It’s been 15 years since the first technique for immunoPCR was described, but it still hasn’t gotten the kind of traction other antigen-detection technologies have enjoyed. But as improvements to PCR — most importantly, the advent of real-time PCR — have made quantification more effective, and as clinical monitoring of disease is increasingly based on the detection of antigens and other protein biomarkers, immunoPCR’s star might be rising at last.
For those of you still unfamiliar with it, immunoPCR is a technique similar to ELISA whereby an antibody is used to detect and quantify a specific antigen in a mixed sample. In immunoPCR, however, the antibody is coupled to a piece of DNA, which is then amplified using real-time PCR. The technique is markedly more sensitive than ELISA, primarily because it combines the detection specificity of an antibody with the amplification power of real-time PCR for a nucleic acid. The technique was first described using PCR in 1992, but since coupling the technique to real-time PCR has become the standard in recent years, it’s now possible to harness the real heft of immunoPCR; lower limits of detection can surpass ELISA by 100- to 10,000-fold. Once the standardization hurdles are overcome, immunoPCR could well be an effective, and highly sensitive, diagnostic tool to test for protein infectious agents, biomarkers, and toxins.
A Method Is Born
In 1992, Takeshi Sano, Cassandra Smith, and Charles Cantor first described an original immunoPCR method of using an antibody sandwich coupled to a DNA molecule to capture and measure the amount of antigen present. While similar to ELISA, the assay was not enzyme-linked; instead of correlating the amount of antigen captured to the intensity of the antibody-linked enzymatic reaction, the label on the detection antibody was a DNA molecule, which was then amplified and quantified using PCR.
Sano, Smith, and Cantor’s method used a linker molecule to attach the piece of DNA to the detection antibody. Taking advantage of the fact that biotin has a high affinity for streptavidin, they created a streptavidin-protein chimera that had binding affinity for both a biotinylated DNA molecule and the antibody. The chimera bound on both ends: streptavidin to the biotinylated DNA and the protein to the antibody. Using bovine serum albumin as antigen, they were able to detect a mere 580 molecules, or a 105 enhancement over ELISA detection, and several orders of magnitude more sensitive than radioimmunoassays.
Most of the limitations that arose back then, and that still exist in current methods, were centered around the difficulties of creating the antibody-DNA conjugate. “The real challenge in ’92 was making the antibody-nucleic conjugate in a clean way,” Cantor says, adding that the issue of background noise was also a potential problem. Background, he says, comes from the “antibody sticking to [the] glass, binding to the wrong protein, [and] binding to each other.”
Mark Pandori, who is now chief microbiologist at San Francisco’s Department of Public Health Laboratory, remembers the hurdles of this first research as a postdoc in Cantor’s lab. “Basically, the problem at hand [was], how do you connect the DNA molecule to the antibody?” Pandori says. “When we started out, there wasn’t a real easy way to do that. We had not thought through the chemistry of it — and nobody had, because nobody could imagine that you would want to conjugate a DNA molecule to an antibody.”
When it came to analyzing whole blood specimens, says Pandori, competitive binding became a major problem using the original conjugation technique. In blood samples, for instance, there’s more than just the antigen of interest; there are all sorts of analytes, including other proteins and antibodies. When coupling streptavidin to what they called protein A, Pandori says, “what happens is that the protein A moiety will bind fairly tightly in IgG — and it really depends on what species as far as how tightly it will bind.”
Fast-forward 15 years and the process has evolved significantly, thanks to improved conjugation techniques. Typically, current methods directly couple a biotinylated DNA molecule to a biotinylated detection antibody via a streptavidin linking bridge; this conjugation method decreases the risk of nonspecific binding and has made it a suitable assay for both basic and clinical research.
“The stumbling block was developing the reagents, and now there are a number of different approaches to make these reagents in a very efficient way,” Cantor says. “So that’s going to change things abruptly.”
The Latest Protocols
A decade and a half after the first go at assembling a DNA-antibody conjugate, there still is no standard lab protocol for making a linker; each lab must determine its own methods and optimize them as best they can.
According to Ron Wacker, product manager at Chimera Biotec of Dortmund, Germany, today’s conjugates typically fall into one of two categories: biotin/streptavidin conjugates or covalent conjugates. Wacker says that the most difficult step is conjugating the DNA to the antibody. “I have to say it’s not the easiest technology,” he says. “During optimization, there are a lot of difficulties to set up the assay. You just have to know where to switch something to get a better result, so it’s more or less experience you need to establish immunoPCR assays in your lab.”
Mikael Kubista is a leader in the field of real-time PCR. He founded the TATAA Biocenter in Sweden in 2001, which is now the leading provider of real-time PCR services and hands-on courses in Scandinavia and central Europe. In a 2005 paper, he and graduate student Kristina Lind described results of comparing three different immunoPCR assemblages to the standard ELISA test for PSA, or prostate specific antigen.
