Building
an assay for the diagnosis of the R117H mutation using allele specific PCR in
IB3 cystic fibrosis cells
By: Rohini Shrivastava,
Kaycie Campbell, Michella
McCormick, Chelsey Klein
http://www.msu.edu/~shrivas4
Abstract:
Cystic Fibrosis is a genetic disease that
affects the lungs, liver, pancreas, intestines, reproductive tract, and skin.
It is caused by mutations in the CFTR gene, leading to a malfunction in the
CFTR chloride channel in epithelial cells (Welsh and Smith, 1995). One of the
mutations is a missense mutation in CFTR, called R117H, replacing arginine with
histidine at amino acid 117 on exon 4 (Sheppard et
al., 1993). The purpose of this study was to recreate a diagnostic assay
that accurately identifies R117H. Allele specific polymerase chain reaction was
used to amplify the region of the CFTR gene containing R117H and analysed using gel electrophoresis with the Yaku primers (Yaku et al,
2008) and primers designed by Ferrie et al
(1992). We hypothesized that the designed primers containing an intentional
mismatch at the third base pair from the 3’ end will anneal better than the Ferrie et al. published primers that contain an
intentional mismatch on the second base pair from the 3’ end, resulting in a
more accurate diagnosis of the R117H CFTR mutation because the intentional
mismatch on the third base pair will produce a bump that will either drive the
oligonucleotide away from the DNA if the final base pair doesn’t match the DNA
template, or not interfere with annealing if the final base pair does match the
template. The Yaku primers and the published primers
are predicted to have a band length of 692 and 237 base pairs respectively. A
control PCR using lambda virus DNA was conducted to test equipment and
determine the PCR cocktail, successfully resulting in a band 419 base pairs
long. DNA was extracted from IB3-1 cells to be used in the PCR experiments for
the published and Yaku primers, and was analyzed
using spectrophotometry and gel electrophoresis. The DNA was determined to be
pure with an average 260/280 ratio of 1.89 and total yield of 48.5 ug. Better detection of the R117H mutation is
significant because of its frequency in the overall population, and because it
will help physicians diagnose it with more certainty (Lilley et al, 2010).
Discussion:
Experiment Summary
The cystic fibrosis
transmembrane conductance regulator (CFTR) is a membrane protein located in
epithelial cells, that functions as a chloride
channel. The protein is 1480 amino acids in length, which are coded for by the
CFTR gene, a 230,000 base pair gene located on chromosome 7 (Awasthi et al., 2012). Mutations in each allele of
the CFTR gene lead to the autosomal recessive disease called cystic fibrosis,
which affects many organs including the lungs, liver, pancreas, intestines, and
reproductive tracts (Welsh and Smith, 1995). The R117H mutation is a G to A
point mutation at nucleotide 482, resulting in amino acid 117 coding for histidine instead of arginine. Although the R117H-CFTR
protein is still transported to the membrane, it has a reduced chloride current
making it a class IV mutation (Awasthi et al,
2012). PCR using allele specific primers has been shown to be an effective
method for detecting the CF R117H mutation (Ferrie et
al, 1992). We hypothesized that the designed primers containing an
intentional mismatch at the third base pair from the 3’ end will anneal better
than the Ferrie et al. published primers that
contain an intentional mismatch on the second base pair from the 3’ end,
resulting in a more accurate diagnosis of the R117H CFTR mutation because the
intentional mismatch on the third base pair will produce a bump that will
either drive the oligonucleotide away from the DNA if the final base pair
doesn’t match the DNA template, or not interfere with annealing if the final
base pair does match the template.
Original Predictions
The control allele
specific primers contain an intentional mismatch on the second base from the 3’
end with the first base on the 3’ end matching either the mutant or wild type
sequence (Ferrie et al, 1992). The published
annealing temperature was 60oC however,
our calculations using OligoAnalyzer3.1 by IDT DNA Technologies indicated that
this temperature is over the melting temperature of the primers and would
therefore be unsuccessful. An annealing temperature gradient of 50o-60oC was used.
