Transport in Plants and in Root Nodules

DEVELOPMENT OF SPECIALIZED TRANSPORT PATHWAYS IN PISUM

SATIVUM ROOT NODULES

Cherish A. Warner

A dissertation submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences

Fall 2018

DEVELOPMENT OF SPECIALIZED TRANSPORT PATHWAYS IN PISUM

SATIVUM ROOT NODULES

Cherish A. Warner

Approved: ______E. Fidelma Boyd, Ph.D. Chair of the Department of Biological Sciences

Approved: ______John Pelesko, Ph.D. Interim Dean of the College of Arts and Sciences

Approved: ______Douglas J. Doren, Ph.D. Interim Vice Provost for Graduate and Professional Education

I certify that I have read this dissertation and that in my opinion it meets the academic and professional standard required by the University as a dissertation for the degree of Doctor of Philosophy.

Signed: ______D. Janine Sherrier, Ph.D. Professor in charge of dissertation

Signed: ______Jeffrey L. Caplan, Ph.D. Member of dissertation committee

Signed: ______Diane Herson, Ph.D. Member of dissertation committee

Signed: ______Jia Song, Ph.D. Member of dissertation committee

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Janine Sherrier for her advising, encouragement and invaluable contribution on my thesis, my projects, and my research in both general and specific terms. I thank all the members of Delaware Biotechnology

Institute for creating such an inclusive and wonderful supportive community. I thank the people of the Bioimaging Center of Delaware Biotechnology Institute, especially Dr. Jeffrey Caplan and Shannon Modla for providing outstanding training and assistance on the imaging instruments and techniques. I thank all my friends and family for their unwavering support. In others eyes I was the least likely to come this far but the most stubborn of us to do it. I want to dedicate this piece of writing to Ta and Gran (Norman and Loretta Gilbreath) whose work ethic and life story was inspirational in times of hardship. To you Pete, my best friend, I thank you for everything and love you with all my heart. I thank my Dad Doug, for always listening. I would like to thank my Mom Kathy, who raised three wonderful girls, myself and my sisters, Daisha and January. A special thanks to my second Dad, Steve Belcher who cared for me like I was his own. I love you all and couldn’t have done this without you. To my nephews Luis and Scott, and nieces August, Vivian and Lillian, I hope your dreams come true and I believe in you. I thank all my current and previous labmates for giving me friendship and support, both professionally and personally. I want to thank all those not named above. Your support and kindness made this possible. I’m tremendously grateful for your guidance and counsel during the course of my graduate studies.

TABLE OF CONTENTS

LIST OF TABLES ...... ix LIST OF FIGURES ...... x ABSTRACT ...... xii

Chapter

1 INTRODUCTION ...... 1

1.1 Statement of Research Question ...... 1 1.2 The Importance of Legumes ...... 3

1.2.1 Feeding Mankind; Chemical versus biological nitrogen fixation .. 4

1.3 An overview of the L-R Symbiosis and Organ Formation ...... 9

1.3.1 In the Beginning: Evolution of L-R Symbiosis ...... 9 1.3.2 “Sweet talking”; Initiation of Symbiosis ...... 10 1.3.3 Organogenesis ...... 11 1.3.4 Maintenance of a Functional Root Nodule ...... 13

1.4 Transport Pathways in Plants ...... 14

1.4.1 Transport and Structure of Vascular System ...... 14 1.4.2 Two types of ‘cell to cell’ transport in Plants and in Root Nodules ...... 16 1.4.3 The Plant Wall Barrier ...... 19 1.4.4 Formation and Structure of PD ...... 19 1.4.5 Mechanisms of Function ...... 20

1.5 Transport and Metabolism To and From Infected Cells ...... 22 1.6 Supporting Literature and Research Approach ...... 23

2 MATERIALS AND METHODS ...... 26

2.1 Materials ...... 26

2.1.1 Bacterial Strains ...... 26

2.1.2 Plant Species ...... 26 2.1.3 Reagents and Enzymes ...... 26 2.1.4 Microscopy Stains and Dyes ...... 27 2.1.5 Standard Solutions ...... 28

2.2 Methods...... 29

2.2.1 Bacterial Culture ...... 29 2.2.2 Pisum sativum Plant Growth ...... 30

2.2.2.1 Seed Germination...... 30 2.2.2.2 Pisum sativum Plant Growth ...... 30

2.2.3 Medicago truncatula Plant Growth ...... 32

2.2.3.1 Seed Germination...... 32 2.2.3.2 Plant Growth ...... 33 2.2.3.3 Rhizobial Inoculation of Caisson ...... 33

2.2.4 Harvest of Root Nodules ...... 34 2.2.5 Optical Clearing of Nodule Tissue ...... 34

2.2.5.1 Sample Preparation and Infiltration with Clearing Reagent ...... 34 2.2.5.2 Data Acquisition for Confocal Microscopy ...... 35 2.2.5.3 Data Processing for Three-Dimensional Reconstruction and Segmentation...... 36

2.2.6 Transmission Electron Microscopy- Plasmodesmata Quantification ...... 36

2.2.6.1 Sample Preparation ...... 36 2.2.6.2 Data Acquisition ...... 37 2.2.6.3 Plasmodesmata Scoring for Quantification ...... 38

2.2.7 Transmission Electron Microscopy- Plasmodesmata Tomography ...... 38

2.2.7.1 Sample Preparation ...... 39 2.2.7.2 Data Acquisition ...... 40 2.2.7.3 Tomogram Reconstruction and 3D Modeling ...... 40

2.2.8 Mapping of Uninfected Cell Arrays and Vasculature Tissue ...... 41 2.2.9 Dye Loading Technique ...... 41

3 SYMPLASMIC FLOW IN CENTRAL NODULE TISSUE...... 43

3.1 Background ...... 43

3.1.1 Nitrogen transport in determinate and indeterminate nodules ..... 46 3.1.2 Sucrose transport and tissue organization in determinate and indeterminate nodules ...... 47 3.1.3 Other aspects of transport involved in nodulation ...... 47 3.1.4 The role of PD in symplasmic transport ...... 48 3.1.5 Transport and Tissue Organization in the Indeterminate Nodule 50

3.2 Results ...... 52

3.2.1 Nodule vasculature is branched and peripheral ...... 52 3.2.2 Patterns of uninfected cells are present from bundles to central tissue ...... 55 3.2.3 Callose antibody localized primarily to uninfected cell wall interfaces ...... 56 3.2.4 Uninfected cell wall interfaces contained more PD channels ...... 61 3.2.5 PD distribution within each zone differed depending on wall interface...... 63 3.2.6 PD Density increased along uninfected cell walls through development ...... 69 3.2.7 PD along uninfected cell walls become more complex in structure...... 70 3.2.8 Uninfected cells contain more clustered PD ...... 73 3.2.9 PD ultrastructure revealed simple PD could be complex ...... 77 3.2.10 Symplasmic transport occurs directly into developmental zones through uninfected cells ...... 79

3.3 Discussion ...... 82

3.3.1 Uninfected tissue and vascular mapping support current literature ...... 82 3.3.2 Uninfected cells are more connected ...... 83 3.3.3 Distribution across developmental gradient suggests secondary PD formation ...... 83 3.3.4 Formation of secondary PD is accompanied by branching ...... 84 3.3.5 PD simple in structure could be complex ...... 85 3.3.6 Uninfected cells are active in symplasmic transport into the central tissue...... 86

vii

4 SCALEP OPTICALLY CLEARS WHOLE PLANT ORGANS ALLOWING FOR DEEP TISSUE IMAGING AND USE OF FLUORESCENT MICROSCOPY TECHNIQUE ...... 89

4.1 Background ...... 89 4.2 Results ...... 91

4.2.1 ScaleP clearing increases tissue transparency in multiple tissue types ...... 91 4.2.2 ScaleP clearing increased depth of imaging by 3-fold ...... 94 4.2.3 ScaleP clearing effective for studying plant microbe interactions94 4.2.4 ScaleP is compatible with commonly used fluorescent markers, probes, stains and dyes ...... 94

4.3 Discussion ...... 98

5 IMPLICATIONS AND FUTURE WORK ...... 99

REFERENCES ...... 101

Appendix

A SUPPLEMENTAL MATERIALS...... 110 B PERMISSION FOR REPRINT AND SUPPLEMENTAL MATERIALS FOR CHAPTER 4 ...... 113

viii

LIST OF TABLES

Table 1.1 World Population Milestones ...... 5

Table 2.1 Stains and Dyes Utilized in Confocal Microscopy ...... 27

Table A.1 Plasmodesmata Analysis Totals: Measurements and Counts...... 111

LIST OF FIGURES

Figure 1.1 Estimated world population by the end of the Century...... 6

Figure 1.2 Picture of legume model Medicago truncatula...... 8

Figure 1.3 Developmental Zones of a Mature Indeterminate Root Nodule...... 12

Figure 1.4 Structure and Fluid Movement of Plant Vasculature...... 15

Figure 1.5 Diagram of the Apoplast and Symplast...... 17

Figure 1.6 Structure of a Plasmodesma...... 18

Figure 3.1 Structure of the HPTS Acetate used in Dye Loading Experiments.. 52

Figure 3.2 Nodule vasculature branches radially at zone II and the apical meristem...... 54

Figure 3.3 Uninfected cell arrays link periphery to central cortex...... 56

Figure 3.4 Callose antibody localized along uninfected cells...... 58

Figure 3.5 Distribution changes between cell types but not developmental zones...... 62

Figure 3.6 Differences in PD connections between cell types start in zone II. ... 64

Figure 3.7 Differences in PD distribution between cell types in zone II/II...... 66

Figure 3.8 Differences in PD distribution between cell types within the nitrogen fixation zone III...... 68

Figure 3.9 Differences in PD frequencies across development...... 70

Figure 3.10 The wall interface between adjacent uninfected cells have more complex PD...... 72

Figure 3.11 Percentage of complex PD increased in uninfected cells during development...... 73

Figure 3.12 Uninfected cells have more clusters of PD...... 75

Figure 3.13 Clusters between infected-uninfected cells have more channels...... 77

Figure 3.14 Plasmodesmata with simple structures could be complex and branched...... 79

Figure 3.15 HPTS unloaded from vascular bundles enters the central tissue via uninfected cell networks...... 81

Figure 3.16 Schematic of HPTS flow in Mature Indeterminate Root Nodules. ... 87

Figure 4.1 ScaleP optically clears whole plant organs increasing the depth of imaging possible...... 93

Figure 4.2 Compatibility of ScaleP cleared specimens with fluorescent markers...... 97

Figure A.1 Flask growth and nodule harvest method...... 110

Figure A.2 Method of Plasmodesmata Quantification...... 112

Figure B.1 Clearing effects in monocot and dicot leaf tissues...... 113

Figure B.2 Permission to Reprint Published Works...... 114

ABSTRACT

Legumes, from the family Fabaceae, represent 30% of global primary crop produced, with grain legumes contributing approximately one third of humans’ dietary protein needs. Legumes have the unique ability to form a symbiotic relationship with nitrogen-fixing soil bacteria called rhizobia, to obtain one of the most limiting nutrients for plant growth, nitrogen. Of particular importance is their ability to replenish nitrogen in the soil with an estimated nitrogen sequestration greater than 40 million tons annually. The study of the Legume-Rhizobium symbiosis could help restore the Earth’s carbon-nitrogen balance and potentially allow us to induce the symbiosis in non-legume crops.

Legume-rhizobia symbioses result in formation of a novel plant organ called a root nodule which houses the nitrogen-fixing bacteria. Components necessary for the formation, maintenance, and persistence of this organ must be transported to and from the nodule with great efficiency. Mature nodules are considered sink organs and require large amounts of imported sugar derived from photosynthesis to support the endosymbionts. In addition, the plant effectively manages export of assimilated organic nitrogen compounds from the organ.

Few studies have examined transport within nitrogen-fixing nodules formed during the L-R symbiosis, particularly with regard to symplasmic transport, which is

xii

short-distance transport from cell-to-cell through intercellular channels called plasmodesmata (PD). The transport capacity of a functional mature nodule must support phloem loading and unloading of materials into and out of the organ. Paths of transport to and from infected cells must also be established for a productive symbiotic relationship. The recognized importance of symplasmic connectivity in sink organ formation and function suggests indeterminate nodules must continue to develop PD beyond nodule initiation to further support the symbiosis.

Symplasmic connectivity of cells within the nodules’ central tissue was evaluated to further elucidate how the organ handles the fluxes of nutrient transport taking place. Connectivity of infected and uninfected cells was determined by quantification of the symplasts’ functional unit, plasmodesmata. TEM analysis of the frequency and distribution of these channels showed increased connections between adjacent uninfected cells and between uninfected and infected cells. These connections increased in frequency and complexity through the developmental gradient of an indeterminate nodule. The connections between two adjacent infected cells did not follow the same trend of increased PD connections nor channel complexity, even though large amounts of nitrogen are exported and sugar requirements are higher in these cell types. Interestingly, the greatest change in PD complexity was seen between infected-uninfected cell types and not adjacent uninfected cells, which suggests these cells become more interconnected as the organ matures.

To determine if uninfected cells are active in symplasmic transport, a fluorescent tracer molecule was used to map the movement of water. Assessment of symplasmic

xiii

flow using HPTS (8-Hydroxypyrene-1,3,6-Trisulfonic Acid, Trisodium Salt) revealed transport from the nodule vascular phloem into the parenchyma cell layer adjacent to the bundles. Dye then entered the vascular endodermal layer and radiated into the central nodule tissue, primarily through bands of uninfected cells. HPTS fluorescence was shown entering directly into each developmental zone and not exclusively into the nodule meristem. This suggests an additional symplasmic route is established during organ maturation.

To map the nodule vasculature and uninfected cell networks, a novel clearing technique was developed to image deep into plant material. The goal was to develop and fine tune a method to facilitate deep-tissue imaging of intact plant organs or whole plants. Specifically, we aimed to develop an approach to retain fine cellular features in specimens, closely match the refractive indices of plant tissues, enhance light transmission through the sample, and preserve the ability to use common fluorescent stains and proteins. The plant-tissue clearing agent presented here overcomes all the major obstacles that have limited plant imaging to date, and it is, therefore, of broad use to the plant scientific community.

The application of the clearing technique called ScaleP revealed a novel view of nodule vasculature branching. Imaging of cleared nodules showed uniform branching along the periphery of the nodule. In addition, rays of uninfected cells seen radiating from the periphery to the central tissue occurred in all developmental zones of the organ. Further studies using this clearing technique could reveal a developmental link

xiv

with timing of vascular branching and would provide insight into the role uninfected tissue in the indeterminate root nodule.

Chapter 1

INTRODUCTION

1.1 Statement of Research Question As plant cells mature, they differentiate to perform or facilitate specialized functions. Examples of cellular changes that occur during differentiation include variations in cell shape or cell size, signaling cascade and enzymatic activation, modification of structural components, and accessibility to neighboring cells. Since plant vasculature provides the bulk flow of water, sugars, proteins, and signaling molecules needed for the above functions, plants have established mechanisms to rapidly repair and reorganize vascular networks to facilitate both long and short- distance transport needs. Cell-to-cell transport between tissue systems and the vasculature occurs via the apoplast and/or symplast. This short-distance transport helps distribute or remove the necessary molecules for cellular functions connecting these tissue systems with long-distance transport pathways. These efficient mechanisms of restructuring have been well documented in a range of processes from grafting to establishment of symbiotic relationships. In the legume-rhizobia symbiosis, a mature nitrogen-fixing root nodule must establish and later modify transport pathways to accommodate the functional requirements of the organ through development. In plants that form indeterminate nodules, the plant must provide resources to support continual meristematic activity,

further organ formation, sustain the endosymbionts, and facilitate transport of assimilated nitrogen compounds from the nodule to the main plant body. To better understand nodule transport, it is necessary to identify spatially and temporally the points during organ development where cells transport abilities become specialized. Transport specialization in other tissue systems during organ formation occurred through changes in the vasculatures gross structure and in the functional unit of cell-to- cell transport, the plasmodesmata. Therefore, it is likely plants modify these structures in a similar fashion to accommodate nodulation. Current knowledge of transport through nodule vascular suggests sugar derived from photosynthesis is transported via the phloem and other required material is transported through the xylem or phloem (Abd-Alla et al., 2000; Complainville et al., 2003). After unloading into the organ, the short-distance transport of some solutes occurs through the functional units of the symplast known as plasmodesmata. In early nodule development, cells which form the nodule primordium were shown to be connected symplastically, via plasmodesmata, through the nodule vasculature to the root phloem with the ability to transport bulky macromolecules, such as GFP (27kD) (Complainville et al., 2003). Plasmodesmata (PD) are channels which connect adjacent cells cytoplasm to create a continuous pathway for cell-to-cell flow of solutes. Studies of the differences in frequency and complexity of these channels between cell types throughout nodule development provide insight into the overall flow capacity and establishment of symplasmic pathways. A particularly interesting aspect of short-distance transport in nodules, is what happens to PD during cellular differentiation and specialization. In the

nodule, two cell types are established, one functioning as hosts to the symbiont while the other remains uninfected. This study focuses on specialized pathways involved in short-distance transport through the symplast in the indeterminate nodule of Pisum sativum. We hypothesis that one cell type supports the majority of symplasmic transport within the nodule and not only do these cells have more PD but the frequency and complexity of the channels changes throughout development.

1.2 The Importance of Legumes The word legume is derived from the Latin term legumen roughly translated as “anything that can be gathered”. There are more than 18,000 different species of legumes found in a wide range of environments, which include shrubs, bushes, trees, herbs, and even vines. Their high nutrient value and protein content make them an important food for many animals including humans. The ability of legumes to obtain their own source of nitrogen and replenish depleted nitrogen in soils make them integral to many ecosystems and in green agriculture. Legumes such as peas, soybean, or alfalfa were among the first cultivated plants in the world (Harris, R. David, Hillman, 2014), and continue to be grown widely today. Approximately 15% of Earth’s arable land is allocated to legume crops (Graham and Vance, 2003), representing 30% of global primary crop production annually (FAO 2012), with an estimated 60 million tons at the global market (Chaudhry, 2011). Grain legumes alone contribute one third of dietary protein needs of humans (Vance et al.,

2000) and under subsistence conditions, this can reach twice that value (Graham and Vance, 2003).

The unique qualities of legumes and the role they play in both past and present civilizations have made them mandatory organisms for research. Today two types of legumes are cultivated: Grain legumes and forage legumes. Grain legumes such as peanuts, peas, and lentils are grown for seed collection. They are used for oil production, human consumption, and animal feed. Forage legumes, such as clover and alfalfa, are grown for grazing and feeding of livestock. Advances in our understanding of legumes over the past century have improved our cultivation techniques of these and many other vital crops helping to feed the ever increasing human population, (Charpentier and Oldroyd, 2010; Roser and Ortiz-Ospina, 2018).

