.D. & Kunkel T.A., 2008).
4.1.5.2 Nucleotide excision repair (NER) & Base excision repair (BER)
NER can remove and recognise different helix-distorting adducts, like those caused by
radiation, UV, etc. Such distorted regions are recognized and the damaged DNA is excised by
15
making two incisions on either side of the adduct resulting in a 27-29 nucleotide gap that is
filled by polymerases δ and ε (Lange et al., 2011).
BER is another part of the "DNA caretakers" that include base excision repair. It typically
mediates the removal and replacement of single base residues. The substrates for this
reaction are uracil residues in DNA and damaged bases caused by reactive oxygen species,
methylation or hydrolytic reactions. Damaged bases are removed by a specific DNA
glycosylase. Afterwards the resulting abasic site is incised by an “apurinic” or “apyrimidinic”
(AP) endonuclease. After the “5’-deoxyribose-phosphate” (dRP) residue is removed by a dRP
lyase, the resulting one nucleotide gap is filled by DNA polymerase β (also known as POLB)
(Lange et al., 2011). An example of a BER encoding gene is MUTYH (de la Chapelle, 2004).
Biallelic germline MUTYH mutations, cause defective base excision repair, what predispose to
adenomatous colorectal polyposis and CRC (Heitzer E. & Tomlinson I., 2014).
4.1.5.3 Other stability genes
Other stability genes are involved in the control of mitotic recombination and chromosomal
segregation (Vogelstein B. & Kinzler W.K., 2004). Furthermore, there are DNA repair proteins
that are involved in the regulation of dsDNA-breaks; i.e. CHEK2, BRCA1, BRCA2 and BLM; and
mutations in these are associated with a high risk of mainly breast but also colon cancer (Grady
W.M. & Carethers J.M., 2008). Another DNA repair gene is TP53, which is involved in cell
checkpoint control that prohibit cells to replicate DNA in presence of DNA breaks. Loss of this
gene it is frequently present in cancer cells (Lange et al., 2011).
4.1.5.4 Mutations in stability genes
When the caretaker genes are inactivated, mutations in other genes occur at a higher rate,
whereby somatic mutations are produced a 100 times the normal rate in the affected cells
leading to an increased potential risk of tumorigenesis, especially when the affected genes are
oncogenes or tumor suppressor genes (Vogelstein B. & Kinzler W.K., 2004 & knudson A.G.,
2002). If mispaired bases are not corrected by MMR, they may cause nucleotide transitions or
transversions, allowing a novel base to alter the authentic genetic sequence. Such point
mutations in genes that regulate cell growth can accumulate in cells with defective MMR and
may promote neoplastic growth (Grady W.M. & Carethers J.M., 2008 & de la Chapelle, A.,
2004). One wild-type allele of an MMR gene is generally sufficient to maintain normal MMR
function (Grady W.M & Carethers J.M., 2008). The occurrence of germline mutations in the
MMR genes system such as MLH1, MSH2, MSH6 and PMS2 family of proteins, is most relevant
in LS. They can predispose to cancer as the somatic mutations produces loss or mutation of
the remaining wild-type allele in a cell, rendering it defective for MMR (Boland et al., 2008 &
de la Chapelle A., 2004 & Knudson A.G., 2002 & Grady W.M & Carethers J.M., 2008). Multiple
point mutations and “microsatellite instability” (MSI) present in a tumor is often referred to
as "hypermutator phenotype”. It was recognised that MSI occurs in 15%-20% of sporadic
16
human colorectal tumors and in > 95% of colon cancers arising in patients with LS. It was
shown that biallelic inactivation of members of the human DNA MMR gene family was the
cause of microsatellite unstable CRC. The presence of MSI can be a hallmark for hereditary
CRCs (Grady W.M & Carethers J.M., 2008).
In 15% - 20% of colon cancers, inactivation of MMR system either through aberrant CpG island
methylation of MLH1 or germline point mutations in MLH1, MSH2, MSH6, PMS2 or other
members of the MMR family leads to MSI (Grady W.M. & Carethers J.M., 2008 & de la
Chapelle, A., 2004). Microsatellites, stretches of DNA with a repetitive sequence of
nucleotides, are particularly susceptible to acquire errors if MMR gene function is impaired
(Kohlmann, W. & Gruber, S.B., 2012). MSI is termed with the occurrence of shorter or longer
repetitive nucleotide sequences (microsatellites) in tumor cells such as An or CAn, where the n
stands for the number of repeats. The prolongation of the repetitive nucleotide sequences
takes place during replication through nucleotide pairing slippage on the newly synthesized
strand, whereas the shortened repetitive nucleotide sequence occurs through slippage of the
template strand, both alterations define MSI. Such repetitive nucleotide sequences are also
present in coding region (exons) of critical growth regulatory genes, like tumor suppressor
genes. Mutation in these coding microsatellite occurs with defective MMR, resulting in length
changes in the microsatellite within the gene producing a frameshift mutation that renders
the protein inactive, what can cause tumor promoting effects (Grady W.M. & Carethers J.M.,
2008). It has been noted that longer repetitive elements are more prone to mutation than
shorter ones (Bolande et al., 2008).
17
4.2 Colorectal Cancer (CRC)
One of the main types of cancer is CRC accounting for 608 000 deaths a year, which is the
second largest leading cause of cancer deaths in the Western world (Cancer, 2013 & Ranplex
CRC Array, 2007). CRC is a cancer that is located in the intestine and is being responsible for
55% of all cancer-cases, in the more developed regions. Men are affected more frequently
than women and two third of the colon cancers develop in the proximal colon (Kohlmann, W.
& Gruber, S.B., 2012 & All Cancers, 2012). According to “world health organisation” (WHO)
the five leading risks for cancer are high body mass index, low fruit and vegetable intake, lack
of physical activity, tobacco use and alcohol use. Also nutritional factors play an important
role in cancer development, especially in CRC (de la Chapelle A., 2004 & Mach, M., 2013 &
WHO, 2008). Studies have established a positive correlation between CRC risk and the intake
of fat, red meat and alcohol, as well as smoking. In contrast there exists an inversed correlation
between CRC risk and the intake of vegetables and “non-steroidal inflammatory drugs”
(NSAID), “hormone-replacement therapy” (HRT) and physical activity (de la Chapelle A., 2004).
CRC is a complex and heterogeneous disorder identified
through analysis of precursor lesions and hereditary forms of
the disease (Shi C. & Washington K., 2012). Most CRCs arise
from adenomatous polyps, the cells of which have a clear
growth advantage leading to a benign neoplasm (Knudson
A.G., 2001 & de la Chapelle A., 2004 & Fearnhead et al.,
2002). This transformation is resulting from the progressive
accumulation of genetic and epigenetic alterations leading to
the transformation of normal colonic epithelium to colon
adenocarcinoma, which is called the “polyp to cancer
progression sequence” (Grady W.M. & Carethers J.M., 2008
& Bardhan K. & Liu K., 2013). This sequence of events from
aberrant crypt proliferation or hyperplasia to benign
adenomas, to carcinomas in situ and finally to metastatic Figure 6. A multi-hit model for CRC,
showing mutational events that
carcinoma, attests the step ways progression of the cancer, correlate with each step in the
sequence
which often occurs over many years. Many somatic mutations adenoma-carcinoma
(Knutson A.G., 2001)
occur in several genes during this sequence of events, and
occur, at least to some extent, in a predictable order (de la Chapelle A., 2004).
This sequence of events associated with both, hereditary cancer and with benign precursors
of malignant tumors, includes mutations in APC and TP53, respectively increasing cell birth
rate and reducing cell death rate. Mutation and/or loss of APC gene is the most earliest genetic
change associated with adenomatous polyps, whereas TP53 mutations occur more frequently
in high-grade dysplastic polyps and are thought to mark the transition from adenoma to
18
carcinoma (Fearnhead et al., 2002 & Knudson A.G., 2001). Also mutations of the RAS oncogene
family is frequently observed. RAS mutations are found in approximately 50% of the CRCs and
thought to be a relative early event that correlates histologically with early to late adenomas.
This well-known ‘adenoma-carcinoma’ sequence in CRC has made this disease a popular
model for a “multi-hit” cancer pathway (Figure 6) (Knudson A.G., 2001 & de la Chapelle A.,
2004 & Fearnhead et al., 2002).
4.2.1 Sporadic versus familial cases
CRC can be divided into sporadic and familial (hereditary) cases. The sporadic CRC cases
represents the largest group of CRCs. To date three different mechanism have been described
to promote tumorigenesis of sporadic CRC. In the first mechanism APC has been implicated to
cause “chromosomal instability” (CIN), whereby CIN tumors account for 60% of the sporadic
CRCs. Another mechanism is MSI resulting from mutations in the MMR genes, which are
observed in both, sporadic and LS-associate tumors. The third mechanism is DNA
hypermethylation causing “CpG island methylator phenotype” (CIMP) (Mach, M., 2013).
In about 20-25% of all CRCs cases a
family history of tumor occurrence is
reported. However, a genetic
predisposition can only be identified
in 5-10% of all cases. CRC is one of the
best characterized cancers in terms
of gene mutation contributing to
tumorigenesis (de la Chapelle A., Figure 7. The difference between a hereditary cancer and a nonhereditary cancer The ‘one hit’ is a precursor to the tumor in non2004 & Martinez, J.D., 2005). When hereditary cancers, whereas all cells are one-hit clones in hereditary
CRC is detected in a young patient, cancers. The “two-hit” formation causes cancer (Knudson et al., 2002).
present as synchronous or metachronous CRC or as adenomatous polyps, or when there is a
strong family history of CRC, it can be presumed that the patient has a hereditary
susceptibility, explained by the “two-hit” model from Knudson (Figure 7) (Fearnhead et al.,
2002 & Knudson A.G. Jr., 1971).
19
4.2.2 Penetrance of hereditary colorectal cancer syndromes
Cancer-causing mutations can be high-penetrance mutations or low-penetrance mutations.
The high-penetrance mutations confer predisposition to CRC mainly in LS and FAP, accounting
for 5% or less of all cases of CRC (de la Chapelle A., 2004). LS and FAP are dominantly inherited
conditions with 80% and 100% life-time risk of developing CRC, respectively (Fearnhead et al,
2002).
Although
the
mutations involved in LS and
FAP are high penetrant, the
exact level of penetrance is
various, e.g. the difference in
age of onset (Figure 8). Lowpenetrant mutations account
for a high proportion of all
attributable risks of CRC,
rather than being a true Figure 8. Representation of colorectal tumor progression in sporadic and high-risk
genetic syndromes. By which a normal colonocyte stem cell that has sustained
predisposition
and
are genetic damage due to the local environment and any germline genetic mutation
therefore more difficult to that has been inherited causes tumor progression (Grady W.M. & Carethers J.M.,
2008).
identify (de la Chapelle A.,
2004).
4.2.3 Genetic pathways of CRC
Due to analysis of precursor lesions and hereditary forms of the disease, it has been concluded
that CRC is not a single disease entity that shares the same clinical characteristics, cause and
treatment results. At least 3 distinct molecular pathways to CRC have been described,
including the conventional suppressor pathway characterized by mutations in APC and
exemplified by FAP, the hereditary MSI pathway found in LS and the serrated pathway
characterized by aberrant crypt CpG island methylation. Different genes can be mutated or
altered in carcinomas arising via the same genetic pathway. Exome sequencing studies
revealed that individual CRCs harbour an average of 76 gene mutations. Furthermore, the
spectrum of mutated genes in two tumors only overlap in a small extend, whereby a few
genes, like APC, are mutated in a high frequency (Shi C. & Washington K., 2012).
The hereditary gastrointestinal polyposis syndromes can be divided into those were the polyps
are predominantly adenomatous, FAP and LS, and in those were the polyps are predominantly
hamartomatous (Zbuk K.M. & Eng C., 2007). Recently a possible new CRC class, characterised
by multiple colorectal adenomas and carcinomas, was defined as “polymerase proofreading
associated polyposis” (PPAP). Although, the pathogenic polymerase EDM cancers, causing
PPAP, form a rare subtype of tumors that is seemingly restricted to the colorectum and
20
endometrium, but there is no reason to regard these cancers as an unimportant group (Heitzer
E. & Tomlinson I., 2014).
4.2.4 Familial adenomatous polyposis (FAP)
FAP is a well understood familial autosomal-dominant CRC syndrome. In 1861 it was first
reported as a hereditary from of CRC and in 1925 Lockhart-Mummery described it as a
dominantly inherited Mendelian trait.
FAP is caused by high-penetrance mutations with a penetrance of nearly 100%. FAP is
accounting for about 0.5% of all CRCs and is characterized by the presence of hundreds to
thousands of adenomas in late childhood or adolescence, throughout the colorectum. 70-80%
of these tumors occur on the left side of the colon (de la Chapelle A., 2004 & Fearnhead et al.,
2002 & Galiatsatos P. & Faulkes W.D., 2006 & Lynch et al., 2009). Most patients develop polyps
at the age of 20-30 that progress to tumors at the age of 35-45. These polyps and the cancer
arises in adolescence and are usually located in the lower gastrointestinal tract. 90% of FAP
patients develop polyps in the upper gastrointestinal by the age of 70 years. And roughly 7080% of FAP patients show congenital hypertrophy of the retinal pigment epithelium, referring
to the presence of characteristic pigmented fundus lesions.
Other clinical presentations of FAP include desmoid tumors, thyroid cancer, hepatoblastomas
and other extracolonic malignancies. In addition, there exist also some distinct subtypes of
FAP including Gardner Syndrome, Turcot Syndrome and Hereditary Desmoid Disease
(Galiatsatos P. & Faulkes W.D., 2006).
4.2.4.1 Adenomatous polyposis coli (APC)
FAP is caused by mutations in the “adenomatous polyposis coli” (APC) gene, a tumor
suppressor gene located on chromosome 5q21 (Papou E.P. & Ahuja N., 2010 & Galiatsatos P.
& Faulkes W.D., 2006). This gene encodes for a 312 kDa protein consisting of 2.843 amino
acids and is involved in a variety of cellular processes including migration, adhesion,
proliferation, etc. More than 95% of the germline mutations in APC are nonsense or truncating
mutations, which are typical deletions or insertions leading to an altered reading frame (de la
Chapelle A., 2004 & Fearnhead et al., 2002 & Galiatsatos P. & Faulkes W.D., 2006).
Almost all individuals who carries a germline pathogenic mutation of APC, which applies for
the first hit” in accordance to Knudson’s “two-hit hypothesis”, eventually develop FAP (Grady
W.M. & Carethers J.M., 2008 & de la Chapelle A., 2004). Tumors will inevitable, if untreated,
develop in patients with FAP after somatic inactivation of the wild-type APC allele, which
applies for the second hit (Smith et al., 2013 & Fearnhead et al., 2002). The risk of cancer in
FAP is generally correlated with polyp number. Germline mutations in APC are spread fairly
evenly between codons 200 and 1600. The type of germline APC mutation in FAP appears to
21
determine the nature of the second somatic hit to APC (Fearnhead et al., 2002). To date, over
700 different disease-causing APC mutations have been reported, whereby the most common
germline mutation, caused either by a nonsense mutation, frameshift mutation, or large
deletion, leads to truncation of the protein product in the C-terminal region, involving the
introduction of a premature stop codon. These germline mutations are often clustered at the
5’ end of exon 15, otherwise referred to as the mutation cluster region (Galiatsatos P. &
Faulkes W.D., 2006).
4.2.4.2 Affected pathways
APC is a part of the β-catenin destruction complex that controls Wnt signalling, and thereby
regulates the proliferation of colonic cell (Smith et al., 2013). The Wnt signalling pathway
regulates the phosphorylation and degradation of β-catenin, an intracellular protein that
binds to the cell adhesion molecule E-cadherin and links E-cadherin to the actin cytoskeleton.
