Gene doping is the misuse of gene therapy to enhance athletic performance. It has recently been recognised as a potential threat and subsequently been prohibited by the World Anti-Doping Agency. Despite concerns with safety and efficacy of gene therapy, the technology is progressing steadily. Many of the genes/proteins which are involved in determining key components of athletic performance have been identified. Naturally occurring mutations in humans as well as gene-transfer experiments in adult animals have shown that altered expression of these genes does indeed affect physical performance. For athletes, however, the gains in performance must be weighed against the health risks associated with the gene-transfer process, whereas the detection of such practices will provide new challenges for the anti-doping authorities.
In the great construction that is the human body our genes could easily be described as the architect’s plans. Encoded in the nuclei of all cells are the genetic blueprints for the manufacture of the proteins that constitute all the tissues and organs of the body. This blueprint takes the form of an array of nucleotides that make up DNA. It is our ability to subtly redraw these plans by manipulating DNA that has, in recent years, formed the basis of new strategies for treating a number of diseases, a process known as gene therapy. In contrast with its use as a medicinal tool, gene ‘doping’ is where DNA is manipulated in otherwise healthy individuals for the purposes of improving certain physiological functions and ultimately athletic performance. The principles underlying gene doping and gene therapy are essentially the same and thus, before going on to address gene doping specifically, it is necessary to outline the concepts involved in the development of gene therapy as a legitimate therapeutic tool.
As our understanding of molecular biology has increased, the possibility of manipulating our genes to correct genetic disorders has begun to be realized. Genetic disorders are those resulting from mutations, or errors in the sequence of a gene. Some mutations are inconsequential to the function of a protein, but others can result in production of a non-functional protein or one with aberrant function. Cystic fibrosis is an example of a hereditary condition which results from mutations in the CFTR (cystic fibrosis transmembrane conductance regulator) gene. In other diseases, such as cancers, mutations can accumulate during the life of the organism due to environmental mutagens or simply errors made in duplicating DNA during the process of cell proliferation.
The best candidate diseases for gene therapy are monogeneic diseases, those caused by defects in a single gene. The cure is to insert a correct copy of the gene thus restoring the ability to produce a functional protein. This strategy may not be suitable for diseases resulting from proteins with aberrant function or deregulated expression. In these cases, the strategy would be to deliver a gene encoding a protein that can specifically inhibit gene expression, protein expression or protein function.
The process of gene expression can be regulated at several levels: the transcription of a gene from DNA to pre-mRNA; the processing of the pre-mRNA to mature mRNA; the translation of mRNA to protein; and the stability of the mRNA or protein. In many cases, the amount of protein can be regulated by inhibiting or stimulating the gene transcription to produce more or less mRNA. In theory, this can be achieved easily by introducing extra copies of the gene or by altering the levels of the factors that regulate the transcription of a particular gene. These are transcription factors; proteins that bind to regulatory sequences in the DNA and interact with the enzymes of the transcription machinery, thereby activating or inhibiting gene transcription. Other regulatory factors are involved in the process of translation, whereas post-translational modifications of a protein (proteolytic cleavage, glycosylation, phosphorylation) regulate its activity and stability.
Currently the major obstacle to gene therapy is inefficient delivery and expression of the gene of interest into the target tissue. Therefore the focus of much research is the optimization of gene-delivery methods. It is no surprise that the most efficient method of gene delivery to mammalian cells involves the use of viruses, as their natural function is precisely to infect cells and utilize the protein and DNA synthesis machinery of the host to replicate. Most viruses used for gene therapy are modified to make them replication incompetent. There are several categories of viral vectors that are currently being used, each with distinct advantages and disadvantages .
Non-viral vectors include plasmid DNA (pure DNA containing the gene of interest and sequences required for expression) and complexes formed by DNA and synthetic vectors such as liposomes (lipid vesicles), cationic polymers and cell-penetrating peptides. The disadvantage in their use is the much lower efficiencies of transfection compared with viral vectors. A variety of methods are being developed to improve delivery. These include application of electric fields to the site of injection (electroporation), ultrasound, laser magnetic fields, mechanical massage and pressurized vascular delivery where local occlusion of blood vessels increases pressure in the vascular bed at the site of injection. The cause of improved transfection efficiency using these techniques is unclear but it is probably due to increased permeability of cell membranes .
