The Regulation of Ontogenetic Diversity in Papaveraceae Compound Leaf Development

Home , Argemone, Argemone mexicana, Cardamine californica, Chelidonium, Eschscholzia, Eschscholzia californica

A thesis presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Master of Science

Alastair R. Plant

August 2013

This thesis titled

The Regulation of Ontogenetic Diversity in Papaveraceae Compound Leaf Development

ALASTAIR R PLANT

has been approved for

the Department of Environmental and Plant Biology

and the College of Arts and Sciences by

Stefan Gleissberg

Assistant Professor of Environmental and Plant Biology

Robert Frank

Dean, College of Arts and Sciences 3

ABSTRACT

PLANT, ALASTAIR R., M.S., August 2013, Plant Biology

The Regulation of Ontogenetic Diversity in Papaveraceae Compound Leaf Development

Director of Thesis: Stefan Gleissberg

The leaf is almost ubiquitous throughout land plants but due to its complex and

flexible developmental program is highly morphologically variable between taxa.

Description of the functions of regulatory genes key to leaf development in different

evolutionary lineages allows the study of changes in developmental mechanisms through

evolutionary time as a means for anatomical and morphological diversification. The

roles of homologs of CINCINNATA-like TCP family genes, ARP genes, and Class I

KNOX genes were investigated in two members of the Papaveraceae, a basal eudicot lineage positioned in between major angiosperm groups, by phylogenetic analysis, in situ hybridization, expression profiling by quantitative polymerase chain reaction, and virus- induced gene silencing in Eschscholzia californica and Cysticapnos vesicaria.

Expression data were similar to those for homologous genes in core eudicot species, however, some gene functions found in core eudicots were not associated with basal eudicot homologs, and so have either been gained or lost from the ancestral state. This reflects the dynamism of the leaf developmental plan and its diversification through evolution. 4

DEDICATION

This thesis is dedicated to curiosity.

ACKNOWLEDGMENTS

I would like to thank those who have contributed to this body of research: Anandi

Bhattacharya cloned the TCP domain of Eschscholzia californica CINCINNATA

(EcCIN), Stefan Gleissberg cloned EcPHAN and performed the EcPHAN expression profile RT-PCR, and Andrea Scholz produced the Eschscholzia californica

PHANTASTICA (EcPHAN) VIGS construct.

I would also like to thank my advisor Stefan Gleissberg and my committee members Harvey Ballard, and Sarah Wyatt for their assistance and guidance, Oriane

Hidalgo and Conny Bartholmes for training and advice, and undergraduate researchers

Celeste Taylor, Avery Tucker, Brooke Johnson, Ben Imbus, Chi Zhang, Emily Usher,

Jennifer Leetch, and Abby Pugel for their help in data collection and in being a source of new ideas and inspiration. This research was funded in part by an Ohio University

Research Committee Grant.

TABLE OF CONTENTS

Page

Abstract ...... 3 Dedication ...... 4 Acknowledgments...... 5 List of Tables ...... 8 List of Figures ...... 9 Chapter 1: Introduction ...... 12 References ...... 20 Chapter 2: Regulation of dissected leaf architecture in Eschscholzia californica By a CIN-TCP gene ...... 28 Abstract ...... 28 Introduction ...... 29 Methods ...... 33 Results ...... 44 Discussion ...... 56 References ...... 59 Chapter 3: Eschscholzia californica PHANTASTICA (EcPHAN) regulates petal morphogenesis in the California Poppy ...... 65 Abstract ...... 65 Introduction ...... 65 Methods ...... 69 Results ...... 71 Discussion ...... 77 References ...... 80 Chapter 4: Duplicated STM-like KNOX I genes act in floral meristem activity in Eschscholzia californica (Papaveraceae) ...... 86 Authors ...... 86 Abstract ...... 86 Introduction ...... 87 7

Materials and Methods ...... 91 Results ...... 95 Discussion ...... 110 Acknowledgements ...... 117 Author Contributions ...... 117 References ...... 118 Supplementary Data ...... 128 Chapter 5: Laser microdissection of Eschscholzia californica leaf primordia for comparison of gene expression between developmental stages ...... 134 Abstract ...... 134 Introduction ...... 134 Method ...... 138 Results ...... 141 Discussion ...... 143 Acknowledgements ...... 146 References ...... 146 Chapter 6: Discussion ...... 149 References ...... 151 Appendix I: Cloning of putative microRNA319 homologs from Eschscholzia californica ...... 153 References ...... 156 Appendix II: Cloning of gibberellic acid and cytokinin metabolic genes from Eschscholzia californica ...... 158 References ...... 162

LIST OF TABLES

Page

Table 1: Primers used for cloning, in situ hybridization probe preparation and quantitative

PCR for CIN-TCP genes ...... 43

Table 2: Primers used for amplification of KNOX genes ...... 132

Table 3: Sequence IDs for KNOX genes used in phylogenetic analyses ...... 133

LIST OF FIGURES

Page

Figure 1: Summary of interactions between selected major regulators of leaf development ...... 17

Figure 2: Eschscholzia TCP gene structures ...... 44

Figure 3: Phylogeny of amino acid sequences putatively translated from Papaveraceae

and Arabidopsis thaliana CIN-TCP genes ...... 46

Figure 4: Expression profiles of EcCIN, EcTCP2-LIKE and EcTCP5-LIKE in wild-type

Eschscholzia californica tissues ...... 47

Figure 5: In situ hybridization of EcCIN mRNA in wild type Eschscholzia californica ...... 48

Figure 6: Box and violin plots of total leaflets per node in plants infiltrated with pTRV2- empty (control), pTRV2-3’-specific and pTRV2 –TCP-domain constructs

...... 49

Figure 7: Phenotypic effects of silencing EcCIN with TCP-domain-specific and unconserved 3’-end-specific VIGS constructs ...... 50

Figure 8: pTRV2-control and pTRV2-CvCIN plants...... 52

Figure 9: Box and violin plots for leaflets per node in pTRV2-empty and pTRV2-CvCIN plants ...... 53

Figure 10: Quantitative PCR shows decreased expression of CvCIN in pTRV2-CvCIN- treated VIGS plants ...... 54

Figure 11: Box and violin plots of total leaflets per node for pTRV2-EcCIN-3’, pTRV2-

EcTCP2-LIKE, and pTRV2-EcTCP5-LIKE plants...... 55 10

Figure 12: Phylogeny of ASYMMETRIC LEAVES 1, ROUGH SHEATH 2 and

PHANTASTICA and their homologs in the Papaveraceae and in the basal angiosperm

Amborella trichopoda ...... 72

Figure 13: RT-PCR expression profile of EcPHAN ...... 73

Figure 14: VIGS-induced EcPHAN knockdown phenotypes ...... 75

Figure 15: Scanning electron microscopy of petals exhibiting the EcPHAN knockdown phenotype ...... 76

Figure 16: Domain structure of the hypothetical proteins encoded by the four class I

KNOX genes in Eschscholzia californica...... 96

Figure 17: Phylogram of angiosperm KNOX I genes ...... 98

Figure 18: Expression profiles of Eschscholzia californica class I KNOX genes using semi-quantitative RT-PCR...... 99

Figure 19 Floral phenotypes of EcSTM1, EcSTM2, and EcSTM1+2 silenced Eschscholzia californica plants ...... 101

Figure 20: Scanning electron micrographs and longitudinal sections of EcSTM-VIGS floral buds...... 102

Figure 21: Reduced gynoecium and flowers in EcSTM-silenced Eschscholzia plants.

...... 103

Figure 22: Stamen numbers in EcSTM-silenced Eschscholzia flowers...... 104

Figure 23: Spectrum of floral organ initiation defects in EcSTM-VIGS flowers...... 107

Figure 24: Homeotic organ transformations in EcSTM-VIGS plants...... 108 11

Figure 25: Additional pseudowhorls in Eschscholzia shoots...... 109

Figure 26: Phylogram of selected angiosperm KNOX nucleotide sequences (supplemental) ...... 128

Figure 27: Phylogram of poppy and few selected other eudicot KNOX I deduced amino acid sequences (supplemental)...... 129

Figure 28: Semi-quantitative RT-PCR of STM-like genes in floral terminal flower buds of VIGS-treated Eschscholzia californica (supplemental)...... 129

Figure 29: Distribution of stamen numbers in EcSTM-silenced and control flowers (supplemental)...... 130

Figure 30: Degree of leaf dissection in KNOX I-silenced and control leaves

(supplemental)...... 131

Figure 31: Laser microdissection enables profiling of gene expression at different stages

of leaf development...... 142

Figure 32. 2-D centroid model for an Eschscholzia californica premiRNA319 sequence

...... 154

CHAPTER 1: INTRODUCTION

The establishment of macroscopic form and structure from a multitude of interactions between the molecular agents of the cell is a remarkable, if not elegant process. For evolution to render but a single form would be considered an achievement, yet morphological and anatomical diversity extend far beyond the single plan. The source of such diversity is development, the four-dimensional product of genetic, molecular and biomechanical interactions at the cellular, tissue and organismal levels, and its malleability. Modification of the hereditary aspects of developmental processes through evolution has borne the array of phenotypic diversity visible today, and is the subject of the field entitled evolutionary-developmental biology, or ‘evo-devo’. A primarily genetic approach has been taken; changes to the expression, functions and/or interactions between genes involved in development have the potential to modify existing architecture or to produce entirely new arrangements. The broader question asks which changes were necessary to the evolution of new forms and structures, the more specific asking what subtle changes must occur to produce diversity over short periods of evolutionary time, even at the species level. This original thesis places emphasis on the latter, and explores how the differing employment of certain genes may result in development differences between taxa and lineages, commenting upon the importance of regulatory genes in development.

The development of the plant leaf shares with other developmental processes an abundance of characteristics: founder cells arising from a source of indeterminate cells must divide, expand and differentiate along axes (proximodistal, mediolateral, and 13

dorsiventral), and assemble into gross morphological structures (the blade and any

marginal appendages, and the petiole) while establishing histological patterns within, to

render an adaptive structure to the benefit of its host organism. Its roles in gas exchange

and transpiration, thermoregulation, and photosynthesis doubtlessly offer selective

advantages, to the extent that the bladed, vascularized leaves often termed megaphylls

have evolved independently in multiple lineages of the plant kingdom (Tomescu, 2009).

In the majority of plant species, leaves are determinate structures, and the morphology of

the leaf is sculpted in its primordial stages. Primordia that will become compound leaves

must not only pattern a simple lamina but must also initiate perhaps multiple orders of

dissections that give rise to lobes, leaflets, and serrations at maturity from the margins of

the blade where morphogenetic competency to produce new structures is retained, i.e.

expansion and differentiation are postponed. Formation of the specialized histology of

the leaf begins comparatively late in development during histogenesis, to which

morphogenesis eventually defers.

A host of genes are involved in the initiation of the incipient primordium, the

establishment and maintenance of the axes of growth, the morphogenetic patterning of

the blade, the transition to histogenesis and loss of morphogenetic competency,

histogenesis itself, and the expansion of the leaf to its size at maturity.

At the start of development, accumulation of auxin and downregulation of Class I

KNOX (KNOTTED homeobox) genes at a phyllotactically determined site indicates the induction of the incipient primordium in the peripheral zone of the shoot apical meristem

(SAM) (Bharathan et al., 2002; Blein et al., 2010). These genes, of which there are four 14

in Arabidopsis, perform partially redundant functions in shoot apical meristem

maintenance (Hay and Tsiantis, 2010). KNOX downregulation by the ARP genes

(named after ASYMMETRIC LEAVES 1, ROUGH SHEATH 2 and PHANTASTICA) diverts the cells from the meristematic indeterminacy of the SAM (Guo et al., 2008).

Induction is followed by morphogenesis, in which phase cell division is the main mode of morphological change. In compound leaves, the expression of KNOX genes resumes after establishment of the primordium (Bharathan et al., 2002, Hay and Tsiantis, 2006).

Proximodistal, dorsiventral and mediolateral axes are established and maintained by the expression of mutually antagonistic genes in adjacent domains, promoting region- specific cell identities. In Zea mays, the Class I KNOX gene KNOTTED1 has been implicated in proximodistal patterning (Ramirez et al., 2009) and mutants of the kinase

LIGULELESS NARROW have disrupted proximodistal and mediolateral patterning

(Moon et al., 2013), but a broad understanding of the genetic specification of these axes is lacking. In contrast, dorsiventral patterning is better described. The ARP genes contribute to the specification of adaxial identity (Eckhardt, 2004), although this is achieved primarily through stable regulation of KNOX gene expression (Guo et al.,

2008); the Class III homeodomain leucine zipper (HD-ZIP) genes PHABULOSA,

PHAVOLUTA and REVOLUTA are more specific contributors (Prigge et al., 2005), and counteract the abaxial identity genes that include the YABBY family (which also promote lamina expansion – Sarojam et al., 2010) and KANADI (Kerstetter et al., 2001).

Auxin remains concentrated at the primordial apex as a result of polar auxin transport, mediated by the arrangement of auxin efflux proteins of the PIN-FORMED 15

family (although pin1 mutants are capable of leaf formation, indicating the involvement

of other mechanisms – Guenot et al., 2012). In compound leaves, auxin also accumulates

at leaflet initiation sites (Kawamura et al., 2010). After such sites are marked, CUP-

SHAPED COTYLEDONS (CUC) genes, first associated with boundary formation in

embryonic development are expressed in the distal region adjacent to the sinus (Blein,

2010). In numerous species, the CUC genes promote resumed expression of the Class I

KNOX genes that were downregulated upon induction of the leaf primordium. KNOX

expression may maintain the indeterminate fates of those cells at the margin and prolong

cell division (Hay and Tsiantis, 2010), providing additional cells that are funneled into

the developing structures at the margin (Kawamura et al., 2010). An alternate

mechanism for promoting leaflet development exists in the Inverted Repeat Lacking

Clade of the Fabaceae, where UNIFOLIATA, a homologue of the Solanum lycopersicum

gene FALSIFLORA, fulfills a similar function to the KNOX genes (Champagne et al.,

2007; Molinero-Rosales et al., 1999).

Marginal elaboration is not indefinite. Cessation of new leaflet formation and

progression into histogenesis requires the genetically orchestrated reduction in expression

of cytokinins and increased expression of gibberellins to limit cell division and promote

cell elongation and differentiation. One key actor in these events is CINCINNATA

(Antirrhinum majus), homologous to LANCEOLATE in Solanum lycopersicum and TCPs

3, 4, and 10 in Arabidopsis thaliana (Crawford et al., 2004). It is a member of the TCP transcription factor family, named after TEOSINTE BRANCHED 1, CINCINNATA, and

PROLIFERATING CELL FACTORS 1 & 2. CINCINNATA promotes maturation of leaf 16

primordium via several regulatory pathways: antagonism of cell cycle progression from

G1 to S phase to limit cell division (Aggarwal et al., 2011); prevention of leaflet and

sinus formation by suppression of the CUC genes through miRNA 164 (Koyama et al.,

2010); direct interaction with the ARP genes (Koyama et al., 2010) and their interacting

partner ASYMMETRIC LEAVES 2 to downregulate multiple Class I KNOX genes (Li et al., 2012); and, hypothetically, limitation of the auxin response via maintenance or upregulation of INDOLE-3-ACETIC ACID3/SHORT HYPOCOTYL2 (IAA/SHY2) and

SMALL AUXIN UP RNA (SAUR) (Koyama et al., 2010). CINCINNATA itself is

negatively regulated by microRNA 319, which also represses its relatives in the CIN-TCP

(Class II) subclade of the TCP family. A gradual increase in CINCINNATA expression

and the release of miRNA repression herald the onset of maturation (Shleizer-Burko et

al., 2011).

The sustenance of indeterminacy and cell division in the margins of the leaf,

promoted by KNOX but opposed by CINCINNATA and the ARP genes (Figure 1), may

determine the degree of dissection in that leaf, permitting higher orders of dissection or

the formation of more frequent structures (Hagemann and Gleissberg, 1996), and may

therefore offer an explanation for the breadth of leaf morphologies in existence. The

genetic regulatory networks governing leaf development would certainly have been

subject to selective pressure when species expanded into new ranges; numerous authors

have speculated that, in combination with anatomical and morphological characteristics

such as venation pattern and blade depth, dissection of the leaf blade to form a compound

leaf provides an adaptive advantage in new environs where, for example, dissection may 17 reduce mechanical stress in high winds, or conserve moisture by removing tissue that is more distant from veins (reviewed in Nicotra et al. 2011). Subtle changes in the expression of these key players may have radical effects. For example, alteration of

PHANTASTICA expression can change the patterning of dissection from pinnate, palmate, to simple (Kim et al., 2003). Numerous genera exhibit an equal or greater range of morphological diversity (Nicotra et al., 2011, Jones et al., 2009), despite the similarity of their developmental plans inherited from their common ancestry, due to differential integration between genetic regulatory networks and modules (Klingenberg, 2008;

Klingenberg et al., 2011).

YABBYs

Sarojam et al. Ori et al. (2007) (2010) miRNA319 CINCINNATA

Koyama et al. Koyama et al. (2010) (2010) miRNA164a

Nikovics et al. (2006) Guo et al. (2008) CUP-SHAPED COTYLEDONS AS1/RS2/PHAN + AS2

Timmermans et al. (1999) Waites et al. Kawamura et al. (1998) (2010) SHOOTMERISTEMLESS & Class I KNOX

Figure 1. Summary of interactions between selected major regulators of leaf development. 18

While biologically interesting, such developmental flexibility does not lend itself

to the study of the original application of a given gene regulatory network to a

developmental process such as leaf development. To give an example, the existence of

KNOX genes precedes not only the existence of the leaf but also the existence of the

shoot apical meristem, and the time of the first employment of KNOX in leaf

development is unknown (Bharathan et al., 2002).

One means by which to trace the evolutionary histories and importance of genes

and the importance of their regulatory networks and modules throughout evolutionary

history is to establish new species as phylogenetic landmarks at which the roles of those

genes are described, so that an image of their diversification can be constructed. The basal angiosperm Amborella trichopoda is an example, being used to describe the recruitment of genes to angiosperm-specific innovations such as floral organs (Zuccolo et al., 2011; Vialette-Guiraud et al., 2011), with the caveat that characteristics may be derived rather than ancestral (Posluszny and Tomlinson, 2003).

The Papaveraceae is the most basal family of the Ranunculales clade of

angiosperms, with the exception of the Eupteleaceae, and is the most basal herbaceous

eudicot family (Kadereit et al., 1997; Worberg et al., 2007 Wang et al., 2009). In

addition to their economic and scientific significance as synthesizers of alkaloids,

particularly opiates (Cahlíková et al., 2012; de Luca et al., 2012), their phylogenetic

position makes the poppies (subfamily Papaveroideae) and fumitories (subfamily

Fumarioideae) important for the study of eudicot evolution and development, and have 19

been the subjects of studies into leaf and flower development in basal eudicots

(Bartholmes et al., 2012; Lange et al., 2013).

This thesis explores the genetic regulation of the development of lateral organs, leaves and floral organs, in the Papaveraceae. Characterization of the leaf developmental program in the California poppy Eschscholzia californica and the fumitory Cysticapnos vesicaria will allow comparisons to be made with the core eudicots, into which several model species such as Arabidopsis thaliana and Antirrhinum majus fall, as well as the early-diverging basal angiosperms and the basal monocots (Floyd and Bowman, 2007).

Several candidate genes known for roles in lateral organ development in core eudicot and monocot models were studied and their functions in basal eudicots compared to their homologs in model species, with a view to expanding our understanding how development of the leaf has changed and diversified through evolution.

Chapter 2 investigates the role of homologs of CINCINNATA and related TCP genes in the leaf development of Eschscholzia californica and Cysticapnos vesicaria, while Chapter 3 explores the role EcPHAN, the ARP homolog in Eschscholzia californica petal development. These chapters will be prepared for publication as research articles. Chapter 4 presents the cloning, phylogenetic, and expression analysis of four KNOX I genes in Eschscholzia californica, and reports phenotypic effects following virus-mediated silencing of two STM-like KNOX genes. This chapter has been accepted for publication in Development Genes and Evolution. Chapter 5 explores the potential of Laser Microdissection Microscopy (LMD) in the study of multiple gene expression in various stages of leaf development, using Eschscholzia californica, and 20 shall function as a technical reference. In Chapter 6, the conclusions drawn from this work are presented with comments on the outlook for evolutionary-developmental research.

References

Aggarwal P., Padmanabhan B., Bhat A., Sarvepalli K., Sadhale P.P., and Nath

U. (2011). The TCP4 transcription factor of Arabidopsis blocks cell division in yeast at

G1 → S transition. Biochem. Biophys. Res. Commun. 410, 276-281.

Bartholmes C., Hidalgo O., and Gleissberg S. (2012). Evolution of the YABBY gene family with emphasis on the basal eudicot Eschscholzia californica

(Papaveraceae). 14, 11-23.

Bharathan G., Goliber T.E., Moore C., Kessler S., Pham T., and Sinha N.R. (2002).

Homologies in leaf form inferred from KNOXI gene expression during development.

Science 296, 1858-1860.

Blein T., Hasson A., and Laufs P. (2010). Leaf development: what it needs to be complex. Curr. Opin. Plant Biol. 13, 75-82.

Cahlíková L., Kučera R., Hošt'Álková A., Klimeš J., and Opletal L. (2012).

Identification of pavinane alkaloids in the genera Argemone and Eschscholzia by GC-

MS. 7, 1279-1281. 21

Champagne C. and Sinha N. (2004). Compound leaves: equal to the sum of their parts?

Development 131, 4401-4412.

Crawford B.C.W., Nath U., Carpenter R., and Coen E.S. (2004). CINCINNATA controls both cell differentiation and growth in petal lobes and leaves of antirrhinum.

Plant Physiol. 135, 244-253.

De Luca V., Salim V., Atsumi S.M., and Yu F. (2012). Mining the biodiversity of plants: A revolution in the making. Science 336,1658-1661.

Eckardt N.A. (2004). The Role of PHANTASTICA in Leaf Development. The Plant

Cell Online 16, 1073-1075.

Floyd S.K. and Bowman J.L. (2007). The ancestral developmental tool kit of land plants. Int. J. Plant Sci. 168, 1-35.

Guenot B., Bayer E., Kierzkowski D., Smith R.S., Mandel T., Zadnikova P.,

Benkova E., and Kuhlemeier C. (2012). PIN1-independent leaf initiation in Arabidopsis thaliana. Plant Physiology 159(4), 1501-1510. 22

Guo M., Thomas J., Collins G., and Timmermans M.C.P. (January 2008). Direct

Repression of KNOX Loci by the ASYMMETRIC LEAVES1 Complex of Arabidopsis.

The Plant Cell Online 20, 48-58.

Hagemann W. and Gleissberg S. (1996). Organogenetic capacity of leaves: The significance of marginal blastozones in angiosperms. Plant Syst. Evol. 199, 121-152.

Hay A. and Tsiantis M. (2010). KNOX genes: versatile regulators of plant development and diversity. Development 137, 3153-3165.

Jones C.S., Bakker F.T., Schlichting C.D., and Nicotra A.B. (2009). Leaf shape evolution in the South African genus Pelargonium L’ Hér. (Geraniaceae).

Evolution 63, 479-497.

Kadereit J.W., Schwarzbach A.E. and Jork K.B. (1997). The phylogeny ofPapaver s. l. (Papaveraceae): Polyphyly or monophyly? Plant Syst. Evol. 204, 75-98

Kawamura E., Horiguchi G., and Tsukaya H. (2010). Mechanisms of leaf tooth formation in Arabidopsis. Plant J. 62, 429-441.

Kerstetter R.A., Bollman K., Taylor R.A., Bomblies K., and Poethig R.S. (2001).

KANADI regulates organ polarity in Arabidopsis. Nature 411, 706-709.

Kim M., McCormick S., Timmermans M., and Sinha N. (2003). The expression domain of PHANTASTICA determines leaflet placement in compound leaves.

Nature 424, 438-443.

Klingenberg, C. P. 2008. Novelty and ‘‘homology-free’’ morphometrics: What’s in a name? Evolutionary Biology 35, 186–190.

Klingenberg, C. P. 2010. Evolution and development of shape: integrating quantitative approaches. Nature Reviews Genetics 11, 623–635.

Koyama T., Mitsuda N., Seki M., Shinozaki K., and Ohme-Takagi M. (2010). TCP

Transcription Factors Regulate the Activities of ASYMMETRIC LEAVES1 and miR164, as Well as the Auxin Response, during Differentiation of Leaves in Arabidopsis. The

Plant Cell Online 22, 3574-3588.

Lange M., Orashakova S., Lange S., Melzer R., Theißen G., Smyth D.R., and Becker

A. (2013). The seirena B class floral homeotic mutant of California poppy (Eschscholzia californica) reveals a function of the enigmatic PI motif in the formation of specific multimeric MADS domain protein complexes. Plant Cell 25, 438-453. 24

Li Z., Li B., Shen W., Huang H., and Dong A. (2012). TCP transcription factors interact with AS2 in the repression of class-I KNOX genes in Arabidopsis thaliana. Plant

J. 71, 99-107.

Molinero-Rosales N., Jamilena M., Zurita S., Gómez P., Capel J., and Lozano R.

(1999). FALSIFLORA, the tomato orthologue of FLORICAULA and LEAFY, controls flowering time and floral meristem identity. 20, 685-693.

Moon J., Candela H., and Hake S. (2013). The Liguleless narrow mutation affects proximal-distal signaling and leaf growth. 140, 405-412.

Nicotra B., Leigh A., Boyce A., Jones C.S., Niklas K., Royer D., and Tsukaya

H. (2011). The evolution and functional significance of leaf shape in the angiosperms.

Functional Plant Biology 38, 535-552.

Nikovics K., Blein T., Peaucelle A., Ishida T., Morin H., Aida M., and Laufs P.

