A sustained CYCLINB1;1 and STM expression in the neoplastic tissues induced by Rhodococcus fascians on Arabidopsis underlies the persistence of the leafy gall structure

ABSTRACT

Rhodococcus fascians

is a gram-positive phytopathogen that infects a wide range of plant species. The actinomycete induces the formation of neoplastic growths, termed leafy galls, that consist of a gall body covered by small shoots of which the outgrowth is arrested due to an extreme form of apical dominance. In our previous work, we demonstrated that in the developing gall, auxin drives the transdifferentiation of parenchyma cells into vascular elements. In this work, with the use of transgenic Arabidopsis thaliana plants carrying molecular reporters for cell division (pCYCB1;1:GUS) and meristematic activity (pSTM:GUS), we analyzed the fate of cells within the leafy gall. Our results indicate that the size of the gall body is determined by ongoing mitotic cell divisions as illustrated by strong CYCB1;1 expression combined with the de novo formation of new meristematic areas triggered by STM expression. The shoot meristems that develop in the peripheral parts of the gall are originating from high ectopic STM expression. Altogether the presented data provide further insight into the cellular events that accompany the development of leafy galls in response to R. fascians infection.

Introduction

Generally, under optimal growth conditions, plant development follows a predetermined path that leads to a morphology that is specific for a particular species. Nevertheless, as sedentary organisms, plants have the capacity to modify their development in response to a diversity of abiotic stresses assuring their survival under adverse environmental situations. This developmental plasticity of plants is also exploited by a variety of organisms, including other (parasitic) plants, diverse animals (such as nematodes, insects, mites, and rotifers), protista, fungi, oomycetes, viruses, and, bacteria, that trigger the plant into forming atypical neoplastic outgrowths that serve as a food source and/or a niche for the inciting organism.^1,2 The morphology of plant galls resulting from such biotic interactions is for the most part determined by the gall-inducer rather than by the plant. However, overall, gall formation is established through particular conserved processes that include localized changes in cell cycle activity, cell growth, cell differentiation, and general reprogramming of morphogenesis. The modification of these pathways results from alterations in plant gene expression and accompanying induced hormonal imbalances triggered by the signals from the gall-inducers .^1,2 The implicated phytohormones are not only causal for gall initiation and development, they also play an important role in connecting the tumor to the plant vasculature and for the subsequent vascularization of the neoplasm, assuring water and nutrient provision to the gall cells and the gall-inducers.³ The molecular basis of these cellular events has been investigated for several galling systems in model plants like Arabidopsis thaliana taking advantage of diverse tools, such as reporter lines [e.g. 4–9].

A. thaliana is also a host for Rhodococcus fascians, a gram-positive bacterial phytopathogen that affects a wide range of mostly dicotyledonous herbaceous plants.^10,11 Infection with this actinomycete triggers a variety of plant responses, including stunting, leaf deformations, fasciations, witches’ brooms or other forms of excessive shoot formation, and ectopic adventitious shoot meristem development^12–14 However, the most characteristic phenotypic output of R. fascians infection is the formation of a highly differentiated neoplastic outgrowth, termed a leafy gall. Although the detailed appearance of a leafy gall can vary depending on the plant host and the infection conditions, its overall structure is comparable amongst hosts and consists of a gall body that is covered with small shoots, the outgrowth of which is arrested, giving the neoplasm a very compact look [11, 13, Figure 1a]. Interestingly, even when the plant host is senescing, the leafy gall structure persists.¹³ The leafy phenotype of the R. fascians-induced galls results from the bacterial production of regular and unique, methylated isoprenoid cytokinins^15–17 which directly affect plant development as well as stimulate the host to produce auxins and polyamines that aid in symptom development [reviewed in 18, 19]. Making use of an A. thaliana reporter line carrying a synthetic DR5 auxin-responsive element (pDR5:GUS), we have recently shown that changes in auxin signaling drive the neovascularization of leafy galls.²⁰ The extensive vascular network that surrounds the interior of the leafy gall extends in various directions toward the protruding developing shoots at the outer surface of the gall.²⁰

Figure 1.

