Download Applications of Molecular genetics in - e

Applications of Molecular genetics in
Personalized Medicine
Chapter: Application of Microsatellite Marker Analysis
Edited by: Korkut Ulucan
Published by OMICS Group eBooks
731 Gull Ave, Foster City. CA 94404, USA
Copyright © 2013 OMICS Group
All book chapters are Open Access distributed under the Creative Commons Attribution
3.0 license, which allows users to download, copy and build upon published articles
even for commercial purposes, as long as the author and publisher are properly credited,
which ensures maximum dissemination and a wider impact of our publications. However,
users who aim to disseminate and distribute copies of this book as a whole must not seek
monetary compensation for such service (excluded OMICS Group representatives and
agreed collaborations). After this work has been published by OMICS Group, authors have
the right to republish it, in whole or part, in any publication of which they are the author,
and to make other personal use of the work. Any republication, referencing or personal use
of the work must explicitly identify the original source.
Notice:
Statements and opinions expressed in the book are these of the individual contributors and
not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy
of information contained in the published chapters. The publisher assumes no responsibility
for any damage or injury to persons or property arising out of the use of any materials,
instructions, methods or ideas contained in the book.
Cover OMICS Group Design team
First published October, 2013
A free online edition of this book is available at www.esciencecentral.org/ebooks
Additional hard copies can be obtained from orders @ www.esciencecentral.org/ebooks
Application of Microsatellite Marker
Analysis
Lan Kluwe*
Laboratory for Tumour Genetics, Department of Neurology, Department
of Maxillofacial Surgery, University Hospital Eppendorf, Hamburg,
Germany
*Corresponding author: Laboratory for Tumour Genetics, Department
of Neurology, Department of Maxillofacial Surgery, University Hospital
Eppendorf, Building O48, 4th floor, Martinistr 52, 20246 Hamburg,
Germany, Phone: +49 40 7410 58267; Fax: +49 40 7410 5 9665;
E-mail: [email protected]
Abstract
Microsatellites are DNA sequences consisting of short repeats of highly variable number. Hundreds thousands of such pieces of
sequences have been found throughout the human genome. Microsatellites are extremely powerful markers for distinguishing alleles
at linked loci and thus can be used in linkage studies, segregation analysis and in investigating somatic loss of heterozygosity or allele
imbalance in tumours. In this chapter, several applications of microsatellites will be introduced:
1.
2.
3.
4.
Tracing parental origin of de novo mutations
Indirect pre-symptomatic / prenatal diagnosis for at-risk offspring
Determining cellular origin of tumour cells
Discriminating tumour cells from non-tumour cells derived from a tumour.
Basics of microsatellite markers
What is a microsatellite?
Originally, satellite DNA was a term for large centromeric tandem repeats because they were found in a “satellite” peak in density
gradient fractionation of sheared DNA. Later, shorter tandem repeats (10-30 bp) were analogically called minisatellites and finally,
sequences containing very short repeats of 1 to 7 bp were termed microsatellites. Microsatellites with di-nucleotide repeats, for example
CA and GT repeats, are most frequent. Also known as short tandem repeats or simple sequence repeats, microsatellites are polymorphic
markers. An example of microsatellite containing “CA” repeats is given as following:
Name: D22S430
Location: chromosome 22, bp 28936331 - 28936414
agtgcagcagatccttcttaaccatagtatttatttggaatttttttctaattgcagatg
taaattgtgcagaatgcagaaaggtaggactcacctgtgcatgcgtgtgtgcctctatgc
acacacacacacacacacacacacacacaacgggaggaaaagccaagatcatttatagct
cactaccctgggtagccctggttaacccctcagagtacctccctccagtctttcatttat
tccctgaaggggttgattccaagtgctgtaagatagggaaacccagggaggtgatgggga
ccagggaagtcgggatggccttggggagaaggtgagattcagactcagtcttaaaagagg
In the above example, the number of the “CA” repeats is 15. This number varies to large extent in each allele and thus gives rise to
a polymorphism in length. The major genetic mechanism resulting in size-polymorphism of microsatellites is slippage of one or more
repeat units during DNA replication. D22S430 has 8 different alleles, reaching a heterozygosity of 70%, meaning that 70% individuals
have two different alleles of this microsatellite, which appear as two bands on a gel or two peaks on an electrogram after amplification
of this marker. In the other 30% of cases, the two alleles of this marker in an individual have identical lengths which appear as a single
band on a gel or a single peak on the electrogram upon amplification. Generally, microsatellite markers with more alleles have higher
the heterozygosity.
