Molecular Ecology (1999)

8

, 1749–1752

© 1999 Blackwell Science Ltd

Blackwell Science, Ltd

TECHNICAL NOTES

17491750Graphicraft Limited, Hong Kong

Noninvasive methods for collecting fresh hair tissue

X . VALDERRAMA,* W. B . KARESH,† D. E . WILDMAN‡ and D. J . MELNICK*§

*

Department of Anthropology, Columbia University, New York, New York 10027, USA,

†

Wildlife Conservation Society, 185th St. & Southern Blvd, Bronx, New York 10460, USA,

‡

Department of Anthropology, New York University, New York, New York 10003, USA,

§

Center for Environmental Research and Conservation, Columbia University, New York, New York 10027, USA

Keywords

: hair tissue, microsatellite DNA, mitochondrial DNA

Received 12 March 1999; revision accepted 18 May 1999

Correspondence: D. E. Wildman. Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA. Fax: +01-212-995-4014; E-mail: dew7676@is.nyu.edu

Collecting genetic material noninvasively from mammalsincludes plucking fresh hair (Higuchi

et al

. 1988), collectingshed hair (Morin

et al

. 1993), and collecting fecal material(Hoss

et al

. 1992).The bulb at the end of a fresh hair provides high-quality

DNA for amplification. Although shed hair is more access-ible, it contains only a fraction of DNA isolated from freshhair (Gagneux

et al

. 1997). Reliably amplifying DNA fromfecal material can be difficult, because of DNA degradation(Golenberg

et al

. 1996), DNA concentration (Taberlet

et al

.1996), or PCR-inhibiting compounds in the DNA extract(Constable

et al

. 1995). Here, we describe four novel, noninvas-ive methods of collecting fresh hairs from wild and captivemammals. In most cases, these methods can be used to targetspecific individuals, an advantage that is often absent whencollecting shed hair.

Method one involves shooting a rolled strip of duct tape,pressed onto the flat tip of a plastic syringe, from an air-powered dart pistol (Telinject USA, Inc.). The pistol is fittedwith a barrel of variable length. Optimal shooting distancewas

<

15

m from subjects. Pressure from a CO

2

cartridge canbe controlled so that the syringe bounces harmlessly offthe subject while the tape pulls fresh hairs. The syringemay be filled partially with water as a balance. Thismethod was developed for sampling capuchin monkeys,

Cebus olivaceus

, most of which are too small to dart safely ortrap efficiently.

The second sampling method involves making a corral byenclosing a small area with duct tape. When used on baboons,

Papio hamadryas

, this area was approximately 1

m

2

. Hori-zontal rails of tape are spaced every 30

cm for a maximumheight of 1.5

m, and bait is placed inside the corral. Typi-cally, several animals approach the corral and squeeze theirbodies between tape rails to reach the bait, and thereby leavehairs. This method was designed to obtain hairs simultane-ously from several baboons.

Method three requires that target animals handle bait thathas been wrapped with duct tape. Examples of this methodinclude attaching bait to a tree limb by wrapping tape aroundit, making tape-covered food baskets, and wrapping fooddirectly with inverted tape. Bait wraps were used success-fully on capuchins and baboons of varying ages.

Finally, captive animals may be sampled by wrapping invertedtape around the tip of a stick, which is then inserted into thecage to touch an animal directly so that the tape pulls hairs.Successful hair sampling with this method was carried out withcarnivores (

Hyaena hyaena

,

Panthera leo

,

Felis caracal

), and baboons.The number of hairs collected varies across different methods,

as presented in Table 1. Nevertheless, one to three hair bulbsyield sufficient DNA for amplification of nuclear and mito-chondrial loci. DNA was isolated from one to four hairs accord-ing to Higuchi

et al

. (1988) or with QIAamp tissue kits (QIAGEN,Inc.). It is critical to use sufficient DNA in each PCR reactionto avoid autosomal genotyping errors resulting from disparateallelic proportions in the DNA extract (Taberlet

et al

. 1996).Further, the annealing temperature should be sufficiently lowto enhance detection of alleles with mutations in the primingsites (Pemberton

et al

. 1995). Incorporating these precautionsyielded consistent PCR products (Fig. 1a,b). Additionally, wefound that an unexpected advantage of using nuclear DNAfrom hair was that it generally produced fewer artefacts duringamplification than did DNA isolated from other tissues.

