New Technologies in Human Genetic Analysis

Human genetics has benefited in the last decade from spectacular technological advances in DNA analysis. The rate of data acquisition has increased by orders of magnitude. To cope with this flood of information, semiautomated methods have been developed for primary data collection and interpretation of genotyping and direct DNA sequencing. Here the authors examine several improvements in experimental design and analysis, expediting disease gene discovery in family-based genetic studies.

Microsatellite genotyping

Microsatellite markers, also known as short tandem repeats (STRs), arise from consecutive repeat units such as CACACA . . . or GATAGATAGATA . . . embedded within unique sequences.¹ The human genome contains many thousands of such repeats of di-, tri-, and tetranucleotide types.²  PCR primers designed in the surrounding unique sequence allow unambiguous amplification of particular microsatellites. Many of these sites are polymorphic within the human population, segregating multiple alleles with different repeat lengths. These PCR amplicons can serve as genetic markers for the corresponding chromosomal segments as they segregate through families. Microsatellite marker length is normally conserved over several meiotic generations, making these effective tools for pedigree-based linkage analysis.³

Microsatellites offer several technical challenges.⁴  PCR enzymes tend to stutter at the repeats, leading to a series of smaller-than-full-length peaks whose lengths usually differ by one repeat unit.⁵ More serious is the tendency of these enzymes to add an additional, nontemplated nucleotide at the 3′ end of the product. When the extent of such addition is variable, this leads to the phenomenon of peak splitting, which can render dinucleotide markers practically useless. One way to reduce variability in peak splitting is to add a specific sequence tag at the 5′ end of one of the PCR amplification primers. When fluorescent genotyping is being performed, this should be the nonlabeled primer (often though not always formalized as the reverse primer). Several such sequence tags have been reported,^6,7 although their exact mechanism of action remains uncertain.

The authors have verified the effectiveness of such a sequence tag, using either specialized plusA or nominally standard PCR conditions. As shown in Figure 1a and b, two different dinucleotide markers varied in the sensitivity of splitting to the PCR protocol employed when default primer design was used. However, addition of the 5′ tag on the unlabeled primer minimized variability for both markers, making genotype calling less problematic in each case. This suggests that the tag should routinely be included on all customized microsatellite markers as a preventive measure.

Figure 1 -  Microsatellite genotyping before/after reverse 5′ sequence tag. a) Fluorescent primers for D2S133 were used to amplify DNA from one randomly selected sample, which is homozygous for a particular allele of this marker. Amplification was performed under two different PCR protocols, using either the tagged or untagged reverse primer. From top to bottom: minus tag, plusA PCR; minus tag, standard PCR; plus tag, plusA PCR; plus tag, standard PCR. In the absence of the reverse tag, use of plusA PCR conditions biases products toward nontemplated nucleotide addition, as seen by comparing the first and second panels. The addition of the reverse tag increases amplicon size, as shown in the two bottom panels, and reduces variability in peak splitting under different PCR conditions. Note that the stutter peaks at 2-bp intervals are independent of PCR conditions and presence of reverse tag. GeneMarker genotyping software was used to identify and call the allele as 297 bp in the bottom panels (gray vertical bar). b) D16S520, from top to bottom: minus tag, plusA PCR; minus tag, standard PCR; plus tag, plusA PCR; plus tag, standard PCR. Similar to panel (a), but this marker is heterozygous for alleles at 196 and 198 bp, as called by GeneMarker (gray vertical bars) for the reverse tagged version of the marker. Peak splitting is much more severe for this marker, but addition of the reverse tag still reduces the variation significantly. As above, addition of the reverse tag increases the size of the amplicon.

SNP genotyping

Single-nucleotide polymorphisms (SNPs) provide an alternative to microsatellites for genetic mapping. Individual SNPs are overwhelmingly biallelic, hence intrinsically less informative than microsatellites. However, multiple SNPs analyzed in tandem have the potential to carry the same or greater information content than microsatellites. Several millions of SNPs have been genotyped as part of the HapMap project, providing allele frequencies in several different populations.⁸ High-density SNP panels are now commercially available. When used appropriately, these panels can be used for genetic mapping of monogenic disorders in the traditional family-based paradigm.

