Bench philosophy: Isothermal DNA Amplification
PCR's Smart Isothermal Cousins
by Steven D. Buckingham, Labtimes 05/2015
Most researchers associate DNA amplification with PCR. Nucleic acids may also be amplified isothermically, avoiding PCR's time-consuming temperature cycles and sloppy specificity.
Where would we be without PCR? Whenever you find yourself wanting to amplify up a bit of DNA, that famous TLA (three-letter acronym) springs to mind in a flash. It stands for Polymerase Chain Reaction, by the way, and it is the cornerstone of molecular biology, allowing tiny amounts of DNA to be amplified.
And as with PCR, so with many other tools in our molecular toolbox: it all comes down to a gift of Nature. All the clever stuff in PCR is done by the DNA polymerase enzyme, which grabs nucleotides kindly chaperoned by the transfer RNA and attaches them to a template strand. The in-built complementarity of Nature's four nucleotides ensures (at least with reasonable fidelity) that the new strand is a fair copy of the original – only in mirror image, so to speak. This strict complementarity means PCR is sensitive and specific. The biology takes care of it all.
But much as we love it, PCR has its limitations and one of those is the awkward regime of temperature cycles it needs to get it to work. Indeed you need to get up to some sweltering temperatures, all because, once the polymerase has done its job, you have to pull the complementary strands off the target. In the lab this is done by heating the DNA up: when things get hot, the two DNA strands part company as the DNA simply melts apart.
The problem is, this happens at about 70 °C, which is more than warm enough to cook most enzymes (just try putting your finger in water that hot), including the polymerase enzyme. However, rescue came from a thermophile bacterium, which kindly gave us a super, heat-resistant polymerase called Taq, which made PCR easy and Cetis Corporation very rich.
So thank you, Taq, and your small band of similar, tough-skinned polymerases. You have solved the heat problem. Well, sort of. The temperature issue rules PCR out for a lot of applications. First of all, forget about applying PCR to living tissues. And that's a big loss, considering the huge interest in detecting DNA or RNA sequences in living cells. Take microRNAs for example. Given our comparative ignorance of the biological roles of miRNAs and their possible roles in disease, it is vital we find ways of accurately measuring what miRNAs are being expressed in tissues.
So wouldn't it be nice to be able to do the whole thing at the same temperature? That would make the amplification much more efficient and save a bit of time – after all, amplification only takes place at one part of a PCR thermocycle. And while we're at it, another real plus would be if that temperature happened to be one that doesn't cook tissues. Can this be done? Yes it can – indeed, isothermal amplification is a fast-moving and exciting field that has been somewhat eclipsed by PCR and its variants.
Isothermal amplification refers to amplification methods that take place at a constant temperature. The lack of cycling means that all the time is being spent on amplification, not just part of a cycle. And it means cheaper equipment, such as a water bath – good news for the lab budget. On top of all that, isothermal strategies are often more sensitive and more specific. So what isothermal amplification strategies are available? How good are they and do they compare well against good ol' PCR? Furthermore, who is using it and for what?
The answer to the first question is simple – there is a whole host of them. Their sheer variety defies any attempt to classify them. Often almost byzantine in complexity, they are an illustration of the ingenuity of their developers and show what can be done with some imaginative cobbling together of different molecular tricks and tools. And yet, there is just one thing they all have in common – they all incorporate a trick to get the newly-synthesised DNA strand to move off out of the way and allow further rounds of extension to take place.
So how do they do this? Take a method called Strand Displacement Amplification (SDA) as an example. This uses a primer against your DNA target that also includes a HincII recognition site. You let the DNA polymerase do its work and extend the target but you include in your mix some deoxyadenosine 5'-[α-thio]triphosphate (dATP[αS]). The new strand incorporates the HincII site but the HincII can't nick the dATP[αS]. The result is that the primer, and not the target, gets nicked. This leaves a 3' hydroxyl end, which gets extended by the DNA polymerase, displacing the original primer. And so the cycle continues. Clever.
And that was just plain vanilla SDA. There are also some new, more elaborate variants on the block, including "SDA-Ligase-G-quadruplex peroxidase detection" strategy. This is so specific that it has been used to detect variations of a single nucleotide with incredible sensitivity. How does it do this? To start with, three probes are used. Two of them are specific to either wild-type or mutant (called the discriminating probes) and a third can bind to either. If the discriminating probe binds perfectly with the target, it ligates with the third probe and rounds of SDA can proceed. If it doesn't match perfectly (because of a polymorphism or mutation), ligation doesn't occur and neither does SDA.
This approach has been used to identify specific miRNAs in single cells and in breast cancer patients – RuiXue Duan and colleagues at the Huazhong University of Science and Technology, China, could detect just nine strands of miRNA in a 15 μL sample (Duan et al., 2013, J. Am. Chem.Soc.).