How to analyse microsatellites
A microsatellite can be analysed by PCR amplification and subsequent electrophoresis-separation by size. The primers for
amplification are chosen in the sequences flanking the repeats, one of a primer-pair is be labelled with fluorescence-dye to visualize
one strand of the amplification product on a fluorescence sequencer. By choosing proper size and dye for each microsatellite, multiple
microsatellites can be amplified and analysed simultaneously (Figure 1). Alternatively, each microsatellite can also be amplified separately
and pooled together for the subsequent analysis. Using internal size standard, absolute size of each microsatellite allele can be calculated
by an integrated program (Figure 1).
OMICS Group eBooks
In contrast, a single nucleotide polymorphism (SNP) has only two alleles. For example, an SNP allele for an A/C polymorphism
can only be either A or C. Most of such markers have only very low heterozygosity rate, meaning large proportion of individuals has
identical allele of this SNP.
003
Figure 1: Electrogram of 7 microsatellites run simultaneously. Microsatellites with overlapping size-ranges are labelled with different fluorescence dyes,
those with non-overlapping size-ranges can be labelled with a same dye (markes 4 and 5). The red peaks are internal size standard run together with the
amplified markers; they enable calculation of the absolute size of the microsatellites.
Figure 2: Patterns of a di-nucleotide microsatellite. A: a homo/hemizygous microsatellite consisting of a main peak (array) and three smaller shadow peaks
with decreasing intensities. B: a heterozygous microsatellite with two alleles differing in 6 base pairs. Multiple peaks from each allele make the pattern
complex. The largest (rightmost) peaks in each group are the real alleles (arrays) while the smaller shadow peaks should not be counted. Notice in a
heterozygous case, the smaller allele (allele 1) has higher intensity than the larger allele (allele 2). Two factors contribute to this phenomenon: (1) smaller
fragments amplify more efficiently than larger ones and (2) the peak of the larger allele (in the case of B, the 3rd one) adds extra intensity to the smaller
allele. The second effect is especially profound when the two alleles differ only in on repeat unit, like in the case of C: where a heterozygous microsatellite
with two alleles differing in 2-bp. Allele 2 is not read as a shadow peak of allele 1 because it is larger than allele 2 while a shadow peak must be smaller
than the main peak of an allele. Allele 1 was not read as a shadow peak of allele 2 because it has higher intensity than allele 2 while all shadow peaks must
have low intensities than the main peak of an allele.
These shadow peaks complicate the pattern of an amplified microsatellite. Its interpretation thus demands extensive experience and
knowledge. As a rule, the true allele for a microsatellite is the main peak on the right which is the largest in size and in amount (Figure 2A,
array). A heterozygous microsatellite will give two groups of peaks whereas the largest and highest two peaks in each group represent the
two alleles (Figure 2B, arrays). More difficult is the pattern of two alleles differing only by a single repeat-unit. This is because the shadow
peak of the larger allele is at the same position of the smaller allele (Figure 2C). Therefore, a major peak followed by several shadow peaks
OMICS Group eBooks
Each a heterozygous microsatellite has two alleles differing in lengths and each allele has two strands differing in nucleotide
composition. Fortunately, since only one primer is labelled with fluorescence, only one strand of each allele of a microsatellite will be
visible on an electrogram. However, even the single strand of a single allele does not appears as a single peak on an electrogram, but
often as a group of peaks consisting of a main peak with the largest size and highest intensity, and several “shadow” peaks (Figure 2).
These shadow peaks are shorter than the main peak by one, two or three-folds of the repeat-unit, for example, in the case of di-nucleotide
repeats, by 2-, 4- or 6-bp (Figure 2A). This “ladder” is caused by “slipping” during the amplification of a microsatellite, in which the strand
in synthesis slips one repeat-unit forward, resulting in fragment which is shorter by one-repeat. The extent of the “slipping” effect varies
from microsatellite to microsatellite and from amplification to amplification. In microsatellites consisting of tri- and tetra-nucleotide
repeats, the “slipping” effect is much less prominent.
004
with decreasing intensity is the pattern of a homozygous allele whereas the largest peak followed by a peak with higher intensity and
several additional shadow peaks is the pattern of a heterozygous allele differing by a single repeat unit.
Applications of microsatellite markers
Standard applications of microsatellites such as linkage and segregation analysis are not the main focus of this chapter. Instead, several
specific applications will be introduced, which are all based on loss of heterozygosity in tumours.
