Maximising Potential for Zinc Fingers in DIYbio
10 views
Skip to first unread message
Cathal Garvey
Dec 31, 2009, 9:20:36 AM12/31/09
to diybio
Hi all,
When I started my PhD in cancer research, it was an immediate dive into the exciting field of Zinc Finger Nucleases and site-directed repair/mutagenesis! Unfortunately, my project took a turn for the boring shortly thereafter, and I never got to work with the things. Looking at the ridiculous prices Sangamo are charging for sets of Zinc Fingers now, it looks like commercial fingers will be out of the DIYbio price range for years to come.
You can make your own Zinc Fingers using techniques, tools and kits provided kindly by the Zinc Finger Consortium, but the hit-and-miss reliability requires lots of prototyping, and the steps needed to test new fingers take ages. You could order fingers from a gene synthesis company, and probably save your money's worth in time, but it's still expensive when it turns out not to work.
But don't fret, DIYbioers! All is not lost. Sure, Zinc-Finger Nucleases for your-site-of-interest work are currently impractical. So what? That still leaves a whole spectrum of use open to you for these fantastic, tiny proteins. And that's what I'd like to start a discussion on here, after offering some suggestions for cunning workarounds.
Workaround #1: Bring the Mountain to Mohammad
So, it's hard to make Zinc Fingers to target DNA reliably. But, others have already made and proven Zinc Finger sequences that target their own DNA very reliably indeed! As nucleases, this mightn't be very handy unless their target sequence happens to exist in your species of interest's DNA. However, ditch that FokI cleavage domain, and replace it with a transcriptional activator domain, and you've got a transcription factor that binds reliably to a known site.
Now, for your next project requiring transcriptional control, you can add the 20ish bp needed to the upstream region of your gene, and that's all the promoter you'll need to be certain of what's going on. No more guesswork with natural promoters, you've just turned someone else's carefully designed Zinc Finger on its head and used it as part of a sophisticated control system.
This scales up nicely with synthetic biology; if you want to add lots of complex systems to a species, you can do so as a branch of its normal transcriptional control rather than relying on the existing, complicated-and-difficult-to-predict systems. If your transcription factor is ordered with is own binding region in advance of itself, you've invented a new constitutive control region that can be reliably, predictably used for any downstream projects you care to try. I'd like to do this with Cyanobacteria!
Workaround #2: Mixing and Matching
Zinc Finger Nucleases work by conjugating the FokI cleavage domain to a pair of zinc fingers that each bind either up or downstream of the target site. Because FokI cleaves only as a dimer, the ZFN only cuts at the target site because either end won't cut alone. If they'd used a protein that could cut on its own, then it wouldn't matter how long the single-ZF nuclease was, because partial affinity to bits of DNA would let it cut up the whole genome in no time.
The nice side effect of this is that, although it's unlikely your ~20bp site matches exactly that of someone else's site previously done and demonstrated, because their carefully made Zinc Fingers are actually a pair of ~10bp targeting fingers, someone may have already targeted half of your region and come up with a working Zinc Finger Array for the job.
So, if you look for the full sequence, you're unlikely to get much. But if you search for subsections of your sequence, you might get lucky! I seem to recall that the ZFC's search tool could compare known ZFNs to a given sequence to find binding sites. This could make this job easy, but failing that a search for "Zinc Finger" plus a bit of dna code might turn something up in the literature.
Workaround #3: Stretching Your Fingers (Theoretical!)
Zinc Fingers twist as they bind DNA, and much of the annoying optimisation of getting a ZFN to cut isn't into the binding affinity of the Zinc Fingers themselves in the array, but they spacing between fingers, or between the fingers and the FokI domain. FokI needs to meet another of its kind *across* the DNA helix from itself, so correct spacing and orientation are essential.
This is annoying, but noting this property also suggests another potential avenue to re-using an existing finger, although some prototyping may still be needed. If you happen to find that your target sequence is partially covered by one prior-art ZFN array, but with another one a few nucleotides down (so close! Damn!), you might be able to bridge that gap by throwing a few spacer amino acids into the father-away ZFN, so the FokI domain is looser and stretchier. If you space things just right (I don't pretend to be an expert at this sort of math), you should be able to get the FokI domains to meet at the point you want to cut. Your accuracy will be lower, but it's saved you a lot of time if it works!
