Molecular composition and function of the spiral ganglion neuron peripheral synapse in mice
Reijntjes, Daniël Onne Jilt
DOI:
10.33612/diss.93524048
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Reijntjes, D. O. J. (2019). Molecular composition and function of the spiral ganglion neuron peripheral synapse in mice. University of Groningen. https://doi.org/10.33612/diss.93524048
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 6
General discussion
The spiral ganglion neuron (SGN) peripheral dendrite is the most vulnerable part of the auditory system (Liberman and Kujawa, 2017) and is thought to be lost follow-ing strong activation of the auditory system. The loss of the SGN peripheral dendrite is likely induced by over-exposure to the neurotransmitter glutamate that is released from the inner hair cells (IHCs) in response to acoustical stimulation (Ruel et al., 2007). Furthermore, this process is thought to affect mostly a specific subset of the SGN peripheral dendrites, those of the low spontaneously active SGNs (Furman et al., 2013). How exactly this glutamate exposure would lead to loss of the periph-eral dendrite is unclear. In addition, it is unclear what determines the differences in spontaneous activity between the SGNs. Both the spontaneous activity and the vulnerability to glutamate exposure are likely determined by specific proteins in the peripheral dendrite that regulate specific cellular processes. Unfortunately, many proteins that form the molecular architecture of SGNs and the molecular processes that occur in the spiral ganglion SGN peripheral dendrite are poorly understood both under normal and pathological conditions. Therefore, the goal of this thesis is to fur-ther increase our understanding of both the molecular architecture and the molecular processes in the SGN peripheral dendrites.
6.1. Molecular architecture and molecular processes
in the spiral ganglion neuron peripheral dendrite
In chapter 2, the known proteins and processes of the SGN peripheral dendrites are reviewed in order to map out known proteins that shape SGN activity and vul-nerability and indicate knowledge gaps. These proteins and processes are integrated into the afferent signaling complex that outlines the most important contributing pro-cesses underlying SGN activity and vulnerability: glutamate signaling, ion channel function and lateral efferent innervation. Besides knowledge gaps in the known pro-teins that contribute to these processes, there is still a major disconnect in how these processes interact with each other. A prime example is the function of the lateral efferent system. This system is thought to provide feedback from the midbrain back to the cochlea that can both inhibit and stimulate the SGNs (Reijntjes and Pyott, 2016). However, how these effects are brought about is unknown, although it is likely that the lateral efferent system affects the function of either glutamate signaling or ion channel related processes (Valdés-Baizabal et al., 2015). The described afferent signaling complex provides a framework for future investigations into specific parts of the afferent signaling complex and how they interact with each other.
6. General discussion 127
6.2. Novel protein expression and function
To start filling in the gaps in the afferent signaling complex, the presence of volt-age gated ion channels in the SGN peripheral dendrite is investigated in chapter 3. The RNA expression of these voltage gated ion channels is examined in the SGNs. The rationale for a primary focus on examining voltage gated ion channels is twofold. First, voltage gated ion channels are largely responsible for determining passive and active membrane properties such as the resting membrane potential. These proper-ties are crucial in setting the physiological responses of neurons such as their spon-taneous activity. Therefore, voltage gated ion channels likely help shape the differ-ent SGN subgroups. Second, in the CNS, neuronal damage due to glutamate over-exposure depends on increases of both intracellular Na+and Ca2+(Kostandy, 2012).
Although it is unclear whether these mechanisms are the same in the SGNs, there are indications that similar mechanisms for neuronal damage are present in the SGN (Chen et al., 2009, 2007). In this case, voltage gated ion channels would likely con-tribute to either the influx or efflux of Na+and Ca2+and therefore contribute to this
process (see chapter 2).
RNA expression in the peripheral dendrite indicates that many voltage gated ion channels that are known in the CNS are also expressed in the SGN peripheral den-drite. Many of the proteins for which these genes encode are known (Oak and Yi, 2014; Rusznák and Szucs, 2009), but importantly, some of these genes indicate the expression of novel proteins in the SGN peripheral dendrite (see chapters 2 and 3). Although the expression of RNA within the SGNs indicates that these proteins are likely expressed in the SGN, this information has little meaning without evidence that these transcripts are translated into proteins and that these proteins are ex-pressed in the peripheral dendrite. Therefore, the function of these proteins within the peripheral dendrite needs to be examined. A novel specific subgroup of voltage gated ion channels, sodium-activated potassium channels from the KN a1 family is
further examined for expression and function in the SGN peripheral dendrite. These voltage gated ion channels are chosen specifically because their activation and open-ing is co-dependent on Na+influx (Kaczmarek, 2013). This co-dependent activation
makes these ion channels ideal to regulate both SGN activity and possibly, to affect how Na+influx triggers death of the SGN peripheral dendrite.