(2006). The balance between the MIR164A and CUC2 genes controls leaf margin serration in Arabidopsis. Plant Cell 18, 2929-2945.

Ori N., Cohen A.R., Etzioni A., Brand A., Yanai O., Shleizer S., Menda N.,

Amsellem Z., Efroni I., Pekker I., Alvarez J.P., Blum E., Zamir D., and Eshed Y. 25

(2007/06//print). Regulation of LANCEOLATE by miR319 is required for compound- leaf development in tomato. Nat Genet 39, 787-791.

Posluszny, U. and P. B. Tomlinson. 2003. Aspects of inflorescence and floral development in the putative basal angiosperm Amborella trichopoda (Amborellaceae).

Canadian Journal of Botany 81:28-39.

Prigge M.J., Otsuga D., Alonso J.M., Ecker J.R., Drews G.N., and Clark S.E. (2005).

Class III Homeodomain-Leucine Zipper Gene Family Members Have Overlapping,

Antagonistic, and Distinct Roles in Arabidopsis Development. The Plant Cell

Online17, 61-76.

Ramirez J., Bolduc N., Lisch D., and Hake S. (2009). Distal expression of knotted1 in maize leaves leads to reestablishment of proximal/distal patterning and leaf dissection.

Plant Physiol. 151, 1878-1888.

Sarojam R., Sappl P.G., Goldshmidt A., Efroni I., Floyd S.K., Eshed Y., and

Bowmana J.L. (2010). Differentiating Arabidopsis shoots from leaves by combined

YABBY activities. Plant Cell 22, 2113-2130.

Shleizer-Burko S., Burko Y., Ben-Herzel O., and Ori N. (2011). Dynamic growth program regulated by LANCEOLATE enables flexible leaf patterning.

Development 138, 695-704.

Timmermans M.C.P., Hudson A., Becraft P.W., Nelson T. (1999). ROUGH

SHEATH2: a Myb protein that represses knox homeobox genes in maize lateral organ primordia. Science 284:151–53

Tomescu A.M.F. (2009). Megaphylls, microphylls and the evolution of leaf development. Trends Plant Sci. 14, 5-12.

Vialette-Guiraud A.C.M., Adam H., Finet C., Jasinski S., Jouannic S., and Scutt

C.P. (2011). Insights from ANA-grade angiosperms into the early evolution of CUP-

SHAPED COTYLEDON genes. 107, 1511-1519.

Waites R., Selvadurai H., Oliver I., and Hudson A. (1998). The PHANTASTICA gene encodes a MYB transcription factor involved in growth and dorsoventrality of lateral organs in Antirrhinum. Cell 93, 779-789.

Wang W., Lu A.-., Ren Y., Endress M.E., and Chen Z. (2009). Phylogeny and classification of Ranunculales: Evidence from four molecular loci and morphological data. Perspect. Plant Ecol. Evol. Syst. 11, 81-110. 27

Worberg A., Quandt D., Barniske A., Löhne C., Hilu K.W., and Borsch T. (2007).

Phylogeny of basal eudicots: Insights from non-coding and rapidly evolving DNA. 7, 55-

77.

Zuccolo A., Bowers J.E., Estill J.C., Xiong Z., Luo M., Sebastian A., Goicoechea

J.L., Collura K., Yu Y., Jiao Y., Duarte J., Tang H., Ayyampalayam S., Rounsley S.,

Kudrna D., Paterson A.H., Pires J.C., Chanderbali A., Soltis D.E., Chamala S.,

Barbazuk B., Soltis P.S., Albert V.A., Ma H., Mandoli D., Banks J., Carlson J.E.,

Tomkins J., dePamphilis C.W., Wing R.A., and Leebens-Mack J. (2011). A physical map for the Amborella trichopoda genome sheds light on the evolution of angiosperm genome structure. Genome Biol. 12 (5), R48.

CHAPTER 2: REGULATION OF DISSECTED LEAF ARCHITECTURE IN

ESCHSCHOLZIA CALIFORNICA BY A CIN-TCP GENE

Abstract

Dissected leaves are characterized by a prolonged organogenetic phase that allows

leaflets and other marginal structures to form before maturation terminates organogenetic

competency of the primordial leaf margins. Dissected leaves are widespread in many

eudicot lineages and are morphologically diverse. In core eudicots, the CINCINNATA gene promotes tissue differentiation and maturation and antagonizes factors responsible for marginal organogenesis. Here we present an analysis of EcCIN, a CINCINNATA homolog in the basal eudicot species Eschscholzia californica. Silencing of EcCIN using virus-induced gene silencing resulted in increased dissection of the leaf blade. Additional leaf segments formed particularly in proximal and central portions of the blade, while the acropetalous initiation of leaflets along the rachis was not affected, suggesting a role for

EcCIN in the termination of later phases of leaf organogenesis. Silencing of a

CINCINNATA homolog in the fumarioid poppy Cysticapnos vesicaria had similar effects, suggesting conservation of CIN gene function among dissected members of the poppy family. Further, silencing of the related EcTCP2-LIKE and EcTCP5-LIKE genes in

Eschscholzia indicated that these CIN-TCP genes may contribute to the maturation- promoting role of EcCIN.

Introduction

The morphological diversity of plants is evidence for the flexibility of the genetic pathways that direct the development of plant tissues and organs. Intensive study of development in core eudicots such as Arabidopsis thaliana has identified many of the genetic regulatory pathways responsible for organogenesis, and it is now possible to establish other species as phylogenetic landmarks for evolutionary-developmental studies, as well as to compare between close relatives with divergent morphological features.

Leaf development can be considered as the sum of two phases: morphogenesis, the establishment of growth axes and patterning of the leaf blade, and histogenesis, the fixation of anatomical structures and the determination of cell fates (Hagemann and

Gleissberg 1996; Shleizer-Burko et al., 2011). Dissected leaves produce lobes, leaflets and/or other marginal structures during primary morphogenesis by controlled but indeterminate growth at the undifferentiated edges of the leaf blade, the marginal blastozone (Hagemann and Gleissberg, 1996).

The duration for which the primordium maintains an organogenetic state strongly influences leaf morphology (Shleizer-Burko et al. 2011). At the threshold between morphogenesis and histogenesis, the rate of mitosis falls and cell expansion begins

(Tsukaya, 2006), precluding the formation of additional and/or higher orders of marginal dissection (Hagemann and Gleissberg, 1996).

The duration for which the primordium maintains an organogenetic state strongly influences leaf morphology (Shleizer-Burko et al. 2011). Retention of organogenetic 30 competence at the margins of the blade (the marginal blastozone) permits formation of complex structures. Therefore, prolonged expression of genes promoting and maintaining cell division would be expected to enhance the complexity of the margin while expression of those favoring the termination of division and cell expansion would simplify the margin. Interspecific heterochronic expression of genes controlling cell division and cell expansion may underlie morphological variation between the dissected leaves of sister taxa (Hagemann and Gleissberg 1996, Efroni et al., 2008).

Shleizer-Burko et al. (2011) demonstrated that differing spatial-temporal expression patterns for the gene LANCEOLATE in the Solanaceae were associated with alternative developmental programs and final leaf morphologies; simple-leafed species expressed LANCEOLATE early (or maintained their incipient stage for longer before a brief morphogenetic period), while dissected-leaf species expressed it later or restricted its activity at the leaf margins for longer. In tomato, dominant mutations preventing microRNA-mediated degradation render the leaf margins simple (Ori et al., 2007; Nag et al., 2009).

LANCEOLATE is a member of the TCP (TEOSINTE BRANCHED1,

CINCINNATA and PROLIFERATING CELL FACTORS 1 and 2) family of non-canonical basic helix-loop-helix (bHLH) transcription factors (Ori et al. 2007, Cubas et al. 1999,

Martin-Trillo and Cubas, 2010). TCP genes have reported roles in stem branching, zygomorphic flower development, and leaf morphogenesis, and are segregated into two classes based upon TCP domain-specific differences, as well as the presence or absence of a binding site for miR319 (Nag et al., 2009) and a conserved, coiled-coil R domain, 31 which are found in Class II (CIN and CYC subfamily) genes (Martin-Trillo and Cubas,

2010). LANCEOLATE is a member of the CIN-like subgroup of Class II.

Studies into LANCEOLATE and its homologs in Arabidopsis thaliana (TCPs 3, 4 and 10) and Antirrhinum majus (CINCINNATA ) indicate a conserved role in lateral organ development via the suppression of cell division and the promotion of maturation

(Koyama et al., 2007; Nath et al., 2003; Crawford et al., 2004). Loss-of-function mutations of CINCINNATA that permit uncontrolled mitosis distort surface curvature in simple Antirrhinum leaves but conversely reduces flower petals (Nath et al. 2003) while suppression or (multiple) mutations of TCPs 3, 4 and 10 causes disruption the normally smooth surfaces of Arabidopsis leaves (Koyama et al. 2010) and petals (Koyama et al.

2011). Not only is the rate of cell division affected, but so is the process of differentiation; in dominant TCP4 mutants, leaf epidermal cell shape is modified

(Sarvepalli and Nath, 2011).

The multiple targets of these genes include: ASYMMETRIC LEAVES 1 (Koyama et al. 2010) and ASYMMETRIC LEAVES 2 (Li et al. 2012), which are upregulated, form the main components of a regulatory complex that suppresses the pro-cell division Class I

KNOX genes (Byrne et al., 2002); miR164 (Koyama et al. 2010), suppressor of boundary genes CUP-SHAPED COTYLEDONS 1 and 2 (Laufs et al. 2004; Koyama et al., 2007) that are involved in the specification of leaflets and sinuses (Nikovics et al., 2006), is upregulated; INDOLE-3-ACETIC ACID3/SHORT HYPOCOTYL2 (IAA/SHY2) which negatively regulates auxin signaling (Tian et al., 2002) and the early auxin response gene

SMALL AUXIN UP RNA (Koyama et al.. 2011, Hagen and Guilfoyle, 2002) are also 32

upregulated. Cellular responses to cytokinin are diminished; Efroni et al. (2013) found

that in Arabidopsis, TCP4 interacts with SWI3C and BRAHMA, a SWI/SNF family

ATPase involved in chromatin remodeling. These act in concert to bind to the promoter

and induce activation of ARR16, an auxin response regulator that suppresses responses to cytokinin (Efroni et al., 2013).

CINCINNATA more directly affects the cell cycle by targeting Pcl5, which is

downregulated (Aggarwal et al., 2011). Pcl5 is a component of the complex that represses the cyclin‐dependent kinase inhibitor Sic1 by phosphorylating it, thereby preserving the CDK complex Cln‐Cdc28, and promoting cell cycle progression

(Aggarwal et al., 2011). TCP4 (and by inference CINCINNATA and LANCEOLATE proteins also) obstructs cell division by stabilizing Sic1.

Therefore, TCP4/CINCINNATA/LANCEOLATE act upon multiple targets in different regulatory networks to promote a transition to maturation, and interspecific differences in mature morphology may be the products of subtle modifications of their expression, or specificity for and interactions with upstream or downstream partners.

One may hypothesize that developmental control via this regulatory network may be an ancestral characteristic of plants, repeatedly recycled. A lack of involvement by

CINCINNATA would contradict the notion that dissection is the ancestral characteristic of the eudicot leaf.

Discovery of the target specificity, expression pattern and regulation of the homologous genes in basal eudicots (as opposed to the core eudicots Arabidopsis thaliana or Antirrhinum majus) would be informative as to the evolutionary history of 33

this gene family and its role in patterning the leaf blade. The basal eudicot species

Eschscholzia californica (Papaveraceae, Papaveroideae) and the fumarioid Cysticapnos

vesicaria (Papaveraceae, Fumarioideae) are the subjects of recent evolutionary

developmental studies of lateral organ formation (Wege et al. 2007, Hidalgo et al. 2012).

The species and members of neighbouring subfamilies yet differ in growth habit and leaf

morphology; Eschscholzia produces polyternately dissected leaves with leaflets indistinct

from their petiolules, while Cysticapnos leaves are once-pinnately dissected, and develop

terminal leaflets that are sometimes branched. In both cases, leaflet development is

acropetalous.

We describe the role of a development-regulating gene, CINCINNATA, in these

two species, and make comparisons with core eudicots homologs, and with other CIN-

TCP homologues found in Eschscholzia. Gene structures and virally induced silencing

phenotypes are described for Eschscholzia californica CINCINNATA (EcCIN), EcTCP2-

LIKE and EcTCP5-LIKE genes, and for Cysticapnos vesicaria CINCINNATA (CvCIN), and are discussed in the context of the evolutionary-developmental biology of the leaf.

Methods

Sequence identification

Basal eudicots DNA sequences with homology to Class II CIN-TCP genes were obtained for Eschscholzia californica, Chelidonium majus, Glaucium flavum, Argemone mexicana, Papaver somniferum, Papaver bracteatum, Papaver setigerum, and Papaver rhoeas from the Phytometasyn and 1KP transcriptome databases using exhaustive blastn 34 and tblastx searches with Arabidopsis sequences and searches by annotation using

“TCP”, “CINCINNATA”, “LANCEOLATE”, and related key words. Nucleotide sequences encoding encoding the TCP domain as delineated in Navaud et al. (2007) were translated using Expasy TRANSLATE (Gasteiger et al., 2003) then aligned by MAFFT v7 G L-ins-

I (Katoh and Standley, 2013) for maximum accuracy to confirm similarity to known TCP genes. Verified DNA sequences sequences were aligned and the consensus sequences produced were used to design degenerate primers for cloning of CIN homologs from

Cysticapnos vesicaria and Hypecoum procumbens.

Expression profiling

Eschscholzia californica seeds were stratified at 4°C for three days before germination under constant light at 22°C. Root, vegetative shoot tip (including shoot apical meristem and young leaves), mature leaf, floral bud and whole flower tissue was flash frozen in liquid nitrogen and homogenized, with three biological replicates per tissue type. RNA was extracted using an RNeasy RNA extraction kit (Qiagen) and quantified using a Nanodrop ND-1000. Reverse transcription of 200ng per sample of

RNA to cDNA was performed with MMLV reverse transcriptase (Promega). Expression levels of EcCIN, EcTCP2-LIKE, and EcTCP5-LIKE were assessed by quantitative polymerase chain reaction by comparison to the control gene actin-2 at 55°C respectively

(one minute elongation) using the primers EcCIN-21F and EcCIN-22R, EcTCP2-6F and

EcTCP2-7R, EcTCP5-6F and EcTCP5-7R, and ACTIN-2-fwd and ACTIN-2-rev.

In situ hybridization

Shoot tips from Eschscholzia californica seedlings with two leaves were fixed in

4% paraformaldehyde in PBS buffer, dehydrated by washing with ethanol : DEPC-treated

water solutions of increasing ethanol concentration up to 100%, cleared with Histoclear

and infiltrated with paraffin wax (Tissue-Tek). Infiltrated shoot tips were embedded in

blocks, cut into 10μm sections with a rotary microtome and mounted on poly-lysine

coated glass slides. The sections were probed for the expression of EcCIN using a

digoxygenin-labeled RNA probe. Primers EcCIN-15F and EcCIN-16R, the latter

prefixed with a T3 RNA polymerase binding site, were used to transcribe an antisense

RNA probe in vitro using T3 polymerase (Fisher Scientific). This sequence corresponds

to the EcCIN TCP domain. Signal detection was performed using alkaline phosphatase- conjugated Fab fragments with BCPIP as the substrate.

Phylogenetic analysis

Full-length nucleotide coding sequences for Class II CIN-TCP genes from taxa

with multiple unique representatives were translated into amino acid sequences with

Expasy Trasnslate. These were aligned on the MAFFT v7 server using the G L-ins-I

algorithm (Katoh and Standley, 2013). The alignment was checked in MacClade 4 and

converted to Phylip 3.6 format. Model testing with ProtTest 2.4 (Abascal et al., 2005)

recommended the JTT+G model according to the AIC (Akaike Information Criterion),

AICc, and BIC. A consensus phylogenetic tree with branch posterior probabilities were 36

inferred with MrBayes 3.2 (350,000 generations, standard deviation of split frequencies

<0.01; sump burnin = 1000; sumt burnin=1000) (Ronquist et al., 2012).

Cloning and VIGS construct preparation

The TCP domain of EcCIN was amplified by touchdown polymerase chain reaction (PCR) from oligo-dT primed complementary DNA (cDNA) derived from

Eschscholzia vegetative shoot tips using degenerate primers derived from Antirrhinum majus CINCINNATA and Arabidopsis thaliana TCP4 sequences (forward primer and

reverse primer) (denaturation at 94°C; initial annealing at 64°C, decreasing 0.5°C per

cycle to 55°C then maintained at 50°C for an additional 25 cycles; one minute elongation

per cycle with a ten minute final elongation period). The size of the amplified fragment

was confirmed by gel electrophoresis, excised, and purified using a Gel Extraction Kit

(Denville), then ligated into pGEM-T overnight at 4°C with the pGEM-T Easy Vector

System I kit (Promega). A 200μl aliquot of chemically competent JM109 E. coli was transformed with the ligation by heat shock transformation, incubated for one hour with

800μl antibiotic-free LB (37°C, 200rpm shaking) then plated onto LB agar plates containing 50μg/ml ampicillin as well as X-Galactose and IPTG. Colonies positive for the transformed plasmid were selected by blue-white screening and PCR with plasmid specific primers T7 and SP6 (55°C annealing, one minute elongation, 35 cycles), cultured overnight at 37°C in 5ml LB with 50μg/ml ampicillin (37°C, 200rpm shaking) and purified using an EZNA Plasmid isolation Kit II (Omega Bio-tek). The sequence of the plasmid insert was verified by Sanger sequencing (Ohio University Genomics 37

Facility) and is supported by the Eschscholzia californica transcriptome contig

NKJC0015470 (1KP).

Additional downstream sequence data was obtained by rapid amplification of cDNA ends using a 3'/5' RACE kit (Promega) and the nested, gene specific primers

EcCIN-F8 and EcCIN-15F to ensure the specificity of the product.

Two EcCIN fragments were inserted into a vector derived from the tobacco rattle virus, pTRV2. One construct comprises the EcCIN TCP domain in the vector and the other instead containing a downstream region that, based on phylogenetic analysis, is less conserved. The latter construct for comparison and ensured that cross-silencing of related TCP genes containing the conserved TCP domain was not the source of any phenotype. The TCP domain was amplified by PCR (55°C, one minute elongation, 35 cycles) with nested primers tailed with EcoRI and BamHI restriction sites (EcCIN-8F and

EcCIN-R9 for the TCP-specific construct, EcCIN-17F and EcCIN-18R for the downstream construct).

The product was verified for its size by gel electrophoresis, excised from the gel and purified using an EZNA Gel Extraction Kit (Omega bio-tek). The product was digested at 37°C for one hour with EcoRI and BamHI (Promega). The restriction enzymes were heat inactivated at 70°C for ten minutes. Empty pTRV2 vector was similarly digested with EcoRI and BamHI and purified using an EZNA Gel Extraction kit

(Omni Bio-Tek) protocol for enzymatic reaction cleanup. The two products were ligated together overnight at 4°C with T4 DNA ligase (Promega), transformed into JM109 cells by heat shock transformation and cultured overnight at 37°C on LB agar plates 38

containing 50μg/ml kanamycin. Colonies containing plasmids with the desired insert were identified by PCR with the primers pTRV2-fwd and pTRV2-rev (55°C annealing,

35 cycles) and cultured overnight in 5ml LB containing 50μg/ml kanamycin. Plasmids

were purified with the EZNA Plasmid Isolation kit II. Possession of the desired insert by

the plasmid was verified by Sanger sequencing.

The pTRV2 vector containing the insert was transformed into 100μl

Agrobacterium tumefaciens strain GV3101 by electroporation at 1800mV and incubated

at 28°C with 900ul antibiotic-free LB for one hour before being plated on LB agar plates

containing 50μg/ml kanamycin and 50μg/ml gentamycin. Positive colonies were selected

by PCR with the primers pTRV2-fwd and pTRV2-rev (55°C annealing, one minute

elongation, 35 cycles) and restreaked on a fresh antibiotic plate. The second construct

was similarly prepared from the pGEM-T vector containing the 3' RACE product. A

region of the insert was amplified with the nested primers EcCIN-17F and EcCIN-18R

(annealing at 55°C, one minute elongation, 35 cycles), purified and appropriately digested along with the pTRV2 vector before cloning into E. coli and A. tumefaciens.

The contig Eschscholzia californica RKGT-0016406 (1KP) is homologous to

Arabidopsis thaliana TCPs 2 and 24. A 308bp section of the sequence was amplified using the KpnI- and XbaI-flanked primers EcRKGT-0016406-3F and EcRKGT-0016406-

4R by 35 cycles of PCR at 58°C (one minute elongation) then cloned into pGEM-T

(JM109 E. coli) and into pTRV2 (JM109 E. coli and GV3101 A. tumefaciens) as described for EcCIN. Likewise, the contig Eschscholzia californica UNTP-0023803

(1KP) is homologous to Arabidopsis thaliana TCPs 5, 13 and 17, and a 301bp section of 39

this sequence spanning the TCP domain was similarly obtained using primers with

flanking KpnI and XbaI sites (primers EcUNTP-0023803-3F and EcUNTP-0023803-4R

respectively). These amplificates were cloned into pGEM-T and pTRV2 and finally

transformed into A. tumefaciens.

Partial sequences of the CIN homologues of Cysticapnos vesicaria and another

basal eudicot, Hypecoum procumbens, were obtained using degenerate primers based

upon TCP domain sequences identified during the phylogenetic analysis (primers

PapCIN-1F and PapCIN-2R) using vegetative shoot tip cDNA from the respective species as a template. After cloning into pGEM-T, VIGS constructs were prepared as

previously described with nested primers flanked with enzymes XbaI and KpnI (CvCIN-

1F and CvCIN-2R). A Hypecoum VIGS constructs for the positive control PHYTOENE

DESATURASE (PDS) was produced by the same approach with the generic primers PDS-

5F and PDS-6R and the specific nested primers HpPDS-1F (with XbaI site) and HpPDS-

2F (with SmaI site). An HpCIN construct was prepared with HpCIN-1F and HpCIN-2R

(with XbaI and KpnI sites respectively). PDS construct preparation for Eschscholzia and

Cysticapnos has been described in Wege et al. (2007) and Hidalgo et al. (2012).

Infiltration technique

To prepare the VIGS infiltration mixture, single colonies containing the pTRV2

construct of interest were cultured for 24 hours in 5ml LB containing 5ml LB containing

50μg/ml kanamycin and 50μg/ml gentamycin. Colonies containing the pTRV1 plasmid

(Wege et al., 2007) were cultured identically. Cells of each type were separated from 40

their media by centrifuging 1ml of each culture at 5000 x g for 30 seconds then

resuspended together in 1ml 5% w/v sucrose. Seedlings with between one and three

leaves were mechanically wounded at the hypocotyl and 2µl of the infiltration mixture

was pipetted onto the wound.

Photobleaching of control pTRV2-PDS plants in Eschscholzia and Cysticapnos

arising between eight and fifteen days after infiltration indicated the efficacy of VIGS.

Hypecoum procumbens was found to be unamenable to VIGS with pTRV2 constructs

derived from Hypecoum procumbens PDS (HpPDS) and CIN (HpCIN) genes (pTRV2-

HpPDS n = 16 plants; pTRV2-HpCIN n = 16).

Plant culture

Seeds were sown in standard flat trays with either 48 or 32 larger wells in a

standard potting soil with good drainage and covered with clear lids. After sowing, the

seeds were stratified at 4°C and in darkness for between three days and one week before

transfer to constant light (concentration) at 22°C. Plants were watered with tap water for

two weeks after germination, after which the water was supplemented with 250µl/l

‘Grow 7 - 9 - 5’ fertilizer (Dyna-Gro).

RNA extraction, reverse transcription, and demonstration of CvCIN downregulation by quantitative PCR

RNA was isolated from lateral shoot tips of pTRV2-empty and pTRV2-CvCIN

VIGS plants. pTRV2-empty plants showed no abnormal phenotypes. To select control 41

plants, the number of segments per leaf was averaged for nodes 7 – 12 for each pTRV2-

empty plant, and those closest to the mean were selected. Three pTRV2-CvCIN plants

with strong silencing were selected from those whose average number of segments per

leaf was more than one standard deviation greater than the average of the pTRV2-empty group.

Total RNA was isolated with Tri Reagent (Sigma), chloroform, and isopropanol, then washed with 70% ethanol in DEPC-treated distilled water and resuspended in

DEPC-treated distilled water. Isolated RNA was assessed for concentration using a

Nanodrop ND-1000 and for quality on a Bioanalyzer Nano chip, and then was stored at -

80°C until use. Two micrograms of RNA per sample was reverse transcribed into cDNA at 37°C using MMLV-reverse transcriptase (Promega) and random hexamer primers in a

1:1 ratio to RNA (Promega). The cDNA was diluted 1/10 with nuclease-free water.

The cDNA was used as a template for quantitative polymerase chain reaction

(QPCR) for CvCIN and the control gene ACTIN-2 using the primers ACTIN-2-fwd and

ACTIN-2-rev. Each reaction (1μl cDNA sample, appropriate primer pair, and SYBR master mix) was performed in triplicate, alongside negative controls (nuclease-free water) and RNA to confirm the absence of DNA contamination.

Preservation of voucher specimens

Voucher specimens for cultivated Eschscholzia californica, Cysticapnos vesicaria, and Hypecoum procumbens were deposited at the Bartlett Herbarium (BHO) at

Ohio University, Athens, Ohio 45701 (accession numbers pending). 42

Documentation of phenotypes

The number of leaflets per leaf was counted for nodes six through twelve in VIGS experiment plants. Phenotypes were documented with a Canon 7D digital SLR camera equipped with a 100mm f/2.8 macro lens or 65mm MP-E f/2.8 macro lens and a Canon

MT-24EX twin-light flash system.

