Forensic Science International 137 (2003) 74 84
Multiplex PCR and minisequencing of SNPs
a model with 35 Y chromosome SNPs
Juan J. Sancheza,*, Claus BÅ‚rstinga, Charlotte Hallenberga, Anders Bucharda,
Alexis Hernandezb, Niels Morlinga
a
Department of Forensic Genetics, Institute of Forensic Medicine, University of Copenhagen, 11 Frederik V s Vej,
DK-2100 Copenhagen, Denmark
b
Departamento de Canarias, Instituto Nacional de Toxicología, Campus de Ciencias de la Salud, 38320 La Laguna, Tenerife, Spain
Received 22 January 2003; received in revised form 2 July 2003; accepted 7 July 2003
Abstract
We have developed a robust single nucleotide polymorphism (SNPs) typing assay with co-amplification of 25 DNA-fragments
and the detection of 35 human Y chromosome SNPs. The sizes of the PCR products ranged from 79 to 186 base pairs. PCR
primers were designed to have a theoretical Tm of 60 5 8C at a salt concentration of 180 mM. The sizes of the primers ranged
from 19 to 34 nucleotides. The concentration of amplification primers was adjusted to obtain balanced amounts of PCR products
in 8 mM MgCl2. For routine purposes, 1 ng of genomic DNAwas amplified and the lower limit was approximately 100 pg DNA.
The minisequencing reactions were performed simultaneously for all 35 SNPs with fluorescently labelled dideoxynucleotides.
The size of the minisequencing primers ranged from 19 to 106 nucleotides. The minisequencing reactions were analysed by
capillary electrophoresis and multicolour fluorescence detection. Female DNA did not influence the results of Y chromosome
SNP typing when added in concentrations more than 300 times the concentrations of male DNA. The frequencies of the 35 SNPs
were determined in 194 male Danes. The gene diversity of the SNPs ranged from 0.01 to 0.5.
# 2003 Elsevier Ireland Ltd. All rights reserved.
Keywords: Y chromosome; Single nucleotide polymorphism; Multiplex PCR; Minisequencing; Genotyping
1. Introduction parentage testing and forensic casework in the future. The
advantage of SNPs in forensic casework is that small DNA
A large number of single nucleotide polymorphisms fragments of 40 50 bps from e.g. heavily degraded DNA can
(SNPs) have been identified [1]. Investigations of SNPs be SNP typed. Furthermore, the SNP technology has a high
on the Y chromosome in various populations have given potential for automation. Although the genetic information
us important information on the history of the human male obtained by a SNP, in average, is much lower than that
populations (e.g. [2 8]). Due to the low mutation rates of obtained by an STR system, typing of 50 100 selected SNPs
SNPs, the information relates to longer periods of time would be sufficient for forensic casework [14]. The low
compared to the information obtained with e.g. short tandem mutation rate of SNPs [15,16] makes these markers an
repeat (STR) [9 11] and minisatellite markers as, for exam- attractive tool for parentage testing.
ple MSY1 [12,13]. Genetic markers on the Y chromosome are valuable tools
Presently, typing of selected short tandem repeat (STR) in forensic casework in special situations, e.g. in cases with
systems is the state of the art in forensic routine casework. It mixtures of DNA with a dominant amount of female DNA
is, however, anticipated that SNP typing will be used for and a very small amount of male DNA. In such cases, the
DNA profile of the autosomes of the male cannot be
*
obtained, but the Y chromosome markers can usually be
Corresponding author. Tel.: þ45-35-32-62-25;
typed, even in situations with a very large relative amount of
fax: þ45-35-32-61-20.
E-mail address: [e-mail] (J.J. Sanchez). female DNA [17]. In special cases of parentage testing, e.g.
0379-0738/$  see front matter # 2003 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/S0379-0738(03)00299-8
J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84 75
if the alleged father is unavailable for testing while close Hoechst 33258 (Molecular Probes Inc., Eugene, OR) using
male relatives are available, investigation of genetic markers a Hoefer DyNA Quant 200 instrument (Molecular Vision).
on the Y chromosome are valuable. Calibration reference curves were established using a calf
If SNP typing is going to be used in forensic casework, it thymus DNA standard (Sigma Aldrich, Missouri, USA).
is essential that the investigations can be performed on small
amounts of DNA, if possible, <1 ng DNA. If the polymerase 2.2. Selection of PCR amplification primers
chain reaction (PCR) is used, the amplifications of all DNA
fragments to be investigated must be done in one or very few The Y chromosome SNPs selected (Table 1) included
amplification reactions. those used by Semino et al. [21] for a study of the distribu-
We decided to explore a SNP typing method that is based on tion of Y chromosome SNPs in European populations. In
multiplex PCR and multiplex minisequencing. We chose SNP addition, we included SNPs that were reported to be poly-
markers on the Y chromosome because these markers, in morphic in other ethnic groups.
forensic genetics, offer additional information to the informa- DNA segments including the SNPs selected were identi-
tion obtained by STR typing. Furthermore, the Y chromosome fied and complementary primers were designed so that the
SNPs are useful tools for the study of genetics of populations. lengths of the amplified genomic Y chromosome DNA
In the last years, a number of multiplex PCR Y chromo- fragments would range from 79 to 186 nucleotides. Some
some SNP analyses have been reported. Most of them SNPs were situated very closely to each other and it was
included a limited amount of SNPs (often 3 10 SNPs) in decided to include a number of amplification targets with
each PCR (e.g. [2,3,18]) although larger multiplexes have two or three SNPs (Table 1).
been reported [19,20]. The sequence of each locus was obtained from GenBank1
We selected Y chromosome SNPs that were reported to be (http://www.ncbi.nlm.nih.gov) using a nucleotide basic local
polymorphic in European and other populations [4,21,22]. alignment search tool (BLAST). Published PCR primers
However, the main purpose of the study was to explore the were initially used as the reference sequence for each Y
technical issues related to multiplexing a larger number of SNP locus, but all of them needed to be redesigned.
