As protein crystals are at the bases of protein crystallography, the sufficient amounts of pure protein are at the beginning of such crystals growth. And even though astrophysicists are today theorizing about the negative energy of empty space and implicating that whole universes could arise from nothing [1, 2], modern era of structural biology is still depending on heterologous recombinant protein expression as a primary source of well folded proteins.Thus although sometimes quite costly enterprise, as a primary source of material for biomedical applications and research, it is a necessary one. Unfortunately, there is no single recipe for expression of all protein constructs that would guarantee high yields. While some proteins are quite easily expressed in prokaryotic cells (either in soluble or renaturable insoluble state), for many we need more „sophisticated” approaches and some we label as the „difficult targets “. Moreover, as the “easy targets” ale less abundant in these days the strategies for more “sophisticated” approaches are more and more in demand.
Here, on a klr and clec gene families [3] – a families of lymphocyte receptors that deserve the label of “difficult targets”, we would like to showcase our efforts and different expression strategies that we employed in overcoming such difficulties. Our first efforts to study the ectodomains of these proteins on a structural level via recombinant expression in E. coli have shown that many aggregate to inclusion bodies and their refolding is inefficient. Although some perform well in transiently transfected human embryonic kidney 293 (HEK 293) cells, in some cases, an addition (restoring disulphide bond) or removal of cysteine residue is necessary for high yield of well folded product. For optimization of expression construct of klrb1 we have used a high-throughput approach – combining 26 different constructs with 8 different affinity tags and tested the for expression both in prokaryotic, insect and mammalian cells without major yield improvement.
Finally, as a tool of last resort, we have tested stably transfected pools of HEK293S GnTI-[4] cells with either stochastic or controlled integration of expression cassette into the host genome. While the randomly integrated stable cell pools provided and excess of 10-fold improvement yield-wise, the TetOn inducible expression system employing the controlled integration via the piggyBac DNA transposase [5] is performing even better with roughly 20-fold improvement in yield, compared to transient transfection of HEK293S GnTI- cells.