Research Undertaken

For my PhD I am studying neuropeptide secretion. Understanding when and how peptide hormones are secreted is of fundamental importance in understanding animal physiology. The aim of this project was to create a novel, FRET-based, peptide detection system to allow us to follow peptide secretion in real time.
The use of fluorescent proteins to follow biological processes is a very powerful and rapidly expanding area of research. Many of the most useful fluorescent biosensors utilise a conformational change that follows ligand binding to create FRET-based reporter systems. A limiting factor in producing these novel biosensors is the difficulty in finding fluorophore insertion positions that optimise the signal without disrupting protein function. To address this issue, Professor Thom Hughes has developed a technique for rapidly generating libraries of fluorescent fusion proteins. Using this method many insertion positions can be examined, thus allowing optimisation.

My biosensor was designed to utilise the large conformational change which the protein DppA undergoes on binding di-alanine. The examination of crystal structures of this protein both with and without ligands bound show that regions of DppA will undergo large confirmation changes upon ligand binding (Figure 1 Upper panel).  If fluorescent proteins can be inserted into these regions without disrupting function they should undergo large FRET changes.  However, the best places to insert fluorescent protein sequences without destroying protein function are hard to predict (Giraldez et al., 2005).  Indeed when a similar approach was taken to creating glutamate biosensors the signal-to-noise ratio was unexpectedly low, apparently because the fluorescent proteins did not move as anticipated (Okumoto et al., 2005).  In order to avoid this problem it therefore seemed simplest to use a trial and error approach.


Figure 1.

Principle of FRET peptide sensor. Upper panel, crystal structures of DppA showing the open (left hand) and bound/closed (right hand) structures. White arrows indicate loops that undergo large movements upon peptide binding while the yellow arrow indicates the binding pocket. Lower panel, scheme of intended fluorescent fusion protein. To detect these motions we intend to add two fluorescent proteins (ECFP and Citrine). The first will be attached to the N-terminus of DppA. The optimum placement of a second fluorophore will determined by creating a library of constructs with Citrine randomly inserted into different positions.

Professor Hughes’s technique for creating fluorescent fusion protein libraries is based on Tn5 transposon technology (Sheridan et al., 2002).  DNA sequences flanked by the recognition sequence for the transposase enzyme: mosaic ends (MEs), can be inserted via the transposase into other DNA regions.  Insertion occurs at most positions in a near random manner. Using this approach, and by inserting a fluorescent protein sequence between MEs, libraries of “standard proteins” labelled at a variety of positions with fluorescent proteins, can be created.  Once created, these libraries can be screened in order to find the best fluorescent clones for particular purposes such as an optimised FRET pair.

 To create my biosensor I first inserted the cyan fluorescent protein (ECFP) at the N-terminus of DppA and then used the transposition technique to randomly insert a FRET partner, the yellow fluorescent protein citrine, within the DppA sequence (Figure 1 Lower panel). The resulting fluorescent protein library may now be screened to find which clone gives the greatest signal change on binding di-alanine and is therefore the most sensitive detector.

Although di-alanine is not biologically relevant as a neuropeptide it is possible to use it as a reporter of neuropeptide secretion. To do this we will use a feature of the way neuropeptides are produced. Neuropeptides are genetically encoded as a long pre-prohormone which codes for several peptides; either different neuropeptides or multiple copies of the same neuropeptide. During protein production these peptides are expressed as one long chain which is packaged into a single large dense core granule. This long chain is later cleaved into individual peptides inside the granule. Thus peptides which are encoded for on a single pre-prohormone will be packaged and released from the same granule. Therefore by inserting the di-alanine sequence into the pre-prohormone sequence of a particular neuropeptide we can use our di-alanine sensor to follow secretion of that neuropeptide. A similar rationale has been used to follow NPY release via tagging with FMRFamide (Whim and Moss, 2001) and is one of the techniques I have used during my PhD.

Currently a library of 36 unique in-frame insertion positions of citrine has been created. Further screening will be done to determine whether a sufficiently sensitive detector has been made.


  1. Giraldez T, Hughes TE, Sigworth FJ. Generation of functional fluorescent BK channels by random insertion of GFP variants. J Gen Physiol. 2005 Nov;126(5):429-38.
  2. Okumoto S, Looger LL, Micheva KD, Reimer RJ, Smith SJ, Frommer WB. Detection of glutamate release from neurons by genetically encoded surface-displayed FRET nanosensors.  Proc Natl Acad Sci U S A. 2005 Jun 14;102(24):8740-5.
  3. Sheridan DL, Berlot CH, Robert A, Inglis FM, Jakobsdottir KB, Howe JR, Hughes TE. A new way to rapidly create functional, fluorescent fusion proteins: random insertion of GFP with an in vitro transposition reaction. BMC Neurosci. 2002 Jun 19;3:7.
  4. Whim M D and Moss G W J. A novel technique that measures peptide secretion on a millisecond timescale reveals rapid changes in release. Neuron. 2001 Apr;30(1):37-50

Page last modified on 20 jan 10 19:46