Electrophysiological properties of outer hair cells and the localization of Ctbp2 and Shank1

(1)

Electrophysiological properties of outer hair cells

and the localization of Ctbp2 and Shank1

 Student: Saša Peter, 5644844

 Supervisor: Dr. Johanna Montgomery

 co-assessor / UvA representative: Dr. Harm J Krugers  50 ECs

 15-03-2011 / 15-12-2011 (final date 24-4-2012)  2nd

Research Project, Auckland, New Zealand at the Centre for Brain Research  MSc in Brain and Cognitive Sciences, University of Amsterdam, Neuroscience Track

(2)

2

Introduction

In the mammalian cochlea, the process of hearing involves transduction of the acoustic signal through optimal innervation of the sensory hair cells (Spoendlin 1975;

Defourny, Lallemend et al. 2011). The tonotopically organized sensory epithelium, the organ of corti (OC), contains two functionally distinct neuron populations which convey the sound information from the cochlea to the central nervous system. In the OC each inner hair cell (IHC), regarded as the primary functional conveyer of the acoustic signal, is innervated by multiple type I spiral ganglion neuron (SGN) afferent fibers (Nienhuys and Clark 1978; Pickles 1988). This type I SGN innervation to the IHCs accounts for approximately 95% of the total neuron population in the OC. By contrast, the afferent innervation of the outer hair cells (OHCs) is conveyed by the SGN type II fibers, which share “en passant” connections to the OHCs and account for approximately 5% of the remaining neuron population (Keithley and Feldman 1982; Simmons and Liberman 1988). Even though the physiological function of this type II SGN innervation has not yet been clearly defined, there is evidence that it could be part of a sensory control loop that amplifies cochlear sensitivity through feedback to the inhibitory olivocochlear efferent innervation of both the OHCs and the postsynaptic region of the type I afferents at the IHCs (Pickles 1988; Jagger and Housley 2003; Darrow, Maison et al. 2007).

The neonatal development towards this mature configuration of afferent innervation in the mouse cochlea goes through roughly three different stages: neurite growth and extension of both type I and type II afferents to all hair cells (E18-P0); neurite refinement, with more advanced formation of spiral bundles innervating OHCs (P0-P3); and finally neurite

retraction, which eliminates type I SGN innervation to the OHCs while preserving the type I innervation of the IHCs (P3-P6) (Huang, Thorne et al. 2007). The study of this naturally occurring synapse elimination during neonatal afferent development is an interesting one for several reasons. The mammalian nervous system involves many similar synchronous events where formation, plasticity and retraction of neurites to appropriate targets take place (Malenka 2003; Montgomery and Madison 2004) . In this sense the cochlea can be used as a model system to study neurite formation, plasticity and elimination in the central nervous system. In the current report, emerging evidence will be discussed on the

differential molecular mechanisms that define the permanent versus transient state of type I fibers. Furthermore, for the first time, data is presented of the postsynaptic shank1 protein

(4)

4

expression in the synapses of hair cells. The shank1 protein, being a postsynaptic “master regulator”, binds many different scaffold proteins involved in the structure, maintenance and plasticity of the post synaptic density. And is for these reasons, an interesting target to look at when discussing possible regulatory mechanisms of the permanent versus transient state of the type I fibers. In the following paragraphs, a more detailed picture is given on the transient versus permanent synapses and the ideas behind the molecular and functional approaches that were taken to investigate the differential synaptic profiles between them. Transient vs. permanent

In the mouse cochlea, type I afferent fibers innervate both IHC and OHC before the mature configuration at the end of the postnatal week is established where these fibers only innervate the IHCs (Huang, Thorne et al. 2007). From these observations it follows that the transient type I fibers that innervate the OHCs exhibit different molecular mechanisms from the permanent type I fibers. It is still unknown what these mechanisms are, and what the drive is to differentiate in the first place. However, a recent study by Huang (et al. submitted) has provided new insights into the synaptic profiles of these transient and permanent

synapses. This study revealed that the type I transient synapses on OHCs have synaptic ribbons that are anchored by the bassoon protein and directly oppose AMPA receptor subunits and synaptic scaffold proteins, despite the temporary nature of these synapses. Ribbon synapses are specialized electron dense structures that are capable of synchronous multi-vesicle release at the type I permanent synapses of IHCs (fig. 1) (Sterling and Matthews 2005). However, between P3 and P6 ribbons in the OHCs disperse from the active zone towards the soma instead. This dispersion of ribbons seems to be of significance as it has been shown to coincide with the elimination of the type I afferents from the OHCs (Huang et al. submitted). The time dependant dispersion of ribbons in the OHCs is preceded by a decrease in bassoon that is thought to anchor the ribbons to the active zone. These types of synaptic assemblies suggest that the transient type I synapse at the OHCs could be a

(5)

5 Fig. 1: A schematic drawing of the organ of corti, where you can see the one row of IHCs and the three rows of OHCs. In this representation, only afferents to the IHCs are pictured. A closer look at the synapses of the IHC afferent synapse is presented, including a schematic representation of the presynaptic ribbon responsible for synchronized multi vesicle release. Adopted from (Meyer and Moser 2010).

