This story originally ran on June 30.
By Tony Fong
It isn't only baseball players and track stars who dope up to gain an upper hand on the playing field. Horse trainers and owners do also. But if researchers in the UK are successful, an improved biomarker-based approach to test for steroids and other banned performance-enhancing substances may be in horse racing's future.
Unlike existing methods, the approach being investigated by the scientists at Quotient Bioresearch looks not for pharmaceutical agents in a horse's system but for biological changes that could be indications of doping.
Any kind of doping test resulting from the work, described in a study published in the June issue of Proteomics, is still far off, but in an age when no sport seems immune from the threat of doping, the effort could open a new door to steroid testing.
The approach developed by the Quotient Bioresearch team uses LC-MS/MS for the targeted analysis of proteotypic peptides. Peptides produced by enzymatic hydrolysis of a protein act as quantitative surrogates. Liquid chromatography separates the peptides, which are then monitored by tandem mass spectrometry using multiple reaction monitoring.
Because protein identification and quantitation are performed separately and only a well-defined set of peptides is studied in each quantitation experiment, the approach leads to higher quality data, according to the researchers. In combination with isotopically labeled peptide internal standards, the approach also was shown to be "highly accurate and reproducible with quantitative results comparable to that of clinical analyzers," the researchers added.
And because the instrument used for the approach is a triple-quadrupole mass spectrometer, "commonly used for detection of small molecules in doping control laboratories across the world," their method can be "easily implemented" in most sports-testing regimes, they said.
Juiced Up Ponies
The study comes in the wake of a seemingly unending current of news reports linking athletes — most prominently major league baseball players and track-and-field athletes — with steroid use. Though more quietly, equine sports also is having its doping issues.
Recently, the international press has been peppered with stories of admitted or suspected equine doping. Last week, Germany's Isabell Werth, a multiple Olympic gold medalist in dressage, was suspended from competition after a 10-year-old gelding she rode tested positive for the sedative fluphenazine at an event in May. Earlier in June, trainer Matt Gingell admitted to the British Horseracing Authority to routinely injecting his horses with sodium bicarbonate mixed with water to prevent them from being fatigued, according to UK newspaper the Daily Express.
And during the same month the International Equestrian Federation announced it was investigating the stepson of its president, Princess Haya of Jordan, after a horse he rode in a race tested positive for the anabolic steroid stanozolol. The princess' husband is also being investigated after his horse tested positive for stanozolol in a different race.
Horse doping is not a recent phenomenon either. In 1933, the arrest of seven men in the Chicago area in connection to the doping of more than 200 horse races led to a full-scale federal investigation of doping in the thoroughbred industry.
What is new and continually evolving, however, are the substances being used. In their study, the researchers said that recent advances in doping, "including designer drugs and recombinant protein therapeutics (e.g. growth hormone, insulin-like growth factor-I and erythropoietin) have required changes to a doping regime previously targeted at small molecules."
In an interview with ProteoMonitor last week, Chris Barton, senior scientist at Quotient Bioresearch and the corresponding author of the Proteomics study, pointed to the BALCO steroid case as one example showing the ineffectiveness of traditional testing methods in the face of new doping methods.
[ pagebreak ]
The BALCO case involved tetrahydrogestrinone, a steroid that until 2003 could not be detected by conventional tests. Among those who were implicated in the case are major league baseball players Barry Bonds and Jason Giambi, and runner Marion Jones who eventually landed in jail for lying to two grand juries about her steroid use.
"What has happened has been described … as almost an arms race between the potential [for doping] and the techniques for the detection of doping," Barton said.
Whereas small molecule-based methods for detecting steroids look for the presence of the drugs themselves either in urine or blood, a biomarker-based strategy, such as that being developed by Quotient Bioresearch, identifies biological changes that suggest steroid use.
"Biomarkers ... are looking for the effects of a drug rather than the drug itself," Barton said, and "are a way of looking for a whole class of substances at the same time. So the concept behind biomarker screening is [that] instead of looking for one steroid, you look for the effects of any steroid."
In an e-mail to ProteoMonitor, Scot Waterman, executive director of the Racing Medication & Testing Consortium, said that he expects to see further development of new testing methods, such as biomarker-based approaches, "due to drugs hitting the market with a long duration (making detection of the parent drug difficult) or issues such as gene doping where there is no 'drug' administered at all."
The RMTC was formed in 2000 to address "issues relating to the medication and post-race testing of racehorses," and to develop "uniform national medication policy for racehorses," according to its website.
