A quite popular bioanalytical approach to quantify nucleic acid therapeutics for pharmacokinetic research is the sandwich hybridisation ligand binding assay, also called the dual probe assay, or capture probe assay. Originally developed for diagnostic purposes (viral and bacterial human pathogen detection, you know …), pioneering companies like Isis Inc. or NeXstar Inc. (now part of Gilead Sciences, ) explored its real potential for their nucleic acid drug development programs.
The sandwich principle employs a capture probe and a detect probe both bearing complementary sequences to hybridise to the same single-stranded nucleic acid molecule (2). At the same time, it does play a role whether both probes are added to the ligand simultaneously (a liquid phase assay par excellence), followed by immobilization to multiwell plates or beads (3); or the ligand/detect probe complex is immobilized to capture probes (4) which previously were coupled to multiwell plates or beads (a solid phase assay). Both ligand binding assay assay variations seem to be quite similar, though bearing small but nice stumbling stones.
Subsequent Biotech companies with research activities in the field of nucleic acid therapeutics all applied this kind of bioanalytical ligand binding assays, here are at least some of them:
1. Sirna/Elli Lilly, single-stranded ribozyme as ligand (3),
2. Genta, single-stranded antisense as ligand (4),
3. Coley/Pfizer OTC, single-stranded CpG oligo as ligand (5),
4. Gilead/Eyetech/Pfizer, single stranded aptamer as ligand (6),
5. Noxxon Pharma, single stranded spiegelmer aptamer (7),
6. Silence Therapeutics, double-stranded liposomal siRNA (8),
7. Isis, phosphorothioate antisense oligo (9),
8. Calando, double-stranded siRNA (10),
9. Marina Biotech, double-stranded siRNA (11),
10. and finally, …
… Santaris Pharma uses the sandwich hybridisation ligand binding assay to quantify Miravirsen, the first microRNA antagonist currently being tested in a Phase 2 clinical trial (12, 13); the story behind this 15mer LNA/DNA chimeric microRNA-122 antagomir can be found here;
and many more …. Some of the companies mentioned above tried to generate IP on the assays, but basically failed, so this is why there should be a freedom to operate for all who plan to setup their own nucleic acid drug program.
Certainly there are other hybridisation assays which do not belong to the sandwich ones, see the comment on this topic (added on May 06, 2011).
(1) Gilead Sciences Press Release, dated March 01, 1999.
(2) Virtanen M et al 1983, The Lancet 1(8321): 381-3.
(3) Brown-Augsburger P et al 2004, J Pharm Biomed Anal: 129-39.
(the famous Sirna Heptazyme® …)
(4) Dai G, et al 2005, Clin Cancer Res 11(8): 2998-3008.
(the famous G3139 Genasense® …)
(5) Efler SM et al (2005), Oligonucleotides 15: 119-31.
(6) Drolet DW et al (2000), Pharm Res 17(12): 1503-1510.
(the famous Macugen® to be the first FDA approved nucleic acid therapeutic worldwide…)
(7) Helmling S et al (2004), Proc Natl Acad Sci USA 101(36): 13174-9.
(8) Aleku et al (2008), Cancer Res 68(23): 9788-98.
(9) Yu RZ et al (2002), Anal Biochem 304: 19-25.
(10) Heidel JD et al (2007), Proc Natl Acad Sci USA 104(14): 5715-21.
(11) Seth S et al (2011), Mol Ther doi:10.1038/mt.2011.21.
(12) Lanford RE et al (2010), Science 327: 198-201.
(13) Santaris Press Release, dated September 23, 2010.
Genasense® is a trademark of Genta Inc.
Macugen® is a trademark of Pfizer Inc.
Heptazyme® is a trademark of Sirna Inc.
Raising antibodies against nucleic acids, nucleic acid therapeutics, or nucleic acid hybrids can have superior advantages over conventional sandwich hybridisation ligand binding assays. Because these antibodies are often underestimated, in most cases they even are not recognized. Typically they are a sort of waste product observed in antibody detection assays (ADA assays), which nucleic acid biotech has to conduct in the course of their preclinical drug development programs (though most of these companies don´t want to see an immune response to their nucleic acid therapeutics, I guess…). It offers not only the opportunity to increase selectivity in terms of ligand binding, but also can increase specificity to the ligand.
One of the most interesting applications of solid phase ligand binding assays – at least from the bioanalytical point of view – are the Hybrid Capture® Assays once created by US-based Digene Inc. (1). These binding assays detect human pathogens in biological samples, with the pathogen´s genomic DNA to be the ligand, and are currently marketed by Qiagen Inc., who acquired Digene in 2007 (2).