1.2.1 Feeding Mankind; Chemical versus biological nitrogen fixation William Crookes was among the first to describe the problem of nitrogen availability limiting the world’s crop yields (Crookes, 1898). His 1898 paper pointed out that elemental nitrogen, which constitutes the majority of our atmosphere as dinitrogen, is only accessible to plants as a water-soluble compound such as ammonia or nitrates. The continued rise in human population and insufficient crop yields to keep pace was cause for concern and spurred research into synthetic chemical fertilizers. In 1909, two chemists by the name Fritz Haber and Carl Bosch, developed a method allowing for artificial nitrogen fixation. Known today as the Haber-Bosch process. The first industrial application of the Haber-Bosch process for large scale ammonia production was in 1913 and is still used today in the industrial production of chemical nitrogen fertilizers. These gentlemen were later awarded the Nobel Prize for their work on artificial nitrogen fixation. By the mid 1900’s, production and use of such synthetic fertilizers was steadily increasing and was viewed as the solution for increasing food production to keep pace

with the world’s increasing human population. As the use of industrially-produced nitrogen fertilizer became more common and widespread, the resulting boost to crop yields allowed millions of people to obtain food security. However, rather than keeping pace with population growth, widespread use of chemical nitrogen fertilizer correlates with the unanticipated acceleration in growth of the world population (Table 1.1), which is expected to reach an estimated 9.3 billion by 2050 (Roser and Ortiz-Ospina, 2018).

Table 1.1 World Population Milestones. Data Source: World population over 12,000 years-Various Sources (2016), Medium Projection-UN Population Division (2015 Revision). Published online at OurworldInData.org. Retrieved from: ‘https://ourworldindata.org/world-population-growth’ (Roser and Ortiz-Ospina, 2018).

World population reached: Year Time to add 1 billion 1 billion 1804 2 billion 1927 123 years 3 billion 1960 33 years 4 billion 1974 14 years 5 billion 1987 13 years 6 billion 1999 12 years

Figure 1.1 Estimated world population by the end of the Century. Graphic obtained from https://www.businessinsider.com/un-world-population-projections-2015- 7.

Industrially-produced fertilizers, in particular nitrogen, are energy-intensive to make and transport. Environmental impacts of such large-scale production and the application of large quantities of fertilizers are rife with both pros and cons (Crews and Peoples, 2004; Helming et al., 2014; Qiao et al., 2014). For instance, crops, including legumes, are inefficient at the uptake of nitrogen when present at high levels in the soil, as often happens when soil is amended with nitrogen fertilizer (Morel et al., 2012; Helming et al., 2014). Approximately half of all applied chemical nitrogen fertilizer is utilized by plants. The remaining fertilizer remains in the environment. This excess

nitrogen causes eutrophication of waterways and has negative impacts on the environment and human health. Decades of chemical nitrogen fertilizer overuse has contributed to the formation of dead zones along coastal shorelines, massive marine die- offs, and frequent algal blooms. Some countries have started addressing these problems by reducing chemical fertilizer use through education, regulations, and the implementation of other viable alternatives. The most promising alternative to reduce dependence on chemical nitrogen fertilizer comes in the form of biomimicry of biological nitrogen fixation. Unique among all plants, legumes can form root nodules (Figure A.1 and Figure 1.2), which are novel organs that form on the root systems to host the mutualistic bacteria, rhizobia. Rhizobia are nitrogen-fixing bacteria that, when in symbiosis with a host plant, convert atmospheric nitrogen into a form that is bioavailable to plants. This naturally occurring partnership, called the legume-rhizobia symbiosis, provides approximately 40 million tons of atmospheric nitrogen yearly (Herridge et al., 2008). This is more than half of the

70 million tons of chemical nitrogen fertilizer produced yearly (https://ourworldindata.org). Formation of this symbiosis used in conjunction with crop rotation could help to rebalance carbon and nitrogen cycling within the environment.

Figure 1.2 Picture of legume model Medicago truncatula. (A) M. truncatula forms an endosymbiotic relationship with the nitrogen-fixing bacterium, Sinorhizobium meliloti. (B) Insert displays root nodules at 14 days post inoculation. Bar = 1 mm. Figure curtesy of Joan K. Gisiora Sherrier Lab.

The identification of root nodules on leguminous plants is credited to Hellriegel and Wilfarth in 1888, but the advantage this organ gives legumes was apparent to mankind much earlier. The Egyptians were probably the first to use legumes in intercropping (Malleson, 2016). While the microfloral underpinnings of the symbiosis were unknown, the macrobiotic benefits were quite evident, even to some of the earliest agronomists with a written history.

The integral role legume crops have in our society is evident in the story of Dr. Johanna Döbereiner, a legume researcher (Hirsch, 2009). The OPEC embargo of the early 1970’s had significantly increased the price of chemical fertilizer production, incentivizing research into alternative means of crop nutrition. Dr. Döbereiner was among the first to apply the then recently-developed acetylene reduction assay (ARA), to quantify biological nitrogen fixation activity in legumes. The push for alternative sources of nitrogen and the ability to measure the natural biological acquisition of nitrogen led to the discovery and isolation of many nitrogen fixing-bacterial species. These species, in turn were used as bio inoculants reducing the use of industrially-produced nitrogen fertilizers (Smith, 1992; Clayton and Rice, 2004; Crews and Peoples, 2004; Bala et al., 2011; Rahman et al., 2014). These successes further fueled research into L-R symbiosis and other forms of biological nitrogen-fixation, with the hopes of nodulating non-legumes or perfecting nitrogen sequestration (Santi et al., 2013; Rogers and Oldroyd, 2014; Remigi et al., 2015).

1.3 An overview of the L-R Symbiosis and Organ Formation

1.3.1 In the Beginning: Evolution of L-R Symbiosis The legume-rhizobia relationship probably started off adversarial. It is possible that rhizobia initially attacked plants roots, colonizing the surfaces and penetrating the root. In response, the legume would start defensive measures to reduce the extent of infection. In doing so, the rhizobia would become walled-off or trapped, unable to progress further into the root, ultimately leading to the formation of the nodule. Within the root nodule environment, rhizobia, then reverted to a diazotroph. This was made possible within the nodules ideal micro-aerobic environment, which allowed for activity

of the bacterially-encoded enzyme nitrogenase. Over millions of years this infection slowly evolved into a facultative symbiotic relationship. This is a gross over simplification describing years of evolutionary trial and error. There are several interesting in-depth reviews on the evolution of the legume-Rhizobium symbiosis abbreviated L-R symbiosis, each with unique approaches to examine the development of this intricate relationship (Sprent and Embrapa, 1980). Even today, this complex relationship could “sour” turning into a pathogenic infection at any point during nodulation. As a result, many legumes and rhizobial strains can be highly selective with whom they associate. It would have to be an obligate symbiosis for just any old strain to become a legumes endosymbiont, having resulted out of necessity for survival rather than being mutually beneficial. This is why one of the most unique aspects of the L-R relationship is its selectivity. Sometimes described as locks and keys, but better described as ‘‘handshaking’’ (Thompson, 2005; Herre, 2006) and ‘‘sweet talking’’(Cullimore and Dénarié, 2003) these initial interactions have been studied for decades in order to decipher the complex interplay necessary to start the L-R symbiotic relationship.

1.3.2 “Sweet talking”; Initiation of Symbiosis The “sweet talking” begins when a legume plant is in need of nitrogen. Nitrogen-bereft roots release signaling molecules called flavonoids into the rhizosphere to initiate the relationship. If the correct rhizobial strain is present in the surrounding soil, the flavonoids initiate expression of nodulation genes within the bacteria and subsequently interacts with the corresponding Nod proteins (transcriptional activator proteins). This starts production of Nod Factors (lipooligosaccharide signaling molecules) which are released by rhizobia and recognized by the host plant.

A later step in these early signaling events is the perception of rhizobia- generated Nod factors (NF) by the host root. Upon perception of NF, root nodule organogenesis begins (Reviewed by Brewin, 1991; Hirsch, 1992). The root hair curls and deforms, and the inner cortical cell layer divides establishing the nodule primordium. Rhizobia colonize the root surface before entering one of two ways; either by entry through cracks between cells, or more commonly in the Pisum-Rhizobium symbiosis, becoming trapped in a curled root hair. Only two or three Rhizobium get trapped, where they then enter the root through a plant derived tubular infection thread (IT), which traverses intracellularly through the epidermis and outer cortical cell layers of the root.

1.3.3 Organogenesis A mature indeterminate nodule can be described by the cells developmental stage. Each stage is considered a zone according to cellular function which is describe in detail below (Figure 1.3). The cells of a nodule’s meristem are undifferentiated. As organogenesis progresses, post-meristematic cells encounter IT filled with rhizobia. The cells continue to differentiate in developmental zone II of the nodule, where ITs invade and release bacteroids into some cells, but not others. We still do not understand the catalyst for differentiation into potential host cells. The IT is encapsulated by a thick plant derived matrix, made of cell wall and plasma membrane material (Newcomb et al., 1979). Within the IT, bacteria continue to grow and divide. ITs traverse through and between the cells’ walls maneuvered by the cells’ cytoskeleton (Timmers et al., 1999;

Kitaeva et al., 2016). As rhizobia are released from the IT, each bacterium is provided with a plant-derived outer membrane, called the symbiosome membrane. After which,

they continue to grow and divide within these infected cells’ cytoplasm. The infected cells subsequently fill with differentiated rhizobia referred to as bacteroids.

Figure 1.3 Developmental Zones of a Mature Indeterminate Root Nodule. Infection thread networks (yellow) traverse newly divided cells within zone II. Starch granules (small light blue circles) accumulated within infected cells (dark blue rectangles) indicate zone II/III. Infected cells filled with nitrogen fixing bacteroids are much larger than uninfected cells (white rectangles) within the nitrogen fixation zone III.

The newly-infected cells have now extended further away from the point of origin, within a developmental zone described as the intermediate zone II/III. Here rhizobia undergo significant metabolic and morphological modifications distinguishing them from their free-living counterparts (Karunakaran et al., 2009). Their morphology changes from a traditional rod shaped gram-negative bacterium to more of a “Y” shape, but have not begun nitrogen fixation because their metabolism has not adjusted to accommodate metabolism of carbohydrates within the micro-aerobic environment within the root nodule. Plant cells within zone II/III convert unutilized sugars into starch, since it can no longer be utilized by the bacteroids, resulting in a distinct starch band seen in several cell layers. Maturation of these cells signify the next stage of development referred to as the nitrogen fixation zone III where the infected cells have grown to twice their original size while uninfected cells’ size remains relatively unchanged.

1.3.4 Maintenance of a Functional Root Nodule Maintenance of zone III requires leghemoglobin, a hemoprotein like hemoglobin in color, structure, and chemical properties, is produced by both the plant and bacterium (Brian et al., 1987; Santana et al., 1998) to establish and maintain low levels of oxygen. This structure effectively buffers the free oxygen content in the cells’ cytoplasm, increasing the activity of the oxygen-sensitive enzyme nitrogenase which catalyzes nitrogen fixation, while maintaining the symbionts respiratory oxygen needs (Appleby, 1984). This hypoxic environment and the presence of leghemoglobin results in the signature pinkish color indicative of an active nitrogen-fixing root nodule. Integral to the maintenance and persistence of a mature nodule includes rapid removal of fixed nitrogen and delivery of provisions necessary for upkeep of the

symbionts and meristematic region to maintain homeostasis (Walsh, 1990; Roberts and Oparka, 2003; Guinel, 2009a; Udvardi and Poole, 2013). Though there are many molecular and physiological aspects of nodule homeostasis, it can effectively be described as the plants way of mediating signaling cascades and transport of macromolecules between cells within a nodule (Pickard, 2003; Knox and Benitez- Alfonso, 2014). Efficient control of cellular cues and transport within one region of tissue independent of another, allows the plant to compartmentalize. The two types of cell-to-cell transport in plants is briefly described below.

1.4 Transport Pathways in Plants

1.4.1 Transport and Structure of Vascular System The vascular system consists of phloem and xylem, each comprised of many cell types (Figure 1.4). The xylem consists of dead cells with thick rigid walls made of lignin. These cells transport water and minerals from the root system of plants to the leaves. The phloem, comprised of multiple living cell types, contain thin permeable walls made of cellulose. The unidirectional flow carries water, nutrients, and minerals to and from storage organs and growing plant parts. The phloem contains a specialized cell type known as sieve elements which are responsible for transporting sugars produced during photosynthesis. At maturity these conducting cells lack nuclei and rely on companion cells or albuminous cells for metabolic support. The sieve tube cell retains their endoplasmic reticulum (ER) at maturity which localizes along the cell walls. The ER is often associated with PD that connect sieve elements to their companion or albuminus cells. All sieve cells have groups of PD that overtime, become enlarged and modified creating large porous areas known as sieve areas.

Figure 1.4 Structure and Fluid Movement of Plant Vasculature. (a) Phloem is comprised of cells called sieve-tube elements. Phloem sap travels through perforations called sieve tube plates. Neighboring companion cells carry out metabolic functions for the sieve-tube elements and provide them with energy. Lateral sieve areas connect the sieve-tube elements to the companion cells. (b) Sucrose is actively transported from source cells into companion cells and then into the sieve-tube elements. This reduces the water potential, which causes water to enter the phloem from the xylem. The resulting positive pressure forces the sucrose-water mixture down toward the roots, where sucrose is unloaded. Transpiration causes water to return to the leaves through the xylem vessels. Diagram available from open source: OpenStax CNXhttp://cnx.org/contents/[email protected]

Companion cells are specialized parenchyma cells, which carry out all cellular functions for the sieve-tube element. Companion cells which have dense cytoplasm are connected to the adjacent sieve-tube elements by PD. There are two types of companion cells, ordinary companion cells and transfer cells (Pate et al., 1969). Transfer cells contain secondary wall ingrowths lined with plasma membrane creating large surface areas specialized for solute transport. These companion cells are usually located at “bottlenecks” within nutrient pathways where the wall ingrowth/plasma membrane complexes are located in the direction of solution flow. Transfer cells may be involved in formation of symbiosis with mycorrhizal and rhizobium-root nodules (Gunning et al., 1974; Allaway et al., 1985; Berry et al., 1986).

1.4.2 Two types of ‘cell to cell’ transport in Plants and in Root Nodules Plants facilitate cell to cell transport via two routes, the apoplast or symplast (Figure 1.5). The apoplast comprises the extracellular spaces where transport occurs. Whereas the symplast is the intercellular connectivity between cytoplasm’s of neighboring cells.

Figure 1.5 The Apoplast and Symplast. Water and dissolved nutrients can move from soil to the xylem by two pathways-apoplast or symplast. Diagram from Introduction to Plant Physiology, Copyright John Wiley & Sons.

The term apoplast is defined by the space outside the plasma membrane where materials can diffuse freely. This can be impeded by air spaces or plant cuticle between tissues. In roots, apoplasmic transport is impeded by the Casparian strip which consists of suberin and sometimes lignin. Unique from the rest of the cell wall, these specialized wall structures are found in radial and transverse walls of the endoderm, normally around vascular bundles. They prevent leakage back into the cortex. These strips were first described by Robert Caspary in the early 1800s and since found to play a role in many tissue systems including root nodules, See review (Pate et al., 1969; Brown and Walsh, 1994). Once solutes in the apoplast reach a Casparian strip they are forced into the symplast.

The symplast describes the space where intercellular transport of solutes between adjacent cell’s cytoplasm occurs. The flow of this pathway is facilitated and largely controlled by the properties of channels crossing the cell wall, called PD (Figure 1.6). Plasmodesma (singular) or plasmodesmata (plural), abbreviated PD, allow for small molecules such as signaling molecules, sugars, or amino acids to flow directly from cell to cell. This creates a cytoplasmic continuity between cells within a tissue system and facilitates compartmentalization for a multitude of processes. These compartments are characterized by cells whose PD have similar size exclusion limits (SEL). The SEL is controlled by symplasmic microdomains. This term describes the walls surrounding PD, which are modified to affect the channels diameter, thus restricting movement. For example at interfaces where cells with different functions are connected by PD (Christensen et al., 2009).

Figure 1.6 Structure of a Plasmodesma. Schematic representation of a simple stranded PD. ER, Endoplasmic Reticulum; PM, Plasma Membrane; PD-PM, Plasma Membrane lining PD. Diagram source: (Salmon and Bayer, 2013)

1.4.3 The Plant Wall Barrier The wall is the first point of entry of materials into cells and must provide the mechanical strengthening factors needed to withstand mechanical stresses while still allowing cell growth. Thus, cell wall modifications must provide flexibility and plasticity. Little is known about cell wall modifications made during nodule organogenesis except modifications must be extensive to accommodate cells’ morphological differences seen between cell types within each functional/developmental zone. Transport across cell walls is facilitated by the functional unit of the symplast, the PD. The formation, modification, and regulation of these cytoplasmic channels is discussed below. Followed by our current understanding of symplasmic transport activity within a mature nodule.

1.4.4 Formation and Structure of PD Plasmodesmata formation can be described as primary or secondary. Primary PD form during cell division or cytokinesis when endoplasmic reticulum (ER) is present in the middle lamella space and the new wall material is being deposited. The ER essentially gets trapped in this space and is repurposed into one of three components of PD: the plasma membrane, cytoplasmic sleeve, or desmotubule (Robards and Lucas, 1990; Heinlein, 2015). These PD are normally simple in structure and interconnect adjacent cells cytoplasm by the plasmalemma channel crossing the cells wall. This creates a continuous path, joining two adjacent cells’ plasma membranes. Formation of PD after cytokinesis are considered secondary in nature. In general, their structures are more complex and serve a multitude of functions, with a large range of molecules able to pass through. They are indistinguishable from primary PD based on structure alone. Secondary PD formation have been described in several

processes such as graft unions, where the mechanism of secondary-plasmodesma formation at the graft interface revealed localized thinning and loosening of cell walls. This allowed for adjacent cells plasma membranes to interact and fuse together before becoming trapped during further modification of the wall materials (Kollmann, R., 1991; Ehlers and Kollmann, 2001). Primary and secondary PD can be simple or complex in structure. The most observable complex structures are distinct types of branching. These include Y- or X- shaped PD and higher-order combinations of both. Formation of complex branching is not well-understood and can occur by modification of an existing PD or formation of new channels altogether. One reported mechanism suggests PD branching occurs in order to maintain cell-to-cell communication during wall modifications such as thickening (Ehlers and Kollmann, 1996). It is important to note PD can be degraded or walled off at any time by the plant, permanently closing the channel. However, there are several other temporary modifications that can be made to PD which affect its connectivity and functionality, which is known as plasmodesmal gating (Oparka et al., 1997; Knox and Benitez- Alfonso, 2014; Otero et al., 2016). The mechanisms that change PD function are briefly discussed below.

1.4.5 Mechanisms of Function The roles of PD are numerous and facilitate intercellular signaling, short and long-distance communication, coordination of developmental processes, and adaptation during changing conditions. The size exclusion limit (SEL) of these channels determines the maximum size of a molecule able to pass through and is controlled by cell wall components deposited around the opening. This area of wall is referred to as

the sphincter collar and can rapidly modify the SEL of PD by dilating or constricting its aperture (Knox and Benitez-Alfonso, 2014). The ability for molecules to pass through PD also depends on the molecules’ size and shape, along with the cell type and/or developmental stage of the tissue. For many macromolecules, transport is highly selective, with exchanges only occurring between certain cell types (Lucas and Lee, 2004; Zambryski, 2004). The regions of cell wall surrounding PD are described as micro-domains. The wall components and proteins that localize to this region have been studied for their effects on PD function (Simpson et al., 2009; Gutierrez et al., 2010; Knox and Benitez- Alfonso, 2014). Callose, a ß-1,3-glucan, is deposited by plants to manipulate PD aperture, changing the shape and function of the channels (Oparka et al., 1997; Radford et al., 1998; De Storme and Geelen, 2014). Callose is one of few well-studied mediators of PD, with several mutations and conditions affecting its presence around PD involved in modifying transport capacity (Simpson et al., 2009; De Storme and Geelen, 2014).

