University of Florida Thesis Or Dissertation Formatting

Home , Acacia pruinosa, Cephalotus, Cunonia capensis, Dischidia, Elaeocarpaceae, Pseudoweinmannia

EVOLUTION OF THE GENETIC CONTROL OF LEAF DEVELOPMENT WITH AN EMPHASIS ON CARNIVOROUS PITCHER PLANTS

NICHOLAS WILLIAM MILES

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2013

To those who believed in me, especially my Mom and Dad, Heather-Rose, and Susan Spaulding

ACKNOWLEDGMENTS

I would like to thank my advisors, Pam and Doug Soltis, for their tireless work and support. This dissertation is testament to the saint-like patience they have had with me over the years, and especially this year. I would also like to thank my committee members, Drs. David Oppenheimer and Mark Settles, for the integral part they have played in the forming of my ideas. Both of them and their labs have shared resources that I cannot thank them for enough. Drs. Tsukaya and Yamaguchi, who showed genuine rectitude as hosts to me in a foreign country and enabled me to collect data that I had strived to do for 6 years, I am forever indebted to. The members of the Soltis

Lab, who I now view in likeness of family members, I will remember for the rest of my life with great sentiment.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 7

LIST OF FIGURES ...... 8

ABSTRACT ...... 11

CHAPTER

1 INTRODUCTION ...... 12

Angiosperm Leaves as a System for the Study of Evolutionary Development (Evo-Devo) ...... 12 Nucleotide Sequencing Technology Leafs Out ...... 15 Evo-Devo Studies of Carnivorous Pitcher Plants ...... 16

2 NEO- AND SUB-FUNCTIONALIZATION OF THE PLANT BODY PATTERNING GENE CLASS III HD-ZIP CORRELATES WITH LAND PLANT EVOLUTION ...... 19

Introductory Remarks...... 19 Methods ...... 24 Sequence Retrieval ...... 24 Alignment ...... 25 Phylogenetic Analyses ...... 25 Results ...... 26 BLAST results ...... 26 Phylogenetic analyses ...... 26 Discussion ...... 29

3 GENOMIC COEVOLUTION OF HD-ZIP III RELATED PATHWAY GENES ...... 46

Introductory Remarks...... 46 Methods ...... 51 Sequence retrieval ...... 51 Alignment ...... 53 Phylogenetic analyses ...... 53 Arabidopsis microarray expression ...... 53 Results ...... 54 Paralog numbers ...... 54 Pathway proportionality ...... 54 Lineage evolution ...... 55 Discussion ...... 56

4 LEAF EVO-DEVO OF THE OF CARNIVOROUS PITCHER PLANT, CEPHALOTUS FOLLICULARIS (CEPHALOTACEAE) ...... 77

Introductory Remarks...... 77 Methods ...... 80 RNA extraction and gene amplification ...... 80 Sequence alignment and phylogenetic analysis ...... 81 In situ hybridization ...... 81 Probe construction ...... 81 Tissue fixation ...... 82 Hybridization ...... 82 Results ...... 83 Discussion ...... 83

5 CONCLUSION ...... 96

APPENDIX

A FURTHER CARNIVOROUS PLANT EXPERIMENTS ...... 99

AS2 ...... 99 VIGS Development ...... 100

LIST OF REFERENCES ...... 107

BIOGRAPHICAL SKETCH ...... 116

LIST OF TABLES Table page

2-1 GenBank accession numbers of sequences from previous studies of Floyd et al. (2006) and Prigge and Clark (2006) ...... 33

2-2 Sampled Gymnosperm Taxa from 1KP Data...... 34

2-3 Sampled Monilophyte taxa from 1KP data ...... 35

3-1 Number of paralogs of the 4 gene families in the 6 sampled angiosperm genomes...... 60

3-2 Proportions of the number of paralogs for each of the 4 gene families in the genomes of the 6 sampled angiosperm genomes...... 61

3- 3 Places in the genomes of the sampled angiosperm genomes that contain the tandem precursor miR166 sequences...... 61

4-1 Nucleotide sequence of primers used to amplify HistoneH4 and YABBY genes in Cephalotus follicularis...... 86

LIST OF FIGURES Figure page

1-1 Genetic model controlling leaf development and polarity in angiosperms...... 18

2- 1 Evolution of HD-ZIP III and ZPR in land plants...... 36

2-2 Protein domain structure of HD-ZIP III and ZPR...... 38

2-3 Phylogenetic tree of the topology found to have the highest likelihood by RAxML for sampled vascular plant HD-ZIP III sequences...... 39

2-4 Subtree of HD-ZIP III vascular plant phylogeny containing lycophtye and PaleoHDZ3 clades...... 40

2-5 Subtree HD-ZIP III vascular plant phylogeny containing gymnosperm ZPR clade...... 42

2-6 Subtree HD-ZIP III vascular plant phylogeny containing monilophyte RBVC8 as well as gymnosperm and angiosperm C8 clades...... 43

2-7 Gymnosperm RBVC8 as well as gymnosperm and angiosperm RBV subtree of of HD-ZIP III vascular plant phylogeny...... 44

2-8 Alignment of gymnosperm ZPR sequences and the Homeodomain-Leucine- Zipper Region of HD-ZIP III...... 45

3-1 Phylogeny of major Angiosperm lineages and eudicot orders with the placement of sampled taxa...... 62

3-2 Phylogeny and ancestral genome evolution of sampled angiosperms and the number of copies of each of the sampled genes in their genome...... 63

3-3 Bar graphs comparing the number of copies of HD-ZIP III with the number of copies of the genes that regulate it in the genomes of the six sampled species...... 64

3-4 Gene trees and species trees of the four leaf polarity genes in the six angiosperm genomes...... 66

3-5 Phylogram of sampled HD-ZIP III genes...... 67

3-6 Unrooted phylogeny of sampled GARP and KANADI genes from green plants and a protozoan...... 68

3-7 Unrooted phylogram of sampled KANADI genes...... 69

3-8 Alignment of miR166 from the genomes of Amborella, Aquilegia, Vitis, and Arabidopsis...... 70

3-9 Syntenic regions of the Amborella and Aquilegia genomes that contain a discovered miR166 copy...... 71

3-10 Unrooted phylogram of sampled ZPR genes...... 72

3-11 Expression from 454 read counts from Ancestral Angiosperm project (http://ancangio.uga.edu/) of Amborella HD-ZIP III paralogs...... 73

3-12 Expression from microarray intensities of Arabidopsis HD-ZIP III paralogs...... 74

3-13 Expression from microarray intensities of Arabidopsis ZPR paralogs...... 75

3-14 Expression from microarray intensities of Arabidopsis KANADI paralogs...... 76

4-1 Front view of Cephalotus follicularis pitcher and basal rosette of flat leaves...... 88

4- 2 Photographs of Cephalotus pitcher, flat leaves, intermediate leaf, inflorescence, and flowers...... 89

4-3 Longitudinal section drawing of Cephalotus mature pitcher and shoot tip...... 90

4-4 Alignment of Cephalotus and Arabidopsis HistoneH4 amino acid sequences. ... 91

4-5 Phylogram of angiosperm and isolated carnivorous pitcher plant YABBY paralogs...... 92

4-6 In-situ hybridization of HistoneH4 in developing Cephalotus pitcher...... 93

4-7 In-situ hybrizations of CephFIL in serial longitudinal sections of a developing Cephalotus pitcher...... 94

4-8 Stipules in the seedling of Geissois pruinosa (Cunoniaceae)...... 95

4-9 Expression of TAS3 in the vasculature, adaxial domain of developing leaves, and stipules of Arabidopsis thaliana...... 95

A-1 Phylogeny of Arabidopsis AS2/LOB paralogs and AS2/LOB BLAST results form 1KP (www.onekp.com) transcripts for Nepenthes alata (abbreviated Nep), Cunonia capensis (Cun), and Cephalotus follicularis (Cep)...... 101

A-2 Longitudinal section of in situ hybridization of Nepenthes alata AS2...... 102

A-3 Longitudinal sections of fluorescent In situ hybridization of Nephenthes alata AS2...... 103

A-4 Longitudinal sections of fluorescent in situ hybridization of Cephalotus AS2. .. 104

A-5 Longitudinal sections of fluorescent in situ hybridization of Cephalotus AS2. .. 105

A-6 Electrophoresis gel visualization of TRV1 and TRV2 PCR’s of Agrobacterium infected plants...... 106

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

EVOLUTION OF THE GENETIC CONTROL OF LEAF DEVELOPMENT WITH AN EMPHASIS ON CARNIVOROUS PITCHER PLANTS By

Nicholas William Miles

August 2013

Chair: Pamela S. Soltis Cochair: Douglas E. Soltis Major: Botany

In this dissertation I explore, with various methods, the evolution of the genes that control the development of leaves. First, the evolutionary history of the adaxial identity gene Class III Homeodomain-Leucine Zipper (HD-ZIP III) and the microProtein

LITTLE ZIPPER (ZPR) is reconstructed, and ZPR is shown to be derived from HD-ZIP

III through a duplication event prior to the origin of extant seed plants. Second, a genome-scale sampling of diverse angiosperms for four genes whose interactions control leaf polarity is conducted. This approach allows predictions of subfunctionalization of paralogs and tests a hypothesis of stoichiometry of genes in the genome. The carnivorous plant, Utricularia gibba, is also sampled and shows extreme amounts of duplications and losses of gene lineages. Third, the expression of an abaxial identity gene, YABBY, is characterized in the carnivorous pitcher plant

Cephalotus follicularis and informs on the evolution of the leaf into a pitcher structure.

CHAPTER 1 INTRODUCTION

Angiosperm Leaves as a System for the Study of Evolutionary Development (Evo-Devo)

“Alles ist Blatt.”

Johann Wolfgang von Goethe (1790)

As one of the greatest thinkers of the 18th Century, Goethe had an undeniably large effect on the field of evolutionary botany, and his famous quotation, translated as

“All is leaf”, highlights the special place that the leaf has in the evolution and development of plants. Many of the adaptations in angiosperms that are thought to be responsible for them being the most specious group of plants living, evolved as developmental changes to leaves. These include the evolutionary transitions to most of the reproductive organs, i.e., sepals, petals, stamens, carpels, the ovule, and most fruits

(reviewed in Mathews and Kramer, 2012).

Leaves also have taken on a multitude of forms in land plants, as they have adapted to the biotic and abiotic interactions in the environment. As one of the central questions of biology is how variation arises, leaves are especially suited to answer this question because of the vast amounts of variation between clades, within clades, and even within an individual.

Some of the most amazing forms that never cease to produce wonder in humans are the extreme shapes adapted for interactions with animals. For instance, some plants modify their leaves into pouches that serve as nests for ants, as seen in the bullate leaves of Dischidia (Apocynaceae) or the modified stipules of some Acacia species (Fabaceae). In Acacia, the stipules of some species have been transformed into sharp spines to protect the plant from herbivores. Modified leaves in the form of

spines are also highly adapted in the succulent, arid-adapted Cactaceae and

Euphorbiaceae. The cup-shaped leaves of Ficus benghalinensis var. krishnae seem to have the unique function of producing an animal interaction that has been adapted as a religious symbol in the Hare Krishna religion (Juniper et al., 1989). And finally, in carnivorous plants, all of the trapping mechanisms involve morphological and structural changes to leaves (Juniper et al., 1989).

All of the above-mentioned instances of extreme structural and functional leaf changes were studied in great detail by plant morphologists, like Goethe, and these aberrations have served as interesting challenges to the models of plant morphology.

Today, the field of Evolutionary-Development, or “Evo-Devo”, typically applies a model of development that involves the interaction of genes to developmental aberrations to investigate how changes in those genes account for the changes in development and evolutionary transitions in morphology. In the case of bizarre leaf shapes, many of the changes involve a change from bifacial to unifacial polarity. The field of developmental genetics has produced a model of bifacial leaf development (Figure 1-1), and applying that model to the origin and development of various leaf forms was the main aim of this dissertation work.

The model of leaf polarity consists of genes that have mutually exclusive adaxial and abaxial expression domains. The initial cue for the orientation of these domains comes from a signal in the meristem so that the adaxial domain is positioned next to the meristem and the abaxial away. The juxtaposition of the expression domains in the leaf primordia is essential for defining a plane a growth to create a flat leaf from the initially radial primordia. The model is also predicated on the idea that the expression domains

of the genes in a bifacial leaf define the anatomical differentiation as well as the morphological differentiation. When a gene of the model is mutated the disruption of the repulsing behavior of the network usually leads to the domination of one domain over the other and the subsequent leaf is radial in morphology with a uniform distribution of the cell types that are characteristic of the dominate domain. Therefore, because carnivorous pitcher plants have unifacial leaves without a flat growth plane, it is expected that the leaf polarity model is acting differently during their development.

Plant leaves also show morphological and anatomical polarity in the proximal- distal axis of the leaf. Most leaves have proceeding from the proximal to distal end base, petiole, and lamina zones. The distance from the shoot apical meristem defines this axis of polarity. One of the molecular signals for positioning the proximal-distal zones is the class I knotted1-like homeobox (KNOX) gene (Freeling, 1992). Gain of function mutations in KNOX cause changes to the proximal-distal placement of leaf characteristics, like the leaf ligule of the maize (Zea mays) (Freeling, 1992). The gene

BLADE ON PETIOLE (BOP) is also known to control the proximal-distal axil, specifically the boundary between the lamina and petiole (Ha et al., 2007). There is evidence that the proximal-distal and adaxial-abaxial genes interact so that the two axes are interlinked genetically (Ha et al., 2007). Carnivorous pitcher plants also show distinct developmental changes in the proximal-distal axes. Nepenthaceae is interpreted morphologically to have a very large base region that is expanded for light capture, a twinning tendril-like petiole, and a pitcher lamina (Franck, 1976). The lamina region of

Sarraceniaceae can be very elongated to make a tall pitcher (Franck, 1976).

Nucleotide Sequencing Technology Leafs Out

Recent advances in nucleotide sequencing technology has allowed an unprecedented amount of sequence data to be collected from diverse plant groups, with most of the data being freely available to the public. Understanding the homology and sequence evolution of the genes involved in the model of bifacial leaf development is a fundamental first step in the study of leaf Evo-Devo and is the motivation for Chapters two and three.

Chapter two describes the evolutionary history of two plant organ polarity genes,

Class III Homeodomain Leucine-Zipper (HD-ZIP III) and LITTLE ZIPPER (ZPR), in land plants. These two genes have an interesting feedback regulatory interaction, in that,

HD-ZIP III is a transcription factor that enhances the transcription of down-stream adaxial development genes, but also enhances ZPR, a microprotein whose protein binds to HD-ZIP III proteins and inactivates their DNA bind ability (Wenkel et al., 2007).

My study provides new insights into the evolution of these genes in gymnosperm and monilophytes, where the number of copies and history of the two genes is not well understood because of the lack of fully sequenced genomes from taxa in those groups.

The presence or absence of ZPR outside of the Angiosperms was not known prior to this study. By fully sampling from all major groups of land plants with data from the One

Thousand Plant Transcriptomes Project (1KP; www.onekp.com), this study was able to better understand the make up of the main clades of these genes and the relations among them. Perhaps, most importantly, this study was able to provide evidence that

ZPR evolved from HD-ZIP III just prior to the divergence of gymnosperms and angiosperms.

Chapter three describes the evolution of HD-ZIP III and three of its regulatory genes in the sequenced genomes of six angiosperm species, including the newly available draft of the first carnivorous plant genome. My description of three regulatory genes of HD-ZIP III includes a post-transcriptional microRNA, a post-translational micro-

Protein, and a repulsing transcription factor. With this sampling of a network of interacting genes, I set out to find out if the network is behaving stoichiometrically by quantifying the number of coding copies in the genomes and their proportions to the gene they regulate, HD-ZIP III. My findings show that, indeed, they do have similar proportionalities in all six sampled genomes, even in the fractionalized genome of the carnivorous plant. My findings also show that the process of fractionalization in the carnivorous plant genome has removed major lineages of the genes, which might be related for the major loss of roots in the plant. The possibility of the lost lineages being specialized for root development is further corroborated by the quantified expression of those lineages in Arabidopsis and Amborella.