DNA fragments and simultaneous detection of a large The primers for the genomic segments spanning one or more
number of SNPs. The intention was not to make a final Y chromosome markers were designed with the Primer
panel for typing of major Y chromosome haplogroups. In 3.0 program v. 0.2 (http://www-genome.wi.mit.edu/cgi-bin/
order to assess the technical performance of the SNP typing primer/primer3_www.cgi). All primers were selected to have
system, we included four pairs of SNPs each of which pair theoretical melting temperatures of 60 5 8C at a salt con-
was expected to give concordant results (e.g. M40 and M96). centration of 180 mM and a purine:pyrimidine content close
Here, we describe a method for typing 35 SNPs on the Y to 1:1, when possible. The lengths of the primers ranged
chromosome. The typing was performed by (1) multiplex between 19 and 34 nt. Primers with four or more bases at the
PCR amplification of 25 Y chromosome DNA fragments, (2) 30 end complementary to another part of the primer were
multiplex primer extension reactions of 35 SNPs with discarded or redesigned to avoid artefacts due to hairpin
fluorescence labelled nucleotides, and (3) detection of the formation. Each primer pair was tested for primer primer
35 SNPs by capillary electrophoresis and multicolour fluor- interactions, and the primer sequences were checked to avoid
escence detection. similarities with repetitive sequences or with other loci in the
genome. The primers were checked for homology to other
amplicons in the pool of 25 primer pairs. Table 1 shows the
2. Materials and methods sequences of the amplification primers selected.
2.1. Donors and DNA preparations 2.3. PCR conditions
A total of 194 unrelated males and 15 unrelated female HPLC purified primers for amplification were purchased
Danes donated blood samples or buccal cells. DNA was from TAG A/S (Copenhagen, Denmark). A primer stock
isolated from 200 ml of peripheral blood using QIAamp solution was prepared by dissolving the lyophilized primers
DNA Blood Mini Kit according to the manufacturer s pro- in Tris/EDTA buffer (10 mM Tris, 100 mM EDTA, pH 7.5;
tocol (Qiagen, Hagen, Germany). Alternatively, 1.2 mm Sigma Aldrich) to a final DNA concentration of 100 pmol/ml.
(diameter) FTA1 paper (Whatman International, Cam- Each primer pair was tested in singleplex PCR. Ten ng
bridge, UK) soaked with blood or buccal cells was used. template was amplified by PCR in a 25 ml reaction volume
Mixtures of DNA from males and females were prepared in containing 1 PCR buffer, 1.5 mM MgCl2, 200 mM of each
checker board with three concentrations of male DNA (0.16, dNTP, 0.4 mM of each primer, and 0.6 units of AmpliTaq
0.8 and 1.6 ng) and female DNA ranging from 0 to 60 ng. Gold DNA polymerase at 94 8C for 5 min followed by 30
Fluorometric measurement of DNA concentration was cycles of 30 s at 95 8C, 30 s at 60 8C, 30 s at 72 8C, and a
done by SYBR Green I and analysed in a LightCycler final extension for 5 min at 72 8C. The products were
instrument (Roche Diagnostics GmbH, Germany) and analysed by electrophoresis in 11% polyacrylamide gels.
76 J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84
Table 1
Y chromosome SNPs and primer sequences for PCR amplification of 25 Y chromosome DNA fragments with SNPs
Locus GenBank or Mutation PCR primers (50 ! 30) mM Amplicon
dbSNPs accesion size (bp)
Forward primer Reverse primer
M2/sY81 Rs3893 A/G acggaaggagttctaaaattcagg aaaatacagctccccctttatcct 0.15 128
M9a Rs3900 C/G aggaccctgaaatacagaactg aaatatttcaacatttcacaaaggaa 0.36 186
M17a Rs3908 4G/3G cctggtcataacactggaaatc agctgaccacaaactgatgtaga 0.09 170
M18a Rs3909 2 bp insertion cctggtcataacactggaaatc agctgaccacaaactgatgtaga 0.09 170
M19a Rs3010 T/A cctggtcataacactggaaatc agctgaccacaaactgatgtaga 0.09 170
M32a AC009977 T/C tgaccgtcataggctgagaca ttgaagcccccaagagagac 0.07 160
M33a AC009977 A/C tgaccgtcataggctgagaca ttgaagcccccaagagagac 0.07 160
M35 Rs1179188 G/C agggcatggtccctttctat tccatgcagactttcggagt 0.42 96
M40/SRY4064 AC006040 G/A tggtctcaatctcttcaccctgt catttcagtaaatgccacacaaga 0.18 119
M45a Rs2032631 G/A gagagaggatatcaaaaattggcagt tgacagtggcaccaaaggtc 0.03 138
M46/Tat AC002531 T/C tatatggactctgagtgtagacttgtga ggtgccgtaaaagtgtgaaataatc 0.46 115
M52 AC009977 A/C cctcaacttcccagagtgttg gacgaagcaaacatttcaagagag 0.03 152
M78a AC010889 C/T tgcattactccgtatgttcgac tggaagcttaccatctttttatga 0.08 132
M81a Rs2032640 C/T catctcttaacaaaagaggtaaattttgtcc cattgtgttacatggcctataatattcagt 0.24 179
M89 Rs2032652 C/T tggattcagctctcttcctaaggttat ctgctcaggtacacacagagtatca 0.03 135
M96 AC010889 G/C tgccctctcacagagcactt ccacccactttgttgctttg 0.27 143
M123 AC010889 G/A gttgcccaggaatttgcat cacagagcaagtgactctcaaag 0.02 88
M139a AC010137 5G/4G ccccgaaagttttattttattcca ttctcagacaccaatggtcctatc 0.06 113
M151a AC010889 G/A catctcttaacaaaagaggtaaattttgtcc cattgtgttacatggcctataatattcagt 0.24 179
M153a AC010137 T/A ccccgaaagttttattttattcca ttctcagacaccaatggtcctatc 0.06 113
M154a AC010889 T/C catctcttaacaaaagaggtaaattttgtcc cattgtgttacatggcctataatattcagt 0.24 179
M157a AC010889 A/C gagagaggatatcaaaaattggcagt tgacagtggcaccaaaggtc 0.03 138
M163a AC009977 A/C aggaccctgaaatacagaactg aaatatttcaacatttcacaaaggaa 0.36 186
M167/SRY2627 AC006040 C/T cggaaccactaccagcttca agttaaggccccacgcagt 0.03 113
M170 Rs2032597 A/C cagctcttattaagttatgttttcatattctgtg gtcctcattttacagtgagacacaac 0.07 119
M172 Rs2032604 T/G tgagccctctccatcagaag gccaggtacagagaaagtttgg 0.16 179
M173 Rs2032624 A/C ttttcttacaattcaagggcatttag ctgaaaacaaaacactggcttatca 0.10 81
M175 Rs2032678 5 bp gatttaaactctctgaatcaggcacat ttctactgatacctttgtttctgttcattc 0.02 79
M212a Rs2032664 C/A ccatataaaaacgcagcattctgtt tggagagaacttgagaaaaagtagagaa 0.12 176
M213a Rs2032665 T/C ccatataaaaacgcagcattctgtt tggagagaacttgagaaaaagtagagaa 0.12 176
M224a AC010889 T/C tgcattactccgtatgttcgac tggaagcttaccatctttttatga 0.08 132
SRY10831/SRY1532 Rs2534636 A/G tcatccagtccttagcaaccatta ccacataggtgaaccttgaaaatg 0.06 150
12f2 AC005820 Present/absent cactgactgatcaaaatgcttacagat ggatcccttccttacaccttataca 0.06 90
92R7 Rs2535813 GA/A ttaaatccctcctatttgtgctaacc aatgcatgaacacaaaagacgtaga 0.04 89
P25 Rs150173 C/CA tggaccatcacctgggtaaagt ggcagtataaggttgtcacatcacat 0.01 109
a
SNP markers on the same DNA fragment: (M9 and M163), (M17, M18 and M19), (M32 and M33), (M45 and M157), (M78 and M224),
(M81, M151 and M154), (M139 and M153), (M212 and M213). All primers were redesigned compared to previously published primers.