Workaround #4: Really pushing it with Adaptamers (Pure, possibly nonsense theory)
I don't recall if ZFNs are used to bind RNA, or if they can. But, assuming for a minute that they bind RNA too and can or have been used for that purpose, there are ways one might exploit that property to apply a totally unsuitable ZFN to a target region.
By having a stretch or RNA transcribed that matches the ZFN at one point and a point on your target DNA/RNA on the other, you may be able to get the ZFN to bind to the RNA, and have the RNA target either other RNA molecules (watch out for dsRNA-targeting machinery if done in relevant species) or open DNA helices, for example a replication fork or a highly transcribed region.
More promisingly, if you can convince a ZF-Array to bind RNA and then mimic the structure of a tRNA/Ribosome complex, you can start to reverse-engineer your own Ribosomes, potentially opening up a whole field of crazy-science Synthetic biology. Reverse-engineering and redesigning one of the oldest and most stable proteins in nature would be a pretty neat coup for DIYbio, don't you think?
Share and Enjoy!
If you've read this far, you must be hardcore. Share your own notions or your thoughts on what I've written above. Making the most of ZInc Fingers as this technology matures could be a great way for DIYbio to help shape the next generation of genetic technologies!
-Cathal
--
letters.cunningprojects.com
twitter.com/onetruecathal
When I started my PhD in cancer research, it was an immediate dive into the exciting field of Zinc Finger Nucleases and site-directed repair/mutagenesis! Unfortunately, my project took a turn for the boring shortly thereafter, and I never got to work with the things. Looking at the ridiculous prices Sangamo are charging for sets of Zinc Fingers now, it looks like commercial fingers will be out of the DIYbio price range for years to come.
You can make your own Zinc Fingers using techniques, tools and kits provided kindly by the Zinc Finger Consortium, but the hit-and-miss reliability requires lots of prototyping, and the steps needed to test new fingers take ages. You could order fingers from a gene synthesis company, and probably save your money's worth in time, but it's still expensive when it turns out not to work.
But don't fret, DIYbioers! All is not lost. Sure, Zinc-Finger Nucleases for your-site-of-interest work are currently impractical. So what? That still leaves a whole spectrum of use open to you for these fantastic, tiny proteins. And that's what I'd like to start a discussion on here, after offering some suggestions for cunning workarounds.
Workaround #1: Bring the Mountain to Mohammad
So, it's hard to make Zinc Fingers to target DNA reliably. But, others have already made and proven Zinc Finger sequences that target their own DNA very reliably indeed! As nucleases, this mightn't be very handy unless their target sequence happens to exist in your species of interest's DNA. However, ditch that FokI cleavage domain, and replace it with a transcriptional activator domain, and you've got a transcription factor that binds reliably to a known site.
Now, for your next project requiring transcriptional control, you can add the 20ish bp needed to the upstream region of your gene, and that's all the promoter you'll need to be certain of what's going on. No more guesswork with natural promoters, you've just turned someone else's carefully designed Zinc Finger on its head and used it as part of a sophisticated control system.
This scales up nicely with synthetic biology; if you want to add lots of complex systems to a species, you can do so as a branch of its normal transcriptional control rather than relying on the existing, complicated-and-difficult-to-predict systems. If your transcription factor is ordered with is own binding region in advance of itself, you've invented a new constitutive control region that can be reliably, predictably used for any downstream projects you care to try. I'd like to do this with Cyanobacteria!
Workaround #2: Mixing and Matching
Zinc Finger Nucleases work by conjugating the FokI cleavage domain to a pair of zinc fingers that each bind either up or downstream of the target site. Because FokI cleaves only as a dimer, the ZFN only cuts at the target site because either end won't cut alone. If they'd used a protein that could cut on its own, then it wouldn't matter how long the single-ZF nuclease was, because partial affinity to bits of DNA would let it cut up the whole genome in no time.