Ctbp2 and Shank1

The ribbon synapse can be found in the retina and the inner ear where it maintains tonic neurotransmitter release so that a broad range of stimulus frequencies can be represented (Schmitz 2009). These high rates of neurotransmitter release require both functional and structural adaptations within the synapse. The synaptic ribbon, an electron dense structure that binds many synaptic vesicles closely to the active zone, is a crucial part of the ribbon synapse. The main structural part of the synaptic ribbon is the RIBEYE domain, which is responsible for the buildup of the ribbon through RIBEYE-RIBEYE interaction. The structure of RIBEYE consists of two domains (A and B), which are both quite distinct in their buildup (fig. 2). Domain A consists of a unique aminoterminal, proline-rich A-domain to which no known homologous proteins exist, whereas domain B is largely identical to the co-repressor protein Ctbp2 (Schmitz, Königstorfer et al. 2000; Alpadi, Magupalli et al. 2008).

(6)

6

Therefore domain B makes Ctbp2 a good antibody target as a means for localization of the synaptic ribbon.

Fig. 2: A representation of the domain structure of RIBEYE. Consisting of two domains (A and B), the B domain points towards the cytoplasm while the A domain is thought to anchor with other RIBEYE A domains to create the main rigid structure of the ribbon. The B domain has several isomers, of which ctbp2 is a good antibody target to localize the synaptic ribbon. Adopted from (Sheng and Kim 2000).

The postsynaptic density (PSD) is a complex structure central in excitatory glutamatergic synapses, which is composed of hundreds of different proteins connecting surface

transmitter receptors and their signal transduction machinery (Kreienkamp 2008). Because of its compact size and complexity, scaffold proteins are needed that can keep this assembly together through protein interaction sites. The shank1 protein, almost exclusively expressed in the brain, is part of the shank family of postsynaptic scaffold proteins that interact with many different membrane and cytoplasmic proteins (fig. 3) (Lim, Naisbitt et al. 1999; Yao, Hata et al. 1999; Sheng and Kim 2000). Because of its size and many interaction sites, it has been often called the ‘master regulator’ (fig. 4). The presence of shank1 is for these reasons, a good indicator for the formation of the PSD and consequentially a functional synapse which makes it a good target to investigate the transient vs. permanent type I afferent synaptic profiles.

Fig. 3: Schematic domain structure representing the interaction domains (in color) and alternative splicing sites (arrow heads) of shank1. Adopted from (Sheng and Kim 2000).

(7)

7

Fig. 4: A schematic representation of the postsynaptic density, involving many different scaffold proteins connecting receptors and their associated messenger systems. Here a central role for the shank protein is pictured representing its ability to anchor and connect different scaffold proteins through a variety of binding sites. Adopted from (Sheng and Kim 2000).

In the current study, focus on pre and postsynaptic expression of ctbp2 and shank1

respectively has given more insight into the synaptic profiles of the OHCs and IHCs. By means of co-localization of these proteins different expression patterns have been revealed that might help explain the differential neonatal synaptic development in permanent vs. transient afferent fibers.

Electrophysiological properties of hair cells

Having discussed several molecular mechanisms that could help explain the

differential synaptic profiles between transient en permanent type I afferents, the question remains how this is expressed in a functional manner involving pre synaptic spike generation and post synaptic EPSCs. In order to understand the difference between transient (OHCs) vs. permanent (IHCs) type I afferent synapses, we need to consider the functional

differentiation which occurs during the neonatal period. At about embryonic day E14-E15 in mice, IHCs and OHCs already start to differentiate in their expression of voltage gated ion channels (Marcotti, Johnson et al. 2003; Helyer, Kennedy et al. 2005). During the next few weeks, both type of hair cells undergo significant changes that transform these cells into mature sound transducers (around P12) with a distinct population of voltage gated ion channels (fig. 5). One significant difference between the two types of neonatal hair cells, is that IHCs are capable of generating rapid spontaneous Ca2+ action potentials and the OHCs

(8)