The ligand binding assay principle is quite simple: genomic DNA ligands are captured by specific RNA probes to form a DNA:RNA hybrid, which afterwards is bound to microplates. The immobilization is mediated by specific antibodies raised against these DNA:RNA hybrids. The immobilized complex again is detected by additional antibodies, the final binding assay readout is like a common ELISA.
Another exciting example was explored by one of the pioneering antisense oligonucleotide therapeutics R&D companies, who succeeded in raising monoclonal antibodies against phosphorothioate antisense therapeutics (3). This tools detect antisense therapeutics in immunohistochemical specimen, as an alternative to in situ hybridisation. Reference (4) reflects basic research on this topic. Well, it strongly depends on the affinity, selectivity, and specificity of the antibody, but applying these antibodies to solid phase ligand binding assays could also be an option.
And finally, there is a monoclonal antibody available specific for the naturally occuring 5-methylcytidine (m5C) nucleoside modification. This antibody was recently developed for immunoprecipitation of methylated genomic DNA for pre-transcriptional gene silencing studies.
(1) Cullen AP et al (1997), J Clin Microbiol 35(9): 2275-8.
(2) Qiagen Press Release, dated June 03, 2007.
(3) Myers, KJ et al (2005), J Neuroimmun 160 (1-2): 12-24.
(4) Butler M et al (1997), Lab Invest 77(4): 379-88.
Hybrid Capture® is a registered trademark of Qiagen Inc.
The design of branched DNA probes by Chiron scientists in 1995 really marked a breakthrough in assay sensitivity (1, 2). This ligand binding assay worked basically like a sandwich hybridisation assay employing detect probes with a single terminal biotin for detection. Branched DNA sandwich hybridisation assays use a ravel of biotinylated DNA oligonucleotides, and due to their unique sequence, these oligos form a branched structure bearing numerous biotins for downward detection (3, 4). As you can guess, the sensitivity is dramatically increased even to the single transcript level, which forces Chiron to develop several diagnostic ligand binding assay kits for viral pathogen detection in human blood specimen.
Bayer AG aquired Chiron´s in vitro diagnostics business unit in 1998 (5), and built up the branched DNA ligand binding assay business. Genospectra Inc. exclusively licenced branched DNA assay technology from Bayer in 2003 (6). As times went by, Chiron Corp. vanished, Novartis acquired the company in 2005 (7). The same year Genospectra acquired and became Panomics Inc. (8), which again was aquired by Affymetrix in 2008 (9). And yes, the diagnostics market is still on the move …
Nowadays Panomics markets its Luminex® multiplex transcript quantification platform employing branched DNA sandwich hybridisation assay technology on beads. Panomics also distributes custom-made branched DNA assays for specific target transcript quantification as an alternative to Real-Time RT-qPCR liquid phase ligand binding assays.
(1) Pachl C et al (1995), J Acquir Immune Defic Syndr Hum Retrovirol 8(5): 446-54.
(2) Collins ML et al (1997), Nucleic Acid Res 25(15): 2979-84.
(3) Canales RD et al (2006), Nature Biotech 24(9): 1155-22.
(4) Shi L et al (2006), Nature Biotech 24(9): 1151-61.
(5) Bayer Press Release, dated September 21, 1998.
(6) Genospectra Press Release, dated October 31, 2003.
(7) Novartis Press Release, dated October 31, 2005.
(8) Genospectra Press Release, dated February 10, 2006.
(9) Panomics Press Release, dated November 11, 2008.
Luminex® is a trademark of Panomics Inc.
The preclinical/clinical drug development process of novel nucleic acid therapeutics requires not only species-specific pharmacokinetic studies (which means how the drug is biodistributed and cleared), but also pharmacodynamic examinations like the detection of any immune response to siRNA or microRNA drug administration. Most preferably those bioanalytical ligand binding assay studies are applied to blood samples, because this compartment is easily accessible, and sample collection is reasonable even for cancer patients.
The primary immune response to siRNA therapeutics can be triggered by the innate immune system (1), which means basically, that e.g.
1. an inflammation arises as a reaction to “drug” infection,
2. the complement system is activated, or
3. Toll-like receptors mediate an immune response.
An innate immune response to nucleic acid therapeutics like siRNA, aptamer, microRNA, etc administration could be easily detected by
1. detection of specific immunoglobulin levels (2) in the blood (IgG, IgA, and IgM),
2. C3a or C5a serum levels (3) leading to the release of thromboxane A2 (TXA2),
3. the release of cytocines like TNF-alpha or IFN-alpha (4) to the extracellular medium (cell culture medium in vitro, blood in vivo).