The nice side effect of this is that, although it's unlikely your ~20bp site matches exactly that of someone else's site previously done and demonstrated, because their carefully made Zinc Fingers are actually a pair of ~10bp targeting fingers, someone may have already targeted half of your region and come up with a working Zinc Finger Array for the job.
So, if you look for the full sequence, you're unlikely to get much. But if you search for subsections of your sequence, you might get lucky! I seem to recall that the ZFC's search tool could compare known ZFNs to a given sequence to find binding sites. This could make this job easy, but failing that a search for "Zinc Finger" plus a bit of dna code might turn something up in the literature.
Workaround #3: Stretching Your Fingers (Theoretical!)
Zinc Fingers twist as they bind DNA, and much of the annoying optimisation of getting a ZFN to cut isn't into the binding affinity of the Zinc Fingers themselves in the array, but they spacing between fingers, or between the fingers and the FokI domain. FokI needs to meet another of its kind *across* the DNA helix from itself, so correct spacing and orientation are essential.
This is annoying, but noting this property also suggests another potential avenue to re-using an existing finger, although some prototyping may still be needed. If you happen to find that your target sequence is partially covered by one prior-art ZFN array, but with another one a few nucleotides down (so close! Damn!), you might be able to bridge that gap by throwing a few spacer amino acids into the father-away ZFN, so the FokI domain is looser and stretchier. If you space things just right (I don't pretend to be an expert at this sort of math), you should be able to get the FokI domains to meet at the point you want to cut. Your accuracy will be lower, but it's saved you a lot of time if it works!
Workaround #4: Really pushing it with Adaptamers (Pure, possibly nonsense theory)
I don't recall if ZFNs are used to bind RNA, or if they can. But, assuming for a minute that they bind RNA too and can or have been used for that purpose, there are ways one might exploit that property to apply a totally unsuitable ZFN to a target region.
By having a stretch or RNA transcribed that matches the ZFN at one point and a point on your target DNA/RNA on the other, you may be able to get the ZFN to bind to the RNA, and have the RNA target either other RNA molecules (watch out for dsRNA-targeting machinery if done in relevant species) or open DNA helices, for example a replication fork or a highly transcribed region.
More promisingly, if you can convince a ZF-Array to bind RNA and then mimic the structure of a tRNA/Ribosome complex, you can start to reverse-engineer your own Ribosomes, potentially opening up a whole field of crazy-science Synthetic biology. Reverse-engineering and redesigning one of the oldest and most stable proteins in nature would be a pretty neat coup for DIYbio, don't you think?
Share and Enjoy!
If you've read this far, you must be hardcore. Share your own notions or your thoughts on what I've written above. Making the most of ZInc Fingers as this technology matures could be a great way for DIYbio to help shape the next generation of genetic technologies!
-Cathal
--
letters.cunningprojects.com
twitter.com/onetruecathal
Lee Nelson
Dec 31, 2009, 9:50:31 AM12/31/09
to diy...@googlegroups.com
region that can be reliably, predictably used for any downstream projects you care to try. I'd like to do this with Cyanobacteria!
Workaround #4: Really pushing it with Adaptamers (Pure, possibly nonsense theory)
I don't recall if ZFNs are used to bind RNA, or if they can. But, assuming for a minute that they bind RNA too and can or have been used for that purpose,
Yes, these things are possible.
Nathan McCorkle
Dec 31, 2009, 2:20:01 PM12/31/09
to diy...@googlegroups.com
Very, very cool... I could easily find the background info online, but you seem to have some level of expertise or significant awareness of the area, so maybe you could find the time to post some info on the mechanism of action of these proteins, how they're structured and how that structure is built and what makes it engineerable, or rather what elements can be engineered and how/ to what purpose (in the sense of if you change this DNA, such and such changes, and then the final structure changes in this way, which makes the final mechanics do this differently).
Great read, this is the beginning of a great FAQ or wiki page!
--
letters.cunningprojects.com
twitter.com/onetruecathal
--
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To post to this group, send email to diy...@googlegroups.com.
To unsubscribe from this group, send email to diybio+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/diybio?hl=en.
--
Nathan McCorkle
Rochester Institute of Technology
College of Science, Biotechnology/Bioinformatics
Reply all
Reply to author
Forward
0 new messages