Monday, May 5, 2014

The human machine: replacing damaged components

The previous post in this series can be found here.

The major theme of my 'human machine' series of posts has been that we are, as the name suggests, machines; explicable in basic mechanical terms. Sure, we are incredibly sophisticated biological machines, but machines nonetheless. So, like any machine, there is theoretically nothing stopping us from being able to play about with our fundamental components to suit our own ends. This is the oft feared spectre of 'genetic modification' that has been trotted out in countless works of science fiction, inexorably linked to concepts of eugenics and Frankenstein-style abominations. Clearly genetic modification of both humans and other organisms is closely tied to issues of ethics, and biosafety, and must obviously continue to be thoroughly debated and assessed at all stages, but in principle there is no mechanistic difference between human-driven genetic modification and the mutations that arise spontaneously in nature. The benefit of human-driven modification, however, is that it has foresight and purpose, unlike the randomness of nature. As long as that purpose is for a common good and is morally defensible, then in my eyes such intervention is a good thing.

One fairly obvious beneficial outcome of genetic modification is in the curing of various genetic disorders. Many human diseases are the result of defective genes that can manifest symptoms at varying times of life. Some genetic disorders are the result of mutations that cause a defect in a product protein, others are the complete loss of a gene, and some are caused by abnormal levels of gene activity - either too much or too little.  A potential means to cure such disorders is to correct the problematic gene within all of the affected tissue. The most efficient means to do that would be to correct it very early in development, since if you corrected it in the initial embryo then it would be retained in all of the cells that subsequently develop from that embryo. This is currently way beyond our technical limitations for several reasons. Firstly, we don't routinely screen embryos for genetic abnormalities and so don't know which ones might need treatment. Secondly, the margin for error in this kind of gene therapy is incredibly narrow as you have to ensure that every single cell that the person has for the rest of their life will not be adversely affected by what you do to the embryonic cells in this early stage - we're not there yet. Thirdly, our genetic technology is not yet sophisticated enough to allow us to remove a damaged gene and replace it with a healthy one in an already growing embryo - the best we can do it stick in the healthy gene alongside the defective one and hope it does the job. There is certainly no fundamental reason why our technology could not one day reach the stage where this kind of procedure is feasible, but we are a long way off yet.

So, for the time being what can we do? Well instead of treating the body at the embryonic stage, the next best approach is to treat specifically the affected cells later on in life.  This involves identifying the problematic gene and then using a delivery method to insert the correct gene into whatever tissues manifest the disease, preferably permanently. This is broadly known as gene therapy, and is one of the most promising current fields of 'personalised' medicine.  

From humble beginnings

The idea of gene therapy has been around for quite a while now, but the clinical side of the field got off to a tragic start in 1999 with the death of 18 year old Jesse Gelsinger during a trial for gene therapy to treat his ornithine transcarbamylase deficiency. The idea behind the treatment was fairly simple, the disorder was the result of a missing gene encoding an enzyme in ammonia metabolism so Jesse was given a virus engineered to return the gene to his liver cells. Unfortunately the specific virus used, a form of adenovirus, had the severe side effect of causing a massive immune response that fatally damaged Jesse's already struggling liver. Just four days after receiving treatment Jesse died in the full glare of the media spotlight and the field of gene therapy hit what was to be the first of a number of severe stumbling blocks.

This case highlighted the difficulties in developing safe delivery systems for gene therapy. Viruses are inherently good at getting foreign genes into cells - it's what they've evolved to do - but they come with associated dangers. What's more, even if you manage to get the correct gene into the correct cells without any problems, there can still be dramatic unforeseen problems. This was brought to prominent attention in 2003, when gene therapy for severe combined immunodeficiency (SCID) caused leukaemia in a group of French children. SCID is a disorder of immune cells that essentially leaves the patient with no effective immune system. The gene therapy in this case was designed to insert a working copy of a defective gene, IL2RG, into the children's immune cells, thereby rendering them functional and restoring immune activity. Unfortunately, when the IL2RG was inserted into the genomes of these cells, it got positioned next to another gene, LMO2, and activated it. LMO2 is involved in the development of immune cells as is not normally active in fully developed cells. Its activation by IL2RG cause the cells to begin dividing in an uncontrolled manner, thereby causing leukaemia. It was a sobering reminder to the field that current gene therapy techniques can be a bit of a shot-gun approach.

