PEDS Advance Access originally published online on August 18, 2006
Protein Engineering Design and Selection 2006 19(10):479-481; doi:10.1093/protein/gzl032
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Short Communication |
Engineering aggregation-resistant proteins by directed evolution
1 Centre for Protein Engineering, Medical Research Council Centre Hills Road, Cambridge CB2 2QH, UK 2 Laboratory of Molecular Biology, Medical Research Council Centre Hills Road, Cambridge CB2 2QH, UK
3To whom correspondence should be addressed. E-mail address: winter{at}mrc-lmb.cam.c.uk
 |
Abstract |
|---|
 Top Abstract AcknowledgementsReferences |
|---|
A method was recently described for selecting aggregation-resistant antibody domains. A repertoire of domains displayed on filamentous bacteriophage were heated/cooled and selected for binding to the affinity ligand protein A specific for the folded domains. Here we describe a generalization of the method based on the selection for retained phage infectivity and for the binding of an appended sequence tag, and applicable to any protein displayable in multivalent form on phage.
Keywords: directed evolution/protein aggregation/phage display
Protein aggregation is an obstacle in biotechnology, as many proteins aggregate when removed from their physiological context. Directed evolution has proved a powerful means to engineer protein properties (Clackson et al., 1991
; Forrer et al., 1999; Farinas et al., 2001), including resistance to aggregation in vivo (Waldo, 2003; Roodveldt et al., 2005). Recently a method was developed for the selection of aggregation resistance under non-physiological conditions. A repertoire of antibody heavy chain variable domains (dAbs) of the VH3 family was displayed multivalently on the pIII coat proteins on the infective tips of fd bacteriophage, the phage transiently exposed to 80°C in phosphate-buffered saline (PBS) (thereby promoting aggregation between the aggregation-prone dAbs), and then the phage carrying aggregation-resistant dAbs were selected by using an immobilized ligand specific to the natively folded dAbs (protein A) (Jespers, 2004a). As aggregation occurs in vitro non-physiological conditions can be used; indeed the phage can withstand high temperatures (>80°C), denaturants (10 M urea or 8 M guanidine hydrochloride) and extreme pH (pH 2–12) (Kristensen and Winter, 1998; Holliger et al., 1999). However, a major limitation in generalizing this method is the requirement for a ligand for the folded protein, and here we describe an alternative solution.
Aggregation on phage is a self-specific, two-step process akin to the nucleation-growth mechanism by which many proteins aggregate in solution (Fink, 1998). First, the displayed proteins form aggregate nuclei on the phage tips, cross-linking the pIII coat proteins to which they are fused. The aggregates then grow by joining such nuclei, generating tip-to-tip phage clusters (Jespers et al., 2004a). Functional pIII proteins are essential for pilus adhesion and initiation of infection of Escherichia coli cells (Riechmann and Holliger, 1997), and we have previously proposed that their cross-linking and burial in phage clusters upon aggregation lead to reduced phage infectivity (Jespers et al., 2004a). Here, we exploited the link between loss of phage infectivity and aggregation.
The effect of heating phage that displayed ß-lactamase (BTL) was first examined since this protein (Iwai and Plückthun, 1999), like some dAbs (Jespers et al., 2004a), is known to aggregate in solution upon heating. As with the dAb DP47d, this protein aggregated on phage as manifested by extensive formation of tip-to-tip clusters detected by transmission electron microscopy (Figure 1a–b) and had reduced infectious titres compared with wt phage (Figure 1c). In contrast wt phage and phage dAbs selected previously from protein A based selections (Jespers et al., 2004a) retained phage infectivity (Figure 2a).
|
|
We used the loss of infectivity as the basis for our selections. Phage [1 x 1012 Transducing Units (TU) per ml] were heated at 80°C for 10 min in PBS and then used to infect exponentially growing E.coli TG1 cells. In a test selection from a 1:1000 mix of wt phage to BTL phage, wt phage was enriched 16-fold per round after heating; similarly from 1:1,000 mix of HEL4 phage (HEL4 is an aggregation-resistant dAb, Jespers et al., 2004b
) to DP47d phage, the HEL4 phage were enriched 26-fold per round. This suggested that phage infectivity alone could be used to select proteins that resist heat aggregation from those that do not.
To enable a direct comparison of the infectivity selection with the protein A-based selections described earlier (Jespers et al., 2004a), we attempted to isolate aggregation-resistant dAbs from the same DP47d-based library as before, using three rounds of selection. All the isolated clones largely retained infectivity on heating, but several only recovered a fraction of their protein A binding (Figure 2b). Western blots of the phage suggested this was due to a decrease in display levels (blots not shown; for methods see Jespers et al., 2004a). On average for the phage clones isolated from infectivity selections (Figure 2b), 34% of the pIII proteins were decorated with dAbs, contrasting with the 61% for those isolated from the protein A selections (Figure 2a). The earlier work (Jespers et al., 2004a) had shown that the aggregation of the phage dAbs is strongly dependent on the valency of display; it appears that some phage with aggregation-prone dAbs get through our infectivity selections by loss of dAbs from the phage tip (probably due to proteolytic cleavage of the dAb from pIII before and/or during phage assembly). In contrast the protein A selections (Figure 2a) maintain pressure both for the retained native dAb fold and for good display levels.
