Altering our proteome.

Several of the SENS interventions rely on our ability to introduce new proteins (or an RNA, but for brevity we'll stick to proteins) into a person's body, or to remove those already present. There are four fundamentally distinct ways to do this: transplantation, cell therapy, somatic gene therapy and somatic protein therapy. In addition we consider germline gene therapy, and explain why this is unlikely to be a viable solution to the specific question of aging, despite its potential for use in other areas of medicine.

Transplantation

This is the simplest of the four methods and involves introducing cells or organs from someone who already expresses the protein – in other words, the kind of transplantation of organs and tissues that we already use to replace the blood of chemotherapy patients or the kidneys of dialysis patients. This is clearly limited to proteins that at least some humans already express (a relevant limitation, as discussed here, for example), and it generally carries the risk of terrible complications when the immune system rejects and attacks the foreign tissue – a side effect that must be suppressed by lifelong drug administration, impairing the recipient's resistance to infections and possibly to cancer.

A crucially important issue in transplant-based therapies is the source of the organs or cells for transplantation. Fortunately, there have been a great many encouraging recent results in tissue engineering, the science of growing organs for transplant in an artificial, biodegradable scaffold outside the body. Although the field is still young, it is developing at a frantic pace – many of the key achievements have been reported at the SENS conferences.

Tissue engineering holds the promise of solving the problem of organ shortages and potentially also eliminating the need for the unpleasant and risky process of lifelong immunosuppression (since the cells used to produce the engineered tissue can be derived from the patient themselves, avoiding the prospect of immune rejection).

Cell Therapy

In this case, the idea is to take "autologous" cells (ones taken from the patient), modify them in the required way (for example, by inserting genes for new proteins), and reintroduce the modified cells into the patient, thereby avoiding an immune response. However, this approach still has limitations: the original cells that the engineered cells are designed to replace must be eliminated. Sometimes the problem to be rectified is precisely that the cells that we need are already gone, and in other cases they are present but failing. In the latter situation, the failing cells can theoretically be eliminated and replaced with healthy new cells, but there are technical difficulties in doing this smoothly enough to maintain tissue function throughout the procedure. Thus, many SENS interventions will require the use of options three and four: somatic gene therapy and somatic protein therapy respectively.

Somatic Gene Therapy

Getting engineered DNA into specific places in the genome.

Several of the strategies for repairing the seven SENS targets will almost certainly require, in some tissues, somatic gene therapy – altering the DNA of cells whilst they are still in the body. This is much harder than taking cells from the person, altering the DNA in the laboratory, and then putting them back, because altering the DNA of cells is very error-prone. If you do the alteration in the lab, you can check whether the correct alteration happened (and that nothing else happened) and only put back cells that pass that test. The error-prone nature of all existing approaches to somatic gene therapy makes it still highly experimental and risky, which is the main reason why it is still in its infancy.

Therefore, one of the things SENS needs most badly is an improvement in our techniques for doing what we want to do to our genomes in situ, and not accidentally doing other things at the same time. Luckily, several methods are currently under intense investigation; here we look at two of the problems and their possible solutions in a little more detail.

Inserting genes in a safe place: the adeno-associated virus and beyond.

The problem. For most applications, all we need to do is get a new gene or genes into our chromosomes, and it probably doesn't much matter where it goes in - unless it disrupts the genes we already have. Unfortunately, that "unless" can't be neglected, because we have so many cells. If we're trying to get the DNA into all (or even most) of our cells then it's going to hit genes in a fair proportion of them, and in a few of those it's more or less certain to hit genes involved in cell cycle control, which means, of course, that it may disrupt the normal function of these genes and thereby promote cancer.

The solution. The first big breakthrough in solving this problem was the discovery that one virus, the adeno-associated virus (AAV), preferentially inserts genes into a particularly safe place, located on human chromosome 19. This is good, but not good enough, because in order to make the virus carry the useful genes that we want to put into our cells, we have to take out the very stuff that gives it its site-specificity. But there are various approaches being explored for hybrid viruses that get the best of both worlds: enough carrying capacity to be useful, without loss of site-specificity.

Also, even the unmodified AAV doesn't always insert in this particular spot. This is largely because its DNA has a habit of "invading" double-stranded DNA and sometimes undergoing a random mixing of its genes with those in the host organism. Much excitement is therefore currently surrounding a new type of virus (actually a bacterial virus) called a phage, which has only a very low tendency to mix into other DNA at random. It does get into DNA, but only when it expresses a particular enzyme called an integrase. Better yet, it only goes into a few specific places in the genome, although unfortunately these do not appear to be as safe as the site preferred by AAV. However, some success has been achieved in "evolving" these enzymes in the lab so that they prefer safer sites, so there is strong hope that these engineered phages will function as safe gene therapy vectors soon.

When a safe place won't do: targeted gene disruption.

The problem. Nearly everything that we would like to do with gene therapy, whether for aging or for any other condition, can probably be done pretty well by introducing new genes into cells in a safe place. Sometimes we want to stop a gene from expressing its product, because the product is toxic (such as the mutation that causes Huntington's disease), but even then we can probably achieve the desired effect just by putting in a gene, because we can use an amazing phenomenon called RNA interference (RNAi) to cause the original gene's transcript to be destroyed before it is translated.