Callose has also been implicated in defense responses associated with infection and symbiotic interactions (Oa et al., 2006; Rigano et al., 2007; Hofmann et al., 2010). Similar to callose, multiple polysaccharides such as pectin, also localize to these wall domains greatly limiting our ability to quantify and study their roles in mediating PD function (Steele et al., 1997; Schubert et al., 203; Demchenko et al., 2014). Several studies of symplasmic transport in determinate nodules elude to PDs role in the nodule organ maintenance (Pate et al., 1969; Brown and Walsh, 1994; Schubert et al., 2011;

Courty et al., 2014).

1.5 Transport and Metabolism To and From Infected Cells Root nodules are one of few organs in higher plants which are highly active in metabolism and transport (Pate et al., 1969). They are a sink for carbon and source of nitrogen. A thorough review of the molecular aspects on this topic, including bacterial transport and metabolism, by Michael Udvardi and Philip Poole was published in 2013 and should be referred to for in-depth discussions of genetic components (Udvardi and Poole, 2013).

Previous findings show a legume nodule can turnover 3-10 times its own nitrogen content per day without significant accumulation within the organ (Bond, 1936; Pate, 1957). In tropical and subtropical legumes, nitrogen is assimilated into ureides, allantoin, and allantoic acid for export into the plant body. In temperate legumes, nitrogen is assimilated into amides or amino acids, as glutamine and asparagine (Sprent and Embrapa, 1980; Brown et al., 1995). It has been shown that uninfected cells in determinate nodules play a role in nitrogen assimilation and transport. Here, ureides are produced and moved through other uninfected cells via the symplast (Newcomb et al., 1985). In indeterminate nodules, the role of uninfected cells in transport of nitrogen containing molecules is unknown. It has been shown assimilated nitrogen transport into the vascular bundles occurs through transfer cells in the pericycle (Pate et al., 1969; Mylona et al., 1995). Transport of reduced carbon in the form of dicarboxylic acids such as malate to the nodule, which is likely the primary form of carbon provided to bacteroids, must equal that needed by the nodule. This is estimated to be the equivalent of 3-19mg of carbohydrates consumed or respired for each milligram of nitrogen fixed depending on the type of nodule (Bond, 1948; Gibson, 1966).

1.6 Supporting Literature and Research Approach Legume-rhizobia symbioses result in formation of a novel plant organ called a root nodule which houses nitrogen-fixing bacteria. Components necessary for the formation, maintenance, and persistence of this organ must be transported to and from the nodule with great efficiency. Mature nodules are considered sink organs and require large amounts of imported sugar derived from photosynthesis to support the endosymbionts. In addition, the plant effectively manages export of assimilated organic nitrogen compounds from the organ. In higher plants, carbon is transported from source organs to sinks via the phloem, mostly as the simple sugar, sucrose (Zimmermann MH, 1975). Unloading from the phloem and subsequent transport of carbohydrates occurs symplastically through PD or apoplastmically through sugar transporters in the plasma membrane (Patrick, 1997). Few studies have examined transport within mature indeterminate nodules formed during the L-R symbiosis, particularly symplasmic transport. Disparate results from studies in symbiotic nodules, suggest symplasmic connectivity influences the formation and function of the organs (Benitez-Alfonso, 2014). Assessment of symplasmic domains formed during nodule initiation in Medicago truncatula roots infected with rhizobia revealed transport between phloem and adjacent tissue is positively regulated, with unloading from the phloem occurring in the nodule meristematic region (Complainville et al., 2003). Expression of a viral movement protein, which dilates PD aperture thus improving permeability, significantly increased nodule formation, suggesting PD regulation is important in nodule development and/or function. Interestingly, formation, frequency, and complexity of PD changes drastically in uninfected tissue within symbiotic nodules and in the adjacent roots (Brown et al.,

1995; Schubert et al., 2011; Schubert et al., 2013), suggesting uninfected cells have a significant role in supporting symplasmic transport during nodulation. The transport capacity necessary for a functional mature indeterminate nodule formed during the L-R symbioses must involve phloem loading and unloading of materials into and out of the organ. Paths of transport to and from infected cells must also be established for a productive symbiotic relation to occur. The recognized importance of symplasmic connectivity for proper nodule formation and function suggests indeterminate nodules must establish such a pathway beyond nodule initiation to support the symbioses. This dissertation investigates the structural and physiological roles of nodule vasculature, uninfected tissue, and PD form and function in regards to symplasmic transport, particularly within the nodule cortex. The following studies provide insight into the symplasmic transport system throughout the developmental zones of mature Medicago truncatula and Pisum sativum root nodules, along with novel views of the organ and intricate relationship taking place within. An optical clearing technique was developed to allow imaging and characterization of the organ’s vasculature and uninfected tissue within intact root nodules. Plasmodesmata Quantification of the frequency and distribution of PD along the wall interfaces between adjacent cell types in the nodule cortex, within each developmental zone and across zones in mature indeterminate nodules. These results show increased connections between adjacent uninfected cells and between uninfected and infected cells. These connections increased in frequency and complexity through the developmental gradient. The connections between two adjacent infected cells did not follow the same trend of increased PD connections nor channel complexity, even

though large amounts of nitrogen are exported and sugar requirements are higher in these cell types. Interestingly the greatest increase in PD complexity was seen between infected-uninfected cell types and not adjacent uninfected cells. To determine if the uninfected cells were indeed active in nodule transport a fluorescent symplasmic tracer molecule was used to map the movement of water. Assessment of symplasmic flow using HPTS revealed an independent route from the nodule vasculature. HPTS fluorescence was observed within the parenchyma cell layer adjacent to the vascular bundles, within the vascular endodermal layer, and radiated into the central nodule tissue primarily through bands of uninfected cells. HPTS fluorescence was shown entering directly into each developmental zone and not exclusively into the nodule meristem. This suggests an alternative symplasmic route than through the vasculature and organ maturation involves changes to the symplasmic route established during nodule initiation. Taken together the above findings provide additional support and a novel view of indeterminate nodule vasculature structure. The utilization of our novel clearing technique provided a truly 3D representation of the multiple vascular bundles containing multiple points of branching which wrap around the periphery of the nodule. Observed increases in PD formation and morphology between the different cell types occurs in developmental zone II, suggesting “form” is established early as cells’ gain their unique function. Uninfected cells develop greater numbers of PD, suggests these cells serve the function of “support cells” as a unique and independent symplasmic transport route developed to maintain a persistent and functional nitrogen fixation zone.

Chapter 2

MATERIALS AND METHODS

2.1 Materials

2.1.1 Bacterial Strains Rhizobial strains Sinorhizobium meliloti 2011 (Meade et al., 1982) and wild type Rhizobium leguminosarum bv. viciae 3841 (Wood et al., 1989) were utilized to induce nodule formation in Medicago truncatula and Pisum sativum respectively. For long-term storage, bacterial strains were resuspended in 50% glycerol and stored at -20 ºC.

2.1.2 Plant Species Medicago truncatula (Jemalong A17) and/or Pisum sativum (Early Alaska Pea) were used as host plants in these studies.

2.1.3 Reagents and Enzymes Laboratory grade reagents were acquired from a variety of companies including Bio-Rad Laboratories (Hercules, CA), Electron Microscopy Sciences (Hatfield, PA), Fisher Scientific (Pittsburg, PA), and Sigma-Aldrich (Saint Louis, MO). Cell-wall digesting enzymes Cellulase-RS (#2019, Lot#3030) and Pectolyase-Y23 (#8006, Lot#5154) were obtained from Karlan Research

Products (Santa Rosa, CA). For experiments, enzymes were dissolved in PME buffer with 0.1% (w/v) bovine serum albumin fraction V (Sigma-Aldrich, St. Louis, MO).

2.1.4 Microscopy Stains and Dyes Diverse stains and dyes were used for confocal and electron microscopy studies Table 2.1.

Table 2.1: Stains and Dyes Utilized in Confocal Microscopy

Name Marker Excitation/Emissi Working Source & specificity on Wavelengths Concentratio Catalog n Number Calcofluor White cellulose 440/500-520 1:500 Sigma-Aldrich M2R 18909

SYTO 13 nucleic acids 488/509 1:1000 ThermoFisher S7575

SYTO 59 nucleic acids 622/645 1:10,000 ThermoFisher S11341

HPTS; pyranine (8- water- 355/465 10mM in ThermoFisher Hydroxypyrene- soluble sterile water H348 1,3,6-Trisulfonic membrane- Acid, Trisodium impermeant Salt) pH indicator

2.1.5 Standard Solutions All solutions were made using sterile, filtered RO water, unless otherwise noted. Working solutions were made fresh from concentrated stock solutions. Concentrated Stock Solutions:

1. 300 mM PIPES buffer; piperazine-N, N′-bis (2-ethanesulfonic acid) (Sigma P7643), pH 7.0.

2. 10X Tris-buffered saline (TBS); 250 mM Tris base, 1.36 M NaCl, 26.8 mM KCl, pH 7.0.

3. 1000X Fahraeus Nitrogen-free medium (FP) trace elements; 46.25 mM H3BO3, 9.1 mM MnSO4 4H2O, 0.769 mM ZnSO4 7H2O, 0.32 mM CuSO4 5H2O, 0.579 mM Na2MoO2 2H2O (Nutman, 1970; based on Fahraeus, 1957).

Working Solutions:

4. PME buffer; 80 mM PIPES, 5 mM MgSO4, and 10 mM EGTA (ethylene glycol-bis (β-aminoethyl ether)-N, N, N’, N’-tetraacetic acid), pH 7.0 (Harrison et al. 2009).

5. PHEM buffer; 60mM PIPES, 25 mM HEPES, 10 mM EGTA, 2 mM MgCl2, pH 6.8.

6. TBST; Tris-buffered saline with 0.05% (v/v) Tween®-20 (Fisher BP337-100).

7. Blocking buffer for immunolabeling; 3% bovine serum albumin fraction V (w/v), 1% normal goat serum (w/v), in TBS, filter sterilized through a 0.22 µm disc filter on a sterile disposable syringe.

8. FP, Fahraeus nitrogen-free medium; 0.68 mM CaCl2, 0.49 mM MgSO4, 0.77 mM KH2PO4, 0.84 mM Na2HPO4, 0.02 mM ferric citrate, 0.8 ml/L 1000x trace elements stock (Nutman 1970; based on Fahraeus, 1957).

9. FPA, Fahraeus nitrogen-free medium agar; 0.68 mM CaCl2, 0.49 mM MgSO4, 0.77 mM KH2PO4, 0.84 mM Na2HPO4, 0.02 mM

ferric citrate, 0.8 ml/L 1000x trace elements stock, 1% Bacto® agar (Nutman 1970; based on Fahraeus, 1957).

10. FP, Fahraeus medium with nitrogen; 0.68 mM CaCl2, 0.49 mM MgSO4, 0.77 mM KH2PO4, 0.84 mM Na2HPO4, 0.5 mM NH4NO3, 0.02 mM ferric citrate, 0.8 ml/L 1000x trace elements stock (Nutman 1970; based on Fahraeus, 1957).

11. TY, tryptone yeast broth; 5 g/L tryptone peptone, 3 g/L yeast extract, and 0.9 g/L CaCl2 2H2O (Vedam et al., 2003) autoclaved.

12. TYA, Tryptone yeast agar: 5 g/L tryptone peptone, 3 g/L yeast extract, and 0.9 g/L CaCl2 2H2O, 1.5% agar (w/v) (Vedam et al., 2003) autoclaved.

13. Lullien solution nitrogen free; 0.52 M K2SO4, 0.25 mM MgSO4, 50 µM Na2EDTA (Ethylenediaminetetraacetic acid), 30 µM H3BO3, 10 µM MnSO4, 0.7 µM ZnSO4, 0.2 µM CuSO4, 1 µM NaMoO4, 0.04 µM CoCl2, 1 mM CaCl2, 5.5 mM phosphate buffer, and 50 µM FeSO4 (Lullien et al., 1987).

14. Lullien solution with nitrogen; 0.52 M K2SO4, 0.25 mM MgSO4, 50 µM Na2EDTA (Ethylenediaminetetraacetic acid), 30 µM H3BO3, 10 µM MnSO4, 0.7 µM ZnSO4, 0.2 µM CuSO4, 1 µM NaMoO4, 0.04 µM CoCl2, 1 mM CaCl2, 5 mM NH4NO3, 5.5 mM phosphate buffer, and 50 µM FeSO4 (Lullien et al., 1987).

2.2 Methods

2.2.1 Bacterial Culture One loop of bacterial strains taken from a thawed glycerol stocks was plated and cultured on TYA in 10 mm Petri dishes, each containing approximately 20 ml TYA 28°C for three days. An individual bacterial colony was used to inoculate 10 ml of TY in a sterile, lidded 50 ml conical tube, and the cultures were grown at 28ºC at 260 rpm for approximately 24-30 h when the culture reached mid-to late log phase (108 cells/ml). Bacterial density was quantified in a spectrophotometer, using light absorbance at a wavelength of

600nm. Cultures were spun at 1800xg for 10min to pellet the bacterial cells and the supernatant was discarded. The pellet was suspended in 20 ml of sterile water and 10 ml was used to inoculate one caisson. This equates to 10 x 108 bacterial cells per caisson.

2.2.2 Pisum sativum Plant Growth

2.2.2.1 Seed Germination Early Alaska peas were surface sterilized for 30 seconds in ethanol, rinsed three times with sterile water, soaked for 3min in full strength Clorox® bleach and rinsed thoroughly with sterile water under sterile conditions until bleach smell dissipated. Seeds were transferred to sterile 1% water agar petri dishes with a thin film of water over the bottoms. The plates were sealed with Parafilm and placed in dark at room temperature for 48 to 72 hours to promote germination.

2.2.2.2 Pisum sativum Plant Growth Depending on the planned final use of tissues, P. sativum “Early Alaska” plants were grown using FP nutrient solution with one of three methods: 1) Aeroponic growth for fluorescent and TEM microscopy; 2) Flask growth for ScaleP studies; 3) Growth pouch for dye loading experiments.

Aeroponic Growth Seedlings were transferred into a “caisson” an aeroponic growth unit manufactured in-house at UD per established specs (Gallardo et al., 2006). Briefly, a dark colored 55-gallon polyethylene trash barrel was modified to accommodate a sealed Defensor 505S humidifier which generated a fine mist of

nutrient solution (Defensor, Switzerland) (http://www.medicago.org/documents/Protocols). To physically support plant growth, a 0.5 cm black plexiglass sheet was sealed onto the lid; the plexiglass was drilled with 0.5 cm holes to allow the roots to reach the aerosolized growth medium, support the germinated seedlings and to exclude light. The maximum volume of FP media defined by the

Defensor 505S humidifier fill line was added to the caisson and maintained throughout nodulation. Roots of germinated seedlings were threaded through the lid of the caisson unit. The plants were misted with sterile water, and then covered with a plastic dome to increase relative humidity before transferring unit into a Conviron® growth chamber on a 14h light, 22ºC /10h, dark cycle at 55% relative humidity. FP media without nitrogen was used one week prior to inoculation to establish nodulation.

Growth in Flasks Sowing of pea seeds in flasks used the above sterilization method however individual seeds were sown directly into 500mL autoclaved Erlenmeyer flasks containing 200 mL of FPA media without nitrogen or sterilized Hydroton® clay pebbles. Flasks were sealed with foam bungs and wrapped in black plastic bags before placing them in a dark cabinet for a three to five-day germination period. After which, etiolated pea hypocotyls were gently pulled out of the mouth of the flask, replacing the foam bungs. Seedlings were then inoculated. After which the foam bungs were replaced to support the shoot, with roots kept sterile inside flask. Black plastic bags were replaced and plants

were transferred to Conviron® growth chamber. The conditions of the Conviron® growth chamber for plant growth were maintained at 55% relative humidity with a 14 h day, 22ºC /10 h night cycle, 18ºC light cycle. Units were inoculated with wild type R. leguminosarum bv. viciae 3841 in log phase at 10 x108 bacterial cells per caisson, by addition directly to flask.

Growth in CYG™ Pouches To prepare growth pouches (CYG™, Newport, MN), a sheet of filter papers cut to size were inserted into back of pouches to maintain moisture and prevent contamination. Pouches were then wrapped in foil and autoclaved. Seedlings geminated on sterile water agar were transferred to pouches, with foil remaining to exclude light from root systems. Plants were kept moist and watered with sterile FP without nitrogen media alternating with sterile water. After two to three days, plants were inoculated with wild type R. leguminosarum bv. viciae 3841 in log phase growth at 1 x 108 bacterial cells per plant by direct application to the roots.

2.2.3 Medicago truncatula Plant Growth

2.2.3.1 Seed Germination M. truncatula seeds were extracted from pods by manually breaking pods between two corrugated rubber mats (fine grooved) attached to wooded support pieces. Ten to twenty pods were crushed by pressing down with a circular movement and seeds were removed from pod remnants and debris. Seeds were chemically scarified by treatment with sulfuric acid before seed imbibition. Seeds were soaked in just enough sulfuric acid to cover seeds

for 3 min with gentle agitation. Careful of seeds, H2SO4 was removed and seeds were rinsed thoroughly (6-10) times with large volumes of sterile water to prevent overheating of seeds. Seeds were covered with sterile water and imbibed at room temperature in light for 4 h on an orbital shaker (~120 rpm). To synchronize germination, seeds were vernalized at 4°C in the dark overnight. Subsequently, seeds were rinsed in sterile water with rotation (120 rpm) for 6 h at room temperature to remove inhibitory compounds, with water changes once every hour. Seeds were then transferred to moistened Petri dishes, sealed with parafilm and inverted. To promote germination seeds were transferred into a dark cabinet for approximately 48 h.

2.2.3.2 Plant Growth Seedlings were transferred to a caisson with 0.1 cm holes containing

0.5X Lullien nutrient solution with nitrogen for one week, then 0.5X Lullien nutrient solution without nitrogen for one with before inoculation. Concentration of media was maintained by adding one liter of 0.5X Lullien nutrient solution without nitrogen to every four liters of water, maintaining the water level line. After one week of nitrogen starvation caisson was inoculated with

Sinorhizobium meliloti 2011 (Meade et al., 1982)

2.2.3.3 Rhizobial Inoculation of Caisson A single colony of S. meliloti 2011 was picked from a freshly grown agar plate (7-day old TY plate grown from a glycerol stock) and used to inoculate 50 ml of TY broth. The culture was grown at 28ºC, 220 rpm for approximately 24- 30 H until the culture reached mid-log phase (approximately 108 cells/ml TY

media). For caisson inoculation, 20 ml of the overnight culture (OD600=0.7-1.5) was spun at 1800xg for 15 min to pellet the bacterial cells and the supernatant was discarded. The pellet was resuspended in 20 ml of sterile water and 10 ml of the resuspended culture was used to inoculate one caisson. This equates to approximately 10 x 108 bacterial cells per caisson.

2.2.4 Harvest of Root Nodules Nodules were harvested as intact organs or middle central tissues were excised using following method (See Figure A.1). Fourteen or twenty-one day post inoculation (dpi) nodules grown in flasks were hand sectioned using a double-edged razor blade or collected whole and prepared according to protocol (Blancaflor et al., 2001; Harrison et al., 2002). Nodule sections were fixed at 28°C with rotation overnight in 4% paraformaldehyde, 1% glutaraldehyde, and 80 mM PIPES buffer, pH 7. Samples were either used for ScaleP clearing experiments or embedded in resin (See sections 2.2.5 & 2.2.6).