8

are not (Housley, Marcotti et al. 2006). In the IHCs this spontaneous spike generation stops before the onset of hearing (Tritsch, Yi et al. 2007), after which spikes can only be generated following depolarizing current injections (Glowatzki and Fuchs 2000). It is important to consider, even though the exact role of spontaneous spiking activity in the neonatal IHCs has not yet been revealed (however see (Tritsch, Yi et al. 2007)), that the spontaneous spiking could contribute to post-synaptic type I afferent synapse refinement. This would not be the first time that similar spontaneous activity is linked to refinement and maturation of

synapses, in fact, exactly this type of correlation has been found in several different neural circuits (Zhang and Poo 2001). In this sense, the disappearance of the spontaneous activity in IHCs before the onset of hearing is to some extent in line with the morphological type I afferent maturation to the IHCs and may help to explain its ‘permanent’ state. In addition, spontaneous activity has been shown to be of great importance for the maturation of auditory pathways (Tritsch, Yi et al. 2007). The OHCs on the other hand do not possess this spontaneously generated spiking, and could in this sense be limited in their refinement and enhancement of their type I afferent connection towards a more stable and mature state. And exactly this lack of ‘enhancement’ from the pre-synaptic site might trigger, or be involved in, molecular mechanisms which initiate the down regulation of bassoon and subsequent dispersion of the ribbons towards the soma which coincides with the retraction and elimination of the type I transient afferent fibers.

The nonexistence of spontaneous activity in the OHCs, can to an extent, be explained by the absence of the SK current (fig. 5) in the OHCs during the neonatal period and its importance in the generation of spontaneous spike generation in the IHCs (Marcotti, Johnson et al. 2004). In addition to the absence of SK current in the OHCs, differential concentrations of Ca2+ buffering proteins between IHCs and OHCs might play a role in Ca2+ dependent

intracellular mechanisms that would be needed in the generation of sustained spontaneous activity (Hackney, Mahendrasingam et al. 2005).

(9)

9 Fig. 5. Changes in the expression of multiple currents during development in the mammalian cochlea (mainly mouse, but also rat and gerbil with similar auditory development). The width of the bars indicates the maturation of the size of the currents. Adopted from Housley & Marcotti (2006).

So even if this spontaneous spike generation proves to be critical in the maturation of the type I afferent fibers, the question remains if the type I transient afferents at the OHCs develop fully and have the morphological and physiological properties to be functional during their short P0 to P3 neonatal existence. A good way to investigate this would be to use dual whole-cell recordings by stimulating the pre-synaptic component (hair cell) and measuring the post-synaptic currents of the type I afferent synapse. This would be an excellent approach towards more insight in the functional profiles of the transient vs. permanent type I afferents. In the current study, the focus lies on the preparation of such a setup for the near future. Here the pre synaptic possibility of spike generation in the OHCs is investigated during the three different stages of neonatal afferent fiber development.

(10)

10

Methods

Animals

All procedures performed for experiments were approved by the University of Auckland Animal Ethics Committee. The animals used were C57 mice, aged postnatal day 0, 3 and 6 (P0, P3, P6). The cochlear tissue was obtained after the animals were killed with pentabarbitone sodium (60 mg/kg; Nembutal, Virbac Laboratories Ltd. New Zealand) by intra-peritoneal injection.

Immunohistochemistry

The cochlea were dissected and perfused with 1 % paraformaldehyde in 0.1 M phosphate buffer (PFA; pH 7.4) via the round and oval windows and fixed in PFA solution for one hour at room temperature. The cochlea were the prepared for cryoprotection by 30% sucrose incubation for 48 hours. In preparation of cryosectioning the cochlea were mounted in an O.C.T. sucrose solution (50:50 ratio) and sectioned using a cryostat microtome

(CM1900, Leica, Germany) at 50 µm thickness after which the sections were placed in 0.01 M phosphate buffer saline (PBS; PB with 0.9% NaCl at pH 7.4). All cochlea tissue were

incubated in endogenous mouse IgG blocking solution (Fab fragment Donkey anti-mouse IgG (H+L) and whole Donkey anti-mouse IgG (H+L) Jackson immunoresearch laboratories, INC. PA, USA; 1:50 ratio in 0.01 M PBS with 5% normal horse serum (NHS), 1% bovine serum albumin (BSA; Gbico) and 0.4 % tritonX-100 overnight at 4oC. The following day three washes were performed of 5, 15 and 30 minutes respectively using 0.01M PBS with 0.25% TritonX-100. After the washing procedure sections were incubated in a blocking and permeablisation solution (5% NHS, 5% normal goat serum (NGS), 1% BSA and 0.4 % TritonX-100 in 0.01 PBS) for one hour at room temperature. The primary antibodies (CtBP2 mouse monoclonal BID- Biosciences (1:1000 ratio) and Shank-1 rabbit polyclonal Neuromics (1:500 ratio)) were then diluted in the blocking solution and applied to the sections for 48 hour at 4oC. In order to remove non bound primary antibody four washes were performed of respectively 5, 10, 15 and 30 minutes. The secondary antibodies (Alexa-488 goat anti-rabbit IgG (1:500 ratio) and Alexa-647 goat anti-mouse IgG1 (1:500 ratio)) were diluted in the blocking solution and applied to the sections for two hours at room temperature. The secondary antibodies wash protocol was 5, 10, 15, 30 and 60 minutes respectively after which the cochlear sections

(11)

11

were mounted in 20 µl Citifluor (AF1, Citifluor Ltd, UK) on slides and covered with a coverslip.