Another isothermal method is best suited for detecting RNA and it is called Nucleic Acid Sequence-Based Amplification (NASBA). This starts with a primer (RNA+) that is complementary to the target RNA(+) but also contains a T7 promoter sequence. The primer directs synthesis of a single strand of DNA(-), resulting in a DNA-/RNA+ hybrid. Meanwhile, RNAse H digests the original RNA+ primer, leaving the single-stranded DNA exposed as a template for reverse transcriptase generation of double-stranded DNA.
This dsDNA is the entry point for the amplification cycle, as its T7 promoter sequence (it was in the original primer, remember?) directs the synthesis of lots of new RNA(-) molecules. These, in turn, are bound by a reverse primer and, once again, the DNA+/RNA- hybrid is produced, which, when RNAse H has digested the RNA(-) strand, leaves another single stranded DNA(+). Just like before, this is turned into dsDNA, which directs the synthesis of RNA(-), and so the cycle goes on generating RNA(-) exponentially.
SMART – Signal Mediated Amplification of RNA Technology (see what they did there?) – uses a three-way junctional complex. The idea is to have two primers that bind flanking regions of the target DNA. Only when the two primers bind do they form a stable bond with each other. One of the primers is longer than the other, and contains a T7 promoter and a transcription template (used for recognising the RNA). Once the three-way junctional structure has thus formed, the longer primer directs elongation of the shorter one. The T7 promoter drives RNA synthesis from the resulting dsDNA, resulting in many copies of RNA, which can be detected by many different possible methods.
One of the most popular isothermal amplification strategies is called LAMP, which stands for Loop-Mediated Amplification. LAMP works using two pairs of primers (two forward and two backward primers). Each pair consists of an inner primer and an outer primer. The inner primer also contains a sequence (F1c) complementary to a region (F1) slightly further along the target sequence. As this strand gets synthesised (we'll call this strand S1) it picks up a second F1c from the F1 on the target sequence. Meanwhile, the outer primer starts extending, pushing S1 off, while the F1c originally present in the primer, folds over and binds to the new F1c. This gives a dumbbell shape at the 5'-end.
The same happens at the 3'-end thanks to the work of the backward pair of primers. Result: a double dumbbell structure looking a bit like a closed staple. A bit more work by the DNA polymerase turns the dumbbell into a stem loop structure and this stem loop undergoes a proliferative recycling phase. If all that was just too much to follow, the Eiken Chemical company has a pretty animation to show the whole process (http://loopamp.eiken.co.jp/e/lamp/anim.html).
So how does LAMP compare with its older cousin, PCR? First of all, it is faster and cheaper. It is faster because there is no thermocycle involved and cheaper for the very same reason. But many labs have also found it to be more robust and more sensitive than PCR, and this is credited to the fact that it recognises no less than six separate sequences in the target. In the original paper, in which LAMP was announced, its inventors showed evidence that LAMP amplified just six copies of target molecule (Notomi et al. 2010, Nucleic Acids Research).
LAMP is also claimed to have higher specificity than PCR. Indeed, one of the problems with PCR is it is often thrown off the scent by irrelevant DNA present in a sample. Again, in the original paper, just six pieces of target were successfully amplified up, despite the contaminating presence of 100 ng of human DNA in the LAMP reaction mixture. This robustness against interfering materials makes LAMP particularly appealing in clinical diagnostic applications. And as far as simplicity goes, all you need to do LAMP is have the four primers, DNA polymerase and a laboratory water bath.
No wonder then that LAMP has established itself as a standard tool for molecular diagnostics in the field, such as identifying species. Most often, it has been used for identifying pathogens, either in the field or in biological samples. LAMP has been shown to be successful in signalling the presence of Neisseria meningitidis, the bacterium that causes meningitis and other serious illnesses, in cerebrospinal fluid (Lee et al., 2015, PLoS ONE).
But the prize for the most amazing application of LAMP must surely go to George Whitesides and his group at Harvard. In a recent issue of Analytical Chemistry, Whitesides announced a hand-held, paper-based (yes, you read that right – paper based!) isothermal DNA amplification device that can detect E. coli bacteria in a sample, and all it needs is a UV source and a camera phone (Connelly et al., Anal. Chem., 87, 7595-601).
It uses paper microfluidics and a multilayer gadget that allows you to add the sample, push in a slider, and then add buffer and LAMP master mix again just by pushing in another slider. After an hour's incubation at 65 °C, you add a bit of SYBR Green, shine your UV onto it and take a snap with your camera phone. And with as few as five E. coli in the sample, it glows bright green to give a positive signal.
Isothermal DNA amplification is not new but it seems to be going through its own proliferative amplification stage, as researchers are finding more and more novel applications, especially in contexts where either low temperatures are required (such as in vivo applications) or where economic (when you can't afford a thermocycler) or strategic (developing hand-held, in-field devices) considerations are important.
Handheld "Paper machine" for easy molecular diagnostics in four steps, applying loop-mediated DNA amplification. Foto: American Chemical Society
Last Changed: 14.09.2015