To compare the effectiveness of dense SNP panels to microsatellite markers for pedigree-based genetic mapping, the authors sought to verify a known genetic linkage, using a large Nova Scotia Acadian family segregating the recessive trait Niemann-Pick type D.^9,10 The underlying causal gene was identified as NPC1 (OMIM #257220) by positional cloning, ¹¹ and homozygous mutations in NPC1 were confirmed in affected individuals in the Nova Scotia kindred.¹² The authors genotyped two distantly related affected individuals plus one unaffected sibling from this family (see Figure 2) using the Xba 50K chip (Affymetrix, Santa Clara, CA). As seen in Table 1, the longest stretch of SNPs with shared homozygosity for the same allele in the two affecteds was approx. 7.7 Mbp on chromosome 18, including 71 consecutive SNPs around the NPC1 locus located at position 19.4 Mbp. Unaffected sibling 2244 was discordant (either heterozygous or else homozygous for the other SNP allele) for 30 of these markers distributed across the interval (data not shown). Thus, the SNP chip successfully replicated the linkage in this subset of samples from the entire pedigree with substantial savings in time, labor, and cost.

Figure 2 - Niemann-Pick type D (now called type C) pedigree. Individuals 2245 and 2247 (affected) and 2244 (unaffected) were genotyped using the Xba 50K SNP chip (pedigree simplified from Greer et al.¹⁰).

Subtle discrepancies between absolute length and number of consecutive SNPs across multiple chromosomal regions led the authors to examine the distribution of the entire 116,000 marker set, including both the Xba and Hind chips. The distribution of gap length is bimodal, with peaks at approx. 400 and 22,000 bp. Excluding the centromeres, there are three gaps between 2 and 3 Mbp in length and 33 gaps between 1 and 2 Mbp in length. Some of these gaps are telomeric, and some are relatively gene-poor, but others are in gene-dense regions presumably lacking appropriate restriction sites near informative SNPs. The number of gaps in the 1–3 Mbp range suggests that caution should be used in interpreting high-density linkage disequilibrium (LD) experiments. The newest generation of ultrahigh-density 500K SNP panels are now becoming available and these may be preferable for some experimental designs.

Mutation detection

The final step in a gene discovery experiment is mutation detection. Despite the use of indirect physicochemical detection methods such as denaturing HPLC (dHPLC), DNA sequencing still provides the gold standard for sequence variant (i.e., mutation) detection. Manual review of large amounts of sequence data is inefficient and at risk for missing mutations through human error. Thus, several semiautomated methods have been developed to serve this growing need in molecular genetic analysis.^13–17

The authors evaluated Mutation Surveyor (MutSurv, SoftGenetics, Inc., State College, PA) for sequence variant detection, validating the software using several samples containing known variants including single base changes and insertion/deletions. The software aligns and compares sample traces to reference or virtual consensus sequence traces and performs the detection algorithm, reporting potential mutations/polymorphisms with quality scores, difference chromatograms, and graphic outputs. The software imports and utilizes genomic annotation if provided, including location of exons and introns, open reading frames, and known variants.

The software detected all previously identified variants in the samples tested (Figure3a and b). In the case of single-nucleotide variants, the software flagged the mutation and conveniently interpreted the expected effect on the protein coding potential for the gene. In several cases, MutSurv has identified known SNPs annotated in dbSNP. The graphic alignment along the exons in the software made it very simple to identify these known SNPs by comparison to the UCSC (University of California, Santa Cruz) genome browser interface. In the case of insertion/deletions, it not only detected several different mutations automatically, but also deconvoluted them and correctly described the exact sequence inserted or deleted (Figure 3c). All of these results were verified by manual review.