Imaging and analysis

Image acquisition was done using a confocal microscope (Olympus FV1000, Japan) and processed using Image J plus software. The pixel size was 90 nm x 90 nm x 200 nm, while making sure that there was no saturation so that the information coupled to the signal intensity was not lost. The optical resolution of the confocal microscope used was 240 nm in lateral and 500 nm in axial direction, which covers the size of the pre-synaptic ribbon of 380 nm (Meyer, Frank et al. 2009) so that most puncta would be detected.

In order to increase the signal-to-noise ratio the confocal fluorescence images were

deconvolved before 3D reconstructions with image-Pro plus 3D suite (MediaCybernetics Inc.) for quantitative measures was performed. The deconvolution processing was done using the Huygens Essential Software (Scientific Volume Imaging, Hilversum, The Netherlands). The stack of images obtained after deconvolution were then processed using the 3D constructor of Image-Pro plus. The first step was to ensure that objects that occupied more than two pixels in each direction passed the filter so that background fluorescence was not included. Colocalization for CtBP2 and Shank1 was then determined using 3D object based analysis. First the coordinates of the center of the objects were obtained, after which the distance between the center of presynaptic objects and the center of their neighboring postsynaptic objects smaller than 0.5 µm was identified for those two objects to be colocalised. The formula used to calculate the distance between two objects (Pythagorean Theorem) was:

Statistical analysis was done using SPSS (19.0, SPSS Inc, U.S.A.) determining the mean and standard error of the mean. For the electrophysiological data where there were less than 3 groups compared student’s T-test was used. For multiple comparisons, analysis of variance (ANOVA) was used followed by Tukey’s test with P < 0.05 as the criterion for statistical significance.

Electrophysiology

After the removal of the cochlea’s, the Organ of Corti was transferred to the microscope chamber (Olympus BX51WI with camera from Dage MTI – model NC-70) and

(12)

12

held under an insect pin on a small coverslip. The chamber was super fused (2 ml min-1) with artificial perilymph (AP) consisting of (mM): 142 NaCl, 5.8 KCL, 1.3 CaCl2, 0.9 MgCl2, 0.7

NaH2PO4, 2 sodium pyruvate, 5.6 C6H12O6 (D-Glucose), 10 HEPES. The pH of the AP was

adjusted to 7.4 (NaOH) with osmolality around 307 mosmol kg-1. Membrane currents of outer hair cells (OHC) were studied at room temperature ( 21oC) using the whole cell patching technique. The pipettes used for patching were pulled from glass capillaries with a resistance of 5 to 6 MΩ in the bath solution. The intracellular solution used contained (mM): 20 KCL (or TEA.CL), 0.1 CaCl2, 5 MgCl2, 5 HEPES, 5 EGTA, 4 2 ATP, 0.3 GTP, and 5

Na-phosphocreatine. The pH of the intracellular solution was adjusted to 7.2 (KOH) with osmolality around 294 mosmol kg-1. Data was acquired using pCLAMP software (version 9, Axon instruments, USA) and stored on the pc for offline analysis. Membrane potentials were automatically corrected for series resistance (compensation up to 70 % just before

oscillations occurred). The voltage clamp protocols were conducted from a -85 mV holding potential. The current clamp protocols were conducted with a 300 pA or a 500 pA injection, while lasting for a short (26.5 ms) or a long (326.5 ms) period of time.

(13)

13

Results

The data presented is gathered during the neonatal development towards the mature configuration of afferent innervation in the mouse cochlea. Namely, neurite refinement (P0-P3); and there after neurite retraction, which eliminates type I afferent innervation to the OHCs while preserving the type I innervation of the IHCs (P3-P6) (Huang, Thorne et al. 2007).

Imaging pre and postsynaptic proteins during synapse formation and elimination

The pre-synaptic ribbons and the post-synaptic shank1 protein were localized using the Ctbp2-RIBEYE and Shank-1 immunolabeling respectively. Both Ctbp2 and Shank1 puncta were observed in the P0, P3 and P6 neonatal ages (fig. 6.).

Fig. 6: Immunolabeling of Ctbp2 (red) and Shank1 (green) in the neonatal (P0-P6) OHCs and IHCs of the mouse cochlea. Notice the dispersal of Ctbp2 from P3 to P6, which reflects the previously reported movement of the ribbons from the active zone towards the soma (Huang et al. submitted).

