Patrick Shire

CASE – Mechanisms of endosomal permeability and release: a route for enhancing cytoplasmic delivery of endocytosed therapeutics

My project

Introduction and methodology

Cells internalize macromolecules such as antibodies byreceptor-mediated endocytosis which results in the macromolecules being sequestered withinmembrane-bound endosomes and lysosomes. Normally, endosomal leakage is minimal, andcytoplasmic delivery is profoundly inefficient (Gilleron et al., 2013). However efficient release hasbeen observed following binding of adenovirus (Ad2) particles to cells (Meier et al., 2002). Thisstimulates macropinocytosis whereby cells engulf large volumes of extracellular fluid by extendingprotrusions from the cell surface which fuse together to form macropinosomes which are leaky inAd2-treated cells.We have developed an assay for Ad2-mediated release of endosomal contents in HeLa cells. Cellsurface binding of Ad2 is sufficient to stimulate macropinocytosis of GFP with a nuclear importsequence (NLS-GFP) and the resulting macropinosomes can burst, delivering NLS-GFP into thecytoplasm. NLS-GFP can then translocate into the nucleus and be detected byimmunofluorescence.

Experimental plan

The first aim is to carry out a genome wide siRNA screen of the kinome,within the Sheffield RNAi Screening Facility, to identify kinases involved in Ad2-mediatedendosomal release. Kinome screens are likely to result in many potential hits, some of which maybe indirect. This assay will also select for kinases required for macropinocytosis and for nuclearimport. There are surprisingly few kinases implicated in regulation of the latter, however we doexpect to obtain hits that interfere with macropinocytosis (e.g. PAK1, PI3kinase). A screen usingphorbol ester-induced macropinocytosis will be used to eliminate those kinases that function in theinitial budding off of macropinosomes.We will then choose 4 to 6 candidate kinases and:

  1. Validate these hits using alternative siRNAs to eliminate off-target effects.
  2. Generate siRNA resistant tagged versions of each kinase in an appropriate FLP in/CRISPrcell line to confirm rescue of the knockdown phenotype. The advantage of these lines isthat the kinases will be expressed at endogenous levels.
  3. Where appropriate, further validation will be carried out to confirm that the phenotype canbe recapitulated using known kinase inhibitors.We will use the tag on the siRNA resistant construct to perform subcellular localisation studies +/-Ad2. This will test whether ‘hits’ are recruited to macropinosomes or the plasma membrane.Recruitment to the plasma membrane would suggest that intermediate components are required tomediate endosomal permeability. In this case and if we have identified well-characterised signalling cascades, we will test potential downstream signaling molecules using both siRNA andpharmacological inhibitors where available. It is also likely that the siRNA screen will yield other‘hits’ for downstream targets which will be characterised as above. If we identify several signalling molecules, we will test whether they act in the same pathway and characterise the hierarchy of thepathway using knockdown/rescue analysis.

Industrial placement

The student will then visit MedImmune to identify those hits which are likelyto lead to a therapeutic/commercial impact. MedImmune has reagents which affect key intracellularsignalling pathways relevant to cancer. The student will address whether the identified candidatescan enhance the delivery and subsequent function of these intracellular antibodies in tumour cells.

Mechanistic studies

We will determine the substrates of the strongest candidates by quantitativephosphoproteomics in SILAC labelled cells at the biOMICsfacility, focusing especially on anymacropinosomal membrane proteins. For relevant substrates, we will determine whether thesephospho sites are affected by Ad2 binding and investigate the consequences of disrupting thesemodifications by generating cell-lines inducibly expressing phosphorylation- or phospho-mimeticmutants.


We have chosen the kinome because it is likely to yield druggable targets. In the unlikelyevent that we identify no validated candidate hits, we will use other siRNA collections within theSRSF. e.g. ubiquitome, lipid metabolizing enzymes, ion channels.