Essential role for plasma membrane glutamate transporters in stimulus intensity coding in auditory neurons

Reviewer #1 (Public review):

In this article, the authors investigate how glutamate transporter function regulates excitability and synaptic coding in T-stellate cells in the mouse ventral cochlear nucleus. They test this in acute brain slices using whole-cell electrophysiology and artificially raise the relative local concentration of glutamate via pharmacological inhibition of transporter proteins. The main finding is that when sub-saturating doses of DL-TBOA are applied, cells become much more sensitive to synaptic input, diminishing the normally high fidelity of EPSP-spike coupling in these neurons. Notably, high-frequency stimulation in the presence of DL-TBOA reveals a large and slowly decaying AMPA receptor component that underlies persistent/rebound firing in earlier recordings. These effects are not seen in other ventral cochlear neurons, suggesting that rapid glutamate clearance in T-stellate cells, particularly, is important for auditory intensity coding. Overall, these experiments are well-performed, and the findings are robust, though there are some aspects that could be expanded to make the work more impactful. These include a better understanding of the relative contribution of neuronal vs glial transporters and an ability to separate the relative contributions of tonic glutamate concentrations in the cleft vs changes in membrane potential in action potential output. Additionally, there were some minor issues of clarity in both the figure presentation and the main text language that should be addressed.

Major Points:

(1) Given the dramatic effect of saturating DL-TBOA on tonic leak/RMP and that the sub-maximal concentration used in most of the experiments still varied between 25-50 uM, Figure 1 would be strengthened substantially by a dose-response curve. Ideally, 5 or 6 concentrations, plotting the effect on tonic current or RMP increase.

(2) Examining the contribution of glial (EAAT1/2) vs. neuronal (EAAT3) transporters (Fig 8) is intriguing but comes across as incomplete here, especially given the small number of recordings. Using a different non-selective EAAT inhibitor (TFB-TBOA) to chase the EAAT1/2 blocker combo seems like an odd choice, given that you have already characterized the effects of DL-TBOA well. One could also try a lower concentration (~50-100 nM) of TFB-TBOA since it is somewhat selective itself for glial EAAT1/2. Given the data presented, neuronal transporters (presumably EAAT3) appear to dominate the rapid clearance of glutamate at this synapse, but this point isn't emphasized or explored sufficiently.

(3) Separating the effects of depolarization vs. glutamate clearance was never explored. What effect does depolarizing the cell ~10 mV in control conditions (i.e., without TBOA) have on AP number/fidelity during synaptic stimulation experiments? The authors state that submaximal DL-TBOA generally causes no more than a 5 mV change in RMP, but tonic depolarization could also influence spike fidelity. This experiment could demonstrate that the increase in excitability during/after stimulation is not due to increased engagement of voltage-gated channels.

Reviewer #2 (Public review):

Summary:

This manuscript addresses an important and mechanistically interesting question: whether plasma membrane glutamate transporters contribute only to slow clearance of ambient glutamate or whether they can rapidly shape synaptic signaling during high-frequency auditory activity. This manuscript provides important evidence that EAAT-mediated glutamate uptake is not merely a slow background clearance mechanism but is essential for maintaining reliable synaptic transmission and linear stimulus-intensity coding in ventral cochlear nucleus T-stellate cells during sustained auditory nerve activity.

Strengths:

The finding that EAATs may be required for rapid, local control of glutamate during high-frequency auditory nerve activity is interesting and could have broad relevance to auditory processing. The electrophysiological evidence is generally strong, particularly the use of patch-clamp recordings, stimulus trains, partial versus complete EAAT blockade, and comparison with bushy cell/endbulb synapses. The comparison between T-stellate cells and bushy cells/endbulb synapses strengthens the manuscript. The authors demonstrate that EAAT blockade disrupts coding in T-stellate cells but has little effect on bushy cell spike transmission, supporting a cell-type- and synapse-specific role of glutamate uptake.

Weaknesses:

However, some mechanistic conclusions, especially the specific contribution of neuronal versus glial EAATs and the absence of glutamate crosstalk between auditory nerve inputs, rely mainly on pharmacological and indirect electrophysiological inference and would be strengthened by additional anatomical, genetic, or direct glutamate-sensing evidence.

(1) Clarification of DL-TBOA concentration.

The authors used bath application of 200 µM TBOA and 25-50 µM in the other experiments, stating that "sub-maximal concentrations (25-50 µM)". The authors should provide a clearer rationale for why different concentrations were used across experiments rather than a fixed concentration.