Promoter activity of pCYCB1;1:GUS and pSTM:GUS in A. thaliana infected with R. fascians D188

(a) Leafy gall developed on the inflorescence at 3 weeks after inoculation. (b, c) pCYCB1;1:GUS (b) and pSTM:GUS (c) expression in the axil of the oldest side branch of the inflorescence without inoculation (left panels) and in the well-developed leafy galls 21 days after inoculation with R. fascians (right panels); black arrowheads: shoots protruding from the leafy gall; white arrowheads: side branch and subtending leaf; black arrows: main stem. (d) View of a central section through a leafy gall on pCYCB1;1:GUS plants (made along the axis of the main stem as indicated by the black arrow in b); white arrow: circular GUS signal in the leafy gall interior; black asterisk: GUS signal at the leafy gall periphery. (e) Detail of (d) showing differentiated vascular strands (pointed with black arrows). (f) Magnification of a longitudinal section through a newly formed shoot outgrowth on a pCYCB1;1:GUS plant, with a clearly distinguishable shoot apical meristem (marked with black asterisk) and flower primordia. (g-i) Images of successive cross sections through a shoot outgrowth, extending perpendicularly to the plane at which the leafy gall section was made (forming in the region marked with a black square in d). (g) No pCYCB1;1:GUS activity is detected where the vasculature is well-differentiated at the position where the outgrowth is connected to the leafy gall body (black arrow: differentiated vascular strands). (h-i) Strong pCYCB1;1:GUS activity is detected toward the more apical part of the extending outgrowth, where the vasculature just started to differentiate (h) or has not developed yet (i). (j) View of the central section through a leafy gall on pSTM:GUS plants (made along the axis of the main stem as indicated by the black arrow in c); white arrows: circular GUS signal in the leafy gall interior; black asterisk: GUS signal at the leafy gall periphery). (k) Detail of (j) showing differentiated vascular strands (marked with black arrows). (l) Magnification of a longitudinal section through a newly formed shoot outgrowth on a pSTM:GUS plant, with a clearly distinguishable shoot apical meristem (marked with black asterisk) and leaf primordia. (m-o) Images of successive cross sections through a shoot outgrowth, extending perpendicularly to the plane at which the leafy gall section was made (forming in the region marked with a black square in j). (m) No pSTM:GUS activity is detected where the vasculature is well-differentiated at the position where the outgrowth is connected to the leafy gall body (black arrow: differentiated vascular strands). (n-o) Strong pSTM:GUS activity is detected toward the more apical part of the extending outgrowth, where the vasculature just started to differentiate (n) or has not developed yet (o). Scale bars: 100 µm (a-c); 50 µm (d-o).

Here, we wanted to gain insight into the cellular events that underlie the growth of the gall body and the localized formation of shoots, taking a similar approach as in our study on leafy gall vascularization.²⁰ The meristematic character of the shoot apical meristem in Arabidopsis is maintained as a consequence of the activity of KNOTTED-like homeobox (KNOX) genes, such as SHOOTMERISTEMLESS (STM), and cell cycle-related genes, including CYCLINB1;1 (CYCB1;1),^20–23 Importantly, both these genes are induced upon R. fascians infection of A. thaliana and other hosts.^12,13,24,25 Thus, axils of side branches of inflorescences of pCYCB1;1:GUS and pSTM:GUS reporter lines were infected with R. fascians strain D188. The expression of both markers was analyzed at the tissue level by histochemical staining to provide a cellular map of dividing and meristematic cells within 3-week-old leafy galls. Since CYCB1;1 and STM expression in the R. fascians-Arabidopsis pathosystem is associated with the bacterial cytokinins, ^26–29 the results of this study will complement the view of the cellular auxin landscape in a leafy gall obtained with the pDR5:GUS reporter line.²⁰ Even more, given their role in the meristematic identity of tissues, analysis of the expression of CYCB1;1 and STM could also provide insight into the delayed senescence of leafy galls.

Materials and methods

Biological material and growth conditions

Transgenic Arabidopsis thaliana, Ler and Col-0 lines carrying, respectively, a pCYCB1;1:GUS (kind gift of Prof. Dengler – University of Toronto, Toronto, Canada) or pSTM:GUS (kind gift of Prof. Werr – University of Cologne, Cologne, Germany) reporter construct were used for the experiments. Seeds were surface sterilized, germinated, and grown in vitro using 0.5× Murashige and Skoog medium (MS) in long day conditions (LD, 16 h day/8 h night, 22°C). Flowering plants with a 10-cm high floral stalk were spot-inoculated at the axil of the oldest side branch junction with R. fascians D188 as described.²⁰ After 21 days the developed leafy galls were analyzed.

