Characterization of Chemosensory Proteins from the Red Imported Fire Ant, Solenopsis Invicta

Material Information

Characterization of Chemosensory Proteins from the Red Imported Fire Ant, Solenopsis Invicta
Wanchoo, Arun
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (173 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Microbiology and Cell Science
Committee Chair:
Committee Co-Chair:
Committee Members:
Graduation Date:


Subjects / Keywords:
Amino acids ( jstor )
Ants ( jstor )
Chemicals ( jstor )
Female animals ( jstor )
Fire ants ( jstor )
Insect antennae ( jstor )
Insects ( jstor )
Ligands ( jstor )
Sensilla ( jstor )
Signals ( jstor )
Microbiology and Cell Science -- Dissertations, Academic -- UF
chemosensory -- dichroism -- invicta -- ligand -- odorant -- solenopsis
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Microbiology and Cell Science thesis, Ph.D.


Ants are one of the most successful animals on our planet, yet remarkably little is known about how they perceive chemical compounds, a process essential for their survival. Solenopsis invicta, the red imported fire ant employs a set of twenty one chemosensory proteins (CSPs) to aid in olfaction, chemical reception and or chemical transport. Our central hypotheses are that (a) CSPs display differential binding to a wide range of chemical and odorant molecules and (b) CSPs are differentially regulated in tissue and developmental specific patterns. Knowledge concerning the ligand specificity and gene expression patterns of CSPs can in turn can be used to make predictions regarding the role(s) of specific CSPs in chemoreception and other processes. Fourteen of the S. invicta chemosensory proteins (SiCSPs) were expressed and purified in recombinant form for the purposes of probing the activity and structural features of these proteins. Ligand binding profiles testing 78 different environmental and/or physiologically relevant compounds were determined for each protein. In addition, since CSPs are predicted to be predominantly alpha helical, circular dichroism polarimetry was used to test whether these proteins undergo any secondary structural conformational changes upon ligand binding. Real-Time quantitative polymerase chain reaction (RTQPCR) was used to examine the tissue and developmental gene expression profiles of all twenty-one CSPs. The ligand binding and gene expression profiles indicate complex patterns of chemical compounds perceived and tissue/developmental distribution. These data suggest that a subset of CSPs are multifunctional proteins with a subset involved in olfaction, while others potentially participate in chemical transport, release, and or sequestration. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2014.
Co-adviser: KIMA,PETER EPEH.
Electronic Access:
Statement of Responsibility:
by Arun Wanchoo.

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
Embargo Date:
Resource Identifier:
907379558 ( OCLC )
LD1780 2014 ( lcc )


This item has the following downloads:

Full Text




2 2014 Arun Wanchoo


3 To my parents and brother, w ithout their help and support I would not be here


4 ACKNOWLEDGMENTS I would like to t hank my family for supporting me during my graduate studies. Without their help and support I would not have achieved what I have. Ever since I was in elementary school I had an interest in science, participating in science fairs and learning about natur e. My parents always pushed me to do more than adequate and to truly have fun in what I create. I feel this push was what kept me working hard and making it more than just work. I also would like to thank my committee and advisor for assisting me to get to the point I am in education When we take classes in our undergraduate education, our interactions with our professors are transient, often never seeing them upon the completion of a course. My professors have stuck with me and helped me through the obstacles I faced in my research. Thank you Drs. Nemat Keyhani, Peter Kima, Byungho Kang, Vol ker Mai, and Barry Ache. Finally I would like to acknowledge my peers and those who worked with me. We studied together and worked tog ether regardless of how p ertinent our work was to each other. The undergraduate students in our la b were of great help and really allowed me to get my thesis completed in a timely fashion. Thank you Dr. Almudena Ortiz, Jonathan Boswell, Matthew Harbour Mia Choi, Jar e d H o mayouni Juri Na, Susan Rifai, Melanie Brewster and Collette Stelter.


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIG URES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 14 Solenopsis invicta ................................ ................................ ................................ ... 14 Morphology and Colony Development ................................ ................................ .... 15 Chemoreception in Insects ................................ ................................ ..................... 19 Olfactory Targets of Reception ................................ ................................ ............... 28 Biochemical Binding Assays ................................ ................................ ................... 30 Circular Dichroism ................................ ................................ ................................ .. 32 2 IDENTIFICATION, EXPRESSION AND PURIFICATION OF RECOMBINANT S. INVICTA CHEMOSENSORY PROTEINS ................................ ............................... 46 Introduction ................................ ................................ ................................ ............. 46 Materials and Methods ................................ ................................ ............................ 47 Strains and Reagents ................................ ................................ ....................... 47 Cloning of Recombinant CSPs ................................ ................................ ......... 48 Expression and Purification ................................ ................................ .............. 48 Phylogenetic analysis ................................ ................................ ....................... 50 Results ................................ ................................ ................................ .................... 51 Identification of S. invicta CSPs and Bioinformatic Analysis ............................. 51 Expression and Purification of S. invicta CSPs ................................ ................ 52 Discussion ................................ ................................ ................................ .............. 53 3 GENE EXPRESSION ANALYSIS OF S. INVICTA CHEMOSENSORY PROTEINS ................................ ................................ ................................ ............. 66 Introduction ................................ ................................ ................................ ............. 66 Materials and Methods ................................ ................................ ............................ 69 Primer Design ................................ ................................ ................................ ... 69 RNA Preparation ................................ ................................ .............................. 70 cDNA Synthesis ................................ ................................ ............................... 70


6 Real Time Polymerase Chain Reaction ................................ ............................ 71 Results ................................ ................................ ................................ .................... 71 Discussion ................................ ................................ ................................ .............. 75 4 LIGAND BINDIND SPECIFICITIES OF SiCSPs ................................ ..................... 87 Introduction ................................ ................................ ................................ ............. 87 Materials and Methods ................................ ................................ ............................ 89 Chemicals ................................ ................................ ................................ ......... 89 1 NPN Binding Assay ................................ ................................ ....................... 89 1 NPN Competition Assays ................................ ................................ .............. 90 Results ................................ ................................ ................................ .................... 91 Discussion ................................ ................................ ................................ .............. 94 5 PROBING THE SECONDARY STRUCTURE OF The SiCSPS ........................... 115 Introdu ction ................................ ................................ ................................ ........... 115 Materials and Methods ................................ ................................ .......................... 118 Phyre 2 structural predictions ................................ ................................ ........... 118 Protein and Ligand Solutions ................................ ................................ .......... 118 Circular Dichroism ................................ ................................ .......................... 118 Results ................................ ................................ ................................ .................. 119 Phyre 2 protein structure predictions ................................ ................................ 119 CD polarimetry ................................ ................................ ............................... 120 Discussion ................................ ................................ ................................ ............ 122 6 CONCLUSIONS ................................ ................................ ................................ ... 155 Ligand Binding Specificity ................................ ................................ ..................... 161 P rotein Secondary Structure ................................ ................................ ................. 162 LIST OF REFERENCES ................................ ................................ ............................. 164 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 173


7 LIST OF TABLES Table page 1 1 List of ligands. ................................ ................................ ................................ ..... 43 1 2 S. invicta pheromones. ................................ ................................ ....................... 45 2 1 CSP genes from the genome of S. invicta. ................................ ......................... 56 2 2 Native and Reco mbinant SiCSP Properties. ................................ ...................... 58 2 3 Overview of CSPs in S. invicta. ................................ ................................ .......... 59 3 1 SiCSP Primers Used for RT PCR. ................................ ................................ ...... 79 3 2 Reference Ge ne primers used for RT PCR. ................................ ....................... 80 3 3 Validation of RTPCR Primer sets. ................................ ................................ ...... 81 4 1 1 NPN binding affinities of fourteen SiCSPs. ................................ .................... 112 5 1 Overview of SiCSP secondary structure analysis. ................................ ............ 153 5 2 Phyre 2 Predictions. ................................ ................................ ........................... 154


8 LIST OF FIGURES Figure page 1 1 Organization of the order Insecta. ................................ ................................ ...... 34 1 2 The anatomy of an ant. ................................ ................................ ....................... 35 1 3 Worker ant polymorphism. ................................ ................................ .................. 36 1 4 Scanning electron microscopy of different sensilla. ................................ ............ 37 1 5 Within the sensillum. ................................ ................................ ........................... 38 1 6 The odorant receptor and coreceptor. ................................ ................................ 39 1 7 The ne urological organization of insects ................................ ............................ 40 1 9 Circular dichrosim spectra. ................................ ................................ ................. 42 2 1 Strategy for Recombinant CSPs. ................................ ................................ ........ 60 2 2 Purified Recombinant SiCSPs. ................................ ................................ ........... 61 2 3 ESI MS profile of SiCSP5. ................................ ................................ .................. 64 2 4 Phylogenetic analysis of the 21 CSPs from Solenopsis invicta. ......................... 65 3 1 Cycle Thresholds of Reference Genes. ................................ .............................. 82 3 2 SiCSP Expression of SiCSPs in Developing Fire Ants. ................................ ...... 83 3 3 Expression of SiCSPs in Fire ant Worker tissues. Expression of SiCSPs in worker ant s. ................................ ................................ ................................ ........ 84 3 4 Expression of SiCSPs in Fire ant Alate Females. Expression of SiCSPs in alate female ants. ................................ ................................ ............................... 85 3 5 E xpression of SiCSPs in Fire ant Alate Males. Expression of SiCSPs in alate male ants. ................................ ................................ ................................ ........... 86 4 1 NPN Assay of SiCSPs NPN Assay of SiCSPs. SiCSPs were reacted at a concentration of 1 uM in 10 mM HEPES 50 mM NaCl pH 7.5 against concentrations of NPN ranging from 0 uM to 20 uM NPN in 2% methanol. Curves were fit to the Hill equation to estimate a one site binding model. .......... 98


9 4 2 NPN Assay of SiCSPs NPN Assay of SiCSPs. SiCSPs were reacted at a concentration of 1 uM in 10 mM HEPES 50 mM NaCl pH 7.5 against concentrations of NPN ranging from 0 uM to 20 uM NPN in 2% methanol. Curves were fit to the Adair equation to estim ate a two site binding model. ... 105 4 3 Heat map of Apparent K d of SiCSPs Against Numerous Odorants. .................. 113 5 1 S iCSP1 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 125 5 2 SiCSP2 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 127 5 3 SiCSP3 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 129 5 4 SiCSP4 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 131 5 5 SiCSP5 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 133 5 6 SiCSP6 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 135 5 7 SiCSP7 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 137 5 8 SiCSP8 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 139 5 9 SiCSP9 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 141 5 10 SiCSP10 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 143


10 5 11 SiCSP11 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 145 5 12 SiCSP12 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 147 5 13 SiCSP13 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 149 5 14 SiCSP14 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ................................ ............................... 151


11 LIST OF ABBREVIATIONS BLAST Basic Local Alignment Search Tool CD Circular Dichroism polarimetry CSP Chemosensory Protein DEET N N Diethyl meta toluamide MS Mass Spectrometry NPN 1 N Phenyl Naphthylamine OBP Odorant Binding P roteins OR Odorant Receptor ORN Olfactory Receptor Neuron RIFA Red Imported Fire Ant RTPCR Real Time Polymerase Chain Reaction SiCSP S. invicta derived CSP


12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERIZATION OF CHEMOSENSORY PROTEINS FROM THE RED IMPORTED FIRE ANT, SOLENOPSIS INVICTA By Arun Wanchoo May 201 4 Chair: Nemat Keyhani Major: Microbiology and Cell Science Ants are one of the most successful animals on our planet, yet remarkably little is known about how they perceive chemical compounds, a process essential for their survival. Solenopsis invicta, the red imported fire ant employs a set of t wenty one chemosensory proteins (CSPs) to aid in olfactio n, chemical reception and /or chemical transport Our central hypotheses are that (a) CSPs display differential bind ing to a wide range of chemical and odorant molecules and (b) CSPs are differentiall y regulated in tissue and developmental specific patterns. Knowledge concerning the ligand specificity and gene expression patterns of CSPs can in turn can be used to make predictions regarding the role ( s ) of specific CSPs in chemoreception and other proce sses Fourteen of the S. invicta chemosensory proteins (SiCSPs) were expressed and purified in recombinant form for the purposes of probing the activity and structur al features of these proteins. Ligand binding profiles testing ~75 different environmenta l and/or physiologically relevant compounds were determined for each protein. In addition, since CSPs are predicted to be helical c ircular dichroism polarimetry was used to test whether these proteins undergo a ny secondary structural conf ormational changes upon ligand binding Real Time q uantitative p olymerase c hain r eaction (RT


13 QPCR) w as used to examine the tissue and developmental gene expression profiles of all twenty one CSPs The ligand binding and gene expression profiles indica te complex patterns of chemical compounds perceived and tissue/developmental distribution. These data suggest that a subset of CSPs are multifunctional proteins with a subset involved in olfaction, while others potentially partic ipate in chemical transport, release, and/or sequestration.


14 CHAPTER 1 LITERATURE REVIEW Solenopsis invicta Solenopsis invicta Buren also known as the Red Imported Fire Ant, was originally endemic to Northern Argentina, Southern Brazil and parts of Para guay. Sometime between 1933 and 1945 this ant found its way aboard ship(s) carrying exotic fruits, lumber, and other goods from South America to the United States ( Mobley D, 2005 ) For some unknown reason, the ant first established its beachhead in North America at Mobile, Alabama. It is likely that fire ants were introduced into North America on several occasions, however, what is clear is that they have spread through Pensacola, Florida to the Eastern seaboard and much of the low er Southern United States These ants have reached as far as parts of Tennessee, westwards throughout Texas and more recently have been found in California. In addition to the United St ates, the red imported fire ant has had a dramati c worldwide spread including reaching S outhern China, Australia New Zealand and the Philippines likely through trade routes via the United States. Solenopsis invicta, referring to the Latin word for undefeated aptly describes the curre nt reign of these ants in the United States. species, with S. invicta being particularly aggressive, for purposes of simplicity and S. invicta environmental context, S. invicta has been observed to attack other species of insects as well as waging intra and inter species aggression/warfare. These traits of aggression make the fire ant a pa rticularly problematic pest in farmlands, pastures, and semi populated areas. Fire ants will often dig colonies in the root systems of plants,


15 disrupting their growth and leaving an unpleasant surprise for any unwary farmers examining and/or harvesting th eir crops ( Adams et al. 1983 ) As omnivores, fire ants will also swarm and kill land borne animals including small mammals, lizards, baby birds, as well as any other insects in their search for food ( Seagraves and McPherson, 2006 ) ( Allen et al. 1994 ) ( Bencki ser, 2010 ) Problematic especially for children, fire ants have potent venom and can sting multiple times, with victims finding themselves covered in a rash of fire ant bites. Within hours the bites become inflamed and swollen with pus, causing pain, a nd even more severe reactions can be seen in allergic victims. Thus, a lthough it is not a known disease vector, it can inducing strong allerg enic responses and even anaphylaxis in a portion of the population ( ACAAI, 2010 ) Despite its often vivid demonization in popular press articles, however, fire ants are more of a nuisance than a significan t pest problem, and after decades of failed eradication efforts, current control efforts rely on spot treatment in particular locations, e.g. lawn, active pastures, near residential areas as needed ( Hedges, 1997 ; Bextine and Thorvilson, 2002 ) ( Bextine and Thorvilson, 2002 ) ( Buhs, 2002 ) Morphology and Colony Development Belonging to the order Hymenoptera ants share ancestral and hence a degree of structural (body plan) similarit y to bees and wasps (also Hymenoptera) (Figure1 1) Hymenoptera are c haract erized by the presence of two pairs of wings ( present in reproductive male and female alate ants ), which are attached to the thorax by a hamulus, or membranous hook. Within the Family subclassification ants belong to the Forimicidae (Latreille 1809 ) and are characterized by having 10 segment jointed antennae, metapleural glands, and petiole (s) ( one or two bulbous node s between the alitrunk and gaster) (Figure 1 2 ) Many (but not all) a nt species exhibit polymorphism s in


16 size and shape sometimes cla castes that typically includ e soldier s major worker s minor worker s drone s (male s all of which are haploid ), and one or more queen (s) S. invicta castes are typically divided into minor workers, foragers, and soldier an ts, althou gh there are significant variations in ant sizes ranging from ~2 to 6 mm with >25 fold differences possible in body weights between the smallest and largest members (Figure 1 3 ) ( Araujo and Tschinkel, 2010 ) The life span of the ants appears to roug hly correlate with body size, with minor workers living for 30 60 days, mid sized workers 60 90 days, major workers 90 180 days, and queens capable of living 3 6 years, and even up to 7 9 years. In fire ants (there is a great deal of variability amongst a nts species in almost all aspects of their physiology and behavior), worker roles appear to be differentiated by the size of the ant, with the smaller workers remaining within the colony tending to eggs, larvae and pupae, while larger workers forage also assuming the role of soldiers. It should be noted that there is a significant degree of behavioral plasticity, and that larger workers can also tend brood and smaller workers will forage and/or fight under appropriate circumstances. The development of work er castes has been attributed directly to the circumstances in which the eggs were tended and nutrient intake of the larvae In general, i f the workers have limited and/or restricted food intake as larvae, they develop into smaller colony workers while l arger forager and/ or soldier ants receive greater nourish ment as larvae. In fire ants (again, there is significant variation in this between different ant genera) w orkers (non reproductives) have only rudimentary ovaries and are unable to reproduce. ( Carey and Carlson, 2011 )


17 As the fire ant colony grows and matur es flights of winged male and female alate s leave the colony to find a mate (typically 1 2 years after founding) Upon successful pairing and mating a female will break off and consume her wings, burrow to found her nest, and lay the first few eggs of a new colony M ales that have mated (as well as those that went on a mating flight but failed to m ate) die shortly after The first few worker larvae that emerge are tended to by the queen and fed by her nutrient stores. The first few workers are often unspecialized and develop the mound and scavenge for food while the queen remains the sole reproduc tive unit ( Tschinkel, 2006 ) Alate s develop when a colony is mature (1 2 years) and prepare for a new flight to found new colonies and hence close the life cycle. Two distinct social organizations that correlate with specific genetic patterns have been found in fire ants. In one form, fire ant colonies have a single virile, ovipositing queen (monogyny) whereas in the other the colony contains multiple queens (polygyny) colonies have been found ( Ross and Keller, 1995 ) The two social polymorphs display important difference s in organization and even behavior includ i ng their levels of aggression towards other fire ants and their method s of initiating new colonies Monogyne colonies spread during a mating flight only while polygyne co lonies can group of mated (most likely within the colony itself) females alates emigrate with a group of workers to start a colony usually relatively nearby One o f the more striking differences are their aggression responses; members of monogyne colonies are aggressive to other colonies whether they are mongyne or polygyne in social structure. s while the polygyne phenotype is inhospitable only to monogyne colonies but will


18 cooperate with other polygyne colonies. The polygyne phenotype is often attributed to the Gp 9 allele encoding an odorant binding protein (OBP). The presence of two different Gp 9 alleles associated with the two phenotypes implied a selective ro le based on chemical communication ( Krieger and Ross, 2005 ) However, the Gp 9 allele by itself has been suggested to be unlikely to be able to account for the monogyne versus polygyne variation and more rece nt studies have revealed that the phenotype extends far beyond just the Gp 9 locus to a non recombining 13 megabase portion of a chromosome (Social or S chromosome) propagated similar to a Y chromosome like mechanism, although it is only a part of a larger chromosome ( Wang et al. 2013 ) These studies have shown that there exists an inversion in this S chromosome that prevents recombination between the chromosomes of the two social isomorphs. It has yet to be confirmed what t he Gp 9 OBP recognizes and it may contribute to recognition of polygyne queens by her workers ( Krieger and Ross, 2005 ) however there is also some evidence that Gp 9 is actually not likely to be involved in determined social order (Leal and Ishida, 2008, PLoS One, 3:e3762), although this suggestion has been challenged (Gotzek and Ross, PLoS One, 2009, 4:e7713) Given t he variability of their social organization, the varied tasks that need to be performed, and the needs of the colony for proper cohesion and development there is a ne e d some modality of communication between members that is not required for non social animals ( Smith et al. 2009 ) At least part of t he complexity of an a nt olfactory chemoreception. Though fire ants have eyes, and also respond to tactile sensation (by means of hairs distributed upon the cuticle), these seem to be secondary


19 to olfaction /chemical reception, which consti tutes the main modality for communication within the colony and reception of and response to environmental cues ( Sato et al. 2008 ; Sato and Touhara, 2009 ) Pheromones are perhaps the best known chemical signals in insects. Trail pheromones have been isolated and shown to be used by foraging ants to develop a path to food sources and back to the colony providing cu es for orientation and navigation ( Vander Meer, 1981 ; Vander Meer, 1983 ; Vander Meer, 1988 ; Vander Meer and Alonso, 1998 ) Chemoreception in Insects Insect chemoreception is achieved through two major organs: the maxillary palps (structures near the mouth parts) and the antennae (Figure 1 2) The mouthparts and maxillary palps are considered to participate more in contact chemoreception of food particles and thus are thought to play a limited role in olfaction A ntennae can vary greatly be tween insect species in structure a lthough they share broad general function s Moths for example have fan like antennae that are finely attuned to detect aerial semiochemicals and pheromones to find mates and food sources. Ants however have smaller club shaped antennae that can participate in both airborne and contact chemoreception ( Renthal et al. 2003 ) In addition, within the social context of ant societies they help mediate interactions including self/non self recognition, coo peration in task performance, e.g. brood care, nest maintenance, food acquisition, and defense, and in hygienic/palliative interactions, e.g. grooming and trophyallaxis (transfer of food or other fluids amongst nestmates) ( Tschinkel, 2006 ) In ants, as in most insects, t he maxillary palps and antennae are covered with filamentous projections called sensilla. Insects ( de Brito Sanchez et al. 2014 ) also have contact chemosensory organs (essentially gustatory or gans) on their f eet (that also


