On the path from proteomics to function: GABA_AR trafficking takes a turn

Abstract

There are significant challenges in identifying receptor-specific functional interactors in vivo. In this issue of Neuron, Ge et al (2018) identify a novel GABA_A receptor (GABA_AR)-interacting protein, Clptm1 that regulates forward trafficking of GABA_ARs and inhibitory transmission.

Main text

Interactions between neurotransmitter receptors and regulatory proteins control synaptic transmission and plasticity through various pathways. In order to fulfill their function, neurotransmitter receptors first exit the endoplasmic reticulum (ER) and travel to the cell surface. During this process, interactions between the receptor and various binding partners promote and/or regulate receptor assembly and trafficking. Through these interactions, a mature receptor complex is formed, enabling its proper localization and function at the plasma membrane and synapses (Figure 1). Indeed, protein complexes between a neurotransmitter receptor and regulatory, interacting-proteins (‘interactors’) emerge as a common theme in receptor biology. For example, excitatory AMPA and kainate glutamate receptors create a stable complex with their auxiliary proteins TARPs and Neto, respectively, and these interactions are required for proper synaptic transmission (Jackson and Nicoll, 2011; Yan and Tomita, 2012). Recently, GABA_ARs were found to form a stable tripartite complex with GARLH and Neuroligin-2 (Yamasaki et al., 2017), which is required for synaptic localization of GABA_ARs, but not forward trafficking (Davenport et al., 2017; Yamasaki et al., 2017).

FIGURE 1. Stable and transient interactors accompany and control neurotransmitter receptors.

During the process of forward trafficking, receptors form complexes with both transient and stable interactors that either limit or facilitate their activity. Transient interactions can support receptor maturation and transport from ER to plasma membrane, whereas stable interactions (such as with auxiliary subunits) can also modulate membrane localization and functional properties. By their nature, functional transient interactors are more challenging to be identified by proteomic methods.

Not only stable, but also transient interactions help shape the receptor landscape on the neuronal membrane (Figure 1). These transient interactions can take place along the various stages of forward receptor trafficking, including complex assembly, ER exit, trafficking, and exocytosis. Uncovering the specific interactors controlling receptor function is critical to our mechanistic understanding of neuronal signaling, yet the biochemical isolation and identification of transient interactors in vivo has proven technically demanding. A major challenge is the fact that transient complexes are (by definition) less abundant, and are therefore harder to detect. Consequently, proteomic approaches to reveal transient interactions may lead to large sets of candidates, with dozens of potential interactors. In some cases, rather than relying solely on classical proteomic approaches, ‘fishing’ for specific transient interactors can be achieved by adopting complementary methods such as high-throughput functional assays. Indeed, taking such a combined approach has led in the past to a functional confirmation of ten AMPA-receptor interactors and identified Porcupine as a transient interactor that modulates AMPAR forward trafficking (Erlenhardt et al., 2016; Schwenk et al., 2012).

A further complicating factor in studying neurotransmitter receptors is heterogeneity in receptor composition. In the case of GABA_AR, the receptor assembles as a hetero-pentamer, containing at least one of six α subunits, one of three β subunits and one of ten non-α/β subunits. Various combinations of these subunits give rise to a high degree of GABA_AR molecular diversity in the brain. Therefore, identifying specific molecules that control GABA_AR trafficking may become a daunting task if trying to account for the many possible subunit combinations.

In this issue of Neuron, overcoming many of the challenges inherent to proteomic-based identification of GABA_ARs-specific transient interactors, Ge et al. (Ge et al., 2018) applied an optimized proteomic approach to identify Clptm1 as a GABA_AR-interacting protein that controls receptor forward trafficking in neurons. To circumvent potential caveats in specificity of antibodies in proteomics, the authors applied a transgenic approach. Using knockin mice that express His-FLAG-YFP-GABA_AR γ2, the authors purified the receptor complex through a tandem purification technique with nickel-charged affinity resin and an anti-GFP antibody. This has led to the identification of 39 proteins as well as previously identified associated proteins of GABA_AR subunits and neuroligin-2. Next, Ge et al focused on six proteins for further analysis of their interaction in HEK293 cells, and found that three transmembrane proteins (Clptm1, Itm2C and Glgl1) could interact with GABA_ARs, but not with glycine receptors.

The authors next assessed the functional role of these three proteins and found that only Clptm1 altered GABA_ARs-mediated currents. This was demonstrated by a substantial reduction in GABA-evoked currents, with no apparent effect on glycine receptor activity. Over-expression of a Myc-tagged Clptm1 resulted in substantial accumulation of Clptm1 and GABA_AR in the ER and reduced surface expression of GABA_ARs, suggesting a role of Clptm1 in intracellular retention. Perhaps most remarkably, knocking down Clptm1 increased GABA-evoked currents in primary cultured neurons, and overexpression and knockdown of clptm1 in vivo reduced and increased miniature inhibitory postsynaptic current (mIPSC) amplitudes, respectively, without altering excitatory miniature excitatory postsynaptic currents (mEPSCs). This set of experiments establishes Cltpm1 as an intracellular molecule for controlling GABA_AR forward trafficking, acting likely via ER retention.