Microscopy observations

Control uninoculated plants and plants with developed leafy gall were stained for β‐glucuronidase activity. Plant material was pre-treated with 90% (v/v) ice-cold acetone for 30 min at room temperature and then rinsed with 50 mM sodium phosphate buffer (pH 7.2). The GUS staining procedure was performed with 2 mM 5-bromo-4-chloro-3-indolyl ß-D-glucuronide cyclohexamine (X-Gluc) for 16 h in the dark at 37°C. Then the samples were treated with an ethanol series (10, 30, 50%), fixed with an FAA solution (formaldehyde-acetic acid-ethanol), and stored in 50% ethanol. Tissue for sectioning was dehydrated in an increasing gradient of ethanol solutions (from 50% up to 100%) and embedded in paraffin.³⁰ Material was sectioned using a Leica RM2135 rotary microtome (Leica Instruments GmbH) for 10 μm sections along the axis of the main stem, gradually deparaffinized and mounted in Euparal. The images were captured using an epi-fluorescence microscope (BX51; Olympus Optical Co.) or a stereomicroscope (SZX9, Olympus Optical Co.) equipped with a DP71 camera (Olympus Optical Co.). The images shown are representative for three plants analyzed per treatment and reporter line.

Results

Upon infection of the axils of the oldest side branch of the inflorescence of A. thaliana with R. fascians D188, plant responses became macroscopically apparent after 4–5 days. The developmental alterations started with the activation of axillary meristems, followed by tissue swelling and the formation of adventitious shoot meristems [data not shown; 12, 20]. Eventually, 3 weeks after infection, these processes led to the development of a relatively large, globular, protruding structure of about 0.5–1 cm in diameter, the leafy gall (Figure 1a). Because leafy galls are characterized by the continuous and chaotic formation of multiple, precociously terminated shoots, at a given time-point, these neoplasms contain multiple outgrowths at different developmental stages.

In uninfected control pCYCB1;1:GUS and pSTM:GUS plants, GUS activity in the inflorescence side branch was typically only detected at the base of the branch (Figure 1b,c; left panels). In contrast, in 3-week-old leafy galls, a very strong GUS signal from both the pCYCB1;1 and pSTM promoters was detected in the central part of the outgrowths, signifying a continuation of cell division and a sustained meristematic activity at this location (Figure 1b,c; right panels). GUS activity was lower in the larger lateral protrusions likely reflecting the advanced developmental stage of their shoots/leaves (indicated by black arrowheads in Figure 1b,c). Just as in the control plants, no GUS activity was detected in the side branch and subtending leaf (indicated by white arrowheads in Figure 1b,c).

To better visualize the spatial pattern of ongoing cell division and of the distribution of meristematic cells within the gall structure, a series of sections was made through the leafy galls along the axis of the main stem (indicated by black arrows in Figure 1b,c). pCYCB1;1 promoter activity was widespread but not uniformly distributed throughout the gall tissues: GUS expression occurred in the peripheral regions of the gall coinciding with newly initiated shoot meristems (indicated by a black star in Figure 1d) and was also detected in circular regions in the interior part of the gall (indicated by a white arrow in Figure 1d). In developmentally more advanced shoot outgrowths, pCYCB1;1 promoter activity was more uniformly distributed throughout the tissues indicating that cell division and hence development had not arrested yet (Figure 1f). In serial cross sections from the most basal toward the apical part of new shoot outgrowths, a gradient of vascular differentiation could be discerned from highly differentiated at the base where the outgrowth is connected to the gall body (black arrow in Figure 1g), over developing but not yet fully differentiated in the mid part (Figure 1h), to still absent at the apex (Figure 1i). Whereas GUS activity was not detected in tissues surrounding fully differentiated vascular strands (Figure 1e,g), it was associated with regions of developing vasculature (Figure 1h) and regions without vascular tissue (Figure 1i).

In leafy galls formed on pSTM:GUS plants, similar as for pCYCB1;1 plants, promotor activity was detected in regions where new shoot meristems were forming (Figure 1d,j). Nevertheless, especially at the gall periphery, pSTM promoter activity was more pronounced and more elaborate than that of the pCYCB1;1 promoter (Figure 1d,j). In shoot outgrowths that were developmentally more advanced, the pSTM promoter was mainly active in the meristematic dome and in the region just beneath it, whereas in the leaves its activity diminished toward the tips (Figure 1l). pSTM:GUS expression in the leafy gall body was more restricted compared to pCYCB1;1 expression, and was detected in smaller dispersed circular regions (indicated by a white arrows in Figure 1j), suggesting the generation of multiple sites in which undifferentiated cells are being organized into self-sustaining meristems. Serial cross sections of new shoot outgrowths revealed that, just as for pCYCB1;1:GUS, pSTM:GUS expression was absent in tissues associated with fully differentiated vasculature (Figure 1k,m), whereas it was detected when the vasculature was still undergoing differentiation (Figure 1n,o).

