6.1.1. Multiplexed ARMS-PCR Approach for the Detection of Common MECP2
Mutations
In the present study, we have described the development and validation of a
multiplexed multiplex amplification refractory mutation system (ARMS) - PCR assay for
identification of seven common mutations that accounts for almost 65 per cent of all
MECP2 gene mutations (RettBASE). So far, these mutations could be investigated using
PCR-RFLP or DNA sequencing techniques. Although appropriate, DNA sequencing is not
well suited for routine use in a clinical laboratory; it is cumbersome, time-consuming, and
technically demanding. Carvalho et al. (2006) have described a multiplex minisequencing
technique for the detection of most common 10 mutations. This assay allows detection of
p.R106W, p.A140V and p.G269fs mutations in addition to our assay. However, reagent
and equipment costs of minisequencing PCR limit the implementation of this assay into
research laboratories. Our multiplex ARMS-PCR assay offers a straightforward,
inexpensive, and accurate alternative method. Each mutation-specific primer pair was
designed to produce a different-sized fragment, so that the mutation in the sample can be
identified unambiguously after agarose gel electrophoresis.
121
The seven MECP2 mutations included in this assay were selected based on their
frequencies in RettBASE database. It is possible that the prevalence of the included
mutations may vary among RTT patients from different populations. However, the assay
could be improved by the inclusion of primer sets for detection of additional pathogenic
mutations. For example, Kim et al. (2006) and Yamada et al. (2001) did not identify the
mutation p.R106W in Korean and Japanese patients, respectively, but the same mutation
accounted for 5-6 per cent of the mutations in French and German RTT patients,
respectively (Bienvenu et al., 2002; Laccone et al., 2001). An ARMS assay was also
designed and tested the for the p.R106W mutation, however, it could not be implemented
to the panels. This mutation can be tested independently from the multiplex assay.
A well-designed ARMS multiplex requires compatible primer sequences in
appropriate concentrations. Optimization of the PCR reaction, started with equimolar
primer concentrations (10 pmol), revealed uneven amplification of the products, with some
of the alleles barely visible. This problem was overcome by increasing the concentration of
primers for the ‘weak’ loci and decreasing for the ‘strong’ loci. The specificity of the
primer sets used was also found to be highly critical to ensure clear discrimination between
target and non-target genomic sequence to avoid any ambiguity of detection. Primer
dimers and nonspecific allelic noise was observed with primers for p.T158M and p.R168X
mutations. These nonspecific bands could be eliminated by introducing a mismatch at the
3′ end of the allele-specific primers. As for any multiplex PCR, we recommend that users
optimize the conditions, especially the primer concentrations for each set, in their
laboratory.
In conclusion, the multiplex ARMS test is an efficient and cost-effective screen for
molecular genetic testing of patients with RTT. It is a simple and reliable test that does not
require any specialized equipment and can be completed in 1–2 days upon receiving the
sample.
122
6.1.2. The Effect of DNA Concentration on Reliability and Reproducibility of SYBR
Green Dye-based Real Time PCR Analysis to Detect the Exon Rearrangements
Up to date more than 200 different mutations of MECP2 have been reported in
patients with classical and atypical RTT (RettBASE). MECP2 exon rearrangements are
very frequently observed and identified in 2.9- 14 per cent of patients with RTT.
Previously, the identification of MECP2 gene rearrangements was carried out by
traditional methods including Southern blotting, fluorescent in situ hybridization (FISH),
and Long-Range PCR and subsequent DNA sequencing analysis. So far, no
rearrangements of the MECP2 gene have been identified by FISH and these approaches are
time-consuming and may suffer from a limited sensitivity (e.g., the size of
rearrangements). To overcome these problems, rapid and sensitive PCR based assays
including quantitative fluorescent PCR (QF-PCR), robust dosage PCR (RD-PCR), and
multiple ligation-dependent probe amplification (MLPA) have been developed. In these
methods, end-point PCRs are accomplished in 20–25 cycles, when amplification is
supposed to be in its exponential phase such that a linear relationship between quantities of
template DNA and PCR products is maintained. These assays should be considered semi-
quantitative since the actual amplification profile of the reaction is based on a theoretical
assumption. Real-time PCR was specifically developed to quantify specific DNA targets
through the monitoring of product formation. This technology has been successfully
applied for the detection of hemizygous deletions or duplications in different genetic
disorders. Recently, quantitative real-time PCR assays based on SYBR Green I dye and
TaqMan probe has been developed for the detection of deletions/duplications of the
MECP2 gene (Ariani et al., 2004; Laccone et al., 2004).