RNA for KN a1 ion channels was found to be expressed in the SGN cell bodies,
but the KN a1ion channel proteins could not be specifically localized. However, loss
of these KN a1channels in a KO mouse model resulted in an altered phenotype both
in vitroand in vivo, which suggests that these ion channels are indeed expressed in
the SGN peripheral dendrite. Specifically, cultured SGN soma from KNa1 KO mice revealed a reduction in activation threshold and a reduction in action potential la-tency of the SGNs. In KN a1KO mice, ABR wave I amplitude slopes were markedly
of the ABR wave I amplitude was not correlated to an increased loss of the SGN pe-ripheral dendrites as observed in hidden hearing loss (Liberman and Kujawa, 2017). These results are indicative of a role for KN a1 channels in shaping the activity of
the SGN peripheral dendrite. As the spike generating zone for the SGNs is located near the terminal ending of the peripheral dendrite (Hossain, 2005), this is further proof that these KN a1channels are located in the peripheral dendrite. Furthermore,
these results are indicative of a kind of hearing loss where SGN function is altered due to perturbations of normal ion channel function that does not directly lead to loss of the SGN peripheral dendrite. This appears to be a novel kind of hearing loss that has received little previous attention. Thus, importantly, these experiments have re-vealed the expression of RNA for novel proteins in the SGN peripheral dendrite and suggest new candidate proteins that could shape the function of the SGN peripheral dendrite. In addition, these experiments indicate the presence of KN a1channels in
the peripheral dendrite and suggest that they are involved in shaping the activity of the SGN peripheral dendrite. This latter finding may have implications for both subgroup differentiation and SGN peripheral dendrite vulnerability. In addition, loss of these KN a1 channels resulted in a kind of hearing loss that is distinct from
the current ideas of the normal development of hearing loss (Liberman, 2017). The contribution of these KN a1channels to loss of the peripheral dendrites and shaping
SGN subgroups needs to be further investigated. In addition, the other novel candi-date proteins could be studied in the future.
To further examine the contribution of these KN a1 channels to loss of the
pe-ripheral dendrites, the effect of KO of these KN a1channels on age related and noise
induced hearing loss is studied in chapter 4. Due to their interesting phenotype, studying KN a1KO mice could provide information on the progression of hearing loss.
First, SGNs lacking KN a1channels were shown in chapter 3 to have reduced
activa-tion thresholds. One difference between the subgroups of SGNs is a difference in activation threshold that is suggested to correlate to an increased vulnerability for SGNs with a high threshold (Furman et al., 2013; Liberman, 1978). Therefore, if activation threshold is causally related to SGN vulnerability, an overall decrease in SGN excitability may affect the vulnerability of SGNs to damage. Second, the loss of ABR wave I amplitude in KN a1KO mice prior to the loss of SGN peripheral dendrites
suggests a different development of hearing loss than has been proposed in a previ-ous model (Liberman and Kujawa, 2017). In this model, the onset of hearing loss is the destruction of the SGN peripheral dendrite that results in a reduction of wave I ABR amplitude followed by outer hair cell loss and increases in ABR wave I thresh-olds. The peripheral dendrite is lost with increasing age even without the presence of loud sound in the environment (Sergeyenko et al., 2013). In addition, moderate sound exposure results in the loss of the peripheral dendrites without obvious signs of damage to other structures surrounding the SGN peripheral dendrite (Kujawa and Liberman, 2009).
6. General discussion 129
To examine the effect of KN a1channels in the SGNs on the development of
noise-induced and age-related hearing, KN a1KO mice were either allowed to age or were
exposed to moderate and loud sounds to induce hearing loss. The KN a1 KO mice
exhibited a similar development of acquired hearing loss to control mice albeit with an accelerated pathology with regard to both ABR wave I threshold shifts and loss of the SGN peripheral dendrite. The exception was the previously observed reduction in ABR wave I amplitude that was lower in KN a1KO mice than in WT mice except at
old ages or after high sound intensity noise exposure. Therefore it remains unclear whether the loss of these KN a1channels accelerates the progression of acquired
hear-ing loss or whether this acceleration depends on a different process. These results do indicate that mice lacking these KN a1channels are more vulnerable to acquired
hearing loss. Future studies should focus on studying the mechanisms underlying this increased vulnerability and whether the acceleration of SGN peripheral dendrite loss is specific to a particular subgroup of SGNs.