The phosphorylation of β-catenin will attract ubiquitin ligases, leading to its destruction at the
proteasome (Galiatsatos P. & Faulkes W.D., 2006). Normally Apc prevents β-catenin to bind
to the “T cell factor” (Tcf) family of transcription factors, like Tcf-4, that alters the expression
of various genes affecting the proliferation, differentiation, migration, and apoptosis of cells.
Mutations in APC cause accumulation of β-catenin in the cytoplasm and overexpression of
oncogene c-myc, thereby promoting cell growth, which provided the first connection between
the loss of a tumor suppressor gene and the activation of an oncogene (Papou E.P. & Ahuja
N., 2010 & Galiatsatos P. & Faulkes W.D., 2006). APC also plays a role in controlling the cell
cycle, by inhibiting the progression of cells from the G0/G1 to the S phase, which helps to
suppress tumorigenesis. Furthermore, APC stabilizes microtubules, promoting chromosomal
stability (Galiatsatos P. & Faulkes W.D., 2006).
4.2.4.3 Attenuated FAP (AFAP) & MYH associated polyposis (MAP)
“Attenuated FAP” (AFAP), is a phenotypically distinct variant of FAP caused by mutations in
APC at the extreme proximal or distal portions of the gene. AFAP is typically characterized by
dozen to hundred polyps, in contrary with the thousand adenomas in FAP, a more proximal
colonic location of polyps and the delayed age of onset (Galiatsatos P. & Faulkes W.D., 2006
& de la Chapelle A., 2004)
Inactivation of MUTYH, also known as MYH, causes the occurrence of an autosomal recessive
form of adenomatous polyposis, “MYH associated polyposis” (MAP) syndrome, characterised
by the development of colorectal adenomas and carcinomas (de la Chapelle, 2004 & Smith et
al., 2013 & Galiatsatos P. & Faulkes W.D., 2006 & Grady W.M. & Carethers J.M., 2008). MAP
is associated with inactivation of the BER gene, MYH, causing multiple adenomatous polyposis
through genomic instability at the base pair level. Consistent with a recessive trait the
penetrance of MYH mutations are very high when both germline alleles are affected. These
MYH germline mutations are characterised by somatic G:C > A:T mutations without any
22
detectable germline APC mutations in neoplasms of patients with adenomatous polyposis
(Grady W.M. & Carethers J.M., 2008 & de la Chapelle A., 2004). MAP occurs in up to 5-10% of
individuals who have adenomatous polyposis syndromes (Grady W.M. & Carethers J.M.,
2008).
4.2.5 Lynch syndrome (LS)
LS is named after the oncologist who pioneered the study of this disease. It was previously
called “hereditary non-polyposis colon cancer syndrome” (HNPCC), because of the absence of
polyposis. Although, polyps are found at a low rate and they appeared to convert rapidly to
carcinomas. HNPCC is a misnomer as the syndrome involves predisposition not only to CRC,
but also to cancers of at least seven other organs. LS is a cancer syndrome with an increased
risk in developing colon and extracolonic cancers with an unusual early age of onset of 30 to
45 years (de la Chapelle A., 2004 & Boland et al., 2008 & Lynch et al., 2009 & Knudson A.G.,
2002).
LS is an autosomal dominant inherited condition with an 80% life time risk of CRC, and 50-60%
life-time risk of EC in women (Fearnhead et al., 2002 & Kohlmann, W. & Gruber, S.B., 2012). It
is the most common hereditary CRC predisposing syndrome accounting for approximately 3%
of all CRCs, although LS only explains 10-25% of familial CRC (Lynch et al., 2009). The majority
of individuals diagnosed with LS have inherited a predisposing allele from a parent. However,
because of incomplete penetrance, variable age of cancer development, early death, etc., not
all individuals with a mutation in a LS-predisposing gene have a parent who had cancer
(Kohlman, W. & Gruber, S.B., 2012).
4.2.5.1 Pathogenesis of LS
In the early 1990s, it was found that LS is caused by germline mutations in four MMR genes
including MSH2, MLH1, MSH6 and PMS2 (Grady W.M. & Carethers J.M., 2008 & Heitzer E. &
Tomlinson I., 2014 & Boland et al., 2008 & Kohlmann, W. & Gruber, S.B., 2012). As well
germline deletions of EPCAM predispose to LS leading to a loss of expression of the MSH2
allele, which is identified in about 1% of the individuals with LS (Kohlmann, W. & Gruber, S.B.,
2012).
The two ‘major’ DNA MMR genes, MSH2 and MLH1, are stabilized by interactions between
several ‘minor’ DNA MMR genes, including MSH6 and PMS2. The expression of these ‘minor’
DNA MMR genes is depending on the binding partners (Lynch et al., 2009). This can be
explained by the fact that MMR gene products works in dimers, whereby MSH2 protein can
complex with MSH6 or MSH3 protein and MLH1 protein can complex with PMS2 or PMS1
protein. Notably, MSH6 and PMS2 proteins are unstable when not pared in complex, which
can be determined in “immunohistochemistry” (IHC). Generally a germline mutation in MSH2,
or, more rarely, a deletion in EPCAM, results in loss of expression of the proteins MSH2 and
23
MSH6. Similarly a germline mutation in MLH1 results in a loss of expression of the proteins
MLH1 and PMS2. In addition, germline mutations in MSH6 and PMS2 only show isolated loss
and do not result in loss of MSH2 or MLH1 expression as they are still present in other pairings
(Kohlmann, W. & Gruber S.B., 2012 & Lynch et al., 2009 & Steward, A., 2013). Loss or
malfunction of the MMR system leads to numerous unrepaired errors in the genome, whereof
some errors are manifested as changes in the lengths of short repeated sequences known as
microsatellites, causing MSI (Steward, A., 2013).
4.2.5.2 Mutation spectra in LS
Numerous germline mutations in MMR
genes, i.e. nonsense, splice-site, truncating,
frameshift,
missense,
in
frame
insertion/deletion or other kind of
mutations are associated with LS (de la
Chapelle A., 2004 & Grandval et al., 2013 &
Thompson et al., 2014). The distribution of
mutations in the three most frequently
mutated MMR genes in LS is shown in Figure
9. Nonsense, splice-site and truncating
mutations are always considered to be
deleterious, whereas missense mutations Figure 9. Distribution of mutations in the three main MMR
are not a priori disease causing. For genes associated with LS (de la Chapelle A., 2004).
missense mutations the clinical relevance is often not obvious and therefore they are classified
as “unknown variants” (UV). Large deletions are common in MSH2, accounting for 15-20% of
all mutations, but rare in MLH1 and even less common in MSH6 (de la Chapelle A., 2004).
The best known recurrent mutation causing LS, occurs worldwide and may account for as
much as 5-10% of all LS, includes an A to T transversion. Which is presumably caused by the
slippage of DNA polymerase during replication in the splice donor site of intron 5 of MSH2
leading to transcriptional skipping of exon 5 causing loss of MSH2 expression. The recurrent
nature of this mutations is explained by the fact that the concerning adenine is the first of a
whole stretch of 26 adenines creating a “hotspot” for this particular change (de la Chapelle A.,
2004 & Lynch et al., 2009). The cancer risk varies among the four MMR genes, by which
heterozygosity for an MSH2 mutation is linked with the greatest risk for extracolonic cancers
in contrary with heterozygosity for other MMR genes (Kohlmann, W. & Gruber, S.B., 2012).
Although MSH2 and MLH1 mutations account for most cases of LS, it is recommended to
evaluate a suspected LS patient on mutations in one of the ‘minor’ MMR genes (Lynch et al.,
2009).
24
4.2.5.3 Tumorigenesis and presence of MSI in LS
Germline mutations in one of the MMR genes results in an increased risk in tumorigenesis due
to the fact the person only needs one more mutation in the wild-type allele of the MMR gene
affected. The MMR genes behaves like a tumor suppressor, therefore a second somatic
mutation must occur in the intact allele to develop a tumor. This second hit can occur as a
deletion (LOH), another mutation or methylation of CpG islands in the MHL1 promotor (Grady
W.M. & Carethers J.M., 2008 & de la Chapelle A., 2004 & Kohlmann, W. & Gruber, S.B., 2012).
It was demonstrated in 1991 that tumors occurring in LS patients had a characteristic
molecular change named MSI, previously called ‘ubiquitous somatic mutations in simple
repeated sequences’, or a ‘replication error’ phenotype (Lynch et al., 2009). In 15-20% of CRCs,
inactivation of MMR system through point mutations in MLH1, MSH2 or other members of
the MMR family or through aberrant CpG island methylation of MLH1 leading to the
occurrence of MSI (Grady W.M. & Carethers J.M., 2008).
4.2.6 Hamartomatous Polyposis Syndromes
Hamartomatous polyps are composed out of normal cellular elements of the
“gastrointestinal” (GI) tract, but contains a distorted architecture. The hamartomatous
polyposis syndromes are a heterogeneous group of inherited autosomal-dominant disorders
and can be distinguished by their extracolonic manifestations and by their hamartomatous
rather than adenomatous pathology in several conditions, including “Peutz-Jeghers
syndromes” (PJS), “Juvenile Polyposis Syndromes” (JPS) and “PTEN hamartomatous tumor
syndrome”(PHTS). Most of these syndromes increase the risk of developing CRC, however
they account for less than 1% of all CRC cases. In fact these syndromes are examples of genetic
predisposition that falls between ‘high’ and ‘low’ penetrance, and between ‘familial’ and
‘sporadic’ (Zbuk K.M. & Eng C., 2007 & de la Chapelle A., 2004 & Kohlmann, W. & Gruber, S.B.,
2012).
4.2.6.1 Peutz-jeghers syndromes (PJS)
PJS is an autosomal dominant condition characterized with the presence of hamartomatous
polyps in the gastrointestinal tract, and is associated with mucocutaneous pigmentation.
Germline mutations in STK11, also known as LKB1, are detected in 70-80% of patients with
PJS. STK11 encodes a serine-threonine kinase that modulates cellular proliferation, controls
cell polarity, and has an important function in responding to low cellular energy levels (Zbuk
K.M. & Eng C., 2007 & Kohlmann, W. & Gruber, S.B., 2012).
4.2.6.2 Juvenile polyposis syndromes (JPS)
JPS, an autosomal dominant disorder, is characterized by the occurrence of multiple juvenile
polyps at any age. Almost 50% of juvenile polyposis is caused by germline mutations in SMAD4,
although in some cases mutations in BMPR1A or ENG are found. These 3 genes encodes for
proteins with a role in the TGF-β signalling pathway, which is an important modulator of many
25
cellular processes such as proliferation, differentiation and adhesion. TGF- β itself has a major
role in the control of colonic epithelial growth (de la Chapelle A., 2004 & Zbuk K.M. & Eng C.,
2007 & Kohlmann, W. & Gruber, S.B., 2012). SMAD4 and BMPR1A also play a role in “bone
morphogenetic protein” (BMP)-mediated signalling. Biallelic inactivation of SMAD4,
consistent with it function as a tumor-suppressor gene, occurs in a substantial proportion of
“advanced CRC” (aCRC) (smith et al., 2013).
4.2.6.3 PTEN hamartomatous tumor syndromes (PHTS)
PHTS is a term developed to unify a heterogeneous group of disorders that is caused by
mutations in the tumor suppressor gene, PTEN. PHTS includes Cowden disease, “BannayanRiley-Ruvalcaba syndrome” (BRRS), Proteus syndrome and all other syndromes where
germline PTEN mutations are detected. Pten protein is a ubiquitously expressed dual
phosphatase, which contains for as well lipid as protein a phosphatase activity, of which has
an important role in controlling cell growth, proliferation and angiogenesis (de la Chapelle A.,
2004 & Zbuk K.M. & Eng C., 2007 & Kohlmann, W. & Gruber, S.B., 2012).
4.2.7 Polymerase proofreading associated polyposis (PPAP)
Previously yeast models
have
shown
a
base
substitution
mutator
phenotype with variable
severity,
caused
by
inactivating missense EDMs
in polymerase δ and
polymerase ε homologues.
Though less is known about
the impact of these
proofreading-defective DNA
polymerase mutations in
higher eukaryotes. Recently,
Palles et al identified
germline EDMs in human
POLD1 and POLE, impairing
polymerase proofreading,
Figure 10. The genomic structure of POLD1 (a) and POLE (b) genes, including the
that
predispose
to functional domains in their protein products. The domains in red are indicating for
exons corresponding to the 3’ exonuclease, that is, proofreading, domains
“polymerase proofreading the
(Yoshida et al., 2011).
associated polyposis” (PPAP)
(Figure 10). Segregation analysis confirmed a dominant, high-penetrance predisposition of
these mutations to colorectal adenomas (Smith et al., 2013 & Heitzer E. & Tomlinson I., 2014).
26
Generally PPAP is a disease, with high penetrance and dominant inheritance, characterised by
multiple colorectal adenomas and carcinomas. Church et al has shown that germline POLE and
POLD1 EDMs predispose to CRC and, in the latter case also to “endometrial cancer” (EC).
Patients who carry EDMs in POLD1 and POLE show variable phenotypes, whereby some had
tens of adenomas that did not appear to progress rapidly to cancer, whereas others had a
small number of large adenomas or early-onset carcinomas that resembled LS. So the disease
is called PPAP as the phenotype overlaps with those who carry germline mutations in MUTYH
and the MMR genes, resulting in MAP and LS indications, respectively.
Notably, in CRCs with EDMs, the spectrum and/or frequency of known driver mutations, is
unusual. Recurrent mutations are frequently observed in the known CRC driver genes, by
which these are often of types and at position other than the common hotspots. These include
mutations in KRAS, APC and TP53. There is considerable evidence that specific POLE mutations
have different effects on the somatic mutation spectrum. And it is noted that secondary
somatic mutations, resulting from defective proofreading, tend to occur at sites on the
“positive” DNA strand flanked by an A base, rather than by a T, G or C. This could be explained
by the lower helix ‘melting‘ temperatures of A:T tracts. In addition, it has been suggested that
somatic POLE mutations occur very early during colorectal tumorigenesis, as the frameshift
mutations often found at APC in unselected CRCs are not seen in tumors with EDMs (Heitzer
E. & Tomlinson I., 2014).
EDM tumors are “microsatellite stable” (MSS) and show an “ultramutator” phenotype with a
dramatic increased tendency for somatic base substitutions of all types, with about 5000
substitutions in the coding regions alone, compared to POLE wild-type tumors. Whereas C:G
> T:A changes generally remained the most common type of mutations (Church et al., 2013 &
Palles et al., 2013 & Heitzer E. & Tomlinson I., 2014). In addition, several “short nucleotide
polymorphisms” (SNPs) located at conserved sites within the polymerase or exonuclease
domains of POLE and POLD1 have been identified, but without any association to date (Heitzer
E. & Tomlinson I., 2014). The mechanism underlying POLE and POLD1-associated colorectal
tumorigenesis is not yet full established (Palles et al., 2013 & Smith et al., 2013).
27
4.2.8 Short summary of genes associated with a high susceptibility of CRC
Table 3. High susceptibility genes for colorectal cancer (Colorectal cancer genes, 2013).
Gene
Syndrome
Tumor suppressor genes:
Hereditary pattern
Predominant cancer
APC
TP53
FAP
Li-Fraumeni
Dominant
Dominant
Colon, intestine, etc.
