|
|
||||||||
Articles |
1
Unité de Neurogénétique Moléculaire, Laboratoire de Biochimie, Hôpital de lAntiquaille, F-69005 Lyon, France.
2
Département de Génétique
Moléculaire et de Biochimie Clinique, Faculté de Pharmacie,
F-69008 Lyon, France.
3
Département de Génétique
Médicale, Hôpital dEnfants de la Timone, F-13005
Marseille, France.
aAddress correspondence to this author at: Laboratoire de Biochimie, Hôpital de lAntiquaille, 1, rue de lAntiquaille, 69321 Lyon CEDEX 05, France. Fax 33-04-7238-5110; e-mail philippe.latour{at}chu-lyon.fr.
| Abstract |
|---|
|
|
|---|
Methods: We tested 10 STRs located within the duplication for polymorphic behavior. Three STRs were selected and used to test a set of 130 unrelated CMT1A patients and were compared with nonduplicated controls. The study was then extended to a larger population of patients. Alleles of interest were sequenced. A manual protocol using polyacrylamide electrophoresis and silver staining and an automated capillary electrophoresis protocol to separate fluorescently labeled alleles were validated.
Results: We identified three new STRs covering 0.55 Mb in the center of the CMT1A duplication. One marker, 4A, is located inside the PMP22 gene. The two others, 9A and 9B, more telomerically positioned, have the highest observed heterozygosity reported to date for CMT1A markers: 0.80 for 9A, and 0.79 for 9B. Tetra- and pentanucleotide repeats offered clear amplification, accurate sizing, and easy quantification of intensities.
Conclusions: Combined use of the three STRs allows robust diagnosis with almost complete informativeness. In our routine diagnosis for CMT1A, they have replaced the use of other polymorphic markers, either in a manual adaptation or combined with fluorescence labeling and allele sizing on a DNA sequencer.
| Introduction |
|---|
|
|
|---|
1 in 2500 (1). Two major
forms of CMT can be identified electrophysiologically: one form shows
defects in the formation or maintenance of myelin (CMT1) and the other
primary axonal degeneration (CMT2) (2). A microduplication
of 1.5 Mb containing the gene for peripheral myelin protein 22
(PMP22) on chromosome 17p11.2 is responsible for 75% of
cases of the demyelinating form (CMT1A) (3)(4)(5)(6)(7)(8)(9). Diagnosis
with markers located inside the duplication usually is carried out
using restriction fragment length polymorphism (RFLP) probes
(3)(4)(10) and poly(AC) repeats
(11)(12)(13), and less frequently with fluorescence in situ
analysis (14) or pulse-field gel electrophoresis
(15). Molecular diagnosis often relies on the interpretation
of differences in allele intensities, even with the most polymorphic
markers that have been reported to date. When poly(AC) repeats are used for molecular diagnosis, artifact bands produced by slippage of the polymerase enzyme may lead to difficulties in interpretation of dosage for different alleles. Other useful diagnostic methods have been developed after extensive investigation of the two repeated 30-kb sequences (REPs) flanking the duplicated region (16). Recombination events leading to the CMT1A duplication occur almost entirely in four adjacent regions within 7.8 kb of the CMT1A-REP sequences, with a "hot spot" of recombination located in a 3.2-kb central segment defined by two restriction enzyme sites, EcoRI and SacI (17)(18). Characterization of the junction fragment resulting from recombination allows positive identification of the duplication with RFLP probes on Southern blots (17)(18).
Several methods with direct allele-specific amplification of the junction fragments by PCR have also been reported (19)(20)(21)(22). However, to date, no PCR strategy covering all of the possible recombination events has been reported; thus, the overall efficiency of this approach is difficult to evaluate. Moreover, infrequent variation in the CMT1A-REP sequences (23)(24)(25), chimerism of the duplication junction fragments (26), and possible recombination outside the CMT1A-REP sequences (27)(28) make the methods based on identification of the CMT1A junction fragments prone to misinterpretation. We have developed a new strategy with polymorphic short tandem repeats (STRs) located inside the CMT1A duplicated region that allow amplification with very low or no stuttering, accurate sizing, and visual quantification of allele intensity, which offers the choice between inexpensive detection without labeling, using nondenaturing polyacrylamide gels and silver staining, or the use of fluorescent-labeled primers and separation on an automated sequencing apparatus.
