Abstract

It
is widely believed that most xenobiotics cross biomembranes by diffusing
through the phospholipid bilayer, and that the use of protein transporters is
an occasional adjunct. According to an alternative view, phospholipid bilayer
transport is negligible, and several different transporters may be involved in
the uptake of an individual molecular type. We recognise here that the
availability of gene knockout collections allows one to assess the contributions
of all potential transporters, and flow cytometry based on fluorescence
provides a convenient high-throughput assay for xenobiotic uptake in individual
cells [1].

We
used high-throughput flow cytometry and a PAA S-LAB^TM automated
plate handler to assess the ability of individual gene knockout strains of E
coli to take up two membrane-permeable, cationic fluorescent dyes, namely the carbocyanine
diS-C3(5) and the DNA dye SYBR Green. Individual strains showed a large range
of distributions of uptake. The range of modal steady-state uptakes for the carbocyanine
between the different strains was 36-fold. Knockouts of the ATP synthase α- and
β-subunits greatly inhibited uptake, implying that most uptake was ATP-driven
rather than being driven by a membrane potential. Dozens of transporters
changed the steady-state uptake of the dye by more than 50% with respect to
that of the wild type, in either direction (increased or decreased); knockouts
of known influx and efflux transporters
behaved as expected, giving credence to the general strategy. Many of the
knockouts with the most reduced uptake were transporter genes of unknown
function (‘y-genes’). Similarly, several
overexpression variants in the ‘ASKA’ collection
had the anticipated, opposite effects. Similar results were obtained with SYBR
Green (the range being approximately 69-fold). Although it too contains a
benzothiazole motif there was negligible correlation between its uptake and
that of the carbocyanine when compared across the various strains (although the
membrane potential is presumably the same in each case). We also show the
uptake of several fluorescent dyes in mammalian cells (e.g K562 and KU812 cell
lines) as well as competition assays using these and pharmaceutical drugs.

Overall,
we conclude that the uptake of these dyes may be catalysed by a great many
transporters of putatively broad and presently unknown specificity, and that
the very large range between the ‘lowest’ and
the ‘highest’ levels of uptake, even in
knockouts of just single genes, implies strongly that phospholipid bilayer transport
is indeed negligible. This work also casts serious doubt upon the use of such
dyes as quantitative stains for representing either bioenergetic parameters or
the amount of cellular DNA in unfixed cells (in vivo). By contrast, it opens up
their potential use as transporter assay substrates in high-throughput
screening.