Whole-mount confocal imaging of nuclei in giant feeding cells induced by root-knot nematodes in Arabidopsis


  • Paulo Vieira,

    1. Institut National de la Recherche Agronomique, UMR 1355 ISA, 400 route des Chappes, Sophia-Antipolis, France
    2. Centre National de la Recherche Scientifique, UMR 7254 ISA, 400 route des Chappes, Sophia-Antipolis, France
    3. Université de Nice-Sophia Antipolis, UMR ISA, 400 route des Chappes, Sophia-Antipolis, France
    Search for more papers by this author
  • Gilbert Engler,

    1. Institut National de la Recherche Agronomique, UMR 1355 ISA, 400 route des Chappes, Sophia-Antipolis, France
    2. Centre National de la Recherche Scientifique, UMR 7254 ISA, 400 route des Chappes, Sophia-Antipolis, France
    3. Université de Nice-Sophia Antipolis, UMR ISA, 400 route des Chappes, Sophia-Antipolis, France
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Janice de Almeida Engler

    1. Institut National de la Recherche Agronomique, UMR 1355 ISA, 400 route des Chappes, Sophia-Antipolis, France
    2. Centre National de la Recherche Scientifique, UMR 7254 ISA, 400 route des Chappes, Sophia-Antipolis, France
    3. Université de Nice-Sophia Antipolis, UMR ISA, 400 route des Chappes, Sophia-Antipolis, France
    Search for more papers by this author
    • These authors contributed equally to this work.

Author for correspondence:
Janice de Almeida Engler
Tel: +33 492 386 459
Email: janice.almeida-engler@sophia.inra.fr


  • Excellent visualization of nuclei was obtained here using a whole-mount procedure adapted to provide high-resolution images of large, irregularly shaped nuclei. The procedure is based on tissue clearing, and fluorescent staining of nuclear DNA with the dye propidium iodide.
  • The method developed for standard confocal imaging was applied to large multicellular root swellings, named galls, induced in plant hosts by the root-knot nematode Meloidogyne incognita.
  • Here, we performed a functional analysis, and examined the nuclear structure in giant feeding cells overexpressing the cell cycle inhibitor Kip-related protein 4 (KRP4). Ectopic KRP4 expression in galls led to aberrant nuclear structure, disturbing giant cell expansion and nematode reproduction. In vivo live-cell imaging of GFP-KRP4 demonstrated that this protein co-localizes to chromosomes from prophase to late anaphase during cell cycle progression.
  • The data presented here suggest the involvement of KRP4 during mitotic progression in plant cells. The detailed results obtained using confocal analysis also demonstrate the potential utility of a rapid, easy-to-use clearing method for the analysis of the nuclei of certain Arabidopsis mutants and other complex plant nuclei.


Root-knot nematodes (Meloidogyne spp.) are plant-parasitic nematodes capable of inducing large feeding sites within the root vascular tissue of plants. Feeding sites contain five to seven hypertrophied giant cells resulting from the substantial developmental reprogramming of host root cells caused by the pathogen. Giant cells undergo multiple rounds of nuclear division without cytokinesis, giving rise to large multinucleated cells (Sijmons et al., 1994; de Almeida Engler et al., 1999). These cells are highly metabolically active, and provide the nematode with nutrients necessary for its development and life cycle completion (Jones & Payne, 1978; Davis et al., 2000). As the development of the nematode feeding site (NFS) progresses, multiple rounds of DNA synthesis occur in the nuclei of giant cells, coupled with nuclear and cellular expansion (de Almeida Engler et al., 1999). Parallel with giant cell development, neighbouring root cells proliferate to form a multilayered cellular coat surrounding the giant feeding cells. Hyperplasia of neighbouring cells and hypertrophy of giant cells form the typical root-knot gall enclosing the nematodes (Jones & Northcote, 1972).

Plant–nematode interactions are often studied by performing detailed cytological analyses of transgenic lines compared with wild type in order to better understand major changes occurring during NFS development. Most of these cytological analyses are based on the physical production of gall sections using classical microtomy techniques. Although these sections can be easily stained with various dyes and microscopically observed, the method remains limited. A major drawback when using sectioned gall tissues for studying nuclei in giant cells is the loss of spatial perspective, especially when studying complex-shaped nuclei within feeding cells. Furthermore, reconstructing a 3D view from individual tissue sections from a complex architecture remains quite difficult and time consuming.