To promising results

Whilst cases such as these are both tragic on a personal scale and damaging on a scientific level, they are nonetheless informative, and so the field has continued to edge forward with ever improving approaches. The French children, whilst suffering from leukaemia, did indeed have their immune function restored, and so in that sense the treatment was a partial success. We are now at the stage where exciting, apparently safe gene therapy is becoming a clinical reality. The most recent success story was published earlier this year in a Lancet article from Robert MacLaren's group in Oxford. This study aimed to treat patients with choroideremia, a degenerative eye disease that generally causes total blindness by middle age. Choroideremia is causes by a defect in the protein REP1, encoded by the gene CHM. By using a virus engineered to specifically deliver the corrected gene to the retinal cells without eliciting a strong immune response, the trial was able to significantly improve the light sensitivity of all of the treated patients. Some were even able to read letters on a standard eye chart, and showed significant improvements in the structure of the affected tissues.

Degenerative eye disorders are rich hunting for gene therapists due to the fact that they are usually caused by problems with a single gene, and only affect one specific tissue - the eye. Unsurprising, then, that another recent success story has come in the form of a treatment for Leber's congenital amaurosis, which also results in blindness. In some cases this is caused by a defect in  the RPE65 gene, which leads to a deficiency in vitamin A within cells of the retina. Insertion of normal RPE65 into the retina has successfully increased light sensitivity, and sometimes vision clarity, in a number of patients in clinical trials.

Both of the above trials have used recombinant 'adeno-associated viruses' to deliver the target gene. The advantage of this virus is that the gene doesn't actually get incorporated into the host genome and so there is no chance of the same kind of off-target effects experienced by the French SCID children. The downside is that this isn't a permanent fix. Because the DNA exists within the target cells but not in their genome, it can be lost over subsequent rounds of cell division and so repeated treatment might be required. That's not to say that it isn't necessarily long-lived, it can last years, but it's not permanent. A better solution would be to have a way of integrating the gene into the genome without affecting other genes.

This is difficult because the way to do this is by using a type of virus called a retrovirus that fully inserts its genome into that of the target cell, but unfortunately has a tendency to do this at exactly the points where other genes are present. This is because viruses want to replicate quickly, and they will have a better chance of doing that if their genes are in a highly active area of the genome. It has therefore taken a lot of research to come up with a retroviral vector that is safe to use in human cells, but we are finally making progress. Ironically, this progress has come in the form of a modified relative of HIV, thereby turing one of the modern world's biggest medical problems into a potential wonder cure. That said, the use of these retroviruses in humans is still tricky, and so at the moment we are generally limited to removing cells from patients, treating them, and then putting them back. This has proved effective in the treatment of metachromatic leukodystrophy, which causes severe nerve damage resulting in progressive motor and cognitive impairment. Replacement of the defective ARSA gene in stem cells removed from affected patients allowed the gene to be stably introduced into their cerebrospinal systems when the cells were put back. Although only three patients were involved in this study, all of them showed either a halt in disease progression or a failure to manifest the disease in the first place. Promising stuff.

Questions of inheritance

The use of gene therapy to treat debilitating conditions such as those described above is undoubtedly a good thing, but it does come with a important additional concern. The current treatments that we have available do not actually cure the genetic disorder, they simply mask it. Insofar as the patient is concerned this doesn't make a blind bit of difference, but where their children are concerned it does. Because the treatment does not correct the genes within the patient's germline cells (i.e. either sperm or ova) they are still likely to pass the defect on to their children. This has the worrying implication that over time such disorders might become more and more common within the population as there is no longer any negative selection pressure against them (i.e. people no longer die before they can pass them on). You could say that this is fine since we essentially have a cure for them, but it is far from desirable to have a large proportion of the population requiring gene therapy in order to survive. Aside from the logistical and safety implications of this, what would happen in the event of periodic breaks in the supply chain of medicine? Something like a major war might claim far more lives if we were so heavily dependent on a constant stream of therapeutic gene treatments.

That said, it is equally unthinkable that we would deny use of successful therapies to those who need it, or that we should stop striving to further eradicate such diseases, as doing either would be little different that killing those people ourselves. Instead, we will need a careful system of screening for people undergoing such treatment to try to minimise the potential for inheritance of their disorder by their children. Eventually it may be possible to treat the germline cells of patients to completely eradicate the faulty genes from their family, and many believe that this is our best bet for overcoming genetic disorders in the long term. Whatever happens, this kind of medicine is only going to become more and more prominent and widespread, and I for one welcome that.