20 contain sensilla ) essentially allowing them to taste with their limbs. U p to 2000 sensilla are found per antennae ; divided amongst f ive main classes of sensilla: trichodea, trichodea curvata (curved sensilla), basiconica, am pullacaea, and coeloconica ( Figure1 4 ) ( Renthal et al. 2003 ) Trichodea and trichodea curvata are both slim sensilla that taper down to a point (with trichodea curvata bearing a curve) with pores located laterally throughout the s ensillum. The presence of pores that would allow chemicals to permeate into the sensillar lymph could indicate a role in general aromatic odorant reception. Sensilla basiconica are stockier, shorter sensilla with pores located distally from the sensillum base and likely play a role in contact chemosensation. Coeloconica are champagne cork shaped sensilla with a single pore and are also thought to be involved with non specific odorant reception. Coeloconic sensilla number in six per antenna implying tha t they may play a minor role in chemosensation for ants. Ampullacaea are uniquely shaped sensilla with an internalized pore adapted for sensing CO 2 levels. Fire ants cannot directly detect O 2 however the ability to differentiate different CO 2 levels may be important to identify conditions in which the nest requires better ventilation and/or in proper nest enlargement and maintenance ( Renthal et al. 2003 ; Kaupp, 2010 ; Dietrich Gotzek1*.¤, 2011 ; Gu et al. 2011 ) As mentioned above, t he sensillu m is a cuticular projection extending from the insect (ant) exoskeleton that is covered in pores to allow the entrance and release of semiochemicals contains within its structure the sensillar lymph, an aqueous solution rich in protein and cations to fac ilitate chemoreception and the olfactory neurons themselves ( Figure 1 5 ) At the base of the sensillum are two types of cells, basal or accessory cells, and as mentioned, the odorant receptor neurons (ORNs). The


21 accessory cells are responsible for produ cing the sensillum lymph and maintain the environment in the sensilla. Typically two ORNs inhabit the base of the sensilla and extend dendrites into the sensillum. Along these dendrites a signal cascade stemming from membrane receptors initiates the acti on potential that leads to odorant perception. ( Renthal et al. 2003 ; Rutzler and Zwiebel, 2005 ) There exist th ree main classes of receptors on the ORNs: gustatory receptors (GRs), ionotropic glutamate like receptors (IRs), and odorant receptors (ORs) ( Kaupp, 2010 ; Getahun et al. 2013 ) Despite the naming conventio n seeming exclusive, ORs and GR s are also ionotropic receptors in contrast to vertebrate metabotropic receptors In addition, ORs, IRs, and GRs, are passive io n channels that when opened allow ions to flow down the ion gradient ( Kaupp, 2010 ) Ionotropic glutamate like rece ptors (IRs) consist of a class of proteins with three transmembrane helices that appear to show little specificity for ligands but react with broad classes of molecules ( Abuin et al. 2011 ) This may imply a role in general olfactory reception with low binding sensitivity. IRs display homology to vertebrate glutamate receptors, however, they lack the glutamate binding domain and do not operate by a metabotropic (t he use of a second messenger) mechanism. The coeloconic sensilla exclusively house IR expressing ORNs, which may number up to five in each sensilla (in contrast to the typical two ORNs in trichodea sensilla) ( Kaupp, 2010 ) To add to the complexity, multiple different IRs can also be expressed in the same ORN that may results in different combinations of heteromeric complexes to facilitate odorant reception. Two broadly expressed IRs (IR25a and IR8a is Drosophila sp. ) appear to act as coreceptors necessary for electrophysiological activity ( Abuin et


22 al. 2011 ) Unlike the ORs, IRs seem to be less sensitive and the degeneracy in ligand binding may allow them to react in consortium to add context to odorant reception ( Benton et al. 2006 ; Benton et al. 2009 ) Because there are only six coeliconic sensilla present in each antenna of fire ants ( Renthal et al. 2003 ) it is unclear how much of a contribution IRs play in fire ant chemosensation, unless IRs are also expressed in other sensilla types in this organism (a phenomenon that has yet to be observed). Gustatory receptors (GRs), like ORs are membrane prote ins that have seven transmembrane domains and a ligand binding domain. GRs are thought to bind hydrophilic compounds including taste perceived molecules such as sugars (sucrose, glucose, etc.). GRs also work in heterodimers with a co receptor (i.e. GR64a In Drosophilla sp .) necessary for an electrophysiological response ( Jiao et al. 2007 ) GRs are expressed in the basiconic sensilla on the maxillary palps in num erous insects and even on the tarsi in at least the bee, Aphis mellifera ( de Brito Sanchez e t al. 2014 ) Odorant receptors (ORs) are the membrane proteins that respond to chemical signals. Insect genomes code for multiple ORs and S. invicta contains 297 different OR genes ( Wurm et al. 2011 ) Each ORN typically expresses only one OR as well as the general odorant co receptor Orco which is considered to form a heterodimer with all ORs. Odorant receptors are a class of membrane proteins with 7 transmembr ane helices, a putative extracellular ligand binding domain and a protein protein interaction domain. Although they share some homolog and originally were considered thus, insect ORs are distinct from G protein coupled receptors (GPCRs) seen in vertebrate olfaction and likely operate via a different biochemical mechanism than vertebrate ORs. One significant compelling evidence that insect ORs do not operate via a G protein coupled


23 mechanism has been derived from topology modeling which shows that the doma in likely to interact with the G proteins is extracellular for the insect proteins. More detailed biochemical evidence has shown that together with the general purpose transmembrane protein, OR83b ( Orco ), ORs form a heterodimeric, ionotropic membrane chann el which when activated allows potassium ions into the cell and sodium ions out, resulting in the initiation of the action potential and activation of the neuron (Figure 1 6). As mentioned, the extracellular domain of the OR contains a ligand specific bin ding pocket and the current hypothetical mechanism is that upon the binding of a ligand, the OR interacts with OR83b (Orco) to allow for the K + /Na + counter flow, triggering the action potential. Though this model is not completely substantiated, if the O rco gene is knocked out, the organism becomes much less sensitive to odorant stimuli and the localization of the OR to the ORN dendrites is greatly reduced ( Benton et al. 2006 ) The downstream insect nervous system includes olfactory glomeruli, which consist of numerous neurons networked on the receiving end of ORNs. Signals relayed to the olfactory glomeruli are cross linked and interpreted in the mushroom bod y or the main nerve cluster in the insect (Figure 1 7). An example of the application of antennal chemical perception to behavior can be shown in the trail pursuing activity of ants ( Holldobler and Wilson, 1990 ) When both anten nae receive the trail pheromone evenly, the motion of the ant remains essentially linear, while if the left antenna receives the pheromone more than the right, the ant will turn left until perception of the trail pheromone is even again, and vice versa for the right antenna and rightward movement. ( Ozaki et al. 2005 ; Sanchez Gracia et al. 2009 ; Touhara and Vosshall, 2009 ; Vieira and Rozas, 2011 ; Kulmuni, 2013 ) To further study the ORs one group has


24 even produced an empty neuron model system in Drosophila sp. This has allowed the cloning of any OR into this knockout fly and testing a panel of odors and using electroantennogram to determine which odors elicit an electrophysiological response (Car ey, Carlson, 2010). Odorant Binding Proteins (OBPs) and Chemosensory Proteins (CSPs) are the two main classes of soluble binding proteins found in the insect sensillum. Both proteins are small (10 15 kDa) binding proteins, some of which are thought to act as solubilizers for hydrophobic molecules (odorants) ( Vogt, 2005 ) OBPs consist of a class of helical proteins that have bee n described as either belonging to the general OBP class or to the pheromone binding protein group (PBP). PBPs have been extensively studied in the Bombyx mori (BmorPBP1) ( Grosse Wilde et al. 2006 ) and Drosophila melanogast er (L USH ) ( Kruse et al. 2003 ) systems. BmorPBP1 is a PBP that binds bombykol, a sex pheromone in a specific fashion in order to direct mating behaviors. Though el imination (by means of a gene knockout) of the BmorPBP1 has shown that the odorant receptor (BmOR1) ca n still respond to the pheromone, systems using BmorPBP1 to solubilize bombykol were two to three times more sensitive ( Gro sse Wilde et al. 2006 ) The L USH PBP found in D. melanogaster was originally thought to be a means of detecting alcohols (the original mutant was described as losing sensitivity to ethanol), which can be toxic to flies, as a means of sensing and avoidi ng ( Swarup et al. 2011 ) However, because of the relatively low binding affinity to alcohols and the subsequent finding that the affinity for LUSH to the thereafter discovered sex pheromone, 11 cis vaccenyl acetate, it is now believed that the Lush protein is in fact a PBP involved in mating response (This is an example of


25 why it is necessary to test several compounds). Incidentally the knockout of OBPs as observed in the Drosophila system ca n lead to behavioral change. The knockouts of OBP57d and OBP57e in D. melanogaster resulted in an increase in preference for short chain fatty acids, something the flies usually avoid ( Harada et al. 2008 ) The genome of the fi re ant includes 18 OBPs which still require characterization ( Gotzek et al. 2011 ) In contrast to the literature on OBPs, CSPs have remained less studied (except for phylogenetic analyses). The three dimensional structures of a number of CSPs have been solved ( Mamestra brassicae MbraACSP6 and Schistocerca gergaria, SgreCSP4) and CSPs a re predicted to be helical proteins (containing six helices) and contain four oxidized (paired) cysteines, two of which bear the motif CXXC ( Picone et al. 2001 ; Mosbah et al. 2003 ) The CSPs have been predicted to bind multiple ligands as opposed to displaying strong ligand specificity. The fire ant genome contains 21 CSPs ( Gonzalez et al. 2009 ; Kulmuni and Havukainen, 2013 ; Kulmuni et al. 2013 ) SiCSP1, or the first CSP from S. invicta to be isolated. This CSP was present in the antennae in high concentration and is predicted to be involved with olfaction. Though there is no evidence supporting the direct interaction between the OBPs or CSPs with ORs (or IRs and GRs), the proposed role of the OBPs and CSPs are to solubilize hydrophobic odorants to facilitate the transport through the sensillum to the ORs (or GRs and IRs) outlined in Figure 1 6 ( Vogt, 2005 ; Sanchez Gracia et al. 2009 ) In this study we will focus on the CSPs from S invicta In addition to roles in chemoreception and odorant sensing, we bel ieve CSPs may be involved in other roles involving ligand binding including the transport and release of odorants.


26 The role and method of action of the CSPs are outlined in Figure 1 6 but this represents speculation instead of confirmed mechanism. Though the odorant receptors are extensively studied, there had to be a solubilizing agent to allow hydrophobic odorants to traverse the lymph and the proposed models of OBPs and CSPs were borne of this necessity. Since, experiments have shown the effect of knoc king out PBPs leading to a lower response to pheromone, it is possible the soluble players in the sensilla either decrease the amount of odorant needed to reach a threshold required for physiological response, or makes the system as a whole more sensitive to an odorant. The possibility that CSPs are general hydrophobic binding proteins, unlike the PBP (ligand specificity) may assist in this threshold/sensitivity concept by acting as solubilizers, increasing the concentration of hydrophobic odorant present in the lymph. This would relinquish the need for CSPs to directly interact with the odorant receptors (there is no confirmed interaction between OBPs and CSPs with ORs, IRs, or GRs). If CSPs lower the threshold for response then can they help relieve the response? If the concentration of odorant is greater outside of the sensillum lymph than within, naturally, the CSPs will help absorb the odorant. Likewise this will reciprocate. When the insect is removed from the odorant source, the concentration of o dorant will be higher in the lymph than in the air surrounding the sensillum and the odorant will diffuse out on its own. Why then is the CSP necessary in this model? Hydrophobic molecules have a knack of depositing on surfaces surrounding a hydrophilic solution, potentially sticking to the inner surface of the sensillum cuticle, or to lipids/lipoproteins on the ORN membrane. CSPs may play a role in keeping the odorants from depositing in these areas and facilitating release as ligand concentration decre ases. It has also


27 been thought that CSPs may interact with degrading enzymes, however this would require a protein protein interaction that has yet to be observed. (anywhere bet ween four in the case of D. melanogaster and twenty one in S. invicta ) and OBPs (fifty one in D. melanogaster and eighteen in S. invicta ). Clearly CSPs and OBPs do not always operate in a one odorant, one binding protein (CSP or OBP), and one receptor per odorant model, and cases like PBPs become the exception. For this reason it is thought that the CSPs (and some OBPs) are general binding proteins, binding to anything that could be categorized as an odorant (hydrophobic ligand binding protein). In this case then why must there be so many copies? In the fire ant genome (as well as other ant genomes) there are instances of positive selection where expansions from a common ancestral CSP may have occurred. The selective pressure for these expansions is unc lear however having more CSPs may be related to the ground dwelling and foraging nature of ants as these CSP expansions are not present in flies, or even bees. Social insects depend greatly on communication, which in turn consists of receiving and sendin g signals. We already know that the fire ants receive signals through their antennae and secondary sensory organs (maxillary palps and tarsi) but the whole insect is a means of sending signals. The cuticle is rich in hydrophobic compounds created by the ant as well as molecules acquired from their environment from daily contact. The ant also releases numerous signals in the form of pheromones from a variety of glands. CSPs may play an important role outside olfaction, facilitating the release of hydroph obic signal molecules.


28 Olfactory Targets of Reception One approach to understand ing how ( fire ) ant s sense and respond to their environment is to look a t what may attract or repel the ant. In addition, the context an odor is received and the specificity of the odorant receptor interaction is likely to be important. The ant encounters a wide range of environmental ligand sources that include diverse plan ts, animals (non fire ant), nestmates and other fire ants microoganisms, and soil and airborne chemical compounds Chemicals perceived by ants can be separated into two broad classes; (a) intrinsic or endogenous, i.e. those produced by the ants themselv es, e.g. pheromones, surface cuticular compounds, etc, and (b) extrinsic or exogenous i.e. those derived from other organisms and the environmental. It should be noted, however, that many chemicals can originate both from the ant as well as can be found from other environmental sources, i.e. microbes, plants, other insects, etc. For many ants, p lant derived odorants may act as dietary signals during foraging activities. However, some plant extracts also have natural repellant properties. Even if a plant itself is not a nutrient source for a particular ant species, specific plants may harbor (insects and other animals) that are food sources, thus discrimination of different plants can be important even when not consumed by the ant. Plant derived odorants are commonly called essential oils and include compounds with the chemical structure of terpenes and terpenoids. These molecules are aromatic hydrocarbons (terpenes) or oxidized aromatic hydrocarbons with functional groups (terpenoids). These are often f ound in the saps of trees and plants and may provide the plant with a means of defending against parasites, as well as fulfilling other physiological functions in the plant (Table 1 1 )


29 Amongst the intrinsic chemicals perceived by ants, a subclass of o dora nts are the better known (though still not always completely characterized) pheromones which often e l i cit specific responses. Pheromones provide a means by which ants communicate with other members of the colony and coordinate complex actions including t he triggering of mating, foraging trail formation, and defense responses However, even when described, few pheromones are commercially available, although hydrocarbon precursors to pheromones are often readily available. ( Table 1 2 ). Two additional points about intrinsic perceived chemical compounds should be noted. First pheromones and other ant derived chemicals may also require (in additional to being perceived) means for transport throughout the body and release from the body. Second, there also exists the possibility (as seen from observational studies) of discrimination of intrinsic signals not only between ants of different species, but also between different colonies of the same species. For fire ants, which are territorial, i.e. they can be aggress ive towards other fire ant colonies; a mechanism for (odorant) discrimination between colonies must exist ( Vander Meer and Alonso, 1998 ; Tschinkel, 2006 ) Response to divergent animal species, i.e. mammals may trigger a generalized frenz ied response to defend the colony and may only nominally involve olfaction However, it is possible that odors given off as products of metabolic breakdown and waste are perceived by ants and help them to identify imminent invaders. It is unknown the extent to which other animals species (besides ants and insects) are recognized, although the nature of the potential threat c ould alter the response of a colony Regardless, understanding the range and nature of chemical compounds that an ant can perceive would aid in our understanding of the trigger s fire ants to respond to and


30 the behavioral range that they display. Knowledge concerning the chemical compounds perc eived would also lead to a better understanding of invertebrate olfaction and the environmental success of ants. This information also has implications in the evolution of invertebrate olfaction and comparison and distinction from vertebrat e processes. Ultimately, this information can also be used to model the behavioral dynamics of ants as social insects and may lead to novel approaches at insect control. Biochemical Binding Assays As part of this work, fourteen S. invicta CSPs were express ed and purified and the binding of the purified proteins to a range of chemical ligands tested. The classes of chemical molecules and their origin s are listed in Table 1 1. These 78 compounds represent different chemical classes and different origins of t he odorant ( whether related to intrinsic or extrinsic chemcals ). Screening using this library of odorants was used to determine how each SiCSP binds to different chemical compounds to provide insight into the preference s and possible function s of each CSP Chemical binding affinities for each CSPs was examined via determination of dissociation constant s (K d ). The K d is a measure of affinity between a macromolecule and a ligand and c an be described as the concentration of liga nd at which half of the protein is complexed This means that as the ligand concentration increases the binding protein becomes saturated and at the ligand concentration that allows the bound p opulation of the protein to equal the free population of protein we have K d The significance of K d values include their use in pharmacology to estimate ligand selectivity and tightness of binding ( Copelans et al. 2006 ; Hulme and Trevethick, 2010 ) For this study the K d will be pres ented to quantify and compare the protein ligand interaction between CSPs and between ligands in regards to a CSP.


31 1 N Phenylnaphthylamine (NPN) is a fluorescent ligand used for probing hydrophobicity Upon binding in a hydrophobic pocket of a protein, NP N will fluoresce at 415 nm (when excited with 350 nm light). By detecting this fluorescence the direct binding of NPN to the protein can be examined and a dissociation constant can be determined ( in vitro ) ( Ban et al. 2002 ; Ban et al. 2003 ; Gonzalez et al. 2009 ; Gong and Pl ettner, 2011 ) The simplest model for the interaction of one protein molecule with one ligand (one site binding) would use the Hill equation: (1 1) Where Bmax is the maximum detected value at saturation of the protein, K d is the X value at of Bmax and n is the Hill coefficient ( n = 1 for one site binding models) (F igure 1 8 ). In multiple binding site models, this fit estimates the proteins overall binding efficiency rather than yielding multiple diss ociation constants for the multiple sites. However if a protein has multiple binding sites with equal binding affinities, the appearance would be identical to a single site binding curve. In order to determine multiple dissociation constants present whe n proteins have multiple binding sites the more complex Adair model becomes pertinent. The presence if t wo binding sites can be estimated as follows : (1 2) The Adair equation/ fit allows us to take into account binding affinities for both sites when analyzing the binding data


32 Because the binding of most physiologically relevan t odorants is difficult to measure directly (i.e. they have n o detectable spectroscopic shift upon binding unlike NPN) a competition method wa s used to determine the binding affinities for the chemical compounds examined Essentially, u pon priming /loading a protein with NPN the NPN probe is displaced by any compet ing chemicals that can bind in the binding pocket, thus with increasing amounts of competing ligand a loss of fluorescence signal will be observed. This signal loss can be converted to a percent value with the initial fluorescence value set to 100% bindi ng. The resultant curve can be estimated using a one site binding competition fit if the protein binds one molecule of ligand for each molecule of protein: (1 3) Where Y max an d Y min correspond to the maximum and minimum values of the data. EC50 is the concentration of competitor at 50% competition. In addition the two site binding competitions can be fit to another estimation: (1 4) Where Bottom represents the Y value at complete competition (baseline), Top represents the initial 100% signal, and fraction1 represents the range of data for the first binding site. Circular Dichroism Circular dichroism polarimetry is a form of spectroscopy involving illuminating a sample with polarized light of left and right handed orientation and observi ng the differential absorption of one beam compared to the other ( Greenfield, 2006 ; Clarke, 2011 ) Circula r dichroism has commonly been used to determine the chirality of a mix


33 of isomers in solution for chemical applications as well as for probing protein secondary structures. CD spectroscopy, though unable to visualize discreet three dimensional structure s can reveal the presence of prominent structural motifs such as alpha helices and beta sheets as well as portray real time structural change s of proteins under different conditions The presence of alpha helices results in characteristic bimodal spectra wh ile, beta sheets and unstructured polypeptides result in unique single modal spectra ( Figure 1 9 ). Since p roteins are examined in solution ligands can be added to proteins to examine effects on secondary structure and various parameter s i.e. pH, ion, and temperature stability can readily be examined. In this study, CD spectroscopy was used to examine dynamic change s in secondary structure as a fu nction of temperature and protein ligand interaction ( Yu et al. 2012 )


34 Figure 1 1 Organization of the order Insecta. Informal distribution and relationships of various insects. Hymenoptera includes the social insects, e.g. ants and bees ( Evans and Gundersen Rindal, 2003 )


35 Figure 1 2. The anatomy of an ant. Like many insects, ants develop in segments. The fireant develops in ten segments. The head and an tennae form the first, then the thorax forms the nex t three, and the abdomen forms the remaining six. The first two segments in the abdomen form petioles (in the pedicel or region bet ween the gaster and the thorax ) and allows the ant flexibility for the g aster The antennae and maxillary palps (labeled in the image) are the primary chemosensory organs whi le hairs on the legs play a role in contact chemonsensati on. (


36 Figure 1 3 Worker ant polymorphism Fire ant wo rkers range in size from about 2 mm to 6 mm sometimes having a mass difference of greater than twenty fold The size and mass difference is attributed to th e amount of nutrition fed to them as larvae. Two queens are shown on the right for size comparison I mage: Dr. Samford Porter, USDA CMAVE, Gainsville, FL.


