Exclusive Interview

Hartley: I had just spent a long time constructing an expression clone for the 10 kDa protein ladder, and it expressed very poorly. Mike Brasch and I had an interest in using site-specific recombination (SSR) in vitro, and we agreed that there must be some way to use SSR to move genes into new vectors. Our initial version was crude, but over the next year various permutations were tried that led pretty much to the present Gateway configuration. We must also acknowledge the wonderful generosity of Howard Nash at the National Institutes of Health, who provided samples of Int, Xis, and IHF proteins that enabled us to test our ideas.

Hartley: We thought moving genes into new vectors would be the primary benefit of Gateway. As the technology developed and we used it at NCI, another aspect turned out to be really important: the simplicity and reliability of Gateway cloning. Cloning PCR products with the BP reaction works >99% of the time, probably because the pDONR™ vectors are transcriptionally silent and high-copy plasmids. That was first demonstrated by my former colleague Gary Temple, who is now running the Mammalian Gene Collection at the NIH’s National Human Genome Research Institute. And the subcloning LR reaction always works. It is somewhat painful to go back to restriction enzymes and ligase.

Esposito: Recombinational cloning has really become the technology of choice for both small cloning projects and especially high-throughput genomics cloning projects. Gateway cloning has a number of advantages over other recombinational cloning technologies, particularly with regard to the efficiency of the reactions and the ease of use and flexibility of the destination vector. As Jim points out, the simplicity can’t be beat, either. It takes a technician about 60 seconds to set up a Gateway reaction. The overall hands-on time for a BP or LR reaction from the initial reaction mix setup to picking colonies from a plate is about 10 minutes. Coupled with the fact that you almost never need to redo a Gateway reaction—unlike the multiple attempts commonly necessary for restriction enzyme cloning—it’s really a no-brainer.

Hartley: Led by Marc Vidal, the ORFeome community is a major user of Gateway cloning. People working with C. elegans use a Gateway version of Andrew Fire’s vector to test hundreds of ORFs for RNAi effects by feeding worms with E. coli. Because it is easy to convert vectors to Gateway compatibility, dozens of specialized viral—retroviral, lentiviral, adenoviral—and plant Gateway destination vectors have been described. Jonathan Melnick profiled 90 human kinase activities in vivo using a Gateway retroviral vector. Maghsoud Pazhouhandeh made two split YFP destination vectors so that protein-protein interactions resulted in fluorescence by bimolecular complementation.

Hartley: We use Gateway cloning almost exclusively to make protein expression constructs for our NIH client investigators. For our own R&D we use a pooling strategy to test hundreds of human ORFs for clones that express well. In the course of meeting our internal and external needs we have accumulated over 150 destination vectors for expression in bacterial, insect, yeast, and mammalian cells.

Esposito: A common pathway for us is that an NIH investigator needs to clone a gene or genes for various purposes, whether these be structural studies, antibody generation, in vivo localization, pulldowns, two-hybrid, or biochemical studies. Each of these studies requires different approaches to the cloning and expression of proteins. For instance, someone interested in pulldowns needs a GST fusion, and perhaps an epitope tag for visualization. But he or she may also want to purify protein, and therefore needs to have an affinity tag and maybe a protease cleavage site to remove the tag. The investigator may also want to try different hosts for expression in case one of them doesn’t pan out. Instead of going through multiple cloning and sequencing steps to generate five to 10 different vectors, we can make one or two carefully constructed entry clones, and in a single day generate expression clones for all of these purposes. The time savings are dramatic, and the cost savings to the investigator are even greater. Instead of waiting to see if Pathway A works before spending time cloning a construct for Pathway B, we can now do it all up front and have all bases covered. Given the increasing number of choices for expression hosts, constructs, and fusion tags, this makes Gateway cloning even more essential.

Another common pathway that we focus on is enhancing soluble expression in E. coli. This often requires testing a number of solubility tags coupled with affinity tags and protease cleavage sites to remove the tags after purification. Quite often there is no a priori way to know which tags are better for a given protein, and Gateway cloning allows us to quickly manufacture expression clones side by side with four to eight different tag combinations, and with the exact same backbone. This produces a very fair test of the effect of the solubility tag and, again, allows us to make all of these clones in a single day without the need for sequencing. This parallel processing can mean the difference between trying one or two tags that fail to make soluble protein and having the ability to test eight clones to find the one tag that does make a difference.

Hartley: We’ve used MultiSite Gateway cloning a number of times to assemble complex plasmids. We made an NCI client very happy recently when a MultiSite Gateway construct coexpressing two interacting proteins yielded a soluble, purifiable complex that crystallized and diffracted well. The structure turned out to have a nice biological implication. We also find Multisite Gateway a handy way to assemble reporter constructs used in making mice with glowing tissues. David Cheo deserves the bulk of the credit for developing the MultiSite Gateway technology.

Esposito: MultiSite Gateway cloning offers a large amount of flexibility. We have made reporter constructs by combining promoter fragments and standard reporter entry clones into nontraditional destination vectors, and have used these to make transgenic mice. But we have also used nonstandard entry clones and other tags in standard Gateway lentiviral vectors to generate reporter lentivirus clones. In both cases the power of MultiSite Gateway is really limitless, and the technology is nearly as efficient and easy to use as regular Gateway cloning. See more information about MultiSite Gateway cloning.

Hartley: MultiSite Gateway cloning will be more and more important for expression in mammalian cells and live animals. This will be driven by the diversity of promoters and reporters needed to answer important biological questions.

Esposito: There needs to be more of an effort to make a centralized repository of Gateway destination vectors where researchers can go to share and see what vectors other people have made, how well they work, and what they can be used for.

The other area that needs to be developed quickly is ORF libraries. Right now there are a number of sources of Gateway-compatible ORF libraries. The protein expression field would probably prefer to see clones with a protease site up front or perhaps a 6xHis tag at the C-terminus. A concerted effort to get everyone to standardize a set of clones would be extremely useful for all scientists working on protein expression.