6.3. Spiral ganglion neuron subgroup detection
To examine the loss of specific subgroups of SGN peripheral dendrites, these den-drites need to be classified according to their functional subgroup. However, it is technically challenging to record physiological responses of individual SGN periph-eral dendrites in mice in addition to examining the loss of SGN periphperiph-eral dendrites in response to age-related or noise-induced damage. The different subgroups of SGN are thought to preferentially target specific regions of the inner hair cell (Merchan-Perez and Liberman, 1996). Specifically, high spontaneously active SGNs are thought to target the pillar side of the IHCs, whereas the low spontaneously active SGNs are thought to target the modiolar side of the IHC. Therefore, an approximation of the subgroup to which a SGN peripheral dendrite belongs can be made by assessing the location on the IHC where the SGN peripheral dendrite makes contact (Liberman et al., 2011). In addition, the volume size of specific pre- and postsynaptic proteins and the ratio between the volume size of these proteins is thought to be further indica-tive of the SGN subgroups and to underlie in part the functional differences in SGN activity (Liberman et al., 2011). Specifically, the volume of presynaptic C-terminal-binding-protein 2 (CTBP2), that forms most of the presynaptic ribbon, was found to be larger in synapses on the modiolar side corresponding to the low spontaneously ac-tive SGNs than on the pillar side corresponding to high spontaneously acac-tive SGNs. In contrast, the volume of the glutamate receptor subunit GluA2 was found to be inversely correlated with the CTBP2 volume so that GluA2 volumes were larger on the pillar side and smaller on the modiolar side (Liberman et al., 2011). In order to examine whether the increased vulnerability to synapse loss observed in KN a1 KO
on previous methods (Liberman et al., 2011; Liberman and Liberman, 2016; Yin et al., 2014; Zhang et al., 2018) to assess the localization of SGN peripheral dendrites on the IHC surface and examine the volume of pre- and postsynaptic proteins (chapter 5).
During the development of this method, irregularities with previous reports led to a further investigation of the volume size of specific pre- and postsynaptic proteins. Volume differences between the pillar and modiolar side were found to correspond with those found previously but this was not so for volume differences of the post-synaptic GluA2. Subsequently, these proposed differences in volume size for CTBP2 and GluA2 between pillar and modiolar synapses were examined in three different strains of mice. Overall, presynaptic volume differences in CTBP2 were common across the three strains examined, but only in the strain where volume differences in GluA2 were originally found could volume differences in GluA2 be detected. In the other two strains, no differences in GluA2 volume between synapses on the pillar and modiolar side were observed. In combination with a study in gerbils where the volume differences in GluA2 were positively correlated to the volume differences in CTBP2 (Zhang et al., 2018), these results suggest variation in GluA2 volume differences be-tween synapses on the pillar and modiolar side across strains of mice and species of rodent. Therefore the idea that these volume differences underlie the functional dif-ferences of SGN subgroups needs careful re-examination. In contrast, CTBP2 volume differences were found across all strains and all studies and thus, volume gradients in CTBP2 may still be indicative of SGN subgroup. Importantly, this study shows that specific differences thought to underline the functional differences of the SGN subgroups may depend on the strain or species studied. This has important implica-tions for the implicaimplica-tions of further discoveries of differences in proteins between the SGN subgroups. These volume differences would need to be examined across species to be considered as a generalizable mechanism across mammals.
6.4. Conclusions
The goal of this thesis was to further increase our understanding of both the molec-ular architecture and the molecmolec-ular processes in the SGN peripheral dendrites. In chapter 2 the known architecture was mapped and important gaps were identified. In chapter 3 the presence of in particular voltage gated ion channels was assessed in SGNs which yielded several new candidate proteins that could regulate processes in the SGN peripheral dendrite. Of these proteins, KN a1ion channel proteins were
fur-ther examined and found to be present in the SGN peripheral dendrite. In addition, these proteins were found to have an effect on setting SGN excitability. Further-more, loss of these KN a1ion channel proteins increased the vulnerability to hearing
6.6. General discussion 131
loss of KN a1 ion channel proteins resulted in an obscure type of hearing loss that
reduces the ABR wave I amplitudes without affecting the number of SGN peripheral dendrites in young adult animals. Therefore, examination of these ion channels and possibly similar ion channels could provide further information on the development of hearing loss. Volume differences in proteins that form essential parts of the synapse between the inner hair cell and the spiral ganglion peripheral dendrite were studied in an attempt to examine differences between the SGN subtypes. These experiments indicated that volume differences in GluA2 are strain and species dependent. There-fore, the specific processes dependent on these proteins may be strain and species specific.
6.6. Future outlook
Currently there are still many mysteries concerning SGN function and vulnera-bility with many questions remaining unanswered. We still do not know what mech-anisms define the different SGN subgroups, what causes the destruction of the SGN peripheral terminals, and the processes that are ongoing in the SGN peripheral ter-minal. Direct examination of these issues is made difficult by technical difficulties in examining the SGNs in vivo. Therefore, these issues will need to be investigated in creative ways that are being made possible by developments in experimental method-ology.