TBE (1 ) (89 mmol/l Tris base, 89 mmol/l boric acid, The concentrations of the primers in the multiplex reac-
2 mmol/l EDTA, pH 8.3) was used as electrophoresis buffer. tion were adjusted in order to obtain equal amount of each
The gels were stained with 0.5 mg/ml ethidium bromide. The PCR product. The primer concentrations ranged from 0.01 to
10 bp ladder from invitrogen (Groningen, The Netherlands) 0.42 mM(Table 1).
was used to assign the sizes of the fragments. The PCR products were analysed on 11% polyacrylamide
The final setup of the PCR amplification included 1 ng gels as described later (Fig. 1).
DNA in a 50 ml reaction volume containing 1 PCR buffer, In order to eliminate the excess of primers and dNTPs, the
8 mM MgCl2, 400 mM of each dNTP, 0.01 0.42 mM of each PCR products was purified on a MinElute PCR purification
primer, and 2.5 units of AmpliTaq Gold DNA polymerase spin column (Qiagen, Hagen, Germany) following the man-
(AB, Foster City, CA). ufacturer s protocol. The DNAwas eluted in 30 ml of Milli-Q
All DNA amplifications were performed in a GeneAmp water.
9600 thermal cycler (Perkin-Elmer, Wellesley, USA) using E. coli exonuclease I (Exo I) and shrimp alkaline phos-
the following programme: denaturation at 94 8C for 5 min phatase (SAP) was also used to remove primers and unin-
followed by 33 cycles for 30 s at 95 8C, 30 s at 60 8C, and 30 s corporated dNTPs (USB Corporation, Cleveland, USA). Six
at 65 8C, followed by a final extension for 7 min at 65 8C. microliters ExoSAP-IT kit or 5 units of SAP and 2 units of
J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84 77
Fig. 1. Multiplex PCR products of 25 Y chromosome DNA fragments. Ethidium bromide stained polyacrylamide gel with
PCR products obtained from various sources of blood. A negative control with DNA from a female was included. (L) 10 bp ladder
from invitrogene.
Exo I were added to 15 ml of PCR product, mixed, and 50-AACTGACTAAACTAGGTGCCACGTCGTGAAAGT-
incubated at 37 8C for 1 h. The enzymes were inactivated at CTGACAA-30, is a random sequence that did not match
75 8C for 15 min. with any human sequence in the NCBI non-redundant
database [19].
2.4. Design of PCR minisequencing primers For each 4 bp DNA fragment size interval of the detection
primers, two SNP loci were detected. This was done by
Table 2 shows the genotyping primers designed for each selecting two SNP loci with different nucleotide polymorph-
SNP. Primers for detection of deletions and insertions were ism. One SNP could be, e.g. an A/T SNP and the other a C/G
designed with the 30, base corresponding to the last base SNP. Thus, the minisequencing primers for the two SNPs
before the possible deletion or insertion. For each SNP could have the same length and the two polymorphisms
system investigated in the present study, the following base would still be detectable. Primers for minisequencing were
would identify the polymorphism. The sequences of the HPLC purified (DNA-Technology A/S, Aarhus, Denmark
primers were checked for the possibility of primer dimer and Proligo France SAS, Paris, France).
and hairpin formation and investigated in PCR without
template ( self-extension reaction ). In order to distinguish 2.5. Minisequencing reaction and capillary
between the sizes of the detection primers, the primers electrophoresis
were synthesized with lengths between 19 and 106 nucleo-
tides with intervals of four nucleotides for the great major- Multiplex PCR minisequencing was performed in 8 ml
ity of the primers (Table 2). The lengths of the template reactions with 0.2 ml purified PCR product (6 10 ng equiva-
specific parts of the primers ranged from 16 to 29 nucleo- lent to 5 8 fmol of each fragment), 4 ml of SNaPshotTM
tides. The desired length of a primer was adjusted at the reaction mix and 0.01 0.5 mM of the primers (Table 2). The
50 end by addition of a piece of a  neutral sequence thermal cycling was performed with a rapid thermal ramp to
and, if necessary, a poly-C tail. The neutral sequence, 96 8C for 10 s, 50 8C for 5 s, and 60 8C for 30 s for 25 cycles.
