Uncommon ways of looking at common problems- gene therapy for mitochondrial DNA diseases

I’m always very interested in less conventional approaches to conventional problems and a recent review found while looking for a paper for a journal club provided a good example.

As you know, gene therapy aims at correcting a genetic deficiency by the introduction and/or substitution of the mutated gene by a functional copy into relevant cells. The classic examples for gene therapy are normally diseases such as Cystic Fibrosis- monogenic, recessive, autosomal diseases, as these are, for obvious reasons, those that theoretically offer higher chances of success. There are, however, other types of genetic diseases that offer more space for original, unconventional gene therapy approaches.

Mitochondrial DNA diseases are a good example. The human mitochondrial genome is small (16kb) and is housed in the mitochondrial matrix. As most of its genes encode proteins involved in oxidative phosphorylation necessary for cellular respiration and ATP production, is not too surprising that mutations in such genes can cause a variety of diseases, many of which with symptoms associated with progressive defects in muscle and nerve systems.

There are two obvious strategies to tackle this problem using gene therapy. One is to introduce the functional copy of the mutated gene into the nucleus. The protein, with an appropriate mitochondrial targeting sequence, can then be transported to the mitochondria. This is the so-called allotropic expression of the gene, and has been shown to produce promising results in culture. Other option is to target the plasmid containing the functional gene into the mitochondria itself, as these organelles contain all the ribosomal RNA and tRNAs necessary for intra-mitochondrial protein synthesis (assuming that these are not the mutated genes). These strategies have been tried out with some success both using a wild type sequence of the mutated gene, or a gene from another species, known to perform its function in the pathway more efficiently (xenotopic expression).

However, as you might have already noticed, a mitochondrial DNA disease is characterised by the fact that not all the mitochondrial genome copies in a cell are necessarily mutated (the presence of different genome types within the same cell is known as heteroplasmy). This has two main implications: firstly, mtDNA diseases are threshold diseases: only when around 60-95% of copies in a cell are mutated will symptoms appear. This threshold will obviously vary depending on the disease. Secondly, rather than trying to introduce a functional copy into the mutated mitochondria, a much more interesting approach can be to just manipulate the relative amounts of the mutated and the wild type populations of mitochondria.

One approach is to promote specific types of exercise training to promote mitochondrial biogenesis. However, as the geneticist that I am, the alternative strategy- to manipulate the mutated mt genome- is much more interesting. A brilliant idea, in my point of view, is to take advantage of the fact that some mutation types introduce restriction sites. That way, rather than trying to express a functional copy of the mutated gene, it may be more successful to express a specific restriction enzyme that will degrade the mutated genome altogether. This is limited, obviously, to those diseases in which a unique new site is introduced, or in those where the number of restriction sites is higher enough in the mutated population to have a significant effect.

Another clever idea is to methylate the mutated mt genome so that it is not expressed, allowing the wild type population to proliferate and overcome the mutated one with time. The use of mitochondrial targeted PNAs (peptide nucleic acids) were a potential method, but were shown not to cross the inner membrane of the mitochondria. However, the use of zinc-finger peptides with DNA sequence specific binding capacities have been used to successfully methylate differently the mutated mitochondrial genome. If these domains are linked to a nuclease, it may be possible to degrade the mutated genome as well.

These are all new strategies, only attempted in cell culture and their usefulness in in vivo models and ultimately in patients is yet to be determined. Nevertheless, I think these novel approaches just show that even standard problems can be approached in innovative ways.

References:

Kyriakouli, D., Boesch, P., Taylor, R., & Lightowlers, R. (2008). Progress and prospects: gene therapy for mitochondrial DNA disease Gene Therapy, 15 (14), 1017-1023 DOI: 10.1038/gt.2008.91

Alexeyev, M., Venediktova, N., Pastukh, V., Shokolenko, I., Bonilla, G., & Wilson, G. (2008). Selective elimination of mutant mitochondrial genomes as therapeutic strategy for the treatment of NARP and MILS syndromes Gene Therapy, 15 (7), 516-523 DOI: 10.1038/gt.2008.11

(the most recent paper I could find in this field, using the restriction enzyme approach in culture)

Bioluminescence

In many biological studies it is essential to be able to translate into an observable and measurable signal the specific process under analysis. This is normally possible through the use of assay systems based on radioactivity, photon absorption or photon emission. Radioactivity and photon emission are less and less used nowadays, the first due to the restrictions and cares associated with the use of such hazardous materials, the second due to its low sensitivity.

Photon emission can occur by two different processes- either fluorescence or chemiluminescence. Both processes involve the emission of photons associated with transitions between energy states, but the way the excited state is generated varies. In fluorescence, the absorption of light (i.e. of photons) produces the excited state, while in chemiluminescence this is due to exothermic chemical reactions (see diagram- fluorescence is on the right, luminescence on the left).