Several studies have shown that DNA extraction methods, presence of inhibitors and
inefficient homogenization of the sample may lead to false negative results. This may lead
to reduced efficiency of real time PCR reaction and underestimation of the quantity of the
target DNA upon reaction. The purpose of our study was to analyse the effect of DNA
concentration on reliablity and reproducibility of SYBR green dye-based real time PCR
analysis to detect the exon rearrangments.
123
We have tested normal DNA samples along with duplication and deletions using six
different DNA concetrations (ranging from 0.1 to 200 ng) in triplicate measurements. The
expected copy number ratio (± 2SD) was obtained in all cases when DNA concentration is
between 1 to 50 ng. The ratio was out of expected range in six and four of 15
measurements with 0.1 and 100 ng DNA, respectively, and led to misgenotyping of the
samples. Expected copy number could never be obtained with 200 ng DNA. These results
suggested that Real Time PCR analysis might not be reliable for determination of the exon
copy number with DNA in the range of 1-50 ng.
The effect of the DNA concentration on the amplification efficiency and specifity
could be observed in amplification curve and melting curve analysis. Using high DNA
concentrations (100–200 ng) resulted in inhibition of the amplification and/or nonspecific
product formation in some cases. For dilute DNA samples several hypotheses have been
proposed to explain misgenotyping. The first issue is the apparent labiality (instability) of
the DNA upon prolonged storage periods at low concentrations. Ellison et al. (2006) has
shown that plastic tube had a strong effect on measured DNA concentration at
concentrations below 100 genome equivalents. Teo et al. (2002) has observed fluctuations
in the concentration of standard solutions even after short time storage in eppendorf tubes
due to binding of DNA to the tube walls. Secondly, due to the stochastic distribution of
molecules at very low copy number, a sampling error can be introduced when pipetting
aliquots of DNA. Measurement variability at low DNA concentration has been
demonstrated by the observation that the confidence intervals (representing the
measurement uncertainty) associated with amplification from low initial copy numbers of
template are much greater than those with high initial copy numbers (Peccoud and Jacob,
1996).
Additionally, we have compared the results obtained using comperative Ct and
standard curve methods. A paired sample t-test showed that the results by both methods
were significantly similar (P = 0.451>0.05). In addition, the Pearson correlation coefficient
of 0.996 (P = 1E-6 <0.01) demonstrated the equivalence of both methods.
Our study shows that data obtained by the standard curve method and by the
comperative Ct method are equally reliable and correlate extremely well. However, the
124
comperative Ct method appers to be more convenient and efficient to analyze the exon
rearrangments in real time PCR experiments. First, it improves the productivity by
eliminating the time and effort required to prepare the standards and set up the standard
curves. Second, the Ct method reduces the overall cost of assays by reducing the number
of reactions run each time and thereby sparing expensive reagents. The only disadvantage
of comperative Ct method is use of a single calibrator (reference) DNA sample and the
contamination of target DNA with salt, phenol, chloroform and/or ethanol that may cause a
low PCR efficiency and miscalculations. In case of standard curve method, serial dilution
of the reference sample will also dilute these inhibitors and decrease its effect on the PCR
reaction, thereby increasing the PCR efficiency with each dilution step.
6.2. Methylation Analyses of the Putative Promoter Region of hHR23 Genes in
Breast Tumor Tissues
Carcinogenesis is a multistep process composed of genetic and epigenetic alterations
involving proto-oncogenes, tumor suppressor genes, cell-cycle regulator genes, tissue-
invasion-related genes, or mismatch repair genes. Aberrant cytosine methylation of CpG-
rich sites was regarded as an epigenetic mechanism for the transcriptional silencing of
several repair genes (BRCA1, hMLH1, O6-MGMT, TDG, and WRN) in different types of
cancer including breast carcinoma. Recently, Peng et al. (2005) has shown that hHR23B
gene, a key component in nucleotide excision repair pathway, was epigenetically silenced
in Interleukin-6-responsive Multiple Myeloma KAS-6/1 cells lines. This latter finding
prompted us to investigate the methylation status of hHR23B and its homolog hHR23A
gene in tumor tissues. Additionally, the following cellular functions of both hHR23A and
hHR23B make them candidate tumor susceptibility genes: (1) hHR23A/B were also shown
to participate not only in NER but also in base excision repair (BER) pathway; (2) The
hHR23B
KD
cell lines and hHR23A/B KO mice exhibit severe UV sensitivity and NER
deficiency; (3) hHR23A/B are involved in induction and stability of the damage-signaling
tumor-suppressor protein p53; and (4) hHR23B was shown to be required for genotoxic-
specific activation of p53 and apoptosis.