Intriguingly, the bi-directional effects of Clptm1 knockdown and overexpression on mIPSCs resemble the process of homeostatic synaptic scaling. In primary neuronal culture preparations, homeostatic synaptic scaling can be defined as a compensatory response in which silencing GABAergic transmission leads to an increase in synaptic GABA_AR activity, while silencing overall neuronal firing reduces GABA_AR activity. Ge et al. demonstrated that bidirectional manipulation of GABA_ARs by Clptm1 knockdown and overexpression can either accentuate or occlude homeostatic scaling. This raises the possibility that rather than playing a static role, Clptm1 may participate in the dynamic adjustment of GABA_ARs number at the synapse.

Overall, the study of Ge et al adds an important step on our way to understanding the cellular machinery that controls inhibitory transmission. It identifies a negative regulator, Clptm1, which acts to trap GABA_ARs within intracellular compartments and thus limits their expression on the neuronal membrane. Receptor-specific interactors that restrict surface expression highlight the importance of precise and tight control of GABA_AR number on the plasma membrane.

Several important questions remain open for future studies. Interestingly, although Clptm1 was originally identified using a tagged GABA_AR-γ2, Ge et al detected its interaction (in transfected heterologous cells) with all tested GABA_AR subunits (namely γ2, α1, α2, β2, β3). This is surprising, considering the relatively low sequence homology between these different subunits. Future studies can elucidate the precise molecular interaction of Clptm1 with the various subunits of GABA_ARs. Traditional molecular approaches to characterize interaction domains may be challenging in this case. Typically, deletion mutants or chimeric proteins between functional and non-functional homologous proteins are used for mapping interaction domains. However, applying this approach to study the interaction between two transmembrane proteins is notoriously difficult because of the relatively low tolerance of transmembrane proteins to mutations that affect their structure. With recent advances in cryoelectron microscopy, however, solving the structure of the receptor complex at the atomic level should provide important insights on the interaction between the receptor and its binding partners.

An additional outstanding question concerns the physiological roles that Clptm1 serves by accumulating GABA_ARs at the ER. GABA_ARs at the ER may function as a reservoir for surface expression. In such a scenario, Clptm1 or its associated proteins could serve as a dynamic sensor of neuronal activity by releasing GABA_ARs from the ER upon demand. Indeed, knockdown or overexpression of Clptm1 altered the number of GABA_ARs at the cell surface, mimicking the effects of homeostatic synaptic scaling in vitro. Future endeavors should reveal whether Clptm1-induced retention of GABA_ARs plays a physiological role in the dynamic control of inhibitory transmission in vivo. An alternative (though not mutually exclusive) physiological role of Clptm1 might be to delay GABA_AR exit from the ER in order to allow proper assembly with other stable interactors. Mature GABA_ARs form a tri-partite complex with GARLH and Neuroligin2 (Yamasaki et al. 2017), but the underlying mechanisms remain unknown. Interestingly, Clptm1 was identified using a tagged GABA_A-γ2 subunit, while GARLH requires the same subunit in order to interact with GABA_ARs (Martenson et al., 2017; Yamasaki et al., 2017). Thus, Clptm1 may act as a checkpoint for GABA_AR complexes upon exiting the ER. Analysis of GABA_AR complex assembly and localization in Clptm1 knockout mice can be especially beneficial in revealing such potential roles. Furthermore, complementary proteomic analysis of Clptm1 (as done for GABA_AR in the current work by Ge et al) can provide important information on the molecular makeup of GABA_AR in complex with Clptm1, and help discover new binding partners involved in Clptm1-induced retention.

Finally, the proteomic analysis by Ge et al identified 39 proteins and extensively studied 6 of them. It is common for proteomic studies to generate relatively large lists of potentially interesting targets. Indeed, different proteomic approaches have been successful in identifying additional sets of GABA_AR-associated proteins that may serve important roles in the cellular and physiological regulation of these receptors (Heller et al., 2012; Nakamura et al., 2016). Considering the many functions that GABA_AR-associated proteins can serve in receptor maturation, trafficking, membrane localization and biophysical properties, further analysis of individual interactors and their in vivo functions should provide important insights. Currently, sifting through large amounts of candidates to identify the most promising ones remains labor intensive. Establishing high-throughput pipelines to systematically evaluate the functional contribution of all proteomic hits is therefore of high priority to the field and should bring us closer to a complete mechanistic understanding of a receptor’s life cycle in the living neuron, and the consequences for brain function.