Multiple (including colon)
STK11 (LKB1)
PJS
Dominant
Multiple (including intestine)
PTEN
Cowden
Dominant
Multiple (including intestine)
BMPR1A
SMAD4 (MADH/DPC4)
JPS
JPS
Dominant
Dominant
Gastrointestinal
Gastrointestinal
LS
Dominant
MAP
PPAP
Recessive
Dominant
Multiple (including colon, uterus,
and others)
Colon
Colon endometrial
Repair/Stability genes:
MLH1, MSH2,
MSH6,PMS2
MYH (MUTYH)
POLD1, POLE
4.3 Genetic Testing
It is estimated that about 5 -10% of all cancers are inherited (WHO, 2008 & Garber J. & Offit
K., 2005). If a familial predisposition is suspected due to a family history of cancer or an early
onset of the disease, genetic testing can be considered. Genetic testing includes screening of
high-penetrance tumor susceptibility genes allowing the identification of individuals at high
risk for cancer (Pasche, B., 2010 & WHO, 2008).
There is a difference between somatic genetic profiling and germline testing of cancer tissue
to predict prognosis or treatment response. Somatic genetic profiling includes the DNA
analysis of the tumor. In contrast, germline testing involves the DNA analysis of blood or saliva
for testing inherited mutations in specific genes linked with the type of cancer seen in the
individual’s family. The identification of such high-penetrance mutations can predict higher
risks for cancer, as these mutations result in significant function alterations of the
corresponding gene product. The identification of these high-penetrance mutations often
justifies clinical care, tight surveillance or preventive surgery (Robson et al., 2010). Identifying
hereditary cancer syndromes is the current indication for standard-of-care molecular testing
in colorectal carcinomas (Shi C. & Washington K., 2012). This testing for germline mutations
should occur in the setting of appropriate genetic counselling and offer predictive testing for
family members (Zbuk K.M. & Eng C., 2007). Nevertheless additional research is needed to
better understand and to ultimately prevent and/or treat these diseases (McCulloch S.D. &
Kunkel T.A., 2008).
28
4.3.1 Routine mutation screening of CRC
Research on CRC is leading the way in understanding cancer genetics. Genetics plays an
important role in predisposition to CRC and in understanding its initiation and progression.
The identification of genetic alterations can help to understand the cancerogenic process, and
favour the development of diagnostic, preventive and therapeutic strategies (de la Chapelle
A., 2004). The process of genetic testing includes identification of affected and at-risk
individuals, genetic counselling, laboratory testing, and accurate interpretation of results.
Genetic testing should be carried out in an individual suspected of having a hereditary form of
CRC, or in the relatives of a known genetic carrier, by which the proband (original patient
presenting the disease) should be genotyped first. This has the advantage to identify the
responsible mutation and to increase the sensitivity of detection in at-risk relatives
(Fearnhead et al., 2002). Genetic counselling is also important as it provides the patient and
family the important details about genetic risk for cancer and the consequences of a positive
test. Genetic testing should only be done when there is a clear benefit that can be provided
for individuals with an inherited susceptibility of cancer (Fearnhead et al., 2002 & Lynch et al.,
2009).
4.3.1.1
Germline detection of CRC
The detection of germline mutations is performed by a multitarget assay containing several
markers (Pappou E.P. & Ahuja N., 2010). To date mutations in at least 10 genes have been
identified that are responsible for Mendelian syndromes associated with CRC tumorigenesis.
Different genes are associated to different colorectal cancer syndromes, APC and MUTYH are
the primary predisposition genes linked to multiple adenomas, which are benign precursors
of many CRCs. In addition other genes, like STK11-LKB1, SMAD4, BMPR1A and GREM1,
mediate a CRC risk through the development of hamartomas or mixed polyps. In contrast
mutations in the MMR genes (MSH2, MLH1, MSH6 and PMS2) show a typical presentation of
early-onset CRC or endometrial cancer without a great number of polyps. All these known
Mendelian genes, with varied function, are responsible for CRC (Lange et al., 2011 & Palles et
al., 2013).
4.3.1.2
Genetic testing for FAP
Genetic testing for FAP causing mutations is performed by sequencing the whole APC gene.
As a correlation of phenotype and genotype of FAP is known, targeted genetic testing in FAP
instead of sequencing whole the gene might be considered. A variety of methods for mutation
detection in FAP are current in use including direct sequencing, mutations screening with
either “single stranded conformational polymorphism” (SSCP) or “denaturing gradient gel
electrophoresis” (DGGE), and the “protein truncation test” (PTT) (Fearnhead et al., 2002). 20%
to 30% of classical FAP patients do not harbour any detectable APC germline mutation despite
genetic testing by routine screening methods. This could be explained by the possibility that
29
the PTT misses the presence of a nontruncating missense mutation. Nowadays, PTT has been
largely replaced by DNA sequencing and newer diagnostic methods, like “monoallelic
mutation analysis” (MAMA), whereby it is possible to examine the two APC alleles
independently (Galiatsatos P. & Faulkes W.D., 2006).
It is recommended to screen every adolescent and young adult with a family history of FAP by
genetic testing, as it allows clinical surveillance procedures to focus only on those at real risk
(de la Chapelle A., 2004).
4.3.1.3
Genetic testing for LS
The diagnosis of LS can be made by the molecular genetic testing of one of four MMR genes
in an individual or family with colon cancer (de la chapelle A., 2004). Specific criteria applies
that the a priori probability for the detection of a deleterious mutation in one of the four
implicated MMR genes is about 70% in LS (Lynch et al., 2009). Nevertheless, a germline
mutation in a MMR gene is detected in only 50% of all traditionally defined LS cases (de la
Chapelle A., 2004). Step-wise approach to identify LS is widely used, clinical suspicion of LS, by
early onset or family history, is followed by MSI testing (Lynch et al., 2009). MSI can occur in
2 different forms, including a low frequency of instability (MSI-L) or a high frequency of
instability (MSI-H). MSI-H is characterized when the microsatellites in the tumors are unstable
for more than 30%. MSI-L is characterized when 1%-30% of the microsatellites are unstable.
Germline mutation testing is recommended if the tumor is MSI-H. However, most CRCs show
no mutated microsatellite sequences and are MSS (Stewart, A., 2013 & Lynch et al., 2009 &
Bolande et al., 2010). MSI-L is, in contrast to MSI-H, not associated with germline mutations
in DNA MMR genes (Lynch et al., 2009). Additionally, the use of IHC is recommended to help
predicting germline mutations in the MMR genes (Kohlmann, W. & Gruber, S.B., 2012).
These MSI and IHC tests are sensitive and specific in predicting mutations in the MMR genes,
although determination of MSI alone is not sufficient to identify LS as 10-15% of sporadic CRCs
exhibit MSI, caused by methylation-induced silencing of MLH1 (Stewart, A., 2013). Only 2025% of MSI-H tumors are associated with germline mutations in a DNA MMR gene. Evaluation
of MSI status is still the first test to identify tumor characterises predictive of an underlying
MMR mutation (Lynch et al., 2009 & Kohlmann, W. & Gruber, S.B., 2012). In some cases
molecular genetic testing of the tumor for methylation and somatic BRAF mutations is
performed to determine if tumors are more likely to be sporadic than hereditary. If there is
indication of a hereditary form, genetic testing of the MMR genes can be performed
(Kohlmann, W. & Gruber, S.B., 2012 & Stewart, A., 2013 & Lynch et al., 2009).
30
MSI testing
MSI is defined by
the occurrence of
shorter or longer
repetitive
nucleotide
sequences in the
coding regions of
genes and is an
excellent marker Figure 11. MSI testing is used for identifying tumors caused by defective MMR by comparing
for
MMR nucleotide repeats in a panel of markers in normal tissue with the number in tumor tissue
from the same individual. MMS is present if as well tumor as normal tissue show the same
deficiencies.
number of repeats, whereas MSI is present if tumor and normal tissue show different number
However, it is not of repeats (Kohlmann, W. & Gruber, S.B., 2012).
highly specific as it can also occur in tumors with APC gene mutations (Grady W.M. & Carethers
J.M., 2008). MSI testing is still used to identify tumors caused by defective MMR performed
by comparing PCR-amplified microsatellite fragment lengths between tumor and normal
tissue from the same individual (Figure 11).
To include MSI testing in routine testing it was required to standardize and select the optimal
microsatellite sequences. First an initial panel of five microsatellite markers, according to the
Bethesda guidelines, including three dinucleotide repeats and two mononucleotide repeats,
was recommended in 1998. Recently, a new panel of five microsatellite markers has been
identified which can run in a single pentaplex PCR. This panel was shown to be more sensitive
and specific than the original markers, and did not require analysis of normal tissue (Lynch et
al., 2009 & Stewert, A., 2013).
Immunohistochemistry (IHC)
The IHC testing is performed to help identifying germline mutations in the MMR gene by
detecting the presence or absence of the protein products expressed by MMR genes
(Kohlmann, W. & Gruber, S.B., 2012 & Stewart, A., 2013). This can be performed by using
antibodies directed against the common MMR genes, which are commercially available. This
test is of value to supplement MSI testing and is probably 95% sensitive for DNA MMR
deficiency, although IHC cannot distinguish mutant proteins commonly resulting from
missense mutations from wild type polypeptides, what makes the IHC results ambiguous
(Lynch et al., 2009 & Shia J., 2008).
31
Sequencing method
The MLH1 and MSH2 mutations are seen in all regions of these genes, for which it is necessary
to sequence whole the gene for mutations (de la Chapelle A., 2004). There are two possible
DNA sequencing methods, conventional and conversion analysis, respectively. Some MMR
mutations are not detected by conventional DNA sequencing, as the wild-type allele can
“mask” the mutant one. Therefore, a diploid to haploid conversion analysis was developed,
by which the paternal and maternal alleles could be separated, to unmask certain mutations.
This made it possible to detect heterogeneous mutations (Lynch et al., 2009).
4.3.2 The search for new predisposition genes
In order to increase the detection rate during genetic testing new candidate target genes are
needed. In this context the DNA replication apparatus itself came into point of interest,
especially DNA polymerases δ and ε. Mutations in these genes can cause and enhance the
mutation rate during DNA replication as mutations in the proofreading domain may cause a
reduction of the fidelity of DNA replication (Flohr et al, 1999).
Recently Palles et al found possible predisposition genes, POLE and POLD1, which encode for
the exonuclease domains of DNA polymerase ε and δ, respectively. In addition Church et al
found somatic and germline EDMs in these genes that predispose to CRCs or/and ECs (Palles
et al., 2013 & Church et al., 2013 & Heitzer E. & Tomlinson I., 2014). Inclusion of these genes
in routine genetic diagnostics could contribute to an increased detection rate of predisposing
mutations. This can lead to a better understanding of the causes of CRC and to reduce the
morbidity and mortality (de la Chapelle A., 2004).
32
5. Experimental research
5.1 Material and Method
5.1.1 Patient samples
The Institute of Human Genetics, Medical University Graz has a large collection of DNA
samples from patients that underwent routine molecular diagnostics. In the last ten years
more than 200 patients with a putative predisposition to colorectal cancer have been analysed
at the institute. However, in 59 patients no disease causing mutation has been identified in
one of the most common known predisposition genes. These patients were selected for the
mutational screening of the exonuclease domain of POLD1 and POLE that was performed in
this study.
5.1.2 DNA quantification
DNA was quantified using two different methods, i.e. absorbance or optical density
(NanoDrop™-1000, Thermo Scientific) and fluorescent DNA-binding dyes (Qubit® 2.0
Fluorometer, Catalogue no. Q32850).
The “NanoDrop” (ND) spectrophotometer (ND-1000, Thermo Scientific) measures the
absorbance of the heterocyclic rings of the nucleotides by ultraviolet light (260nm). The more
concentrated the DNA solution is, the more UV light will be absorbed according to the law of
Lambert-Beer. However, there are also other components, such as proteins, that can absorb
UV light, and therefore falsify the measurement. Another drawback is that DNA and RNA can
hardly be distinguished from each other by this quantification method (The NanoDrop
Spectrophotometers and Fluorospectrometers, 2008).
The Qubit® 2.0 Fluorometer (Life Technologies) utilizes specifically designed fluorometric
technology using Molecular Probes® dyes. The fluorescent dye will only emit a signal when
bound on DNA. The fluorescence signal is then converted into a DNA concentration
measurement using DNA standards of known concentration (Qubit® 2.0 Fluorometer, 2014).
5.1.2.1
Reagents
Measurement of DNA concentrations with the NanoDrop does not require special reagents.
Reagents for the Qubit are commercially available as Qubit® Assay Kits and provide
concentrated assay reagent, dilution buffer, and prediluted standards. Depending on the
expected concentrations of the samples different assays are recommended. For
concentrations between 10pg/µL–100ng/µL a “high sensitivity” (HS) assay (Qubit® dsDNA HS
assay kit) and for concentration 100pg/µL–1ng/µL the “broad range” (BR) assay (Qubit®
dsDNA BR assay kit) should be used. As the available DNA samples were extracted form whole
33
blood and therefore a high concentration was expected, the Qubit® dsDNA BR Assay Kit (Life
technologies, Carlsbad, USA Catalogue no. Q32850) was used. For an overview of the required
reagents see Table 4.
Table 4. Required reagents for measurement of DNA concentration.
NanoDrop
ND-1000 (Thermo Scientific)
No special reagents needed
Qubit
Qubit® 2.0 Fluorometer (Life Technologies)
Qubit® dsDNA BR Assay Kit (Life Technologies)
 Qubit™ dsDNA BR reagent
= Fluorescent dye 200x concentrated in
“dimethylsulfoxide” (DMSO)
 Qubit™ dsDNA BR buffer
 Standard #1
 Standard #2
5.1.2.2
Protocol
NanoDrop:
For DNA quantification the required analysis module “Nucleic Acid” and “DNA” has to be
chosen in the NanoDrop Software. Before measuring the samples, a blank must be measured
to establish the non-fluorescent signal reaching the detector. Then 1µl of the sample is loaded
on the clean pedestal and analyzed.
Qubit:
For DNA concentration measurement, the Qubit® dsDNA BR Assay was used according to the
manufacturer’s protocol. First, the Qubit working solution was freshly prepared on the day of
use by diluting Qubit™ dsDNA BR reagent 1:200 in Qubit™ dsDNA BR buffer (10µl Qubit™
dsDNA BR reagent with 1990µl of Qubit™ dsDNA BR buffer) and was mixed well by vortexing.
The final volume in each assay tube must be 200μl, therefore the required volume of working
solution was prepared depending on the amount of samples analysed. As the Qubit™ dsDNA
BR reagent is light sensitive, it needed to be light protected to avoid photo degradation.
Second, 10µl of the Qubit™ dsDNA BR standards (#1 respectively #2) were diluted with 190µl
Qubit working solution. Third, 1 µl of the samples were diluted with 199µl Qubit working
solution. Diluted standards and samples were vortexed and incubated for 2 minutes at room
temperature.
Afterwards the standards, for calibrations, and the samples were analysed with Qubit® 2.0
Fluorometer (Life Technologies), using dsDNA BR assay type. Upon completion of the sample
measurement, the concentration is calculated by taking the dilution factor into account.
Accordingly, the volume of the original sample that has been added to the assay tube was
selected under “Calculate Stock Conc“.
34
5.1.3 Polymerase chain reaction (PCR)
“Polymerase chain reaction” (PCR) is a molecular technique that allows selective amplification
of specific DNA regions of interest. The first step is the chemical synthesis of primer DNA
oligonucleotides, which are complementary to the region that need to be amplified, known as
the target sequence. One primer is complementary to the first strand of DNA double helix and
the other one to the second strand. They start at the opposite end of the sequence, through
which they initiate amplification by the enzyme DNA polymerase.
PCR amplification occurs by repeated cycles (usually 30-40 cycles) consistent out of three
temperature dependent steps. In the first step a brief heat treatment, between 94 – 96°C, is
performed to separate the double stranded DNA. This step is called the denaturation step, by
which the hydrogen bonds, that holds the two strands together, brake. In the second step
called annealing the temperature
lowers, which allows the primers to
hybridize to the complementary
single-stranded DNA. This is provided
when the primer oligonucleotides are
present in large excess and when the
temperature cools down until a
temperature that is usually 5°C cooler
than the primer’s melting point.