Evo-Devo Studies of Carnivorous Pitcher Plants

Chapter four describes at the expression of a paralog of the abaxial leaf identity gene, YABBY, in the carnivorous pitcher plant Cephalotus follicularis. This investigation relies on laborious molecular techniques for assessing the exact areas of expression in the leaf. However, this data allows for the comparisons with the expression domain in bifacial leaves and other non-carnivorous unifacial leaves. The development of carnivorous pitcher plants has been most recently understood as being homologous with the development of the unifacial peltate leaves (Franck, 1976) and expression of

YABBY has been previously described in a peltate leaf (Gleissberg, 2005). Therefore, I test the hypothesis that the carnivorous pitcher plant would have a similar expression

domain of the homologous YABBY paralog as the peltate leaf. The findings from this chapter allow for the tying together of morphological studies two centuries old with genetic models of the past two decades.

Appendix A of this dissertation is a description of other laboratory experiments done with carnivorous plants to try to better understand how the leaf polarity genes are working in the three instances of convergent carnivorous pitcher plant evolution. Some of these experiments include describing the expression domain of the adaxial leaf identity gene, ASSYMENTRIC LEAF2 (AS2) in two of the pitcher plant families,

Cephalotaceae and Nepenthaceae. Other experiments tried to develop a protocol for knocking down the expression of target genes in carnivorous plants. Knocking down the expression of leaf polarity genes was the ultimate goal because it would have provided better evidence for the role of leaf polarity genes in pitcher plants.

Figure 1-1. Genetic model controlling leaf development and polarity in angiosperms. Cross-section of a Hoya sp. leaf showing typical angiosperm leaf polarity (Left). The vein in the middle shows polarity, with xylem on the adaxial side and phloem on the abaxial side. The dermal tissues also show polarity, with large epidermal cells and a thick cuticle on the adaxial side while the abaxial epidermis has smaller cells with interspersed stomata. The ground tissue is polarized with palisade parenchyma on the adaxial side and spongy parenchyma on the abaxial side. The genetic model of leaf polarity, based on developmental genetics of angiosperm model organism, developed during the past two decades (Right). Gene families are shown in their area of expression (adaxial or abaxial), and their interactions with other gene families are depicted (repression or activation arrows).

CHAPTER 2 NEO- AND SUB-FUNCTIONALIZATION OF THE PLANT BODY PATTERNING GENE CLASS III HD-ZIP CORRELATES WITH LAND PLANT EVOLUTION

Introductory Remarks

Class III Homeodomain-Leucine Zipper (HD-ZIP III) is a plant transcription factor that plays an important role in patterning much of the plant body, including embryos, vascular cambium, lateral organ polarity, and apical meristem maintenance. These parts of the plant body have undergone significant adaptations during the evolution of land plants (Embryophyta), producing characteristics of ecological and physiological importance such as plant height, wood, seeds, and leaves. These traits also play integral roles as the targets of agricultural crop improvement. Therefore, understanding the evolution and regulation of HD-ZIP III is closely connected to both fundamental knowledge about plant development and evolution, as well as the use of seeds, grains, wood products, and lateral organs such as fruits and leaves by humans.

Specific actions of HD-ZIP III in plant organs include a role in patterning of early development from zygote to young embryo. Normally developing embryos show a polarized expression pattern of HD-ZIP III, while mutants of the gene fail to develop important early body plan characteristics properly (Bowman and Floyd, 2008). In a certain Arabidopsis thaliana HD-ZIP III mutant, the first embryonic seed leaves, cotyledons, fail to develop (Prigge et al., 2005). This is of economic interest as cotyledons are important products from many agricultural plants, such as seeds of the legume family, Fabaceae (e.g., soybean and peanut). In peanut, the HD-ZIP III regulator, miR166, is normally expressed in the shoot apical meristem, root, and leaves, but lowly expressed in the cotyledons (Chi et al., 2011), providing evidence that HD-ZIP

III may have higher expression in the large peanut cotyledon and may contributes to its size.

Another region of the embryo in which HD-ZIP III is expressed, in addition to the adaxial side of the cotyledons, is the area that develops into the shoot apical meristem

(SAM) (Prigge et al., 2005). The SAM is a region of undifferentiated stem cells that give rise to all the aboveground organs of the plant, including the stem and lateral organs.

The SAM shows increasing size and organization from the earliest land plants to angiosperms (Floyd and Bowman, 2007). HD-ZIP III gene family members control SAM patterning in the angiosperm model Arabidopsis thaliana, with mutations generating plants with reduced or completely absent SAMs (Byrne, 2006).

The vascular cambium is the area of undifferentiated cells that undergo secondary growth to produce phloem and xylem, the latter maturing into wood. True secondary growth is only found in members of Embryophyta. HD-ZIP III has been shown to be involved in vascular cambium growth in the rosid Arabidopsis thaliana

(Zhong and Ye, 2007), the rosid timber tree Populus trichocarpa (Robischon et al.,

2011; Zhu et al., 2013), the asterid Zinnia elegans (Ohashi-Ito and Fukuda, 2003), and the conifer Picea glauca (Côte et al., 2010). Differential development of xylem over phloem by the vascular cambium, and therefore overproduction of wood, has been shown to be promoted by an HD-ZIP III paralog in Populus trichocarpa and is seasonally regulated (Ko et al., 2006).

HD-ZIP III genes also play an important part in controlling the shape and patterning of leaves and other lateral organs by patterning the adaxial side of those developing organs. The horizontal growth of the leaf lamina is attributed to the

juxtaposition of the adaxial and abaxial expression domains in developing leaf primordia. Triple mutants of the HD-ZIP III paralogs phabulosa, phavulota, and revoluta generate Arabidopsis plants with leaf laminar growth abnormalities. Strong mutant plants produce symmetrical needle-like leaves, and weak mutants have a rounded lamina described as a trumpet or bell (McConnell et al., 2001).

The important role of HD-ZIP III in leaf development has made it a popular choice for studies on the evolution of this highly adaptive organ of vascular plants. The main photosynthetic, light-capturing, and respiratory organs, leaves are also commonly co- opted for novel ecological and physiological functions. In vascular plants, leaves are thought to have evolved multiple times (Figure 2-1). Lycophytes first evolved rudimentary microphylls more than 400 million years ago (Kenrick and Crane, 1997).

Later, larger megaphylls evolved independently multiple times in the monilophytes and once in seed plants. These lateral organs have also been modified into reproductive structures, including floral organs (sepals, petals, stamens, and carpels) in angiosperms

(Mathews and Kramer, 2012). The question of homology and developmental mechanisms in the variety of lateral organs in land plants has been debated for some time (Kenrick and Crane, 1997; Kaplan, 2001). Molecular tools and data can be used for interpreting the homology and evolution of lateral organs, facets of these organs that have been hard to interpret based on the morphology of fossil and extant species

(Boyce and Knoll, 2002).

To understand the evolution of plant form, various studies have been conducted on the evolution of molecular developmental mechanisms, with the evolution of leaf developmental mechanisms widely investigated (reviewed in Yamaguchi et al., 2012).

The genome sequence of the lycophyte Selaginella moellendorffii and developmental genetic data for the angiosperm model Arabidopsis thaliana have supplied the most information about the genetic differences between the disparate types of leaves. In general, a genome-wide comparison has revealed a general trend of increasing number of gene families, with 1350 gene families being gained and 442 gene families lost during the evolution between a hypothesized basal vascular plant gene set to the hypothesized basal angiosperm gene set (Banks et al., 2011). Focusing just on the gene families known to be involved in lateral organ initiation and patterning, Floyd and Bowman

(2007) found that only one of the 20 gene families they investigated was missing from the Selaginella genome, but there was a general increase in the number of gene family members between Selaginella and angiosperms. The patterning controlled by the interplay of two of those gene families, KNOX and ASYMMETRIC LEAVES1/ROUGH

SHEATH2/PHANTASTICA (ARP), is similar between Selaginella and Arabidopsis, indicating a possible independent recruitment of the same pathway for similar purposes

(Harrison et al., 2005). HD-ZIP III has also undergone independent diversification and subfunctionalization of paralogs to pattern microphylls in Selaginella and megaphylls in seed plants (Floyd et al., 2006). The micro-RNA-mediated post-transcriptional regulation of HD-ZIP III was also found to be a mechanism of patterning that is shared across all green plants (Floyd and Bowman, 2004).

HD-ZIP III proteins act as transcription factors whose patterns of expression dictate the localized expression of downstream genes that promote the specific cell growth shape and differentiation of the tissue type. The protein regulates downstream genes either as a homo- or heterodimer (Sessa et al., 1993). There are four known

domains in HD-ZIP III proteins (Figure 2-2). Near the N terminus is the homeodomain, a helix-loop-helix motif that binds to the DNA strand of the gene it is targeting for regulation. Directly next to the homeodomain is the leucine-zipper domain where the two HD-ZIP III proteins bind together to make a dimer (Staudt and Wenkel, 2011). The

START (STeroidogenic Acute Regulatory protein–related lipid Transfer) domain is a putative lipid-binding site in the middle of the protein (Mukherjee et al., 2009). Next to that is the HD-SAD domain, which is known to be associated with the START domain of

HD-ZIP III and HD-ZIP IV, but its activity is unknown (Mukherjee and Burglin, 2006). At the C terminus is the MEKHLA domain, which is a PAS-like domain that regulates the activity of the protein-protein dimer. It is the target of an unknown cellular signal that binds to the domain and affects conformation of the HD-ZIP III peptide. This domain’s evolution is curious as it is only found in cyanobacteria and chloroplast genomes and is hypothesized to have fused to an HD-ZIP IV copy to produce the HD-ZIP III gene at some time after the divergence of red algae from green plants but before the origin of land plants (Mukherjee and Burglin, 2006; Magnani and Barton; 2011). Floyd et al.

(2006), in their sampling of charophycean algae, found one copy of HD-ZIP III in Chara carollina (Charales) but no copies in Coleochaete scutata (Coleochaetales). Lack of expression of the gene in the abaxial side is achieved through the post-transcriptional degradation initiated by miR165/166. Mutational studies in Arabidopsis have shown that mutations in the micro-RNA binding site cause a gain of function so that HD-ZIP III is also expressed in the abaxial side and the subsequent organs are adaxialized.

Mutations in other areas of HD-ZIP III cause a loss of function, and the organ is abaxialized (McConnell et al., 2001).

LITTLE ZIPPER (ZPR) is a small “microProtein” (Staudt and Wenkel, 2011), with a conserved leucine-zipper region whose expression is activated by HD-ZIP III (Wenkel et al, 2007). Once translated, the Leucine-Zipper region of the ZPR protein binds to the

Leucine-Zipper region of an HD-ZIP III protein, forming a non-functional heterodimer.

This negative feedback pathway has been found in Arabidopsis, and the paralogs have been found in the genomes of rice (Oryza sativa) and corn (Zea mays), but little is known about its evolution or presence in other plants (Staudt and Wenkel, 2011). The evolutionary history of HD-ZIP III has been well studied in land plants (Prigge and Clark,

2006; Floyd et al., 2006), in the genomes of rice (Jain et al., 2008) and poplar (Populus trichocarpa) (Hu et al., 2012), and in the context of the whole super-class of

Homeodomain genes in plants (Mukherjee et al., 2009). However, most studies of HD-

ZIP III have sparsely sampled the monilophyte clade and gymnosperms. With the recent sequencing of a large sampling of green plant transcriptomes by the 1KP project (One

Thousand Plant Transcriptomes; www.onekp.com), we are now able to understand better the evolution of HD-ZIP III and its regulator, ZPR. Here we investigate the evolutionary history of HD-ZIP III and ZPR with extensive sampling across land plants, and test the hypothesis that ZPR evolved from an HD-ZIP III paralog by truncation of the gene down to the lone Leucine-Zipper region.

Methods

Sequence Retrieval

Nucleotide sequences from previous studies of HD-ZIP III (Prigge and Clark,

2006; Floyd et al., 2006) were downloaded from GenBank

(http://www.ncbi.nlm.nih.gov/genbank), and ZPR sequences were downloaded from The

Arabidopsis Information Resource (www.arabidopsis.org; Lamesch et al., 2011) (Table

2-1). Those nucleotide sequences were translated to amino acid sequences and used as a query in tBLASTn searches of the Illumina-generated transcriptome sequence data produced by the 1KP (One Thousand Plant Transcriptomes; www.onekp.com) project for the taxa listed (Table 2-2; Table 2-3). TBLASTn searches were also performed on the EST assemblies of Zamia vazquezii from the Ancestral Angiosperm Genome

Project (www.ancangio.uga.edu) and the assembled genome of Amborella trichopoda

(www.amborella.org).

Alignment

Previously published sequences and large 1KP BLAST hits with characteristic

HD-ZIP III motifs were translated and included in the final alignment using MAFFT

(Katoh et al., 2002) in Geneious version 5.6.5 (Biomatters, www.geneious.com). The alignments were subsequently refined manually. The leucine-zipper regions of the ZPR genes were aligned with leucine-zipper regions of the HD-ZIP III genes.

Phylogenetic Analyses

The amino acid alignment was phylogenetically analyzed using RAxML (RAxML-

VI-HPC) (Stamatakis, 2006) with the GTR + I + Gamma model of amino acid sequence evolution. The topology with the highest maximum likelihood score was generated, and support for that topology was assessed by 1000 bootstrap replicate searches using the thorough bootstrap algorithm.

Amino acid sequences were used instead of nucleotide to account for the long evolutionary distance of the sampled taxa. Amino acid data can be favored because the

20 characters of the amino acid data can code for more evolutionary information per site than the 4-letter code of nucleotide data and amino acid data have less compositional bias than DNA (Brocchieri, 2001). Past studies of HD-ZIP III evolution compared the

phylogenetic results of the same data represented in amino acid characters and nucleotide characters and found few discrepancies (Prigge and Clark, 2006).

Results

BLAST results

Searches of the HD-ZIP III genes found putative copies in the transcriptomes of the sampled lycophytes, monilophytes, and gymnosperms. The Amborella genome was also queried with HD-ZIP III using tBLASTn, and three areas of the genome showed characteristic HD-ZIP III motifs and intron-exon structure. TBLASTn searches with ZPR found two copies in the genome of Amborella and generally one putative copy in each of the gymnosperm taxa. However, no suitable hits were found in the transcriptomes of the monilophytes, and the closest hits were of HD-ZIP III monilophyte transcripts. The reverse searches with gymnosperm ZPR genes as a query of GenBank found only angiosperm HD-ZIP III and ZPR genes. The leucine-zipper regions of the HD-ZIP III and

ZPR genes showed distinct motifs differing from the leucine-zipper regions of other HD-

ZIP III genes (Figure 2-4).

Phylogenetic analyses

Phylogenetic analyses of the combined data of Floyd et al. (2006) and Prigge and Clark (2006) revealed that there is another lineage of HD-ZIP III in plants that was not evident in either of the individual data sets. This new lineage is present in monilophytes and gymnosperms but does not appear to exist in angiosperms (Figure 2-

1, Figure 2-3). This previously undetected lineage is sister to another exclusively monilophyte + gymnosperm lineage termed “PaleoHDZ3” by Prigge and Clark (2006).

This lineage appears to have arisen from a duplication that occurred prior to the origin of monilophytes but after the divergence of lycophytes from the remaining

tracheophytes (and thus is present in extant gymnosperms). This means that two separate duplications of this gene occurred prior to the divergence of monilophytes and seed plants. No HD-ZIP III paralogs that belonged to the “PaleoHDZ3” lineages were found in angiosperms, making it appear that both copies were lost after the divergence of angiosperms from extant gymnosperms but prior to the diversification of extant angiosperms.