Table 1. Primers used for cloning, in situ hybridization probe preparation and quantitative PCR for CIN-TCP genes

Primer name Sequence

EcCIN-21F TTCAAGACTTGGGGTAGTAAGAGG EcCIN-22R AACAGTAGATGCAGTTGGTCTCC EcTCP2-6F AAGGAAAAACCCGAAGAACC EcTCP2-7R TTGAGCTTGAACCGAAAAGC EcTCP5-6F GATCCAAACCTCCATCTTCG EcTCP5-7R CCAAAAGTACGGGAAACACG CvCIN-3F CTCAGCGAGTTCATCAATGG CvCIN-4R CAAGGTGGTCACATTGTTCG EcCIN-8F AAGAATTCCCAAGAGATCGAAGAGTTCGTCTTTCAGC EcCIN-9R AAAGGATCCGTGTCGGCAATGGAATCAGAGTCC EcCIN-17F CAAAGAATTCCAGAAATGGGTAGGTTTCAGAG EcCIN-18R GCAAAGGATCCAACAGGAATGAAAACC EcRKGT-0016406-3F AAATCTAGAGTTGGTGGTTTTCATGTTGG EcRKGT-0016406-4R AACGGTACCTTAGTTGCTTTGGGGTTTCG EcUNTP-0023803-3F AAAGGTACCTGGCTACGACAAGATCATCG EcUNTP-0023803-4R AAGTCTAGACCATTGATTGAGAATACTGACC PapCIN-1F AAGAATTCGAGGTACAAGGWGGYCACATTG PapCIN-2R TTCCTAGGGAGTTACTGGACTGAAGGGGTC CvCIN-2R ATCGGTACCGGCGATTAAGTTGGGTAACG PDS-5F AATCTAGACGAGTAACTGATGAGGTGTTTATTGC PDS-6R AACCATGGAGCATGGTTCCAAGATGGC HpPDS-1F AAATCTAGAGTAACCCTCCTGAGAGACTTTGCATG HpPDS-2R AAACCCGGGGAGGGGACTTCTGCTGAAGAGTAG HpCIN-1F ATCTCTAGATGATCAAGAAGGCCAAATCC HpCIN-2R ATCGGTACCGGTCAGGTGGGTAATTCTGG ACTIN2-Fwd TTACAATGAGCTTCGTGTTGC ACTIN2-Rev TCCAGCACAATACCTGTAGTA

Results

CIN is conserved between basal and core eudicots

Eschscholzia Class II TCP genes were identified by PCR amplification from

cDNA using degenerate primers based on model plant species and from the online

transcriptome databases 1KP and Phytometasyn. A single Eschscholzia sequence was

identified for each of the three CIN-TCP subclades delineated in the phylogeny

constructed by Martin-Trillo and Cubas (2010), which Arabidopsis thaliana TCP 3, 4 and

10, 2 and 24, and 13, 5 and 17 respectively. Partial sequences from single copies of CIN

were obtained from Cysticapnos vesicaria and Hypecoum procumbens. Gene

composition of the homologs was similar to those of Arabidopsis, with EcCIN (Figure 2)

and EcTCP2-LIKE containing similar miRNA319 binding regions and EcTCP2-LIKE

containing a somewhat diverged and lengthened R domain, while EcTCP5-LIKE

contained neither. Interestingly, the TCP domain of EcTCP5-LIKE includes the sequence

IDEL at its 3’ end, a feature that while common in other TCP genes has been lost from

Arabidopsis TCP5.

Figure 2. Eschscholzia TCP gene structures. EcCIN, EcTCP2-LIKE and EcTCP5-LIKE contain a conserved TCP domain; EcCIN and EcTCP2-LIKE contain a 3’ miR319 binding site, the latter also containing an R domain. Note the different positions of the miR319 binding site.

Sequences representative of the TCP3/4/10 clade were also found for species

from Chelidonium, Argemone, Sanguinaria, Corydalis and Papaver. Within these, the

only variable amino acid residues in the TCP domain were at positions 54 (Alanine

(hydrophobic) or Serine (polar neutral)), 55 (Serine in Eschscholzia versus Alanine in

other) and 59 (polar neutral Glutamine in Argemone versus acidic glutamic acid in others). The exchange of histidine with proline at residue 25 that is evident in TCP3 but not TCP4 in Arabidopsis thaliana, as well as in the TCP13/5/17 clade, has not occurred in the Papaveraceae homologs; however, substantial DNA and amino acid sequence variation between genera can be found outside of these conserved regions.

Gene duplications have occurred for Argemone mexicana (CIN-like) and Papaver rhoeas (TCP2-like) genes. Further investigation of Argemone indicated that other important developmental genes have been duplicated, suggesting chromosomal or whole genome duplications in that lineage (Figure 3). For example, 1KP identifies two homologs of PHANTASTICA (‘GOQJ-2014003-Argemone_mexicana-stem’ and ‘Ame-

CCHG-2003316-Argemone_mexicana-flower_bud’). It is unclear if this is also the case for P. rhoeas, as the absence of additional sequenced homologs in the database provides no certainty of their absence.

Figure 3. Phylogeny of amino acid sequences putatively translated from Papaveraceae and Arabidopsis thaliana CIN-TCP genes, supporting the three clades outlined by Martin-Trillo and Cubas (2010).

EcCIN is most highly expressed in shoot tips that include young leaves and the shoot apical meristem. Homologs of TCPs 2 & 24 and TCP5, 13 & 17 show similar expression.

Eschscholzia vegetative and floral tissue were harvested from at least three plants and expression levels for CIN-TCP genes were quantified in triplicate and normalized

against the expression of ACTIN-2. The expression profiles of EcCIN and the other CIN-

TCP homologs (Figure 4, Figure 5) are largely consistent with those of Arabidopsis

TCP4 (Koyama et al., 2007). Expression of EcCIN is low in root tissue compared to

maturing lateral organs, but interestingly, apparently mature organs continue to express 47

these genes, albeit at low levels. This is consistent with prolonged expression of

LANCEOLATE in compound-leafed Solanum species. In situ hybridization of EcCIN reveals its expression in leaf primordia, especially in immature leaflets. EcCIN is absent from mature stem, leaf and petiole tissue.

Figure 4. Expression profiles of EcCIN, EcTCP2-LIKE and EcTCP5-LIKE in wild-type Eschscholzia californica tissues. Expression of all genes was elevated in shoot tips and flower buds, with lower expression in roots and mature leaf and flower tissue. Note that amplification efficiency is not uniform between plots.

Figure 5. In situ hybridization of EcCIN mRNA in wild type Eschscholzia californica demonstrates that EcCIN is expressed in the immature leaf blade (BL) and developing leaflets (A – E) but not in mature stem tissue (STEM), the shoot apical meristem (SAM; A – C), or in the petioles of mature leaves (PET; A – E). Arrows indicate regions of intense expression in leaflets.

VIGS of Eschscholzia and Cysticapnos TCP genes increases the dissection of the leaf blade but floral organs are unaffected

Virally-induced silencing of EcCIN was effected with both the 3’-specific (60 plants, complete data sets for 27) and TCP –specific (64 plants, complete data sets for 16) constructs, rendering phenotypic plants (Figure 6) with increased marginal dissection 49

concentrated proximally, the distalmost leaflets remaining normally sized, compared to

pTRV2-empty control plants. The TCP-specific construct significantly increased

(Kruskal-Wallace test, p=<0.05) in nodes 8 – 12, corresponding with primordia in their incipient or morphogenetic stages at the time of infiltration. Note that while infiltration efficiency profoundly impacts statistical significance and that low efficiency of infiltration and silencing promotes Type II errors, the leaflet counts from phenotypic plants obviously skew the normal distribution of the data or form a bimodal distribution on a histogram; Kruskal-Wallace tests may therefore be unsuitable, however, a significant difference between distributions is indicated by the Kolmogorov-Smirnov test

(Figure 6).

Figure 6. Box and violin plots of total leaflets per node in plants infiltrated with pTRV2- empty (control), pTRV2-3’-specific and pTRV2 –TCP-domain constructs.

EcCIN knockdown phenotypes (exemplified in Figure 7) ranged from a moderate increase in dissection of the leaf blade to a dramatic increase that excluded only the most distal leaflets. Variation between plants along this continuum likely reflects the degree of TRV-mediated silencing. The bias towards proximal dissection and the absence of increased terminal dissection, for example, the formation of additional orders of ternate

dissection, is a consequence of the early maturation of the terminal leaflet particular to

the species, which is resistant to EcCIN silencing even in plants exhibiting a strong

phenotype, and occurs early in primordial development (Gleissberg, 2004). The order of

dissection was not obviously altered, although this has not been statistically assessed.

Figure 7. Phenotypic effects of silencing EcCIN with TCP-domain-specific and unconserved 3’-end-specific VIGS constructs. (A) Wild type leaf, node 10; (B-E) phenotypic range of silenced plants at node 10. Distal leaflets exhibit normal size and morphology while proximal leaflets are more severely dissected. This is more pronounced in the distal flanks of the primary leaflets. Leaf blade area is not conspicuously increased as a product of dissection, however, and the number of primary leaflets is not increased.

No changes in petal surface curvature, pigmentation or dissection were observed

in either treatment group, even in plants exhibiting strong leaf phenotypes, contrasting 51

with observations from Arabidopsis (Koyama et al., 2011), Antirrhinum (Crawford et al.,

2004), and Cyclamen (Tanaka et al., 2011).

Eight pTRV2-CvPDS, 17 pTRV2-empty, and 17 pTRV2-CvCIN Cysticapnos vesicaria plants were assessed for phenotypes. Silencing of CvCIN in Cysticapnos

significantly increases leaf blade dissection but tendril development is unaffected (Figure

8). VIGS knockdown of CvCIN subtly increased leaf blade dissection from node eight

through twelve (Kolmogorov-Smironov tests, P<0.05), superficially increasing proximal

leaflet dissection the most. As the terminal leaflet of Cysticapnos is replaced by a tendril

there was no conspicuous marker of differential dissection, the tendril developing

normally and without a significant change in branching or any alteration in climbing

habit. No changes in floral morphology were visible, although the number of flowers

available for study was small (<6).

Figure 8. pTRV2-empty and pTRV2-CvCIN plants (A & B respectively) and leaves from the same plants from nodes 7 – 10 (wild type = C – F; pTRV2-CvCIN = G – H). VIGS plants exhibit increased leaflet dissection. 53

Figure 9. Box and violin plots for leaflets per node in pTRV2-empty and pTRV2-CvCIN plants. Distributions differed significantly at nodes 8 and 9 (not shown), 10, 11, and 12 (Kolmogorov-Smirnov, P<0.05). PDS silencing in pTRV2-CvPDS control plants was visible at node 6.

CvCIN expression was compared between select pTRV2-empty plants (those with average leaflet counts most similar to the mean for the treatment) and select pTRV2-

CvCIN plants (those most different from the control treatment, average leaflet counts more than one standard deviation greater). CvCIN expression was lower in pTRV2-

CvCIN plants with the increased leaf dissection phenotype (Figure 10).

Figure 10. Quantitative PCR shows decreased expression of CvCIN in pTRV2-CvCIN- treated VIGS plants. Numbers appended to columns are average leaflet numbers per node for nodes 7-12, which differed significantly between the pTRV2-empty and pTRV2-CvCIN groups (Kolmogorov-Smirnov test, p<0.05). Average leaflet counts for pTRV2-CvCIN plants selected differed from the pTRV2-empty control group average by more than one standard deviation.

Silencing of additional Eschscholzia CIN-TCP genes produces similar phenotypes to

silencing of EcCIN

* * * *

Figure 11. Box and violin plots of total leaflets per node for pTRV2-empty, pTRV2- EcCIN-3’, pTRV2-EcTCP2-LIKE, and pTRV2-EcTCP5-LIKE plants. Asterisks indicate a significant difference in distribution to the pTRV2-empty group.

A second VIGS experiment to individually silence EcCIN (3’ unconserved region construct), EcTCP2-LIKE and EcTCP5-LIKE genes reiterated the inefficacy of the

EcCIN 3’ construct while revealing the importance of EcTCP2-LIKE, which is similar to

EcCIN in its regulation and has been shown to have similar effects when knocked down in Arabidopsis (Koyama et al., 2007). A small number of pTRV2-EcTCP5-LIKE plants bore leaves with greatly increased dissection, but it was unclear whether these plants were outliers or strongly silenced plants in a batch with few successfully infiltrated plants; the size of the pTRV2-EcTCP5-LIKE batch was smaller than that of the other 56 constructs (total plants = 26, complete data sets for 19, versus pTRV2-empty = 34/29, pTRV2-EcCIN-3’ = 36/35; pTRV2-EcTCP2-LIKE 34/23).

Discussion

The discovery and experimental verification of the DNA sequences described in

Eschscholzia californica supports a hypothetical early origin of three CIN-TCP subfamily genes, with roughly homologous function conserved between basal and core eudicots. As in Solanum and Antirrhinum, a single copy of a CINCINNATA homolog was identified; database mining and PCR amplification and cloning from CINCINNATA-based primers yielded one sequence. This contrasts with TCP3, 4, and 10 in Arabidopsis and PCF 5 and 7 in Oryza sativa, species that have undergone additional whole genome duplications

(Blanc et al., 2003; Guyot and Keller, 2004). This lack of redundancy precludes the modulation of expression level and domain between copies observed in TCP 3, 4, and 10

(Koyama et al., 2007).

The phenotypic changes observed in Eschscholzia and Cysticapnos VIGS plants are in accordance with reported knockdown phenotypes observed in core eudicots species, for example, in Solanum lycopersicum la-6 plants (Ori et al., 2007), in that the basic ground plan of the leaf blade is unchanged; dissection of the margins of the distal leaflets is increased but additional pairs of primary leaflets do not develop.

CINCINNATA and its homologs have been associated with maturation and the cessation of dissection of the leaf blade via downregulation of pro-morphogenesis factors such as the CUC and Class I KNOX genes. Recently, Efroni et al. (2013) showed Arabidopsis 57

TCP4 to directly promote ARR16, an auxin response regulator that decreases cell sensitivity to cytokinins, which promote cell division, rather than affecting cytokinin production per se. The absence of CINCINNATA expression does not increase baseline

cytokinin concentrations and increase cell division rate, instead cytokinin sensitivity is

maintained for a longer period, allowing continuance of dissection that ordinarily ceases

at the onset of CINCINNATA activity (i.e. when CINCINNATA is expressed in the

diminished presence or the absence of miR319) (Efroni et al., 2013).

Evidence provided by Yanai et al. (2011) indicates that in tomato, LANCEOLATE

also directly or indirectly stimulates production of gibberellic acids by SlGAox4,

providing a link to pro-maturation signalling, i.e. cell differentiation and expansion are

promoted. Involvement of LANCEOLATE and its homologs in a range of convergent pathways support its status as a master regulator of leaf development.

Given the evidence for the non-uniform expression of CIN throughout the blade,

transitioning basipetally in the simple-leafed Antirrhinum (Nath et al., 2003) while being

gradually restricted to the margins of tomato’s compound leaves (Ori et al., 2007), it is

unsurprising that distal leaflets remain comparatively undissected in Eschscholzia and

Cysticapnos and that the tendrils of the latter, to which cell expansion contributes

substantially anyway, are unchanged. An interesting comparison could be made if the

CINCINNATA homolog of a related, compound-leaved species with a basipetalous dissection programme (rather than acropetalous, as in Eschscholzia and Cysticanos) were

knocked out, wherein we might find dissection to be intensified in different parts of the

blade. Caution must be exercised, however, as the sequence of maturation of leaflets is 58

likely more influential that their sequence of initiation; in Eschscholzia, pairs of leaflets

are induced acropetally, but the terminal leaflet consistently matures first.

The absence of floral phenotypes even in the presence of changes in leaf

morphology is an indicator that the involvement of CIN-TCP genes in petal development

in species such as Antirrhinum majus is a derived characteristic, where regulators of cell division and maturation involved in formation of the leaf lamina are co-opted into development of the petal lamina; however, the expression of EcCIN in floral buds and

mature flowers opposes the idea that this is a novelty of the core eudicots. Furthermore,

we cannot assume that CIN is important only in zygomorphic flowers that are dissimilar

from those Eschscholzia, as Arabidopsis TCP4 knockdown mutants possess wavy petals.

We might therefore accept Eschscholzia and other basal eudicots as the exception rather than the rule (although the limited availability of Cysticapnos flowers invokes caution).

Observations of VIGS phenotypes support a role for CIN-TCP genes in the

patterning of the leaf blade. Furthermore, recent identification of targets of

CINCINNATA and its homologs show it to have a broad regulatory role in promoting

maturation. Functional experiments of the kind described herein indicate that the

imposition of repressive genes and genetic modules such as miR319 silencing and

expression of Class I KNOX genes onto the CIN expression pattern could act as a

mechanism to shape the leaf blade. This mechanism may be recycled in many lineages to

provide a means for diversification of form throughout eudicot evolution.

References

Abascal F., Zardoya R. and Posada D. (2005). ProtTest: Selection of best-fit models of

protein evolution. Bioinformatics: 21(9):2104-2105.

Aggarwal P., Padmanabhan B., Bhat A., Sarvepalli K., Sadhale P.P., and Nath

U. (2011). The TCP4 transcription factor of Arabidopsis blocks cell division in yeast at

G1 → S transition. Biochem. Biophys. Res. Commun. 410, 276-281.

Blanc, G., Hokamp, K., and Wolfe, K.H. (2003). A recent polyploidy superimposed on

older large-scale duplications in the Arabidopsis genome. Genome Res. 13, 137–144.

Crawford B.C.W., Nath U., Carpenter R., and Coen E.S. (2004). CINCINNATA

controls both cell differentiation and growth in petal lobes and leaves of antirrhinum.

Plant Physiol. 135, 244-253.

Cubas P., Lauter N., Doebley J., and Coen E. (1999). The TCP domain: a motif found

in proteins regulating plant growth and development. Plant J. 18, 215-222.

Efroni I., Blum E., Goldshmidt A., and Eshed Y. (2008). A Protracted and Dynamic

Maturation Schedule Underlies Arabidopsis Leaf Development. The Plant Cell Online

20, 2293-2306.

Efroni I., Han S., Kim H., Wu M., Steiner E., Birnbaum K., Hong J., Eshed Y., and

Wagner D. (2013). Regulation of Leaf Maturation by Chromatin-Mediated Modulation of Cytokinin Responses. 24, 438-445.

Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D. and Bairoch A. (2003).

"ExPASy: The proteomics server for in-depth protein knowledge and analysis". Nucleic

Acids Research 31 (13): 3784–8.

Gleissberg S. (2004). Comparative analysis of leaf shape development in Eschscholzia californica and other Papaveraceae-Eschscholzioideae. Am. J. Bot. 91, 306-312.

Guyot R, and Keller B. (2004). Ancestral genome duplication in rice. Genome. 47:610–

614.

Hagen G. and Guilfoyle T. (2002). Auxin-responsive gene expression: genes, promoters and regulatory factors. Plant Mol. Biol. 49: 373-385

Hagemann W. and Gleissberg S. (1996). Organogenetic capacity of leaves: The significance of marginal blastozones in angiosperms. Plant Syst. Evol. 199, 121-152.

Hidalgo O., Bartholmes C., and Gleissberg S. (2012). Virus-induced gene silencing

(VIGS) in Cysticapnos vesicaria, a zygomorphic-flowered Papaveraceae (Ranunculales, basal eudicots). Annals of Botany 109, 911-920.

Katoh K. and Standley (2013). MAFFT multiple sequence alignment software version

7: improvements in performance and usability. Molecular Biology and Evolution 30:772-

780.

Koyama T., Furutani M., Tasaka M., and Ohme-Takagi M. (2007). TCP transcription factors control the morphology of shoot lateral organs via negative regulation of the expression of boundary-specific genes in Arabidopsis. Plant Cell 19, 473-484.

Koyama T., Mitsuda N., Seki M., Shinozaki K., and Ohme-Takagi M. (2010). TCP

Transcription Factors Regulate the Activities of ASYMMETRIC LEAVES1 and miR164, as Well as the Auxin Response, during Differentiation of Leaves in Arabidopsis. The

Plant Cell Online 22, 3574-3588.

Koyama T., Ohme-Takagi M., and Sato F. (2011). Generation of serrated and wavy petals by inhibition of the activity of TCP transcription factors in Arabidopsis thaliana. 6,

697-699.

Laufs P., Peaucelle A., Morin H., and Traas J. (2004). MicroRNA regulation of the

CUC genes is required for boundary size control in Arabidopsis meristems. Development

131, 4311-4322.

Li Z., Li B., Shen W., Huang H., and Dong A. (2012). TCP transcription factors

interact with AS2 in the repression of class-I KNOX genes in Arabidopsis thaliana. Plant

J. 71, 99-107.

Martín-Trillo M. and Cubas P. (2010). TCP genes: a family snapshot ten years later.

Trends Plant Sci. 15, 31-39.

Nag A., King S., and Jack T. (2009). miR319a targeting of TCP4 is critical for petal growth and development in Arabidopsis. Proceedings of the National Academy of

Sciences 106, 22534-22539.

Nath U., Crawford B.C.W., Carpenter R., and Coen E. (2003). Genetic control of

surface curvature. Science 299, 1404-1407.

Navaud O., Dabos P., Carnus E., Tremousaygue D., and Hervé C. (2007). TCP

transcription factors predate the emergence of land plants. J. Mol. Evol. 65, 23-33.

Nikovics K., Blein T., Peaucelle A., Ishida T., Morin H., Aida M., and Laufs P.

(2006). The balance between the MIR164A and CUC2 genes controls leaf margin serration in Arabidopsis. Plant Cell 18, 2929-2945.

Ori N., Cohen A.R., Etzioni A., Brand A., Yanai O., Shleizer S., Menda N.,

Amsellem Z., Efroni I., Pekker I., Alvarez J.P., Blum E., Zamir D., and Eshed Y.

(2007/06//print). Regulation of LANCEOLATE by miR319 is required for compound- leaf development in tomato. Nat Genet 39, 787-791.

Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Höhna S, Larget B,

Liu L, Suchard MA, and Huelsenbeck JP. (2012). MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol. 61(3),

539-42.

Sarojam R., Sappl P.G., Goldshmidt A., Efroni I., Floyd S.K., Eshed Y., and

Bowmana J.L. (2010). Differentiating Arabidopsis shoots from leaves by combined

YABBY activities. Plant Cell 22, 2113-2130.

Sarvepalli K. and Nath U. (2011). Interaction of TCP4-mediated growth module with phytohormones. Plant Signal Behav 6 (10), 1440-1443. 64

Shleizer-Burko S., Burko Y., Ben-Herzel O., and Ori N. (2011). Dynamic growth

program regulated by LANCEOLATE enables flexible leaf patterning. Development 138,

695-704.

Tanaka Y., Yamamura T., and Terakawa T. (2011). Identification and expression

analysis of the Cyclamen persicum MADS-box gene family. Plant Biotechnology 28,

167–172.

Tian Q., Uhlir N.J., and Reed J.W. (2002). Arabidopsis SHY2/IAA3 inhibits auxin-

regulated gene expression. Plant Cell. 14(2):301-19.

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

477-496.

Wege S., Scholz A., Gleissberg S., and Becker A. (2007). Highly efficient virus- induced gene silencing (VIGS) in california poppy (Eschscholzia californica): An evaluation of VIGS as a strategy to obtain functional data from non-model plants. Ann.

Bot. 100, 641-649.

Yanai O., Shani E., Russ D., and Ori N. (2011). Gibberellin partly mediates

LANCEOLATE activity in tomato. 68, 571-582. 65

CHAPTER 3: ESCHSCHOLZIA CALIFORNICA PHANTASTICA (ECPHAN)

REGULATES PETAL MORPHOGENESIS IN THE CALIFORNIA POPPY

Abstract

Throughout evolution, genes have been co-opted from one aspect of plant development into another. The ARP genes are associated with adaxial-abaxial polarization and control of KNOX gene expression in the leaf blade and in flower petals.

In the basal eudicot Eschscholzia californica, virally-induced silencing of the homologous gene EcPHAN modifies petals to produce ectopic growth axes, partial radialization, and/or dissection of the distal edge. In contrast to mutants identified in core eudicot species, no leaf phenotype was observed, possibly due to decoupling of EcPHAN from leaf development, efficient compensation by other adaxial specifiers in the leaf, or the presence of a subfunctionalized paralog in Eschscholzia.

Introduction

The morphological diversity of plants is evidence for the flexibility of the genetic pathways that direct the development of structures, organs, and tissues. Intensive study of development in core eudicots such as Arabidopsis thaliana has identified many of the genetic regulatory pathways responsible for organogenesis, and it is now possible to establish other species as phylogenetic landmarks for evolutionary-developmental studies, as well as to compare between close relatives with divergent morphological features. 66

The plant leaf is the product of a complex four-dimensional pattern of genetic interactions. Key events that must occur during leaf development are the designation of the incipient primordium as a region distinct from the shoot apical meristem (SAM), the specification of dorsiventral (adaxial-abaxial), proximodistal, and mediolateral axes, patterning of the leaf blade, and the differentiation of cells to form the anatomy of the leaf (Blein et al., 2010). The processes may be broadly divided into three stages, initiation, morphogenesis and histogenesis, wherein the general morphology of the leaf

(especially the blade) is established by dividing, undifferentiated cells, and then the leaf ceases to change in its overall shape as the frequency of cell division slows and cells differentiate and expand. Altered regulation of the maintenance of cells in an undifferentiated state, the transition from morphogenesis to histogenesis, and the subsequent regulation of cell division and expansion creates a variety of mature morphologies and, therefore, the expression, regulation, and function genes relevant to these processes are of interest to evolutionary-developmental biologists.