Finally, during elongation the
temperature is raised to 70-74°C.
Hereby the DNA polymerase activates
the DNA synthesis, by synthesizing the
complementary strand of DNA by
which the primer is the starting point.
DNA and primers are surrounded by
the four dNTPs, which are added to
the PCR reaction, and will incorporate
in the newly synthesized DNA strand
by the polymerase. After the first PCR
cycle is finished the next cycle starts in
the same way, composed out of the Figure 12. The principle of Polymerase chain reaction (PCR) (A
described three steps. Afterwards the revolution in the microbiology laboratory, 2009).
newly synthesize DNA strands can act as new templates for the further amplification.
Eventually, DNA will be cloned several times, depending on the total number of cycles (Figure
12) (Koch, M., 2013 & Polymerase chain reaction (PCR), 2014).
35
5.1.3.1
Reagents
For the performance of a PCR reaction materials according Table 5 were required.
Table 5. Reagents acquired for performing PCR reactions on DNA samples of patients.
Reagents for PCR reaction
Dyad Disciple Peltier Thermal cycler (Biorad)
“Forward” (FWD) and “Reverse” (REV) primers (AG Microsynth)
Nuclease free water
50ng/µl patients DNA samples
80ng/µl Human genomic female DNA (Promega)
Hotstart Taq™ mastermix (Qiagen), containing polymerase, dNTPs, buffer, etc.
5.1.3.2
Primer design
Primers are short strands of nucleic acids that are complementary to the targeted sequence
and are key components to enable selective and repeated amplification of DNA. Primers
specific for the exons encompassing the exonuclease domain of the POLD1 and POLE genes
were designed at the Wellcome Trust Centre of Human Genetics, University of Oxford (Table
6). In total twelve primer combinations, for the different exons of the POLD1 and POLE genes,
were obtained from Microsynth AG.
Table 6. Summary available primer combinations.
Primername and
combination
Primer sequence
Target
POLD1_E8_F
POLD1_E8_R
FWD: 5'-AGG CTG AAG GAG AAG GTG-3' Exon 8 of POLD1 in DNA polymerase ε
REV: 5'- CCT CTG TGC TCA CTC CAT -3'
POLD1_E9_F
POLD1_E9_R
POLD1_E10_F
POLD1_E10_R
POLD1_E11_F
POLD1_E11_R
POLD1_E12_F
POLD1_E12_R
POLD1_E13_F
POLD1_E13_R
POLE_E9_F
POLE_E9_R
POLE_E10_F
POLE_E10_R
POLE_E11_F
POLE_E11_R
POLE_E12_F
POLE_E12_R
POLE_E13_F
POLE_E13_R
5'- TGC TGT GTT GGG AGT GA -3'
5'- CAA GCC TAA TTG CAC AGG -3'
5'- AAC ATT CTG GAA GTA GGG G -3'
5'- ATG TGG AGA GGG AGT GG -3'
5'- AAT CTC CGT TCT TCA GGC-3'
5'- AGC TTG TAC TCC CGC AG -3'
5'- GAC ACC AAG GTT GTC AGC AT -3'
5'- GAA AAA GTG GGC GTC AGG -3'
5'- AGG CTA CCT CAC CCT GAC -3'
5'- GGA GTG GGT ATA GGT GGG -3'
5'- CTA ACA GTG GGG CAG ATG CT -3'
5'- GGT GTT CAG GGA GGC CTA AT -3'
5'- TGA GGC CTT GGA AAG ATC C -3'
5'- TGC TGT GGA CTT CTT TGT AGT GA -3'
5'- CCT AAG TCG ACA TGG GAA GC -3'
5'- GAG GCT GCT GCT TCT GAA CT -3'
5'- TCT GCA AGA GGC CTT CAG AT -3'
5'- GGA GAA AGG GGC ATT AGA GC -3'
5'- CTG CTC CGT GGC CAT CT -3'
5'- TGC CCA GTT TTG CCA GTT -3'
Exon 9 of POLD1 in DNA polymerase ε
Exon 10 of POLD1 in DNA polymerase ε
Exon 11 of POLD1 in DNA polymerase ε
Exon 12 of POLD1 in DNA polymerase ε
Exon 13 of POLD1 in DNA polymerase ε
Exon 9 of POLE in DNA polymerase δ
Exon 10 of POLE in DNA polymerase δ
Exon 11 of POLE in DNA polymerase δ
Exon 12 of POLE in DNA polymerase δ
Exon 13 of POLE in DNA polymerase δ
36
POLE_E14_F
POLE_E14_R
5'- CAG GCT TTG CTT TCT GTG -3' Exon 14 of POLE in DNA polymerase δ
5'- CAC TCC TGG GAC ATC CAC -3'
The obtained lyophilized primers were dissolved in nuclease free sterile water to a final
concentration of 100µM. The PCR primers were then further diluted to a working solution of
10µM.
5.1.3.3
Protocol PCR reaction
All PCR reactions were first tested on human genomic female DNA (Promega), by using all
twelve primer combinations. For a single PCR reaction nuclease free sterile water, 6µl 2x
Hotstart Taq™ Mastermix (Qiagen), 0.5µl of 10µM primer pairs (FWD & REV primers of one
exon, obtained from Microsynth AG, mixed together) and 0.5µl of 80ng/µl human genomic
female DNA (Promega) were mixed to a final volume of 12µl. In addition, two “no template
controls” (NTC) were included, which did not contain human genomic female DNA (Promega).
One NTC contained all POLD1 primer pairs and the other all POLE primer pairs. All PCR tests
were carried out in the Bio-Rad Dyad Disciple Peltier Thermal cycler. By which PCR reactions
were amplified with the cycling conditions displayed in Table 7.
Table 7. PCR cycling conditions.
Phase
Hotstart
Denaturation
Annealing
Extension
Temperature (°C)
Time (sec, min)
95
15 min
95
45 sec
57
45 sec
72
45 sec
35Cycles in total
Final extension
72
7 min
End
8
∞
Since all primer pairs gave sufficient PCR product, no optimisation steps were necessary and
the same PCR protocol was applied to the patient samples. First the patient DNA samples were
diluted to 50ng/µl genomic DNA in a final volume of 100µl nuclease free water. Afterwards a
mastermix was prepared for the PCR reactions on patient samples, of which the composition
is displayed in Table 8.
Table 8. Mastermix preparation for PCR reaction on patient samples.
Component
1x reaction (µl)
2 x Hotstart Taq™ Mastermix (Qiagen)
6
Primer mix (10µM)
0.5
Nuclease free water
4.5
Total volume
11
For every PCR reaction 11µl of mastermix was added to 1µl of 50ng/µl patient DNA sample
and were carried out using PCR program shown in Table 7. In addition, a NTC and a “positive
control” (PC), containing 40ng/µl human genomic female DNA (Promega) instead of patients
37
DNA sample, were included in every run. Afterwards, PCR products were quality checked using
gel electrophoresis.
5.1.4 Gel electrophoresis and Agilent 2100 Bioanalyzer
Gel electrophoresis and Agilent 2100 Bioanalyzer (Agilent technologies) can be used to assay
the DNA amplicons derived from the PCR reactions. Both methods works by a different
principle.
Gel electrophoresis, also known as conventional electrophoresis, is a method based on the
migration of DNA through the gel under influence of an electric field. The negative loaded DNA
fragments migrates through the gel to the positive pole. Separation of fragments with
different size is based on the speed of migration whereby smaller DNA fragments will migrate
faster than bigger DNA fragments. To make the DNA visible, the gel is treated with a
fluorescent dye that emits light under UV, when bound to DNA. To determine the size and
quantity of testing DNA fragments a molecular-weight size marker in the form of a 100bp DNA
ladder (Invitrogen) is carried along.
The Agilent 2100 Bioanalyzer (Agilent Technologies) is a microfluidics-based platform for
quantifying and sizing DNA. Performed according to the manufacturer’s protocol based on the
same principle as gel electrophoresis, by which the DNA fragments are separated depending
on their size. But samples are analyzed, in contrast to conventional electrophoresis, on a
microchip, also known as lab-on-a-chip. The Bioanalyzer kit contains a DNA ladder from 100
to 7500bp, DNA markers, DNA dye concentrate and DNA gel Matrix (Pischler, C., 2013).
5.1.4.1
Reagents
For the PCR quantification using conventional gel electrophoresis or the Agilent 2100
Bioanalyzer (Agilent Technologies) reagents according Table 9 were required.
Table 9. Reagents acquired for performing a PCR quantification by gel electrophoresis and by the Agilent 2100
Bioanalyzer (Agilent Technologies).
Gel electrophoresis
Agarose (C12H18O9 from Biozym)
Red nucleic acid stain (10,000x, Biotium)
1x “tris-boraat-EDTA” (1x TBE)
Dye buffer
PCR sample
100bp ladder (Invitrogen)
A chamber
An electrophoresis machine
Agilent 2100 Bioanalyzer
(Agilent Technologies)
The 2100 Bioanalyzer instrument
A chip priming station
A chip vortexer
2100 expert software for instrument
control and data analysis
The DNA 7500 kit
PCR samples
38
5.1.4.2
Protocol
Gel electrophoresis:
For quality check of PCR products a 1% agarose gel was used. For this reason 0.8g agarose
(C12H18O9 from Biozym) were added to 80-85ml 1x TBE. After cooking the mixture for 2 to 3
minutes into the microwave, 8µl of gel Red nucleic acid stain (10,000x from Biotium) was
added to the lukewarm solution. Right after the gel was hardened it was brought into the gel
chamber and immersed into 1x TBE.
PCR samples were prepared by adding 3µl dye buffer to 5µl of the PCR reactions. 4-5µl of
these preparations were loaded on the gel together with 5µl of the 100bp ladder (Invitrogen).
The gel ran at room temperature for 40 minutes at 130V. After gel electrophoresis the PCR
products were visualized with standard UV illumination without additional post-run staining.
Bioanalyzer:
According to the manufacturer’s protocol it was indicated to incubate the Bioanalyzer kit for
30 minutes at room temperature. The DNA chip was loaded on the chip priming station and
9µl of gel dye mix was pipetted in the marked G well. After that the chip priming station was
closed and the 1ml plunge was put into position by pressing it for 30 seconds on 1 ml, the
plunger was slowly pulled back over 5 seconds. Then the other G marked wells were filled with
9µl of gel-dye mix. Then 5µl of marker was pipetted in sample and ladder wells. For wells with
no samples 6µl of marker was loaded. Subsequently, 1µl of the ladder was pipetted in the
ladder well and 1 µl of sample in the sample wells. After vortexing the chip for 1 minute, by
gently increasing the “rounds per minute” (rpm) to 2400rpm, it was analyzed by the Agilent
2100 Bioanalyzer (Agilent Technologies) within 5 minutes.
5.1.5 Sanger sequencing
Sanger sequencing, also called the didesoxynucleotide or termination method, is an
oligonucleotides sequencing method via enzymic polymerization. It is considered as a ‘first
generation’ sequencing technology, discovered by Sanger and his co-workers in the seventies
(Metzker M. L., 2010 & França et al., 2002 & BigDye® Terminator v3.1 Cycle sequencing Kit
from Applied biosystems, 2006).
The principle of this method is based on the ability of DNA polymerase to incorporate
analogous of nucleotide bases, i.e. 2’, 3’ – “didesoxyribonucleoside triphosphate” (ddNTPs) in
the new DNA strand. When ddNTPs are incorporated at the 3’ end of the chain, the chain
elongation is selectively terminated due to the lack of the OH-group in ddNTPs. Many
improvements, to detect the incorporated ddNTP, were made. For example the fluorescent
labelling of each ddNTP which made it possible to determine the incorporated ddNTP at the
end of the chain. Each ddNTP contains another dye and are added sequentially to the primer
39
through a cycle of sequencing reactions (BigDye® Terminator v3.1 Cycle sequencing Kit from
Applied Biosystems, 2006).
However, before the sequencing
reaction can take place the DNA
needs to be amplified by PCR.
Sequencing reaction, itself is also
PCR-based and is performed
during a ‘cycle sequencing’ Figure 13. ‘Cycle sequencing’ reaction (What is Sanger sequencing/ Dye terminating
reaction, in which cycles of sequencing, 2013).
template denaturation, primer annealing and primer extension are performed resulting in
different sized “single stranded” (ss) DNA (Figure 13) (Shendure J. & Ji H., 2008 & França et al.,
2002).
This reaction can take place in a single tube,
containing dNTPs, fluorescent dyed ddNTPs,
buffer, primers (either FWD or REV). By which the
primers are complementary to the sequence
immediately flanking the region of interest. These
provide a starting point for DNA synthesis, as
isothermal DNA polymerase can bind hereupon.
Taq polymerases were not suitable for sequencing
as they prefer incorporation of ddNTPs rather than
dNTPs. In addition, the concerning polymerase
needed to lack proofreading ability to avoid uninterpretable data, explaining the choice of
isothermal T4 or T7 DNA polymerases in the DNA
sequencing methods. All these products, except
the primers are present in the BigDye® Terminator
v3.1 kit, which needs to be kept cold for
maintaining polymerase function. The sequencing
reaction will produce a series of molecules of
different lengths, each one terminated and Figure 14. Depicture of automated Sanger
labelled at a different base (Figure 14). This is sequencing procedure (DNA sequencing, 2011).
possible by the fact that the ratio dNTPs/ddNTPs in the reaction is calculated so that the
termination happens at least once for every position of the template.
Purification was performed using Sephadex columns. Sephadex™ G-50 Superfine (GE
Healthcare) is a powder that needs to be dissolved in nuclease free water. The solution needed
to be incubated at room temperature to let the Sephadex™ G-50 Superfine powder absorb
40
the water and thicken. Afterwards it became gelatinous, ideal to filter the sequence reactions
and to concentrate the DNA strands.
The purified sequencing reactions were analysed using the 3130 xl Genetic Analyzer (Life
Technology) by which the different DNA strands were separated by “capillary
electrophoreses” (CE) using a CE-based polymer gel. This technique is based on the separation
of the DNA strands in a narrow-bore fused silica capillary under influence of a high electric
field. Small fragments will leave the capillary first and pass a laser, which emit a spectra
depending on which fluorescent labelled ddNTP the fragment is terminated. Afterwards a
software (SeqScape) can translate these sequencing traces into DNA sequences (Shendure J.
& Ji H., 2008 & França et al., 2002 & BigDye® Terminator v3.1 Cycle sequencing Kit from
Applied biosystems, 2006).
5.1.5.1
Reagents
For Sanger sequencing and clean-up of the sequence reactions, reagents according Table 10
were required.
Table 10. Reagents acquired for performing a Sanger sequencing and clean-up of the sequence reactions
Sanger sequencing reaction
Thermocycler
Big dye 3.1 ( “Applied biosystems” (AB))
5x sequencing buffer (AB)
FWD & REV primers
PCR samples
Nuclease free water
5.1.5.2
Clean-up
Sephadex™ G-50 Superfine (GE Healthcare)
Strip of columns or column plate
Elution eppis or elution plate
Centrifuge
Protocol
Sanger sequencing:
Depending on the amount of available PCR products, either 0.5µl or 1µl of the PCR product
was used for sequencing. As well FWD as REV primers were used in different reactions.
Composition of the mastermix is shown in Table 11.
Table 11. Preparation mastermix for Sanger sequence reaction on a high (and a low) concentrated PCR product.