In summary, fresh hair tissue may be collected using non-invasive methods and without the need for restraint ortrauma to animals. Advantages for researchers using freshhairs include the ability to target specific individuals andobtain high-quality DNA. The combination of the safety andsuccess of these noninvasive methods supports their applica-tions in genetic studies across a wide variety of mammals,including small, arboreal, or endangered taxa.

The authors are grateful to the American Institute forYemeni Studies, Cecilia and Tomás Blohm, Todd Disotell,and Clifford Jolly.

References

Constable JJ, Packer C, Collins DA, Pusey AE (1995) NuclearDNA from primate dung.

Nature

,

373

, 393.Gagneux P, Boesch C, Woodruff DS (1997) Microsatellite scor-

ing errors associated with noninvasive genotyping based onnuclear DNA amplified from shed hair.

Molecular Ecology

,

6

,861–868.

Table 1 Comparative success of methods for collecting freshhair tissue

Sampling methodTypical and (maximum) number of hair bulbs collected per sample

Shooting tape 5–8 (89)Bait wrap 1–2 (6)Corral 1–4 (9)Sticky stick 100 (150)

MEC738.fm Page 1749 Wednesday, September 29, 1999 10:18 AM

1750

TE C H N I C A L N O T E S

© 1999 Blackwell Science Ltd,

Molecular Ecology

, 8, 1749–1752

Golenberg EM, Bickel A, Weihs P (1996) Effect of highly frag-mented DNA on PCR.

Nucleic Acids Research

,

24

, 5026–5033.Hayasaka K, Fujii K, Horai S (1996) Molecular phylogeny of

macaques: implications of nucleotide sequences from an 896-base pair region of mitochondrial DNA.

Molecular Biology andEvolution

,

13

, 1044–1053.Higuchi R, von Beroldingen CH, Sensabaugh GF, Erlich HA

(1988) DNA typing from single hairs.

Nature

,

332

, 543–546.Hoss M, Kohn M, Pääbo S, Knauer F, Schröder W (1992) Excre-

ment analysis by PCR.

Nature

,

359

, 199.Morin PA, Wallis J, Moore JJ, Chakraborty R, Woodruff DS (1993)

Non-invasive sampling and DNA amplification for paternityexclusion, community structure, and phylogeography in wildchimpanzees.

Primates

,

34

, 347–356.Pemberton JM, Slate J, Bancroft DR, Barrett JA (1995) Nonampli-

fying alleles at microsatellite loci: a caution for parentage andpopulation studies.

Molecular Ecology

,

4

, 249–252.Taberlet P, Griffin S, Goossens B

et al

. (1996) Reliable genotypingof samples with very low DNA quantities using PCR.

NucleicAcids Research

,

24

, 3189–3194.

8101999770

TE C H N I C A L N O T E STE C H N I C A L N O T E S

17501752Graphicraft Limited, Hong Kong

The accuracy of heterozygous base calling from diploid sequence and resolution of haplotypes using allele-specific sequencing

MATTHEW P. HARE and STEPHEN R. PALUMBI

Organismic and Evolutionary Biology, Harvard University, 16 Divinity Avenue, Cambridge, Massachusetts, USA 02138

Keywords

:

cis–trans

phase, PCR, nuclear gene tree, phylogeny

Received 7 March 1999; revision received 27 April 1999; accepted 15 June 1999

Correspondence: M. P. Hare. Fax: +1-617-495-1958; E-mail: mhare@oeb.harvard.edu

For a diploid locus, a rapid way to identify variable DNAsequence positions and estimate their population frequenciesis to directly sequence diploid polymerase chain reaction(PCR) products and use doublets in the sequence data todetect heterozygous sites. Here we provide one of the firstpublished measures of accuracy for this increasingly commonprocedure. For genealogical analyses of these data, hetero-zygous diploid sequences must be further resolved into pairsof haplotypes that reveal the relative orientation (

cis–trans

phase) of multiple heterozygous polymorphisms. We describehere the rapid determination of 2000 bp haplotypes by aligningdiploid sequences, designing allele-specific primers at hetero-zygous sites and directly sequencing the original diploidtemplate with those primers.