Total IgG, IgA, or IgM levels specific to microRNA, siRNA, or aptamer ligands can be detected, and even quantified by simple multiplate ligand binding assays. This procedure involves (a) coupling of the ligand to the surface of 96well or 384well plates, (b) incubation of antisera collected from a species at different time points for immunoglobulin immobilization, and (c) detection by species-specific anti-immunoglobulins coupled to a reporter protein like alkaline phosphatase.
Well, for understandable reasons the nucleic acid Biotech scene is not keen on detecting any immune response to this kind of drugs. This is why the final ligand binding assay readout is absorbance rather than chemiluminescence, because of its less sensitivity …
From our experience, there are some important aspects to be addressed before starting bioanalytical ligand binding assay development and validation in an Good Laboratory Practice (GLP) environment:
1. What could be a reasonable positive control?
2. What does the term “sample stability” mean and what might be the consequence for assay validation?
3. Which ligand is offered to the immunoglobulins to bind to? This is not as trivial as it seems to be, because microRNA or siRNA usually is packed into liposomal shuttles for intracellular delivery. And you can´t sort it out completely the secondary structure of nucleic acids is influenced by the ionic interaction with the (cationic) lipids.
4. Is it only important (not) to detect antibodies to the nucleic acid therapeutics itself, but also to the liposome vehicles (5), as the surface of the same is basically presented to the organism?
The ligand binding assay validation to detect immune response to biologicals or nucleic acid therapeutics is discussed elsewhere, but reference (6) is a good paper to start with.
Certainly the antibody detection assay (ADA assay) is also available with incredible sensitivity: a combination of enzyme-linked ligand binding assay with real-time qPCR, also called the immuno qPCR (iq-PCR).
(1) Judge A et al (2008), Human Gene Ther 19: 111-24.
(2) Wang XY et al (2007), J Controlled Rel 119: 236-44.
(3) Moghimi SM et al (2006), FASEB J 20: E2057-67.
(4) Tluk S et al (2009), Int Immunol 21(5): 607-19.
(5) Judge A et al (2006), Mol Therapy 13(2): 328-37.
(6) Gupta S et al (2007), J Immunol Methods 321: 1-18.
please be patient, coming soon
The latest development of ultra-sensitive multiplate readers produced an interesting enzyme-linked ligand binding assay which is also based on the sandwich ligand binding principle: the immunosorbent spot assay. It works basically as the classic enzyme-linked immunosorbent assays (ELISA), the ELISpot assay is also performed in 96-well plates, even the readout can be the same (absorbance or fluorescence).
However, the ELISpot assay is one of the most sensitive assays for the detection of specific proteins or peptides on the single cell level. The innovative sensitivity of the ELISpot assay is achieved by a simple modification of the ELISA: usually a polyvinyl-difluoride (PFDV) membrane is employed as a solid phase at the bottom of the wells (1).
A first capture molecule is then immobilized on the membrane using covalent coupling chemistry. As the ELISpot ligand binding assay was originally developed for screening experiments, each well can present a different capture molecule, e.g. like an antibody or a peptide.
A most common application of ELISpot ligand binding assays is the screening of a cytokine panel secreted from immune cells with or without external stimulus. In this case each well is covered with an antibody specific for a cytokine ligand. If immune cells like PBMCs are grown in the wells, they might secrete – lets say – TNF-alpha upon the stimulation with CpG oligonucleotides (2,3). The secreted cytokine is captured by a specific antibody immobilized on one of the wells. After incubation is complete, the cell supernatant is removed, and the TNF-alpha ligand which remains in the plate wells is detected by another secondary capture antibody coupled to a reporter enzyme.
The ELISpot ligand binding assay then proceeds like an ELISA. Finally the sensitive reader can detect single secretion events, which are visible as isolated spots on the membrane circles at the bottom of the multiplate wells. It is assumed that each spot represents a single cell which secreted the detected cytokine.
The ELISpot assay can be applied to almost any screening experiment to address distinct scientific questions. A nice application is the epitope mapping to characterize specific T-cell triggered immune response by means of IFN-gamma secretion (4), e.g. to oral grain challenge (5). In patients suffering from the celiac disease, peripheral blood T cells were isolated and grown on IFN-gamma ELISpot plates. Cells were stimulated by a peptide library simulating proteolytic degradation products of gluten. The degree of stimulation is then monitored by the amount of IFN-gamma secretion of the T cells.
(1) Czerkinsky CC et al (1983): J Immunol Methods 65: 109-121.
(2) Hartmann G & Krieg AM (1999), Gene Therapy 6: 893-903.
(3) Tluk S et al (2009), Int Immunol 21(5): 607-19.