| Materials and Methods |
|---|
|
|
|---|
5 repeats
and (b) the possibility of finding unique specific primers
for amplification (Table 1
|
|
manual adaptation
PCR amplification was carried out in standard conditions in a
final volume of 25 µL in the presence of 20 pM of each primer and 0.5
U of EuroblueTaq® ADN polymerase (Laboratoires
Eurobio) for 35 cycles. Final reaction conditions were 67 mM Tris-HCl,
16 mM
(NH4)2SO4,
and 0.1 mL/L Tween 20. STRs showing locus-specific amplification and a
polymorphic pattern for five unrelated persons on a 2% agarose gel
were chosen for subsequent analysis. Each PCR product (50 ng) was
loaded on a nondenaturing 8% polyacrylamide gel. Gels 0.75 or 1 mm
thick were used indifferently. Gels were stained using the Silver Stain
reagent set from Bio-Rad Laboratories, according to the manufacturers
instructions.
evaluation of STRs
Samples from 130 unrelated patients carrying a CMT1A duplication
identified with probes pVAW409R3 and pEW401 were tested with the
selected STRs. For each of the duplications, the recombination event
had been assigned to one of the four recombination regions in the
CMT1A-REP sequences using probe pNEA102 on
EcoRI+SacI digestions and probe pJ5P on
EcoRI+HincII digestions
(17)(18).
A complementary study was carried out with the three selected STRs for nine unrelated CMT1A patients with absence of the EcoRI+SacI junction fragment and with no dosage differences between the alleles using the pNEA102 CMT1A-REP probe. These patients, correctly identified with markers located inside of the duplication, would have been misdiagnosed with the pNEA102 probe. They represent 9 of 480 unrelated CMT1A patients (1.9%) identified in our laboratory
Seventy controls showing two alleles with RFLP probes D17S122 and D17S61 were also evaluated. We also analyzed 50 patients with hereditary neuropathy with liability to pressure palsies (HNPP), who were hemizygous for the PMP22 region (34), as confirmed by the observation of a deletion-specific junction fragment obtained with probe pNEA102.
sequencing of alleles
Sequencing of alleles was performed mainly on genomic DNA from
HNPP patients. For some alleles, genomic DNA from healthy controls was
amplified and cloned in pGEM®-T Easy (Promega
Corporation). Sufficient clones were randomly selected and sequenced to
obtain the sequence of both alleles. Sequencing was carried out with
plasmid-specific T7 and SP6 primers and using fluorescently labeled
dideoxynucleotide triphosphate substrates
(PRISMTM Ready Reactions Rhodamine Terminator
reagent set; Applied Biosystems). Allele numbering was according to
proposed guidelines (35).
allele frequencies
For the three selected STRs, the distribution of alleles for
unrelated duplicated patients was compared with that observed for the
same number of alleles in controls (
2 test;
210 alleles for each sample). Because there was no significant
difference (P = 0.05), all available results were
pooled to calculate allele frequencies, with 486 (4A), 705 (9A), and
775 (9B) alleles, respectively.
fluorescent labeling and sizing on an automated sequencer
The forward primers for 9A and 9B were labeled with fluorescent
6-carboxyfluorescein (6-FAM), and 4A was labeled with
hexachloro-6-carboxyfluorescein (HEX; MWG Biotech A.G.). PCR
amplification conditions were as described above for unlabeled primers.
Products were analyzed on an ABI Prism 310 using a 47-cm capillary
filled with POP4 and using GeneScan® 2.1
software (all products from Applied Biosystems).
| Results |
|---|
|
|
|---|
Complete concordance (100%) was found between results obtained with the combination of RFLP probes pVAW409R3/pEW401/pNEA102 and with the three-STR (4A, 9A, and 9B) combination. In particular, all nine duplications for which EcoRI+SacI junction fragments were absent with the CMT1A-REP probe pNEA102 and that showed no dosage differences between alleles with this probe were also correctly identified with the three-STR combination.