Effective diploid sequencing requires faithful amplifica-tion of both alleles in heterozygotes and accurate identi-fication of heterozygous sites. We determined the frequencyof null PCR alleles among presumed ‘diploid’ sequencingtemplates and tested the accuracy of heterozygous base callsin diploid fluorescent cycle-sequence (Applied BiosystemsFS dye-terminators). From a sample of five individuals ineach of two closely related dolphin species,

Lagenorhynchusobliquidens

and

L. obscurus

, we sequenced 2000 bp introns

Fig. 1 (a) Nuclear tetranucleotide locus D14S302 (human MapPairs™, Research Genetics, Inc.) amplified from DNA of four distinctcapuchin individuals. Lanes: 1, molecular weight marker øX174 DNA digested with HinfI; 2, template DNA isolated from blood;3–5, template DNA isolated from hairs. Samples were run on a 6%, 8.0 cm polyacrylamide gel and stained in SYBR Green (MolecularProbes Inc.). (b) Partial sequence from an 896-bp region of mitochondrial DNA (Hayasaka et al. 1996). Template DNA was isolated frombaboon hairs. The DNA was sequenced on an ABI Prism™ 310 genetic analyser using a dRhodamine terminator cycle sequencing readyreaction kit.

MEC738.fm Page 1750 Wednesday, September 29, 1999 10:18 AM

TE C H N I C A L N O T E S

1751

© 1999 Blackwell Science Ltd,

Molecular Ecology

, 8, 1749–1752

in three nuclear genes. In the total data set there were 37variable positions. Seventeen (46%) of these positions werewithin a restriction site, allowing confirmation of diploidPCR and sequence heterozygosity using restriction fragmentlength polymorphism (RFLP) analysis. Restriction digestionswere applied to independent PCR products amplified fromgenomic DNA using internal primers (Table 1).

RFLPs showed that amplification of complete intronsproduced diploid product in 28 out of 30 locus-by-individualreactions, but amplified only one allele from a heterozygotein two cases. Evidence for these PCR null alleles underscoresthe importance of confirming diploid sequencing results usingRFLPs in alternately primed PCR products.

Restriction assays also demonstrated that out of 161 con-firmable chromatogram base calls (excluding individuals

with ‘null’ alleles), 154 were correct with respect to homo- orheterozygosity (Table 1). All seven errors resulted from scoringtrue heterozygotes as homozygotes. The high overall accuracyachieved (96%) may have been enhanced by spacing sequen-cing primers only 350 bp apart such that the two strandswere compared over 98–100% of the sequence. Additionaloverlapping sequences were also available for each strandover approximately 75% of each intron. Heterozygositywas recorded only when two or more sequencing reactions,including both strands, showed doublet chromatogram peaks.However, in some cases sequence from different directionsaccentuated different nucleotides at a heterozygous site, withthe alternative minor nucleotide indistinguishable from noise.For example, a position might show an unambiguous A inone strand and a G in the other strand. These sites were

RFLP–checkable total

Locus Total bp Total het positions het positions het bases* No. correct % correct

CAMK 2144 8 4 36 34 94BTM 1900 11 5 45 40 89HEXB 1945 16 8 80 80 100Total 5988 37 17 161 154 96

*Checkable het bases = (number of checkable het positions) (number of individuals with diploid sequence).

Table 1 RFLP testing of heterozygous (het)sequence accuracy

Table 2 Alignments of variable nucleotides in CAMK and BTM introns. Restriction enzymes confirming heterozygous base calls areshown above nucleotide positions. Priming sites for allele-specific sequencing are shown as arrows above the template sequences.Nucleotides with phase determined are shown in bold; R = A or G, Y = T/C, W = A/T, S = G/C, K = G/T (correct phase in GenBanksequence accessions AF140557–140574 and AF140803–140821). Allele-specific primers had 3′ terminal nucleotide at the heterozygous site,length > 15 bp and Tm near 50 °C. Sensitivity of allele-specific sequencing to different 3′ mismatches was tested by making the alternateprimer at three of the allele-specific priming sites (shown as double-headed arrows) in order to sequence the alternate allele. In everycase the alternate primer confirmed the phase deduced from the first allele sequence