Such tests would usher in a new era of steroid testing that potentially could be filled with legal minefields. Whereas a traditional small-molecule approach is a direct and definitive indication of doping, a biomarker test is not: A change detected by a biomarker may be but is not necessarily the result of steroid use, and an absolute link between the two would need to be made.
Waterman said tests like the one being developed by Quotient Bioresearch are entering uncharted waters and "the use of these techniques will need to pass the litmus test of the court of law."
The new sophisticated doping strategies necessitate new detection methods, he said, and "the question for our regulatory bodies … will be whether we can shift from a prima facie evidence (meaning drug that was detected by the lab is evidence that the horse was carrying the drug in its system during the running of the race) to the use of what is called a 'non-analytical' positive test," Waterman said.
For their study, Barton and his co-researchers first set out to identify the most abundant proteins in horse plasma. Despite plasma being an "ideal matrix for equine doping research," they wrote, its study has been "hindered by a lack of knowledge of equine plasma's protein composition and a shortage of proven techniques for their analysis."
In an earlier study by other researchers, a 2D approach was used to identify horse plasma proteins, but Barton and his team said that the technique is "constrained in its sample-throughput capacity."
Immunological approaches such as ELISAs, they added, are complicated by the cross-species variation in protein structure in horses.
Performing a 2D LC-MS/MS peptide analysis, they identified 70 proteins in undepleted horse plasma, based on the newly annotated horse genome. Among the proteins they identified are carrier proteins, apoliproproteins, immunoglobulins, and plasma glycoproteins. They also detected abundant plasma carboxyesterase and a hypothetical protein of unknown function, "showing distant homology to mammalian beta-defensins." The list also includes proteins that other studies have reported to be regulated by doping agents in other species.
[ pagebreak ]
As in human plasma, the dynamic range of horse plasma presented problems and Quotient Bioresearch is currently working on extraction techniques to get better access to the lower-abundance proteins, Barton said.
The list of proteins, he and his colleagues said, is a 2.5-fold increase in the number of previously identified horse plasma proteins and served as a "suitable starting point" for the development of a targeted quantitative approach using LC-MS/MS.
For that, they selected proteotypic peptides from each protein for detailed analysis and multiple reaction monitoring design. Peptides that were identified in the 2D LC-MS-MS study were searched using Blast against the annotated horse genome "to identify proteotypic peptides which corresponded to a single gene product only," according to the study. Two proteotypic peptides for each protein were selected for further analysis.
Four MRM transitions were performed for each peptide and "optimal MRM transitions were chosen based on the most intense y ions in the MS/MS spectra, with preference given to product ions with m/z values above the precursor ions to increase MRM specificity," the authors said. A direct link from each MRM to its peptide target was then evaluated experimentally.
They then developed an assay based on isotope dilution mass spectrometry, using istopically labeled peptide internal standards. An experiment to establish baseline protein levels and establish the variance of each protein within a population of 40 racehorses was also performed.
Finally, Barton and his colleagues tested the utility of their method on two horses treated with a single long-acting intramuscular does of testosterone ester. Plasma testosterone levels did not return to pre-administration levels by the time of sample collection, and the majority of proteins showed no change over time courses.
The levels of peptide from clusterin and leucine-rich alpha-2-glycoprotein increased, however, confirming "the utility of the LC-MS/MS technique to study longitudinal changed induced by doping in horse plasma."
According to Barton, the use of proteomics was part of a broader effort applying other -omics technologies and methods. But, he said, "The thing we get from proteomics is a causal link. …For biomarker research, in general, I think a very important thing is a causal link between a change and the agent that causes that change."
While the paper is a proof of concept and the study didn’t have the statistical power to put forth candidate biomarkers, the researchers noted some especially interesting observations, including the identification of plasma protein carboxyesterase in horse plasma, supporting similar results from other studies and excluding it as a potential marker for doping.
In addition, the increase in clusterin and leucine-rich alpha-2-glycoprotein levels even toward the end of the time course, when testosterone levels in plasma were low but still detectable, "highlights one potential advantage of a biomarker approach to detection of doping: an increase in the window of detection of a doping event," the authors added.
Creating a biomarker-based doping test for horses may be easier than developing one for humans, Barton said, because racehorses are largely isogenic. But because changes in protein biomarkers may occur for any number of reasons, "all of which are very hard to control for or study in one experiment," larger studies are needed.
Nonetheless, the method developed by the Quotient researchers, "allows us to very quickly rule out proteins that might be of interest to us and not study them further," Barton said. "Any protein that does work well in an assay such as this, we can build more targeted assays looking at those proteins in more detail. This is the first step in that process."