Health risks and ethical dilemmas are inherent to using gene therapy for human enhancement. Sports is currently the only field where active measures against genetic enhancement, also known as gene doping, are desired. Present gene doping detection methods like nested Polymerase Chain Reaction (PCR) and the privacy invasive whole genome sequencing (WGS) lack versatility to detect the plethora of genetic differences or require large storage capacities and high processing power. Therefore, we developed a comprehensive package (ADOPE) consisting of a cheap and scalable gold nanoparticle based visual prescreening, a newly designed, produced and characterized fusion protein for targeted sequencing and a self-learning algorithm for data analysis to anticipate future gene doping development. ADOPE is versatile, efficient and reliable and can detect low copy number DNA in samples, whereby it not only facilitates gene doping detection, but also blood banks and food safety institutes around the world.
1. Gene Doping: The Threat
Gene doping is a true problem. It is the first doping that could not only find interest in the athlete population, but also in the rest of society. Thereby it is not only a global, but especially an intergenerational problem, given the possibility of (accidental) germ line infections, as well as the extension it finds in the designer baby concept. However, there are many health risks related to gene doping use, amongst which are acute immune responses upon the viral infection, as well as cancer and many unforeseen consequences.
Gene doping as a result of the desire to be perfect is a substantial dilemma, making a strong focus on gene doping and the debate about it absolutely indispensable. The video below gives a summary of our expert discussion at the University of Stirling where we addressed the eminent threat of gene doping.
Sports Coach Stirling
Due to lacking figures on actual gene doping use, unfortunately we cannot say gene doping is happening for sure. However, in this case the question should not be whether it is already happening, it should be whether we can afford not to worry. That is what we wanted to solve, because responsible innovation starts with anticipation.
Gene doping is a real problem. As the science improves the usage will expand. Testing is going to have an impact, in lots of diverse ways. This group have expertise and are willing to engage in transparent, open debate. I think @wada_ama and many others should talk to them. https://t.co/zrO6GQFSVz— Paul Dimeo (@pauldimeo2) August 30, 2018
In the meantime many (conventional) doping cases keep being revealed. Doping use is not only a risk for athlete health, but also has a big impact on his or her competitors. We interviewed Moniek Nijhuis, a former professional swimmer, who participated in many competitions including the Olympic Games, winning many medals. In the video we made, she tells about her experiences with doping and what it did to her when two years after a competition it was discovered that one of her competitors had been using doping. All these reasons sum up to ADOPE, since every athlete deserves a healthy and secure future and a fair chance.
2. Current Gene Doping Detection
Up until now no gene doping detection methods have been actively implemented. However, several research groups have been working on the topic, both with nested PCR and WGS.
2.1 Nested PCR
Last year, an Australian research group published one of the first detection method for gene doping (Wilkin, T.et al, 2017). This method relies on a technique called nested PCR. Nested PCR is PCR modified to reduce non-specific amplification due to unexpected primer binding. To do this, nested PCR requires two sets of primers that are used in two successive rounds of PCR. Here, the second primer set allows for amplification of a secondary target within the product produced in the first round.
However, gene doping consists of a plethora of possible changes in a plethora of possible genes. Currently more than 200 genes have been identified that are involved in the development of physical performance (Moran et al, 2017). Together, they form a technical challenge that requires enormous versatility in the detection method design. This is where nested PCR, requiring 4 known primers per target site, is hard to extend and a main point of improvement for gene doping detection methods.
2.2 Whole Genome Sequencing
A more extensive alternative to nested PCR is whole genome sequencing as an approach to tackle gene doping. One of the laboratories working on the use of this technique against gene doping is the group of professor Hidde Haisma at Groningen University. With the current technologies however, sequencing all genomes of the participating athletes in the 2016 Rio Olympic Games would cost up to 37 years (Jain, M. et al, 2018), making WGS into a not very realistic solution. Therefore, efficiency is an important design requirement that we improve on.
3. ADOPE: the Ultimate Solution
Our product has been shaped under the influence of many stakeholders. In this way we integrated all design requirements in our comprehensive gene doping method that has the potential to combat gene doping now and in the future. During the complete process we have been working closely together with the Dutch Doping Authority, that amongst others led us to include our prescreening method.
We also created a VR explanation of the prescreen in use to inform future users of our method.
4. ADOPE’s Impact on Lives
Strengths, Weaknesses, Opportunities and Threats (SWOT)
As a result of our flourishing inclusion process, we had much valuable input from many different backgrounds, from athletes and the Dutch Doping Authority to Computer Scientists and Ethicists. This prompted us to refine the analysis of our strengths, weaknesses, opportunities and threats (SWOT) (Hill et al, 1997) to distill our core values from our challenges. Subsequently, we integrated all values and feedback into our design requirements and we reflected on future implications and applications of our method based on our SWOT analysis. Here we describe our strengths, and how we overcome our weaknesses and threats, leaving the opportunities for our Entrepreneurship page.