37 Figure 1 4 Scanning electron microscopy of different sensilla. SEM w as performed on sensilla on the apical club of worker fire ant antennae (A) The last segment of the antennal club is littered with the three main type of sensilla: sensilla basiconica (b) sensilla trichodea (t), and trichodea curvata (tc). Close ups of sensilla basiconica (B) and trichodea curvata (C) reveal pores for the entry of odorants. Images: ( Renthal et al. 2003 )


38 Figure 1 5 Within the sensillum The typical sensillum consists of between two and four sensory neuron s each expressing one odorant receptor gene along its dendrites The sensllum lymph is produced by the support cells and likely the location of CSP production. Image: ( Sanchez Gracia et al., 2009 )


39 Figure 1 6 The o dorant r eceptor and c oreceptor. The current proposed mechanism for the action of CSPs is characterized by the chaperoning of hydrophobic od orants to and from the OR/ORCO heterodimeric receptor complex. It is predicted that CSPs undergo a conformational change upon binding the odorant allowing it to interac t with the OR/ORCO receptor complex. In turn this complex is thought to act a s the ion otropicchannel (Na + Ca 2+ in and K + out) to propogate the signal transduction. In order to terminate the reception event the CSP may be involved with relieving the receptor complex of the odorant and facilitating the relaxation of the response. Image: ( Sanchez Gracia et al., 2009 ) CSP CSP


40 Figure 1 7 The neurological organization of insects. The axons from the ORN s in the antennae converge at the antennal lobe. This glomerulus acts as a relaying point for the numerous signals comi ng from the antennae and form synapses with intermediary projection neurons. Projection neurons play a role in amplifying the respo nses from the ORNs and synapse with neurons in the mushroom body (Kenyon cells) and the lateral horn (Lateral horn neurons) The lateral horn is involved with innate responses. The mus hroom body is the center of higher processing and learning responses. Image: ( Carey and Carlson, 2011 )


41 Figure 1 8 Typical b inding p rotein c haracteristics. Simple one site binding proteins by accepting the ligand into the bin ding pocket and can be quantified by their dissociation constant k d (Top). This reflects on how tight the interaction is and can be explained by the Hill equation. The two site binding model (Bottom) in this case is presented as cooperative (when the Hil l coefficient is greater than 1) and can be modeled using two k d s where the binding of the ligand in the first binding pocket positively facilitates the binding of the second ligand.


42 Figure 1 9 Circular dichrosim spectra. The CD spectra of proteins represent the unequal absorption of polarized light. The typical alpha helical protein (black) show a bimodal pattern in their spectra with a unique peak at 222 nm. Beta sheet (in red) and disordered proteins (green) spectra each are single peak curves and are both unique with respect to alpha helical proteins Image: ( Greenfield, 2006 )


43 Tab le 1 1 List of ligands. Ligand Chemical Classification Origin Z 3 Hexenol Alcohol Exogenous 2 Octanol Alcohol Endogenous 1 3 Octen ol Alcohol Endogenous 2 Ethyl 1 Hexenol Alcohol Endogenous 1 Nonanol Alcohol Endogenous Trans Trans Farnesol Alcohol Endogenous Nonanal Aldehyde Endogenous Decanal Aldehyde Endogenous Lauric aldehyde Aldehyde Exogenous Valeraldehyde Aldehyde Exogenous Linalyl acetate Ester Endogenous Ethyl butyrate Ester Endogenous Nonyl acetate Ester Endoge nous Butyl Butyrate Ester Endogenous Hexyl butyrate Ester Endogenous 3 Methoxyacetophenone Ketone Exogenous 3 Nitroacetophenone Ketone Endogenous 2 Nonanone Ketone Both 6 methyl 5 heptenone Ketone Endogenous m Cresol Aromatic Compound Exog enous p Cresol Aromatic Compound Exogenous DEET Aromatic Compound Exogenous Amyl Cinnamaldehyde Aromatic Compound Exogenous Butyl Benzoate Aromatic Compound Endogenous Dimethoxyacetophenone Aromatic Compound Endogenous Dimethoxybenzyl chloride Aromat ic Compound Endogenous 2' 6' Nitrotoluene Aromatic Compound Exogenous Dimethyl Pyrazine Aromatic Compound Endogenous Dimethylethyl Pyrazine Aromatic Compound Endogenous Indole Aromatic Compound Endogenous Eugenol Terpene/Terpenoid Endogenous D Limone ne Terpene/Terpenoid Exogenous Myrcene Terpene/Terpenoid Endogenous Geraniol Terpene/Terpenoid Exogenous Linalool Terpene/Terpenoid Exogenous Citranellal Terpene/Terpenoid Exogenous Nerolidol Terpene/Terpenoid Exogenous Occimene Terpene/Terpenoid Exo genous Caryophylene Terpene/Terpenoid Exogenous N Farnescene Terpene/Terpenoid Endogenous


44 Table 1 1 Continued Ligand Chemical Classification Origin Sucrose Sugar Exogenous Glucose Sugar Exogenous Trehalose Sugar Both Nonanoic Acid Carboxylic aci d Both Lauric Acid Carboxylic acid Both Glyceryl trioleate Carboxylic acid Both Palmitic Acid Carboxylic acid Both Oleic Acid Carboxylic acid Both Isovaleric Acid Carboxylic acid Exogenous Dimethoxybenzoic acid Carboxylic acid Exogenous Lactic Acid Carboxylic acid Both Benzoic Acid Carboxylic acid Both 12 Hydroxlauric Acid Carboxylic acid Endogenous C9 Alkane Endogenous C10 Alkane Endogenous C11 Alkane Endogenous C12 Alkane Endogenous C13 Alkane Endogenous C14 Alkane Endogenous C15 Alkane En dogenous C16 Alkane Endogenous C17 Alkane Endogenous C18 Alkane Endogenous C19 Alkane Endogenous C20 Alkane Endogenous C21 Alkane Endogenous C22 Alkane Endogenous C23 Alkane Endogenous C24 Alkane Endogenous C26 Alkane Endogenous C27 Alkane Endog enous C28 Alkane Endogenous C29 Alkane Endogenous C30 Alkane Endogenous C31 Alkane Endogenous C32 Alkane Endogenous C33 Alkane Endogenous C41 Alkane Endogenous


45 Table 1 2 S. invicta pheromones. Pheromone Gland Identity Reference Alarm Mandibula r gland, Poison Sac 1 Nonanone (octanone, octanol, 3 decanol other sp.) ( Holldobler, 1995 ; Vander Meer and Alonso, 1998 ) Brood Produced in Larvae Triolein ( Bigley and Vinson, 1975 ) Colony Odor Cuticle Mix of odor ants ( Vander Meer, 1983 ) Female Sex Pheromone C11, C13, Z 4 Tridecene ( Vander Meer, 1983 ; Vander Meer and Alonso, 1998 ) Primer Poison sac, (Not yet isolated) ( Vargo and Hulsey, 2000 ) Queen Recognition Poison sac E 6 (1 pentenyl) 2H Pyrazone ( Rocca et al. 1983 ) Trail Dufour s gland Farnescene ( Vander Meer, 1983 ) Trail Recruitment Dufour s gland Heptadecane ( Vander Meer, 1988 )


46 CHAPTER 2 IDENTIFICATION, EXPRESSION AND PURIFICATION OF RECOMBINANT S. INVICTA CHEMOSENSORY PROTEINS Introduction Our goals are to characterize the ligand binding specificity of the CSP repertoire in the fire a nt. An initial survey of the S. invicta genome and EST collections resulted in the identification of fourteen CSP gene s ( Xu et al. 2009 ) A subsequent re analysis of updated genomic data identified an additional set of seven proteins, yielding a total of twenty one CSPs for the fire ant, the most CSPs found in an insect (genome) to date ( ( Kulmuni et al. 2013 ) ). The de novo isolation of native proteins from the fire ant faces significant obstacles especially due to low/variable expression, difficulties in the acquisition of adequate tissue difficulty in developing purif ication protocols, and the overall processing time need ed to isolate enough protein ( ten s of mg range) needed for the proposed studies As one example purifying proteins from the chitinous insect cuticle c ould require complicated isolation procedures to achieve a homogenous solution of the target protein and would likely yield only very small amounts of protein ( Gonzalez et al. 2009 ) In contrast, p roduc tion of recombinant SiCSPs in a microbial system affords a simple method for producing CSPs with advantages including the ease of production, simplified purification protocols, and high protein yield potentials We have used a synthetic biology approach to clone f ourteen SiC SP genes The accession numbers for the SiC SP s are given in Table 2 1 Nucleotide sequences were used to determine the amino acid sequences of each respective protein. The amino acid sequences were then used to determine the codon optimized nucl eotide sequence for maximal expression in E. coli The overall strategy of the expression system is shown in Figure 2 1.


47 One CSP has already been partially characterized from S. invicta. SiCSP1 the first fire ant CSP characterized and isolated from the a ntennae was shown to bind 1 NPN as well as polar cuticular lipids Th ese data w ere used as a reference point to cross validate our data. Recombinant purified SiCSPs were also analyzed by m ass s pectroscopy ( ElectroSpray Ionization MS ) to determine whether a ny endogenously derived ligand from the E. coli host was bound to any of the purified proteins. Successful expression and purification of recombinant SiCSPs will allow for the production of the amounts of protein necessary to accomplish the planned ligand binding and circular dichroism polarimetry assays. The expression and study of recombinant CSPs is not something new however to produce the complete (in our case pending completion) repertoire of CSPs from any one organism has not been done. This could have important implications in building a deep er unde rstanding of the role of CSPs to the organism as a whole. By studying the family of CSPs as a whole we will be able determine the ligand range and specificity of all of the CSPs. Materials and Methods Strains and R eagents Growth cultures were maintained in Luria Bertani Broth (Fisher Scientific, ) The chemicals used in this study were purchased from Simga Aldrich ( ) unless otherwise noted. The r ecombinant SiCSP constructs were synthesized by Genewiz (South Plainfield New Jersey) inserted in the p UC57 vector. E. coli T7 SHuffle lysY from New England Bio labs ( ) was used as the exp ression strain for all of the SiCSPs.


48 Cloning of Recombinant CSPs Solenopsis invicta CSP sequences were identified from genomics and ESTs databases ( Gonzalez et al. 2009 ) The original fire ant nucleotide sequences of the CSP open reading frames (ORFs) were translated into corresponding amino acid sequences and then subsequently reconverted to E. coli expressio n codon optimized nucleotide sequences. Constructs we re designed in silico sequences that include a high expression T7 promoter, an optimized Shine Delgarno sequence for high efficiency translation, and a lacI repressor binding site in order to be able to tightly n a protease ( thrombin or TEV ) cleavage site, a 10 amino acid histidine tag and then the stop codon (Figure 2 1 ). As mentioned, t he intervening sequence for each cons truct is then based upon the amino acid sequence of the CSP converted to a codon optimized nucleotide sequence. The synthesized construct is ready for expression in a suitable E. coli strain with no further modification needed All c lones were by GENEWIZ (South Plaintfield, New Jersey ) and inserted into pUC57 plasmid (ampicillin resistance marker) Expression of the proteins in E. coli via the T7 promoter is IPTG (isopropylthiogalactopyranoside) inducible. The Genbank accession number and the translated amino acid sequences are given in Table 2 1 Theoretical properties of the proteins including molecular weight and pI are given in Table 2 2 Expression and Purification All expression constructs were transformed into E. coli T7 SHuffle lysY host cells. This strain is engineered for the production of target proteins, especially those with disulfide bridges and includes the express ion of chaperones and proteins that facilitate proper disulfide bond formation in the proteins. Once transformed, clon es were single


49 colony isolated via select ion for ampicillin resistance on LB/ampicillin plates (100 g/ml) and frozen for long term storage with 10% glycerol. Typically, for protein expression and purification, p reculture tubes (8 mL) were inoculated by a septic transfer into LB with 100 ug/mL ampicillin and incubated overnight ( 16 18 h ours ) at 30 C. Precultures tubes were used to inoculate 800 mL of LB broth containing 100 ug/mL ampicillin and incubated with shaking at 30 C. Cultures were grown until O D 600 0.5 and then induced via the addition of 0.5 mM IPTG (final concentration) after which flasks were incubated at 18 C overnight (12 16 h) Cells were harvested by centrifugation ( 4000 x g for 10 minutes ), washed twice with 10 mM HEPES 200 mM NaCl 10 mM i midazole pH 7.5 buffer and finally resuspended in the wash buffer (15 25 ml) Exceptions to this protocol are noted in Table 2 3. Cells were l ys ed via passage through a French Press (2 3 X) Cell biomass was lysed at 1000 PSI repeate d 3 times per sample followed by a 45 second burst using an ultrasonic needle probe to further disrupt insoluble cellular debris /shear DNA Lysates were subjected to high speed centrifugation at 1 2 ,000xg, 30 min to separate soluble proteins (supernatant ) from the insoluble fraction (pellet) Purification of proteins wa s achieved using nickel and/or c obalt affinity chromatograph y Columns were packed with resin (Ni 2+ or Co 2+ ) to the needed capacity (typically 3 5 ml of packed resin capable of binding 30 50 mg of protein & used at predicted half capacity) The divalent cation chemistry of the nickel/ cobalt column s maintain coordinate ionic interaction s with the 10 amino acid h is tail added to C termini of the SiCSP proteins. Briefly, purification was performed as follows: (a) crude supernatant extract (20 ml) was passed over a column equilibrated in wash/lysis buffer


50 as above, (b) column was washed (20 ml) with wash/lysis buffer, and (c) proteins eluted in several fractions using increasing concent rations; lysis buffer containing 100 500 mM imidazole (increments of 100 mM/ fraction with each elution 10 15 ml, unless otherwise noted). Fractions were analyzed by SDS PAGE analysis and fractions >90% pure were pooled The pooled fractions were concent rated via PEG water absoption in dialysis tubing ( Thermo Scientific 3000 kDa pore size), typically to 8 12 ml, 0.5 5 mg/mL protein, and then further purified/desalted via passage on Sepharose size exclusion column ( GE HiPrep 26/60 Sepha cryl S 200 HR ) u sing a Pharmacia Fast Protein Liquid Chromatography system. The column was equilibrated and run in 10 mM HEPES 20 mM NaCl pH 7.5 buffer. Samples were concentrated as needed using 3000 k Da dialysis tubing set in polyethylene glycol as above Protein samples were an a l y zed by SDS PAGE and protein concentrations were quantified and aliquots flash frozen in liquid nitrogen for long term storage. Phylogenetic a nalysis The amino acid seque nces of the twenty one SiCSP genes as well as CSP genes found in the ant species of Acromyrmex echinator Camponotus floridanus, and Herpegnathos saltitor were aligned to form a phylogenetic tree using the maximum likelihood fit and the amino acid substitu tion model (JTT) In addition to ant species, Drosophila melanogaster and Drosophila grimshawi were also included for comparative resolution of the ant CSPs HS like fits were used to determine the closeness between homologs.


51 Results Identification o f S i nvicta CSP s a nd B ioinformatic A nalysis The first 14 CSPs were found by using a patterned search from the Lausanne fire ant EST library and a BLAST search using the previous sequences to find any omitted sequences ( Gonzalez et al. 2009 ) By using the predicted protein models of the previous SiCSPs seven more SiCSPs have been annotated from the fire ant genome (while removi ng potential pseudogenes based on interrupting stop codons and frame shifts) ( Xu et al. 2009 ; Kulmuni et al. 20 13 ) Signal peptide motifs were identified on all 21 SiCSP sequences and were retained in the phylogenetic analysis For the purpose of this research the naming convention used here is identical to the original fourteen SiCSPs ( Gonzalez et al. 2009 ) SiCSP15, SiCSP20 and SiCSP21 match the numbering convention in the more recently discovered SiCSPs but SiCSP16, 17, 18, and 19 have been arbitrarily labeled ( Kulmuni et al. 2013 ) Amino acid sequences and accession numbers are listed below in Table 2 1. The phylogenetic analysis revealed a broad clade of SiCSPs that shared homology with CSPs from other species of ants ( S. echinator, H. saltitor and C. floridanus ) as well as t hose from non ant species ( D. grimshawi and D. melanogaster ). The six SiCSPs that belonged to this broad insect general CSPs included SiCSP6, 7, 16, 12, 9, and 11. The relationships between these six SiCSPs were not very apparent. A small grouping arose having homology to only other ant species (while not having homologs derived from Drosophila sp. ) included SiCSP4 19, and 17. SiCSP4 bore ho mology to AeCSP11 and CfCSP12 SiCSP17 had homology with AeCSP4, HsCSP7, and CfCSP7. SiCSP18 paired with AeCSP9


52 Two clades were identified containing only CSPs derived from S. invict a. The first clade contained SiCSP3, 18, 2, 10, 5, 1, and 20 Within this clade CSPs 3 and 18, 2 and 10, and 5, 1, and 20 were found to be more closely related and indicated that t he most recent gene expansion events involved two gene duplication and one triplication event. The s econd S. invicta specific group contained SiCSP 14, 15, 8, 13, 21 showing a nested duplication even with SICSP13 and 21 being the most closely related, then SiCSP8, followed by SiCSP15 and SICSP 14. These S. invicta specific clades had higher homology to each other relative to the ant specific and the insect general clades. The other ant species also had small groupings of their own CSP expansions that were more closely resembling the other ant species specific expansions in contrast to insect general CSPs. For example, the group of A.echinator derived AeCSP15, 7, and 10 and C. floridanus : Cf CSP10, 9, 8, and 11 had small expansions towards one end of the ph ylogenetic tree. Four groupings of SiCSPs have been identified from the phylogenetic analysis. Two separate S. invicta specific expansions, one expansion seen only in ants, and a broad expansion that bore homology to CSPs in other insects. Expression and P urification of S. invicta CSPs Fourteen of the twenty one SiCSPs were synthesized as expression clones and expressed using an E. coli system as detailed in the methods section . All constructs were transformed into E. coli T7 SHuffle which expresses chaperone proteins that assist in the proper folding of disulfide bridge containing proteins. Typically cells were grown at 30 C and induced with IPTG for production of the desired recombinant protein, at 18 C for 12 16 h. In some cases, in or der to avoid includion body production and/or increase production, cells were induced at 30 37 C for 3 5 h before harvesting. Cell lysis was typically achieved using a French Press and clarified lysates were run on Co 2+ or Ni 2+


53 columns. Proteins were elu ted with imidazole. Final purity (and desalting) was typically achieved by passage on an sepharose S 200 gel filtration column. A summary of the expression results is given in Table 2 2. An example of the purification process and the purity of resulting fr actions are given in Figure 2 2 SDS PAGE analysis of the purified proteins is given in Fig ure 2 2. With the exceptions of SiCSP3, 6 and 9 all proteins were produced by an optimized protocol while the aforementioned SiCSPs used an alternative purificatio n procedure. Origin all y SiCSP3 was being produced in inclusion bodies, however altering the inducing IPTG concentration to 250 M, increasing the induction temperature to 3 7 C, and decreasing the induction time to 5 hours allowed a shift towards producing soluble SiCSP3. SiCSP6 was not expressing under the standard protocol but increasing the inducing IPTG concentration to 1 mM and increasing the induction time to 48 hours allowed for protein production. It appeared that the high pI of the recombinant SiCSP9 protein interfered with the standard purification process, causing it to elute form the ion exchange column much earlier t han other proteins as well as precipitating on the FPLC column. After an altered elution protocol noted in Table 2 3 SiCSP9 was collected. S iCSP 5 w as run through Mass Spectroscopy and no endogenous ligands were found in samples ( Figure 2 3) B ecause of the outlines of sample preparation for Mass Spectrometry (sample should be isolated highly pure and in water) and time constraints only this sample was run. Discussion Phylogenetic analyses revealed that the SiCSPs could be separated into t wo broadly observed groups and several distinct subgroups. Broad grouping of SiCSPs showed division between those shared with other non ant insect species (i.e. SICSP6, 7, 9, 11,12, and 16 ) and those found exclusively in ants ( Si CSP 4 and 17 ). Within the


54 a nt specific CSPs, two clear sets of gene duplication events were seen. It is likely that with other insects. The observation of two S. invicta specific CSPs clades indicates positive selection for gene duplication events for these CSPs. Although speculative, these data suggest several possible scenarios; (a) this may indicate a fitness advantage gained by expression of additional CSPs via greater sensitivity and/or recogniti on of a larger set of odorants, (b) functionality within the social context of ants (not present for most other insects which are not social, (c) novel roles for CSPs in ants in addition to olfaction. In the fire ant, CSPs are expressed as secreted prote ins due to the presence of signal peptides. For expression in E. coli, the signal peptide sequences were removed and the nucleotide sequence codon optimized for the maximal protein yields in the bacterial system. All fourteen recombinant SiCSPs were produc ed successfully in the E. coli bacterial host. In several cases, expression of s ome SiCSPs was low during 18 o C overnight expression regime and had to be modified (noted in Table 2 3) SiCSP3 was originally being produced exclusively in incl usion bodies ( identified in the insoluble fractions). This was addressed by decreasing the inducing IPTG concentration and growing the cells for only 5 hours at 30 C. SiCSP 6 was not being expressed under standard conditions so the protocol was modified so that it wou ld be in duced wit h 1 mM IPTG and gro w for 48 hours at 16 C. This allowed expression at a slower growth rate than the standard conditions but made up for the yield with the longer in duction time. SiCSP9 was eluting very early on the Cobalt 2+ column so t he purification strategy was modified to use less eluting imidazole ( 20, 30, 50 mM imidazole elutions) and avoided


55 the FPLC for isolation and desalting (only relying on the dialysis against 10 mM HEPES 100 mM NaCl pH 7.5 for 2 days at 4 C). The differenc es in purification protocols remain unresolved but are likely due to effects derived from minor differences in structure and sequenc e of each protein. The mass spectrometry analysis found a single peak at the molecular weight predicted of SiCSP1. This en sured that there was no endogenous ligand embedded in the binding pocket that could potentially interfere with the binding analysis. This however was performed only on one sample because of setbacks in protein yields of other samples and financial feasibi lity. Thanks to the the University of Texas; we were able to compare the binding activity of our recombinant SiCSP1 to that of the originally isolated SiCSP1. This however was performed only on SiCSP1 a s we had previous binding data ( Gonzalez et al. 2009 ) that would provide a comparison to a native SiCSP. In conc lusion, all fourteen SiCSPs were successfully expressed and purified from an E. coli host. These data form the basis for characterization of the proteins.