2.2.5 Optical Clearing of Nodule Tissue

2.2.5.1 Sample Preparation and Infiltration with Clearing Reagent For optical clearing and confocal microscopy analysis, nodules in fixative were rinsed three times 5min durations with 80 mM PIPES buffer and placed in ScaleP, an optical clearing reagent containing: 6 M urea, 30% glycerol, and 0.1% Triton X-100 in 80 mM PIPES buffer modified for plant material (Warner et al., 2014) from a mammalian tissue technique (Hama et al.,

2011). Nodules cleared for two to four weeks at room temp with rotation before immunolabeling or fluorescent staining. Preparation of samples for uninfected cell network mapping used 14 dpi whole nodules harvested and fix as above, rinsed three times with H2O and placed in 10% KOH (w/v) for 12 hours at room temp with gentle agitation. Afterwards, nodules were thoroughly rinsed and placed in ScaleP clearing reagent for one to two weeks until optically clear. For multi-photon microscopy, nodules were stained with SYTO13® (Molecular Probes, Eugene, OR) nucleic acid stain, dilution of 1:1000, and Calcofluor® White staining cellulose, dilution 1:500, added directly to clearing solution. Samples covered with foil and stained for 2 h prior to imaging. Samples were placed in two-chambered Nunc™ Lab-Tek™ II Chamber slide system (Thermofisher, Wilmington, DE) or in double coverslip chambers approximately 0.8 mm deep with clearing solution. Chambers were sealed with quick dry clear nail polish and allowed to dry for at least 1 h prior to imaging. 30% glycerol for mounting media was used with multi-immersion lens with correction collar adjusted appropriately.

2.2.5.2 Data Acquisition for Confocal Microscopy Imaging related to optical clearing publication were acquired on an inverted Zeiss LSM 510 NLO laser scanning microscope (Carl Zeiss, Inc.,

Germany) using a Zeiss LD 25x C-Apochromat lens (NA 0.8). Nodules were visualized using 740 nm excitation with a Coherent Mira 900 titanium sapphire laser. Images were captured as single optical sections (2D) or as a tiled z-series of optical sections (3D).

Imaging of intact nodules was performed on a Zeiss 880 multiphoton microscope (NIH Grant #1 S10 OD016361 and Delaware INBRE Grant #P20 GM103446) using a Zeiss LD 25x C-Apochromat lens (NA 0.8). Three by three tiled z-series were acquired from one half of nodule, chamber flipped, sample allowed to settle, then imaged second half of nodule. Acquisition included 10% overlap of tiles and approximately 100 µm overlap of central tissue for merging both halves.

2.2.5.3 Data Processing for Three-Dimensional Reconstruction and Segmentation Stitching of tiled z-series were performed in Carl Zeiss software ZEN with 10% overlap. Merging of two z-series were performed in Figi (Schindelin et al., 2012) ImageJ (Schneider et al., 2012) using Pair-wise stitching plugin (Preibisch et al., 2009) before importing into Amira 6 3D software (Delaware

INBRE Grant #P20 GM103446).

2.2.6 Transmission Electron Microscopy- Plasmodesmata Quantification

2.2.6.1 Sample Preparation Early Alaska pea plants grown to 21dpi in caissons and harvested into 4% (w/v) paraformaldehyde, 1% (v/v) glutaraldehyde in 100 mM PIPES pH 7.0 using a double-edged razor blade and prepared according to a modified protocol

(Blancaflor et al., 2001; Harrison et al., 2002). Samples were fixed overnight at

4ºC with rotation. The following day, samples were post fixed with 1% OsO4 for 2 h, and dehydrated with a graded dilution series of acetone 10%, 30%, 50%, 70%, 90%, 100%, and 100%. Each dehydration step was performed for 30 min

at room temperature with rotation. Tissues were then embedded by a graded series in Epon 812-Araldite resin (Electron Microscopy Sciences, Hatfield, PA): 10%, 30%, 50%, 70%, 90%, 100%, and 100%, each step was performed for 30 min, except last step which infiltrated overnight, at room temperature with rotation. Samples were then transferred to aluminum weigh boats containing fresh Epon 812-Araldite resin and heat cured at 70°C for 48 h (Mollenhauer,

1964). Embedded nodules were cut with glass or diamond knives (Delaware Diamond Knife, Wilmington, DE) on a Reichert Ultracut E ultramicrotome. To obtain longitudinal sections through the middle of the nodule, sections were collected onto gelatin-coated glass slides, and counter stained with 0.1% (w/v) basic fuschin or 0.1% (w/v) basic fuschin and 1% toluidine blue in 1% NaBO2

4H2O. After correct orientation and depth of embedded nodules were assessed using a Zeiss Axioscope 2 microscope with a mounted Nikon digital camera, sample blocks were trimmed down and 90 nm thick sections were collected onto 200 mesh formvar coated nickel grids. Several sections were discarded between collections onto grids to ensure same PD were not counted twice. Grids were then post stained with 2% aqueous uranyl acetate and Reynolds’ lead citrate. Rinsed on 8 droplets of filter sterilized water and allowed to dry before imaging.

2.2.6.2 Data Acquisition Samples were imaged using a Zeiss Libra 120 Plus transmission electron microscope at 120 kV. Digital images were acquired with a Gatan Ultrascan 1000, 2k CCD camera. Files were saved as tiff for later analysis.

Low magnification images were used to map regions of interest and measure lengths of cell wall interfaces. Count data was logged into an excel spreadsheet before analysis in JMP Pro13.

2.2.6.3 Plasmodesmata Scoring for Quantification Acquired low magnification images of grid squares for quantification were used to map regions of interest and measure length of cell wall interfaces using ImageJ software. See appendix figure A.3 for outline and example of quantification method. Collected data compiled in Microsoft excel included the following information; developmental zone, length measurements of cell wall segments, adjacent cell types, and number of PD. Quantification criteria followed that outlined by Brown, 1995 and Gunning, 1976 (Gunning and Robards, 1976;

Brown et al., 1995). PD were scored using the following criteria. Only PD with longitudinal orientation which extended across the cell walls middle lamella were counted. PD with branching structure e.g. “H” or “Y” shaped were counted as complex. Clusters of PD were characterized by punctate cross sections in groups greater than two. Each punctate were counted individually for number of PD per cluster.

2.2.7 Transmission Electron Microscopy- Plasmodesmata Tomography To examine PD three-dimensional ultrastructure at the TEM level, a technique known as tilt-series tomography was utilized (Wagner et al., 2012). All acquired tomograms were obtained from epoxy embedded 14dpi pea nodules.

2.2.7.1 Sample Preparation Samples were fixed overnight at 4ºC with rotation. The following day, samples were post fixed with 1% OsO4 for 2 h, and dehydrated with a graded dilution series of acetone 10%, 30%, 50%, 70%, 90%, 100%, and 100%. Each dehydration step was performed for 30 min at room temperature with rotation. Tissues were then embedded by a graded series in Epon 812-Araldite resin

(Electron Microscopy Sciences, Hatfield, PA): 10%, 30%, 50%, 70%, 90%,

100%, and 100%, each step was performed for 30 min, except last step which infiltrated overnight, at room temperature with rotation. Samples were then transferred to aluminum weigh boats containing fresh Epon 812-Araldite resin and heat cured at 70°C for 48 h (Mollenhauer, 1964). Semi-thick sections (200nm) from epoxy embedded nodules were cut with a diamond knife (Delaware Diamond Knife, Wilmington, DE) on a

Reichert Ultracut E ultramicrotome. Before collection 200 mesh formvar- carbon-coated copper grids were floated, shiny side down, on a droplet of 1% (w/v) poly-l-lysine for 15 min, rinsed three times with distilled water and allowed to air dry. Grids were then floated onto droplets of 20 nm colloidal gold particles (1:10 dilution) in water for 10 min, rinsed three times and allowed to air dry. After collection of semi-thick sections, grids were counterstained with saturated methanolic uranyl acetate and Reynolds’ lead citrate and allowed to dry. Grids were floated, shiny side down, on a droplet of 1% (w/v) poly-l-lysine for 15 min, rinsed three times with distilled water and allowed to air dry. Grids were then floated onto droplets of 15 nm colloidal gold particles (1:10 dilution)

in water for 10 min, rinsed three times and allowed to dry completely before imaging.

2.2.7.2 Data Acquisition Dual axis tilt series, collected from +60° to −60° at 1° increments, were acquired in regions of interest containing PD simple in structure running perpendicular to the middle lamella. Regions of interest included combinations of adjacent cell types (I-I, I-U, and U-U) from developmental zones II and III. Images were collected at a magnification of 8000 with a pixel size of 1.4nm on a Zeiss Libra 120 Plus transmission electron microscope at 120 kV. Digital images were acquired with a Gatan Ultrascan 1000, 2k CCD camera.

2.2.7.3 Tomogram Reconstruction and 3D Modeling Utilizing the colloidal gold particles applied to the sections, the tilt series was reconstructed using a R‐ weighted back projection in IMOD 4.1 (Boulder Lab. for 3‐ D Electron Microscopy of Cells) (Kremer et al., 1996). The tilt series was examined with the same software. Two representative areas of interest were selected for video analysis and computer modeling and converted to TIFF image formats. Two TIFF stacks of reconstructed tomograms were converted to a 3D data set (voxelated) before importing into Amira 6 3D software (Delaware

INBRE Grant #P20 GM103446) (Visage Imaging Inc., San Diego, CA, USA). Segmentation was performed with Amira before final presentation in 3D space was generated.

2.2.8 Mapping of Uninfected Cell Arrays and Vasculature Tissue 14dpi nodules were harvested whole into 4% paraformaldehyde, in 80 mM PIPES and placed in 4°C with rotation overnight. Samples were rinsed 3x with PHEM buffer before clearing with ScaleP for 1-2 weeks. After nodules become transparent, those approximately 800 µm thick were stained with 1:500 Calcofluor White and 1:1000 Syto13 directly into clearing solution, subsequently rinsed three times. Samples were transferred into fresh clearing solution before sealing in a double coverslip chamber with clear nail polish. Whole nodules were imaged one side at a time using an LSM880 multiphoton confocal microscope (NIH Grant #1 S10 OD016361 and Delaware INBRE Grant #P20 GM103446). Images of nodules were acquired as tiled z-stacks. Tiles were stitched together with a 10% overlay and the two sides combined using ImageJ software plugin pair-wise stitching (Preibisch et al., 2009).

Combined z-stacks were loaded into Amira© 3D modeling software on workstation (Delaware INBRE Grant #P20 GM103446) for segmentation of uninfected cell networks and vasculature tissue. Segmentation of combined z- stacks were done by hand in Amira©.

2.2.9 Dye Loading Technique To assess the symplasmic continuity within nodule tissue hydroxy- pyranine-tri-sulphonate (HPTS) was used to map cell connectivity within mature pea nodules. Pea plants between 14 and 21dpi grown in pouches had tips removed from lateral root below healthy nodules. Roots were placed in 10 mM solution of HPTS in water for approximately 3 h. Samples were thoroughly rinsed with PHEM buffer before fixation with 4% (w/v) paraformaldehyde, 2%

(v/v) glutaraldehyde in PHEM buffer for 24 hours. Fixed samples were rinsed three times with PHEM buffer and placed in fresh 0.2 M glycine for 8 hours to quench any autofluorescence from fixation. Samples were rinsed three times with PHEM buffer and stained with 1 µM of Syto59, red fluorescent nucleic acid stain before imaging. Samples were imaged on LSM 880 multiphoton confocal microscope (NIH Grant #1 S10 OD016361 and Delaware INBRE

Grant #P20 GM103446) with a 25x/0.8 LD LCI PlanApochromat multi- immersion lens mounted with 30% (v/v) glycerol. Spectral imaging coupled with image analysis using linear un-mixing was used to segregate mixed fluorescent signals arising from artifacts because of natural autofluorescence and use of fixatives. Dye loading was repeated on four separate occasions. Each time 15 to 20 nodules were loaded. Out of 73 loaded nodules, 56 were imaged. Out of the

56 nodules imaged 37 nodules contained dye.

SYMPLASMIC FLOW IN CENTRAL NODULE TISSUE

3.1 Background Nitrogen-fixing root nodules are unique plant organs in regard to transport. Legumes establish nodules as an efficient way of obtaining nitrogen when insufficient amounts are present in the surrounding soil. Establishment of transport to this novel organ utilizes both long distance transport through the vasculature and short cell-to-cell transport via the apoplast or symplast (Pate et al., 1969). In higher plants, the vasculature is a network of specialized tissues that supports the long-distance transport of water, minerals, and other compounds through the xylem and photosynthates through the phloem (Hartig, 1837). The xylem transports material via capillary action from the root to the shoot through vessel elements and tracheid, which are dead at maturity. The phloem consists of multiple specialized cell types carrying photosynthates and other molecules from the source (photosynthetic tissues, e.g., leaves) to sink tissues, such as root nodules (Zimmermann, M.H., Milburn, J.A. and Eschrich,

1975; van Bel, 1993). This process is referred to as translocation. Once material leaves the phloem or xylem it is transported locally from cell to cell within the apoplast or symplast.

The apoplast refers to the space outside the plasma membrane, including the extracellular matrix and expanded middle lamellae. In roots, solutes move through this space by diffusion until it reaches a Casparian strip. Casparian strips are found in endodermal cells and are comprised of suberin and sometimes lignin, which prevents further translocation of materials through the extracellular matrix (Oparka and Cruz, 2000). Once solutes in the apoplast reach the

Casparian strip they must cross the plasma membrane to enter the symplast. The symplast describes the interconnected cytoplasm of adjacent cells. The flow of this pathway is facilitated and largely controlled by channels crossing cell walls, called plasmodesma (singular) or plasmodesmata (plural) abbreviated as PD, allow for small molecules such as signaling molecules, sugars, or amino acids to flow directly from cell to cell (Esau and Thorsch, 1985; Roberts and Oparka, 2003). This creates a cytoplasmic continuity between cells within a tissue system and facilitates compartmentalization for a multitude of processes. Nitrogen-fixing root nodules formed during the rhizobia-legume symbiosis are active in metabolism and transport (Pate et al., 1969). Nodule development begins in the root pericycle and cortical cells. Cortical cells de- differentiate and start the formation of the nodule primordium which eventually forms zone I, the nodule meristem. The nodule meristem contains actively dividing and growing cells, which require substantial amounts of nutrients transported into the newly formed root organ (Łotocka et al., 2012; Xiao et al., 2014). In determinate nodules, this transport requirement is only necessary for the first few days of organogenesis until the meristem is no longer active. In

indeterminate nodules, these provisions must be provided continually since the meristematic region remains active throughout nodulation. As the symbiosis progresses, two cell types in the nodule central tissue emerge, those that remain uninfected and those that become infected (Gadd, 1888). Those now infected cells must accommodate the new symbiont within its cytoplasm, which are actively dividing and differentiating. This stage of development is referred to as the infection or invasion zone II. Large amounts of carbon, mainly sugar, are transported into zone II to support the rhizobial population and provide the material necessary for host cells continued growth. The next developmental stage of these cells is referred to as the intermediate zone II/III. Comprised of these newly infected cells and those that remained uninfected, zone II/III marks the point when the rhizobia differentiate into forms capable of nitrogen fixation. Bacteroids lose their ability to metabolize the sugar provided by the plant and the excess is stored within these infected and uninfected cells in the form of starch (Walsh, 1990). The final developmental stage of these cells before senescence is the nitrogen fixation zone III. Within zone III, infected cells’ cytoplasms are filled with thousands of nitrogen-fixing bacteroids, to which the plant provides a continuous supply of sugar in exchange for the bioavailable form of nitrogen.

The enzymatic conversion of N2 into ammonium is an energy intensive process, and this energy is derived photosynthetically. As a result, there are significant fluxes of bioavailable nitrogen moving out of this zone into the plant body and of sucrose from the plant body into the nodule to support nitrogen fixation

(MacRobbie, 1971; Brown et al., 1995).

All cells within determinate nodules are developmentally the same since the organ does not maintain an active meristem like indeterminate nodules. Within an indeterminate nodule, the transport requirements of each developmental zone described above must be maintained for the remainder of the organs life. Researchers have explored various aspects of nitrogen and sucrose transport in both nodule types together, with frequent cross comparisons

(Guinel, 2009b; Schubert et al., 2011; Terpolilli et al., 2012). Though disparate at times, these comparisons have led to major advances in our understanding of beneficial plant-microbe interactions and biological nitrogen fixation. With that in mind, the similarities and differences in the biological processes of nitrogen and sugar transport in both nodule types are discussed below.

3.1.1 Nitrogen transport in determinate and indeterminate nodules Early studies estimated that nodules have the capacity to chemically reduce 3-10 times its own nitrogen content per day, averaging daily fixation rates between 30-100 mg N/g fresh weight of nodules (Pate, 1958). In determinate nodules, it is estimated materials synthesized from nitrogen fixation must travel anywhere from 0.3-1.3mm before reaching the nodule’s vascular tissue (Allen K et al., 1955; Pate et al., 1969). In indeterminate nodules such as those formed by Medicago truncatula and Pisum sativum, the fixed nitrogen is incorporated into amide-containing molecules or the amino acids asparagine and glutamine and mobilized from the nodule. In most cases, there is little or no accumulation of nitrogen in the nodule (Bond, 1936; Pate, 1962), suggesting that the chemically-reduced nitrogen is rapidly assimilated into the above nitrogen- containing molecules and transported out of the nodule into the host plant.

3.1.2 Sucrose transport and tissue organization in determinate and indeterminate nodules Organization of the nodule tissue must also facilitate rapid transport and metabolism of sucrose to the infected cells. This is true for both determinate and indeterminate nodules, with both types evolving unique ways to accommodate differences in tissue differentiation. In determinate nodules, sucrose and water enter the nodule through the phloem and unloads into the adjacent pericycle.

Here both the symplasmic and apoplasmic transport pathways are utilized to access the infected cells through the inner cortex (Walsh et al., 1989). In determinate nodules, soluble sugars are shown to reach infected cells through the apoplast or through the symplast via infected or uninfected cells, where a higher frequency of PD were found between adjacent uninfected cells (Brown et al., 1995). It is likely since infected cells require a way to receive and deliver the bulk of nutrients in exchange for nitrogen that the nodule would need an adequate pathway to support the influxes as a sink organ. In determinate nodules this role is supported by arrays of uninfected cells where every infected cell is in direct contact with an uninfected cell (Newcomb et al., 1985; Selker, 1988). The role of uninfected cells in indeterminate nodules is poorly understood. Though they are implicated in facilitating transport associated with nitrogen fixation.