To maintain the selection pressure for display, we incorporated a sequence tag (His6) as an N-terminal fusion to two model phage dAbs (HEL4 and the clone INF18 with 70 and 2%, respectively, of pIII proteins decorated) and also to the DP47d-based library. In selections, the (His6)-dAb phage were first bound to a nickel-nitriloacetic acid (Ni-NTA) matrix, washed with 50 mM imidazole to remove weakly bound (low display level) phage and then eluted with 500 mM imidazole (further details in figure legend). When we mixed the two model (His6)-dAb phage 1:1000 (HEL4 to INF18), the well-displayed (His6)-HEL4 phage were enriched 320-fold in the eluate from the affinity matrix. We operated the display and infectivity selections sequentially, thus enriching for phage that display dAbs well, and then for those that do not aggregate on heating. After three such rounds on the (His6)-DP47d-based library, all the isolated phage retained binding to protein A after heating (Figure 2c), and the average display level improved to 51% with no clone under 30%.
Five of the phage dAbs emerging from the selection were expressed as soluble free (untagged) dAbs by secretion from E.coli, and their aggregation propensities determined in solution at 80°C in PBS. As readout for aggregation upon heating, the solution turbidity was followed over time (Figure 2d). The selected dAbs showed no sign of aggregation during the first 10 min (the exposure time in the selections), in contrast to DP47d, which aggregated within 1 min. Further, three of the five selected dAbs had longer lag-times than HEL4, and one showed no sign of aggregation even after 15 h at 80°C.
Our results suggest that infectivity-based selections can provide a method to evolve aggregation resistance of any protein that can be displayed multivalently at the tips of filamentous phage. Like the earlier method (Jespers et al., 2004a), the method developed here uses an affinity ligand, but this ligand is directed to a sequence tag rather than the folded protein, and this should facilitate its wider application. As the method does not select for folded protein per se, it should also be applicable to the evolution of aggregation resistance in natively unfolded proteins. Although the method could generate aggregation-resistant proteins that are unfolded, we have not seen this with the dAbs; furthermore, this could be eliminated by use of proteolysis and/or functional screens (Finucane et al., 1999; Riechmann and Winter, 2000).
|
Footnotes |
|---|
Edited by Jane Clarke
|
Acknowledgements |
|---|
|
TopAbstractAcknowledgementsReferences |
|---|
The authors thank Daniel Christ and Laurent Jespers for helpful discussions on the design of the selection process.
|
References |
|---|
|
TopAbstractAcknowledgementsReferences |
|---|
Clackson T., Hoogenboom H.R., Griffiths A.D., Winter G. (1991) Nature 352:624–628.[CrossRef][Medline]
Farinas E.T., Bulter T., Arnold F.H. (2001) Curr. Opin. Biotechnol. 12:545–551.[CrossRef][Web of Science][Medline]
Fink A.L. (1998) Fold. Des. 3:R9–R23.[CrossRef][Web of Science][Medline]
Finucane M.D., Tuna M., Lees J.H., Woolfson D.N. (1999) Biochemistry 38:11604–11612.[CrossRef][Medline]
Forrer P., Jung S., Plückthun A. (1999) Curr. Opin. Struct. Biol. 9:514–520.[CrossRef][Web of Science][Medline]
Holliger P., Riechmann L., Williams R.L. (1999) J. Mol. Biol. 288:649–657.[CrossRef][Web of Science][Medline]
Iwai H. and Plückthun A. (1999) FEBS Lett. 459:166–172.[CrossRef][Web of Science][Medline]
Jespers L., Schon O., Famm K., Winter G. (2004a) Nat. Biotechnol. 22:1161–1165.[CrossRef][Web of Science][Medline]
Jespers L., Schon O., James L.C., Veprintsev D., Winter G. (2004b) J. Mol. Biol. 337:893–903.[CrossRef][Web of Science][Medline]
Kristensen P. and Winter G. (1998) Fold. Des. 3:321–328.[CrossRef][Web of Science][Medline]
Riechmann L. and Holliger P. (1997) Cell 90:351–360.[CrossRef][Web of Science][Medline]
Riechmann L. and Winter G. (2000) Proc. Natl Acad. Sci. USA 97:10068–10073.
Roodveldt C., Aharoni A., Tawfik D.S. (2005) Curr. Opin. Struct. Biol. 15:50–56.[CrossRef][Web of Science][Medline]
Waldo G.S. (2003) Curr. Opin. Chem. Biol. 7:33–38.[CrossRef][Web of Science][Medline]
Received July 3, 2006; accepted July 10, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Dudgeon, K. Famm, and D. Christ Sequence determinants of protein aggregation in human VH domains Protein Eng. Des. Sel., March 1, 2009; 22(3): 217 - 220. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Liu, A. V. Mack, M.-L. Tsao, J. H. Mills, H. S. Lee, H. Choe, M. Farzan, P. G. Schultz, and V. V. Smider Protein evolution with an expanded genetic code PNAS, November 18, 2008; 105(46): 17688 - 17693. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||