In the IHCs, for all postnatal ages, the ribbons were located in the basolateral region of the cell where the afferent fiber synapses are located. The shank1 total puncta count (n=3)

P0

P3

P6

IHC IHC IHC

OHC OHC OHC

OHC OHC

OHC

(14)

14

increased significantly from P0 to P6 in the IHCs, while the ctbp2 puncta remained stable during the same period (fig. 7 A + B). The ctbp2 co-localization with shank1 is stable and relatively high during the P0 to P6 neonatal period (fig. 7 D). However, when looking at the shank1 co-localization with ctbp2 (fig. 7 C) a decreasing trend is visible which could be due to the branching of the IHC fibers that occurs during this period of time (Echteler 1992).

Fig. 7: Shank1 and ctbp2 immunolabeling puncta count (A & B) and co-localization (C & D) in the IHCs of the mouse cochlea for P0, P3 and P6 (n=3 for each age). A statistically significant difference at the *p<0.05 level (in graph A) was found using ANOVA followed by Tukey’s test. The error bars represent the S.D. values.

In the OHCs, the location of the ribbons differentiated across the postnatal ages

investigated. Between P0 and P3 the ribbon puncta were mostly found in the basolateral area of the OHCs (fig 6). However, in P3 to some extent, and even more in P6; the ribbons are moving away from the basolateral active zone towards the soma. The shank1 and Ctbp2 total puncta count (n=3) has an increasing trend from P0 to P6 in the OHCs (fig. 8 A + B). In contrast with the IHCs, the Ctbp2 co-localization with shank1, and vice versa, decreases dramatically from P3 to P6 (fig. 8 C + D) in the OHCs. Even though the results are not significant due to the large deviation in P0-P3 compared to P6 (more n is needed); this

C

D

(15)

15

absence of co-localization reflects the previously reported dispersion of ribbons from the active zone towards the soma (Huang et al. submitted).

Fig. 8: Shank1 and ctbp2 immunolabeling puncta count (A & B) and co-localization (C & D) in the OHCs of the mouse cochlea for P0, P3 and P6 (n=3 for each age). Even thou no statistically significant difference was found due to the large deviations within the population, a relatively strong trend is present (figures C & D) from P3 to P6 reflecting the dispersion of ribbons to the soma. The error bars represent the S.D. values.

Electrophysiological characterization of hair cells during synapse formation and elimination

For the electrophysiological part of the report, involving the functional synaptic profiles of permanent versus transient synapses, it would be ideal to have a dual whole-cell recording setup where stimulation of the pre-synaptic component (hair cell) is induced and the postsynaptic current is measured. Here we make the first steps towards such an

approach by characterizing OHC outward currents (voltage clamp) and spike induction (current clamp). In the current setup experiments were performed on OHCs alone as the IHCs were not visible enough, which made the whole cell recordings for these cells not possible at the time.

A

B

D

C

(16)

16

The mean resting membrane potential of the cells recorded (n=8 for each age) was -52mv  6.9 for P0, -50mv  4.3 for P3 and -48mv  4.1 for P6 with no significant difference between them. The capacitance found was 7,2pF  2.3 for P0, 8,6pF  1.8 for P3 and 8,2pF  2.1 for P6 with also no significant difference between the ages. These values differ slightly from previous reports (Marcotti and Kros 1999; Helyer, Kennedy et al. 2005), which might be due to the slight differences in the APF and internal solutions used.

K+ _{outward currents}

Depolarizing voltage steps in the OHCs of the mouse (P0, P3 and P6) caused slowly activating voltage dependent outward currents; an example of the voltage clamp protocol can be seen in (fig 9).

Fig. 9: Example of a voltage clamp protocol in a P6 OHC (B), demonstrating the outward currents (A) when depolarizing voltage steps are induced. The membrane potential was held at -85mV, and from there 10mv voltage steps were applied.

Using 20 mM TEA (selective K+ receptor antagonist) instead of 20mM KCL in the internal solution, the outward currents reduced up to 90% (n=2, P6) confirming the identity of the outward currents as K+ currents (Marcotti and Kros 1999). In fig. 10 the outward K+ current, which increased in size with postnatal age (P0, P3 and P6), was induced using the

A

(17)

17

voltage protocol described. From these results it is clear that significant development of the K+ outward current takes place during the first neonatal week.

Fig. 10: I/V relationship of the outward K+ currents in the OHCs (P0, P3 and P6) induced in voltage clamp mode with 10mv depolarizing voltage steps from a -85mv holding potential. A statistically significant difference between all ages at the **p<0.01 level was found using ANOVA followed by Tukey’s test. The error bars represent the S.D. values.