In this study, we have analyzed the methylation status of 5’ flanking regions
(including the CpG islands and putative promoter sequence) of hHR23A and hHR23B
125
genes in primary breast tumor, tumor adjacent tissues, and normal breast tissues in order to
investigate their possible involvement in breast carcinogenesis.
First of all, we have characterized the CpG islands and the putative promoter region
in the 5' flanking region of the hHR23A and hHR23B genes using web-based analysis.
MethPrimer and PROSCAN softwares revealed two CpG islands (at positions -580/-454
and -341/-55 nts) and a putative eukaryotic Pol II promoter region within the second CpG
island (position -48 to -298 nts) with a score of 98.79. The promoter region lacks the
CCAAT and TATA-like elements, a common feature of the house-keeping genes. The
analysis revealed potential binding sites for the transcription factor Sp1 (two sites),
ATF/CREB, Elk-1 and two M22 motifs (5’-TGCGCANK-3’). The analysis of 5’ flanking
region of hHR23B gene revealed a CCAAT and TATA- box lacking putative Pol II
promoter sequence (from position -14 to -264 nts with a score of 217.35) containing four
Sp1 binding sites.
The lack of TATA and CAAT boxes, the presence of high C+G content and Sp1
binding sites in the hHR23A and hHR23B promoters are typical features of house-keeping
genes. Yang et al. (2007) has shown that three transcription factor binding motifs, Sp1,
Elk-1, and M22 are preferentially found in promoters that lack TATA elements. TATA-
less promoters are generally enriched in the Sp1 motif that can direct weak transcription
initiation from Transcription Start Site in vitro from core promoters.(Smale and Kadonaga,
2003). Elk-1, an ETS domain transcription factor of the TCF (ternary complex factor)
subfamily, is known to be involved in the regulation of immediate-early genes such as c-
fos
upon mitogen activation, and thus commonly implicated in cell proliferation. The M22
motif (5’-TGCGCANK-3’) is the most intriguing one because its potential role in
regulation of TATA-independent transcription is not known. The ATF family of
transcription factors can form either homodimers or heterodimers with c-Jun and
subsequently bind to the cyclic AMP response element (CRE) (5’-TGACGTCA-3’) (van
Dam et al., 1993). The transcription factor ATF-2 is a nuclear target of stress-activated
protein kinases (such as p38), that are activated by various extra-cellular stresses, including
UV light, osmotic stress, hypoxia, and inflammatory cytokines (Morrison et al., 2003).
ATF-2 plays a critical role in hypoxia- and high-cell-density-induced apoptosis, growth
control, and the development of mammary tumors. The ATF-2 mRNA levels in human
126
breast cancers were lower than those in normal breast tissue (Maekawa et al., 2007). Since
hHR23A was shown to interact with stress-related factors (eEF1A, Hsp70, and Hsp71)
(Chen and Madura, 2006), it might be a novel target of stress-activated protein kinases via
the transcription factor ATF.
The methylation status of the CpG islands was investigated by bisulfite-sequencing
method. Semi-nested PCR strategy was used to amplify the bisulfite modified DNA
samples. PCR fragments could be produced in 38 samples for hHR23A gene and 58
samples for hHR23B gene out of 61 samples tested. The failure in amplification might be
due to poor-quality of DNA samples. It is known that the extraction of high-quality nucleic
acid from formalin-fixed tissues might not be possible because of cross-linking between
proteins and DNA. The fixation and paraffin embedding processes might also damage the
DNA. Additionally, formalin-fixed tissues undergo degradation possibly because of an
inadequate neutralization of the formalin, eventually resulting in acid depurination. The
acid is known to depurinate the DNA and destroy its structure, thus preventing
amplification (Goelz et al., 1985). In retrospective studies using fixed tissues, the primers
must be carefully chosen to generate smaller amplification products, because larger DNA
fragments are more difficult to amplify (Rivero et al., 2006). This is in accordance with our
results that amplification of hHR23B gene was more successful than hHR23A gene (95%
vs 62%) since the hHR23B primers generated smaller PCR products. The agarose gel
electrophoresis showed that paraffin embedded tissue DNAs are more fragmented when
compared to DNAs isolated from peripheral blood supporting the findings of poor
preservation and high degradation of DNA extracted from fixed tissues.
The methylation status of hHR23A and hHR23B genes could be determined in 35 of
38 and 51 of 58 samples, respectively. Briefly, methylation analysis of 5’ flanking region
of hHR23A gene revealed cytosine methylation in 12 tumor and 2 tumor adjacent tissues.
