8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 1/6

nature methods | VOL.7 NO.5| MAY 2010| 351

technology feature

C v r PCR: or o yp , a r voMonya Baker

With microfluidics and multiplexing, researchers can get more information from PCR products in lesstime and with fewer reagents.

l 354

t b 1: P p v dna 354

B 1: Pcr p b p 352

“Where PCR is really going,” says OlivierHarismendy at the University of California,San Diego, “is parallelization and minia-turization.” Indeed, researchers are mak-

ing use of a wide variety of materials andapplications toward these goals.

Miniaturizing PCR protocols offers arange of benefits, says Bruce Gale, whodirects the Utah State Center of Excellencefor Biomedical Microfluidics. “As soon as you go to microfluidics, you can bump upthe speed and precision because you usesmall volumes,” he says. Other advantagesinclude portability plus the ability to workwith smaller samples and fewer reagents.Growing enthusiasm for shrinking vol-umes is being felt in the industr y, says

Jeremy Gillespie, group product managerat Thermo Fisher Scientific. Scientists areincreasingly interested in buying smaller

amounts of reagents because they are usingsmaller volumes in their assays, he says.

Most microfluidics chips conduct PCRusing minute volumes of highly dilute

samples, relying on fluorescence (ormelting-point temperatures) to determinethe presence and quantities of a geneticsequence in a sample. Many applicationsinvolve finding some version of a needleinside some version of a haystack: raremutations among lots of wild-type DNAor fetal DNA in a maternal blood sample,for example. Some applications lookat relative amounts of DNA sequences,such as copy-number variation, alleleratios or microbe sampling. Newer PCRapplications use microfluidics techniques

as a preparative technology for next-generation sequencing. These sequesterDNA in tiny reaction vessels for PCR

amplification, then collect the ampliconsfor subsequent analysis.

m -

With each run of a sequencer potentially generating data for 100 gigabases, “you’vegot to think about how to design experi-ments so as not to waste dat a, becauseit’s costly,” says Daniel J. Turner, head of sequencing technology development atthe Wellcome Trust Sanger Institute. “Thetwo things that affect how much sequenc-ing you have to do to see what you need tosee are how specific your sequence enrich-ment is and how uniform your data are,”explains Turner.

Sequencing capacity is wasted whensome genomic regions are amplifiedmore than others. “When you say youwant 20-fold, you want 20-fold every-where,” says Harismendy. “If you want tosequence all the exons at 20-fold cover-age, you don’t want some at 100-fold orat fivefold ,” he adds. If some exons in astudy are only present at fivefold coverage,then an analysis can only claim fivefoldcoverage. Compared with other technolo-gies for massively paral lel PCR, he says,microfluidics platforms excel at unifor-mity because the products of PCRs do notcompete with each other.

Fluidigm launched one such microflu-idic platform, its Access Array, in the fallof 2009. The chip uses a matrix that guidessamples and reagents into tiny chambers.“You take 48 samples and 48 primer pairs,and it makes every combination and doesPCR, and then you can pump the samplesback out,” explains Ken Livak, technol-ogy developer at Fluidigm Corporation.During the PCR a DNA barcode is addedto each sample so that the 2,304 amplifiedThe Access Array chip from Fluidigm can amplify 48 genomic regions from 48 samples.

F l u i d i g m

C o r p o r a

t i o n

8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 2/6

352 | VOL.7 NO.5| MAY 2010| nature methods

technology feature

allowing his team to decide whether ornot to run a sample.

RainDance Technologies has com-mercialized an instrument that usesmicrodroplets to efficiently prepare sam-ples for next-generation sequencing. Thecore of RainDance’s technology is the cre-

ation and combination of two large sets of tiny, precisely aliquoted drops, each aboutthe size of a typical eukaryotic cell. In oneset, each drop contains a specific pair of PCR primers designed to target a regionof interest. In the other set, each drop con-tains a large portion of genomic DNA tobe amplified along with the enzymes andother reagents necessary for PCR. A spe-cially designed chip merges the drops ata rate of about 3,000 per second, all thewhile ensuring that the drops combine ina constant one to one ratio.

Harismendy recently completed a col-laboration with RainDance showing howthe technology can be used for target cap-ture. He and colleagues reported usingthis approach to examine 435 exons in

can be pooled and sequenced together.The Access Array Chip excels in produc-ing uniform numbers of the desired ampl-icons, says Livak. “You get a ‘tight’ range sothat you don’t have to go way deep [intothe sequence] without missing anything,”he adds.

Johan den Dunnen heads the GenomeTechnology Center at Leiden University Medical Center in the Netherlands, wherehis team uses Fluidigm’s products as wellas an approach called hybridization cap-ture ( Box 1). The larger or more complexthe targeted regions are, he says, the morelikely he is to go with hybridization, butthe Fluidigm PCR approach offers severaladvantages. Even without any optimiza-tion, he says, it has “nice uniform enrich-ment” compared to hybridization tech-niques. Plus, with hybridization capture,

researchers only find out that targetedregions were missed after a sequencingrun has been completed. In contrast,Fluidigm’s arrays can show whether a tar-get region was enriched before sequencing,

products can be pooled and shuttled intoany next-generation sequencer. Separatechips can provide unique sets of barcodesso that PCR products from multiple chips

The stand-alone thermocycler from Fluidigmconducts PCRs on a microfluidics chip.

