Application and modification of combinatorial domain hunting.
Doctoral thesis, UCL (University College London).
Combinatorial Domain Hunting (CDH) is a recently introduced technology that addresses the problem of obtaining solubly expressed protein domains, or more correctly polypeptide truncates, from a larger target. CDH relies upon enzymatic fragmentation of a template gene, non-directional cloning of these fragments into a set of bespoke expression vectors, and screening of the resulting library of clones for the expression of soluble, affinity-tagged, protein fragments. Prior to the work described in this thesis the CDH concept had only been described in outline. The thesis describes various aspects of the application of CDH to a model target, incorporating modifications to the CDH pipeline that were adopted on the basis of this application, as well as development of an extension to the CDH concept that introduces the potential to screen for soluble expression of high-affinity protein-protein complexes. CDH was applied to Hsp90, a well-characterised molecular chaperone with predicted modular structure. Following fragmentation of the Hsp90 cDNA, a library of 15,000 clones was produced and 11,500 colonies were screened for expression of soluble Hsp90 truncates, resulting in 107 preliminary hits, of which seven were duplicates. Assessment of a subset of these hits indicated a significant potential for soluble protein aggregates to survive the affinity purification/ filtration stage of CDH. Implementation of an additional centrifugation step, although initially at some cost to the overall timescale of a CDH screen, was shown to reduce the number of putative false-positive, soluble protein aggregates identified (by 45% in the case of Hsp90, resulting in 55 unique hits). The advantage derived from this modification is that fewer preliminary hits would be taken forward for biochemical and structural validation. The original CDH concept invoked the use of NMR spectroscopy for the assessment of truncate foldedness. In the context of the real world application of CDH to Hsp90, it was found that such assessment was inefficient in terms of the amount of material required per preliminary hit. The thesis describes the application of differential scanning fluorimetry (DSF, also known as ‘Thermofluor’) analysis to the collection of Hsp90 CDH hits, demonstrating significant efficiency savings in the assessment of foldedness, most dramatically in that much smaller amounts of protein were required than for NMR. For the Hsp90 screen, DSF analysis indicated that four out of the 41 (14 of the 55 did not express protein) preliminary CDH hits corresponded to folded truncates (validated by 1D 1H NMR). The 41 hits were also characterised by size exclusion chromatography, where 12 eluted mainly in the void volume, and 11 others with greater than 40% in the void volume, possibly indicating that the centrifugation step applied at small scale, does not reliably eliminate those truncates with the potential to self-associate at a preparative scale. A CDH screen, including the modifications derived in the Hsp90 example, was applied to the Mycobacterium tuberculosis histidine kinase DosS. 52,000 colonies were screened of a 100,000 clone library which resulted in 23 preliminary hits. Bioinformatic analysis indicated that one of the hits covered the putative tandem histidine kinase and histidine kinase acceptor domains. The thesis describes biochemical and biophysical validation of the function and foldedness of this DosS truncate, and enumerates initial attempts at crystallisation and NMR analysis. For a given protein target, identification of folded domains may require the presence of an obligate partner chain: with this in mind, this thesis introduces a novel adaptation of the CDH paradigm, dubbed 2C-CDH, in which the fragment library of one partner protein isolated with one affinity tag, is screened on the basis of immunodetection of an orthogonal tag appended to the other partner. The design and implementation of an appropriate two plasmid orthogonal expression-and detection system is described. A proof-of-principle application of the 2C-CDH concept is provided, using the complex formed between Hsp90 and Cdc37 proteins. Two screens were performed, using a known soluble fragment of either Hsp90 or Cdc37 (as bait) screened against a fragment library (as prey) derived from Cdc37 and Hsp90 respectively. In each case the 2C-CDH procedure identified one or more truncates in the prey library with amino acid sequence boundaries consistent with the known characteristics of the of the Hsp90-Cdc37 structure. Moreover, for each hit, size exclusion chromatography analysis validated maintenance of a well-behaved 1:1 complex at preparative scale. This thesis concludes with a commentary upon the CDH concept in the context of related current and possible future paradigms for soluble protein expression.
|Title:||Application and modification of combinatorial domain hunting|
|Additional information:||Authorisation for digitisation not received|
|UCL classification:||UCL > School of Life and Medical Sciences > Faculty of Life Sciences > Biosciences (Division of) > Structural and Molecular Biology|
Archive Staff Only