2014-08-12 Progress Report

Williams

There are a couple of very informative papers by James Williams from Nature Technology Corporation (NTC): (Williams et al. 2009, Williams 2013). For some reason the 2013 paper ("Vector Design for Improved DNA Vaccine Efficacy, Safety and Production") is not on pubmed so I am including the title here.

Williams points to pVAX1 and gWIZ as "first-generation" DNA vaccine vectors. pVAX1 an extremely minimalist vector, much more so than other "expression vectors" I've seen for sale. It's basically the vector I've been expecting but unable to find. However, Williams claims that the performance of first-generation vectors is poor due to several poor design choices. Williams' company (NTC) makes a "third-generation" vector that has a number of claimed advantages: mainly increased expression but also adherence to regulatory guidelines.

One modification NTC has made is that their vector is not selected for using antibiotics like pVAX1. There is a good reason for this (see below), but it means that the vector apparently must be grown up in a specialized E. coli...

The European Medicines Agency (EMA) has further concluded that kanamycin and neomycin are of importance for veterinary and human use and cannot be classified as having minor therapeutic relevance due to current use in critical clinical settings. To address these regulatory concerns, alternative non-antibiotic selection methods are needed.

Software

I went back to Teselagen and Benchling to see if either had a straightforward way to embed a protein in a pre-defined expression vector. If there was, I did not find it... (In retrospect, Genome Compiler may be better at this.) I think the problem is that these programs are really designed for operating at a lower level than I want. If I want to manipulate DNA sequence, I'm not sure I want to use a GUI at all. Rather, I want something more like a wizard that solves my specific use-case and assures me that all the hard-to-check things (protein phase, RNA secondary structure) are ok. I'm sure I'll come back to try these programs again though...

The best software I have found so far for my needs is DNA 2.0's website. Unfortunately, to my knowledge, DNA 2.0 does not have a DNA vaccine vector and the cost per base is prohibitively high.

Commercial vectors

Most vectors seem to be designed for producing and harvesting recombinant protein, so there's a focus on tagging and protein recovery. I don't need that, and it could even be detrimental to protein function. After reading Williams' paper, it turns out these are just the wrong vectors and that there are better choices.

These Sigma Aldrich vectors, which are a good example of the typical vector product page. Confusingly, there are "transient" or "stable" types — the "stable" expression vectors have an additional neo gene (neomycin), which allows you to select cells where rare integrations have occurred. Every vector has 1-3 N- or C-terminus FLAG peptides (22-66 amino acids) and/or c-Myc to allow for protein purification.

DNA 2.0 also has a nice tool to select from their mammalian expression vectors. These vectors are very inexpensive at $200 but again they are not designed for use as DNA vaccines.

One commenter notes that he does not like Life's pSecTag vector because it includes NcoI cleavage sites, which means he cannot use this restriction enzyme to clone into the MCS. That's something I didn't think about...

Next steps

I'll need to reread the Williams papers, and then pick one of the few available DNA vaccine-specific vectors, and see if it's possible to use it in combination with my protein.