Each of the three HD-ZIP III sequences found in the genome of Amborella appears in the phylogeny as sister to one of the three main lineages of HD-ZIP III found in other angiosperms, including a “C8”, “REV”, and “PHX” copy (following the nomenclature of Prigge and Clark (2006)). This means that of the five copies present in the gymnosperms, the ancestor of angiosperms retained only two copies and that a duplication of one of those two copies prior to the origin of the angiosperms produced the “REV” and “PHX” lineages. The fact that no PaleoHDZ3 sequences could be found in the genome of Amborella allows for a better understanding of the timing of the loss of

PaleoHDZ3 lineages in angiosperms. This loss seems to have taken place between 150 and 300 million years ago, oddly enough, when a major paleopolyploid event is believed to have taken place before angiosperm diversification (Jiao et al., 2011). However, our phylogenetic results show that a minimum of three duplication events happened prior to the diversification of the seed plants, lending support to the presence of the whole- genome duplication event before the diversification of the clade.

The inclusion of 110 gymnosperm HD-ZIP III sequences, 48 monilophyte HD-ZIP

III sequences, and 27 gymnosperm ZPR sequences further clarified and supported the presence of an unknown monilophyte + seed plant lineage (Figure 2- 3). With this larger

sampling it is clear that the clade denoted as PaleoHDZ3 by Prigge and Clark (2006) actuality represents two separate gene lineages in vascular plants, denoted here as

PaleoHDZ3a and PaleoHDZ3b. The monophyly of both the PaleoHDZIP3a and

PaleoHDZIP3b clades was supported by 82% and 92% bootstrap support, respectively.

On a finer scale, a paleo-duplication event in the PaleoHDZ3b lineage is evident in leptosporangiate ferns. The Angiopteris evecta PaleoHDZ3b sequence is sister to the two lineages of leptosporangiate PaleoHDZIP3b sequences with 94% bootstrap support. This indicates the duplication happened after the divergence of leptosporangiate and eusporangiate ferns and is consistent with a hypothesized whole- genome duplication event prior to the diversification of leptosporangiate ferns (Barker,

2013).

The backbone of the “RBVC8” seed plant clade was poorly resolved (Figure 2-3) with the Gymno-RBCVC8 clade sister to the Gymno-RBV clade but with only 21% bootstrap support. The relationship between Gymno-RBVC8, Gymno-C8, and Gymno-

RBV has been a difficult relationship to resolve with high support in our study, as well as in previous studies (Prigge and Clark, 2006; Floyd et al., 2006), possibly because of the number of and time since the duplication events. We have chosen to depict the relationship in Figure 2-1 with Gymno-RBVC8 sister to the RBV + C8 clades because of the low support for our inferred relationship and the precedence of the two previous studies (Prigge and Clark, 2006; Floyd et al., 2006). However, both topologies, with

Gymno-C8 sister to RBVC8 and RBV, or with Gymno-RBVC8 sister to C8 and RBV, require an equally parsimonious number of duplication and loss events to explain the presence in angiosperms.

The ZPR sequences of gymnosperms and angiosperms were consistently grouped together with >98% bootstrap support when included in the phylogenetic analyses with the HD-ZIP III sequences, indicating that gymnosperm and angiosperm

ZPR arose as the result of a single evolutionary event. When the ZPR sequences were aligned with the HD-ZIP III sequences, the ZPR sequences grouped consistently with the paleoHDZ3 lineages. The gymnosperm sequences were easier to align to the HD-

ZIP III sequences presumably because the amino acid residues around the leucine- zipper were more conserved than the angiosperm ZPR sequences. When only gymnosperm ZPR genes were included in the ZPR-HD-ZIP III alignment, the ZPR sequences grouped with the paleoHDZ3b sequences with 84% bootstrap support and grouped with the gymno-paleoHDZ3b sequences with 70% bootstrap support. This is evidence that ZPR evolved from the duplication of a gymno-paleoHDZIP3b sequence and was subsequently passed on to angiosperms.

Discussion

In previous studies (Floyd et al., 2006; Prigge and Clark, 2006), the number of

HD-ZIP III lineages in some angiosperms, lycophytes, and mosses was known with a high degree of certainty because of the genome sequences available for species from each group. However, no genomes or genomic resources were available for any monilophytes or gymnosperms. The utility of recent sequences obtained from the transcriptome sequencing project, 1KP , was well illustrated here. Two main factors of the project contributed to the successful use of the sequences they generated. First, the

1KP sampling was of considerable breadth, including many species from each of the main lineages of gymnosperms and monilophytes. Second, for this study, the partial, random sequencing of the transcriptome employed by 1KP, rather than generating

traditional Expressed Sequence Tags (ESTs), allowed for better sampling of the highly conserved regions of the HD-ZIP III genes. ESTs, in contrast, capture mostly 3’-UTR regions of the gene that are rapidly evolving and difficult to align in large taxonomic samplings. The 1KP transcriptome data were therefore more useful for large-scale phylogenetic studies, like this one, and made it easier to find homologs with tBLASTn searches queried with homologs from other lineages.

The 1KP transcriptome data detected more paralogs than did the two previous studies based on Sanger sequencing (Floyd et al., 2006; Prigge and Clark, 2006). Our findings also differ from the EST sequence study of Côte et al. (2010), whose Neighbor-

Joining analyses differ in the description of the relationship of the “PaleoHDZIP3” lineage to the other seed plant lineages. This discrepancy may be due to the greater susceptibility of Neighbor-Joining methods than maximum likelihood to long-branch attraction (Felsenstein, 1978), which could be expected in data sets that have limited taxonomic sampling (Graybeal, 1998; Soltis and Soltis, 2004).

The truncation of an HD-ZIP III paralog to the ZPR gene is reminiscent of the other lateral organ development gene family, AUXIN RESPONSE FACTOR (ARF), which also functions as a dimer and has evolved a diversity of isoforms mainly by truncation and alternative splicing (Finet et al., 2013). In addition, a similar gene truncation event happened in the Homeodomain gene KNOX, where the truncation also involved the loss of the homeodomain region but a conservation of the dimer-binding domain as a way to regulate the activity of the KNOX genes (Magnani and Hake, 2008).

This study helps explain not only when ZPR evolved, but also the circumstances that may have necessitated its evolution. As a regulator of HD-ZIP III, ZPR origin may

be related to the fact that there was an unprecedentedly large number of functional copies in the genome when ZPR originated. Given that HD-ZIP III evolved somewhere in multicellular green algae, additional copies of HD-ZIP III have accumulated in the genomes of mosses, and still more in lycophytes, and monilophytes. These groups evolved more specialized organs, and patterning of those organs became more complex as they adapted to conditions on land; concomitantly, the number of copies of

HD-ZIP III increased with an increase in the complexity of function. For instance, there is no known copy of HD-ZIP III in Coleochaetales, and there is only one copy of HD-ZIP

III in Chara coralline (Charales), which represents one of the first lineages to evolve apical growth (Floyd et al., 2006). In mosses there may be multiple copies but only one that was shared between the most recent common ancestor of mosses and vascular plants, and there is only one copy reported in hornworts (Floyd et al., 2006). Lycophytes have two copies of HD-ZIP III, and these are subfunctionalized into two distinct expression domains in response to two major adaptations, vasculature and microphylls

(Floyd et al., 2006). In monilophytes, there are three distinct lineages of HD-ZIP III and this increase in the number of paralogs looks to have happened at the time as the evolution of megaphylls in the group. In gymnosperms, there is a large jump to six HD-

ZIP III lineages, and happens around the same time as the evolution of secondary growth. The origin of the micro-protein ZPR in gymnosperms may have been advantageous for controlling the more complicated subfunctionalized actions of HD-ZIP

III. In angiosperms, an evolutionary trend is observed with duplications of the microprotein and a decrease in the number of copies of the transcription factor, with three

HD-ZIP III lineages and two ZPR lineages in the early angiosperm Amborella

trichopoda. In Arabidopsis thaliana, the ratio is similar, with five HD-ZIP III lineages and four ZPR lineages.

This study is also germane to discussions on the homology of lateral organs. The data suggest that the microphylls of lycophytes and the megaphylls found in euphyllophytes are evolutionarily very different (Kenrick and Crane, 1997; Kaplan,

2001). There appears to be only one copy of HD-ZIP III that is found in the most recent common ancestor of lycophytes and euphyllophytes. The gene lineage in each of these two clades of vascular plants has independently diversified and subfunctionalized and interacted with other genes to produce a leaf organ. With regard to the homology of megaphylls of monilophytes and seed plants, the data do not disagree with the possibility that these two groups co-inherited their HD-ZIP III leaf developmental mechanism. There is no major loss of HD-ZIP III copies prior to the origin of extant gymnosperms or in the various monilophyte clades, making it highly likely that the most recent common ancestor of euphyllophytes had a leaf organ that was inherited by both lineages. However, the fact that seed plants evolved the ZPR gene and monilophytes did not suggests that the lateral organ developmental tool kit in seed plants is more specialized and that ZPR represents a mechanism of control that monilophytes did not evolve.

Table 2-1. GenBank accession numbers of sequences from previous studies of Floyd et al. (2006) and Prigge and Clark (2006) GenBank Species Sequence Accession Species Sequence Sequence Name Number Name Name Selaginella kraussiana SkHDZ31 DQ657196 Cycas revoluta CreHDZ32 DQ657218 Selaginella kraussiana SkHDZ32 DQ657197 Ginkgo biloba GbHDZ31 DQ657215 Selaginella moellendorffii SmHDZ31 DQ657198 Ginkgo biloba GbHDZ32 DQ657216 Selaginella moellendorffii SmHDZ32 DQ657199 Ginkgo biloba GbHDZ33 DQ657217 Selaginella moellendorffii SmHDZ32 DQ657197 Taxus globosa TgC3HDZ1 DQ385530 Ceratopteris richardii CrC3HDZ1 DQ385524 Taxus globosa TgC3HDZ2 DQ385531 Ceratopteris richardii CrC3HDZ1 DQ385524 Pinus taeda PtaHDZ31 DQ657210 Marsilea minuta MmHDZ31 DQ657207 Pinus taeda PtaHDZ32 DQ657211 Marsilea minuta MmHDZ32 DQ657209 Pinus taeda PtaHDZ33 DQ657212 Psilotum nudum PnC3HDZ1 DQ385521 Pinus taeda PtaHDZ34 DQ657213. Psilotum nudum PnC3HDZ2 DQ385522 Pinus taeda PtaHDZ35 DQ657214 Psilotum nudum PnC3HDZ3 DQ385523 Pseudotsuga menziesii PmC3HDZ1 DQ385528 Ceratopteris richardii CriHDZ32 DQ657206

Table 2-2. Sampled Gymnosperm Taxa from 1KP Data Family Species Family Species

Cupressaceae Athrotaxis cupressoides Pinaceae Pinus parviflora Chamaecyparis Pinus ponderosa lawsoniana

Cryptomeria japonica Pinus jeffreyi

Cunninghamia Pinaceae Pinus radiata lanceolata

Fokienia hodginsii Podocarpaceae Podocarpus rubens

Glyptostrobus pensilis Prumnopitys andina

Juniperus scopulorum Sundacarpus amarus

Papaucedrus papuana Sciadopityaceae Sciadopitys verticillata

Thuja plicata Taxaceae Taxus baccata

Thujopsis dolabrata Torreya nucifera

Glyptostrobus pensilis Cycadaceae Cycas micholitzii

Austrocedrus chilensis Zamia vazquezii

Calocedrus decurrens Gnetaceae Gnetum montanum

Cedrus libani Ginkgoaceae Ginkgo biloba

Pinaceae Keteleeria evelyniana Araucariaceae Wollemia nobilis

Cephalotaxus Pseudotsuga wilsoniana Cephalotaxaceae harringtonia

Ephedraceae Ephedra sinica Welwitschiaceae Welwitschia mirabilis

Table 2-3. Sampled Monilophyte taxa from 1KP data Family Species

Asplenium nidus Aspleniaceae Asplenium platyneuron Equisetaceae Equisetum diffusum Marattiaceae Angiopteris evecta Ophioglossaceae Ophioglossum vulgatum Polypodium amorphum Polypodiaceae Polypodium hesperium Pteridaceae Adiantum aleuticum Cheilanthes arizonica Myriopteris eatonii Woodsiaceae Cystopteris fragilis Cystopteris protrusa Gymnocarpium dryopteris Woodsia scopulina

Figure 2- 1. Evolution of HD-ZIP III and ZPR in land plants. Black outline phylogeny represents a chronogram of land plant evolution (species tree) based on inferred dates of Jiao et al. (2011). Colored internal lines represent HD-ZIP III lineages (gene tree). Black arrows mark major land plant morphological adaptations. Green arrows mark currently accepted evolution of microphyll and megaphyll leaves. Starbursts represent gene duplication events.

Figure 2-2. Protein domain structure of HD-ZIP III and ZPR. Dashed lines represent protein-protein interaction. a. HD-ZIP III protein bound to ZPR protein to make a non-functional transcription factor. b. HD-ZIP III proteins form a homodimer that acts as a functional transcription factor (Wenkel et al., 2007).

Figure 2-3. Phylogenetic tree of the topology found to have the highest likelihood by RAxML for sampled vascular plant HD-ZIP III sequences. Numbers at nodes represent the percentage of 1000 bootstrap analyses that support with this topology. The presence of two clades of paleoHDZ3 was not previously known. This phylogeny also represents the first placement of ZPR phylogenetically with HD-ZIP III.

Figure 2-4. Subtree of HD-ZIP III vascular plant phylogeny containing lycophtye and PaleoHDZ3 clades. Tips are labeled with their species name, 4-letter 1KP species identifier, and the scaffold number of the sequence. Tips that represent the sequences form Floyd et al. (2006) and Prigge and Clark (2006) are labeled with the sequence names from those publications.

Figure 2-5. Subtree HD-ZIP III vascular plant phylogeny containing gymnosperm ZPR clade. Tips are labeled with their species name, 4-letter 1KP species identifier, and the scaffold number of the sequence.

Figure 2-6. Subtree HD-ZIP III vascular plant phylogeny containing monilophyte RBVC8 as well as gymnosperm and angiosperm C8 clades. Tips are labeled with their species name, 4-letter 1KP species identifier, and the scaffold number of the sequence. Tips that represent the sequences form Floyd et al. (2006) and Prigge and Clark (2006) are labeled with the sequence names from those publications.

Figure 2-7. Gymnosperm RBVC8 as well as gymnosperm and angiosperm RBV subtree of of HD-ZIP III vascular plant phylogeny. Tips are labeled with their species name, 4-letter 1KP species identifier, and the scaffold number of the sequence. Tips that represent the sequences form Floyd et al. (2006) and Prigge and Clark (2006) are labeled with the sequence names from those publications.

Figure 2-8. Alignment of gymnosperm ZPR sequences and the Homeodomain-Leucine-Zipper Region of HD-ZIP III. Yellow dots mark Leucine sites in HD-ZIP III that have become Isoleucine sites in ZPR. Black dots mark homologous leucine sights in HD-ZIP III and ZPR Leucine-Zipper domains. Red bar marks a highly conserved motif between ZPR and HD-ZIP III.

CHAPTER 3 GENOMIC COEVOLUTION OF HD-ZIP III RELATED PATHWAY GENES

Introductory Remarks

Polyploidy or whole-genome duplication (WGD) is a common phenomenon in plants and is concurrent with the major diversifications in angiosperms (e.g., Soltis et al., 2009; Van de Peer, 2011; Jiao et al., 2011). Polyploidy is known to be a source of genetic innovations with the main mode of the evolution of genetic complexity in angiosperms coming from gene and whole-genome duplications (Otto & Whitton, 2000;

Mable, 2003; Gregory & Mable, 2005; Freeling and Thomas, 2006; Soltis et al., 2010;

Buggs et al., 2011). The diversification of transcription-associated proteins, in particular, has been implicated in the increase in complexity of land plants (Richardt et al., 2007) and more recently the evolution of such structures as the seed and flower following ancient polyploidy in the early evolution of seed plants and angiosperms (Zahn et al.,

2005; Jiao et al. 2011).