78 J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84
Table 2
Minisequencing primer sequences for typing of 35 Y chromosome SNP markers
Locus Poly Neutral Sequence Target specific sequence Orientationa mM Primer
(dC) (50 ! 30) (50 ! 30) size (nt)
M170 None None caacccacactgaaaaaaa Reverse 0.02 19
M45 None caa ctcagaaggagctttttgc Reverse 0.02 22
M139 None aa taatctgacttggaaagggg Forward 0.01 22
M2/sY81 None gacaa ctttatcctccacagatctca Reverse 0.28 26
M46/Tat None None gctctgaaatattaaattaaaacaac Reverse 0.25 26
M167/SRY2627 None tgaaagtctgacaa aagccccacagggtgc Forward 0.35 30
M213 None tgacaa tcagaacttaaaacatctcgttac Reverse 0.02 30
M52 None tctgacaa aatatcaagaaacctatcaaacatcc Reverse 0.02 34
P25 None tcgtgaaagtctgacaa tgcctgaaacctgcctg Forward 0.04 34
M78 None gaaagtctgacaa cttattttgaaatatttggaagggc Reverse 0.02 38
92R7 None gtgaaagtctgacaa catgaacacaaaagacgtagaag Reverse 0.01 38
M89 None cacgtcgtgaaagtctgacaa aactcaggcaaagtgagagat Reverse 0.09 42
M123 None acgtcgtgaaagtctgacaa atttctaggtattcaggcgatg Reverse 0.03 42
M35 None ggtgccacgtcgtgaaagtctgacaa tcggagtctctgcctgtgtc Reverse 0.25 46
M153 None ggtgccacgtcgtgaaagtctgacaa gctcaaagggtatgtgaaca Forward 0.02 46
M40/SRY4064 None aaactaggtgccacgtcgtgaaagtctgacaa tccaccctgtgatccgct Reverse 0.08 50
M154 None gccacgtcgtgaaagtctgacaa gttacatggcctataatattcagtaca Reverse 0.03 50
M32 None taggtgccacgtcgtgaaagtctgacaa agacaagatctgttcagtttatctca Forward 0.50 54
M151 None aggtgccacgtcgtgaaagtctgacaa caatctactacatacctacgctatatg Forward 0.02 54
M17 None actaaactaggtgccacgtcgtgaaagtctgacaa ccaaaattcacttaaaaaaaccc Reverse 0.02 58
M96 None aactgactaaactaggtgccacgtcgtgaaagtctgacaa ggaaaacaggtctctcataata Forward 0.15 62
M172 7 aactgactaaactaggtgccacgtcgtgaaagtctgacaa caaacccattttgatgctt Forward 0.10 66
M173 3 aactgactaaactaggtgccacgtcgtgaaagtctgacaa tacaattcaagggcatttagaac Forward 0.03 66
M19 4 aactgactaaactaggtgccacgtcgtgaaagtctgacaa aaactatttttgtgaagactgttgta Forward 0.10 70
M224 7 aactgactaaactaggtgccacgtcgtgaaagtctgacaa aattgatacacttaacaaagatacttc Forward 0.13 74
SRY10831/SRY1532 10 aactgactaaactaggtgccacgtcgtgaaagtctgacaa ttgtatctgactttttcacacagt Forward 0.03 74
M18 17 aactgactaaactaggtgccacgtcgtgaaagtctgacaa gtttgtggttgctggttgtta Forward 0.05 78
M157 18 aactgactaaactaggtgccacgtcgtgaaagtctgacaa caccaaaggtcatttgtggt Reverse 0.20 78
M81 14 aactgactaaactaggtgccacgtcgtgaaagtctgacaa cttggtttgtgtgagtatactctatgac Reverse 0.03 82
M163 25 aactgactaaactaggtgccacgtcgtgaaagtctgacaa cacaaaggaattttttttgag Reverse 0.51 86
M212 20 aactgactaaactaggtgccacgtcgtgaaagtctgacaa gcattctgttaatataaaacacaaaa Forward 0.20 86
M9 22 aactgactaaactaggtgccacgtcgtgaaagtctgacaa catgtctaaattaaagaaaaataaagag Reverse 0.40 90
12f2 29 aactgactaaactaggtgccacgtcgtgaaagtctgacaa aacatgtaagtctttaatccatctc Forward 0.02 94
M33 29 aactgactaaactaggtgccacgtcgtgaaagtctgacaa cagttacaaaagtataatatgtctgagat Reverse 0.18 98
M175 46 aactgactaaactaggtgccacgtcgtgaaagtctgacaa cacatgccttctcacttctc Forward 0.28 106
a
The detection orientation has been probed relative to the YCC information reported in [32].
A positive control (provided with the kit) and negative injections. GeneScan-120 LIZTM was used as internal size
control (sterile water or PCR product from a female), was standard. The data were analysed using GeneScan Analysis
performed for each batch of 44 samples. software v. 3.7 (Applied Biosystems). After background
The homogeneity of each primer was checked in single- substraction and colour separation, peaks were sorted
plex minisequencing. The occurrence of extra peaks one or into bins according to sizes by comparison to the internal
more nucleotides smaller than the expected size indicated size standard. Peaks above 400 relative fluorescence units
heterogeneity of the minisequencing primer. were considered positive signals and a SNP type was
After the minisequencing reaction, 1 Unit of SAP was assigned.
added and the tube was incubated at 37 8C for 1 h in order to
remove the 50 phosphoryl groups of the unincorporated 2.6. Reproducibility studies
[F]ddNTPs. SAP was inactivated by incubation at 75 8C
for 15 min. DNA samples from 194 unrelated male Danes were typed
One ml of the purified minisequencing PCR product was twice with the minisequencing technique and assigned SNP
analysed on an AB Prism 3100 Genetic Analyser with a types for the 35 SNP systems. The assignments of SNP types
36 cm capillary array, POP-4 polymer and 10 s at 3000 V of the duplicate testing were compared.
J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84 79
2.7. Statistical methods with  GC at the 30 end with successful results. Twenty-one of
the 25 primer pairs worked satisfactorily at the first design.
Gene diversities and standard errors were calculated
according to the methods of Nei [23]. 3.3. PCR buffer and efficiency of multiplex PCR
amplification
3. Results
We found that the best results of amplification of all 25
3.1. DNA purification methods DNA targets were obtained by increasing the concentration
to 8 mM MgCl2. Higher concentrations inhibited the ampli-
DNA purified with Qiagen columns and DNA from FTA1 fication (data not shown).
paper with bloodstains in all cases gave satisfactory results
(Fig. 1). DNA from buccal cells on FTA1 paper gave 3.4. Quality of DNA primers for template PCR
variable intensities of the results of samples. amplification
3.2. Design of primers Unpurified primers could be combined into multiplexes
up to seven systems while HPLC purified primers could be
When no band or only a very weak band was observed, combined to amplify at least 25 templates in one reaction.
suggesting that the affinities of the primers were suboptimal,
the primers were redesigned. In one case, the PCR amplifica- 3.5. Titration of primer concentrations in PCR
tion was very weak and four different sets of primers were amplification
tried before an acceptable yield was obtained. It was not
possible to understand the reason since the primer set best It was necessary to titrate primer concentrations to obtain
suited from a theoretical point gave the lowest yield. In three a balanced PCR multimix for minisequencing. The final
cases with unsatisfactory yields, the primers were redesigned concentrations of primers ranged from 0.11 to 0.46 mM.