58 Table 2 2 Native and Recombinant Si CSP Properties Properties of the original amino acid sequences (without signal peptides) as well as th e recombinant SiCSPs The molecular weights and pI values were predicted using Ex PAS Compute pI/ MW tool ( ). Protein Amino Acid Length Number of Cysteines Number of Tr yptoph ans Molecular Weight (Da) pI Fusion amino acid length Fusion Molecular weight (Da) Fusion pI SiCSP1 92 4 3 10702.84 4.19 115 13289.63 5.16 SiCSP2 91 4 2 10659.25 5.86 114 13 246.04 6.82 SiCSP3 90 4 3 10730.17 4.64 113 13316.96 6.0 2 SiCSP4 99 4 2 12050.60 5.21 122 146 37 .38 6.40 SiCSP5 92 4 3 11145.70 5.09 115 13736.49 6. 24 SiCSP6 109 4 1 12411.04 5.32 132 15071.96 6.35 SiCSP7 109 4 1 12734.48 7.85 131 1 539 5 .40 8.3 8 SiCSP8 98 4 2 11425.87 6.14 120 143 17 01 6. 89 SiCSP9 90 4 1 10245.84 9.87 112 12 906.76 10.05 SiCSP10 91 4 2 10133.35 4.44 113 12 564 .0 5 6.01 SiCSP11 101 4 1 11906.45 8.86 123 14 567.37 9.03 SiCSP12 111 4 1 131 21.86 5.68 133 1 6 013.00 6. 52 SiCSP13 100 4 2 11793.41 8.64 122 14454.33 8.86 SiCSP14 101 4 2 11811.43 6.77 123 14 472 34 7.77


59 Table 2 3 Overview of CSPs in S. invicta. The average final concen trations of isolated SiCSPs are provided below and achieved by using the parameters for induction. The concent rations of imidazole at which CSPs eluted were noted following SDS PAGE analysis. Protein Average Y ield (mg/mL) IPTG (mM) Induction T emperature ( C) Induction Time (Hrs.) Elution Fraction (mM Imidazole) SiCSP1 6.77 0.5 18 20 1 00 3 00 SiCSP2 1.52 0.5 18 20 300 500 SiCSP3 1.20 0.25 37 5 2 00 300 SiCSP4 1.39 0.5 18 20 200 300 SiCSP5 2.59 0.5 18 20 300 500 SiCSP6 0.96 1.0 18 48 200 500 SiCSP7 0.9 1 0.5 18 20 200 500 SiCSP8 4.57 0.5 18 20 200 500 SiCSP9 0.35 0.5 18 20 20 50 SiCSP10 0.79 0.5 18 20 200 500 SiCSP11 0.84 0.5 18 20 200 500 SiCSP12 1.68 0.5 18 20 200 500 SiCSP13 1.18 0.5 18 20 200 500 SiCSP14 2.46 0.5 18 20 200 500


60 Figure 2 1 Strategy for Recombinant CSPs. The construction of the recombinant SiCSPs was achieved using the T7 promoter under the control of the lacI operator to achieve efficient protein production. The nucleic acid sequence was codon optimized to facilitate production in E. coli. A 10 His tag was added to the C terminus of the construct to allow for purification by a Cobalt 2+ ion exchange column and a thrombin recognition site was added between the target gene and the His tag to allow the thrombin digestion to remove the His tag yielding the native polypeptide of each CSP Thrombin cut site


61 Figure 2 2 P urified Recombinant SiCSPs. Purified SiCSPs were electrophoresed through 16% polyacrylamide gel. Electrophoresis was performed on SiCSP1 SiCSP 2 SiCSP 3 SiCSP 4 SiCSP 6 SiCSP 7 SiCSP1 0 SiCSP1 3 SiCSP 1 4 10.5 kDa 14 kDa 22 kDa 10.5 kDa 14 kDa 22 kDa SiCSP 5 10.5 kDa 14 kDa 22 kDa SiCSP 8 SiCSP 9 SiCSP 11 SiCSP 12


62 multiple different gels with the final images spliced. Samples were stained with Bulldog Bio AcquaStain (Bulldog Bio, Portsmouth, New Hampshire). Each lane was loaded with 15 ug of protein. Proteins labeled on image (Top). Note the less pure status of SiCSP9 due to the altered purification process. The typical purification process of SiCSP1 from left to right: pellet, supernatant, c olumn flowthrough, 100 mM imidazole, 100 (second elution), 200, 300, 500, Protein Standard, crude lysate.


63 Figure 2 2 Continued


64 Figure 2 3 ESI MS profile of SiCSP 5 Electrospray ionization mass spectrometry was performed on SiCSP 5 to establish the expected molecular weight and the absence of an E. coli derived non native endogenous ligand, which would be indicated by mul tiple primary peaks The predicted molecular weight was 13736.49 Da.


65 Figure 2 4 Phylo genetic analysis of the 21 CSPs from Solenopsis invicta. The phylogenetic distribution of CSPs from Solenopsis invicta (SiCSP), Harpegnathos saltator (HsCSP) C amponotus floridanus (CfCSP), Acromyrmex echinator (AeCSP), Drosophila melanogaster (DmCSP), and Drosophila grimshawi (DgCSP) were assembled using a maximum likelihood tree and amino acid substituion model (JTT) The numbers indicate HS like branch suppor t SiCSPs are highlighted in red for visualization.


66 CHAPTER 3 GENE EXPRESSION ANALYSIS OF S. INVICTA CHEMOSENSORY PROTEINS Introducti on Chemosensory proteins have been canonically defined as playing a role in olfaction and chemoreception This role can be as a shuttle factor bringing perceived chemicals to the appropriate odorant receptor(s) or as a means for removing the signal from t he receptor in order to modulate the signal, relaxing the response to an odorant. However, the existence of expanded gene families (especially in ants) and detection of CSPs in non sensory tissues and imply functions(s) other than exclusively in chemorecep tion and/or broader complementary function related to chemoreception. For example, kin /nestmate recognition and identification is linked to the presence of endogenously derived hydrocarbons on the surface of the ant ( Vander Meer, 1983 ) The not produced and deposited on t social interactions including grooming, sharing of bodily fluids, and surface cleaning and deposition of cuticular compounds by other ants. In this manner, a colony is thought to contain as signals to identify friend, caste, or stage of development relationships to the other ants in the colony. Thus, CSPs may function as a means to sequester chemicals as they are exchanged and depo sited on the surface of ants. In addition, CSPs may function in endogenous transport of ligands throughout the body potentially targeting chemical molecules to appropriate tissues. The fire ant has a number of glands used for producing semiochemicals and various regions of their anatomy may be used for chemoreception ( Holldobler and


67 Wilson, 1990 ; Vargo and Hulsey, 2000 ) The an tennae are the primary odorant reception organs however, recent studies have shown that sensilla on the tarsi and limbs may play secondary roles in contact chemoreception ( Renthal, 2003 ; de Brito Sanchez et al. 2014 ) Insect heads contain four main glands: the mandibular, the maxillary, the propharygeal, and the post pharyngeal glands. The mandibul ar gland produces alarm and defense pheromones and the mandibles are also covered in sensilla associated with the gustatory response. The propharyngeal and maxillary glands produce digestive enzymes while the postpharyngeal gland produces the colony odor involved in identification, and cues for food odors to assist in foraging as well as feeding of the larvae (Table 1 2) important in regards to maintenance of social cohesion and nestmate recognition The postpharyngeal gland acts as a reservoir for cuticular surface compounds (hydrocarbons, fatty acids, alkyl a ldehydes, and alkylketones ) that are produced by each individual but also as acquired during grooming and other contact behaviors. Thus, dur the postpharyngeal gland), to be redistributed in further grooming/contact events to The thorax contains the salivary enzyme producing labial gland and the metapleural gland that produces alarm pheromones as well as antimicrobial and antifungal compounds. In fire ants, the abdomen contains the Dufours gland which releases the trail pheromone and a poison gland, which secretes and injects venom into prey as well as performs a role in the disinfection of the colony and nestmates. The fire ant can aim the abdominal region like a miniature cannon to spray the contents of the


68 poison gland as an aggres sion response but also to coat the walls of the nest as an act of house cleaning. Although proteomic techniques can allow for the analysis of proteins found in different tissues; sample preparation, issue of protein yields (especially from an ant), and th e associated cost s may limit applicability of the method ( Jungblut et al. 2008 ) Quantitative r eal time PCR, i.e. the measurem ent of transcript levels, can provide sensitivity in the femtogram of RNA range and allows for the simultaneous measurement of d ifferent genes in a wide variety of tissue sources ( Pfaffl, 2001 ) Knowledge concerning the gene expression levels of CSPs, as a function of tissue and developmental stage, can shed important insights into the dynam ics of CSP expression and would point to those tissues/stages in which the protein can be expected to be found. Furthermore, information concerning CSP gene expression levels can be integrated with the ligand specificity of the proteins to provide a more c omplete picture of the potential functions of the CSPs Primers for QRT PCR were designed to the exonic regions of all twenty one SiCSP genes as well as ten housekeeping genes for potential use as references. Developmental specificity of expression was exa mined by pre p aring RNA samples from workers, female alate s, male alate s, eggs, early instar larvae, late instar larvae, and pupae. For adult ant (workers, female and male alate s) samples were further subdivided into specific anatomical pa rts that included the antennae, heads, thoraces, and abdomens. It should be noted that fire ants develop in four stages of expansion (instars) before pupation. Due to difficulty in clear discrimination between instars, in our study, the ants were separate d only into two instar stages termed early and late. The early


69 instar set represents all potential castes of fire ant, however as they develop, minor worker larvae do not expand as greatly in size as the rest of the castes. The late instar larvae represen t developing major workers, male alate s, and female alate s, which look identical in the larval stages but are distinct from minor workers By correlating the SiCSP gene expression levels to tissue and developmental stages we can track trends o f expression. This information can be used to help make functional predictions. For example, it is possible that the ligand binding profiles of several CSP may be very similar but that they are differentially expressed in different tissues. Thus, two CSPs may both bind hexadecane, however if one is expressed in the antennae and one in the abdominal region, one would predict that former plays a role in the binding of hexadecane during olfaction (antennael specific) wheras the latter may play a role in trans port or dispersal of hexadecane (abdominal specific). Materials and Methods Primer Design Primers for the twenty one S icsp genes w ere designed to the expressed sequence tags using Primer BLAST ( blast/ ). Each prim er was optimized to be 20 bases long and have melting temperatures greater than 57 C. Primers were initially validated in silico for specificity against each CSP and the S. invicta genome. Primers sequences are shown in Table 3 1 and Table 3 2 Primers w ere purchased from Life Technologies ( ) Upon receipt, the primers were diluted to 25 M and aliquot into several tubes in order to minimize freeze thaw cycles when used


70 RNA Preparation Individual S. inv icta ants (derived from a colony maintained by Dr. R. Pereira, University of Florida Dept. of Entomology and Nematology) were sorted by castes and developmental stages for RNA extraction Of the adult castes, ants were dissected into four sections: antenn ae, head, thorax (alitrunk), and abdomen (gaster). ( Wang et al. 2007 ) Tissues (between 50 and 100 mg) were frozen in liquid nitrogen and homogenized using a micr o pestle and mortar. Tissues were preserved in RNA Later (Ambion) and immediately added to 500 uL TriReagent per 100 mg tissue. Homogenates were centrifuged for 1 min at 16,000 xg to remove cutic ular/ exoskeleton and other insoluble debris The remaining soluble portion was applied to ZymoSpin RNA Prep columns and processed as per instructions ( Zymo Research, Irvine, CA ) After column washing, total RNA was eluted using pre warm ed (37 o C) RN a se free water T otal RNA concentration was det ermined spectroscopically cDNA Synthesis Reverse transcription from total RNA samples was achieved using the Applied Bio systems cDNA Synthesis Kit Applied Biosystems, en/home.html ) Reactions were adjusted to start with 1 2 ug of purified RNA Sampl e sizes were adjusted to produce 12.5 n g cDNA per RTPCR reaction Reverse transcription reactions w ere performed in a Biorad T1000 thermocycler with the typical cycling con ditions: 25 C for 10 minutes 37 C for 120 minutes and 85 C for 5 minutes


71 Real Time Polymerase Chain Reaction Real time PCR was performed using the Illumina ECO (Illumina, San Diego, California) platform that allows for rapid thermocycling and reduced reaction sizes. Typical reaction mixtures contained (in a total volume of 15 l): 12.5 ng of template cDNA 2 5 M of each primer and 7.5 l iQ Sybr green super mix (Biorad, Samples were run through 40 cycles (30 seconds at 95 o C, followed by 30 seconds at 58 C ) of PCR, followed by a 1 minute melting curve (20 90 o C) Cycle t hresholds (C T ) we re calculated using the ECO software ( Illumina, San Diego, California) ) Controls included no primers, each primer individually, no template, and the RNA preparation (no cDNA conversion, genomic contamination control) Primer efficiency experiments were performed in order to determine the rate at which template cDNA is amplified from each pair of primers to facilitate the quantification of the gene ( S i C SP and reference) fragments in the isolated RNA preparations. Primer effic iency experiments were carried out by establishing a standardization curve. Dilutions of cDNA: 10ng/uL, 1 ng/uL, 0.1 ng/uL, 0.01 ng/uL, and 0.0 1 ng/uL were used to establish the standard curve against each primer pair in the proportions : 2.5 uM:2.5uM (1:1 ) 5.0 uM:2.5 uM (2:1) and 2.5 uM:5.0 uM (1:2) A linear regression was determined when the results were plotted on a logarithmic scale. Using the slope value, the Efficiency (E) value was calculated using the formula: (3 1) Results A set of preliminary experiments were performed to validate each set of primers (listed in Table 3 1 and Table 3 2 ). PCR Amplification using the primer sets resulted in


72 single bands after analysis on agarose ge ls corresponding to the predicted size of each amplicon (Table 3 3). In addition, melting curves of the PCR products showed a single peak for each reaction product. Of the initial ten reference gene primer sets, five resulted in amplification efficiencies and TBP. Previous study has shown that two reference genes are adequate to establish relative expression in RTPCR ( Daifeng Cheng et al. 2013 ) The primer efficie ncies f or the C S P primer pairs are given in Table 3 3 all of which were >95%. With regards to reference genes, a low C T value is needed to establish consistent expression/normalization across samples using the reference gene. Analysis of the C T values for the five reference genes which showed acceptable primer efficiencies versus RNA/cDNA sample revealed significant varia tion, i.e. 10 15 log units (Figure 3 1A). Note that each point represent the C T value of the referenced gene as derived from a different sampl e, i.e. biological duplicates, within sample variation was significantly smaller (Figure 3 1B). TBP expression was low (highest C T values) and could not be detected in some tissues. The lowest mean C T values observed belonged to GAPDH a poor candidate for a reference gene. Despite their low means however, as mentioned, the variation in signal (derived from developmental stages as well as dissections of each caste of ant) r endered any of the reference genes examined inappropriate for normalization. Since no reference gene proved to remain invariant across the samples examined, the data with respect to the SiCSP s were normalized to input RNA concentration, with each figure c ontaining the expression level of Ef1 for comparison.


73 Only nominal expression of the CSPs (as for Ef1 ) was seen in eggs, likely indicating a general low transcription rate i n this developmental stage (Figure 3 2). In contrast, strong EF1 expression was seen in early and late instar larvae as well as pupae. SiCSP7 and to a lesser degree CSP14 appear to be the ones that are expressed at high levels early, remaining highly expressed in late instar and pupae. Two CSPs (in addition to CSP7) are highly expressed (above CSP14 for example) at the late instar stage, both of which decrease somewhat in expression levels in the pupae. The notable emergence of CSP1 expression in pupae (but not other stages examined) can be seen. In general, moderate expression of the other C SPs was seen from the early instar stages onwards. Exceptions to this include CSP2 and to a lesser extent CSP5 whose expression levels were low in the samples examined. CSP10 was notably absent in the early instar (aside from eggs in which all CSP expression was low), CSP15 was absen t in pupae but moderately expressed in early and late instar larvae. CSP6 was somewhat unique in that it was found in early instar and pupae but not late instar larvae. Almost all twenty one CSPs, with the exception of CSP6, were expressed in the antennae of fire ant workers (non reproductive female), with CSPs 1, 3, and 21 levels the highest (Fig ure 3 3). Lower CSP transcript levels were generally seen in head and thorax tissues, with CSPs 1, 18, 20, 15, 17, 19, 6, 9, 11, 12, and 16 levels particularly lo w. CSP3 retained moderate expression levels in the head but lower in the thorax, whereas CSPs 20, 13, 19, 12, and 16 levels appeared higher in the thorax compared to the head. In contrast the expression of CSPs in abdomen appeared robust, with the excepti on of CSPs 6, 9, 11 and to a lesser extent 1.


74 In sharp contrast to the workers, female alate s appeared to only produce CSP8 and to a lesser extent 9 in their antennae (Figure 3 4). It is further noteworthy that expression was also almost n on existent, implying dramatically lowered transcription rates in these tissues. levels in female alate head, thorax, and abdomen sections were comparable to their worker counterparts. In the female alate head section, CSPs 3, 17, and 6 were highly expressed, with the rest moderately expressed, with the exceptions of CSPs 9, 11, and 16. The thorax of female alate s contained moderate transcript levels of approximately half of the fire ant CPS repertoire (CSPs 3, 5, 10, 18, 4, 8, 1 3, 14, 15, 21, 7, and 12) with the remaining (CSPs 1, 2, 20, 17, 19, 6, 9, 11, and 16) expressed at lower levels. Robust expression of CSPs was seen in female alate abdomen sections, with high levels of CSPs 3, 5, 10, 4, 8, 13, 14, 21, and 7, moderat e levels of CSPs 1, 2, 18, 20, 15, 9, and 12, and low levels of CSPs 17, 19, 6, 11, and 16 transcripts detected. Expression levels in male alate antennae w ere moderate for most CSPs, all of which appeared to be expressed in these tissues (Figure 3 5). In contrast to the antennae (as well as the heads and thoraces of workers and females) very little CSP expression was seen in these tissues in the male alate s, with the exception of CSP17 in the head, and CSPs 17, 19, and 21 (the latter two high ly expressed) in the abdomen. It should be noted that as had been seen for the female alate antennae but in this case the male head, thorax, and to a lesser extent abdomen, expression levels were low. Expression of CSPs 2, 5, 10, and 11 were high in the male abdomen, with moderate levels of CSPs 8 and 13, lower levels of CSPs 1,3, 18, 14, 15, 17, 19, 21, 6,


75 and 9 seen. In these tissues, very little transcript corresponding to CSPs 20, 4, 7, 12, and 16 were detected. Discussion Q RT PCR was u sed to examine the tissue and development distribution of the SiCSPs. Ten reference genes were examined, with five showing acceptable primer efficiencies. These five included: Quantification of transcript levels of these five genes, however, did not result in the identification of any that remained stable with adequate expression (i.e. low C T value) over all of the different samples examined. Primers designed to the TATA binding protein (TBP) transcript could not detect any appreciable transcript in several samples. Actin expression appeared bimodal, and a wide variation was seen with , and GAPD H Due to this variation, all samples were normalized to input cDNA concentration with respective values to and GAPD included for comparison. Primers designed to each CSP displayed specificity as observed by single bands of appropriate molecular weight and single peaks in PCR product me lting curves. It should be furth er noted that CSPs share low (<5 0%) homology at the nucleotide sequence level even within related clades and primers were cross referenced with other CSPs as well as the S. invicta genome. Primer efficiencies were also verif ied for each CSP (Table 3 3) The data show strong expression of most CSPs in the antennae of workers, moderate expression in the antennae of male alate s, and little to no expression in female alate s, although for the latter, even Ef1 could barely be detected. These data suggest the somewhat surprising finding that female alate antennae may be partly transcriptionally silent.


