Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
Genomic analysis—whether sequencing, genotyping, haplotyping, microarray analysis, comparative hybridization or what have you—often requires more DNA than you have on hand. Typical next-generation sequencing protocols, for example, call for microgram quantities of starting material, far more than may be obtained from many precious clinical or forensic samples.
Researchers have long had techniques for amplifying small amounts of DNA, of course. Even before PCR, they could bulk-produce DNA segments by chopping them up and cloning them in bugs—a laborious, time-consuming and not-always-successful process when trying to evenly reproduce an entire genome. More recently, researchers have turned to so-called whole-genome amplification (WGA). Like PCR, these test-tube techniques use polymerases to generate multiple copies of starting (template) DNA. But they’re not necessarily based on thermal cycling alone, and they generally are far less prone to amplification bias.
WGA today
Several techniques were developed in the early 1990s that used random or partially degenerate oligonucleotides to prime single-stranded genomic DNA at multiple sites. These primer-template complexes were then subjected to multiple rounds of PCR to geometrically amplify the DNA. At roughly the same time, researchers introduced techniques that first ligated specific primer-binding sites to restriction-digested genomic DNA and amplified the resulting material using PCR. Variations, improvements and combinations of these methods are still being used today, both as homebrew protocols and commercial kits.
Yet because PCR is exponential, it doesn’t afford even amplification. “It’s fine for detecting one gene—it’s very sensitive. But if you want to detect all the genes in the genome, then the sequence-dependent bias is severe, meaning that some genes get amplified thousands of millions of times higher than other genes,” says Xiaoliang “Sunney” Xie, Mallinckrodt, professor of chemistry and chemical biology at Harvard University. One way to correct for this problem, Xie says, is to “sequence really deep. But then that’s very expensive.”
Multiple displacement amplification (MDA) was developed in part to solve the problem of amplification bias [1]. The isothermal technique uses exonuclease-resistant random primers and a polymerase that can displace overlapping products from the template. Nascent strands act as substrates for further priming, seeding additional rounds of synthesis, resulting in a hyperbranched DNA product.
The Phi29 polymerase employed in the original studies, and frequently incorporated into commercial MDA kits, is highly processive and has 3’ → 5’ exonuclease-proofreading ability. Amplicons produced are typically greater than 10 kb, with error rates perhaps 1,000 times lower than those created using standard Taq polymerase.
MDA allows for “more even amplification than PCR,” says Xie. But because it is exponential rather than linear, any difference between the amplification efficiency of different sequences is amplified as copies are made from copies. “The error will propagate.”
Eliminating bias
Various modifications to PCR- and MDA-based WGA methods attempt to minimize amplification bias further. Qiagen’s MDA-based REPLI-g® Single Cell Kit, for example, uses a proprietary version of the Phi29 enzyme—which supposedly improves amplification uniformity and reduces chimerism (artifacts formed when amplicons hybridize with each other or template to prime further reactions)—and UV-treats the kit’s ingredients to eliminate contaminating DNA.
Sigma-Aldrich's GenomePlex® whole-genome amplification kits make use of long, quasi-degenerate primers containing a 5' universal adapter sequence, which are allowed to anneal to fragmented DNA. The primers are extended isothermally to produce fragments with the adapter sequences on both the 3' and 5' ends. This is followed by a limited number of PCR cycles using the universal primers. By initially using degenerate primers, much of the priming bias of sequence-specific PCR is avoided, notes Savita Bagga, global product manager for epigenetics and whole genome/transcriptome technologies at Sigma-Aldrich. (The company has also recently introduced SeqPlex WGA kits, designed to prepare the product to seamlessly enter next generation sequencing workflows.)
The single primer isothermal amplification (SPIA) technique, too, begins its WGA process by annealing degenerate primers with a unique adapter sequence to template DNA. In the case of SPIA (marketed by NuGEN as the Ovation® WGA System), the primers are composed of a chimeric DNA/RNA which is elongated to produce double-stranded DNA with a DNA/RNA heteroduplex at one end. RNAse H and a strand-displacing DNA polymerase degrade the RNA portion of the heteroduplex and expose the binding site for another primer for subsequent MDA synthesis. The amplicons are not themselves templates for the chimeric DNA/RNA primer, thus avoiding the pitfalls of exponential synthesis.
Xie and his colleagues have come up with another way to avoid amplicons becoming templates. The multiple annealing and looping-based amplification cycles (MALBAC) technique uses random primers containing a common 27-base sequence to hybridize to the template DNA [2]. Amplifying these with a strand-displacing polymerase at 65°C generates “semiamplicons,” which are then melted off at 94°C. Subsequent cycling yields full amplicons that form a hairpin loop, preventing them from being used as templates. “We don’t make copies of the product. We only use the original template,” Xie explains. MALBAC is currently offered in kit form from Yikon Genomics.
Size matters
The quality, length and quantity of starting material, as well as the expected length and concentration of product, can make a difference when selecting a WGA method or kit.
Some techniques use damaged, degraded, formalin-fixed paraffin-embedded (FFPE), sheared or similarly fragmented DNA as their input, making use of the frequent breaks for priming or ligation. Others, like MDA, require longer stretches of intact DNA and yield longer products—which are necessary for certain downstream applications such as Southern blotting or restriction fragment length polymorphism (RFLP) analysis.
Similarly, some protocols require more input DNA than others. Active Motif’s GenoMatrix™ Whole Genome Amplification Kit, for example, requires at least 10 ng of starting material. “It’s probably not for single cells,” says product manager Kyle Hondorp. Meanwhile, Xie and his collaborators have applied MALBAC to amplify single human oocytes and circulating tumor cells, and several companies offer kits specifically for single cells, as well.
Ultimately, says Hondorp, the hope is to develop downstream techniques sensitive enough that whole genome amplification is no longer even required. In the meantime, if you need more DNA than your sample allows, you have plenty of options to get the job done.
References
[1] Dean, F.B., et al., “Comprehensive human genome amplification using multiple displacement amplification,” Proc Natl Acad Sci USA, 99:5261-6, 2002. [PubMed ID: 11959976]
[2] Zong, C, et al., “Genome-wide detection of single-nucleotide and copy-number variations of a single human cell,” Science, 338:1622-6, 2012. [PubMed ID: 23258894]