Bench philosophy: The Next Big Thing?

by Steven buckingham, Labtimes 07/2012

Nature is notorious for hiding her secrets from the probing scientist. But every so often she seems to throw a clue and tells us a tale.

Remember RNAi? There we were working late shifts to knock down genes, then along comes Andy Fire with a note from Nature to “try double stranded RNA”. Easy knockdowns for us, Nobel Prize for Andy.

Now it seems Nature has tossed us another bone. Thanks to TALENs, (transcription activator effector nucleases), it seems we have been endowed with a tool that can cut and paste the genome at will.

We owe it all to the pernicious habits of Xanthomonas. This nasty bug infects crops by secreting a protein called transcription activator-like effectors (TALEs). As their name suggests, they act like the host plant’s own transcription factors – the regulatory elements in the DNA that control gene expression. With sequences that are designed to bind to the gene regulatory site, they fool the plant’s DNA to start expressing genes favourable to the invader. They are genome hijackers – effectively reprogramming the host cells by imitating the cells’ transcription factors. How convenient. Now, let’s see if we can do a bit of hijacking of our own.

A plague for pepper and tomatoes but a boon for life scientists trying to edit genes: Xanthomonas transcription activator-like effectors.

Let’s take it one step further. TALENs are an adaptation of TALEs that allow DNA to be mutated at just about any location. All you do is stick a nuclease onto the end of your TALE (the temptation of reporter puns is strong...but I promise to remain professional). What you get is a construct that targets any point in the genome you like and then cuts it where you like. This is a whole step up from, say, zinc-finger based mutagenesis, which restricts cutting to locations near a limited range of sequence motifs.

Versatile repeat motif

So, how are they able to do this? The secret is in the versatility of their DNA-binding motif. All naturally occurring TALEs (the ones Xanthomonas uses to make pepper plants sick) have a 32-34 repeat motif that does the DNA binding. Interestingly, these amino acid repeat sequences are almost identical – all, that is, except for a single pair of highly variable residues in each repeat.

This is the secret to TALE specificity; the key that researchers have hacked to their advantage. Three years ago, two papers appeared in the journal, Science, announcing that the secret to TALE specificity had been discovered (Moscou and Bogdanove, Science, 326, 1501; Boch et al., Science, 326, 1509). And yes, you guessed it; it is those non-conformist pairs of amino acids. They are known as the hypervariable pair: “the awkward couple” in each repeat.

Jens Boch and his colleagues from the University of Halle, Germany described what happened when they took a close look at a TALE called AvrBs3. Now, AvrBs3 targets a gene regulatory sequence that controls the expression of a number of genes, including the Bs3 resistance gene in pepper plants. This target stretch is known as the UPA box (UPA? Oh, that stands for ‘upregulated by AvrBs3’ – don’t worry, it all makes sense). What struck Boch and his team was that the number of repeats in the TALE tail matched the number of base pairs in the UPA.

So, could it actually be as simple as one awkward couple for each base pair? When Boch’s team looked more closely, they noticed there was indeed a correspondence, between which two amino acids made up the awkward couple and the identity of the base pair.

But could it just be a coincidence? The team then looked up the sequences of other TALEs, paying special attention to their awkward couples. From that they predicted, if their theory is right, there should be promoters somewhere in the target genomes that match these pairs. And there were. Better still, these promoter sequences only showed up in the alleles that are induced by the corresponding TALE.

Interestingly, there was a variation in the specificity of each awkward pair for their corresponding DNA base pairs – some were highly specific, others less choosy. Most useful. So, just how are we going to use this to our advantage? Having a way to make a DNA-binding protein to order, to any sequence you like, is exciting. First of all, how about putting a cutting sequence on the end? Then you can cut the DNA wherever you like – not just at restriction sites. TALENs are just that: TALES with a nuclease on the end. (See? No puns, just as I promised).

Exploiting Nature’s errors

That in itself opens up all sorts of possibilities. Just cutting the DNA causes some pretty useful reactions to kick in. First of all, the genome reacts by trying to fix the break. But the convenient thing is, it often gets it wrong, adding in an extra base pair by mistake or perhaps lopping off a base pair from the ends before stitching everything back together again.

Bad for the genome, good for the scientist. Indels, they call them: insertions or deletions that induce a frame shift or nonsense mutation, effectively taking that gene down.

So, now I can put an indel in the genome anywhere I want. In the ‘old’ days, (i.e. before 2009 – yes, even that far back) it was a bit trickier – some places were easy to squeeze an indel into, others impossible.

The research community was quick to latch on to this. Only this year, David Langenau’s group at the Harvard Stem Cell Institute showed just how easy it is to design TALENs to target zebrafish genes (Moore et al., PLoS ONE 7(5): e37877). Far easier, in fact, than the hitherto established methods. As has become the custom, they adapted the TALEN approach to make it even more specific. The nuclease they stuck on the end of their TALE was one called “FokI”. This domain only works as a dimer. So all you have to do, is to make two TALEs (I really do feel a cheesy pun coming on), flanking both sides of the site you want to cut. That adds extra stringency, reducing the chances of a random, erroneous cut caused by imperfect specificity of the TALE sequence.

Outperforming zinc finger proteins

Langenau’s main purpose in their study was to compare the TALEN approach with the existing standard technique that uses zinc finger nucleases (ZFNs). At first sight, ZFNs and TALENs work pretty much the same way. ZFNs have repeat domains that bind specific sequences and you just stick a nuclease on the end, just like for TALENs. But Langenau showed that TALENs outperform ZFNs in many ways. For one, they are more versatile. You can only make ZFNs effectively for about one in every 500 base pairs of random DNA, restricting the places you can cut. Secondly, ZFNs don’t seem to do the job as well as TALENs, giving lower mutation rates on average and a more variable performance.

It is not just about taking genes out – the prospects are growing for using TALEs to put genes in. This exploits another feature of how genomes repair double-stranded breaks and it has to do with homologous recombination. During the cell cycle, if you remember, chromosomes get cut and do a little swapping around. The same mechanism kicks in when a TALEN makes a double-strand cut. All we have to do is to make sure the right bit of recombinant DNA just happens to be around at the time. The recombination machinery incorporates your custom-made transgene at the site of the DNA break, fixes it all up for you and off you go.

Designer TALENs

Now, I don’t want to make it look like life is always that easy. You have to design your TALEN first and that has issues of its own. First of all, that long string of repeats causes problems when it comes to annealing. The trick is to design your repeats with care. There is a host of commercial and freely-available programmes to help do that. TALEN-pioneer Adam Bogdanove and his colleagues at the Iowa State University recently published a set of guidelines, based on natural sequences of TALEs (Nucl. Acids Res. (2011) 39 (12): e82.). What is more, they have encapsulated these guidelines in a freely available programme online (

Or you can pay someone else to do the hard work. A quick Google of “TALEN design” will bring up a list of companies that will design and synthesise TALENs to order, with all the extras you can expect when you pay the experts.

Either way, since the arrival of the TALEN­ technique, RNAi and zinc finger proteins are starting to look sooo 2011. And thereby hangs the TALE (sorry, I couldn’t help it).

Last Changed: 07.02.2013