3.1.3 Other aspects of transport involved in nodulation Another noteworthy aspect of transport lies at the interface of the bacteroids and the surrounding plant cell cytosol, organelles, and intercellular structures. This interface is maintained by the symbiosome membrane (Catalano, 2005; Maunoury et al., 2010). Requirements for setting up the symbiont within the hosts cytoplasm and further interaction between the two

partners is established and maintained by the symbiosome membrane (Udvardi and Day, 1997). The symbiosome membrane which is plant derived, surrounds the bacteroids within an infected cells cytoplasm acting as the main mediator between the host and the symbiont. All exchanges and communication must occur across this membrane which regulates the transport of metabolites to and from the plant (Udvardi and Poole, 2013). The symbiosome membrane is essential for an effective relationship and its degradation is linked to early senescence of the organ (Werner et al., 1985). In addition, proliferation of the symbiosome membrane within the infected cells requires massive synthesis of lipids and proteins (Verma, 1992; Verma et al., 1995). The primary metabolic exchange between the plant and bacteroids at this interface is reduced carbon for reduced nitrogen. Studies of this interface and the exchanges occurring constitute one of the most well researched aspects of nodulation to date with several comprehensive reviews available on this topic (Udvardi and Day, 1997; Terpolilli et al., 2012; Udvardi and Poole, 2013; Coba et al., 2018).

3.1.4 The role of PD in symplasmic transport Coordination of organ formation and tissue patterning is based on the ability of cells to communicate with each other. In plants, this coordination occurs in part symplastically (Zambryski, 2004; Otero et al., 2016) through PD. Plasmodesmata channels, which traverse the cell wall connecting the cytoplasm and endoplasmic reticulum of two adjacent cells are responsible for transport of an array of solutes. When needed, cell networks can create symplasmic micro- domains controlling intercellular transport by changing the frequency and distribution of PD, along with the size exclusion limit (SEL) of these channels to

facilitate tissue specific developmental processes and transport needs (Knox and Benitez-Alfonso, 2014). Callose a β 1, 3, glucan polysaccharide, plays a significant role in many plant developmental processes including the regulation of PD SEL (Kohle et al., 1985; Brown, R.C. and Lemmon, 2009; De Storme and Geelen, 2014). Most notably, it was identified as having an integral role in scaffolding of the cell plate during cytokinesis (Brown, R.C. and Lemmon, 2009). At this stage, walls are structurally bare and lack the rigid structure provided by interlinked cellulose, callose fills that role until it is eventually replaced with cellulose and other wall components. However, at the site of PD along these walls, callose remains undegraded (Brown, R.C. and Lemmon, 2009) and forms a sheath around the collar of the channel creating a sphincter-like structure (Olesen, 1979; Olesen and Robards,1990). Callose degradation or deposition at this region in the wall participates in the regulation of PD opening and closure (Wolf et al., 1991; Sivaguru et al., 2000; Radford and White, 2001; Heinlein and Epel, 2004). The diameter of PD, which determines the size of molecule able to pass through, is referred to as the size exclusion limit (SEL) of PD (Levy and Epel, 2009). Studies of PD frequency and distribution in determinate nodules have provided some insight into symplasmic transport (Robards and Lucas, 1990;

Brown and Walsh, 1994). Previous findings in determinate nodules of soybean estimated that uninfected cells, which are significantly smaller than infected cells, contain 2.5 times more PD (Brown et al., 1995). These results suggested uninfected cells play a greater role in symplasmic transport during plant to

nodule nutrient exchange. Other studies are of early time points in nodule development, before nitrogen fixation is at full capacity and PD quantification only included counts from a few cells or samples, giving little weight to their statistical analysis (Brown et al., 1995; Abd-Alla et al., 2000). One study of young nodules suggested the organ is symplastically isolated from the root (Complainville et al., 2003). They concluded that transport of sucrose from the phloem likely enters through the nodule vasculature and is unloaded into the nodule meristem before passing into zone II. Interestingly, the results from this study only showed transport up to the intermediate zone II/III in nodules without an established nitrogen fixation zone. One of two explanations are possible; one the experiment was terminated too early, not allotting enough time for transport to occur through this intermediate zone. Two, symplasmic transport is indeed impeded by gating of transport through PD within zone II/III. Though Complainville’s work shows establishment of symplasmic continuity during nodule primordia initiation within the root, bulk transport is not yet occurring and sink requirements equivalent to that of a mature nitrogen-fixing organ have not been established.

3.1.5 Transport and Tissue Organization in the Indeterminate Nodule The organization of an indeterminate nodules central tissue must facilitate rapid transport and metabolism at different times during the organs development. For instance, all cells are derived from the nodule apical meristem, as newer cells differentiate, those already present are maturing either into infected cells or remain uninfected and serve as “support cells” (Selker, 1988;

Udvardi and Poole, 2013). Transport requirements of cells in both developmental stages must then be supported by the plant. In initial stages of nodule development before the nitrogen fixation zone III is established, solutes up to 27kD are transported directly from the nodule vasculature throughout the apical meristem before entering zone II and II/III (Brown et al., 1995). Of particular interest though, is what happens in zone III of the central tissue. This zone is established by 7dpi at which point bacteroids within infected cells are actively fixing nitrogen. The possible routes of symplasmic transport within zone III has yet to be elucidated. This study focuses on cell-to-cell transport pathways through the symplast in established nitrogen-fixing root nodules of the indeterminate type. Using correlative microscopy techniques, the possible routes for transport of water soluble molecules less than 520 D within the central tissue were assessed in indeterminate nodules of Pisum sativum. The differences in connectivity between nodule cell types were evaluated by quantification of the functional units of the symplast, the PD. PD distribution were first assessed with immunolabeling of callose, which is found at the neck region of the channel and is commonly used as a marker for PD, then further quantified at the TEM level with wall walking and counting of the channels. Just because PD are present does not indicate the channel is functional and active in a transport pathway. To determine if the results showing differences in PD distribution between cell types, in particular uninfected cells, are functional and correlate with a symplasmic pathway, a small phloem-mobile fluorescent tracer was used to map flow into the central tissue of the organ.

HPTS (8-hydroxypyrene 1,3,6 trisulfonic acid; 520 D) is a fluorescent probe commonly used as a phloem-mobile, symplasmic tracer (Wright and Oparka, 1996). It has been used in studies of plant development (Zambryski, 2004) and as an indicator of PD function involved in transport of macromolecules and signaling (Heinlein and Epel, 2004; Schubert et al., 2013; Knoblauch et al., 2015). HPTS-acetate loaded into the xylem of the root is subsequently off- loaded into the phloem where it enters the nodule vasculature. When HPTS enters a symplasmic space, it is cleaved by esterases and subsequently starts fluorescing. HPTS is now retained in the symplast and can only move through PD. The retention of HPTS in the symplast as a fluorescent molecule, its relatively small size, and the fact its water soluble make it an ideal tracer for investigating PD activity and pathways of transport through the symplast.

Figure 3.1 Structure of the HPTS Acetate used in Dye Loading Experiments (Knoblauch et al., 2015).

3.2 Results

3.2.1 Nodule vasculature is branched and peripheral Due to current microscope limitations, imaging of whole nodules all the way through was not possible. To map the vasculature and central tissue in

ScaleP cleared nodules, two z-stacks were acquired, one from each side and merged together. The vasculature connecting the root to an established nitrogen fixing nodule does not enter at a central point, rather is seen as two distinct bundles (Figure 3.2 A). The two bundles then wrap around the periphery of the nodule on opposite sides adjacent to nitrogen fixing zone III (Figure 3.2 D). Branching of the two bundles are shown in Figure 3.2 B &C. The complete vasculature of the mature pea nodule forms a three-dimensional oblong spherical network lying only a few cell layers in the periphery of the organ (Figure 3.2 E).

Figure 3.2 Nodule vasculature branches radially at zone II and the apical meristem. Two independent vascular bundles (magenta) enter the nodule from the root (A) and proceed around the periphery of zone III where nitrogen fixation is occurring (D). The bundles branch (B) with uniform distribution around the nodule, then again at the apical meristem (C) with six final bundles observed. The overall vascular network resembles that of a cradle with the central tissue within (E). A, B, & C scale bar =250µm. D & E scale bar = 500µm.

3.2.2 Patterns of uninfected cells are present from bundles to central tissue Cells adjacent to the nodule vasculature into the developmental zones II, II/III, and III were easily observed in ScaleP cleared nodule stacks. Representative bands of uninfected tissue in zones II and III can be observed in Figure 3.3 B, C. Clear bands of uninfected cells were observed from the uninfected periphery radiating towards the central tissue where uninfected tissue appeared denser and disorganized (Figure 3.3 A, arrowheads). Difficult to capture in two-dimensional views are subtler cell distribution characteristics which are noticeable with repeated viewings of three-dimensional z-stacks of intact nodules. Uninfected tissue seemed to narrow, with cells shapes becoming more oblong particularly further into the nodule center of the nodule (Figure 3.3 A, large arrows), while ovular infected cells were observed throughout the tissue (Figure 3.3 A, arrowheads) this observation was not quantified in the study. Attempts to map uninfected bands proved challenging even with our current technology. Segmentation of uninfected tissue revealed elaborate organization with rays sometimes spanning several cell layers running perpendicular to the root (Figure 3.3 A, large arrows).

Figure 3.3 Uninfected cell arrays link periphery to central cortex. An optical slice representative of an average 14dpi Pisum Sativum root nodule taken from the central tissue (A) Scale bar = 250µm. All developmental zones are present, the apical meristem (M) Followed by zone II and zone III. Cell walls are stained with Calcofluor White (Blue) and nucleic acid stain Syto 13 (Green). Large white arrows mark the first and last cell observed in the longest band of uninfected cells (grey). Small arrowheads mark a few uninfected cell arrays amongst the central tissue where uninfected cell density appears the greatest. White boxes outline areas presented in A & B where depth is added showing uninfected bands from representative zones. Red arrows indicate rotation from orientation of A.

3.2.3 Callose antibody localized primarily to uninfected cell wall interfaces To evaluate the distribution of PD in mature pea nodules, 21dpi nodules were probed with an anti-(l-3)-β-glucan specific monoclonal antibody, which is a commonly used marker for callose associated with PD (Oparka et al., 1997; Dahiya and Brewin, 2000). To optimize visualization of immunolabeling,

ScaleP cleared nodules were used in addition to resin embedded. This clearing technique provided novel views of the callose antigen within cell walls of central nodule tissue. To show the advantages gained using the ScaleP clearing method (Warner et al., 2014) versus the commonly used technique for immunohistochemistry where resin embedded samples are sectioned and subsequently surface labelled, both techniques are included for comparison

(Figure 3.4). The results from both labeling methods showed the same distribution patterns of callose antigen, confirming ScaleP’s versatility. However, the three-dimensional view obtained by clearing provided insight into callose antigen distribution within entire cells and along cell networks (Figure

3.4, F). Callose antigen was present along uninfected cell wall interfaces with lesser amounts labeled along infected cell wall interfaces. This pattern was observed in all zones and no discernable differences across zones was discernable (Figure 3.4 B-F).

Figure 3.4 Callose antibody localized along uninfected cells. (A-C) Resin embedded surface labelling of anti-callose (red & magenta) compared to labeling of ScaleP cleared samples (D-F). Scale bar =200 µm (A), Scale bar= 50 µm (B-F). Syto13 nucleic acid stain (Green), Calcofluor White stain for cell wall (Blue). Clearing allowed labelling of entire cells (F) providing a more representative view of distribution along cell networks (D & E).

The callose antibody labelling along uninfected cell walls suggests the presence of PD, confocal microscopy lacks the resolution to distinguish individual PD. Therefore, this could be a single callose-rich PD, or multiple PD clustered together. To discern any differences between the zones, the total number of PD by cell type, zone, and cell types within zones were quantified using transmission electron microscopy (Figure 3.5).

Previous quantification of PD is limited, particularly PD quantification between known cell types in legume nodules (Abd-Alla et al., 2000; Complainville et al., 2003; Schubert et al., 2013). To the best of our knowledge, this is the most comprehensive analysis of PD in legume root nodules to date. This study included extensive sampling for quantification of PD, three biological replicates were used to collect 12 technical replicates per zone. The number and total length of cell wall interfaces and biological replicates analyzed far exceeded previously reported counts. This study also took into consideration PD orientation, density, and complexity. The collection of data followed stringent parameters for inclusion of PD. For instance, only PD which was seen from one cells cytoplasm crossing the wall at least meeting the middle lamella was counted. Parameters to standardize the collection of PD count data followed strict requirements and are detailed in the methods sections of this document (see S2.2 and Methods section on PD quantification). The lengths of cell wall were measured from low magnification images (S2.2 C) and PD (S2.2 B) were counted along the wall interfaces between uninfected cells, infected cells, and adjacent infected and uninfected cells at high magnification (S2.2 D, E, F).

However, depending on the cut through the tissue, two different views of PD are observed. Longitudinal views of PD are those that crossed the middle lamella from one cells cytoplasm and cross-sectional views of PD, seen as punctate spots between the two walls. Each provide unique insight into PD and were quantified separately. Measurements and PD count totals are summarized in appendix Figure A.2.

Previous studies report the central nodule tissue is comprised of approximately 20% uninfected tissue (Selker, 1988; Davidson and Newcomb, 2001). This was evident during our quantification of PD and every effort was made to measure a statistically significant amount of uninfected wall interfaces. Even so, adjacent uninfected cells were not prevalent and the total length measured was less than other adjacent cell wall interfaces (Figure A.2). Data transformations are an important tool for the proper statistical analysis of biological data sets and it is suggested to use those transformations common to the field of study (McDonald, 2009). Ecological count data are often log-transformed to satisfy parametric test assumptions, but many such data sets contain zeros and create bias when removed or replaced by adding a value in its place (O’Hara and Kotze, 2010). After several sessions with a statistician, Dr. Tom Illvento from the Department of Applied Economics and Statistics at the University of Delaware, it was decided that data presented without statistical log-transformations would be most representative of the true numbers found in our biological data set.

3.2.4 Uninfected cell wall interfaces contained more PD channels The average number of total PD running longitudinal for each combination of adjacent cell types’ wall interfaces is presented in Figure 3.5 along with the total number of PD quantified within each zone. Overall, uninfected cells adjacent to another uninfected or an infected cell have more PD per µm. Adjacent uninfected cells wall interface average about 3 PD per 10 µm compared to less than 1 PD per 10µm found along adjacent infected cell wall interfaces (Figure 3.5, A). Regardless of cell type, each zone seemed to have approximately the same number of PD, with only a slight increase occurring from zone II to zone III.

Figure 3.5 Distribution changes between cell types but not developmental zones. PD distribution per 10 µm between cell wall interfaces (top) adjacent to representative micrograph, I= infected, U=uninfected. PD distribution per 10 µm within each developmental zone (bottom) adjacent to representation of zone distribution in indeterminate root nodule.

3.2.5 PD distribution within each zone differed depending on wall interface Within each zone, the number of PD between the different cell types’ interfaces were statistically different. Figure 3.6 A, shows an example of the targeted location for zone II relative to the root. The data sets were obtained from at least three different nodules with twelve different slices from the central tissue. Figure 3.6, B is a representative slice of the invasion zone II. Figure 3.6

C & D are examples of PD counted from this zone. The longitudinal view shows the orientation of the channels which extends passed the middle lamella and lies perpendicular to the wall. In zone II, adjacent uninfected cell wall interfaces had a significant amount of connections (Figure 3.6, graph), with an average of 2 PD per 10 µm, where the wall interface between adjacent infected cells had just over 1 PD per 10 µm. Wall interfaces of infected cells next to uninfected cells were more connected than when infected cells are adjacent to on other.

Figure 3.6 Differences in PD connections between cell types start in zone II. Within invasion zone II (A, inlay) location of PD quantification between cell wall interfaces. Uninfected cells have more PD connections than wall interface between adjacent infected cells (graph). B is a representative area where quantification took place and C & D are examples of PD quantified. Scale Bar=500 µm (A), Scale Bar=10 µm (B), and Scale bar=0.2 µm (C&D). (I) infected, (U) uninfected cells.

The intermediate zone II/III is comprised of infected cells filled with dividing and differentiating bacteroids that rely solely on the host plant at this point in development. Starch, characteristic of this zone, is seen in both infected and uninfected cells. The quantification of PD by cell types within the intermediate zone II/III is presented in Figure 3.7. Figure 3.7 A, shows an example of the targeted location for zone II/III PD quantification relative to the root. The data sets were obtained from at least three different nodules with twelve different slices from the central tissue. Figure 3.7, B is a representative slice of the intermediate zone II/III where “bubble like” starch granules are visible. Figure 3.7 C & D are examples of PD quantified from this zone. The longitudinal view shows the orientation of the channels which extends passed the middle lamella and lies perpendicular to the wall. In zone II/III, adjacent uninfected cell wall interfaces had significantly more connections than the other two types (Figure 3.7, graph). Wall interfaces of adjacent infected cells had an average of 1 PD per 10 µm but when an infected cell is adjacent to an uninfected cell the number of connections at the wall interface increased to more than 2 PD per 10 µm. Wall interfaces between adjacent uninfected cells were found to contain the most cell to cell connections in this zone, with an average of 3 PD per 10 µm of adjoined wall.

Figure 3.7 Differences in PD distribution between cell types in zone II/II. Within intermediate zone II/III (A, inlay) shows location of PD quantification between cell wall interfaces. Adjacent uninfected cells have more PD connections than wall interfaces between adjacent infected cells or infected-uninfected wall interfaces (graph). B is a representative area where quantification took place and C & D are examples of PD quantified. Scale Bar=500 µm (A), Scale Bar=10 µm (B), and Scale bar=0.2 µm (C&D). (I) infected, (U) uninfected cells.

Zone III consists of large infected cells with thousands of nitrogen-fixing bacteroids with the cytoplasm. Nutrient demand is high for these cells and they must efficiently remove the newly acquired bioavailable nitrogen and exchange it for sugar. Less is known about the role uninfected cells have in this exchange, except that ureide synthesis in determinate nodules occurs within the cytoplasm of uninfected cells. The quantification of PD between cell types in zone III is presented in Figure 3.8. Figure 3.8 A, shows an example of the targeted location for zone III PD quantification relative to the root. The data sets were obtained from at least three different nodules with twelve different slices from the central tissue. Figure 3.8, B is a representative slice of the nitrogen fixation zone III where infected cells filled with bacteroids are much larger than uninfected cells. Figure 3.8 C & D are examples of PD quantified from this zone. The longitudinal view shows the orientation of the channels which extends passed the middle lamella and lies perpendicular to the wall. In zone III, adjacent uninfected cell wall interfaces had significantly more connections than the other two types (Figure 3.6, graph). Adjacent infected cells had a relatively sparse number of connections per micron of wall interface, averaging 0.5 PD every 10 µm. In contrast, infected cells with a wall interface neighboring an uninfected cell, had three times as many PD connecting them. Wall interfaces between adjacent uninfected cells had significantly higher numbers of PD, with

3.5 PD per 10 µm and contained the most PD on average in this zone. This is a sevenfold difference in connections compared to adjacent infected cells in zone III.

Figure 3.8 Differences in PD distribution between cell types within the nitrogen fixation zone III. Within zone III, (A, inlay) shows location of PD quantification between cell wall interfaces. Adjacent uninfected cells have more PD connections than wall interfaces between adjacent infected cells or infected-uninfected wall interfaces (graph). B is a representative area where quantification took place and C & D are examples of PD quantified. Scale Bar=500 µm (A), Scale Bar=10 µm (B), and Scale bar=0.2 µm (C&D). (I) infected, (U) uninfected cells.