76 In contrast two CSPs largely absent in heads of workers and males were highly expressed in the head of female alate s, namely CSPs 17 and 6. In general, males heads appeared similar to female alate antennae, i.e across the board low expression, including for Ef1 with the possible exception of CSP 17 (which as indicated above was highly express ed in female alate heads). Worker heads showed moderate to little expression of the CSPs. Within the thorax only subtle differences were seen between workers and female alate s, whereas male thoraces displayed generally low expression (including for Ef1 ), with the notable exception of CSPs 17, 19, and 21. Finally, the abdomens of female alate s appeared rich in CSP transcripts with many highly expressed, and somewhat lower, but similar in overall pattern in workers. CSP transcripts in the abdomen o f male appeared somewhat mixed. Ef1 levels were low, however, a number of CSPs showed relatively high expression including CSP s 2 and 3, and to a lesser extend 10, 11, and then 8, 13, 17, 19, and 21. Within the developmental stages examined a number of patterns emerged. First it should be noted that early instar samples include all future differentiated forms, i.e. small to large workers, and male and female alate s, although are likely dominated by workers. Late instar larval samples include CSP 20 21, 7 CSP expression (as well as Ef1 ) was almost absent in eggs. Ef1 expression was robust in the other stages examined. Early instar larvae notably expressed high levels of SiCSP7 transcript that persisted in the other developmental stages. Late instar larvae appeared to add high le ve l s o f expression of CSP20 and 21, with a sharp drop in CSP6 expression. Pupae began to express high levels of CSP1 and 14, but little CSP2 or 15.


77 A number of observations can be drawn from the data, suggesting directions for further experimental testing. (1) Male specific CSPs : N o male specific CSP s were obvious in either the antennae or the head, however, CSP 19 (and possibly 21) were specific to the male abdomen (CSP21 was also highly expressed in late instar larvae). In the male abdomen, CSP2 was high (although also found in moderate levels throughout workers and parts of female alate s) as was CSP11, the latter absent in female alate s, found in moderate levels in worker antennae and in pupae. (2) Female alate specific CSPs CSP8 a nd to a lesser extend 9 and 20, were essentially the only CSPs found in female alate antennae, however, they were also found widely distributed in many other sample. CSP6 and possibly 17 appeared to be mostly specific to the female alate head, expressed at high levels. CSP expression in the thorax of female alate s was mostly similar to that of workers, as was the CSP expression in the abdomen. (3) Workers displayed a general expression of most CSPs in the antennae and abdomen, with decreas ed levels seen in the head and thorax regions. With respect to developmental stages; (1) eggs appeared largely dormant in CSP expression, CSP7 initiating in the early larval stages, added with CSPs 20 and 21 in the late instar larval stages, and then with CSP1 in pupae. Overall, these data show a complex pattern of CSP expression related to caste and development. These data show that CSPs are not constitutively expressed in caste antennae (or body part or developmental stage) indicating specialized function al roles that may be able to partially account for ligand binding redundancies. Our data support the idea that for some CSPs that might share ligand binding profiles might indeed perform similar functions but are differentially regulated in order for prope r development


78 programs to occur. The abundance of CSPs in the antennae of workers likely mirrors the wide range of chemicals perceived by these ant, whereas the lack of CSPs in female alate antennae suggests poor olfactory responses in the ants. It i s intriguing to speculate the CSP8, the sole well expressed CSP in female alate antennae may be linked to either mate finding to a crucial post mating process, i.e. finding and/or excavating a suitable nest or interaction with the first brood. The ob servation that workers and female alate abdomen are rich in CSPs suggests potential roles in either ligand dispersal and/or sequestration. The finding of CSPs in early and late larval, as well as pupal stages was unexpected since these stages are not considered to require olfactory response (as opposed perhaps to gustatory), however, our data indicate that some aspects of chemical sensing occurs in these stages. Interestingly, CSP1 (highly expressed in worker antennae and the first CSP characterized) was also highly expressed in pupae. The implications of high levels of CSPs in pupae are unclear, but may reflect priming of chemosensory responses. With respect to development CSPs may be produced to assist in nursing of the larvae. They could act as ide ntification signals for care including receiving food and the colony odor. These studies provide a basis for knowing which CSPs to look for in different tissues and developmental stages and future experiments confirming protein production in the various t issues (via antibody or isolation) will provide a link between the gene expression studies and actual CSP content.






81 Table 3 3 Validation of R TPCR Primer sets. Primers for CSPs and targeted reference genes we re validated both on agarose gel and through RTPC R to confirm proper amplicon size and the propagation of a single amplicon. *Primers were pre viously validated ( Cheng et al. 2013 ) **Primers failed to prime against RNA samples and did not produce usable E values (less than 90% efficiency) and subsequently omitted from the study. Primer Pair Amplicon Size (bp) Size on Gel (bp) Ratio (uM:uM) T M ( C) E % value SiCSP1 138 150 2 .5:2.5 60 93.84 SICSP2 179 200 2.5:2.5 60 100.04 SiCSP3 78 100 2.5:2.5 60 90.60 SiCSP4 143 150 2.5:2.5 60 98.52 SiCSP5 149 150 2.5:5 60 96.14 SiCSP6 1 67 150 2.5:2.5 61 112.86 SiCSP7 80 100 2.5:2.5 59 93.80 SiCSP8 144 150 2.5:2.5 60 91.01 SiCSP9 165 150 2.5:2.5 60 101.14 SiCSP10 187 200 2.5:5 60 113 .81 SiCSP11 221 200 2.5:5 60 95.77 SiCSP12 192 200 2.5:2.5 60 102.70 SiCSP13 1 13 100 2.5:5 60 112.28 SiCSP14 121 100 2.5:2.5 61 97.43 SiCSP15 139 150 2.5:2.5 60 104.12 SiCSP16 125 100 2.5:5 60 84.91 SiCSP17 106 100 2.5:2.5 61 102.21 SiCSP18 169 150 2.5:2.5 60 1 11.45 SiCSP19 128 100 5:2.5 61 91.96 SiCSP20 143 150 2.5:2.5 59 104.21 SiCSP21 93 100 2.5:5 60 108.44 101 100 5:2.5 60 87.52 Tubulin 95 100 ** ** ** 125 100 ** ** ** 79 100 2.5:5 60 95.76 301 2.5:2.5 95.60 GAPDH 189 4 2.5:2.5 104.00 RpI32 174 200 ** ** ** RpS20 90 100 ** ** ** TBP 524 2.5:2.5 101.00 18srRNA 14 6 150 ** ** **


82 Figure 3 1. Cycle Thresholds of Reference Genes. Of the 10 reference gene candidates, 5 pairs of primers were determined to be nearly 100% efficient in amplification. Samples were tested against different tissues to determine th e spread of CT values for each reference gene (A). EF1 CT values presented for each tissue. Data is representative of 3 independent assays with standard error mean shown as error bars. A B


83 Figure 3 2 SiCSP Expression of SiCSPs in Developing Fire Ants RT PCR was performed on RNA isolated form Pupae, Late instar larvae, early instar larvae, and eggs. Warmer colors represent higher expression and are relative to the log of the C T values. Data are representative of three independent preparations. Col or scale shown below in log base ten of the C T Pupae Late Instar Larvae Early Instar Larvae Eggs


84 Figure 3 3 Expression of SiCSPs in Fire ant Worker tissues Expression of SiCSPs in worker ants. RT PCR was performed on RNA isolated form dissections of the antennae, heads, thoraxes, and abdomen s. Warmer colors represent higher expression and are relative to the log of the C T values. Data are representative of three independent preparations. Color scale shown below in log base ten of the C T Antennae Head Thorax Abdomen


85 Figure 3 4 Expression of SiCSPs in Fire an t Alate Females. Expression of SiCSPs in alate female ants. RT PCR was performed on RNA isolated form dissections of the antennae, heads, thoraxes, and abdomens. Warmer colors represent higher expression and are relative to the log of the C T values. Data are representative of three independent preparations. Color scale sho wn below in log base ten of the C T Antennae Head Thorax Abdomen


86 A Figure 3 5 Expression of SiCSPs in Fire ant Alate Males. Expression of SiCSPs in alate male ants. RT PC R was performed on RNA isolated form dissections of the antennae, heads, thoraxes, and abdomens. Warmer colors represent higher expression and are relative to the log of the C T values. Data are representative of three independent preparations. A) Color s cale shown in log base ten of the C T Antennae Head Thorax Abdomen


87 CHAPTER 4 LIGAND BINDIND SPECIFICITIES OF SiCSPs Introduction One of the putative functions of CSPs is to shuttl e odorants to and from the odorant receptor through the aqueous sensillar lymph. This may be of partic ular importance for volatile and/or hydrophobic chemicals with low aqueous solubility. Though odorant receptors from both vertebrates and invertebrates have been shown to be able to sense and initiate signal transduction events in the absence of additional proteins including CSPs, CSPs may add context sensitivity and/or attenuation events to odorant reception. It should also be noted that the direct response of odorant receptors to various chemicals has typically been performed in heterologous cell expre ssion systems and tissue culture and may not completely reflect the subtler organismal context of the ORs and their interactions with odorants. Our hypothesis is that the CSP repertoire (of an insect) acts as a means for binding a wide range of chemical li gands, both of extrinsic and intrinsic origin involved in chemoreception, with each CSP or groups of CSPs showing specificity to discrete ligands. Thus CSPs may show multiple ligand specificity with overlap between proteins. Our objectives are to: (a) de termine the ligand binding specificity for each CSP, (b) correlate specificity with the phylogeny, and (c) correlate specificity with gene expression data. These data will address several critical issues concerning CSPs including: (a) what is the nature of ligand bound, i.e. do some CSPs show broad substrate specificity and other narrow or are most CSPs similar in binding profiles? (b) Do CSPs within the same clade show similar ligand binding profiles or have they evolved specialized binding capabilities? ( c) If CSPs are general ligand binding proteins, are their in vivo biological activities mainly


88 determined by expression and localization? The first step in addressing these questions and developing a comprehensive picture of CSP activity is to determine th e ligand binding specificity. Knowledge concerning the ligand binding affinities can also shed light on the relative importance of perception, transport, and/or dispersal of specific chemical compounds. In order to screen a wide range of ligands for bindin g to each CSP, a relatively facile assay is needed. The chemical compound 1 N Phenyl Naphthyl amine (1 NPN) is a small fluorescent ligand that displays a characteristic shift in emission when excited in a hydrophobic environment such as the binding pocket o f a protein, specifically producing a peak at 415 nm when excited at 350 nm. To date, all Si CSPs examined bind 1 NPN. Since physiologically relevant chemicals typically do not have usable excitation/emission spectra when bound to proteins ( CSPs ), 1 NPN can be used in competition assays to determine apparent association/disassociation constants Thus, by loading the protein with a semi saturating concentration of 1 NPN competition experiments can be performed. From the competition data the apparent K d can be calculated using experimentally determined IC 50 (inhibitory concentration to compete for 50% of bound protein) values. The expression for K d Apparent ( K d App) as a function of the IC 50 is as follows: ( 4 1) K d App was used to quantify the binding affinity of the CSP for each ligand. However, since K d App is and indirect measurement dependent on the CSP in NPN


89 Drosophila LUSH protein was originally thought to have an af finity for ethanol, and potentially other alcohol compounds, but upon further analysis it was revealed that the protein display ed stronger binding to vaccenyl ( Kruse et al. 2003 ) This is an example of why it is nece ssary to test several compounds. Additionally, by establishing the ligand binding specificity or range for each CSP we can estimate roles for each CSP in the fire ant. CSPs which bind numerous odorants may play roles in the reception and release of general odorants and subject to transcriptional or translational control to modify sensitivity to signals while CSPs that have a narrow range or specific ligand may play a m ore directed role in increasing sensitivity to that particular ligand (or pheromone) and be indicative of selective pressure. Materials and Methods Chemicals Chemicals used in this study and their sources are listed in Table 1 1 All Chemicals were purcha sed from Sigma Aldrich company ( 1 NPN Binding Assay Purified proteins were isolated as described in Chapter 1. To determine the binding constant ( K d [ NPN ] ) 1 N Phenlynaphthylamine (NPN) was used at increasing concentrations from 0 to 20 M with 1.0 M purified SiCSP (in 10mM HEPES 100 mM NaCl buffer pH 7.5). Stock solutions of each c oncentrations of 1 NPN w ere prepared in methanol ( 100 %) Typical reaction mixtures contain ed ( 200 L total volume); 1 M SiCSP ( 196 l ) and 4 l respective 1 NPN stock solution bringing the total methanol concentration to 2% Each protein sample w as blanked against buffer (no protein)


90 containing the corresponding concentration of 1 NPN. Reactions were performed in triplicate using Greiner Bio (Monroe, North C ar olina) high adhesion black Fluotrac 600 fluorescence microtiter plates (Greiner, location) and read on a Spectramax Gemini XS spectrophotometer ( Molecular Devices Sunnyvale, California ) with 350/415nm ex/em. Each experiment was performed at least twice and then subsequently repeated with each independent batch of purified protein. Results were averaged and fit to the Hill equation ( Hulme and Trevethick, 2010 ; Ortiz, 2013 ) for determining whether the data could be fit to a one site bi nding model or to the Adair equation ( Simoni et al. 2002 ; Stefan and Novere, 2013 ) for determining w hether the optimal fit corresponded to a two site binding model In addition, to confirm the fits a Sca tchard transformation was performed on each set of data by graphing the solved values in terms of Bound vs Bound/Free (or in our case transform the RFU values into X values and then plot RFU/ [NPN]). A single site binding protein will appear to have a linear regression with a slope equal to 1/Kd while a two site cooperative binding would show an inverted parabolic regression ( Bordbar et al. 1996 ) As indicated, the K d was for 1 NPN binding was measured for each batch of purified protein to confirm the integri ty of the protein binding activity 1 NPN Competition Assay s The apparent dissociation constant s of test ligands were determined via competition with 1 NPN binding. Typically, reaction mixtures contained 1 M SiCSP + 20 M NPN. Stock solutions of test compounds from 0.1 n M to 5 m M ( in 100 % hexane for all hydrophobic ligands ) were prepared such that reaction mixtures received identical volumes of the test ligand (with afinal concentration of 2% hexane for hy drophobic


91 ligands) in order to reach the desired ligand concentrations in order to control for any solvent effects. Experiments were performed and examined as above for the 1 NPN assays Data w ere converted to percent reduction of the relative fluorescent units ( RFU s) and fit to one site or two site competition curves. This prov ided for an internal reference as data was relative to the starting point containing only 10 mM HEPES buffer, CSP and 1 NPN. From the predicted IC 50 (inhibitory concentration at 50% competition) the apparent K d was determined using equation 4 1. Results All SiCSPs bound 1 NPN with the binding curves displaying optimal fits to either a single or double binding site models. SiCSPs 2, 4, 8, 10, 12, 13, and 14 showed classical one sit e 1 NPN binding curves (Fig ure 4 1 ). Analysis of the binding curves for SiCSPs 1, 3, 5, 6, 7, 9, and 11 indicated a best fit using a tw o site 1 NPN binding model (Fig ure 4 2 ). A summary of the 1 NPN K d s in given in Table 4 1 Of the CSPs that shared simila rity with CSPs from other insect (beyond ants), i.e. SiCSPs 6, 7, 9, 11, 12, and 16, all but SiCSP12 showed a best fit to two site binding models (note that the data for SiCSP16 is from others in the lab and are mentioned here for comparative purposes). Of the ant specific CSPs, three, SiCSP4, 17, and 19 were dispersed within other ant lineages, and SiCSP4 showed best fit to a single binding site, whereas SiCSPs 17 and 19 to two binding sites (again the latter two data are included for comparative purposes, data generated by others in the Keyhani lab). As previously mentioned, phylogenetic analysis indicates two fire ant specific CSP gene expansions. One comprising SiCSPs 8, 13, 14, 15, and 21, all displayed best fits to single binding site models (SiCSP15 a nd 21, data Keyhani, lab). The other fire ant specific CSP clade consisting of CSPs 1,2, 3, 5, 10, 18, and 20 showed mixed one and two binding site fits;


92 SiCSPs 2 and 10, which grouped more closely to each other than to the other proteins in the clade show ed best fits to one site binding, whereas the remaining proteins (SiCSPs 1, 3, 18 and 20) showed best fits to two site binding models. 1 NPN binding was used as a reporter for competition experiments against a wide range of ligands (> 70) belonging to dif ferent chemical families. These included alkanes (C 9 C 41 ) that constitute important cuticular compounds, fatty acids, e.g. lauric, palmitic, and oleic acids, alcohols, e.g. hexenol, octanol, nonanol, aldehydes, actetates, various odorants, plant derived te rpenes, pheromones/pheromone precursors, various chemical repellents, carbohydrates, and other compounds. A number of SiCSPs appeared to possess exceptionally broad ligand binding ranges. These included SiCSPs 1, 5, 6, and 9. These proteins bound a wide ra nge of alkanes, however, important differences were noted even between these broad range hydrocarbon binders. SiCSP9 did not bind to either dimethyoxybenzoic, benzoic, lactic, or hydroxylauric acids, whereas SiCSP6 bound all except dimethoxybenzoic acid, S iCSP1 did not bind to either lactic or benzoic acids, and SiCSP5 bound all four compounds. SiCSP1 did not bind to palmitic or oleic acids, which the three other proteins were capable of binding. SiCSPs 1 and 5 also did not bind nonanal, whereas SiCSPs 6 an d 9 could. SiCSP6 could also bind alcohols (hexenol, octanol, etc) with only SiCSP1 sharing hexenol binding but not binding to the other alcohols examined. CSP11 also bound a wide range of the alkanes as well as the fatty acids, but showed a more patchy bi nding to alcohols, acetates terpenes, and other odorants, although strong responses to linalool, citranellal, and caprophylene were noted. Intriguingly, SiCSPs 5, 6, 9, and 11 shared binding to trehalose, although SiCSP6 also bound sucrose and


93 glucose. Si CSPs 3 and 7 also showed broad ligand binding ranges, although not as much as those described above. Notably, SiCSP3 did not appear to bind low chain alkanes C 13 19 whereas SiCSP7 did not bind the higher chain alkanes (C 26 41 ). As mentioned, the remainin g seven SiCSPs examined were modeled to a best fit line of a single binding site protein These proteins (as opposed to the two site) displayed in general, mu c h more restricted ligand binding profiles. SiCSP12, which grouped along with other insect CSPs sh owed strong binding to isovaleric acid, short chain alkanes (C 11 ) and moderate binding to the insect repellent DEET (to which most CSPs appeared to also respond to). SiCSP4 showed strong binding to 2 nonanone, as well as to several alcohols, butyrates. SiC SPs 2 and 10, which grouped together within an S. invicta CSP clade, never the less showed different ligand binding profiles. With respect to the hydrocarbon tested, SiCSP2 showed strong binding essentially to mid length alkanes (C 16 C 19 ) whereas SiCSP10 b ound shorter chain alkanes, C 10 and C 11 and to lesser extents up to C 15 In addition CSP2 showed strong binding to 2 octanol, farnesol, nonanal, decanal, lauric acid, linalyl acetate, heptanone, and cinnamaledhyde, whereas CSP10 showed strong binding to 2 nonanone, DEET, and to a lesser extent citranellal, nerolidol, and N farnescene. As previously mentioned, all members of the second fire ant specific clade (SiCSPs 8, 13, 14, 15, and 21) showed single (1 NPN) binding sites. SiCSPs 8, 13, and 14 (the ones under consideration in this work) showed relatively narrow (compared to the CSPs above) and different ligand binding profiles. SiCSPs 8 and 14 showed some binding to shorter chain alkanes (C 9 C 15 ), but SiCSP13 did not appear to appreciably bind most alkan es. In contrast to SiCSP8 (and 13), however, SiCSP14 also showed


94 strong binding to 1 nonanol, nonanol, decanal, lauric aldehyde, and hexyl butyrate. SiCSP13 on the other hand, showed strong binding to 2 nonanone and DEET. Discussion The fluorescent ligand 1 NPN was used as a reporter for examining the ligand binding range of the CSPs. All CSPs examined to date show binding to this compound. Our data show two discrete patterns of binding to 1 NPN, namely those proteins whose 1 NPN binding curves could readi ly be fit to single binding site models, and those that appeared to display better fits to two site models in which cooperative binding was seen It should be emphasized, however, that in a number of cases the r values of the fits between single and double binding site models were not that different and a Scatchard transformation (Bound vs Bound/Free Trasnformation) was performed to determine if a true one site pattern was observed. With this caveat, for purposes of further discussion, the two site models shall be considered as our best current model for the indicated proteins. Displacement of 1 NPN was used to measure apparent affinity constants for the test ligands examined. As an indirect assay, a number of points of caution need to be taken into accoun t. (1) Only a pparent affinity constants are obtained. (2) Our data cannot differentiate between the direct displacement of the 1 NPN from the protein NPN from its bindin g site. The ligand binding experiments revealed several general aspects of the SiCSPs. Several phylogenetically disparate CSPs showed similar (although important difference in specific compounds were always noted) binding profiles especially for the broad ligand distribution binders. SiCSPs 1, 5, 6, 9, and 11 representing two members of a fire