The ARP genes, named after ASYMMETRIC LEAVES 1 (Arabidopsis thaliana;

Byrne et al. 2000), ROUGH SHEATH 2 (Zea mays; Schneeberger et al. 1998) and

PHANTASTICA (Antirrhinum majus; Waites and Hudson, 1995) are transcription factors of the MYB (Waites et al. 1998) family. In several eudicot species, ARP genes are involved in the specification of the incipient leaf primordium where they are involved in the initial downregulation of class I KNOX genes. ARP genes also seem to have a conserved role in the promotion and maintenance of the adaxial leaf domain (Kidner and

Timmermans 2010). Promotion of adaxial cell fates is performed in concert with the 67

Class III homeodomain leucine zipper (HD-ZIP) genes PHABULOSA, PHAVOLUTA and

REVOLUTA (Prigge et al., 2005) and other factors. KANADI and YABBY genes instead promote abaxial identity (Sarojam et al., 2010), and mutual repression interactions between adaxial and abaxial factors establish a boundary along which the primordial leaf margin forms and blade outgrowth occurs. Consequently, mutants in whom either of the two identities is not established, fail to form a margin and develop into bladeless, often radialized structures (Waites and Hudson 1995). The first such mutant characterized was

PHANTASTICA in Antirrhinum majus, in which loss of AmPHAN results in radial leaves with only abaxial surface (Waites and Hudson, 1995). Some leaves develop a margin and blade and are bifacial; however, they may produce ectopic patches of abaxial tissue on their adaxial surface, suggesting local loss of adaxial identity. These patches are surrounded by an ectopic margin marking the boundary between the two identities.

ARP activity requires interaction with homologs of the Arabidopsis thaliana genes

ASYMMETRIC LEAVES 2 and HISTONE DEACETYLASE 6 (HDA6) (Byrne et al. 2000,

Luo et al. 2012) to form a regulatory complex that represses the expression of Class I

KNOX genes (Kim et al., 2003). Rather than directly determining cell fate, the primary function of ARP genes is likely the repression of KNOX genes, which maintain cell indeterminacy in the shoot meristem and in nascent lateral organs. Recently, Koyama et al. (2010) reported activation of ASYMMETRIC LEAVES 1 by CINCINNATA-like TCP genes, which are known to promote cell fate determinacy and the transition of leaf primordia from a morphogenetic state to maturation (Nath et al,. 2003; Ori et al., 2007) 68

The role of ARP genes has also been studied in core eudicot species with

dissected leaves. In Solanum lycopersicum, ARP genes appear to control the positioning

of leaflets along the leaf axis (Kim et al., 2004). Evidence supports the notion that the

size and shape of the adaxial domain as a component of the leaf primordium dictates the

patterning of dissection (e.g. pinnate or palmate) and the placement of leaflets (Zoulias et

al. 2012). In Pisum sativum, the crispa mutation represents a loss-of-function of the pea

ARP homolog and results in abaxialized leaflets and ectopic stipules (Tattersall et al.,

2005).

ARP genes also influence the development of petals, which like leaves develop

dorsiventral, proximodistal and mediolateral axes. Mutant petals of Antirrhinum

PHANTASTICA exhibit radialized needle-like petal lobes and the establishment of

ectopic margins surrounding patches of abaxial tissue growth similar to those seen in

leaves (Waites and Hudson, 1995). Similarly, floral organs in Pisum sativum crispa

mutants exhibit polarity defects (Tattersall et al., 2005).

Although some aspects of ARP function are conserved between species, there is

considerable variation in both expression patterns and mutant phenotypes between

lineages. For instance, mutants of the Zea mays ARP gene ROUGH SHEATH 2 exhibit

no effects of leaf polarity (Timmermans et al. 1999). This study characterizes the knock-

down phenotype of a PHAN ortholog, EcPHAN, in the basal eudicot Eschscholzia

californica (Papaveraceae). Virus-induced gene silencing resulted in an abnormal petal

phenotype with ectopic growth axes; in contrast, leaf development appeared to be

unaffected. 69

Methods

Phylogenetic analyses

Putative basal eudicots and basal angiosperm homologs of AS1, RS2 and

PHANTASTICA were obtained from the 1KP transcriptome database using blastn with

EcPHAN as the search query. Similar sequences were aligned with MAFFT v7 (Katoh and Standley, 2013) and divergent sequences eliminated. Remaining, complete coding sequences were translated with the Expasy Translate tool (Gasteiger et al., 2003) and the resultant amino acid sequences were again aligned with MAFFT v7. A Phyml 4 – format file was submitted to the ProtTest 2.4 server (Abascal et al., 2005) to select an appropriate evolutionary model for Bayesian analysis. JTT + G (Jones’ model with a gamma distribution of variable sites) was selected under AIC, AICc and BIC criteria.

Phylogenetic tree form and branch posterior probabilities were inferred with MrBayes 3.2

(50,000 generations, standard deviation of split frequencies <0.001; sump burnin = 400; sumt burnin=400) (Ronquist et al., 2012).

Cloning and sequencing of EcPHAN

A 334 bp sequence of EcPHAN (AY228766.1) downstream of the conserved

MYB domain was amplified from cDNA and cloned into pTRV2 using XbaI- and SacI- tailed primers EcPh14F (TCTAGATACTTCCACCTTGGCTTT) and EcPh15R

(GAGCTCCTCTGACTCGAGTTGTAG). The resultant pTRV2-EcPHAN vector was confirmed by Sanger sequencing and cloned into Agrobacterium tumefaciens strain

GV3101.

RT-PCR expression profile

To study expression of EcPHAN, total RNA was isolated from flowers, stems

(peduncles), roots, mature leaves and vegetative shoot tips and reverse-transcribed with

AMV reverse transcriptase (Roche) and an oligo-dT adaptor primer

(GACTCGAGTCGACATCGATTTTTTTTTTTTTTTT). cDNA concentration was normalized by inspecting band brightness after gel electrophoresis of total RNA.

EcPHAN was amplified with primers EcPh399F

(AGGAAGAACAACGACTTGTAATCCGTC) and EcPh1028R

(TTCAACCCTTTTAAGCCTCCAAGCTGC), yielding a 629 bp product. Gel band brightness was inspected after 30, 35, and 40 cycles to determine a suitable pre-saturation stage of amplification, and 31 cycles were used to assess relative expression in different tissues.

VIGS Infiltration technique

To prepare the VIGS infiltration mixture, single colonies containing the pTRV2 construct of interest were cultured for 24 hours in 5ml LB containing 5ml LB containing

50ug/ml kanamycin and 50ug/ml gentamycin. Colonies containing the pTRV1 plasmid

(Wege et al., 2007) were cultured identically. Cells of each type were separated from their media by centrifuging 1ml of each culture at 5000 x g for 30 seconds then resuspended together in 1ml 5% w/v sucrose. Seedlings with between one and three leaves were mechanically wounded at the hypocotyl and 2µl of the infiltration mixture was pipetted onto the wound. The experiment was performed twice. In the first replicate 71

21 pTRV2-empty and 44 pTRV2-EcPHAN produced at least one flower. In the second replicate 13 pTRV2-empty and 88 pTRV2-EcPHAN plants flowered.

Plant culture

Seeds were sown in trays with either 48 x n cm2 or 32 x n cm2 wells in a standard potting soil with good drainage and covered with clear lids. After sowing, the seeds were stratified at 4°C and in darkness for between three days and one week before transfer to constant light (concentration) at 22°C. Plants were watered with tap water for two weeks after germination, after which 250µl/l ‘Grow 7 - 9 - 5’ fertilizer (Dyna-Gro) was added.

Scanning electron microscopy

Petals with abnormal phenotypes from pTRV2-EcPHAN plants were allowed to expand fully before removal and preservation in 70% ethanol before drying using a carbon dioxide critical point dryer (Balzers CPD 030, Bal-tec (now Leica

Microsystems)). Dried material was dissected with a razor blade as required before mounting on aluminum stubs and sputter coating (Balzers SCD 050). The specimens were viewed and photographed using a Zeiss EVO-50XVP (University of Dayton) at

15kV.

Results

EcPHAN is an ARP homolog in the Papaveraceae

EcPHAN (AY228766.1) is homologous in structure and conserved protein domains to ASYMMETRIC LEAVES 1, ROUGH SHEATH 2 and PHANTASTICA, having 72 a similar N-terminal MYB domain. Cloning and mining of the 1KP database uncovered single homologs of PHANTASTICA in Eschscholzia and other Papaveraceae species, with the exception of a duplication in Argemone mexicana. Another important leaf developmental gene, CINCINNATA, has undergone duplication in Argemone mexicana, and it is possible that a whole genome duplication has occurred. The amino acid sequence encoded by EcPHAN clusters closely with putative Papaveraceae

PHANTASTICA homologs, and is more closely related to ASYMMETRIC LEAVES 1 than to PHANTASTICA, found in the less derived species Antirrhinum majus.

Figure 12. Phylogeny of ASYMMETRIC LEAVES 1, ROUGH SHEATH 2 and PHANTASTICA and their homologs in the Papaveraceae and in the basal angiosperm Amborella trichopoda. Numbers indicate posterior probabilities. EcPHAN is differentially expressed in Eschscholzia organs 73

RT-PCR profiling of EcPHAN revealed strong expression in shoot tips with developing leaf primordia, as well as in elongated stem tissue of reproductive stems and in flowers

(Figure 13). Weak expression was found in root tissue and in mature leaf tissue.

Figure 13. RT-PCR expression profile of EcPHAN. Expression is strongest in shoot tips containing the SAM and developing leaf primordia (SAM), and in mature elongated pedicels (stem). Expression was also strong in floral tissue. Weak expression was detected in root tissue and in mature leaves.

EcPHAN VIGS plants exhibited changes in petal morphology and produced additional petaloid laminae on the adaxial surfaces of petals

Silencing of EcPHAN induced a range of abnormal petal phenotypes (phenotypic frequencies were 12/44 (27%) and 8/88 (5%) in independent experiments). In mildly affected flowers, irregular distal dissection and cleavage of petals occurred (Figure 14 B) along with irregular longitudinal ridges and folds, suggesting perturbed growth. Petals were often narrowed (Figure 14 C) and sometimes involuted (Figure 14 E, third from left). . The most conspicuous phenotype involved flaps of additional petal laminae on the adaxial surfaces (Figure 14 C). These ectopic petal flaps varied in number and width, but often extended almost as far along the proximo-distal petal axis as the regular petal 74 lamina, broadening distally. Epidermal coloration of these petal flaps was mostly pale yellow, more similar to the wild-type abaxial petal surface, contrasting with the orange color of the adaxial petal surface not covered by ectopic flaps. Ectopic petal flaps emerged from or near the base of petals, remaining or becoming fused along some or most of the petals’ lengths.

In some cases, ectopic petal flaps were connected with the regular petal along their margins, forming a double layered and partially hollow tube that superficially resembled a corollary tube in synpetalous species (Figure 14 D). The outer surface had a glossy yellow epidermis resembling the abaxial epidermis in wild-type petals, whereas the silky-orange color of wild-type adaxial epidermis was evident in the open distal part of the trumpet-shaped petal. Sepals, stamens, and gynoecia developed normally.

Phenotypes were manifest in the terminal and in lateral flowers.

Scanning electron microscopy provided details of the surface of the ectopic laminae and fusion between the ectopic and regular laminae (Figure 15). In studied examples, the superficial tissue flaps resembled normal petals in terms of cell sizes, shapes and apparent identities, however, the longitudinal ripples and ridges seen in Figure

14 (B and C) were clear on the ‘adaxial’-type faces (Figure 15 A and B). Flaps were attached at the bases of petals and remained partially fused (Figure 15 C and D) and sometimes became fused to the normal petal again at other points (Figure 14 A). 75

Figure 14. VIGS-induced EcPHAN knockdown phenotypes. In contrast to those of the wild type flower (A), the petals of pTRV2-EcPHAN plants exhibited dissection (star in B), narrowing (C, D), ectopic petaloid laminae on their adaxial surfaces (C), or hollow, tubular petals (D, E). 76

Figure 15. Scanning electron microscopy of petals exhibiting the EcPHAN knockdown phenotype. (A) An additional, ectopic lamina (right) overlaps and is partially fused to the adaxial surface of the petal below (left) . Scale = 1mm. The distal edges of both are uneven and have deep incisions. (B) Distortion of the normally smooth petal surface was observed on the adaxial (AD) faces of petals (P) and also on the near face of ectopic laminae (L). Scale = 100µm. (C) A longitudinal section through a petal with ectopic tissue showing the site of initiation (arrow) of that tissue. Scale = 200µm. The ectopic tissue remains or has become fused to the petal along part of its length. (D) Enlarged view of the initiation site (arrow). Scale = 100µm.

No abnormal phenotype was observed in the stems and leaves of pTRV2-

EcPHAN plants, even in those exhibiting strong floral phenotypes. Counts of leaflets on leaves at nodes six through ten indicated that pTRV2-EcPHAN plants had the same levels of leaf dissection as empty vector control plants (data not shown).

Discussion

The changing role of PHAN in leaf development.

PHAN genes were first studied in plants with simple leaves, Antirrhinum majus and

Arabidopsis thaliana, where expression of ARP genes appear mutually exclusive with class I KNOX gene expression. Loss-of-function of PHAN in Antirrhinum produces a range of leaf phenotypes that suggest complete or partial loss of adaxial leaf identity

(Waites and Hudson, 1995). In Nicotiana tabacum, comparable unifacial and peltate phenotypes occur, but were interpreted as resulting from delayed tissue differentiation due to extended KNOX expression when PHAN is silenced (McHale and Koning 2004).

Several loss-of-function alleles of Arabidopsis thaliana AS1 retain a bifacial, margined leaf. In maize, rough sheath2 mutants maintain adaxial-abaxial polarity (Timmermans et al., 1999). This suggests variable evolutionary roles of PHAN in simple-leaved species.

In dissected-leaved species of the core-eudicot lineage, leaf expression of PHAN has been shown to co-occur with KNOX genes. In Solanum lycopersicum, an asterid core eudicot with basipetal-pinnate leaf architecture, silencing PHAN caused a change in leaf architecture to palmate or peltate (Kim et al., 2003). In Cardamine hirsuta, a rosid core eudicot with basipetal-pinnate leaves (Hay and Tsiantis 2006), mutation in a PHAN homolog also resulted in the compression of the leaf axis and altered arrangement of leaflets. Leaves of crispa mutants in Pisum sativum, a rosid core eudicot acropetal- pinnate leaves, showed compression of the leaf axis and peltate leaflets (Tattersall et al.,

2005). These data suggest that that leaf development is affected in a species and morphology-dependent way in various core eudicots. This study presents the first knock- 78

down study of an ARP gene in a basal eudicot. Silencing of EcPHAN in Eschscholzia

californica did not produce any leaf phenotype. Leaves were bifacial and had normal

architecture. No ectopic tissue was observed, and the leaf axis was not compressed.

Levels of leaf dissection did not differ between pTRV2-empty and pTRV2-EcPHAN plants. The RT-PCR data provided in this study are consistent with co-expression of

EcPHAN in either the shoot apical meristem and/or leaf primordia in Eschscholzia, where class I KNOX genes have been shown to be expressed (Groot et al., 2005, Stammler et al., in press). Higher-resolution expression analyses of EcPHAN by in situ hybridization would be necessary to determine whether co-expression with KNOX genes occurs in this species. Together, these results may indicate that the regulation by PHAN of adaxial leaf identity, mesophyll differentiation, or leaflet patterning along the leaf axis evolved in core eudicots after the split from Eschscholzia ancestors. Alternatively, the role of

EcPHAN in these processes may be masked by the activity of redundant adaxial identity factors, such as class III HD-ZIP genes. Experiments in which Eschscholzia HD-ZIP orthologs are silenced together with EcPHAN would allow us to investigate this possibility.

EcPHAN specifically affects petal development and morphology

The formation of ectopic petal laminae from the adaxial petal surface in pTRV2-

EcPHAN flowers resembles the patches of ectopic tissue that develop in Antirrhinum phantastica mutant corollas, where they arise from a defect in the establishment or maintenance of adaxial identity, followed by the formation of a boundary-induced ectopic 79

margin. It is likely that the ectopic petal flaps reflect a role of EcPHAN as a determinant

of adaxial identity, that is conserved between basal and core eudicots. Ectopic lamina

flaps in pTRV2-EcPHAN petals can be interpreted as the induction of an entirely new lamina by the ectopic expression of KNOX genes that are normally suppressed by PHAN.

Class I KNOX genes are expressed in the Eschscholzia floral meristem but downregulated

in floral organ primordia (Groot et al., 2005). Petals in Eschscholzia plants silenced with

two KNOX I genes, SHOOTMERISTEMLESS1 (EcSTM1) and EcSTM2 does not affect

petal growth. The ectopic outgrowths of pTRV2-EcPHAN petals have their own adaxial-

abaxial polarity. The pale, abaxial-like pigmentation of the upper surface of ectopic flaps suggests homology to the wild type abaxial face. The inner lower face derives context and presumably identity from its continuity with the adaxial face of the regular petal.

Weaker phenotypes with narrowed petals with ridges and distal incisions likely reflect uneven longitudinal expansion due to adaxial tissue identity defects. Narrowed petals with incisions were also observed in EcYABBY VIGS petals that are likely due to defects in abaxial tissue identity (Bartholmes et al., in prep.).

Some petals in which the ectopic surface developed in continuum the regular petal exhibited a ridge along the center of their adaxial surface, suggesting that the margins were rolled inwards and subsequently fused; however, this contrasts with the clear separation of most ectopic laminae from the lateral edges of the petals. All ectopic flaps originated close to the proximal end of the petal, suggesting that ectopic growth is initiated early in petal development at the base of the petal primordium, prior to the expansion of the petal. The fact that ectopic flaps often expanded to almost the same 80 extent as the regular petal underneath suggests that the ectopic growth axes were initiated early, both growth axes developing over a similar time period and at similar rates.

The striking phenotype observed in pTRV2-EcPHAN VIGS plants suggests that

ARP gene function in petal morphogenesis is conserved between basal eudicots, rosid, and asterid eudicots. However, the developmental outcome of compromising PHAN function is quite different between Eschscholzia and Antirrhinum. This illustrates that the role of species or lineage-specific morphogenetic contexts in specifying specific outcomes of a conserved function, necessitating a characterization of the developmental context in the species under study. Further, the absence of a leaf phenotype in pTRV2-

EcPHAN plants suggests that EcPHAN functions have diverged from those of

PHANTASTICA in Antirrhinum and AS1 in Arabidopsis.

References

Abascal F., Zardoya R. and Posada D. (2005). ProtTest: Selection of best-fit models of protein evolution. Bioinformatics: 21(9), 2104-2105.

Blein T., Hasson A., and Laufs P. (2010). Leaf development: what it needs to be complex. Curr. Opin. Plant Biol. 13, 75-82.

Byrne M.E., Barley R., Curtis M., Arroyo J.M., Dunham M., et al. (2000).

Asymmetric leaves1 mediates leaf patterning and stem cell function in Arabidopsis.

Nature 408, 967–971. 81

Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D. and Bairoch A. (2003).

"ExPASy: The proteomics server for in-depth protein knowledge and analysis". Nucleic

Acids Research 31 (13), 3784–3788.

Groot E.P., Sinha N., and Gleissberg S. (2005). Expression patterns of STM-like

KNOX and Histone H4 genes in shoot development of the dissected-leaved basal eudicot plants Chelidonium majus and Eschscholzia californica (Papaveraceae). Plant Mol. Biol.

58, 317-331.

Hay A. and Tsiantis M. (2010). KNOX genes: versatile regulators of plant development and diversity. Development 137, 3153-3165.

Katoh K. and Standley (2013). MAFFT multiple sequence alignment software version

7: improvements in performance and usability. Molecular Biology and Evolution 30,

772-780.

Kidner C.A. and Timmermans M.C. (2010). Signaling sides adaxial-abaxial patterning in leaves. Curr Top Dev Biol. 91, 141-168.

Kim M., McCormick S., Timmermans M., and Sinha N. (2003). The expression

domain of PHANTASTICA determines leaflet placement in compound leaves. Nature

424, 438-443.

Kim S., Soltis D.E., Soltis P.S., Zanis M.J., and Suh Y. (2004). Phylogenetic

relationships among early-diverging eudicots based on four genes: were the eudicots

ancestrally woody? Mol. Phylogenet. Evol. 31, 16-30.

Koyama T., Mitsuda N., Seki M., Shinozaki K., and Ohme-Takagi M. (2010). TCP

Transcription Factors Regulate the Activities of ASYMMETRIC LEAVES1 and miR164,

as Well as the Auxin Response, during Differentiation of Leaves in Arabidopsis. The

Plant Cell Online 22, 3574-3588.

Luo M., Yu C., Chen F., Zhao L., Tian G., Liu X., Cui Y., Yang J., and Wu K.

(2012). Histone Deacetylase HDA6 Is Functionally Associated with AS1 in Repression

of KNOX Genes in Arabidopsis. 8, e1003114.

McHale N.A. and Koning R.E. (2004). PHANTASTICA regulates development of the adaxial mesophyll in Nicotiana leaves. Plant Cell 16, 1251–1262.

Nath U., Crawford B.C.W., Carpenter R., and Coen E. (2003). Genetic control of surface curvature. Science 299, 1404-1407. 83

Ori N., Cohen A.R., Etzioni A., Brand A., Yanai O., Shleizer S., Menda N.,

Amsellem Z., Efroni I., Pekker I., Alvarez J.P., Blum E., Zamir D., and Eshed Y.

(2007/06//print). Regulation of LANCEOLATE by miR319 is required for compound-

leaf development in tomato. Nat Genet 39, 787-791.

Prigge M.J., Otsuga D., Alonso J.M., Ecker J.R., Drews G.N., and 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,

61-76.

Tattersall A.D., Turner L., Knox M.R., Ambrose M.J., Ellis T.H.N., Hofer J.M.I..

(2005). The mutant crispa reveals multiple roles for PHANTASTICA in pea compound

leaf development. Plant Cell 17, 1046–1060.

Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Höhna S, Larget B,

Liu L, Suchard MA, and Huelsenbeck JP. (2012). MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol. 61(3),

539-542.

Sarojam R., Sappl P.G., Goldshmidt A., Efroni I., Floyd S.K., Eshed Y., and

Bowmana J.L. (2010). Differentiating Arabidopsis shoots from leaves by combined

YABBY activities. Plant Cell 22, 2113-2130. 84

Schneeberger R., Tsiantis M., Freeling M., Langdale J.A.(1998). The rough sheath2 gene negatively regulates homeobox gene expression during maize leaf development.

Development 125, 2857–2865.

Timmermans M.C.P., Hudson A., Becraft P.W., Nelson T. (1999). ROUGH

SHEATH2: a Myb protein that represses knox homeobox genes in maize lateral organ primordia. Science 284, 151–153.

Waites R. and Hudson A. (1995). Phantastica - a Gene Required for Dorsoventrality of

Leaves in Antirrhinum-Majus. Development 121, 2143-2154.

Bot. 100, 641-649.

Zoulias N., Koenig D., Hamidi A., McCormick S., and Kim M. (2012). A role for

PHANTASTICA in medio-lateral regulation of adaxial domain development in tomato and tobacco leaves. Ann. Bot. 109, 407-418.

CHAPTER 4: DUPLICATED STM-LIKE KNOX I GENES ACT IN FLORAL

MERISTEM ACTIVITY IN ESCHSCHOLZIA CALIFORNICA (PAPAVERACEAE)

This chapter has been published by the journal Genes, Development and Evolution. The contribution of this author was the selection and alignment of angiosperm KNOX I DNA sequences, their alignment, and the assembly of a phylogenetic tree of those sequences by

Bayesian inference.

Authors

Angelika Stammler • Sandra S. Meyer • Alastair R. Plant • Brad T. Townsley • Annette

Becker • Stefan Gleissberg

Abstract

In angiosperms, the shoot apical meristem is at the origin of leaves and stems and is eventually transformed into the floral meristem. Class I knotted-like homeobox (KNOX

I) genes are known as crucial regulators of shoot meristem formation and maintenance.

KNOX I genes maintain the undifferentiated state of the apical meristem and are locally downregulated upon leaf initiation. In Arabidopsis, KNOX I genes, especially

SHOOTMERISTEMLESS (STM), have been shown to regulate flower development and the formation of carpels. We investigated the role of STM-like genes in the reproductive development of Eschscholzia californica, to learn more about the evolution of KNOX I gene function in basal eudicots. We identified two orthologs of STM in Eschscholzia,

EcSTM1 and EcSTM2, which are predominantly expressed in floral tissues. In contrast, a 87

KNAT1/BP-like and a KNAT2/6-like KNOX I gene are mainly expressed in vegetative

organs. Virus induced gene silencing (VIGS) was used to knockdown gene expression,

revealing that both EcSTM genes are required for the formation of reproductive organs.

Silencing of EcSTM1 resulted in the loss of the gynoecium and a reduced number of

stamens. EcSTM2-VIGS flowers had reduced and defective gynoecia and a stronger

reduction in the number of stamen than observed in EcSTM1-VIGS. Co-silencing of both

genes led to more pronounced phenotypes. In addition, silencing of EcSTM2 alone or

together with EcSTM1 resulted in altered patterns of internodal elongation and sometimes

in other floral defects. Our data suggest that some aspects of STM function present in

Arabidopsis evolved already before the basal eudicots diverged from core eudicots.