Loss or gain of chromosome material is frequent in tumours. The 10 most frequent deletions across 17 types of solid tumours are listed
in (Table 1). By examining these regions using microsatellite markers, it is highly likely that one or more LOH will be found for a tumour.
Chromosome and band
Peak region
9p21.3
chr9:21489625-22474701
GISTIC q value
0,00E+00
16q23.1
chr16:76685816-78205652
1,40E-176
9p24.1
chr9:7161607-12713130
1,01E-83
20p12.1
chr20:14210829-15988895
9,69E-74
6q26
chr6:161612277-163134099
8,13E-65
13q14.2
chr13:46362859-48209064
3,90E-63
2q22.1
chr2:138479322-143365272
1,10E-60
4q35.2
chr4:186684565-191273063
1,44E-58
5q11.2
chr5:57754754-59053198
1,52E-55
16p13.3
chr16:5062786-7709383
4,98E-49
Table 1: The 10 most frequent deletions across 17 cancer types (Beroukhim et al. 2010).
Loss of allele can be loss of one allele or copy number neutral loss of heterozygosity (LOH). The latter corresponds to isodisomy due
to somatic recombination. Microsatellite typing using paired blood/tumor DNA is a powerful tool for detecting LOH in tumours. It is
simple, fast and inexpensive. Furthermore, microsatellite typing can detect copy-number neutral LOH while CGH, FISH and digital PCR
cannot.
It is desirable that paired blood/tumour samples are typed simultaneously using multiple markers for regions of interest. For cases
where no blood is available, DNA can also be prepared from non-tumour area in the resected tumour specimen. Evaluating LOH demands
extensive knowledge and experience. As mentioned above, patterns of amplified microsatellite on an eletrogram is complex and patterns
of LOH of a tumour with considerable proportion of non-tumour cells are more complex. Studies often set some criteria based ratio of
the peak intensities of the two samples. However, such ratios vary largely, depending on various factors including the absolute intensity
of the peaks and the setting of base line. We have good experience with a manual direct inspection of the electrogram. Figure 3 shows an
example of LOH analysis using 4 microsatellites linked to the NF2 gene on chromosome 22. Markers 1, 3 and 4 have two peaks in blood
and therefore are heterozygous for this patient. Marker 2 has only one peak and is therefore homozygous. In tumours 1 and 3, one peak of
each of the three heterozygous markers is lost or obviously reduced, clear pattern of complete and incomplete LOH, respectively.
Incomplete or partial LOH is due to presence of non-tumour cells in the tumour, for example, blood cells and fibroblasts. These nontumour cells do not have the LOH and therefore account for reduced or trace amount of the lost alleles. Large proportion of non-tumour
tissues can obscure the LOH in the tumour. Therefore, for LOH-analysis, the content of tumour cells should be checked in the tumour
part from which DNA was prepared. For example, we prepare a series of sections from a tumour block and ensure sufficient tumour-cell
content on one section pathologically. Should the section contain large areas of non-tumour cells, such areas can be avoided. However, the
obtained DNA is often of inadequate quality and insufficient quantity, resulting in frequent allele drop-out or allele imbalance of amplified
microsatellites which have the identical appearance of LOH or partial LOH and therefore mistakenly interpreted as such [1].
When using DNA of poor quality or quantity, for example, those extracted from achieved specimen or from single tumours captured
OMICS Group eBooks
Figure 3: An example of loss of heterozygosities (LOH) in tumours. 4 microsatellites linked to the NF2 tumor suppressor gene were amplified from
DNA of blood and tumours 1 through 4 of a patient. Marker 1, 3 and 4 are heterozygous in this patient as can be seen in the two peaks of each of them.
Heterozygosity of all the three markers was lost in tumours 1 and 3 (arrays) but not in tumours 2 and 4.
005
by laser-microdissection, the amounts of two microsatellite alleles amplified can be out of balance or one of the allele can drop out, with
same appearance of partial or complete LOH. This can be checked by analyzing several microsatellites from other loci where not LOH is
expected. If those microsatellites show apparent partial LOH, then it is due to imbalance of the alleles in the starting DNA, but not true
LOH. Limiting PCR cycles below 30 provides another strategy to exclude samples of inadequate quality. Also re-extraction of DNA from
tumour and carrying out PCR in replicates minimize wrong interpretation. This is because a true lost of or reduced microsatellite allele
in a tumour must be the same in repeatedly extracted DNA and in all multiple PCR replicates while artefact LOH can randomly occur
on both alleles [1].
