The Olfactory Receptor Associated Proteome

INTERNATIONAL GRADUATE SCHOOL OF NEUROSCIENCES (IGSN)

RUHR UNIVERSITÄT BOCHUM

THE OLFACTORY RECEPTOR ASSOCIATED PROTEOME

Doctoral Dissertation

David Jonathan Barbour

Department of Cell Physiology

Thesis advisor: Prof. Dr. Dr. Dr. Hanns Hatt

Bochum, Germany (30.12.05)

ABSTRACT

Olfactory receptors (OR) are G-protein-coupled membrane receptors (GPCRs) that comprise the largest vertebrate multigene family (~1,000 ORs in mouse and rat, ~350 in human); they are expressed individually in the sensory neurons of the nose and have also been identified in human testis and sperm. In order to gain further insight into the underlying molecular mechanisms of OR regulation, a bifurcate proteomic strategy was employed.

Firstly, the question of stimulus induced plasticity of the olfactory sensory neuron was addressed. Juvenile mice were exposed to either a pulsed or continuous application of an aldehyde odorant, octanal, for 20 days. This was followed by behavioural, electrophysiological and proteomic investigations. Both treated groups displayed peripheral desensitization to octanal as determined by electro-olfactogram recordings. This was not due to anosmia as they were on average faster than the control group in a behavioural food discovery task. To elucidate differentially regulated proteins between the control and treated mice, fluorescent Difference Gel Electrophoresis (DIGE) was used. Seven significantly up-regulated and ten significantly down-regulated gel spots were identified in the continuously treated mice; four and twenty-four significantly up- and down-regulated spots were identified for the pulsed mice, respectively. The spots were excised and proteins were identified using mass spectrometry. Several promising candidate proteins were identified including potential transcription factors, cytoskeletal proteins as well as calcium binding and odorant binding proteins. We propose that the dominant desensitizing factor in the continuously treated mice was down-regulation of odorant binding proteins. In the pulsed group no principle factor was evident, however, in terms of the number of proteins and degree of post-translational modifications, the pulsed group displayed greater plasticity.

Secondly, in order to reveal which olfactory receptors are expressed in human spermatozoa, and alternative proteomic strategy was developed. The optimised method employs an ‘affinity two phase partition’ system in conjunction with multi- enzymatic digestion in the presence of an organic solvent. The resultant peptides were identified using Multidimensional Protein Identification Technology (MudPIT). 222 integral membrane proteins were identified including 57 GPCRs, 35 of which were chemoreceptors, consisting of 32 ORs and 3 taste receptors. Notably, most of the peptides which resulted in GPCR identification were cleaved from transmembrane domains, thus demonstrating the efficacy of this strategy in membrane proteomics. In addition to the chemoreceptors, 23 neuronal proteins were also detected suggesting that both cell types may have more in common than usually perceived.

Both proteomic strategies afford a powerful means whereby novel protein candidates can be elucidated and thereby provide greater insight into plasticity of the olfactory receptor, its associated proteins, and the role of olfactory receptors in reproduction. TABLE OF CONTENTS

1 Introduction ______1 1.1 The Olfactory System ______1 1.1.1 The olfactory epithelium ______3 1.1.2 The olfactory receptor protein (OR) ______3 1.2 Project Aims______7 2 Proteomic investigation of the olfactory epithelium ______7 2.1 Olfactory stimulus induced plasticity – a background ______7 2.2 Proteomics: an introduction______11 2.2.1 Technologies for large scale proteomics______12 2.2.1.1 2D-electrophoresis ______12 2.2.1.2 Multi-dimensional protein identification technology MudPIT _____ 14 2.2.2 Neuro-proteomics______16 2.3 Conclusion ______18 2.4 Materials and Methods ______18 2.4.1 Animal and Tissue Preparation ______18 2.4.2 Behavioural study ______19 2.4.3 Microdissection of olfactory epithelium ______20 2.4.4 Electro-olfactogram (EOG) recording______20 2.4.5 Sample preparation for DIGE______21 2.4.6 Protein Labelling - Fluorescent DIGE minimal labelling ______21 2.4.7 IEF and 2-DE______22 2.4.8 Gel scanning, digitising and analysis______22 2.4.9 2D-SDS PAGE Gel total protein staining ______23 2.4.10 Trypsin digestion ______23 2.4.11 Mass Spectrometry ______24 2.4.12 Statistics ______25 2.5 Results ______25 2.5.1 Regions of interest______29 2.5.2 Pulsed Gel Analysis ______33 2.5.3 Continuous ______39 2.5.4 Protein identification ______43 2.5.5 Behavioural study ______49 2.5.6 EOG______51 2.6 Discussion ______54 2.6.1 Mice physiology______54 2.6.2 Regulation of proteins: functional categorisation & comparison______55 2.6.2.1 The cytoskeleton ______57 2.6.2.2 Intermediate early genes / transcription ______60 2.6.2.3 Calcium binding proteins ______62 2.6.2.4 Chaperones ______68 2.6.2.5 Lipocalins ______70 2.6.2.6 Xenobiotic & anti-oxidant metabolism______76 2.6.2.7 Energy metabolism ______77 2.6.3 Consolidating the findings: Protein to Phenotype ______78

2.6.3.1 Protein regulation in the continuously treated mice ______79 2.6.3.2 Protein regulation in the pulsed treated mice ______79 2.6.4 Conclusion ______80 3 Proteomic investigation of the human sperm membrane______82 3.1 A correlation between chemosenses and reproduction? ______82 3.2 The challenge of membrane proteomics ______83 3.2.1 In-gel methods ______84 3.2.2 In-Solution shotgun approach ______85 3.2.3 Fractionation______86 3.3 Sperm proteomics: a background ______87 3.4 Conclusion ______89 3.5 Materials and Methods ______90 3.5.1 LC/LC-MS/MS and Protein Identification ______90 3.5.2 Sperm preparation______91 3.5.3 Ca2+ imaging ______91 3.5.4 Immunocytochemistry______91 3.5.5 Sample optimisation - main strategies ______92 3.5.5.1 Gel Based Approach ______92 3.5.5.2 In Solution Strategy ______94 3.6 Results ______97 3.6.1 Gel Based Strategy______97 3.6.2 In Solution Strategy ______100 3.6.2.1 Solubilisation using organic acid______100 3.6.2.2 Solubilisation using organic solvent ______101 3.6.2.3 Vectorial labelling ______102 3.6.2.4 Lipase & Affinity Enrichment ______102 3.6.3 Validation of data ______107 3.6.3.1 Functional validation - calcium imaging ______108 3.6.3.2 Immunocytochemistry______109 3.7 Discussion ______110 3.7.1 Strategy development ______110 3.7.1.1 Gel Based Strategy______110 3.7.1.2 ‘In-solution’ based strategies ______113 3.7.2 The Solution: A final strategy______117 3.8 Identified membrane proteins ______118 3.8.1 A technical discussion on the identified proteins ______119 3.8.2 The identified proteins from a biological perspective ______120 4 Final Conclusion ______124 4.1.1 Olfactory receptor plasticity______124 4.1.2 Spermatozoa membrane proteome ______125 4.1.3 The olfactory receptor proteome ______125

LIST OF FIGURES

Figure 1.1 Diagramatic representation of the rodent olfactory system Error! Bookmark not defined. Figure 1.2 Diagram depicting initial olfactory receptor signal transduction ______5 Figure 2.1 Schematic representation of potential stimulus induced plasticity ______11 Figure 2.2 DIGE fluorescent CyDye minimal labelling______13 Figure 2.3 Schematic of 2D-Electrophoresis workflow ______14 Figure 2.4 Schematic of MudPIT Workflow ______15 Figure 2.5 Picture of mice odorant exposure apparatus ______19 Figure 2.6 The structure and properties of Octanal______26 Figure 2.7 Schematic representation of parameters used to calculate spot data______27 Figure 2.8 Example of two DIGE gels Upper______28 Figure 2.9 DIGE scan from 5 gels comparing the control and pulsed treated mice___ 29 Figure 2.10 Biological variation – DIGE gels ______30 Figure 2.11 Biological variation - Histogram from DIGE gels ______31 Figure 2.12 Master gels showing significantly regulated spots ______32 Figure 2.13 ‘Troubleshooting’ protein identification______44 Figure 2.14 Chart showing interaction between Treatment and Substance ______50 Figure 2.15 Frequency of found substance as function of treatment ______51 Figure 2.16 Chart showing EOG amplitudes in response Octanal ______52 Figure 2.17 Functional classification of identified proteins ______55 Figure 2.18 Alignment of identified unknown candidate protein ______72 Figure 2.19 Probability of residue phosphorylation (y-axis______73 Figure 2.20 Comparison of MUP 5 binding pheromones with octanal______75 Figure 2.21 Proposed xenobiotic clearance of octanal ______77 Figure 3.1 Organisational chart showing experimental strategies ______97 Figure 3.2 Sequence coverage of Semenogelin II and Lactotransferrin______98 Figure 3.3 Functional classification of identified 222 Integral Membrane Proteins _103 Figure 3.4 Cleaved peptides that identified GPCRs ______104 Figure 3.5 Scatter plots showing pI,MW,hydrophobicity correlation ______105 Figure 3.6 Virtual 2D gel of identified IMPs ______106 Figure 3.7 Calcium imaging of human sperm ______108 Figure 3.8 Immunocytochemistry validation of a section of neuronal proteins _____ 109 Figure 3.9 The lipid composition of human sperm ______112 Figure 3.10 Coupling of affinity ligand (WGA) to dextran ______117 Figure 3.11 Strategy to identify olfactory receptors in human sperm ______118