But there is one case where we probably will really need somatic gene therapy targeted at a specific place in the genome, and that's for the anti-cancer therapy, WILT. WILT relies on treating cancer in such a way that it would be impossible for the cancer to mutate its way out of the problem, by removing the genes that cancer cells need to sustain the reproduction of their DNA when the cells divide. We could use RNAi to disrupt the expression of these genes, but this really isn't a strong enough defence, because cancer cells have a very powerful ability to mutate their way out of problems. As long as the relevant genes themselves remain intact, there would always be the threat that the RNAi disruption would be subverted by such a mutation, allowing cancer to progress again.

The solution. At present, there are several approaches to targeted gene disruption (or "gene targeting", as it's usually called) that can be customised to attack anywhere in the genome, but they're almost all highly error-prone, disrupting other locations a great deal. Recently, some very encouraging progress with artificial enzymes called zinc-finger nucleases (ZFNs) has given reason for optimism; these enzymes can be targeted to genomic sites far more consistently than viral vectors, and their site-specificity is so well understood that they can be engineered to exclusively target single locations in the human genome. ZFNs can achieve a 40% or better "hit rate" in human cells without any selection process being applied (exactly as would be the case in the body) and this efficiency is being improved all the time.

Fortunately, scientists are vigorously pursuing such methods, because of their recognized value for treating a variety of inherited diseases. Once the technology has been perfected for these disorders, it will be available for SENS interventions, too.

Somatic Protein Therapy

Getting proteins into cells to avoid gene therapy.

Many components of SENS entail alteration of our genetic codes in lots of different cell types. For cells that are constantly renewed from stem cells, this is relatively easy, because we can extract cells from the individual, do what we want to those cells in the laboratory, check that we've done exactly what we wanted to do, and put them back in. This is not easy, let me stress – especially hard is doing all this without the cells losing their "stemness" – but it'll probably be a lot easier than the alternative, somatic gene therapy. However, tissues that do not have continuous renewal - such as most of the brain - can't be altered in this way, so at first it might seem that somatic gene therapy is the only option. There are two potential solutions to this problem.

Putting proteins into cells.

The basic reason we want to change the genome of cells is so that those cells will start to make different proteins. For most SENS purposes (and indeed for most biomedical purposes generally), we want the cell to have proteins that it didn't have before, as opposed to lacking ones that it used to have. In principle, therefore, this can be done by introducing the proteins themselves, rather than the genes encoding them.

The obvious problem with such an approach is scale. Most proteins are rather short-lived, so the cell needs to make them again and again in order to have them around in the required abundance. Thus, it would be impractical to introduce enough protein. When the physician and biologist Roscoe Brady first sought to explore this approach, he was soundly dismissed for this reason.

However, it turns out that there are plenty of cases where this is not a showstopper. The class of proteins in which Brady was (and still is) interested are enzymes that break things down in lysosomes; these enzymes are congenitally absent in sufferers of lysosomal storage diseases. Brady eventually succeeded in developing methods to make enough enzyme and target it to the right cells to be able to give many such people a normal life when they would otherwise certainly have died in childhood. One of the most important SENS strands, lysosomal enhancement, may well be able to work this way for many tissues.

Using tissues that are renewed from stem cells.

Another way out of the protein scale problem is to introduce the genes for the desired proteins into one tissue and arrange for them to be exported from the cells that make them and imported by the ones that need them. This makes sense because genes can be introduced safely in stem cells in vitro much more easily than somatically, as noted above. It is quite easy to modify genes so that their encoded protein will be secreted, and there are also techniques for targeting proteins in the circulation to particular organs. An especially important one is the brain, which is protected from the circulation by a special system that stitches the cells of the blood vessel lining together much more tightly than elsewhere in the body: this is called the blood-brain barrier. Some native proteins need to be transported into the brain from the blood, and we have a growing understanding of how this happens and how we could exploit that system to get our chosen proteins across.

Germline Gene Therapy

Finally, a word about "germline" gene therapy - that is, changing the genome of either a gamete (sperm or egg) or a zygote (a single cell formed either by fertilisation or by somatic cell nuclear transfer, a.k.a. cloning) so that people are born with a designed genetic alteration already present in their cells.

Some people think this would always be far too dangerous to be useful, while others have argued persuasively that these dangers can be overcome. However, the attractiveness of this approach is limited – not only by the timescales involved before the effects of an intervention become measurable (since aging only starts being bad for us after we reach 50 or so), but because any solutions involving germline therapy would be of no use to those of us who have already been born and thus cannot benefit from the intervention.

For these reasons germline gene therapy is quite likely to become an important biomedical procedure in the future for a variety of diseases – but not for combating aging.

Resources

Talks on this topic at IABG 10: Capecchi, Kmiec, Sefton

At SENS2: Porteus, Larrick, Mavilio, Margison, Brady, Atala (abstract only), Ratner

At SENS3: Calos, Gardiner, Holmes, Laurencin (abstract only), Mason, Morgan, Philpott