Discussion

By combining our analysis of leafy gall neovascularization,²⁰ the data presented here, and previous studies in A. thaliana and tobacco in which the expression of cell cycle and KNOX genes has been analyzed in response to R. fascians infection, we can get a better insight into the leafy gall ontogeny. In tobacco, CYCB1;1 expression has been detected in differentiated cortical cells as early as 3 days after infection with R. fascians D188.²⁵ Auxin signaling, as evidenced by pDR5:GUS expression, in infected A. thaliana is increased within the same timeframe leading to the activation of the fascicular meristems.²⁰ These observations indicate that both a reentry into the cell cycle and a stimulation of vascular development are at the basis of the initiation of the gall structure. The CYCB1;1 expression data presented here further demonstrate that cell proliferation is the main factor that drives the gall size, because many cells within the R. fascians-infected tissues continue to be committed to mitotic cycling. Importantly, in Arabidopsis, prior to CYCB1;1 expression, R. fascians signals were shown to activate CYCA2;1 gene expression, implying that not only cell division but also the competence to divide is activated.¹³ This finding is supported by the STM expression pattern detected in this study, which showed that although many cells of the gall peripheral tissues have the potential to divide, they are not dividing yet, as evidenced by the lack of CYCB1;1 expression. Thus, the gall body and its peripheral tissues represent a reservoir of cells with (shoot) meristem characteristics that are contributing to the expansion of the gall and are assuring a continuous formation of new shoots responsible for the unique leafy phenotype of R. fascians-induced galls. Connecting the newly formed shoots to the vasculature of the plant is possibly mediated by auxin signaling, as besides CYCB1;1 and STM expression, also pDR5:GUS activity was observed in the shoot tips formed in the peripheral regions of the gall.²⁰

Previous transcriptome analysis revealed that R. fascians leaves a characteristic mark on A. thaliana indicative of a very strong cytokinin reponse²⁸ and STM activation in leaves was situated downstream of this response.²⁶ Hence, bacterially produced cytokinins are the most probable driving force for the observed activation of cell proliferation and meristematic activity within the gall tissue. It thus appears that the observed longevity of the leafy gall is a consequence of the presence of the bacterial cytokinins, the concomitant sink characteristics of the infected tissues,^27,28 and the cellular profile of the neoplasm. Also, in other pathosystems involving prokaryotic and eukaryotic gall-inducing organisms, elevated cytokinin levels or increased phytohormone sensitivity, induction of cell proliferation, and modifications in meristematic activity appear to be common processes involved in gall initiation.¹ For instance, during crown gall formation induced by Agrobacteriumtumefaciens, cytokinin and auxin overproduction by the transgenic cells, but also expression of T-DNA plast genes lead to ectopic expression of CYCB1;1 and class I KNOX genes.^31,32 Nematode feeding site development is triggered by nematode-derived cytokinins and depends on the subsequent activation of the cell cycle, including induced expression of CYCA2;1, CYCB1;1 and CDC2a, and of KNOX expression.^33–35 Upon inoculation with Plasmodiophora brassicae, clubroot initiation is accompanied by the induced expression of CYCB1;1 and other cell cycle genes and ANT (a marker for meristematic activity), in part as a result of plasmodial-produced cytokinins.^5,7,8 Also, at the onset of insect gall development, class I KNOX genes are activated most likely as a consequence of auxin and cytokinin secretion by the inciting insect.³⁶

In conclusion, despite the variety in the morphology of galls induced by different organisms, the disturbance of the phytohormonal balance and the acquisition of meristematic characteristics by differentiated cells are shared causes of the transition of the plant cells into tumorous growth. However, it is the blend of phytohormones and/or plant modifying morphogens, such as peptides or effectors, produced by these organisms that determine the phenotype of the induced galls.¹ It will be interesting to analyze the impact of regular isoprenoid cytokines and that of the methylated forms on A. thaliana development as a means to understand the cellular reprogramming that leads to the differentiated state of the leafy galls and that gives R. fascians its status of a somewhat unusual bacterial gall-inducer.

Funding Statement

This work was supported by the Ministerstwo Nauki i Szkolnictwa Wyższego [no. 501/73/ZBRR/20].

Author contribution statement

AD and DV conceived the research. AD, AB and DV designed the research. AD and AB conducted experiments. AD and DV wrote the manuscript. All authors read and approved the manuscript.

Disclosure of interest

The authors report no conflict of interest