Fig. 2. Electropherogramme with 35 Y chromosome SNP profiles from a male donor. GeneScan analysis of SNaPshotTM minisequencing of
the Y chromosome SNP multiplex.
80 J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84
25 3.10. Annealing temperature of minisequencing
primers
20
Annealing temperatures from 50 to 60 8C gave almost the
same overall results in the 35 SNP multiplex when judged by
inspection of the peak areas.
15
3.11. Y chromosome SNP typing results
10
Fig. 2 shows a representative electropherogramme of
typing of 35 Y chromosome SNPs in an individual. In
5
one of the 194 males, typing could no reaction was obtained
in M81. The same lack of reaction in M81 was found in the
son of the investigated man. The remaining 34 Y chromo-
0
50 100 250 500 1000 2000 4000 8000
some SNPs were detected in the man and his child. All other
male samples gave a full 35-Y-SNP profile.
DNA (pg)
Fig. 3. Sensitivity of the 35 Y chromosome SNP typing assay.
3.12. Reproducibility of Y chromosome SNP typing
For each DNA concentration, the relative fluorescence units
with minisequencing
(RFUs from GeneScan) of investigations of four SNPs detected
with each of the four dyes: blue, green, yellow and red were
SNP typing was performed twice in all 194 male Danes
collated from typing of two individuals. For each DNA
and the duplicate types were consistent. In each minisequen-
concentration, the median RFU value of the two individuals
cing experiments, at least one sample with known types for
was calculated for each dye, and for each concentration the
all 35 SNPs was included, and concordant assignments of
median RFUs were normalized as a percentage of the total RFUs
SNP types were obtained in all cases.
of all the RFUs for the dye in question. Finally, for each DNA
concentration, the median of the normalized RFU values for all Four samples were typed for the 11 SNPs SRY2627,
four dyes was calculated as a percentage of the sum of all
M213, M35, M153, SRY4064, M17, M18, M9, SRY10831,
normalized median RFU values of all concentrations. Thus, the
92R7, and P25 as part of an interlaboratory exercise of the
sum of RFUs in the figure sum up to 100%. The error bars
European DNA Profiling Group, and correct results were
indicate the standard error of the mean (S.E.M.).
obtained.
3.6. Sensitivity of the target multiplex PCR amplification
In our hands, the lower limit for reproducible results was
approximately 100 pg DNA with a range up to approxi-
mately 10 ng and an optimum at 1 2ng (Figs. 2 and 3).
3.7. Purification of the PCR template amplification
product
Both spin column and enzymatic purified PCR amplifica-
tion products gave satisfactory minisequencing typing reac-
tions. The recovery with the Exo I-SAP was 100% while the
column purification had a recovery of approximately 80%
(data not shown).
3.8. Design of DNA primers for minisequencing
None of the 35 detection primers had to be redesigned.
Fig. 4. Effect of excess DNA from females on the 35 Y
chromosome SNP typing assay. The relative fluorescence units
3.9. Quality of DNA primers for minisequencing
(RFUs from GeneScan) of mixtures of male DNA and female
DNA in great excess. The RFUs were calculated as indicated in
Clear, homogeneous peaks were obtained only if the
Fig. 3. In general, there was a dose response relation between the
purity of the primers was higher than approximately 90%.
concentration of male DNA and the RFU signal strength, while
If the purity was less, the signal from degenerated primers
female DNA had practically no influence on the RFU signal in
(n 1, n 2, etc.) would decrease the discrimination. the concentration range investigated.
Relative Fluorescence Units
(%)
J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84 81
Table 3
Frequencies of 35 Y chromosome SNP markers in male Danes
Locus Fragment numbera Polymorphismb Frequency (number) Frequency (%)
M2/sY81 1 A/G 194/0 100.0/0.0
M9 2 C/G 85/109 43.8/56.2
M17 3 4G/3G 162/32 83.5/16.5
M18 3 No ins./2 bp ins. 194/0 100.0/0.0
M19 3 T/A 194/0 100.0/0.0
M32 4 T/C 194/0 100.0/0.0
M33 4 A/C 194/0 100.0/0.0
M35 5 G/C 190/4 97.9/2.1
M40/SRY4064 6 G/A 190/4 97.9/2.1
M45 7 G/A 86/108 44.3/55.7
M46/Tat 8 T/C 193/1 99.5/0.5
M52 9 A/C 194/0 100.0/0.0
M78 10 C/T 192/2 99.0/1.0
M81c 11 C/T 193/0 100.0/0.0
M89 12 C/T 4/190 2.1/97.9
M96 13 G/C 190/4 97.9/2.1
M123 14 G/A 193/1 99.5/0.5
M139 15 5G/4G 0/194 0.0/100.0
M151 11 G/A 194/0 100.0/0.0
M153 15 T/A 194/0 100.0/0.0
M154 11 T/C 194/0 100.0/0.0
M157 7 A/C 194/0 100.0/0.0
M163 2 A/C 194/0 100.0/0.0
M167/SRY2627 16 C/T 194/0 100.0/0.0
M170 17 A/C 119/75 61.3/38.7
M172 18 T/G 189/5 97.4/2.6
M173 19 A/C 89/105 45.9/54.1
M175 20 No del./5 bp del. 194/0 100.0/0.0
M212 21 C/A 194/0 100.0/0.0
M213 21 T/C 4/190 2.1/97.9
M224 10 T/C 194/0 100.0/0.0
SRY10831/SRY1532 22 A/G 32/162 16.5/83.5
12f2 23 Present/absent 189/5 97.4/2.6
92R7 24 GA/Ad 86/108 44.3/55.7
P25 25 C/CAd 124/70 63.9/36.1
a
Some PCR products contain more than one SNP in the same fragment.
b
Following the Y chromosome consortium nomenclature system [32].
c
One male gave no reaction in minisequencing of M81.
d
Two signals were detected in some individuals [24,33].