3.2.6 PD Density increased along uninfected cell walls through development In indeterminate nodules, developmental zones have specific biological functions that have specific transport requirements. Within each zone, infected and uninfected cells contained observable differences in the number of PD connections depending on the adjacent cell type (Figures 3.6-8), implicating uninfected cells as “support cells” capable of higher rates of symplasmic transport. Indeterminate nodules provide a rare opportunity to study all physiological stages of cell development within the same longitudinal slice of the organ. To determine if the observed changes in PD distribution within each zone correlate with changes throughout the developmental gradient, PD quantified in cell types for each zone were compared to the other zones. The results would elucidate if a developmentally linked change occurs in different cells symplasmic transport capabilities. The number of PD connecting adjacent uninfected cells significantly increased from zone II to zone II/III (Figure 3.9). By zone III, the number of connections had increased from 2 up to 3.5 PD per 10 µm of wall interface. Though, between adjacent infected cells the average went down through development. It started with an average of 1 PD per 10 µm in newly infected cells but dropped to 0.5 PD per 10 µm in cells who had reached full maturity in zone III. Interestingly, the frequency of PD between infected and uninfected cells remained relatively the same throughout the developmental zones, with an average of 2 PD per 10 µm of wall and with a slight increase by zone III.

Figure 3.9 Differences in PD frequencies across development. Average number of PD found perpendicular to the middle lamella by cell type and developmental zone. Image inlay shows three simple PD and one branched PD for a total of four PD.

3.2.7 PD along uninfected cell walls become more complex in structure In higher plants, PD formation and further structural modifications such as branching. have been linked to cellular differentiation and specialization (Ehlers and Kollmann, 1996; Ehlers and Kollmann, 2001; Fitzgibbon et al., 2013). In addition, accommodation of increased symplasmic flow is commonly

accompanied by PD increased PD complexity, particularly during establishment of sink-source (Fitzgibbon et al., 2013; Schulz, 2014). To determine if and when PD become more complex in structure during nodule development, analysis of PD structure was factored into the quantification methods. Longitudinally viewed PD were reclassified into simple or complex. Those seen as a single channel were considered simple and those with an X or Y shaped structure were considered branched or complex. Within zone II, already a significant difference was observed in the number of complex PD between cell types (Figure 3.10). Wall interfaces of adjacent uninfected cells had more than four times the number of complex PD on average compared to adjacent infected cells and infected-uninfected cell wall interfaces had almost three times the amount on average. As the cells develop into zone II/III the number of complex PD drastically increased along wall interfaces between uninfected cells, with at least one branched PD observed every 10 µm by zone III when compared to 1 every 50 µm in zone II. The average number between adjacent uninfected and infected cells plateauing by zone II/III, remaining at 0.3 branched PD every 10 µm into zone III. No significant differences were found between the adjacent infected cells throughout the developmental zones.

Figure 3.10 The wall interface between adjacent uninfected cells have more complex PD. The average number of PD that have branching structures per 10 µm of cell wall interfaces.

Complex PD were then assessed as a percentage of the total number of longitudinal PD observed (Figure 3.11). The percent of complex PD along wall interfaces of adjoining infected cells remained relatively low, with 3-4% branched. The percent between adjacent infected and uninfected cell wall interfaces increased 2-fold by zone III, going from 5% to 15% of total observed PD between these cell types. Adjacent uninfected cells had the biggest increase with 30% of all PD observed between these cells within zone III, up from 9% in zone II.

Figure 3.11 Percentage of complex PD increased in uninfected cells during development.

3.2.8 Uninfected cells contain more clustered PD Quantification thus far only considered those PD observed longitudinal to the wall. Cross-sectional views of PD which appeared as punctate spots, were excluded. In higher plants, both primary and secondary PD formation and further structural modifications have been linked to cellular differentiation and specialization (Ehlers and Kollmann, 1996; Ehlers and Kollmann, 2001; Fitzgibbon et al., 2013). To determine if the PD quantified are likely primary or secondary in formation and to better understand the observed PD distribution

patterns in regard to frequency and structural complexity, further assessment of PD connections took into consideration PD oriented in the wall displaying a cross-section view. Views of cross-sectioned PD are seen as punctate spots between the cell walls, each representing a channel. Primary PD, which are formed during cytokinesis, are straight, single-stranded channels (Hepler, 1982). Modifications of this structure are possible during subsequent cellular differentiation and development one of two ways. PD trapped in thickening cell walls are forced to branch, or adjacent PD fuse together producing an H-shape after the cell wall is disintegrated in the mid-region of the channel (Ehlers and Kollmann, 2001). Secondary PD formed when plant cells undergo differentiation are reportedly inserted into extending walls, sometimes in clusters referred to as pit fields (Esau and Thorsch, 1985). Channels viewed in this orientation provide information about secondary PD formation and changes in PD complexity regardless of the type. Punctate PD found clustered in groups of three or more are presented in (Figure 3.12). The average number of clusters were not significantly different in zone II, regardless of cell types. The number of clusters significantly decreased along adjacent infected cell wall interfaces from zone II to zone III by 50%, started out with an average of 2 clusters per 100 µm to less than 1 cluster per

100 µm. The number of clusters observed along adjacent infected-uninfected cells wall interfaces slightly increased between developmental zones, with zone II averaging just under 2 clusters per 100 µm up to 3 by zone III (Figure 3.12).

Interestingly, there were no notable changes in the number of clusters between

adjacent uninfected cells wall interfaces as were observed in longitudinal views of PD.

Figure 3.12 Uninfected cells have more clusters of PD. Average number of PD per 100 µm of cell wall. Clusters are cross-section views of PD channels.

To further investigate the differences between PD viewed as a cross- section and those longitudinal, the number of channels was considered as a variable in considering clusters of PD. If the criteria of having at least three punctate spots in order to count as a cluster was eliminating simple PD (i.e. one

punctate spot), then perhaps differences in the number of channels per cluster would reveal the same trend in complexity seen in the longitudinal view. To investigate further the changes in PD complexity, the average number of channels per cluster was quantified (Figure 3.13). These results are the most intriguing of the findings since the greatest difference was not found along adjacent uninfected cells wall interfaces but between adjacent infected and uninfected cells. Even though fewer clusters were present between adjacent infected and uninfected cells (Figure 3.12) the density of PD within each cluster significantly increased. They Started with 9 puncta per cluster in zone II and more than doubled to about 23 channels per cluster by the nitrogen fixation zone III (Figure 3.13). Adjacent uninfected cells wall interfaces, which had the greatest number of clusters in all zones, had a smaller increase in the average number of channels per cluster, which increased from an average of 4 channels in zone II to about 9 channels per cluster by zone III.

Figure 3.13 Clusters between infected-uninfected cells have more channels. Punctate spots greater than three were counted representing cross- sections of PD channels.

3.2.9 PD ultrastructure revealed simple PD could be complex Though the average total number of PD presented here does not factor in structure, many of the simple PD observed could be branched further down the middle lamella. Two areas in the quantification process could be skewed by this fact. First, simple PD adjacent to each other in the longitudinal orientation, would suggest a greater number of individual connections, each counted separately. Two, the increase in clusters of PD and the significant differences in

number of channels per cluster could represent varying numbers of PD with simple or complex structures. To determine if the number of channels per cluster and those PD simple in structure could indeed be branched PD, a technique called tilt series tomography was utilized. This type of tomography allowed for a novel view of PD ultrastructure and enabled us to assess if PD appearing simple in structure, could indeed be branched and if those clustered could be highly branched or clusters of many simple PD. However, the overall processing time for tilt-series tomography is lengthy, each required 3 hours of acquisition time with 2.5 hours for reconstruction. Further segmentation and three-dimensional rendering required an additional 15-20 hours per tomogram. This limitation prevents its use as a quantifiable method for modeling PD structure and branching. A total of 28 tomograms were acquired and reconstructed across zones II and III. Two of the 28 were segmented and rendered requiring an additional 15- 20 hours each. Figure 3.14 shows these two tomograms from the data set, which provides insight into the three-dimensional structures of PD below the 2D surface view seen in standard TEM micrographs. Plasmodesmata which appear simple in the cross-section (Figure 3.14 A, C arrows) indeed branch lower in the middle lamella (Figure 3.14 B, D, arrows). Between the adjacent uninfected cells (Figure 3.14 C, D) a few PD with simple structures remain as single channels.

Figure 3.14 Plasmodesmata with simple structures could be complex and branched. Tomograms representing complexity of apparent simple PD. Two adjacent infected cells (A & B) seen with several simple PD (A, Arrow) are complex in structure lower in the cell wall (B, arrow). Two adjacent uninfected cells (C & D) seen with several simple PD (C, Arrow) are complex in structure lower in the cell wall (D, arrow). In B and D, false colors were given to the Cell wall (green) and PD channels (yellow).

3.2.10 Symplasmic transport occurs directly into developmental zones through uninfected cells PD quantification showed a clear increase in the number of connections between adjacent uninfected cells, particularly in zone III. To assess if these channels are actively involved in transport of solutes unloaded from the phloem and if there exists a size exclusion limit barrier at zone II/III, as seen with free GFP, a smaller phloem-mobile symplasmic tracer HPTS was used to map transport from the phloem into mature pea nodules 14 to 21dpi. The following figure 3.15 are representative images of dye loading experiments conducted on 4 separate occasions. Each time 15 to 20 nodules

were loaded. Out of 73 loaded nodules, 56 were imaged. Out of the 56 nodules imaged 37 nodules contained dye. HPTS, loaded into the xylem from cut root 3 cm below the nodule, entered the phloem and subsequently the nodule (Figure 3.15, A & B). HPTS is observed in uninfected tissue adjacent to the nodule vascular bundles (Figure 3.15, C) and entered the central tissue through bands of uninfected cells (Figure

3.15, D & E). Eventually, dye was sequestered in vacuoles (Figure 3.15, E).

Figure 3.15 HPTS unloaded from vascular bundles enters the central tissue via uninfected cell networks. A & C Scale bar =500µm, B, D, & E scale bar=50µm. Nodule meristem (NM), vascular bundle (VB). HPTS (green) is seen entering the nodule from the root. Dye is observed from the root all the way to the meristem (A) in uninfected tissue adjacent to the bundles (C). HPTS enters the central tissue through bands of uninfected cells (D & E) and eventually sequestered into vacuoles (E, arrowheads). HPTS was observed occasionally in infected cells (E, arrow). Dye loading experiments were performed on four separate occasions. Each time 15 to 20 nodules were loaded. Out of 73 loaded nodules, 56 were imaged. Out of the 56 nodules imaged 37 nodules contained dye.

3.3 Discussion The indeterminate root nodule is a unique model for biological studies in several ways. The resulting de novo organ formation on mature plants and the developmental gradient established within the organ allow every stage of the cells development to be viewed in a single nodule. Each developmental stage has unique transport requirements and cellular functions easily defined within zones, which are visibly identifiable in a transverse cut through a mature nodule.

It has a persistent meristematic region, where new cells continually divide and start to differentiate eventually making up the nitrogen fixation zone III. This means cells in zone II are merely earlier stages in the developmental process. These features of the indeterminate nodule allow for simpler studies such as this one, to have greater impact on our understanding of developmental processes, such as cell specialization for transport needs. The following discussion of our results takes this developmental aspect into consideration as is commonly done in nodulation studies.

3.3.1 Uninfected tissue and vascular mapping support current literature Figures 3.2 and 3.3 are excellent examples of an indeterminate nodules’ vasculature and networks of uninfected cells respectively. This was made possible by clearing with ScaleP, which is discussed in depth in the next chapter. ScaleP allowed three-dimensional views of previously described

(Guinel, 2009a) vascular structures located along the periphery of indeterminate nodules. In addition, uninfected cell networks or arrays which were previously described (Selker, 1988) could be tracked several cell layers in depth.

Application of the ScaleP clearing technique has great potential in future studies of vascular development and the role uninfected tissue has in nodulation.

3.3.2 Uninfected cells are more connected Qualitative analysis of callose, known to localize to PD neck regions in the adjoining wall, suggested uninfected cell walls contain more PD than infected cell wall interfaces within the same developmental zone. Quantification of PD at the TEM level supported this finding and provided additional information on how distribution changes throughout nodule development. Within zone II, uninfected cell interfaces had more PD than adjacent infected cells wall interfaces. This finding suggests that by the time cells become infected, individual cell fate is already determined. An interesting finding since the underlying mechanism for determining which cells become infected or remain uninfected is unknown. The fact uninfected cell’s role in the indeterminate nodules is still largely unknown, this finding could provide a starting point for further studies of uninfected cell differentiation. By zone II/III the difference in the number of PD along uninfected cell wall interfaces increased compared to adjacent infected cells wall interfaces in the same zone.

This difference can be explained by changes in cell size. As infected cells become larger, the existing PD move further apart. This is also a plausible explanation for the same observed pattern seen in zone III.

3.3.3 Distribution across developmental gradient suggests secondary PD formation What is more revealing about the specialization of cell types for transport is when PD distribution is compared across zones. In Figure 3.8 the

decreased number of PD between adjacent infected cells is explained by expansion of the cell wall. When an infected cell is adjacent to an uninfected cell, the distribution of PD stayed relatively the same even though both cell types increased in size. This suggests more PD are being incorporated into the wall for cellular specialization. The increase in PD connecting adjacent uninfected cells, which do slightly increase in size during development, suggests the change is a result of more PD being incorporated into the walls to support influxes of symplasmic transport that occur in zone III. These results do not reflect those previously reported in determinate nodules, where PD were 2.5 times greater in infected cells compared to those uninfected (Brown et al., 1995). Though the same conclusions can be inferred here, that the apparent higher frequency of PD are secondary in formation (Robards and Lucas, 1990) and occurred after infection zone II.

3.3.4 Formation of secondary PD is accompanied by branching This idea is supported by numerous studies showing secondary PD formation, accompanied by changes in complexity occur when plant tissues undergo the sink/source transition (Oparka et al., 1999; Ehlers and Kollmann,

2001; De Storme and Geelen, 2014). Oparka et al. demonstrated that PD in leaf mesophyll decreased in permeability as the leaves underwent the sink/source transition and this was accompanied by simple PD becoming branched. To determine if the increase in PD observed between developmental zones is also accompanied by changes in PD complexity PD quantification included structure of the channel.

The number of channels observed with branched structures was significantly higher in uninfected cells by zone III (Figure 3.8), with a large spike occurring between zones II and II/III. This supports the current hypothesis that increased symplasmic transport requirements is accompanied by simple PD becoming branched.

3.3.5 PD simple in structure could be complex The tomography data brings to light three areas of advancement or areas that need for further assessment. First, the data presented above is the most extensive to our knowledge that examines changes in PD during indeterminate nodule development. Secondly, Initial analysis could be strengthened with development of more efficient tools with shorter processing to assess ultrastructure. For the 28 tomograms, each required 3 hours of acquisition time, with 2.5 hours each for reconstruction, and two of the 28 were segmented and rendered requiring 10-20 hours each. Lastly, the complexity of PDs ultrastructure revealed by this method, suggests there is much more to PD than meets the eye. Branching appeared to be common and PD recorded as simple could actually be complex in structure elsewhere in the middle lamella. This does not invalidate the above findings however. Indeed, some PD counted as single could in fact be branched, but the amount of sampling was extensive and is likely representative of the actual frequencies and distribution found within indeterminate nodules.

3.3.6 Uninfected cells are active in symplasmic transport into the central tissue The presence of PD in any form provides the structural component for symplasmic transport. However, PD are dynamic structures and since the cell is capable of altering PD size exclusion limits in response to stimuli thus restricting molecules of varied sizes from passing through the channel (Gunning and Robards, 1976; Robards and Lucas, 1990; Wolf et al., 1991; Roberts and

Oparka, 2003; Christensen et al., 2009), this study assessed PD function using a small fluorescent tracer molecule. Few studies examining symplasmic transport in legume root nodules have shown both PD presence and function. In determinate nodule of Glycine max cv., Brown et. al. proposed observed bands of uninfected cells forming rays may act as a preferred symplasmic pathway leading to infected cells (Brown et al., 1995). From a small sample size (one nodule), they quantified PD along the bands of uninfected tissue and found a higher frequency of PD between uninfected cells compared to infected cells. They used a membrane-impermeant fluorochrome Lucifer Yellow-CH as a symplasmic transport tracer molecule. However, the dye did not show a preferential pathway through the bands of uninfected cells. Another study focused on initial stages of organogenesis in the indeterminate nodule of Medicago truncatula (Complainville et al., 2003). Their study quantified PD, in a small sample size (one nodule), in the newly formed nodule primordia and used free GFP (27kDa) as a symplasmic tracer. Interestingly, GFP was unloaded preferentially into the nodule meristem and stopped at zone II. No GFP was observed unloading from the phloem anywhere else and was not present in the intermediate zone II/III.

Our results suggest HPTS follows the path of water. After loading the dye into the xylem from the cut root tip, it enters the vasculature along the nodule periphery. After which, the dye leaks into to the phloem and eventually enters bands of uninfected cells (Figure 3.16).

Figure 3.16 Schematic of HPTS flow in Mature Indeterminate Root Nodules. HPTS (yellow) loaded into cut root tips enters the nodule vasculature through the xylem (orange). HPTS then enters the phloem (light green) before entering the symplast. Only able to pass through PD, HPTS flows primarily along networks of uninfected cells (blue rectangles).

The entry into the central tissue through uninfected tissue suggests they play a role in solute transport through the symplast. The increased number of PD channels found along uninfected cells supports this hypothesis. These findings differ from the pathway characterized by Complainville, which showed GFP expressed in the phloem unloading exclusively into the nodule meristematic region. The idea that all symplasmic transport supporting a functional nitrogen

fixation zone must move away from the root through two developmental zones, one of which highly regulated due to active infections occurring, just to be exported through the vasculature is highly unlikely. Our data suggests mature nodules establish more direct symplasmic pathways to the nodule vascular bundles and uninfected cells provide the route.

Chapter 4

SCALEP OPTICALLY CLEARS WHOLE PLANT ORGANS

ALLOWING FOR DEEP TISSUE IMAGING AND USE OF

FLUORESCENT MICROSCOPY TECHNIQUE

This work was published as a collaborative effort and is copyrighted by the American Society of Plant Biologists. My contributions include development of technique, writing and editing of manuscript, and image acquisition and processing. Journal URL: www.plantphysiol.org (Warner et al., 2014).

4.1 Background Fluorescence-based microscopy and three-dimensional imaging have revolutionized plant science research, leading to unique insights into phenotypic variation (Meijón et al., 2014), plant-microbe interactions (Dagdas et al., 2012), and gene expression patterns (Truernit et al., 2008). However, the potential of these technologies to help tackle a range of fundamental questions in plant biology has not been fully realized because of limitations caused by plants’ intrinsic properties and current methods used for sample preparation. The transmission of light through plant cells is impeded by the heterogeneous refractive indices of the cell wall and cytoplasm, and light is absorbed by pigments. In addition, the fluorescence of pigments and aromatic molecules further reduces the signal-to-noise ratio in images. Collectively, these intrinsic

properties restrict the imaging depth for plant samples to approximately 30 µm, representing the distance of only a few layers of cells, a depth much lower than the theoretical limits of modern confocal microscopes and result in poor image quality (Feijó and Moreno, 2004; Paddock and Eliceiri, 2014). To visualize internal plant structures, samples may be sectioned mechanically or processed using a clearing agent before imaging. Sectioning is labor and time intensive and can be complicated by the loss or damage of structures. It also requires alignment of serial images for interpretation within the context of the whole tissue, and this can also lead to inaccuracies in the representation of a specimen. Alternatively, solvents, such as ethanol, are frequently used to clear thick tissue samples (Phillips and Hayman, 1970; Gardner, 2009). Although this approach yields greater tissue transparency and light transmission, it unfortunately also results in the extraction of cellular constituents and the disruption of subcellular structures. Chloral hydrate preserves cellular features better than many other solvents (Lersten, 1986), but its alternative use as a sedative or hypnotic drug prevents widespread adoption in routine laboratory studies. In the past few years, rapid advances in mammalian tissue clearing techniques have facilitated microscopic analysis of whole tissues and intact organisms (Hama et al., 2011; Chung et al., 2013; Ke et al., 2013). Treatment of whole mouse embryos with one such reagent facilitated microscopic analysis of entire neuronal networks tagged with GFP (Hama et al., 2011). The goal of this work was to develop and fine tune a method to facilitate deep-tissue imaging of intact plant organs or whole plants. Specifically, we

aimed to develop an approach to retain fine cellular features in specimens, closely match the refractive indices of plant tissues, enhance light transmission through the sample, and preserve the ability to use common fluorescent stains and proteins. The plant-tissue clearing agent presented here overcomes all the major obstacles that have limited plant imaging to date, and it is, therefore, of broad use to the plant scientific community.