95 ant specific clade (1 and 5) and three members shared with insects at large were similar in binding ranges, suggesting either conserved ligand binding pockets or exam ples of potentially convergent evolution in binding ranges. SiCSPs 2 and 10, which were more closely related to SiCSPs 1 and 5 (as opposed to 7, 6, and 11), however, displayed narrow binding specificity, suggesting the evolution of narrowed (and single bin ding site) specificity within this clade. Overall, i t appears that although some of the SiCSPs have structural homology members of this family have diverged in binding specificities whereas other have converged DEET is a well known insect repellent. T he mechanism of repellence by DEET has been shown include binding to an odorant binding protein (OBP) in mosquitoes ( Murphy et al. 2013 ) Our results show that a number of SiCSPs bind DEET suggesting broad targets for DEET mediated activity and supporting observations that DEET acts as a repellent for (fire ant) ants as well ( Anderson et al. 2002 ) A number of compounds appeared to bind to most SiCSPs. The most notable was amyl cinnamaldehyde, a cinnamon tree odorant, which showed low apparent K d values for most SiCSPs. Other broadly bound chemicals included nonaoic acid (excepting SiCSPs 5 and 10), lauric acid (excepting SiCSPs 1 and 10), C9 (excepting SiCSPs 12, 2, and 13), decanal (excepting SiCSPs 5, 10 and 13), and non yl acetate (excepting SiCSP 14) SiCSP8 and SiCSP14 showed high affinity for low molecular weight alkanes and aldehydes (only SICSP14). SiCSP13 showed high affinity for 2 nonanone, which may show specif icity to this alarm pheromone. However SICSP13 also shows a high binding affinity for DEET. One may speculate that DEET could mask effect mediated by


96 SiCSP13, thus potentially dampening alarm responses by a colony. SiCSP2 seems to show specificity for al dehydes as wel l as linalyl acetate, an ester and 2 octanol. These molecules are environmental odorants and the similarity in the structure of these molecules may imply some specificity of this CSP towards aldehydes and a few other odorants. In addition, however, SiCSP2 seems to also be able t o bind hexadecane and other mid molecular weight alkanes, indicating the potential for complex specificities that cannot easily be reconciled with the structure of the ligands. Similarly, SiCSP10 binds a few low mole cular weight alkanes as well as 2 nonanone, ligands that are structurally dissimilar. SiCSP4, an ant specific but not exclusively S. invicta specific CSP, seems to have specificity for 2 nonanone and Z 3 hexenol in addition to weaker binding to other alc ohols. These are typically environmentally derived odorants produced by plants ( Billen and Morgan, 1998 ; Trapp and Croteau, 2001 ) SiCSP12, the uni que single site binding CSP from the general insect group, seems to bind two ligands tightly, undecane and isovaleric acid (both compounds, however, are also bound by most other CSPs). Undecane acts as a mild sex attractant for a number of cockroach and m oths species, but had also been implicated as an alter signal in various ants ( Holldobler and Wilson, 1990 ) In contrast, isovaleric acid is a mammal derived odorant, which may allow for the sensing of such organisms. For most o f the two site binding SiCSPs, binding was appeared cooperative, with the K d for the second site lower than for the first. Although probing the biological significance of these observations requires additional experimentation this could provide for layere d sensitization (and hence response) to the chemicals detected and/or means


97 for gauging chemical concentrations before response. Although multi chemical mixtures were not tested, the potential for two binding sites also raises the intriguing possibility of integrating two signals (chemicals) within the response of a single protein. The binding profiles of the SiCSPs paint a complex picture revealing that closely related homologs diverge in specificity, whereas more distantly related CSPs appear to have som e convergence in specificity. The chemical binding profiles described can now be used to form the basis for making predictions on the functional roles of some of these proteins


98 Figure 4 1 NPN Assay of SiCSPs NPN Assay of SiCSPs. SiCSPs were reacted at a concentration of 1 uM in 10 mM HEPES 50 mM NaCl pH 7.5 a gainst concentrations of NPN ranging from 0 uM to 20 uM NPN in 2% methanol. Curves were fit to the Hill equation to estimate a one site binding model K d s were determined from the fit and used in the competition assays that followed. Curves represent on e of three independent productions of each SiCSP. Error bars are of Standard error in one experiment run in triplicate The Scatchard transformation is inse t.


99 Figure 4 1 Continued

PAGE 100

100 Figure 4 1 Continued

PAGE 101

101 Figure 4 1 Continued

PAGE 102

102 Figure 4 1 Continued

PAGE 103

103 Figure 4 1 Continued

PAGE 104

104 Figure 4 1 Continued

PAGE 105

105 Figure 4 2 NPN Assay of SiCSPs NPN Assay of SiCSPs. SiCSPs were reacted at a concentration of 1 uM in 10 mM HEPES 50 mM NaCl pH 7.5 against concentrations of NPN r anging from 0 uM to 20 uM NPN in 2% methanol. Curves were fit to the Adair equation to estimate a two site binding model K d s were determined from the fit and used in the competition assays that followed. Curves represent one of three independent produc tions of each SiCSP. Error bars are of Standard error in one experiment run in triplicate The Scatchard transformation in inset

PAGE 106

106 Figure 4 2 Continued

PAGE 107

107 Figure 4 2 Continued

PAGE 108

108 Figure 4 2 Continued

PAGE 109

109 Figure 4 2 Continued

PAGE 110

110 Figure 4 2 Continued

PAGE 111

111 Figure 4 2 Continued

PAGE 112

112 Table 4 1 1 NPN binding affinities of fourteen SiCSPs. Below are the binding affinities (K d ) of fourteen SiCSPs with 1 NPN. Only one K d value presented for single site binding SiCSPs ( signified by an empty K d2 field entry) and two site binding SiCSPs values are presented in response to a cooperati ve binging Scatchard transformation Goodness fit are represented as R 2 values to the Hill (one site) or Adair (two site) fit s SiCSP K d 1 M) K d2 ( M) R 2 value SiCSP1 18.06 5.513 0.9846 SiCSP2 1.311 0.9 831 SiCSP3 15.97 8.731 0.9887 SiCSP4 3.587 0.9856 SiCSP5 11.83 4.055 0.9523 SiCSP6 4.881 14.39 0.9811 SiCSP7 1.746x10 12 7.212 0.9733 SiCSP8 2.590 0.9784 SiCSP9 10.51 11.77 0.9913 SiCSP10 0.7532 0.9821 SiCSP11 26.67 2.560 0.9902 SiCSP12 11.53 0 .9900 SiCSP13 1.114 0.9882 SiCSP14 1.133 0.9849

PAGE 113

113 Figure 4 3 Heat map of Apparent K d of SiCSPs Against Numerous Odorants. SiCSPs were primed with 20 uM NPN and then reacted in triplicate with odorants in concentrations ranging form 0.0001 uM to 500 uM. Resultant competition curves were used to solve for IC50. Data is representative of the solved apparent K d derived from IC50 data. SiCSPs with a or b substituents re present the first and second binding sites on their respective proteins. The hotter colors re present tight, specific binding. Groups of Ligands are as follows: Alcohols Blue, Aliphatic Aldehydes Red, Esters Green, Ketones Violet, Aromatic Compounds Orang e, Terpenes and Terpenoids Yellow, Sugars Solid Black, Carboxylic Acids Dashed Black, Alkanes Dotted Black. The s cale is shown on the previous page in M Insect Specific Ant Spec. S.i. Specific 1 S.i. Specific 2

PAGE 114

114 Figure 4 3 Continued Insect Specific Ant Spec. S.i. Specific 1 S.i. Specific 2

PAGE 115

115 CH APTER 5 PROBING THE SECONDARY STRUCTURE OF The S i CSPS Introduction A number of OBPs and CSP from other organisms have been crystallized and their three dimensional structures determined ( Lartigue et al. 2002 ; Campanacci et al. 2003 ; Kruse et al. 2003 ; Lartigue et al. 2003 ; Mosbah et al. 2003 ; Tegoni et al. 2004 ; Pelosi et al. 2006 ) The CSP isolated from the moth, Mamestra brassicae, was crystalized and revealed a structure consisting of two pairs of disulfide bonds and six helices. MbraCSPA6 as this CSP was named bares SiCSP12 as its closest hom olog from S. invicta Latrigue et al predicted that MbraCSPA6 would bind alkanes C 12 C 18 but of interesting note is that SiCSP12 tightly binds C 11 ( Lartigue et al. 2002 ) The PBP from D. melanogaster as well as the PBP from Leucophaea maderae both had a simila r six helical structure as predicted CSPs ( Kruse et al. 2003 ; Lartigue et al. 2003 ) Further more CSPsg4 isola ted from proved to be a helical in nature measured by CD polarimetry ( Picone et al. 2001 ; Ban et al. 2002 ; Tomaselli et al. 2006 ) Overall these data indicate conserved three dimensional structures that are helical in nature These algorithms can yield predictions of the protein structure including the binding pocket and can be used a models for theoretical docking experiments with (experimentally determined) ligands. However the any structur al changes resulting from alterin g the proteins conditions ( t emperature or pH) or the binding of ligands would not easily be seen. These data can serve as a guide, however, the physiological status and structure of the protein may not match predictions. Although X ray crystallography is a powerful tool to probe the structure of a protein, the procedure can be time co n suming and costly. Circular d ichroism (CD)

PAGE 116

116 p olarimetry affords a simple and convenient method that uses polarized light of two opposing spin s (considered left handed and right handed polarization) to assess protein secondary structure. The differential absorption of a spectrum of lights between left and right handed polarized light by a protein can reveal motifs in secondary structure. The ma in motifs seen in the CD are helices and sheets, each of which give s characteristic spectra (Figure 1 9) Though the results only elucidate the predominant structural motifs and not discreet 3D structure, CD can be done in solution and in real time. Furthermore experimental cond itions such as temperature, pH and ligand can be easily manipulated, and especially with regards to the latter, observing the change in the spectral reading will give insight into the interaction between the proteins and small molecules. It is well known that for some proteins, change s in structure and/or stability can occur upon binding of small molecule ligands A drastic conformational change in the structure of one OBP has been observed in the presence of its pheromone ligand ( Campanacci et al. 2003 ) By using CD, we can confirm that the SiCSPs share structural similarities with homologs as well as determine a way to quantify the thermodynamics of CSPs when bound to their ligand. Ligands may create a stabilizing or a destabilizing effect during interactions with target protein s Using CD, the effect of temperature can be examined while observing the output of the CD polarimeter at 222 nm, which represe specific to the presence of helical structure s (Figure 1 9) A temperature melt profile results when the protein solution is heated and the protein begins to unfold or lose structure which results in loss of the helix signal (at 222 nm) whi ch can be plotted as a function of the temperature. Typically, the readout forms

PAGE 117

117 a sigmoid curve and t he temperature at which the function is half maximal represents the midpoint melting temperature (T m ). In the presence of ligand, any shifts in T m would represent a quantifiable indicator to compare protein ligand interactions. By observing the shifts in T m in the CSPs in the presence and absence of ligand solutions we can look into the protein ligand interaction an d any structural/stability consequences that might occur when a ligand interact s with a protein providing an independent (to the ligand binding assays) means of testing specificity and nature of the ligands bound to a CSP Thus CD offers a means to observe structural changes upon binding the ligand, which can potentially reveal different interaction s between the CSP proteins and the different ligands that they may bind By assaying the secondary structure of these proteins we can gain insight in how the proteins bind a ligand and how specific ity may arise. During binding assays the apparent dissociation constants (K d ) were determined for fourteen of the twenty one CSPs found in the fire ant. In the experiments described here, secondary structure as temperature stability was monitored for each CSP using CD assays CSPs w ere incubated in the absence and presence of ligands at a concentration greater than their K d In addition, three dimensional structure predictions were made for each protein using the Phyre 2 ( html/page.cgi?id=index ) platform ( Kelley and Sternberg, 2009 )

PAGE 118

118 Materials and Methods Phyre 2 structural predictions The amino acid sequence of each SiCSP were inpu t into the Phyr e 2 e ngine ( ) and the resultant .pdb file was edited in Chimera ( ) to orient the proteins Protein and Ligand Solutions Proteins were purified and stored at 70 C until use as described in Chapter 2 Protein s olutions were thawed and adjusted ( diluted ) to a concentration of 0. 2 mg/mL (~ 15 M) in a volume of 1 mL in 10 mM HEPES 100 mM NaCl pH 7.5 buffer. For addition of ligands, controls contained 20 L of solvent (hexane) added to the protein mixture and samples consisted o f 20 L of ligand (dissolved in hexane) added to the protein rea ction volume In addition, CD spectra were recorded for HEPES buffer with solvent and HEPES buffer with solvent and ligands (no protein) as controls to determine if there was any significant absorption by the buffer /ligands Circular Dichroism CD analyses w ere performed using AVIV 400 CD polarimeter ( assistance kindly provided by Dr. Robert McKenna University of Florida). Sampling was performed in a 10 mm quartz cuvette with a 1 mm path length and 350 L of the sample was pipetted into the cuvette for each assay. Spectra were measured every nanometer between 200 and 260 nm with an averaging time of 3 seconds per reading. Data was recorded in ed using Dichroweb ( and fits were determined by using the K2D

PAGE 119

119 analysis program. Fits represented a smooth line estimation of secondary structure as determined by the CD data. For temperature stability experimen ts, a measured, the CD instrument was maintained at a single detecti on wavelength ( 222 nm ) and the temperature of the sample elevat ed from 25C to 100C within a 10 minute time interval Readings were made ever degree and averaged for 3 seconds each. All d ata were collected and graphed using GraphPad Prism 6.1. Data w ere fit to a Boltzman Sigmoidal regression from which the T M s w ere calculated Results Phyre 2 protein structure predictions The P hyr e 2 platform was used to predict the three dimensional structure of fourteen SiCSP s using their amino acid sequences The amino acid query is aligned against a database of close as well as remote homologs Using three programs (Psi Pred, SSPro and JNet) the Phyre 2 system to pr edict the secondary structure. This is then aligned with the top scoring alignments to predict the 3D structure ( Kelley and Sternberg, 2009 ) All fourteen SiCSPs had predicted st ructures containing six helices and typically unstructured termini All structures had 100% confidence as reported by Phyre 2 (Table 5 2) The two closes t homologous CSPs that have characterized 3D structures were from the cabbage moth ( Mamestra brassicae MbraCSPA6) and the de sert locust ( Schistocerca gregaria SgreCSP4) ( Mosbah et al. 2003 ; Tomaselli et al. 2006 ) Their coverage ranged from 75 to 88 amino acids of about 110 ini tial amino acids and had percent identities between 18 and 51 (Table 5 1).

PAGE 120

120 CD polarimetry CD spectra were obtained for fourteen SiCSPs at room temperature (22 C). SiCSPs 1, 2, 3, 5, 6, 7, 8, 9, 12, 13, 1 corresponding helical signals from ~220 230 nm (inverted peak ~222 225), although for a number of proteins (SiCSPs 2, 5, 6, 7, 9, and 12) the response was somewhat shallow (Figures 5 2, 5 5, 5 6, 5 7, 5 9, 5 12 ). Atypical spectra were seen for SiCSPs 4, 10, and 11 (Fig ure s 5 4 5 10 and 5 11 ). For SiCSP4, a small trough can be seen, however, the signal rapidly rises from ~225 nm onwards. For both SiCSPs 10 and 11, although the helical trough can been (~215 nm) and dip (~ 210 nm) are absent. The thermal stability of the CSPs was determined by monitoring the signal at 222 nm as a function of increased temperature. In these experiments, SiCSP1 showed a chara cteristic melting profile with a calculated T m of 72.25 C (Fig ure 5 1). Addition of cinnamald ehyde ( 15 M) or benzoic acid ( 50 M) resulted in distorted profiles and loss of the characteristic shape including the helix signal for the protein, resulting in thermal melt profiles that could not be quantified. Similar anomalous effects we re seen when most ligands were added to a CSP. Addition of cinnamald ehyde or oleic acid to SiCSP2 resulted in distorted CD spectra ( Figure 5 2 ). For SiCSP3, however, the CD spectrum, although altered, retained similar curve pattern as without the ligand ( Figure 5 3 ). In this case, T M s with and without ligand could be calculated, with a 2.2 C increase in thermal stability seen in the presence of ligand. For SiCSP4, addition of various ligands produced different alte rations in the CD spectrum (Fig ure 5 4). Addition of cinnamaldehyde to SiCSP5 resulted in some distortion of the C D spectrum,

PAGE 121

121 although measurements at 222 nm appeared to be stable and reflect the state of the protein and a small de stabilizing shift in T M was observed (84.9 to 83.1 C; no lig and, ligand, r espectively) (Fig ur e 5 5). For SiCSP6, the T M (in the absence of ligand) was high, 93.4 C (Fig ure 5 6). In this case, addition of various ligands had significantly different effects on the T M of the protein. Of four ligands that showed low apparent K d (i .e high affinity) values from the 1 N PN competition experiments, Z 3 h e xe n o l, 2 octanol, and DEET, resulted in stabilization of the protein by ~ 1, 3.1, and 6.1 C, respectively. In contrast, addition of DEET to the protein resulted in ~7.9 C decrease in t he T M Both SiCSP7 and SiCSP8 maintained CD spectra indicative of proteins helical in structure with and without ligands (Fig ure 5 7 and 5 8). Indeed, unlike for most of the CSPs examined, additional of amylcinnamaldehyde to SiCSP7 did not appreciably change the CD spectrum, whereas for SiCSP8, lower signals (upward shift) was noted, however, the shape of the curve was unchanged. A small shift towards stabilization was seen for SiCSP7 in the presence of ligand (77.7 to 78.8 C), whereas all three ligan ds ( amylcinnamaldehyde, oleic acid, and palmitic acid) tested with SiCSP8 resulted in decreased T M values (~1 3 C). In contrast, despite giving a reasonable CD spectrum in the absence of any ligand, signals from SiCSP9 in the presence of amylcinnamaldehyd e showed such large variation that the data are difficult to interpret ( Figure 5 9). Similarly, a T M could be calculated for SiCSP10 in the absence of ligand (71.6 C), but addition of any of the tested ligands resulted in large variations in the spectra ( Fig ure 5 10). For SiCSP11, no T M could be derived in the abs ence or presence of ligand (Fig ure 5 11). Addition of ligands to SiCSP12 resulted in anomalous CD

PAGE 122

122 spectra, however, the signal (at 222 nm) could still be followed and showed destabilization of str ucture in the presence of cinnamaldehyde (T M from 79.9 to 77.4 C) but stabilization in the presence of oleic acid (T M from 79.9 to 83.4 C). Similar to SiCSP10, addition of ligand to SiCSP13 or SiCSP14 resulted in high variability in the data such that T M s could not be calculated, although in the absence of ligand a T M of 73.5 C and 70.5 C was determined for SiCSP13 and SiCSP14, respectively. Discussion All fourteen SiCSPs examined were successfully modeled using Phyre 2 for 3 D predictions and all ret urned a 100% confidence value to previously documented CSPs These analyses indicate that despite low amino acid homology between the CSPs (documented by the range of coverage and the lower percent identity) a significant (predicted) structure is shared b etween the proteins. Other than secondary structure similarities, the class of CSPs contains two disulfide bonds (oxidized cysteines) with conserved pairings These data also indicated the essential helicity of the proteins. Homologous CSPs imply that S binding pocket could contai n one or more tryptophan These residues could potentially be used to examine intrinsic fluorescent signals from the proteins. SiCSPs 4, 10, and 11, gave distorted CD spectra even in the absence of any ligands. The re ason( s) for these observations is unclear and no distinct pattern (ligand or phylogenetic) could be seen Addition of ligands had differential effects on the proteins. Data for SiCSPs 1 2, 9, 10, 11, 13, and 14 showed large variation and were not interpretab le. Addition of ligands to SiCSPs 3 and 7, had stabilizing effects (increase in T M ), whereas addition of ligands to SiCSPs 5 and 8, had destabilizing effects (decrease in T M ). For SiCSP4 (despite resulting in an anomalous CD spectrum in the absence of liga nd, the signal at

PAGE 123

123 222 nm could still be followed), only minor effects were seen in the T M in the presence of ligands. In contrast, for SiCSPs 6 and 12 both stabilizing and destabilizing effects were seen depending upon the ligand tested. These data indicat e at least one important conclusion, namely that CSPs may have dynamic flexibility and respond differently to different ligands. The observation that different ligands may stabilizing or destabilize CSP structure could be important for downstream interacti ons (presumably to the membrane receptor) that the proteins interact with. Another possibi lity is that upon binding a high concentration of ligand, some CSPs may undergo polymerization or agglutination Although speculative, it is possible that the effe ct on structure (as perceived by changes in the T M ) may reflect conformational states needed for CSPs to interact with different partner. Future work identifying downstream partners is needed. T he stabilization/ destabiliz ation effects seen for the protein s can likely be attributed to the change in helic al structures since these appear to predominate CSPs appear to have an outer, hydrophilic shell, and an inner hydrophobic pocket (similar to most such protein) Keeping the hydrophobic pocket internalized in the protein allows for high solubility of the proteins a nd protection of a hydrophobic core binding pocket needed to accommodate the hydrophobic molecules detected by these proteins Destabilization and loss of helicity may be evidence of a conformatio nal change that disrupts the hydrophobic pocket to allow flexibility in mediating binding and ret ention but possibly also eventual release of the ligand. For CSPs destabilized in the presence of ligand, one model to account for the data is that t hey rema in stable and available to bind ligands near pores or close to the air lymph (aqueous) interface in the sensillum After binding of the ligand, the protein

PAGE 124

124 undergoes a conformational change that allows the CSP to target to the neuronal cell membrane to int eract with the odorant receptor C onverse ly some CPSs may work in the opposite direction to release and transport ligand s to the surface for dispersal either to other internal structures or to the environment/cuticle. Alternately, some CPSs may function to dispose of ligand s by targeting them for degradation as part of attenuation circuits Although at this stage, there has yet to be a demonstration of the specificity of a CSP for any particular odorant receptor, this mechanism can potentially explain th e role of some CSPs as being chaperones for hydrophobic odorants. Future work examining additional ligands as well as different concentration of ligands and/or effects of salt, pH, and other factors is needed.

