Rebuttal from James Howe

Abstract

MacLean (2013) suggests an alternative approach to explain the effects of transmembrane AMPA receptor regulatory proteins (TARPs), namely that TARPs may promote closed-cleft ligand binding domain (LBD) conformations that are distinct from those frequented without TARPs. Evidence for multiple closed LBD conformations exists (Birdsey-Benson et al. 2010; Landes et al. 2011), and if they are correlated with structural differences elsewhere in the receptor this might lead to changes in multiple rate constants. Mechanistically, such a scenario could be represented with tiered reaction mechanisms similar to those employed to describe features of NMDA receptor gating (Zhang et al. 2008b; Vance et al. 2012). Both MacLean and I agree that kinetic models provide a useful focus for thinking about mechanisms. Although structural data are required to lend physical reality to the discrete states in such models, relating functional kinetic data to explicit mechanisms gives insight into the temporal sequence of conformational changes the receptor undergoes and the energy landscape through which it transits. Such an approach also avoids the danger of drawing mechanistic conclusions from measurements of operationally defined changes in population responses. For example, everyone agrees that TARPs slow the rate at which ensemble currents decay in response to a rapid application of near-saturating glutamate, which by definition represents a slowing of desensitization. However, the available evidence argues strongly against a significant effect of TARPs on δ, the rate at which receptors enter desensitized states (Fig. 1 in Howe (2013). If this rate primarily reflects the rate at which the dimer interface separates (Sun et al. 2002), then by extension TARPs do not primarily act by stabilizing this interface. The T686 mutants MacLean refers to, which disrupt a molecular latch that engages when the LBD closes, are an interesting case in point in this regard. The mutations both slow the rate at which ensemble currents decay during a sustained application of glutamate and increase the amplitude of the steady-state current as a fraction of the initial peak current (Robert et al. 2005). However, single-channel data show that both effects result directly from destabilization of the closed-cleft LBD (increases in CO; Fig. 1 in Howe (2013)), which prolongs latencies-to-first-opening and desynchronizes activation (Zhang et al. 2008a). Desynchronizing activation both slows the decay of ensemble currents and substantially reduces peak Popen (Colquhoun & Hawkes, 1995), and the larger relative sustained currents in the T686 mutants are not due to an increase in open probability late in the application but rather a reduction in the amplitude of peak currents.