Introduction

Apical meristems are at the origin of growth and development in plants. Shoot apical meristems (SAMs) produce leaves from their periphery, while maintaining a central pool of initials, or stem cells, that allow for principally indefinite shoot development. A family of plant-specific homeobox genes, the class I knotted-like (KNOX

I) homeobox genes, is crucial for maintaining the continuous organ-producing activity of

SAMs. Loss-of function causes the failure to replenish the pluripotent pool of cells in the central zone of the SAM, so that further shoot growth and consequently leaf initiation are arrested (Endrizzi et al. 1996; Long et al. 1996). The central role of KNOX I genes in

SAM functioning is demonstrated by the fact that overexpression of KNOX I genes is sufficient to form ectopic SAMs in leaves (Chuck et al. 1996; Brand et al. 2002). 88

Exclusion of KNOX I gene products from leaf founder cells is required for a leaf-specific

determinate developmental program to be established, and the phytohormone auxin as

well as the transcription factors ASYMMETRIC LEAVES 1 (AS1) and FILAMENTOUS

FLOWER (FIL) contribute to the downregulation of KNOX I genes in incipient leaf primordia (Hamant and Pautot 2010). The separation of the leaf from the SAM is further defined by KNOX I interaction with CUP SHAPED COTYLEDON-LIKE (CUC) genes

(Aida et al. 1999; Hay and Tsiantis 2010). KNOX I gene function in the SAM is mediated through regulation of the phytohormone cytokinin which promotes cell proliferation in the SAM, and gibberellin, which promotes a determinate leaf program in the absence of

KNOX I (Jasinski et al. 2005). However, in many species KNOX I gene expression is re- established in the context of dissected leaf development, allowing leaflet initiation from the primordial margins for a brief period of time (Hareven et al. 1996; Bharathan et al.

2002; Champagne and Sinha 2004; Shani et al. 2009).

Duplication of angiosperm KNOX I genes has led to a differentiation of their general function in SAM maintenance and organ differentiation (Hay and Tsiantis 2010).

In Arabidopsis, the SHOOTMERISTEMLESS (STM) gene assumes its role as early as the establishment of the embryonic SAM, and remains the major KNOX I contributor to

SAM function throughout vegetative, inflorescence, and flower development (Hamant and Pautot 2010). BREVIPEDICELLUS (BP), also known as KNAT1, and two recent

Arabidopsis duplicates, KNAT2 and KNAT6, have more restricted expression domains, and loss-of-function mutations do not lead to SAM arrest, indicating that their role in meristem function is less crucial. Mutations in the KNAT1/BP gene impair internode and 89

pedicel growth, and thus affect the function of intercalary meristems located outside and

below the SAM and the floral meristem (Venglat et al. 2002).

KNOX I proteins form complexes with BELL-like homeobox proteins, and

various dimer combinations between KNOX I and BELL are implicated in most KNOX I functions. In addition to SAM maintenance and the control of dissected leaf development,

KNOX I and BELL genes control phyllotaxy and internode spacing in Arabidopsis

(Douglas et al. 2002; Venglat et al. 2002; Smith and Hake 2003; Kanrar et al. 2006;

Hamant and Pautot 2010). The function of KNAT1/BP in the pedicel and in the inflorescence is mediated through interaction with other KNOX I and with BELL-like homeodomain proteins and involves downregulation of KNAT6 and KNAT2 (Ragni et al.

2008).

BELL and KNOX I genes contribute to flower meristem specification by regulating the floral meristem identity genes LEAFY (LFY) and APETALA 1 (AP1)

(Smith et al. 2004; Kanrar et al. 2008). Double mutants in the BELL genes PENNYWISE

(PNY) and POUNDFOOLISH (PNF) do not produce flowers, and LFY and AP1 levels are reduced. In addition, STM-PNY/PNF act together with FLOWERING LOCUS T (FT) to specify flower identity (Smith et al. 2011). Mutant combinations between stm-10 and ft/fd produced more cauline leaves and coflorescences and only few flowers, or produced a terminal cluster of leaves after which the meristem was aborted, and both LFY and AP1 fail to initiate expression (Smith et al. 2004; Kanrar et al. 2008).

A specific role of STM and other KNOX I genes in carpel formation has been demonstrated in Arabidopsis (Endrizzi et al. 1996; Scofield et al. 2007; Alonso- 90

Cantabrana et al. 2007). All KNOX I genes participate in meristematic activity of the ovule-bearing replum/placenta region of the gynoecium (Alonso-Cantabrana et al. 2007).

STM plays an essential role in carpel initiation, which fails if STM is silenced (Scofield et al. 2007). STM overexpression can activate de novo carpel formation and can induce the homeotic conversion of ovules to carpels (Scofield et al. 2007). No comparable effects were observed in KNAT2 or KNAT1/BP mutants (Pautot et al. 2001, Chuck et al. 1996;

Lincoln et al. 1994; Belles-Boix et al. 2006).

Only very limited information is available about floral KNOX I loss-of-function phenotypes in other species. In Zea mays, knotted1 loss-of-function results in a delayed or a complete loss of gynoecium development, but may also cause an increase in carpel number (Kerstetter et al. 1997). Hence more ancestral floral roles for KNOX I genes outside of core eudicots remain inconclusive. We were interested in assessing the role of

KNOX I genes in the basal eudicots, a lineage positioned between Arabidopsis and other core eudicots and monocot model systems in the grass family. Here we present the cloning, phylogenetic, and expression analysis of four KNOX I genes in the California poppy, Eschscholzia californica. We report phenotypic effects following virus-mediated silencing of two STM orthologs in this poppy species. We demonstrate that STM genes control gynoecium and androecium development in Eschscholzia, and have some additional roles in the development of reproductive shoots. Our results suggest that the specific role of STM genes in floral meristem activity during production of the fertile whorls arose prior to the evolution of core eudicots.

Materials and Methods

Isolation of KNOX I genes

Genomic DNA from Eschscholzia californica was isolated with the DNeasy Plant

Mini Kit (QIAGEN GmbH, Hilden). Total RNA was isolated from Eschscholzia

californica shoot tips during late rosette stage by using the RNeasy Plant Kit (Qiagen,

Hilden, Germany). cDNA syntheses were carried out following the instructions for the

SuperScript III reverse transcriptase (Invitrogen, Karlsruhe, Germany) using an oligo

(dT) anchor primer (suppl. Table 2). To obtain the full open reading frame (ORF) of

EcSTM1, sequence information of a published partial clone (Groot et al., 2005) and

promoter sequences were used to design gene-specific primers. PCR with genomic DNA

using forward primer EcKn17F in the 5'UTR and reverse primers EcKn18R and

EcKn19R in the coding region completed the 5' end of EcSTM1. The 5' part of EcSTM2

was obtained with a similar approach using the primers EcKn30F and EcKn31R, and the

EcSTM2 3' end was obtained using 3'RACE (suppl. Table 2). Based on partial coding

regions, the 3' end of EcKNAT1 was cloned using 3'RACE with the primers EcKN21F

and EcKN27F; and the 3'end of EcKNAT2 was cloned using TAIL-PCR with primers

EcKN50F and the degenerate primers AD2-2 and AD5. TAIL PCR was also used to

clone the 5'ends of EcKNAT1 (using AD3 with EcKn38R and EcKn39R) and EcKNAT2

(using AD1-2 with EcKn42R and AD2-1 with EcKn43R). Subsequently, complete ORFs of the four genes were isolated with restriction site-tailed primers located in the 5'UTRs and 3'UTRs of the four genes (suppl. Table 2) and cloned into pGEM-T (Promega,

Madison WI, USA) or pART7 (Gleave, 1992). Restriction enzymes from Promega, 92

Madison WI, USA and Fisher Scientific, Pittsburgh PA, USA were used in the cloning

process. Each plasmid was sequenced in both directions.

Multiple sequence alignments and phylogenetic analysis

For the phylogenetic analysis of Eschscholzia KNOX I genes, nucleotide

sequences of selected asterid, rosid, basal eudicot, and monocot species were retrieved

from GenBank (http://www.ncbi.nlm.nih.gov) and The Gene Index Project

(http://compbio.dfci.harvard.edu/). Additional Papaveraceae sequences were retrieved

from PhytoMetaSyn (http://www.phytometasyn.ca/; suppl. Table 2). Nucleotide

sequences were aligned with MAFFT 6 (Katoh and Toh, 2008) using the G-INS-i strategy with manual editing in MacClade 4.08 (suppl. Table 3). Poorly conserved areas

5' to the MEINOX domain, and 5' to the ELK/Homeodomain encoding DNA sequence, as well as the 3' end of the ORFs were removed, and incomplete sequences that did not cover the conserved areas were excluded. A concatenated data set of 582 nucleotides was used for Bayesian analyses with MrBayes v3.2. GTR was used as the least specific evolutionary model implemented in MrBayes since models suggested by JModelTest2

(Darriba et al. 2012) were not available in MrBayes. The algorithm was run for 1,000,000 generations with sampling every 100 generations, giving a final standard deviation of less than 0.01. The burnin for parameter values and for trees was set to 2000. Two nucleotide datasets were run, one that included class II KNOX genes as outgroup (62 genes from 23 genera) and one with class I KNOX genes only (53 genes from 22 genera). A smaller dataset with translated amino acid sequences that focused on Papaveraceae sequences 93

was also run. For this analysis with 31 sequences from ten species, the JTT + I + G

evolutionary model (Jones et al. 1992) was selected using ProtTest 2.4 (Abascal et al.

2005) according to AIC, AICc and BIC. For all analyses, consensus trees with the posterior probabilities of bifurcations were viewed and analysed in Archaeopteryx v1.

RT-PCR

For expression profiles, total RNA was isolated from various tissues of several stages of plant development using the RNeasy Plant Kit (Qiagen GmbH, Hilden,

Germany) or Plant-rna-OLS (OLS OMNI Life Science GmbH & Co. KG, Hamburg,

Germany). cDNA was generated by reverse transcription of 200 ng RNA by SuperScript

III Kit (Invitrogen, Karlsruhe, Germany) in combination with the Oligo-T anchor primer

AB05. To test for VIGS-induced downregulation of KNOX I genes, total RNA was extracted from individual terminal flower buds (diameter less than 2 mm) of nine plants per experiment, three weeks after inoculation with the VIGS construct, using the RNeasy

Plant Kit Micro (Qiagen GmbH, Hilden, Germany), and reverse transcribed.

Amplification of a 191-bp fragment of EcActin2 was used to normalize the cDNA pools for RT-PCR. RT-PCR for KNOX I genes was conducted with gene specific primers

(EcSTM1-RT-fw, EcSTM1-RT-rev EcSTM2-RT-fw, EcSTM2-RT-rev). Primers were designed to yield 799-bp and 244-bp fragments for EcSTM1 and EcSTM2, respectively, as well as 829-bp and a 848-bp fragments for EcKNAT1 and EcKNAT2/6. PCR was performed in 25µl volumes containing 1x GreenGoTaq Flexi buffer, 0,2 mM dNTPs,

10µl of each primer, 2 mM MgCl2, 0,5 U GoTaq DNA polymerase and 10µl 1:5 to 1:200 94

diluted cDNA with the following cycling program: 96 °C for 2 min, 94 °C for 30 s (step

2), 54° C for 60 s (step 3), 72 °C for 60 s (step 4) and a final extension of 10 min at 72

°C. Steps two through four were repeated 29 to 37 times. EcSTM1 was amplified for 35

cycles, EcSTM2, EcKNAT1, and EcKNAT2/6 for 37 cycles, and EcActin2 for 30 cycles.

Virus induced gene silencing (VIGS)

VIGS was performed using the tobacco rattle virus system (Liu et al. 2002). To

silence EcSTM1, a 387-bp fragment of EcSTM1 between the KNOX and ELK domains

was directionally cloned into pTRV2 using XbaI and SacI restriction sites (pTRV2-

EcSTM1b). A construct containing the full-length EcSTM1 coding region (1111 bp) was

also cloned using BamHI and SacI restriction sites (pTRV2-EcSTM1c). To silence

EcSTM2, a 604 bp fragment containing the ELK/homeodomain and the 3'-end of

EcSTM2 was cloned using EcoRI restriction sites yielding pTRV2-EcSTM2f. To target

both EcSTM1 and EcSTM2, the gene fragments used in pTRV2-EcSTM1b and pTRV2-

EcSTM2f were combined into plasmid pTRV2-EcSTM1b+2f. Tailed primers used for

directional cloning are listed in suppl. Table 2. Each pTRV2-STM construct as well as pTRV1 and empty control vector pTRV2-E were transformed into Agrobacterium tumefaciens strain GV3101. Inoculation of Eschscholzia plants with a mixture of A. tumefaciens strains pTRV1 and one specific pTRV2 plasmid was conducted as described previously (Wege et al. 2007) with the modification that the infection solution was applied to the hypocotyls of 2-3 weeks old plants using a 2 ml syringe combined to a 0.45 x 25 mm needle. Growing conditions were the same as described previously (Wege et al. 95

2007), and flowers were manually cross-pollinated. In each experiment, a small number

of plants inoculated with pTRV2-EcPDS were included to confirm successful silencing

through the associated photobleaching phenotype (Wege et al. 2007). Scoring was based

on the about 70-80% of plants that survived inoculation, and focused on the terminal

flower and the two flowers opening after the terminal flower. The number of flowers to

be scored was further reduced by flower bud abortion (up to 20% of buds) and

individuals that did not commence flowering within a nine week period after inoculation

(up to 20% of plants). There was no apparent difference between the experimental and

control batches regarding lethality, bud abortion, or non-flowering plants.

Results

Orthology of Eschscholzia class I KNOX I genes

All four class I KNOX genes isolated from shoot cDNA contained the MEINOX

domain with the containing KNOX A, KNOX B domains, as well as an ELK and

homeodomain typical of KNOX genes (Figure 16). BLAST searches indicated that

Eschscholzia has two STM-like paralogs that are co-orthologous with Arabidopsis STM.

The open reading frame (ORF) of EcSTM1 (GenBank accession nr. HQ337629) is 1089 bp long and codes for a 362 aa protein, while the ORF of EcSTM2 (GenBank accession

nr. HQ337630) contains 1158 bp coding for a 385 aa protein. A partial clone reported by

Groot et al. (2005) corresponds to EcSTM1. The ORF of a KNAT1/BP-like gene,

EcKNAT1 (GenBank accession nr. HQ337627), contains 1218 bp encoding 405 aa and

matches with GenBank entry DQ133604. The ORF of EcKNAT2 (GenBank accession nr. 96

HQ337628) consists of 1068 bp coding for 355 aa, and matches a partial sequence

deposited in GenBank DQ012434.

Figure 16. Domain structure of the hypothetical proteins encoded by the four class I KNOX genes in Eschscholzia californica. Domains are shown in darker shading and defined according to Kimura et al., (2008), the amino acid positions of the domains are indicated above.

Bayesian phylogenetic analyses of angiosperm KNOX I genes revealed four distinct gene clades (Figure 17). When rooted with class II KNOX genes, three clades received highest branch support. STM-like genes were sister to KNAT1/BP-like genes, which together were sister to KNAT2/6-like genes. The forth group of class I genes,

OSH6-like, formed a basal grade within class I genes, with monocot members more basal than eudicot genes (suppl. Figure 26). Analyses without class II KNOX genes recovered all four groups, including OSH6-like, as monophyletic clades with high support (Figure

2). Further, all four clades contained species from monocots, basal eudicots, rosids, and asterids. 20 sequences from seven Papaveraceae species were found in all four clades, 97

however, an OSH6-like gene is missing from Eschscholzia. A Bayesian analysis using

amino acid sequences of all available 20 Papaveraceae sequences and from Solanum

lycopersicon, Arabidopsis thaliana, and Aquilegia formosa × pubescens confirmed the gene clade affiliations of Eschscholzia and other Papaveraceae genes (suppl. Figure 27).

All analyses revealed the existence of two distinct STM clades that contained only

Papavearaceae genes. The STM2 clade contained EcSTM2 and sequences from chelidonoid poppies (Sanguinaria canadensis, Stylophorum diphyllum) and from papaveroid poppies (Argemone mexicana, Papaver bracteatum). The STM1 clade contained EcSTM1 and likewise sequences from chelidonioids (Sanguinaria canadensis,

Glaucium flavum) and a papaveroid species (Argemone mexicana).

Figure 17. Phylogram of angiosperm KNOX I genes. Consensus tree of Bayesian analysis of 53 genes from 22 genera, rooted with OSH6+KNAT2/6. The four gene clades are indicated. Genes are identified by accession numbers or gene names (see suppl. Table 3). Monocot genes are indicated in blue, poppies and Aquilegia in orange, rosids in green, and asterids in purple. Eschscholzia genes isolated in this study are marked in bold. Branch support (posterior probabilities) of branch nodes over 80% are indicated. Poppy- specific duplicated STM clades are marked with orange circles.

Class 1 KNOX genes are differentially expressed during Eschscholzia shoot development

To analyze the expression patterns of Eschscholzia KNOX I genes, RNA

accumulation was visualized using RT-PCR (Figure 18). The two STM paralogs EcSTM1

and EcSTM2 were weakly expressed in vegetative shoot tips and developing leaves, and

were sharply upregulated upon floral transition and in early flower buds. Expression was

lower in older flower buds. EcSTM1 showed a second peak in 4-5 mm flower buds that

may coincide with ovule development (Becker et al. 2005). In all analyzed tissues, the 99 expression observed for EcSTM1 was stronger than that of EcSTM2. In contrast to the two STM-like genes, EcKNAT1 and EcKNAT2 exhibited strong expression during vegetative development, in shoot tips and developing leaves. Both genes were also expressed in developing flower buds, but at lower levels. Expression of EcKNAT2 increased in older flower buds, and was also detected in cotyledons. Taken together, all four Eschscholzia class 1 KNOX genes were expressed in both vegetative and reproductive stages, but STM-like genes were expressed at higher levels in floral tissues, whereas EcKNAT1 and EcKNAT2 were stronger in vegetative apices including developing leaves.

Figure 18. Expression profiles of Eschscholzia californica class I KNOX genes using semi-quantitative RT-PCR. The four genes are expressed in both vegetative and floral tissues, STM-like genes are expressed at higher levels in early floral meristems. SAM early: less than 5 leaves per rosette, SAM late: more than 8 leaves per rosette, byl blade of young leaves, fb floral bud, FM floral meristem, cot cotyledon, LM length marker, nc negative control, SAM shoot apical meristem, yl young leaf. EcActin2 was used for normalization. PCR cycle numbers and gene fragment lengths are indicated.

100

Downregulation of EcSTMs affects gynoecium formation

VIGS-mediated silencing of both Eschscholzia EcSTM genes resulted in various

flower and shoot phenotypes. Semi-quantitative RT-PCR confirmed effective silencing in

a proportion of the treated plants, consistent with the proportion of plants showing

phenotypes (suppl. Figure 28). Silencing of EcSTM1 and EcSTM2 was not fully specific

as intended by the respective single gene construct, suggesting some extent of cross-

silencing between the two duplicated genes. EcSTM-VIGS flowers were severely

impaired in gynoecium initiation or development, whereas empty vector-treated plants

were normal (Figure 19). Longitudinal sections of untreated and pTRV2-E flowers

exhibited the wild-type morphology with a central, bicarpellate gynoecium inserted at the

bottom of the receptacle cup (Figure 19A,C). Four petals and numerous stamens are

inserted at the top of the floral cup that extends outwards into a floral rim, and the hood-

like calyx dehisces at anthesis. Viewed from above, the four filamentous yellow stigma

rays extend from the floral cup, surrounded by many broader stamens (Figure 19B,D). In

EcSTM1-silenced flowers, perianth and floral cup formed as in untreated plants, but the

floral cup did not contain a gynoecium (Figure 19E,F). Scanning electron microscopy of developing EcSTM1-VIGS flowers revealed that carpel initiation at stage 5 of

Eschscholzia flower development (Becker et al. 2005; Figure 20A) did not occur (Figure

20B). The central region of the floral meristem which in wild-type plants is consumed in

the process of gynoecium initiation remains flat in EcSTM1-VIGS treated plants (Figure

20C). Observations of stained longitudinal sections through equivalent stages confirmed

the lack of initiation of a gynoecium in EcSTM1-VIGS flowers (Figure 20E) in about 101

71% (n=14), as compared to untreated floral buds that had carpels initiated (Figure 5D).

Scored at anthesis, about 52% of all developed EcSTM1-VIGS flowers lacked a

gynoecium (Figure 21A). Two pTRV2 constructs containing a 387 bp fragment (pTRV2-

EcSTM1b) respectively the entire coding sequence (pTRV2-EcSTM1c) resulted in similar

frequencies. (Figure 21A). Post-anthesis, an empty floral cup and rim persist after petals

and stamens have abscised (inset in Figure 19E).

Figure 19. Floral phenotypes of EcSTM1, EcSTM2, and EcSTM1+2 silenced Eschscholzia californica plants. Longitudinally bisected and central views of flowers at anthesis are shown, the calyx has fallen off. Untreated (A,B) and pTRV2-E empty vector controls (C,D) flowers show wild-type morphology. The ovary arises from the bottom of the floral cup, and four stigmatic rays extend beyond the floral cup (A,C) and are seen from above among numerous stamens (B,D). E,F pTRV2-EcSTM1b flowers had an empty floral cup lacking a gynoecium. Inset in (E) shows the empty cup of a post- anthesis flower. (G,H) pTRV2-EcSTM2f flowers had a rudimentary gynoecium that sometimes extended beyond the cup post-anthesis (inset in G). A petaloid stamen is seen in (H). Combined silencing in pTRV2-EcSTM1b/EcSTM2f plants had flowers with no or with a rudimentary gynoecium at anthesis, and fewer stamens (I,J). Arrows in (E), (G) and (I) point to the bottom of the floral cup. Asterisks in (B) indicate stigmatic rays. Asterisk in (H) indicates petaloid stamen.

102

Figure 20. Scanning electron micrographs and longitudinal sections of EcSTM-VIGS floral buds. A,D untreated Eschscholzia californica. B,C,E pTRV2-EcSTM1b treated. The center of the floral meristem lacks initiation of carpels and remains flat (red arrow in (E)). (C) is a magnified view of the floral meristem center shown in (B). F pTRV2- EcSTM2f treated. The center of the floral bud comprises a rudimental structure (red arrow). g gynoecium, p petal, s stamen, se sepals, le leaf. In (A) and (B), stamen primordia of consecutive androecial whorls are marked with numbers. Scale bars are 100µm.

103

Figure 21. Reduced gynoecium and flowers in EcSTM-silenced Eschscholzia plants. (A) About 52% of EcSTM1 silenced flowers (light grey columns) had no gynoecium, while about 24% of EcSTM2-silenced flowers had an aborted or reduced gynoecium (dark grey columns). 68% of flowers co-silenced with EcSTM1 and EcSTM2 (black column) lacked or had a reduced gynoecium. (B) Reduced flowers consisting only of a calyx were found in 5.6% of flowers silenced in EcSTM2. This phenotype was rare in EcSTM1-silenced plants. Reduced flowers in EcSTM1/2-co-silenced plants occurred at 3.3%. (B) Reduced flowers consisting only of a calyx were found in 5.6% of flowers silenced in EcSTM2. This phenotype was rare in EcSTM1-silenced plants. Reduced flowers in EcSTM1/2-co-silenced plants occurred at 3.3%. 104

Figure 22. Stamen numbers in EcSTM-silenced Eschscholzia flowers. Compared to untreated plants and plants inoculated with the empty vector pTRV2-E (white columns), silencing of all STM constructs resulted in reduced stamen numbers. Silencing of EcSTM2 with the pTRV2-EcSTM2f construct (dark grey column) had a stronger effect than two EcSTM1-containing constructs (light grey columns). In plants inoculated with pTRV2-EcSTM1b, flowers without gynoecium (- gyn) had fewer stamens than those with gynoecium (+ gyn). Standard deviation and sample numbers are shown.

Silencing of EcSTM2 had effects that were overlapping but distinguishable from

EcSTM1. At anthesis, many EcSTM2-VIGS treated flowers appeared to lack a gynoecium like plants treated with EcSTM1-VIGS. However, closer examination often revealed a rudimentary gynoecium at the bottom of the floral cup (Figure 19G) at a frequency of about 24% (Figure 21A). At stage 12 of flower development when petals and stamens abscise (Becker et al. 2005), about 39% of EcSTM2-VIGS flowers (n=90) had strongly reduced capsule sizes, shorter than 7mm in length. Stage 5 flower primordia showed a 105

rudimentary gynoecium structure in about 59% of longitudinally sectioned specimens

(Figure 20F; n=17). Taken together, gynoecium development in EcSTM2-VIGS flowers was arrested at an early stage, or showed a delayed development that resulted in shorter and mostly infertile capsules (inset in Figure 19G).

Combined silencing of EcSTM1 and EcSTM2 resulted in flowers with aborted or arrested gynoecia at higher frequencies (68%), likely reflecting a combinatorial effect

(Figure 20A). Silencing of EcKNAT1 had no effect on gynoecium development, and flowers appeared normal (data not shown).

EcSTM silencing results in reduced stamen numbers

The androecium in Eschscholzia californica consists of variable numbers of hexamerous whorls, except the first whorl that has only four stamens (Karrer 1991;

Becker et al. 2005). Untreated and control plants inoculated with pTRV2-E had an average of 27 and 28 stamens, respectively (Figure 21), corresponding to five stamen whorls, but ranged between four to six whorls (suppl. Figure 29). Silencing of both

EcSTM genes individually resulted in reduced stamen numbers, but silencing with the pTRV2-EcSTM2f construct had a stronger effect than with either of the two EcSTM1

specific constructs (Figure 22, suppl. Figure 29). pTRV2-EcSTM2f-mediated silencing

resulted in an average of only twelve stamens, and a range of zero to four whorls.

Silencing with pTRV2-EcSTM1b resulted in an average of about 21 stamens, or four

whorls, and a range of two to six whorls. When EcSTM1b-silenced flowers with and

without gynoecium were analysed separately, the latter had fewer stamens (18), which 106

may reflect a stronger silencing effect when compared to flowers with a gynoecium

(average of 24 stamens). Co-silencing of both EcSTM genes using pTRV2-

EcSTM1b/EcSTM2f were similar in effect to EcSTM2-VIGS flowers (Figure 22, suppl.