Clinical trials are being conducted for diseases such as cystic fibrosis, Duchenne muscular dystrophy, immunological blood disorders, cardiac ischaemia and cancer. The results have on the whole been disappointing, as the clinical responses expected from animal models have not been consistently observed in humans. Gene therapy, however, has been used successfully to treat immunodeficient children and adults by ex vivo treatment of haemato-poietic stem cells [1,3], the precursors of mature lymphocytes and of other blood cells that are found in the bone marrow. These studies also highlighted one of the major risks of viral-based gene therapy, insertional mutagenesis. In 2 out of 14 children suffering from X-linked severe combined immuno-deficiency, insertion of the virus near a proto-oncogene locus led to development of T-cell proliferative disease, a type of leukaemia . Another risk of viral-based gene therapy is an immune response to the viral vector, and it was a severe immune system reaction that caused the death of 19-year-old Jesse Gelsinger in a 1999 clinical trial. When less severe, the immune response can still limit the number of administrations and, ultimately, the efficacy of treatment. Generation of autoimmune responses against self-antigens is also a concern, particularly in trials that involve the use of immunostimulatory molecules. Injection of vectors expressing T-cell stimulatory molecules into melanoma lesions to boost response against tumour antigens resulted in development of autoimmune vitiligo in 12–14% of patients, probably as a result of a reaction to antigens also present in normal melanocytes .
Recent developments suggest that it may be possible to take advantage of a natural DNA repair process called homologous recombination to replace sequences of DNA, and thus substitute existing genes with the desired variant . This process has two advantages over viral and non-viral gene therapy: gene expression is controlled by endogenous regulatory sequences and there is no risk of insertional mutagenesis. Homologous recombination occurs when chromosomes align during mitosis and meiosis. Identical sequences of DNA can crossover from one molecule to the next and the DNA segments can be exchanged between chromosomes. Engineered gene constructs can be introduced into the cell and recombination takes place within the homologous sequences. The end result is that a single targeted locus has been replaced with the engineered construct.
Despite the setbacks that have occurred in gene therapy trials (Table 1), the fact that more than a thousand different clinical trials have taken place with a relatively small number of reported complications suggests that risks, though potentially severe, may be sufficiently infrequent to justify use of gene therapy where no alternative is available. It is noteworthy, however, that there may be an association between lack of observed complications and limited efficacy of delivery. Safer and more efficient delivery vectors and methods are constantly being developed, and some success has been obtained in large animals, prompting optimism in the field.
As with many medical advances in the past, gene therapy has not gone unnoticed by sections of the athletic and body-building communities looking to gain an unfair advantage. The WADA (World Anti-Doping Agency) defines gene doping as the “non-therapeutic use of genes, genetic elements and/or cells that have the capacity to enhance athletic performance”. In 2003 WADA amended its Prohibited List of Substances and Methods to include gene doping. What genes might athletes be interested in manipulating? To understand this we need to consider the key physiological mechanisms limiting different athletic events and the key proteins involved. The demands of marathon running differ markedly from those of sprinting, jumping, throwing and weightlifting. The last four events require rapid development of high muscle forces (i.e. generation of explosive power). Endurance events, on the other hand, require the ability to sustain a sub-maximal level of power output for a prolonged period of time. Although in theory every protein that is known to affect physical performance and has been cloned can be a target for genetic manipulation, this chapter will focus on a few of the key genes/proteins that are probable candidates for gene doping, and in which there has been considerable recent interest (Figure 1). Other reviews have recently been published which may be of interest to the reader [6–8].
Gene doping for increased strength and power
All other things being equal, muscle force is determined by its size, specifically its cross-sectional area. Athletes abusing drugs have traditionally used anabolic steroids, analogues of the male hormone testosterone, to increase muscle mass. More recently, GH (growth hormone) and IGF-1 (insulin-like growth factor 1) have also been used. For the athlete, the use of gene doping provides a considerable advantage over endogenous administration of synthetic drugs. This is because of the simple fact that manipulation of a gene will result in the production (or over-production) of a given hormone that is equal in sequence to its endogenous counterpart, making detection extremely difficult. Skeletal muscle is an attractive target for gene therapy because of the large size of the muscle cells and the fact that muscle is a non-proliferating tissue, leading to long-lived gene expression. In addition, skeletal muscle is particularly amenable to gene delivery using plasmid DNA vectors which are low risk, can be easily quality controlled, and can be produced in large amounts at a relatively low cost.