Both of these processes have their advantages and disadvantages. Fluorescence (of which the most famous example is probably GFP) normally shows a much stronger signal, as the source of the excited signal are photons which can be introduced in the sample at very high rate. However, this jamming of photons tends to create a high background signal which influences the sensitivity of the assay. Chemiluminescence does create lower intensity signals, but as no photons need to be provided there is hardly any background signal. Whichever of the two systems is used will depend on the machinery required in each assay. If the efficiency of light collection is limited, then fluorescence is the best option, this being the reason why until recently fluorescence was the main assay system used. The development of more sensitive machinery, however, has made the background noise the most important problem to solve, therefore justifying the increased interest in chemiluminescence systems.

Bioluminescence is a form of chemiluminescence that naturally occurs in certain living organisms. The enzymes that catalyse this reaction are called luciferases and their substrates luciferins. Note, however, that these are very general terms as bioluminescence seems to have evolved several times independently. This is why the several existing luciferases have such different molecular structures. In fact, luciferases have been cloned from several different types of organisms, namely jellyfish (Aequorea), sea copepod (Gaussia princeps), corals (Tenilla), click beetle (Pyrophorus plagiophthalamus) and several bacterial species. The most successfully luciferase, however, is the widely used firefly luciferase, from the firefly Photinus pyralis. It is commonly used in its humanized variant (codons optimized for mammalian expression), requiring ATP and magnesium in the oxidation of its substrate luciferin, and yielding a yellow-green light with a maximum luminescence of 560 nm.

Firefly luciferase (FLuc) is normally used as a reporter gene, which may mean very different systems of reporting. The most obvious is, of course, the insertion of the luciferase gene in the same plasmid and under the same regulatory promoter as the gene of interest. In my current department (Gene Therapy department, NHLI), for example, firefly luciferase is the main reporter gene used in CFTR transfections. However, FLuc can report more than the expression levels of transfected genes. As it depends on ATP for its activity, it is ideal in assays of ATP concentration. It can also, for example, be an indicator of how well G-protein coupled receptors (GPCR) work. This is possible by using a cAMP response element (CRE) upstream of the luciferase gene. GPCR activation causes an increase in intracellular cAMP concentration. This increase leads to Protein Kinase A activation, which in turn phosphorylates CRE binding protein. The CRE binding protein will bind to the CRE upstream of the luciferase gene, leading to increased transcription. Overall, an increase in CPCR activity leads to increased luminescence signal. This type of report system, for example, is useful in the study of GPCR agonists. The complexity of the system, however, makes it prone to interference, which can lead to false positive results. To prevent this, a dual-reporter system can be used involving a second luciferase, Renilla Luciferase (RLuc), which is an internal control to detect aberrant data. A dual luciferase system can also be used to detect two different processes or expressed genes simultaneously (as FLuc and RLuc have different spectral maximuns, namely 560 and 480 nm).

As a reporter of transfected gene expression, FLuc shows a few limitations, namely the fact that it is an intracellular protein. This means that measuring of transfected gene expression requires, in in vitro cell culture the lysis of the transfected cells, and in in vivo studies the killing of the animal in order to access the transfected tissue. In patients under clinical trial, luciferase assays imply invasive obtainment of tissues (I am still trying to find out how this is done exactly). This is why the development of new luciferase, namely a secreted luciferase is important. Until now, around 4 types of secreted luciferases have been found, but the only one commercially available for cell supernatant assays is that produced by the copepod Gaussia princeps (see figure). Gaussia luciferase (GLuc) is naturally secreted from cells and used by this animal as a defence mechanism. As Gaussia princeps lives at a depth of between 350 and 1,000 m, the sudden production of light is a good distractive mechanism against dark-adapted predators. As a secreted luciferase, Gaussia is reported to give very good signals when medium supernatants are assayed. The important question right now (at least for me) seems to be if it can be assayed in something more… well, you will have to wait for my year-away talk next year for more information!

References

Fan F., Wood K. (2007). Bioluminescence Assays for High-Throughput screening, Assay and Drug Development Technologies, 5. pp 127-136

Markova S., Golz S., Frank L., Kalthof B., Vysotski E. (2004). Cloning and Expression of cDNA for a luciferase from the marine copepod Metridia longa, The Journal of Biological Chemistry, 5. pp 3212-3217

Roda A., Pasini P., Mirasoli M., Michelini E., Guardigli M. (2004). Biotechnological applications of bioluminescence and chemiluminescence, Trends in Biotechnology, 22. pp 295-303

Serganova I., Moroz E., Moroz M., Pillarsetty N., Blasberg R. (2006). Non-invasive molecular imaging and reporter genes, Central European Jornal of Biology, 1. pp 88-123

Tannous B., Kim D., Fernandez J., Weissleder R., Breakfield X. (2004). Codon-optimized Gaussia Luciferase cDNA for Mammalian Gene expression in culture and in vivo, Molecular Therapy, 11. pp 435-443

Wiles S., Ferguson K., Stefanidou M., Young D., Robertson B. (2005). Alternative Luciferase for monitoring bacterial cells under adverse conditions, Applied and Environmental Microbiology, 71. pp 3427-3432