Many functionally characterized developmental genes in angiosperms have been highly influenced by gene duplications stemming from ancient polyploidy events. These genes and their duplication events have been implicated in helping to generate morphological innovations in angiosperms such as flowers and fruits (Irish and Litt,

2005; Zahn et al., 2005). The process of subfunctionalization or neofunctionalization of developmental gene duplicates has been shown or hypothesized to spur many of these morphological novelties. A duplication of the leaf shape gene Asymmetric

Leaves1/Rough Sheath 2/Phantastica (ARP) gene has been hypothesized to be involved in the morphologically diverse leaf shapes in the stone plants, Ruschioideae

(Iliing et al., 2009). A gene duplication event and subsequent subfunctionalization of the

leaf development gene PHANTASTICA have been implicated in the evolution of the compound leaf in Lotus japonicus (Luo et al., 2005). Duplications in the organ polarity- controlling gene family, YABBY, have also been involved in patterning angiosperm- specific organs (Bartholmes et al., 2011). By sampling the genome of Utricularia, Ibarra-

Laclette et al. (2012) found an abundance of duplications in TCP (TEOSINTE

BRANCHED1/CYCLOIDEA/PCF) genes and hypothesize that they may be responsible for the large amount of branching diversity seen in the genus.

Even though individual gene family duplications and polyploidy have been implicated in a broad range of evolutionary development studies of land plants, few studies have investigated at the effects of duplication events on pathways or sets of interacting genes. Many questions still remain about what happens to interacting genes after polyploidy. For instance, do other genes in a pathway become sub- or neofunctionalized when a gene acquires a new developmental task? If a gene takes on a new function leading to a new phenotype, do the other interacting genes subfunctionalize to interact with that gene to produce that new phenotype? Also, are interacting genes maintained in stoichiometric relationships? These same questions are just as poorly understood when considering the opposite of gene duplication, gene loss.

Considering how many polyploidization events have occurred in angiosperm history

(Soltis et al., 2009; Jiao et al., 2012), but the relatively stable number of genes in the angiosperm genome (Table 3-1) (Galbraith et al., 2011; Soltis and Soltis, 2013), gene loss is a prevalent process, but the rules governing loss are poorly understood (Soltis et al., 2010).

Here I investigate these questions by characterizing the duplications and losses in perhaps the best functionally characterized family of transcription factors involved in plant body plan, Class III Homeodomain-Leucine Zipper (HD-ZIP III) (Prigge et al.,

2005), and three of its closely interacting genes, Little Zipper (ZPR), KANADI (KAN), and microRNA 166/165 (miR166/165) in six fully sequenced angiosperm genomes.

The six angiosperm species compared here, Amborella trichopoda

(Amborellaceae), Aquilegia coerulea (Ranunculaceae), Vitis vinifera (Vitaceae),

Arabidopsis thaliana (Brassicaceae), Mimulus guttatus (Phyrmaceae), and Utricularia gibba (Lentibulariaceae), span a large swath in the phylogenetic and morphological diversity of angiosperms and have varying numbers of inferred paleopolyploidy events in their history (Figure 3-1, Figure 3-2). Amborella trichopoda is restricted to the island of New Caledonia (Thien et al., 2003) and represents a monotypic lineage that is sister to all other extant angiosperms (Soltis et al., 1999). It has only the single polyploidy event that is shared among all angiosperms, thought to have taken place prior to the divergence of all extant angiosperms (Jiao et al., 2012) and displays many putatively ancestral characteristics of angiosperms, including small, gradually differentiated floral organs (Endress and Igersheim, 2000; Buzgo et al., 2004) and a lack of vessels

(Carlquist and Schneider, 2001). Aquilegia coerulea is small herb in the diverse

Ranunculales, a clade that occupies a phylogenetically important place as sister to the rest of the eudicots (Soltis et al. 2011), having originated after the divergence of magnoliids and monocots. Ranunculales are also important because this lineage diverged from other eudicots prior to a paleopolyploid event that characterizes core eudicots (Jiao et al., 2012). Vitis vinifera, grape vine, is a member of the Vitales, an-

early diverging rosid. Arabidopsis thaliana is an herbaceous annual in Rosidae that is a well-developed genetic model plant with two inferred paleopolyploid events in the history of its order, Brassicales (Simillion et al. 2002; Soltis et al., 2009). Mimulus guttatus is a model plant for ecological diversification and an asterid with few paleopolyploid events. Except for the paleohexaploid event (or two WGD events) inferred to have happened early in eudicot evolution (Jiao et al., 2012), Mimulus may have only one other recent paleopolyploid event inferred to share with the recently sequenced genome of the carnivorous plant Utricularia (Ibarra-Laclette et al., 2013).

Utricularia gibba is an aquatic, and morphologically ambiguous, carnivorous bladderwort

(Rutishauser and Isler, 2001) that lacks true roots. It has one of the smallest nuclear genomes of all sampled angiosperms, with an estimated size of only 88 megabase pairs

(Greilhuber et al., 2006), but has undergone two polyploidy events since its most recent common ancestor with Mimulus. Based on genomic synteny, the contradictory number of genome duplications and small genome size of Utricularia have been explained by extreme fractionation of its genome (Ibarra-Laclette et al., 2013).

Perhaps one of the plant gene families whose paralogs are best functionally characterized in Arabidopsis and other model plants is HD-ZIP III (Prigge et al., 2005). It is also part of a pathway that is functionally and evolutionarily well described, with both post-transcriptional and post-translational regulation understood (Kinder and

Timmermans, 2007; Bowman, 2004; Wenkel et al., 2007).

HD-ZIP III is a transcription factor that controls the polar patterning of many plant developmental stages, including the embryo, meristem, lateral organs, and vascular cambium (Prigge et al., 2009). It acts as a homodimer that binds to target genes and

has three protein domains that act as sites for upstream chemical and physical stimuli

(Magnani and Barton, 2011). In lateral organs and vascular cambium it is expressed on the adaxial side (Ohashi-Ito and Fukuda, 2003).

ZPR is a microprotein that has a lone leucine-zipper domain. It is activated by

HD-ZIP III, and its mature proteins in turn bind to HD-ZIP III proteins, forming a nonfunctional heterodimer, a process that seems to be an adaptation for finely regulating the action of HD-ZIP III.

KANADI is a GARP-domain transcription factor that is abaxially expressed by a mutually antagonistic interaction with HD-ZIP III, and it is this juxtaposition of expression domains that is thought to define the plain for lamina outgrowth (Kinder and

Timmermans, 2011). Few studies have investigated the evolutionary history of the

KANADI gene, although it has been suggested that its evolution has played a fundamental role in the evolution of the plant body plan (Yamaguchi et al., 2012).

MicroRNA166/165 is a regulator of HD-ZIP III that has a complementary 21- nucleotide binding site that acts with ARGONAUTE1 and other microRNA machinery to direct cleavage of target HD-ZIP III mRNAs (Kinder and Timmermans; 2011). It is expressed abaxially, thereby post-transcriptionally excluding HD-ZIP III expression from the abaxial domain (Bowman, 2004). MicroRNA165 differs from 166 by one nucleotide, and separate functions of the two microRNAs are not highly evident (Maher et al., 2006)

By fully sampling the genomes of multiple species for genes that interact together in a pathway, the following specific questions can be addressed:

1. Are duplications of the genes noted above correlated with known paleopolyploid events?

2. Do genes in a pathway show stoichiometry of paralogs in the genome?

3. Can co-subfunctionalization among different gene families be deduced from similarly timed duplications or losses?

Also, by sampling the genome of the carnivorous plant, Utricularia gibba, a plant with highly aberrant leaf, embryo, and shoot development and a debated absence of roots, these questions can be addressed:

4. Is a plant with an extremely different morphology and development constrained to the same gene stoichiometry as other, more morphologically ‘typical’ plants?

5. Can any gene losses be attributed to the loss of typical root development?

6. Do the sampled gene families inferred to have a role in either carnivory or the highly modified morphology of U. gibba show similar fractionation patterns as the whole genome?

Methods

Sequence Retrieval

Arabidopsis sequences were downloaded from The Arabidopsis Information

Resource (TAIR) based on the curated names of CORONA (CNA), HOMEOBOX8 (HB-

8), REVOLUTA (REV), PHABULOSA (PHB), PHAVULOTA (PHV). TBLASTn searches of the other angiosperm genomes were performed with queries of conserved domains of the three genes and the binding site of the microRNA. Amborella trichopoda and

Utricularia gibba searches and sequence downloading were performed at CoGe genome sequence database (genomevolution.org). Vitis vinifera, Aquilegia caerulea, and Mimulus guttatus searches and sequence downloading were performed at

Phytozome (phytozome.org).

HD-ZIP III tBLASTn searches were performed with the conserved homeodomain- leucine zipper region of the Arabidopsis HD-ZIP III paralogs. Matching sequences were further pared down by accessing the results by sets of genes with similar e-values

(around -15) and correct HD-ZIP III exon-intron structure.

MicroRNA 166/165 searches were performed with the ultra-conserved 19- nucleotide sequence “TCGGACCAGGCTTCATTCC” at high e-value settings of -0.1 and -0.01 to account for the short sequence. Matching sequences were further pared down by the criteria of not being in predicted coding regions, as microRNAs should not code for proteins, and having a loosely conserved matching sequence within 200 bp that would correspond to the complementary binding site in the preprocessed RNA hairpin stem.

ZPR tBLASTn searches were performed with the conserved leucine zipper domain, “IIRENEKLKKKALLLHQENKTL”.

KANADI tBLASTn searches were performed with the first half of the conserved

GARP domain, “APRMRWTTTLHAHFVHAVELLGGHE”, located on the N-terminus of an exon-intron splice site in the middle of the GARP domain. E-value parameters were set to E=0.001 in CoGe and -1 in Phytozome. Matching sequences were further filtered down by accessing the results by sets of genes with similar e-values and correct

KANADI exon-intron structure.

Deeper relationships of KANADI and GARP-domain genes of other green plants were investigated by searching for homologues in the genome of the moss,

Physcomitrella patens, on the Phytozome website, as well as the transcriptomes of the green algae, Chara vulgaris and Coleochaete irregularis, from the One Thousand Plant

Transcriptome project (1KP; www.onekp.com) data set. Previously published GARP sequences from the protozoan, Giardia lamblia (Sun et al., 2006), were downloaded from GenBank.

Alignment

Translation of the nucleotide sequences of the three sets of genes, HD-ZIP III,

ZPR, and KANADI, to amino acid sequences was done Geneious version 5.6.5

(Biomatters, www.geneious.com) with the Geneious progressive pairwise alignment algorithm and then manually aligned by highly favoring homology based on amino acid polarity and conservatively using gaps based on the assumption that amino acids evolve by transmutations of whole codons rather than insertions/deletion of individual sites.

Phylogenetic Analyses

The amino acid alignments were phylogenetically analyzed using RAxML

(RAxML-VI-HPC) (Stamatakis, 2006) with the GTR + I + Gamma model of amino acid sequence evolution to account for the evolutionary distance of the sampled taxa

(Brocchieri, 2001). The topology with the highest maximum likelihood score was generated, and support for that topology was assessed by 1000 bootstrap replicates using the thorough bootstrap algorithm.

Arabidopsis Microarray Expression

Expression patterns of 10 of the 13 paralogs from the three coding genes are present on the Arabidopsis AtGenExpress Visualization Tool (Schmid et al., 2005).

Locus identifiers for HD-ZIP III (At5g60690, At4g32880, At1g30490, At1g52150,

At2g34710); ZPR (At2g45450, At3g52770); KANADI (At4g17695, At5g42630,

At5g16560) genes were entered into the website, and expression charts were generated and printed from the website.

Results

Paralog Numbers

The number of paralogs of each gene increases with the known paleopolyploid events in the species history. Amborella, the plant with the fewest paleopolyploid events, has the smallest number of each of the genes, and Utricularia, the plant with the most paleopolyploid events, has the largest number of genes, despite its small genome size (Table 3- 2). The number of copies of HD-ZIP III and miR166 in the genome of the six sampled species correlates well, with a correlation coefficient of 0.961 (p-value =

0.0021) (Figure 3-3A). The number of copies of HD-ZIP III and KANADI correlated equally well, with a correlation coefficient of 0.961 (p-value = 0.0022) (Figure 3-3B).

However, the number of copies of HD-ZIP III and ZPR does not correlate, with a correlation coefficient of 0.195 (p-value = 0.7106) (Figure 3-3C). Mimulus shows a reduction in the number of ZPR and KANADI paralogs with one paralog of KANADI showing signs of pseudogenesis. The number of miR166/miR165 precursor copies found in the genomes did show a steady increase in genomes (Figure 3-3D).

Pathway Proportionality

The proportions of each of the genes, defined as the total number of copies of a gene, divided by the total number of copies in the pathway, in the six sampled plant genomes are very similar (Table 3-3, Figure 3-2). The microRNAs show the highest proportion in each of the genomes, varying between 0.38 - 0.43, with an average of 0.4.

KANADI and HD-ZIP III have the exact same number of paralogs in four of the six genomes (Amborella, Aquilegia, Vitis, and Utricularia) with proportions between 0.17 and 0.27, and an average of 0.22 and 0.2, respectively. ZPR had the smallest

proportion of the genes in all six genomes, varying between 0.14 - 0.24, with an average of 0.17.

Lineage Evolution

HD-ZIP III was inferred to have undergone nine duplication events and three losses in the sampled lineages, with four of the duplications and all three of the losses happening in the Utricularia lineage (Figure 3-4) and one loss occurring in the CNA lineage, leaving only one paralog in its genome from the C8 lineage (Figure 3-5).

ZPR was inferred to have undergone 10 duplication events and four losses; one of the losses was inferred to be a loss of a major lineage at the base of the sampled asterids, with only one loss in the Utricularia lineage (Figure 3-4).

The sampled KANADI sequences were inferred to be a separate lineage from other GARP-domain genes based on phylogenetic distance from other Arabidopsis

GARP-domain genes and similar green algae and protozoan sequences (Figure 3-6).

KANADI was inferred to have experienced 10 duplication events and two losses, with one duplication at the base of the rosids and a subsequent loss of one duplicate paralog in the Arabidopsis lineage. Five of the duplications that were inferred to have occurred in the Utricularia lineage were restricted to the Group II KANADI lineage. The one loss in the Utricularia lineage left the genome without any paralogs of the Group II KANADI lineage (Figure 3-7).

MiR166/165 sequences show high amounts of variation between the two binding sites in the prepossessed hairpin structure, prohibiting an alignment with confident homology assessments and subsequent phylogenetic analyses (Figure 3-8). It was evident, however, that the miR165 sequence evolved recently in the Arabidopsis lineage. Homology between the genomic regions containing the microRNAs was

possible by comparing the synteny of the surrounding genomic regions (Figure 3-9). In every genome sampled, there was evident homology between an area of the genome that contained a tandem duplication of two of the microRNA 166 precursors (Table 3-3).

In Arabidopsis this area was where miR166C & D are found on chromosome 5, separated by 1.8 kb.

The ZPR phylogeny showed two main lineages in angiosperms with a loss of one of those lineages in the sampled species from the asterids (Figure 3-10).

Discussion

This study has shown that polyploidy was indeed acting as a driver of genetic innovations in this sampling of interacting genes, especially in the HD-ZIP III gene family. Many of the duplicate paralogs that came about as a result of paleopolyploid events have persisted in angiosperm genomes. The other gene families, ZPR and KAN, show more duplications that do not correlate with paleopolyploid events, specifically the duplications in the common ancestors of the Aquilegia, rosid, and Mimulus lineages.

However, the phylogenies that were used to infer the evolution of those genes were not well supported due to the limited sizes of the conserved domain used in the analysis.

Other methods of assessing homology and evolutionary events in these genes, like gene synteny and phylogenetic analysis of introns, may help better resolve the relationships.