F l u i d i g m

C o r p o r a

t i o n

Several techniques allow researchers to do massively parallelsequencing without using microfluidics 4,5. These often use long,cleverly designed targeting primers or probes aimed at capturingrather than amplifying desired regions.

In the circularization approach, primers have varying target-specific sequences and common linking sequences. When addedto genomic DNA along with ligase, these primers circularizedesired DNA regions (leaving non-circularized bits to be clearedaway by nucleases) and also add common DNA tags that allowevery sequence to be read by a single set of primers. “We’vedone 100,000 simultaneous captures in the same tube, andthey don’t interfere with each other,” says Jay Shendure of theUniversity of Washington. Though Shendure uses molecularinversion probes (also called padlock probes) that he helpeddevelop as a graduate student, conceptually similar reagents willsoon be available commercially. Olink Genomics, for example, isdeveloping what it terms the Selector Technology, which workson DNA that has first been cleaved by restriction enzymes.

Other widely used methods are based on hybridization,which can occur either on arrays or in solution. In thisapproach, desired genomic regions are captured either insolution followed by PCR or on arrays and fed into appropriatenext-generation sequencers.

Researchers led by Nick Papadopoulos and Bert Vogelsteinat Johns Hopkins University recently reported surprisingheterogeneity in different tissues of the same individual 6.To obtain their results, the researchers used a combinationof hybridization and primer design to accomplish 1,000-foldcoverage of the mitochondrial genome. In one set of experiments,researchers used three different sets of primers (segregated into

separate vessels to avoid interference during PCRs) to amplifyseparate but overlapping regions of the mitochondrial genome.In a second set of experiments, the researchers made biotinylatedDNA probes specific for the mitochondrial genome, used these to

pull out desired sequences from whole genome DNA, amplifiedthose and fed them into an Illumina sequencer. “If you want toget the mitochondrial DNA, sequencing the whole genome is justwasteful,” says Papadopoulos. High-throughput microfluidicstechniques seem like good technologies, he says. “However, formtDNA they can be an overkill,” he adds. Nonetheless, complete,thorough coverage is essential; with the appropriate primers onhand, Papdopolous thinks his team will continue to use both thebiotinylation capture and standard PCR approaches: “When youhave the same results from two different methods it’s reallymore believable.”

Most researchers, however, will end up using just oneapproach, and each has pluses and minuses. Hybridizationapproaches may capture many unwanted sequences and offerless uniformity than PCR. However, they are readily scalable,and there is no need to use more DNA even when larger regionsof the genome are captured. The circularization approachesusually capture more than 90% of the desired sequences and cancapture at least 25-fold more sequences than currently availablemicrofluidic approaches from RainDance and Fluidigm, but somesequences are amplified to a much greater degree than others.

As technology develops, of course, optimal approacheschange. Eventually, the cost of capturing targets for sequencingmight outstrip the costs of sequencing the whole genome, orsingle-molecule sequencing could make some target-capturestrategies obsolete.

BOX 1 PCR PROBes fOR multiPleXing

8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 3/6

technology feature

nature methods | VOL.7 NO.5| MAY 2010| 353

says Darren R. Link, a co-founder of RainDance.“They can do the [whole-genome amplification], useRainDance technology andstill have sample remain-ing,” he says. RainDance’s

platform currently allowsabout 4,000 primer pairsto be used at one time, butimproved efficiencies to belaunched this summer willexpand the instru ment’srange to around 20,000primer pairs.

Nonetheless, researcherslike Wellcome’s Turner arecurrently using multiplex-ing rather than microflu-idics approaches because

the former are capable of pulling downtens of thousands of genomics regions atonce ( Table 1 ). When RainDance expandsits primer capacity in the summer of 2010,the technology will start to become com-petitive with multiplexing, Turner says,though he thinks the projects he is run-ning will likely require coverage of many more regions. He is quick to point outthat the decision about which preparativetechnology works best depends on thenumbers of samples as well as the numberand type of sequences to be studied.

samples from six people 1. They choseregions that included repetitive ele-ments, varying amounts of G+C con-tent and other sequence features thatoften cause variation in amplification.The results showed accuracy and cover-age equivalent to what they would haveobtained if they ran all the PCRs in 1.5million separate 20-microliter tubes, butthey used a fraction of the reagents anddisposables. “It was so obvious that it wasa big savings and it’s also much easier,”says Harismendy, who is not affiliated

with RainDance. “There’s littlehandling. You just p ut your[primer] library on one sideand your template [DNA] onthe other s ide,” he adds. Thedroplets for each sample arethen transferred into a singlePCR tube that is placed into athermocycler to allow PCRs torun their course; after a suffi-cient number of amplificationcycles, the oil-water emulsionthat keeps droplets separate isbroken and barcodes can beligated to PCR products, whichare then ready to go into any of the next-generation sequenc-ers.