Figure 3 - Mutation detection with Mutation Surveyor. a) Automated detection of heterozygous missense SNP within FZD4 gene. Sequence chromatograms from potentially affected patients were aligned to a virtual consensus trace created by Mutation Surveyor, and the mutation detection algorithm was run using default parameters. Top panel displays the nucleotide sequences of the virtual reference (ZD4.gbk) and patient samples. Conceptual translation sequence is given for the reference and for sample 372Ec-R with heterozygous missense variant predicted. Middle panels show chromatogram traces for virtual reference sequence and sample 372Ec-R showing heterozygous missense change. Bottom panel shows difference trace between reference and missense change. b) Automated detection of homozygous SNP during positional cloning. As in panel a, Mutation Surveyor output showing alignment of virtual reference sequence of a candidate gene (anonymized) to sequence of an affected patient. Lower panels show chromatograms and difference trace at site of homozygous single-nucleotide change. c) Automated detection and deconvolution of 15-bp deletion in von Hippel-Lindau (VHL) Syndrome gene in affected patient. Mutation Surveyor output showing virtual reference trace, affected patient trace, inferred component traces of heterozygous deletion, and position of deletion within annotated coding exon.

MutSurv flagged several false positives resulting from poor-quality sequence, particularly near the ends of reads. New versions of the software allow for end trimming, which should reduce the extent of this problem. The incidence of false positives in low-quality sequence can also be reduced by requiring mutation detection in both directions, although acquiring sequence in both directions can be problematic with some amplicons.

New genetic initiatives

The authors have recently embarked on a population-wide genetic discovery effort to ascertain and molecularly characterize many monogenic human disorders in Eastern Canadian provinces (Figure 4). Such an effort requires the increased efficiencies possible with the technologies described here. Future improvements are likely to include even higher-density SNP chips; reduced costs for SNP chips; lower DNA sequencing costs; and ultimately cost-effective whole-genome or whole-exon resequencing of individual probands.

Figure 4 - The four Atlantic Canadian provinces of New Brunswick, Newfoundland and Labrador, Nova Scotia, and Prince Edward Island have a total population of approx. 2.3 million. The region contains numerous subpopulation isolates defined by geography and/or ethnic heritage, with historically large families and a few large tertiary-care medical referral centers. These characteristics are optimal for ascertainment and genetic characterization of monogenic disorders. As many as 15 new genetic conditions may arise in the region every year by sporadic mutation; thus, the region provides a unique source of genetic information for the Human Genome Project.

Methods

Samples

DNA was prepared from whole blood using standard methods, or from saliva. Historically alternative methods to venipuncture have included buccal swabs and blood spots, but both of these methods are variable and give very low yields. The Oragene saliva kit (DNA Genotek, Inc., Ottowa, Ontario, Canada) was tested. From collections on six different individuals, genomic DNA yield varied from 20 to 320 μg per individual (for 2 mL of saliva collected). Undigested DNA appeared to be high molecular weight by agarose gel electrophoresis. The A260/A280 ratios (corrected for A320) ranged from 1.4 to 1.6, and some samples appeared slightly turbid, possibly due to residual carbohydrate or lipid components. This ratio can be improved by slight modifications to the protocol, including a 70% ethanol wash. All samples worked well for microsatellite genotyping and DNA sequencing with 10 ng input to PCR. The authors have not systematically evaluated the DNA for long-term storage or use in whole genome amplification, but it should be noted that this protocol has been validated by the manufacturer for use in Affymetrix high-density SNP chip experiments after repurification using Qiagen (Valencia, CA) kits.

Appropriate institutional ethics approval and informed consent from patients were obtained for use of all samples.

Microsatellite genotyping

Primers for annotated microsatellite markers were taken from the GDB database. Custom fluorescent microsatellite markers were developed from primary genomic sequence using Tandem Repeat Finder¹⁸ in combination with the UCSC genome browser,^19,20 Repeat Masker,^21,22 and Primer3.²³ Each amplicon was developed with one labeled forward primer and paired unlabeled reverse primers with or without a sequence tag of 5′-GTTTCTT-3′ on the 5′ end. Two different PCR cycling conditions were tested. Nominal “standard” conditions were: 95 °C 3 min (1 cycle); 95 °C 1 min, 55–60 °C 1 min, 72 °C 1 min (30 cycles); 72 °C 2 min (1 cycle). Specialized “plusA” conditions were: 95 °C 5 min (1 cycle); 94 °C 15 sec, 55 °C 15 sec, 72 °C 30 sec (10 cycles); 89 °C 15 sec, 55 °C 15 sec, 72 °C 30 sec (20 cycles); 72 °C 30 min (1 cycle).⁷ Electrophoresis was on 6% polyacrylamide gels using the ABI 377; chromatograms were generated using ABI GenScan software (ABI 377 and GenScan from Applied Biosystems, Foster City, CA). Genotype chromatograms were analyzed using GeneMarker (SoftGenetics, Inc.). Genotype calls were exported in text format for inheritance verification using PedCheck.²⁴