IGF-1 is a 70-amino-acid polypeptide synthesized primarily in the liver under the control of GH. The GH/IGF-1 axis is extremely important in regulating postnatal growth and development. In addition to the liver, other tissues, including skeletal muscle, can produce IGF-1. The IGF-1 gene comprises six exons and a process of alternative splicing at the 5′ and 3′ ends can generate different isoforms. Evidence from viral gene transfer studies in mice have shown that the two murine 3′ splice variants IGF-1Ea and IGF-1Eb (also termed mechano-growth factor or MGF) can induce significant local muscle hypertrophy in young mice, but only IGF-1Ea works on adult mice . Transgenic approaches have also confirmed the potency of IGF-1Ea as an anabolic agent. Use of a myosin light chain promoter restricting overexpression of IGF-1Ea to muscle tissue, leads to significant muscle hypertrophy in the affected animals . Interestingly, when viral IGF-1Ea gene transfer is combined with high-resistance exercise in adult animals, significantly greater increases in muscle size and strength occur than when gene transfer or resistance exercise are considered separately  (Figure 2). Although such gene studies have not been performed in human beings, the mRNA for IGF-1Ea and MGF have been shown to be expressed in skeletal muscle and are increased following muscle-building, high-resistance, strength-training exercise .
IGF-1 acts through the PKB (Akt; protein kinase B) signalling pathway. In vivo transfection of myofibres with a Ras double mutant construct that selectively activates PKB resulted in marked hypertrophy of these cells . Thus hypertrophy can be achieved by overexpressing IGF-1 itself or activating proteins in its signalling pathway.
In contrast with IGF-1 and anabolic steroids, which stimulate protein synthesis, myostatin, a member of the TGF-Β (transforming growth factor Β) family of growth factors, acts as a negative regulator of muscle growth. The significance of this protein came to light following work on the Belgian Blue and Piedmontese breeds of cattle which exhibit a ‘double-muscled’, highly hypertrophied phenotype. These animals possess a mutation in their myostatin gene which results in production of an inactive protein. Transgenic mice specifically bred to knockout this gene have a highly hypertrophied phenotype that is remarkably similar to that of the IGF-1-overexpressing mice, even though the mechanisms of action of these two proteins are fundamentally different . Myostatin inhibits satellite cell proliferation, whereas IGF-1 does not. Gene doping strategies would thus aim to inhibit production of myostatin or interfere with the function of the endogenous protein.
Manipulation of fibre type
In addition to force of contraction, the other determinant of power is the speed at which a muscle shortens. It is the myosin heavy chain isoform, the molecular motor contained in each fibre, that primarily determines speed of shortening. Human muscle fibres may contain the slow-contracting MHC-I (where MHC is myosin heavy chain), the fast-contracting MHC-IIA or even faster contracting MHC-IIX isoforms. In sprinters it is the MHC-IIa isoform which dominates, whereas in marathon runners MHC-I predominates. It is possible to alter MHC expression through modifying the neural input delivered to a muscle or its mechanical loading status. These physiological manipulations ultimately affect the expression of specific transcription factors which regulate expression of genes associated with fast or slow myosins. Understanding of the regulation of myosin expression opens the possibility for these transcription factors to be genetically altered potentially allowing athletes to alter their natural endowment for a given population of fast and slow fibres. For example, altering the activation of the transcription factor NFAT (nuclear factor of activated T cells) in mice, by plasmid injection of either a specific peptide inhibitor or a constitutively active isoform, can alter MHC mRNA expression in fast and slow muscles, overriding neural input .
Gene doping for increased endurance
In contrast with sprinting, which is primarily determined and limited by the mechanical properties and anaerobic metabolism, endurance performance requires a high oxidative capacity within the muscle and good cardiovascular/respiratory function. There is a constant demand for oxygen to be supplied to the mitochondria so that it can act as the final electron acceptor in the electron transport chain for the generation of ATP. Delivery of oxygen to working muscles occurs through a complex interaction of the cardiovascular and respiratory systems. Ultimately oxygen is transported in the blood primarily bound to haemoglobin contained in red blood cells.
Athletes have long used training at altitude to stimulate the bone marrow to produce more red blood cells and thus increase the oxygen-carrying capacity of the blood. In times of hypoxia, EPO promotes red blood cell production by stimulating CFU-E (erythroid colony-forming unit) progenitor cells to proliferate and differentiate into mature erythrocytes. Recombinant human EPO has been available for human use since 1989. Not surprisingly, this drug was readily identified by endurance athletes for its ergogenic potential. Randomized controlled trials provided evidence of its efficacy in improving endurance performance , and it has subsequently gained particular notoriety among endurance athletes, most notably cyclists in the Tour de France. As methods for the detection of EPO become more refined, athletes may well be tempted to use gene doping techniques to increase its endogenous expression. Following a single intramuscular injection of a vector containing the EPO gene under control of a rapamycin-sensitive promoter, regulated, inducible expression of EPO has been obtained in macaques for a period of six years . In this study, two problems previously identified in EPO gene therapy were overcome: polycytaemia (life-threatening increases in blood viscosity owing to high erythrocyte numbers) and anaemia (owing to development of autoimmunity to endogenous EPO).