Though time was
insufficient to test them, the designed Yaku primers, Forward primer 1 and Forward primer 2, were designed
to anneal specifically to either the wild type DNA (1) or the mutant (2). The
primers were designed with an intentional mismatch for both sequences on the
third base from the 3’ end, a match for both sequences on the second base from
the 3’ end, and a match for either the mutant or wild type sequence on the
first base from the 3’ end. Because pyrimidine/pyrimidine pairing are more
destabilizing than purine/pyrimidine pairings, the 3rd base mismatch was
intentionally made a T in order to force the primer further apart from the DNA
template (Ferrie et al, 1992). With this
primer design, Forward primer 1 should anneal only to the wild type sequence
and Forward primer 2 should anneal only when the R117H mutation is present. The
predicted annealing temperatures are 59oC for Forward primer 1
(wild type) and 57oC for Forward primer 2 (R117H), which were
determined using the OligoAnalyzer3.1 by IDT DNA
Technologies for melting temperatures (see methods).
A successful PCR and gel
electrophoresis that results in a band 692 base pairs long indicates a positive
test, while no band will indicate a negative test (Ferrie
et al, 1992) (Awasthi et al, 2012). DNA
lacking the R117H mutation (an individual homozygous for the wild type) is
therefore predicted to have one band at 692 base pairs when Forward primer 1 is
used. Homozygous DNA with the R117H mutation should have a band at 692 base
pairs when Forward primer 2 is used. (Awasti et al,
2012).
A heterozygous
individual is predicted to have a faint band of 692 base pairs that shows up
using either forward primer. This is because each primer will successfully
anneal, but only to half of the DNA present since there is one of each allele.
The final result is a PCR product with a DNA concentration of half the amount
of a homozygous individual for each primer set (Ferrie
et al, 1992).
Results and Ultimate
Findings
A preliminary control
PCR experiment on the lambda virus was used to test the PCR equipment and gel
electrophoresis. Using the protocol outlined in the methods section, bands were
produced at 418 base pairs (Figure 6). This showed that the equipment and
products used were functional and gave a baseline PCR recipe to troubleshoot
future CFTR research off of.
Cultured IB3-1 CF cells
were obtained from Dr. Douglas Luckie at Michigan
State University. Cells cannot be tested for cystic fibrosis in a whole state;
only pure DNA can be tested. The Generation Capture Column Kit from Qiagen was used and pure DNA was obtained from the cells
(Figure 3). The concentration of DNA in 14 samples from the purified DNA was
measured at 260/280 by a spectrophotometer and determined to have an average
value of .097 at 260 and a value of .05157 at 280. DNA is measured at a 260
wavelength and proteins are measured at a 280 wavelength. The ratio of
wavelengths is 1.89, which is considered ‘clean’ DNA (Ahn
et al, 1996). With the pure DNA, we were then able to test for the R117H
mutation using the published primers (Fahale and
Fischer, 2000).
Two experiments were
conducted using the published primers (see Methods). The first run resulted in
bands 1,250 base pairs in length that were no longer visible when the gel was
run longer to improve the separation of the 1 Kb plus ladder (Figure 7). It is
predicted that the bands were slightly different lengths and dissipated when
the gel was run longer and rephotographed. The
faintness of the bands could be a result of nonspecific binding. The second
experiment resulted in bands at 61 base pairs and was likely primer dimers
(Figure 8). We believe that the primers annealed to each other instead of to
the DNA.
Future Directions
Given more time,
troubleshooting could have been conducted to try to resolve the issues in the
published primer gels (Figure 7 and Figure 8). The first gel (Figure 7)
contained faint bands that were 1,250 base pairs long. In this gel, not only
were the bands not present at 237 base pairs as expected of properly replicated
DNA, they also disappeared after the gel was run longer to lengthen the 1 Kb
ladder. It most likely did not work because there was not enough DNA in the
mixture and the annealing temperature was too high. When redoing this experiment,
we would double the concentration of DNA (from 0.485 ug
to 0.97 ug) and decrease the annealing
temperature gradient from 50oC-60oC to 45oC-55oC. In the second gel
(Figure 8), a band length of 61 base pairs was found, which was not expected.