3.13. Male female mixtures of DNA signals in theses systems was investigated with STR-tech-
nique. Only one STR-profile was obtained in each individual
Female DNA did not influence the results of Y chromo- demonstrating that contamination of DNA was not the
some SNP typing when added in concentrations more than reason for the two signals in P25 and 92R7.
300 times the concentrations of male DNA (Fig. 4).
3.14. Y chromosome SNP population data in Danes 4. Discussion
Table 3 shows the frequency distribution of the 35 SNPs We have developed a PCR multiplex-based system for
investigated in 194 male Danes. No SNP signal was obtained typing of a large number of SNPs using Y chromosome
in 15 female Danes. A total of 19 SNPs showed variation SNPs as an example. An important part of the work was to
while 16 SNPs were monomorphic in the male Danes explore the various aspects of the multiplex PCR methods.
studied. The 35 Y chromosome SNPs presented here are not our final
Two signals were obtained in P25 and 92R7 in some set of Y chromosome SNPs for population studies or forensic
individuals (cf discussion). DNA from individuals with two genetic applications.
82 J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84
Successful PCR multiplexing depends on a number of approximately 103 times more than the concentration of
factors. Below, we present some of our considerations the template.
concerning the selection of the SNPs and the generation We stored dNTPs in small aliquots at 20 8C for up to 8
of the multiplex PCRs for amplification and minisequen- months. However, we observed that dNTPs were sensitive to
cing. repeated freezing and thawing. As a rule of thumb, the
At an early stage, it was decided to use the multicolour multiplex PCR would fail if the dNTPs have been frozen
fluorescence electrophoresis technique combined with PCR and thawed more than four times. The amount of time in
multiplexing at approximately 60 8C in high concentrations freezer was less important as it has been reported by others
of MgCl2. The spacing between minisequencing primers authors [28].
was decided to be four nucleotides because we wanted to The enzymatic purification method is obviously easy, has
obtain reliable separation in the electrophoresis. an almost 100% recovery and a very limited risk of con-
We attempted to avoid SNPs situated in regions reported tamination.
to be replicated. Two exceptions were the P25 and 92R7 We chose to adjust the length of the minisequencing
SNPs that are situated in a region that most probably is part primers by means of (1) a part of a neutral sequence of
of a duplication [24]. Both SNPs seem to discriminate up to 40 nt and for the longer primers (2) an additional poly-
between European and other populations [25]. C part. The neutral sequence was selected in order to obtain a
Multiplex PCR amplification primers between 19 and 34 more balanced base composition. We chose poly-C for the
bases pairs long were selected because it was anticipated that tail because, in theory, poly-G would give a higher molecular
such long primers would work well under multiplex condi- mass, poly-A would have a risk of depurination during
tions [26]. synthesis, and poly-T tails may interfere with the addition
Qiagen purified DNA from blood samples and blood of 30 ddA in the minisequencing reaction (SNaPshotTM
stains on FTA1 paper worked equally well in the assay. protocol recommendation, Applied Biosystems).
Chelex treated blood samples worked as well (data not The quality of minisequencing primers is important
shown). Optimal multiplex SNP typing results were obtained because primer batches with heterogeneous primer
with 1 ng DNA (range 0.1 20 ng DNA). Thus, quantifica- sequences consisting of the intended DNA sequence of
tion of DNA is not mandatory for the SNP assay. It should,  n nucleotides plus a spectrum of shorter nucleotides
however, be noticed that the balance of the amounts of (n 1, n 2, n 3, etc.) in many cases will destroy the
amplification products of the DNA fragments is changed minisequencing reaction. In addition, we observed amplifi-
with increasing amounts of templates. With increasing cation failure due to a heterogeneous primer batch in the
concentrations of PCR amplified fragments, small, fluores- PCR multiplex with seven systems even though each of the
cent adenosinnucleotide peaks with sizes of PCR amplified seven works in singleplex reactions. Therefore, we recom-
fragments plus one nucleotide were seen, most likely do to mend that each primer batch is tested before the multiplex
non-template addition of a single adenosin molecules to the PCR and subsequent analyses, e.g. by minisequencing or
30 end of some PCR amplified fragments. At low amounts of mass spectrometry. Purification of the primers with e.g.
template DNA, loss of signal will occur due to stochastic HPLC or gel purification techniques can to some extent
phenomena [27]. solve these problems.
Commonly used PCR buffers include only KCl, Tris and The minisequencing system was rather insensitive to the
MgCl2. It has been reported that many primer pairs produ- annealing temperature. It was necessary to adjust primer
cing short amplification products (<200 bp) work better at concentrations from 0.01 to 0.50 mM in the minisequencing
higher salt concentration (KCl) in multiplex systems [26]. multimix.
Increasing the concentration of KCl in the PCR buffer 1.6 The longer extension products had electrophoretic mobi-
and 2-fold in our 35-plex did not increase the yield of PCR lities corresponding to those predicted by the number of
product significantly and had no effect on the synthesis of bases. The mobility of shorter extension products with the
fragments >150 bp. Increase of MgCl2 concentration from 2 same number of bases varied to some extent. This is most
to 8 mM increased the yield of amplicons; higher MgCl2 probably due to the fact that differences in the masses of the
concentration inhibited the amplification (data not shown). various fluorochromes used and in the exact composition of
We used AmpliTaq Gold DNA polymerase (Applied purines and pyrimidines have a relatively high influence on
Biosystems) because this enzyme minimizes primer dimer the mobility of short DNA molecules.
formation. Even with a 4-base 30 overlap between two The SNP-typing results were highly reproducible. A total
primers we obtained homogeneous PCR products (data of 194 males were SNP typed in duplicate and no discre-
not shown). The most efficient enzyme concentration pancies were observed. Furthermore, five of the most poly-
seemed to be around 2.5 U/50 ml reaction volume. morphic SNPs were analysed by a DNA hybridisation assay
In our hands, primer concentrations below 0.01 mM were using the Nanogen technology [29]. Concordant results were
insufficient and concentrations above 0.5 mM seemed to obtained for all 194 individuals (data not shown).
inhibit multiplex PCRS probably by inducing dimer dimer In one father-child combination, no allele of M81 was
formation. Primer concentrations were adjusted to be detectable. An amplified fragment was present in the first
J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84 83
C.M. Rice, Z. Ning, J. Rogers, D.R. Bentley, P.Y. Kwok, E.R.