Voltage responses to current injections

It has been known for quite some time that IHCs are capable of generating

spontaneous neonatal spiking activity (Housley, Marcotti et al. 2006). However, the same is not true for the OHCs (Marcotti and Kros 1999). Using the gap-free recording option in pCLAMP, a continuous current clamp recording was made (5 min; not shown here) to confirm that there was indeed no spontaneous spike generation in the OHCs.

Spikes were generated using current clamp with injections of 300pA or 500pA with stimulation lasting for a short (26.5 ms) or a long (326.5 ms) period of time (fig 11). The short stimulation was intended to induce a single spike; the long stimulation would do the same initially and it would also reveal the possibility of additional spikes being generated and show any voltage fluctuations due to the prolonged current injection. From these results it becomes clear that there are differences in spiking properties across the neonatal ages. For all three neonatal ages studied, peak/antipeak amplitude and time have been determined and compared. One striking observation is that P3 and P6 OHCs are capable of generating a spike. For P0 however, it was clear that there was no spike as there was no trace of

(18)

18

Because of these observations P0 has been excluded from comparison in spike properties with P3 and P6 in the following paragraph.

Fig. 11: Spikes generated in the OHCs using either a short 26.5ms (A) or a long 326.5ms (B) current injection. The current injected was either 300pA or 500pA. Here it is visible that P0 has no

repolarization or hyperpolarization during the current injection and thus shows no recognisable spike shape.

(19)

19

Differential spike properties across postnatal ages

An overview of the peak and antipeak amplitude and timing, generated using the described current clamp protocols, can be seen in table 1. Both the (anti) peak timing and amplitudes show a significant development from P3 to P6.

Table 1: Overview of the spike data collected using the current clamp protocols in P3 and P6 OHCs of the mouse cochlea.

Even though there was no clear P0 spiking data to compare with, it is assumable that a similar significant development of spike properties takes place from P0 to P3. The spike development, between P3 and P6, can be seen in fig. 12. From these results it becomes clear that between P0 and P3 OHCs become capable of depolarization, repolarization and

subsequent hyperpolarization, and that the voltage dependent currents change in amplitude (fig. 12 A and C) as well as timing (fig 12 B and D).

(20)

20 Fig. 12: The amplitude and timing of peaks and antipeaks in OHCs generated with a 300pA current injection. A statistically significant difference at the **p<0.01 level was determined using a student’s t-test. The error bars represent the S.D. values.

A

C

B

(21)

21

Discussion

In the mammalian cochlea, transduction of the acoustic signal is achieved through optimal innervation of the sensory cells; the inner and outer hair cells (Spoendlin 1975; Defourny, Lallemend et al. 2011). The type I afferent fibers have been shown to innervate both the IHCs and OHCs up to neonatal age P3 in the mouse cochlea, after which the type I fibers innervating the OHCs become eliminated, and are thus defined as transient, while the type I fibers at the IHCs become permanent (Huang, Thorne et al. 2007). In the current study, different techniques have been employed to provide additional insights into the synaptic profiles of both hair cell types as a means towards more understanding on the differentiation between the transient and permanent states of the type I afferents.

Changes in the shank1 and ctbp2 expression during neonatal development Here for the first time, we describe the postsynaptic shank 1 expression in the IHCs and OHCs of the mouse cochlea. In the IHCs, shank1 increases significantly from P0 to P6, and shows high co localization with presynaptic ribbons. The presence of shank1, being a good indicator of the PSD, together with the co localization of the ribbons shows how the development progresses towards functional maturation for the afferent type I permanent fibers innervating the IHCs. In the OHCs on the other hand, a dispersion of the ribbons takes place (very clear in P6) from the active zone towards the soma which confirms previous findings reporting the correlation between the distribution of synaptic ribbons and the elimination of the transient type I afferent fibers (Huang et al. submitted). Additionally, here we can clearly derive the dispersion of the ctbp2 labeled ribbons from the absence of co-localization with the post-synaptic shank1 scaffold protein from P3 to P6 in the OHCs. An important argument for the elimination of the transient type I fibers comes from previous studies that have shown that synapses with lower neurotransmitter release probability, compared to other synapses with higher release potential, become eliminated and the neurites retracted (Kopp, Perkel et al. 2000; Tashiro, Dunaevsky et al. 2003). In this sense, the importance of the ribbon synapse in sustaining large synchronized periods of neurotransmitter release can be of significant importance for the maturation and survival of the synapse. This also underlines the importance of the shank1 protein as an architectural component of the PSD and indicator of a functional synapse. Because shank1 is so important in its role to anchor and connect receptors and their associated messenger systems in the

(22)

22

PSD, the down regulation of shank1 at a given synapse could be an important factor for destined elimination, and remains for these reasons an interesting target when discussing permanent versus transient synaptic profiles.

