Gene Therapy for Retinal Disease: What Could It Mean for Usher Patients? Part II
October 14, 2009
By Jennifer Phillips, Ph.D.
In Part I, I described the gene replacement therapy used in a pilot experiment to treat progressive retinal degeneration in LCA patients. Without rehashing the details, the major finding from this study to keep in mind is this: delivering a healthy copy of the gene via viral vector into the retinas of LCA2 patients resulted in improved vision in most of the subjects. So, while this is great news indeed for LCA2 patients, what might the relevance be to our community? How might this finding apply to a cure for Usher's down the road?
Well, there are both good and...I won't say bad, I'll say challenging aspects of this pilot therapy as it applies to our cause. I'll go through a few of the major points below, comparing the LCA study with what we could expect from a hypothetical, similar study of Usher gene therapy:
1. Retinal degeneration in LCA vs. Usher Syndrome:
The Good: The progressive nature of these diseases is very similar. LCA patients who already had clinically significant vision loss were helped by this gene therapy, so it's a reasonable hope that a similar treatment directed at an Usher syndrome gene might give similar results. The advantage of slow progressive vision loss is that it allows time within the lifetime of a patient for medical intervention to be administered. Note that this is distinct from a congenital defect, as is the case with the hearing loss in most forms of Usher syndrome.
The Challenging: The cause of the retinal degeneration in LCA2 is quite well understood. That is, researchers have a pretty good idea of what is going wrong at the molecular and cellular level that ultimately causes the death of photoreceptor cells. This is not yet the case with Usher syndrome. We've been able to identify a respectable number of the molecules involved, and have some good indications of how they might be working together, but we don't currently have the full picture of what their jobs inside the cells entail, and why the loss of these functions, whatever they might be, results in cell death.
Silver lining: While a better understanding of this process will undoubtedly enhance research and treatment options for Usher syndrome down the road, it's not an absolute prerequisite for attempting gene therapy. The most important question is: will it work? We don't have to know every minute, mechanistic detail of how it works before attempting it (beyond thorough safety testing, of course).
2. Use of the viral vector delivery system for an LCA gene vs. an Usher gene:
The Good: The LCA trial gives us pretty compelling evidence that this method of gene delivery is safe and effective. The replacement parts (complete, healthy copies of the gene) are being delivered to the right addresses (the retinal cells) and being put to good use (restoring cellular function), while causing no detectable damage to the eye or any other part of the patient. An Usher gene delivered in similar fashion could thus be expected to enter the appropriate cells and result in a similar functional rescue. Furthermore, since we now have a measure of the success of using this viral vector, introduced in this particular way, that could minimize the research time and effort put into Usher-specific gene therapy.
The Challenging: Remember the careful selection process that went into choosing this particular LCA gene for the trial? The researchers were able to select a target with a high chance of success, but, as with Usher syndrome, there are multiple genes that cause LCA, and there were clearly some less-than-ideal choices that were passed over for the purposes of this pilot experiment. Some genes are definitely more suited to viral vector delivery than others. In the case of Usher syndrome, one of the biggest obstacles is going to be the size of some of the genes. Why does size matter? Because there's an upper limit to the length of the DNA code that existing viral vector can take on.
Time remember back to Bio101 again. A 'base' is the essential building block of a nucleic acid like DNA. There are four nucleotide bases in DNA, abbreviated A, G, C, and T. Sound familiar? The sequence in which these bases appear in the gene gives the gene its identity. And, since the principal role of most genes is to provide instructions for building a protein, the base sequence dictates the order in which the building blocks of a protein are assembled. So when we're talking about 'gene size', we're talking about the number of sequential bases that comprise the full set of information, the 'recipe', if you will, for building a complete, functional protein. The upper size limit for most viral vectors is a nucleotide cargo of about 10 kilobases (10,000 bases), and even that is a bit dicey, as the efficiency tends to diminish as cargo size increases. Of the known Usher genes, the smallest ones are about 2 kilobases, and the largest one is close to 20 kilobases. Five Usher genes out of the nine we've identified so far come in under the 10 kb size limit.