iii

LIST OF TABLES

Table 2.1 Most commonly encountered detection methods used in 2D-PAGE ______12 Table 2.2 Statistically significant (P<0.05) changes for pulsed experiment ______33 Table 2.3 Statistically significant (P<0.05) changes for continuous experiment _____ 39 Table 2.4 Protein identification from pulsed experiment ______45 Table 2.5 Proteins identified from continuous experiment ______47 Table 2.6 Fisher least significant difference post hoc results behavioural study _____ 50 Table 2.7 Fisher least significant difference post hoc results EOG______53 Table 2.8 List of candidate proteins ______56 Table 3.1 Olfactory proteins identified using gel based strategy ______100 Table 3.2 Integral membrane proteins identified using organic solvent strategy ____ 101 Table 3.3 Identified chemoreceptors ______107

This work was financially supported by Graduiertenkolleg for “Development and Plasticity of the Nervous System: Molecular, synaptic and cellular mechanisms”. Additional support was given by the International Max Plank Research School in Chemical Biology (IMPRS-CB).

This dissertation has been completed and written independently without external assistance. It has never been submitted in this or a similar form at this or any other domestic or foreign institution of higher learning as a dissertation. The "Guidelines for Good Scientific Practice” (Leitlinien guter wissenschaftlicher Praxis und Grundsätze für das Verfahren bei vermutetem wissenschaftlichen Fehlverhaltens) according to § 9, Sec. 3 of the Promotionsordnung der International Graduate School of Neuroscience der Ruhr-Universität Bochum were adhered to.

ACKNOWLEDGEMENTS

First and foremost I would like to thank both of my supervisors Prof Hanns Hatt and Dr Eva Neuhaus, for giving me the opportunity to work on this project. I am grateful to Prof Hatt for his ‘open door policy’ and encouragement throughout. A special thanks goes to Dr Neuhaus for her unwavering guidance and continual support.

This project would not have been possible without the advice and direction of our collaboration partners Dr Dirk Wolters, Jun Prof Bettina Warscheid and Prof Helmut Meyer. I appreciate their enthusiasm and willingness and for giving me the opportunity to receive excellent ‘hands-on’ experience in proteomics.

Due to my collaborative experience in two proteomic laboratories I had the opportunity to work with and meet many people. In particular I would like to thank Dr Andreas Kyas for his endless support; I am also very grateful to Barbara Sitek and Nadine Stoepel for their enthusiasm and stoicism in teaching me 2D-electrophoresis. In fact, I am indebted to all of the staff working at the Medical Proteome Centre 2D-electrophoresis facility for their patience and help, including Dr. Kai Stühler.

A special thanks to Mr Harry Bartel for his expertise in constructing various application systems – his skills and affability are truly an asset to the laboratory. I am indebted to Mr Thomas Lichtleitner for his readiness to help me with computing problems and especially for his advice in desk top publishing packages. I am also grateful to Weiyi Zhang, Anastasia Mashukova and Ruth Dooley, not only for their scientific help, but also for their friendship over the past three years.

Some experiments involved demanding preparations and without the help of several people would have been arduous. In particular I am appreciative to Daniela Brunert and Heike Piechura for their abetment. I am also grateful to Dr Heike Benecke for assisting with the behavioural study.

I would also like to take this opportunity to thank the staff and students of the Cell Physiology department. When I first arrived to Bochum I was immediately impressed by a welcoming and friendly attitude from everyone. I will undoubtedly have fond memories of my time here.

Last and certainly not least I thank my family for their unconditional love, patience and support.

v i

ABBREVIATIONS

1D-SDS PAGE One dimentional sodium dodecylsulfate-polyacrylamide gel electrophoresis 2D-PAGE Two-dimensional polyacrylamide gel electrophoresis ACN Acteonitrile Ambic Ammonium bicarbonate ANOVA Analysis of variance between groups CNG Cyclic-nucleotide-gated channel DIGE Difference gel electrophoresis DTT Dithiothreitol EEG Electroencephalogram EOG Electro-olfactogram ESI Electrospray ionisation GABA Gamma-aminobutyric acid GPCR G-protein coupled receptor LC-MS/MS Liquid Chromatography-Tandem Mass Spectrometry M/T Mitral/tufted cells MALDI Matrix-assisted laser desorption ionisation MudPIT Multi-dimensional protein identification technology OB Olfactory Bulb OE Olfactory epithelium OR Olfactory receptor ORN Olfactory receptor neuron OSN Olfactory sensory neuron pI Isoelectric point RMS Rostral migratory stream SDS Sodium Dodecyl Sulfate SVZ Subventricular zone TRIS Tris(hydroxymethyl) methylamine TRP Transient receptor protein VNO Vomeronasal organ

v ii

1 Introduction The chemosenses are considered to be phylogenetically the oldest sense providing an important means whereby the chemical environment can communicate with an organism. This diversity ranges from the relatively primitive orientation of a unicellular organism to a chemical gradient (chemotaxis/chemokinesis), to complex behaviour demonstrated by vertebrates. Vertebrates have developed sophisticated means to detect and neuronally process a myriad of air-borne (olfaction) and liquid-borne (taste) signals facilitating emotional (anxiety, pleasure), physiological, sociological and sexual behavioural responses. Most notably, the olfactory sense is responsible for detecting aromas and flavours not to mention the detection of chemical cues integral to survival of a species e.g. danger signals (presence of toxin or predator) and pheromones involved in reproduction. Arguably the olfactory system provides an ideal model to investigate neurobiological questions pertaining to development and plasticity because of sustained neurogenesis throughout life. In terms of plasticity, its unique ability to distinguish and respond to a plethora of odours is regulated both at the behavioural and neuronal level. This dynamic processing is dependent on previous experience, the current environment and internal state. Ultimately these changes involve protein regulation, which in terms of odorant induced signal transduction, must begin at the olfactory receptor protein. The intricate mechanisms underlying receptor plasticity such as receptor activation, internalization, trafficking and recycling are poorly understood. Therefore, in order to unravel potential new mechanisms, this investigation has adopted state of the art proteomic techniques.

1.1 The Olfactory System The anatomy of the olfactory system has been well described for many species which in principle share a basic common primary olfactory pathway (Hildebrand and Shepherd 1997). This common pathway includes the olfactory receptor neurons (ORNs) in the nose (or insect antenna) which project to second order neurons located in the olfactory bulb (or antennal lobe) which ultimately output to the olfactory cortex (or mushroom bodies) (Wilson, Best et al. 2004). In mammals this direct pathway is networked to higher-brain regions including the orbitofrontal cortex and limbic system (amygdala, hippocampus, and perirhinal cortex). Interestingly, recent evidence suggests that conscious perception in mammals may not necessarily involve the thalamus, but instead, a direct pathway between the olfactory cortex and prefrontal cortex (Shepherd 2005) - this raises new questions with regard to central olfactory processing. It has also been proposed that the circuitry of

the olfactory bulb is functionally similar to the circuitry found in the retina and in parts of the neocortex (Mori, Nagao et al. 1999; Dietz and Murthy 2005). It is possible that similar odours are mapped close together in the olfactory bulb and that granule cells create centre-surround fields similar to those in the retina (Mori, Nagao et al. 1999; Aungst, Heyward et al. 2003).

In mammals olfaction is mediated by two anatomically distinct systems-:

i) the accessory olfactory system which receives inputs from the vomeronasal organ (VNO) and encodes for social and sexual behaviour – VNO receptors bind pheromones.

ii) the main olfactory system which receives inputs from the main olfactory epithelium (recognises air-borne odorants) and projects to the olfactory bulb for central processing.