Ultimately, to understand the processes within the SGN peripheral dendrite, the proteins involved need to be characterized. Fortunately more and more contributions to discovering these proteins are made largely because of the rapid development of next generation sequencing techniques. Since our publication on the afferent sig-nalling complex much work has been done by normal RNA sequencing experiments, like our work in chapter 3, and single cell RNA sequencing experiments to discover the various genes that contribute to the cellular processes at the SGN peripheral dendrite. In particular, single cell RNA sequencing experiments have revealed the existence of at least three functional subgroups of SGNs based on the genetic make-up of individual cells in mice (Petitpré et al., 2018; Shrestha et al., 2018; Sun et al., 2018). As suggested in our review (chapter 2), many of these genes that are differen-tially expressed between these subgroups encode for proteins related: to 1) glutamate signalling such as glutamate receptors, 2) voltage gated ion channel function, such as voltage gated potassium channels (KV) and hyperpolarization-activated cyclic nu-cleotide–gated (HCN) channels, and 3) lateral efferent stimulation, such as receptors involved in cholinergic responses. These SGN subgroups appear to be formed during development dependent on activity of the IHC, thus suggesting an important presy-naptic aspect in the formation of these subgroups. Important next steps will be char-acterizing the function of the proteins encoded for by these genes and determining
their function in shaping the heterogeneous responses of SGNs. Other studies have used extensive screening protocols of over a 100 knockout mouse models for various proteins to study their contribution to hearing pathology (Bowl et al., 2017). Such studies could be tailored to specifically study proteins that are part of the afferent signalling complex and contribute to SGN heterogeneous responses. Our work in chapter 3 is an example of one such a “tailored” study but limited to a single knock-out mouse model. Future studies could make use of multiple other knockknock-out mouse models using a similar approach.
To some extent the proteins that determine heterogeneous responses in SGNs will likely be directly or indirectly related to pathways that result in excitotoxic processes that result in the loss of the peripheral terminal or apoptosis of the SGN. However, the precise mechanisms that result in the loss of the SGN peripheral terminal are unclear. Therefore, comprehension of how proteins that shape SGN functional hetero-geneity impact SGN vulnerability can only come from examination of the molecular pathways resulting in the loss of the SGN peripheral terminals. Over-exposure by glutamate and resulting excitotoxicy are considered the prime mechanisms resulting in loss of the peripheral terminals, yet we understand very little about these processes (Pujol and Puel, 1999). In the CNS, glutamate over-exposure results in cellular dam-age either through Na+ influx related pathways or Ca2+ influx related pathways.
But whether glutamate over-exposure follows similar pathways in the SGN periph-eral terminals is unclear. First, the main glutamate receptor is thought to be largely Ca2+ impermeable due to the composition of the subunits within these glutamate receptors (see chapter 2). Therefore, it is currently unclear to what extent Ca2+
in-flux into the SGN occurs, where this inin-flux would come from, which ion channels are involved, and whether this influx then influences SGN peripheral terminal loss. Second, size gradients in synaptic volumes are thought to underlie, at least in part, the functional differences of the SGNs (see chapter 5). The more vulnerable SGNs are suggested to have larger ribbons implying increased glutamate release and fewer glutamate receptors implying a reduced influx of Na+ compared to the less
vulner-able SGNs. If glutamate over-exposure involves GluR-mediated Na+ influx into the
SGNs, then the idea of more vulnerable SGNs with fewer glutamate receptors seems paradoxical. In addition, the generalizability of these size gradients across mammals is unclear (see chapter 5). Therefore, it remains unclear how differences in Na+influx
might contribute to differences in SGN vulnerability. Third, in cultured SGN soma, glutamate receptors are internalized following exposure to glutamate via a Ca2+
de-pendent mechanism in vitro (Chen et al., 2009, 2007). This complicates matters fur-ther as the number of glutamate receptors may be dynamically regulated in vivo. To fully understand how the SGN peripheral terminals are lost and how physiological differences between SGNs arise, the molecular machinery and processes within the peripheral dendrites need to be elucidated. This includes further identification of gene and protein expression in the SGN. In addition, the response to glutamate
ex-6.6. General discussion 133
posure of the different types of SGN need to be studied. Much work has focussed on stringent assessment of SGN peripheral dendrite survival and identification of the SGN subtypes in a variety of experiments. These same techniques should in future be applied to study the effect of titrated exposure of glutamate to the SGN peripheral dendrites. Furthermore, recordings from the peripheral dendrites are technically challenging but have recently been performed in tissue from adult animals (Wu et al., 2016). In addition, optogenetic tools allow control of the release of glutamate from the inner hair cells to provide a titrated, physiologically relevant, exposure of glutamate to the SGNs. Combined, studies applying these techniques can provide information on the processes that are ongoing within the SGN peripheral dendrite during exposure to glutamate.
