| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Institute of Veterinary Medicine, Georg-August-University, Burckhardtweg 2, 37077 Göttingen, Germany;
aauthor for correspondence: fax 49-551-39-3392, e-mail eschuetz{at}mac.com
Abstract
Background: Scrapie is the transmissible spongiform encephalopathy in sheep. Because genetic variants of the ovine PrP gene (PRNP) can be associated with disease risk, the European Union initiated programs to eradicate high-risk PRNP genotypes from sheep livestock. For this purpose, reliable and cost-effective genotyping is needed.
Methods: We amplified DNA to cover the 3 risk codons in exon 3 encoding amino acids 136, 154, and 171. Amplicons were mixed with dye-labeled probe sets, and melting curves were recorded in a LightCycler by use of color and temperature multiplexing. Probe design was based on thermodynamic calculations to ensure unequivocal results for the 3 codons of interest, taking the additional F141 and T137 sequence variants into account.
Results: The fluorescence resonance energy transfer (FRET) method, when compared with sequencing, gave exactly the predicted melting temperatures for all possible genotypes. When we validated the method with samples from official certification programs, it showed completely matching results. Turnaround time was 
5 h after receipt of a whole-blood sample. The method detected the rare sequence variants T137 and F141, which were clearly distinguishable from the other known genotypes by melting curve analysis. One scrapie sheep was ARR/ARR, which is considered the haplotype with the lowest risk.
Conclusions: The FRET-based PRNP genotyping method for sheep is rapid and can differentiate all genotypes at each locus in 1 capillary. The assay is fast and has lower costs than restriction fragment length polymorphism analysis or sequencing.
Scrapie was the first transmissible spongiform encephalopathy described, with reports going back to the 18th century. Although direct transmission to humans is considered impossible, scrapie has been hypothesized as the natural source of the transmissible infectious agent, the prion protein (PrP), in the epidemic of bovine spongiform encephalopathy in the United Kingdom. This raised concern for human health and led to a European Union–wide program for eradication of scrapie from sheep flocks (1)(2).
In contrast to bovine spongiform encephalopathy (3), but like human transmissible spongiform encephalopathy(4), scrapie displays a strong risk dependency on PRNP genotype(4). Three variants in the gene have been described, encoding amino acid changes at positions 136, 154, and 171, that are used for risk prediction(5). Risk groups (see Table 1 in the Data Supplement that accompanies the online version of this Technical Brief at http:www.clinchem.org/content/vol52/issue7) are used in the decision-making process for culling schemes in scrapie herds and for breeding programs(1).
Because of the small profit margins on sheep, the price for genotyping of the 3 loci must be no more than 10 
per animal. Methods available for ovPRNP genotyping are based on sequencing (6), restriction fragment analysis(5), or primer extension(7), which require several technical steps. In contrast, the technology of fluorescence resonance energy transfer (FRET) used in the LightCycler avoids additional steps, and the test can be done directly on extracted DNA(8). Our goal was to develop an inexpensive assay to identify the 15 known genotypes at the 3 positions.
We extracted genomic DNA from sheep, using the QIAamp DNA Blood Mini Kit (cat. no. 51106; Qiagen). Approximately 20 ng of DNA was used for PCR (total reaction volume, 25 µL) with 0.8 µM forward primer (5'-GTTGGCTACATGCT*GGGAAG-3', where * is fluorescein), 0.4 µM reverse primer (5'-TGTTGACACAGTCATGCACAAA-3'), and the PuReTaq PCR reagent set. PCR was performed in a Personal cycler (Biometra) thermocycler and included 5 min of initial denaturation at 95 °C, followed by 38 cycles of 95 °C for 1 min, 55 °C for 30 s, and 72 °C for 2 min (Amersham Biosciences). The product (19 µL) was transferred to the sample receptacle of a LightCycler capillary containing a premade probe mixture consisting of 0.25 µM of each sensor oligonucleotide (as given in Table 1
) and 0.3 µM of anchor oligonucleotide. We used nearest-neighbor calculations in assay design, using MeltCalc 2.3 (9)(10) to develop an ovPRNP genotyping assay to identify all 15 known haplotypes at the 3 positions. This was accomplished by use of a combination of color and temperature multiplexing(11)(12) using in silico–designed oligonucleotide probes and anchors(13).
|
The placements of probes, anchor, and primers are given in Fig. 1 of the online Data Supplement. For codon 136 and 171 alleles, we used 6-carboxy-X-rhodamine (ROX)-labeled probes with one channel (channel F2: 640 nm) with temperature multiplexing, whereas the 154 allele was detected in the second channel (channel F3: 705 nm) with a Cy5.5-labeled probe (see Table 1
).
The melting curve was recorded after 1 min of denaturation (95 °C) followed by reannealing at 35 °C for 1 min and detection during heating to 75 °C at 0.2 °C/s. For comparison and validation purposes, we also analyzed PRNP genotypes by use of Big-Dye Terminator sequencing on a Model ABI3100 sequencer (Applied Biosystems), as described elsewhere. Sequencing of the whole coding region of exon 3 was accomplished with primers specific for the 3' and 5' introns.