In gene doping detection it is important to be able to detect many genetic differences in a plethora of performance enhancing genes. Proteins such as Zinc fingers and TALEN’s require extensive amino acid sequence design and optimalisation for each genetic difference in gene doping and thus are no feasible method for the detection of gene doping. Therefore, we decided to use a dxCas9 with flexible guide RNA libraries to find the gene doping sites effectively. In this way we do not only guarantee versatility within gene doping detection, but might also soon improve viral detection, fetal DNA screening and food safety maintenance. These are only a few examples of areas where we have identified interested parties as can be read more about on the Entrepreneurship page.
4.2 Efficient, High throughput, Limiting costs
As became apparent from a visit to the doping authority, initial development costs are not the big problem. It is the costs that are inherent to the up-scaling that matter to them. Therefore, we made sure our method is efficient, low cost, and high throughput. Professor Hagan Bayley from Oxford University, cofounder of Oxford Nanopore Technologies, pointed us at the value of barcoding and subsequent multiplexing to enhance the sequencing efficiency. Also, we decided to add our gold nanoparticle prescreen based on the visit to the Dutch Doping Authority. In this way we reduce the amount of samples that will need sequencing, reducing the costs. Lastly, our fusion protein makes targeted sequencing possible that not only significantly speeds up the analysis, but also greatly reduces the amount of data that needs to be stored.
4.3 Accuracy and Reliability
Limiting the risk for false positives has been one of our most prominent focuses, because false accusations can ruin athlete’s lives. Therefore, we adhere much value to the build-in double verification. After a prescreen, we use sequence verification to provide athletes with more direct proof. As the Oxford Nanopore MinION device that we use has a misread percentage, we simulated actual read-outs and made sure the separation between gene doping and misread normal DNA is clearly present. In this way we have an accurate and reliable gene doping detection method.
4.4 Privacy in Genomic Data Analysis
From our surveys it appeared that people generally do not care about genomic privacy. Only 22.6% of the general Dutch population thinks we should be utterly careful about the genomic privacy of athletes based on 181 respondents. We however do believe that guaranteeing athletes’ privacy in gene doping detection is one of the most important values to take into consideration within our design.
So, why is genomic privacy important? From one of our initial talks with Professor Dimeo it became apparent that there is not only a threat of hacking attempts on databases. In addition, all the people that work in the regulatory cycle and get to see the results give rise to an opportunity for blackmailing athletes, a real risk that Professor Dimeo is afraid of.
The fear of genome database hacking is not fiction. Scandals are already happening. Big companies known for their ancestry analyses for example have their databases hacked, leaving 92 million people with their genetic code up for grabs (Shaban, June 5 2018).
Why should people be afraid of hacking? Apart from the fact that companies can earn much money with your genetic code, in time your genetic information can be used against you. An example is in health and life insurances. If this continues, in many countries only the good risks can still have a cheap insurance once their health insurer knows their genetic disease predispositions. The same applied for job and sponsoring offers and loans and mortgages. This can lead to large social problems and possibly unethical situations that we will then need to find other solutions for. For our method at least it means we highly value athlete privacy.
Current data storage and safety
Recently it became apparent that around the 2018 Olympic Winter Games in PyeongChang, the medical information of 250 athletes from 30 countries who have a dispensation to use certain medication was hacked (NOS, October 5 2018). The year before, the same hackers had already obtained the information of several other athletes. We make sure that no more data is stored than is absolutely necessary, thereby guaranteeing that no personal information will be leaked by a hack.
As a team, we believe biosafety is more than biocontainment, which we achieve by having a cell free method. Additionally we created an infographic for the Dutch Ministery of Health, Welfare and Sport on how safety has been integrated throughout our project, from start to finish.
-  Ginn SL, Amaya AK, Alexander IE, Edelstein M, Abedi MR. Gene therapy clinical trials worldwide to 2017: An update. J Gene Med. 2018;20:e3015. https://doi.org/10.1002/jgm.3015.
-  Gao, G. (2004). Erythropoietin gene therapy leads to autoimmune anemia in macaques. Blood 103:3300-3302. doi: https://doi.org/10.1182/blood-2003-11-3852.
-  World Anti-Doping Agency (2018). Accredited Laboratories. Retrieved on 23-9-2018 from: https://www.wada-ama.org/en/what-we-do/science-medical/laboratories/accredited-laboratories.
-  Alkilany, A.M. et al (2010). Toxicity and cellular uptake of gold nanoparticles: what we have learned so far? J Nanopart Res. Sep; 12(7):2313-2333. doi:10.1007/s11051-010-9911-8.
-  Singh, K. et al (2018, June 6). Security breach at MyHeritage website leaks details of over 92 million users. Reuters Cyber Risk. Retrieved on 23-9-2018 from: https://www.reuters.com/article/us-myheritage-privacy/security-breach-at-myheritage-website-leaks-details-of-over-92-million-users-idUSKCN1J1308
-  Glaxosmithkline (2018, July 25). GSK and 23andMe sign agreement to leverage genetic insights for the development of novel medicines. Retrieved on 23-9-2018 from: https://www.gsk.com/en-gb/media/press-releases/gsk-and-23andme-sign-agreement-to-leverage-genetic-insights-for-the-development-of-novel-medicines/.