The reversibility of DL-TBOA effects should be demonstrated by washout experiments. In addition, potential off-target effects of DL-TBOA on postsynaptic receptors, intrinsic membrane excitability, or presynaptic release (e.g., PPR measurement) should be carefully considered. It would also be useful to test the effects of the submaximal DL-TBOA concentrations (25-50 µM) on membrane potential and inward currents, shown in Figure 1, to determine whether these concentrations depolarize the membrane potential in current-clamp mode or induce inward currents under voltage-clamp conditions.

(2) Potential contribution of altered intrinsic excitability.

In Figures 3B and 3C, DL-TBOA appears to induce additional action potentials even immediately after the first stimulation, whereas Figures 6 and 7 suggest that the first EPSC is not substantially altered. This raises the possibility that the enhanced firing may partly result from a modest depolarization caused by background glutamate accumulation or from other changes in intrinsic membrane properties after drug treatment. To address this, the authors should provide a quantitative analysis of physiological parameters under submaximal DL-TBOA conditions, including spontaneous action potential frequency, resting membrane potential, input resistance, and spike threshold.

(3) Spillover/ crosstalk between AN-fiber-synpases.

The authors should provide more explanation of how altering the number of active auditory nerve fibers demonstrates the absence of glutamate spillover/crosstalk between bouton synapses. Strong stimulation likely recruits more AN fibers, but it may also change release probability, axonal synchrony, or stimulation spread. The authors should more clearly justify the interpretation that strong stimulation recruits additional independent AN fibers rather than altering release probability or activating fibers with different intrinsic properties.

(4) Interpretation of glial versus neuronal EAAT contributions.

The authors claim that both neuronal and glial transporters contribute to rapid uptake using pharmacological approaches. The pharmacological data demonstrate that glial EAATs play a major role in glutamate clearance at T-stellate cell synapses. The strong increase in EPSC decay time and synaptic charge after UCPH-101/DHK application supports the conclusion that glial transporters contribute substantially to limiting glutamate accumulation during sustained auditory nerve activity. However, the conclusion that neuronal EAATs contribute directly should be stated with some caution. The evidence for neuronal EAAT involvement is indirect and depends on the pharmacological specificity and completeness of glial EAAT blockade. The conclusion would be strengthened by additional evidence, such as EAAT subtype expression/localization in T-stellate cells or auditory nerve terminals, transporter current recordings, immunohistochemistry, or genetic manipulation of neuronal EAATs. In addition, fitting the decay phase with a double-exponential model may help determine whether glial and neuronal EAATs contribute over distinct temporal windows.

Author Response:

We are grateful for the careful and extensive reviews, and are pleased that the reviewers found the work of broad interest to sensory processing. Please find our proposal for revision based on public reviews:

Reviewer 1

1) Request for dose-response curve for DL-TBOA and leak current or RMP. We can provide this, at least for the initial phase of the curve relevant to the concentrations used for synaptic experiments. Prolonged exposure to higher concentrations leads to very large cationic currents (through AMPAR) which appear to be damaging to membrane integrity.

2) We will increase the N for glial vs neuronal block with the blockers we already used; this seems more practical than doing new experiments with different concentrations of TFB-TBOA. 

3) We can include data to test the effect of blockers or small depolarizations on excitability.

Reviewer 2

1) We differentiated experiments with “25-50 uM” from 200 uM DL-TBOA because the higher concentration clearly led to massive AMPAR activation and depolarization block, as shown in Fig 1. We then chose lower concentrations to minimize background current while allowing glutamate build-up during exocytosis.  We felt we were clear on this point. 

As to reversibility and “off target effects” like synaptic changes, we will provide this information. See also response to Reviewer 1, comment 1.

2) See response to Reviewer 1, comment 3.

3) We are certain that increasing stimulus strength increases the number of stimulated fibers, and this is well accepted. The stimulus electrode is placed in the auditory nerve root, well away from recorded cell and synapses, minimizing current spread to synapses. We can compare PPR for weak and strong stimuli in our current dataset to confirm no effects on release probability. As to variations in the intrinsic properties of myelinated auditory nerve fibers and their sensitivity to stimulation, there is no information about this, and do not understand how it would be relevant, particularly in as much as we report a negative result: no difference in blocker effect with small or large numbers of fibers active. The 3 main types of myelinated auditory nerve fiber, Type 1a,b,c, are known to respond to different sound thresholds, but that is a synaptic issue in the inner ear, and apparently not related to the myelinated axon bundle.

4) We appreciate the reviewer's caution about a role for neuronal transporters and will revise accordingly.  We cited molecular evidence for expression of subtypes in the pre and postsynaptic neurons and in glial cells. Of course, given how ubiquitous such expression is across the brain, we suspect the kinds of experiments we provided offer more direct evidence for function.