PCR because two other SNPs (M151 and M154) on the
Mardis, R.T. Yeh, B. Schultz, L. Cook, R. Davenport, M.
fragment were detected, but no reaction of M81 was detected
Dante, L. Fulton, L. Hillier, R.H. Waterston, J.D. McPher-
in the minisequencing reaction. Work is in progress in order
son, B. Gilman, S. Schaffner, W.J. Van Etten, D. Reich, J.
to determine the nature of the variant.
Higgins, M.J. Daly, B. Blumenstiel, J. Baldwin, N. Stange-
A total of 19 of 29 SNPs reported to be polymorphic in
Thomann, M.C. Zody, L. Linton, E.S. Lander, D. Altshuler,
Europeans in a previous study [4] and 9 of 10 SNPs reported
A map of human genome sequence variation containing 1.42
in another study [21] turned out to be polymorphic in the
million single nucleotide polymorphisms, Nature 409 (2001)
male Danes studied. The gene diversity for the loci showing
928 933.
polymorphism ranged from 0.01 to 0.5 (Table 3). M173,
[2] S. Paracchini, B. Arredi, R. Chalk, C. Tyler-Smith,
M45, 92R7 and M9 were the most polymorphic markers in Hierarchical high-throughput SNP genotyping of the human
Y chromosome using MALDI-TOF mass spectrometry,
Danes. The data were described as frequencies of individual
Nucleic Acids Res. 30 (2002) e27.
SNPs and not as Y chromosome haplogroups because the
[3] M. Raitio, K. Lindroos, M. Laukkanen, T. Pastinen, P.
study was a technical study and the Y chromosome multi-
Sistonen, A. Sajantila, A.C. Syvanen, Y-chromosomal SNPs
plex is not ideal for typing of Y chromosome haplogroups. A
in Finno-Ugric-speaking populations analyzed by minise-
larger study of Y chromosome haplogroups in Danes and
quencing on microarrays, Genome Res. 11 (2001) 471 482.
other populations will be published elsewhere.
[4] P.A. Underhill, G. Passarino, A.A. Lin, P. Shen, M. Mirazon
P25 and 92R7 were previously reported as SNPs [30,31].
Lahr, R.A. Foley, P.J. Oefner, L.L. Cavalli-Sforza, The
However, the P25 and 92R7 minisequencing primers were
phylogeography of Y chromosome binary haplotypes and
extended with two different dideoxynucleotides during the
the origins of modern human populations, Ann. Hum. Genet.
minisequencing reaction of numerous samples. This indi- 65 (2001) 43 62.
[5] M.A. Jobling, C. Tyler-Smith, Fathers and sons: the Y
cates that at least two different, almost identical fragments
chromosome and human evolution, Trends Genet. 11 (1995)
were amplified during the PCR reaction. Hurles et al. [33]
449 456.
previously observed that SNP typing of 92R7 gave two
[6] H. Oota, W. Settheetham-Ishida, D. Tiwawech, T. Ishida, M.
results in some individuals. Further studies have confirmed
Stoneking, Human mtDNA and Y chromosome variation is
that P25 and 92R7 are paralogous sequence variants and that
correlated with matrilocal versus patrilocal residence, Nat.
at least one of the sequence variants in each group of loci is
Genet. 29 (2001) 20 21.
polymorphic [24].
[7] M. Seielstad, Asymmetries in the maternal and paternal
The multiplex PCR SNP typing format presented here
genetic histories of Colombian populations, Am. J. Hum.
seems to be useful for forensic casework because small
Genet. 67 (2000) 1062 1066.
amounts of DNA (100 pg DNA) can be reliably typed. [8] G. Passarino, G.L. Cavalleri, A.A. Lin, L.L. Cavalli-Sforza,
A.L. Borresen-Dale, P.A. Underhill, Different genetic com-
The multiplex presented is not our final package for Y
ponents in the Norwegian population revealed by the analysis
chromosome SNPs for forensic purposes. The way forward
of mtDNA and Y chromosome polymorphisms, Eur. J. Hum.
would go either through (1) the development of SNP
Genet. 10 (2002) 521 529.
packages optimised for an initial screening plus further
[9] P. de Knijff, Messages through bottlenecks: on the combined
packages optimised for the major populations or (2) the
use of slow and fast evolving polymorphic markers on
development of a large multiplex package that include Y
the human Y chromosome, Am. J. Hum. Genet. 67 (2000)
chromosome SNPs that can discriminate between individual
1055 1061.
lineages in all populations.
[10] M.A. Jobling, Y-chromosomal SNP haplotype diversity in
forensic analysis, Forensic Sci. Int. 118 (2001) 158 162.
[11] J.L. Mountain, A. Knight, M. Jobin, C. Gignoux, A. Miller,
A.A. Lin, P.A. Underhill, SNPSTRs: empirically derived,
Acknowledgements
rapidly typed, autosomal haplotypes for inference of popula-
tion history and mutational processes, Genome Res. 12 (2002)
We thank Dr. Rebecca Reynolds, Roche Molecular Sys-
1766 1772.
tems, for advice concerning the design of the multiplex PCR
[12] M.A. Jobling, E. Heyer, P. Dieltjes, P. de Knijff, Y
for template generation in the initial phase of the project. We
chromosome-specific microsatellite mutation rates re-exam-
thank Ms. AnneMette Holbo Birk for technical assistance.
ined using a minisatellite, MSY1, Hum. Mol. Genet. 8 (1999)
The work was supported by grants to Juan Sanchez from
2117 2120.
Ellen and Aage Andersen s Foundation and Manuel Morales
[13] M. Brion, R. Cao, A. Salas, M.V. Lareu, A. Carracedo, New
Foundation.
method to measure minisatellite variant repeat variation
in population genetic studies, Am. J. Hum. Biol. 14 (2002)
421 428.
References
[14] P. Gill, An assessment of the utility of single nucleotide
polymorphisms (SNPs) for forensic purposes, Int. J. Legal
[1] R. Sachidanandam, D. Weissman, S.C. Schmidt, J.M. Kakol, Med. 114 (2001) 204 210.