Component
Big dye 3.1 (AB)
5x sequencing buffer (AB)
1x FWD reaction (µl)
0.5 (1.0)
1.4
1x REV reaction (µl)
0.5 (1.0)
1.4
FWD or REV Primer (10µM)
0.3 (FWD)
0.3 (REV)
Nuclease free water
Total volume
7.3 (6.8)
9.5 (9)
7.3 (6.8)
9.5 (9)
41
For every Sanger sequence reaction with high concentrated PCR products 0.5µl of PCR samples
were added to 9.5µl of mastermix, whereas for low concentrated PCR products 1µl of PCR
samples were added to 9µl of mastermix. The cycling condition is shown in Table 12.
Table 12. Sequencing conditions for Sanger sequencing.
Phase
Denaturation
Annealing
Extension
End
Temperature (°C)
96
50
60
25 Cycles in total
6
Tim (sec, min)
30 sec
15 sec
4 min
∞
Clean up:
Sephadex was prepared by mixing 6-7 spoons of Sephadex™ G-50 Superfine (GE Healthcare)
in 45-50ml nuclease free water. This solution was incubated for 30 minutes at room
temperature. Then 20µl of nuclease free water was added to 10µl sequence reaction samples.
Depending on the sample size two slightly different protocols were used including strips with
8 columns per strip or column plates containing 96 columns in one plate.
The column strips were prepared by filling 2/3 of the columns with Sephadex, by which the
columns always needed to be featured with elution strips. These were centrifuged for 7
seconds at 750 “rotational centrifuge force” (rcf). Elution strips were emptied in the sink.
Afterwards the columns were centrifuged again for 2min at 750rcf. Hereafter the columns
were featured with new elution strips and the whole volume of the sequence reaction samples
were loaded on the columns. Elution was performed by centrifuging the columns for 2 minutes
at 750rcf. These elutates were then transferred to the wells of a 96-well plate, special for the
3130 xl Genetic Analyser (Life Technology).
For the column plates a similar protocol was performed. The columns of the plate were filled
with 300µl Sephadex solution and featured with an elution plate. The plate was centrifuged
for 2 minutes at 1500rcf, after which the plate was featured with a new elution plate (special
for 3130 xl Genetic Analyser, Life Technology). Then the whole volume of sequence reaction
samples were loaded on the columns and elution was performed by centrifuged the plate for
2 minutes at 1500rcf. It was necessary to take into account that the elutates volumes in the
elution plate were at least 7.5µl and did not contain any air bells. Afterwards the used columns
were cleaned out as Sephadex dries out and becomes sticky, which can silt the columns.
Then the plate was subjected to the 3130 xl Genetic Analyser (Life Technology). After checking
the presence of enough polymer, buffer and water, the analyser could be programed to
analyse the samples. After analysis the results could be interpreted by using the program
SeqScape.
42
5.1.6 Statistics
Two sided Fisher’s exact tests was used to compare the single SNP genotype frequencies
between patients and controls, to check for non-random associations. The fisher exact test is
more accurate for small population tests than the Chi-Square test. This involves an exact
calculation of the probability (p-value) by which two hypothesis are prepared including the
“null hypothesis” (H0), which defines that populations are not significant, and the “alternative
hypothesis” (H1), which defines that populations are significant. The H0 can be rejected when
p < 0.05, by which it is 95% probable that the population is significant (Weisstein, E. W., 2014).
43
5.2 Results
5.2.1 DNA quantification
For all samples NanoDrop concentration values were on average 1.58 times higher than Qubit
concentrations. Linear regression was used to compare dilutions of the different DNA samples,
considering each measurement as an individual point. Results of the NanoDrop are plotted on
the x-axis and the results of the Qubit are plotted on the y-axis. The concordance between
measured values for both technologies contained a correlation coefficient (R2) of 0.773 (Figure
15).
NanoDrop VS Qubit (2)
200
180
C(DNA) Qubit (ng/µl)
160
y = 0,5739x + 10,416
R² = 0,773
140
120
100
80
60
40
20
0
0
50
100
150
200
250
300
350
C(DNA) NanoDrop (ng/µl)
Figure 15. Comparison of DNA concentration measurement with NanoDrop and Qubit, C = concentration (ng/µl).
44
5.2.2 PCR quality check
5.2.2.1
Gel electrophoresis
For the establishment of the PCR reactions with the different primer combinations, test
reactions with human genomic female DNA were performed. By which the PCR products were
quality checked on a 1% agarose gel. All PCR reactions produced specific PCR products and no
further optimization of PCR conditions were necessary (Figure 16).
Figure 16. Gel electrophoresis of test PCR
45
5.2.2.2
Agilent 2100 Bioanalyzer
The same PCR products were additionally analysed on an Agilent Bioanalyzer (Agilent
Technologies). Results were highly concordant with those of the conventional electrophoresis
(Figure 17). However, quantity and size could be determined more accurately compared to
conventional electrophoresis (Table 13). Every individual result is showed in appendix p. I-V.
Figure 17. Overview results Bioanalyzer.
Table 13. Overview individual samples.
Sample
1
2
3
4
5
6
7
8
9
10
11
12
Definition
PCR with POLD1_E8
PCR with POLD1_E9
PCR with POLD1_E10
PCR with POLD1_E11
PCR with POLD1_E12
PCR with POLD1_E13
PCR with POLE_E9
PCR with POLE_E10
PCR with POLE_E11
PCR with POLE_E12
PCR with POLE_E13
PCR with POLE_E14
Bp Size (bp)
Concentration (ng/µl)
386
387
269
329
411
425
262
303
271
263
286
297
30,49
19,62
23,00
3,53
19,86
27,88
19,00
14,42
18,41
9,94
13,86
14,19
46
5.2.3 Sanger sequencing
To test whether the quality of the PCR products is sufficient for Sanger sequencing, seven PCR
product of the human genomic female DNA (Promega) were Sanger sequenced. This included
3 POLD1 (E8, E9 and E12) and 3 POLE (E10, E11 and E14) PCR samples with a high concentration
and one POLD1_E11 PCR sample with a low concentration of amplicon. The results were
analysed by the program SeqScape, by which NM_001256849 and NM_006231.3 were used
as reference for POLD1 and POLE genes, respectively. Sanger sequencing of all samples, except
for POLD1 exon 8, succeeded. Every succeeded sequence product of the samples were
compared with the reference of the related gene. None of these first trial samples showed a
mutation (for results see appendix p. V-VI).
5.2.4 Patient characteristics
The mean age of patients was 50 years old, ranging from 15 to 83 years of age. 61.02% of the
patients were male and 38.98% were female. All these patients had clear indication for genetic
testing with respect to familial colon cancer. Initially 57.63% of the patients were tested for
LS and 42.37% for FAP. In FAP 60% of the patients were male and 40% of the patients were
female, as for LS, 61.76% of the patients were male and 38.24% of the patients were female.
For LS, immunohistochemistry of available tumor samples were performed before molecular
genetic testing of special MMR-genes. Depending on which MMR-protein (MSH2, MSH6,
MLH1 or PMS2) showed decreased expression, the corresponding gene was tested by Sanger
sequencing. Furthermore in most patients MSI analysis was performed on tumor samples to
predict a deficient MMR-system. In patients that showed a FAP indication, the APC gene was
initially screened by Sanger sequencing for mutations.
5.2.5 Mutation detection of patient samples
PCR and Sanger sequencing of patient samples were performed on human genomic DNA
samples extracted from whole blood, to identify mutations in the exons encompassing the
exonuclease domain of the POLD1 and POLE genes. For both genes, POLD1 and POLE, six exons
were screened. Five patients of the 59 were excluded from research population as their DNA
samples were not found in the large collection of the Institute of Human Genetics, University
of Graz.
In total, two sequence alterations causing amino acid substitutions, one in exon 8 of POLD1
and one in exon 13 of POLE gene, respectively, were detected. In addition, one very common
“short nucleotide polymorphism” (SNP) was identified in exon 12 of POLD1 together with a
total of six intronic SNPs. An overview of all identified mutations and SNPs in patients are
shown in Table 14.
47
5.2.5.1
Overview mutations found in patients
Table 14. Sample information and detected mutations in patients.
Patient
No.
DNA
No.
Sex
Age
(y)
Indication
Genes
sequenced
with Sanger
POLD1
POLE
1
7858
M
23
FAP, aFAP
APC, MUTYH
c.1243-50T>G
2
8275
F
41
aFAP
MUTYH
c.1243-50T>G
3
8294
M
39
FAP
APC, MUTYH
c.1243-50T>G
4
8925
F
41
aFAP
MUTYH
5
9696
M
39
6
10888
M
54
NA
10941
M
48
MLH1, MSH2,
MSH6
MLH1, MSH2,
MSH6
MSH2
NA
7
8
10957
F
43
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
NA
c.910-6G>C,
c.1359+43G>A
NA
9
11071
M
59
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
aFAP
c.1137+53G>A,
c.1243-50T>G
NA
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
NA
10
11
11374
11728
M
M
33
15
FAP, aFAP
FAP
APC, MUTYH
APC
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
c.1359+43G>A
12
11944
F
51
MSH2, MSH6
13
12155
F
57
14
14178
F
73
MLH1, MSH2,
MSH6
MSH6
c.1137+53G>A,
c.1243-50T>G
c.1243-50T>G
15
14305
M
27
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
MSH2, MSH6
16
14566
F
26
aFAP
MUTYH
p.Arg306Ile307delinsLeu
/c.917-919Del.,
c.1243-50T>G
c.1243-50T>G
17
14844
F
82
MHL1
c.1243-50T>G
18
14981
M
46
15300
M
18
MLH1, MSH2,
MSH6
MLH1, MSH6
c.1243-50T>G
19
20
21
15482
16771
F
M
45
32
MUTYH
MLH1
c.1243-50T>G
c.1243-50T>G
22
16772
M
45
MLH1, PMS2
23
16866
M
38
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
aFAP
Lynch
syndrome
Lynch
syndrome
aFAP
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
MLH1, MSH2,
MSH6
MUTYH
MUTYH
NA
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
c.910-6G>C,
c.1359+43G>A
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
NA
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
c.910-6G>C,
c.1359+43G>A
48
Patient
No.
DNA
No.
Sex
Age
(y)
Indication
Genes
sequenced
with Sanger
24
16930
M
38
Lynch
syndrome
MLH1, PMS2
c.1686+47G>A,
c.1243-50T>G
25
26
17145
17355
M
F
55
70
aFAP
Lynch
syndrome
MUTYH
MLH1, MSH2,
MSH6
c.1243-50T>G
c.1137+53G>A,
c.1243-50T>G
27
17888
M
38
Lynch
syndrome
MLH1, MSH2,
MSH6
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.910-6G>C,
c.1359+43G>A
28
18587
M
42
Lynch
syndrome
MLH1, MSH2,
MSH6
c.1137+53G>A,
c.1243-50T>G
c.910-6G>C,
c.1359+43G>A
29
18720
M
72
Lynch
syndrome
MLH1, MSH2,
MSH6
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.910-6G>C,
c.1359+43G>A
30
18818
F
59
PMS2
19449
F
56
c.1243-50T>G,
c.1686+47G>A
c.1243-50T>G
c.910-6G>C
31
32
19812
F
38
33
20080
M
79
34
20299
F
45
35
36
20833
21307
M
M
47
50
37
21802
F
57
38
22795
F
65
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
Lynch
syndrome
FAP
Lynch
syndrome
Lynch
syndrome
aFAP
39
22886
M
44
40
23028
M
42
41
23061
M
74
42
23166
M
43
23432
44
45
MLH1
Msh2
POLD1
POLE
p.Ala428Thr/
c.1282G>A,
c.910-6G>C
c.910-6G>C
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.1359+43G>A
MLH1
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
MLH1
c.1243-50T>G
APC
MSH6
c.1243-50T>G
c.1243-50T>G
MLH1, PMS2
MUTYH
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
Lynch
syndrome
FAP
MLH1, MSH2
NA
APC
c.910-6G>C
MLH1, PMS2
64
Lynch
syndrome
FAP, aFAP
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
F
43
FAP
APC
23616
F
56
FAP
APC
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
23657
M
54
Lynch
syndrome
MLH1
c.1243-50T>G
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1226+45C>T,
c.1359+43G>A
c.910-6G>C
APC, MUTYH
c.910-6G>C
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
c.910-6G>C,
c.1226+45C>T
c.910-6G>C
c.910-6G>C,
c.1359+43G>A
NA
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
49
Patient
No.
DNA
No.
Sex
Age
(y)
46
23898
M
67
47
23899
M
43
48
24090
M
65
49
24564
M
83
50
24611
M
64
51
24625
F
62
52
24659
F
53
25178
54
Indication
Genes
sequenced
with Sanger
POLD1
POLE
Lynch
syndrome
Lynch
syndrome
FAP, aFAP
MLH1, PMS2
c.1243-50T>G
c.910-6G>C
MSH2, MSH6
c.910-6G>C
Lynch
syndrome
aFAP
MLH1, PMS2
c.1243-50T>G,
c.1686+47G>A
c.1243-50T>G,
c.1686+47G>A
c.1243-50T>G
MLH1, PMS2
25
Lynch
syndrome
FAP, aFAP
APC, MUTYH
c.1137+53G>A,
c.1243-50T>G
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.1243-50T>G
M
55
FAP
APC
c.1243-50T>G
25347
M
72
MSH6
55
56
25606
25638
M
F
41
65
Lynch
syndrome
aFAP
aFAP
57
26021
F
66
aFAP
MUTYH
c.1137+53G>A,
c.1243-50T>G
c.1243-50T>G
c.1243-50T>G,
c.1686+47G>A
c.1243-50T>G
58
26132
M
60
FAP
APC
59
26149
F
45
APC, MUTYH
MUTYH
MUTYH
MUTYH
Lynch
MLH1
syndrome
M = male, F = female, y = years old, NA = not analysed
c.1137+53G>A,
c.1243-50T>G
p.Thr495Thr/ c.1485C>T,
c.1243-50T>G
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C
c.910-6G>C
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.1359+43G>A
c.1359+43G>A
c.910-6G>C,
c.1226+45C>T
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
c.910-6G>C,
c.1359+43G>A
50
5.2.5.2
Putative mutations
POLD1 mutation exon 8 patient 15
In patient 15 a deletion of three nucleotides in exon 8 of POLD1 was identified (c.917919delGCA) (Figure 18). The deletion results in a change of the reading frame and causes a
deletion from two amino acids, i.e. arginine and isoleucine, and an insertion of the amino acid
leucine indicated in Table 15 (p.Arg306Ile307delinsLeu). The presence of the mutation was
confirmed in a second sequencing run.
Figure 18. Deletion in exon 8 of POLD1 in patient 15 analyzed and identified by SeqScape.
Table 15. Description of the mechanism of the mutation found in patient 15.
Protein
Wild-Type
R
I A
P
CGC ATT GCG CCC
Mutant protein
Mutation
L
A P
CTT GCG CCC
R = arginine, I = isoleucine, A = alanine, P = proline, L = leucine
As the mutation is neither reported in the literature nor in publically available mutations
databases it can be considered as novel. Furthermore the mutation has not been reported in
dbSNP or the 1000 genome project. To get more insight into the functional impact of this
mutation we checked whether the amino acids were conserved among different species.
Therefore, the protein conservation was checked on the DNA level by UCSC genome browser.
Both were highly conserved, although the arginine on position 306 was less conserved then
the isoleucine on position 307. This was also confirmed by 100 vertebrates’ base wise
conservation and the in silico GERP-Score (Appendix page VII).
Patient 15 is currently 27 years old and developed an osteosarcoma at the age of 11 and colon
cancer at the age of 22. The early onset and the development of two different cancers strongly
51
indicates a high penetrance germline mutation although there is no clear family history of
cancer (Figure 19).
Figure 19. Pedigree patient 15.
POLE mutation exon 13 patient 24
In patient 24 a heterozygous missense mutation in exon 13 of the POLE gene was identified.