Parental origin of mutation
New mutations causing genetic diseases can occur either on paternal or maternal alleles. Study parental origin of new mutations
contributes to our understanding of molecular mechanism underlying such mutations. Traditionally, parental origin of new mutations
is investigated via identification of mutation-bearing allele by segregation analysis in families with multiple affected and unaffected
members. In a tumour suppressor gene, parental origin can also be traced via identification of the in tumour retained allele which carries
the mutation. By compare patterns of microsatellites in this tumour suppressor gene region of the parents, we can find out if the retained
allele is paternal or maternal (Figure 4). Using this approach, we have determined parental origin of mutation-bearing alleles in 45
neurofibromatosis 2 (NF2) patients bearing new NF2 mutations [2].
Figure 4: Tracing parental origin of the mutant allele in a de novo patient based on identification of the retained allele in a NF2-associated tumour. A
microsatellite linked to the NF2 gene is heterozygous in blood of the patient and hemizygous in tumour from the patient. The larger (right) allele was lost
in the tumour (array) while the smaller allele was retained. The retained allele carries the constitutional mutation. Comparison with alleles of the parents
reveals that this retained allele is from the father. A paternal origin of the mutation is therefore inferred.
Indirect Presymptomatic/Prenatal Diagnosis
Similarly as tracing parental origin, identification of the lost allele in tumour also enables identification of carriers of the mutation in
asymptomatic family members of a tumour suppressor gene disorder.
Conventional presymtomatic/prenatal diagnosis is based on identification of mutation in the patient. However, mutation analysis is
time-consuming and costly, especially for large genes. In addition, mutation can not be found in all cases. In the case of neurofibromatosis
2 (NF2), mutations can only be found in about 50% of blood of the patients. A major reason for the missing mutations in blood of NF2
patients is the frequent mosaicism.
Also LOH-based presymptomatic diagnosis is an indirect approach based on identification of the mutation-carrying allele in the
patient. However, in contrast to segregation analysis, LOH-based approach does not rely on the availability of multiple family members.
In stead, tumours of the patient will be analysed to identify the lost allele. Identification of a lost allele is also the identification of the
mutation-carrying allele which is retained in the tumour [2]. Subsequently, asymptomatic family members can be tested for having
inherited the mutation-carrying allele (Figure 5).
Since the LOH-based diagnosis is relies solely on typing of several microsatellite markers, is simple, quick and inexpensive. Only very
small amount of DNA is needed and many achieved material can be used.
OMICS Group eBooks
An alternative approach is segregation analysis using linked markers, based on identification of the disease-bearing allele which is
shared by affected family while not by unaffected family members. For this analysis, at least two ascertained affected individuals and at
least one unaffected individuals are necessary. However, in reality, many patients carry new or de novo mutations and therefore no other
family member is affected. Even in inherited cases, it is often painful and more often impossible to obtain DNA samples from multiple
affected and unaffected for various reasons. Thus, segregation analysis remains mostly theoretical and hardly applied in practice.
006
Figure 5: Principle of using LOH data for determining the mutant allele in presymptomatic / prenatal diagnosis. The lost allele of a tumour suppressor gene
during tumorigenesis is the healthy allele while the in retained allele carries the mutated gene, resulting in bi-allelic inactivation of this gene. Identification
of the retained allele therefore enables identification of the mutated allele which can be checked in family members. In the illustrated case, child 1 inherited
the mutated allele (red) from the affected father and child 2 inherited the healthy allele (green). The other alleles (green) were inherited from the unaffected
mother.
Determining Cellular Origin of Tumour Cells
A tumour usually consists of multiple types of cells. For example, neurofibromas contain Schwann cells, fibroblasts, mast cells,
perineurial cells and neurons. Schwann cells have been considered to be the progenitors of these tumours. However, there were also data
showing involvement of fibroblasts. Direct evidence of the cellular origin for this kind of tumours has been missing.
Figure 6: LOH for determining the cellular origin in a tumour. A partial LOH (arrays) was found in a neurofibroma which is composed mainly of Schwann
cells and fibroblasts. In selectively cultured Schwann cells derived from this neurofibroma, the LOH was nearly complete. In contrast, it was seen in the
fibroblasts derived from same tumour. These data demonstrate that the somatic alteration occurred only in Schwann cells, providing solid genetic evidence
that Schwann cells are the tumour cells.