We assume that the ends of the two primers annealed to each other, creating a
primer dimer. We predict that this happened because the PCR mix was not kept at
a cold temperature, causing the DNA to degrade so the primers annealed to each
other instead of the DNA. In order to test the primer dimer theory, we would
create another PCR mix with the same ingredients outlined in the Methods
section, but without any DNA. The mix would then be run through the same
thermal cycler program. We would then compare the two results using a semi-log
plot.
The 1 Kb plus ladder did
not separate correctly in Figure 6. We believe this happened because we were
using too high of a concentration of the ladder. To troubleshoot this we would
use a recipe with 1 ul of the 10x 1Kb
plus ladder and 5ul of SSDD water instead of 2.5 ul
ladder and 5 ul loading dye. It would
then be pipetted into a 1% TBE agarose gel with 6 ul of the solution and 1 ul
of loading dye as well as the lambda virus PCR mix and run at 150 Volts for
30 minutes, in order to ensure the ladder spreads out fully.
With more time, the
published primers could be tested further to successfully replicate wild type
DNA, as well as DNA containing the R117H mutation. The PCR diagnostic assay
designed using the Yaku method could be tested and
optimized using DNA with the R117H mutation and with wild type DNA. We predict
that the primers designed using the Yaku method will
produce fewer false positives because the last base pair on the 3’ end will
fully anneal, allowing better extension (Yaku et
al, 2008). If the control tests using the Forward primer 1 and Forward
primer 2 resulted in fewer false positives than the experiments run using the
control primers, then it would support the hypothesis that the Yaku design is more accurate in detecting the R117H
mutation using allele specific PCR (Yaku et
al, 2008). Multiple tests would be done to determine the accuracy of both
the published and designed primer sets on wild type and mutant DNA. A
chi-squared test for independence would then be conducted to determine the
accuracy of each primer. A p-value of less than 0.05 would indicate that the
results are statistically significant, thereby rejecting the null hypothesis
that the designed primers are equally or less accurate than the published
primers.
Further Research
Previous studies
indicate that the R117H mutation works in cis
with the Poly-T tract. The Poly-T tract is a string of thymidine bases position
in intron 8 of the CFTR gene (Moskowitz et al,
2008). When the R117H mutation is combined with the 5T mutation on the same
allele (as well as another CF mutation), the phenotypic result is often classic
cystic fibrosis. Under the same conditions, with the exception being the 7T
mutation instead of the 5T, the patient is more likely to suffer from mild CF
or less severe CF diseases for instance, congenital bilateral absence of the
vas deferens (CBAVD) (Thauvin-Robinet et al,
2009). Further research could be done by designing a PCR assay that diagnoses
both the R117H and the Poly-T tract mutation, in order to predict the most
probable phenotype of an individual when combined with a second allele with a
CFTR mutation.
Figures:
Figure 8: Published
Primers PCR with Short Cycle: A PCR was run using the published primers designed by Ferrie et al (1992). A PCR mix with 1 ul of the purified DNA, 40 ul
of water, 5 ul of 10x buffer, 1 ul of 10 mM dNTPs, 1 ul of the forward
primer, 1 ul of the reverse primer, and 1 ul of the Taq
polymerase was pipetted into a PCR tube and were then put in a thermocycler. Denaturation occurred at 94oC for
the control primers, annealing temperatures were between 51oC and 58oC,
and extension took place at 72oC. The thermal cycler went through 17
cycles for each run of PCR. Well 1 is the 1 Kb+ ladder (with 1 ul ladder, 5 ul
water, and 2 ul dye), which
contains known band lengths of DNA to measure the lengths of bands in the PCR
product and serves as a control. Wells 2 – 14 contain 10 ul
of the PCR product and 2 ul of dye.
Wells 2-7 contain the PCR mix with wild type reverse primers and wells 8-13
contain the mix with mutant reverse primers.Faint
bands were found at a base pair length of 61 base pairs. Though a test was not
run, we expect there to be a primer dimer. Air bubbles were found in almost
every well. There were also formations of crystals in the gel.