In Arabidopsis, whole-mount (WM) techniques are well established for studying the nuclei of different plant tissues (e.g. root meristems, endosperm or leaves) as well, for DNA quantification, nuclei in situ hybridizations, and others (Bauwens et al., 1994; Boisnard-Lorig et al., 2001; Willemse et al., 2008; Tirichine et al., 2009). These methods, coupled with confocal microscopy and computer-aided modelling software, provide detailed qualitative and quantitative information on the nuclear changes occurring in different plant tissues. Alternatively, an elegant WM procedure has been established for the model plant Arabidopsis based on the specific fluorescent staining of plant cell walls, using covalently attached propidium iodide (PI) fluorochromes. This method allows visualization of the plant cellular organization in WM to depths of > 200 μm, and was successfully used to generate excellent 3D reconstructions of protophloem development (Truernit & Haseloff, 2008). Although excellent for staining the cell walls of WM samples, this procedure does not stain nuclear DNA, even in the presence of the double-stranded DNA intercalating dye PI. Therefore, to stain nuclear DNA with PI in cleared WMs, we took advantage of established protocols (Bauwens et al., 1994; Willemse et al., 2008), and adapted the methodology to visualize nuclei in galls induced by root-knot nematodes. This allowed us to study a cell cycle inhibitor, Kip-related protein 4 (interactor/inhibitor of CDK (inhibitor/interactor of cyclin-dependent kinase, ICK7/KRP4)), in Arabidopsis tissue that showed altered nuclear morphology. The WM method presented facilitates the study of nuclear organization in complex plant tissues, such as galls, while avoiding laborious tissue sectioning. Although the molecular mechanism behind the formation and development of the NFSs is still not well understood, the characterization of several core cell cycle genes highly expressed during NFS development implies the involvement of the host plant cell cycle machinery during this plant-pathogen interaction (Niebel et al., 1996; de Almeida Engler et al., 1999, 2011, and de Almeida Engler et al., unpublished data).

One of the gene families involved in the regulation of the cyclin-dependent kinase (CDK)/cyclin (CYC) complexes during the cell cycle are the ICK/KRPs (Wang et al., 1997; De Veylder et al., 2001; Inzé & De Veylder, 2006). ICK/KRP genes have been associated with the switch from mitosis to endocycle and have been suggested to be involved in the various steps of cell division by binding to and inactivating CDK/CYC complexes. Previous studies have shown that ectopic expression of ICK/KRP genes affects nuclear ploidy levels. Low ICK/KRP levels block cell division and induce endoreduplication, while overexpression of ICK/KRP genes negatively affects cell division and endoreduplication cycles, resulting in a decrease in nuclear DNA content (Wang et al., 2000; De Veylder et al., 2001; Zhou et al., 2002; Verkest et al., 2005). Here we present data illustrating that expression of ICK7/KRP4 (hereafter termed KRP4) in galls strongly affects nuclear morphology in giant cells. This phenotype is unique among members of the ICK/KRP family in Arabidopsis. We show that a simplified WM protocol can be efficiently used to study nuclear morphology throughout giant cell development, permitting the study of nuclear rearrangements that occur in complex organs such as galls induced by root-knot nematodes in plant roots.