STR 4A is ideally located in the first intron of the PMP22
gene, 3.5 kb upstream of exon 2. It has a complex structure (Table 2
). The distribution of alleles was bimodal (Fig. 2
) with a maximum value for the 14.2-repeat
(ATCT)3AT(ATCT)6(ACCT)5
with length 118 bp, and another for the 17-repeat
(ATCT)11(ACCT)6 with length
128 bp. The eight alleles characterized to date had sizes between 116
and 136 bp. Sequencing of four alleles revealed that at least two
different sequences underlie the most common allele, 14.2 (118 bp),
probably with equal frequency. This was also the case for the rarer
allele, 15 (Table 2
). From allele 16 on, the shorter
(ACCT)6 tract seemed to remain constant, whereas
the longer ATCT repeat was variable. Because of an insertion of a
dinucleotide in the ATCT tract, the difference in length of the shorter
alleles was only two nucleotides, but these were well resolved by
electrophoresis on polyacrylamide gel (Fig. 3
) and by capillary electrophoresis (not shown). The
heterozygosity of this locus for healthy controls was 0.69.
|
|
|
For STR 9A, the sequenced alleles had a uniformly repeated pentamer
(CAATA)n motif. Eleven alleles were observed in a
discrete symmetrical distribution around allele 13 (162 bp), starting
from 8 repeats (137 bp) up to 18 (187 bp; Table 2
and Fig. 2
). The
heterozygosity for healthy controls was 0.80. An example of
transmission of STR 9A alleles in a CMT1A family is given in Fig. 4
.
|
The repeat motif of STR 9B was composed of a tetranucleotide,
(TTTC)n, and variation gave a continuous set of
11 alleles. The number of repeats ranged from 7 (95 bp) to 17 (135 bp).
The maximum frequency was found for allele 12 (115 bp), followed by
alleles 11 and 14 (Table 2
and Fig. 2
). Observed heterozygosity for
healthy controls was 0.79. Examples of fragments obtained for this
locus are presented in Fig. 5
.
|
An example including all three loci is given in Fig. 6
in a family study where both parents and two children carry the
duplication and where the third child carries two duplications. The
family has been described in more detail elsewhere (36). The
histograms in Fig. 2
summarize the distributions for the three STRs. An
example of semiautomated separation is presented in Fig. 7
.
|
|
| Discussion |
|---|
|
|
|---|
The STRs had maximum allele distributions of 14 repeats (4A and 9A) and 12 repeats (9B), respectively. For the three STRs, the lowest repeat number was 7 or 8 and ranged up to 1719 repeats. By contrast, STRs 1A (five repeats), 5A (eight repeats), 6B (nine repeats), 8A (six repeats), 10A (six repeats) and 12A (seven repeats) did not show a large variation of alleles. Because the sequenced clones in the databank most probably represent a frequent allele, candidate STRs should be selected with 1214 repeats to find other highly polymorphic STRs. The most polymorphic STRs have simple motif repeats (9A and 9B), but compound repeats such as 4A must also be considered when testing for polymorphisms.
The three markers cover
0.55 Mb at the center of the CMT1A region,
including the PMP22 gene sequence. The localization of STR
4A inside the genomic sequence of PMP22 is particularly
valuable. At least two different sequences underlie the most frequent
allele, 14.2 (118 bp), probably with equal frequency. Sequencing of
this STR may help to establish the presence of two copies of the
PMP22 gene when no informativeness is found at the
PMP22 gene locus. On the distal side, STRs 9A and 9B are
separated by only 20 kb and are located 130150 kb upstream of the
group of markers D17S125, D17S839, D17S1358, and D17S61.
The observed heterozygosity of 0.69 at locus 4A is close to the highest heterozygosities reported for poly(AC)s of the CMT1A region: D17S122, 0.74 (3); D17S793, 0.70; D17S921, 0.74 (11)(12); and D17S1358, 0.74 (13). STRs 9A and 9B are to date the most polymorphic loci within the CMT1A duplication, with observed heterozygosities of 0.80 and 0.79. When we studied 130 unrelated duplications, 78% showed, at least once, three alleles for one of the three markers 4A, 9A, and 9B. An efficiency of 85% was achieved using a combination of six poly(AC) markers (13). Overall, taking into account dosage differences between fragments, 96% (125 of 130) were informative for two or more STRs, and none was noninformative for the combination of the three markers 4A, 9A, and 9B.