This mutation includes a G to A transition, on nucleotide position 1282 (c.1282G>A). This
mutation, on the first nucleotide of the codon, causes an amino acid change from alanine to
threonine at residue position 428 of the protein (p.Ala428Thr). The mutation was also
confirmed in a second independent sequencing run. Although the mutation was listed in
dbSNP (rs150032060) the mutation was not present in 2276 individuals (4552 chromosomes).
Therefore, this mutation does not turn out to be a SNP, outlined in Table 16. However the
biological significance of the amino acid change is not clear. In silico analysis to determine the
impact of the mutation on the protein function revealed inconclusive results. However, as the
affected amino acid position is not conserved between POLE and POLD1 and it is located right
on the edge of the Exo domain, the mutation might not be a true pathogenic variant.
52
Table 16. Mutation in exon 13 of POLE in patient 24 analyzed and identified by SeqScape with the corresponding results
according dbSNP on NCBI.
Results SeqScape
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs150032060
 NM_006231.2: c.1282G>A
 NP_006222.2: p.Ala428Thr
 “Minor allele frequency” (MAF) : NA
The protein conservation was checked on DNA level by UCSC genomes, by which the Alanine
was shown to be conserved among different species. After checking 100 vertebrates Basewise
conservation and the in silico GERP-Score it showed that the first nucleotide from the affected
codon was not conserved (Appendix page VII).
Patient 24 developed colon cancer at the age of 33 which indicates a predisposition for
colorectal cancer. Except the grandmother that died of breast cancer at the age of 42 there
was no clear family history of cancer (Figure 20).
Figure 20. Pedigree of patient 24
53
5.2.5.3
Single nucleotide polymorphisms (SNP)
A single nucleotide polymorphism, or SNP, includes the variation in a DNA sequence among
individuals at a single position. If a variant is present in more than 1% of a population the
variation can be classified as a SNP. SNPs are found in both, coding and noncoding regions of
the DNA (SNP, 2013).
A total of 7 SNPs were detected in both POLE and POLD1 genes as outlined in Table 17. The
only exonic SNP (rs2230245) was detected in POLD1. All other 6 SNPs were intronic sequence
alterations in POLD1 and POLE. In addition, the allele frequencies from the general population
as indicated in dbSNP are presented in Table 17. Two sided Fisher’s exact tests were used to
compare the single SNP genotype frequencies between patients and controls (general
population). However, no significant changes in allele frequencies were observed as the Pvalue turned out to be almost 1 for all SNPs, which indicates acceptance of H0, i.e. no
significance between populations.
Table 17. Overview of the defined SNPs found in exonic and intronic sequences of the POLE and POLD1 genes in patients.
SNP
c. Position
p. position
Patient’s genotype
# patients
(percentage patients (%))
WT/WT
rs2230245
rs1673043
rs1673041
rs3219399
rs4077170
rs5744761
rs4883555
c.1485C>T
p.Thr495Thr 41
(75.93)
c.1137+53G>A NA
47
(87.04)
c.1243-50T>G NA
0
(0.00)
c.1686+47G>A NA
50
(92.60)
c.910-6G>C
NA
7
(12.96)
c.1226+45C>T NA
51
(94.44)
C.1359+43G>A NA
16
(29.63)
WT/MUT
MUT/MUT
13
(24.07)
7
(12.96)
23
(42.60)
0
(0.00)
17
(31.48)
3
(5.55)
29
(53.70)
0
(0.00)
0
(0.00)
31
(57.41)
4
(7.41)
30
(55.55)
0
(0.00)
9
(16.67)
Genotype dbSNP
Percentage (%)
WT/WT
WT/MUT
MUT/MUT
84.70
13.60
1.70
68.50
25.80
5.60
3.50
33.60
62.80
97.80
2.20
0.00
13.30
45.00
41.70
81.60
17.20
1.10
15.60
46.70
37.80
# = quantity, WT = wild type, MUT = mutant, NA = not analysed
In appendix p. VIII - X more detailed results of these SNPs can be found, including the SeqScape
result and the corresponding allele frequencies of control population according dbSNP.
54
5.3 Discussion
5.3.1 Technical considerations
DNA quantification
For efficient PCR and sequencing performance, accurate DNA quantification is necessary.
Therefore, the DNA samples included in this study were quantified using two different
methods, i.e. NanoDrop and Qubit. Results obtained using NanoDrop revealed higher
concentrations compared to the Qubit fluorimeter which was not unexpected as
spectrophotometric methods do not measure nucleic acids directly, but the absorbance of
everything in the sample. The two methods were correlated (R2=0.773), i.e. 77.30% of the
variance of Y is explained or predicted by the variable X. This is not a perfect- but still a high
positive correlation. The correlation implies that both methods gave similar results and are
useful for estimating DNA concentrations. Taken together, the NanoDrop was less specific
then the Qubit. Therefore, the Qubit method is elected for obtaining more specific DNA
concentrations, whereas the NanoDrop is elected for quick determinations. As the results
from the Qubit were more specific, due to the fluorescence method, these results were used
in further calculations in this project.
PCR quality check
DNA gel electrophoresis is a technique used for the detection and separation of DNA
molecules. In this study PCR products of the test PCRs were quality checked with conventional
gel electrophoresis and with the Agilent 2100 Bioanalyzer (Agilent Technologies). In
conventional gel electrophoresis only qualitative analysis of the PCR product is possible, which
means either presence or absence of the PCR product can be observed. However, an
estimation of a concentration can be performed by reviewing the strength of the band signal.
In contrast with the Agilent 2100 Bioanalyzer (Agilent Technologies) (Figure 17) more accurate
and specific results can be obtained in determining the bp size and amplicon concentrations
(Table 13 and Appendix p. I-V). Generally, both methods have their own pros and cons. The
Agilent 2100 Bioanalyzer (Agilent Technologies) was much more specific, but is very expensive
and could only measure 12 samples at a time. Although gel electrophoresis is less specific,
samples were analysed with gel electrophoresis due to its more efficient throughput and
simplicity.
55
Sanger sequencing reaction
For the mutational screening of the exonuclease domains of POLD1 and POLE, Sanger
sequencing was selected, even when it is slipping out of the limelight as next-generation
sequencing came onto the scene. Although panel sequencing with new sequencing
technologies are becoming more and more important, Sanger sequencing is still routinely
used in research and clinical labs, mainly for SNP genotyping or mutation detection of small
genes (Curtin, C., 2013). The key strength of Sanger sequencing is that it is the most available
technology today and that its well defined chemistry makes it the most accurate method for
sequencing, especially for small DNA fragments of 500bp to 1kb in length (Rizzo J.M. & Buck
M.K., 2012). Amplifying and Sanger sequencing of a small part of the genome is still cheaper
than performing a library preparation for next-generation sequencing (Curtin, C., 2013).
Because in our project it is only needed to amplify and sequence several exons from POLD1
and POLE, Sanger sequencing was preferred as method of use.
5.3.2 Interpretation of mutational screening results
Molecular testing in CRC is important to identify individuals with mutations in one of the
cancer predisposition genes. In this way the disease can be identified in an early, localised and
often symptomless stage thus increasing the survival rate (de la Chapelle A., 2004 & Palles et
al., 2013). However, in 70% of the patients, no germline mutations can be identified despite a
high a priori probability that they carry a high-penetrance germline mutation attributing to
cancer development.
Therefore, there is a high need for identifying new predisposition genes for cancer in routine
diagnostic, in order to ensure an early diagnosis of the disease in at high risk-cancer patients.
For CRC there are ten predisposition genes known, which are tested in routine diagnostics.
When patients are positively tested for a germline mutation in one of these genes, it is certain
that these persons have a higher risk for developing cancer. However, a negative result does
not mean the patient has no increased cancer-risk at all, it is possible that the mutated gene
is not known or has not been identified yet. For this reason intensive research in detecting
predisposition genes is needed to cover a greater percentage of patients to explain their
cancer. Recently, mutations identified in POLD1 and POLE genes became a point of interest in
predisposing CRC and EC.
The aim of the present thesis was a mutational screening the exons encoding the exonuclease
domain of the POLD1 and POLE genes. The 3’ exonuclease domain is responsible for the
proofreading activity of polymerase δ and polymerase ε, respectively. In total 59 patients were
initially included in the study, though only 54 could be analysed due to missing DNA samples.
All patients previously underwent routine molecular genetic diagnostics due to clinical
indications resembling LS or FAP, which are high-penetrance CRC phenotypes. However, after
56
sequencing of the most common disease causing genes, no mutation were found. These
patients belong to the 70% by which no germline mutations could be identified despite the
high a priori probability that their cancer may attribute to a high penetrance germline
mutation.
We chose our patient population for mutational analysis of the EDMs in the POLD1 and POLE
genes because of its resemblance with the population from Palles et al (2013). It has already
been confirmed that EDMs in these genes confer a high risk of multiple colorectal adenomas
and carcinomas, proven by the mapping of previous reported mutations onto a hybrid
structure of yeast DNA polymerase and T4 polymerase. This mapping implicated that the
EDMs mostly lie along the DNA binding pocket of the exonuclease domain (Heitzer E. &
Tomlinson I., 2014 & Palles et al., 2013 & Church et al, 2013 & Briggs S. & Tomlinson I., 2013).
Yoshida et al showed that mice carrying artificial alleles with substitution at essential amino
acids residues in the proofreading domains of polymerases ε and δ contained a mutator
phenotype. Lange et al confirmed the presence of this mutator phenotype in POLD1 and POLE
exo-deficient mice (Lange et al., 2011 & Yoshida et al., 2011 & McCulloch S.D. & Kunkel T.A.,
2008).
Therefore, the aim of this thesis was to identify new mutations, or to confirm previous found
mutations in these genes. This in order to underpin the hypothesis that these EDMs could
predispose for CRCs and ECs. Results of this study should help to evaluate whether inclusions
of POLD1 and POLE as predisposition genes in routine molecular diagnostics might contribute
to an increased percentage of detection and a decreased development of these common CRCassociated tumor syndromes. After screening 54 patients a total of two putative EDMs were
identified, one in POLD1 and one in POLE. In addition several common SNPs were detected.
Putative EDMs
Two putative EDMs in both POLD1 and POLE genes were identified. These mutations were
both found in the exons of the exonuclease domains of the concerning polymerase genes and
were not defined as SNPs according dbSNP. Patients 15 and 24 both showed indication of LS
but did not harbour predisposition in MSH2 and MSH6 genes.
POLD1 EDM in exon 8 of Patient 15
Patient 15 developed an osteosarcoma at the age of 11 and colon cancer at the age of 22
which strongly indicates a possible predisposition for a cancer syndrome. Although he did not
show a clear family history for a predisposition to cancer (Figure 19). The brother of the
grandfather developed pancreas cancer at the age of 55, and the daughter of this brother,
developed abdominal and colon cancer at the age of 46. Except those two cases, no other
family member had any sign of cancer.
57
Assuming that the identified mutation in POLD1 was indeed pathogenic, the mutation would
be inherited from the mother’s family branch, as the cancer cases are seen in this part of the
family. The fact that no other family member suffered from cancer could be attributed to
incomplete penetrance, whereby the genotype could be present but the phenotype is not
observable. Anything that shows less than 100% penetrance is an example of incomplete
penetrance (Mika I., 2008). This could be explained by the “two-hit” model of Knudson, by
which the first hit, the putative pathogenic mutation in POLD1, is inherited and is the precursor
to the tumor whereby all cells of the body are one-hit clones. The second hit still needs to
occur in one of these cells causing the occurrence of cancer (Knudson et al., 2002). However,
when a second hit does not occur, the individual will not show a cancer phenotype but carries
the mutated allele and can inherit them to the descendants. The probability of inheritance for
a mutated allele is 50%. Nevertheless, penetrance of a pathogenic mutation for colorectal
cancer may be highly variable. In other words, half of the children of the affected parent could
carry the gene and in case of a high-penetrance there is a high chance that these kids will
develop cancer after a second hit. If this was the case in the respective family of patient 15, it
would be very likely that at least some of the 5 children of the grandfather, whose brother
had cancer, carried a mutated allele. However, this was not the case what impeaches this
theory together with the fact that POLE and POLD1 are thought not to act as classical “two
hit” tumor suppressor genes (Heitzer E. & Tomlinson I., 2014). It would be enlightening to test
more family members to elucidate whether the mutation segregates with the disease. In view
of the fact that all other affected family members are deceased, it remains unclear whether
the mutation is disease-causing or rather a neutral variant.
Nevertheless, to gain more insight in the segregation of the mutation, both mother and father
of patient 15 should be tested for the presence of the mutation. If the mutation was inherited
from the mother, it may suggest that the mutation might be disease-causing but has a low
penetrance as there were only two cases with early onset cancer in this family branch. If the
mutation was inherited from the father’s family branch, one could conclude that the mutation
might not be pathogenic as the father and his family did not develop cancer. So, for now the
found mutation in patient 15 cannot be linked to an inherited trait. Another possibility is that
patient 15 inherited several low penetrance mutations from both sides that contributed to his
cancer onset.
A further explanation would be that the putative cancer-causing mutation of patient 15 is de
novo. If this was the case, one of the germ cells of the parents could be affected with the
mutation and inherited it to patient 15. Nevertheless, in addition to a segregation analysis
further research, especially in terms of the functional impact of this mutation, is needed.
The mutation in patient 15 was found in exon 8 of POLD1. The deletion of three nucleotides,
c.917-919del, lead to the deletion of two amino acids, arginine and isoleucine, and an insertion
58
of another amino acid leucine (p.Arg306Ile307delinsLeu). Most deletion lead to frameshift
resulting in a premature stop of the protein synthesis. However, such mutation would not be
expected to be pathogenic for tumor syndromes as an inactive polymerase would rather lead
to disturbed cell proliferation and replication leading to apoptosis (Heitzer E. & Tomlinson I.,
2014 & Griffiths, A.J.F et al, 1999). For tumorigenesis the polymerase protein needs to remain
his function to keep the mutant cell alive, and only then a mutator phenotype can be
presented. Although previously Weedon et al reported an in-frame deletion in POLD1, causing
a multisystem disorder that includes lipodystrophy, deafness, etc. Also in this patient the
polymerase protein did encounter a deletion but stayed in frame, which largely remained the
protein function, i.e. the survival of the cell.
Arginine and isoleucine are both hydrophobic amino acids and are replaced for one
hydrophobic leucine by this mutation, i.e. no large changes in the protein would happen as
the change stays hydrophobic. This matter of hydrophobicity is very important in maintaining
or altering the protein structure as hydrophobic amino acids stick to other hydrophobic amino
acids and hydrophilic amino acids to other hydrophilic amino acids. So when a mutation
includes also a change from hydrophobic to hydrophilic amino acid it could dramatically
change the protein structure, as the corresponding amino acid might bind to some other
amino acid. In addition, conservation of the amino acids was checked in different species, by
which isoleucine turned out to be more conserved then arginine. Conserved amino acids have
such important function in the protein structure that an amino acid change could inactivate
or disturb protein function, whereas a low conserved amino acid causes a harmless amino acid
change (Heitzer E. & Tomlinson I., 2014).
The affected amino acid in residue 306-307 lies just at the beginning of the exonuclease
domain of POLD1, therefore a contribution to a deficient proofreading is not yet clear. For
some of the POLD1 EDMs, found by other researchers, the mechanism of polymerase EDMdriven tumorigenesis is known. Yoshida et al checked five and four exons of human POLD1
and POLE genes, respectively, and did not find sequence alterations such as insertions or
deletions. They did report two POLD1 sequence alterations, i.e. p.Arg506His and p.Val312Met,
of which the first was reported previously in DLD-1 cells. Both residues, respectively arginine
and valine, at codons 506 and 312 in POLD1 are highly conserved from yeast to mammals,
therefore these changes might be possible cancer causing mutations (Yoshida et al., 2011).
Palles et al identified a POLD1 mutation at Ser478, which turned out to be highly conserved.
The p.Ser478Asn mutation was not present in nearly 7000 UK controls or in public databases
of controls, and additionally showed a greatly increased risk of EC in females. Similar to the
POLE p.Leu424Val mutation, the p.Ser478Asn mutation packs together at the interface
between two helices that form the base of the exonuclease activity site, by which mutations
are predicted to distort the packing of the helices, affecting nuclease activity and causing
59
cancer (Heitzer E. & Tomlinson I., 2014 & Palles et al., 2013 & Briggs S. & Tomlinson I., 2013).
Additional screening of the exonuclease domain of POLD1 (codons 304-517) in 469 cases with
multiple colorectal adenomas and/or familial CRC identified three further heterozygous
POLD1 variants including Ser370, Gly426 and Pro327. The first two were unlikely to be
involved in direct DNA binding, whereby alterations in these residues were predicted to have
only moderate functional effects. The latter POLD1 variation, Pro327, was shown to have
functional consequences as the residue is fully evolutionarily conserved and lies on the DNA
interface of the POLD1 exonuclease domain. This residue is analogous to Pro286 in POLE,
which is an amino acid that sits on the edge of the DNA binding pocket of POLE adjacent to
the exonuclease active site and showed to be close to the nascent ssDNA. This was confirmed
by structural modelling, by which substitutions in this residue results in substantial disruption
of the DNA binding pocket, predicting significant effects on protein function. The p.Pro327Leu
mutation was also described by Heitzer E. & Tomlinson I. as a probable pathogenic mutation
found in a further patient with multiple adenomas (Briggs S. & Tomlinson I., 2013 & Palles et
al., 2013 & Church et al., 2013).
POLE EDM in Exon 13 of Patient 24
The second mutation was identified in patient 24, who developed colon cancer at the age of
33. Similar to patient 15 with the mutation in POLD1, this patient did not have a clear family
history of cancer (Figure 20). There were only two other cases of cancer in the mothers family
branch, i.e. the grandmother who died of breast cancer at the age of 42 and her brother who
suffered from leukaemia. To better evaluate whether the identified mutation may be a
predisposing mutation segregation analysis would be of advantage. However, as all affected
family members were deceased, it would not be very informative. It would still be indicated
to test the parents to check whether the mutation was inherited from a parent’s family branch
or whether the mutation occurred de novo.
The missense mutation, C.1282G>A, was located in exon 13 of POLE gene and led to an amino
acid change, p.Ala428Thr. Although the mutation was listed in dbSNP, it was not found in a
cohort of 2276 controls suggesting that the variant is rather a mutation than a rare SNP.
Alanine and threonine are both hydrophobic amino acids, which implies that this amino acid
change probably does not alter the protein structure. In silico analysis on the impact of this
mutation on protein function also indicated a minor effect on the protein. Furthermore, the
affected amino acid alanine at position 428 is not highly conserved which additionally suggests
that the mutation may be a neutral variant (Heitzer E. & Tomlinson I., 2014).
This EDM at residue 428 is located in the exonuclease domain of POLE and was not reported
previously. However, other researchers have found other POLE EDMs contributing to cancer
development. Yoshida et al reported the first POLE mutation, p.Phe367Ser, in human cells and
first identified POLE mutation in human diseases. This phenylalanine residue is highly
60
conserved, which suggests that this mutations is dysfunctional and possibly pathogenic
(Yoshida et al., 2011).
Palles et al identified a germline mutation in POLE p.Leu424Val, which lies not far from the
mutation we identified, and that was not present in nearly 7000 UK controls or in public
databases of controls. The Leu424 is highly conserved and the leucine-to-valine change is
predicted to have severe functional consequence for protein activity as Leu424 maps directly
to the active site of the exonuclease domain, whereby it is only approximately 5 amino acids
away from the ssDNA in the polymerase. Mutations in Leu424 are likely to distort the packing
of the helices, what causes a change that propagates through the active site, affecting
nuclease activity, which suggests a decreased fidelity of replication-associated polymerase
proofreading, leading to an increased mutation rate by base substitutions.
Irrespective of the work that discovered germline POLE and POLD1 mutations there have also
been found somatic POLE mutations in CRC by “The Cancer Genome Atlas” (TCGA) exome
sequencing project. It was proven that POLE was the target of recurrent somatic mutations in
MRR-proficient, but ‘ultramutated’ CRCs, demonstrating that the EDMs were causative for this
so-called ‘ultramutator’ phenotype. In the initial TCGA cohort, 7 POLE EDMs out of a total of
226 CRCs were identified. All of the identified mutations were missense mutations although
the germline p.Leu424Val change was absent. All of the analysed tumors were MSS, i.e. prima
facie having normal MMR (Heitzer E. & Tomlinson I., 2014). In the TCGA data set, two more
recurrent somatic changes were found, p.Val411Leu and p.Ser459Phe, whereas another
recurrent mutation p.Pro286Arg was identified in a project called ‘CRC exome sequencing
project’ (Heitzer E. & Tomlinson I., 2014 & Briggs S. & Tomlinson I., 2013 & Palles et al., 2013).
The most common somatic POLE mutation, p.Arg286His, is localised in the DNA binding pocket
adjacent to the exonuclease active site, with its side chain very close to the nascent ssDNA.
Substitutions at this site probably perturb the structure of the DNA binding pocket, which is
confirmed by the equivalent residue mutation, p.Pro123Leu, in T4 bacteriophage producing a
strong mutator phenotype. Structural analysis has shown that POLE amino acid 297 interacts
with exonuclease catalytic site residue 275 and mutations at this position would probably alter
active site conformation. In contrast POLE residue 411 is not predicted to interact with DNA
or catalytic site residues, which suggests that the increased mutation rate may result from
secondary effects on the binding pocket (Heitzer E. & Tomlinson I., 2014 & Briggs S. &
Tomlinson I., 2013). Recently released, but unpublished data from TCGA confirmed that these
codons 286, 411 and 459 are mutation hotspots (Briggs S. & Tomlinson I., 2013). Different
studies found somatic POLE EDMs but only few somatic POLD1 EDMs in sporadic ECs and CRCs
(Church et al., 2013 & Palles et al., 2013 & Briggs S. & Tomlinson I., 2013).
It is thought that EDMs result in defective polymerase proofreading and may therefore
contribute to the mutator phenotype as a ‘booster’ of MMR deficiency and can cause the
61
accumulation of mutations in various genes like tumor suppressor genes and oncogenes
(Yoshida et al., 2011). Some have speculated that MMR deficiency may be required for the
EDM mutator phenotype to be manifested. However, genetic studies in proofreadingdeficient, haploid yeast strains carrying a MMR-defect on the one hand and mice studies with
loss of both proofreading and MMR, have shown a synthetically lethal phenotype and led to
embryonic lethality, respectively. This indicates a synergistic effect on the mutation rate of
proofreading and MMR (Heitzer E. & Tomlinson I., 2014). A possible explanation for the
colorectal cancer phenotype in these patients is the presence of mutations in APC, KRAS and
TP53; as these genes play a crucial role in the step-by-step tumorigenesis in the human
colorectal epithelia. Additionally, these mutations showed an unusual spectrum and
frequency by which they often occurred of types and at position other than the common
hotspots. That is why it could be recommended to test these patients on APC, KRAS and TP53
mutations which could explain presence of CRCs secondary to the POLE or POLD1 EDMs
(Yoshida et al., 2011).
Different theoretical considerations identified that it is highly plausible that a variety of
germline POLE and POLD1 mutations affecting the exonuclease domain can influence CRC risk,
but the effect of these mutations might vary depending on the degree to which they impair
proofreading. Also pathogenic mutations in the non-proofreading domains of POLE and POLD1
could occur. They act a little different from those affecting the proofreading domains as they
cause tendency for the polymerase to incorporate mispaired bases into the growing strand
(Palles et al., 2013).
Palles et al concluded that germline POLD1 EDMs predispose as well for CRCs as for ECs,
whereas germline POLE EDMs predispose more to CRCs (Palles et al., 2013). In general,
germline as well as somatic DNA polymerase EDMs might cause an ‘ultramutated’, MMS type
of cancer. This sometimes leads to over a million base substitutions and to 5000 substitutions
in the coding regions of an EDM tumor compared to the POLE wild-type tumors. All types of
base substitutions are increased in frequency, with C:G > T:A changes being the most common
ones. It has also been noticed that there is a particular increase in the proportion of G:C > T:A
and A:T > C:G transversions. In addition, specific mutations could have different effects on the
mutation spectrum (Heitzer E. & Tomlinson I., 2014 & Briggs S. & Tomlinson I., 2013). To
further evaluate the pathogenicity of the identified mutations of this study, analysis of the
tumor samples could be helpful. An ultramutator phenotype in the tumor samples would
indicate a definite implication in tumorigenesis.
Although it is unclear if hypermutation is the only tumor-promoting consequence of these
mutations, it is intriguing to determine whether proofreading deficiency has any effect on
polymerase processivity, as negative effects on this function may be selectively deleterious
for the cell. POLE and POLD1 play additional roles in new strand synthesis as part of BER and
62
MMR. POLE is also involved in break-induced replication, which is a form of ds break repair
where the homologous chromosome is used as a template resulting in copy-neutral LOH. It is
not yet clear if these aspects of DNA polymerase function are important for tumor
development. It is however striking that defects in at least three pathways, involved in repair
of basepair-level mutations, can predispose to colorectal tumors (Briggs S. & Tomlinson I.,
2013).
Taken together, testing the patient samples for these previously mentioned hypotheses is
contemplated for further researches. These include testing the tumors on the presence of
somatic mutations, probably causing LOH, in the POLE and POLD1 genes according to Church
et al. Another argument to analyse the tumors from these patients is the fact that tumor
analysis could explain the method of tumorigenesis. In agreement with EDM phenotype, the
tumors from mutation carriers are thought to be MSS and tend to acquire base substitution
mutations, as confirmed by yeast functional assays (Palles et al., 2013). So, it is recommended
to test these patients for MSI and the presence of a mutator phenotype. In addition, it can be
presumed that pathogenic EDMs are selectively haploinsufficient and it can be noted that
somatic MSH2 and MSH6 mutations secondary to the EDM are common and may contribute
to tumorigenesis (Heitzer E. & Tomlinson I., 2014). So, another possibility is to test these
tumors on somatic MSH2 and MSH6 mutations secondary to EDM. Flohr et al used some
different methods to analyse sequence alterations and that could also be used in further
research. This includes mRNA testing of tumor cells by making cDNA and the performance of
a protein truncation assay to confirm mutations in the protein (Flohr et al., 1999).
SNPs
In addition to the two putative mutations several SNPs have been identified. Although an SNP
may not cause a disorder, as they appear as a common variation in the population, some SNPs
are associated with certain diseases. If certain SNPs are known to be associated with a trait,
the scientists may examine stretches of DNA near these SNPs in an attempt to identify the
gene or genes responsible for the trait (SNP, 2013). However, in our analysed patient cohort
no differences to data of the general population were observed.
This is in-line with data in the literature, where several SNPs located at conserved sites within
the polymerase or exonuclease domains of POLE and POLD1 were studied by genome-wide
association studies and targeted studies by which no association with cancer risk were found.
However, another common SNP within POLD3 has been found to be associated with an
increased risk of CRC in the general northern European population, although the mechanism
of action is unknown (Heitzer E. & Tomlinson I., 2014). Also Matakidou pointed out a
correlation between a POLE SNP and poor patient outcomes in lung cancer (Matakidou et al.,
2007), indicating that SNPs still should be analysed.
63
6. Conclusion
Mutational screening, by PCR and Sanger sequencing, of EDMs in POLE and POLD1 was
performed on 54 patients to identify causative mutations predisposing to CRC. These 54
patients showed a high-penetrance CRC phenotype but did not harbour mutations in the
known predisposition genes. However, despite the fact that no mutation was found, these
patients had a high a priori probability of a predisposing mutation. Therefore, the patient
cohort seemed to be ideal for testing on polymerase mutations.
Finally two putative EDMs were identified in two patients that were originally tested for LS.
However, it remains unclear whether the mutations are indeed disease causing or rather
neutral variants. Both EDMs have not been identified previously and were according to dbSNP
and the 1000 genome project not present in the general population. Further research of these
mutations is needed to fully understand how they might contribute to tumorigenesis. Testing
the relatives of these patients could give an insight in the segregation of the mutation and
should therefore be recommended to the patients and the family members. In addition,
analysis of the respective tumor samples might help elucidating the potential impact of the
EDMs on tumor formation. However, to fully characterise the identified mutations functional
analysis is needed.
Our study confirmed that EDMs in POLD1 or POLE could predispose to CRC. As evidence
already exists for POLE and POLD1 as important contributors to the pathogenesis of CRC and
EC. This concludes that it is definitely worthwhile to include the screening of these genes into
routine molecular genetic diagnostic for hereditary cancer syndromes. Although pathogenic
polymerase EDM cancers form a rare subtype of tumor, which is apparently restricted to the
colorectum and endometrium, there is no reason to disregard their importance (Heitzer E. &
Tomlinson I., 2014 & Briggs S. & Tomlinson I., 2013 & Valle et al., 2014 & Seshagiri S., 2013).
64
7. References

















All Cancers (2012). Consulted on 14 January 2014 via
http://globocan.iarc.fr/Pages/fact_sheets_cancer.aspx
A revolution in the microbiology laboratory (2009). Consulted on 13 February 2014 via
http://www.foodsafetywatch.org/features/a-revolution-in-the-microbiology-laboratory/
Balmain et al. (2003, March). The genetics and genomics of cancer. Nature genetics
supplement, 33, PP. 238-244.
Bardhan K. & Liu K. (2013). Epigenetics and Colorectal cancer pathogenesis, cancers, 5, pp.
676-713.
BigDye® Terminator v3.1 Cycle sequencing Kit from Applied Biosystems (2006). Consulted
on 3 February 2014 via http://www.biocompare.com/Product-Reviews/41093-BigDyeTerminator-v3-1-Cycle-Sequencing-Kit-From-Applied-Biosystems/
Boland et al. (2008, 31 March). The biochemical basis of microsatellite instability and
abnormal immunochemistry and clinical behaviour in Lynch Syndrome: from bench to
bedside. Fam Cancer, 7, nr: 1, pp. 41-52.
Briggs S. & Tomlinson I. (2013). Germline and somatic polymerase ε and δ mutations
define a new class of hypermutated colorectal and endometrial cancers. Journal of
pathology, 230, pp. 148-153.
Cancer (2013). Consulted on 8 January 2014 via
http://www.who.int/mediacentre/factsheets/fs297/en/
Cancer (2014). Consulted on 30 January via
http://www.medicinenet.com/cancer/article.htm#what_is_cancer
Colon Cancer genes (2013). Consulted on 30 January via
http://www.cancer.gov/cancertopics/pdq/genetics/colorectal/HealthProfessional/page2
Chan V.T.W. & McGee J.O'D (1987). Cellular oncogenes in neoplasia. Journal clinical
pathology, 40, pp. 1055-1063.
Church et al. (2013, March). DNA polymerase ε and δ exonuclease domain mutations in
endometrial cancer. Human molecular genetics, 22, nr. 14, pp. 2820-2828.
Curtin, C. (2013). GenomeWeb Feature: Sanger sequencing finds a spot in the small scale.
Genomeweb. Consulted on 13 March 2014 via http://www.genomeweb.com/sequencing/
Davis, C.P. (2011). Cancer. Medicine Net. Consulted on 9 January 2014 via
http://www.medicinenet.com/cancer/article.htm#what_is_cancer
de la Chapelle A. (2004, October). Genetic predisposition to colorectal cancer. Nature, 4,
pp. 769-780.