To advance our understanding of intercellular transport in indeterminate root nodule tissues the above clearing technique was used to analyze uninfected cell arrays and those results can be found within the following chapter.

4.2 Results

4.2.1 ScaleP clearing increases tissue transparency in multiple tissue types Clearing efficacy in photosynthetic and root-derived samples from monocot and dicot species was assessed in preserved samples treated with clearing solution. The formulation developed for this study was highly effective at clearing leaves and roots of all tested plant species, including Arabidopsis (Arabidopsis thaliana), Medicago truncatula, pea (Pisum sativum), Nicotiana benthamiana, and maize (Zea mays) (Figure B.1). Clearing substantially increased tissue transparency compared with untreated tissue Figure 4.1. The method optically cleared specimens in as little as 1 to 3 weeks, making them ready for staining and complete traverse imaging. Samples of mature monocotyledonous leaves became transparent after treatment (Figure 4.1A; Figure B.1), even in areas with thickened secondary wall development, such as the veins. Similarly, intact samples from dicotyledonous plants, such as leaves

and root nodules from pea, became transparent after treatment (Figure 4.1, C and D; Figure B.1, G and F). We assessed the depth of imaging achievable in cleared tissues using thick sections from root nodules (Figure 4.1, E and F).

Figure 4.1 ScaleP optically clears whole plant organs increasing the depth of imaging possible. Photos of a maize leaf punch and intact pea root nodule taken before (A and C) and after (B and D) treatment with clearing solution. Grid scale = 15 µm A and B. Bars = 500 µm in C and D. Depth of imaging achievable in the z axis of untreated and treated nodule tissue (E and F). Root nodule tissue was stained with the cellulose stain Calcofluor White (blue) and nucleic acid stain SYTO13 (green) and imaged using a multiphoton confocal microscope. Bars = 100 µm.

4.2.2 ScaleP clearing increased depth of imaging by 3-fold Cleared samples and uncleared controls were stained with SYTO13 and Calcofluor White and imaged using a multiphoton confocal microscope and a 25x/0.8 LD LCI PlanApochromat multi-immersion lens mounted with 30% (v/v) glycerol. Tissue clearing increased the depth of imaging by more than 3-fold compared with untreated tissue, with the depth of the z axis more than 350 µm in treated samples compared with 100 µm in untreated samples (Figure 4.1, E and F). The notable increase in depth of imaging after clearing allowed for three-dimensional analysis of multiple cell layers in different specimens.

4.2.3 ScaleP clearing effective for studying plant microbe interactions Transverse optical images of the entire thickness of a cleared maize leaf were rendered by z-stack reconstruction using Huygen’s Professional deconvolution and Amira three-dimensional visualization software (Figure 4.2,

A and B). This analysis allowed cell layers, cell junctions, and tissue structures to be visualized in the three-dimensional models, highlighting the mesophyll and vasculature structural features through the maize leaf (Figure 4.2A). Furthermore, the spatial relationship of these structures to a fungal pathogen, Cochliobolus heterostrophus, could be determined from whole-thickness samples (Figure 4.2B).

4.2.4 ScaleP is compatible with commonly used fluorescent markers, probes, stains and dyes Using the same rendering methods, an optical section (200 µm thick) was collected and reassembled from the central zone of an intact pea root nodule (Figure 4.2D). Key features could be analyzed without physical disruption of

the intricate association between pea and its symbiont Rhizobium leguminosarum bv viciae, including infection threads traversing multiple cells (Figure 4.2E). To evaluate the preservation of subcellular features and the compatibility of clearing with the use of fluorescent marker proteins, we cleared samples of N. benthamiana leaves expressing a fluorescent marker protein. The fluorescent property of the marker for actin, mTalin (Kost et al., 1998) tagged with the yellow GFP variant Citrine, was retained, allowing for visualization of both fine and thick actin bundles (Figure 4.2C). Consistent with the results reported from studies using standard microscopic techniques with fluorescent reporter proteins in fixed tissues, the marker retained fluorescence after fixation (Knapp et al., 2012). The preservation of the marker protein fluorescence in addition to fine subcellular structure opens the possibility to study protein distribution within every cell type using intact specimens treated with the clearing solution. Although fluorescent tags are commonly used to determine subcellular distribution of proteins, immunocytochemical methods are used to visualize the distribution of proteins and nonproteinaceous cellular constituents. To determine if immunocytochemical methods could be used as well, cleared nodule tissues were probed with anti-β-1,3-glucan antibodies using standard methods (Catalano et al., 2007) to determine the distribution of callose. Cleared samples were washed to remove clearing solution, immunolabeled, and subsequently re- infiltrated with the clearing solution for imaging. As expected, callose was detected on the cell walls within the central zone of nodule tissue (Figure 4.2F, arrows). Although clearing did not result in a measurable increase in

penetration of a large antibody probe into this tissue, it is possible that antibody penetration could be enhanced using established methods (Blancaflor et al., 2001; Harrison et al., 2002). These results, however, unequivocally show compatibility of clearing with immunolocalization methods and the preservation of fluorescent protein tags.

Figure 4.2 Compatibility of ScaleP cleared specimens with fluorescent markers. A and B, Three-dimensional rendering of a cleared maize leaf (approximately 200 µm thick). A, Cell walls were stained with Calcofluor White (brown). B, Fungal hyphae was stained with Alexa Fluor 594 wheat germ agglutinin (green). C, Three- dimensional maximum intensity projection of N. benthamiana leaf expressing protein marker mTalin-Citrine labeling actin filaments (green). D to F, Pea root nodules stained with Calcofluor White (brown) and SYTO13 (blue). D, A 200 µm -thick three- dimensional rendering reconstructed from multiple z stacks. E, an infection thread traversing multiple cells (arrows). F, Cell walls with cellulose microfibril lattice structures (brown) labeled with anti-b-1, 3-glucan and Alexa Fluor 568 antibodies (arrows; magenta). Bars = 20 µm in C, E, and F.

4.3 Discussion In conclusion, our goal was to develop and demonstrate the potential range of applications of this plant clearing technique. This work provides a useful method and a guide for plant researchers to achieve optical clarity in tissues, allowing increased imaging depths in plant samples. This optical clearing method enables complete imaging of intact plant nodules. The tissue- clearing formulation can be utilized in concert with diverse fluorescent stains, common fluorescent marker proteins, and antibody-based detection of epitopes (i.e., indirect immunolabeling methods), opening a range of unique possibilities for microscopic investigation in plant.

This chapter is an excerpt from the collaborative publication; An Optical Clearing Technique for Plant Tissues Allowing Deep Imaging and Compatible with Fluorescence Microscopy (Warner et al., 2014). It is included in this thesis with publisher’s permission.

Chapter 5

IMPLICATIONS AND FUTURE WORK

The callose immunolabelling patterns observed along uninfected cells suggested a greater number of PD are present. Quantification of PD at the TEM level confirmed that pattern and revealed density and complexity of these channels does change along the developmental gradient. The ability to clearly view callose distribution by clearing with ScaleP has many implications for future studies of plant development and cellular functions, such as regulation of PD.

Callose deposition is suggestive of multiple biological processes including modification of PD size exclusion limit (SEL) and plant defense responses. Provided here is correlative and suggestive data of callose as an indicator of PD presence. The fact that callose is predominantly found along uninfected cells could also suggest that the SEL is being actively regulated to control symplasmic transport in these areas. This might indicate formation of symplasmic micro-domains within the zones, which are defined by groups of cells that have the same SEL of their PD. The ability of plants to regulate PD size exclusion limit is not exclusively controlled by the reversible deposition of callose. An interesting follow up study to the callose immunolabelling performed here in conjunction with ScaleP clearing would test the hypothesis proposed by 99

Demchenko et al. that PD without callose and calreticulin are open channels for fast symplasmic transport (Demchenko et al., 2014). They theorize that high transport areas in mature tissue might not be tightly regulated. In fact, they suggest PD lose their desmotubules, and become open channels for fast symplasmic transport. Their theory is based off of studies of another symbiosis, the nitrogen-fixing actinorhizal root nodules of the Australian tree Casuarina glauca (Schubert et al., 2013; Demchenko et al., 2014). Perhaps in zone II were callose was frequently associated with infection threads (IT), the plant uses it to help navigate threads as they traverse the cell layers by controlling PD aperture. Callose has previously been implicated in the orchestration of such events in other symbiotic and non-symbiotic plant microbe interactions (Van Wees et al., 2008; Hofmann et al., 2010; Zamioudis and Pieterse, 2012). It would be interesting to further investigate callose’s role in infection thread progression during nodulation.

100

REFERENCES

Abd-Alla MH, Koyro H-W, Yan F, Schubert S, Peiter E (2000) Functional structure of the indeterminate Vicia faba L. root nodule: implications for metabolite transport. J Plant Physiol 157: 335–343 Allaway WG, Carpenter JL, Ashford AE (1985) Amplification of inter-symbiont surface by root epidermal transfer cells in the Pisonia mycorrhiza. Protoplasma 128: 227–231 Appleby CA (1984) leghemoglobin and rhizobium respiration. Annu Rev Plant Physiol 35: 443–478 Bala A, Karanja N, Murwira M, Lwimbi L, Abaidoo R, Giller K (2011) Production and use of Rhizobial inoculants in Africa. wwwN2Africa.org 21 van Bel AJE (1993) The Transport Phloem. Specifics of its Functioning BT. In H-D Behnke, U Lüttge, K Esser, JW Kadereit, M Runge, eds, Prog. Bot. Struct. Bot. Physiol. Genet. Taxon. Geobot. Springer Berlin Heidelberg, Berlin, Heidelberg, pp 134–150 Benitez-Alfonso Y (2014) Symplastic intercellular transport from a developmental perspective. J Exp Bot 65: 1857–63 Berry AM, Mcintyre L, Mccully ME (1986) Fine structure of root hair infection leading to nodulation in the Frankia - Alnus symbiosis. Can J Bot 292–305 Blancaflor EB, Zhao L, Harrison MJ (2001) Microtubule organization in root cells of Medicago truncatula during development of an arbuscular mycorrhizal symbiosis with Glomus versiforme. Protoplasma 217: 154–65 Bond G (1936) Quantitative Observations on the Fixation and Transfer of Nitrogen in the Soya Bean , with Especial Reference to the Mechanism of Transfer of Fixed Nitrogen from Bacillus to Host. Ann Bot 50: 559–578 Bond L (1948) Origin and developmental morphology of root nodules of Pisum sativum. Bot Gaz 109: 411–434. Brewin NJ (1991) Development of the legume root nodule. Annu Rev Cell Biol 7: 191–226 Brian MRO, Kirshbom PM, Maier RJ (1987) Bacterial heme synthesis is required for expression of the leghemoglobin holoprotein but not the apoprotein in soybean root nodules. Proc Natl Acad Sci 84: 8390–8393 Brown, R.C. and Lemmon BE (2009) Callose in cell division. Chem. Biochem. Biol. (1→3)-β-glucans Relat. polysaccharides. pp 425–437 Brown SM, Oparka KJ, Sprent JI, Walsh KB (1995) Symplastic transport in soybean root nodules. Soil Biol Biochem 27: 387–399

101

Brown SM, Walsh KB (1994) Anatomy of the legume nodule cortex with respect of nodule permeability. Aust J Plant Physiol 21: 49–68 Catalano CM (2005) Symbiosome membrane specialization in Medicago truncatula Root Nodules. University of Delaware Catalano CM, Czymmek KJ, Gann JG, Sherrier DJ (2007) Medicago truncatula syntaxin SYP132 defines the symbiosome membrane and infection droplet membrane in root nodules. Planta 225: 541–50 Charpentier M, Oldroyd G (2010) How close are we to nitrogen-fixing cereals? Curr Opin Plant Biol 13: 556–564 Chaudhry M (2011) Green Gold: Value-added pulses, 1st ed. Quantum Media Christensen NM, Faulkner C, Oparka K (2009) Evidence for unidirectional flow through plasmodesmata. Plant Physiol 150: 96–104 Chung K, Wallace J, Kim S-Y, Kalyanasundaram S, Andalman AS, Davidson TJ, Mirzabekov JJ, Zalocusky K a, Mattis J, Denisin AK, et al (2013) Structural and molecular interrogation of intact biological systems. Nature 497: 332–7 Clayton G, Rice W (2004) Inoculant formulation and fertilizer nitrogen effects on field pea: Crop yield and seed quality. Can J … 84: 89–96 Coba T, Peña D, Fedorova E, Pueyo JJ, Lucas MM (2018) The Symbiosome : Legume and Rhizobia Co-evolution toward a Nitrogen-Fixing Organelle ? 8: 1–26 Complainville A, Brocard L, Roberts I, Dax E, Sever N, Sauer N, Kondorosi A, Wolf S, Oparka K, Crespi M (2003) Nodule Initiation Involves the Creation of a New Symplasmic Field in Specific Root Cells of Medicago Species The Plant Cell. Plant Cell … 15: 2778–2791 Courty PE, Smith P, Koegel S, Redecker D, Wipf D (2014) Inorganic nitrogen uptake and transport in beneficial plant root-microbe interactions. CRC Crit Rev Plant Sci 34: 4–16 Crews T., Peoples M. (2004) Legume versus fertilizer sources of nitrogen: ecological tradeoffs and human needs. Agric Ecosyst Environ 102: 279–297 Crookes W (1898) Address of the President Before the British Association for the Advancement of Science, Bristol. Sci New Ser 8: 561–575 Cullimore J, Denarie J (2003) How legumes select their sweet talking symbionts. Science (80- ) 302: 575–578 Dagdas YF, Yoshino K, Dagdas G, Ryder LS, Bielska E, Steinberg G, Talbot NJ (2012) Septin-mediated plant cell invasion by the rice blast fungus, Magnaporthe oryzae. Science 336: 1590–5 Dahiya P, Brewin NJ (2000) Immunogold localization of callose and other cell wall components in pea nodule transfer cells. Protoplasma 214: 210–218 Davidson a. LL, Newcomb W (2001) Changes in actin microfilament arrays in developing pea root nodule cells. Can J Bot 79: 767–776 Demchenko KN, Voitsekhovskaja O V, Pawlowski K (2014) Plasmodesmata without callose and calreticulin in higher plants - open channels for fast symplastic transport? Front Plant Sci 5: 74 Ehlers K, Kollmann R (2001) Primary and secondary plasmodesmata: structure,

102

origin, and functioning. Protoplasma 216: 1–30 Ehlers K, Kollmann R (1996) Formation of branched plasmodesmata in regenerating Solanum nigrum-protoplasts. Planta 199: 126–138 Esau K, Thorsch J (1985) Sieve Plate Pores and Plasmodesmata , the Communication Channels of the Symplast : Ultrastructural Aspects and Developmental Relations. Katherine Esau and Jennifer Thorsch. Am J Bot 72: 1641–1653 Feijó J a, Moreno N (2004) Imaging plant cells by two-photon excitation. Protoplasma 223: 1–32 Fitzgibbon J, Beck M, Zhou J, Faulkner C, Robatzek S, Oparka K (2013) A developmental framework for complex plasmodesmata formation revealed by large-scale imaging of the Arabidopsis leaf epidermis. Plant Cell 25: 57–70 Gadd CH (1888) the nodules of leguminous plants. 17–21 Gallardo K, Lesignor C, Darmency M, Burstin J, Küster H (2006) Medicago truncatula handbook. Medicago truncatula Handb 1–23 Gardner RO (2009) An Overview of Botanical Clearing Technique. Stain Technol 50: 99–105 Gibson AH (1966) The Carbohydrate Requirements for Symbiotic Nitrogen Fixation: A “Whole-Plant” Growth Analysis Approach. Aust J Biol Sci 19: 499–515 Graham PH, Vance CP (2003) Update on Legume Utilization Legumes : Importance and Constraints to Greater Use. Plant Physiol 131: 872–877 Guinel FC (2009a) Getting around the legume nodule: II. Molecular biology of its peripheral zone and approaches to study its vasculature. Botany 87: 1139–1166 Guinel FC (2009b) Getting around the legume nodule: I. The structure of the peripheral zone in four nodule types. Botany 87: 1117–1138 Gunning BE, Pate JS, Minchin FR, Marks I (1974) Quantitative aspects of transfer cell structure in relation to vein loading in leaves and solute transport in legume nodules. Symp Soc Exp Biol 87–126 Gunning BES, Robards AW (1976) lntroduction to plasmodesmata. In Inter cellular Communication in Plants: Studies on Plasmodesmata. Intercell. Commun. Plants Stud. Plasmodesmata Gutierrez R, Grossmann G, Frommer WB, Ehrhardt DW (2010) Opportunities to explore plant membrane organization with super-resolution microscopy. Plant Physiol 154: 463–6 Hama H, Kurokawa H, Kawano H, Ando R, Shimogori T, Noda H, Fukami K, Sakaue-Sawano A, Miyawaki A (2011) Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse brain. Nat Neurosci 14: 1481–8 Harris, R. David, Hillman CG (2014) Foraging and Farming: The Evolution of Plant Exploitation. Routledge Harrison MJ, Dewbre GR, Liu J, Samuel T, Noble R, Parkway SN (2002) A Phosphate Transporter from Medicago truncatula Involved in the Acquisition of Phosphate Released by Arbuscular Mycorrhizal Fungi. Am Soc Plant Biol 14: 2413–2429