Hypermethylation was observed in four tumors (from patients Rad21, 27T, 31T, and 32T)
and one tumor adjacent tissue sample (Rad31N). Two patients (Rad17 and 28T) have
single CpG methylation and seven patients (Rad4, 6, 15, 29T, 29N, 33T, and 48) have non-
CpG methylation. Cytosine methylation (CpG and non-CpG) in the upstream of hHR23B
gene was observed in 15 tumor tissues. CpG methylation was observed in six patients
(Rad9, 13, 16, 20, 22, and 27T) and non-CpG methylation was also present in three of
127
them (Rad9, 20, and 22). Nine patients have only non-CpG methylation. The methylated
cytosine residues were within the Sp1 binding sites in patients Rad9, 17, 22, 27T, 32T, and
50. The methylated C*CpWGG motif was present in four samples (Rad2, 9, 25T, and
30T).
The methylation analysis revealed either hypermethylation or cytosine methylation
of single CpG and non-CpG methylation in the putative promoter region of hHR23 genes.
It is known that hypermethylation of CpG island in the promoter region is generally
associated with transcription silencing. Intriguingly, Pogribny et al. (2000) and Veerla et
al
. (2008) have shown that single CpG methylation could down regulate the expression of
the p53 and IRF7 genes, respectively. It was also reported that methylation of single non-
CpG dinucleotides (CpA or CpC) within the transcription binding site could affect the
expression pattern of the genes (Veerla et al. 2008).
Until recently, a few studies reported cytosine methylation of non-CpG dinucleotides
in genomic DNAs from human carcinomas and its involvement in the carcinogenesis. Two
recent studies have reported non-CpG methylation pattern in B cell-specific B29 gene in
Primary Effusion Lymphoma, and p53 gene in non-small cell lung carcinoma (Malone et
al
., 2001; Kouidou et al., 2005 and 2006). Analysis of the p53 exon 5 mutation spectrum in
mutation databases for lung cancer revealed frequent G:C > A:T transitions, several of
which occur at non-CpG methylated sequences. Additionally, non-CpG methylation was
observed in the tissues adjacent to the tumor in the lung, which indicates that non-CpG
methylation may appear in the early stage of carcinogenesis and serve as a useful tool for
early cancer detection (Kouidou et al., 2005). Thus, the findings of present study might be
the additional support for the involvement of non-CpG methylation in carcinogenesis.
A specific pattern of non-CpG methylation in the C*CpWGG motif was reported in
plants and few human genes including p53 and B29 genes. The cytosine methylation of
C*CpWGG motif in the promoter region of p53 and B29 genes resulted in transcriptional
silencing (Malone et al., 2001; Agirre et al., 2003). We have identified methylated
C*CpWGG motif in hHR23B in four tumor samples (Rad2, 9, 25T, and 30T). The
methylated C*CpWGG motif was at positition -212/-208 in Rad2 and 30T whereas
C*CpWGG motif at positition -151/-147 was methylated in samples Rad9 and 25T. Based
128
on the above observations, we could speculate that the presence of methylated C*CpWGG
motif in the promoter region might affect the transcription of the hHR23B gene in the
tumor tissues.
We have observed that the methylated cytosine residues were within the Sp1 binding
sites in hHR23B gene in patients Rad9, 17, 22, 27T, 32T, and 50. The affected Sp1 binding
site (at position -135/-126) is the same in five of them whereas the Sp1 site at position -
270/-261 was methylated in tumor sample Rad27T. Butta et al. (2006) has shown that the
in vitro transcription of the human podocalyxin (Podxl) promoter is dependent on the
presence of Sp1 sites. The progressive rise in the promoter activity directly correlates with
the number of recognition sites for Sp1. The methylation or deletion of Sp1 element
resulted in repression of Podxl gene (Butta et al., 2006). The findings of Butta et al. (2006)
and Veerla et al. (2008) prompted us to speculate that the cytosine methylation in Sp1
binding site might interfere the binding of Sp1 and down regulate the transcription of
hHR23B in tumor tissues Rad9, 17, 22, 27T, and 32T.
In five samples, hHR23A gene was partially hypermethylated. However, the extent
and the position of the methylation site were different in each sample. The down-stream
region (positions +6/+117) of TSS site was hypermethylated in tumor adjacent tissue
Rad31N whereas hypermethylation spread to the upstream region (between -249/+117 nts)
of TSS site in tumor tissue of the same sample (Rad31T). The upstream region was
hypermethylated in sample Rad21 whereas both upstream and downstream regions were
hypermethylated in samples Rad27T, 31T, and 32T. The observed non-uniformity in
hypermethylation pattern suggest that the methylation initiation site is probably similar in
four samples and within downstream of TSS (position +1/+117) whereas it is different in
sample Rad21 probably being between -249/-27 nts. It is known that the spreading of
methylation from the foci of methylated CpG sites is a common event in tumor tissues.