PAGE 125

125 Figure 5 1 SiC SP1 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were d etermined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 126

126 Figure 5 1 Continued. B C

PAGE 127

127 Figure 5 2 SiCSP2 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuri ng at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 128

128 Figure 5 2 Continued B C

PAGE 129

129 Figure 5 3 SiCSP3 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 130

130 Figure 5 3 Continued B C

PAGE 131

131 Figure 5 4 SiCSP4 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm t o 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A B

PAGE 132

132 Figure 5 4 Continued B C

PAGE 133

133 Figure 5 5 SiCSP5 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars A

PAGE 134

134 Figure 5 5 Continued B C

PAGE 135

135 Figure 5 6 SiCSP6 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 136

136 Figure 5 6 Continued B C

PAGE 137

137 Figure 5 7 SiCSP7 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 24 0 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by determined by a sigmoid curve fit (C). CD was p erformed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 138

138 Figure 5 7 Continued B C

PAGE 139

139 Figure 5 8 SiCSP8 Secondary Structure. The Phyre 2.0 algorithm was employed to pred ict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean repo rted as error bars) A

PAGE 140

140 Figure 5 8 Continued B C

PAGE 141

141 Figure 5 9 SiCSP9 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein ( A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 142

142 Figure 5 9 Continued B C

PAGE 143

143 Figure 5 10 SiCSP10 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm t o 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 144

144 Figure 5 10 Continued B C

PAGE 145

145 Figure 5 11 SiCSP11 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 n m to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 146

146 Figure 5 11 Continued B C

PAGE 147

147 Figure 5 12 SiCSP12 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars) A

PAGE 148

148 Figure 5 1 2 Continued B C

PAGE 149

149 Figure 5 13 SiCSP13 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. (Standard error mean reported as error bars)

PAGE 150

150 Figure 5 13 Continued B C

PAGE 151

151 Figure 5 14 SiCSP14 Secondary Structure. The Phyre 2.0 algorithm was employed to predict the three dimensional structure of the SiCSP using the amino acid sequence of the recombinant protein (A). CD Spectra observed from 200 nm to 240 nm, and the predicted curves were determined using Dichroweb using a K2D algorithm (B). Midpoint melting temperate as determined by measuring at 222 nm between a temperature range of 25 C to 100 C was determined by a sigmoid curve fit (C). CD was performed using 0.2 mg/mL protein and represents one of three independent assays. Ligands chosen were ones shown to bind nominally with their res pective SiCSP during ligand binding assays. Each ligand was reacted at their IC50 as previously determined (Not shown). (Standard error mean reported as error bars) A

PAGE 152

152 Figure 5 1 4 Continued B C

PAGE 153

153 Table 5 1 Overview of SiCSP secondary structure a nalysi s. The midpoint melting temperatures of 14 of the SiCSPs were determined by CD. Each SiCSP was additionally reacted with ligands at a concentration equal to their IC50 to observe any changes in the thermodynamic properties of the SiCSPs. Ligands were de termined to be nominal binders from the ligand binding studies. T M T amyl Cinnamaldehyde T M DEET T M Oleic a cid T M Palmitic a cid T M Nonanol T M 2 O ctanol T M Z 3 Hexenol T M Nonyl a cetate T M Tritriacontane (C33) SiCSP1 72.25 0 .88 268.7 SiCSP2 80.92 2.59 79.80 80.53 1.3 SiCSP3 67.45 1.28 69.57 2.28 SiCSP4 77.95 0 .35 77.31 .41 76.33 0 .28 76.66 0 .35 76.69 0 .30 SiCSP5 84.93 1.62 83.10 1.40 SiCSP6 93.41 16.0 8 99.52 13.64 94.46 19.45 94.42 8.71 85.52 13.88 SiCSP7 77.68 0 .65 78.76 0 .56 SiCSP8 74.64 0 .36 73.72 0 .27 71.76 0 .29 73.70 0 .28 SiCSP9 61.20 4.13 61.16 2.96 SiCSP10 71.64 0 .73 SICSP11 SiCSP12 79.87 1.81 77.37 3.74 83.42 2.84 SiCSP13 73.52 1.23 SiCSP14 70.46 0 .36

PAGE 154

154 Table 5 2 Phyre 2 Predictions. The recomb inant SiCSP amino acid seq uences were inputted to the Phyre 2 engine for the prediction of the 3D structure. Reference homolog indicates the closes t related homolog from the Phyre 2 search of the SiCSP using the amino acid sequence. % I dentity and % coverage are of the amino acid s equences betwee n the reference homolog and the SiCSP Confidence values are represented in percent reflecting the repeatability of the 3D model. MbraCSPA6 is a CSP isolated from Mamestra brassica e (cabbage moth) and SgreCSP4 from Schistocerca gregaria ( desert locust). SiCSP Reference Homolog % Identity % Coverage Confidence SiCSP1 MbraCSPA6 20 100 SiCSP2 MbraCSPA6 26 100 SiCSP3 MbraCSPA6 30 80 100 SiCSP4 MbraCSPA6 30 82 100 SiCSP5 MbraCSPA6 23 83 100 SiCSP6 MbraCSPA6 36 76 100 SiCSP7 Mbra CSPA6 50 77 100 SiCSP8 MbraCSPA6 31 83 100 SiCSP9 MbraCSPA6 18 88 100 SiCSP10 MbraCSPA6 18 88 100 SiCSP11 MbraCSPA6 44 81 100 SiCSP12 MbraCSPA6 51 75 100 SiCSP13 MbraCSPA6 31 82 100 SiCSP14 SgreCSP4 38 84 100

PAGE 155

155 CHAPTER 6 CONCLUSIONS Our central h ypothesis is that CSPs differ in expression pattern and specificity for ligands. In order to test this hypothesis, (1) quantitative RT PCR was used to examine the gene expression profiles of the twenty one identified S. invicta CSPs in various ant castes a nd developmental stages, and (2) fou rteen SiCSPs were expressed their ligand binding specificities determined, and aspects of their structure were characterized. The results presented here indicate a complex picture with respect to gene expression, ligan d binding, and structural features of the CSPs examined. A number of models concerning the functions of CSPs were examined, and the critical questions addressed in this study include; (i) are CSPs general binding proteins that share similar binding profile s but differ in expression, i.e. is the main level of biologically relevant control on the gene expression level with CSPs essentially interchangeable in function? The alternative extreme to this model is that CSPs are each unique in their ligand binding p rofiles and essentially constitutively expressed (or antennal specific) in the ant. (ii) Does phylogeny predict the ligand binding range and/or expression pattern of the CSPs? (iii) Does ligand binding affect protein structure? Our data indicate that both ligand binding differences and gene expression variation are important parameters that are likely to ultimately define CSP functions. A wide dynamic range in gene expression profiles, ligand binding, and structural aspects was seen, although a number of pa tterns emerged. First, neither CSP expression nor ligand binding profiles could be predicted (or correlated) with their phylogenetic distribution, suggesting that gene duplication events have subsequently led to diversification in function. Indeed, SiCSPs 1 and 5 (closely related in a fire ant specific

PAGE 156

156 CSPs), all show very similar binding profiles; binding a very diverse range of ligands from alkanes to alcohols, aliph atic aldehydes ketones, esters, and aromatics (although potentially important differences in specific chemical tested was noticed). However, these CSPs show different expression patterns. SiCSP1 was highly expressed in the antennae of workers (from where it was first isolated) and males (but not female alate s, In contrast SiCSP5 was found predominately in adult abdomens, SiCSP6 in the heads of female alate s (but not anten nae), SiCSP9 in various developmental stages but low in adults, and SiCSP11 found in antennae, abdomen of males, and in early and late instar larvae as well as pupae. These data suggest that this group of CSPs may mediate general binding to a wide range of compounds with biological function determined by the tissue/developmental distribution. In contrast to the group described above, SiCSPs 2 and 10, are closely related phylogenetically, belong to a fire ant specific clade, but have narrow and different li gand binding specificities, although they have similar gene expression/tissue distribution profiles. Thus, for these two proteins, their narrow and specific binding ranges may indicate mediation of different chemical signals in the various tissues (includi ng antennae) in which they are expressed. An important caveat to this discussion is that we did not measure actual protein levels in the various tissues, and that transcript levels may not fully translate to protein levels and/or that CSPs may be expressed and translated into proteins in one tissue but secreted/transported to others within the organism. Within this context, it should ne noted that all CSPs contain protein secretion

PAGE 157

157 signal sequences (with respect to the antennae, presumably to localize to th e sensillar lymph). SiCSPs 8, 13, and 14, part of a fire ant specific CSP clade, showed narrow binding ranges to different compounds, with limited overlap (SiCSPs 8 and 14 both appear to bind shorter chain alkanes ( C 9 C 15 C 11 is of note since it is an ale rt signal in some ants), but also showed some differential tissue distribution. Other CSPs showed partially overlapping but discrete ligand binding and gene expression profiles. SiCSPs 3 and 7, the former within a fire ant specific clade, and the latter pa rt of the greater insect CSP group, displayed relatively large binding ranges but differed in a number of chemical compounds. In particular, SiCSP3 did not bind mid range shorter chain alkanes (C 13 19 ) whereas SiCSP7 did not bind the longer chain alkanes ( C 26 onwards). There expression patterns in adults (workers, male and female alate s) were similar, however, significant differences were seen in the developmental stages. SiCSP3 was not expressed to an appreciable extent in eggs or early/late instar larvae, and moderately in pupae. In contrast, SiCSP7 was highly expressed in early/late instar larvae and pupae (not in eggs). These data suggest an important (added?) role for SiCSP7 in development. It should be noted that SiCSPs proved difficult to standardiz e as none of the genes used for normalizing was consistent across the various tissues and developmental stages examined. However, most CSPs were produced in adults worker and male antennae. Intriguingly female alate s showed poor CSP expressed but also ver and GAPDH. Only SiCSP8 (and to a lesser extent SiCSP9 and possibly 20) were detected in these samples. These data suggest that female alate antennae appear to be transcriptionally quiescent,

PAGE 158

15 8 implying li mited response/utility of the antennae in this caste. A possible explanation is that female alate s, as potential future queens, do not require significant chemical perception as workers and other castes. Experiments to confirm the quiescent nature of femal e alate antennae via electroantenography may help confirm the biological implications of these results. SiCSP1, the first CSP characterized in fire ant (from worker antennae) was also expressed in the abdomen of female alate s, though to a much lesser degre e. Highly expressed antennal CSPs included, SiCSP1 SiCSP3, SiCSP8, SiCSP12, SiCSP17, and SiCSP21. Interestingly, the majority of these CSPs highly belong to one of the S. invicta specific groups. These may play sensory as well as other roles since SiCS Ps 3, 8, and 21 in particular are also found else where in the body. The observed tissue/developmental diversification of CSP expression raises several issues. First, it should be noted that transcript levels may not correlate to protein, and further that where the protein is made may not correlate to final localization especially since these proteins have secretion signal sequences and thus are exported. Export, however, may not be to the antennae but to any number of glands, tissues, and/or even the hemol ymph where the CSPs may mediate binding/transport like duties. However, assuming that the gene expression results do provide a rough approximation of tissue distribution, the question arises as to whether CSPs perform different functions in the different t issues that they are expressed. This could easily be the case for those CSPs that show broad ligand binding, such that depending upon the tissue/developmental stage they would mediate chemical detection or transport of different chemicals. Thus, for some C SPs, their role might be contextual, possibly acting as general ligand

PAGE 159

159 chaperone proteins. For CSPs with narrow substrate specificities, the implication is that those compounds bound must also be present in the various tissues in which the CSP is expresse d. With respect to chemical perception, the sensillum lymph is a heterogenous mixture of proteins and cations Many of the CSPs could act as general purpose binding proteins in order to increase the solubility of perceived chemicals (odorants) to facilit ate the odorant to odorant receptor interactions. Some SiCSPs may play more dedicated roles in the antennae, and aside from acting as chemical carriers; CSPs may function to eliminate odorants in processes that results in olfactory attenuation and/or habi tuation. Our results do not provide evidence for such roles, but does provide clues and candidates for future work. In particular, it would be interesting to examine downstream interacting partners with the CSPs. A good place to start would be to see if an y CSPs interact with each other. Gustatory receptors are associated with the mouthparts of the insects, so CSPs located in the head may have a role in mediating chemical detected during tasting. This may include food, but also behavioral aspects such as gr ooming and pathogen removal. In addition to the sense of taste, the fire ant also has glands that produce signals that can trigger alarm and defense response. SiCSPs 3, 6, 17, the former two with broad specificities were highly expressed in female alate heads (recall that female alate antennae were poor in CSP expression). Furthermore, overall, female alate s expressed a wide range of CSPs. Although quite speculative, it is possible that for these future queens the senses localized to the head (maxillary palps) play an important determinant role in successful colony founding. For example, for the first few clutches of larvae, the

PAGE 160

160 queen feeds the larvae herself to nurse the nascent workers. These CSPs may play a role in interactions with the queen or for f eeding larvae before the colony matures. Workers expressed a range of CSPs but not as much as seen in their antennae, alate s, in contrast, showed poor CSP (and reference gene) expression, with SiCSP17 p redominating. The substrate specificity of SiCSP17 has yet to be determined, however, there appears to be a clear reduced reliance on CSPs in the head region for male alate fire ants. SiCSP3 has a very broad substrate specificity, is rather widely expresse d, and as suggested above may play different roles in different tissues. SiCSP19 and SiCSP21 are notably expressed in the thorax of adults (with the exception of SiCSP19 which is absent in the female alate thorax). The specificities of these two CSPs is u nknown. Functions of glands in the thoracic region may include aiding in the release of antimicrobial agents from the metapleural gland or salivary components from the labial gland. It is unclear what role CSPs would play in either of these glands however contact sensilla have been found on the tarsi and legs of ants (REF). It is possible that these SiCSPs may play a role in chemical contact chemoreception. The abdominal region was found to be rich in SiCSP expression (although somewhat lower in males). or the poison gland might be responsible for this. Although the trail pheromone has been identified from the Dufours gland, it is likely this is not the only pheromone released from the hind sec tion of the ants. These CSPs may play a role as general holding and release factors. Because of the broad range of SiCSPs expressed in these

PAGE 161

161 tissues, it is difficult to identify potential functions. As mentioned, it is also possible that the fire ant pr oduces SiCSP transcript (and potentially translated polypeptides) in the abdomen before exporting a particular CSP elsewhere in the body. Regarding m ale alate s expression of CSPs in non sensory organs, especially the abdomen, may allude to potential roles release of odorants for mating and kin identification. Indeed, some CSPs have been initially isolated and identified as ejaculatory bulb proteins in Drosophila sp (Dyanov et al ., GenBank Accession No. U08281) via proteomic approaches, although their role (s) in these structures remained unknown. Ligand Binding Specificity One S. invicta specific CSP clade (SiCSPs 8, 13, 14, 15, and 21), appeared to consist of single site binding model proteins, with narrow specificity (data acquired for SiCSPs 8, 13, and 14, and predicted based upon these results for 15 and 21). All other SiCSP groups contained proteins that modeled for both one and two site models. Within the other S. invicta specific clade, comprising of SiCSPs 1,2,3,5, 10, 18, and 20, CSPs 2 and 10 (mos t closely related to each other) were single binding site proteins, whereas the rest modeled best to two site binding. From the phylogenetic tree it would appear as if the ancestral protein to this group would have had two binding sites and been of broad s ubstrate specificity and that SiCSPs 2 and 10 have narrowed their specificity (and ligand binding pocket). Others are currently testing this prediction in the lab. In general, cooperative two site binding of ligands correlated to broad binding, whereas th e single site proteins displayed narrower substrate specificities. The two site binding CSPs appear to have a sensitive but non specific initial binding site followed by a second binding site that shows lower level affinity to bind a second ligand molecule

PAGE 162

162 (sometimes more than 5 orders of magnitude lower affinity). It is possible that the second binding site may add sensitivity or context to ligand binding. These CSPs could also bind two different ligands and modulate chemoreception in the context of the two ligands present. Future work examining this issue is warranted There was no clear correlation between expression (location) and the number of putative binding sites. Thus, these proteins likely evolved to have a wide range of binding to fill multip le roles in ligand binding for the insect. Although the majority of substrates tested were hydrophobic, several CSPs (5, 6, 11, and 9) could bind hydrophilic ligands (sugars, lactic acid or benzoic acid). However, these proteins, with the exception of SiC SP6 found in the heads of female alate s, were not found in the head/mouthparts. Binding to trehalose, the major carbohydrate constituent of insect hemolymph, however, may imply a role in the hemolymph for some of these CSPs. Thus, our data suggest that in itial consideration that CSPs are mainly or exclusively involved in chemoreception is inaccurate, and that CSPs have a potential for more than just allowing the insect to sense the environment but can impact other aspects of insect communication, feeding, and sensation, and may mediate a range of other physiological processes. Protein Secondary Structure Secondary structural analysis via CD polarimetry was performed for all the SiCSPs (with a n exception to SiCSP11) and indicated the a helical nature of the CSPs largely as predicted from the primary amino acid sequence. Thermal melting profiles further showed that t he SiCSPs proved to be quite thermostable proteins, some not reaching full denaturation even at 100 C. A number of CSPs displayed anomalous CD spectra, and currently the reason for this is unknown. It may reflect significant flexibility of the proteins or else the presence of oligomers and/or aggregates that might have

PAGE 163

163 obscured the readings. Examining the proteins via other techniques such as lig ht scattering is warranted. Addition of ligands produced a wide range of structural re s ponses. For some CSPs, presence of ligand appeared to stabilize the protein whereas for other a destabilization effect was noted. In some cases, some ligands stabilized the protein whereas other ligands appeared to destabilize the protein. Such effects may indicate differing protein conformations with respect to different ligands. This could in turn, affect the functions of the CSP. If a CSP assumes different conformation s upon binding to specific ligands this could provide for a mechanism for interacting with different downstream partners, i.e. different odorant receptors, thus addressing the question of how a general binding protein could provide specificity in signaling Other possibilities may have to do with ligand release; ligands that destabilize the protein structure may allow for easier release in the presence of some other signal. For some CSPs addition of ligand resulted in CD spectra that could not be interprete d. This could reflect intrinsic dynamics of the proteins, oligomerization/aggregation effects, or issue dealing with the ligand tested. Regarding the latter, it should be noted that only one concentration of ligand was tested for most proteins, thus perfor ming a titration could yield better data. All of the SiCSPs shared very similar protein three dimensional structure predictions using Phyre 2 The SiCSPs contained six helices, with rather unstructured C and N termini. Future work solving the crystal struc tures of these proteins to confirm and expand upon these findings is needed.

