The Molecular Composition of Inner Hair Cell Ribbon Synapses

Ribbon synapses, as well as synapses at neuronal nerve terminals, are designed for Ca²⁺-induced neurotransmitter release. Developmentally, IHCs and neurons are formed from different embryonic compartments, though. IHCs derive from placodes formed within the epithelial part of the ectoderm, whereas neurons of the central nervous system originate from the neural tube (Graham and Shimeld, 2013). Therefore, it is plausible that the molecular composition of the AZ and of the release machinery partially differ between IHC ribbon synapses and neuronal synapses.

SCAFFOLDING PROTEINS

The cytomatrix protein RIBEYE represents the main organizational compound of synaptic ribbons not only at IHC ribbon synapses but also in the retina. RIBEYE consists of an A domain involved in

Introduction | What Makes Ribbon Synapses so Special?

18

the assembly of the ribbon and a B domain with enzymatic activity (Khimich et al., 2005; Schmitz et al., 2000; Schwarz et al., 2011). The B domain is structurally almost identical to the transcription repressor C-terminal binding protein 2 (CtBP2) and may facilitate the tethering of SVs to the ribbon (Schmitz et al., 2000). Knockout of the RIBEYE A domain was found to severely impair synaptic function in the retina, whereas IHC synaptic transmission was shown to be rather mildly affected (Becker et al., 2018; Jean et al., 2018; Maxeiner et al., 2016). Impressive developmental modifications - including the re-shaping of the AZ with several small presynaptic densities that were shown to tether SVs in RIBEYE-knockout (KO) IHCs - may here compensate in part for the absence of the synaptic ribbons (Jean et al., 2018).

The scaffolding protein bassoon has originally been described as one major component of the CAZ at central synapses (Südhof, 2012; tom Dieck et al., 1998). Immunogold electron microscopy revealed that bassoon is also present at IHC and retinal AZs and here forms a component of the presynaptic density anchoring the synaptic ribbon (Dick et al., 2003; tom Dieck et al., 2005; Wong et al., 2014).

Consequently, ribbons are not attached to the AZ in bassoon-KO mice but free-floating in the cytosol (Dick et al., 2003; Frank et al., 2010; Khimich et al., 2005). In addition, bassoon was shown to be required for the stabilization of the RRP and for the clustering of Ca²⁺-channels (Frank et al., 2010;

Jean et al., 2018; Jing et al., 2013; Khimich et al., 2005; Neef et al., 2018). At central synapses, bassoon and piccolo mostly act together in the assembly of the AZ and in scaffolding, as well as in SV replenishment (Butola et al., 2017; Fenster et al., 2000; Mendoza Schulz et al., 2014; Südhof, 2012;

tom Dieck et al., 1998), but seem to have distinct functions at ribbon synapses. At retinal and IHC ribbon synapses, only a truncated form of piccolo called piccolino could be substantiated, which is missing the C-terminal binding sites for bassoon and RIM (Regus-Leidig et al., 2013). Immunogold labeling revealed the localization of piccolino exclusively around the ribbon (reminiscent of RIBEYE staining); and knockdown of piccolo/ piccolino in the retina was shown to result in impaired ribbon formation (Limbach et al., 2011; Regus-Leidig et al., 2014). It remains to be investigated if piccolino is involved in ribbon assembly, if it has a function in organizing CAZ or fusion proteins, or if piccolino contributes to the formation of tethers. Based on the localization, piccolino could be involved in tethering SVs to the ribbon. In contrast, RIMs were shown to play a role in the tethering of SVs to the AZ (Jung et al., 2015a). Moreover, RIMs can interact with CaV1.3 Ca²⁺-channels and, accompanied by RIM-binding protein (RIM-BP), regulate the clustering of these channels (Jung et al., 2015a; Krinner et al., 2017, 201; Picher et al., 2017).

Introduction | What Makes Ribbon Synapses so Special?

19 CALCIUM CHANNELS

In IHCs, CaV1.3 is the almost exclusively present subtype of Ca²⁺-channels (Brandt et al., 2003; Platzer et al., 2000). Approx. 50-80% of these channels in a cell form dense clusters at the AZ, with each cluster consisting of 80-120 individual channels on average (Brandt et al., 2005; Neef et al., 2018;

Roberts et al., 1990; Wong et al., 2014; Zampini et al., 2013). Still, numbers of channels and cluster lengths can vary greatly depending on the localization of the AZ within the IHC (Frank et al., 2010;

Neef et al., 2018; Ohn et al., 2016; Wong et al., 2014). In general, Ca²⁺-channels are organized in a comparably higher density within the clusters at IHC ribbon synapses than at central synapses, and provide ultrafast activation and very slow inactivation kinetics (Neef et al., 2018; Zampini et al., 2013). Mostly, these clusters form a stripe- or double stripe-like pattern that closely conforms the distribution of bassoon (Frank et al., 2010; Neef et al., 2018; Wong et al., 2014). Ca²⁺-channels are tightly coupled to SVs, and it seems as if a single channel controls the release of one SV (Brandt et al., 2005; Wong et al., 2014).

THE MULTI-C2DOMAIN PROTEIN OTOFERLIN

In IHCs after the onset of hearing, Ca²⁺-sensing for exocytosis apparently does not require the neuronal Ca²⁺ sensor synaptotagmin-1 (Beurg et al., 2010; Reisinger et al., 2011; Safieddine and Wenthold, 1999). However, the IHC-specific protein otoferlin harbors 6-7 C2 domains, which are structurally related to those of synaptotagmin-1 (Helfmann et al., 2011; et al., 2012;

Reisinger et al., 2011; Roux et al., 2006). Apart from the C2 domains, most of which can bind Ca²⁺, otoferlin possesses a C-terminal trans-membrane domain, as well as a Fer domain and a coiled-coil domain (Johnson and Chapman, 2010; et al., 2012; Roux et al., 2006). Even though otoferlin has been suggested to act as a Ca²⁺-sensor for exocytosis in IHCs (Johnson and Chapman, 2010;

Michalski et al., 2017; Roux et al., 2006), transgenic expression of synaptotagmin-1 in otoferlin-KO mice could not restore exocytosis (Reisinger et al., 2011). These and further studies on otoferlin-KO mice led to the assumption that otoferlin is moreover crucial for subsequent steps of exocytosis like SV tethering and priming ( et al., 2010; Roux et al., 2006; Strenzke et al., 2016; Vogl et al., 2015). Even though otoferlin seems not to be a main component of short tethers, it may aid their formation potentially as a priming factor (Vogl et al., 2015). In this regard, multiple binding sites for Ca²⁺-channels, phospholipids and adaptor proteins may allow otoferlin to link fusion-competent SVs and exocytic proteins in close proximity to the release sites (Hams et al., 2017; Padmanarayana

Introduction | What Makes Ribbon Synapses so Special?

20

et al., 2014; Ramakrishnan et al., 2009). Alternatively or in addition, functions of otoferlin in SV replenishment and/ or release site clearance have been suggested (Chakrabarti et al., 2018; Jung et al., 2015b; et al., 2010). Direct interactions of otoferlin and the adaptor protein AP-2 have been interpreted as indicators for a role of otoferlin in SV reformation, in endocytosis and/ or in the coupling of exocytosis and endocytosis (Duncker et al., 2013; Jung et al., 2015b; Strenzke et al., 2016).

Introduction | IHC Synapses Show Adaptive Neurotransmission

21