Two samples have cytosine methylation within upstream region of hHR23A in both
tumor and adjacent tissues. Patient Rad29 has CpC methylation at position -114/-113 in
tumor adjacent tissue (Rad29N) whereas CpT and CpA di-nucleotides were methylated at
positions -133/-132 and +11/+12 in the tumor tissue (Rad29T). The region between
nucleotide positions +6/+117 was hypermethylated in tumor adjacent tissue Rad31N
129
whereas hypermethylation spread to the region between -249/+117 nts in tumor tissue of
the same sample (Rad31T). The observation of cytosine methylation in tissues adjacent to
tumor is consistent with previously reported findings (Kouidou et al., 2005). The cytosine
methylation pattern in samples Rad25T, 27T, 28T, 30T, 32T, and 33T was not present in
corresponding tumor adjacent tissues suggesting that the observed de novo methylation is
specific to tumor formation. Additionally, cytosine methylation was not observed in five
DNA samples isolated from peripheral blood tissue.
We could not obtain statistical differences among patients when compared with
respect to the presence, type, and position of methylation in hHR23A or hHR23B. On the
other hand, correlations based on histopathological features of the samples implicate some
preliminary features. For example, in all patients (Rad21, 27T, 31T, and 32T) showing
hypermethylation of hHR23A gene, lymph node metastasis (LNM) was positive and
showed high grade (III) and stage (pT2Nx) tumor progression. The Estrogen Receptor
(ER) expression was positive in all samples. c-erbB-2 expression was negative in tumor
samples Rad27T, 31T, and 32T, however,it was expressed in tumor tissue Rad21 showing
different methylation initiation site in the upstream region of TSS. Among the patients
(Rad9, 17, 22, 27T, 32T, and 50) presenting methylated cytosine residues within the Sp1
binding sites in hHR23B gene, there is no uniformity in histopathological features of tumor
tissues. However, the bisulfite sequencing revealed that three of them (samples Rad17,
27T, and 32T) have cytosine methylation in both hHR23A and hHR23B genes. The sample
Rad27T showed hypermethylated region (between nts -226/+117) in hHR23A and
methylated cytosine residue within the Sp1 binding site (-270/-261) in hHR23B gene.
Similarly, upstream region (between -121/+117 nts) of hHR23A gene was
hypermethylated along with the cytosine residue within the Sp1 binding site (-135/-126) in
hHR23B gene in patient Rad32T. Sample Rad17 has CpC methylation at position -116/-
115 and methylated cytosine residue within the Sp1 binding site (-135/-126) in hHR23A
and hHR23B genes, respectively. Among the patients (Rad2, 9, 25T, and 30T) having the
methylated C*CpWGG motif in the hHR23B promoter region, the histological grade of
tumor showed differences according to the position of methylated C*CpWGG motif. The
tumor grade was II in samples Rad2 and Rad30T whereas the grade is higher in samples
Rad9 and Rad25T.
130
The methylated cytosine residues were not seem to be within the conserved motifs or
transcription binding sites in samples Rad4, 6, 15, 17, 28T, 29T, 29N, 33T, and 48 for
hHR23A and Rad13, 14, 16, 20, 46, and 51 for hHR23B genes. Bisulfite sequencing of
HHR23A and hHR23B revealed no methylation in 21 and 36 tissue samples, respectively.
There was no significant histopathological difference among these samples.
In conclusion, there are no known reports investigating the role of methylation of
hHR23A and hHR23B genes in the tumor tissues (breast cancer). The observations of the
hypermethylation of hHR23A gene and the presence of methylated conserved motifs and
transcription binding sites in hHR23B gene among tumor tissues suggested the
involvement of methylation of hHR23 genes in the breast carcinogenesis. However, the
expression pattern of methylated hHR23 genes should be investigated in fresh or frozen
tumor tissues. If the promoter methylation correlates with loss of protein expression,
methylation status of hHR23 genes could be a marker in breast carcinomas. It is known
that the genetic and epigenetic alterations that initiate and drive cancer can be used as
targets for detection of neoplasia in bodily fluids. Several studies showed that tumor cell-
specific aberrant promoter hypermethylation can be detected in nipple aspirate and ductal
lavage from breast cancer patients and the results were concordant between tumor and
circulating DNA methylation (Dulaimi et al., 2004; Mirza et al., 2007). Hypermethylation-
based screening of serum, a readily accessible bodily fluid and pre-invasive method, may
enhance early detection of breast cancer.
Dostları ilə paylaş: |