SNP genotyping

Genotypes were collected for the Xba 50K SNP chip at the Microarray Facility of the University of Toronto Hospital for Sick Children (Canada). Genotype call data were generated using Affymetrix software and provided in spreadsheet format from the facility. Data were collected on 58,960 SNPs, of which 58,494 had unique locations in the human genome. Lengths of homozygous SNP alleles identical by state were derived by direct query of the database, and sorted either by physical position or number of consecutive homozygous IBS markers. Centromeric gaps lacking sequence and marker information were deleted manually for each chromosome.

Mutation detection

Fluorescent DNA sequencing trace files were obtained following electrophoresis on the ABI 377. Traces were imported into Mutation Surveyor and analyzed for sequence variants. Genomic exon/intron and protein-coding annotation was provided from the manufacturer’s database site or from the National Center for Biotechnology Information (NCBI). Synthetic wild-type reference sequence traces were generated by the software from the consensus genomic sequence.

On-line databases and tools

National Center for Biotechnology Information (NCBI)
Genome Home Page: http://www.ncbi.nlm.nih.gov/
UCSC Genome Home Page: http://genome.ucsc.edu
Online Mendelian Inheritance in Man (OMIM): http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIM
Genome Canada: http://www.genomecanada.ca/GCprogrammesRecherche
Repeat Masker: http://www.repeatmasker.org
Tandem Repeat Finder: http://tandem.bu.edu/trf/trf.html
Primer3: http://frodo.wi.mit.edu/
Applied Biosystems: http://www.appliedbiosystems.com
Affymetrix:  http://www.affymetrix.com
SoftGenetics: http://www.softgenetics.com
GDB: http://www.gdb.org