PAGE 164

164 LIST OF REFERENCES Abuin, L., B. Bargeton, M.H. Ulbrich, E.Y. Isacoff, S. Kellenberger and R. Benton, 2011. Funct ional architecture of olfactory ionotropic glutamate receptors. Neuron, 69(1): 44 60. Available from DOI 10.1016/j.neuron.2010.11.042. ACAAI, 2010. Fi re ant sting allergy. sting allergies/Pages/fire ant stings allergy.aspx Adams, C.T., W.A. Banks, C.S. Lofgren, B.J. Smittle and D.P. Harlan, 1983. Impact of the red imported fire ant, solenopsis invicta (hymenoptera, formicidae), on the growth and yield of soybeans. Journal of Economic Entomology, 76(5): 1129 1132. Available from ://A1983RV56700035. Allen, C.R., S. Demarais and R.S. Lutz, 1994. Red imported fire ant impact on wildlife an overview. Texas Journal of Science, 46(1): 51 59. Available from ://A1994NB78100008. Anderson, J.T., H. Thorvilson and S. Russel, 2002. Landscape materials as repellents of red imported fire ants. Southwest Entomol, 27(2): 156 163. Araujo, M.B. and W.R. Tschinkel, 2010. Worker allometry in relation to colony size and social form in the fi re ant solenopsis invicta. Journal of Insect Science, 10(94): 1 10. DOI 10.1371/journal.pone.0057718.t001. Ban, L., A. Scaloni, A. Brandazza, S. Angeli, L. Zhang, Y. Yan and P. Pelosi, 2003. Chemosensory proteins of locusta migratoria. Insect molecular bi ology, 12(2): 125 134. Available from Ban, L., L. Zhang, Y. Yan and P. Pelosi, 2002. Binding properties of a locust's chemosensory protein. Biochemica l and biophysical research communications, 293(1): 50 54. Available from DOI 10.1016/S0006 291X(02)00185 7. Benckiser, G., 2010. Ants and sustainable agriculture. A review. Agronomy for Sustainable Development, 30(2): 191 199. Available from ://000276769800001. Benton, R., S. Sachse, S.W. Michnick and L.B. Vosshall, 2006. Atypical membrane topology and heteromeric function of drosophila odor ant receptors in vivo. PLoS biology, 4(2): e20. Available from DOI 10.1371/journal.pbio.0040020. Benton, R., K.S. Vannice, C. Gomez Diaz and L.B. Voss hall, 2009. Variant ionotropic glutamate receptors as chemosensory receptors in drosophila. Cell, 136(1): 149

PAGE 165

165 162. Available from DOI 10.1016/j.cell.20 08.12.001. Bextine, B.R. and H.G. Thorvilson, 2002. Field applications of bait formulated beauveria bassiana alginate pellets for biological control of the red imported fire ant (hymenoptera : Formicidae). Environmental Entomology, 31(4): 746 752. Availab le from ://000177610500022. Bextine, B.R. and H.G. Thorvilson, 2002. Monitoring solenopsis invicta (hymenoptera : Formicidae) foraging with peanut oil baited, uv reflective beauveria bassiana alginate pellets. Southwest Entomol, 27(1): 31 36. A vailable from ://000176309400004. Bigley, W.S. and S.B. Vinson, 1975. Isolation of a brood pheromone isolated from the sexual brood of the fire ant, solenopsis invicta Annals of the Entomological Society of America, 68: 301 304. Billen, J. and E.D. Morgan, 1998. Pheromone communication in social insects: Sources and secretions. In: Pheromone communication in social insects. WestviewPress. Bordbar, A.K., A.A. Saboury and A.A. Moosavi Movahedi, 1996. The shapes of scatchard plots for systems wit h two sets of binding sites. Biochemical Education, 24(3): 172 175. Buhs, J.B., 2002. The fire ant wars nature and science in the pesticide controversies of the late twentieth century. Isis, 93(3): 377 400. Available from ://000179866000001. Campanacci, V., A. Lartigue, B.M. Hallberg, T.A. Jones, M.T. Giudici Orticoni, M. Tegoni and C. Cambillau, 2003. Moth chemosensory protein exhibits drastic conformational changes and cooperativity on ligand binding. Proceedings of the National Academy of Sciences of the United States of America, 100(9): 5069 5074. Available from DOI 10.1073/pnas.0836654100. Carey, A.F. and J.R. Carlson, 2011. Insect ol faction from model systems to disease control. Proceedings of the National Academy of Sciences of the United States of America, 108(32): 12987 12995. Available from ed/21746926 DOI 10.1073/pnas.1103472108. Cheng, D., Z. Zhang, Z. He and G. Liang, 2013. Validation of reference genes in solenopsis invicta in different developmental stages, castes and tissues. Plos One, 8(2): 1 6. DOI 10.1371/journal.pone.0057718.t001 Clarke, D.T., 2011. Circular dichroism and its use in protein folding studies. Methods Mol Biol, 752: 59 72. Available from

PAGE 166

166 DOI 10.1007/978 1 60327 223 0_5. Copelans, R.A., D.L. Pompliano and T.D. Meek, 2006. Drug target residence time and its implications for lead optimization. Nature, 5: 730 739. Daifeng Cheng, Z.Z., Xiaofang He, and G.L. 2013. Validation of reference genes in solenopsis invic ta in different developmental stages, castes and tissues. PLoS ONE, 8(2). DOI 10.1371/journal.pone.0057718.t001. de Brito Sanchez, M.G., E. Lorenzo, S. Su, F. Liu, Y. Zhan and M. Giurfa, 2014. The tarsal taste of honey bees: Behavioral and electrophysiolo gical analyses. Frontiers in behavioral neuroscience, 8: 25. Available from DOI 10.3389/fnbeh.2014.00025. Dietrich Gotzek1*.¤, H.M.R., Yannick Wurm1, DeWayne Shoemaker3, 2011. Odorant binding proteins of the red imported fire ant, solenopsis invicta: An example of the problems facing the analysis of widely divergent proteins. PLoS ONE, 6(1): 1 9. DOI 10.1371/journal.pone.0016289.t001. Evans, J.D. and D Gundersen Rindal, 2003. Beenomes to bombyx: Future directions in applied insect genomics Genome biology, 4(3): 1 4. Getahun, M.N., S.B. Olsson, S. Lavista Llanos, B.S. Hansson and D. Wicher, 2013. Insect odorant response sensitivity is tuned by metabotr opically autoregulated olfactory receptors. Plos One, 8(3): e58889. Available from DOI 10.1371/journal.pone.0058889. Gong, Y. and E. Plettner, 2011. E ffects of aromatic compounds on antennal responses and on the pheromone binding proteins of the gypsy moth (lymantria dispar). Chemical senses, 36(3): 291 300. Available from DOI 10.1093/chemse/bjq130. Gonzalez, D., Q. Zhao, C. McMahan, D. Velasquez, W.E. Haskins, V. Sponsel, A. Cassill and R. Renthal, 2009. The major antennal chemosensory protein of red imported fire ant workers. Insect molecular biolo gy, 18(3): 395 404. Available from DOI 10.1111/j.1365 2583.2009.00883.x. Gotzek, D., H.M. Robertson, Y. Wurm and D. Shoemaker, 2011. Odorant binding p roteins of the red imported fire ant, solenopsis invicta: An example of the problems facing the analysis of widely divergent proteins. Plos One, 6(1): e16289. Available from http://www.ncbi.nlm.nih .gov/pubmed/21305009 DOI 10.1371/journal.pone.0016289.

PAGE 167

167 Greenfield, N.J., 2006. Using circular dichroism spectra to estimate protein secondary structure. Nature protocols, 1(6): 2876 2890. Available from DOI 10.1038/nprot.2006.202. Grosse Wilde, E., A. Svatos and J. Krieger, 2006. A pheromone binding protein mediates the bombykol induced activation of a pheromone receptor in vitro. Chemical senses, 31 (6): 547 555. Available from DOI 10.1093/chemse/bjj059. Gu, S.H., W.X. Wang, G.R. Wang, X.Y. Zhang, Y.Y. Guo, Z. Zhang, J.J. Zhou and Y.J. Zhang, 2011 Functional characterization and immunolocalization of odorant binding protein 1 in the lucerne plant bug, adelphocoris lineolatus (goeze). Arch Insect Biochem Physiol, 77(2): 81 99. Available from DOI 10.1002/arch.20427. Harada, E., D. Haba, T. Aigaki and T. Matsuo, 2008. Behavioral analyses of mutants for two odorant binding protein genes, obp57d and obp 57e in drosophila melanogaster Genes Gentic S ystems(83): 257 264. Hedges, 1997. Fire ants. Handbook of Pest Control, 8th Edition( ): 531 535. Holldobler, B., 1995. The chemistry of social regulation multicomponent signals in ant societies. P Natl Acad Sci USA, 92(1): 19 22. Available from ://A1995QB23800006. DOI Doi 10.1073/Pnas.92.1.19. Holldobler, B. and A.J. Wilson, 1990. The ants. Harvard University Press, Cambridge, Massachusetts. Hulme, E.C. and M.A. Trevethick, 2010. Ligand binding assays at equilibrium: Validation and interpre tation. British Journal of Pharmacology, 161: 1219 1237. DOI 10.1111/j.1476 5381.2009.00604.x 10.1111/bph.2010.161.issue 6. Jiao, Y., S.J. Moon and C. Montell, 2007. A drosophila gustatory receptor required for the responses to sucrose, glucose, and malto se identified by mrna tagging. Proceedings of the National Academy of Sciences of the United States of America, 104(35): 14110 14115. Available from DO I 10.1073/pnas.0702421104. Jungblut, P.R., H.G. Holzhutter, R. Apweiler and H. Schluter, 2008. The speciation of the proteome. Chemistry Central journal, 2: 16. Available from http://www.ncbi.nlm. DOI 10.1186/1752 153X 2 16. Kaupp, U.B., 2010. Olfactory signalling in vertebrates and insects: Differences and commonalities. Nature reviews. Neuroscience, 11(3): 188 200. Available from DOI 10.1038/nrn2789.

PAGE 168

168 Kelley, L.A. and M.J. Sternberg, 2009. Protein structure prediction on the web: A case study using the phyre server. Nature protocols, 4(3): 363 371. Available from DOI 10.1038/nprot.2009.2. Krieger, M.J. and K.G. Ross, 2005. Molecular evolutionary analyses of the odorant binding protein gene gp 9 in fire ants and othe r solenopsis species. Molecular biology and evolution, 22(10): 2090 2103. Available from DOI 10.1093/molbev/msi203. Kruse, S.W., R. Zhao, D.P. Smith a nd D.N. Jones, 2003. Structure of a specific alcohol binding site defined by the odorant binding protein lush from drosophila melanogaster. Nature structural biology, 10(9): 694 700. Available from DOI 10.1038/nsb960. Kulmuni, J., 2013. Comparative genomics of chemosensory protein genes reveals rapid evolution and positive selection in ant specific duplicates. Heredity: 1 10. Kulmuni, J. and H. Havukain en, 2013. Insights into the evolution of the csp gene family through the integration of evolutionary analysis and comparative protein modeling. Plos One, 8(5). Available from ://000319733000019. DOI ARTN e63688 DOI 10.1371/journal.pone.0063688. Kulmuni, J., Y. Wurm and P. Pamilo, 2013. Comparative genomics of chemosensory protein genes reveals rapid evolution and positive selection in ant specific duplicates. Heredity, 110(6): 538 547. Available from DOI 10.1038/hdy.2012.122. Lartigue, A., V. Campanacci, A. Roussel, A.M. Larsson, T.A. Jones, M. Tegoni and C. Cambillau, 2002. X ray structure and ligand binding study of a moth chemosensory prot ein. The Journal of biological chemistry, 277(35): 32094 32098. Available from DOI 10.1074/jbc.M204371200. Lartigue, A., A. Gruez, S. Spinelli, S. Riv iere, R. Brossut, M. Tegoni and C. Cambillau, 2003. The crystal structure of a cockroach pheromone binding protein suggests a new ligand binding and release mechanism. The Journal of biological chemistry, 278(32): 30213 30218. Available from DOI 10.1074/jbc.M304688200. Mobley D, R.J., 2005. Usda amends imported fire ant quarantine. In: USDA APHIS Press Releases. USDA APHIS Press Releases, Mosbah, A., V. Campanacci, A. Lartigue, M. Tegoni, C. Cambillau and H. Darbon, 2003. Solution structure of a chemosensory protein from the moth mamestra brassicae.

PAGE 169

169 The Biochemical journal, 369(Pt 1): 39 44. Available from DOI 10.1042/BJ20021217. Murphy, E.J., J.C. Booth, F. Davra zou, A.M. Port and D.N. Jones, 2013. Interactions of anopheles gambiae odorant binding proteins with a human derived repellent: Implications for the mode of action of n,n diethyl 3 methylbenzamide (deet). The Journal of biological chemistry, 288(6): 4475 4 485. Available from DOI 10.1074/jbc.M112.436386. Ortiz, A., 2013. Derivation of hill's equation from scale invariance. Journal of Uncertain Systems, 7 (3): 198 202. Ozaki, M., A. Wada Katsumata, K. Fujikawa, M. Iwasaki, F. Yokohari, Y. Satoji, T. Nisimura and R. Yamaoka, 2005. Ant nestmate and non nestmate discrimination by a chemosensory sensillum. Science, 309(5732): 311 314. Available from DOI 10.1126/science.1105244. Pelosi, P., J.J. Zhou, L.P. Ban and M. Calvello, 2006. Soluble proteins in insect chemical communication. Cell. Mol. Life Sci., 63( 14): 1658 1676. Available from 005 5607 0 DOI 10.1007/s00018 005 5607 0. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real time rt pcr. Nucleic Acids Research 29(9): 2002 2007. Picone, D., O. Crescenzi, S. Angeli, S. Marchese, A. Brandazza, L. Ferrara, P. Pelosi and A. Scaloni, 2001. Bacterial expression and conformational analysis of a chemosensory protein from schistocerca gregari a. Eur J Biochem, 268(17): 4794 4801. Available from ://000170903500021. Picone, D., O. Crescenzi, S. Angeli, S. Marchese, A. Brandazza, L. Ferrara, P. Pelosi and A. Scaloni, 2001. Bacterial expression and conformational analysis of a chemosens ory proteinfrom schistocera gregaria Eur J Biochem, 268: 4794 4801. Renthal, R., 2003. Discovering pheromones of the red imported fire (ant solenopsis invicta buren) a review and proposed new target for pheromone disruption. Journal of Agricultural and U rban Entomology, 20(3): 113 121. Renthal, R., D. Velasquez, O. David, H. Joan and P.W. William, 2003. Structure and distribution of antennal sensilla of the red imported fire ant. Micron, 34. Available from 4328(03)00050 7 DOI 10.1016/s0968 4328(03)00050 7. Renthal, R., D. Velasquez, D. Olmos, J. Hampton and W.P. Wergin, 2003. Structure and distribution of antennal sensilla of the red imported fire ant. Micron 34(8): 405 413. DOI 10.1016/s0968 4328(03)00050 7.

PAGE 170

170 Rocca, J.R., J.H. Tumlinson, B.M. Glancey and C.S. Lofgren, 1983. The queen recognition pheromone of solenopsis invicta: Preparation of (e) 6 (1 pen tenyl) 2 h pyran 2 one. Tetrahedron Letters, 24(18): 1889 1892. Ross, K.G. and L. Keller, 1995. Ecology and evoltion of social organization: Insights from fire ants and other highly eusocial insects. Annu. Rev. Ecol. Syst., 26: 631 656. Rutzler, M. and L.J. Zwiebel, 2005. Molecular biology of insect olfact ion: Recent progress and conceptual models. Journal of comparative physiology. A, Neuroethology, sensory, neural, and behavioral physiology, 191(9): 777 790. Available from http://www.ncbi.nlm.nih. gov/pubmed/16094545 DOI 10.1007/s00359 005 0044 y. Sanchez Gracia, A., F.G. Vieira and J. Rozas, 2009. Molecular evolution of the major chemosensory gene families in insects. Heredity, 103(3): 208 216. Available from DOI 10.1038/hdy.2009.55. Sato, K., M. Pellegrino, T. Nakagawa, T. Nakagawa, L.B. Vosshall and K. Touhara, 2008. Insect olfactory receptors are heteromeric ligand gated ion channels. Natu re, 452(7190): 1002 1006. Available from DOI 10.1038/nature06850. Sato, K. and K. Touhara, 2009. Insect olfaction: Receptors, signal transduction, and behavior. Results and problems in cell differentiation, 47: 121 138. Available from DOI 10.1007/400_2008_10. Seagraves, M.P. and R.M. McPherson, 2006 Monitoring red imported fire ant (hymenoptera : Formicidae) foraging behavior and impact on foliage dwelling arthropods on soybean produced under three ant suppression regimes. Journal of Entomological Science, 41(4): 374 384. Available from : //000241270500009. Simoni, R.D., R.L. Hill and M. Vaughan, 2002. The structure and function of hemoglobin: Gilbert smithson adair and the adair equations. Journal of Biological Chemistry, 277(31). Smith, C.R., A. Dolezal, D. Eliyahu, C.T. Holbrook and J. Gadau, 2009. Ants (formicidae): Models for social complexity. Cold Spring Harbor protocols, 2009(7): pdb emo125. Available from DOI 10.1101/pdb.emo125 Stefan, M.I. and N.L. Novere, 2013. Cooperative binding. PLOS Computational Biology, 9(6). DOI 10.1371/journal.pcbi.1003106.g001 10.1371/journal.pcbi.1003106.g002.

PAGE 171

171 Swarup, S., T.I. Williams and R.R. Anholt, 2011. Functional dissection of odorant bindin g protein genes in drosophila melanogaster. Genes, brain, and behavior, 10(6): 648 657. Available from DOI 10.1111/j.1601 183X.2011.00704.x. Tegoni, M ., V. Campanacci and C. Cambillau, 2004. Structural aspects of sexual attraction and chemical communication in insects. Trends in biochemical sciences, 29(5): 257 264. Available from http://www.ncb DOI 10.1016/j.tibs.2004.03.003. Tomaselli, S., O. Crescenzi, D. Sanfelice, E. Ab, R. Wechselberger, S. Angeli, A. Scaloni, R. Boelens, T. Tancredi, P. Pelosi and D. Picone, 2006. Solution structure of a chemosensory protein from the desert locust schistocerca gregaria. Biochemistry, 45(35): 10606 10613. Available from DOI 10.1021/bi060998w. Touhara, K. and L.B. Vosshall, 2009. Sensing odorants and pher omones with chemosensory receptors. Annual review of physiology, 71: 307 332. Available from DOI 10.1146/annurev.physiol.010908.163209. Trapp, S.C. an d R.B. Croteau, 2001. Genomic organization of plant terpene synthases and molecular evolution implications. Genetics, 158: 811 832. Tschinkel, W.R., 2006. The fire ants. Belknap Press of Harvard University. Vander Meer, R.K., 1981. Hydrocarbon components of the trail phermone of the red imported fire ant, solenopsis invicta Tetrahedron Letters, 22(18): 1651 1654. Vander Meer, R.K., 1983. Semio chemicals and the red imported fire ant (solenopsis invicta buren)(hymenoptera:Formicidae). The Florida Entomol ogist, 66(1): 139 161. Vander Meer, R.K., 1988. Isolation of the trail recruitment pheromone of solenopsis invicta. Journal of Chemical Ecology, 14(3): 825 838. Vander Meer, R.K. and L.E. Alonso, 1998. Pheromone directed behavior in ants. Pheromone Commu nication in Social Insects: 159 192. Vargo, E.L. and C.D. Hulsey, 2000. Multiple glandular origins of queen pheromones in the fire ant solenopsis invicta Insect Physiology, 46: 1151 1159.

PAGE 172

172 Vieira, F.G. and J. Rozas, 2011. Comparative genomics of the odor ant binding and chemosensory protein gene families across the arthropoda: Origin and evolutionary history of the chemosensory system. Genome biology and evolution, 3: 476 490. Available from http:/ / DOI 10.1093/gbe/evr033. Vogt, R.G., 2005. Molecular basis of pheromone detection in insects. Comprehensive Insect Physiology, Biochemistry, Pharmacology and Molecular Biology, 3: 753 804. Wang, J., S. Jemielity, P. Uva, Y. Wurm, J. Graff and L. Keller, 2007. An annotated cdna library and microarray for large scale gene expression studies in the ant solenopsis invicta. Genome biology, 8(1): R9. Available from DOI 10.1186/gb 2007 8 1 r9. Wang, J., Y. Wurm, M. Nipitwattanaphon, O. Riba Grognuz, Y.C. Huang, D. Shoemaker and L. Keller, 2013. A y like social chromosome causes alternative colony organization in fire ant s. Nature, 493(7434): 664 668. Available from DOI 10.1038/nature11832. Wurm, Y., J. Wang, O. Riba Grognuz, M. Corona, S. Nygaard, B.G. Hunt, K.K. Ingr am, L. Falquet, M. Nipitwattanaphon, D. Gotzek, M.B. Dijkstra, J. Oettler, F. Comtesse, C.J. Shih, W.J. Wu, C.C. Yang, J. Thomas, E. Beaudoing, S. Pradervand, V. Flegel, E.D. Cook, R. Fabbretti, H. Stockinger, L. Long, W.G. Farmerie, J. Oakey, J.J. Boomsma P. Pamilo, S.V. Yi, J. Heinze, M.A. Goodisman, L. Farinelli, K. Harshman, N. Hulo, L. Cerutti, I. Xenarios, D. Shoemaker and L. Keller, 2011. The genome of the fire ant solenopsis invicta. Proceedings of the National Academy of Sciences of the United Sta tes of America, 108(14): 5679 5684. Available from DOI 10.1073/pnas.1009690108. Xu, Y.L., P. He, L. Zhang, S.Q. Fang, S.L. Dong, Y.J. Zhang and F. Li, 2009. Large scale identification of odorant binding proteins and chemosensory proteins from expressed sequence tags in insects. BMC genomics, 10: 632. Available from bmed/20034407 DOI 10.1186/1471 2164 10 632. Yu, Y., F. Ma, Y. Cao, J. Zhang, Y. Zhang, S. Duan, Y. Wei, S. Zhu and N. Chen, 2012. Structural and functional difference of pheromone binding proteins in discriminating chemicals in the gypsy moth, lymantria dispar. International journal of biological sciences, 8(7): 979 991. Available from DOI 10.7150/ijbs.4557.

PAGE 173

173 BIOGRAPHICAL SKETCH Arun Wanchoo was bor n and brought up in the suburbs of Lauderhil l, Florida in the mid 1980s. His father is from New Delhi, India and his mother from Trinidad and Tobago with Indian descent. He grew up with little league sports, playing i n the front yard, and school. In his younger school years he p articipated in science fairs, math and engi neering competitions. The enjoyment of higher level thinking as well as working with his hands lead him towards a science and engineering direction in life. Though he enjoyed the phys ical and life sciences, he was led in the direction of mechanical engineering because of an interest in physics and math and the love to create However, after taking Advanced Placement Biology with a very enthusiastic Dr. Anthony Arico an interest in bi ology sparked Dr. Arico introduced the fascinating topic of molecular biology and engineering involving streptavidin and biotin. From this, grew an interest for molecular biology and biological engineering When Arun cam e to the University of Florida he had already declared his major ed back. His first opportunity in the lab that expanded to a Research Experience for Undergraduates internship opportunity As an undergraduate Arun was awarded an NSF REU research and a graduate education become his goal. He was accepted into the Fellowship

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EPF8DCL46_J5CQPR INGEST_TIME 2014-10-03T22:16:46Z PACKAGE UFE0045413_00001