DNA
sequencing
(2011).
Consulted
on
13
March
2014
via
http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/D/DNAsequencing.html
Fearnhead et al. (2002). Genetics of colorectal cancer: hereditary aspects and overview of
colorectal tumorigenesis. British Medical Bulletin, 64, pp. 27-43.
65

















Fidler, I.J. (2003). The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis
revisited. Nature, 3, pp. 453-458. Consulted on 10 January 2014 via
http://www.nature.com/nrc/journal/v3/n6/abs/nrc1098.html
Flohr et al. (1999). Detection of mutations in the DNA polymerase δ gene of human
sporadic colorectal cancers and colon cancer cell lines. Int. J. Cancer, 80, pp. 919-929.
França et al. (2002). A review of DNA sequencing techniques. Quarterly Reviews of
Biophysiscs, 35, nr. 2, pp. 169-200.
Galiatsatos P. & Faulkes W.D. (2006). Familial Adenomatous Polyposis. American Journal
of Gastroenterology, 101, pp. 385-398.
Garber J. & Offit K. (2005, 10 January). Hereditary cancer predisposition syndromes.
Journal of Clinical Oncology, 23, nr. 2, pp. 276–292.
Grady W.M. & Carethers J.M. (2008). Genomic and Epigenetic instability in colorectal
cancer pathogenesis. Reviews in basic and clinical gastroenterology, 135, pp. 1079-1099.
Grandval et al. (2013). UMD-MLH1/MSH2/MSH6 databases: description and analysis of
genetic variations in French Lynch syndrome families. Database, 2013. Consulted on 7
March 2014 via http://database.oxfordjournals.org/content/2013/bat036.full
Griffiths et al. (1999). Modern genetic analysis. Consulted on 17 March 2014 via
http://www.ncbi.nlm.nih.gov
Heitzer E. & Tomlinson I. (2014). Replicative DNA polymerase mutations in cancer. Current
opinion in Genetics & Development, 24, pp. 107-113.
Hereditary and cancer (2013). Consulted on 26 February 2014 via
http://www.cancer.org/cancer/cancercauses/geneticsandcancer/heredity-and-cancer
Hoffbrand, A.V. & Moss, P.A.H. (2011). Essential Haematology. (6th edition). USA: WileyBlackwell.
Knudson A.G Jr. (1971, April). Mutation and Cancer: Statistical Study of Retinoblastoma.
Proc. Nat. Acad. Sci. USA, 68, nr.4, pp. 820-823.
Knudson, A.G. (1996). Hereditary cancer: two hits revisited. J Cancer Res Clin Oncol, 122,
pp.
135-140.
Consulted
on
28
February
2014
via
http://link.springer.com/article/10.1007%2FBF01366952
Knudson A.G. (2001, November). Two genetic hits (more or less) to cancer. Nature
Revieuws,
1,
pp.
157-162.
Consulted
on
28
February
2014
via
http://www.nature.com/nrc/journal/v1/n2/abs/nrc1101-157a.html
Knudson A.G. (2002). Cancer genetics. American Journal of Medical Genetics, 111, pp. 96102.
Koch, M. (2013). Establishing a method for rapid identification of chromosomal copy
number changes [Master]. FH Joanneum, University of Applied Sciences.
Kohlmann, W. & Gruber, S.B. (2012). Lynch syndrome. Consulted on 5 February 2014 via
http://www.ncbi.nlm.nih.gov/books/NBK1211/
66

















Lange et al. (2011, February). DNA polymerases and cancer. Nature, 11, pp. 96-110.
Lynch et al. (2009, April). Review of the Lynch syndrome: history, molecular genetics,
screening, differential diagnosis, and medicolegal ramifications. Clinical genetics, 76, pp.
1-18.
Mach, M. (2013). Identification of prognostic methylation markers in stage II colorectal
cancer patients [Master]. Karl-Franzens University Graz, Faculty of Natural Science.
Malumbres M. & Barbacid M. (2001, December). To cycle or not to cycle: A critical decision
in cancer. Nature, 4, pp. 222-231.
Martinez, J.D. (2005). Horizons in cancer research: Focus on colorectal cancer research.
(First edition). New York: Nova Biomedical books.
Matakidou et al. (2007). Genetic variation in the DNA repair genes predictive of outcome
in lung cancer. Human Mol Genet, 16, pp. 2333-2340.
McCulloch S.D. & Kunkel T.A. (2008, January). The fidelity of DNA synthesis by eukaryotic
replicative and translesion synthesis polymerases. Nature, 18, nr: 148, pp. 148-161.
Mehlen P. & Puisieux A. (2006, June). Metastasis: a question of life or death. Nature
reviews cancer, 6, pp. 449-458.
Metzker M.L. (2010, January). Sequencing technologies- the next generation. Nature
reviews, 11, pp. 31-46.
Michor et al. (2004, March). Dynamics of cancer progression. Nature reviews, 4, pp. 197206.
Consulted
on
14
January
2014
via
http://www.nature.com/nrc/journal/v4/n3/pdf/nrc1295.pdf
Palles et al. (2013, February). Germline mutations affecting the proofreading domains of
POLE and POLD1 predispose to colorectal adenomas and carcinomas. Nature genetics, 45,
nr. 2, pp. 136-146.
Pappou E.P. & Ahuja N. (2010, December). The role of oncogenes in gastrointestinal
cancer. Gastrointestinal cancer research, pp. 2-15. Consulted on 10 January 2014 via
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3047044/
Pasche, B. (2010). Cancer Genetics. Consulted on 28 January 2014 via
http://books.google.be.
Pavlov et al. (2006, 24 January). Evidence that errors made by DNA polymerase α are
corrected by DNA polymerase δ. Curr Boil, 16, pp. 202-207.
Peto J. (2001, 17 May). Cancer epidemiology in the last century and the next decade.
nature, 411, pp. 390-395
Pischler, C. (2013). New applications in prenatal diagnosis [Master]. Karl-Franzens
University Graz, Faculty of Natural Science.
Polymerase chain reaction (PCR) (2014). Consulted on 13 February 2014 via
http://amrita.vlab.co.in/?sub=3&brch=186&sim=321&cnt=2
67















Qubit® 2.0 Fluorometer (2014). Consulted on 14 January 2014 via
http://www.lifetechnologies.com/at/en/home/brands/product-brand/qubit/qubitfluorometer.html
RanplexCRC
Array
(2007).
consulted
on
8
January
2014
via
http://www.randox.com/brochures/pdf%20brochure/lt141.pdf
Rizzo J.M. & Buck M.J. (2012, 22 May). Key principle and clinical applications of “NextGeneration” DNA sequencing. Cancer Prev Res, 5, pp. 887-900.
Robson et al. (2010, 10 February). American Society of Clinical oncology policy statement
Update: Genetic and Genomic Testing for cancer susceptibility. Journal of clinical
Oncology, 28, nr: 5, pp. 893-901.
Schwarte-Waldhoff et al. (2000, 15 August). Smad4/DPC4-mediated tumor suppression
through suppression of angiogenesis. PNAS, 97, Nr: 17, PP. 9624-9626.
Seshagiri S. (2013, February). The burden of faulty proofreading in colon cancer. Nature
genetics, 45, nr. 2, pp. 121-123.
Shendure J. & Ji H. (2008, October). Next-generation DNA sequencing. Nature
biotechnology, 26, nr. 10, pp. 1135-1145.
Shi C. & Washington K. (2012). Molecular Testing in Colorectal Cancer. American Journal
for clinical pathology, 137, pp. 847-859.
Shia J. (2008, 4 July). Immunohistochemistry versus Microsatellite Instability Testing for
screening colorectal cancer Patients at risk for Hereditary Nonpolyposis Colorectal cancer
syndrome. Journal of Molecular diagnostics, 10, nr. 4, pp. 293-300.
Smith et al. (2013, April). Exome resequencing identifies potential tumor-suppressor genes
that predispose to colorectal cancer. Hum Mutat 2013, 34, nr. 7, pp. 1026-1034.
SNP
(2013).
Consulted
on
11
March
2014
via
http://www.nature.com/scitable/definition/snp-295
Steitz T.A. (1999, 18 June). DNA Polymerases: Structural Diversity and Common
Mechanisms. The Journal of Biological Chemistry, 274, nr. 25, pp. 17395-17398.
Stewart, A. (2013, 16 September). Genetic Testing Strategies in Newly Diagnosed
Endometrial Cancer patients Aimed at Reducing Morbidity or Mortality from Lynch
Syndrome in the Index case or Her Relative. PLOS currents evidence on genomic tests.
Consulted
on
7
March
2014
via
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3775889/?report=classic
Su et al. (1999, June). Germline and Somatic Mutations of the STK11/LKB1 Peutz-Jeghers
Gene in Pancreatic and Biliary cancers. American journal of pathology, 154, nr: 6, pp. 18351840. Consulted on 20 January 2014 via http://www.ncbi.nlm.nih.gov/pubmed/10362809
The NanoDrop Spectrophotometers and Fluorospectrometers (2008). Consulted on 16
January 2014 via http://www.nanodrop.com/Library/NanoDrop-TechnologyOverview.pdf
68















Thompson et al. (2014, February). Application of a 5-tiered scheme for standardized
classification of 2.360 unique mismatch repair genes variants in the InsiGHT locus-specific
database. Nature genetics, 46, nr. 2, pp. 107-117.
Valle et al. (2014, 18 February). New insight into POLE and POLD1 germline mutations in
familial colorectal cancer and polyposis. Hum mol genet. Consulted on 17 March 2014 via
http://hmg.oxfordjournals.org
Vogelstein B. & Kinzler W.K. (2004, August). Cancer genes and the pathways they control.
Nature medicine, 10, nr: 8, pp. 789-799.
Vogelstein et al. (2010). p53: The most frequently altered Gene in Human Cancers. Nature
Education, 3, nr: 9, PP. 6. Consulted on 14 January 2014 via
http://www.nature.com/scitable/topicpage/p53-the-most-frequently-altered-gene-in14192717
Waters, J.D. (2010). Frontiers in brain repair: Stem cell origin of brain tumors. (First
edition). USA: Springer.
Weedon et al. (2013, August). An in-frame deletion at the polymerase active site of POLD1
causes a multisystem disorder with lipodystrophy. Nature genetics, 45, nr. 8, pp. 947-951.
Weisstein,
E.W.
(2014).
Consulted
on
28
March
2014
via
http://mathworld.wolfram.com/FishersExactTest.html
What is a tumor? (2012). Consulted on 25 February 2014 via
http://www.medicalnewstoday.com/articles/249141.php
What Is Cancer? What Causes Cancer? (2013). consulted on 9 January 2014 via
http://www.medicalnewstoday.com/info/cancer-oncology/
What is Carcinoma in Situ? (2014). Consulted on 25 February 2014 via
http://www.wisegeek.com/what-is-carcinoma-in-situ.htm
What is Sanger sequencing/Dye terminating sequencing (2013). Consulted on 13 March
2014 via http://www.symposcium.com/2013/05/sanger-dye-terminating/
WHO
(2014).
Consulted
on
11
March
2014
via
http://www.who.int/mediacentre/factsheets/fs297/en/
WHO. (2008). Cancer Control Knowledge into Action: WHO guide for effective
programmes. (First edition) Switzerland: WHO.
Yoshida et al. (2011, December). Concurrent Genetic alterations in DNA polymerase
proofreading and mismatch repair in human colorectal cancer. European Journal of Human
Genetics, 19, pp. 320-325.
Zbuk K.M. & Eng C. (2007, September). Hamartomatous polyposis syndromes. Nature
clinical practice Gastroenterology & Hepatology, 4, nr: 9, pp. 492-502.
69
8. Appendix
8.1. PCR quality check result from Agilent 2100 Bioanalyzer
Table I. Quality checks of PCR products from different exons of POLD1 and POLE on human genomic female DNA by
Agilent 2100 Bioanalyzer (Agilent Technologies).
Sample
Definition
1
PCR with POLD1_E8
2
PCR with POLD1_E9
Result Agilent 2100 Bioanalyzer (Agilent Technologies)
I
3
PCR
POLD1_E10
with
4
PCR
POLD1_E11
with
5
PCR
POLD1_E12
with
II
6
PCR
POLD1_E13
with
7
PCR with POLE_E9
8
PCR with POLE_E10
III
9
PCR with POLE_E11
10
PCR with POLE_E12
11
PCR with POLE_E13
IV
12
PCR with POLE_E14
8.2. Sequencing results human genomic female DNA
6.3.1 Sequence results of POLD1 exons
Table II. Overview results Sanger sequencing of different exons of POLD1 of human genomic female DNA (Promega).
Exon
E8
E11
Result (SeqScape)
Failed
E11
V
E12
6.3.2 Sequence results of POLE exons
Table III. Overview results Sanger sequencing of different exons of POLE of human genomic female DNA (Promega).
Exon
Result (SeqScape)
E10
E11
E14
VI
8.3.Amino acid conservation
Table IV. Amino acid conservation of the 2 putative EDM's.
Mutation
Amino acid conservation
p.Arg306Ile307DelinsL
eu
c. 917-919Del
p. Ala428Thr
c. 1282G>A
VII
8.4.Detailed information of identified SNPs
8.4.1 SNP in POLD1 intron 9
Table V. Heterozygous SNP in intron 9 of POLD1 in patients analyzed and identified by SeqScape with the corresponding
results according dbSNP on NCBI.
Results SeqScape
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs1673043
 NM_001256849.1: c.1137+53G>A
 MAF: A=0.244/531 according 1000 genomes
6.4.2 SNP in POLD1 intron 10
Table VI. Heterozygous and homozygous SNP in intron 10 of POLD1 in patients analyzed and identified by SeqScape with
the corresponding results according dbSNP on NCBI.
Results SeqScape
Homozygous
Mutant:
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs1673041
 NM_001256849.1: c.1243-50T>G
 MAF: T=0.304/662 according 1000 genomes
VIII
6.4.3 SNP in POLD1 intron 13
Table VII. Heterozygous SNP in intron 9 of POLD1 in patients analyzed and identified by SeqScape with the corresponding
results according dbSNP on NCBI.
Results SeqScape
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs3219399
 NM_001256849.1: c.1686+47G>A
 MAF: A=0.002/4 according 1000 genomes
6.4.4 SNP in POLE intron 9
Table VIII. Heterozygous and homozygous SNP in intron 9 of POLE in patients analyzed and identified by SeqScape with
the corresponding results according dbSNP on NCBI.
Results SeqScape
Homozygous
Mutant:
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs4077170
 NM_006231.2: c.910-6G>C
 MAF: C=0.443/965 according 1000 genomes
6.4.5 SNP in POLE intron 12
Table IX. Heterozygous mutation in intron 12 of POLE in patients analyzed and identified by SeqScape with the
corresponding results according dbSNP on NCBI.
Results SeqScape
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs5744761
 NM_006231.2: c.1226+45C>T
 MAF: A=0.062/135 according 1000 genomes
IX
6.4.6 SNPs in POLE intron 13
Table X. Heterozygous and homozygous SNP in intron 13 of POLE in patients analyzed and identified by SeqScape with
the corresponding results according dbSNP on NCBI.
Results SeqScape
Homozygous
Mutant:
Heterozygous:
Results according dbSNP
The mutation according dbSNP on NCBI
 (RefSNP) Cluster report: rs4883555
 NM_006231.2: c.1359+43G>A
 MAF: A=0.441/961 according 1000 genomes
X
XI