Materials and Methods

Nematode inoculation tests

For both transgenic line and wild-type Arabidopsis thaliana (L.) Heynh Columbia (Col-0), c. 60 sterilized seeds were placed in sterile Petri dishes containing 1% Murashige and Skoog (MS) germination medium (Duchefa, Haarlem, the Netherlands), 1% sucrose and 0.8% plant cell culture-tested agar (Sigma-Aldrich), supplemented with the appropriate antibiotics. After 2 d of cold treatment in the dark, seeds were kept in a growth chamber with a 16-h light : 8-h dark photoperiod at 21°C : 18°C, respectively. Fourteen-day-old plants were transferred to MS medium, and five plants per plate were placed in eight replica plates per line. One week later, root tips were inoculated with c. 100 surface-sterilized, freshly hatched, second-stage juveniles (J2s) of Meloidogyne incognita. All infection tests were performed in three independent experiments for both transgenic and wild-type lines. Transgenic T1 lines overexpressing ICK7/KRP4, driven by the cauliflower mosaic virus (CaMV) 35S promoter (35S::GFP-ICK7/KRP4, designated KRP4OE hereafter), were kindly provided by E. Russinova (Department of Plant Systems Biology, Flanders Institute for Biotechnology, Gent, Belgium). Kanamycin-resistant T3 KRP4OE lines were generated and evaluated based on KRP4 gene expression levels, protein localization and plant phenotype.

RNA isolation and semiquantitative RT-PCR amplification

To examine the expression of KRP4 in wild-type and KRP4OE Arabidopsis plants, total RNA was extracted with Trizol reagent (Invitrogen) from 7-d-old whole seedlings, using the RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. The RNA was treated with RQ1 RNase-free DNase (Promega) before reverse transcription. One microgram of treated RNA was added to RT reactions using a Bio-Rad iScript first-strand synthesis kit with random hexamer primers to make cDNA, according to the manufacturer’s instructions. Semiquantitative RT-PCR amplification was performed using KRP4 forward (Fw, 5′-AACCCCGGGATCGTCTAC-3′) and reverse (Rv, 5′-GCACCCGAGAAAAACTCG-3′) primers. The Arabidopsis Elongation Factor 1-α (EF1-α) gene was used as a control (Fw, 5′-GCTCTATGGAAGTTCGAGACCAC and Rv, 5′-GGTGTGGCAATCGAGAACTG).

Whole-mount analysis of giant cell nuclei

Nuclear analyses of cleared feeding sites were performed based on a WM mRNA in situ hybridization protocol described by Bauwens et al. (1994) and Willemse et al. (2008), and adapted with the following modifications. Four-week-old, uninfected and nematode-infected roots from 1 to 40 d after inoculation (DAI) were fixed in phosphate-buffered saline (PBS), pH 7.2, containing 1% formaldehyde and 10% dimethyl sulfoxide (DMSO) for 30 min at 20°C. Plant material was subsequently dehydrated twice with absolute methanol, and four times with absolute ethanol for 10 min for each step. Samples were stored at −20°C for 2–4 d for tissue clearing. Subsequently, seedlings were rinsed twice for 10 min with ethanol and for 40 min in ethanol : xylene (1 : 1). Samples were then washed twice with absolute ethanol and absolute methanol, for 5 min each, and post-fixed for 5 min in PBS, containing 0.1% Tween 20 (PBT) and 1% formaldehyde. Following post-fixation, roots were washed five times with PBT for 5 min. Samples were then washed twice for 5 min with PBS containing 50% formamide, and boiled twice for 3 min followed by quenching on ice. After samples had been stored overnight at room temperature (RT), they were washed four times in PBS for 15 min and stained with 0.5 μg ml−1 propidium iodide (PI) in PBS for 30 min at RT. Stained samples were mounted in 90% glycerol on a microscope slide and cover-slipped.

Confocal microscopy

Image acquisition was performed using a Zeiss LSM 510 META confocal laser scanning microscopy. The excitation wavelength for PI-stained samples was 543 nm, and the fluorescence emission was collected between 560 and 650 nm. High-resolution optical sections were acquired to distances > 200 μm inside the cleared galls using a ×40 1.3 oil objective lens. By combining the high fluorescence levels of the PI-stained nuclei and a high refraction index mounting medium, it was possible to collect extended z-series images with good resolution, allowing appropriate 3D projection of nuclei. Z-stacks were processed using the LSM 510 software to create maximum brightness projections and 3D view representations of nuclei. To visually improve image contrast, the red fluorescence signal of PI was converted to white.