References

1. Weber, J.L. Human DNA polymorphisms and methods of analysis. Curr. Opin. Biotechnol.1990, 1, 166–71.
2. Tamiya, G.; Shinya, M.; Imanishi, T.; Ikuta, T.; Makino, S.; Okamoto, K.; Furugaki, K.; Matsumoto, T.; Mano, S.; Ando, S.; Nozaki, Y.; Yukawa, W.; Nakashige, R.; Yamaguchi, D.; Ishibashi, H.; Yonekura, M.; Nakami, Y.; Takayama, S.; Endo, T.; Saruwatari, T.; Yagura, M.; Yoshikawa, Y.; Fujimoto, K.; Oka, A.; Chiku, S.; Linsen, S.E.; Giphart, M.J.; Kulski, J.K.; Fukazawa, T.; Hashimoto, H.; Kimura, M.; Hoshina, Y.; Suzuki, Y.; Hotta, T.; Mochida, J.; Minezaki, T.; Komai, K.; Shiozawa, S.; Taniguchi, A.; Yamanaka, H.; Kamatani, N.; Gojobori, T.; Bahram, S.; Inoko, H. Whole genome association study of rheumatoid arthritis using 27 039 microsatellites. Hum. Mol. Genet. 2005, 14, 2305–21.
3. Ghosh, S.; Karanjawala, Z.E.; Hauser, E.R.; Ally, D.; Knapp, J.I.; Rayman, J.B.; Musick, A.; Tannenbaum, J.; Te, C.; Shapiro, S.; Eldridge, W.; Musick, T.; Martin, C.; Smith, J.R.; Carpten, J.D.; Brownstein, M.J.; Powell, J.I.; Whiten, R.; Chines, P.; Nylund, S.J.; Magnuson, V.L.; Boehnke, M.; Collins, F.S. Methods for precise sizing, automated binning of alleles, and reduction of error rates in large-scale genotyping using fluorescently labeled dinucleotide markers. FUSION (Finland–U.S. Investigation of NIDDM Genetics) Study Group. Genome Res. 1997, 7, 165–78.
4. Samuels, M.E.; Dubé, M.P. Linkage Mapping. In Encyclopedia of Genetics, Genomics, Proteomics, and Bioinformatics; John Wiley & Sons, Inc.: New York, 2005.
5. Perlin, M.W.; Lancia, G.; Ng, S.K. Toward fully automated
   genotyping: genotyping microsatellite markers
   by deconvolution. Am. J. Hum. Genet.1995, 57,1199–1210.
6. Brownstein, M.J.; Carpten, J.D.; Smith, J.R. Modulation of non-templated nucleotide addition by Taq DNA polymerase: primer modifications that facilitate genotyping. Biotechniques1996, 20, 1004–6, 1008–10.
7. Magnuson, V.L.; Ally, D.S.; Nylund, S.J.; Karanjawala, Z.E.; Rayman, J.B.; Knapp, J.I.; Lowe, A.L.; Ghosh, S.; Collins, F.S. Substrate nucleotide-determined non-templated addition of adenine by Taq DNA polymerase: implications for PCR-based genotyping and cloning. Biotechniques1996, 21, 700–9.
8. Altshuler, D.; Brooks, L.D.; Chakravarti, A.; Collins, F.S.; Daly, M.J.; Donnelly, P. A haplotype map of the human genome. Nature2005, 437, 1299–1320.
9. Greer, W.L.; Riddell, D.C.; Byers, D.M.; Welch, J.P.; Girouard, G.S.; Sparrow, S.M.; Gillan, T.L.; Neumann, P.E. Linkage of Niemann-Pick disease type D to the same region of human chromosome 18 as Niemann-Pick disease type C. Am. J. Hum. Genet. 1997, 61, 139–42.
10. Greer, W.L.; Riddell, D.C.; Murty, S.; Gillan, T.L.; Girouard, G.S.; Sparrow, S.M.; Tatlidil, C.; Dobson, M.J.; Neumann, P.E. Linkage disequilibrium mapping of the Nova Scotia variant of Niemann-Pick disease. Clin. Genet. 1999, 55, 248–55.
11. Carstea, E.D.; Morris, J.A.; Coleman, K.G.; Loftus, S.K.; Zhang, D.; Cummings, C.; Gu, J.; Rosenfeld, M.A.; Pavan, W.J.; Krizman, D.B.; Nagle, J.; Polymeropoulos, M.H.; Sturley, S.L.; Ioannou, Y.A.; Higgins, M.E.; Comly, M.; Cooney, A.; Brown, A.; Kaneski, C.R.; Blanchette-Mackie, E.J.; Dwyer, N.K.; Neufeld, E.B.; Chang, T.Y.; Liscum, L.; Strauss III, J.F.; Ohno, K.; Zeigler, M.; Carmi, R.; Sokol, J.; Markie, D.; O’Neill, R.R.; van Diggelen, O. P.; Elleder, M.; Patterson, M.C.; Brady, R.O.; Vanier, M.T.; Pentchev, P.G.; Tagle, D. Niemann-Pick C1 disease gene: homology to mediators of cholesterol homeostasis. Science1997, 277, 228–31.
12. Greer, W.L.; Riddell, D.C.; Gillan, T.L.; Girouard, G.S.; Sparrow, S.M.; Byers, D.M.; Dobson, M.J.; Neumann, P.E. The Nova Scotia (type D) form of Niemann-Pick disease is caused by a G3097-->T transversion in NPC1. Am. J. Hum. Genet. 1998, 63, 52–4.
13. Nickerson, D.A.; Tobe, V.O.; Taylor, S.L. PolyPhred: automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing. Nucl. Acids Res.1997, 25, 2745–51.
14. Flood, E.M.; Tang, F.; Horvath, M.M.; Pertsemlidis, A.; Garner, H.R. SNPCEQer: detecting SNPs in sequences generated by the Beckman CEQ2000 DNA Analysis System. Biotechniques2002, 33, 814–20.
15. Weckx, S.; Del-Favero, J.; Rademakers, R.; Claes, L.; Cruts, M.; De Jonghe, P.; Van Broeckhoven, C.; De Rijk, P. novoSNP, a novel computational tool for sequence variation discovery. Genome Res. 2005, 15, 436–42.
16. Manaster, C.; Zheng, W.; Teuber, M.; Wachter, S.; Doring, F.; Schreiber, S.; Hampe, J. InSNP: a tool for automated detection and visualization of SNPs and InDels. Hum. Mutat. 2005, 26, 11–19.
17. Crowe, M.L. SeqDoC: rapid SNP and mutation detection by direct comparison of DNA sequence chromatograms. BMC Bioinform. 2005, 6, 133.
18. Benson, G. Tandem repeats finder: a program to analyze DNA sequences. Nucl. Acids Res.1999, 27, 573–80.
19. Karolchik, D.; Baertsch, R.; Diekhans, M.; Furey, T.S.; Hinrichs, A.; Lu, Y.T.; Roskin, K.M.; Schwartz, M.; Sugnet, C.W.; Thomas, D.J.; Weber, R.J.; Haussler, D.; Kent, W.J. The UCSC Genome Browser Database. Nucl. Acids Res. 2003, 31, 51–4.
20. Kent, W.J.; Sugnet, C.W.; Furey, T.S.; Roskin, K.M.; Pringle, T.H.; Zahler, A.M.; Haussler, D. The human genome browser at UCSC. Genome Res. 2002, 12, 996–1006.
21. Jurka, J. Repbase Update: a database and an electronic journal of repetitive elements. Trends Genet. 2000, 16, 418–20.
22. Jurka, J.; Kapitonov, V.V.; Pavlicek, A.; Klonowski, P.; Kohany, O.; Walichiewicz, J. Repbase Update, a database of eukaryotic repetitive elements. Cytogenet. Genome Res. 2005, 110, 462–7.
23. Rozen, S.; Skaletsky, H. Primer3 on the WWW for general users and for biologist programmers. Meth. Mol. Biol. 2000, 132, 365–86.
24. O’Connell, J.R.; Weeks, D.E. PedCheck: a program for identification of genotype incompatibilities in linkage analysis. Am. J. Hum. Genet. 1998, 63, 259–66.