Figure 1.1 Diagramatic representation of the rodent olfactory system Taken from Lledo (2005) et al. The main olfactory system is in green; the accessory in red

1.1.1 The olfactory epithelium The olfactory epithelium (OE) is situated dorsocaudal to the respiratory epithelium in the roof of the two nasal cavities of the nose. In adult mice it measures around 170mm2 and comprises of three principle cell types - sensory neurons, sustentacular cells and basal cells. Adjacent sustentacular cells provide metabolic and structural support to ORNs including the detoxification of xenobiotic compounds. Basal cells, on the other hand, are considered to be progenitor (stem) cells which continually regenerate the OE, in fact ORNs are continually regenerated throughout the murine lifespan (Beites, Kawauchi et al. 2005). The epithelium is kept moist by the secretions of olfactory glands which also include odorant binding proteins. These bind to hydrophobic odorants and ‘present’ them to the olfactory receptors.

1.1.2 The olfactory receptor protein (OR) In a landmark paper (Buck and Axel 1991) the olfactory receptor proteins were identified as a superfamily of G-protein coupled receptors (GPCRs). According to the latest genomic analysis it was estimated that between 1,000-1,300 ORs are expressed in mice (Zhang, Rodriguez et al. 2004) with up to 425 in humans (Armbruster and Roth 2005). They have conserved motifs consisting of around 5-10 amino acids and transmembrane (TM) variable regions which may function as ligand- binding sites. Recent molecular modeling of a mouse olfactory receptor (mOR-EG) revealed that the most critical residues involved in ligand binding are hydrophobic amino acids which form a binding pocket in the space formed by TM3, TM5 and TM6 (Katada, Hirokawa et al. 2005). Previous molecular modeling of the rat I7 olfactory receptor also predicted a similar ligand binding site and in particular, identified the residue lysine 164 as having the highest affinity for octanal (Singer 2000). It was proposed that Lysine 164 (TM4) forms an electrostatic interaction with the carbonyl oxygen of octanal.

Most vertebrate investigations have examined rodent olfaction which over time has lead to the establishment of several dogmas-:

1. One receptor one-neuron hypothesis – from an battery of independent investigations it is generally accepted that each olfactory sensory neuron expresses only one olfactory receptor (Mombaerts 2004). Recently this dogma has been challenged by convincing work which showed that in Drosophila, two functional odorant receptors are co-expressed in one

neuron (Goldman, Van der Goes van Naters et al. 2005). Nevertheless, in mice this hypothesis is still persuasive.

2. Zonal organization of olfactory receptors - incontestable evidence demonstrated that ORs are organised into four zones (Mori, von Campenhaus et al. 2000) which also converge into specific subsets of glomeruli (Treloar, Feinstein et al. 2002).

3. Adult neurogensis of olfactory sensory neurons – classical experiments have demonstrated that the principle regions of constitutive adult mammalian neurogenesis are the subventricular zone/olfactory bulb and hippocampal dentate gyrus (Emsley, Mitchell et al. 2005). The olfactory epithelium neural stem cell (basal cell) produces two main lineages responsible for regeneration of sustentacular cells and olfactory ensheathing cells (Beites, Kawauchi et al. 2005).

4. In the main olfactory epithelium, OR signal transducion is mediated by Golf/AC/cAMP – the signal transduction mechanism of the ORs is schematically represented below (Figure 1.2). In brief, odorant binding with the olfactory receptor leads to GPCR activation and

dissociation of the Gαolf sub unit from βγ (Menini 1999). Gαolf in turn stimulates a type III adenylate cyclase (AC) which catalyses the conversion of ATP into cAMP. The elevation in cAMP opens a cyclic-nucleotide-gated channel (cAMP increases the channel ‘open probability’), leading to an influx of Na2+ and Ca2+ and a membrane depolarisation. This depolarisation is amplified through the opening of calcium activated chloride channels leading to a Cl- efflux. At the same time the Ca2+ influx decreases cAMP synthesis and the effective affinity for CNG-cAMP, producing channel adaption (Bhandawat, Reisert et al. 2005). Ultimately, this Ca2+ influx in combination with receptor phosphorylation and hydrolysis of cAMP into 5’-AMP, leads to a ‘reset’ of the signal transduction cascade (Breer 2003). It is worth noting that in the vomeronasal organ (VNO), neurons appear to use the PtdIns(4,5)P2/phospholipase C/Ins(1,4,5)P3 pathway for signal transduction (Lucas,

Ukhanov et al. 2003). The classical olfactory transduction elements (Golf, ACIII and CNG-

channels) are replaced with Go- Gi2- subunits and gating is facilitated by a TRP (transient receptor) channel (Breer 2003).

Figure 1.2 Diagram depicting initial olfactory receptor signal transduction

5. Olfactory receptors are only functionally expressed in the Olfactory Epithelium – after the identification of the olfactory receptome it was generally accepted that expression of ORs is restricted to the OE. This dogma was later disproved, now evidence exists that ORs are expressed in additional tissues including prostate, skin and spermatozoa. Parmentier suggested that ORs are expressed in mammalian germ cells (Parmentier, Libert et al. 1992) and it was purported that they may play a chemosensory role in sperm development, chemotaxis or in oocyte/sperm interaction. At least one OR protein was shown to be localized to the flagellum of mature dog sperm (Vanderhaeghen, Schurmans et al. 1993). Human sperm have the ability to detect and respond to chemotactic signals secreted by the egg through the odorant receptor hOR 17-4 (Spehr, Gisselmann et al. 2003). Several other components

of the odorant signalling cascade of olfactory neurons are present in sperm which include cyclic nucleotide-gated (CNG) channels (Vanderhaeghen, Schurmans et al. 1993; Weyand, Godde et al. 1994; Walensky, Roskams et al. 1995) suggesting analogous transduction mechanisms.

Over the last few years there have been significant advancements in understanding olfactory processes, nevertheless, in spite of this, there is still a salient lack in understanding some fundamental mechanisms. These include-:

• How olfactory receptor expression is controlled at the ‘zonal level’ in the OE.

• The mechanism of olfactory sensory neurons (OSN) convergence to glomeruli.

• The mechanism of olfactory receptor trafficking.

• How odours are encoded and cogitatively perceived.

• Plasticity at a cellular level – activity dependend modulation of OSN thresholds.

In order to identify novel proteins and potential protein-protein interactions which could shed new light on OR regulation, we devised a proteomic strategy. The proteomic investigation of the olfactory receptor per se considered two aspects. Firstly, identifying novel candidate proteins which are either up- or down-regulated in the olfactory epithelium. These changes are in response to odorant activation of the olfactory receptor and therefore reflect plasticity of the olfactory sensory neuron. Secondly, we paralleled this with an investigation into identifying novel candidates in human sperm. One might expect to find a population of olfactory receptors in human sperm which may be modulated by similar mechanisms as in olfactory sensory neurons. Moreover, findings from these parallel investigations could not only help to understand mechanisms in neuroscience, but also may be applied to reproductive medicine.

Both strategies require independent method development and independent experimental design, thus, both approaches are discussed in separate chapters. This dichotomy converges in the ‘final

conclusion section’ (chapter 4) which gives an overview of both strategies and highlights principle findings.

1.2 Project Aims The project aim was the analysis of the olfactory receptor associated proteome in order to elucidate novel protein(s) which may play a role in OR regulation in response to odorant stimulation. This goal was realised by addressing two main questions:

1) What changes occur to protein regulation in the olfactory receptor neuron as a consequence of receptor stimulation? Stimulus induced plasticity of the OSN analysing long term affects on protein regulation. Refer to chapter 2.

2) Are olfactory receptors expressed in mature human sperm? Can a parallel be identified between sperm and olfactory chemoreceptors? Characterisation of the human sperm membrane proteome with a particular emphasis on chemoreceptors. Refer to chapter 3.

2 Proteomic investigation of the olfactory epithelium 2.1 Olfactory stimulus induced plasticity – a background Investigations into olfactory plasticity have usually focused on the olfactory bulb and higher cortical regions with less of an emphasis being placed on the periphery. Moreover, the vast majority of these studies were electrophysiological. It is now generally accepted that the characteristic electrical oscillations observed in the olfactory system are gamma (30–80 Hz), beta (15–40 Hz) and theta rhythms (3–12 Hz), with the former two frequencies being associated with odour stimuli (Lagier, Carleton et al. 2004; Lledo, Gheusi et al. 2005). These oscillations are known to take part in encoding of sensory information before their transfer to higher subcortical and cortical areas. For example, there is a clear correlation between odorant induced spatial-temporal formatting in the bulb to a variety of behavioural states, such as memory or perception (Laurent, Stopfer et al. 2001; Laurent 2002).

Doving et al (Doving and Pinching 1973) were one of the first to report stimulus induced plasticity in vertebrates (rats) and revealed that a long exposure (several months) to an odorant led to specific

patterns of mitral cell degeneration in the OB of adult rats. Later work (Vanderwolf and Zibrowski 2001) demonstrated that repeated ‘short term’ exposure to odour caused a gradual sensitisation of olfactory ß-waves in the pyriform cortex. This phenomenon occurred whilst the mucosal receptor potential was adapting (decreasing). More recently, an even shorter exposure regime (<8hrs) caused changes in the surface markers of olfactory receptor neurones (Yoshihara, Katoh et al. 1993) and expression of immediate-early genes in cells in the olfactory bulb (Sallaz and Jourdan 1993; Guthrie and Gall 1995; Baba, Ikeda et al. 1997). Buonviso et al (Buonviso and Chaput 1999) administered isoamyl acetate for 20 min pulses to rats for six consecutive days. This resulted in a decrease in responsiveness of the olfactory bulb, not only to the familiar odorant (demonstrating habituation or adaptation), but also to unfamiliar odorants. This effect was observed for up to 10 days after exposure and was attributed to an elevation in the inhibition of mitral/tufted (M/T) cell responses by periglomerular cells. Buonviso postulated that this inhibition was mediated by dopamine, or by GABA from granular cells (Buonviso and Chaput 1999). Inhibition was not only effected by different odorants, but also to a greater extent by the odorant concentration. This odorant mediated M/T cell inhibition was also observed in rat pups and was correlated to positive reinforcement in an associative learning task (Wilson and Leon 1988). Similar evidence exists in rabbits (Skarda and Freeman 1987) and, as with the Wilson study, electrophysiological recordings were associated with a learning task. We can therefore conclude that mitral-tufted cells have odorant receptive fields which are dynamic and can be modulated by odour experience (sensitisation and habitation). Arguably the dynamic nature of MT cells could directly affect OB spatiotemporal dynamics as well as cortical odour processing (Fletcher and Wilson 2003).

Several investigations have also examined the effect of odorant stimuli on ontogenetic development. Schmidt (Schmidt and Eckert 1988) compared the effect of the odorant geraniol in three groups of mice at varying stages in development. All datasets were referenced to an unexposed control. The groups were comprised of different age ranges in development, namely; Group 0-13 (postnatal day 0 to postnatal day 13 i.e. p0-p13); Group 0-6 (p0-p6) and Group 6-12 (p6-p12) – these groups received a continuous application of geraniol during the indicated periods. EEG recordings from the OB showed that Groups 0-13 and 6-12 demonstrated a significant response to geraniol whereas the control and Group 0-6 did not. Group 0-13 mice were also recorded at p70 (adults), the marked response was still evident suggesting that this period in murine olfactory development is plastic and can facilitate long lasting changes. Moreover, exposure and

deprivation of odours to rat pups led to a respective increase, (Rosselli-Austin and Williams 1990) and decrease, (Frazier and Brunjes 1988) in mitral and granule cell layer neurons. In fact, the spine density of granule cells was shown to be maximal at postnatal day 21 in rat pups. Continuous exposure to odourless air led to a reduction in spine density on the medial and lateral sides of the bulb whereas exposure to a single odorant, cyclohexanone, only reduced spine density on the lateral side (Rehn, Panhuber et al. 1988). Rather surprisingly, the chronic exposure of a single odorant for two months caused extensive cell shrinkage in the adult rat olfactory bulb when compared to the normal control. The shrinkage in mitral cells was even more pronounced than in rats exposed to an odourless environment. This suggests that an absence of stimulus is not as damaging as a single dominant odour (Panhuber, Mackay-Sim et al. 1987). Similar findings showed that continuous exposure of the odourant geraniol accelerated normal granule cell spine loss in the developing ferret olfactory bulb (Apfelbach and Weiler 1985). On the contrary, it was also shown that if adult mice (two months old) were exposed to an odorant enriched environment, twice as many newborn neurons survived in the main olfactory bulb. This was accompanied with an improved olfactory memory, however, it was not associated with an increased proliferation in the rostral migratory stream (RMS) and subventricular zone (SVZ) (Rochefort, Gheusi et al. 2002).

It was also shown, that repeated exposure to odorants increased peripheral olfactory sensitivity in mice (Wang, Wysocki et al. 1993) showing stimulus-induced plasticity in the receptor cell. Wang et al used inbred strains of NZB/B1NJ and CBA/J mice which are anosmic for, and sensitive to, androstenone, respectively. Shifts in sensitivity were electro-physiologically quantified by measuring field potentials from the olfactory epithelium – the electro-olfactogram (EOG) (Knecht and Hummel 2004). The male, adult mice were exposed to either androstenone or isoamyl acetate for 16 hours, daily for 2-6 weeks. EOG recordings from androstenone stimulated NZB/B1NJ mice demonstrated a 2.2-fold increase in amplitude, relative to the control. This sensitisation was not observed when the OE was stimulated with isoamyl acetate and was not demonstrated in the CBA/J androstenone stimulated mice i.e. they did not demonstrate increased sensitivity. It remains to be shown whether this effect is an androstenone specific phenomenon, nevertheless, this plasticity may be related to the continuous turnover of neurons, or alternatively it may be a form of stimulus-controlled gene expression. It is now known that the mechanisms of heptahelical receptor signalling are more diverse than previously thought, and that receptor stimulation can have various physiological consequences, like cytoskeletal rearrangements and gene expression (Lefkowitz 2000;

Hoyer, Hannon et al. 2002). These effects do not seem to be mediated by G-protein activation, but require the association of the receptors with a variety of other intracellular partners like arrestins, GRKs, SH2 proteins, small GTP-binding proteins, and PDZ domain–containing proteins (McDonald and Lefkowitz 2001; Rebois and Hebert 2003). Moreover, it is becoming increasingly clear that these proteins regulate the localization of the receptor and other key proteins in different subcellular compartments. Organization of GPCRs in supramolecular complexes was extensively studied using photoreceptor cells of the fruitfly (Harris and Lim 2001). The multivalent PDZ domain scaffolding protein INAD is required here to ensure the sensitivity and temporal resolution of the signal transduction process, most likely by coordinating the rhodopsin and other signaling molecules such as phospholipase Cb, protein kinase C, and the TRP ion channel into a multimeric complex (Harris and Lim 2001). In the last years, interactions with PDZ domain-containing proteins turned out to represent a generalized mechanism for modulating GPCR-mediated signal transduction (Zhong, Wade et al. 2003). Considering the high sensitivity and temporal resolution in vertebrate olfaction, one might expect to find a similar arrangement of olfactory receptors in supramolecular complexes in the sensory cilia of olfactory sensory neurons.

More recently there has been an increased interest in another family of scaffolding proteins, the arrestins. Two family members, β-arrestin 1 and β-arrestin 2, are ubiquitously expressed and are responsible for GPCR densensitization and internalization (Pouyssegur 2000). Arrestin binds to phosphorylated uncoupled GPCRs (βγ subunits) facilitating receptor internalization and interaction with mitogen-activated protein kinases (MAPKs). MAP kinases operate in modules performing a mini-cascade of sequential phosphorylations - MAP kinase kinase kinase (MKKK) activates MAP kinase kinase (MKK), which then activates MAP kinase which in turn has a myriad of intracellular effects including activation of c-Jun N-terminal kinase 3 (JNK3). JNK3 directly translocates to the nucleus instigating transcription (McDonald, Chow et al. 2000). Until very recently arrestin was considered to be exclusively located in the cytosol. However, exciting work has also demonstrated that β arrestin can directly translocate to the nucleus upon ligand stimulation of odorant receptors (Neuhaus et al, submitted). Nuclear translocation evidently results in histone H4 acetylation and gene transcription (Jiuhong Kang 2005).

To summarise, there have been previous reports which have investigated stimulus induced plasticity, however, these were mainly electrophysiological studies which focused on the olfactory

bulb. To date, there are relatively few investigations concerned with odorant induced plasticity in the olfactory epithelium and no proteomic reports. Therefore in order to address the question of differential protein regulation induced by OR activation, we applied a recent proteomic technology, fluorescent difference gel 2D-electrophoresis.

Figure 2.1 Schematic representation of potential stimulus induced plasticity of the olfactory sensory neuron What potential changes occur to protein regulation as a consequence of receptor stimulation? This question was addressed using a proteomic strategy.

2.2 Proteomics: an introduction The term “proteome” was initially coined by Wilkins who defined it as “…the entire PROTEin complement expressed by a genOME, a cell or tissue type, at a specific time in the development of the organism under specific physiological conditions” (Wilkins, Sanchez et al. 1996). Large-scale proteomics technologies have been developed to generate comprehensive, cellular protein-protein interaction maps, although “functional proteomics” in neuroscience is a relatively novel idea. Until relatively recently large scale proteomics was somewhat neglected in favour of using mass spectrometry (MS) to identify less complex samples. Interaction partners of the N-methyl-D-

aspartic acid (NMDA) receptor (Husi, Ward et al. 2000), of the P2X7 ATP receptor (Kim, Jiang et al. 2001), of the 5-HT2C receptor (Becamel, Alonso et al. 2002) and more recently, of mGluR5 protein complexes (Farr, Gafken et al. 2004) have been reported. The combination of classical biochemical techniques (e.g. pull-down, co-immunoprecipitation, cross-linking) with MS is clearly powerful, however, to look for changes in protein regulation in the olfactory epithelium clearly a large scale technology is required.

2.2.1 Technologies for large scale proteomics 2.2.1.1 2D-electrophoresis Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) was first described in 1975 and is still the ‘work horse’ in proteomics (Klose 1975; O'Farrell 1975). 2D-PAGE is a reproducible, reliable technique which produces high resolution protein separations. Proteins are separated in the first dimension according to protein charge as determined by a pH gradient (isoelectric point (pI)), and in the second dimension by molecular weight (SDS-PAGE). The technique has been refined (Klose and Kobalz 1995) and can now resolve up to 10,000 spots on one gel. Unfortunately, there are several criticisms of conventional 2D-PAGE. These include sensitivity, the linear dynamic range, reproducibility and accuracy of visualization methods. Moreover, it is well recognised that hydrophobic proteins, highly basic or acidic proteins cannot be easily identified using this technology.

Table 2.1 Comparison of the most commonly encountered detection methods used in 2D-PAGE NB: The silver stain provides only qualitative information about the proteins due to its narrow 10-fold linear dynamic range, while the fluorescent stain provides quantitative capabilities extending over 3-logs.

Stain Sensitivity Linear Range Coomassie Brilliant Blue G-250 30-100 ng ~10-30 X Colloidal Coomassie Brilliant Blue G-250 8-10 ng ~10-30 X Silver ~1 ng ~10 X SYPRO (fluorescent stain) ~3 ng ~103 X

Problems with reproducibly can be minimised with good standard operational procedures and handling technique, nevertheless, using conventional stains a large dataset is required. This is labour intensive and can also introduce inter-gel variability. Many of these limitations have been

relatively recently overcome with the emergence of a new fluorescent labelling technology, namely, DIGE - Difference gel electrophoresis (Marouga, David et al. 2005). This propriety technology (Amersham Biosciences) employs three fluorescent succinimidyl esters to specifically label proteins. These “CyDyes” (Cy2, Cy3 and Cy5) are spectrally resolvable and can therefore be mixed and run on one gel. Any differences between protein samples are directly reflected in the different ratios of the fluorescent signal. It is recommended that this method is used to compare two different biological samples, for example, treated versus control. The third dye is used to label an internal standard which allows accurate inter-gel matching. Protein solubility is maintained as this is a ‘minimal labelling’ technique with only ~1-2% of the lysine residues of the proteins being fluorescently modified. Furthermore, the singly charged CyDye ‘replaces’ the single lysine charge via a covalent NHS-ester linkage.

Figure 2.2 Fluorescent DIGE CyDye minimal labelling The CyDye minimal dyes have an NHS-ester reactive group and covalently couple to the ε-amino group of lysine of proteins via an amide linkage.

In addition, this technique is compatible with mass spectrometry, the penultimate step in protein identification. Either matrix-assisted laser desorption ionisation (MALDI) (Karas and Hillenkamp

1988) and/or electrospray ionisation (ESI) (Whitehouse, Dreyer et al. 1985; Fenn, Mann et al. 1989) mass spectrometry may be used to ‘fingerprint’ or sequence the protein. The resultant spectra are integrated into an algorithm which searches a protein database e.g. MASCOT or SEQUEST. Before these ‘soft-ionization’ techniques were reported, the only reasonable alternative was Edman sequencing which, due to the time and cost, was prohibitive. Proteins can now be reliably resolved using 2D-electrophoresis and subsequently identified using ESI and/or MALDI.

Figure 2.3 Schematic of 2D-Electrophoresis workflow

2.2.1.2 Multi-dimensional protein identification technology MudPIT An interesting technique for the analysis of complex protein mixtures is 'shotgun proteomics'. This approach has been facilitated by the use of multidimensional protein identification technology (MudPIT), which incorporates multidimensional high-pressure liquid chromatography (LC/LC), tandem mass spectrometry (MS/MS) with database-searching algorithms (Washburn, Wolters et al. 2001) and facilitates the concomitant identification of soluble and membrane associated proteins (Wu, MacCoss et al. 2003). In MudPIT a complex protein mix (e.g. cell lysate) is digested ‘in-

solution’ and separated by multidimensional capillary chromatography which is connected on-line to an ion trap mass spectrometer. The term ‘multidimensional’ is attributed to the fact that there are two dimensions in the chromatography step, however, rather than using conventional 2D-gel electrophoresis, MudPIT separates peptides in 2D-liquid chromatography. Thus, the column can be directly interfaced to a mass spectrometer for electrospray ionisation tandem MS i.e. LC/LC- ESI-MS/MS. The first dimension consists of strong cation exchange (SCX) material. This is packed ‘back-to-back’ with the second dimension, a reversed phase (RP) material inside fused silica capillaries. Peptides are therefore resolved on the column according to their charge (SCX) and hydrophobicity (RP).

Despite the power of this technique it does have limitations. There are some criticisms over the scalability and reproducibility of MudPIT (Graumann, Dunipace et al. 2004). In addition, concerning the quantification of subtle changes in protein regulation and/or in post-translational modifications, undoubtedly LC-MS techniques lag behind gel based techniques.

Figure 2.4 Schematic of MudPIT Workflow

2.2.2 Neuro-proteomics The proteomic methods described above can be easily applied to global profiling of many samples including neurological probes. Already early proteomic studies have attempted to map protein expression in various brain regions including the adult cerebellum from mouse (Beranova- Giorgianni, Pabst et al. 2002), rat (Taoka, Wakamiya et al. 2000) and pig (Friso and Wikstrom 1999; Morrison, Kinoshita et al. 2002). An international concerted effort (HUPO Brain Proteome Project) is now underway to differentially analyse normal and neurodegenerative pathologies, with a particular emphasis on Morbus Alzheimer, Morbus Parkinson and senescence (Michael Hamacher 2005). Several studies have used human tissue, a recent example is a comparative study of the prefrontal white matter of human alcoholics referenced to normal subjects (Alexander-Kaufman, James et al. 2005). 60 statistically different (p<0.05, ANOVA) proteins were identified, several of which were novel findings. The vast majority of neuro-proteomic studies have opted to use samples surgically removed from individual subjects. An additional approach is the use of cultures which is gaining in popularity. Recently Sitek et al stably transfected a neuroblastoma cell line (SY5Y) with TrkA or TrkB receptors (Sitek, Apostolov et al. 2005). The cells were stimulated with their respective ligands (TrkA-NGF and TrkB-BDNF) and using fluorescence two-dimensional difference gel electrophoresis (DIGE), differences in protein regulation were quantified over a variety of post-stimulation time points. Employing the same technology, another elegant study demonstrated the importance of the cytosolic protein stathmin in the migration of newborn neurons in the adult brain (Jin, Mao et al. 2004). In this case, a primary neuronal culture from p15 CD1 mice was established. ‘Immature’ neuronal cultures (cells harvested 5-6 days after plating) were compared with the ‘mature’ population (four passage cycles) – stathmin was up-regulated 7.6- fold in immature cultures. This evidence was corroborated by western blots and immunocytochemisty. Confocal images of sagital sections from neuroproliferative zones of normal adult rat brain exhibited prominent stathmin staining. Moreover, stathmin colocalized with the immature neuronal marker DCX in recently divided cells (BrdU positive). Conversely, stathmin did not colocalize with the astroglial marker GFAP, the neuropeithelial cell marker nestin, or the mature neuronal marker NeuN. Knockdown of stathmin expression with an antisense oligodeoxynucleotide (ODN) inhibited the migration of DiI-labeled cells from subventricular zone (SVZ), via the rostral migratory stream (RMS), to the olfactory bulb (OB). This comprehensive

study clearly exemplified the power of fluorescent DIGE in identifying candidate proteins (Jin, Mao et al. 2004).

Only one differential proteomic study has been reported with regard to the olfactory system. Recently, Poon examined the differential expression of proteins in the aging murine olfactory system (Poon, Vaishnav et al. 2005). Olfactory epithelium and olfactory bulb from C57BL/6 ‘young’ mice (6-week-old) were compared to OE and OB from ‘old’ mice (80-week-old) using conventional 2D electrophoresis. In the OE and OB 9 and 20 significantly regulated proteins were identified, respectively. These cytosolic proteins were categorised into three functional groups, namely, metabolic (primarily mitochondrial metabolism); transport/motility; and stress response. The alterations in the protein profile of the aging olfactory epithelium and olfactory bulb are specific to pathways involving mitochondrial and cytosolic metabolism, intracellular transport and altered response to stress. Some of these mitochondrial protein changes were already reported in microarray analysis (Kopsidas, Kovalenko et al. 2000) reinforcing the argument that oxidative stress plays an integral role in aging. This study did identify several candidate proteins; however, most of these findings were not novel with regard to senescence. Moreover, conventional 2D lacks the sensitivity and dynamic range of fluorescent labelling and therefore subtle changes in lower abundant proteins could not be identified by this methodology. Tun-Tzu Yu, on the other hand, used quantitative reverse transcriptase-PCR in order to identify differentially expressed transcripts in OSNs (Tun-Tzu Yu 2005). OSNs were purified from juvenile mice using OMP-LacZ3 transgenic mice - OSNs specifically express β-galactosidase due to LacZ being ‘under’ the promoter of olfactory marker protein – (OMP). The FACS (fluorescence-activated cell sorting) purified OSNs were compared to neighbouring cells in the OE. Of the differentially expressed transcription products, 33% had no known function, the remaining encoded for proteins involved in signalling, xenobiotic metabolism, protein metabolism and vesicle transportation. This study identified more interesting candidates than the Poon investigation, however, the paradigm of exclusively characterising proteins using their sequences is inadequate. To exemplify, with only approx. 25,000 genes in the human genome, the proteome is clearly much more complicated. Variable gene expression may result from mRNA alternative splicing, frame-shifting and the use of alternate translation start or stop sites. Proteins may undergo an array of modifications including phosphorylation, glycosylation, acetylation, nitrosation, farnesylation and sulfation. Up to 300 different post-translational modifications have been reported (Morrison, Kinoshita et al. 2002).

Such modifications will affect protein conformation and functionality. Furthermore, it is generally accepted that transcription does not necessary directly correlate to protein expression and protein functionality.

2.3 Conclusion The aim of this project is the analysis of stimulus induced plasticity of the olfactory epithelium (OE). To address this question, mice were exposed to either a continuous or pulsed application of an odorant, octanal, for 20 days. Juvenile mice were selected for this study (p0-p30) as this stage in olfactory development is sensitive to odorant triggered plasticity. These changes may even last into adulthood. In order to elucidate changes in protein regulation which are due to the treatment conditions, a proteomic technique compared the OE from treated mice to a sham treated control. We used fluorescent difference gel electrophoresis (DIGE) which utilises spectrally resolvable fluorescent dyes to differentially label proteins from the OE of treated and control mice. Significantly up- or down-regulated proteins can be subsequently identified using mass spectrometry.

2.4 Materials and Methods 2.4.1 Animal and Tissue Preparation All procedures were in accordance with German animal welfare act (1998). 15 timed pregnant C57BL/6 mice were purchased from Charles River, Germany. All animals were held in standard rat cages at an animal facility at room temperature (22oC) and maintained on 12-hour light/dark cycle. Water and food was given ad libitum. The dams were separated into three groups of five mice per cage – Continuous Group A, Pulsed Group B and Control Group C. Each cage was housed in a customised hood designed to deliver an isolated air stream, the ‘exhaust’ air was extracted to a fume cupboard. A picture of the apparatus follows (Figure 2.5).

The apparatus was designed in order to ensure that the vapour phase concentration of the odorant remained consistent throughout the experiment. All tubes were silicon coated to minimise odorant adsorbance and the hood was constructed from Perspex glass (Biology Werkstatt, RUB). The day of pup delivery was considered P0. The odorant regime commenced at P10 and was terminated at

P31 – the aldehyde Octanal (Sigma-Aldrich) was exclusively used at a 1:10,000 dilution in ddH20. Freshly made odorant solution was prepared every morning. For Group A (Continuous): Octanal

was applied for 24 hours/day for the entire duration of the experimental period. For Group B (Pulsed): Octanal was applied four times daily at the following time periods - 09:00-09:15; 12:00- 12:15; 15:00-15:15; 18:00-18:15. Group C (Control): were sham exposed to water for the entire duration of the experimental period. It was essential that handling was minimised to reduce stress, however, when absolutely necessary (e.g. new bedding required) each group was handled to the same extent. At P31 pups were taken for behavioural analysis and EOG recordings (see below). At P31 the total mean body weights for each group was as follows: pulsed 17.25g; continuous 16.18g; control 17.76g.

Figure 2.5 Picture of mice odorant exposure apparatus

2.4.2 Behavioural study P31 continuously (n=38), pulsed (n=33) and control (n=30) mice were starved for 8 hours and subjected to an exploration test as previously reported (Yamada, Wada et al. 2001; Dawson, Steane et al. 2005), with minor modifications. One control mouse was removed and placed into a test arena (rat cage containing 3 cm freshly laid bedding; Goldspan) which contained either a hidden 1.5ml vial of cheese, (stilton blue cheese, Globus, Germany) or octanal (1:500; Sigma). The vial was randomly placed in one of the four cage corners at a depth of 2 cm and the time to find the hidden food/octanal was measured. If the exploratory time exceeded 600 sec, the mouse was removed

and ‘failure to find’ was recorded as “600 sec” for that trial. Each control mouse (male and female) was tested for octanal and cheese; 50% of the control group (n=15) were first tested with octanal and after a short interval, cheese. The remaining 50% were first tested with cheese followed by octanal - this design eliminated any statistical bias due to the test sequence. The trials were repeated for all mice in each of the treatment groups (continuous and pulsed). The bedding was renewed ad hoc to ensure that ‘distracting’ odours such as urine and faeces were eliminated.

2.4.3 Microdissection of olfactory epithelium At P32 all pups were sacrificed by cervical dislocation followed by decapitation. The OE was dissected as previously described (Spehr, Wetzel et al. 2002). Briefly, the septal bone with the intact olfactory epithelium was dissected from the head and either used intact for EOG recording, or the epithelium was dissected from the septal bone and directly placed into 100μl ice-cooled dissection buffer in a 1.5ml eppendorf. The buffer consisted of ringers [NaCl 140 mM, KCl 5 mM, MgCl2 1 mM, CaCl2 2 mM, 10 mM Hepes, 10 mM glucose] with Roche Complete® protease inhibitor cocktail and phosphatase inhibitors [β-Glycerophosphate 10 mM, Na-orthovanadate 1 mM, NaF 9.5 mM, Na-pyrophosphate 10 mM – Sigma Aldrich]. After the sample weight was recorded the OE was snap frozen in liquid nitrogen and stored at -80oC until preparation for 2D-electrophoresis.

2.4.4 Electro-olfactogram (EOG) recording For EOG recording, the exposed nasal septum and its associated epithelium was placed in a recording chamber continually perfused with oxygenated mammalian ringer (95% O2/5% CO2) (in mM) 120 NaCl, 25 NaHCO3, 5 KCl, 5 BES, 1 MgSO4, 1 CaCl2, and 10 mM glucose. Field potentials were recorded using glass pipettes (1 MΩ; tips filled with extracellular solution in 1% agar) which was placed on the airface of the epithelium. Electrophysiological signals were transduced by a unipolar silver /silver chloride (Ag/AgCl) electrode ‘housed’ in the glass pipette; the DC output was differentially amplified (P18C, Grass Instrument Co.) and referenced to an identical electrode (Ag/AgCl) grounding the bath. Odorants were delivered in the vapour phase as described previously (Spehr, Wetzel et al. 2002). A 1:10,000 dilution of octanal was prepared in ringers and soaked on a filter paper which was inserted into a motor-controlled perfusion syringe (Braun Melsungen). The odorant was injected into a constant stream of humidified air that continuously superfused the epithelium. Signals were recorded on a chart recorder and their peak amplitude was measured manually. The peak amplitudes from 1:20,000, 1:2,000 and 1:200 octanal

dilutions were normalised to a final application of 1:100 Henkel 100 (complex mixture of 100 odorants).

2.4.5 Sample preparation for DIGE

The OE from one mouse (single sample) was ground using a pestle and mortar under l N2 with sample buffer (30mM Tris Base, 2M Thiourea, 7M Urea, 4% CHAPS (w/v) [Biorad, Germany] – pH8.5). The amount of lysis buffer was critical –1.43 times (w/v) the sample weight. The mixture was thoroughly ground, the frozen power transferred to a 1.5mL eppendorf tube and sonicated in an ice-cooled sonication bath for 10 seconds. The sample was immediately placed in ice for another 10 seconds before repeating this procedure five more times. The sample was allowed to stand on ice for 2 minutes before centrifugation at 16,000 X g; 15min in a 4oC cooled microcentrifuge (Centrifuge 5415R, Eppendorf, Hamburg, Germany). After standing on ice for 1 min the supernatant was removed and stored on ice for protein concentration. The sediment was resuspended in X 1 (w/v) lysis buffer (i.e. equal to the weight of the original OE sample) and was deposited on the lN2 cooled mortar. The homogenisation procedure was repeated as before. The protein concentration from both aliquots was determined using the Biorad Bradford assay according to the manufactures instructions (Biorad, Munich, Germany). If a significant concentration (>15μg/μl) of protein was measured in the second ‘sediment’ supernatant the samples were pooled, otherwise, it was discarded. The typical total protein yield was between 20-35 μg/ul in the first preparation, the second preparation was usually <10 μg/ul. The samples were stored at -80oC until fluorescent DIGE labelling.

2.4.6 Protein Labelling - Fluorescent DIGE minimal labelling The samples were minimally labelled according to the manufactures instructions (Amersham Biosciences, Freibrug, Germany). Each sample, which represents the cell lysate from one mouse, was individually labelled. For statistical significance the total sample size was 5 (n=5). Briefly, stock cyanine dyes (1 nmol/μl) were diluted in freshly prepared anhydrous dimethyl formamide (DMF) p.a. to 400 pmol/μl (Sigma, St. Louis, MO), and 8 pmol dye was added per μg protein in the cell lysate. The sample was vortexed, centrifuged briefly, and left on ice for 30 min in the dark. No primary amines, DTT or carrier ampholytes were included in the lysis buffer as such components could potentially react with the NHS esters of the cyanine dyes. The labelling reaction was quenched by adding 1μl per 400 pmol dye of 10 mM L-lysine (Sigma). The treated (pulsed or

continuous) and untreated (control) samples were labelled with Cy5 and Cy3, respectively; the internal reference (mixture of control and treated) was Cy2 labelled. After further vortexing and centrifugation the sample was left on ice for 10mins in the dark. Thereafter, the Cy5, Cy3 and Cy2 labelled samples were mixed 1:1:1 (equating to 150μg total protein) and prepared for iso-electric focussing. To summarise, one CyDye mix comprised of: i) 50 μg treated OE from one mouse (pulsed or continuous) – Cy5; ii) 50 μg untreated control OE from one mouse – Cy3; iii) 50 μg OE from a pool of ALL sample (internal reference) – Cy2. Comparisons were made between continuous versus control and pulsed versus control. For each comparison n=5 i.e. a total of 10 gels were run.

2.4.7 IEF and 2-DE To the mixed labelled samples 10% (v/v) DTT (1,08 g/ml; BioRad) and 10% (v/v) Ampholine 2-4 (Amersham Bioscience) was added. IEF was performed using tube gels (20 cm X 0.9 mm) containing carrier ampholytes (CA-IEF) and applying a voltage gradient in an IEF-chamber produced in-house (Klose and Kobalz 1995). After IEF, the ejected tube gels were incubated in equilibration buffer (125 mM Tris, 40% (w/v) glycerol, 3% (w/v) SDS, 65 mM DTT, pH 6.8) for 10 min, after which they were washed threefold using SDS-PAGE running buffer (25 mM Tris, 192 mM Glycine, 0.2% SDS at room temperature) before transferring to the second dimension. The second dimension (SDS-PAGE) was performed on (15.2% total acrylamide, 1.3% bisacrylamide) polyacrylamide gels using a Desaphor VA 300 system. IEF tube gels were placed onto the polyacrylamide gels (20 cm X 30 cm X 0.7 mm) and fixed using 1.0% (w/v) agarose containing 0.01% (w/v) bromphenol blue dye (Riedel de-Haen, Seelze, Germany). The temperature of the running buffer was thermally regulated for the duration of the run (typically 5 hours at 200v).

NB: For protein identification by mass spectrometry a larger amount of protein was required. 500μg of total protein was subjected to 2D-electrophoresis and subsequently stained using an MS compatible (see below) stain. For protein identification DIGE staining was not employed.

2.4.8 Gel scanning, digitising and analysis The resultant 10 sample gels (5 pulsed Vs control; 5 continuous Vs control) were wiped with MEK ethanol and subsequently scanned using a Typhoon 9400 scanner (Amersham Biosciences). The appropriate excitation and emission wavelengths were chosen specifically for each of the dyes as

recommended by the manufacturer. Images were scanned at a resolution of 100μM and pre- processed using ImageQuantTM software (Amersham Biosciences). Processed, cropped images were imported into the Differential In-gel Analysis (DIA) software module (Amersham Biosciences). The DIA software module provides consistent and accurate co-detection of image pairs from the same gel using novel algorithms. Background subtraction, quantification, normalization and first level matching within one gel are automated allowing for high-throughput analysis with low experimental variation. The subsequent DIA analysed dataset and images were imported into the Biological Variation Analysis (BVA) module. This facilitates batch-processing, up to several hundred gel pair-images can be processed unattended for significant time savings. Images are then matched between gels using the BVA software module that looks for consistency of differences between samples across all the gels and applies statistics to associate a level of confidence for each of those differences.

Spot intensities were normalized to the internal standard (Cy2 labelled) and significantly regulated proteins were identified by Student T-test (p<0.05). Protein spots differentially expressed between the treated and untreated groups were flagged for spot excision and protein identification by MS.

2.4.9 2D-SDS PAGE Gel total protein staining In the course of optimisation a mass spectrometry incompatible stain was employed (Swain and Ross 1995). This simple sensitive procedure includes the cross-linker glutaraldehyde which is incompatible with MS. For protein identification an MS compatible silver stain (“Blum”) was employed as previously reported (Nesterenko, Tilley et al. 1994). The staining protocols were performed as described in the literature.

2.4.10 Trypsin digestion Spots which were significantly regulated by a ratio of 1.5 or more were manually excised from the preparative identification gel using a clean spot picker and transferred to clean glass mini-tubes. Spots were destained and digested with trypsin (Promega, Madison, WI) as previously described (Gharahdaghi, Weinberg et al. 1999). Briefly, spots were destained with a 1:1 working solution of 30mM potassium ferrocyanide and 100mM sodium thiosulphate (Merck, Darmstadt, Germany). A sufficient volume to cover the spots was added (30-40μL) and allowed to sit for 1-2 min. The spot was equilibrated with 10mM ammonium bicarbonate – ‘Ambic’ (Merck, Darmstadt, Germany) for

10 min, followed by dehydration solution (Ambic:acteonitrile [ACN] 1:1). This was repeated once more and subsequently dried by speedvac (Savant). Trypsin digestion was performed with 5± 10 ng/mL of trypsin and 10 mM ammonium bicarbonate and incubated overnight at 37oC. The peptides were extracted in 20μL extraction solution (5% Trifluoroacetic acid in ACN) and placed in a sonication bath for 10 min. This was repeated twice, the pooled extracted extracts were briefly reduced by speedvac and analysed by ESI-MS/MS.

2.4.11 Mass Spectrometry Materials and Reagents. Acetonitrile (ACN, HPLC-S gradient grade) was purchased from Biosolve (Valkenswaard, Netherlands). Formic acid (FA, Baker Analyzed) was obtained from J.T. Baker B.V. (Deventer, Netherlands), and trifluoroacetic acid (TFA, for protein analysis) was obtained from Merck (Darmstadt, Germany). Sequencing grade modified trypsin was purchased from Promega (Madison, WI, USA). Mass calibration standards were obtained from Applied Biosystems (Foster City, CA, USA). Water was used from an ELGA Labwater filtration system (Vivendi Water Systems, Ransbach-Baumbach, Germany). All other reagents were acquired from commercial sources.

Mass Spectrometric Analysis. Tandem MS (MS/MS) analyses were performed on a Bruker Daltonics HCT plus Ion Trap (Bremen, Germany) system operated in the sensitive mode. A nano electrospray ionization (nano-ESI) source (Bruker Daltonics, Germany) equipped with distal coated SilicaTips (FS360–20–10-D; New Objective) was used in conjunction with online capillary HPLC with the following mass spectral parameters: Capillary Voltage (CV), 1400 V; End Plate Offset (PO), 500 V; Dry Gas (DG), 10.0 l/min; Dry Temperature (DT), 160°C; aimed Ion Charge Control (ICC) 150000; maximal fill-time 500 ms. Online reversed-phase capillary HPLC separations were performed with the Dionex LC Packings HPLC systems (Dionex LC Packings, Idstein, Germany) as previously described by Schäfer et al (Heike Schaefer 2004).

