The liver-type (GLUT2) and brain-type (GLUT3) human facilitative glucose transporters exhibit distinct kinetics (Km values for deoxyglucose transport of approximately 11 mM and approximately 1.5 mM, respectively) and patterns of substrate transport (GLUT2 is capable of D-fructose transport, while GLUT3 is not). Using a range of chimeric glucose transporters comprised of regions of GLUT2 and GLUT3 studied by expression in Xenopus oocytes after microinjection of cRNA, we have proposed that the seventh putative transmembrane helix is intimately involved in the selection of transported substrate and that this region plays an important role in determining the Km for 2-deoxyglucose [Arbuckle, M. I., Kane, S., Porter, L. M., Seatter, M. J., and Gould, G. W. (1996) Biochemistry 35, 16519-16527]. Inspection of the predicted amino acid sequence of this region reveals that GLUTs 1, 3, and 4 (high-affinity glucose transporters) contain a conserved QLS motif in this helix (residues 277-279 in human GLUT3). In the glucose/fructose transporter (GLUT2) this motif is replaced by HVA. To study the role of the QLS motif in substrate selection, we have engineered substitutions in this region between GLUT2 and GLUT3. GLUT3 (QLS > HVA) exhibits a Km for deoxyglucose transport identical to that of native GLUT3 but increased sensitivity for inhibition of deoxyglucose transport by D-fructose. However, unlike native GLUT3, this species is capable of transporting D-fructose. Compared to wild-type GLUT2, GLUT2 (HVA > QLS) exhibits a lower Km for deoxyglucose transport (approximately 3 mM vs approximately 11 mM), the ability to transport D-fructose is reduced, and D-fructose is a less efficient inhibitor of deoxyglucose transport. Analysis of the ability of a range of glucose epimers and analogues to inhibit transport by these species suggests that the QLS motif interacts with the incoming D-glucose at the C-1 position; this may be a key interaction in the high-affinity recognition of the transported substrate. We further argue that this interaction acts as a molecular filter that is involved in the selection of the transported substrate.

BackgroundThere are limited effective prophylactic treatments for SARS-CoV-2 infection, and limited early treatment options. Viral cell entry requires spike protein binding to the ACE2 receptor and spike cleavage by TMPRSS2, a cell surface serine protease. Targeting of TMPRSS2 by either androgen blockade or direct inhibition is already in clinical trials in early SARS-CoV-2 infection.MethodsThe likely initial cells of SARS-CoV-2 entry are the ciliated cells of the upper airway. We therefore used differentiated primary human airway epithelial cells maintained at the air-liquid interface (ALI) to test the impact of targeting TMPRSS2 on the prevention of SARS-CoV-2 infection.ResultsWe first modelled the systemic delivery of compounds. Enzalutamide, an oral androgen receptor antagonist, had no impact on SARS-Cov-2 infection. By contrast, camostat mesylate, an orally available serine protease inhibitor, blocked SARS-CoV-2 entry. However, camostat is rapidly metabolised in the circulation in vivo, and systemic bioavailability after oral dosing is low. We therefore modelled local airway administration by applying camostat to the apical surface of the differentiated ALI cultures. We demonstrated that a brief exposure to topical camostat is effective at restricting SARS-CoV-2 viral infection.ConclusionThese experiments demonstrate a potential therapeutic role for topical camostat for pre- or post-exposure prophylaxis of SARS-CoV-2, which can now be evaluated in a clinical trial.