The finding of conserved ratios of the 4 gene families in the sampled angiosperm genomes is surprising considering the genomic rearrangements, evolutionary distance, and morphological differences of the sampled taxa. In Utricularia this is especially unexpected because of the fractionation that has led to massive genomic shuffling. One could have also expected that the asterid species, Mimulus and Utricularia, would have

been unconstrained in HD-ZIP III pathway stoichiometry because mutants in the model asterid species Antirrhinum majus have been hypothesized to rely on an alternative pathway for leaf polarity (Kinder and Timmermans, 2007).

Deducing subfunctionalizations of interacting genes and previously unknown gene functions was possible in this data set with these methods. Perhaps the most exciting and obvious of these deductions is the role of certain lineages in the development of roots. Because Utricularia lacks true root development, we can hypothesize that the loss of the CNA lineage of HD-ZIP III, Clade I of ZPR, and Group II of KANADI is related to the loss of a root-specific developmental pathway. I further corroborated this hypothesis by examining the expression levels of these genes in unduplicated forms in Amborella and in the better sampled Arabidopsis microarray data.

The preduplicated homolog of CNA and HB8 is the CNAHB8 homolog in Amborella

(and Aquilegia). In the transcriptome data of Amborella, CNAHB8 has by far the highest number of reads sampled from the root tissue (Figure 3-11). CNAHB8 is also highly expressed in the stem of Amborella, suggesting that the preduplicated homolog is patterning vascular development in both the stem and roots. In the Arabidopsis expression data, CNA is the most highly expressed HD-ZIP III paralog in root tissues

(Figure 3-12). Clade I of the ZPR family was also lost in the genomes of both Mimulus and Utricularia. Only two of the four ZPR paralogs were present in the Arabidopsis microarray sampling, but those two paralogs were from Clades I and II of ZPR, with the

Clade I paralog more highly expressed in the root and stem tissues (Figure 3-13).

Group II of KAN was present in all sampled genomes except Utricularia. In the

Arabidopsis expression data, three of the four paralogs were present, and each of the

paralogs was from a separate group. The lost Utricularia paralog of KAN corresponds to the most highly expressed root paralog in the Arabidopsis expression data. The expression of these KAN paralogs shows a clear case of subfunctionalized expression of this group in the roots of Arabidopsis (Figure 3-14); however, the mutant of this paralog, aberrant testa shape, is well characterized as an ovule developmental mutant

(McAbee et al., 2006).

The genomic losses and Arabidopsis expression patterns of those paralogs provide evidence of subfunctionalization of those lineages for root development. This is a substantial finding, as root development has not been described for any of these gene families, even the well-characterized HD-ZIP III (Prigge et al., 2005), and the description of root development is commonly ignored in mutant screens. However, root development is controlled by a meristem just as shoots are, and it would follow that similar genes have been subfunctionalized to control development of this important land plant organ. This finding also carries great weight on the age-old debate of morphological homology in Lentibulariaceae. The morphological definitions of the organs of Utricularia have been debated by every major plant morphologist from the beginning of modern botany (see Rutishauser and Isler, 2001). A modern “fuzzy” definition of its organs has been proposed, with a mix of leaf, stem, and root structural definitions used to explain the organs in the genus. Based on this study, it seems that the developmental genetics of roots has been lost and that the root-like stolons found in the genus are not produced by the conserved angiosperm root-specific paralogs of the studied developmental genes. The modern morphological definition of Utricularia does seem to be correct in assuming that Utricularia has undergone an extensive

canalization and a subsequent decanalization (much like the “fractionation” of Ibarra-

Lacette et al. (2013)) of its genome that is reflected in a “decanalization” of its morphology (Rutishauser and Isler, 2001).

This study has shown the utility of whole genome sequencing from phylogenetically important taxa for understanding gene function and evolution, particularly of genes controlling aberrant morphologies, such as those in carnivorous plants. The many copies of genes in angiosperm genomes can make it difficult to find a specific mutant that shows a phenotypic difference and therefore implicate it in the development of a specific part of the plant because of functional overlap of gene duplicates. This limitation makes the approach used in this study a valuable alternative for identifying genes with specific phenotypic effects. This study has also shown that a more holistic view of gene evolution, by including interacting genes and investigating losses as well as duplications, can lead to a better understanding of plant evolutionary development

Table 3-1. Number of paralogs of the 4 gene families in the 6 sampled angiosperm genomes. Total number of genes, calculated by adding the number of all 4 gene families for that plant, is reported in column 6. The number of paleopolyploid (PP) events known in each species’ history is given in column 7, and the total number of genes in each species’ genome is given in column 8. HD- ZPR KANADI miR166 TOTAL # of Total

ZIP III PP # of

events genes

Amborella 3 2 3 6 14 1 26.8k

Aquilegia 3 4 3 7 17 1 24.8k

Vitis 4 4 4 8 20 2 26.3k

Arabidopsis 5 4 4 9 22 4 27.4k

Mimulus 7 3 6 10 26 3 25.5k

Utricularia 7 4 7 11 29 5 28.5k

Table 3-2. Proportions of the number of paralogs for each of the 4 gene families in the genomes of the 6 sampled angiosperm genomes. Proportions were calculated by dividing the number of paralogs for each gene family by the total number of genes in the 4 gene families. The average proportions among the 6 genomes are reported in row 8. HD-ZIP III ZPR KANADI miR166

Amborella 0.21 0.14 0.21 0.43

Aquilegia 0.17 0.24 0.17 0.41

Vitis 0.2 0.2 0.2 0.4

Arabidopsis 0.23 0.18 0.18 0.41

Mimulus 0.27 0.16 0.23 0.38

Utricularia 0.24 0.14 0.24 0.4

Average 0.22 0.17 0.2 0.4

Table 3- 3. Places in the genomes of the sampled angiosperm genomes that contain the tandem precursor miR166 sequences. Species Location in genome assembly

Amborella Scaffold 128

Aquilegia Scaffold 69

Vitis Chromosome 7

Arabidopsis Chromosome 5

Mimulus Scaffold 76

Utricularia Scaffold 127

Figure 3-1. Phylogeny of major Angiosperm lineages and eudicot orders with the placement of sampled taxa. Based on Angiosperm Phylogeny Group III (APG III) (Bremmer et al., 2009)

Figure 3-2. Phylogeny and ancestral genome evolution of sampled angiosperms and the number of copies of each of the sampled genes in their genome. Starbursts represent whole genome duplication events.

Figure 3-3. Bar graphs comparing the number of copies of HD-ZIP III with the number of copies of the genes that regulate it in the genomes of the six sampled species. A) HD-ZIP III compared with miR166. The number of copies of the two genes correlates well, with a correlation coefficient of 0.961 (p-value = 0.0021). B) HD-ZIP III compared with KANADI. The number of copies of the two genes correlates well, with a correlation coefficient of 0.961 (p-value = 0.0022). C) HD-ZIP III compared with ZPR. The number of copies of the two genes does not correlate well, with a correlation coefficient of 0.195 (p-value = 0.7106). D) Stacked bar graph showing the sum of all four genes added together.

Figure 3-4. Gene trees and species trees of the four leaf polarity genes in the six angiosperm genomes. Gold stars represent gene duplication events and red octagons represent gene losses.

Figure 3-5. Phylogram of sampled HD-ZIP III genes. Blue lines represent Amborella genes; red represents Aquilegia; purple represents Vitis; green represents Arabidopsis; orange represents Mimulus, and yellow represents Utricularia. The labeling of the main clades adapted from Prigge and Clark (2006).

Figure 3-6. Unrooted phylogeny of sampled GARP and KANADI genes from green plants and a protozoan.

Figure 3-7. Unrooted phylogram of sampled KANADI genes. Blue dots represent Amborella genes; red represents Aquilegia; purple represents Vitis; green represents Arabidopsis; orange represents Mimulus. Yellow represents Utricularia.

Figure 3-8. Alignment of miR166 from the genomes of Amborella, Aquilegia, Vitis, and Arabidopsis. Identity graph displays the agreement between sequences. The mature microRNA binding site is in 100% agreement at sites 408 – 421 of this alignment. The complimentary binding site in the precursor hairpin structure is in high agreement at sites 245 – 276.

Figure 3-9. Syntenic regions of the Amborella and Aquilegia genomes that contain a discovered miR166 copy. Part of the genome assembly of Amborella is plotted on the top half and the genome assembly of Aquilegia on the bottom half. Horizontal arrows represent areas of coding genes. Vertical red line between genomes represent highly similar regions of coding sequence. The two starbursts connected by a vertical line represents the areas of the genome containing the discover miR166 copy. The area of the genome containing the microRNA is not coding (intergenic).

Figure 3-10. Unrooted phylogram of sampled ZPR genes. Blue dots represent Amborella genes; red represents Aquilegia; purple represents Vitis; green represents Arabidopsis; orange represents Mimulus. Yellow represents Utricularia.

Figure 3-11. Expression from 454 read counts from Ancestral Angiosperm project (http://ancangio.uga.edu/) of Amborella HD-ZIP III paralogs. The sampled tissues are grouped on the x-axis. The number of reads in the assembled HD-ZIP III paralogs found in Amborella is plotted on the y-axis. The number of reads is indicative of the expression level of that paralog in the tissue. The CNAHB8 paralog is the highest expressed paralog in the root, stem, and female flower tissue, however Utricularia has lost the CNA paralog.

Figure 3-12. Expression from microarray intensities of Arabidopsis HD-ZIP III paralogs. Different tissue samples are plotted on the x-axis. The intensity of light signal given off by the binding of copies of the expressed gene to the gene probe on the microarray is plotted on the y-axis. The intensity of light signal is representative of gene expression level in that sample. The expression of the five paralogs of HD-ZIP III is here shown to vary depending on the tissue type. CNA is highly expressed in the root and stem tissues and is the paralog lost in the Utricularia genome.

Figure 3-13. Expression from microarray intensities of Arabidopsis ZPR paralogs. Different tissue samples are plotted on the x-axis. The intensity of light signal given off by the binding of copies of the expressed gene to the gene probe on the microarray is plotted on the y-axis. The intensity of light signal is representative of gene expression level in that sample. The expression of the two paralogs of ZPR is here shown to vary depending on the tissue type. The ZPR paralog from Clade I is highly expressed in the root and stem tissues, however there are no Mimulus or Utricularia Clade I paralogs in their genome.

Figure 3-14. Expression from microarray intensities of Arabidopsis KANADI paralogs. Different tissue samples are plotted on the x-axis. The intensity of light signal given off by the binding of copies of the expressed gene to the gene probe on the microarray is plotted on the y-axis. The intensity of light signal is representative of gene expression level in that sample. The expression of the three paralogs of KANADI is here shown to vary depending on the tissue type. The KANADI paralog from Group II is the highest expressed paralog in the root, however there are no Utricularia Group II paralogs in their genome.

CHAPTER 4 LEAF EVO-DEVO OF THE OF CARNIVOROUS PITCHER PLANT, CEPHALOTUS FOLLICULARIS (CEPHALOTACEAE)

Introductory Remarks

Charles Darwin saw the phenomena of carnivorous plants as an evolutionary puzzle in which natural selection of structures in plants could explain their evolution. He extensively studied the natural variation in glands of carnivorous plants and the adaptations of their structure (Darwin, 1875). Today we have the opportunity to understand the evolution of carnivorous plants from the point of view of the genetic changes that produced their odd structures. With genetic research in model plants and current genetic techniques, we can test hypotheses about what genetic regulatory changes took place during the evolution of carnivorous plants and better understand the homology of structures.

Because of the extensive morphological change that occurred during the evolution of carnivorous plants, placing them in plant classification and knowing their closest non-carnivorous relatives has been historically difficult. Prior to classifications based on explicitly constructed phylogenetic hypotheses, deducing the morphological homology of structural features of carnivorous plants to those of non-carnivorous plants proved challenging (Franck, 1976). With molecular phylogenetics we now know that carnivory evolved a minimum of four times in angiosperms (Albert et al., 1992; Soltis et al., 2005). Despite their separate origins from distantly related angiosperms, these instances of independent evolution reveal strong similarities in native habitats and trapping mechanisms among carnivorous plants.

Although the different groups of carnivorous plants are found in almost all areas of the globe, the type of habitats they live in is very similar. Carnivorous plants typically

grow in mesic soils that are low in nutrients, particularly nitrogen (D’Amato, 1998).

Carnivorous plants obtain nutrients from insects that they capture, while still carrying out photosynthesis for energy (D’Amato, 1998). The multiple, independent origins of carnivorous lineages can therefore be seen as an adaptation to similar environmental pressures.

The change to carnivory typically has involved large structural changes to leaves, resulting in flypaper traps, snap traps, lobster pot traps, and pitchers (Soltis et al.,

2005). A given trap type may be found in multiple distantly related clades, with pitcher traps having originated in three different groups including Sarraceniaceae in the

Ericales, Nepenthaceae in the Caryophylalles, and Cephalotaceae in the Oxalidales.

Thus, the structural evolution of carnivorous plants is an interesting study system because there is both convergence producing similar yet separate carnivorous groups and divergence producing variation within the groups.

The modern interpretation of all carnivorous pitchers, including Cephalotus, is that the unifacial epiascidiate tubular structure of the pitcher (Figure 4-1, Figure 4-2,

Figure 4-3) is achieved through a developmental sequence similar to that of peltate leaves with outgrowths of the margins of the leaf growing towards the apex and uniting to form a ring in cross section (Franck, 1976). Various interpretations of the homology of the pitcher lid have been made, including as an extension of the collar (Arber, 1941), as a congenitally fused pair of pinnae of a pinnately compound leaf (Troll, 1939; Froebe and Baur, 1988), and as a ligule (Eichler, 1881).

Molecular genetics of the past two decades have provided insights into the control of the development of leaf shapes (Tsukaya, 2006; Yamaguchi et al., 2012), with

the mechanism of the bifacial polarity of the leaf being closely connected to the lamina’s morphological shape and marginal growth (Waites and Hudson, 1995; Kinder and

Timmermans, 2010). Gleissberg et al. (2005) examined this model of the genetic control of leaf polarity in the peltate leaf of Tropaeolum majus (Tropaeolaceae) and found that the marginal outgrowth that unites the leaf edges and forms the peltate structure shows expression and shows expression of the abaxial identity gene, YABBY. YABBY is a gene family with five well-characterized paralogs in Arabidopsis, including CRABS

CLAW (CRC), YABBY2 (YAB2), YABBY5 (YAB5), INNER NO OUTER (INO), and

FILAMENTOUS FLOWER (FIL), and all five have been shown to control the abaxial identity of lateral organs in Arabidopsis (Siegfried et al., 1999).

Based on the leaf polarity model and evidence of how the model works in other unifacial leaves, I hypothesize that the epiascidiate leaves of carnivorous pitchers are also produced by marginal outgrowths associated with abaxial identity gene expression.

I have chosen to test this hypothesis by investigating the expression of a YABBY gene homolog in Cephalotus follicularis. Cephalotus belongs to the monotypic family,

Cephalotaceae, which is a member of the Oxalidales, making it the most closely related pitcher plant to Brassicales, the order to which both Arabidopsis thaliana (Brassicaceae) and Tropaeolum majus (Tropaeolaceae) belong.

To test the hypothesis that the pitcher development in Cephalotus is associated with a changed YABBY expression domain, I first isolated and described the expression domain of the HistoneH4 gene as a common positive control for the in-situ protocol

(Gleissberg et al., 2005; Yamaguchi et al., 2012). As this is the first reported in situ study for Cephalotus and carnivorous plants are known to be difficult material for

molecular work (Fleischmann and Heubl, 2009), this is an important first step for collecting and interpreting in situ hybridization data. Next, YABBY genes were isolated and their orthology deduced from the phylogenetic placement with previously identified

Arabidopsis and Oryza sativa orthologs (Yamaguchi et al., 2012). The expression of the isolated Cephalotus FIL ortholog (CephFIL) was then used to identify the adaxial genetic identity domain in the serial longitudinal section of the developing Cephalotus pitcher, as was done in the peltate leaf of Tropaeolum (Gleissberg et al., 2005). If

CephFIL expression is seen in the outgrowths that fuse to form the lid and the adaxial region of the petiole, then I can conclude that the pitcher is genetically similar to peltate leaves. However, if no CephFIL expression is seen in the lid or adaxial petiole region, then I would conclude that the lid is an outgrowth of the adaxial region and that it and the petiole are not formed by the same genetic causes as peltate leaves.