DNA was extracted from cells of the two cultures and subjected to microsatellite analysis (Figure 6). Two intra NF1 microsatellite
markers were heterozygous because two peaks are present in the blood DNA for each marker. Partial loss of heterozygosity (LOH) of the
markers was visible in the tumour. The same LOH was found in the enriched Schwann cells, and the loss was nearly complete indicating
that this culture contains mostly tumour cells. In contrast, this LOH was not in the fibroblasts grown under standard cultural condition,
indicating that these cells do not harbour somatic alteration and they are therefore not the tumour cells. This result provided solid genetic
OMICS Group eBooks
When a neurofibroma is dissected and cultured under standard condition, fibroblasts grow predominantly. However, under special
cultural condition, it is also possible in some cases to selectively grow Schwann cells from neurofibromas. We cultured cells from a
neurofibroma under these two conditions separately. Cells obtained under condition for Schwann cells exhibited morphology of Schwann
cells: small, bi-poplar and often shiny cell bodies. In contrast, cells grown under standard condition exhibited typical features of fibroblasts:
flat and thin (Figure 6).
007
evidence that only the Schwann cells carry the somatic alteration in the NF1 gene and they are thus the progenitor tumour cells in
neurofibromas [3].
Verifying Tumour Cells in Culture
Cells in culture derived from a tumour are not necessarily always tumour cells [4]. For example, in vitro, fibroblasts often grow faster
even than malignant tumour cells. Also here, LOH in the original tumour can be used to verify the authenticity of the derived cells.
Figure 7 shows an example of two malignant nerve sheath tumours from a patient (one primary and one recurrent). A microsatellite
marker linked to the NF1 gene (a tumour suppressor gene) is heterozygous in blood of the patient. This heterozygosity was lost in both
the primary and the recurrent tumours. In the cells derived from the primary tumour, LOH was visible (grey array), providing genetic
evidence that these cells are indeed tumour cells. In contrast, cells derived from the recurrent tumour recovered the heterozygosity which
was lost in the original tumour, indicating that they are of non-tumour cells which have overgrown the tumour cells in the culture.
Closing remark
These specific applications introduced in the chapter demonstrated the value and power of microsatellite-typing. Microsatellite-typing
is simple, easy, quick and inexpensive and thus will remain a useful powerful genetic tool in the future.
References
1. Kluwe L (2012) Allelic drop-out, allele imbalance, or loss of heterozygosity? Genes Chromosomes Cancer 51: 521-522.
2. Kluwe L, Mautner V, Parry DM, Jacoby LB, Baser M, et al. (2000) The parental origin of new mutations in neurofibromatosis 2. Neurogenetics 3: 17-24.
3. Kluwe L, Friedrich R, Mautner VF (1999) Loss of NF1 allele in Schwann cells but not in fibroblasts derived from an NF1-associated neurofibroma.
Genes Chromosomes and Cancer 24: 283-285.
OMICS Group eBooks
4. Frahm S, Mautner VF, Brems H, Legius E, Debiec-Rychter M, et al. (2004) Genetic and phenotypic characterization of tumor cells derived from
malignant peripheral nerve sheath tumors of neurofibromatosis type 1 patients. Neurobiol Dis 16: 85-91.
008
Sponsor Advertisement
TIF Publications
TIF Publications cater to the needs of readers of all ages and educational
backgrounds, and provide concise up-to-date information on every aspect of
thalassaemia - from prevention to clinical management. TIF’s publications have
been translated into numerous languages in order to cover the needs of the
medical, scientific, patients and parents communities and the general community.
List of Publications - ORDER YOUR BOOKS!
N E W ! Ju
s t R e le a
se
d!
N E W ! Ju
s t R e le a
sed
Hard copies and CD-ROM or DVD versions can be
ordered directly from TIF and are distributed free of charge.
Place your order at [email protected]
The translation of TIF’s educational publications into
various languages continues in 2013.
All translated publications are or will become available on
our website. Check with us to get updated on the latest
translations!
UPCOMING TIF PUBLICATIONS
• Community Awareness Booklets on α-thalassaemia, β-thalassaemia & Sickle Cell Disease (Greek)
(Eleftheriou A)
• Sickle Cell Disease: A booklet for parents, patients and the community, 2nd Edition
(Inati-Khoriaty A)
• Guidelines for the Clinical Management of Transfusion Dependent Thalassaemias, 3rd Edition
(Cappellini M D, Cohen A, Eleftheriou A, Piga A, Porter J, Taher A)
Please visit our website at
http://www.thalassaemia.org.cy/list-of-publications
Free of charge
All our publications are
available as PDF files on
our website, completely
free of charge.
!