BTM Sequence position and restriction sites CAMK Sequence position and restriction sitesMnlI HinfI AluI TspRI TspRI MnlI DpnII ApoI BsmAI BsrI

1 1 1 1 1 1 1 1 1 21 2 3 4 4 9 2 3 6 3 0 2 2 4 4 7 0

4 9 6 2 5 7 9 9 5 1 8 1 2 5 6 5 8 3 6Sample 7 6 7 9 2 0 9 3 2 4 1 Sample 9 2 8 9 5 2 7 4

Lobl 143 C C A C G C A T T C C Lobl 1137 G G A T T T T CLobl 1137 C C G T A C T G T C C Lobl 1495 G G A T T T T C

←Lobl 701 Y C R Y R C A T T C C Lobl 1477 G G A T T T Y C

2 ← →Lobl 1454 C C R Y R C W K T C C Lobl 143 G R R Y T Y T CLobl 1289 C C A C G C A T T C C → ←Lobs NZ11 C G A T G C A T T C Y Lobs K99 G A R Y Y C T S

2 ← ← Lobs 178 R A G C C C T CLobs 178 C Y R T R Y A T T Y C Lobs 236 R A A T T C T C

→ ← ← Lobs NZ8 R A A T T C T CLobs 236 C Y R T R Y A T T C C 2 1

→ ← ← Lobs NZ11 R A R Y Y C T CLobs K99 C C R T R Y A T Y C C

MEC738.fm Page 1751 Wednesday, September 29, 1999 10:18 AM

1752

TE C H N I C A L N O T E S

© 1999 Blackwell Science Ltd,

Molecular Ecology

, 8, 1749–1752

scored as heterozygous if both sequences had well-resolvedsignal, or if there were overlapping sequences confirming theheterozygous pattern. Obviously, comparison of both strandswas essential for these assignments.

For phylogenetic analysis of alleles, diploid sequenceswith

≥

2 heterozygous sites must be resolved into theircomponent haplotypes by determining the

cis–trans

phase ofheterozygous positions. With diploid sequence data in hand,allele-specific PCR (Bottema

et al

. 1993) becomes a viablemethod of haplotype determination but requires the produc-tion of new sequencing template. We reasoned that allele-specific PCR primers might also act specifically in a sequenc-ing reaction with the original diploid sequencing template.

Allele-specific sequencing primers were designed atheterozygous positions if there was one or more additionalheterozygous site(s) within 500 bp. All allele-specific primersgenerated well-resolved haploid sequence or showed stronglydominant (> 50% difference) chromatogram peaks withinheterozygous doublets. Using these primers, allele-specificsequencing distinguished phase for 9 of 16 (56%) hetero-zygous bases at the calmodulin-dependent kinase (CAMK)locus and 19 of 22 (86%) heterozygous bases at butyrophilin(BTM; Table 2). Among a total of eight diploid sequencescontaining two or more heterozygous sites, phase was com-pletely resolved in three. The other five sequences each hada single heterozygous site still unresolved, mostly because

these sites were outside the range of allele-specific sequencingprimers (Table 2). Thus, the efficacy of allele-specific sequen-cing depends on the distribution of heterozygous sites. Whenheterozygous sites are clustered, allele-specific sequencingof diploid template provides a more efficient method thanallele-specific PCR followed by sequencing.

We have demonstrated that diploid sequencing can bean accurate method to rapidly obtain population data onnuclear polymorphisms. Coupled with direct allele-specificsequencing of the same template to resolve haplotypes, thesemethods can facilitate more rapid collection of diploid nucleardata for population and genealogical analyses.

Acknowledgements

DNA samples were provided by F. Cipriano. T. Duda suppliedhelpful comments. This work was funded by NSF grant DEB 97–96061 to SRP.

References

Bottema CD, Sarkar G, Cassady JD, Ii S, Dutton CM, Sommer SS(1993) Polymerase chain reaction amplification of specific alleles:a general method of detection of mutations, polymorphisms,and haplotypes.

Methods in Enzymology

,

218

, 388–402.

MEC738.fm Page 1752 Wednesday, September 29, 1999 10:18 AM