As shown in Fig. 1A
, the genotype call for codons 136 and 171 in channel F2 of the instrument is simple because the melting curves are well separated. The grouping into 2 high-temperature peaks for codon 136 and 3 low-temperature peaks for codon 171 leads to a reliable visual genotype call on a prima vista basis. This holds true for codon 154 as well (Fig. 1B
), for which the 2 alleles are easily distinguishable. The time required to evaluate a run of 32 samples in a LightCycler rotor was 5 min.
|
We compared the results obtained with the LightCycler assay with sequencing in a random set of 100 samples representing all genotypes. In all cases, the LightCycler results agreed with sequencing. The assay was assessed in national and international external PRNP genotyping quality assurance schemes, with complete agreement as to the underlying PRNP genotypes. The overall frequency of repeat tests for samples that failed (e.g., insufficient amplification) was 1% (e.g., 7 of 823 samples done in 1 month). The between-run imprecision for the melting temperature (Tm) is given in Table 1
. The within-run Tm variation was negligible.
Two additional sequence variants, T137 and F141 [the latter is often referred to as NOR98 (14)], linked to so-called atypical scrapie cases(15) need to be considered because they are covered by the diagnostic probe for the A136 allele and can occur in considerable frequencies in specific breeds(15)(16)(17)(18). As shown in Fig. 1C
, the T137 variant can easily be distinguished from both wild-type (A136) and the risk-associated variant V136. We observed T137 in a German Bentheim sheep.
The F141 variant leads to a penultimate mismatch with the codon 136 probe with a calculated (homozygous) Tm that is 2 °C lower than the perfectly matched probe. As shown in Fig. 1C
, in the heterozygous state, the F141 variant gives a Tm that is 1 °C lower than the Tm for A136 and considerably higher than the Tm for V136. The F141 cases were confirmed by use of a F141 matching probe (see Fig. 2 in the online Data Supplement). To date, we have found this variant in several sheep as well as in one confirmed scrapie sheep (ARR/AF141RQ). The herd to which that sheep belonged was completely genotyped. Of 845 sheep, we where able to find the F141 variant in 24 animals (allelic frequency =1.4%), linked to the ARQ genotype (ARQ/ARQ, n = 9; AHQ/ARQ, n = 2; ARR/ARQ, n = 13).
One confirmed scrapie sheep had the homozygous ARR genotype, which is considered to be the lowest risk group. For this sheep, we sequenced coding exon 3 of ovPRNP, including the 5' and 3' intron boundaries for confirmation. We found no deviation from the wild-type ovPRNP, which makes this one of the rare scrapie cases in the low risk (G1) genotype class (19).
Because prices are important in ovPRNP genotyping, a thorough calculation of the costs is mandatory. Using the parameters given in Table 2 in the online Data Supplement, we calculated that the cost for genotyping of 1 sample by sequencing is 7 (controls and overhead not included), whereas the LightCycler assay can be performed for approximately one half this cost. The costs of restriction fragment length polymorphism analysis are intermediate between of these two. An overview of the cost analysis is given in Table 1
.
Genotyping must meet several key criteria, including reliability, fast turnaround, and cost-effectiveness. The latter holds true in particular for diagnostics in livestock because of the small profit margins. With sequencing, the cost for diagnostic genotyping in a laboratory is almost the same as the reimbursement values per animal in the European Union (1). We therefore evaluated a method on the LightCycler with the primary aim of genotyping all 3 codons in one capillary. We now demonstrate that this is possible by use of in silico probe design guided by physical principles combined with color and temperature multiplexing(11)(12). High throughput is achieved by use of the LightCycler only for the diagnostic melting, whereas the specific PRNP PCR is done in a conventional thermocycler at lower cost.
One disadvantage of FRET analysis is the risk of erroneous results attributable to additional sequence variations that may be covered by the diagnostic probe. Here we are facing a special situation because one haplotype (ARQ) is known to have a high linkage to variants within the open reading frame (17)(20)(21), of which 2 (T137 and F141) are in the probe regions. Both can be clearly distinguished. In temperature multiplexing, probe construction is mainly sequence driven, which in this case leads to a high Tm probe for A136 and a lower Tm probe for R171. In the latter case, a low temperature ensures good separation of the Q171 and H171 genotypes. The routine use of this assay revealed the first T137 case in a German Bentheim sheep. This sequence variant had not been described in this breed. The second rare sequence variation, F141, which affords a high odds ratio for scrapie(15), is also known as NOR98(14) and was seen in so-called atypical scrapie accompanied by a distinct PrP deposition pattern in the brain. The performance of the A136 probe could be improved by use of an internally labeled primer (see Fig. 1 in the online Data Supplement). This gives a wider choice of probes because an overlap of 3 bp is well tolerated in the assay, as calculated with the nearest-neighbor model(13). The F141 variant could be set to the penultimate position of the probe, with a minor destabilizing effect compared with V136 but with sufficient difference to be seen the LightCycler. We saw this sequence variant in one sheep with confirmed scrapie, again showing the importance of F141, which was present in 2.8% of that herd. The case of 1 homozygous ARR sheep suffering from scrapie shows again that G1 sheep are not resistant to scrapie, as was thought previously.
In summary, this assay shows that evidence-based in silico probe design is capable of providing genotyping methods that meet the present needs of good reliability coupled with cost-effectiveness.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. T. McKay, T. A. Brigner, B. E. Caplin, K. S. McCurdy, and R. L. Forde A real-time polymerase chain reaction assay to detect single nucleotide polymorphisms at codon 171 in the prion gene for the genotyping of scrapie susceptibility in sheep J Vet Diagn Invest, March 1, 2008; 20(2): 209 - 212. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Vaccari, C. D'Agostino, R. Nonno, F. Rosone, M. Conte, M. A. Di Bari, B. Chiappini, E. Esposito, L. De Grossi, F. Giordani, et al. Prion Protein Alleles Showing a Protective Effect on the Susceptibility of Sheep to Scrapie and Bovine Spongiform Encephalopathy J. Virol., July 1, 2007; 81(13): 7306 - 7309. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