VEGF (vascular endothelial growth factor)
In addition to a better oxygen-carrying capability, an optimized vascular system would also promote oxygen and nutrient delivery to working muscles. VEGF is an important factor in stimulating the formation of new blood vessels. Phase I and II clinical trials have already been undertaken using VEGF viral gene therapy techniques to induce angiogenesis in patients with ischaemic heart disease .
PGC1α [PPAR (peroxisome-proliferator-activated receptor) γ co-activator] and PPARδ
In addition to oxygen delivery, metabolic characteristics of muscle fibres are important for endurance. Studies in mice show that transgenic animals for either PGC1α or PPARδ have an increase in type I fibres as assessed by oxidative enzyme expression, muscle colour (which reflects myoglobin content), sarcomeric protein expression and mitochondrial content [19,20]. Importantly, these mice outperform littermate controls in running endurance and muscle fatigue resistance. The question remains whether elevated expression of these factors in adult animals, as opposed to during embryonic development, would have similar effects.
Endorphins, endurance and pain
The death of Tommy Simpson during the Tour de France in 1967 brought to the fore the use of amphetamines as a tool to dull the sensations of pain associated with fatigue. The sensation of pain is an adaptive mechanism that helps us to avoid noxious stimuli, prevents us from pushing ourselves too far and protects an injury while healing takes place. The nature of physical activity required during high-level competitive sports often means that athletes push their bodies through barriers of discomfort not usually tolerated. They may also live with chronic pain. The endogenous opioid peptides, such as Β-endorphins, affect both local neurons and the pain-modulating circuitry in the central nervous system leading to pain relief.
Strenuous exercise causes a release of Β-endorphin from the pituitary into the bloodstream. Motor neurons are able to synthesize Β-endorphin and receptors for this substance are present on the membrane of skeletal muscle fibres. In addition to pain relief, a direct effect of Β-endorphin on skeletal muscle has been proposed since this substance has been shown to reduce fatiguability and increase glucose consumption in isolated muscle preparations [20a]. Thus this factor could have a double effect in terms of modulating physical performance: the first in alleviating pain and helping an athlete push through sensations of discomfort, the second in enhancing muscle endurance. The administration of endogenous opioids using gene therapy approaches is being investigated to provide relief for chronic pain, such as might occur after trauma or in advanced stages of cancer. This has the advantage that the peptides could be synthesized in the vicinity or inside target cells and thus avoid some of the complications associated with long-term administration of opioids. In addition to their analgesic effect, endorphins may have roles in the response to stress, determining mood, and regulating the release of hormones from the pituitary gland, notably GH and the gonadotropin hormones.
Can gene doping be detected?
The major attraction of gene doping for sport is the theoretical difficulty in distinguishing the gene product of the exogenous gene from the product of its endogenous counterpart. Nevertheless there is hope for the authorities that detection methods will not lag behind the adoption of gene therapy technologies for doping. Detection strategies have been reviewed recently  and include detection of the vectors used for gene delivery, evidence of an immune response to these vectors, alteration of various parameters (level of mRNA and protein, relative mRNA or protein isoform expression, post-translational modifications) which may differ between transcripts and proteins produced from endogenous and exogenous genes. A critical factor in gene delivery is the size of the DNA that can be introduced into a vector. Some of the sequences that regulate gene expression can be very distant from the protein-coding regions of the DNA molecule and may not even be known. Thus the manipulation of a gene for insertion into a vector is likely to affect some of the regulatory elements and its ‘natural’ expression may never be accurately mimicked at all levels.
Among the most promising applications is the use of complex molecular signatures to identify normal, exercise-regulated and aberrant levels of a given biomarker. For example, external hormone administration or disease states disturb the composition of blood and urine. Application of high-throughput analyses of protein spectra in these body fluids will allow identification of specific profiles indicative, at minimum, of disturbance in normal physio-logy. The challenge will be to relate these disturbances to gene doping. At present, the application of this technique is mainly in cancer diagnosis, but advances in this area will aid the fight against gene doping. In this regard, it is encouraging to note that the pattern of post-translational modifications of proteins or production of specific protein isoforms may require physiological signals that are bypassed by overexpression of the transgene. In addition, the processing of the gene may differ in different tissues. Thus targeting the transgene to skeletal muscle might result in production of protein products distinct from those occurring naturally if the major source of protein is another organ (for example, the liver). Last but not least, it is important to note that it may not be possible for gene-doped athletes to arrest gene expression, unless they use inducible vectors. In this way, supraphysiological levels of protein or the use of drugs required for inducible gene expression may be sufficient to indicate doping. From the point of view of the gene-doped athlete, it is worthwhile to note that deregulated protein expression is a risk not only for detection but also for continuing health. The body is a fine-tuned machine and the deregulation of gene expression in a healthy person may lead to long-term complications even if it is beneficial in the short-term. For example, IGF-1 may be an anabolic substance, but it is also a potent mitogen thus increasing the risk of tumour development. In transgenic mice, chronic EPO overexpression leads to multiple organ degeneration and reduced life expectancy . In people suffering from disease or poor health, the benefits outweigh the risks, but the same risk/benefit calculation does not necessarily apply to a healthy individual.