The next post in this series can be found here.


MacLaren, R., Groppe, M., Barnard, A., Cottriall, C., Tolmachova, T., Seymour, L., Clark, K., During, M., Cremers, F., Black, G., Lotery, A., Downes, S., Webster, A., & Seabra, M. (2014). Retinal gene therapy in patients with choroideremia: initial findings from a phase 1/2 clinical trial The Lancet, 383 (9923), 1129-1137 DOI: 10.1016/S0140-6736(13)62117-0

Bennett J, Ashtari M, Wellman J, Marshall KA, Cyckowski LL, Chung DC, McCague S, Pierce EA, Chen Y, Bennicelli JL, Zhu X, Ying GS, Sun J, Wright JF, Auricchio A, Simonelli F, Shindler KS, Mingozzi F, High KA, & Maguire AM (2012). AAV2 gene therapy readministration in three adults with congenital blindness. Science translational medicine, 4 (120) PMID: 22323828

Biffi A, Montini E, Lorioli L, Cesani M, Fumagalli F, Plati T, Baldoli C, Martino S, Calabria A, Canale S, Benedicenti F, Vallanti G, Biasco L, Leo S, Kabbara N, Zanetti G, Rizzo WB, Mehta NA, Cicalese MP, Casiraghi M, Boelens JJ, Del Carro U, Dow DJ, Schmidt M, Assanelli A, Neduva V, Di Serio C, Stupka E, Gardner J, von Kalle C, Bordignon C, Ciceri F, Rovelli A, Roncarolo MG, Aiuti A, Sessa M, & Naldini L (2013). Lentiviral hematopoietic stem cell gene therapy benefits metachromatic leukodystrophy. Science (New York, N.Y.), 341 (6148) PMID: 23845948


  1. Aren't many disorders somewhat genetic anyway. Therefore, this new type of treatment doesn't necessarily raise ethical questions that many other medications don't already. For example chemotherapy might allow someone who has cancer to reproduce and pass on genes that are more prone to cancer and therefore we are producing a society heavily prone to cancer. Or is the point that, although those questions were more or less always in the background, treating a disorder that is 100% inherited, rather than a disorder where the inherited thing is just a proneness, brings the question up much more blatantly?

    It's definitely a frightening question though because there really don't seem to be any obvious ethical ways out. It becomes a clash between what is good for the society and what is good for the individual. Eugenics is also such a clash, but there it is what is "good" for the society vs what is bad for the individual. And here it is what is good for the individual vs what is bad for the society. Eugenics can be clearly argued to be wrong because not doing it doesn't actively damage society, we would arguably still improve and reach the same point in the long run. But with this, there is the genuine risk (as you point out) that treatments such as this, taken to the extreme could make a human species that is incredibly fragile and dependent on medication to survive.

    1. You're quite right that a lot of disorders have a genetic component, but the key difference is that most of these don't tend to kill the sufferer until later in life and so they still have time to reproduce. Cancer, for example, is generally a disease of middle/old age so people tend to have already had kids by this point. Also, as you indicate, many of these disorders are the product of which combination of genes you have, and no single gene is solely responsible for the condition. The prevalence of these kinds of genes is pretty high in the population at large, but most people don't have the specific combination required to cause disease. The mutations responsible for monogenic (i.e. single-gene) disorders are far more rare, and they're the ones that might begin to expand in response to gene therapy.

      I think that this can be managed properly with a strong post-treatment focus for patients receiving gene therapy. A patient who receives treatment for a monogenic disease could have their gamete (i.e. sperm or ova) screened prior to in vitro fertilisation, so that they can be confident their offspring won't inherit the disorder. There is already a screening programme in place to let carriers of recessive genetic diseases know what the risks are of their children being sufferers, and some do choose to go with IVF to reduce the risks. At the minute, though, this is prohibitively expensive and not feasible to cover the entire populous, but one day it might be.

      In some cases, though, this isn't possible. For example, in X-linked recessive disorders (such as haemophilia) an affected mother has no chance of having a son who does not suffer from the disease, since she has 2 copies of the mutated gene and the son can't receive a healthy copy from his father (since boys only inherit the X chromosome from their mother). Similarly, a father with an X-linked dominant disease (such as Rett Syndrome) can't have a daughter without the disease, as she will definitely inherit the mutated X chromosome from him. In these cases screening would limit them to having children of only one gender, which may be preferable to having a child with the disease but that is personal choice that each patient would have to make.