Methods

RNA Extraction and Gene Amplification

Total RNA was isolated from shoot tips and young pitchers of Cephalotus plants using Plant RNA Isolation Reagent (Invitrogen, Carlsbad, USA), with subsequent treatment with DNase I (Invitrogen). Total RNA (1 μg) was used for first-strand cDNA synthesis using the SuperScript III first-strand synthesis system (Invitrogen), and 0.5 μL of this reaction was used as the template for PCR amplification. Target genes were amplified using degenerate primers (Table 4-1). Amplified DNA fragments were gel- extracted and cloned into pCRII-TOPO (Invitrogen), and at least 16 clones per fragment were sequenced.

Sequence Alignment and Phylogenetic Analysis

HistoneH4 and YABBY clones were sequenced with forward and reverse plasmid cloning site primers, M13F and M13R. The forward and reverse sequences of each gene were then aligned together and checked for agreement. The HistoneH4 sequences were then translated and aligned with the translated sequence of

Arabidopsis thaliana HistoneH4 sequence downloaded from TAIR

(http://www.arabidopsis.org, AT2G28740.1). The YABBY sequences were translated and aligned with sequences from Arabidopsis thaliana, Oryza sativa, and Amborella trichopoda, representing the five major lineages of YABBY genes (Bartholmes et al.,

2012) using Geneious version 5.6.5 (Biomatters, www.geneious.com) with the alignment algorithm developed by Biomatters and implemented in Geneious. The

YABBY amino acid alignment was phylogenetically analyzed using RAxML (RAxML-VI-

HPC) (Stamatakis, 2006) with the GTR + I + Gamma model of amino acid sequence evolution. The topology with the highest maximum likelihood score was generated, and support for that topology was assessed by 1000 bootstrap replicates using the thorough bootstrap algorithm.

In situ Hybridization

Probe construction

Plasmids containing the sequenced HistoneH4 and the CephFIL paralogs were amplified using the M13 priming sites in the plasmid and cleaned using a Wizard SV

PCR clean-up kit (Promega, Madison, USA). Appropriate sample dilutions were then used for RNA probe synthesis using a DIG RNA labeling kit with a T3 and SP6 RNA polymerase (Roche, Indianapolis, USA).

Tissue fixation

Shoot apices and young pitchers were fixed in 3% paraformaldehyde and 0.25% glutaraldehyde in 0.1 M sodium phosphate buffer for 16 hours at 4°C. Tissue was then dehydrated in a graded ethanol series, replaced with xylene, and embedded in

Paraplast Plus (Oxford Labware, Mansfield, Massachusetts, USA).

Hybridization

In situ hybridization was performed with a protocol adapted from Kramer (2005) and Kouchi and Hata (1993). Microtome sections of paraplast-embedded material were cut at 7 μm and mounted on Probe-On-Plus Slides (Fisher, Waltham, MA). Slides were de-waxed and re-hydrated in increasing dilutions of ethanol, and proteins were digested in proteinase K (1 μg/mL) at 37 °C for 40-60 minutes in 100 mM Tris-HCl (pH 7.5), 50 mM EDTA and then washed in distilled water for 5 minutes. After washing in 0.2% phosphate-buffered glycine solution and phosphate-buffered saline (PBS: 7 mM

Na2HPO4, 3 mM NaH2PO4, 130 mM NaCl, pH 7.0), the tissue was re-fixed in 4% paraformaldehyde. Positive charges were masked by treatment with 16 mM acetic anhydrate in 0.1 M triethanolamine buffer. DIG-labeled probes were used for hybridization at a concentration of 2.5 ng in hybridization buffer consisting of 50% deionized formamide, 0.3 M NaC1, 10% dextran sulfate, 10 mM TRIS- C1,pH 7.4, 1mM

EDTA, 100mM DTT and 500~tg/ml Escherichia coli tRNA. Sections were incubated at

42° C for more than 16 h with coverslips in hybridization buffer (40-50 gl per slide) containing the probes. The probe was detected by an alkaline-phosphatase-coupled anti-digoxigenin-antibody and visualized by NBT/X phosphate blue color substrate.

Results

The isolated HistoneH4 transcript from Cephalotus that was used as the template for the in situ hybridization probe showed 100% amino acid sequence identity with that of the Arabidopsis thaliana amino acid sequence at 72 amino acid residues which accounts for 68% percent of the 105 amino acid long protein (Figure 4-4).

Isolated YABBY transcripts from Cephalotus were phylogenetically placed in two of the major lineages, FIL and CRC (Figure 4-5). All major lineages of YABBY genes were recovered in the phylogeny, although not with high bootstrap support. In addition to overall sequence-based orthology assessment, the amino acid sequences of the

Cephalotus transcripts show characteristic motifs of the lineages to which they belong.

The HistoneH4 transcripts were localized in areas of active cell division, as this gene is expressed during the S phase of the cell cycle (Gaudin et al., 2000).

Specifically, transcripts were detected as bluish-purplish signal in all major areas of the developing pitcher, including the inner (adaxial) region of pitcher, the tip of the pitcher, the median cells of the lid, the tip of the lid, and the median cells of the keel (Figure 4-

6).

The CephFIL transcripts were localized in the outer (abaxial) region of the pitcher but not in the abaxial keel. There was also no CephFIL expression detected in either the inner (adaxial) region of the pitcher or anywhere in the developing lid (Figure 4-7).

Discussion

These results do not support the hypothesis that the lid and outgrowing margins of the Cephalotus pitcher would show YABBY expression. This is surprising as the lid is the only part of the pitcher that is formed by an outgrowth and congenital fusion of the opposite leaf margins, like the growth that produces peltate leaves and the pitchers of

the other two pitcher plant lineages, Sarraceniaceae and Nepenthaceae (Franck, 1976).

However, these data do agree with previous interpretations of the developmental origins of the pitcher of Cephalotus. The lid has been described by Lloyd (1942) as “an outgrowth from the upper [adaxial] surface of the petiole below the pitcher proper while the pitcher has been produced by a ventro-dorsal invagination of the upper, more distal region.” Because of the CephFIL expression domain, I predict that the lid has genetic adaxial identity given that the abaxial identity gene is not expressed in the lid and abaxial-adaxial identity genes have mutually exclusive expression domains (reviewed in

Kinder and Timmermans, 2010). Eichler’s (1881) interpretation of the lid as a ligule also fits my data, as ligules are adaxial outgrowths.

As reviewed in Bartholmes et al. (2011), the expression domain of FIL has been documented in some of the model angiosperms and variation in the location of that domain is present. In the monocots it is the most labial, most likely because of the multiple paralogs that appear to have come from duplications of the gene prior to the monocot diversification. In Oryza sativa, the three paralogs YAB3, YAB4, and YAB5 have no expression in the leaf, the abaxial side of the leaf, and throughout the whole leaf, respectively. In the only isolated paralog of FIL from Triticum aestivum, expression is throughout the leaf. And in the maize, two of the paralogs are expressed adaxially, and the precise location of the third paralog is not known. In eudicots, however, FIL is expressed in the abaxial side of all taxa sampled including Arabidopsis thaliana,

Anthirrhinum majus, Streptocarpus rexii, and Lotus japonicus. Therefore, because

Cephalotus is a member of the eudicots, I am interpreting the expression of CephFIL to be in the abaxial part of the leaf.

No interpretations of the homology of the Cephalotus pitcher have been made in the years following the studies that have used molecular phylogenetics to place

Cephalotus in the Oxalidales. The exact location of Cephalotus in the Oxalidales has not been recovered with high support (Soltis et al., 2011) but it is related to

Cunoniaceae, Bruneliaceae, and Elaeocarpaceae. The fact that this pitcher plant is a rosid and that the lid arises at the base of the leaf and has adaxial identity would allow for the lid to be interpreted as a pair of fused stipules, as stipules are common in rosids.

Stipules in the Oxalidales are very diverse (Dickinson and Rutishauser, 1990), and there are no other structures on Cephalotus that have been interpreted as stipules. The recently described stipules at different growth stages of Geissois pruinosa

(Cunoniaceae) (Fogliania et al., 2009) show similarities (Figure 4-8) to the morphology of the lid in intermediate pitcher leaves in Cephalotus (Figure 4-2 c-f).

Interpreting the lid as a pair of stipules would also require a change in the proximal-distal axis since stipules are usually located at the base of the leaf. Mutants of the KNOX gene in maize show the gene controls the proximal-distal placement of the ligule with parts of the ligule moving farther towards the distal end of the leaf (Freeling,

1992). A similar change may be happening in Cephalotus to place the lid at the end of the petiole, near the pitcher. The placement of the stipules in the Oxalidales is labial, with stipules at the base of the petiole, as in Geissois pruinosa (Figure 4-8) or at the distal end of the petiole, as in Pseudoweinmannia lachnocarpa (Cunoniaceae)

(Dickinson and Rutishauser, 1990).

The stipules of Arabidopsis have been shown to have genetic adaxial identity with the trans-acting small interfering RNA Gene, TAS3, showing high expression

throughout the whole stipule (Garcia et al., 2006) (Figure 4-9). TAS3 is an indirect suppressor of FIL, acting through the abaxially expressed AUXIN RESPONSE FACTOR

3 (ARF3/ETT) and ARF4 (Garcia et al., 2006). Therefore, based on the model of adaxial-abaxial identity gene interactions and the lack of CephFIL expression in the lid and the predicted adaxial identity of the lid, I am hypothesizing that the lid of Cephalotus is genetically a pair of fused stipules. This interpretation would also mean that the pitcher is not genetically homologous to a peltate leaf and that its unifacial qualities are not produced by a cross-zone formation fusing the two abaxial margins together, as is seen in the other two independent evolutions of pitchers in Sarraceniaceae and

Nepenthaceae (Franck, 1976). In those two families the lid is at the distal end of the pitcher, indicating that they are most likely not of stipular origin, and thus marks a distinctive difference in the evolutionary changes of the leaf that happened during the convergent evolution of the pitcher in the three different angiosperm clades.

Table 4-1. Nucleotide sequence of primers used to amplify HistoneH4 and YABBY genes in Cephalotus follicularis. Primer Name 5’ to 3’ sequence dYABu2 GGGGAAGTGGGCCCARTTYTTNG dYABd1 TGTTCAAGACCGTGACCGTNMGNTGYGG dH4-d1 GGAAGGTGCTGCGGGAYAAYATHCA dH4-u2 CCGCTTCAGGGCGTACACNACRTCCAT

Table 4-2. GenBank sequences used in phylogenetic study of YABBY genes. Species Abbreviation GenBank Accession Numbers Pinus taeda Pinus DR100835 Amborella trichopoda AmbF2 AB168113 Amborella trichopoda AmbCRC AJ877257 Amborella trichopoda AmbF1 AB126654 Arabidopsis thaliana AtYAB2 AF136539 Arabidopsis thaliana AtYAB5 NM179750 Arabidopsis thaliana AtINO AF195047 Arabidopsis thaliana AtCRC AF132606 Arabidopsis thaliana AtYAB3 af136540 Arabidopsis thaliana AtFIL AF136538 Populus trichocarpa PopFIL XM_002329830 Oryza sativa OsDL AB106553 Oryza sativa OsYAB5 AB274017 Oryza sativa OsYAB4 AB274016 Oryza sativa OsYABBY1 AB274013 Oryza sativa OsYAB2 AB274014 Oryza sativa OsYAB6 AB274018 Oryza sativa OsYAB7 AB274019

Figure 4-1. Front view of Cephalotus follicularis pitcher and basal rosette of flat leaves.

Figure 4- 2. Photographs of Cephalotus pitcher, flat leaves, intermediate leaf, inflorescence, and flowers. The orientation of the pitcher (A, B) is depicted in the position that it emerges as a leaf primordium with the keel at the distal end and the lid attached at the proximal base of the petiole.

Figure 4-3. Longitudinal section drawing of Cephalotus mature pitcher and shoot tip. Left-most pitcher represents orientation of pitcher at maturity and the structures on either side of the shoot apical meristem represent developing pitchers (not to scale). Lid (L); Ascidium (A); Pitcher (P); Keel (K); and Shoot Apical Meristem (SAM).

Figure 4-4. Alignment of Cephalotus and Arabidopsis HistoneH4 amino acid sequences.

Figure 4-5. Phylogram of angiosperm and isolated carnivorous pitcher plant YABBY paralogs. Ceph22YAB represents the CephFIL sequence used as a probe for in situ hybridization. Sequence abbreviations explained in Table 4-2. Cephalotus clones represented by Ceph abbreviation. Sarracenia flava clones represented by Sar, and Nepenthes alata clones represented by Nep.

Figure 4-6. In-situ hybridization of HistoneH4 in developing Cephalotus pitcher. Purple staining indicates areas of HistoneH4 expression. Keel (K); pitcher (P); ascidium (A); Lid (L).

Figure 4-7. In-situ hybrizations of CephFIL in serial longitudinal sections of a developing Cephalotus pitcher. The median of the pitcher is shown first (A) and successive images represent serial sections proceeding to the margin of the pitcher (I). The early keel is located at the distal (top) of the pitcher with the ascidium in the center and the lid at the proximal end of the pitcher. Keel (K); lid (L); ventral side of pitcher (VP); dorsal side of pitcher (DP).

Figure 4-8. Stipules in the seedling of Geissois pruinosa (Cunoniaceae). Free stipules (FrSt) and fused stipules (FuSt) are indicated with arrows. Used with permission from Fogliani et al. (2007).

Figure 4-9. Expression of TAS3 in the vasculature, adaxial domain of developing leaves, and stipules of Arabidopsis thaliana. Used with permission from Garcia et al. (2006). Blue staining expression of inserted of GUS reporter- gene 37bp of TAS3 (AT3g17185).

CHAPTER 6 CONCLUSION

In conclusion, this dissertation has used the leaf polarity genetic model to explore the evolution of leaves throughout land plant evolution (Chapter Two), genome dynamics and the loss of roots in angiosperms (Chapter Three), and finally, the evolution of a highly modified and enigmatic carnivorous pitcher plant (Chapter Four).

These chapters reflect the ever-changing methods of science as most of this data was collected outside of the lab and required no funding. This research can be seen as just the beginning of our understanding of the evolution of the genes controlling leaf development and provides a base on which to explore these findings further.

The findings in Chapter Two are based on the newly available gymnosperm and monilophyte transcriptomes from the 1KP project. These data fill a substantial hole in the sequence data available across the spectrum of plants. However, even today genome sequencing of the gymnosperms and monilophytes is underway, and my findings will provide hypotheses to test in those genomes. Specifically, better understanding of the duplication that gave rise to the ZPR gene from HD-ZIP III will be possible with gymnosperm genome sequences. My findings predict that the ZPR gene will show genomic synteny with the PaleoHDZa paralog if ZPR did indeed originate from a duplication of PaleoHDZa. The exact function of the five different HD-ZIP III paralogs in the development of gymnosperms would also be a key question to address in the future. As only three of those paralogs were passed on to angiosperms, one wonders what the other copies do and if their function is similar to or different from HD-ZIP III functions in angiosperms. Answering functional questions, unlike questions of lineage evolution, currently requires a large amount of time and funding, but may be of special

interest to the timber industry as these genes most likely play a role in wood development in gymnosperms.