Electrophysiological properties of OHCs

Using depolarizing voltage steps in the OHCs, evidence is provided which confirms the presence of the previously described slowly activating K+ outward currents in the OHCs (Marcotti and Kros 1999). This K+ current has been shown to increase significantly from P0 to P6, and reflects important channel expression and localization for the facilitation of the Ik.

The increase of the Ik from P0 to P6also reflects the differential spike properties described

here. Significant differences in spike properties between P3 and P6 OHCs are presented in both the peak and antipeak amplitude and timing. The spikes generated in current clamp show clear depolarization, repolarization and subsequent hyperpolarization states. In contrast, P0 shows no repolarization and thus no recognizable spike shape. The absence of the spike in P0 has been confirmed also by prolonged current injection, which was

performed to check whether it might be a delayed spike beyond the duration of the short current injection. These findings, of differential spiking properties in OHCs in early neonatal ages, underline the major development that takes place in the amount and types of

receptors expressed.

One difference between the two types of neonatal hair cells, is the finding that IHCs are capable of generating rapid spontaneous Ca2+ action potentials and the OHCs are not

(Housley, Marcotti et al. 2006). In the current report, this absence of spontaneous activity in the OHCs is confirmed. And exactly this difference could be of great importance in the transient vs. permanent type I afferent determination. Together with the down regulation of the ribbon anchoring protein: bassoon, and the subsequent ribbon dispersal from the active zone towards the soma (Huang et al. submitted), the absence of the spontaneous activity in the OHCs to consolidate and refine the type I afferent synapses altogether could prove to be important factors in the type I fiber elimination. In contrast, the IHCs show spontaneous activity which more than likely helps to facilitate the maturation and refinement of the ribbon synapse, which is reflected also in the high co localization across the neonatal ages of shank1 and the ribbons.

(23)

23

Conclusion

In the current report both molecular and functional data is provided which reflects the developmental differentiation of synapses and adds towards more understanding into the differences between transient and permanent type I afferents in the mouse cochlea. Here it is confirmed that a dispersion of the ribbons takes place from the active zone towards the soma (P3 to P6) in the OHCs during the same time that the type I transient fibers are eliminated. For the first time, the expression of shank1 in the neonatal afferent synapses is investigated. Reduced co-localization of shank1 and ribbons in the OHCs reflects the dispersal of the ribbons from the active zone during the time that the transient fibers are eliminated, while the same co-localization is preserved at the permanent type I fibers

innervating the IHCs. The role of shank1 in this sense can be an important one, as its down regulation could result in less defined PSD, which together with the absence of

multivesicular release from the ribbon synapse could lead to synapse elimination due to significantly reduced consolidation. In order to get an even better profile of these synapses, functional data obtained with a dual whole cell patch setup would be a big step forward. In the current report, data on the electrophysiological characteristics of OHCs is presented which is an important step towards this dual whole cell patch setup. Differential spiking properties found in the neonatal OHCs reflect the significant changes that take place during neonatal development in the expression of voltage gated ion channels. Investigating

additional synaptic targets, that are involved in the stability of the ribbon (like bassoon) and the postsynaptic density (e.g. shank1, GIRK, PSD95) combined with functional data from dual whole cell patching will provide more understanding into the differential synaptic profiles of transient versus permanent synapses. And with these insights, gained through the study of the cochlea as a model system for neurite formation, we can work towards better

understanding of the mechanisms behind formation, elimination and plasticity of neurites towards their appropriate targets in the brain.

(24)

24

References

Alpadi, K., V. G. Magupalli, et al. (2008). "RIBEYE recruits Munc119, a mammalian ortholog of the Caenorhabditis elegans protein unc119, to synaptic ribbons of photoreceptor synapses." Journal of Biological Chemistry 283(39): 26461.

Darrow, K. N., S. F. Maison, et al. (2007). "Selective removal of lateral olivocochlear efferents increases vulnerability to acute acoustic injury." Journal of neurophysiology 97(2): 1775-1785.

Defourny, J., F. Lallemend, et al. (2011). "Structure and development of cochlear afferent

innervation in mammals." American Journal of Physiology-Cell Physiology 301(4): C750-C761.

Echteler, S. M. (1992). "Developmental segregation in the afferent projections to mammalian auditory hair cells." Proceedings of the National Academy of Sciences 89(14): 6324. Glowatzki, E. and P. A. Fuchs (2000). "Cholinergic synaptic inhibition of inner hair cells in the

neonatal mammalian cochlea." Science's STKE 288(5475): 2366.

Hackney, C. M., S. Mahendrasingam, et al. (2005). "The concentrations of calcium buffering proteins in mammalian cochlear hair cells." The Journal of Neuroscience 25(34): 7867-7875.

