Engineered pores for stochastic sensing and single molecule studies

Thesis

Stochastic sensing is a powerful approach for the detection of a range of analytes, yielding information at the single molecule level. Protein pores engineered for the binding of desired analytes are inserted into planar lipid bilayers and analyte binding is detected by observing the modulation of current flowing through the pore under an applied potential.

In this work, two complementary approaches for the detection of protein analytes were investigated. Chapter 2 describes the development of genetically encoded hetero-heptameric αHL stochastic sensors that bear a single Pim kinase consensus substrate sequence at the trans mouth of the pore. These were achieved by insertional fusion in the trans loop of a single subunit of the pore to yield a loop-constrained peptide, or via a novel method wherein site-specific proteolytic cleavage of the trans loop fusion yielded a peptide attached by a single terminus. The affinity of Pim-1 for the proteolytically-cleaved sensor was determined to be 99 nM. Additionally, the addition of MgATP to the sensor in the presence of Pim-1 produced an irreversible change in current behaviour that is consistent with phosphorylation of the sensor.

The sensors and methods outlined here allow for sensitive, label-free detection of Pim-1 kinase. Further, the observation of phosphorylation in the presence of MgATP would suggest that it should be possible to observe catalytic turnover of the sensor substrate if the kinase activity is coupled with an active phosphatase. Further, it should be straightforward to modify the sensor to facilitate the screening of ATP-competitive inhibitors. Hence, the sensors developed here will be useful for the sensitive, label-free detection and assay of kinases by stochastic sensing.

In Chapter 3, a scheme is proposed in which outer membrane protein G (OmpG) pores that bind a desired target are selected from a combinatorial peptide library inserted in an extramembranous loop of OmpG. This would allow the selection of stochastic sensors against targets for which no suitable ligand exists. The suitability of OmpG for bacterial display was first demonstrated, before construction of the combinatorial library. However, OmpG expression had an adverse effect on cell viability that prevented successful screening. Additionally, a rapid method of characterising selected clones using droplet interface bilayers was developed that may be useful for high throughput applications.

Finally, Chapter 4 details efforts to express and engineer the Voltage-Dependent Anion Channel (VDAC) for stochastic sensing. This monomeric β-barrel membrane protein possesses a large lumen of diameter 25Å. VDAC could therefore have potential for the stochastic sensing of substrates too large for the lumen of αHL or OmpG.

Authors: Harrington, L

• DOI: 10.5287/ora-dpnkmqgwa
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Subjects: Kinases; Nanopores