Dr. Samuels is with the Department of Medicine, University of Montreal, Centre de Recherche du Chum, Hopital Notre-Dame, Local Y-3633, 2099 rue Alexandre de-Seve, Montréal QC H2L 2W5, Canada; tel.: 514-890-8000; e-mail: [email protected]. Mr. Marcadier is currently with the Dalhousie Faculty of Medicine, Halifax, Canada. Mr. Higgins and Dr. Bowman are with the Atlantic Genome Centre, Halifax, Canada; Mr. Higgins is also with the National Research Council Institute for Marine Biosciences, Halifax, Canada. Ms. Provost and Dr. Dubé are with the Montreal Heart Institute Research Center, Montréal, Canada. Dr. Blouin is with the Faculty of Computer Science, Dalhousie University, Halifax, Canada. The authors wish to thank Dr. Wenda Greer for the Niemann-Pick DNA samples, and Drs. Andrew Orr and Duane Guernsey for additional DNA samples. DNA sequence chromatograms containing verified mutations are courtesy of Drs. Johane Robitaille, Duane Guernsey, and Christie Riddell. The authors also wish to thank the families and patients who generously contributed their time and materials for this research. Dr. Samuels and Mr. Marcadier were supported by Dalhousie University, the IWK Health Centre (Halifax, Canada), the Dalhousie Medical Research Foundation, Genome Atlantic, Genome Canada, and the Capital District Health Authority. Mr. Higgins and Dr. Bowman were supported by Genome Canada/Genome Atlantic. Dr. Dubé was supported by the Fonds de Recherche en Santé du Québec (FRSQ). Dr. Blouin was supported by Genome Canada and the Natural Sciences and Engineering Research Council of Canada (NSERC).