Beyond the size consideration, another confounding aspect of initiating gene therapy with Usher genes is that many of them use the information in the DNA code to make multiple forms of the corresponding Usher protein. These multiple protein forms are akin to the new-fangled Lego kits that my son is so crazy about, from which one can assemble a helicopter, or a boat, or a motorcycle by recombining the core components in slightly different ways. Some of the Usher proteins have gotten pretty creative with their genetic Lego sets, assembling as many as 20 different forms of the same basic protein by mixing and matching different parts of the genetic code. Thus, if we're endeavoring to devise a gene therapy for a protein with such multiple assembly options (proper science term: isoforms), we need to take the number of know isoforms into account. Do we make up multiple batches of the viral vectors, each containing DNA specifically encoding one of the possible isoforms? Or are there some isoforms we can dispense with, cutting down our workload and minimizing the number of different constructs that actually need to be put into the patient? Not to mention the recalibration of the effective dosage that would have to take place if one were to include multiple isoforms in our liquid solution for subretinal injection, as the presence of multiple different viral vectors and their cargo will obviously effect the concentration of any given vector. This is where the deficit in our understanding of molecular mechanisms comes into play a little bit. These would be difficult decisions to make without knowing more about how (and where, and when) each isoform functions.
Silver linings: As noted above, slightly more than half of the Usher genes would be viable candidates for viral vector delivery, strictly on the basis of gene size. There are some multiple isoform issues on that short list to contend with, but as stated, we can potentially narrow down the essential isoforms to include in our trials as we learn more about the function of these various protein forms.
3. 'Bench to bedside' progression of LCA gene therapy vs. potential for same with Usher syndrome gene therapy:
The Good: As I mentioned above, it's potentially extremely helpful to us to have the LCA clinical trial results in hand, as it means that the vector itself has undergone thorough testing for safety. At the basic cellular level, at least in theory, substituting an appropriately sized Usher gene in place of the previously successful LCA2 gene is just details. The virus should behave the same way in terms of its impact (as distinct from the specific effect of the gene therapy itself) on the overall health of the subject, regardless of the particular gene sequence it's carrying.
The Challenging: One key advantage that the LCA researchers had in their translational research progression was an abundance of animal disease models. They had a mouse model AND a canine model of LCA2-animals with documented retinal degeneration on whom not only safety but effectiveness of the gene therapy could be tested. The significant therapeutic effect of the gene therapy of the LCA2 dogs in particular (due to the enduring improvements to their vision observed over many years) was undoubtedly a factor in getting the clinical trials underway, and in having some sense of what to expect when the treatment was applied to human subjects.
Dog models of human disease are pretty rare, but genetic engineering has made the generation of mouse models of human disease quite commonplace. Unfortunately, with respect to Usher syndrome specifically, there are very few existing mouse models of the disease. I've discussed this in a bit more detail elsewhere (skip down to the third paragraph after the second figure). The bottom line is that, currently, although we could certainly glean sufficient safety data for new or existing vectors bearing Usher genes to be delivered retinally, our options for judging efficacy in pre-clinical studies ("does this therapy reduce or slow retinal degeneration?") are limited.
Silver linings: 'Few existing mouse models' means that some do in fact exist. One pre-clinical study of virally-delivered gene therapy has already been conducted with a mouse model of Usher type 1B, with promising results (see here for an abstract). Other Usher genes have recently been, or are now, being targeted for deletion in mice. Hopefully at least some of these Usher mice will be able to serve as models in which to test gene therapies in the future.
Every day, we gain a greater understanding of the molecular basis of Usher syndrome. By studying the ways in which the various proteins behave and interact within the retina and the inner ear, or investigating similarities and differences between Usher gene function in humans compared to mice or other model organisms like the zebrafish (my personal favorite), we get closer to having enough information and resources to develop treatments. There is most definitely cause for optimism here, and more motivation than ever to write your representatives and urge them to support Usher syndrome research.