Figure 29).

Some EcSTM-VIGS flowers show reduction of perianth organs and development of floral shoots

About 5.6% of EcSTM2-VIGS flowers and 3.3% of double-silenced flowers lacked petals and stamens in addition to an arrested gynoecium (Figure 21B, 23D). Often, the calyx of these flowers assumed largely leaf identity, forming dissected organs above the floral rim (Figure 23E-H). Reduction in calyx identity was accompanied by a reduction of the floral rim (Figure 23H) and the pedicel. Some of these flowers formed additional leafy organs inside the leafy calyx (Figure 23F-G), and the floral meristem occasionally could be seen differentiated into a tiny pin (Figure 23G). This phenotype was rarely observed in EcSTM1-VIGS flowers (Figure 21B).

Premature termination of the floral meristem was sometimes followed by the

formation of a replacement shoot that emerged from within the floral rim and calyx

(Figure 23I). These replacement shoots appeared as the continuation of the main axis and

developed another pseudowhorl and terminal flower (Figure 23J-K). If sepal or rim identity was not obvious, as in Figure 23L, replacement shoots following early floral meristem termination were difficult to detect.

107

Figure 23. Spectrum of floral organ initiation defects in EcSTM-VIGS flowers. A, untreated wild-type flower. B, flower lacks gynoecium. C, flower lacks all fertile organs. D, flower bud consisting of a narrowed calyx only. E-H, flowers in which sepals have largely assumed leaf identity and form as separate organs with various degrees of dissection. A reduced floral rim is present in E-G, but is missing in H. In F and G, additional leafy organs have formed inside the leafy calyx. In G, a pin-like structure has formed (arrowhead). I, a leafy shoot has formed, two sepal-like organs are seen to the left. J, shoot formation from a sepal-bearing flower, sepal is to the right. K, same shoot after ten days, the shoot has formed two basal leaves, a pseudowhorl of two leaves with one axillary shoot, and a terminal flower. L, magnification of K, showing two floral rims carrying a sepal-like organ each, and the base of the floral shoot. Arrows point to the position of the floral rim at the end of the pedicel.

EcSTM silencing can cause homeotic conversions in floral organs

Silencing of both EcSTM genes, singly and combined, led to changes in floral

organ identity in a few flowers. Homeotic conversions of stamens into petals were mostly

observed in flowers that lacked a gynoecium (Figure 24A). Occasionally, petals extended

to the center of the flower (Figure 24C). Sometimes, mosaic sepal/leaf organs formed 108

(Figure 24B, D-F). Sepal sectors with leaf identity were associated with an interruption of

the floral rim (Figure 24E, F), and the calyx failed to abscise at anthesis (Figure 24A, B).

Figure 24. Homeotic organ transformations in EcSTM-VIGS plants. A,B the same flower from above and below showing petaloid stamens and leafy persistent sepals. C petaloid organs occupy the center of the flower. D-F floral buds with leafy calyxes. Interruption of the floral rim is indicated with arrows.

Elongation of stem internodes and the degree of leaf dissection are affected in EcSTM-

VIGS plants

Examination of internodes between stem leaves revealed that EcSTM-silenced

plants frequently formed multiple clusters of leaves separated by longer internodes along

the stem. In contrast, wild type plants have single leaves separated by internodes, and a

single cluster of two or three leaves, called the pseudowhorl, preceding flowers (Becker

et al. 2005; Figure 25). In addition, some EcSTM-VIGS stems exhibited irregular phyllotaxy and fusion of adjacent leaves (data not shown). Finally, overall leaf dissection at higher nodes appeared to be reduced in pTRV2-EcSTM1b stem leaves, albeit 109

variability of this trait was high (suppl. Figure 30). A milder reduction in leaf segment

number was also observed in EcKNAT1-VIGS plants.

Figure 25. Additional pseudowhorls in Eschscholzia shoots. More than one pseudowhorl was rare in control plants (pTRV2-E), but occurred between 19% in pTRV2-EcSTM2f plants and in 45% percent in pTRV2-EcSTM1b/EcSTM2f plants. Shoots with poorly defined pseudowhorls, or where pseudowhorls were not preceded by a single stem leaf, were excluded.

In summary, silencing of the two EcSTM genes had overlapping but distinct effects on the formation of fertile floral organs. In addition to the gynoecium and androecium, silencing of EcSTM genes affected stem internode patterning, and at lower 110 frequencies caused defects in perianth formation, floral specification, and floral organ identity.

Discussion

KNOX I gene evolution

KNOX I genes form a small family of transcription factors that have diversified into four subclades in flowering plants. Most phylogenetic analyses have grouped STM- like genes as sister to KNAT1/BP-like (Golz et al. 2002; Guillet-Claude et al. 2004; Groot et al. 2005; Harrison et al. 2005; Sano et al. 2005; Zluvova et al. 2006; Floyd and

Bowman 2007; Hirayama et al. 2007; Jouannic et al. 2007; Di Giacomo et al. 2008;

Tanaka et al. 2008; Alakonya et al. 2012; Box et al. 2012), while few studies suggest

KNAT1/BP- like are closer to KNAT2/6-like genes than to STM-like (Sakakibara et al.

2008; Magnani et al. 2008; Mukherjee et al. 2009). Our phylogenetic comparisons including the four Eschscholzia KNOX I genes support that STM genes are closest to

KNAT1/BP, which in turn are sister to KNAT2/6 genes). KNAT2/6-like are either sister to

OSH6-like genes (Floyd and Bowman 2007; Jouannic et al. 2007; Hirayama et al. 2007;

Mukherjee et al. 2009; Box et al. 2012), or OSH6-like genes appear as sister to all other

KNOX I genes (Sano et al. 2005; Harrison et al. 2005; Alakonya et al. 2012). KNOX I duplications that produced the four clades precede the divergence of monocots and eudicots, since they contain members of both clades. However, STM-like genes are not found in Poaceae, and no OSH6-like gene occurs in Arabidopsis. Since OSH-like genes 111

are present in two Papaveraceae species and in the Ranunculaceae Aquilegia, an OSH6-

like gene in Eschscholzia either remains undetected, or was lost.

The two Eschscholzia STM-like genes isolated in this study, EcSTM1 and

EcSTM2, have their respective orthologues in other Papaveraceae, but are not found outside the family. Our phylogenetic analyses suggest a STM duplication at the base of subfamily Papaveroideae, since both genes are found in all three clades of this subfamily, in eschscholzioid (Eschscholzia), papaveroid (Papaver, Argemone), and chelidonioid

(Sanguinaria, Glaucium, Stylophorum) poppies. More sequences from subfamily

Fumarioideae, and from other Ranunculales families will be needed to precisely localize the node of this duplication event.

Diversification of expression and function of Eschscholzia KNOX I genes

The two Eschscholzia STM-like genes, EcSTM1 and EcSTM2, like their single

Arabidopsis homologue, are expressed in both vegetative and reproductive meristems, but exhibit a strong upregulation during the floral transition. Expression of EcSTM1 in leaves and floral meristems has previously been demonstrated by RNA in situ hybridization (Groot et al. 2005). In contrast, EcKNAT1, a KNAT1/BP homologue, and

EcKNAT2, a KNAT2/6 homologue, are more strongly expressed in vegetative shoot tips and young leaves, suggesting a partial subfunctionalization among KNOX I genes between vegetative and reproductive stages. In accordance with this, EcSTM silencing phenotypes are primarily floral, while we could not detect flower defects in EcKNAT1-

VIGS plants (not shown). 112

EcSTM genes function in reproductive floral organ development

The spectrum of phenotypes following silencing of EcSTM1 and EcSTM2 suggests that both Eschscholzia STM genes have roles in gynoecium and androecium development. Although single-gene VIGS resulted in some degree of cross-silencing of the other paralogue, limiting an assessment of the degree of redundancy and/or subfunctionalization, a comparison of single and double VIGS plants suggest some extent of subfunctionalization. EcSTM1 silencing has a stronger effect on gynoecium initiation, as gynoecium defects occur less frequently in EcSTM2-VIGS flowers, and lead to arrested or delayed growth rather than complete failure of gynoecium initiation. On the other hand, EcSTM2 silencing affects stamen numbers more frequently and more severely, compared to EcSTM1. Co-silencing of the two paralogues results in stronger phenotypes compared to silencing of each individual gene, indicating additive effects and/or stronger silencing of both genes.

Meristem maintenance and floral determinacy

Reduction of the gynoecium and the androecium in EcSTM-VIGS flowers suggests that EcSTM genes regulate meristem growth specifically during the formation of the fertile floral organs. According to this hypothesis, premature meristem arrest underlies the reduction of gynoecium and androecium in Eschscholzia STM-VIGS plants.

Hence, the general role of KNOX I genes in other species, and STM genes in particular, in maintaining a pool of undifferentiated meristematic cells, is largely restricted to later 113

stages of floral organogenesis following perianth formation. This is in agreement with the

predominantly floral expression of both EcSTM genes. Our RT-PCR data show that both

genes are sharply upregulated during the transition to flowering. It is possible that

EcSTM1 primarily controls meristem maintenance necessary to provide the central

meristem territory from which the gynoecium arises, while EcSTM2 primarily operates in

the androecial ring meristem that gives rise to stamens. However, the flat center of the

floral meristem in gynoecium-less EcSTM1-VIGS flowers and the non-vacuolated,

meristematic character of its cells (Figure 20B,C,E) suggest that central cells do form but

fail to initiate a gynoecium. In situ hybridization data would be needed to correlate these

differences in function with any temporal-spatial differences of expression. Although our

data reveal differential effects of the two EcSTM genes, effects common to plants

silenced with either gene may be due to unintended cross-silencing. Stable single-gene mutants would be necessary to fully differentiate between the functions of the two genes.

The Eschscholzia AGAMOUS (AG) genes EScaAG1 and EScaAG2 confer determinacy in both the central floral meristem territory where carpels form, and in the ring-like androecial meristem surrounding it (Yellina et al. 2010). Virus-mediated silencing of EScaAG genes increases both stamen numbers arising from the ring meristem as well as carpel numbers in the floral meristem center, an effect opposite to

EcSTM silencing that reduces organ numbers in both the gynoecium and the androecium.

In addition, the Eschscholzia CRABS CLAW gene EcCRC contributes to floral meristem determinacy after the gynoecium is initiated but does not act on the ring meristem generating stamen primordia (Orashakova et al. 2009). Hence, a balance of EScaAGs, 114

EcSTMs, and EcCRC expression may regulate floral determinacy during androecium and gynoecium development, determining the correct number of fertile floral organs in

Eschscholzia. Interestingly, both stamen and carpel numbers show prominent variation in other genera of the Papaveraceae family, suggesting these genes as part of a network that might have been imporant in evolutionary changes of fertile floral organ number. In contrast, the FLORICAULA/LEAFY-like gene EcFLO regulates floral determinacy specifically during perianth development, and EcFLO-VIGS flowers produce additional sepal and petal whorls (Wreath et al., 2013).

STM genes as floral specifiers

A number of studies in Arabidopsis suggest that STM also plays a role in flower meristem specification and is required for floral organ formation, particularly of carpels

(Pautot et al. 2001; Endrizzi et al. 1996; Scofield et al. 2007; Yu et al. 2009; Smith et al.

2011). STM cooperates with the BELL-like homeobox genes PNY and PNF to specify flower meristems together with the flowering gene FT (Smith et al. 2011), and STM-BEL complexes may act as additional activators of the flower meristem identity genes LFY and AP1 (Yu et al. 2009; Smith et al. 2011). Occasionally, EcSTM-VIGS flowers showed homeotic conversions between floral organs, indicating a compromised floral organ identity. Indeed, leafy sectors in the calyx of EcSTM-VIGS flowers (Figure 23A,B) are similar to EcFUL-VIGS flowers in which an AP1/FUL-like gene is silenced (Pabon-Mora et al., 2012), and a reiterated floral rim (Figure 23L) is reminiscent of the EcFLO-VIGS phenotype (Wreath et al., 2013). Occasionally, EcSTM-VIGS flowers consisted only of 115

sepals or leafy sepals, suggesting that EcSTM genes contribute, most likely indirectly via activation of floral meristem identity genes, to the specification of sepal and petal

identity, in addition to their major role in stamen and carpel formation.

EcSTM-VIGS flowers suggest that the role of STM in gynoecium initiation and

growth is conserved between basal eudicots and rosid core eudicots (Pautot et al. 2001;

Endrizzi et al. 1996; Scofield et al. 2007). The reduction of stamen number, along with

carpels, suggest that EcSTMs may have a broader role in fertile organs, compared to

Arabidopsis.

The effects on carpels seen in some EcSTM-VIGS flowers may reflect a role of

STM in floral meristem maintenance, in carpel specification and formation, or both. A

role in meristem maintenance would depend on the time point at which the meristem

ceases in EcSTM-VIGS flowers if the resulting flower is lacking only a gynoecium or,

when it ceases earlier, the flower is lacking a gynoecium and some stamen whorls. The

fact that EcSTM2-VIGS flowers do initiate a gynoecium that is subsequently arrested in

its development might suggest an additional role in gynoecium growth. In Eschscholzia,

carpel formation and differentiation also requires AG (Yellina et al. 2010), while flowers

silenced in EcFLO retain a gynoecium in the center of the flower (Wreath et al., 2013).

Pre-floral roles of EcSTMs

In core eudicots KNOX I genes are implicated in promoting leaf complexity. Our

data indicate that leaf dissection might be reduced in Eschscholzia when EcSTM1 is 116

silenced. More studies, particularly including the silencing of multiple genes are needed

to corroborate a role of KNOX I genes in leaf dissection in basal eudicots.

Arabidopsis stm shoots repeatedly initiate replacement shoots in place of an arrested SAM (Endrizzi et al., 1996). Similar floral replacement shoots were observed in reduced sepal-bearing EcSTM-VIGS flowers. Further, STM-deficient Arabidopsis inflorescences exhibit irregular internode spacing, perturbed phyllotaxy, and fusion of adjacent leaves (Endrizzi et al. 1996; Smith et al. 2011). Similar effects were also seen in

EcSTM-VIGS shoots, suggesting conserved roles in inflorescence shoot patterning despite weaker expression of both EcSTM genes in pre-floral tissues. The occurrence of additional pseudowhorls could, at least in part, also reflect the origin of replacement shoots following prefloral meristem abortion. Alternatively, the shoots emerging from

EcSTM-VIGS sepals could be explained by a reversion of the floral into a vegetative meristem.

We have characterized the role of STM-like KNOX I genes in the basal eudicot species Eschscholzia californica, using virus-induced gene silencing. We show that two paralogues exist that are predominantly expressed during the reproductive phase of shoot development where they play crucial roles in the initiation and development of the fertile inner floral organs. Our data suggest that duplication of an ancestral STM gene in the lineage leading to Eschscholzia may have been followed by partial subfunctionalization of their role in flower development, establishing some organ-type specificity. We could also detect that Eschscholzia STM genes contribute to stem internode patterning, perianth formation, flower specification, and floral organ identity. We conclude that the described 117

aspects of STM function originated in basal eudicots, or earlier, and have been maintained

in core eudicots such as Arabidopsis. Equivalent data from basal angiosperms will be

needed to further unravel the evolution of function of these crucial meristem regulators in

flowering plants.

Acknowledgements

S. Gleissberg received funding from the German Research Foundation (DFG) and

a start-up fund from Ohio University. A. Becker received follow-up funding from the

German Research Foundation (DFG). We thank N. Sinha (Davis) for providing

sequences, Andrea Scholz (Mainz) for cloning of pTRV2-EcSTM1-b, Chi Elsie Zhang

(Athens) for database work, Angelika Trambacz and Werner Vogel (Bremen) for plant

care, Friederike Koenig (Bremen) for discussions, Abdinasir Mohamud and Timothy

Pritchard (Athens) for help with phenotypic scoring, and two anonymous reviewers for

comments.

Author Contributions

S. Gleissberg designed the project and led the research with A. Becker. S. Meyer cloned and characterized the genes with B. Townsley. Angelika Stammler contributed to cloning, carried out RT-PCR profiling, and performed the VIGS experiments with phenotypic scoring, data analyses, and data presentation. S. Gleissberg contributed to data analyses and presentation. A. Plant performed the phylogenetic analyses. A.

Stammler, S. Gleissberg, and A. Becker wrote the manuscript. 118

References

Abascal F, Zardoya R, Posada D (2005) ProtTest: Selection of best-fit models of

protein evolution. Bioinformatics 21, 2104-2105.

Aida M, Ishida T, Tasaka M (1999) Shoot apical meristem and cotyledon formation

during Arabidopsis embryogenesis: interaction among the CUP-SHAPED COTYLEDON and SHOOT MERISTEMLESS genes. Development 126:1563-1570.

Alakonya A, Kumar R,Koenig D,Kimura S, Townsley B, Runo S, Garces HM, Kang

J, Yanez A, David-Schwartz R, Machuka J, Sinha N (2012) Interspecific RNA interference of SHOOT MERISTEMLESS-like disrupts Cuscuta pentagona plant parasitism. Plant Cell 24, 3153-3166.

Alonso-Cantabrana H, Ripoll JJ, Ochando I, Vera A, Ferrandiz C, Martinez-

Laborda A (2007) Common regulatory networks in leaf and fruit patterning revealed by mutations in the Arabidopsis ASYMMETRIC LEAVES1 gene. Development 134:2663-

2671.

Becker A, Gleissberg S, Smyth D (2005) Floral and vegetative morphogenesis in

California poppy (Eschscholzia californica Cham.). Int J Plant Sci 166, 537-555.

119

Belles-Boix E, Hamant O, Witiak SM, Morin H, Traas J, Pautot V (2006) KNAT6: an

Arabidopsis homeobox gene involved in meristem activity and organ separation. Plant

Cell 18, 1900-1907.

Bharathan G, Goliber T, Moore C, Kessler S, Pham T, Sinha N (2002) Homologies

in leaf form inferred from KNOX1 gene expression during development. Science 296,

1858-1860.

Box MS, Dodsworth S, Rudall PJ, Bateman RM, Glover BJ (2012) Flower-specific

KNOX phenotype in the orchid Dactylorhiza fuschsii. J Exp Bot 63, 4811-4819.

Brand U, Grünewald M, Hobe M, Simon R (2002) Regulation of CLV3 expression by two homeobox genes in Arabidopsis. Plant Physiol 129, 565-575.

Champagne C, Sinha N (2004) Compound leaves: equal to the sum of their parts?

Development 131, 4401-4412.

Chuck G, Lincoln C, Hake S (1996) KNAT1 induces lobed leaves with ectopic meristems when overexpressed in Arabidopsis. Plant Cell 8, 1277-1289

Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2: more models, new heuristics and parallel computing. Nature Methods 9, 772 120

Di Giacomo E, Sestili F, Iannelli MA, Testone G, Mariotti D, Frugis G (2008)

Characterization of KNOX genes in Medicago truncatula. Plant Mol Biol 67, 135-150

Douglas SJ, Chuck G, Dengler RE, Pelecanda L, Riggs CD (2002) KNAT1 and

ERECTA regulate inflorescence architecture in Arabidopsis. Plant Cell 14, 547-558

Endrizzi K, Moussian B, Haecker A, Levin JZ, Laux T (1996) The SHOOT

MERISTEMLESS gene is required for maintenance of undifferentiated cells in

Arabidopsis shoot and floral meristems and acts at a different regulatory level than the meristem genes WUSCHEL and ZWILLE. Plant J 10, 967-979

Floyd SK, Bowman JL (2007) The ancestral developmental tool kit of land plants. Int J

Plant Sci 168, 1-35.

Golz JF, Keck EJ, Hudson A (2002) Spontaneous mutations in KNOX genes give rise to a novel floral structure in Antirrhinum. Curr Biol 12, 515-522.

Groot EP, Sinha N, Gleissberg S (2005) Expression patterns of STM-like KNOX and

Histone H4 genes in shoot development of the dissected-leaved basal eudicot plants

Chelidonium majus and Eschscholzia californica (Papaveraceae). Plant Mol Biol 58, 317-

331.

121

Guillet-Claude C, Isabel N, Pelgas B, Bousquet J (2004) The evolutionary implications of knox-I gene duplications in conifers: correlated evidence from phylogeny, gene mapping, and analysis of functional divergence. Mol Biol Evol 21, 2232-2245.

Hamant O, Pautot V (2010) Plant development: a TALE story. C R Biol 333, 371-381.

Hareven D, Gutfinger T, Parnis A, Eshed Y, Lifschitz E (1996) The making of a compound leaf: genetic manipulation of leaf architecture in tomato. Cell 84, 735-744.

Harrison J, Möller M, Langdale J, Cronk Q, Hudson A (2005) The role of KNOX genes in the evolution of morphological novelty in Streptocarpus. Plant Cell 17, 430-443.

Hay A, Tsiantis M (2010) KNOX genes: versatile regulators of plant development and diversity. Development 137, 3153-3165.

Hirayama Y, Yamada T, Oya Y, Ito M, Kato M, Imaichi R (2007) Expression patterns of class I KNOX and YABBY genes in Ruscus aculeatus (Asparagaceae) with implications for phylloclade homology. Dev Genes Evol 217, 363-372.

Jasinski S, Piazza P, Craft J, Hay A, Woolley L, Rieu I, Phillips A, Hedden P,

Tsiantis M (2005) KNOX action in Arabidopsis is mediated by coordinate regulation of cytokinin and gibberellin activities. Curr Biol 15, 1560-1565. 122

Jones DT, Taylor WR, Thornton JM (1992) The rapid generation of mutation data matrices from protein sequences. Comp Appl Biosciences 8, 275-282.

Jouannic S, Collin M, Vidal B, Verdeil JL, Tregear JW (2007) A class I KNOX gene from the palm species Elaeis guineensis (Arecaceae) is associated with meristem function and a distinct mode of leaf dissection. New Phytol 174:551-568

Kanrar S, Onguka O, Smith HMS (2006) Arabidopsis inflorescence architecture requires the activities of KNOX-BELL homeodomain heterodimers. Planta 224, 1163-

1173.

Kanrar S, Bhattacharya M, Arthur B, Courtier J, Smith HMS (2008) Regulatory networks that function to specify flower meristems require the function of homeobox genes PENNYWISE and POUND-FOOLISH in Arabidopsis. Plant J 54, 924-937.

Karrer AB (1991) Blütenentwicklung und systematische Stellung der Papaveraceae und

Capparaceae. PhD dissertation, Universitaet Zuerich

Katoh K, Toh H (2008) Recent developments in the MAFFT multiple sequence alignment program. Briefings Bioinformatics 9, 286-298.

123

Kerstetter RA, Laudencia-Chingcuanco D, Smith LG, Hake S (1997) Loss-of- function mutations in the maize homeobox gene, knotted1, are defective in shoot meristem maintenance. Development 124, 3045-3054.

Lincoln C, Long J, Yamaguchi J, Serikawa K, Hake S (1994) A knotted1-like homeobox gene in Arabidopsis is expressed in the vegetative meristem and dramatically alters leaf morphology when overexpressed in transgenic plants. Plant Cell 6, 1859-1876.

Liu Y, Schiff M, Dinesh-Kumar SP (2002) Virus-induced gene silencing in tomato.

Plant J 31, 777-786.

Long JA, Moan EI, Medford JI, Barton MK (1996) A member of the KNOTTED class of homeodomain proteins encoded by the SHOOTMERISTEMLESS gene of

Arabidopsis. Nature 379, 66-69.

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

875-887.

Mukherjee K, Brocchieri L, Bürglin TR (2009) A comprehensive classification and evolutionary analysis of plant homeobox genes. Mol Biol Evol 26, 2775-2794.

124

Orashakova S, Lange M, Lange S, Wege S, Becker A (2009) The CRABS CLAW ortholog from California poppy (Eschscholzia californica, Papaveraceae), EcCRC, is involved in floral meristem termination, gynoecium differentiation and ovule initiation.

Plant J 58, 682-693.

Pabón-Mora N, Ambrose BA, Litt A (2012) Poppy APETALA1/FRUITFULL orthologs control flowering time, branching, perianth identity, and fruit development. Plant Physiol

158, 1685-1704.

Pautot V, Dockx J, Hamant O, Kronenberger J, Grandjean O, Jublot D, Traas J

(2001) KNAT2: evidence for a link between knotted-like genes and carpel development.

Plant Cell 13, 1719-1734

Ragni L, Belles-Boix E, Günl M, Pautot V (2008) Interaction of KNAT6 and KNAT2 with BREVIPEDICELLUS and PENNYWISE in Arabidopsis inflorescences. Plant Cell

20, 888-900.

Sakakibara K, Nishiyama T, Deguchi H, Hasebe M (2008) Class 1 KNOX genes are not involved in shoot development in the moss Physcomitrella patens but do function in sporophyte development. Evol Dev 10, 555-566.

125

Sano R, Juárez CM, Hass B, Sakakibara K, Ito M, Banks JA, Hasebe M (2005)

KNOX homeobox genes potentially have similar function in both diploid unicellular and multicellular meristems, but not in haploid meristems. Evol Dev 7, 69-78.

Scofield S, Dewitte W, Murray JAH (2007) The KNOX gene SHOOT MERISTEMLESS is required for the development of reproductive meristematic tissues in Arabidopsis. Plant

J 50, 767-781.

Shani E, Burko Y, Ben-Yaakov L, Berger Y, Amsellem Z, Goldshmidt A, Sharon E,

Ori N (2009) Stage-specific regulation of Solanum lycopersicum leaf maturation by class

1 KNOTTED1-LIKE HOMEOBOX proteins. Plant Cell 21, 3078-3092.

Smith HMS, Hake S (2003) The interaction of two homeobox genes,

BREVIPEDICELLUS and PENNYWISE, regulates internode patterning in the

Arabidopsis inflorescence. Plant Cell 15, 1717-1727.