Differences in allele size allow separation on a nondenaturing gel even
for the shortest STR 4A alleles, which differ by only 2 bp (Fig. 2
). In
addition, the absence of stuttering artifactual bands, often seen after
amplification of poly(AC) repeats, makes the reading easy, particularly
when dosage intensities are to be interpreted (
Figs. 24
).
Amplification results for poly(AC) loci D17S122 (3) and
D17S1357 (13) are more difficult to interpret. The three
STRs were tested successfully in a family with one child carrying four
copies of the PMP22 region and illustrating all possible
allele combinations (Fig. 6
).
The PCR primers were selected as close as possible to the repeated sequences. The presently defined optimal PCR amplification conditions are different for each STR and do not allow multiplex amplification. Overlap in fragment length occurs between loci 4A and 9B and does not allow loading on the same slot of a polyacrylamide gel. However, when locus-specific fluorescent primers are used, 4A and 9B can be readily separated and simultaneously recorded by the detection system of automated electrophoresis equipment.
We tried to evaluate this new polymorphic system for diagnostic testing of HNPP. HNPP patients are hemizygous for the region duplicated in CMT1A patients. When samples are available from two or more family generations, a transmission study with RFLPs, microsatellites, or with the three STRs presented here is most often informative. When only the patient is sampled, the methods become uninformative, and the use of a method such as the characterization of a deletion junction fragment is required to separate homozygotes and hemizygotes. In our series, most of the patients noninformative for the three STRs were revealed to be hemizygotes by the characterization of a deletion junction fragment with probe pNEA102. In a group of 213 patients (without deletion or duplication), we found 210 persons who showed two alleles for at least one of the three STRs and 3 persons who were noninformative (1.4%). This result is compatible with the 1.3% noninformative patients expected when absence of haplotype association between the three loci (0.20 x 0.21 x 0.31) is assumed. As a conclusion, HNPP can be tested with the three STRs, and undeleted patients should be detected with a probability >98%.
In conclusion, we present a novel reliable diagnostic method to detect the CMT1A duplication that combines the informativeness of highly polymorphic markers located in the center of the duplication with the simplicity and speed of PCR methods using STRs, without stuttering and interpretation problems. Both laboratories involved in the development of this new technique have chosen to adopt it as their only diagnostic approach.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
Y. C. Lee, C.T.R. Yu, K. P. Lin, M. H. Chang, S. L. Hsu, Y. F. Liu, Y. C. Lu, and B. W. Soong MPZ mutation G123S characterization: Evidence for a complex pathogenesis in CMT disease Neurology, January 22, 2008; 70(4): 273 - 277. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Sanahuja, E. Franco, R. Rojas-Garcia, E. Gallardo, O. Combarros, R. Begue, P. Granes, and I. Illa Central Nervous System Involvement in Hereditary Neuropathy With Liability to Pressure Palsies: Description of a Large Family With This Association Arch Neurol, December 1, 2005; 62(12): 1911 - 1914. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Verny, N. Ravise, A-L. Leutenegger, F. Pouplard, O. Dubourg, S. Tardieu, F. Dubas, A. Brice, E. Genin, and E. LeGuern Coincidence of two genetic forms of Charcot-Marie-Tooth disease in a single family Neurology, October 26, 2004; 63(8): 1527 - 1529. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. De Vos, K. Sermon, M. De Rijcke, V. Goossens, P. Henderix, N. Van Ranst, P. Platteau, W. Lissens, P. Devroey, A. Van Steirteghem, et al. Preimplantation genetic diagnosis for Charcot-Marie-Tooth disease type 1A Mol. Hum. Reprod., July 1, 2003; 9(7): 429 - 435. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. G.B. Leonard Improved Method for Diagnosis of Charcot-Marie-Tooth Type 1A: Patent Pending? Clin. Chem., May 1, 2001; 47(5): 807 - 808. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |