The Pennsylvania State University the Graduate School College of Agriculture

The Pennsylvania State University The Graduate School College of Agriculture

UNDERSTANDING THE SENSORY BASIS FOR ANOPLOPHORA GLABRIPENNIS

ODOR-MEDIATED BEHAVIOR

A Dissertation in

Entomology

Loyal Philip Hall

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2018

The dissertation of Loyal Philip Hall was reviewed and approved* by the following:

Thomas C. Baker Distinguished Professor of Entomology and Chemical Ecology Dissertation Adviser Chair of Committee

Kelli Hoover Professor of Entomology

Melody Keena Adjunct Faculty Research Entomologist, USDA Forest Service Special Member

Laura P. Leites Assistant Professor of Quantitative Forest Ecology

Michael Saunders Professor Emeritus of Entomology

Gary W. Felton, Ph.D. Professor of Entomology Department Head of Entomology

*Signatures are on file in the Graduate School.

ABSTRACT

I have been examining the Anoplophora glabripennis olfactory sensory neurons (OSNs), their axonal targets in the antennal lobe, the sensilla containing them, and their collective activities in an attempt to describe the neurophysiological activities in males and females that may help explain their behavioral responses to male-produced pheromone and other volatiles. This dissertation is a series of investigations that create a clearer picture of the how the olfactory system that is used by these beetles to detect general odorants and pheromone components is constructed. Chapter 2 contains the results of neuroanatomical investigations of the male and female antennal lobe, in which I looked for possible sexual dimorphism that might explain any differences in male vs. female behavioral responses to sex pheromone components. Any sexually dimorphic glomeruli would be likely targets for pheromone-tuned OSNs. In these studies I created a 3-dimensional atlas of male and female antennal lobes, resulting in models for which I labeled glomeruli according to their topographical locations as I looked for any sex-specific glomeruli related to sex pheromone olfaction. Chapter 3 is a report on progress towards a protocol to use a calcium imaging technique to look for activation of specific glomeruli in the antennal lobe that would receive inputs from the axons of sex-pheromone-tuned OSNs. Chapter 4 describes using simultaneously recorded EAGs of antennal sections to try to identify locations of sensilla containing OSNs tuned to different classes of odorants. EAG evidence is provided showing that water vapor is sensed by the antenna and describing the distribution of colocalized OSNs responsive to pheromone components versus those responsive to other volatiles. Chapter 5 describes the ability of the maxillary and labial palps to detect several odorants and water vapor. OSNs that contribute to these EAP responses are located and recorded from using SSR and the type of sensilla they are housed in is characterized. In general, sensing water vapor and acetic acid is indicated from antennal and palp recordings as being important to A. glabripennis.

iii

TABLE OF CONTENTS

List of Tables...... v List of Figures……………………………………………...……………………………...……...vi Acknowledgements……………………………………………………………………………….ix

Chapter 1. Introduction……………………...……………………..……………………………...1 Chapter 1 References...... 14

Chapter 2. Odorant Receptors and Antennal Lobe Morphology Offer a New Approach to Understanding Olfaction in the Asian Longhorned Beetle………………………………………19 Chapter 2 References...... 34

Chapter 3. Calcium Imaging Studies to Identify Glomeruli Activated by Male-Emitted Pheromone……………………………………………...………………………………………..39 Chapter 3 References...... 45

Chapter 4. Understanding the Distribution of Olfactory Sensory Neurons Along the Antenna Using Electroantennogram Recordings………………………………………………………….46 Chapter 4 References...... 77

Chapter 5. Palp Electrophysiology Indicates Well-developed Sensitivity to Moisture Levels, Along With Certain Volatile Chemicals………………………...... ……….…….....81 Chapter 5 References...... 105

List of Tables

Table 3-1. Calcium imaging methodology permutations and results...... 44 Table 4-1. Odorants used in this study, along with their commercial sources and purities...... 49 Table 4-2. Numbers of trichoid, basiconic 1, and basiconic 2 sensilla counted on A. glabripennis antennal flagellomeres using SEM imaging...... 70 Table 5-1. Odorants used in this study, along with their commercial sources and purities...... 88 Table 5-2. Change in pre- to post-stimulus spike frequency (spikes/sec) of OSNs in coeloconic sensilla on maxillary palps of males and females in response to 100ug of odorant...... 100

List of Figures

Figure 1-1. Anoplophora glabripennis checking out the surroundings in Jena, Germany………...………1 Figure 1-2. Scanning electron micrographs of sensilla on the antennae of female A. glabripennis….……5 Figure 1-3. SSR tracings and corresponding SAPD analysis indicating either two reporting OSNs primarily responding to either the pheromone component or general odorant or single reporting OSNs responding to both odorants……………………………………………………………………………...6,7 Figure 1-4. Spike trains recorded from OSNs in A. glabripennis trichoid sensilla in response to successive puffs of the ALB aldehyde pheromone component and different plant volatiles showing a larger spiking OSN responding to the pheromone components regardless of order of stimulation……….8 Figure 1-5. Spike trains recorded from OSNs in A. glabripennis trichoid sensilla in response to successive puffs of the ALB alcohol pheromone component and different plant volatiles showing a larger spiking OSN responding to the pheromone components regardless of order of stimulation………………9 Figure 1-6. Mean amplitudes (± SE) of action potentials recorded from A. glabripennis trichoid sensilla in response to successive puffs of the two ALB pheromone components and citronellal/geraniol...... 10 Figure 1-7. Spike trains recorded from the large spiking OSN in an A. glabripennis trichoid sensillum in response to successive puffs of the ALB aldehyde and ALB alcohol pheromone components showing that this larger spiking OSN responds to both pheromone components……………………….………………11 Figure 1-8. Mean (± SE) action potential frequencies of A. glabripennis female (top) and male (bottom) recorded from the large-spiking OSNs housed in trichoid sensilla responding to a dose-response series of the ALB alcohol pheromone component……………………………………….…………………………12 Figure 1-9. Mean (± S.E.) action potential frequencies of A. glabripennis female (top) and male (bottom) recorded from the large-spiking OSNs housed in trichoid sensilla responding to a dose-response series of the ALB aldehyde pheromone component………………………………………………………………..12 Figure 2-1. Unrooted phylogram illustrating the OR gene family of Anoplophora glabripennis...... 25 Figure 2-2. Unrooted cladogram illustrating relationships among known ORs from five beetle families...... 26 Figure 2-3. Abundance and sex bias of ORs in male and female Anoplophora glabripennis, as measured by the presence of unique reads of each receptor in the transcriptome……………...……………………27 Figure 2-4. Glomeruli in the left AL of adult Anoplophora glabripennis, with orientation given relative to the neuraxis………………………………………………………………………………………………..29 Figure 2-5. Volumes of glomeruli in four female and four male Anoplophora glabripennis...... 30 Figure 3-1. Beetle prepared for calcium imaging. The brain is exposed and a well of ringer solution is kept over the preparation...... 41 Figure 3-2. Detail of exposed antennal lobes (AL), showing the antennal nerve (AN) intact. The protocerebrum (PC) is also visible...... 42 Figure 3-3. ALB positioned for calcium imaging...... 43 Figure 4-1. Quadroprobe electroantennogram probe used to measure simultaneous EAGs of four antennal sections...... 48 Figure 4-2. Electroantennogram tracings from a male A. glabripennis in response to a puff of 100µg of eugenol...... 50

Figure 4-3. EAGs from four different sections of female and male antennae in response to puffs of various amounts of water...... 52 Figure 4-4. The humidifying cylinder used to add water vapor to the puffs and its placement in the set-up...... 53 Figure 4-5. EAGs from female and male antennal sections in response to geraniol that was puffed from the odor cartridge using dry air from the puffing stream or air-puffs that had been humidified by the water-saturated dental wick upstream of the cartridge...... 56 Figure 4-6 Mean (± S.E.) electroantennogram response amplitudes (µV) from both female and male antennae showing the responses from four different antennal sections to various odorants...... 59 Figure 4-7. Mean electroantennogram responses to odorants exhibited by each of the four antennal sections for all odorants tested...... 60 Figure 4-8. Mean EAG amplitudes (± S.E.) from male and female antennal sections in response to different terpenoid volatiles compared to those evoked by the ALB OH and ALB Ald pheromone components...... 61-63 Figure 4-9. Mean EAG amplitudes (± S.E.) from male and female antennal sections in response to different general plant-related and other volatiles compared to those evoked by the ALB OH and ALB Ald pheromone components...... 64-66 Figure 4-10. Lengths of antennal flagellomeres of A. glabripennis ...... 69 Figure 4-11. Widths of antennal flagellomeres...... 69 Figure 4-12. Average numbers of combined trichoid and basiconic sensilla obtained via SEM imaging from two studies...... 70 Figure 4-13. Stubby sensilla with a possible single-pore depression located in small numbers on all segments of A. glabripennis antennae ...... 75 Figure 4-14. Drawing of a transverse (vertical) section of a water-sensing sensillum located on the antennae of the ground beetle, Notiophilus biguttatus ( from Altner et al. 1983)...... 76 Figure 5-1. Palps prepared for Electropalpogram/Single sensillum (EPG/SSR) recordings...... 83 Figure 5-2 The labial and maxillary palps. The apical pit of the maxillary palp can be seen flexed...... 84 Figure 5-3. SEM image of the tip of a female ALB maxillary palp, showing the presumed concave architecture and many sensilla of the apical pit surface...... 85 Figure 5-4. A. Light microscope and SEM images of the maxillary palp, coeloconic sensilla...... 86 Figure 5-5. Tungsten probe placed against the apical pit of the palp for EPG recording...... 87 Figure 5-6. A hyperpolarizing EPG response from a male labial palp in response to a puff containing 100µg of geraniol...... 87 Figure 5-7. Series of EPGs from a single male maxillary palp bathed in a humidified airstream to a succession of puffs using dry, unhumidified air to introduce dosages of water on filter paper ...... 89 Figure 5-8. Box plots of EPGs evoked by puffs from odor cartridges loaded with different dosages of water...... 90 Figure 5-9. Humidifying apparatus for injection line and its placement in the setup...... 91 Figure 5-10. Effect of decreases or increases in moisture levels imparted by the odor cartridge on the responses, showing EPG responses for blank, geraniol, and water injected with and without a humidified injection stream into a humidified constant airstream...... 92

vii

Figure 5-11. EPG responses by male and female maxillary and labial palps to different odorants...... 94,95 Figure 5-12. EPG responses to different odorants from male and female maxillary and labial palps, including responses to acetic acid and butyraldehyde...... 96,97 Figure 5-13. EPG tracing of female maxillary palp response to butyraldehyde and male maxillary palp response to acetic acid showing tonic depolarizations...... 98 Figure 5-14. SSR tracing of an ORN from a male coeloconic sensillum responding to butyraldehyde simultaneously showing the DC current from its EPG...... 99 Figure 5-15. SSR tracing of an ORN from a female coeloconic sensillum responding to butyraldehyde simultaneously showing the DC current from its EPG...... 99

viii

ACKNOWLEDGMENTS

A PhD is an accomplishment. I'm not unaware of the efforts it took on my part to get here. But I did not get here all on my own. Below is an incomplete list of those deserving of recognition:

Tom Baker, my advisor, who is also a mentor and a role model of how to be a rigorous scientist with humility and care for others.

My committee members Kelli Hoover, Melody Keena, Laura Leites, and Mike Saunders, provided valuable and needed advice and made their expertise available to me.

The Kelli Hoover lab provided the beetles used in these studies, along with many of the pheromones.

The beetles.

The Beatles.

My labmates Mike Domingue, Stefanos Andreadis, Jianrong Wei, Qiong Zhou, Haibin Chen, Justin George, Andy Myrick, Sara Hermann-Ali, and Kevin Cloonan gave me advice, bounced ideas around, taught me their techniques, offered their friendship, listened to me whine, got me into and out of trouble, and made the lab a great place to work.

Arash and Azi Maleki are great neighbors both at home and at work and I appreciate their love and friendship. They are quality scientists to boot.

My parents, Tom and Donna Hall, let their strange child be strange and continued to support, advise, and encourage me throughout adulthood.

My children, Felipe, James, and Ester, without whom I would have finished this PhD a lot sooner, but it would have been a lot lonelier and my life less full of joy.

My wife, Rebeca, has been a constant encouragement through the years. Her love has been a source of strength and joy for me. She manages to keep me grounded and drives me crazy, usually simultaneously.

Jesus.

Coffee.

Chapter 1 Introduction

The universe is a thing of beauty and awe at any scale. The scale on which I have been conducting my explorations has been the odorant sensing system of Anoplophora glabripennis, the Asian longhorned beetle (ALB). In particular, I have been examining olfactory sensory neurons (OSNs), their axonal targets in the antennal lobe, the sensilla containing them, and their collective activities in an attempt to describe the neurophysiological activities in males and females that may help explain their behavioral responses to male-produced pheromone.

Figure 1-1. Anoplophora glabripennis checking out the surroundings in Jena, Germany.

A. glabripennis is a longhorned beetle (Fig. 1-1), family Cerambycidae, from eastern Asia, primarily inhabiting areas in the Peoples Republic of China and the Koreas. The primary coloring of the adult beetle is black with white and, on closer inspection, blue and some scattered orange. This species is polyphagous, feeding on many species of hardwood trees from diverse genera. In its native range it feeds primarily on species of poplar. In North America, however, A. glabripennis avoids poplar and will use species of birch, willow, elm, and others as hosts, but primarily attacks maple species (Haack et al., 1997, USDA-APHIS-PPQ-CPHST 2015). Adults feed on the bark of living twigs on the host tree. Larvae feed under the bark at first and later in the heartwood, consuming up to 1000 cm3 of wood during the course of development (Yan and Qin 1992). The large galleries they create retard tree growth and ultimately kill the host tree as

the infestation develops, and the population of pest insects grows and progresses into the main branches and trunk (Meng et al. 2015).

The larvae hatch from eggs laid singly in oviposition pits created by the female biting into the bark of the host tree. Eggs hatch in 2 weeks - 2 months, depending on temperature. Eggs are capable of overwintering if laid too late in the season. Larvae develop over the course of 1-2 years before pupating for 12-50 days, again dependent upon temperature; in total, it may take up to 3 years for an adult beetle to mature from the time the egg is laid. Adult beetles can live for several months, up to 202 days, in laboratory conditions (Keena 2006). In natural conditions adults live about 30 days (Faccoli et al. 2015).

As of 2018, A .glabripennis is limited to only a few hundred square miles in North America, in several scattered infestation sites. It has already been eradicated in several locations and the possibility of eradication on this continent remains. The dispersal ability of this pest, without human intervention such transporting infested wood, is relatively low. The distance traveled can vary with the landscape (Wen et al. 1998, Smith et al. 2001, Sawyer et al. 2011) and temperature (Sawyer et al. 2011, Turgeon et al 2015), but usually is limited to a few hundred meters or less per year. Longer distances of 2-8 km have been recorded, though (Smith et al. 2004, Favaro et al 2015, Trotter and Hull-Sanders 2015). Because of the relatively small area of infestation and the normally short distance these beetles will fly in dispersal, delineating the areas infested will make it possible for more targeted destruction of host trees. At present, the most effective control method is to destroy host trees in an area of an infestation. This is not without challenges. Canada had declared the insect eradicated within its borders, but A. glabripennis was then found again in 2013, five years after the last reported detection (CFIA 2013), highlighting the need for more sensitive detection methods.

Sampling to identify infested trees is a challenging endeavor. Trained spotters using binoculars can search for signs of beetle presence but the point of attack for infestation can be the crown of the tree, meaning years may pass before beetle activity is within view of someone from the ground (Haack et al. 2006). The use of ladders, lifts, and climbing experts to inspect the upper reaches of trees is more likely to find beetles than ground inspections but is time-consuming and costly (Meng et al. 2015). Trained dogs can be used to locate the smell of A.glabripennis frass with 80-90% accuracy in some tests but is largely impractical due to expense and false-positives (Errico 2012). Detecting the larvae acoustically via microphones (Haack et al. 2001) and laser vibrometry are also in development (Zorovic and Cokl 2015).

Trapping methods for detection, monitoring, and mitigation of A. glabripennis populations in areas of invasion in the U.S. have shown promise with some successes thus far (Nehme et al. 2010, Meng et al. 2014). It is my hope that the research described in this thesis will prove to be useful in helping increase the sensitivity and specificity of any pheromone-based trapping systems used for eliminating this pest in areas of infestation. I hope, too, that my research will help in expanding our knowledge of cerambycid olfaction, an area of knowledge that has only

recently expanded explosively, with sex-pheromone-related research results reported now for an increasing number of cerambycid species only in the past 10-15 years (Ginzel and Hanks 2004, Lacey et al. 2004, Crook et al. 2014, Hansen et al. 2015, Mitchell et al 2011, Zhang et al. 2002, Lacey et al. 2007).

Mate-finding in A. glabripennis involves both sexes emitting pheromone. Males emit a pheromone that along with host volatiles has been shown to attract females from long distances, and females produce a pheromone that causes male attraction from a much shorter range via cuticular pheromones deposited on the surface upon which she walks, as described below. These complex A. glabripennis male-female pheromone interactions are not dissimilar to the mate- finding systems of some other cerambycids (Hanks and Millar 2017, Millar et al. 2009, Allison et al. 2004).

A. glabripennis males produce a blend of two components, 4-(n-heptyloxy )-butanal (“ALB aldehyde”) and 4-(n-heptyloxy)- butan-1-ol (“ALB alcohol”) (Zhang et al. 2002). Virgin females are attracted to this blend in a 1:1 ratio when these components are combined with host volatiles (Nehme et al. 2009). Field trapping studies have confirmed that the male-produced blend is most attractive to females with the addition of host volatiles, including (Z)-3-hexen-1-ol, (-)-linalool, and -caryophyllene (Nehme et al. 2009, Nehme et al. 2010, Meng et al. 2014, Kelli Hoover, personal communication). This behavioral synergy of host volatiles and pheromone components dovetails with reports of olfactory pathway synergy of pheromones and plant volatiles in other insect groups. For instance, Trona et al. (2013) showed that the presence of host volatiles increased excitation in an antennal lobe glomerulus of the codling moth that receives inputs from pheromone-sensitive neurons on the antenna. A potential third minor component, (E,E)- - farnesene has also been identified (Crook et al. 2014).

For their part, female A. glabripennis ready to mate produce a cuticular pheromone blend of (Z)- 9-tricosene, 2-methyldocosane, (Z)-7-pentacosene, (Z)-9-pentacosene, (Z)-7-heptacosene, and (Z)-9-heptacosene (Zhang et al. 2003). As females walk along branches, the components (Z)-9- tricosene and 2-methyldocosane are left behind, along with the minor components, (Z)-7- pentacosene and (Z)-9-pentacosene in a ratio of 2:12:1:1 (Hoover et al. 2014). These compounds function as a trail-sex pheromone that causes male attraction along the trail to the female. The use of trail pheromones had been hinted at by other studies of cerambycids (Millar et al. 2009) but had not been demonstrated for A. glabripennis until just recently (Hoover et al. 2014). Recent unpublished work by Fern Graves and Kelli Hoover (personal communication) indicate that only the major components, Z-9-heptacosene and 2-methyldocosane are utilized for the mate finding by the males, with the minor compounds being used as a spacing pheromone to keep eggs from being laid too closely together. Males follow this trail-sex pheromone blend, palpating the surface as they do, indicating the maxillary and labial palps are likely utilized to detect these chemicals (Graves et al. 2017). Upon sighting the female, the male rapidly approaches (He and Huang 1993), uses his antennae to verify the female (via the cuticular pheromone blend) as a conspecific ready to mate, and then initiates mating (Zhang et al. 2003).

Single Sensillum Recordings Leading to Dissertation

I was one of three scientists who contributed single sensillum recordings (SSR) to the study of the sensilla on the A. glabripennis antennae conducted during my PhD work. This study by Jianrong Wei, Qiong Zhou, Loyal Hall, Andrew Myrick, Kelli Hoover, Kathleen Shields, and Thomas C. Baker is in the process of submission for publication to the Journal of Chemical Ecology as "Olfactory Sensory Neurons of the Asian Longhorned Beetle, Anoplophora glabripennis, Specifically Responsive to Its Two Aggregation-Sex Pheromone Components." The findings of this study directly lead to much of this dissertion and for this reason is summarized here in the introduction. We performed single-sensillum recordings from male and female antennae of the Asian longhorned beetle that included as stimuli the two components of the aggregation-sex-pheromone in addition to various general odorants. We compared the aggregation-sex-pheromone-component responses of olfactory sensory neurons (OSNs) to those of OSNs that responded to a variety of plant-related odorants.

We recognized three main types of sensilla from which we tried to record OSN activities (Figs. 1-2A, B). One type was a so-called “blunt-tipped” basiconic sensillum (Fig. 1-2A; “Ba.1”), which we recorded from primarily on the middle flagellomeres nos. 4 through 7. A second type that we tried recording from was what we named a “sharp-tipped” basiconic sensillum (Fig. 1- 2A,B; “Ba.2”). We found that this sensillum type was distributed all over the antenna, and despite numerous attempts, we either failed to gain electrical connections or else any OSNs residing in these sensilla were unresponsive to all of the compounds we exposed them to. A third type was what we designated as “trichoid” sensilla (Fig. 1-2B; “Tr.”) from which we recorded from most easily on the two most distal antennal flagellomeres of both males and females.

Figure 1-2. Scanning electron micrographs of sensilla on the antennae of female A. glabripennis. A. Trichoid and basiconic sensilla on Flagellomere 9. “Ba 2” denotes a sharp-tipped type of basiconic sensillum that yielded no recordings. “Ba 1” denotes blunt-tipped basiconic sensillum. “Tr” denotes a slender, trichoid type of sensillum characterized by a slightly curved and tapered tip. B. Trichoid sensilla (“Tr”) on Flagellomere 10. Scale bar in A and B = 20 μm.

Making contacts with ALB sensilla and gaining good recordings was quite difficult, perhaps due to the hardness and small size of these sensilla. We found that recordings were not possible by puncturing the sensillum with a glass or tungsten electrode. Thus, among the small percentage of successful recordings we were able to obtain, success was gained when an tungsten recording electrode was touched to the base of either a blunt-tipped basiconic or trichoid sensillum apparently without puncturing it.

To differentiate between possibly two or more different OSNs’ responses from within a single sensillum, I used an automated spike-sorting program (SAPID, Smith et al., 1990). A template for each of the two action potential amplitudes and waveforms occurring within the sensillum were formed by the program and further modified by the user after inspecting the temporal occurrence of the spikes that had been sorted according to the two templates. Following this procedure, the templates were finalized and the SAPID program sorted the spikes according to the template; then the time of occurrence of the two different types of spikes was compared against the timing of the two different stimulus deliveries (Fig. 1-3). Additionally, the mean spike amplitudes from the first 10 action potentials generated by each of the two puffs were calculated and the ratios of larger-to-smaller amplitudes were calculated for the two different OSNs from each sensillum. When no significant change in spike amplitude occurred in the second burst of spikes to the second stimulus, it was concluded that there was a single OSN responding to each of the two stimuli. When there was no change in action potential frequency in response to the second stimulus, again it was concluded that the same OSN was responding to the two stimuli, because it had become adapted to the stimulus from the first puff.

Figure 1-3. SSR tracings and corresponding SAPD analysis indicating either two reporting OSNs (A and B), each primarily responding to either the pheromone component or general odorant, or single reporting OSNs (C and D), responding to both odorants.

In recording from the very narrow, delicate, slightly curved and smoothly tapered sensilla we are calling “trichoid” sensilla (Fig. 1-2B) that we sampled only on the terminal flagellomeres of both male and female antennae, we found OSNs exhibiting large amplitude action potentials (spikes) responsive to both the ALB aldehyde and ALB alcohol pheromone components (Figs. 4, 5). We found these OSNs in recordings from 18 female and 6 male A. glabripennis antennae. In these sensilla, there were also smaller-spike-amplitude OSNs co-located with the pheromone- component-responsive OSNs, and these companion OSNs responded to a variety of general odorants, most often either to geraniol or citronellal (Figs. 1-4,5,6).

Cross-adaptation experiments showed clearly that the larger-spiking pheromone-component- tuned OSNs were different OSNs from those responding to plant-related odorants, which always displayed a smaller spike amplitude (Figs. 1-4,5,6). It did not matter whether the ALB alcohol or ALB aldehyde was puffed first or second in the cross-adaptation regime. In either sequence was no cross-adaptation, and the two types of OSNs displayed different sized spikes (Figs. 1-4,5,6).

Figure 1-4. Spike trains recorded from OSNs in A. glabripennis trichoid sensilla in response to successive puffs of the ALB aldehyde pheromone component and different plant volatiles showing a larger spiking OSN responding to the pheromone components regardless of order of stimulation. A, B. The ALB aldehyde pheromone component and citronellal. C,D. The ALB aldehyde pheromone component and geraniol. Distance between baseline and either of the two gray horizontal lines represents 0.8 mV. Time-scale bars represent 0.25 s.

Figure 1-5. Spike trains recorded from OSNs in A. glabripennis trichoid sensilla in response to successive puffs of the ALB alcohol pheromone component and different plant volatiles showing a larger spiking OSN responding to the pheromone components regardless of order of stimulation. A, B. The ALB alcohol pheromone component and citronellal. C,D. The ALB alcohol pheromone component and geraniol. Distance between baseline and either of the two gray horizontal lines represents 0.8 mV. Time- scale bars represent 0.25 s.

Cross-adaptation studies using the ALB alcohol and ALB aldehyde against each other showed that in every case, the ALB aldehyde and alcohol were stimulating the same large-spiking OSN in each of the trichoid sensilla we recorded from (Fig. 1-6). The large-spiking OSN first stimulated by the ALB alcohol was followed by the same large-spiking OSN stimulated by the ALB aldehyde, and vice-versa (Fig. 1-6). These OSNs appeared to be slightly more responsive to the alcohol compared to the aldehyde, because the spike frequency in response to a puff of ALB aldehyde that followed the puff of the ALB alcohol was more likely to be reduced or even adapted than when the alcohol followed the aldehyde. In five of the eight sensilla that we were able to record from using double-puffed pheromone components, the response to the second puff was a negligible in change in spike frequency and with no change in spike amplitude. Thus the first-puffed pheromone component adapted the OSN so it did not respond to the second component, showing that the same OSN was tuned to both components. In these cases amplitudes to the second puff could not be discerned from those to the first puff and were not measured.

Figure 1-6. Mean amplitudes (± S.E.) of action potentials recorded from A. glabripennis trichoid sensilla in response to successive puffs of the two ALB pheromone components and either citronellal or geraniol. Successive puffs of the ALB aldehyde and alcohol are also shown for comparison. The OSN exhibiting the larger amplitude spikes responds to both the aldehyde and the alcohol pheromone components, and the co-compartmentalized OSN exhibiting smaller amplitude spikes responds to the two plant-related compounds, citronellal and geraniol. Data from successive puffs, regardless of order (plant volatile or pheromone component puffed either first or second), were merged. * = P < 0.05; ** = P < 0.01; *** P < 0.001 according to T-tests between odorant pairs. N= 9, 4, 10, 8, and 6 for ALB OH/geraniol, ALB OH/citronellal, ALB Ald/geraniol, ALB aldehyde/citronellal and ALB Ald/ALB OH pairs, respectively.

Dose-response series performed on this type of OSN indicated that there were no differences in the responsiveness between the male and female OSNs to increasing dosages of either the ALB alcohol (Fig. 1-8) or the ALB aldehyde (Fig. 1-9). It appeared that spike frequency of the OSNs in response to the two highest doses of the ALB alcohol reached higher levels compared to the same doses of the aldehyde in both males and females (Figs. 1-8, 9). This was consistent with the impression from the double-puffed studies (Fig.1-7), that this type of OSN is more sensitive to the ALB OH than to the ALB Ald.

Figure 1-7. Spike trains recorded from the large spiking OSN in an A. glabripennis trichoid sensillum in response to successive puffs of the ALB aldehyde and ALB alcohol pheromone components showing that this larger spiking OSN responds to both pheromone components. Distance between baseline and either the highest or lowest of the gray horizontal lines represents 0.8 mV. Time-scale bars represent 0.25 s.

Figure 1-8. Mean (± S.E.) action potential frequencies of A. glabripennis female (top) and male (bottom) recorded from the large-spiking OSNs housed in trichoid sensilla responding to a dose-response series of the ALB alcohol pheromone component. For females, N = 9, 15, 14, 15, and 12 for the 0.01, 0.1, 1, 10, and 100 µg doses, respectively. For males, N = 3 for all dosages.

Figure 1-9. Mean (± S.E.) action potential frequencies of A. glabripennis female (top) and male (bottom) recorded from the large-spiking OSNs housed in trichoid sensilla responding to a dose-response series of the ALB aldehyde pheromone component. For females, N = 9, 15, 15, 15, and 14 for the 0.01, 0.1, 1, 10, and 100 µg doses, respectively. For males, N = 3 for all dosages.

In sensilla located in the distal antennal flagellomeres as well as those located more proximally, i.e., mid-length along the antenna on flagellomere nos. 4-7, we found single OSNs in blunt- tipped basiconic sensilla (Figs. 1-3C,D) that were responsive to general and plant-related volatiles, especially the terpenoids, (E,E)-α-farnesene, (E)-β-farnesene, β-caryophyllene, and eugenol. Some of these same OSNs responded additionally to either of the two sex pheromone components, but because these OSNs also responded to some of the above plant volatiles as shown by cross-adaptation experiments, these OSNs will not be the types that convey sex- pheromone-specific information to the antennal lobe, but may explain the ability of the insect to discriminate between its own blend of components and single components such as is used by the closely related A. chinensis (Hansen et al 2015) and the activity of (E,E)-α-farnesene described by Crook et al (2014).

In the smooth-tipped, tapered, trichoid sensilla on the most distal antennal flagellomeres nos. 10 or 11 of both males and females, we found OSNs with high-amplitude action potentials that were tuned to the aldehyde and alcohol pheromone components and that did not respond to various plant-related volatiles. Because this OSN type responded to both the alcohol and aldehyde components it cannot be considered to be specifically tuned to either component. These large- spiking OSNs were co-compartmentalized in these sensilla with a second, smaller-spiking OSN responding to general volatiles such as geraniol, citronellal, limonene, 1-octanol, nerol and citral. The large-spiking OSNs thus appear to be a type that will be involved in aggregation-sex pheromone pathways targeting a specific glomerulus in the antennal lobe and in generating pheromone-related behavioral responses in A. glabripennis.

Summary of Dissertation Chapters

In the following chapters is a series of investigations that create a clearer picture of the how the olfactory system that is used by these beetles to detect general odorants and pheromone components is constructed. Chapter 2 contains the results of neuroanatomical investigations of the male and female antennal lobe, in which I looked for possible sexual dimorphism that might explain any differences in male vs. female behavioral responses to sex pheromone components. Any sexually dimorphic glomeruli would be likely targets for pheromone-tuned OSNs. In these studies I created a 3-dimensional atlas of male and female antennal lobes, resulting in models for which I labeled glomeruli according to their topographical locations as I looked for any sex- specific glomeruli related to sex pheromone olfaction. Chapter 3 is a report on progress towards use of a calcium imaging technique to look for activation of specific glomeruli in the antennal lobe that would receive inputs from the axons of sex-pheromone-tuned OSNs. Chapter 4 describes using simultaneously recorded EAGs of antennal sections to try to identify locations of sensilla containing OSNs tuned to different classes of odorants. EAG evidence is provided showing that water vapor is sensed by the antenna and describing the distribution of colocalized OSNs responsive to pheromone components versus those responsive to other volatiles. Chapter 5

describes the ability of the maxillary and labial palps to detect several odorants and water vapor. OSNs that contribute to these EAP responses are located and recorded from using SSR and the type of sensilla they are housed in is characterized. In general, sensing water vapor and acetic acid is indicated from antennal and palp recordings as being important to A. glabripennis.

The research performed in this dissertation was funded in part by Cooperative Agreement #16- 8130-1430-CA between The Pennsylvania State University and the USDA/APHIS entitled, "Improvement of Traps and Lures for Exotic Wood-boring Beetles."

References

Allison J, Borden J, Seybold S. 2004. A review of the chemical ecology of the Cerambycidae (Coleoptera). Chemoecology 14:123–150.

(CFIA) Canadian Food Inspection Agency. 2013. Asian long-horned beetle found in Mississauga, Ontario. (http://www.inspection.gc.ca/about-the-cfia/ newsroom/news- releases/2013-09-20/eng/1379685062894/1379685078216)

Crook D, Lance D, Mastro V. 2014. Identification of a potential third component of the male- produced pheromone of Anoplophora glabripennis and its effect on behavior. J. Chem. Ecol. 40: 1241–1250.

Errico M. 2012. Asian longhorned beetle detector dog pilot project, p. 18. in Proceedings, 23rd USDA Interagency Research Forum on Invasive Species, 10–13 January 2012, Annapolis, MD. General Technical Report NRS-P-114. U.S. Forest Service, Newtown Square, PA.

Faccoli M, Favaro R., Smith MT, Wu J. 2015. Life history of the Asian longhorn beetle Anoplophora glabripennis (Coleoptera Cerambycidae) in southern Europe. Agricultural and Forest Entomology 17: 188–196.

Favaro R, Wichmann L, Ravn HP, Faccoli M. 2015. Spatial spread and infestation risk assessment in the Asian longhorned beetle, Anoplophora glabripennis. Entomologia Experimentalis et Applicata 155:95-101.

Ginzel M, Hanks LM. 2004. Role of host plant volatiles in mate location for three species of longhorned beetles. Journal of Chemical Ecology 31: 213-217.

Haack RA. 2006. Exotic bark- and wood-boring Coleoptera in the United States: Recent establishments and interceptions. The Canadian Journal of Forest Research 36: 269–288.

Haack RA, CaveyJ, Hoebeke ER, Law K. 1996. Anoplophora glabripennis: A new tree-infesting exotic cerambycid invades New York. Newsletter of Michigan Entomological Society 41: 1–3.

Haack RA, PolandTM, Petrice TR, Smith C, Treece D. 2001. Acoustic detection of Anoplophora glabripennis larvae and other wood-borers. Proceedings of the 2001 USDA Interagency pp. 74–75.

Hansen L, Xu T, Wickham J, Chen Y, Hao D, Hanks LM, Millar JG, Teale SA. 2015. Identification of a male- produced pheromone component of the citrus longhorned beetle, Anoplophora chinensis. PLoS ONE 10(8): e0134358.

Hoover K, Keena MA, Nehme ME, Wang S, Meng P, Zhang A. 2014. Sex-specific trail pheromone mediates complex mate finding behavior in Anoplophora glabripennis. Journal of Chemical Ecolology 40: 169–180.

Keena MA. 2006. Effects of temperature on Anoplophora glabripennis (Coleoptera: Cerambycidae) adult survival, reproduction, and egg hatch. Environmental Entomology 35: 912–921.

Lacey ES, AM Ray, Hanks LM. 2007. Calling behavior of the Cerambycid beetle Neoclytus acuminatus acuminatus. Journal of Insect Behavior 20:117-128.

Lacey, E. S., Ginzel, M. D., Millar, J. G., And Hanks, L. M. 2004. Male-produced aggregation pheromone of the cerambycid beetle Neoclytus acuminatus acuminatus. J. Chem. Ecol. 30:1493–1507.

Meng PS. 2014. Where's that smell? Trapping and sensory biology of the Asian longhorned beetle, Anoplophora glabripennis. Master's thesis. Pennsylvania State University, University Park, PA.

Millar J, L Hanks, J Moreira, J Barbour, E Lacey. Pheromone chemistry of cerambycid beetles from chemical ecology of wood-boring insects. 2009. Forestry and Forest Products Research Institute. Tsukuba,Ibaraki, Japan.

Mitchell RF, Graham EE, Wong JCH, Reagel PF, Striman BL, Hughes GP, et al. Fuscumol and fuscumol acetate are general attractants for many species of cerambycid beetles in the subfamily Lamiinae. Entomologia Experimentalis et Applicata. 2011; 141: 71–77.

Morewood WD, Neiner PR, Sellmer JC, Hoover K. 2004. Behavior of adult Anoplophora glabripennis on different tree species under greenhouse conditions. Journal of Insect Behavior 17: 215–226.

Nehme ME, Keena MA, Zhang A, Baker TC, Hoover K. 2009. Attraction of Anoplophora glabripennis to male-produced pheromone and plant volatiles. Environ. Entomol. 38: 1745–1755.

Nehme ME, Keena MA, Zhang A, Baker TC, Xu Z, Hoover K. 2010. Evaluating the use of male-produced pheromone components and plant volatiles in two trap designs to monitor Anoplophora glabripennis. Environmental Entomology 39: 169–76.

Nehme ME, Trotter RT, Keena MA, McFarland C, Coop J, Hull-Sanders H, Meng P, De Moraes CM, Mescher MC, Hoover K. 2014. Development and evaluation of a trapping system for Anoplophora glabripennis in the United States. Environmental Entomology 43: 1034–1044.

Sawyer AJ, Panagakos WS, Audra E, Freeman KJ. 2011. Asian longhorned beetle, over the river and through the woods: habitat-dependent population spread. In: McManus KA, Gottschalk KW (eds) Proceedings of 21st U.S. Department of Agriculture interagency research forum on invasive species 2010; 2010 January 12–15; Annapolis. General technical report NRS-P-75.

Smith JJB, Mitchell BK, Rolseth BM, Whitehead AT, Albert PJ (1990) SAPID tools: microcomputer programs for analysis of multi-unit nerve recordings, Chem Senses 15: 253–270.

Smith MT, Bancroft J, Li G, Gao R, Teale S. 2001. Dispersal of Anoplophora glabripennis (Cerambycidae). Environmental Entomology 30:1036–1040.

Smith MT, Tobin PC, Bancroft J, Li G, Gao R. 2004. Dispersal and Spatiotemporal Dynamics of Asian Longhorned Beetle (Coleoptera: Cerambycidae) in China. Environmental Entomology 33:435-442

Trona F, Anfora G, Balkenius A, Bengtsson M, Tasin M, Knight A, Janz N, Witzgall P, Ignell R. 2013. Neural coding merges sex and habitat chemosensory signals in an insect herbivore. Proc R Soc B 280:20130267.

Trona F, Anfora G, Bengtsson M, Witzgall P, Ignell R. 2010. Coding and interaction of sex pheromone and plant volatile signals in the antennal lobe of the codling moth Cydia pomonella. Journal of Experimental Biology 213:4291-4303.

Trotter RT and Hull-Sanders HM. 2015. Quantifying Dispersal of the Asian longhorned beetle (Anoplophora glabripennis, Coleoptera) with incomplete data and behavioral knowledge. Biological Invasions 17:3359–3369.

Turgeon JJ, Orr M, Grant C, Wu Y, Gasman B. Decade-Old Satellite Infestation of Anoplophora glabripennis Motschulsky (Coleoptera: Cerambycidae) Found in Ontario, Canada Outside Regulated Area of Founder Population. The Coleopterists Bulletin, 69:674-678.

(USDA-APHIS) U.S. Department of Agriculture, Animal and Plant Health Inspection Service. 2010. Nursery treatment efficacy study within Worcester County, Massachusetts, to support the Asian longhorned beetle cooperative eradication program. (http://www.aphis.usda.gov/plant_health/ea/downloads/ALB-EA-worcester- nurserystudy.pdf) (accessed April 13, 2016).

(USDA-APHIS) U.S. Department of Agriculture, Animal and Plant Health Inspection Service. 2013a. Asian longhorned beetle cooperative eradication program in Clermont County, Ohio. (http://www.aphis.usda.gov/plant_health/ea/downloads/2013/OHClermontCountyRevise d_EA_May_final.pdf) (accessed April 13, 2016).

(USDA-APHIS) U.S. Department of Agriculture, Animal and Plant Health Inspection Service. 2013b. News release: Federal and state officials announce tree survey efforts in Babylon Township, New York due to the discovery of Asian longhorned beetle. (http://www.aphis.usda.gov/newsroom/ 2013/09/pdf/ny_babylon_township_alb.pdf) (accessed April 13, 2016).

(USDA-APHIS-PPQ) U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Plant Protection and Quarantine. 2010. USDA updates list of regulated articles for Asian longhorned beetle to include Cercidiphyllum (katsura tree). (http://www.aphis.usda.gov/publications/plant_health/content/printable_version/sa_alb_c ercidiphyllum.pdf) (accessed April 13, 2016).

(USDA-APHIS-PPQ). U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Plant Protection and Quarantine. 2014. Asian longhorned beetle response guidelines. (http://www.aphis.usda.gov/plant_health/ plant_pest_info/asian_lhb/downloads/response-guidelines.pdf) (accessed April 13, 2016).

(USDA-APHIS-PPQ-CPHST). U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Plant Protection and Quarantine, Center for Plant Health Science and Technology. 2015. Asian longhorned beetle: Annotated host list. (http://www.aphis.usda.gov/plant_health/plant_pest_info/asian_lhb/downloads/hostlist.pd f) (accessed April 13, 2016).

(USDA-FS) U.S. Department of Agriculture, Forest Service. 2002. Current U. S. forest data and maps. (http://www.fia.fs.fed.us/slides/current-data.pdf) (accessed April 13, 2016).

Wei J, Zhou Q, Hall L, Domingue M, Hoover K, Shields K, Baker T. 2016. Olfactory sensory neurons of the Asian longhorned beetle, Anoplophora glabripennis, specifically responsive to its two sex pheromone components. Journal of Insect Physiology (pending).

Wen J, Li Y, Xia N, Luo Y. 1998. Study on dispersal pattern of Anoplophora glabripennis adults in poplars. Acta Ecologica Sinica 18:269-277.

Yan J and Qin X. 1992. Anoplophora glabripennis (Motsch.), pp. 455–457. In G., Xiao (ed.), Forest insects of China. China, China Forestry Publishing House.

Zhang A, Oliver J, Aldrich J. 1999. Aggregation pheromone for the Asian longhorned beetle, Anoplophora glabripennis (Coleoptera: cerambycidae). Filed Jul. 7, 1999, Appl. No. 347,907. US 6,177,073 B1

Zhang A, Oliver JE, Aldrich JR, Wang B, Mastro VC. 2002. Stimulatory beetle volatiles for the Asian longhorned beetle, Anoplophora glabripennis (Motschulsky). Zeitschrift für Naturforschung 57: 553–558.

Zhang A, Oliver JE, Chauhan K, Zhao B, Xia L, Xu Z. 2003. Evidence for contact sex recognition pheromone of the Asian longhorned beetle, Anoplophora glabripennis (Coleoptera: Cerambycidae). Naturwissenschaften 90: 410–3.

Zorović M and Čokl AJ. 2015. Laser vibrometry as a diagnostic tool for detecting wood-boring beetle larvae. Journal of Pest Science 88: 107–112.

Chapter 2 Odorant Receptors and Antennal Lobe Morphology Offer a New Approach to Understanding Olfaction in the Asian Longhorned Beetle

Published January 2017 Mitchell RF, Hall LP, Reagel PF, McKenna DD, Baker TC, Hildebrand JG. 2017. Odorant receptors and antennal lobe morphology offer a new approach to understanding olfaction in the Asian longhorned beetle. Journal of Comparative Physiology A 203:99–109.

Abstract The Asian longhorned beetle Anoplophora glabripennis (Motchulsky) is an exotic forest pest that has repeatedly invaded North America and Europe from Asia, and has the potential to kill millions of trees and cause billions of dollars in damage. Traps baited with an attractive mixture of volatile organic compounds (VOCs) from hosts have been of limited success in monitoring invasion sites. We propose that lures might be improved through studying the olfactory system of adult beetles, especially the gene family of odorant receptors (ORs) and the structure of the antennal lobes of the brain. Here, we report identification of 132 ORs in the genome of A. glabripennis (inclusive of one Orco gene and 11 pseudogenes), some of which are orthologous to known pheromone receptors of other cerambycid beetles. We also identified three ORs that are strongly biased toward expression in the female transcriptome, and a single OR strongly biased toward males. Three-dimensional reconstruction of the antennal lobes of adults suggested a male-specific macroglomerulus and several enlarged glomeruli in females. We predict that functional characterization of ORs and glomeruli will lead to identification of key odorants in the life history of A. glabripennis that may aid in monitoring and controlling future invasions.

Introduction The Asian longhorned beetle, Anoplophora glabripennis (Motchulsky), is a destructive insect native to Asia that has repeatedly become established in urban areas of North America and Europe (Haack et al. 2010). A. glabripennis is considered to be among the most serious forestry pests on account of its ability to infest and kill healthy trees of numerous genera, including widely planted species of Acer, Salix, and Populus (Hu et al. 2010). To date, invasions have been demarcated and monitored using traps baited with an attractive mixture of volatile organic compounds (VOCs) produced by the host and two components of the male-produced aggregation-sex pheromone, 4-(n-heptyloxy)butan-1-ol and 4-(n-heptyloxy)butanal (e.g., Nehme et al. 2010, 2014). This mixture, however, shows low efficacy of attraction (<100 specimens over multiple seasons; Meng et al. 2014; Nehme et al. 2014) compared to that of host

VOC/pheromone blends that target related genera such as Monochamus (e.g., >2000 specimens, Allison et al. 2012), which suggests that a more optimal mixture or attractant likely remains to be discovered. Female-associated VOCs (Wickham et al. 2012), a trail pheromone (Hoover et al. 2014), and a putative third component of the male-produced aggregation pheromone (Crook et al. 2014) have been recently identified, although their effectiveness in monitoring invasion sites has yet to be tested. We propose that the discovery of additional attractive VOCs might be hastened through studies of the olfactory system of A. glabripennis. Neopteran insects detect VOCs by means of sensory receptor neurons expressing olfactory receptors (ORs), ionotropic receptors (IRs), and a few gustatory receptors (Leal 2013; Missbach et al. 2014). The ORs appear to be primarily responsible for detection of VOCs (Leal 2013), while the IRs appear to be sensitive to VOCs as well as abiotic factors such as humidity and temperature (Benton et al. 2009; Min et al. 2013; Enjin et al. 2016). In insects, ORs are usually expressed in the antennae, where a single OR gene and an Orco co-receptor are expressed on the dendrites of an individual odorant-sensitive receptor neuron (OSN) (Vosshall and Stocker 2007). In some situations, multiple ORs can be expressed on the same OSN (Couto et al. 2005; Fishilevich and Vosshall 2005; Goldman et al. 2005). ach OSN pro ects its axon to the antennal lobe (AL) of the brain, and the axons of each OSN type converge in one of the discrete, condensed neuropil structures in the AL the olfactory glomeruli (Martin et al. 2011). Thus, each glomerulus represents the combined inputs of OSNs expressing the same OR gene, or same sets of OR genes (c.f., Koutroumpa et al. 2014). The size of the glomerulus is thus correlated to the number of associated OSN axons, which in turn may reflect the importance of the odorant information it receives with respect to the life history of the insect (Dekker et al. 2006), especially when the glomerulus is involved in detecting pheromones (e.g., Christensen et al. 1995). By functionally characterizing the ORs and/or glomeruli, we may identify novel volatiles that can influence the behavior of an insect species. Here, we present the OR family of A. glabripennis, annotated from the recent Asian longhorned beetle genome project (McKenna et al. 2016). We compare these OR genes in the transcriptomes of male and female A. glabripennis to identify potential ORs that may be biased toward expression in one sex. Finally, we present an initial map of the AL of male and female A. glabripennis in order to identify enlarged glomeruli, which may be sites of primary neural processing of sensory information about key odorants in the life history of this invasive pest.

Methods and Materials Annotation of ORs Genomic data were made available through the Asian longhorned beetle genome project (McKenna et al. 2016). Using TBLASTN (Altschul et al. 1997), we searched the first draft of the A. glabripennis genome (NCBI: BioProject PRJNA167479) against a database of available coleopteran OR sequences: Tribolium castaneum (Herbst) (Engsontia et al. 2008); Megacyllene caryae (Gahan) (Mitchell et al. 2012); and Ips typographus (L.) and Dendroctonus ponderosae (Hopkins) (Andersson et al. 2013). All of these beetles belong to a single species-rich clade

(series Cucujiformia) containing ~190,000 described extant species (McKenna et al. 2015). Models of ORs were constructed from a combination of predictions by MAKER 2.0 (Canterel et al. 2008), transcriptome support (McKenna et al. 2016), and manual annotation in Geneious R6.1 (Biomatters, Ltd., Auckland, NZ). We discarded most partial models that were <100 amino acids in length or consisted of only a single apparent exon, to avoid designating separated exons of one receptor with two names. Pseudogenes were designated when regions of the genome presented strong sequence similarity to a receptor gene, but with nonsense mutations and/or missing splice site sequences. Peptide sequences of the A. glabripennis O s were aligned using M SCL (10 iterations; gap score 3; Edgar 2004) and subsequent manual adjustment. ORs of A. glabripennis were separately aligned with the three functionally characterized ORs from M. caryae, the OR genes available from the genome of T. castaneum, and published ORs from transcriptomes of three other major beetle families: Chrysomelidae (Ambrostoma quadriimpressum Motchulsky; Wang et al. 2016), Scarabaeidae (Anomala corpulenta Motchulsky; Li et al. 2015), and Curculionidae (D. ponderosae). Only genes longer than 100 amino acids were included in this analysis. Phylogenetic trees were constructed from the alignments using FastTree 2.1 at its default settings (Price et al. 2010) with nodal support values based on a Shimodaira-Hasegawa test (Shimodaira and Hasegawa 1999). Trees were visualized and edited in FigTree v1.4.2 (Rambaut 2014). OR genes from A. glabripennis were designated with the prefix AglaOR and numbered consecutively down the tree so that in most cases, genes similar in number also share a predicted phylogenetic placement. The AglaOR1 designation was reserved for the conserved OR coreceptor (Orco) gene, following Engsontia et al. (2008). Gene names were provided suffix codes of three letters to indicate missing regions of the model (NTE, INT, CTE for N-terminal, internal, or C-terminal gaps), or single letters in the case of multiple missing exons (e.g., NC = NTE + CTE). We assigned OR genes to subfamilies based on their phylogenetic placement and following the naming scheme initiated by Engsontia et al. (2008). This scheme divides ORs among seven major subfamilies (dubbed “Group 1”, “Group 2”, etc.), six of which were annotated from the genome of T. castaneum (Engsontia et al. 2008). The final subfamily in the scheme, Group 7, was subsequently identified from several other beetle families (Mitchell et al. 2012; Andersson et al. 2013; Li et al. 2015) but apparently is absent in T. castaneum.

Expression of ORs We approximated the expression of each OR by searching the unassembled reads of whole-body transcriptomes of an adult male and female A. glabripennis (McKenna et al. 2016). We recorded 100% matches to peptide sequences of each OR and four housekeeping genes (cytoplasmic actin, EF1-alpha, and ribosomal proteins L32 and S3). Hits were recorded only in the event of a perfect match. We took the number of reads as a rough estimate of transcriptional activity. This number was scaled to the read count of each housekeeping gene to generate an average ratio of reads between sexes (female/male), which was expressed as log10 to indicate

female vs. male expression (positive for female, negative for male). Sex-biased transcriptional activity was estimated by observing for genes that were present in a single sex, or, for genes present in both sexes, log ratios that exceeded two standard deviations from the mean activity (exceeding 95% of the expected distribution).

Source of insects for AL morphology A. glabripennis adults were reared separately at the USDA-APHIS CPHST Lab (“Otis Lab”; Otis ANGB, MA, SA) and Pennsylvania State niversity (“PS ”; niversity Park, PA). Beetles at the Otis Lab were reared on a diet similar to that of Dubois et al. (2002) for ~3 mo, chilled at 10 °C for 11 wk, and allowed to complete development on the same diet. Adults were fed twigs of striped maple (Acer pensylvanicum L.) until dissection. Beetles were reared at PSU on a pourable modification of a diet designed for Enaphalodes rufulus (Haldeman) for 90 days (Keena 2005), chilled at 10 °C for 90 days, and allowed to develop until pupation on the same diet. Upon pupation they were transferred to 50 ml Falcon centrifuge tubes and incubated at 27.5 °C until adult eclosion. Adults were fed red maple (Acer rubrum L.) twigs until dissection.

Preparation of ALs We initially attempted to stain glomeruli with nc82 (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, USA), an antibody with an affinity for synaptic neuropils in insects (Wagh et al. 2006). However, the antibody failed to detect a target antigen in the ALs of A. glabripennis. We instead visualized the glomeruli by fixing with glutaraldehyde, which produces fluorescence throughout the AL, but such that the more condensed neuropil of the glomeruli stand out against other tissues (e.g., Sombke et al. 2012). Live, adult beetles of both sexes were decapitated with a razor blade, and the brain was accessed either directly through the exposed occipital foramen, or by cutting along the gena and frons. Brains were carefully dissected from the head capsule under 2% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA) in 0.1M Sorensen’s phosphate buffer (pH 7.4; lectron Microscopy Sciences) and, using forceps, cleaned of trachea, fat deposits, and other accessory tissues that adhered to the surface. Dissections at PSU were conducted in Tucson Ringer (Waldrop et al. 1987), but subsequently incubated in glutaraldehyde solution. Paired ALs were located prominently on the ventral side of the brain relative to the neuraxis, which lies against the dorsal-anterior surface of the head capsule. We discarded samples if these structures were visibly damaged. Successful dissections were fixed in fresh glutaraldehyde solution at 1-3 °C for at least 48 hours, dehydrated by a series of 10-min treatments in solutions of increasing ethanol concentration, and cleared in methyl salicylate (Sigma-Aldrich, St. Louis, MO).

Imaging and analysis of ALs Prepared brains of four male and four female beetles were imaged on a Zeiss Meta 510 LSM confocal microscope (Carl Zeiss AG, Jena, Germany) or an Olympus FluoView FV1000 confocal microscope (Olympus Corporation, Center Valley, PA). Brains were oriented so that

the ventral surface of the brain, and the ALs, were facing the microscope objective and imaged on a transverse plane. Consistent positioning of the brain at this stage was critical for comparing later reconstructions of the AL. This preparation was illuminated by a HeNe laser at 633 nm (Otis) or a combination of 543 nm and 405 nm (PSU). Excitation of the glutaraldehyde background alone was sufficient to visualize glomeruli, which appeared as bright, discrete clusters in the ALs. Image stacks were loaded into the software package Reconstruct (v1.1.0.0; Fiala 2005) and the outline of each glomerulus was manually traced on each section of every stack to create a three-dimensional model of each AL. We compared the resulting three-dimensional models and identified prominent glomeruli of consistent position and size in each brain as landmarks. Some landmark glomeruli appeared consistently greater in size than other glomeruli in the AL, and we noted these as “enlarged” if they exceeded 90% of the range of glomerular size in at least two individuals of either sex, and “macroglomeruli” if they exceeded three times the interquartile range (extremely conservative outliers; Hoaglin et al. 1988). Remaining glomeruli were classified by their spatial relationship to landmark glomeruli. Glomeruli were classified and named following Ghaninia et al. (2007). Landmark glomeruli were designated as Class 1 and the other spatially consistent glomeruli as Class 2, while the remaining structures that varied in location or presence across individuals were designated as Class 3. We named glomeruli by their spatial relationship to the neuraxis, with three capital letters indicating the position relative to each axis: anterior/posterior (A, P), ventral/dorsal (V, D), lateral/medial (L, M), or central (C) (Ghaninia et al. 2007). All subsequent description of the brain is presented relative to the neuraxis unless otherwise specified.

Results Annotation of ORs. We annotated 132 genes with sequences similar to the family of insect ORs (Fig. 2-1, Online Resource 1A), comprising the single expected ortholog to the Orco gene (Vosshall and Hansson 2011; AglaOR1\Orco), eleven apparent pseudogenes (AglaOR22, 23, 34, 67, 75, 83, 88, 104, 111, 119, and 122; designated by suffix PSE), and 120 ORs. Eighty-one of these receptors were annotated with full-length ORFs in the genome. Sixteen additional ORs were nearly complete and lacked only a short initial or terminal exon (Online Resource 1A). The remaining 24 receptors are likely to be complete genes, but with one or more exons missing in unassembled regions of the genome (Online Resource 1A). We were conservative in assigning names to receptor fragments, because fragments with very high sequence similarity to full-length AglaORs may be assembly errors involving different alleles of the same gene. OR models and sequences may be viewed as part of the A. glabripennis genome project (McKenna et al. 2016; NCBI: BioProject PRJNA167479), and FASTA files of peptide and nucleotide sequences are included here as supplementary materials (Online Resources 1B, C).

The Anoplophora ORs could be placed almost entirely within the seven established groups of coleopteran ORs (Figs. 2-1, 2.; Engsontia et al. 2008; Mitchell et al. 2012; McKenna et al. 2016). Groups 1 (AglaOR54-68), 3 (AglaOR73-100), and 7 (AglaOR37-53) were clearly defined with high support values. In contrast, Groups 4, 5, and 6 placed within a single, well- supported lineage of AglaORs (AglaOR104-132), defined by the outgroup of AglaOR101-103. Group 2 was only weakly supported, but consisted of two major subgroups that included AglaOR2-28 and AglaOR29-36. AglaOR69-72 was recovered in a position separate from the seven established groups. When receptors from other beetle lineages were included (Fig. 2-2), AglaOR71-2 resolved as members of the 3A subgroup (Engsontia et al. 2008). AglaOR69-70 apparently were not part of any existing clade of ORs, but grouped with some OR genes in the beetle families Curculionidae (weevils) and Chrysomelidae (leaf beetles), both near relatives of the Cerambycidae.

Fig. 2-1. Unrooted phylogram illustrating the OR gene family of Anoplophora glabripennis. Pseudogenes are indicated by the suffix PSE, and other suffixes indicate missing exons (see text). Different colors indicate the major groups of coleopteran receptors, with Groups 4-6 depicted as a single radiation. Receptors in black did not correspond to a previously established group. Numbers on nodes indicate Shimodaira-Hasegawa support values >0.90.

Fig. 2-2. Unrooted cladogram illustrating the relationships among known ORs from five beetle families: Cerambycidae (Anoplophora glabripennis, “Agla”, red); Chrysomelidae (Ambrostoma quadriimpressum, “Aqua”, teal), Scarabaeidae (Anomala corpulenta, “Acor”, pink), Curculionidae (Dendroctonus ponderosae, “Dpon”, yellow), and Tenebrionidae (Tribolium castaneum, “Tcas”, blue). Receptor sequences with <100 amino acids were excluded from the analysis. Three additional receptors, coded in black, have been functionally characterized from the longhorned beetle Megacyllene caryae (Cerambycidae). Colors and numbers around the perimeter indicate the seven major groups of coleopteran receptors, but with Groups 4-6 labeled only in Tribolium. Large expansions of receptors have been collapsed to improve clarity and are color-coded by beetle family. Numbers alongside major nodes indicate Shimodaira-Hasegawa support values.

Seven AglaOR genes were similar to three functionally characterized ORs from the cerambycid beetle Megacyllene caryae, included in Fig. 2-2. AglaOR29 was orthologous to

McarOR3, a receptor sensitive to the compound 2-methylbutan-1-ol, while AglaOR35-36 were sister to McarOR5, which is sensitive to 2-phenylethanol (Mitchell et al. 2012). Finally, AglaOR65-68 formed a small clade that also included McarOR20, a receptor sensitive to stereoisomers of 2,3-hexanediol and 3-hydroxyhexan-2-one (Mitchell et al. 2012). The above are all volatile compounds produced by male M. caryae as part of this species’ aggregation-sex pheromone (Lacey et al. 2008; Mitchell et al. 2012).

Expression of ORs Of the 120 putative ORs, 89 were present as at least one unique read in the transcriptomes (Fig. 2-3, Online Resource 2). Most ORs were present minimally, with a median of only four reads. Nine ORs were recovered only from the male transcriptome and 36 ORs were recovered only from the female, but only two of these sex-specific ORs (AglaOR99 and AglaOR43; female) were recovered as more than ten reads (Fig. 2-2). Thus, the sex bias of the other receptors is questionable, because it is possible they may have been present but not sequenced from the transcriptome of the other sex. Forty-four receptors were transcribed in both sexes and overall were slightly biased toward the female transcriptome, with an average log ratio of 0.29. Log ratios of 1.62 and -1.02 were calculated as cutoffs for bias in females and males, respectively. AglaOR18, 25, and 28 exceeded the threshold for a bias in females (log ratios 1.85, 1.99, 1.81), and AglaOR16 exceeded the threshold in males (-1.56). AglaOR102, 38, and 39 were not biased toward either sex, but presented a notably high read count (>100) relative to other ORs in the transcriptomes of both sexes. The actual expression of AglaOR38-39 is unclear because many reads corresponded to a region of shared sequence, but unique reads were assigned to both genes. Thirty-two genomic ORs were not recovered by the transcriptomes, and are probably associated with other developmental stages that were not targeted in this study (e.g., larva; Engsontia et al. 2008).

Fig. 2-3. Abundance and sex bias of ORs in male and female Anoplophora glabripennis, as measured by the presence of unique reads of each receptor in the transcriptome. Reads present in males and females are reported according to their normalized log ratio. The solid line above the x-axis is the mean of the distribution of ratios, and dashed lines are two standard deviations above and below the mean. Solid circles are receptors exceeding two standard deviations. Small boxes at the edge of the y-axis include reads that were present only in females (above) or males (below). Prominent receptors are indicated by the number corresponding to their gene name (e.g., 16 = AglaOR16).

Imaging and analysis of ALs Glomeruli were clearly visible in the ventral sections of the ALs of both sexes (Fig. 2- 4A), but the signal strength decreased as imaging proceeded dorsally (Fig. 2-4B), leaving some uncertainty to the form and structure of the glomeruli deeper in the AL. The visible glomeruli presented as a hollow, spherical cluster, consistent with observations of isomorphic glomeruli in other insect species (Rospars and Hildebrand 2000; Dreyer 2010). We identified an average of 60 (range ±2) glomeruli in females, with the median size of glomeruli in an individual beetle ranging between 55,124–79,568 µm3. Males presented 55±3 glomeruli, with a median size range of 64,022–90,570 µm3 (Fig. 2-5). Many glomeruli were inconsistent in location, size, and even presence among individuals in the study, hindering our comparison of the structures between sexes. We ultimately named 18 glomeruli that appeared to be consistent in their large size and/or general location between the sexes, and two additional glomeruli that were apparently unique to females. These glomeruli are

described briefly below and illustrated in Figs. 2-4C and 2-4D. Three-dimensional reconstructions, confocal image stacks, and additional descriptions are available in the supplementary materials (Online Resources 3-5). In both sexes, the AL was clearly bounded by a crest of four class 1 glomeruli along its posterior surface (PCC1-4; Fig. 2-4C,D). The large glomerulus PVL1 was situated directly anterior to this crest and acted as a landmark glomerulus to define the class 2 cluster of PVL2-5. Another large class 1 glomerulus, PVM1, was situated medial to PVL1 and, in females, was tightly appressed to two class 2 glomeruli, PVM2 and 3. The class 1 glomerulus CVC1 was positioned at the ventral apex of the AL, and CVC2-3 were immediately posterior and anterior to CVC1. The final class 1 glomerulus, ACD1, was situated at the anterior margin of the AL, near the antennal nerve. Four class 2 glomeruli, AMD1-4, were defined by their proximity to CVC3 and ACD1.

Fig. 2-4. Glomeruli in the left AL of adult Anoplophora glabripennis, with orientation given relative to the neuraxis (A, anterior; P, posterior; D, dorsal; V, ventral; M, medial; L: lateral; AN: antennal nerve). Ventral sections of the lobe (A) were more distinct than the deeper, dorsal sections (B). Eighteen named glomeruli (colored) are mapped onto three-dimensional reconstructions of the ALs of a male (C) and a female (D). Reconstructions are dorsoventrally doubled in length to better visualize the glomeruli.

Male and female beetles also differed in their apparent macroglomeruli (Fig. 2-5A,B). PVL1 exceeded the threshold for a macroglomerulus in one brain of a female and was greatly enlarged in the remaining three (Fig. 2-5A). This glomerulus, though larger than average in males, was never more than twice the median size. In contrast, we classified ACD1 as a macroglomerulus in males because it greatly exceeded our threshold in all four specimens (Fig.

2-5B), but was mostly indistinct in females (with one exception; Fig. 2-5A, Sample 4). PVM1 exceeded this threshold in a single male, and was enlarged in two other males and a female, suggesting it as a second potential macroglomerulus. Additionally, CVC1 and CVC3 were enlarged in multiple males, while PVL2, PVL3, PVM2, and PCC2 were enlarged in females, and PVL4 was enlarged in one individual of each sex.

Fig. 2-5. Volumes of glomeruli in four female (A) and four male (B) Anoplophora glabripennis. The grey box encompasses the first and third quartiles and the median is indicated by a solid horizontal line. Vertical lines extending from the box indicate 90th and 10th percentiles. Dashed horizontal lines indicate three times the interquartile range. Glomeruli exceeding the 90th percentile in size are indicated as black circles.

Conclusions Our genomic and neuroanatomical survey describes a potentially diverse olfactory repertoire for A. glabripennis and highlights several receptors that may play a key role in the ecology of this invasive pest. Notably, the single receptor AglaOR16 was strongly biased towards males, and a single glomerulus in the ALs of males (ACD1) was vastly expanded relative to other glomeruli. Highly expressed, male-biased receptors may be associated with pheromone production by females (Wanner et al. 2007). Furthermore, a male-specific macroglomerular complex (MGC) in many lepidopteran species receives and processes signals from OR neurons that are specialized to detect the key components of a female-produced pheromone (Christensen et al. 1995; Namiki et al. 2014). Brains of A. glabripennis males did not contain a distinct MGC, in that we did not observe the strong sexual dimorphism in both shape and size that defines this region in Lepidoptera (Rospars and Hildebrand 2000). Instead, each macroglomerulus observed here appears to correspond to a similar but diminished counterpart in the opposite sex, perhaps suggesting a conserved olfactory function, but with differing sensitivity or downstream circuitry (e.g., Datta et al. 2008). Nevertheless, the male-specific macroglomerulus, coupled with the single male-biased OR, hints at a female-produced sex pheromone: perhaps one or more of the oxidized aldehydes that are associated with virgin females and attractive to males (Wickham et al. 2012). Similarly, three receptors were more common in the female transcriptome, and female brains contained a potential macroglomerulus (PVL1). This may be sensitive to the male-produced pheromone components, or to other odorants important to females such as oviposition cues. The size of a glomerulus can correlate to its importance in the life history of an insect (Ibba et al. 2010). We identified several enlarged glomeruli that may thus be responsive to volatiles that strongly influence the behavior of A. glabripennis. Many of the largest glomeruli were located near the surface of the AL, potentially making them more accessible to future characterization via calcium imaging (Galizia et al. 1999) or neural recording and dye injection (Reisenman et al. 2005). A single, unnamed glomerulus deeper in the AL was also enlarged (Fig. 2-5B, Sample 3), but because of limitations in imaging, some dorsal glomeruli were less distinct and we were unable to consistently locate them or assign names. Future surveys of the AL with more powerful imaging equipment (e.g., multiphoton microscopy) or effective antibody stains will be necessary for better characterization of this region. Nevertheless, the total number of glomeruli identified in our analysis was relatively consistent across specimens. The number of glomeruli in the AL can approximate the number of ORs expressed in the antennae (Vosshall and Stocker 2007), and indeed we identified approximately 55 glomeruli in the brains of males, and reads corresponding to 52 ORs. However, although the most complex female AL contained 62 glomeruli, we identified 79 ORs from this transcriptome. One explanation for this is that some glomeruli were not successfully mapped or visualized. In particular, the dorsolateral portion of the AL (Fig. 2-4B, bottom right) transitioned into a region of indiscrete structures, which were occasionally difficult to distinguish from “true” glomeruli. This region probably marks the boundary of the antennal mechanosensory

and motor center (Homberg et al. 1989), but may include additional structures of the olfactory neuropil. Some ORs may also be co-expressed in the same neurons (Couto et al. 2005; Fishilevich and Vosshall 2005; Goldman et al. 2005; Koutroumpa et al. 2014), which is another possible reason for the lack of a strict 1:1 ratio of ORs to glomeruli. Additionally, some OR genes may be expressed elsewhere in the body (e,g, Engsontia et al. 2008), and thus not linked to the deutocerebrum, and such site specificity would have been overlooked by the whole-body transcriptomes conducted here. The Anoplophora ORs are only the second complete set of receptors annotated from a coleopteran genome, and thus offer the first proper comparison to the unusually large olfactory suite of Tribolium castaneum (Engsontia et al. 2008). Our analysis confirms speculation by Engsontia et al. (2008) that the prominent expansions of Tribolium are lineage-specific, and in fact, suggests that the major OR Groups 4, 5, and 6 might be better unified as a single group of receptors, along with the recently-proposed Group 9 (Antony et al. 2016). This unified group is defined by a small outgroup of AglaOR101-103, which also includes conserved representatives from other beetles including TcasOR275 (Engsontia et al. 2008) and McarOR44 (Mitchell et al. 2012). Unfortunately, few members of Groups 4-6 were present in the published transcriptomes of chrysomelid, scarabaeid, and curculionid beetles, so additional beetle genomes must be annotated to further elucidate the extent of this family. The remaining Groups 1, 2, 3, and 7 were generally well-maintained. Group 3 included the largest expansion of AglaORs, including a tandem array of AglaOR89-98 that shared a high degree of sequence similarity and may be recent duplications. This cluster may thus be under active selection in A. glabripennis, although no members were prominent in the transcriptomes. Group 2 is recovered with strong statistical support in other studies (Engsontia et al. 2008; Andersson et al. 2013; Li et al. 2015; RFM unpub. data), so its poor support in the present study is questionable, but it may reflect a division that will become evident as additional beetle genomes are sequenced. The inclusion of TcasOR71-72 in Group 7 is also probably in error, because their placement is inconsistent in studies to date (Engsontia et al 2008; Wang et al. 2016; RFM unpub. data). Some authors have even gone so far as to entirely remove them from analyses (Andersson et al. 2013; Li et al. 2015). It is possible TcasOR71-72 are remnants of a separate, diminished group of receptors in Tribolium that are lost in Anoplophora and the other beetle families included in our analyses. In fact, our phylogeny suggests two such groups of coleopteran ORs that are separate from the seven primary radiations. The enigmatic Group 3A, identified in the original Tribolium genome project by TcasOR167, is joined by AglaOR71-72, and its relationship with Group 3 is poorly supported here. Group 3A also includes a large expansion of scarab (staphyliniform) genes, suggesting it may be a lineage that is diminished in cucujiform beetles. Similarly, we identified a small, orphaned family of receptors including AglaOR69-70, AquaOR11-12, and DponOR15. Again, genomic data from additional families and suborders of beetles will be necessary to properly define these potentially novel receptor groups. We therefore refrain from naming them here.

The ligands of coleopteran ORs remain largely unknown. To date only the three receptors McarOR3, 5, and 20 have been characterized, and they are sensitive to pheromone components produced by the cerambycid beetle M. caryae (Mitchell et al. 2012). McarOR3 and McarOR5 were highly similar to AglaOR29 and AglaOR35-36, suggesting that these genes may be related to pheromone biology. However, the pheromone components in question (2-methylbutan-1-ol and 2-phenylethanol) are also widespread floral compounds (El Sayed 2016) with little structural similarity to the known pheromone components of A. glabripennis. Neither of these receptors appeared to be highly expressed in A. glabripennis, in contrast to what might be expected for a receptor sensitive to pheromone components (Wanner et al. 2008; Mitchell et al. 2012), but quantitative PCR of antennal RNA will be necessary to clearly establish the relative expression of these and other OR genes identified in this study. McarOR20 is sensitive to the conserved pheromone components 2,3-hexanediol and 3- hydroxyhexan-2-one produced by many species in the cerambycid subfamilies Cerambycinae and Prioninae (Millar and Hanks 2016), and it is interesting that McarOR20 shares similarity with AglaOR65-68. These are not known to be produced as pheromone components by species in the cerambycid subfamily Lamiinae (including A. glabripennis); nonetheless, AglaOR65-68 may be tuned to detect similar pheromone components produced by the many sympatric cerambycine and prionine species in Asia (Wickham et al. 2014, 2016), perhaps to enforce mating isolation or aid in identifying host plant material. We employed a promising new approach for elucidating the olfactory biology of a devastating forest pest, and it has yielded numerous candidate receptors that may be associated with the detection of pheromones or other behaviorally-active volatiles. Our data suggest the presence of a key female-produced volatile that is detected by males and has perhaps not been chemically or behaviorally characterized. None of the recently characterized components of the trail sex pheromone produced by females (Hoover et al. 2014) seem to be implicated here as the missing female volatile, because males appear to use mostly their maxillary and labial palps and not their antennae for detecting and responding to this pheromone (Graves 2016). Functional characterization of receptors such as AglaOR16 should be given top priority because they may reveal A. glabripennis pheromone components or other attractants and improve efforts to monitor for invasive populations of the Asian longhorned beetle worldwide.

Acknowledgments We thank Stephen Richards (Baylor College of Medicine Human Genome Sequencing Center) and other members of the Insect 5,000 Genomes Asian Longhorned Beetle Genome Consortium for access to genomic and transcriptomic data. We also thank Patty Jansma, Hong Lei, and Kim Lance at the University of Arizona for assistance with brain imaging. Additional thanks to Carrie Crook and David Lance at the USDA-APHIS CPHST Lab for providing live specimens of Anoplophora glabripennis. Funding was provided through an NIH postdoctoral training grant (5K12 GM000708-15) to RFM, the University of Memphis FedEx Institute, U.S.

NSF grant DEB1355169 and USDA-APHIS cooperative agreement 15-8130-0547-CA to DDM, and USDA-APHIS cooperative agreement 15-8130-1430-CA to TCB.

References Allison JD, McKenney JL, Millar JG, McElfresh JS, Mitchell RF, Hanks LM (2012) Response of the woodborers Monochamus carolinensis and Monochamus titillator (Coleoptera: Cerambycidae) to known cerambycid pheromones in the presence and absence of the host plant volatile α-pinene. Environ Entomol 41:1587-1596 Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389-3402 Andersson MA, Grosse-Wilde E, Keeling CI, Bengtsson JM, Yuen MMS, Li M, Hillbur Y, Bohlmann J, Hansson B, Schlyter F (2013) Antennal transcriptome analysis of the chemosensory gene families in the tree killing bark beetles, Ips typographus and Dendroctonus ponderosae (Coleoptera: Curculionidae: Scolytinae). BMC Genomics 14:198 Antony B, Soffan A, Jakše J, Abdelazim MM, Aldosari SA, Aldawood AS, Pain A (2016) Identification of the genes involved in odorant reception and detection in the palm weevil Rhynchophorus ferrugineus, an important quarantine pest, by antennal transcriptome analysis. BMC Genomics 17:69 Benton R, Vannice KS, Gomez-Diaz C, Vosshall LB (2009) Variant ionotropic glutamate receptors as chemosensory receptors in Drosophila. Cell 136:149-162

Cantarel BL, Korf I, Robb SM, Parra G, Ross E, Moore B, Holt C, Sanchez Alvarado A, Yandell M (2008) MAKER: an easy-to-use annotation pipeline designed for emerging model organism genomes. Genome Res 18:188-96

Christensen TA, Harrow ID, Cuzzocrea C, Randolph PW, Hildebrand JG (1995) Distinct projections of two populations of olfactory receptor axons in the antennal lobe of the sphinx moth Manduca sexta. Chem Senses 20:313-323 Crook DJ, Lance DR, Mastro VC (2014) Identification of a potential third component of the male-produced pheromone of Anoplophora glabripennis and its effect on behavior. J Chem Ecol 40:1241-1250 Couto A, Alenius M, Dickson BJ (2005) Molecular, anatomical, and functional organization of the Drosophila olfactory system. Curr Biol 15:1535-47. Datta RD, Vasconcelos ML, Ruta V, Luo S, Wong A, Demir E, Flores J, Balonze K, Dickson BJ, Axel R (2008) The Drosophila pheromone cVA activates a sexually dimorphic neural circuit. Nature 452:473-477 Dreyer D, Vitt H, Dippel S, Goetz B, el Jundi B, Kollmann M, Huetteroth W, Schachtner J (2010) 3D standard brain of the red flour beetle Tribolium castaneum: a tool to study metamorphic development and adult plasticity. Front Syst Neurosci 4:3

Dubois T, Hajek AE, Smith S (2002) Methods for rearing the Asian longhorned beetle (Coleoptera: Cerambycidae) on artificial diet. Ann Entomol Soc Am 95:223-230 Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32:1792-97 El-Sayed AM (2016) The Pherobase: Database of Pheromones and Semiochemicals. http://www.pherobase.com. Accessed 18 June 2016 Engsontia P, Sanderson A, Cobb M, Walden KKO, Robertson HM, Brown S (2008) The red flour beetle’s large nose: an expanded odorant receptor gene family in Tribolium castaneum. Insect Biochem Mol Biol 38:387-397 Enjin A, Zaharieva EE, Frank DD, Mansourian S, Suh GSB, Gallio M, Stensmyr MC (2016) Humidity sensing in Drosophila. Curr Biol: 26:1352-1358Fiala JC (2005) Reconstruct: a free editor for serial section microscopy. J Microsc 218:52-61 Fishilevich E, Vosshall LB (2005) Genetic and functional subdivision of the Drosophila antennal lobe. Curr Biol 15:1548-53 Galizia CG, Sachse S, Rappert A, Menzel R (1999) The glomerular code for odor representation is species specific in the honeybee Apis mellifera. Nature Neurosci 2:473-478 Ghaninia M, Hansson BS, Ignell R (2007) The antennal lobe of the African malaria mosquito, Anopheles gambiae – innervation and three-dimensional reconstruction. Arthropod Struct Dev 36:23-39 Goldman AL, Van der Goes van Naters W, Lessing D, Warr CG, Carlson JR (2005) Coexpression of two functional odor receptors in one neuron. Neuron 46:661-666 Graves FB (2016) Sensory aspects of trail-following behaviors in the Asian longhorned beetle, Anoplophora glabripennis. Master’s Thesis, Pennsylvania State University Haack RA, Hérard F, Sun J, Turgeon JJ (2010) Managing invasive populations of Asian longhorned beetle and citrus longhorned beetle: a worldwide perspective. Annu Rev Entomol 55:521-546 Hoaglin DC, Iglewicz B, Tukey JW (1986) Performance of some resistant rules for outlier labeling. J Am Stat Assoc 81:991-999 Hoover K, Keena M, Nehme M, Wang S, Meng P, Zhang A (2014) Sex-specific trail pheromone mediates complex mate finding behavior in Anoplophora glabripennis. J Chem Ecol 40:169-180 Homberg U, Christensen TA, Hildebrand JG (1989) Structure and function of the deutocerebrum in insects. Annu Rev Entomol 34:477-501 Hu J, Angeli S, Schuetz S, Luo Y, Hajek AE (2009) Ecology and management of exotic and endemic Asian longhorned beetle Anoplophora glabripennis. Ag For Entomol 11:359-375 Ibba I, Angioy AM, Hansson BS, Dekker T (2010) Macroglomeruli for fruit odors change blend preference in Drosophila. Naturwissenschaften 97:1059–1066 Keena, MA (2005) Pourable artificial diet for rearing Anoplophora glabripennis (Coleoptera: Cerambycidae) and methods to optimize larval survival and synchronize development. Ann Entomol Soc Am 98:536-547

Koutroumpa FA, Kárpáti Z, Monsempes C, Hill SR, Hansson BS, Jacquin-Joly E, Krieger J, Dekker T (2014) Shifts in sensory neuron identity parallel differences in pheromone preference in the European corn borer. Front Ecol Evol 2:65 Lacey ES, Moreira JA, Millar JG, Hanks LM (2008) A male-produced aggregation pheromone blend consisting of alkanediols, terpenoids, and an aromatic alcohol from the cerambycid beetle Megacyllene caryae. J Chem Ecol 34:408-417 Leal WS (2013) Odorant reception in insects: roles of receptors, binding proteins, and degrading enzymes. Annu Rev Entomol 58:373-391 Li X, Ju Q, Jie W, Li F, Jiang X, Hu J, Qu M (2015) Chemosensory gene families in adult antennae of Anomala corpulenta Motschulsky (Coleoptera: Scarabaeidae: Rutelinae). PLoS ONE 10:e0121504 Martin JP, Beyerlein A, Dacks AM, Reisenman CE, Riffell JA, Lei H, Hildebrand JG (2011) The neurobiology of insect olfaction: sensory processing in a comparative context. Prog Neurobiol 95:427-447 McKenna DD, Scully ED, Pauchet Y, Hoover K, Kirsch R, Geib SM, Mitchell RF, Waterhouse RM et al (2016) Genome of the Asian longhorned beetle (Anoplophora glabripennis), a globally significant invasive species, reveals key functional and evolutionary innovations at the beetle-plant interface. Genome Biol 17:227 McKenna DD, Wild AL, Kanda K, Bellamy CL, Beutel RG, Caterino MS, Farnum CW, Hawks DC, Ivie MA, Jameson ML, Leschen RAB, Marvaldi AE, McHugh JV, Newton AF, obertson JA, Thayer MK, Whiting MF, Lawrence JF, Ślipiński A, Maddison D , Farrell BD (2015) The beetle tree of life reveals Coleoptera survived end Permian mass extinction to diversify during the Cretaceous terrestrial revolution. Syst Entomol 40:1-46 Meng PS, Trotter RT, Keena MA, Baker TC, Yan S, Schwartzberg EG, Hoover K (2014) Effects of pheromone and plant volatile release rates and ratios on trapping Anoplophora glabripennis (Coleoptera: Cerambycidae) in China. Environ Entomol 43:1379-1388 Millar JG, Hanks LM (2016) Chemical ecology of cerambycid beetles. In: Wang Q (ed) Cerambycidae of the world: biology and management. CRC Press/Taylor & Francis, Boca Raton (in press) Min S, Minrong A, Shin SA, Sug GSB (2013) Dedicated olfactory neurons mediating attraction behavior to ammonia and armines in Drosophila. Proc Natl Acad Sci USA 110: E1321- E1329 Missbach C, Dweck HKM, Vogel H, Vilcinskas A, Stensmyr MC, Hansson BS, Grosse-Wilde E (2014) Evolution of insect olfactory receptors. eLife 3:e02115 Mitchell RF, Hughes DT, Luetje CW, Millar JG, Soriano-Agatón F, Hanks LM, Robertson HM (2012) Sequencing and characterizing odorant receptors of the cerambycid beetle Megacyllene caryae. Insect Biochem Mol Biol 42:499-505 Namiki S, Daimon T, Iwatsuki C, Shimada T, Kanzaki R (2014) Antennal lobe organization and pheromone usage in bombycid moths. Biol Lett 10:20140096

Nehme ME, Keena MA, Zhang A, Baker TC, Xu Z, Hoover K (2010) Evaluating the use of male-produced pheromone components and plant volatiles in two trap designs to monitor Anoplophora glabripennis. Environ Entomol 39:169-176 Nehme ME, Trotter RT, Keena MA, McFarland C, Coop J, Hull-Sanders HM, Meng P, De Moraes CM, Mescher MC, Hoover K (2014) Development and evaluation of a trapping system for Anoplophora glabripennis (Coleoptera: Cerambycidae) in the United States. Environ Entomol 43:1034-1044 Price MN, Delah PS, Arkin AP (2010) FastTree 2 – Approximately maximum-likelihood trees for large alignments. PLoS ONE 5:e9490 Rambaut A (2014) FigTree v1.4.0, a graphical viewer of phylogenetic trees. Edinburgh: University of Edinburgh. Available at: http://tree.bio.ed.ac.uk/software/figtree/. Accessed 17 June 2016 Reisenman CE, Christensen TA, Hildebrand JG (2005) Chemosensory selectivity of output neurons innervating an identified, sexually isomorphic olfactory glomerulus. J Neurosci 25:8017-8026 Rospars JP, Hildebrand JG (2000) Sexually dimorphic and isomorphic glomeruli in the antennal lobes of the sphinx moth Manduca sexta. Chem Senses 25:119-29 Shimodaira H, Hasegawa M (1999) Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol Biol Evol 16:1114–1116 Sombke A, Lipke E, Kenning M, Müller CHG, Hansson BS, Harzsch S (2012) Comparative analysis of deutocerebral neuropils in Chilopoda (Myriapoda): implications for the evolution of the arthropod olfactory system and support for the Mandibulata concept. BMC Neurosci 13:1 Vosshall LB, Hansson BS (2011) A unified nomenclature system for the insect olfactory coreceptor. Chem Senses 36:497-498 Vosshall LB, Stocker RF (2007) Molecular architecture of smell and taste in Drosophila. Annu Rev Neurosci 30:505-533 Wagh DA, Rasse TM, Asan E, Hofbauer A, Schwenkert I, Dürrbeck H, Buchner S, Dabauvalle M-C, Schmidt M, Qin G, Wichmann C, Kittel R, Sigrist SJ, Buchner E (2006) Bruchpilot, a protein with homology to ELKS/CAST, is required for structural integrity and function of synaptic active zones in Drosophila. Neuron 49:833-44 Waldrop B, Christensen TA, Hildebrand JG (1987) GABA-mediated synaptic inhibition of projection neurons in the antennal lobes of the sphinx moth, Manduca sexta. J Comp Physiol A 161:23-32 Wanner KW, Nichols AS, Walden KKO, Brockmann A, Luetje CW, Robertson HM (2007) A honey bee odorant receptor for the queen substance 9-oxo-2-decenoic acid. Proc Natl Acad Sci USA 104:14383-14388 Wang Y, Chen Q, Zhao H, Ren B (2016) Identification and comparison of candidate olfactory genes in the olfactory and non-olfactory organs of elm pest Ambrostoma quadriimpressum (Coleoptera: Chrysomelidae) based on transcriptome analysis. PLoS ONE 11:e0147144

Wickham JD, Harrison RD, Lu W, Guo Z, Millar JG, Hanks LM, Chen Y (2014) Generic lures attract cerambycid beetles in a tropical montane rain forest in southern China. J Econ Entomol 107:259-267 Wickham JD, Millar JG, Hanks LM, Zou Y, Wong JCH, Harrison RD, Chen Y (2016) (2R,3S)- 2,3-octanediol, a female-produced sex pheromone of Megopis costipennis (Coleoptera: Cerambycidae: Prioninae). Environ Entomol 45:223-228 Wickham JD, Xu Z, Teale SA (2012) Evidence for a female-produced, long range pheromone of Anoplophora glabripennis (Coleoptera: Cerambycidae). Insect Sci 19:355-371

Chapter 3 Calcium Imaging Studies to Identify Glomeruli Activated by Male- Emitted Pheromone

Introduction The characterization made of the existence of sex-specific large glomeruli located at the base of the antennal nerve at the entrance to the antennal lobe (glomerulus ACD1) led to the idea that this glomerulus may be responsible for receiving input from OSNs tuned to one of the sex pheromone components of A. glabripennis. The cobalt-lysine staining technique that is so often used to identify target glomeruli by back-filling single, odorant-specific-characterized OSNs (Lee et al. 2006, Todd et al. 1995) was not possible because due to the anatomy of the sensilla making them both inaccessible and impervious to saline electrodes. Such access would normally allow electrodes to deliver dye to the OSNs for uptake and down their axons to the antennal lobe. Calcium imaging could allow us to observe which glomeruli are active when A. glabripennis uses its antenna to detect different odorants, particularly the ALB aldehyde and ALB alcohol. We collaborated with colleagues at the Max Planck Institute for Chemical Ecology in Jena, Germany to use calcium imaging equipment located there, thus I was able to work with Sonja Bisch-Knaden primarily, and also Silke Sachse, both world-known experts in calcium imaging techniques.

Calcium imaging works by using a molecule that fluoresces upon bonding with calcium ions. The molecule is bonded to a lipophilic molecule that allows the indicator, (“calcium-sensitive dye”) to enter into the neurons one wants to evaluate for its sensitivity to stimulation. Once inside the cell, the indicator is freed from the carrier by cell esterases and the indicator molecule is available to bind to calcium ions. These ions enter the neuron as it depolarizes and releases neurotransmitters. This allows us to observe neurons’ firing activity levels in real time using cameras sensitive to the specific wavelengths of light emitted by the fluorescing molecules, and then use computer-aided enhancements of the very fine changes in luminescence intensity and wavelength (Abel et al. 2001).

In order to observe the activity of neurons in the antennal lobe, the brain of a living, immobilized insect is exposed, bathed in the calcium-sensitive dye to allow uptake into cells, and then washed to remove any extracellular dye remaining on the surface. Odorants are passed over the antenna while a digital microscope records and interprets the changes in luminance in glomeruli in the antennal lobe on the side of the brain connected to that antenna (the ipsilateral antennal lobe) at the wavelength associated with the dye’s fluorescence optima.

This method has been utilized for many insect groups, including members of the orders Diptera, Lepidoptera and Hymenoptera, but at the time of these experiments, had never before been used

successfully on Coleoptera despite many attempts by expert laboratories to do so (B.S. Hansson, personal communication). No reason for this difficulty with beetles was known, but despite this, we deemed that there was a high reward-to-risk ratio to attempting to make this technique work. The possibility to answer such an interesting question about sex pheromone OSNs’ glomerular targets in such a definitive manner provided an impetus to attempt this study.

Twenty-two beetles were tested over the course of two weeks using several variations in technique and dosages of odorants known to elicit OSN responses as reported in the Baker Laboratory's SSR study described in Chapter 1. Unfortunately no glomerular reactions could be detected. I report here the various manipulations in methods tried so that others who might also deign to try calcium imaging on beetles might learn from these efforts and can experiment with variations in technique to perhaps get calcium imaging to work on this important and vast group of insects.

Methods A.glabripennis were shipped as adults from the USDA-APHIS facility in Buzzards Bay, Massachusetts to the Max Planck Institute in Jena, Germany. They were fed on fresh red maple, Acer rubrum, twigs that were also supplied by USDA-APHIS. The adult beetles were supplied with fresh water via cotton wicks and housed in a climate-controlled insectary at the study site.

Specimens to be imaged were immobilized using 50mL centrifuge tubes with the tips removed so that the head could protrude from the end of the tube. Dental wax was used to seal the opening around the head and create a well. Tucson Ringer was applied to the well and the exposed neural tissue was kept in this solution for the duration of experimentation unless otherwise noted (Fig. 3-1).

Figure 3-1. Beetle prepared for calcium imaging. The brain is exposed and a well of ringer solution is kept over the preparation.

An opening was then created to access the brain. This was accomplished using a blade breaker and iris scissors to cut along the edges of the frons and #1 forceps to remove the frons. The musculature and trachea were removed using #5 forceps to expose the antennal lobes (Fig. 3-2).

Figure 3-2. Detail of exposed antennal lobes (AL), showing the antennal nerve (AN) intact. The protocerebrum (PC) is also visible.

A calcium-sensitive dye (calcium green-2-AM dye) was dissolved in 20% Pluronic F-127 in dimethyl sulfoxide (Molecular Probes, Eugene, OR, USA), and diluted in Tucson Ringer solution to 30 mM. 20uL was then applied to the antennal lobes in a dark environment. Beetles were held in a dark, humidified chamber at either room temperature or at 4° C for varied amounts of time (Table 3-1) to allow the dye to gain entrance to neural cells. Ringer solution was used to remove extracellular dye by bathing the brain and then removing the rinsate two times before refilling the well with clean ringer solution.

Recordings were made in vivo after incubation and washing. A continuous charcoal-filtered and moistened airstream (500 ml/min) passed through a glass tube over the antenna (Fig. 3-3). A stimulus controller (SFC-2/b, Syntech, Kirchzarten, Germany) injected a 0.5 s puff (500 ml/min) through odor cartridges, prepared as in Chapter 2, into this glass tube. Odorants were presented in randomized order. Pipettes containing filter paper loaded with 10 ml of solvent were used as control. After presenting odorants, KCl was applied to the brain as a control to verify uptake of dye.

Image data was collected using an Olympus microscope (20x air objective NA 0.50; filter settings: dichroic 500 nm, emission LP 515 nm). The preparation was illuminated at 475 nm.

TILL PHOTONICS imaging software (Grafelfing, Germany) was used to record sequences and identify changes in intensity (Fig. 3-3).

Figure 3-3. ALB positioned for calcium imaging. On the left the tubing delivering clean air with odor cartridges inserted directed at the antennae of the beetle. The cartridges receive measured amounts of air from a computer- operated controller to deliver precise amounts of odorants. Above the beetle is the objective of the microscope extending down into the ringer solution (to reduce refraction). The objective contains fiber optics to deliver the measured amounts of wavelength-controlled light. The objective is focused on the antennal lobe and delivers images to the computer that controls and interprets the data.

Results and Discussion No calcium-sensitive dye responses were observed when different odorants were puffed over the antennae of any of the beetles tested. There was at least one false-positive due to movement of the beetle during recording. Several variations in incubation time, multiple incubations, incubation at room temperature and at 4° C were all attempted as shown in Table 3-1. High filter-paper dosages of the many different odorants were tried that had previously elicited high levels of spike frequency using single sensillum antennal recordings (Chapter 1) were employed. The addition of MK-571 during the incubation process was performed on final individuals tested on the theory that there may be cellular pumps removing the dye from the cells.

Table 3-1. Calcium imaging methodology permutations and results. incubation period incubation KCL effect (intensity Sex (min) temperature Odorants used Additional techniques change) 1 M 120 4 Geraniol 450ug NA 2 M 120 4 Geraniol 450ug NA 3 F 120 21 Geraniol 450ug none 4 M 120 21 Geraniol 450ug failed application 5 F 120 4 Geraniol 1000, 450, 100 ug AL pierced during incubation failed application 6 M 120 4 Geraniol 1000ug AL pierced during incubation 8% 7 F 120 4 Geraniol 1000ug none 8 M 120 4 Geraniol 1000ug 11% 9 F 240 4 Geraniol 1000ug AL pierced during incubation failed application 10 M 240 4 Geraniol 1000ug AL pierced during incubation 5% 11 M 240 4 Geraniol 1000ug 6% 12 F 240 4 Geraniol 1000ug MK571 5% 13 M 15,15 21 Geraniol 1000ug 2x15 min incubations, MK571 18% 14 M 5 21 Geraniol 1000ug MK571 18% 15 F 5 21 Geraniol 1000ug MK571 NA 16 F 10,10,10,10 21 Geraniol 1000ug 4x10 min incubations, Mk571 7% Geraniol 1000ug, ALB ald, ALB OH, 75% (due to 17 F 10,10 21 fuscomol, 100ug each MK571, very active beetle movement, we think) Geraniol 1000ug, ALB ald, ALB OH, 18 F 5 21 fuscomol, 100ug each MK571, very active beetle 10% Pheromone blend, Plant blend, plant MK571, very active beetle even 19 F 20 21 and pheromone blend after ablating abdomen too active Pheromone blend, Plant blend, plant MK571, 4 hours rest after 20 F 20 21 and pheromone blend incubation due to movement 16% MK571; ablation of abdomen ALB ald, ALB OH, plant & pheromone forced matter into head 21 F 20 21 blend destroying antennal nerves NA 22 F 20 21 plant and pheromone mix diluted dye 40%

The KCl control showed minimal dye uptake/retention by the antennal lobe neurons despite the use of techniques that have been shown to help improve uptake in other organisms. The addition of KCl causes all neurons to become excited at once and would normally cause a massive increase in fluorescence from all neurons containing the dye. The relatively small increase in fluorescence when all neurons were active meant that the much smaller increase in fluorescence of dye in a single active glomerulus would be impossible to detect (Trona et al. 2010).

As mentioned before, A.glabripennis is a large, robust beetle. The insects would not keep their antennal lobes perfectly still, confounding efforts to detect minute changes in intensity. They continued to move their mouthparts, and also the movements occurring within the abdomen pushed hemolymph up into the head cavity for hours after dissection. We did begin to try ablating the abdomen after incubation, just prior to testing, in order to alleviate this problem, but this did not solve the problem of negligible dye uptake.

One possible reason given by Max Planck researchers for failure to retain dye in neurons is active transport of the dye from the cells via cellular pumps. The MK-571 was added to prevent these pumps from working. It may be that MK-571 was ineffectual in this regard, or the dosage applied was too low to do this. Using lower levels of dye concentration may be a way to not trigger a removal response- the one trial performed on the final day using a dilute dye and the

MK-571 had the highest control response and variations in protocols in this vein may prove more effective. Different types of calcium-sensitive dyes should also be tested. It may be that a different carrier molecule/active dye will not be removed as easily or may pass through cellular barriers more easily. This leads to another possible reason for dye failure, i.e., other potential barriers to dye uptake such as the hemolymph-brain membrane on the antennal lobes. This layer could be carefully removed prior to incubation.

During development of the protocol, no odors should be tested and only the KCl test for dye uptake should be performed until manipulations that create successful dye uptake can be achieved. Using only the control test, rather than odors, during protocol development would allow more trials in less time and thus less monopolization of equipment by one study. Also during protocol development, movement within the head of the animal should be minimized by completely cutting through mandible musculature and by removing the head just prior to testing.

The failure was disappointing. However, I do not believe this technique should be abandoned. I believe that with more effort, iterations on the protocols I began with MK-571, temperatures, length of incubation, concentrations of dye, and ablation of sources of movement may yield an effective protocol.

References Abel R, Jurgen R, Randolf M. 2001 Structure and Response Patterns of Olfactory Interneurons in the Honeybee, Apis mellifera. The Journal Of Comparative Neurology 437:363–383. Galizia CG, Joerges J, Kuettner A, Faber T, Menzel R. 1997. A semi-in-vivo preparation for optical recording of the insect brain. Journal of Neuroscience Methods 76: 61–69. Lee SG, Vickers NJ, Baker TC. 2006. Glomerular targets of Heliothis subflexa male olfactory receptor neurons housed within long trichoid sensilla. Chemical Senses 31: 821-834 Trona F, Anfora G, Bengtsson M, Witzgall P, Ignell R. 2010. Coding and interaction of sex pheromone and plant volatile signals in the antennal lobe of the codling moth Cydia pomonella. Journal of Experimental Biology 213:4291-4303. Trona F, Anfora G, Balkenius A, Bengtsson M, Tasin M, Knight A, Janz N, Witzgall P, Ignell R. 2013. Neural coding merges sex and habitat chemosensory signals in an insect herbivore. Proc R Soc B 280:20130267.

Todd JL, Anton S, Hansson BS, Baker TC. 1995. Functional organization of the macroglomerular complex related to behaviorally expressed olfactory redundancy in male cabbage looper moths. Physiological Entomology 20: 349-61.

Chapter 4 Understanding the Distribution of Olfactory Sensory Neurons Along the Antenna Using Electroantennogram Recordings

Introduction

Studies reporting the topographical distributions of insect olfactory sensory neurons (OSNs) along or across an antenna that have specific tuning breadths to particular ranges of odorants have only been performed on narrow set of species. The distributions of differentially tuned OSNs on the antennae and palps of Drosophila melanogaster have been elegantly mapped by researchers in the John Carlson laboratory (deBruyne et al. 1999, Shanbhag et al.1999, deBruyne et al. 2001). In Drosophila sechelia Dekker et al. (2006) documented that sensilla housing OSNs tuned specifically to certain key volatiles from this species’ host fruit are more highly represented in abundance and cover more surface area in the D. sechelia antennae compared to these same sensillar types and OSNs on the antennae of D. melanogaster. Such mapping or sampling is performed via single sensillum recordings of OSNs’ responses to a panel of odorants coupled with microscopy and correlating the morphologies of the sensilla with the OSNs types of responses.

In insect sex pheromone research, a few studies on moths have used single sensillum recordings from trichoid sensilla to show how OSNs tuned to major or minor sex pheromone components are differentially distributed. Baker et al. (2004) reported that OSNs tuned to the major pheromone component of two heliothine moth species tend to be located more laterally on antennal flagellomeres than those tuned the minor components. Hansson et al.(1986) reported the opposite distribution, i.e., that the OSNs of Agrotis exclamationis tuned to this species’ minor pheromone component were located on the lateral margins of flagellomeres. Koutroumpa et al. (2014) showed how sensilla housing all three types of OSNs, each differentially tuned to different pheromone-related compounds, have a clumped distribution over this species’ flagellomeres with certain OSNs and sensilla types being found more on certain parts of the antenna than other parts. Lopes et al. (2002, 2004) used electron microscopy to examine the morphology of sensilla on the antenna of the cerambycid Phyoracantha semipunctata, to find that both basiconic (Lopes et al. 2002) and trichoid sensilla (Lopes et al. 2004) are also distributed in a clumped manner along the length of the antenna. Coupled gas chromatographic/single sensillum recordings were conducted on the basiconic sensilla, but no mapping was mentioned of OSN response types.

Single sensillum recordings seem to be more difficult to perform on cerambycid beetle species than on other types of insects such as moths and fruit flies. Cerambycid sensilla are typically small and nestled in amongst thick mats of hair-like scales that present a barrier. Also, the sensilla of many species seem to lie fairly flat in grooves in the antennal cuticle and are not very

accessible to recording electrode probes. In addition, cerambycid cuticle is tough, and so sharp- tipped tungsten electrodes need to be used for recording. There are only a few studies that have successfully accomplished such recordings on cerambycids (Dyer and Seabrook, 1978; Barata et al. 2002; Lopes et al. 2002; MacKay et al. 2015). In one species, Tetropium fuscum, the sensilla are long enough and pro ect out from the antennal cuticle to allow a “cut-tip” method of recording to be performed (MacKay et al. 2015).

Our lab has successfully conducted SSR on the smaller sensilla that have henceforth escaped SSR (Chapter 1; Wei et al. in preparation), but the technique is far more time-consuming and laborious than SSR for other insects and the connections are tenuous and often short-lived. We had made some observations during our SSR work that OSNs tuned to the two major male- produced sex aggregation pheromone components, an alcohol, 4-(n-heptyloxy)-butan-1-ol, and an aldehyde, 4-(n-heptyloxy)-butanal (Zhang et al. 2002), (hereafter referred to as "ALB-OH" and "ALB-Ald," respectively) were more likely to be found in the tip sections of the antenna over others, indicating that A. glabripennis may also have some sort of clumped distribution of OSN types.

It would aid in SSR work to identify areas of the antenna that are more likely to have OSNs more sensitive to particular odorants by quantifying responses to various odorants in subsections of the antennae in order to develop a more objective description of tuned OSN locations and to create a protocol that could aid others working with large antennae to also better locate OSNs of interest.

Materials and Methods

Beetles were reared on a pourable modification (Keena 2005) of a diet designed for Enaphalodes rufulus (Haldeman), which was chilled at 10° C for 90 days, during which time the larvae were allowed to develop until pupation. At this time, the pupae were transferred to 50 ml Falcon centrifuge tubes and incubated at 27.5° C until adult eclosion. Adults were fed red maple (Acer rubrum L.) twigs until preparation for electrophysiological study. Only virgin adults, about 25 days old (+/- 5 days) which had recently had their maturation feeding were used for the study.

To prepare individuals for study, antennae were removed with dissection scissors (World Precision Instruments, FL) and the insect euthanized by being placed in the freezer. The antenna was then cut into a series of two segment-long sections. The tip section was punctured at the tip using an insect pin (Bioquip, CA) to allow the conductive gel to create a contact between the neurons at the tip of the antenna and the probe (Crook et al. 2014).

Electroantennogram recordings (EAGs) were conducted using a Syntech Quadroprobe (Syntech; Germany; Park et al. 2002 Chem. Senses 27:352). Each antennal section was held in place with Spectra 360 electrode gel (Parker Laboratories, Fairfield, NJ, USA) on the probe with the proximal end placed upon a recording electrode and the distal end placed on the shared ground electrode (Fig. 5-1). A custom high-impedance DC-coupled preamplifier and 16 bit analog-to- digital converter with a conversion rate of 46875 Sa/s and voltage range of -1 to +1 volts was

used to make recordings. These data were then re-sampled at 10 KSa/s and stored for later analysis.

Figure 4-1. Quadroprobe electroantennogram probe used to measure simultaneous EAGs of four antennal sections (AS). Each antennal section is embedded at either end in EKG gel between a recording electrode(R) and a ground electrode (G) common to all recording electrodes. The probe can record the activity of each antennal section as various odorants are puffed into a constant clean, humidified air flow.

The odorants used in this study were diluted in hexane in 10-fold steps to different concentrations such that each odor cartridge contained a loading of either 1 µg, 10 µg, or 100 µg of an odorant when a 10 μl aliquot was dispensed onto a 0.2 X 1.5 cm filter paper strip (Whatman). Each strip was inserted into a 15cm-long Pasteur pipette to create each of the odor

cartridges. A constant airflow of charcoal-purified, humidified air was passed across the palps through a glass tube (10 mm diameter) during the experiments. The constant airstream was first filtered through an Ehrlenmyer flask filled with charcoal and then humidified by bubbling the air through an Ehrlenmeyer flask filled with distilled water. Odorants were delivered into this constant air stream via the Pasteur pipette whose tip was inserted through a small hole in the glass tube, 11 cm upstream from its end. A stimulus flow controller (Syntech, Germany) pulsed a 40 ml/s air stream through the cartridge for 0.3 seconds, effectively delivering a puff of volatiles from the odor cartridge into the air stream and onto the antennal sections. Odor cartridges were remade every 30 puffs for lower volatility odorants and every 3 puffs for the highly volatile odorants.

The odorants used were purchased from the companies and contained the purities shown in Table 4-1 below.

Table 4-1. Odorants used in this study, along with their commercial sources and purities. Odorant Supplier Purity (%) citronellal Acros Organics B.V.B.A. (Belgium) 93 4-(n-heptyloxy) butan-1-ol Bedoukian Research Inc. (Danbury, CT, USA) 4-(n-heptyloxy) butanal Bedoukian Research Inc. (Danbury, CT, USA) acetic acid, J.T. Baker, Inc. (Phillipsburg, NJ, USA) 100 (3E,6E)--farnesene Jocelyn Millar Laboratory, University of CA, Riverside 86 trail pheromone, whole blend Kelli Hoover Laboratory, University of PA trail pheromone, major Kelli Hoover Laboratory, University of PA components trail pheromone, minor Kelli Hoover Laboratory, University of PA components isovaleric acid Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 benzoic acid Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 Z-2-hexenal Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 butyraldehyde Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 1-4-diaminobutane Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 β-caryophellene, Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 Z-3-hexen-1-ol Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 linalool, Sigma-Aldrich Corporation (St. Louis, MO, USA) 99 -terpineol Sigma-Aldrich Corporation (St. Louis, MO, USA) 96 geraniol Tokyo Chemical Industry Co., Ltd. (Japan) 96 eugenol Tokyo Chemical Industry Co., Ltd. (Japan) 98

In all recordings, the EAG quadroprobe was placed 2 cm from the mouth of the constant air stream outlet such that the ground probe axis was aligned with the airflow and centered on the middle of the outlet. This insured that each antennal segment was equidistant from the odor

source and would get an equal dose of odorant from a puff into the constant, humidified airstream.

After using the quadroprobe to amplify each of the four EAG signals by a factor of 10 V/V, single pole low-pass filters with 3 dB cutoff frequencies of 16 Hz were applied to reduce powerline interference. Using custom software written using Labview, the signals were acquired with a USB 6212 (National Instruments, TX) using a per-channel sample rate of 50 kSa/s. Following the application of a digital single-pole high-pass filter with a cutoff frequency of 0.1 Hz, the data were lowpass filtered and decimated to 100 Sa/s and written to disk. The EAG recordings were then measured using Labview (National Instruments, TX) by finding the difference between the resting potential and the peak of a depolarization response (Fig. 4-2).

Figure 4-2. Electroantennogram tracings from a male A. glabripennis in response to a puff of 100µg of eugenol. Each of the four tracings is from a different antennal section, with segments 10-11 comprising the tip of the antenna and segments 4-5 being most proximal to the head.

An overall sensitivity to changes in the amount of water vapor present in the constant airstream was found along the length of the entire antennae, strongly suggesting that there are hygroreceptors on all antennal segments. When introducing a puff from the Syntech controller using filtered, ambient air, having a low moisture content due to the winter weather, into the constant humidified air stream, there was a strong reaction to a blank. The addition of water to the filter paper in the cartridge reduced the EAG response appreciably. The need for reducing responsiveness to drier air in the puffs is shown in this initial experiment in which a puff-

humidifying apparatus (explained in more detail later) was not used, only blank tubing upstream of the odor cartridge. Three different loadings of water, from 1µl to 100µl were loaded onto the filter papers of three different odor cartridges, and no water at all was in the fourth, dry-blank cartridge. For both male and female antennal preparations, puffs of dry air through the dry-blank odor cartridge elicited higher EAG amplitudes from the different antennal segments than the odor cartridges loaded with any amount of water (Fig. 4-3). The antennae were responding to the decrease in moisture caused by the dry air being injected into the constant, humidified airstream, apparently due to the activity of hygroreceptors on the ALB antennae.

Figure 4-3. EAGs from four different sections of female and male antennae in response to puffs of various amounts of water using air from the stimulus flow controller employing a stream of non-

humidified air (no saturated dental wick upstream). Section 10-11 (flagellomeres 10 and 11) comprises the tip of the antenna with section 4-5 (flagellomeres 4 and 5) being most proximal to the head. Increasing amounts of water were used on filter papers loaded into odor cartridges, with no water at all (blank, dry puff) used in a fourth cartridge. Capital letters indicate significance between the amounts of water for that section of antenna. The smaller EAGs in response to cartridges containing water-laden filter paper than to the dry blank cartridge indicate that the decrease in water vapor in response to the dry blank puff significantly influenced the EAG amplitudes due to the responses from hygroreceptors on the antennae. Letters indicate significant differences via two-way ANOVA at p ≤ 0.05, N=9.

The puffs of air that were injected into and through the odor cartridge for the body of this experiment were therefore first humidified to reduce any changes in water vapor when the puffed air entered the humidified constant airflow, reducing the variables being measured by the EAGs. This humidification was accomplished by the 0.3 s puff from the flow controller first being passed through a 15 ml centrifuge tube containing a cotton wick saturated with distilled water. The cotton wick was suspended so as not to impede airflow. The centrifuge tube was held in place and made air-tight with “Sugru”, a rubberized, self-setting putty (FormFormForm, UK), This humidification chamber was thus interposed between the odor cartridge and the air-puff injection line immediately upstream (Fig. 4-4). For non-humidified puffs made in only the experiments to demonstrate the effect of dry air on the EAG amplitudes, the humidifier cylinder was swapped out with an equal length of blank tubing.

Figure 4-4. The humidifying cylinder used to add water vapor to the puffs and its placement in the setup.

The removal of this confounding response to water vapor reduction is demonstrated by comparing humidified puffs of geraniol into the humidified airstream vs. unhumidified puffs of geraniol into the humidified airstream. The humidified puffs elicited significantly lower responses than unhumidified puffs of geraniol via paired T-tests at p ≤ 0.05 (Fig. 4-5). (For these non-humidified puffs, the humidifier cylinder with its water-imbued dental wick was swapped out with an equal length of blank tubing.) The difference in water vapor between a humidified puff of a geraniol-loaded cartridge vs. unhumidified puffs was detectable by all antennal sections. The pairs of comparison puffs were conducted on the same day, within an hour of each other, keeping atmospheric humidity and temperature differences to a minimum so that the primary difference was the water vapor added to the puff by the humidifying apparatus. An unknown amount of variance, here sometimes amounting to a more than doubling of the depolarization amplitude, would be added to each stimulus puff in all our experiments due to variations in the humidity in the puff-stream line if it were not humidified. Thus, to preempt this effect on EAGs due to a decrease in water molecules over the antennae, the water-saturated, cotton dental wick was always used upstream of all the odor cartridges for puffs in our stimulus regime throughout this study.

Figure 4-5. EAGs from female and male antennal sections in response to geraniol that was puffed from the odor cartridge using dry air from the puffing stream or air-puffs that had been humidified by the water-saturated dental wick upstream of the cartridge. The response of hygroreceptors to the decrease in

water molecules when dry-air puffs were used was significant and often more than doubled the depolarization amplitude compared to when the puffs were humidified. Capital letters indicate significance for that antennal section. The lesser response to a humidified puff compared to a dry puff indicates that the reduction in humidity elicits a response in addition to a blank alone. Letters indicate significant differences via paired T-tests at p ≤ 0.05, N =14 per sex.

When conducting EAG measurements of the full panel of odorants (Fig. 4-6) the order of odorants used was constant and thus measurements of a control odorant, geraniol, were taken at the beginning, middle, and end of each series of EAGs in order to measure the decay of the tissue/connection for each antennal section of each individual. These three measurements for each antennal section were then used to adjust the EAG measurements of the odorants measured, according to their order of use in the panel, to account for this decay for each antennal section of each individual.

Statistics were conducted as follows:

EAGs of water at various concentrations (Fig. 4-3) A two-way ANOVA was conducted for each antennal section of each sex comparing the various amounts of water and designating individuals as random factors. p ≤ 0.05, N=9 individuals per sex.

Humidified vs unhumidified puffs of geraniol (Fig. 4-5) Paired T-tests were conducted for each antennal section of each sex comparing EAGs of geraniol delivered via humidified vs. unhumidified puffs. p ≤ 0.05, N=14 individuals per sex.

Antennal section EAGs for all odorants combined (Fig. 4-7) Comparing the EAGs between sexes was done using a t-test for each section, comparing the EAGs of the full panel of odorants of each sex. N=294 per section per sex (21 odorants and 14 individuals of each sex). The four antennal sections were compared, separately for each sex, using a Wilcoxon signed rank test using EAG measurements of the entire panel to compare the four sections to one- another. p ≤ 0.05, N=294 per section per sex (21 odorants and 14 individuals of each sex).

EAGs of pheromone components compared to EAGs of other odorants (Figs. 4-8,9) Using the data from the full panel, a two-way ANOVA was conducted for each antennal section of each sex comparing the pheromone components and another odorant and designating individuals as random factors. p ≤ 0.05, N=14 per section per sex.

Results

The levels of EAG-measured activity in the various sections differed from one another in a regular pattern, suggesting that the abundances of responding OSNs in sensilla are not distributed

evenly along the length of the antenna, but that some parts of the antenna have more responding OSNs than others. This potential difference in abundance can be observed according to the pattern of highest-to-lowest EAG responses regardless of the odorant that was tested (Fig. 4-6).

The EAG amplitudes of the female antennae were smaller than the males for all sections per t- tests (Fig. 4-7). For both sexes, flagellomeres 8 and 9 exhibited higher amplitude EAGs than other antennal sections regardless of the odorants they were exposed to. This pattern comes into better focus when the amplitudes of the responses to the different volatiles are simplified by comparing the mean responses of each the four antennal sections of all odorants for each sex using Kruskal-Wallis test followed by a Wilcoxon signed rank test at p ≤ 0.05 (Fig. 4-7). The largest responses were exhibited from flagellomeres 8 and 9, with significantly reduced amplitudes being exhibited from all the other antennal sections, which were especially low in the most proximal section containing flagellomeres 4 and 5. These EAG amplitudes suggest there might be different numbers of responding OSNs – and thus sensilla – distributed along the length of the antennae, which is corroborated by the counts of trichoid and basiconic sensilla that have been made from SEM images showing that sensilla are not evenly distributed along the length of the antenna (see Fig. 4-12 of Discussion). However, the larger EAGs from segments 8-9, do not seem to correlate with the sensillar counts from this section, which are just as high as those from the surrounding segments (Fig. 4-12 of Discussion).

Figure 4-6 Mean (± S.E.) electroantennogram response amplitudes (µV) from both female (A) and male (B) antennae showing the responses from four different antennal sections to various odorants. Section 10- 11 is at the tip of the antenna whereas section 4-5 is most proximal to the head. Section 8-9 regularly exhibited the strongest responses to odorants whereas section 4-5 showed the weakest responses, suggesting that there were differences in the abundances of OSNs responding to the odorants tested that were correlated with these different amplitudes. N=14 individual antennal preparations from each sex.

Figure 4-7. Mean electroantennogram responses to odorants exhibited by each of the four antennal sections for all the odorants tested. Means for each sex having no capitol letters in common are significantly different according to a Wilcoxon signed rank test (performed for each sex, lower case letters denote sex: m = male; f = female). T-tests for each section show that EAG amplitudes for males are significantly greater than EAGs of females, for all antennal sections, p ≤ 0.0001(****). Segments 10- 11 are the most distal, and 4-5 the most proximal antennal segments. N = 294 measurements per section per sex.

Figure 4-8. Mean EAG amplitudes (± S.E.) from male and female antennal sections in response to different terpenoid volatiles compared to those evoked by the ALB OH and ALB Ald pheromone components. A) citronellal; B) geraniol; C) eugenol; D) E,E-α-farnesene; E) β-caryophellene. Means for each antennal section having no letters in common are significantly different according to two-way ANOVA tests between the three odorants for each section, N=14.

Examining the EAG response amplitudes to some of the individual odorants as compared to the responses to the two ALB pheromone components revealed some trends that might be important for understanding how OSNs might be differentially distributed among, or else co-located within, sensilla along the antenna. These comparisons were conducted using the data from the full panel and comparisons made via a two-way ANOVA test of the EAGs of the odorant in question and the two pheromone components for each antennal segment. For the terpenoid class of odorants, it was interesting that EAG amplitudes to these volatiles did not differ very often or very greatly from those to the ALB OH or ALB Ald components (Fig. 4-8). There were only a few differences and these occurred occasionally between males and females to some of these terpenes and the pheromone components. For instance, EAG amplitudes to citronellal were no different than those to the pheromone components, except that the response from segments 6-7 of females was lower than to the ALB Ald, and from males this section’s amplitude was significantly higher than to the ALB OH (Fig. 4-8A). To geraniol, all amplitudes from females were not significantly different from those to the pheromone components, but in segments 6-7 and 8-9 of males they were higher than to the pheromone components (Fig. 4-8B). For eugenol and E,E-α-farnesene, segments 6-7 of both sexes (and segments 8-9 of males for eugenol) exhibited lower amplitudes than to the ALB Ald (Fig. 4-8C,D) . Responses to β-caryophyllene were not significantly different than to the pheromone components along the entire length of the antenna (Fig. 4-8E).

Figure 4-9. Mean EAG amplitudes (± S.E.) from male and female antennal sections in response to different general plant-related and other volatiles compared to those evoked by the ALB OH and ALB Ald pheromone components. A) acetic acid; B) E-2-hexenal; C) Z-3-hexen-1-ol; D) butyraldehyde; E) blank; F) water. Means for each antennal section having no letters in common are significantly different according to two-way ANOVA tests between the three odorants for each section, N=14.

Responses to many of the non-terpenoid odorants did, however, seem to differ more distinctly from the amplitudes evoked by the ALB OH and ALB Ald pheromone components. For these volatiles, the responses were usually higher than those to the pheromone components. The higher amplitudes cannot simply be ascribed to higher volatility of some of these compounds compared to the pheromone components. For instance, whereas female and male antennae both exhibited higher EAG amplitudes to the green leaf volatile E-2-hexanal along the length of the antenna than to the two ALB sex pheromone components (Fig. 4-9B), Z-3-hexen-1-ol, which has a similar volatility, did not (Fig. 4-9C). In fact, Z-3-hexen-1-ol elicited responses that were significantly lower than the pheromone components on section 8-9 and not significantly higher along the other sections than to the pheromone components (Fig. 4-9C). In another comparison, EAG amplitudes in response to acetic acid were significantly higher than to the ALB pheromone components along the length of the antenna of females and section 4-5 of males (Fig. 4-9A), whereas those to butyraldehyde were significantly higher than the pheromone components only on the two most distal sections of females and only on section 8-9 of males (Fig. 4-9D).

As shown in the Materials and Methods section above, along the entire antenna (Fig. 4-3A,B) We found a consistent sensitivity to the amount of water vapor present in the stimulus cartridge, strongly suggesting that there are abundant hygroreceptors on all antennal segments. It can be seen that because we avoided the hygroreceptor response in our stimulus setup that used a humidified air-puff stream, the control treatment stimuli in the odorant panel exhibited significantly lower EAG responses than the pheromone components (Figs. 4-9E,F). These control treatments used either a dry, blank odor cartridge (Fig. 4-9E) or a cartridge loaded with 100 µl of water (Fig. 4-9F).

Discussion

There did seem to be some sexual differences as well as some similarities in the EAG response amplitudes of both females and males to the different odorants in the panel. The responses of both females and males to the three odorants, E-2-hexenal, butyraldehyde, and acetic acid were higher than to many of the terpenoids such as eugenol, E,E-α-farnesene, β-caryophyllene, and citronellal, and the magnitude of this heightened response in female antennae appeared greater than that for male antennae (Fig. 4-6A,B). Conversely, male antennae appeared to be more responsive to the two sex pheromone components ALB OH and ALB Ald, than female antennae, as well as to Z-3-hexen-1-ol, α-terpineol, and linalool compared to females’ responses (Fig. 4- 6A, B).The responses of both females and males to the three odorants, E-2-hexenal, butyraldehyde, and acetic acid were higher than to many of the terpenoids such as eugenol, E,E- α-farnesene, β-caryophyllene, and citronellal, and the magnitude of this heightened response in female antennae appeared greater than that for male antennae (Fig. 4-6A,B). Conversely, male antennae appeared to be more responsive to the two sex pheromone components ALB OH and

ALB Ald, than female antennae, as well as to Z-3-hexen-1-ol, α-terpineol, and linalool compared to females’ responses (Fig. 4-6A, B).

The use of the quadroprobe to compare sections of a single antenna is a novel approach to studying the olfactory abilities of the antennae. In the results section, we stated that the differing EAG responses implied the number of responding OSNs. Alternative explanations are explored below.

These results could be the placement of the antennal sections on the probe and the probe’s placement in the airstream, such that one section was in the primary path of the odor plume, over the other sections. This was a concern during the design of the experiment and for this reason the sections were all placed equidistant from the tip of the probe, and the probe itself was aligned longitudinally such that the ground probe was pointed directly into the effluent and at the center of the air stream exit. Thus all four sections were exposed to the same amount of the odor.

Other possible arguments for the strength of EAGs from section 8-9, compared to the other sections tested, may be related to their physical measurements, e.g. being longer may allow a section to be exposed to more of the odorant-filled airstream or that a thinner segment may be more conductive. In terms of length, section 4-5 is longer than 8-9, and 6-7 and 8-9 are about the same length (Fig. 4-10), yet 8-9 still out-performs those segments with regard to EAG amplitudes. In terms of narrowness, section 10-11 is thinner, and section 4-5 far wider, than 8-9 (Fig. 4-11), and still the EAG amplitudes of 8-9 were greater than those from these other sections. It would appear that size is not a primary factor.

Figure 4-10. Lengths of antennal flagellomeres of A. glabripennis, N=6 per sex (from Meng, 2014).

Figure 4-11. Widths of antennal flagellomeres, N=6 per sex (from Meng, 2014).

The EAG amplitudes might be related to the absolute numbers of trichoid and basiconic 1 sensilla in each segment along with the volume or surface area affecting the conductivity of the segments. Previous work conducted by the Kelli Hoover Laboratory (Meng 2014) and, until now, unpublished work from Kathleen Shields, USDA Forest Service, Northern Research Station, both examined several aspects of A. glabripennis antennal morphology using SEM imaging. Table 4-2 and Figure 4-12, below, show the counts of sensilla that these previous studies individually found. The totals found by these separate studies for each flagellomere were

quite similar (Table 4-2) and, after averaging their numbers for each flagellomere, are graphically displayed (Fig. 4-12).

Table 4-2. Numbers of trichoid, basiconic 1, and basiconic 2 sensilla counted on A. glabripennis antennal flagellomeres using SEM imaging (Shields (this paper); Meng, 2014).

Trichoid and Trichoid Basiconic 1 &2 Basiconic 1 & 2 (Meng) (Meng) (Sheilds) N=6 N=6 N=6 Segment Male Female Male Female Male Female 3 4.8 9.7 2.4 0 2.6 9.7 4 255.7 346.5 207 287 65 60 5 494.5 543.5 421 379 247 123 6 673.5 820.8 495 649 245 172 7 832.6 925.7 661 698 221 227 8 826 798.2 550 597 389 201 9 971.7 892.5 690 474 361 419 10 842.5 773.2 587 474 280 299 11 848.4 1011.5 625 814 428 198

Figure 4-12. Average numbers of combined trichoid and basiconic sensilla obtained via SEM imaging from two studies (Meng 2014; Shields (this study)). N=12.

The numbers of sensilla do differ along the length of the antenna, but EAG amplitudes do not mirror these counts. The section with segments 4-5 has far fewer sensilla than the other sections tested, which all have similar, larger, sensilla counts, while the section with segments 8-9 easily produced the largest EAGs. The counts did not include other sensilla types, such as chetiform or the sparse, stubby, pillow-like sensilla, which could be affecting the EAGs. The higher EAG

amplitudes from segments 8-9 than from 6-7 may also possibly be due to a higher density of sensilla on 8-9. Segments 6-7 is longer than 8-9 and thicker (Figs. 4-10,11), and because the sensillar counts on these two sections are similar (Fig. 4-12), these differences in length and width might mean a higher density of sensilla on segments 8-9.

However the section with segments 10-11 also has a high density of sensilla due to their smaller dimensions compared to other segments (Figs. 4-10,11). The lower EAG amplitudes from segments 10-11 may be explained by these antennal tips receiving wear and tear and damage not incurred by the more proximal segments. Many cerambycids suffer from antennal damage at the very tips (Hanks et al. 1996 and Hanks 1999) over the course of their lives. Finally, some of the reduction may be due to a methodological, electrophysiological issue. In order to record EAGs from the tip of the antennae (flagellomere 11), its rounded end was opened with pin holes rather than being excised with scissors like the other sections, in keeping with previous studies (Crook et al. 2014). This may have possibly resulted in section 10-11 not being as electrically conductive as the other sections that were fully open at both ends due to being fully cut. The tip of flagellomere 11 was not cut because this would have removed substantial numbers of sensilla that populate the rounded end, and we found that creating the pin-holes did allow good EAGs to occur, showing there was good – but maybe not optimal – electrical contact. Or it may be that because the tip segments are shorter, a larger proportion of the segment was covered with electrode gel, blocking more of the sensilla from the odorants.

Segments 8-9 may have more OSNs that are responsive to the odorants tested while the other segments have OSNs only responsive to other odorants and this explains the differing EAGs, but with such a broad spectrum of odorants, the broadly tuned nature of most ORs, and the same pattern being exhibited for each odorant, this is unlikely.

Regardless, there must be some benefit to having long antennae with the majority of sensilla loaded toward the distal end, which is indicated by the actual sensillar counts and the EAG amplitudes from this study.

Cerambycids commonly are found to use their antennae for mate-finding by tactile cues and contact chemoreception, and it has been speculated that longer antennae increase sweeping coverage for mate-finding and detection-aggression against rivals (Hanks et al. 1996, Hanks 1999, Lopes et al. 2004, Morewood et al. 2004). The use of antennae for mate-finding through tactile stimuli and contact chemoreception has also been shown for A. glabripennis (Zhang 2003), which occur at the end of a complex series of steps. These involve initial long-range attraction via their sex-aggregation pheromone plus host odors followed by trail-sex pheromones that bring the sexes into close proximity whereupon these contact cues can function (Graves 2016, Hoover 2014, Crook 2014, Nehme 2009, Zhang 2002). The benefit of long antennae for social interaction is not confined to cerambycidae. Possible convergent evolution has resulted in snapping shrimp, Alpheus spp., also having been selected to have long antennae for mate- finding, courtship, and other conspecific interactions (Vickery et. al. 2012). This antennal

attribute has evolved despite the risk shared with cerambycidae (Hanks et al. 1996 and Hanks 1999) of antennal damage from aggressive interactions (Vickery et al. 2012). Wide-ranging, long antennae that can simultaneously detect left-right differences in chemical and tactile stimuli to allow the animals to perform accurate orientation maneuvers toward a potential mate will be optimized for this purpose if the chemo- and mechanosensors are clustered toward the antennal tips as seen here for A. glabripennis.

The majority of sensilla with their volatile-sensing OSNs are located quite distal from the head, in a location perhaps more vulnerable to damage or loss. The EAG results indicate that even if OSNs at the tips are not responding optimally, the next most proximal segments are reporting odorant abundances quite strongly compared to the more proximal antennal segments. As stated above, it has been surmised that long antennae can help males more efficiently sweep a surface in order to gain contact with females or to detect rivals while mate-guarding. Although males may benefit from the sexual dimorphism of extra antennal length, especially in flagellomere 11, in terms of locating and guarding mates, females continue to have relatively long antennae despite the vulnerability of lengthy antennae to damage. Females do utilize their antennae for exploring their surroundings by antennuating surfaces (personal observations), which implies they are also utilizing mechano- and chemoreception at the tips such that the benefits outweigh the risks. But the likely clustering of responding OSNs in segments 8-9, for both males and females, implies there must be some benefit to having OSNs that are located at some distance from the head despite the increased vulnerability to damage. One such benefit could be for detecting and following odor plumes while walking. Barata and Arauâjo (2001) found that during flight, the cerambycid Phoracantha semipunctata can locate and follow odor plumes to point sources, using what appears to be optomotor anemotaxis, complete with counter-turning, to steer upwind and keep in contact with odor plumes. A. glabripennis spend the majority of their time on foot and although they can fly for dispersal (Meng et al. 2015), they rarely fly if they can walk (Morewood et al. 2004). On the ground, at walking speed, locating and staying in contact with odor plumes would be difficult given the relative speed of walking vs. shifts in wind direction and speed. Long antennae with OSNs that are clustered in sensilla near the ends to create the most widely spaced, simultaneous left-right pheromone detectors possible, and might have been selected for to optimize the precision of orientation decisions made by walking, mate- locating males or females.

We had expected to find some differences in the proportions of OSNs responding to various odorants, and based on initial single cell recordings performed in our lab (Chapter 2; Wei et al. in preparation) we particularly expected that EAGs would imply that there is more of a differential clustering of pheromone-sensitive OSNs toward the antennal tips compared to other types of OSNs. However, although there was an increase in EAG amplitude from the antennal base through section 8-9 and a drop-off in amplitude in section 10-11, the pattern was not unique to pheromone-component-initiated EAGs. EAGs from every odorant in the panel exhibited this same pattern. This implies that sex-aggregation-pheromone-specific OSNs are no more clustered

in specific areas of the antennae than other OSN types. Another possibility is that large numbers of more generalist, broadly tuned OSNs that can respond to the sex-aggregation pheromone components as well as to a large number of plant volatiles might mask the smaller EAG amplitudes contributed by much smaller numbers of pheromone-specific OSNs highly clustered on particular flagella. To determine whether pheromone-tuned OSNs are indeed clumped in some topographic way that was missed in this study would require a study doing the very thing this study was intended to avoid: conducting a large number of single sensillum recordings along the length of the antenna to map the locations of pheromone-component-specific OSNs vs. those more generally responsive to other odorants.

Single sensillum recordings were performed as part of our laboratory’s effort to characterize the tuning profiles of OSNs on male and female A. glabripennis antennae (Chapter 1; Wei et al. in preparation). Thus, some of these EAG results can be interpreted according to what we know about the preferences some OSNs have for various odorant ligands. For instance, the EAG responses to terpenes such as citronellal, eugenol, -caryophyllene, etc. track more closely with the EAGs for the ALB-Ald and ALB-OH male-produced sex-aggregation pheromone components than do EAGs in response to several other types of volatiles (Figs. 4-8,9). In our single-sensillum recordings we found that large-spiking OSNs responding specifically to the ALB-Ald and ALB-OH are co-localized only in trichoid sensilla with a smaller-spiking OSN that responds strongly to several terpenes, including citronellal and geraniol. The correspondence of EAGs (Fig. 4-8) in all flagellomeres to the two pheromone components to those of these terpenes can thus be understood by the co-localization of these two types of OSNs in the same sensilla (Chapter 2; Wei et al. in preparation). In addition, the single sensillum recordings characterized OSNs in basiconic 2 sensilla that were broadly responsive to the ALB-Ald and different types of farnesenes, or else were broadly responsive to the ALB-OH and to other types of terpenes such as eugenol (Chapter 2; Wei et. al in preparation). Such OSNs again explain why the EAG amplitudes along the length of the antenna to these terpenes correspond well with the EAGs to the ALB Ald and ALB OH.

Other results from the single sensillum recordings also make sense in light of the EAG amplitudes along the antenna in response to other classes of odorants. These EAGs did not track as closely with the ALB Ald and ALB OH (Fig. 4-9) and makes sense in that we did not find OSNs responsive to these compounds that either were co-localized with the pheromone- component-tuned OSNs or were broadly tuned to the pheromone OSNs plus these particular odorants (Chapter 2; Wei et al. in preparation). The EAGs suggest that there may be greater (e.g., E-2-hexenal; Fig. 4-9B) or fewer (e.g. Z-3-hexen-1-ol; Fig. 4-9C) numbers of OSNs responsive to these odorants than those responsive to the pheromone components on these antennal sections. Of particular interest is that these differences to such compounds are not consistent along the length of the antennae. Thus, if the EAG amplitudes are correlated with the numbers responding OSNs on certain flagellomeres, then these segmental differences should indicate that the

numbers of OSNs differentially responsive to these odorants do vary topographically along the length of the antennae.

Interestingly, acetic acid elicited very strong EAGs compared to the ALB OH and ALB Ald (Fig. 4-9A). Acetic acid is commonly detected by ionotropic receptors (IRs) expressed in coeloconic sensilla (Benton et al. 2009; Ai et al. 2010), which have not been observed on the antennae of A. glabripennis either by Shields’ S M work or by Meng (2014). There are unusual, stubby, puffy sensilla (Fig. 4-13) that sparsely populate all the antennal segments, but are especially abundant at the terminus of flagellomere 11. In all TEM images, they also seem to have single depression indicative of a possible single pore (Fig. 4-13). Perhaps these sensilla house IR-expressing OSNs. Another possibility is that coeloconic sensilla are hiding in the white, setae-covered areas where the cuticular surface cannot be observed. A. glabripennis antennae have these flattened white setae (sometimes called “scales”) that form white bands alternating between the black areas on the antennae, and even using electron microscopy it is very difficult to observe much of the cuticular surface beneath them. IRs have been found in chetiform sensilla of desert locusts (Guo et. al. 2014). However, histological studies using TEMs have not resulted in any evidence that any type of OSN is housed in chetiform sensilla. Therefore it may be that in the absence of coeloconic sensilla, other types of sensilla on the A. glabripennis antennae house IRs that respond to acetic acid and create these large EAGs. A possible candidate for such a sensillar type might be the basiconic 1 (“sharp-tipped basiconic” sensilla that are so numerous across the length of the antennae, but which we were unable to attain electrical contact with and interrogate any OSNs housed in them in our single sensillum efforts, despite repeated tries over a year of recording experiments (Chapter 2; Wei et al. in preparation).

Figure 4-13. Stubby, puffy sensilla with a possible single-pore depression located in small numbers on all segments of A. glabripennis antennae and larger numbers at the terminus of flagellomere 11 in both sexes.

Also potentially related to OSNs expressing IRs, the EAG recordings from this study show that A. glabripennis antennae house OSNs that are very sensitive to changes in the abundance of water molecules in the air. Water-sensing sensilla were discovered via SSR in basiconic sensilla of Aedes aegypti (Kellogg 1970) and basiconic sensilla are common on A. glabripennis antennae. Changes in the moisture content of the air has been found to be reported by sensory neurons housed in many different kinds of insect sensilla, according to a study that used SSR on 18 species across nine orders by Altner et al. (1983). The study found hygroreceptive abilities in several types of sensilla, including trichoid, basiconic, coeloconic, and styloconic sensilla. One of the findings most relevant to our own study was a hygroreptive sensillum (Fig. 4-14) on the ground beetle, Notiophilus biguttatus, that have a similar appearance the stubby, puffy sensilla found along A. glabripennis male and female antennae (Fig. 4-13).

Figure 4-14. Drawing of a transverse (vertical) section of a water-sensing sensillum located on the antennae of the ground beetle, Notiophilus biguttatus ( from Altner et al. 1983).

The adults of A. glabripennis feed on living trees and females lay their eggs in living wood (Morewood et al. 2004, Meng 2015). To identify moisture content could be a way to insure that there is healthy tissue for feeding, oviposition, and subsequent larval feeding. They many have originally evolved in high-humidity riparian habitats (Williams et al. 2004). Desiccation is a danger for insects and the ability of A. glabripennis to identify favorably moist microenvironments might be important (Enjin et al. 2016). Hygroreceptors have commonly been described during investigations of insect antennal sensilla using SSR (c.f., Altner et al. 1983, Iwasaki et al. 1995, Tichy and Kallina 2010, Enjin et al. 2016). Hygroreception has also been found to be performed by the palps of insect larvae (Eilers et al. 2012) and adults (Chapter 5; Hall et al. in preparation; Chappuis et al. 2013). For cerambycid beetles, the findings presented here may be informative in helping to start finding more hygroreceptive sensing apparatuses in other species in this family and investigating the effects of water on the behavior of this group of beetles.

The primary findings of this study, that EAG amplitudes increase toward the distal end of the antennae combined with sensilla counts increasing toward the distal end of the antennae, imply there is an evolutionary advantage to this arrangement of OSNs. Further study to understand the behavioral reason as to why OSNs are arranged topographically in this way should be rewarding and illuminating. This protocol of EAG recording using four antennal sections simultaneously might be useful for other insects endowed with very long antennae so that likely areas of overabundance of candidate OSNs might be located before single sensillum recordings are employed to more precisely verify such OSNs’ presence in these locations.

References

Ai M, Min S, Grosjean Y, Leblanc C, Bell R, Benton R, Suh GS. 2010. Acid sensing by the Drosophila olfactory system. Nature 468: 691-5.

Baker TC, Ochieng SA, Cosse AA, Lee SG, Todd JL, Quero C, Vickers NJ. 2004. A comparison of responses from olfactory receptor neurons of Heliothis subuflexa and Heliothis virescens to components of their sex pheromone. Journal of Comparative Physiology A 190:155-165.

Barata E, Mustaparta H, Pickett J, Wadhams L, Araujo J. 2002. Encoding of host and non-host plant odours by receptor neurones in the eucalyptus woodborer, Phoracantha semipunctata (Coleoptera: Cerambycidae). Journal of Comparative Physiology A, 188: 121-133.

Benton R, Vannice KS, Gomez-Diaz C, Vosshall LB. 2009. Variant ionotropic glutamate receptors as chemosensory receptors in Drosophila. Cell 136: 149-62.

Chappuis CJF, Béguin S, Vilmant M, Guerin PM. 2013. Water vapour and heat combine to elicit biting and biting persistence in tsetse. Parasites & Vectors 6:240.

Crook, D.J., Lance, D.R. & Mastro, V.C. 2014. Identification of a potential third component of the male-produced pheromone of Anoplophora glabripennis and its effect on behavior. Journal of Chemical Ecology 40: 1241-1250. de Bruyne M, Clyne PJ, Carlson JR. 1999. Odor Coding in a Model Olfactory Organ: The Drosophila Maxillary Palp. The Journal of Neuroscience 19: 4520–4532. de Bruyne M, Foster K, Carlson JR. 2001. Odor Coding in the Drosophila Antenna. Neuron 30: 537-552.

Dekker T, Ibba I, Siju KP, Stensmyr MC, Hansson BS. 2006. Olfactory shifts parallel superspecialism for toxic fruit in Drosophila melanogaster sibling, D. sechellia. Current Biology 16: 101-109.

Eilers EJ, Talarico G, Hansson BS, Hilker M, Reinecke A (2012) Sensing the underground – ultrastructure and function of sensory organs in root-feeding Melolontha melolontha larvae. PLoS ONE 7: e41357.

Enjin A, Zaharieva EE, Frank DD, Mansourian S, Suh GSB, Gallio M, Stensmyr MC. 2016. Humidity Sensing in Drosophila. Current Biology 26: 1352–1358.

Gao, R., and G. Li. 2001. Review and prospect of research on Anoplophora glabripennis in China. Entomological Knowledge 38: 252-258.

George J, Robbins PS, Alessandro RT, Stelinski LL and Lapointe SL. 2016. Formic and acetic acids in degradation products of plant volatiles elicit olfactory and behavioral responses from an insect vector. Chemical Senses 41: 325–338.

Graves F, Baker TC, Zhang A. Keena M, Hoover K. 2016. Sensory aspects of trail-following behaviors in the Asian longhorned beetle, Anoplophora glabripennis. Journal of Insect Behavior 29: 615-628.

Guo M , Krieger J, Große-Wilde E, Mißbach C, Zhang L and Breer H. 2014. Variant Ionotropic Receptors Are Expressed in Olfactory Sensory Neurons of Coeloconic Sensilla on the Antenna of the Desert Locust (Schistocerca gregaria). International Journal of Biological Sciences 10:1-14.

Haack, R. A., F. Herard, J. Sun, and J. J. Turgeon. 2010. Managing Invasive Populations of Asian Longhorned Beetle and Citrus Longhorned Beetle: A Worldwide Perspective. Annual Review of Entomology 55: 521-546.

Hanks LM. 1999. Influence of the larval host plant on reproductive strategies of cerambycid beetles. Annual Review of Entomology 44:483–505.

Hanks, L.M. and Millar, J. G. 2016. Sex and aggregation-sex pheromones of Cerambycid beetles: basic science and practical applications. Journal Of Chemical Ecology 42(7): 631-654.

Hansson, B. S.; Löfstedt, C.; Löfqvist, J.; Hallberg, E. 1986. Spatial arrangement of different types of pheromone-sensitive sensilla in a male moth. Naturwissenschaften73 (5): 269- 270.

Hoover K, Keena M, Nehme M, Wang S, Meng P, Zhang A. 2014. Sex-specific trail pheromone mediates complex mate finding behavior in Anoplophora glabripennis. Journal of Chemical Ecology 40:169–180.

Joseph RM, Devineni AV, King IFG & Heberlein U. 2009. Oviposition preference for and positional avoidance of acetic acid provide a model for competing behavioral drives in Drosophila. Proceedings of the National Academy of Sciences of the USA 106: 11352– 11357.

Kellogg. 1970. Water vapour and carbon dioxide receptors in Aedes aegypti. Journal of Insect Physiology 16: 99-108.

Kontroumpa FA, Karpati Z, Monsempes C, Hill SR, Hansson BS, Jacquin-Joly E, Kreiger J, Dekker T. 2014. Shifts in sensory neuron identity parallel differences in pheromone preference in the European corn borer. Frontiers in Ecology and Evolution 29.

Li D, Tokoro M, Nacashima T. 1999. Mechanism of adult action and mating in Anoplophora glabripennis (Motsch.). Journal of Beijing Forestry University 21:33–36.

Lopes O, Barata EN, Mustaparta H, Araújo J. 2002. Fine structure of antennal sensilla basiconica and their detection of plant volatiles in the eucalyptus woodborer, Phoracantha semipunctata Fabricius (Coleoptera: Cerambycidae). Arthropod Structure and Development 31: 1-13.

MacKay CA, Sweeny JD, Hillier NK. 2014. Morphology of antennal sensilla of the brown spruce longhorn beetle, Tetropium fuscom (Fabr.) (Coleoptera: Cerambycidae). Arthropod Structure and Development 43:469-475.

MacKay CA, Sweeny JD, Hillier NK. 2015. Olfactory Receptor neuron responses of a longhorned beetle, Tetropium fuscom (Fabr.) (Coleoptera: Cerambycidae), to pheromone, host, and non-host volatiles. Journal of Insect Physiology 83: 65-73.

Meng PS, Hoover K, Keena MA. 2015. Asian longhorned beetle (Coleoptera: Cerambycidae), an introduced pest of maple and other hardwood trees in North America and Europe. Journal of Integrated Pest Management 6: 4. Meng, P.S. K. Hoover, and M. A. Keena. 2014. Asian Longhorned Beetle (Coleoptera: Cerambycidae), an Introduced Pest of Maple and Other Hardwood Trees in North America and Europe. ournal of Integrated Pest Management 6: 4.

Meng, PS. 2014. Where's that smell? Trapping and sensory biology of the Asian longhorned beetle, Anoplophora glabripennis. Master's thesis. Pennsylvania State University, University Park, PA.

Mitchell R.F., Hughes D.T., Luetje C.W., Millar J.G., Soriano-Agatón F, Hanks L.M., Robertson H.M. 2012. Sequencing and characterizing odorant receptors of the cerambycid beetle Megacyllene caryae. Insect Biochemistry and Molecular Biology 42: 499-505.

Mitchell RF, Hall LP, Reagel PF, McKenna DD, Baker TC, Hildebrand JG. 2017. Odorant receptors and antennal lobe morphology offer a new approach to understanding olfaction in the Asian longhorned beetle. Journal of Comparative Physiology A 203:99–109.

Morewood WD, Neiner PR, Sellmer JC, Hoover K. 2003. Behavior of adult Anoplophora glabripennis on different tree species under greenhouse conditions. Journal of Insect Behavior 17: 215-226.

Nehme, M. E., M. A. Keena, A. Zhang, T. C. Baker, and K. Hoover. 2009. Attraction of Anoplophora glabripennis to male-produced pheromone and plant volatiles. Environmental Entomology 38: 1745-1755.

Nehme, M. E., M. A. Keena, A. Zhang, T. C. Baker, Z. Xu, and K. Hoover. 2010. Evaluating the use of male-produced pheromone components and plant volatiles in two trap designs to monitor Anoplophora glabripennis. Environmental Entomology 39: 169-176.

Pielou DP. 1940. The humidity behaviour of the mealworm beetle, Tenebrio molitor L. II. The humidity receptors. Journal of Experimental Biology 17:295–306.

Shanbhag SR, Müller B, Steinbrecht RA. 1999. Atlas of olfactory organs of Drosophila melanogaster: 1. Types, external organization, innervation and distribution of olfactory sensilla. International Journal of Insect Morphology and Embryology 28:377-397.

Tichy H and Kallina W. 2010 Insect hygroreceptor responses to continuous changes in humidity and air pressure. Journal of Neurophysiology 103(6): 3274–3286.

Williams DW, Lee HP, Kim IK. 2004. Distribution and Abundance of Anoplophora glabripennis (Coleoptera: Cerambycidae) in Natural Acer Stands in South Korea. Environmental Entomology 33:540-545. Zhang, A. J., J. E. Oliver, J. R. Aldrich, B. D. Wang, and V. C. Mastro. 2002. Stimulatory beetle volatiles for the Asian longhorned beetle, Anoplophora glabripennis (Motschulsky). Zeitschrift fur Naturforschung. Section C, Biosciences 57c: 553-558.

Zhang A, Oliver JE, Chauhan K, Zhao B, Xia L, Xu Z.. 2003. Evidence for contact sex recognition pheromone of the Asian longhorned beetle, Anoplophora glabripennis (Coleoptera: Cerambycidae). Naturwissenschaften 90(9):410-413.

Chapter 5 Palp Electrophysiology Indicates Well-developed Sensitivity to Moisture Levels, Along With Certain Volatile Chemicals

Introduction.

The maxillary and labial palps are insect mouthparts that are usually associated with contact chemoreception, i.e., gustation, and are equipped with sensilla containing gustatory chemoreceptive sensory neurons. Insect palps may also be used for detecting volatile chemicals when they are equipped with sensilla containing olfactory sensory neurons (OSNs) such as are seen in the maxillary palps of malaria mosquitoes (c.f. George et al. 2011). The axons of OSNs originating from their dendrites housed in insect palps terminate and arborize with second-order neurons in the antennal lobe, but in a more ventral (posterior) region than do antennal OSNs (Kent et al. 1986, Dippel et al. 2016). Most notably, dipterans are well known (Ayer and Carlson 1992) and studied (deBrune et al. 1999, Seyed and Leal 2007) for utilizing their maxillary palps for olfaction. Lepidopterans are also known for some species' olfactory utilization of palps (Kent et al. 1986), as are some orthopterans (Blaney 1973). Coleopteran palpal olfaction has likewise not been overlooked by researchers. Morphological (Giglio et al. 2013) and genetic methods (Dippel et al. 2016) have been primarily used to examine coleopteran palps for their olfactory ability.

However, there have been a very few electrophysiological studies of beetle olfaction via mouthparts. Eilers et al. (2012) discovered, using electropalpograms (EPGs) that the labial palps of carabid beetle larvae are olfactory as well as water-sensing. In electrophysiological and behavioral experiments, the palps of adults of the carabid beetle, Siagona europaea, were found by Talarico (2010) to respond to olfactory stimuli. With their impressive antennae, the Asian longhorned beetle (ALB), Anoplophora glabripennis and other cerambycids have attracted electrophysiologial attention at the antennal level (Crook et al. 2013, Toshova et al. 2016, Fan et al. 2007, Hall et al. 2006, Liendo et al. 2005), but to date there has been no study published in searchable western journal databases that has reported electrophysiological examination of palp olfaction in the Cerambycidae. Therefore, the research reported here addresses some relatively uncharted territory for coleopteran, and specifically cerambycid, experiments: interrogating the mouthparts for their ability to detect airborne volatiles. This study focused on the U.S.-invasive cerambycid species, A. glabripennis. Graves et al. 2016 suggested that the maxillary and labial palps of A. glabripennis males were important for detecting this species’ female-deposited sex- trail pheromone blend. However, the trails could seemingly not be successfully followed by males without directly contacting them, indicating that only contact chemoreception might be used for following these pheromone trails. Individuals of another cerambycid, Monochamus

alternatus, have been observed palpating around oviposition sites (Anbutsu and Togashi 1996; 1997a; 2000).

These studies raised a general question about whether the palps of cerambycids had the ability to detect airborne volatile compounds vs. having to use them only for contact chemoreception. It has been observed that A. glabripennis palps are constantly in motion and directly touching surfaces as the beetle walks along (Li et al. 1999). It had been theorized that the trail pheromone, or components thereof, are utilized by gravid females to space out eggs to prevent larval cannibalism (Graves et al. 2016) and also to aggregate on trees to overwhelm defenses (Hoover et al. 2014). But given the use of palps for olfaction in other insects, could the palps be playing a larger role than just contact chemoreception? Might they also be olfactory?

Materials and Methods

Beetles were reared on a pourable modification (Keena 2005) of a diet designed for Enaphalodes rufulus (Haldeman) at 27.5° C, chilled at 10° C for 90 days, then returned to 27.5° C during which time the larvae were allowed to develop until pupation. At this time, the pupae were transferred to 50 ml Falcon centrifuge tubes and incubated at 27.5° C until adult eclosion. Adults were fed red maple (Acer rubrum L.) twigs until preparation for electrophysiological study. Only virgin adults, about 25 days old (+/- 5 days) which had recently had their maturation feeding were used for the study

To prepare individuals for study, the palps were required to be absolutely immobile. In order to keep the tissue functioning longer, the palps were not removed from the head. Rather, the entire head was removed from the body with a razor blade and the antennae removed with scissors. The head was then held in place with dental wax on a glass slide, leaving the mouthparts exposed (Fig. 5-1). A tungsten reference probe was inserted through the open neck cavity into the neural tissue (Fig. 5-1). The opening was then closed with wax to prevent dehydration. A cover slip was cut to fit the size of the head capsule and mouthparts then covered with a piece of 3M 9474LE 300LSE super-strong double-sided adhesive transfer tape (3M, USA). This cover slip was gently interposed between the mandibles and the palps with the lateral edges embedded in the dental wax that held the head. Forceps were then used to maneuver the palps and affix them in the adhesive.

Figure 5-1. Palps prepared for Electropalpogram/Single sensillum (EPG/SSR) recordings. The head is removed from the body and held in place using dental wax with the ventral surface of the head facing upwards. The maxillary and labial palps are affixed to a glass cover slip using double-sided transfer tape. To the left, the grounding electrode is protruding from the neck opening, which is sealed with wax.

Palp Structure The maxillary and labial palps of male and female A. glabripennis consist of four and three palpomeres, respectively. Each palp segment has thick cuticle sparsely populated with setae and possible chemo- and mechanoreceptive sensilla (Fig. 5-2). The terminal segments of both pairs of palps terminate with an apical pit festooned with possible chemoreceptive sensilla (Figs. 5-2, 3).

An unexpected and interesting observation made was a flexing of the apical pits on the tips of the maxillary and labial palps. It had been assumed from scanning electron microscope (SEM) images that the palp tip consisted of a concave apical pit harboring stubby sensilla (Fig. 5-3), but it appears that this fixing of the palp in that concave position was an artifact of preparation for SEM imaging. The apical membranes of living palp tips containing the sensilla were observed to regularly flex from concave to convex, extending the sensilla out into the environment (Fig. 5-2). Videos of this movement can be found in the supplementary materials or viewed at https://youtu.be/3RGu4_aClRo.

Figure 5-2. A. The labial and maxillary palps. The apical pit of the maxillary palp can be seen flexed in. Insert shows an enlarged image of the flexed-in apical pit. B. The labial and maxillary palps showing the apical pits of both palps in a flexed-out position. Inset shows an enlarged image of both apical pits. Arrow points to the array of multiple sensilla that are now visible because the “pit” is now more of a dome during outward flexing of the apical pit membrane. Video of this apical pit activity can be viewed at https://youtu.be/3RGu4_aClRo.

Figure 5-3. SEM image of the tip of a female ALB maxillary palp, showing the presumed concave architecture and many sensilla of the apical pit (ap) surface. From P. S. Meng. 2014, Master’s Thesis, Penn State University.

Figure 5-4. A. Light microscope image of the maxillary palp. Coeloconic sensilla (cs) and apical pit (ap) are indicated. B. SEM image of the tip of a female ALB maxillary palp, arrows indicate coeloconic sensilla. Scale bar = 10 μm. C. SEM high magnification image of a coeloconic sensillum on the maxillary palp. Scale bar = 5 μm. (B and C are from P. S. Meng. 2014, Master’s Thesis, Penn State niversity.)

Electropalpograms.

For electropalpogram recordings (EPGs), tungsten electrodes were prepared by electrolytically sharpening them using a saturated KNO2 solution at 10V. The sharpened tungsten wire was placed into contact with the membrane comprising the apical pit of either the maxillary or labial palp (Fig. 5-5) using a Narishige hydraulic micromanipulator, completing the electrical circuit and allowing the collective neural activity of the palp to be recorded (Fig. 5-6). A custom high- impedance DC-coupled preamplifier and 16 bit analog-to-digital converter with a conversion rate of 46875 Sa/s and voltage range of -1 to +1 volts was used to make recordings. These data were then re-sampled at 10 KSa/s and stored for later analysis.

Figure 5-5. Tungsten probe (TPb) placed against the apical pit of the palp (ap) for EPG recording.

Figure 5-6. A hyperpolarizing EPG response from a male labial palp in response to a puff containing 100µg of geraniol.

The odorants used in this study (Table 5-1) were diluted in hexane in 10-fold steps to different concentrations such that odor cartridges contained loadings of either 1 µg, 10 µg, or 100 µg of an odorant when aliquots of 10 μl of each odorant dilution were dispensed onto a filter paper strip (Whatman; 0.2 X 1.5 cm). Each strip was inserted into a 15 cm Pasteur pipette to create each of the odor cartridges. A constant airflow of charcoal-purified, humidified air was passed across the

palps through a glass tube (10 mm diameter) during the experiments. Odorants were delivered into this constant air stream via the Pasteur pipette whose tip was inserted through a small hole in the glass tube, 11 cm away from its end. A stimulus flow controller (Syntech; Germany) pulsed a 40 ml/s air stream through the cartridge for 0.3 seconds, effectively delivering a puff of volatiles from the odor cartridge into the air stream and onto the antenna. Odor cartridges were remade every 30 puffs for lower volatility odorants and every 3 puffs for the highly volatile odorants.

Single Sensillum Recordings.

Single sensillum recordings (SSR) were conducted similarly except for the recording electrode placement. The electrode was placed against the base of one of the coeloconic sensilla that sparsely populate the outer cuticle along the last segment of the maxillary palps (Fig. 5-4), after which I could extracellularly record the action potentials of olfactory sensory neurons (OSNs) within the sensillum. The output from the electrode could also be configured to now simultaneously record the DC potential from within that sensillum along with SSR action potentials by adjusting the highpass cutoff to 500Hz. In this configuration, the recordings of both the DC depolarizations and the AC action potential spike trains happening within the sensillum could be compared.

Table 5-1. Odorants used in this study, along with their commercial sources and purities.

Odorant Supplier Purity (%) citronellal Acros Organics B.V.B.A. (Belgium) 93 4-(n-heptyloxy) butan-1-ol Bedoukian Research Inc. (Danbury, CT, USA) 4-(n-heptyloxy) butanal Bedoukian Research Inc. (Danbury, CT, USA) acetic acid, J.T. Baker, Inc. (Phillipsburg, NJ, USA) 100 (3E,6E)- -farnesene Jocelyn Millar Laboratory, University of CA, Riverside 86 trail pheromone, whole blend Kelli Hoover Laboratory, University of PA trail pheromone, major components Kelli Hoover Laboratory, University of PA trail pheromone, minor components Kelli Hoover Laboratory, University of PA isovaleric acid Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 benzoic acid Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 Z-2-hexenal Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 butyraldehyde Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 1-4-diaminobutane Sigma-Aldrich Corporation (St. Louis, MO, USA) 97 β-caryophellene, Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 Z-3-hexen-1-ol, Sigma-Aldrich Corporation (St. Louis, MO, USA) 98 linalool, Sigma-Aldrich Corporation (St. Louis, MO, USA) 99 -terpineol, Sigma-Aldrich Corporation (St. Louis, MO, USA) 96 geraniol Tokyo Chemical Industry Co., Ltd. (Japan) 96 eugenol Tokyo Chemical Industry Co., Ltd. (Japan) 98

Sensitivity to Moisture Levels

Both labial and maxillary palps, for both sexes, demonstrated a sensitivity to changes in the amount of water vapor present in the constant airstream, strongly suggesting that there are abundant hygroreceptors on the labial and maxillary palps. When introducing a puff from the Syntech controller using filtered, ambient air, having a low moisture content due to the winter weather, into the constant humidified air stream, there was a large EPG response to a blank. Upon further investigation, adding water to the cartridge reduced the EPG response appreciably. The need for reducing responsiveness to drier air in the puffs is shown in this initial experiment in which a puff-humidifying apparatus (explained in more detail later) was not used, only blank tubing upstream of the odor cartridge. Three different loadings of water, from 1µl to 100µl were loaded onto the filter papers of three different odor cartridges, and no water at all was in the fourth, dry-blank cartridge. Electropalpograms (EPGs) of the labial and maxillary palps in response to various changes in moisture content resulted in altered response levels of the neurons in the palps in a dose-responsive manner according to a Kruskal Wallis test followed by a Wilcoxon signed rank test (p ≤ 0.05). N = 9 per palp type per sex. Injections of drier air caused depolarization whereas injections of more humid air caused hyperpolarization (Fig. 5-7).

Figure 5-7. Series of EPGs from a single male maxillary palp bathed in a humidified airstream to a succession of puffs using dry, unhumidified air to introduce dosages of water on filter paper of 0 µg, 1µg,

10µg, or 100µg (A, B, C, and D, respectively). With increasing amounts of water in the puffs, the responses changed from hyperpolarizations (A, B, C) to depolarizations (D).

The drier or wetter the injection, the greater was the depolarization or hyperpolarization, respectively. Male palps appeared to exhibit shallower dose-response reactions to decreasing moisture levels than did female palps (Fig. 5-8). There was also a difference in responsiveness to moisture-level changes between maxillary and labial palps, with labial palps responding to variations in moisture content in a more pronounced fashion than maxillary palps (Fig. 5-8).

Figure 5-8. Box plots of EPGs evoked by puffs from odor cartridges loaded with different dosages of water. Puffs through the cartridges were made using a dry, unhumidified puffing stream, with the pulse exiting the cartridge then entering into the humidified constant air stream flowing over the palps to evoke an EPG. Positive mV values indicate degree of depolarization evoked by the stimulus that moves the OSNs from their normal negative resting potential values in the palps. Negative mV values indicate an increased hyperpolarization of the OSNs (more negative potential) evoked by the stimulus from their already-negative resting potentials. Letters denote statistically different responses per Kruskal Wallis test followed by a Wilcoxon signed rank test (p ≤ 0.05). N = 9.

For this reason we designed an apparatus to add water vapor to puff of air that was injected into and through the odor cartridge. For the body of this experiment, puffs, when humidified, were first passed through a 15 ml centrifuge tube containing a cotton wick saturated with distilled water and suspended so as not to impede airflow (Fig. 5-9). The tubing was held in place and made air-tight with Sugru (FormFormForm, UK), a rubberized, self-setting putty. This

humidifier cylinder was attached to, and positioned in the air-puff injection line immediately upstream of the odor cartridge. For non-humidified puffs, the humidifier cylinder was swapped out with an equal length of tubing.

Figure 5-9. Humidifying apparatus for injection line and its placement in the setup.

The removal of this confounding response to water vapor reduction is demonstrated by comparing EPGs with and without this humidifying apparatus. The change in water vapor produced some complex combinatorial effects in EPG responses (Fig. 5-10). Puffs from cartridges containing either geraniol, water, or empty filter paper (blanks) – with the puffing airstream being delivered either by passing through the humidifying apparatus or an equal length of plain tubing – were injected into the humidified constant airstream flowing over the palps. An F-test showed that the variances between humidified and unhumidified puffs were unequal in many cases and thus a Wilcoxon signed rank test was used to show differences in EPG magnitude between the two types of puffs for each odorant. Puffs of geraniol or blanks using drier, unhumidified air through the odor cartridge and into the constant humidified airstream produced more hyperpolarized (increased negative mV) EPGs than injections of geraniol or blanks using humidified air for the puffs for both sexes in the labial palps. Importantly, a decrease in moisture passing over the palps from dry air puffed through the cartridges yielded larger variations in responses than when humidified air was puffed through the same cartridges.

Figure 5-10. Effect of decreases or increases in moisture levels imparted by the odor cartridge on the responses, showing EPG responses for blank, geraniol, and water injected with and without a humidified injection stream into a humidified constant airstream. Differing letters denote statistical significance via Wilcoxon signed rank tests performed between the humidified and unhumidified puffs of each odorant, p ≤ 0.05, N=9. Differences in variances between humidified and unhumidified puffs of each odorant, as measured via an F-test, p ≤ 0.05, N=9, are also labeled. Geraniol was presented from cartridges loaded at a dosage of 100μg and water loaded with 100μl. Increasingly positive mV values indicate an increased

degree of depolarization by the stimulus of the normal negative resting potentials of the OSNs in the palps. Increased negative mV values indicate an increased hyperpolarization of the negative resting potentials of the OSNs by the stimulus.

When conducting EPG measurements of the full panel of odorants the order of odorants used was constant and thus measurements of a control odorant, geraniol, were taken at the beginning, middle, and end of the series for every individual palp in order to measure the decay of the tissue/connection for each individual palp. These three measurements were then used to adjust the EPG measurements of the odorants measured, according to their order of use in the panel, to account for this decay for each individual palp.

Results

The above EPG results indicate that the maxillary and labial palps of both sexes are able to respond to increases or decreases in the abundance of water molecules. In additional experiments discussed below, a wide array of odorants was tested, with any changes in moisture levels from the puffed airstream through the odor cartridges tempered through the use of a humidified puffed-air stream through the cartridge that exited into the humidified air constantly flowing over the palps. Several of the odorants produced EPG responses that were significantly different from a blank, as indicated by paired t-tests. These results indicated an olfactory ability by the neurons in the labial or maxillary palps to detect certain odorants, but not others.

Using an enlarged panel of odorants, we found that maxillary palps responded with significant EPGs to the volatiles of only a few odorants. Significant maxillary palp EPGs only occurred in response to acetic acid, butyraldehyde, and water (Figs. 5-11, 12).

Labial palps responded with significant EPGs in response to a greater variety of odorants than did maxillary palps, including several plant compounds (geraniol, eugenol, terpeniol and (E,E)-  -farnesene, also a possible minor sex-aggregation pheromone component (Crook et. al. 2014)). Significant EPGs also were elicited by the blend of sex-trail pheromone (Hoover et al. 2014), in addition to acetic acid, butyraldehyde, (E)-2-hexenal, and water. There were some sexual differences in EPG responses, in that the female labial palps were able to respond significantly to the ALB aldehyde (4-heptyloxy-butanal) component of the male-produced sex-aggregation pheromone blend. There were also some sex-related differences in responses to three other plant compounds, with only the labial palps of males exhibiting significant EPGs to linalool and - caryophyllene and those of females giving significant responses to (Z)-3-hexenol (Figs. 5-11, 12).

Figure 5-11 A-D. EPG responses by male and female maxillary and labial palps to different odorants. "Y" indicates that, yes, the response was significantly different from the blank response via paired t-tests vs the blank, p ≤ 0.05; N = 11. Unless otherwise noted odorants are at 100μg strength. Water is loaded with 100μl. (Though significantly different from the blank, the EPG data for acetic acid and butyraldehyde are not shown on these graphs, due to the relatively large magnitude of those EPGs, but those EPGs are included in the next set of data (Fig. 5-12) comparing amplitudes of EPGs.)

We analyzed the data to statistically look for differences in the amplitudes of EPG responses occurring among all the various odorants, not just between each odorant and the blank using a Kruskal-Wallis test followed by a Wilcoxon signed rank test, p ≤ 0.05. This series also included the odorants, acetic acid and butyraldehyde. For all palp types, acetic acid and butyraldehyde elicited the strongest EPG depolarization responses, followed by the depolarization response amplitude to water. Nearly all the other odorants tested that produced EPGs significantly different than the blank were hyperpolarizations (Fig. 5-12 below).

Figure 5-12 A-D. EPG responses to different odorants from male and female maxillary and labial palps, including responses to acetic acid and butyraldehyde. EPG amplitudes for different odorants having no letters in common are significantly different according to a Kruskal-Wallis test followed by a Wilcoxon signed rank test, p ≤ 0.05; N=11. Unless otherwise noted odorants are at 100µg cartridge loadings and water is loaded at 100μl.

A tonic depolarization resulted both from acetic acid and butyraldehyde in both labial and maxillary palps for both sexes. The initial response was usually a small, brief hyperpolarization that was followed by, or often seemingly masked by, the large tonic depolarization that followed and lasted for sometimes 10 or more seconds (Fig. 5-13). For the EPG analyses, only the initial, peak level of the tonic depolarization was used for an amplitude measurement.

Figure 5-13. A) EPG tracing of female maxillary palp response to butyraldehyde showing tonic depolarization. B) EPG tracing of a male maxillary palp response to acetic acid.

Single Sensillum Recordings

Single sensillum recordings (SSRs) were successfully conducted on twelve coeloconic sensilla on the maxillary palps, four sensilla on males and eight on females. These sensilla were located along the side of the terminal palp segment, slightly proximal from the palp’s terminal apical pit (Fig. 5-4). SSRs had been attempted repeatedly on the sensilla located within the apical pit of the palps, but non-noisy and stable connections were unsuccessful. Representative SSR recordings are shown in Figs. 5-14 and 5-15 below. The simultaneous DC potential tracings obtained along with AC spikes tracings show the depolarizing DC current within the sensillum responsible for generating spikes from the responding OSN (Figs. 5-14, 15).

Figure 5-14. SSR tracing of an ORN from a male coeloconic sensillum responding to butyraldehyde simultaneously showing the DC current from its EPG (or possibly a single-sensillum electrosensillogram (ESG) from this sensillum).

Figure 5-15. SSR tracing of an ORN from a female coeloconic sensillum responding to butyraldehyde simultaneously showing the DC current from its EPG (or possibly a single-sensillum electrosensillogram (ESG) from this sensillum).

100

Table 5-2. Change in pre- to post-stimulus spike frequency (spikes/sec) of OSNs in coeloconic sensilla on maxillary palps of males and females in response to 100ug of odorant. Acetic Butyrald- Acid Blank ehyde Water Trail Males +2 +6 +2 +2 +6 0 +9 0 +2 +28 -1 +12 0 +12 +3 1 +6 -1 +5 Females +6 +14 0 0 +10 +20 -3 0 -1 +2 -4 +3 +2 +4 +10 +12 +2 0 +5 +4 0 -12 0 -9 +8 -6 0 +2 +11 -5 -1

Discussion

The maxillary and labial palps of male and female adult A. glabripennis do have olfactory ability. This, combined with the dearth of coleopteran, and particularly cerambycid, palp olfaction studies indicates that the palps should also receive attention when trying to understand this beetle's olfaction.

The initial impetus for this study stemmed from the results of Graves et al. 2016, which suggested that the maxillary and labial palps of A. glabripennis males were important for detecting this species’ female-deposited sex-trail pheromone blend. They found that the trails could seemingly only be successfully followed by males when directly contacting them with their palps, and not through airborne exposure to the trails from even a few mm away. The EPG results presented here show that although the male labial palps respond with significant, although weak, EPGs to the major trail pheromone components, the palps also respond to many other common general odorants at the same low but significant amplitudes as to the sex-trail pheromone. Thus the EPG responses to this pheromone blend may not represent a specific response to the pheromone, but rather to any of a large array of possible volatiles from plant or detritus sources that have close structural similarity to the pheromone components. To show that there are OSNs on the palps that are specifically tuned to the sex-trail pheromone one would have to perform SSR recordings on sensilla that house such OSNs and interrogate them with general odorants as well as the sex-trail pheromone components.

Despite exhaustive efforts, we were not able to obtain SSR recordings from the large fields of sensilla that populate the terminal pit on either the labial or maxillary palps. Thus, the results of this study in not finding some kind of specific, diagnostic EPG responses in response to the sex-

101

trail pheromone are thus far consistent with the behavioral findings of Graves et al. (2016) that indicate that contact chemoreception, not olfaction, is what is employed by males to detect and follow the trails of females. It may be that the sensilla in the terminal pits are gustatory, requiring contact with the trail pheromone, and because gustatory tip-recording methods (requiring an entirely different setup and equipment) were not used, we were not able to succeed in acquiring SSR recordings from the pit-region sensilla.

Other evidence that the pit sensilla might be involved with contact chemoreception, possibly including sex-trail-pheromone-following, is related to the discovery of the ability of live ALB to flex the membrane of the apical pits outward from a concave to a convex geometry to now expose the sensilla to direct contact with surfaces. This was a quite a surprise and seems to be quite pertinent to the discussion of how ALB detect the sex-trail pheromone. Our observations show that ALB can alternatively flex the membrane of the palp tip from an inward concave, to an outward convex, position to allow the sensilla to project out into the environment for chemosensing. The flexing of the palp-pit membrane strongly implies that it is used for sensing of some kind and leads to a plethora of interesting questions. For instance, the results of Graves et al. (2016) provided evidence that the palps are used to follow the species’ sex-trail pheromone by contact alone. Are the sensilla at the palp termini flexed out for trail pheromone contact-sensing? Also, how often does the flexing of the terminal pit membrane happen when the insect is exploring or at rest, rather than in a highly unnatural and stressful situation of being prepared for electrophysiological testing? How does the convex flexing and subsequent exposure of the sensilla relate to contact chemoreception vs. olfaction, especially with regard to detecting and responding to sex-trail pheromone? Are these terminal sensilla primarily gustatory? Our current results show that at least some of the chemo-sensing neurons on the palps are capable of olfactory responses, but any sensory neurons in the terminal pit remain unrecorded from. Without electrophysiological evidence one cannot say for certain that the apical pit sensilla are used for chemosensing and whether or not that would include olfaction or gustation.

The only direct evidence that OSNs on the palps are capable of olfaction comes from SSR responses from OSNs housed in coeloconic sensilla housed in sensilla on the sides of the terminal segment cuticle, not from sensilla clustered in the terminal pits. Therefore, we are only able to conclude at present that the significant EPGs recorded from the palps were due to the summed depolarizations of OSNs on the sides of the palps, and no definite conclusions can be made about the olfactory ability of the apical pits. Coeloconic sensilla are known to house OSNs that are tuned mostly to aldehydes and acids (Benton et al. 2009; Ai et al. 2010, Yao et al. 2005, van Giesen and Garrity 2017), consistent with the SSR responses found from these sensilla. However, it seems likely that there are other types of chemoreceptive sensilla on the palps other than coeloconic, which will house OSNs responsive to other types of volatiles such as acetates, alcohols, hydrocarbons, etc., and explain the weak but significant EPGs obtained to many different types of odorants tested.

102

The labial palps of A. glabripennis seem to be more responsive to odorants than the maxillary palps, responding with EPGs to a greater array of odorants. This seems to differ from the palps of dipterans, especially mosquitoes, whose maxillary palps have olfactory function. However, like the cerambycid representative A. glabripennis, members of the Lepidoptera (Kent et al. 1986), do utilize labial palps for olfaction as well. Due to the sample size, our study was not very robust concerning the relatively small panel of odorants we tested, and so it is likely that a wider range of odorant molecules might elicit reactions from the palps due to the broadly tuned characteristic of most olfactory receptors (Hallem et al. 2004). Some sexual differences in EPG responses were observed. Male EPGs were greater in amplitude than those of females, despite the known smaller average size of males (Meng et al. 2015). This result could imply that males have more neurons dedicated to odorant detection in their palps than do females. The difference between male and female EPGs might reflect a difference in ability to detect odorants, but it may also simply be an artifact of sample size and thus we are cautious in drawing such a conclusion. A larger study would be necessary to elucidate such a sexual difference more definitively.

The EPG recordings from this study show that palps of A. glabripennis are very sensitive to changes in moisture levels, and SSR recordings identified at least one type of sensillum containing OSNs that respond to water. Also, the multiple short peg sensilla located in the apical pits of the labial and maxillary palps (Fig. 5-3) might also be hygroreceptors and contribute strongly to the EPGs we recorded in response to water concentrations. According to Altner et al. (1983), all hygroreceptors that have been characterized share the feature of being stubby, non- pore-walled pegs of various lengths that always occur in cuticular depressions of some type. The sensilla in the apical pits seem to fit this classification in all respects except that here we have multitudes of pegs in a single, large pit. The EPG results showed that labial palp neurons were the most sensitive to changes in moisture levels, able to detect smaller gradations in humidity than neurons on maxillary palps.

To identify moisture content could be a way to insure that there is healthy tissue for feeding, oviposition, and subsequent larval feeding as well as to avoid dessication. The adults of A. glabripennis feed on living trees and females lay their eggs in living wood (Meng et al. 2015). Desiccation is a danger for insects and the ability of A. glabripennis to identify favorably moist microenvironments might be important (Enjin et al. 2016) especially because it is thought that this species may have initially evolved in moist, riparian habitats (Williams et al. 2004). It could possibly be argued that the response to blanks and to water was the result of mechanoreceptors responding to air pressure changes. However, although it is possible that mechanoreceptors contribute somewhat to the EPGs, the dose-response curves from this study in response to increasing or decreasing amounts of water puffed over the palps indicate that the EPG responses obtained here were significantly elicited by the abundance of water molecules in the puffs from the cartridges. Hygroreceptors have routinely been found on insect antennae through electrophysiological studies (c.f.,Pielou 1940, Altner et al. 1983, Iwasaki et al. 1995, Tichy and Kallina 2010, Enjin et al. 2016) as well as on palps of larvae and adults (Eilers et al. 2012 and

103

Chappuis et al. 2013, respectively). For cerambycid beetles, the findings presented here may be informative in helping to start finding more hygroreceptive sensory neurons in other species in this family.

For herbivorous beetles like A. glabripennis, acetic acid detection could play a role in finding feeding/oviposition sites. Acetic acid is used to judge possible food for suitability by other herbivorous insects (George et al. 2016, Ômura & Honda 2003, Joseph et al.. 2009) and so detection of acetic acid with mouthparts could allow A. glabripennis to gather information about the state of a food source or oviposition site prior to biting into it. This could allow these insects to more effectively select higher quality feeding/oviposition sites and avoid less than ideal resources.

The beetles in this study were able to detect several volatiles associated with plant stress via their labial palps, including well known stress-related volatiles (Z)-3-hexen-1-ol, (E,E)-α-farnesene, and geraniol. It has been theorized that it is advantageous for larval survival for A. glabripennis to overwhelm a tree's defenses with numbers (Hoover et al. 2014). Thus detecting stressed plants when searching for and biting oviposition pits might be a useful adaptation.

The recordings of coeloconic sensilla via SSR show responsiveness to each of the odorants tested against them. However each sensillum tested did not respond to all odorants, only to some of them (Table 5-2). The recordings, themselves, show multiple action potential amplitudes (Figs. 5-14, 15). In sensilla containing multiple OSNs, the amplitudes of each neuron are often distinctive and thus these recordings may indicate more than one neuron in the coeloconic sensilla we recorded from. Perhaps each coeloconic sensilla has the same cadre of neurons and their receptors but we did not record from all neurons in the sensilla at once or, more in line with drosophila research by Yao et al. (2005), each of the sensilla have different sets of neurons their receptors, resulting in different sensilla having different response profiles. The body of knowledge regarding IRs is growing quickly and IRs have been identified responding to a broad range of stimuli including water, acids, general odorants, temperature, and tastes (vanGiesen and Garrity 2017, Hussain et al. 2016). With such a small number of connections and odorants tested via SSR, our study barely scratched the surface of what coeloconic sensilla on the palps are responding to and what sensory neurons are involved.

In the EPG and SSR experiments, my finding of tonic depolarization (Figs. 5-13, 14) and spike initiation by acetic acid and butyraldehyde was a curious thing. Why do these types of molecules activate neural responses so strongly for ten seconds or more but evoke such a short spike train? It could be argued that highly volatile odorants such as these might deliver a mega-dose more quickly than other odorants, but other highly volatile odors tested here did not elicit such long- term reactions. Also, why would such high volatility compound produce a more tonic DC response rather than a highly phasic response? Are there other odorants that elicit this kind of response? Perhaps part of the answer lies in with the fact that OSNs housed in coeloconic sensilla use ionotropic receptors (IRs) (Benton et al. 2009;Guo et al. 2014, Ai et al. 2010 ) and

104

not odorant receptors (ORs), and so the deactivation mechanisms in the perireceptor environment within these coeloconic sensilla may not be designed for fast clearing-out of these acid and aldehyde analytes.

Due to the small number and short duration of SSR connections made, we did not test responsiveness to plant-stress or other general odorants. Since EPGs do show a responsiveness to several of these odorants, there must be sensilla containing OSNs responding to them. Are these responding sensilla the coeloconic sensilla or perhaps to the as yet unrecorded from sensilla in the apical pit? Meng (2014) identified 5 kinds of sensilla on the palps via SEM: coeloconic, the stubby pegs in the apical pit, long chetiform, medium chetiform, and a much smaller "small chetiform" sensilla. The coeloconic and apical pit sensilla are the most likely candidates, though without TEM studies of the cross sections of the small chetiform ruling out pores and encased neurons, we are hesitant to disregard them for olfactory purposes. It could be beneficial to invest in a study utilizing TEM and subsequent SSR on each type of sensilla to identify to source of the EPGs to these other odorants.

Future studies could widen our understanding of what arrays of odorants are being detected by the palps, and how strongly they respond, by using a larger panel of odorants and testing these using gustatory (contact) as well as olfactory chemo-recording techniques to gain robust data sets that include dose-response experiments. A single study like this is a good start, but the possibility for false negatives in olfactory studies demands replication. Also important is how the ability to detect an odorant translates into behavior. More behavioral assays such as the olfactometer studies of Graves et al. (2016) can build on those results and further explore the effect of ablation/silencing of the palps on behavior. Such studies are needed to further tell us how large (or small) of a role that palp olfaction plays in this beetle species. Also, further observational studies of the apical pit movement in a more natural situation might reveal how the flexing of the sensillar fields at the termini of the palps is utilized in nature and therefore provide more information as to the possible function of these terminal sensilla and what stimuli might be used to try to activate the sensory neurons in them.

A major conclusion from this study is that the palps in this species, and possibly those of other cerambycid beetles, are far more involved in sensing volatiles, and particularly moisture levels, than might have been expected. We are getting only a partial picture of olfaction in Anoplophora glabripennis, and likely other cerambycids, when we look only at the antennae, as the palps also have olfactory capabilities waiting to be better understood.

105

References

Ai M, Min S, Grosjean Y, Leblanc C, Bell R, Benton R, Suh GS. 2010. Acid sensing by the Drosophila olfactory system. Nature 468: 691-695.

Altner H, Schaller-Selzer L, Stetter H, Wohlrab I. 1983. Poreless sensilla with inflexible sockets. A comparative study of a fundamental type of insect sensilla probably comprising thermo- and hygroreceptors. Cell and Tissue Research 234: 279-307.Anbutsu H, Togashi K (1996) Deterred oviposition of Monochamus alternatus (Coleoptera: Cerambycidae) on Pinus densiflora bolts from oviposition scars containing eggs or larvae. Applied Entomological Zoology 31: 481–488.

Anbutsu H, Togashi K (1997a) Oviposition behavior and response to the oviposition scars occupied by eggs in Monochamus saltuarius (Coleoptera: Cerambycidae). Applied Entomological Zoology 32: 541–549.

Anbutsu H, Togashi K (2000) Deterred oviposition response of Monochamus alternatus (Coleoptera: Cerambycidae) to oviposition scars occupied by eggs. Agricultural and Forest Entomology 2: 217–223.

Ayer RK and Carlson J. 1992. Olfactory physiology in the Drosophila antenna and maxillary palp: acj6 distinguishes two classes of odorant pathways. Journal of Neurobiology 23: 965-982.

Benton R, Vannice KS, Gomez-Diaz C, Vosshall LB. 2009. Variant ionotropic glutamate receptors as chemosensory receptors in Drosophila. Cell 136:149-162.

Blaney WM. 1973. Electrophysiological responses of the terminal sensilla on the maxillary palps of Locusta migratoria (L.) to some electrolytes and non-electrolytes. Journal of Experimental Biology 60: 275-293.

Chappuis CJF, Béguin S, Vilmant M, Guerin PM. 2013. Water vapour and heat combine to elicit biting and biting persistence in tsetse. Parasites & Vectors 6:240.

Crook DJ, Lance DR & Mastro VC. 2014. Identification of a potential third component of the male-produced pheromone of Anoplophora glabripennis and its effect on behavior. Journal of Chemical Ecology 40: 1241-1250. de Bruyne M, Clyne PJ, Carlson JR. 1999. Odor Coding in a Model Olfactory Organ: The Drosophila Maxillary Palp. The Journal of Neuroscience 19:4520–4532.

Dippel S, Kollmann M, Oberhofer G, Montino A, Knoll C, Krala M, Rexer KH, Frank S, Kumpf R, Schachtner J and Wimmer EA. 2016. Morphological and transcriptomic analysis of a beetle chemosensory system reveals a gnathal olfactory center. Bio Med Central Biology 14:90.

106

Enjin A, Zaharieva EE, Frank DD, Mansourian S, Suh GSB, Gallio M, Stensmyr MC. 2016. Humidity Sensing in Drosophila. Current Biology 26: 1352–1358.

Fan J, Kang L, Sun J. 2007. Role of host volatiles in mate location by the japanese pine sawyer, Monochamus alternatus hope (Coleoptera: Cerambycidae). Environmental Entomology 36:58-63.

Giglio A, Perrotta E, Talarico F, Zetto Brandmayr T and Ferrero EA. 2013. Sensilla on maxillary and labial palps in a helicophagous ground beetle larva (Coleoptera, Carabidae). Acta Zoologica (Stockholm) 94: 324–330.

Graves F, Baker TC, Zhang A. Keena M, Hoover K. 2016. Sensory aspects of trail-following behaviors in the Asian longhorned beetle, Anoplophora glabripennis. Journal of Insect Behavior 29: 615-628.

Hall DR, Cork A, Phythian SJ, Chittamuru S, Jayarama BK, Venkatesha MG, Sreedharan K, Vinod Kumar PK, Seetharama HG, Naidu R. 2006. Identification of components of male- produced pheromone of coffee white stemborer, Xylotrechus quadripes. Journal of Chemical Ecology Volume 32: 195–219.

Hallem EA, Ho MG, Carlson JR. 2004. The molecular basis of odor coding in the Drosophila antenna. Cell 117: 965-979.

Hoover K, Keena M, Nehme M, Wang S, Meng P, Zhang A. 2014. Sex-specific trail pheromone mediates complex mate finding behavior in Anoplophora glabripennis. Journal of Chemical Ecology 40:169–180.

Hussain A, Zhang M, Üçpunar HK, Svensson T, Quillery E, Gompel N, Ignell R, Grunwald- Kadow IC. 2016. Ionotropic Chemosensory Receptors Mediate the Taste and Smell of Polyamines. PLoS Biology 14: e1002454.

107

Iwasaki M, Itoh T, Yokohari F, Tominaga Y. 1995. Identification of antennal hygroreceptive sensillum and other sensilla of the firefly, Luciola cuciata. Zoological Science 12:725- 732.

Kelling FJ, Biancaniello G, and den Otter CJ. 2003. Effect of age and sex on the sensitivity of antennal and palpal olfactory cells of houseflies. Entomologia Experimentalis et Applicata 106: 45–51.

Kent KS, Harrow ID, Quartararo P, and Hildebrand JG. 1986. An accessory olfactory pathway in Lepidoptera: the labial pit organ and its central projections in Manduca sexta and certain other sphinx moths and silk moths. Cell and Tissue Research 245: 237-245.

Li D, Tokoro M, Nacashima T. 1999. Mechanism of adult action and mating in Anoplophora glabripennis (Motsch.). Journal of Beijing Forestry University 21:33–36.

Liendo C, Morillo F, Sánchez P, Muñoz W, Guerra J, Cabrera A, and Hernández JV. 2005. Olfactory behavior and electroantennographic responses of the cocoa beetle, Steirastoma breve (Coleoptera : Cerambycidae). Florida Entomologist 88(2):117-122.

Meng PS. 2014. Where's that smell? Trapping and sensory biology of the Asian longhorned beetle, Anoplophora glabripennis. Master's thesis. Pennsylvania State University, University Park, PA.

Ômura H & Honda K 2003. Feeding responses of adult butterflies, Nymphalis xanthomelas, Kaniska canace and Vanessa indica, to components in tree sap and rotting fruits: synergistic effects of ethanol and acetic acid on sugar responsiveness. Journal of Insect Physiology 49: 1031–1038.

Pielou DP. 1940. The humidity behaviour of the mealworm beetle, Tenebrio molitor L. II. The humidity receptors. Journal of Experimental Biology 17:295–306.

108

Syed Z and Leal WS. 2007. Maxillary Palps Are Broad Spectrum Odorant Detectors in Culex quinquefasciatus. Chemical Senses 32: 727–738.

Talarico F, Giulianini Pg, Brandmayr P, Giglio A, Masala C, Sollai G, Zetto Brandmayr T, Solari P. 2010. Electrophysiological and behavioural analyses on prey searching in the myrmecophagous carabid beetle Siagona europaea (Dejean 1826). Ethology Ecology & Evolution 22: 375–384.

Tichy H and Kallina W. 2010 Insect hygroreceptor responses to continuous changes in humidity and air pressure. Journal of Neurophysiology 103: 3274–3286.

Toshova TB, Subchev M, Abaev V, Vuts J, Imrei Z, Koczor S, Galli Z, Van De Ven R, Tóth M. (2016). Responses of Pseudovadonia livida adults to olfactory and visual cues. Bulletin of Insectology 69: 161-172 van Giesen L and Garrity PA. 2017. More than meets the IR: the expanding roles of variant Ionotropic Glutamate Receptors in sensing odor, taste, temperature and moisture. F1000Research 6:1753.

Williams DW, Lee HP, Kim IK. 2004. Distribution and Abundance of Anoplophora glabripennis (Coleoptera: Cerambycidae) in Natural Acer Stands in South Korea. Environmental Entomology 33:540-545. Yao CA, Ignell R, Carlson JR. 2005. Chemosensory coding by neurons in the coeloconic sensilla of the Drosophila antenna. Journal of Neuroscience 25: 8359-8367.

Zacharuk RY, Albert PJ, Bellamy FW. 1977. Ultrastructure and function of digitifom sensilla on the labial palp of a larval elaterid (Coleoptera). Canadian Journal of Zoology 55: 569- 578.

109

Vita

Family name, First name: Hall, Loyal

Education: Jan 2013- Jan 2018 PhD Department of Entomology Major Professor: Thomas C Baker The Pennsylvania State University, USA Sept 2008- May 2011 Master of Science Department of Entomology Major Professor: Thomas C Baker The Pennsylvania State University, USA Sept 1996- May 2000 Bachelor of Science Department of Biology Grove City College, USA

Participation in Industrial Innovation Creation, revision, and testing of a novel mosquito trap. Hall, Loyal P. (2012) Species-specificity of three commonly used and two novel mosquito field-sampling devices. Master’s Thesis, The Pennsylvania State niversity. https://etda.libraries.psu.edu/catalog/13886

Teaching Activities 2004-2012 County Extension Educator, Vector Management. PSU, USA 2013 Bachelor students- ENT202 Introduction to Entomology (for non-science students). PSU, USA 2014 4thyr Bachelor students- ENT314 Management of Insect Pests of Ornamentals. PSU, USA 2015 3rd-4thyr Bachelor students- ENT313 Introduction to Entomology (for science students). PSU, USA

Awards Time’s 2006 Person of the Year

Publications Hall, Loyal P. et al. (2018) Palpal sensitivity of Anoplophora glabripennis (Coleoptera: cerambycidae) palps to various odorants and to humidity. (in preparation)

Hall, Loyal P. et al. (2018) Odorant sensitivity differentiation along the length of Anoplophora glabripennis (Coleoptera: cerambycidae) antennae indicates OSN grouping. (in preparation)

Mitchell, R.F., Hall, L.P., et al. (2017). Odorant receptors and antennal lobe morphology offer a new approach to understanding olfaction in the Asian longhorned beetle. J Comp Physiol A 203: 99 – 109.

Domingue, Michael J.et al. (2016) Host condition effects upon Agrilus planipennis (Coleoptera: Buprestidae) captures on decoy-baited branch traps. European Journal of Entomology. 113: 438-445.

Domingue, Michael, et al. (2015) "Detecting Emerald Ash Borers ( Agrilus planipennis) Using Branch Traps Baited with 3D-Printed Beetle Decoys." Journal of Pest Science, 88(2): 267.

Domingue, Michael J., et al. (2014) "Bioreplicated visual features of nanofabricated buprestid beetle decoys evoke stereotypical male mating flights." PNAS 111(39): 14106-14111.