Smith HMS, Campbell BC, Hake S (2004) Competence to respond to floral inductive signals requires the homeobox genes PENNYWISE and POUND-FOOLISH. Curr Biol

14, 812-817.

Smith HMS, Ung N, Lal S, Courtier J (2011) Specification of reproductive meristems requires the combined function of SHOOT MERISTEMLESS and floral integrators 126

FLOWERING LOCUS T and FD during Arabidopsis inflorescence development. J Exp

Bot 62, 583-593.

Tanaka M, Kato N, Nakayama H, Nakatani M, Takahata Y (2008) Expression of

class I knotted1-like homeobox genes in the storage roots of sweetpotato (Ipomoea

batatas). J Plant Physiol 165, 1726-1735.

Venglat SP, Dumonceaux T, Rozwadowski K, Parnell L, Babic V, Keller W,

Martienssen R, Selvaraj G, Datla R (2002) The homeobox gene BREVIPEDICELLUS

is a key regulator of inflorescence architecture in Arabidopsis. Proc Natl Acad Sci USA

99, 4730-4735.

Wege S, Scholz A, Gleissberg S, Becker A (2007) Highly efficient virus-induced gene silencing (VIGS) in California poppy (Eschscholzia californica): an evaluation of VIGS as a strategy to obtain functional data from non-model plants. Ann Bot 100, 641-649.

Wreath S, Bartholmes C, Hidalgo O, Scholz A, Gleissberg S (2013): Silencing of

EcFLO, a FLORICAULA/LEAFY gene of the California Poppy (Eschscholzia californica), affects flower specification in a perigynous flower context. Int J Plant Sci

174, 139-153.

127

Yellina AL, Orashakova S, Lange S, Erdmann R, Leebens-Mack J, Becker A (2010)

Floral homeotic C function genes repress specific B function genes in the carpel whorl of the basal eudicot California poppy (Eschscholzia californica). EvoDevo 2010: 1-13.

Yu L, Patibanda V, Smith HMS (2009) A novel role of BELL1-like homeobox genes,

PENNYWISE and POUND-FOOLISH, in floral patterning. Planta 229, 693-707.

Zluvova J, Nicolas M, Berger A, Negrutiu I, Monéger F (2006) Premature arrest of the male flower meristem precedes sexual dimorphism in the dioecious plant Silene latifolia.

Proc Natl Acad Sci USA 103, 18854-18859.

128

Supplementary Data

Figure 26 (supplemental). Phylogram of selected angiosperm KNOX nucleotide sequences. Consensus tree of Bayesian analysis of 62 genes from 23 genera, rooted with class II KNOX genes. Class II KNOX genes and three class I KNOX gene clades are indicated. OSH6-like genes form a basal grade within class I genes and are indicated by a bracket. Genes are identified by accession numbers or gene names (see suppl. Table 3). Monocot genes are indicated in blue, poppies and Aquilegia in orange, rosids in green, and asterids in purple. Eschscholzia genes isolated in this study are marked in bold and underline. Branch support values (posterior probabilities) of bifurcations over 80% are indicated. Poppy-specific duplicated STM clades are marked with orange circles.

129

Figure 27 (supplemental). Phylogram of poppy and few selected other eudicot KNOX I deduced amino acid sequences. Consensus tree of Bayesian analysis of 31 genes from ten species, rooted with OSH6. The four class I KNOX protein clades are indicated. Sequences are identified by accession numbers or gene names (see suppl. Table 3). Monocot genes are indicated in blue, poppies and Aquilegia in orange, rosids in green, and asterids in purple. Eschscholzia genes isolated in this study are marked in bold and underline. Branch support values (posterior probabilities) of bifurcations over 80% are indicated. Poppy specific duplicated STM clades are marked with orange circles.

Figure 28 (supplemental). Semi-quantitative RT-PCR of STM-like genes in floral terminal flower buds with a diameter of less than 2 mm of VIGS-treated Eschscholzia californica. For each of the four VIGS constructs indicated above, three samples are shown along with two negative controls (water, RNA). EcActin2 was used as a positive control. 130

Figure 29 (supplemental). Distribution of stamen numbers in EcSTM-silenced and control flowers. Stamen number of control flowers (pTRV2-E; white columns) ranged between 21 and 33, reflecting four to six stamen whorls. EcSTM1 silencing with pTRV2- EcSTM1b resulted in stamen numbers between seven and 36, or two to six stamen whorls (light gray columns). Stamen numbers in flowers silenced with pTRV2-EcSTM2f ranged between zero and 22, corresponding to zero to four stamen whorls (dark gray columns). Co-silencing of both EcSTM genes using pTRV2-EcSTM1b+EcSTM2f resulted in stamen numbers between zero and 30, or zero to six stamen whorls (black columns). pTRV2-E, n=25; pTRV2-EcSTM1b, n=161; pTRV2-EcSTM2f, n=27; pTRV2-EcSTM1b+EcSTM2f, n=116. 131

Figure 30 (supplemental). Degree of leaf dissection in KNOX I-silenced and control leaves. In control plants (grey bars), the total count of leaf segments increased from leaf node 5 through 17, reflecting leaf heteroblasty. Leaves of EcSTM1-silenced plants were less dissected at nodes 11 through 17 as compared to empty vector-treated plants. EcKNAT1-silenced plants showed a milder reduction in leaflet segment count at the same nodes. Leaves at nodes 5 and 8 showed no difference among treatments. Leaves at these earlier nodes may have surpassed the organogenetic phase of leaf growth before the effect of pTRV2-mediated silencing came into effect. Standard deviation and sample sizes are shown for each data group.

132

Table 2 (supplemental). Primers used for amplification of KNOX genes

Name Sequence 5' to 3' Comments EcKn17F GGA GGG TCC AAG CAG TAA TTT CAT EcSTM1 EcKn18R CTT CTG ACA GTT GAC GTA AGA GGC C EcSTM1 EcKn19R GCA TTT GCT TCC TCT AAT TTA G EcSTM1 EcKn30F CCC GTC CGT CCC CTT AAC ATA EcSTM2 EcKn31R AGG TGC TGC TAC TTG TGG GTT TTG EcSTM2 EcKn21F TAA TGG TCC TAT CCG GGT CTT CAC AGA TG EcKNAT1 EcKn27F CAT CTG AAG AGG ATC AAG AAA ACA GCG CGG GCG EcKNAT1 EcKn50F GGA GAG ATA GAA GTT CAA GAG GTT EcKNAT2 EcKn38R GCC CAG TTA GTT CTT CAC GAT ATT TAA CCA ACA TAT CAT AAT AAG EcKNAT1 EcKn39R CCC A CAA ACT CTT GCC TAG CCT GTG TCA ACC AA EcKNAT1 EcKn42R GGA TGA AGA TTT GTT GTT GCA GAG ATT TGA AAG TTG CAT CTG AAT EcKNAT2 EcKn43R AGCTTC TCA GGA TCA TCA CCA AGA CAA GTA GTA GAA ACA ACA GTT EcKNAT2 Pr1-fw-18Xho1 ATC TCG AGG TCC AAG CAG TAA TTT TCA TG EcSTM1 Pr2-rv +1058Xba1 ATT CTA GAC TTT CAA AGC ATT GGA GTG C EcSTM1 Pr3-fw+3 GAA GTT GGT GGT AGA AGT AGT AGT AGT GA EcSTM2 Pr4-rv+1900 AAG ACG GCT GGG TGT AGT AAT C EcSTM2 Pr5-fw-84EcoRI ATG AAT TCG AGA GAG AGA GTA CTT CTG G EcKNAT1 Pr6-rv+1817XbaI ATT CTA GAA TGA GTC AAG GCC CCA AAC G EcKNAT1 Pr7-fw-6 XhoI ATC TCG AGC AAT CAT CAA TGG AGG ATC TC EcKNAT2 Pr8-rv+1136XbaI ATT CTA GAG GGA CAT TCA GTT TTC GG EcKNAT2 Pr15fw-XbaI TCT AGA AGT TCA TGG AAG CTT ACT GTG AGA TGC pTRV2-EcSTM1b Pr16-rev-SacI GAG CTC TCT TTT GCG ATT CCG AGG pTRV2-EcSTM1b Pr18-fw-BamHI GAA TAG GAT CCG TCC AAG CAG TAA TTT TCA pTRV2-EcSTM1c Pr19rev-SacI GCA TGA GCT CCT TTG AAA GCA TTG GAG TG pTRV2-EcSTM1c Pr22fw CCT GAC ACT CCT CTT ACT AAT TCT C pTRV2-EcSTM2f AB05 GAC TCG AGT CGA CAT CTG TTT TTT TTT TTT TTT TT RACE Oligo(dT) anchor primer AB07 GAC TCG AGT CGA CAT CTG RACE anchor primer AD2-2 AGW GNA GWA NCA WAG G TAIL-PCR AD5 STT GNT AST NCT NTG C TAIL-PCR AD3 WGT GNA GWA NCA NAG A TAIL-PCR AD2-1 NGT CGA SWG ANA WGA A TAIL-PCR AD1-2 NTC GAS TWT SGW GTT TAIL-PCR

133

Table 3 (supplemental). Sequence IDs for KNOX genes used in phylogenetic analyses

Species and Gene Name Sequence ID Taxonomy Antirrhinum majus HIRZINA AY072736 Core eudicots-asterids Antirrhinum majus INVAGINATA AY072735 Core eudicots-asterids Arabidopsis thaliana KNAT1 AY113982 Core eudicots-rosids Arabidopsis thaliana KNAT2 X81353 Core eudicots-rosids Arabidopsis thaliana KNAT3 X92392 Core eudicots-rosids Arabidopsis thaliana KNAT4 X92393 Core eudicots-rosids Arabidopsis thaliana KNAT5 AF306661 Core eudicots-rosids Arabidopsis thaliana KNAT6 AB072362 Core eudicots-rosids Argemone mexicana KNAT1 AMEST1PF_c6638 Papaveroideae-Papavereae Argemone mexicana STM1 AMEST1PF_rep_c1005 Papaveroideae-Papavereae Argemone mexicana STM2 AMEST1PF_rep_c2762 Papaveroideae-Papavereae Corydalis cheilanthifolia KNAT1 CCHRT1PF_c4675 Fumarioideae-Fumarieae Glaucium flavum KNAT1 GFLRT1PF_c4306 Papaveroideae-Chelidonieae Glaucium flavum STM1 GFLRT1PF_rep_c387 Papaveroideae-Chelidonieae Glaucium flavum OSH6 GFLRT1PF_rep_c3987 Papaveroideae-Chelidonieae Papaver bracteatum KNAT1 PBRST1PF_rep_c1541 Papaveroideae-Papavereae Papaver bracteatum STM2 PBRST1PF_rep_c5140 Papaveroideae-Papavereae Pisum sativum PsKn1 AF080104 Core eudicots-rosids Ruscus aculeatus RaSTM AB300055 Monocots Sanguinaria canadensis KNAT2/6 SCARH1PF_c4348 Papaveroideae-Chelidonieae Sanguinaria canadensis STM1 SCARH1PF_rep_c1395 Papaveroideae-Chelidonieae Sanguinaria canadensis KNAT1 SCARH1PF_rep_c2366 Papaveroideae-Chelidonieae Sanguinaria canadensis STM2 SCARH1PF_rep_c2559 Papaveroideae-Chelidonieae Solanum lycopersicon LeT6 AF000141 Core eudicots-asterids Solanum lycopersicon LeTKn1 U32247 Core eudicots-asterids Stylophorum diphyllum KNAT1 SDIST1PF_c10469 Papaveroideae-Chelidonieae Stylophorum diphyllum STM2 SDIST1PF_c15323 Papaveroideae-Chelidonieae Stylophorum diphyllum OSH6 SDIST1PF_rep_c6976 Papaveroideae-Chelidonieae

134

CHAPTER 5: LASER MICRODISSECTION OF ESCHSCHOLZIA CALIFORNICA

LEAF PRIMORDIA FOR COMPARISON OF GENE EXPRESSION BETWEEN

DEVELOPMENTAL STAGES

Abstract

While the development of specific structures and tissues in multicellular organisms requires the expression of specific subsets of genes, organs at different stages of development are likely to differ substantially in the expression of those genes. Proper characterization of a developing tissue in terms of its metabolic, gene expression or proteomic profile necessitates the exclusion of its neighbors when isolating the tissue from the organism to exclude contamination from the other tissues and ensure confidence in the data obtained downstream. Laser microdissection (LMD) enables the isolation of specific regions of tissue from slide-mounted sections with sufficient discriminatory power to isolate structures as small as the shoot apical meristem. We demonstrate the feasibility of isolating leaf primordia at different stages of development, of isolating RNA from that tissue, and of subsequent comparison of expression of a developmental gene,

Eschscholzia californica CINCINNATA, between those extracts by quantitative polymerase chain reaction.

Introduction

Discrimination between tissues and organs allows profiling of messenger RNA transcripts, translated proteins, small interfering RNAs (siRNA) and micro RNAs

(miRNA), and metabolism, genetic sequencing to detect abnormalities, and other tissue- 135

specific factors. Isolation of tissues or structures from their neighbors that are similar or

overlapping with regard to the characteristics of interest is a necessity to ensure the

specificity and validity of the data eventually obtained.

Laser microdissection (LMD) enables the selection of small quantities of tissue or

even single cells in a specific manner. Frozen or paraffin-embedded tissues may be

sectioned with a cryostat or microtome, respectively, and those sections mounted on a

slide-mounted membrane. In the case of paraffin-embedded tissues, the sections are

deparaffinized with an appropriate solvent, retaining the tissue. The slides are viewed

tissue side down using a laser capture microscope, which integrates a computer

numerically controlled (CNC) laser offset from the eyepiece. Power, offset and cutting

path are specified by the user, and the membrane is cut, releasing the membrane and

adherent tissue from within the specified region. The tissue falls directly into a collection

tube. This contrasts with laser capture, where the desired tissues are retained on a

membrane and maintain their relative positions (Nelson et al., 2006). The tube may

contain the first reagent (e.g. lysis buffer) required for extraction of materials of interest,

depending upon the protocol. Multiple membrane fragments may be collected into a

single tube, and the design of the microscope in use may permit harvesting of tissue into

several tubes, selectable via the associated computer software. Once isolated, the range of

techniques applicable to the material includes but is not limited to: sequencing of

nucleotide polymorphisms to identify tissue- or cell-specific mutations; RNA-seq

(Schmid et al., 2012) or RNA extraction for quantitative PCR and microarray analysis

(Wang et al., 2006) to characterize transcription in the tissue; isolation of tissue-specific 136

proteins (Cadron et al., 2009); isolation of tissue-specific components or metabolites

(Korekane et al., 2007).

The advantages of the technique extend beyond target specificity. The use of computational tools during the cutting process allows for the calculation of areas and volumes cut, the measurement of anatomical features, and the capacity to harvest multiple tissue types from a single section. Sections can be photographed and annotated directly so as to accurately record the source of starting materials. There are, however, disadvantages. Accumulation of sufficient quantities of tissue depends upon the amount of tissue available, which may be constrained by the area of the tissue and the maximum

thickness of tissue that can be cut with the laser (although multiple passes are possible),

and choice of tissue thickness is a compromise between yield and transparency for ease

of identification. Preservation method, either cryogenic or embedding, favors the

protection of either RNA integrity or tissue morphology. The number of tissue sections

that must be sampled to obtain useable RNA, for example, may number in the hundreds

for small structures such as leaf primordia, and downstream enrichment or amplification

of desired extracts may be necessary.

LMD is a versatile technique, though, and may be the best or only means to

isolate certain structures and tissues. In development, data describing spatial and

temporal changes in gene expression between stages of organ development or

successively produced organs have become accessible via techniques such as quantitative

PCR, microarray profiling, and RNA-seq, provided that the appropriate tissues are

available. LMD allows the user to study the genetic or metabolic profiles of specific 137

tissues or cells, making it a useful technique in the study of plant lateral organs (Nelson et

al., 2006), wherein manual excision and dissection of those organs in order to isolate

particular parts or stages may be impossible due to the limits of manual dexterity or

inevitable trauma to the tissue, or where separation of one tissue from its neighbor, the

inclusion of which may distort data or contaminate extracts, is essential.

Evolutionary-developmental biology seeks in part to establish how changes in

development produce novel morphologies, thus by comparing developmental patterns

and the gene expression that underlies them between different lineages, the processes by

which different forms have emerged or diverged can be revealed. We sought to test LMD

as a means by which to isolate RNA from the leaf primordia of a compound-leafed plant,

Eschscholzia californica, in order to compare gene expression profiles between

organogenetic and differentiating tissues. We hypothesized that tissue from primordia in

the organogenetic stage would show enhanced expression of class I KNOX genes that

promote the production of leaflets in other species. In contrast, we expected that

maturation-promoting genes, such as the TCP (TEOSINE BRANCHED 1, CINCINNATA,

and PROLIFERATING CELL FACTORs 1 and 2) family transcription factor

CINCINNATA, would be increased in abundance in older tissue that is undergoing

differentiation and maturation. Two temporal stages were isolated by LMD and it was found that several genes could be amplified by quantitative PCR. The choice of fixative, sample size, extract quality and RNA yield is discussed, and LMD is appraised for its value in comparative development studies.

138

Method

Tissue harvesting, embedding, sectioning, and mounting

Eschscholzia californica seeds were sown in trays with 48 x n cm2 wells in a standard potting soil with good drainage and covered with clear lids. After sowing, the seeds were stratified in darkness at 4°C for three days before transfer to constant light at

22°C in growth chambers. Shoot tips were harvested from plants with 1 – 2 fully expanded leaves and immediately fixed in 3:1 ethanol : acetic acid. Kerk et al. (2003) found that this fixative produces higher RNA yields than Prefer (glyoxal) or formaldehyde-acetic acid-ethanol fixatives. Vacuum treatment was used to remove air from the samples to ensure good fixation, after which the tissue was shaken at low speed overnight at 4°C.

Fixed tissue was successively transferred through 70% ethanol, 1 : 1 ethanol :

Histoclear, 100% Histoclear and fresh 100% Histoclear at one hour intervals. A small quantity of paraffin embedding wax pellets (Tissue-Tek) were added to the beaker that contained the samples that was then incubated at 55°C for seven days. Histoclear was drained from the top of the beaker and replaced with paraffin pellets twice per day until all of the Histoclear was replaced by paraffin. The shoot tips were then arranged longitudinally in base molds and embedded with fresh, molten paraffin wax. The paraffin blocks were allowed to solidify overnight at 4°C before trimming and sectioning into

10µm sections with a rotary microtome, section thickness chosen on the basis of RNA yield comparisons by Kerk et al. (2003). Around 15 shoot tips were sectioned. Ribbons of sections were trimmed with a razor blade on an RNase-free surface, floated on DEPC- 139

treated distilled water in a 40°C water bath to allow the sections to expand, and then

mounted temporarily on glass microscopy slides for examination and selection of useful

sections at low magnification under a dissection microscope. Ribbons containing

sections of interest were refloated in the water bath and transferred to the membranes of

PEN membrane slides (Leica), selected to give the largest available membrane area per

slide, thus ten ribbons containing approximately six sections each were arranged on five

PEN membrane slides. The slides were then left on a sterilized, 37°C heat block for approximately 36 hours to ensure attachment of the sections to the membrane before a period of cold storage (4°C) for several days prior microdissection.

Laser microdissection

Laser microdissection was performed using an LM6000 (Leica). Leaf primordia were selected by size from point of attachment to tip as being within the organogenetic phase (<750µm) or the post-organogenetic phase (>750µm) based on the analysis of dissected leaf development in Eschscholzia californica (Gleissberg, 2004). Measurements were made using the inbuilt measurement tool in the LM software suite. Test cuts on shoot tissue were made in order to determine the necessary laser intensity; the default power used was the lowest necessary, to prevent damage to the tissue. Laser offset was minimized to ensure that the marked cutting path was followed as closely as possible.

140

RNA extraction and reverse transcription

Tissue was harvested from approximately 300 sections and pooled into two tubes,

one per developmental stage. Each tube contained 40µl of buffer 1 from an Arcturus

Paradise Plus RNA Extraction and Isolation Kit suitable for paraffin-embedded tissue,

and RNA isolation commenced immediately after all sections were harvested.

Isolated RNA was assessed for concentration using a Nanodrop ND-1000 and for

quality on a Bioanalyzer Nano chip. Two micrograms of RNA per sample was reverse

transcribed into cDNA at 37°C using MMLV-reverse transcriptase (Promega) and

random hexamer primers in a 1:1 ratio to RNA (Promega). The cDNA was diluted 1/10

with nuclease-free water.

Quantitative PCR

cDNA extracted from organogenetic and differentiating-stage leaf primordia was used as a template for quantitative polymerase chain reaction (QPCR) for the genes

ACTIN-2, Eschscholzia californica CINCINNATA (EcCIN), and Eschscholzia californica

TCP2-LIKE, (EcTCP2-LIKE). ACTIN-2 was amplified with the primers ACTIN2-Fwd

(TTACAATGAGCTTCGTGTTGC) and ACTIN2-Rev

(TCCAGCACAATACCTGTAGTA); EcCIN was amplified with EcCIN21F

(TTCAAGACTTGGGGTAGTAAGAGG) and EcCIN22R

(AACAGTAGATGCAGTTGGTCTCC); EcTCP2-LIKE was amplified with EcTCP2-

6F (AAGGAAAAACCCGAAGAACC) and EcTCP2-7R

(TTGAGCTTGAACCGAAAAGC). 141

Each reaction (cDNA sample and appropriate primer pair) was performed in triplicate,

alongside negative controls (nuclease-free water) and RNA (to confirm the absence of

DNA contamination). The quantitative PCR annealing temperature was 55°C (elongation time one minute)

Results

Laser microdissection and RNA quality assessment

Morphogenetic and maturing leaf primordia were harvested from approximately

300 separate sections of shoot apical tissue (Figure 31) taken from 15 plants, tissue from

each stage being pooled. RNA was isolated from the dissected tissue. Quality

assessment using a Bioanalyzer returned the following data: morphogenetic primordia

yielded 3.54µg of total RNA with an RNA Integrity Number (RIN) of 2.2; maturing leaf

primordia yielded 2.44µg of total RNA, RIN = 2.2.

142

Figure 31. Laser microdissection enables profiling of gene expression at different stages of leaf development. Morphogenetic primordia (<750μm, top row) were measured and excised separately from maturing primordia (>750μm, bottom row) and from the shoot apical meristem. Tissues for each stage were pooled for RNA extraction using an Arcturus Paradise kit. Randomly primed reverse transcription has yielded cDNA for quantitative PCR comparison of development-associated candidate gene expression.

Quantitative PCR

The isolated RNA was reverse transcribed into cDNA for comparison of gene expression between the two tissue types by quantitative PCR. Amplification was achieved for ACTIN-2 and EcTCP2-LIKE (Figure 2; Ct ACTIN-2 = 22.51 and 22.27; Ct 143

EcTCP2-LIKE = 23.33 and 22.20 for morphogenetic and maturing primordia respectively), however, amplification of EcCIN failed for both samples. Successful amplification of EcCIN with the same primers in parallel experiments suggests that either low levels of expression or that fragmentation of the RNA for that gene prevented amplification from these cDNAs. For those genes that could be amplified, no significant difference was seen between the samples.

Discussion

Laser microdissection allowed easy dissection of physically close structures attached to the Eschscholzia shoot tip with minimal loss of or damage to tissue of interest

(figure 1), allowing assessment of gene expression at different stages of leaf development. A constitutively expressed gene (ACTIN-2) and a gene with a specific role in leaf development (EcTCP2-LIKE) were successfully amplified and their levels of expression between two developmental stages of leaf primordia were compared by quantitative PCR.

Selection of tissues by size was facilitated by the microscope software, and direct export of images from the software has promise for downstream analysis of, for example, quantification of expression for a given number of cells, by acquiring data such as cell numbers with applications such as ImageJ, or by using built-in features, e.g. recording the area of tissue excised. Such features provide superior accuracy to measurements of the mass of manually excised tissue as they can be used to estimate cell numbers and sizes that may improve between-tissue normalization techniques. 144

The three-dimensional nature of the structures being dissected necessitates some prior knowledge for dissection to be accurate, either from manual dissection of the tissue or a short period of training. The machinery can be operated and tissue selected with relative ease, however, the operating time required is long (approximately six hours to isolate the morphogenetic and maturing primordia from 300 10µm sections) due to the speed of the laser and the time required to select the cutting path. During this time the tissue on the slide remains at room temperature, a concern for RNA stability.

The methods of fixation and embedding proved more than satisfactory morphological preservation, leaving the tissue in better condition than that normally obtained by the authors from paraformaldehyde fixation for in situ hybridization (ethanol

: acetic acid fixation is unsuitable for in situ hybridization (Kerk et al., 2003).

Histological staining was found to be unnecessary, although it is possible to include staining with the method described. The prolonged, harsh conditions of the fixation and embedding processes evidently had a substantial impact upon RNA quality, however, and therefore alternate methods are recommended for the highest yields. Research groups frequently use cryogenic preservation as an alternative to high temperature paraffin infiltration and embedding, despite the necessary compromise of morphological preservation, however, Inada and Wildermuth (2005) detail a method for microwave paraffin embedding that reduces the length of the paraffin embedding process and increases the quality of the RNA obtained while better preserving tissue structure.

Another likely source of damage to RNA is the use of the laser. Inevitably, densely packed tissues are susceptible to heat damage, even when cutting between them. 145

Reduction of laser intensity to a minimum for accurate cutting and reduction of heating

likely improves the quality of the RNA obtained. The intensity necessary is related to the

thickness of the tissue, which in turn corresponds to both the amount of RNA available

per section and to the surface area: volume ratio of the material and thus the possible

exposure of the RNA to external RNases. Thinner sections are, however, less likely to

require multiple cuts. The authors recommend experimentation with a series of tissue

thicknesses.