MS spectra were a sum of seven individual scans ranging from m/z 300 to m/z 1400 with a scanning speed of 8,100 (m/z)/s. For MS/MS analyses, data-dependent software (HCT plus, Esquire Controle, Bruker Daltonics, Germany) was employed to select the two most intense, multiple-charged peptide ions detected in MS spectra. To generate fragment ions, low-energy collision induced dissociation (CID) was performed on previously isolated peptide ions by applying

a fragmentation amplitude of 0.6 V. Generally, MS/MS spectra were a sum of four scans ranging from m/z 100 to m/z 2200 at a scan rate of 26,000 (m/z)/s. Exclusion limits were automatically placed on previously selected mass to charge ratios for 1.2 minutes. The ion trap instrument was externally calibrated with commercially available standard compounds.

For protein identification, uninterpreted peptide MS/MS spectra were correlated with the NCBI protein sequence database (http://www.ncbi.nlm.nih.gov) restricted to Mus musculus applying the SEQUESTTM algorithm (Eng 1994). The reliability of protein identifcation via database searching with SEQUESTTM was checked manually, and sequence coverages reported correspond to the number of amino acids identified in proteins.

2.4.12 Statistics Factorial ANOVA analysis was performed using version 6.0 of Statistica software, StatSoft (Europe) GmbH., Hamburg, Germany. 3D-charts were plotted using the statistical software SPSS, SPSS GmbH Software, Munich, Germany.

2.5 Results A continuous and pulsed odorant exposure regime was designed in order to distinguish differences in protein regulation which are due to stimulus induced plasticity of the olfactory sensory neuron. Two groups of male juvenile C57BL/6 mice were exposed to octanal from postnatal day 10 (p10) for a duration of 20 days. One group received a constant supply of odorant (‘continuous group’), the other group received pulsed odorant applications - 4 x 15 minute daily pulsed treatment (‘pulsed group’). A third control group received sham continuous exposure to water, the ‘control group’. It was hypothesized that the continuous experiment induces long term adaptation whereas the pulsed treatment induces peripheral sensitization to the stimulant.

Octanal* (Figure 2.6) was selected as a stimulant for several reasons. It has been shown to activate the I7 receptor has in both mice and rat (Krautwurst, Yau et al. 1998) and was subsequently functionally characterised. This included molecular modelling of the ligand’s binding site (Singer 2000), elucidation of its receptive field (Araneda, Peterlin et al. 2004) and mapping of the glomeruli response to the stimulus (Johnson, Farahbod et al. 2004). Furthermore, data from previous work

* Octanal has several synonyms – Octylaldehyde, 1-Octanal, Aldehyde C8, Capryl aldehyde & N-Octylaldehyde

from our laboratory indicated that a significant proportion of wild type murine OSNs (approx. 7%) respond to a 1:10,000 dilution of Octanal (Wetzel, Brunert et al. 2005). It was also shown that rat OSNs can respond to several aliphatic aldehydes (carbon chain length C5-C10) with the greatest number of cells responding to octanal (Kaluza and Breer 2000).

Figure 2.6 The structure and properties of Octanal.

In order to identify and quantify differences in protein regulation between treated and control mice fluorescent difference gel electrophoresis (DIGE) was performed. The olfactory epithelium from treated mice were labelled and compared to the OE from control mice using DeCyderTM software suite (Amersham Biosciences); the labelling strategy follows-:

Pulsed Treated Mice (Cy5 Labelled - Red) Vs Control Untreated (Cy3 Labelled - Green)

Continuous Treated Mice (Cy5 Labelled - Red) Vs Control Untreated (Cy3 Labelled - Green)

Internal Reference/Standard (comprising all analysed samples) – Cy2 (yellow)

In each experiment the sample size was five (n=5). Each gel was scanned three times (Cy5, Cy3, Cy2) allowing multiplexing, dataset normalisation and gel matching. Spot volume is normalised using DIATM software. DIA analyses images from a single gel detecting and quantifying spots. Spot volumes are calculated by the pixel intensity relative to the spot boundary; this data also takes into account the volume, slope and peak height of the spot (Figure 2.7). This “co-detection” algorithm acquires spot data from each CyDye scan and reports the ‘spot ratio’ – changes between Cy5 (treated) and Cy3 (untreated) gel images relative to the internal reference (Cy2) i.e. the ratio value of a spot pair. This data is normalised so that the modal peak volume of ratios is 0. The normalized data is referred to as the ‘volume ratio’ – any increase or decrease in protein abundance is consequently reflected as an x-fold change from 1 to ∞ or from -1 to ∞, respectively. To exemplify, a two-fold up-regulation of a spot would be reported as “2.00” i.e. this two-fold increase is relative to the standard; a ‘standardised abundance’.

Figure 2.7 Schematic representation of parameters used to calculate spot data using the DIA (Differential In-gel Analysis) Module

Figure 2.8 Example of two DIGE gels. Upper: Pulsed treated mouse (n=1) Lower: Continuously treated mouse (n=1). In both cases treated and control mice were labelled using Cy5 (red) and Cy3 (green), respectively.

Figure 2.8 above demonstrates two multiplexed images (Cy3, Cy5 & Cy2) from the pulsed and continuous experiments (n=1). Fluorescent green and red spots indicate control and treated labelled samples, respectively. Yellow spots reflect no significant change in protein regulation. Significant changes (+/- 1.5 fold) in spot intensity were manually validated. For the continuous experiment a total of 18 significantly regulated spots were identified: 11 down-regulated and 7 up- regulated. For the pulsed experiment a total of 29 significantly regulated spots were identified: 4 down-regulated and 25 up-regulated

2.5.1 Regions of interest To illustrate the power of DIGE in differential analysis, two regions from the continuously (region A) and pulsed treated (region B) mice were selected for elaboration. To reiterate, 5 gels were run for each experiment (n=5) ensuring that any changes are reproducible and statistically significant.

Figure 2.9 DIGE scan from 5 gels comparing the control (green) and pulsed (red) treated mice. Underneath each fluorescent image is the corresponding 3D-intensity chart which schematically represents changes in intensity.

From Figure 2.9 it can be seen that the magnitude of the control signal amplitude is greater than the treated signal i.e. there was a significant down-regulation of this region as a consequence of pulsed octanal treatment. This striking effect was demonstrated across all gels, however, it was only statistically significant in 4/5 gels. Thus, these spots were ‘flagged’ for excision and identification by mass spectrometry. Area of interest B (Figure 2.10), on the other hand, demonstrates non- reproducible and insignificant changes for one spot across five ‘pulsed’ gels. In 2 of the pulsed gels this spot was up-regulated. This was not the case for the remaining pulsed treated mice where either there was no significant change or a moderate down-regulation.

Figure 2.10 Biological variation. DIGE scan from 5 gels comparing the control (green) and pulsed (red) treated mice. This spot was not significantly regulated

Spot #1 (Figure 2.11) demonstrated the most striking difference, a significant up-regulation. From the standardized log abundance chart this effect was not as marked as in the other pulsed samples.

#1 was analysed by ESI-MS/MS and identified as sulfotransferase family, cytosolic, 1C, member 1 - SULT1C1 (Swiss-Prot Accession Number: Q80VR3). This demonstrates the power of DIGE in detecting subtle changes in protein regulation.

Figure 2.11 Biological variation. Histogram shows differences in standardized log abundance across pulsed gels. Gel images and 3D diagram are from one gel - Spot #1 (region of interest B)

Figure 2.12 Master gels showing statistically significantly up- and down-regulated spots. The spots were arbitrarily numbered & excised for protein identification using mass spectrometry. Magnified area on pulsed gel was significantly up-regulated and identified as two cytoskeltal proteins, K8 & K18, which are clearly subjected to post-translational modifications (PTMs)

2.5.2 Pulsed Gel Analysis Significantly regulated proteins were excised (Figure 2.12), tryptically digested and peptide sequenced using ESI-MS/MS. Using the SEQUESTTM algorithm the resultant spectra were used to search the public NCBI Mus musculus database. The following table (Table 2.2) indicates the spot number and its degree of regulation for the pulsed treated samples. The corresponding gel images and differential abundances are also shown.

Table 2.2 Statistically significant (P<0.05) changes for pulsed experiment.