Morphological analysis

For observation of the morphology of gall tissues, infected roots were collected at 14, 28 and 40 DAI, and fixed in 2% glutaraldehyde in 50 mM PIPES buffer, pH 6.9. Subsequently, samples were dehydrated and embedded in Technovit 7100 (Heraeus Kulzer, Wehrheim, Germany), as described by the manufacturer. Embedded roots and gall tissues were sectioned (3 μm) and stained with 0.05% toluidine blue and mounted in Depex (Sigma-Aldrich). Microscopic observations were performed using bright-field optics and images taken with a digital camera (AxioCam HRc; Zeiss).

In vivo observations of fresh gall root slices and whole-mount roots

Observations of nuclei in galls induced by M. incognita were performed in infected Arabidopsis roots harbouring free GFP (35S::GFP), and in the KRP4OE line. Galls at various time-points after infection (7–40 DAI) were dissected from roots and embedded in 5% agar. Fresh thick sections of 50–100 μm (7–14 DAI) or 150–200 μm (14–40 DAI) were obtained with a HM650V vibratome (Microm, Walldorf, Germany). Whole roots and fresh slices were observed using an inverted confocal microscope (model LSM510 META; Zeiss). As a consequence of the presence of tissue autofluorescence, GFP fluorescence specificity was monitored by spectral fingerprinting, using the Lambda mode tool. Standard GFP fluorescence was captured upon 488-nm excitation using a 500–530-nm emission band pass filter. All observations were performed in three independent experiments.


Protocol set-up for whole-mount clearing and PI staining of roots and galls

The intrinsic thickness and opacity of plant tissues, which can produce inadequate staining and imaging, limit visualization of nuclear morphology in whole galls induced by nematodes. Galls induced by root-knot nematodes in Arabidopsis roots can reach up to 500 μm (or more) in diameter, creating a challenge for the use of WM samples for nuclear imaging.

The principal changes to the original protocols (Bauwens et al., 1994; Willemse et al., 2008) are described hereafter. Whole roots of uninfected 4-wk-old Arabidopsis roots and galls at different developmental stages were tested. In the first series of experiments, we performed the clearing process by leaving the samples in ethanol at −20°C for a minimum of 2 d for thinner root samples, and a maximum of 4 d for thicker galls. Different concentrations of formamide were applied during the different washing steps, and 50% formamide in PBS (pH 7.2) gave the best clearing quality of root tissues. Formamide, also known as methanamide, can be used as a solvent to facilitate double-stranded DNA denaturation upon heating. In order to achieve good preservation and to avoid damage to the root material, a series of alternate short steps of heating and cooling in 50% formamide were tested. Depending on the root thickness, two to four series of 2–3 min at 100°C were applied and then the samples were immediately placed on ice for 5 min. This procedure was found to be best for tissue softening, transparency and nuclear staining. Staining of nuclei was performed by adding 0.5 μg ml−1 of PI in 90% glycerol and roots were mounted onto glass slides.

Slides were kept for several days in this staining mixture at 4°C. Repeated DNA heat denaturation, in the presence of formamide, followed by fast quenching, resulted in the preferential staining of nuclei along the entire root and gall tissue. As seen in Supporting Information Fig. S1, this specificity was lost when the denaturation/quenching step was omitted from the protocol. In Fig. S1(d) nuclei are more clearly visible because of the lack of background autofluorescence in cells filled with a large vacuole, such as epidermal and cortical root cells. By contrast, cells with a dense cytoplasm, such as root meristematic cells and giant cells, are more prone to PI-induced cytoplasmic background fluorescence. We observed that a long incubation in 90% glycerol leads to progressively more translucent roots or gall tissues.

Observation of galls induced by Meloidogyne incognita

Using WM preparations of infected roots, we were able to follow the migratory path of the second-stage nematode juveniles (J2s) within Arabidopsis roots. Nuclei of both roots and migrating nematodes were labelled with PI, whereas a weak PI counterstaining of the cell wall provided information on root cell morphology (Fig. 1a). During NFS induction, the first nuclear divisions occurring in giant cells were detected (Fig. 1b). At 2–4 DAI, several giant cells were generated adjacent to juvenile nematode (J2) heads, showing variable numbers and sizes of nuclei (Fig. 1c–e, Movies S1, S2). At intermediate stages of gall development (7–12 DAI), highly synchronized mitotic events were observed within the same giant cell (Fig. 1f,g). Regularly arranged metaphase-chromosome clusters aligned along a giant cell show that the events propelling mitotic division in giant cells are highly coordinated (Fig. 1f′; Movie S3).