A potential grey area is in the development of gene therapy for the treatment of sporting injuries. Here the aim is to increase the expression of important growth factors in muscle, bone, tendon and cartilage to facilitate the repair processes. A doping issue might arise if the athlete were to carry a permanent marker of treatment which might give a competitive advantage over their peers as a result of the treatment.
Gene polymorphisms and ‘natural gene dopers’
Not all men are created equal, at least not genetically! We all have different physiognomic traits that are genetically determined. It follows therefore that those participating in competitive sport must possess genetically determined traits that confer advantages in physical performance relative to the general population. Indeed the HERITAGE family study provides evidence of a significant genetic component to the ability of individuals to respond to a programme of aerobic training .
There is also evidence of the existence of individuals with naturally occurring gene mutations that confer competitive advantages. Familial erythrocytosis is a condition associated with high haematocrit (red blood cell count) and haemoglobin levels, as well as low serum EPO. The condition results from an EPO receptor mutation that inactivates a negative regulatory region, thereby conferring hypersensitivity of the receptor to EPO. The genetic mutation was detected in a Finnish family which had several members involved in competitive endurance sports, notably Olympic cross-country skier and gold medallist Eero Mantyranta .
More recently a myostatin loss-of-function mutation was reported in a male child who had supranormal muscle mass from birth . At 4½ years of age, he possessed significant skeletal muscle hypertrophy and strength relative to age-matched controls. Other members of his family were exceptionally strong according to clinical records, but they were not available for genetic testing.
Gene polymorphisms rather than mutations may also be associated with increased performance. The difference between a mutation and a polymorphism is basically quantitative. A mutation is considered to be a variation in sequence that is present in less than 1% of the population, whereas a polymorphism is a more common occurrence. ACE (angiotensin-converting enzyme) and ACTN-3 (actinin 3) are genes with polymorphisms that have been correlated with distinct patterns of performance.
ACE catalyses the conversion of angiotensin I into angiotensin II. Angiotensin II is involved in regulating blood pressure primarily by causing vasoconstriction. It is also a potent growth factor for cardiac and vascular tissues. The ACE gene polymorphism is a 287 base pair sequence and the gene copies (alleles) of the ACE gene are classified as insertion (I) or deletion (D) depending on whether they contain this sequence. Humans have two alleles of each gene, one inherited from each parent. There are three possible allelic combinations of the ACE gene: II, DD and ID. Several studies have suggested a link between occurrence of the I allele with endurance and the D allele with strength, but null correlations have also been found . There are also two alleles of ACTN-3: R and R577X or X for short. The R allele results in production of actinin whereas the X does not. Actinin is found only in fast fibres, which are associated with the generation of explosive power. Again the studies showed a correlation between the presence of the R allele and the modality of exercise in elite athletes: 50% of sprinters are RR, 45% are RX and the remaining 5% are XX. Among endurance athletes 31% are RR, 40% are RX and 24% are XX. All studies conducted so far on ACTN-3 and ACE have been correlational in nature, but there is no evidence showing an irrefutable cause/effect relationship between the presence of an allele and performance gains .
As our knowledge of the molecular basis of performance and of the technologies involved in legitimate gene therapy progresses, so do the chances of these techniques being abused by athletes. Whether gene doping is currently practised is not known. The challenge to the anti-doping authorities will be to develop methods which are capable of detecting such practices. This will not be straightforward.
• Gene therapy technology has the potential to be abused for performance gains by athletes.
• Objective clinical responses have rarely been observed in clinical trials.
• There are significant risks associated with viral gene therapy and thus gene doping
• Genes have been identified that affect skeletal muscle size, metabolism and contractile properties
• The cardiovascular/respiratory systems can also be genetically manipulated.
• Development of gene doping detection methods will not be straightforward.
We would like to thank WADA.
- © The Authors Journal compilation © 2008 Biochemical Society