Zinc Fingers: An Emerging Tool for Gene Therapy

Zinc fingers-you're full of them, they're a great, compact molecular tool for binding DNA, and they're my obsession for the summer at Princeton University. Someday, we might see zinc fingers employed in the treatment of a wide array of genetic diseases.

When a new genome is sequenced, the first and obvious features to look for are protein-coding sequences (CDSs). After all, when you get down to business, protein production is the central role of DNA. All CDSs begin with the start codon, ATG, and end with a stop codon, TAA, TAG, or TGA. You can find CDSs (or, at least, candidate CDSs) by looking for start and stop codon pairs with sufficiently long stretches of intervening DNA. However, one is still left wondering when the resultant protein is actually present in the cell, in what quantities, and in concert with which other proteins. All of that information is exterior to the protein-encoding information of the gene, and the most important piece is upstream in the 5' untranslated region (5' UTR). DNA information is transcribed into RNA information by RNA polymerase, which must be recruited to the gene by its 5' UTR. In bacteria, RNA polymerase often binds directly to the DNA from the start, but in eukaryotes other proteins always bind first, called transcription factors (TFs). Zinc fingers, small DNA-binding protein domains, are the heart of the majority of human transcription factors (they are the most common protein binding domain in humans).

As a protein domain, zinc fingers have a particular conserved structure, namely two beta sheets and an alpha helix. The overall zinc finger structure is (unsurprisingly, given its name) stabilized by a zinc ion. It is the amino acids in the alpha helix that bind to DNA and determine the target sequence. One zinc finger can bind three DNA bases, so you can imagine a zinc finger for CGC, one for TAC, another for AAT, and so on for the remaining 61 combinations (this coincidental convergence with the triplet genetic code is, at least, somewhat amusing). The structures at the top of this post depicts a particular zinc finger protein, called Zif268 or EGR1, binding to its DNA target sequence. The spacefilling model shows the tightness of binding between the zinc fingers, in orange, to the DNA, in purple. The cartoon model shows you the conserved zinc finger motif of two beta sheets (yellow arrows) and an alpha helix (pink). Zif268 binds to the sequence GCGTGGGCG, as well as variations on the theme. You can quantitatively understand this variation with a sequence logo (see image, as well as my previous post explaining sequence logos). If you look closely at the structure, you will notice that Zif268 has not one, but three zinc finger domains. By hooking three zinc fingers together, Zif268 is able to extend its recognition site from three bases (what you get for a single zinc finger) to nine. (Notice that the sequence logo depicts eleven bases; apparently Zif268 has some preference as to the bases on either side of its canonical 9-base binding site). A quick Wikipedia search revealed that Zif268 is thought to be involved in controlling gene expression in the brain.

Setting aside their immense importance in regulating our and other species's genes, zinc fingers represent a potentially useful tool for medical applications. It has been a decade since the human genome was published, and still hopes of personalized medicine and therapy for genetic disorders seem far away. There is a big difference between decoding a genome and repairing faults in it, which quickly becomes apparent when one considers the immensity of the problem: an adult human has trillions of cells, all faithfully inheriting the same defect from their parent cells, grandparent cells, and on back to the original zygote. Fixing a mistake in a trillion places is a tall order. Zinc fingers might be the answer to that seemingly insurmountable challenge.

Historically, however, viruses have been the only viable mechanism for faulty gene repair (let's call it gene therapy). That makes sense, since viruses are really good at getting DNA into human cells; viruses that efficiently invade human cells and hijack them with their DNA payload far out-propagate the competition and so are favored by natural selection. Converting a virus's malevolent payload to a benevolent one is as simple as removing the harmful genes and inserting a "good" copy of some gene to replace the faulty copy in the patient's genome. When a cell is infected, the virus splices a copy of the good gene into the genome at a random location. Sounds great, right? Well unfortunately this often causes the patient to develop cancer. That's the problem with sticking a chunk of DNA into the genome at random; sometimes it will interrupt a gene that regulates the cell cycle or cuts off out-of-control growth. Either occurrence is a recipe for disaster.

Zinc fingers might have the answer. Since they bind to a specific sequence of DNA, they can home in on the defect and avoid disturbing harmless bystander genes in the repair process. Research is underway (by many labs around the world, including my summer lab at Princeton) to apply this fact to gene therapy. By attaching a DNA-cutting protein ("nuclease") domain to a series of zinc finger domains, you obtain a protein that can bind to a specified sequence and cut the DNA in two. If a correct copy of the cut region is present in the cell, the DNA repair mechanism will use it as a template to repair the gap in the chromosome. The cell doesn't know that it is using a correct copy; it just happens to be around because of the design of this therapy. But remember that zinc fingers are small proteins; they cannot by themselves carry a DNA payload. For some conditions, such as genetic disorders that affect the blood (sickle cell anemia comes to mind), blood cells could be treated in the lab directly and then re-injected into the patient. Even better, the bone marrow could be treated and reimplanted, providing a source of healthy blood cells for life. However that approach is only viable for a narrow set of genetic disorders. The answer might be, somewhat ironically, to again turn to viruses. But this time we are not interested in viruses that integrate their genetic material into the host genome. Non-integrating viruses could provide the DNA template that zinc fingers, coded in the viral genome, used to repair the genetic defect in the patient.

The realization of this possibility, if it is to come at all, is still far away. However, whether they will turn out to be useful therapies or not, zinc fingers are interesting to study as tools that have emerged and multiplied through evolution for regulating genes. Their simultaneous flexibility and simplicity, all the while in the microcosm of the nano-scale, should be the envy of human engineers the world over.

Image credits: Structures were rendered with RasMol, using coordinates from PDB ID 1A1L. The sequence logo is from the JASPAR database, ID MA0162.1.

References: I believe that the content of this post could fairly be called "common knowledge," which I have obtained from discussions in my lab. If you are interested in learning more about the application of zinc fingers to gene therapy, email me (address listed at the top left) and I will provide you with several good papers to get you started.