Although QPCR amplification was successful for several genes, the low quantity

of RNA isolated, i.e. the small number of transcripts, and the low RNA quality (RIN= 2.2 for both tissues) likely impacted the expression data obtained and possibly prevented detection of a gene (EcCIN) that was expected to be detected in at least one sample.

Improvement of RNA preservation, linear amplification of isolated RNA (Rudnicki et al.,

2004), and/or preferential enrichment of rare transcripts at the reverse transcription stage

with duplex-specific nucleases (Yi et al., 2011) would likely be necessary for broad-scale

comparison of gene expression between tissues, e.g. by microarray analysis or RNA-seq.

In the opinion of the author, the specificity of the technique and its ease of employment

justify the development of improved protocols that add these processes.

In conclusion, while there are unavoidable compromises to the process of laser

microdissection, the ease of isolating small, densely packed structures from the shoot

apex is undeniable, and optimization of fixation and embedding techniques in a tissue-

specific manner will make this a formidable technique in plant evo-devo research.

146

Acknowledgements

We would like to thank Vijay Nadella for instruction in using the laser microscope and in experimental design, and Ahmed Faik and Loren Honaas for discussion of the laser microdissection process. Work was funded in part by an OURC grant to SG.

References

Cadron I., Van Gorp T., Amant F., Vergote I., Moerman P., Waelkens E., Daemen

A., Van De Plas R., De Moor B., and Zeillinger R. (2009). The use of laser microdissection and SELDI-TOF MS in ovarian cancer tissue to identify protein profiles.

Anticancer Res. 29, 1039-1046.

Gleissberg S. (2004). Comparative analysis of leaf shape development in Eschscholzia californica and other Papaveraceae-Eschscholzioideae. Am. J. Bot. 91, 306-312.

Inada N. and Wildermuth M.C. (2005). Novel tissue preparation method and cell- specific marker for laser microdissection of Arabidopsis mature leaf. Planta 221, 9-16.

Kerk N.M., Ceserani T., Tausta S.L., Sussex I.M., and Nelson T.M. (2003). Laser

Capture Microdissection of Cells from Plant Tissues. Plant Physiology 132, 27-35.

147

Korekane H., Shida K., Murata K., Ohue M., Sasaki Y., Imaoka S., and Miyamoto

Y. (2007). Evaluation of laser microdissection as a tool in cancer glycomic studies.

Biochem. Biophys. Res. Commun. 352, 579-586.

Nelson T., Tausta S.L., Gandotra N., and Liu T. (2006). LASER

MICRODISSECTION OF PLANT TISSUE: What You See Is What You Get. Annu.

Rev. Plant Biol. 57, 181-201.

Rudnicki M., Eder S., Schratzberger G., Mayer B., Meyer T.W., Tonko M., and

Mayer G. (2004). Reliability of T7-based mRNA linear amplification validated by gene expression analysis of human kidney cells using cDNA microarrays. 97, e86-e95.

Schmid M.W., Schmidt A., Klostermeier U.C., Barann M., Rosenstiel P., and

Grossniklaus U. (2012). Mechanism of leaf-shape determination. Annu Rev Plant Bio

57, 477-496.

Wang L., Zhu J.-., Song M.-., Chen G.-., and Chen J.-. (2006). Comparison of gene expression profiles between primary tumor and metastatic lesions in gastric cancer patients using laser microdissection and cDNA microarray. 12, 6949-6954.

148

Yi H., Cho Y.-., Won S., Lee J.-., Jin Yu H., Kim S., Schroth G.P., Luo S., and Chun

J. (2011). Duplex-specific nuclease efficiently removes rRNA for prokaryotic RNA-seq.

Nucleic Acids Res. 39.

149

CHAPTER 6: DISCUSSION

Changes in the employment of developmental modules, i.e. groups of genes

acting in concert to produce some function of development, or their integration, the

qualitative or quantitative interactions between or within modules, constitute a driving

force in evolutionary-developmental biology. Modification of an organism’s

developmental program may produce new phenotypes that provide an adaptive

advantage, for example, increasing the likelihood of pollination and fertilization or

facilitating occupation of a different ecological niche, and such changes may be favored

by natural selection and become a part of that species’ evolutionary history.

Examination of the expression and knockdown phenotypes of genes such as

CINCINNATA and PHANTASTICA in separate lineages makes clear the importance of

viewing development as a network of interactions that has been repeatedly modified, for

uniform function cannot be assumed in developmental systems where multiple modules

are partially redundant (e.g. TCP4 with TCP3 and TCP10 (Koyama et al., 2007);

PHANTASTICA with the Class III HD-ZIP genes (Prigge et al., 2005)). The examples

described in chapters two, three and four expand our knowledge of the range of roles for

these genes, and allow us to identify common themes and likely ancestral roles for such

genes. To use the ARP genes as an example, ROUGH SHEATH 2 is required for the

repression of Class I KNOX genes in the leaf of the monocot Zea mays but is apparently

unconnected to specification of leaf polarity (Timmermans et al., 1999), while in the core eudicot Antirrhinum majus the loss of AmPHAN leads to the radialization of leaves.

Inference of EcPHAN function from VIGS in Eschscholzia californica suggests that an 150

expanded role for the ARP genes in the specification of the adaxial domain of the leaf

lamina is a derived characteristic.

Likewise, the absence of floral phenotypes for EcCIN and CvCIN implies co- option of CIN-TCP genes into the development of the petal lamina after the Papaveraceae diverged from their common ancestor with the core eudicots (Crawford et al., 2004, Nag et al., 2009), however, a scarcity of research into the CIN-TCP genes in monocots and basal angiosperms leaves open the possibility of a loss of petal development related functions in Eschscholzia and Cysticapnos. This question could be addressed with a more detailed comparison between species. Comparison of homologs in currently used models as well as in new phylogenetic intermediates, perhaps target range from chromatin immunoprecipitation or upstream regulation at cis regulatory elements, may indicate whether presence or absence of the characteristic is the ancestral state.

Looking forward, the increasing capabilities of next generation sequencing and sequence assembly technologies, in combination with current high throughput techniques such as microarrays, are beginning to permit wide-scale comparisons of genomes and

profiling of gene expression at multiple stages of ontogeny, as discussed in Chapter 5.

Rather than taking a candidate gene approach, evo-devo researchers will be at liberty to

assay multiple genes in a regulatory network, or even genome-level expression; inferred

lineage-specific shifts in regulatory patterns might then be supported with functional

studies using transgenic or post-transcriptional gene silencing technologies (de Bruijn et

al., 2012). With such an approach we will rapidly gain a better understanding of how 151 plant development has evolved to produce the variety of anatomical and morphological forms seen in plants today.

References

Crawford B.C.W., Nath U., Carpenter R., and Coen E.S. (2004). CINCINNATA controls both cell differentiation and growth in petal lobes and leaves of antirrhinum.

Plant Physiol. 135, 244-253.

de Bruijn S., Angenent G.C., and Kaufmann K. (2012). Plant ‘evo-devo’ goes genomic: from candidate genes to regulatory networks. Trends Plant Sci. 17, 441-447.

Nag A., King S., and Jack T. (2009). miR319a targeting of TCP4 is critical for petal growth and development in Arabidopsis. Proceedings of the National Academy of

Sciences 106, 22534-22539.

Prigge M.J., Otsuga D., Alonso J.M., Ecker J.R., Drews G.N., and Clark S.E. (2005).

Class III Homeodomain-Leucine Zipper Gene Family Members Have Overlapping, 152

Antagonistic, and Distinct Roles in Arabidopsis Development. The Plant Cell

Online17, 61-76.

Timmermans M.C.P., Hudson A., Becraft P.W., Nelson T. (1999). ROUGH

SHEATH2: a Myb protein that represses knox homeobox genes in maize lateral organ primordia. Science 284:151–53

153

APPENDIX I: CLONING OF PUTATIVE MICRORNA319 HOMOLOGS FROM

ESCHSCHOLZIA CALIFORNICA

Solanum lycopersicum LANCEOLATE and the homolgous Arabidopsis thaliana genes TCP3, TCP4, TCP10, as well as other CIN-TCP genes TCP2 and TCP24 are negatively post-transcriptionally regulated by the microRNA miRNA319 (Ori et al.,

2007), a relative of miR159 (Li et al., 2011). In Arabidopsis there are multiple RNA precursor genes, silencing functions primarily associated with miR319a (Warthmann et al., 2008; Palatnik et al., 2007), but miR319 family microRNAs have been identified even in moss (Arazi et al., 2005).

Previously, Barakat et al. (2007) conducted broad-scale sequencing of microRNAs from Eschscholzia californica, an herbaceous basal eudicot plant. Six miR319-like sequences were identified, five of which represented subtle variations of the same sequence, and one of which was dissimilar. This dissimilar sequence was inferred to correspond to another region of the pre-miRNA. To clone pre-miRNA319 sequences

(potentially useful for phylogenetic comparison, in situ hybridization or transgenic silencing), the primers EcCINmiRNA319-1F

(GCACTAGTGATTAACTCGAGGCAAATATGG) and EcCINmiRNA319-2R

(CATCTACTCCATATTTGCCTCGAG) were used to amplify DNA from Eschscholzia californica shoot-tip-derived cDNA by PCR. Four unique sequences were obtained from non-exhaustive sequence of transformed plasmid inserts. Prediction of RNA secondary 154 structures for these sequences with the Vienna RNA Websuite (Gruber et al., 2008) supports the hypothesis that these are premiRNA sequences (Figure 32).

Priming sites/ miRNA sites

Figure 32. 2-D centroid model for an Eschscholzia californica premiRNA319 sequence (GLE1388)

155

Sequences

GLE1383 GAGCTCTCTTCAGTCCAGTCACGGAGGTTTCAAGGGTTTGAATTATCTGCCGG CTCATTCATCCAAACACAAAGTAGACACTGGGGAGTACATTTGCTACTGTGA CTGCGTGAATGATACGGGAGATAAATTCCATCCTTTTACCTCTATGATTGGAct GAAGGGAGCTCCCT

GLE1384 GAGCTCTCTTCAGTCCAGTCATGGGTAGCTATGAGATTCAATTAACTGCCGAC TCATTCATCCAAATACTAAGTAAAGAAAAATGATGAAGAACAAGCACAACA AGTACTACTTGCTTGGTAAATGAGTGAATGATGCGGGAGATAAATTGTATCTT TTGCTCCTCTTAATTGGACTGAAGGGAGCTCCCT

GLE1385 GAGCTCTCTTCAGTCCAGTCATGGGTAGCTATGAGATTCAATTAACTGCCGAC TCATTCATCCAAATACTAAGTAAAGAAAAATGATGAAGAACAAGCACAACA AGTACTACTTGCTTGGTAAATGAGTGAATGATGCGGGAGATAAATTGTATCTT TTGCTCCTCTTAATTGGACTGAAGGGAGCTCCCT

GLE1387 GAGCTCTCTTCAGTCCAGTCATGGGTAGCTATGAGATTCAATTAACTGCCGAC TCATTCATCCAAATACTAAGTAAAGAAAAAATGATGAAGAACAAGCACAAC AGGTACTACTTGCTTGGTAAATGAGTGAATGATGCGGGAGATAAATTGTTTCT TTTGCTCCTCTTAATTGGACTGAAGGGAGCTCCCT

GLE1388 GAGCTCTCTTCAGTCCAGTCATGGGTAGCTATGAGATTCAATTAACTGCCGAC TCATTCATCCAAATACTAAGTAAAGAAAAAATGATGAAGAACAAGCACAAC AAGTACTACTTGCTTGGTAAATGAGTGAATGATGCGGGAGATAAATTGTTTCT TTTGCTCCTCTTAATTGGACTGAAGGGAGCTCCCT

GLE1389 GAGCTCTCTTCAGTCCAGTCATGAGTAGTCATAAGATTCAATTATCTGCCGAC TCATTCATCCAAATACTAAGTGAATAAACGATGAATCCCTAACCTATGCAGT ACGTACAACTACTCGCTTGGTAAATGTGTGAATGATACGGGAGATAAATTGA TTCTTTTGCTTCTCGTAATTGGACTGAAGGGAGCTCCCTCCCT

GLE1390 GAGCTCTCTTCAGTCCAGTCATGGGTAGCTATGAGATTCAATTAACTGCCGAC TCATTCATCCAAATACTAAGTAAAGAAAAATGATGAAGAACAAGCACAACA AGTACTACTTGCTTGGTAAATGAGTGAATGATGCGGGAGATAAATTGTATCTT TTGCTCCTCTTAATTGGACTGAAGGGAGCTCCCT

156

References

Arazi T., Talmor-Neiman M., Stav R., Riese M., Huijser P., and Baulcombe D.C.

(2005). Cloning and characterization of micro-RNAs from moss. 43, 837-848.

Barakat A., Wall K., Leebens-Mack J., Wang Y.J., Carlson J.E., and DePamphilis

C.W. (2007). Large-scale identification of microRNAs from a basal eudicot

(Eschscholzia californica) and conservation in flowering plants. 51, 991-1003.

Gruber A.R., Lorenz R., Bernhart S.H., Neuböck R., Hofacker I.L. (2008). The

Vienna RNA Websuite. Nucleic Acids Res. 36, W70-W74.

Li Y., Li C., Ding G., and Jin Y. (2011). Evolution of MIR159/319 microRNA genes and their post-transcriptional regulatory link to siRNA pathways. BMC Evolutionary

Biology 11, 122.

Palatnik J.F., Wollmann H., Schommer C., Schwab R., Boisbouvier J., Rodriguez

R., Warthmann N., Allen E., Dezulian T., Huson D., Carrington J., and Weigel D.

(2007). Sequence and Expression Differences Underlie Functional Specialization of

Arabidopsis MicroRNAs miR159 and miR319. 13, 115-125.

157

Warthmann N., Das S., Lanz C., and Weigel D. (2008). Comparative analysis of the

MIR319a microRNA locus in Arabidopsis and related Brassicaceae. Mol. Biol. Evol. 25,

892-902.

158

APPENDIX II: CLONING OF GIBBERELLIC ACID AND CYTOKININ

METABOLIC GENES FROM ESCHSCHOLZIA CALIFORNICA

Cytokinins

The cytokinins are a family of growth regulators, predominantly

isopentenyladenines but also trans-zeatins (Frebort et al., 2011), that promote cell

division and indeterminacy and regulate processes such as apical dominance and

meristem maintenance. The cytokinin biosynthetic pathway is catalyzed by a series of

enzymes via two alternative pathways, the MVA (isopentenyladenine) and MEP (trans-

zeatin) pathways (Frebort et al., 2011). These use different initial substrates (DMAPP

and HMBDP) that are initially modified by isopentenyltransferases (IPTs) but

intermediate forms from the MVA pathway can be converted to MEP intermediates by

cytochrome p450 monoxygenase. Cytokinin phosphoribohydrolases such as ‘Lonely

Guy’ (LOG) and cytokine N-glucosyl transferases are responsible for producing the

penultimate intermediates and active forms of cytokinin respectively in both pathways.

At several stages of the MEP pathway, intermediates are converted to various cytokinins

by additional enzymes, including zeatin reductase.

Synthesis of the different cytokinin types is compartmentalized, and in

Arabidopsis is performed by IPTs that are spatially and functionally separated (Frebort et

al., 2011). In Arabidopsis, MVA-pathway-specific AtIPT7 is localized to the

mitochondria and AtIPT4 is cytosolic, while the MEP-specific AtIPT1, 3, 5 and 8 are

found in plastids (Frebort et al., 2011). These genes also exhibit different tissue specificities, with AtIPT3, 5 and 7 being expressed ubiquitously (Frebort et al., 2011). 159

Catabolism of active cytokinins is performed by cytokinin dehydrogenases (also known as cytokinin oxidases - CKX) (Frebort et al., 2011). This process is irreversible and is, at present, thought to be performed by this enzyme family only. The CKXs are all glycosylated and share conserved substrate and FAD binding domains, as well as additional conserved motifs associated with the FAD-binding domain and elsewhere

(Frebort et al., 2011). Similarly to the IPTs, different CKXs are found in different subcellular compartments, being vacuolar, cytosolic or apoplastic, and have different substrate specificities and tissue-specific expression (Frebort et al., 2011).

Gibberellins

Gibberellins are tetracyclic diterpenoid growth regulators. Gibberellin biosynthesis requires enzymes such as GA 3 oxidase and GA 20 oxidase to convert transgeranylgeranyl disphosphate to metabolically active gibberellic acids (Thomas and

Sun, 2004). Several essential enzymes are gibberellin, 2-oxoglutarate-dependent dioxygenases, or 2-ODDs. In GA20ox etc. the specific number of course refers to the site of hydroxylation of the substrate (Thomas and Sun, 2004).

2-ODDs commonly have two or three introns. They exhibit homology to one another and share a conserved motif NYYPPCIKP (residues 230-238). Numerous other residues required for prosthetic group binding are conserved, while the 5’ end of the gene is largely unconserved and variable in length (Thomas et al., 1999). An additional short motif LPWKET (residues 148-153) is conserved in GA 20 oxidases only, while many 160 specific residues conserved in GA2Ox are not conserved in GA20Ox and GA3Ox (see

Thomas et al., 1999 for an annotated alignment of this family of genes).

Unlike cytokinin metabolism enzymes, gibberellic acid metabolic enzymes are compartmentalized by function (Olszewski et al., 2002). The final three stages occur in the cytoplasm, where GA20Ox converts the precursors GA12 and GA53 to GA9 and GA20.

GA3Ox converts these to active GA4 and GA1. GA2Ox converts GA9, GA20, GA4 and GA1 to inactive GA34 and GA8 (Olszewski et al., 2002; Sakamoto et al., 2004). The expression of these enzymes is tissue specific, with GA20Ox and GA3Ox sharing common expression patterns (Hedden and Phillips, 2000).

Cytokinin and gibberellin functions

Shani et al. (2010) demonstrated that in tomato, ectopic expression of IPT7 and

CKX3 has opposite effects on leaflet numbers, increasing and decreasing them respectively. The role of cytokinins in promoting cell division, partly through the activation of cyclins (Riou-Khamlichi et al., 1999), as opposed to the promotion of cell expansion and elongation by gibberellins, would suggest a possible role in maintaining the morphogenetic state of developing leaf primordia (Fleishon et al., 2011). This would contrast with the antagonistic roles of gibberellic acids, which favor cell expansion via depletion of DELLA proteins which suppress growth by interfering with basic helix- loop-helix (bHLH) transcription factor binding to DNA targets (Gao et al., 2011). It is not known if TCP transcription factors, which contain a bHLH domain, are targets of

DELLAs, although gibberellins are known to play some part in the regulation of the 161

CINCINNATA homolog LANCEOLATE (Yanai et al., 2011). The possibility raises intriguing questions regarding the integration of multiple developmental maturation pathways. DELLAs also restrict cell division, thus the suppression of their effects broadly promotes organ growth (Achard et al., 2009). Finally, gibberellic acids likely act as suppressors of KNOX gene expression (Singh et al., 2010); Class I KNOX genes are important in the maintenance of cell indeterminacy at the shoot apical meristem and at the margins of developing compound leaves (Hay and Tsiantis, 2010).

Experimental approach

To begin to investigate the role of cytokinin and gibberellin metabolic genes in leaf development in Eschscholzia californica, a species of interest in evolutionary- development research, nucleotide sequences homologous to known cytokinin and gibberellin metabolism genes were identified from Eschscholzia californica transcriptome assemblies in the 1KP database.

A region of an IPT homolog, designated EcIPT, was cloned into pGEM-T using the primers EcIPT-3F (AAACGCGTCGATCAAATGG) and EcIPT-4R

(CCCACCACAAGCTCTTCC) designed multiple 1KP contigs that were similar to

Arabidopsis thaliana IPT. Similarly, part of a GA 20 OXIDASE gene homolog

(EcGA20OX) was cloned with the primers EcGA20Ox-9F

(AACTGGTCCTCATTGTGATCC) and EcGA20Ox-10R

(TGGATCCATTTGGAGAAAGC) that were based upon multiple Eschscholzia californica contigs most similar to Arabidopsis thaliana GA20OX1 and GA20OX2 1KP. 162

The EcGA20OX insert was then amplified using the primers EcGA20Ox11F

(AAGAATTCAACTGGTCCTCATTGTGATCC; contains EcoRI restriction site) and

EcGA20Ox12R (TTCCTAGGGTTGCCCAAGTGAAATCTGG), purified and digested with EcoRI (the sequenc contains an internal EcoRI restriction site) and directionally cloned into pTRV2 (plasmid SG1187) in preparation for Virus-Induced Gene Silencing

(VIGS) in Eschscholzia californica.

Sequences

GLE1385 - EcIPT ATTAAAACGCGTCGATCAAATGGTAAAAATGGGTTTAGTAGAAGAAGTACGA GATATGTTTGAGCCTCACAATAGAGATTATACGCGTGGTATTAGACGCTCAAT TGGTATGCCAGAAATGGACCAATACTTACTACTTGAAGATACTGTTGATGAA GAAACACGCATAAATTTACTCGATATGGCTATTAACGATATCAAAATCAATA CATGTATTTTAGCGTCTCGTCAACTACAAAAAATTTATCGACTTCGTTCCTTA CCGGATTGGAACATACATCGTGTCAGTGCCACAGATGTGTTTTTAATGGAGG GTGGATATTCTTATGATAGATGGGAAGAGCTTGTGGTGGGAATCACTAGTGC GG

GLE1401 – EcGA20OX ATGGCCGCGGGATTAACTGGTCCTCATTGTGATCCAACTTCATTAACAATTCT TCATCAAGATCAAGTTGGTGGTTTACAAGTTTATGTTGATAATCAATGGCATT CTATTGCTCCTAATTCACAAGCATTTGTCGTTAATATCGGCGATACTTTCATG GCTTTATCAAATGGGAGATATAAGAGTTGTTTACACAGAGCAGTAGTAAATA GTGAAACACCAAGAAAATCACTAGCTTTCTTTTTATGCCCTAAAAAAGATAG AAAGGTGTGTCCACCAGAGGAATTGATCAACTTAGAATGTCCAAGAGTTTAC CCAGATTTCACTTGGGCAACTTTTCTTGAATTCACTCAGAAACATTACAGAGC TGATCAAAAGACCCTTGATGCTTTCTCCAAATGGATCCAAATCACTA

References

Achard P., Gusti A., Cheminant S., Alioua M., Dhondt S., Coppens F., Beemster

G.T.S., and Genschik P. (2009). Gibberellin Signaling Controls Cell Proliferation Rate in Arabidopsis. 19, 1188-1193. 163

Fleishon S., Shani E., Ori N., and Weiss D. (2011). Negative reciprocal interactions between gibberellin and cytokinin in tomato. New Phytol. 190, 609-617.

Frébort I., Kowalska M., Hluska T., Frébortová J., and Galuszka P. (2011).

Evolution of cytokinin biosynthesis and degradation. J. Exp. Bot. 62, 2431-2452.

Gao X., Xiao S., Yao Q., Wang Y., and Fu X. (2011). An Updated GA Signaling

‘Relief of Repression’ Regulatory Model. Molecular Plant 4, 601-606.

GarciaMartinez J., LopezDiaz I., SanchezBeltran M., Phillips A., Ward D., Gaskin

P., and Hedden P. (1997). Isolation and transcript analysis of gibberellin 20-oxidase genes in pea and bean in relation to fruit development. Plant Mol. Biol. 33, 1073-1084.

Hay A. and Tsiantis M. (2010). KNOX genes: versatile regulators of plant development and diversity. Development 137, 3153-3165.

Hedden P. and Phillips A.L. (2000). Gibberellin metabolism: new insights revealed by the genes. Trends Plant Sci. 5, 523-530.

Olszewski N., Sun T., and Gubler F. (2002). Gibberellin Signaling: Biosynthesis,

Catabolism, and Response Pathways. The Plant Cell Online 14, S61-S80.

164

Riou-Khamlichi C., Huntley R., Jacqmard A., and Murray J.A.H. (1999). Cytokinin

Activation of Arabidopsis Cell Division Through a D-Type Cyclin. Science 283, 1541-

1544.

Sakamoto T., Miura K., Itoh H., Tatsumi T., Ueguchi-Tanaka M., Ishiyama K.,

Kobayashi M., Agrawal G.K., Takeda S., Abe K., Miyao A., Hirochika H., Kitano

H., Ashikari M., and Matsuoka M. (2004). An Overview of Gibberellin Metabolism

Enzyme Genes and Their Related Mutants in Rice. Plant Physiology 134, 1642-1653.

CORRECTION. (2004). Plant Physiology 135, 1863-1863.

Shani E., Ben-Gera H., Shleizer-Burko S., Burko Y., Weiss D., and Ori N. (2010).

Cytokinin Regulates Compound Leaf Development in Tomato. Plant Cell 22, 3206-3217.

Singh D.P., Filardo F.F., Storey R., Jermakow A.M., Yamaguchi S., and Swain S.M.

(2010). Overexpression of a gibberellin inactivation gene alters seed development,

KNOX gene expression, and plant development in Arabidopsis. Physiol. Plantarum 138,

74-90.

Thomas S.G., Phillips A.L., and Hedden P. (1999). Molecular cloning and functional expression of gibberellin 2- oxidases, multifunctional enzymes involved in gibberellin deactivation. Proceedings of the National Academy of Sciences 96, 4698-4703.

165

Thomas S.G. and Sun T. (2004). Update on Gibberellin Signaling. A Tale of the Tall and the Short. Plant Physiology 135, 668-676.

Yanai O., Shani E., Russ D., and Ori N. (2011). Gibberellin partly mediates

LANCEOLATE activity in tomato. 68, 571-582.

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Thesis and Dissertation Services ! !