103

Heinlein M (2015) Plasmodesmata. doi: 10.1007/978-1-4939-1523-1 Heinlein M, Epel BL (2004) Macromolecular transport and signaling through plasmodesmata. Int Rev Cytol 235: 93–164 Helming J, Kuhlman T, Linderhof V, Oudendag D (2014) Impacts of legume-related policy scenarios. Hepler PK (1982) Endoplasmic reticulum in the formation of the cell plate and plasmodesmata. Protoplasma 111: 121–133 Herre EA (2006) Exploring the Implications of Mosaic Coevolutionary Outcomes. Evolution (N Y) 60: 869–871 Herridge DF, Peoples MB, Boddey RM (2008) Global inputs of biological nitrogen fixation in agricultural systems. Plant Soil 311: 1–18 Hirsch AM (1992) Tansley Review No . 40 Developmental biology of legume nodulation. New Phytol 122: 211–237 Hirsch AM (2009) Brief History of the Discovery of Nitrogen-fixing Organisms Rhizobia Actinomycetes. Situ 1–5 Hofmann J, Youssef-Banora M, de Almeida-Engler J, Grundler FMW (2010) The role of callose deposition along plasmodesmata in nematode feeding sites. Mol Plant Microbe Interact 23: 549–57 Karunakaran R, Ramachandran VK, Seaman JC, East AK, Mouhsine B, Mauchline TH, Prell J, Skeffington A, Poole PS (2009) Transcriptomic analysis of Rhizobium leguminosarum biovar viciae in symbiosis with host plants Pisum sativum and Vicia cracca. J Bacteriol 191: 4002–4014 Ke M-T, Fujimoto S, Imai T (2013) SeeDB: a simple and morphology-preserving optical clearing agent for neuronal circuit reconstruction. Nat Neurosci 16: 1154– 61 Kitaeva AB, Demchenko KN, Tikhonovich IA, Timmers ACJ, Tsyganov VE (2016) Comparative analysis of the tubulin cytoskeleton organization in nodules of Medicago truncatula and Pisum sativum: Bacterial release and bacteroid positioning correlate with characteristic microtubule rearrangements. New Phytol 210: 168–183 Knapp E, Flores R, Scheiblin D, Modla S, Czymmek K, Yusibov V (2012) A cryohistological protocol for preparation of large plant tissue sections for screening intracellular fluorescent protein expression. Biotechniques 52: 31–7 Knoblauch M, Vendrell M, de Leau E, Paterlini A, Knox K, Ross-Elliot T, Reinders A, Brockman SA, Ward J, Oparka K (2015) Multispectral Phloem- Mobile Probes: Properties and Applications. Plant Physiol 167: 1211–1220 Knox JP, Benitez-Alfonso Y (2014) Roles and regulation of plant cell walls surrounding plasmodesmata. Curr Opin Plant Biol 22: 93–100 Kohle H, Jeblick W, Poten F, Blaschek W, Kauss H (1985) Chitosan-Elicited Callose Synthesis in Soybean Cells as a. Plant Physiol 77: 544–551 Kollmann, R. and CG (1991) Studies on graft unions. III. On the mechanism of secondary formation of plasmodesmata at the graft interface. Protoplasma Kost B, Spielhofer P, Chua NH (1998) A GFP-mouse talin fusion protein labels plant

104

actin filaments in vivo and visualizes the actin cytoskeleton in growing pollen tubes. Plant J 16: 393–401 Kremer JR, Mastronarde DN, McIntosh JR (1996) Computer visualization of three- dimensional image data using IMOD. J Struct Biol 116: 71–6 Lersten NR (1986) Modified clearing method to show sieve tubes in minor veins of leaves. Stain Technol 61: 1–4 Levy A, Epel BL (2009) Cytology of the (1-3)-B-Glucan (Callose) in plasmodesmata and sieve plate pores. Chem Biochem Biol 1-3 Beta Glucans Relat Polysaccharides. doi: 10.1016/B978-0-12-373971-1.00013-3 Łotocka B, Kopcińska J, Skalniak M (2012) Review article: The meristem in indeterminate root nodules of Faboideae. Symbiosis 58: 63–72 Lucas WJ, Lee J-Y (2004) Plasmodesmata as a supracellular control network in plants. Nat Rev Mol Cell Biol 5: 712–26 MacRobbie EAC (1971) Phloem translocation. facts and mechanisms: a comparative survey. Biol Cell 46: 429–481 Malleson C (2016) Informal intercropping of legumes with cereals ? A re-assessment of clover abundance in ancient Egyptian cereal processing by-product assemblages : archaeobotanical investigations at Khentkawes town , Giza ( 2300 – 2100 BC ). Veg Hist Archaeobot 25: 431–442 Maunoury N, Redondo-Nieto M, Bourcy M, Van de Velde W, Alunni B, Laporte P, Durand P, Agier N, Marisa L, Vaubert D, et al (2010) Differentiation of symbiotic cells and endosymbionts in Medicago truncatula nodulation are coupled to two transcriptome-switches. PLoS One 5: e9519 McDonald JH (2009) Handbook of Biological Statistics. Sparky House Publ 291 Meade HM, Long SR, Ruvkun GB, Brown SE, Ausubel FM (1982) Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis. J Bacteriol 149: 114–122 Meijón M, Satbhai SB, Tsuchimatsu T, Busch W (2014) Genome-wide association study using cellular traits identifies a new regulator of root development in Arabidopsis. Nat Genet 46: 77–81 Morel MA, Braña V, Castro-sowinski S (2012) Legume Crops , Importance and Use of Bacterial Inoculation to Increase Production. Crop Plant, Dr Aakash Goyal. pp 217–240 Mylona P, Pawlowski K, Bisseling T (1995) Symbiotic Nitrogen Fixation. Plant Cell 7: 869–885 Newcomb EH, Tandon SR, Kowal RR (1985) Ultrastructural specialization for ureide production in uninfected cells of soybean root nodules. Protoplasma 125: 1–12 Newcomb W, Peterson RL, Sippell D (1979) The early morphogenesis of Glycine max and Pisum sativum root nodules. Can J Bot 57: 2603–2616 O’Hara RB, Kotze DJ (2010) Do not log-transform count data. Methods Ecol Evol 1: 118–122 Oa CD, Yun MH, Torres PS, Oirdi M El, Rigano LA, Gonzalez-Lamothe R, Marano R, Castagnaro AP, Dankert MA, Bouarab K, et al (2006) Xanthan

105

Induces Plant Susceptibility by Suppressing Callose Deposition1[OA]. Plant Physiol 141: 178–187 Oparka KJ, Cruz SS (2000) THE GREAT ESCAPE: Phloem Transport and Unloading of Macromolecules. Annu Rev Plant Physiol Plant Mol Biol 51: 323– 347 Oparka KJ, Prior D a, Santa Cruz S, Padgett HS, Beachy RN (1997) Gating of epidermal plasmodesmata is restricted to the leading edge of expanding infection sites of tobacco mosaic virus (TMV). Plant J 12: 781–789 Oparka KJ, Roberts AG, Boevink P, Cruz SS, Roberts I, Pradel KS, Imlau A, Kotlizky G, Sauer N, Epel B (1999) Simple, but not branched, plasmodesmata allow the nonspecific trafficking of proteins in developing tobacco leaves. Cell 97: 743–754 Otero S, Helariutta Y, Benitez-Alfonso Y (2016) Symplastic communication in organ formation and tissue patterning. Curr Opin Plant Biol 29: 21–28 Paddock SW, Eliceiri KW (2014) Laser scanning confocal microscopy: history, applications, and related optical sectioning techniques. Methods Mol Biol 1075: 9– 47 Pate JS (1957) I. The synchronization of host and symbiotic development in the field pea, pisum arvense l. Nodulation Stud. Legum. pp 496–515 Pate JS, Gunning BES, Briarty LG (1969) Ultrastructure and functioning of the transport system of the leguminous root nodule. Planta 85: 11–34 Patrick JW (1997) phloem unloading: Sieve Element Unloading and Post-Sieve Element Transport. Annu Rev Plant Physiol Plant Mol Biol 48: 191–222 Phillips JM, Hayman DS (1970) Improved procedures for clearing roots and staining parasitic and vesicular-arbuscular mycorrhizal fungi for rapid assessment of infection. Trans Br Mycol Soc 55: 158-IN18 Pickard WF (2003) The role of cytoplasmic streaming in symplastic transport. Plant, Cell Environ 26: 1–15 Preibisch S, Saalfeld S, Tomancak P (2009) Globally optimal stitching of tiled 3D microscopic image acquisitions. Bioinformatics 25: 1463–1465 Radford JE, Vesk M, Overall RL (1998) Callose deposition at plasmodesmata. Protoplasma 201: 30–37 Radford JE, White RG (2001) Effects of tissue-preparation-induced callose synthesis on estimates of plasmodesma size exclusion limits. Protoplasma 216: 47–55 Rahman MM, Islam AM, Azirun SM, Boyce AN (2014) Tropical legume crop rotation and nitrogen fertilizer effects on agronomic and nitrogen efficiency of rice. ScientificWorldJournal 2014: 490841 Remigi P, Zhu J, Young JPW, Masson-Boivin C (2015) Symbiosis within Symbiosis: Evolving Nitrogen-Fixing Legume Symbionts. Trends Microbiol xx: 1–13 Rigano LA, Payette C, Brouillard G, Marano MR, Abramowicz L, Torres PS, Yun M, Castagnaro AP, Oirdi M El, Dufour V, et al (2007) Bacterial cyclic beta- (1,2)-glucan acts in systemic suppression of plant immune responses. Plant Cell 19: 2077–89

106

Robards AW, Lucas WJ (1990) Plasmodesmata. Annu Rev Plant Biol 41: 369–419 Roberts AG, Oparka KJ (2003) Plasmodesmata and the control of symplastic transport. Plant, Cell Environ 26: 103–124 Rogers C, Oldroyd GED (2014) Synthetic biology approaches to engineering the nitrogen symbiosis in cereals. J Exp Bot 65: 1939–46 Roser M, Ortiz-Ospina E (2018) “World Population Growth.” OurWorldInData.org, Salmon MS, Bayer EMF (2013) Dissecting plasmodesmata molecular composition by mass spectrometry-based proteomics. Front Plant Sci 3: 1–8 Santana MA, Pihakaski-maunsbach K, Sandal N, Marcker KA, Smith AG (1998) Evidence that the Plant Host Synthesizes the Heme Moiety of Leghemoglobin in Root Nodules. Plant Physiol 116: 1259–1269 Santi C, Bogusz D, Franche C, Perpignan D, Domitia V, Alduy AP, Rhizogene E (2013) Biological nitrogen fixation in non-legume plants. Ann Bot 111: 743–67 Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9: 676–682 Schneider C a, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9: 671–675 Schubert M, Koteyeva NK, Wabnitz PW, Santos P, Büttner M, Sauer N, Demchenko K, Pawlowski K (2011) Plasmodesmata distribution and sugar partitioning in nitrogen-fixing root nodules of Datisca glomerata. Planta 233: 139– 52 Schubert M, Koteyeva NK, Zdyb A, Santos P, Voitsekhovskaja O V, Demchenko KN, Pawlowski K (2013) Lignification of cell walls of infected cells in Casuarina glauca nodules that depend on symplastic sugar supply is accompanied by reduction of plasmodesmata number and narrowing of plasmodesmata. Physiol Plant 147: 524–40 Schulz A (2014) Diffusion or bulk flow: how plasmodesmata facilitate pre-phloem transport of assimilates. J Plant Res 128: 49–61 Selker JML (1988) Three-dimensional organization of uninfected tissue in soybean root nodules and its relation to cell specialization in the central region. Protoplasma 147: 178–190 Simpson C, Thomas C, Findlay K, Bayer E, Maule AJ (2009) An Arabidopsis GPI- anchor plasmodesmal neck protein with callose binding activity and potential to regulate cell-to-cell trafficking. Plant Cell 21: 581–94 Sivaguru M, Fujiwara T, Josef S, Yang Z, Osawa H, Maeda T, Mori T, Volkmann D, Matsumoto H (2000) Trafficking of Molecules through Plasmodesmata . A New Mechanism of Aluminum Toxicity in Plants. Plant Physiol 124: 991–1005 Smith RS (1992) Legume inoculant formulation and application. Can J Microbiol 38: 485–492 Sprent JI, Embrapa (1980) Root nodule anatomy, type of export product and evolutionary origin in some Leguminosae. Plant Cell Environ 3: 35–43 Steele NM, McCann MC, Roberts K (1997) Pectin Modification in Cell Walls of

107

Ripening Tomatoes Occurs in Distinct Domains. Plant Physiol 114: 373–381 De Storme N, Geelen D (2014) Callose homeostasis at plasmodesmata: molecular regulators and developmental relevance. Front Plant Sci 5: 138 Terpolilli JJ, Hood G a., Poole PS (2012) What determines the efficiency of N(2)- fixing Rhizobium-legume symbioses?, 1st ed. Adv Microb Physiol. doi: 10.1016/B978-0-12-398264-3.00005-X Thompson NJ (2005) The Geographic Mosaic of Coevolution. Timmers a C, Auriac MC, Truchet G (1999) Refined analysis of early symbiotic steps of the Rhizobium-Medicago interaction in relationship with microtubular cytoskeleton rearrangements. Development 126: 3617–3628 Truernit E, Bauby H, Dubreucq B, Grandjean O, Runions J, Barthélémy J, Palauqui J-C (2008) High-resolution whole-mount imaging of three-dimensional tissue organization and gene expression enables the study of Phloem development and structure in Arabidopsis. Plant Cell 20: 1494–503 Udvardi M, Poole PS (2013) Transport and metabolism in legume-rhizobia symbioses. Annu Rev Plant Biol 64: 781–805 Udvardi MK, Day DA (1997) Metabolite Transport Across Symbiotic Membranes of Legume Nodules. Annu Rev Plant Physiol Plant Mol Biol 48: 493–523 Vance CP, Graham PH, Allan D (2000) Biological nitrogen fixation. Phosphorus: a critical future need. Vedam V, Kannenberg EL, Haynes JG, Sherrier J, Datta A, Carlson RW, Sherrier DJ (2003) A Rhizobium leguminosarum AcpXL Mutant Produces Lipopolysaccharide Lacking 27-Hydroxyoctacosanoic Acid A Rhizobium leguminosarum AcpXL Mutant Produces Lipopolysaccharide Lacking 27- Hydroxyoctacosanoic Acid. J Bacteriol 185: 1841–1850 Verma DP (1992) Signals in root nodule organogenesis and endocytosis of Rhizobium. Plant Cell 4: 373–382 Verma DPS, Gu X, Hong Z (1995) Biogenesis of the Peribacteroid Membrane in Root Nodules: Roles of Dynamev and Phosphatidyl-Inositol 3-Kinase. In IA Tikhonovich, NA Provorov, VI Romanov, WE Newton, eds, Nitrogen Fixat. Fundam. Appl. Proc. 10th Int. Congr. Nitrogen Fixat. St. Petersburg, Russ. May 28--June 3, 1995. Springer Netherlands, Dordrecht, pp 467–470 Wagner R, Modla S, Hossler F, Czymmek K (2012) Three-Dimensional Analysis and Computer Modeling of the Capillary Endothelial Vesicular System with Electron Tomography. Microcirculation 19: 477–484 Walsh KB (1990) Vascular transport and soybean nodule function. III: Implications of a continual phloem supply of carbon and water. Plant Cell Environ 13–893 Warner CA, Biedrzycki ML, Jacobs SS, Wisser RJ, Caplan JL, Sherrier DJ, Sciences B (2014) An Optical Clearing Technique for Plant Tissues Allowing Deep Imaging and Compatible with Fluorescence Microscopy. Plant Physiol 166: 1684–1687 Van Wees SCM, Van der Ent S, Pieterse CMJ (2008) Plant immune responses triggered by beneficial microbes. Curr Opin Plant Biol 11: 443–8

108

Werner D, Mellor RB, Hahn MG, Grisbach H (1985) soybean root response to symbiotic infection glyceollin-i accumulation in an ineffective type of soybean nodules with an early loss of the peribacteroid membrane. Zeitschrift für Naturforsch C A J Biosci 40: 179–181 Wolf S, Deom CM, Beachy R, Lucas WJ, Wolfia S, Deomib CM, Beachyib R (1991) Plasmodesmatal Function is Probed Using Transgenic Tobacco Plants That Express a Virus Movement Protein. Plant Cell 3: 593–604 Wood EA, Butcher GW, Brewin NJ, Kannenberg EL (1989) Genetic derepression of a developmentally regulated lipopolysaccharide antigen from Rhizobium leguminosarum 3841. J Bacteriol 171: 4549–4555 Wright KM, Oparka KJ (1996) The fluorescent probe HPTS as a phloem-mobile , symplastic tracer : an evaluation using confocal laser scanning microscopy. J Exp Biol 47: 439–445 Xiao TT, Schilderink S, Moling S, Deinum EE, Kondorosi E, Franssen H, Kulikova O, Niebel A, Bisseling T (2014) Fate map of Medicago truncatula root nodules. Development 141: 3517–28 Zambryski P (2004) Cell-to-cell transport of proteins and fluorescent tracers via plasmodesmata during plant development. J Cell Biol 164: 165–168 Zamioudis C, Pieterse CMJCCMJ (2012) Modulation of host immunity by beneficial microbes. Mol Plant Microbe Interact 25: 139–50 Zimmermann, M.H., Milburn, J.A. and Eschrich W (1975) Transport in Plants I. Phloem transport. Encycl plant Physiol New Ser vol 1 245–255 Zimmermann MH ZH (1975) Transport in plants. I. Phloem transport. Encycl. plant Physiol. New Ser. vol 1., Vol. 1. Springer Berlin Heidelberg, pp 480–503 (2012) Food and Agriculture Organization of the United Nations (FAO). line database Agric., http://apps.fao.org/page/collections

109

Appendix A

SUPPLEMENTAL MATERIALS

Figure A.1 Flask growth and nodule harvest method. (A) 21dpi Pisum Sativum plants grown in flasks and removed to show root systems bottom. B is an enlarged root system with nodule formation (white boxes). C is an example of a mature indeterminate root nodule. If bisected along blue dotted lines and rotated (D) each developmental zone can be observed.

110

Table A.1 Plasmodesmata Analysis Totals: Measurements and Counts. I=Infected cell, U=Uninfected. Note increased trends in PD between cell types and developmental zones.

Total # of Avg Total # of PD # of # of # of # of Length Segments Length # of (Simple + Simple Complex Clusters Channels µm Included Per PD Complex) PD in Segment Clusters µm

Developmental Adjacent Sum Sum Mean Sum Sum Sum Sum Sum Sum Zone Cell Types

2 II 2235.57 180 12.529 451 284 274 10 32 167

IU 2439.64 210 11.443 673 461 431 30 35 212

UU 451.66 51 8.269 215 130 118 12 15 85

All 5126.87 441 10.972 1339 875 823 52 82 464

2/3 II 2145.41 122 17.603 243 183 175 8 14 60

IU 3530.04 245 15.077 1116 738 633 105 70 378

UU 692.87 67 11.025 391 209 153 56 35 182

All 6368.32 434 14.821 1750 1130 961 169 119 620

3 II 3844.07 234 16.360 361 219 204 15 32 142

IU 4766.27 307 16.167 1937 1048 894 154 136 889

UU 698.17 60 11.440 557 261 173 88 35 296

All 9308.51 601 14.980 2855 1528 1271 257 203 1327

All All 20803.70 1476 13.855 5944 3533 3055 478 404 2411

111

Figure A.2 Method of Plasmodesmata Quantification. Structure can be simple or complex (branched counted as one (1)). Channel must reach middle lamella PD can be clustered 3 or more punctate grouped µm Suggestive of complexity and branching

112

Appendix B

SUPPLEMENTAL MATERIALS FOR CHAPTER 4

Figure B.1 Clearing effects in monocot and dicot leaf tissues. Images acquired using a Nikon D700 camera of M. truncatula (A, B), N. benthamiana (C, D), A. thaliana (E, F), and P. sativum (G,H), leaves before and after clearing. By week 3 of clearing, samples become optically clear evident against patterned backgrounds. Grid squares equal 15 mm.

113

Figure B.2 Permission to Reprint Published Works. Chapter 4 was published as a collaborative effort and is Copyright by the American Society of Plant Biologists. My contributions include development of technique, writing and editing of manuscript, and image acquisition and processing. Journal URL: www.plantphysiol.org (Warner et al., 2014).

114

Image Designed by Ingrid Stark

115