(4) Wulf M et al (2009), Methods Mol Biol 524: 439-46.
(5) Tye-Din JA et al (2010), Sci Transl Med 2(41): 1-14.
Quantitative immuno polymerase chain reaction (immuno qPCR) provides an interesting combination of the antibody-mediated selectivity with the qPCR assay sensitivity. The method was originally described to improve the limit of detection of commonly used ELISA assays by at least 10.000fold (1). This sandwich ligand binding assay was originally performed in microtiter plate wells, and so was the immuno qPCR, because the ligands to be quantified were proteins or peptides. In fact, virtually any molecule which can be captured by an antibody, could also be detected by immuno qPCR.
The original method comprises an capture antibody homogenously immobilized on the bottom surface of microwell plates, usually O/N at ambient temperature. After washing and blocking, the antigen-containing sample (e.g. human serum with IL-6 as the analyte ligand ), is added, and the specific binding of the antigen is almost complete within 30 minutes. The detection in an ELISA-based sandwich assay employs the use of an labelled, secondary antibody, also specific for the already immobilized antigen. The most sensitive method for labeling is a DNA tag added to the secondary antibody, which is afterwards amplified in an real-time qPCR assay, either comprising beacons or an intercalating dye as the fluorescence reporter in real-time detection.
In the meanwhile, numerous anti-immunoglobulin, species-specific antibodies tagged with DNA for real-time immuno qPCR applications are available (3). This could open the use e.g. of these anti-IgG antibodies to detect immune response to exogenous small RNA/DNA drugs like antisense oligos, microRNA (miRNA) mimics and microRNA (miRNA) antagomirs, short interfering RNA (siRNA), short hairpin RNA (shRNA), etc. However, the immuno qPCR assay sensitivity often is not recommended (desired by the customer, I should better say) for antibody detection assays. We are glad to see that this application is already published (4).
The main two reasons why this sensitive method doesn´t become widely accepted in the drug development process may be (like always):
1. the time-consuming assay development process, that is, the antibody characterization (remember there are two antibodies to be characterized); tagging the secondary antibody with the DNA; illustration of the final assay detection window.
2. the cost of goods to produce the antibodies, the costs for human recources, which might be higher than for conventional ELISAs.
Another limitation of the binding assay is also clear: the immuno qPCR is only as good as the binding affinity of the antibodies to its ligand. And yes, it´s obvious the assay can only be applied if the analyte concentrations in the sample are expected to be beyond the lower limit of quantification (LLOQ), otherwise it does not make sense. Above all these disadvantages, the most affecting aspect is, that endogenous analyte (e.g. Erythropoietin doping most preferred by Tour de France cyclists in 1998 is to be detected in blood samples) may snafu the LLOQ, and therefore may neutralize the main advantage of the whole immuno qPCR assay: its sensitivity.
It should be stated here that another assay for reliable peptide/protein quantification can also comprise the quantitative LC-MS/MS method, a sensitive but expensive bioanalytical instrument (because purchase and maintenance is). Certainly the Good Laboratory Practice (GLP) guidelines and the FDA recommendations on ligand binding assays are also effective, which means: limit of detection (LOD), LLOQ, ULOQ characterization, accuracy, robustness,selectivity, and so on, etc. As an bioanalyst who enjoy to develop ligand binding assays, I am pleased to say that the immuno qPCR assay can sometimes exhibit the infamous Hook effect (7).
Some years ago the immuno qPCR was brought to the market by standardizing the basic components of the assay (5). Despite the main disadvantages the immuno qPCR assay has, it also has great potential in diagnostics of analytes which hardly can´t be detected by routing multiplate assays: peptide and protein doping in sports, microbial pathogen detection in diagnostics, or mycotoxin detection in human food (6).
By the way, the latter paper adopted the originally designed multiwell plate assay to magnetic beads immobilization of the antibody/analyte complex.
Please follow also our recent idea to advance this assay to immuno RT-qPCR format using a RNA tag.
(1) Sano T et al (1992), Science 258: 120-122.
(2) Niemeyer CM et al 2007, Nature Protocols 8(2): 1918-30.
(3) Adler M et al (2003), Biochem Biophys Res Commun 308: 240-50.
(4) Spengler M et al (2009), Biochem Biophys Res Commun 387: 278-82.
(5) Chimera Biotec Press Release, dated January 13, 2005.
(6) Malou N and Raoult D (2011), Trend in Microbiology doi:10.1016/j.tim.2011.03.004.
(7) Babu D and Muriana PM (2011), J Microbiol Methods, doi:10.1016/j.mimet.2011.05.002.