-  Tøndel, C. et al (2012). Safety and Complications of Percutaneous Kidney Biopsies in 715 Children and 8573 Adults in Norway 1988–2010. Clin J Am Soc Nephrol. 7(10): 1591–1597. doi: 10.2215/CJN.02150212.
-  The European Parliament and the Council of the European Union (2016, April 27). Regulation (EU) 2016/679 of the European Parliament and of the Council on the Protection of Natural Persons with Regard to the Processing of Personal Data and on the Free Movement of such Data, and Repealing Directive 95/46/EC (General Data Protection Regulation). Official Journal of the European Union; retrieved on 23-9-2018 from: https://eur-lex.europa.eu/eli/reg/2016/679/oj.
-  Denby, B, & Schofield, D. (1999). Role of virtual reality in safety training of mine personnel. Mining Engineering (Littleton, Colorado): 51; 10: 59-64.
Overcoming Weaknesses and Threats
Doping knows a long history of development. As a rat race, detection methods have for a long time been following up on developments in doping use. We believe we should not be naïve in the sense that this time we are first, since there are certainly clues that gene doping is already happening. However, gene doping is different from more conventional doping in the sense that it is highly versatile. Therefore, we thought of a method to anticipate further developments within gene doping. This we do through our self-learning algorithm for gene doping sequence classification as is described on our software tool page. We challenged engineers at the Cyber Security Week to hack our method in order to become as strong as we can be against gene doping. In this way we are the first to openly use collaborative strengths in the race against doping.
We realize that inventive and knowledgeable people could potentially mess with our system, as with almost every detection method. Therefore, we asked all stakeholders we interviewed how they would circumvent our method in order to map our weaknesses. Most directed us into RNA injections in the samples. This might indeed complicate our targeted sequencing. However, then they would need to have the specific knowledge of the actual guide RNA’s that are used, they must have access to the blood sample afterwards and this will not circumvent our prescreen. When the prescreen is positive and the sequence verification negative, a second blood sample might be requested, which fits in with the standard practice in doping testing.
4.7 Sensitivity, Privacy Invasion and Minimal Invasivity
With our sequencing method we have been able to detect a single molecule of EPO doping DNA form a 3.45 nM sample. Further protocol optimization will thereby make our analysis even more sensitive. Additionally, with some optimization, our prescreen can detect up to 29 fM of doping DNA (Baetsen-Young, et al 2018), making the complete approach into a very sensitive method.
Furthermore, we made an extensive model of the response of the human body to different administration methods of gene doping. From this model we derived the most likely ways in which athletes would be gene doping and established a detection window based on our sensitivity. From the analysis, we found that we will always be able to detect gene doping. Therefore, the gene doping screening could be uptaken in the regular doping testing programs and will not entail further privacy invasion of athletes.
Lastly, we have been focussing on blood for the extraction of cell free DNA. The professional athletes we spoke to however indicated that urine testing is the predominant form of testing. Therefore, we looked into urine based cell free DNA testing. According to Casadio et al 2017, cell free DNA concentrations of up to 138 ng/µL can be found in urine samples. We are therefore confident that in the near future our method can be extended to urine samples with an adapted sample preparation.
- Baetsen-Young, A.M. et al (2018). Direct colorimetric detection of unamplified pathogen DNA by dextrin-capped gold nanoparticles. Biosensors and Bioelectronics. Volume 101, Pages 29-36.
- Casadio, V., Salvi, S., Martignano, F., Gunelli, R., Ravaioli, S., Calistri, D. Cell-Free DNA Integrity Analysis in Urine Samples. J. Vis. Exp. (119), e55049, doi:10.3791/55049 (2017).
- Hill, T. and Westbrook, R. (1997). SWOT Analysis: It's Time for a Product Recall. Long Range Planning. 30 (1): 46–52. doi:10.1016/S0024-6301(96)00095-7.
- Jain, M. et al (2018). Nanopore sequencing and assembly of a human genome with ultra-long reads. Nature Biotechnology, volume 36, pages 338–345.
- Moran, C.N. & Pitsiladis, Y. (2017). Tour de France Champions born or made: where do we take the genetics of performance? Journal of Sports Sciences, 35 (14), pp. 1411-1419.
- NOS (2018, October 5). Medische Gegevens Nederlandse Atleten Gehackt. Retrieved on October 5 2018 from: https://nos.nl/artikel/2253542-medische-gegevens-nederlandse-atleten-gehackt.html.
- Shaban, H. (2018, June 5). DNA testing service MyHeritage says 92 million customer email addresses were exposed. The Washington Post. Retrieved on October 5 2018 from: https://www.washingtonpost.com/news/the-switch/wp/2018/06/05/ancestry-service-myheritage-says-92-million-customer-email-addresses-were-exposed/?noredirect=on&utm_term=.96513a4b72f4.
- Wilkin, T. et al. (2017). Equine performance genes and the future of doping in horseracing. Drug Testing and Analysis: Vol. 9, Iss. 9, p. 1456-1471. https://doi.org/10.1002/dta.2198.