Helyer, R. J., H. J. Kennedy, et al. (2005). "Development of outward potassium currents in inner and outer hair cells from the embryonic mouse cochlea." Audiology and Neurotology 10(1): 22-34.

Housley, G., W. Marcotti, et al. (2006). "Hair cells–beyond the transducer." Journal of Membrane Biology 209(2): 89-118.

Huang, L. C., P. R. Thorne, et al. (2007). "Spatiotemporal definition of neurite outgrowth, refinement and retraction in the developing mouse cochlea." Development 134(16): 2925-2933.

Jagger, D. J. and G. D. Housley (2003). "Membrane properties of type II spiral ganglion neurones identified in a neonatal rat cochlear slice." The Journal of physiology 552(2): 525-533. Keithley, E. M. and M. L. Feldman (1982). "Hair cell counts in an age-graded series of rat

cochleas." Hearing research 8(3): 249-262.

Kopp, D. M., D. J. Perkel, et al. (2000). "Disparity in neurotransmitter release probability among competing inputs during neuromuscular synapse elimination." The Journal of

Neuroscience 20(23): 8771-8779.

Kreienkamp, H. J. (2008). "Scaffolding proteins at the postsynaptic density: shank as the architectural framework." Protein-Protein Interactions as New Drug Targets: 365-380. Lim, S., S. Naisbitt, et al. (1999). "Characterization of the Shank family of synaptic proteins."

Journal of Biological Chemistry 274(41): 29510.

Malenka, R. C. (2003). "The long-term potential of LTP." Nature Reviews Neuroscience 4(11): 923-925.

Marcotti, W., S. L. Johnson, et al. (2003). "Developmental changes in the expression of potassium currents of embryonic, neonatal and mature mouse inner hair cells." The Journal of physiology 548(2): 383-400.

Marcotti, W., S. L. Johnson, et al. (2004). "A transiently expressed SK current sustains and modulates action potential activity in immature mouse inner hair cells." The Journal of physiology 560(3): 691-708.

Marcotti, W. and C. J. Kros (1999). "Developmental expression of the potassium current IK, n contributes to maturation of mouse outer hair cells." The Journal of physiology 520(3): 653-660.

Meyer, A. C., T. Frank, et al. (2009). "Tuning of synapse number, structure and function in the cochlea." Nature neuroscience 12(4): 444-453.

Meyer, A. C. and T. Moser (2010). "Structure and function of cochlear afferent innervation." Current opinion in otolaryngology & head and neck surgery 18(5): 441.

Montgomery, J. M. and D. V. Madison (2004). "Discrete synaptic states define a major mechanism of synapse plasticity." Trends in neurosciences 27(12): 744-750.

(25)

25 Nienhuys, T. G. and G. M. Clark (1978). "Frequency discrimination following the selective

destruction of cochlear inner and outer hair cells." Science 199(4335): 1356-1357. Pickles, J. O. (1988). An introduction to the physiology of hearing, academic Press London. Schmitz, F. (2009). "The making of synaptic ribbons: how they are built and what they do." The

Neuroscientist 15(6): 611-624.

Schmitz, F., A. Königstorfer, et al. (2000). "RIBEYE, a component of synaptic ribbons: a protein's journey through evolution provides insight into synaptic ribbon function." Neuron 28(3): 857-872.

Sheng, M. and E. Kim (2000). "The Shank family of scaffold proteins." Journal of cell science 113(11): 1851-1856.

Simmons, D. and M. Liberman (1988). "Afferent innervation of outer hair cells in adult cats: I. Light microscopic analysis of fibers labeled with horseradish peroxidase." The Journal of comparative neurology 270(1): 132-144.

Spoendlin, H. (1975). "Neuroanatomical basis of cochlear coding mechanisms." International Journal of Audiology 14(5-6): 383-407.

Sterling, P. and G. Matthews (2005). "Structure and function of ribbon synapses." Trends in neurosciences 28(1): 20-29.

Tashiro, A., A. Dunaevsky, et al. (2003). "Bidirectional regulation of hippocampal mossy fiber filopodial motility by kainate receptors: a two-step model of synaptogenesis." Neuron 38(5): 773-784.

Tritsch, N. X., E. Yi, et al. (2007). "The origin of spontaneous activity in the developing auditory system." Nature 450(7166): 50-55.

Yao, I., Y. Hata, et al. (1999). "Synamon, a novel neuronal protein interacting with synapse-associated protein 90/postsynaptic density-95-synapse-associated protein." Journal of Biological Chemistry 274(39): 27463-27466.

Zhang, L. I. and M. Poo (2001). "Electrical activity and development of neural circuits." nature neuroscience 4: 1207-1214.

Electrophysiological properties of outer hair cells and the localization of Ctbp2 and Shank1