Chapter Three makes predictions about previously unreported functions of HD-

ZIP III in root development. Even though this dissertation set out to improve our understanding of leaf development, it is a welcome finding that carnivorous plants, especially Utricularia, might be natural models for root development as well. As a plant that gets its water and nutrients foliarly, Utricularia is, in a way, a root development knock-out. Testing the subfunctionalization of the sampled polarity genes by functionally characterizing their role in the model plant Arabidopsis may be of relative ease given that the mutants are easily and cheaply available from TAIR. Understanding the losses of the supposed root development paralogs in Utricularia may also be a plausible avenue of future research because there is a gradient of root loss in Lentibulariaceae

(the family to which Utricularia belongs). Furthermore, transcriptome or genome sequencing may be financially feasible for this question and many others about genome evolution in the family, as the genome size is very small.

Appendix A, if anything, is a testament to the difficulty that carnivorous plants pose to the techniques of molecular biology, ironically as they also did for morphologists in the past. Their nucleotide-binding secondary compounds, which are well documented to impede even the simplest of molecular biology techniques, including nucleotide extraction and amplification (Bekesiova et al., 1999; Fleischmann and Heubl, 2009), may be exacerbated in Cephalotus follicularis, the only known carnivorous plant in the

Rosidae, a clade of ~70,000 species (Soltis et al., 2005) also known to be difficult molecularly (Zamboni et al., 2008). However, understanding the expression domains of

leaf polarity genes in carnivorous plants is an important endeavor but one that may be better approached using laser dissection RT-PCR (Nelson et al., 2006) as secondary compounds can be removed after dissection during the RNA extraction (Johnson et al.,

2012). New in situ methods may also hold promise (Wang et al., 2013).

APPENDIX A FURTHER CARNIVOROUS PLANT EXPERIMENTS

AS2

ASYMMETRIC LEAVES2 (AS2) is a member of the AS2/LOB gene family and functions in both meristem maintenance and adaxial leaf identity (Iwakawa et al., 2007).

Knowing the expression domain of AS2 in carnivorous pitcher plants would have helped better understand genetically how the pitcher relates to the flat leaves. Phylogenetic analyses with Arabidopsis thaliana sequences Nepenthes alata, Cephalotus follicularis, and Cunonia capensis, a closely related non-carnivorous plant to Cephalotus, were performed to identify homologous transcripts for in situ hybridization probes (Figure A-

1). The sequences placed phylogenetically close to the Arabidopsis AS2 paralog, were used for designing PCR primers for isolating AS2 paralogs from Nepenthes and

Cephalotus, and synthesizing probes for in situ hybridization.

The Nepenthes AS2 in situ hybridizations that used Western Blue alkaline- phospotase conjugation to detect the DIG-labeled probes, showed high background noise and identical staining patterns to the negative control sense probes (Figure A-2).

The only published studies of gene expression in Nepenthaceae used a fluorescent detection of DIG-labeled probes (Schulze et al., 1999; An et al., 2001; An et al., 2002) presumably because the method produces less background noise. This method was used to try to detect the Nepenthes and Cephalotus AS2 probes, however no signal could be detected in Nepenthes meristem, young developing pitcher, or older developing pitchers (Figure A-3). This detection method was also tried on Cephalotus meristem and young leaf primordia (Figure A-4) but no signal was clearly seen. No signal could be detected in the developing pitcher either (Figure A-5). 99

VIGS Development

The best way to know how leaf development pathways are functioning in carnivorous pitcher plants is by knocking down the expression of the pathway genes to see what affect it has on the pitcher development. On of the most common ways for do this in non-model plants is by a method called Virus Induced Gene Silencing (VIGS)

(Gould and Kramer, 2007). This method has not been tried on carnivorous pitcher before. One of the first step is developing this method is finding a strain of

Agrobacterium that can successfully insert a virus vector into the genome of the target plant. I tested the ability of Agrobacterium tumefaciens AGL1 to insert the commonly used VIGS virus vectors, TRV1 and TRV2, into the genome of Nepenthes alata,

Sarracenia flava, Cephalotus follicularis, as well as the close relative of Sarracenia,

Actinidia chinensis, and the close relative of Cephalotus, Oxalis corniculata. The protocol described in Gould and Kramer (2007) was used, specifically the vacuum infiltration method. To test if the vectors were successfully inserted into the genome, a

DNA extraction was performed 5-days after Agrobacterium infection and tested with

PCR for the presence of TVR1 and TRV2 vectors. The PCR results showed that TRV vectors were best inserted into the genomes of Sarracenia Alta, Oxalis corniculata, and

Actinidia chinensis (Figure A-6). The vector was not well inserted into the genomes of

Cephalotus, or Nepenthes (Figure A-6).

100

At LBD42 45 At LBD40 43 At LBD41 Nep2106 At LBD39 91 At LBD37 47 At LBD38 Nep262 34 63 At LBD7 At LBD2 54 87 Cun2373 66 At LBD13 85 At LBD15 Cun6000 97 42 At ASL5 100 At LBD23 At LBD24 Cun8091 65 Nep5918 86 At LBD1 44 At LBD11 49 Cun4539 21 Cep4314 34 Cun6422 34 At ASL9 42 Cep7354 91 Nep05492 47 Nep0549 40 At LBD4 30 Cun0023 Cep6 68 18 At LBD33 49 Nep9509 Nep0352 43 At LBD16 33 82 At LBD14 At LBD29 At LBD20 35 At LBD18 69 38 Nep1498 81 At LBD31 49 At LBD19 13 Cun5843 89 Cun6684 58 At LBD25 76 Cep46191 89 At LOB 21 Cep6724 37 Cun5688 20 Cep8912 Cep6687 94 56 At LBD21 37 Nep674 44 Cep2499 Cun8923 26 At ASL1 38 At LBD10 79 At LBD35 22 At LBD32 At LBD28 Nep8002 58 45 At AS2 20 Cep1993 16 Cun1049 Cep4601 Figure A-1. Phylogeny of Arabidopsis AS2/LOB paralogs and AS2/LOB BLAST results form 1KP (www.onekp.com) transcripts for Nepenthes alata (abbreviated Nep), Cunonia capensis (Cun), and Cephalotus follicularis (Cep). Phylogeny generated by RAxML from an amino acid alignment with GTR+I+G model parameters. Numbers at the node indicate percentage of 1000 bootstrap trees that agree with this topology.

101

Figure A-2. Longitudinal section of in situ hybridization of Nepenthes alata AS2. A) section through the meristem (arrow head) and leaf primordium (arrow) with putative expression signal from anti-sense probe in the epidermal layers of the meristem and the adaxial side of the primodium. B) section through base of a pitcher sheathing around the meristem. C) section through the meristem and developing pitcher on the right side of the meristem. D) section a few section distal of the meristem with expression signal from the sense probe as a negative control

102

Figure A-3. Longitudinal sections of fluorescent In situ hybridization of Nephenthes alata AS2. A-F probe made from anti- sense reverse-transcription. G-J probe made from sense reverse-transcription. A-D) sections through the shoot apical meristem with a leaf primordium on the left side of the meristem. E-F) sections through the tip of a developing pitcher. G-H) sections through meristem. I) section through tip of developing pitcher. J) section through developing tip of pitcher showing the lid, ascidium, and peristome.

103

Figure A-4. Longitudinal sections of fluorescent in situ hybridization of Cephalotus AS2. A-F) sections through the shoot apical meristem and young leaf primordia. Sections probed with anti-sense reverse-transcription CephAS2 probe. Successfully bound probe is expected to be seen as red but in these sections most fluorescence is presumed autofluorescence from the tissue

104

Figure A-5. Longitudinal sections of fluorescent in situ hybridization of Cephalotus AS2. A-L) Developing Cephalotus pitcher in the same developmental stage and position as in Figure 4-7. Sections probed with anti-sense reverse- transcription CephAS2 probe. Successfully bound probe is expected to be seen as red but in these sections most fluorescence is presumed autofluorescence from the tissue

105

Figure A-6. Electrophoresis gel visualization of TRV1 and TRV2 PCR’s of Agrobacterium infected plants. Each well represents a different plant sample from which the DNA was extracted and used as the template for the PCR. Presence of a band in the well indicates presence of the vector in the genome.

106

LIST OF REFERENCES

Albert, V. A., Williams, S. E., & Chase, M. W. (1992, September). Carnivorous plants: phylogeny and structural evolution. American Association for the Advancement of Science.

Alvarez-Ponce, D., & Fares, M. A. (2012). Evolutionary rate and duplicability in the Arabidopsis thaliana protein–protein interaction network. Genome biology and evolution, 4(12), 1263-1274.

An, C. I., Fukusaki, E. I., & Kobayashi, A. (2001). Plasma-membrane H+-ATPases are expressed in pitchers of the carnivorous plant Nepenthes alata Blanco. Planta, 212(4), 547-555.

An, C. I., Fukusaki, E. I., & Kobayashi, A. (2002). Aspartic proteinases are expressed in pitchers of the carnivorous plant Nepenthes alata Blanco. Planta, 214(5), 661- 667.

Arber, A. (1941). On the morphology of the pitcher-leaves in Heliamphora, Sarracenia, Darlingtonia, Cephalotus, and Nepenthes. Annals of Botany, 5(4), 563-578.

Banks, J., Nishiyama, T., Hasebe, M., Bowman, J., Gribskov, M., Depamphilis, C., et al. (2011). The Selaginella Genome Identifies Genetic Changes Associated with the Evolution of Vascular Plants. Science , 332 (6032), 960-963.

Barker, M. S. (2013). Karyotype and genome evolution in pteridophytes. In Plant Genome Diversity Volume 2 (pp. 245-253). Springer Vienna.

Bartholmes, C., Hidalgo, O., & Gleissberg, S. (2012). Evolution of the YABBY gene family with emphasis on the basal eudicot Eschscholzia californica (Papaveraceae). Plant Biology, 14(1), 11-23.

Bekesiova, I., Nap, J. P., & Mlynarova, L. (1999). Isolation of high quality DNA and RNA from leaves of the carnivorous plant Drosera rotundifolia. Plant Molecular Biology Reporter, 17(3), 269-277.

Bowman, J. L. (2004). Class III HD‐ Zip gene regulation, the golden fleece of ARGONAUTE activity?. Bioessays, 26(9), 938-942.

Bowman, J. L., & Floyd, S. K. (2008). Patterning and polarity in seed plant shoots. Annu. Rev. Plant Biol., 59, 67-88.

Boyce, C. K., & Knoll, A. H. (2002). Evolution of developmental potential and the multiple independent origins of leaves in Paleozoic vascular plants. Paleobiology, 28(1), 70-100.

107

Bremer, B., Bremer, K., Chase, M., Fay, M., Reveal, J., Soltis, D., ... & Stevens, P. (2009). An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Botanical Journal of the Linnean Society.

Brocchieri, L. (2001). Phylogenetic inferences from molecular sequences: review and critique. Theoretical population biology, 59(1), 27-40.

Byrne, M. E. (2006). Shoot meristem function and leaf polarity: the role of class III HD– ZIP genes. PLoS genetics, 2(6), e89.

Carlquist, S. J., & Schneider, E. L. (2001). Vegetative anatomy of the New Caledonian endemic Amborella trichopoda: relationships with the Illiciales and implications for vessel origin. Pacific Science, 55(3), 305-312.

Chi, X., Yang, Q., Chen, X., Wang, J., Pan, L., Chen, M., ... & Yu, S. (2011). Identification and characterization of microRNAs from peanut (Arachis hypogaea L.) by high-throughput sequencing. PLoS One, 6(11), e27530.

Côté, C., Boileau, F., Roy, V., Ouellet, M., Levasseur, C., Morency, M. J., ... & MacKay, J. (2010). Gene family structure, expression and functional analysis of HD-Zip III genes in angiosperm and gymnosperm forest trees. BMC Plant Biology, 10(1), 273.

Cuatrecasas, J. (1970). Brunelliaceae. Flora Neotropica, 2, 1-188.

D’Amato, P. (1998). The savage garden: cultivating carnivorous plants. Ten Speed Press.

Darwin, C. (1875). Insectivorous Plants. London: John Murray.

Dickson, A. (1883, January). On the morphology of the pitcher of Cephalotus follicularis. In Transactions of the Botanical Society of Edinburgh (Vol. 14, No. 1-4, pp. 172- 181). Taylor & Francis Group.

Dickison, W. C., & Rutishauser, R. (1990). Developmental morphology of stipules and systematics of the Cunoniaceae and presumed allies. II, Taxa without interpetiolar stipules and conclusions. Botanica helvetica, 100(1), 75-95.

Eichler, A.W. (1881). Uber die Schlauchblatter von Cephalotus follicular (On the trapping leaves of C. follicular). Jahrbuch des Berliner Botanish Garten 1, 193- 197.

Endress, P. K., & Igersheim, A. (2000). The reproductive structures of the basal angiosperm Amborella trichopoda (Amborellaceae). International Journal of Plant Sciences, 161(S6), S237-S248.

108

Felsenstein, J. (1978). Cases in which parsimony or compatibility methods will be positively misleading. Systematic Biology, 27(4), 401-410.

Finet, C., Berne-Dedieu, A., Scutt, C. P., & Marlétaz, F. (2013). Evolution of the ARF Gene Family in Land Plants: Old Domains, New Tricks. Molecular biology and evolution, 30(1), 45-56.

Fleischmann, A., & Heubl, G. (2009). Overcoming DNA extraction problems from carnivorous plants. In Anales del Jardín Botánico de Madrid (Vol. 66, No. 2, pp. 209-215). Real Jardín Botánico.

Floyd, S. K., & Bowman, J. L. (2004). Gene regulation: ancient microRNA target sequences in plants. Nature, 428(6982), 485-486.

Floyd, S. K., & Bowman, J. L. (2007). The ancestral developmental tool kit of land plants. International Journal of Plant Sciences, 168(1), 1-35.

Floyd, S. K., Zalewski, C. S., & Bowman, J. L. (2006). Evolution of class III homeodomain–leucine zipper genes in streptophytes. Genetics, 173(1), 373-388.

Fogliani, B., Hopkins, H. C., Bouraïma-Madjèbi, S., & Medevielle, V. (2009). Morphological development of Geissois pruinosa (Cunoniaceae) from seed to adult, and the expression of plesiomorphic characters in seedlings. Flora- Morphology, Distribution, Functional Ecology of Plants, 204(1), 7-16.

Franck, D. H. (1976). The morphological interpretation of epiascidiate leaves—An historical perspective—. The Botanical Review, 42(3), 345-388.

Freeling, M. (1992). A conceptual framework for maize leaf development. Developmental biology, 153(1), 44-58.

Freeling, M., and Thomas, B.C. (2006). Gene-balanced duplications, like tetraploidy, provide predictable drive to increase morphological complexity. Genome Res. 16, 805–814.

Froebe, H. A., & Baur, N. (1988). Die Morphogenese der Kannenblätter von Cephalotus follicularis Labill (No. 3). Akademie der Wissenschaften und der Literatur.

Galbraith, D. W., Bennetzen, J. L., Kellogg, E. A., Pires, J. C., & Soltis, P. S. (2011). The genomes of all angiosperms: a call for a coordinated global census. Journal of Botany, 2011.

Garcia, D., Collier, S. A., Byrne, M. E., & Martienssen, R. A. (2006). Specification of Leaf Polarity inArabidopsis via the trans-Acting siRNA Pathway. Current Biology, 16(9), 933-938.

109

Gaudin, V., Lunness, P. A., Fobert, P. R., Towers, M., Riou-Khamlichi, C., Murray, J. A., ... & Doonan, J. H. (2000). The expression of D-cyclin genes defines distinct developmental zones in snapdragon apical meristems and is locally regulated by the Cycloidea gene. Plant Physiology, 122(4), 1137-1148.

Gleissberg, S., Groot, E. P., Schmalz, M., Eichert, M., Kölsch, A., & Hutter, S. (2005). Developmental events leading to peltate leaf structure in Tropaeolum majus (Tropaeolaceae) are associated with expression domain changes of a YABBY gene. Development genes and evolution, 215(6), 313-319.

