Nanobody is a term that refers to a single variable domain (VHH) derived from heavy chain only antibodies (IgG) found in camelid species such as alpaca (Hamerscasterman et al., 1993;Muyldermans et al., 1994;Muyldermans & Lauwereys, 1999). These nanobodies are a fraction of the size of conventional antibodies (full length 160 kDa), with a molecular weight of around 15 kDa (Doshi et al., 2014). Their small size, 2.5 x 4.2 nm, gives them certain advantages over antibodies (Revets et al., 2005;Muyldermans et al., 2009), such as a much smaller minimal linkage distance between the tag and the target, which is of particular value for super-resolution imaging and single molecule tracking (Ries et al., 2012;Leduc et al., 2013)
Leduc et al. (2013) have already shown that nanobodies to GFP can be used as a means for labelling proteins of interest with gold nanoparticles (Leduc et al., 2013). This method requires an EGFP fusion of the protein of interest, which we already had available for HIF-2 and was, therefore, appealing to enable the study of HIF-2α at the single molecule level.
Figure 5.17 ǀ Labelling cells expressing EGFP-HIF-2α with anti-GFP nanobodies.HeLa cells transiently transfected with EGFP-HIF-2α were fixed, permeabilised and labelled with anti-GFP nanobodies. i) nuclear stain (Hoechst), ii) EGFP-HIF-2α, iii) ATTO594 conjugated anti-GFP nanobodies, iv) merge of ii & iii. Yellow indicates colocalisation.
Preliminary experiments show that highly specific labelling of EGFP-HIF-2α with GFP nanobodies (conjugated with fluorescent dye ATTO594) can be achieved using standard immuno-fluorescent protocols (Figure 5.15). Figure 5.16 shows an example of a HeLa cells ectopically expressing EGFP-HIF-2α that has been microinjected with the anti-GFP-ATTO594 nanobodies. Nuclear colocalisation between the green and red channels can be observed (Figure 5.16iv.). These promising results suggest that ectopically expressed EGFP-HIF-2α
could potentially be labelled with GNPs via anti-GFP nanobodies in live cells using microinjection as the delivery method.
Figure 5.18 ǀ Microinjection of anti-GFP nanobodies into cells ectopically expressing EGFP-HIF-2α.
Live HeLa cells transiently transfected with EGFP-HIF-2α were injected with anti-GFP nanobodies.i) Brightfield, ii) EGFP-HIF-2α, iii) anti-GFP ATTO594 nanobodies, iv) merge of ii & iii.
Incorporation of the anti-GFP nanobodies onto the surface of the gold nanoparticles requires a different protocol to that of the Halotag-ligand which would need optimisation. As I was approaching the end of my time in the laboratory, and the size and scope of this work was unknown, the GFP-nanobody project was passed onto another member of the laboratory and is currently under development.
5.3
Conclusion
The aim of the work described in this chapter was to develop a method of labelling single molecules of HIF-2α with gold nanoparticles to facilitate their tracking in live cells. The initial plan was to use the Halotag, an enzyme that specifically and irreversibly binds to chloroalkanes. In collaboration with Promega, a Halotag ligand was specifically designed so that it could be incorporated into the ligand shell of a gold nanoparticle but protrude far enough so that it could react with a Halotag fusion protein.
Preliminary ‘proof of principle’ experiments were conducted with fibroblast growth factor-2 (FGF2). Recombinant Halotag-FGF2 was expressed in bacteria and purified using a three- step chromatography protocol. We showed that FGF2 could be coupled to gold nanoparticles via the Halotag technology.
However, when we moved on to the protein of interest, HIF-2α, several issues were encountered. Despite work published in Sun, C. et al (2015) suggesting that having the Halotag at the N-terminal of the fusion protein can aid bacterial expression of recalcitrant
protein and although many different bacterial strains and variations of protocols were trialled, expression of full-length recombinant Halotag-HIF-2α in useful amounts was not achieved. The poor expression observed could be due to the large size of the protein or the large proportion of rare codons in the coding sequence HIF-2α. The latter interpretation is supported by the marginal improvement in expression observed in the Rosetta strain. At present there are no crystal structures of full length HIF-2α available, only structures of the HIF-2α PAS-B domain (Erbel et al., 2003;Card et al., 2005;Scheuermann et al., 2009;Key et al., 2009;Scheuermann et al., 2013;Rogers et al., 2013;Guo et al., 2015), suggesting that others may have also struggled to express full-length recombinant HIF-2α.
Secondly, we encountered problems with reproducibility with regards to nanoparticle production / conjugation. Although it was shown that FGF2 could be conjugated to gold nanoparticles via the Halotag, we were unable to consistently conjugate HIF-2α to HL-GNPs. One possible reason for this is that the Halotag-FGF2 / nanoparticle conjugation reaction occurred in more favourable conditions (in vitro) i.e. as we are able to express & purify recombinant Halotag-FGF2, the reaction occurred in excess of Halotag-FGF2 and with no interference from other proteins. In the case of Halotag-HIF-2α, we seemingly could on express full-length Halotag-HIF-2α in mammalian cell. However, any protein expressed was only visible via western blot (an amplification technique) and not on an SDS PAGE stained with coomassie blue. It is difficult to calculate the concentration of fusion protein that is present in the lysate, but we can be sure that it is far less than that of the purified Halotag- FGF2. In addition to this, there is competition from other endogenous proteins that may non-specifically interact with the surface of the nanoparticle, blocking Halotag-HIF-2a from reacting with the Halotag-ligand. Both of these factors would reduce the chances of successful conjugation.
Another explanation for this could be that the Halotag ligand is inaccessible to the Halotag protein. The Halotag ligand was designed to protrude far enough out of the ligand shell so that it can react with the Halotag protein, however, there are hydrophobic regions along the ligand (for ligand structure see Appendix 1.5) and it is possible that this causes the Halotag ligand to reorient itself so that these regions are immersed in the ligand shell, thus resulting in the Halotag protein being unable to react with the chloroalkane part of the ligand.
Despite lack of reproducibility, the results presented here suggest that it is possible to conjugate HIF-2α to gold nanoparticles via the Halotag protein. With improvements to the design of the Halotag ligand and optimisation of a robust protocol for making Halotag ligand gold nanoparticles, this would be an invaluable technique for labelling proteins of interest for single molecule tracking.
Nanobodies are another promising and versatile method for labelling a protein of interest with gold nanoparticles. The anti-GFP nanobodies can be used with any GFP (and GFP derivative)-fusion proteins. They have been shown to have uses in numerous microscopy techniques, not only for single molecule tracking but also for super-resolution microscopy (Leduc et al., 2013;Ries et al., 2012)