An advantage of microfluid-ics systems is that they can workreliably with whole genome–amplified DNA without becom-ing biased to certain alleles. Thiscould be particularly useful forresearchers who may only havetiny amounts of starting DNA,

Primerlibrary

Genomic DNAtemplate mix

Microfluidicchip

Off-chip

PCR droplets

Break emulsion

DNA clean-up

Sequence

a b

c

a b c

RainDance creates tiny droplets as vessels for PCR amplification.

R a

i n D a n c e

T e c h n o

l o g

i e s

Samples loaded into the Access Array Integrated Fluidic Circuit areamplified, barcoded and pumped out again for analysis.

F l u i d i g m

C o r p o r a

t i o n

8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 4/6

8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 5/6

nature methods | VOL.7 NO.5| MAY 2010| 355

technology feature

The scope of other PCR-based applica-tions is tremendous, Gale says. The hope isthat the advance of new technologies canwork with microfluidic PCR amplificationsin much the same way as it has worked insequencing. In other words, as capacity goesup, costs and size come down.

1. Tewhey, R. et al. Nat. Biotechnol. 27, 1025–1031 (2009).

2. Zeng, Y., Novak, R., Shuga, J., Smith, M.T.& Mathies, R.A. Anal. Chem. advance onlinepublication 1 March 2010 (doi:10.1021/ ac902683t).

3. Sundberg, S.O., Wittwer, C.T., Gao, C. & Gale,B.K. Anal. Chem . 82, 1546–1550 (2010).

4. Turner, E.H., Ng, S.B., Nickerson, D.A. &Shendure, J. Annu. Rev. Genom. Human Genet.

10 , 263–284 (2009).5. Mamanova L.et al. Nat. Methods 7 , 111–118

(2010).6. He, Y. et al. Nature 464 , 610–614 (2010).

Monya Baker is technology editor for Nature and Nature Methods (m.baker@us.nature.com ).

tal samples and bodily fluids. Too young toeven have a website, start-up company Espira

aims to commercialize these disks. The goalis to pair a $20 camera with PCR chips thatcost just a few pennies.

contained. Analysis took just over half anhour, including image analysis. Though the

device is still being optimized, Gale imaginesthat versions of it could be used to find rarecell types or mutations in both environmen-

Microfluidic chip

Dropletencapsulation

Oil

Beads

Cells

Emulsion PCR

Oil Hotstart PCR

Forward primerlinked beads

Dye labeledreverse primers

A microfabricated emulsion generator array from the Richard Mathies lab at the University of California,Berkeley uses PCR to find rare cells.

Y o n g

Z e n g

8/8/2019 Nature Publishing Group 2010 Baker

http://slidepdf.com/reader/full/nature-publishing-group-2010-baker 6/6

technology feature

356 | VOL.7 NO.5| MAY 2010| nature methods

suPPliers guide: comPanies offering Products for Pcr, target enrichment and next-generation sequencing analysisc p W b

454 Life Sciences, a Roche company http://www.454.com/ Affymetrix http://www.affymetrix.com/ Agilent Technologies http://www.agilent.com/ Ambry Genetics http://www.ambrygen.com/ Applied Biosystems (part of LifeTechnologies)

http://www.appliedbiosystems.com/

Biosearch Technologies http://www.biosearchtech.com/ Bioteam http://www.bioteam.net/ Complete Genomics http://www.completegenomics.com/ DNAStar http://www.dnastar.com/ Epicentre Biotechnologies http://www.epibio.com/ Eppendorf http://www.eppendorf.com/ EurekaGenomics http://www.eurekagenomics.com/ Eurofins MWG Operon http://www.eurofinsdna.com/ febit http://www.febit.com/ FlexGen http://www.flexgen.nl/ Fluidigm http://www.fluidigm.com/ Genomatix http://www.genomatix.de/ GenomeQuest http://www.genomequest.com/ GenScript http://www.genscript.com/ Geospiza http://www.geospiza.com/ Helicos Biosciences http://www.helicosbio.com/ Illumina http://www.illumina.com/ Intelligent Bio-Systems http://www.intelligentbiosystems.com/ Invitrogen http://www.invitrogen.com/ LingVitae http://www.lingvitae.com/ OGT http://www.ogt.co.uk/ Olink Bioscience http://www.olink.com/ Open Genomics (part of Agilent) http://www.opengenomics.com/ Oxford Nanopore Technologies http://www.nanoporetech.com/ Partek http://www.partek.com/ Qiagen http://www.qiagen.com/ R&D Systems http://www.rndsystems.com/ RainDance Technologies http://www.raindancetechnologies.com/

Roche NimbleGen http://www.nimblegen.com/ Rubicon Genomics http://www.rubicongenomics.com/ Sigma-Aldrich http://www.sigmaaldrich.com/ ThermoFisher Scientific http://www.thermofisher.com/ USB (sells Affymetrix products) http://www.usbweb.com/