Gould, B., & Kramer, E. M. (2007). Virus-induced gene silencing as a tool for functional analyses in the emerging model plant Aquilegia (columbine, Ranunculaceae). Plant Methods, 3(1), 6.

Graybeal, A. (1998). Is it better to add taxa or characters to a difficult phylogenetic problem?. Systematic Biology, 47(1), 9-17.

Gregory, T. R., & Mable, B. K. (2005). Polyploidy in animals. The evolution of the genome, 171, 427-517.

Ha, C. M., Jun, J. H., Nam, H. G., & Fletcher, J. C. (2007). BLADE-ON-PETIOLE1 and 2 control Arabidopsis lateral organ fate through regulation of LOB domain and adaxial-abaxial polarity genes. The Plant Cell Online, 19(6), 1809-1825.

Harrison, C. J., Corley, S. B., Moylan, E. C., Alexander, D. L., Scotland, R. W., & Langdale, J. A. (2005). Independent recruitment of a conserved developmental mechanism during leaf evolution. Nature, 434(7032), 509-514.

Hu, R., Chi, X., Chai, G., Kong, Y., He, G., Wang, X., ... & Zhou, G. (2012). Genome- wide identification, evolutionary expansion, and expression profile of homeodomain-leucine zipper gene family in poplar (Populus trichocarpa). PloS one, 7(2), e31149.

Ibarra-Laclette, E., Lyons, E., Hernández-Guzmán, G., Pérez-Torres, C. A., Carretero- Paulet, L., Chang, T. H., ... & Herrera-Estrella, L. (2013). Architecture and evolution of a minute plant genome. Nature.

Iwakawa, H., Iwasaki, M., Kojima, S., Ueno, Y., Soma, T., Tanaka, H., ... & Machida, C. (2007). Expression of the ASYMMETRIC LEAVES2 gene in the adaxial domain of Arabidopsis leaves represses cell proliferation in this domain and is critical for the development of properly expanded leaves. The Plant Journal, 51(2), 173- 184.

Jain, M., Tyagi, A. K., & Khurana, J. P. (2008). Genome‐ wide identification, classification, evolutionary expansion and expression analyses of homeobox genes in rice. FEBS journal, 275(11), 2845-2861.

110

Jiao, Y., Leebens-Mack, J., Ayyampalayam, S., Bowers, J. E., McKain, M. R., McNeal, J., ... & W dePamphilis, C. (2012). A genome triplication associated with early diversification of the core eudicots. Genome Biol, 13(1), R3.

Jiao, Y., Wickett, N. J., Ayyampalayam, S., Chanderbali, A. S., Landherr, L., Ralph, P. E., ... & Leebens-Mack, J. (2011). Ancestral polyploidy in seed plants and angiosperms. Nature, 473(7345), 97-100.

Johnson, M. T., Carpenter, E. J., Tian, Z., Bruskiewich, R., Burris, J. N., Carrigan, C. T., ... & Wong, G. K. S. (2012). Evaluating Methods for Isolating Total RNA and Predicting the Success of Sequencing Phylogenetically Diverse Plant Transcriptomes. PloS one, 7(11), e50226.

Juniper, B. E., Robins, R. J., & Joel, D. M. (1989). The carnivorous plants . London, etc.: Academic Press.

Kaplan, D. R. (2001). The science of plant morphology: definition, history, and role in modern biology. American Journal of Botany, 88(10), 1711-1741.

Katoh, K., Misawa, K., Kuma, K. I., & Miyata, T. (2002). MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic acids research, 30(14), 3059-3066.

Kenrick, P., & Crane, P. R. (1997). The origin and early diversification of land plants. A cladistic study (Vol. 560). Smithsonian Institute Press Washington DC.: A cladistic study. Smithsonian Institute Press.

Kidner, C. A., & Timmermans, M. C. (2007). Mixing and matching pathways in leaf polarity. Current opinion in plant biology, 10(1), 13-20.

Ko, J. H., Prassinos, C., & Han, K. H. (2006). Developmental and seasonal expression of PtaHB1, a Populus gene encoding a class III HD‐ Zip protein, is closely associated with secondary growth and inversely correlated with the level of microRNA (miR166). New Phytologist, 169(3), 469-478.

Kouchi, H., & Hata, S. (1993). Isolation and characterization of novel nodulin cDNAs representing genes expressed at early stages of soybean nodule development. Molecular and General Genetics MGG, 238(1-2), 106-119.

Kramer, E. M. (2005). Methods for studying the evolution of plant reproductive structures: comparative gene expression techniques. Methods in enzymology, 395, 617-636.

Lamesch, P., Berardini, T. Z., Li, D., Swarbreck, D., Wilks, C., Sasidharan, R., ... & Huala, E. (2012). The Arabidopsis Information Resource (TAIR): improved gene annotation and new tools. Nucleic acids research, 40(D1), D1202-D1210.

111

Liu, T., Reinhart, B. J., Magnani, E., Huang, T., Kerstetter, R., & Barton, M. K. (2013, May). Of Blades and Branches: Understanding and Expanding the Arabidopsis Ad/Abaxial Regulatory Network through Target Gene Identification. In Cold Spring Harbor symposia on quantitative biology. Cold Spring Harbor Laboratory Press.

Lloyd, F. E. (1942). The carnivorous plants. Waltham, Massachusetts: Chronica Botanica Company.

Luo, J.H., Yan, J., Weng, L., Yang, J., Zhao, Z., Chen, J.H. et al. (2005) Different expression patterns of duplicated PHANTASTICA-like genes in Lotus japonicus suggest their divergent functions during compound leaf development. Cell Res. 15: 665–677.

Mable, B. K. (2003). Breaking down taxonomic barriers in polyploidy research. Trends in plant Science, 8(12), 582-590.

Magnani, E., & Barton, M. K. (2011). A per-ARNT-sim-like sensor domain uniquely regulates the activity of the homeodomain leucine zipper transcription factor REVOLUTA in Arabidopsis. The Plant Cell Online, 23(2), 567-582.

Magnani, E., & Hake, S. (2008). KNOX lost the OX: the Arabidopsis KNATM gene defines a novel class of KNOX transcriptional regulators missing the homeodomain. The Plant Cell Online, 20(4), 875-887.

Maher, C., Stein, L., & Ware, D. (2006). Evolution of Arabidopsis microRNA families through duplication events. Genome research, 16(4), 510-519.

Mathews, S., & Kramer, E. M. (2012). The evolution of reproductive structures in seed plants: a re‐ examination based on insights from developmental genetics. New Phytologist, 194(4), 910-923.

Matthews, M. L., & Endress, P. K. (2002). Comparative floral structure and systematics in Oxalidales (Oxalidaceae, Connaraceae, Brunelliaceae, Cephalotaceae, Cunoniaceae, Elaeocarpaceae, Tremandraceae). Botanical Journal of the Linnean Society, 140(4), 321-381.

McConnell, J. R., Emery, J., Eshed, Y., Bao, N., Bowman, J., & Barton, M. K. (2001). Role of PHABULOSA and PHAVOLUTA in determining radial patterning in shoots. Nature, 411(6838), 709-713.

Moore, R. C., & Purugganan, M. D. (2005). The evolutionary dynamics of plant duplicate genes. Current opinion in plant biology, 8(2), 122-128.

Mukherjee, K., Brocchieri, L., & Bürglin, T. R. (2009). A comprehensive classification and evolutionary analysis of plant homeobox genes. Molecular biology and evolution, 26(12), 2775-2794. 112

Mukherjee, K., & Bürglin, T. R. (2006). MEKHLA, a novel domain with similarity to PAS domains, is fused to plant homeodomain-leucine zipper III proteins. Plant Physiology, 140(4), 1142-1150.

Nelson, T., Tausta, S. L., Gandotra, N., & Liu, T. (2006). Laser microdissection of plant tissue: what you see is what you get. Annu. Rev. Plant Biol., 57, 181-201.

Ohashi-Ito, K., & Fukuda, H. (2003). HD-zip III homeobox genes that include a novel member, ZeHB-13 (Zinnia)/ATHB-15 (Arabidopsis), are involved in procambium and xylem cell differentiation. Plant and Cell Physiology, 44(12), 1350-1358.

Otto, S. P., & Whitton, J. (2000). Polyploid incidence and evolution. Annual review of genetics, 34(1), 401-437.

Prigge, M. J., & Clark, S. E. (2006). Evolution of the class III HD‐ Zip gene family in land plants. Evolution & development, 8(4), 350-361.

Prigge, M. J., Otsuga, D., Alonso, J. M., Ecker, J. R., Drews, G. N., & Clark, S. E. (2005). Class III homeodomain-leucine zipper gene family members have overlapping, antagonistic, and distinct roles in Arabidopsis development. The Plant Cell Online, 17(1), 61-76.

Richardt, S., Lang, D., Reski, R., Frank, W., & Rensing, S. A. (2007). PlanTAPDB, a phylogeny-based resource of plant transcription-associated proteins. Plant Physiology, 143(4), 1452-1466.

Robischon, M., Du, J., Miura, E., & Groover, A. (2011). The Populus class III HD ZIP, popREVOLUTA, influences cambium initiation and patterning of woody stems. Plant physiology, 155(3), 1214-1225.

Schmid, M., Davison, T.S., Henz, S.R., Pape, U.J., Demar, M., Vingron, M., Schölkopf, B., Weigel, D. and Lohmann, J.U. (2005). A gene expression map of Arabidopsis thaliana development. Nat. Genet. 37, 501-6.

Schulze, W., Frommer, W. B., & Ward, J. M. (1999). Transporters for ammonium, amino acids and peptides are expressed in pitchers of the carnivorous plantNepenthes. The Plant Journal, 17(6), 637-646.

Sessa, G., Morelli, G., & Ruberti, I. (1993). The Athb-1 and-2 HD-Zip domains homodimerize forming complexes of different DNA binding specificities. The EMBO journal, 12(9), 3507.

Siegfried, K. R., Eshed, Y., Baum, S. F., Otsuga, D., Drews, G. N., & Bowman, J. L. (1999). Members of the YABBY gene family specify abaxial cell fate in Arabidopsis. Development, 126(18), 4117-4128.

113

Simillion, C., Vandepoele, K., Van Montagu, M. C., Zabeau, M., & Van de Peer, Y. (2002). The hidden duplication past of Arabidopsis thaliana. Proceedings of the National Academy of Sciences, 99(21), 13627-13632.

Soltis, D. E., Albert, V. A., Leebens-Mack, J., Bell, C. D., Paterson, A. H., Zheng, C., ... & Soltis, P. S. (2009). Polyploidy and angiosperm diversification. American Journal of Botany, 96(1), 336-348.

Soltis, P. S., Burleigh, J. G., Chanderbali, A. S., Yoo, M. J., & Soltis, D. E. (2010). Gene and genome duplications in plants. Evolution after Gene Duplication, 269-298.

Soltis, D. E., & Soltis, P. S. (2004). Amborella not a “basal angiosperm”? Not so fast. American Journal of Botany, 91(6), 997-1001.

Soltis, P. S., & Soltis, D. E. (2013). A conifer genome spruces up plant phylogenomics. Genome biology, 14(6), 1-5.

Soltis, P. S., Soltis, D. E., & Chase, M. W. (1999). Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature, 402(6760), 402-404.

Soltis, D. E., Soltis, P. S., Endress, P. K., & Chase, M. W. (2005). Phylogeny and evolution of angiosperms. Sinauer Associates Incorporated.

Soltis, D. E., Smith, S. A., Cellinese, N., Wurdack, K. J., Tank, D. C., Brockington, S. F., ... & Soltis, P. S. (2011). Angiosperm phylogeny: 17 genes, 640 taxa. American Journal of Botany, 98(4), 704-730.

Stamatakis, A. (2006). RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22(21), 2688- 2690.

Staudt, A. C., & Wenkel, S. (2010). Regulation of protein function by ‘microProteins’. EMBO reports, 12(1), 35-42.

Sun, C. H., Su, L. H., & Gillin, F. D. (2006). Novel plant-GARP-like transcription factors in< i> Giardia lamblia. Molecular and biochemical parasitology, 146(1), 45-57.

Thien, L. B., Sage, T. L., Jaffré, T., Bernhardt, P., Pontieri, V., Weston, P. H., ... & Dupre, J. L. (2003). The population structure and floral biology of Amborella trichopoda (Amborellaceae). Annals of the Missouri Botanical Garden, 466-490.

Toriba, T., Harada, K., Takamura, A., Nakamura, H., Ichikawa, H., Suzaki, T., & Hirano, H. Y. (2007). Molecular characterization the YABBY gene family in Oryza sativa and expression analysis of OsYABBY1. Molecular Genetics and Genomics, 277(5), 457-468.

114

Troll, W. (1939). Vergleichende Morphologie der höheren Pflanzen, vol. 1 Gebrüder Borntraeger.

Tsukaya, H. (2006). Mechanism of leaf-shape determination. Annu. Rev. Plant Biol., 57, 477-496.

Van de Peer, Y., Fawcett, J. A., Proost, S., Sterck, L., & Vandepoele, K. (2009). The flowering world: a tale of duplications. Trends in plant science, 14(12), 680-688.

Waites, R., & Hudson, A. (1995). phantastica: a gene required for dorsoventrality of leaves in Antirrhinum majus. Development, 121(7), 2143-2154.

Wang, J., Zhang, G., Li, Q., Jiang, H., Liu, C., Amatore, C., & Wang, X. (2013). In vivo self-bio-imaging of tumors through in situ biosynthesized fluorescent gold nanoclusters. Scientific reports, 3.

Wenkel, S., Emery, J., Hou, B. H., Evans, M. M., & Barton, M. K. (2007). A feedback regulatory module formed by LITTLE ZIPPER and HD-ZIPIII genes. The Plant Cell Online, 19(11), 3379-3390.

Yamaguchi, T., Nukazuka, A., & Tsukaya, H. (2012). Leaf adaxial–abaxial polarity specification and lamina outgrowth: evolution and development. Plant and Cell Physiology, 53(7), 1180-1194.

Zahn, L. M., Kong, H., Leebens-Mack, J. H., Kim, S., Soltis, P. S., Landherr, L. L., ... & Ma, H. (2005). The Evolution of the SEPALLATA Subfamily of MADS-Box Genes A Preangiosperm Origin With Multiple Duplications Throughout Angiosperm History. Genetics, 169(4), 2209-2223.

Zamboni, A., Pierantoni, L., & De Franceschi, P. (2008). Total RNA extraction from strawberry tree (Arbutus unedo) and several other woody-plants. iForest- Biogeosciences and Forestry, 1(1), 122-125.

Zhong, R., & Ye, Z. H. (2007). Regulation of HD-ZIP III genes by microRNA 165. Plant signaling & behavior, 2(5), 351-353.

Zhu, Y., Song, D., Sun, J., Wang, X., & Li, L. (2013). PtrHB7, a class III HD-Zip gene, plays a critical role in regulation of vascular cambium differentiation in Populus. Molecular plant.

115

BIOGRAPHICAL SKETCH

Nicholas William Miles was born in Indianapolis, Indiana on the day of the great mid-west blizzard of 1985. His father was a custodian at the local hospital and his mother a sales associate at an electronics store. As a child he was very active in sports including baseball, soccer, and golf. His first interest in science was at 10 years of age in the areas of astronomy and chemistry, with his favorite book at the time being A Brief

History of Time by Stephen Hawking. He also learned to play his first musical instrument, the violin, at 10, taught to him by his mother. At 14 years of age he started his first job as dishwasher. He was rejected entry to the University of Florida for undergraduate studies and attended Florida State University beginning in 2003. In 2007 he graduated with Honors, Cum Laude, from FSU and started graduate school at the

University of Florida. During his time at the University of Florida he applied for 9 grants,

4 of which were successfully funded, 2 of those from The National Science Foundation.

He mentored 9 undergraduate and 4 high school students from various backgrounds and majors. And he taught 14 semesters of classes, 7 of them upper level plant classes,

2 of them vertebrate centered classes, and one of them attended by a Heisman Trophy

Winner.

116