L.D. Stein, G. Marth, S. Sherry, J.C. Mullikin, B.J. [15] D.E. Reich, S.F. Schaffner, M.J. Daly, G. McVean, J.C.
Mortimore, D.L. Willey, S.E. Hunt, C.G. Cole, P.C. Coggill, Mullikin, J.M. Higgins, D.J. Richter, E.S. Lander, D.
84 J.J. Sanchez et al. / Forensic Science International 137 (2003) 74 84
Altshuler, Human genome sequence variation and the [24] C. BÅ‚sting, J.J. Sanchez, N. Morling, The two Y chromosome
influence of gene history, mutation and recombination, Nat. loci, P25 and 92R7, are polymorphic paralogous sequence
Genet. 32 (2002) 135 142. variants, Forensic Sci. Int. (2003), submitted for publication.
[16] R. Thomson, J.K. Pritchard, P. Shen, P.J. Oefner, M.W. [25] M.F. Hammer, T.M. Karafet, A.J. Redd, H. Jarjanazi, S.
Feldman, Recent common ancestry of human Y chromo- Santachiara-Benerecetti, H. Soodyall, S.L. Zegura, Hierarch-
somes: evidence from DNA sequence data, Proc. Natl. Acad. ical patterns of global human Y chromosome diversity, Mol.
Sci. U.S.A. 97 (2000) 7360 7365. Biol. Evol. 18 (2001) 1189 1203.
[17] M.A. Jobling, A. Pandya, C. Tyler-Smith, The Y chromo- [26] O. Henegariu, N.A. Heerema, S.R. Dlouhy, G.H. Vance, P.H.
some in forensic analysis and paternity testing, Int. J. Legal Vogt, Multiplex PCR: critical parameters and step-by-step
Med. 110 (1997) 118 124. protocol, Biotechniques 23 (1997) 504 511.
[18] N.M. Makridakis, J.K. Reichardt, Multiplex automated [27] B.E. Krenke, A. Tereba, S.J. Anderson, F. Buel, S. Culhane, C.J.
primer extension analysis: simultaneous genotyping of several Finis, C.S. Tomsey, J.M. Zachetti, A. Masibay, D.R. Rabbach,
polymorphisms, Biotechniques 31 (2001) 1374 1380. E.A. Amiott, C.J. Sprecher, Validation of a 16-locus fluorescent
[19] K. Lindblad-Toh, E. Winchester, M.J. Daly, D.G. Wang, J.N. multiplex system, J. Forensic Sci. 47 (2002) 773 785.
Hirschhorn, J.P. Laviolette, K. Ardlie, D.E. Reich, E. [28] P. Markoulatos, N. Siafakas, M. Moncany, Multiplex
Robinson, P. Sklar, N. Shah, D. Thomas, J.B. Fan, T. polymerase chain reaction: a practical approach, J. Clin.
Gingeras, J. Warrington, N. Patil, T.J. Hudson, E.S. Lander, Lab. Anal. 16 (2002) 47 51.
Large-scale discovery and genotyping of single nucleotide [29] P.N. Gilles, D.J. Wu, C.B. Foster, P.J. Dillon, S.J. Chanock,
polymorphisms in the mouse, Nat. Genet. 24 (2000) 381 386. Single nucleotide polymorphic discrimination by an electro-
[20] R. Reynolds, K. Walker, L. Steiner, SNP genotyping using nic dot blot assay on semiconductor microchips, Nat. Biotech.
megaplex PCR amplification and linear probe assays, in: 17 (1999) 365 370.
Proceedings of the 19th International Congress ISFG, [30] M.F. Hammer, A.J. Redd, E.T. Wood, M.R. Bonner, H.
Münster, Germany (P. Gill), FSS, 2002, WS2. Jarjanazi, T. Karafet, S. Santachiara-Benerecetti, A. Oppen-
[21] O. Semino, G. Passarino, P.J. Oefner, A.A. Lin, S. Arbuzova, heim, M.A. Jobling, T. Jenkins, H. Ostrer, B. Bonné-Tamir,
L.E. Beckman, G. De Benedictis, P. Francalacci, A. Kouvatsi, Jewish and middel eastern non-jewish populations share a
S. Limborska, M. Marcikiae, A. Mika, B. Mika, D. Primorac, common pool of Y-chromosome biallelic haplotypes, Proc.
A.S. Santachiara-Benerecetti, L.L. Cavalli-Sforza, P.A. Un- Natl. Acad. Sci. U.S.A. 97 (2000) 6769 6774.
derhill, The genetic legacy of Paleolithic Homo sapiens [31] N. Mathias, M. Bayés, C. Tyler-Smith, Highly informative
sapiens in extant Europeans: a Y chromosome perspective, compound haplotypes for the human Y chromosome, Hum.
Science 290 (2000) 1155 1159. Mol. Genet. 3 (1994) 115 123.
[22] P. Malaspina, F. Cruciani, B.M. Ciminelli, L. Terrenato, P. [32] The Y Chromosome Consortium, A nomenclature system for
Santolamazza, A. Alonso, J. Banyko, R. Brdicka, O. Garcia, the tree of human Y chromosomal binary haplogroups,
C. Gaudiano, G. Guanti, K.K. Kidd, J. Lavinha, M. Avila, P. Genome Res. 12 (2002) 339 3480.
Mandich, P. Moral, R. Qamar, S.Q. Mehdi, A. Ragusa, G. [33] M.E. Hurles, R. Veitia, F. Arroyo, M. Armenteros, J.
Stefanescu, M. Caraghin, C. Tyler-Smith, R. Scozzari, A. Bertranpetit, A. Perez-Lezaun, E. Bosch, M. Shlumukova,
Novelletto, Network analyses of Y-chromosomal types in A. Cambon-Thomsen, K. McElreavey, A. Lopez De Munain,
Europe, northern Africa, and western Asia reveal specific A. Rohl, I.J. Wilson, L. Singh, A. Pandya, F.R. Santos, C.
patterns of geographic distribution, Am. J. Hum. Genet. 63 Tyler-Smith, M.A. Jobling, Recent male-mediated gene flow
(1998) 847 860. over a linguistic barrier in Iberia, suggested by analysis of a
[23] M. Nei, Molecular Evolutionary Genetics, Columbia Uni- Y-chromosomal DNA polymorphism, Am. J. Hum. Genet. 65
versity Press, New York, 1982, p. 512. (1999) 1437 1448.