“We wanted to identify where the bottlenecks [were] in terms of sensitivity, linearity, and reproducibility,” Kubista says. In using the PSA test as a model, they determined that having good antibodies — ones with high specificity and high affinity for the antigen — and protecting the surfaces from nonspecific binding of the antibody are the most important for a successful assay.
Another major consideration they advise taking into account is the number of washing steps. “When we were comparing different ways to assemble these assays, we found that when it comes to reproducibility, what is very, very important is to keep down the number of washing steps as much as possible,” Kubista says. His team found that the assemblage that had the highest sensitivity and the highest reproducibility was the one where the capture antibody was passively adsorbed to the slide surface, but the DNA was covalently conjugated to the detection antibody and premixed with the protein sample before addition to the well. Compared to ELISA, they found immunoPCR to be 100 times more sensitive in detecting PSA.
Kubista, however, recommends caution when embarking on developing a lab protocol for immunoPCR. He suggests starting with a “straightforward procedure, where you actually add the components stepwise, which means that you have a large number of washing steps,” he says. “If the assay looks promising, and you will be using it in routine, then it’s worth [it] to take the time to actually make the conjugate and purify it.” He adds that “it’s not worth doing it unless you’ve tested that the antibody is really good enough.”
Path to Diagnostic
It was Christof Niemeyer who, after finishing his postdoc in Cantor’s lab at Boston University, headed back to Germany and founded Chimera Biotec, the only company on the market today making customized immunoPCR kits. Chimera offers kits that utilize primary and secondary antibody-DNA conjugates, customized services including conjugate synthesis, feasibility studies, optimization protocols, and assay validation. Most of their clients, says Ron Wacker, are large US-based pharmaceutical companies.
“Most assays we develop are for somewhere in the drug development chain,” Wacker says, adding that companies primarily use them in early-stage, preclinical research. In drug design, measuring the pharmacokinetics or toxicokinetics of drug metabolism, he adds, is an area where immunoPCR is especially useful. “In this specific area, there’s a need for [highly] specific [and] sensitive immunoassays.”
While far from an approved diagnostic tool, immunoPCR has also recently started to be used in clinical research for detecting infectious proteins and protein biomarkers of both cancers and viral infectious agents. Niel Constantine’s group at the University of Maryland, Baltimore, has published several papers on the topic, including one that describes the highly sensitive detection of HIV-1 p24 antigen in HIV-1 infected patients using immunoPCR. His team found that using immunoPCR lowered the detection limit to 1,000 p24 antigen molecules, or 0.3 virions, compared to the current limit of 25 virions per milliliter using RT-PCR tests. “We can detect it earlier, days before RNA testing,” Constantine says. “We have a lot more target to go after, so we should be able to detect smaller levels, earlier.”
Though the potential is there for immunoPCR to be used for numerous diagnostic applications — including monitoring disease through biomarkers, and locating low levels of toxins in food, water, or people — the test must be standardized first. “The method is still in the developmental phases and being validated,” Constantine says. “Nobody is using it in any way to diagnose disease or to monitor disease, but that’s the implication. That’s what people would like to be able to use it for, but nobody’s doing it yet.” The reason for this, he says, is that “it has not been standardized. That’s the key.”
For most cases PCR is good enough, because quantification is all that is needed. “But when you get into monitoring cancer, and you’re looking for increases in cancer markers, yes, you want to be able quantify it somehow and [so would] use immuno-qPCR,” Constantine says.
Kubista says the most effective use of immunoPCR comes in situations that require extremely high sensitivity and wide dynamic range, two categories where it handily beats out ELISA-based tests. Though today its use is limited to detecting proteins, “in principle you can detect anything that you can bind with an antibody,” he says. “So you could, in principle, detect different cells, perhaps even metabolites, organelles. And I’m pretty sure that will happen.”
For Kubista’s 2005 study, he and his colleagues chose PSA because it is a very popular model system for protein detection. The most common use of PSA test is to check men for elevated levels of the antigen to detect possible prostate cancer. For that use, “standard ELISA-based tests are good enough and probably difficult to replace in routine,” Kubista says. But indications suggest that very small changes in PSA levels could be worth noting in forensics and cancer diagnostics in women, where minute changes in those levels can indicate breast cancer. To be able to use those for diagnostic purposes, immunoPCR may be a key technology player.
“Our experience is that immuno-qPCR is still mainly used in research, and the first routine applications will be cases where large dynamic range is desired and also higher sensitivity,” Kubista says. “It will take some time before it’s a routine tool, but certainly for these specific special applications, [research] is very important.”