Figure 1.

Whole-mount analysis of nuclei in galls induced by Meloidogyne incognita in Arabidopsis roots stained with propidium iodide. 3D confocal projections of serial optical sections of whole-mount samples were obtained. (a) Second-stage juveniles (J2) migrating within the plant root. (b) First mitotic division (metaphase) occurring in nuclei of an initiating giant cell (white dashed outline). (c) A 2-d-old gall showing young giant cells already containing two or more nuclei (white arrows). (d) A 3-d-old gall showing nuclei of variable sizes within different giant cells (white arrows), and mitotic divisions occurring within a giant cell (red arrows). (e) A 4-d-old gall showing an increased number of nuclei of different sizes and shapes within expanding giant cells. (f, f′) A 7-d-old gall showing synchronized nuclear divisions in one giant cell. (g) A 12-d-old gall showing late-anaphase nuclear divisions occurring in > 30 nuclei in a single giant cell. (h) Fourteen-day-old and (i) 30-d-old galls containing a large number of nuclei in giant cells. Asterisk, giant cell; n, nematode. Bars, (a–d) 10 μm; (e, f, f′) 5 μm; (g–i) 20 μm.

Our main goal was to apply the PI protocol in order to systematically study nuclear arrangements during the later stages of NFS development. It is in these later stages that gall size dramatically increases as a result of giant cell expansion and neighbouring cell proliferation. Therefore, we processed galls at different times (14, 21, 30 and 40 DAI) to optimally image, by confocal optical sectioning, nuclei corresponding to different positions within a giant cell with diameter thicknesses ranging from 100 to 500 μm (Fig. 1h,i). The transparency obtained after clearing allowed us to image deep into the gall tissues, and to generate high-quality data for giant cell nuclei as well as for mature and adult female nematodes. Z-stacks were converted to maximum brightness projections and 3D view representations of nuclei, illustrating their variability and spatial distribution within the different giant cells (Fig. 1i; Movie S4).

The present method preserved the overall morphology of nematodes, allowing us to easily identify their developmental stages during NFS expansion and to visualize gall nuclear morphology simultaneously (Fig. S2a–c). In addition, we were able to follow egg development within the gelatinous matrix secreted by the mature female nematode (Fig. S2d–f).

Ectopic KRP4 expression led to aberrant nuclear configuration in giant cells

During our study of cell cycle inhibitor genes, we observed that seedlings overexpressing KRP4 were slightly smaller than wild type, and presented serrated leaves as described for other KRPs (Wang et al., 2000; Verkest et al., 2005; Bemis & Torii, 2007). Confirmation of ectopic KRP4 expression was obtained by RT-PCR analysis (Fig. S3a) and by nuclear localization of the KRP4 protein (Fig. S3b,c). The effect of KRP4 overexpression during nematode infection was analysed in detail by microscopic analysis of galls. The cellular morphology of both the KRP4OE line and the wild type was analysed and followed by WM clearing and PI staining of KRP4OE galls in order to investigate the overall nuclear morphology. Ultimately, data for fixed cleared tissue were validated by in vivo GFP-KRP4 analysis.

Morphological analysis and observation of 4′,6-diamidino-2-phenylindole (DAPI)-stained giant cells in the KRP4OE line revealed that nuclei displayed irregular shapes compared with the wild type (Fig. 2). Frequently, nuclei in the KRP4OE line were clustered and seemed interconnected, suggesting the occurrence of abnormal nuclear division (Fig. 2e′,f). This atypical nuclear morphology was never observed in wild-type giant cell nuclei (Fig. 2b′,c).

Figure 2.

Morphological analysis of galls overexpressing Kip-related protein 4 (KRP4) and wild type in Arabidopsis roots infected by Meloidogyne incognita. Bright-field micrographs are shown of sections stained with toluidine blue (a, b, d, e) and DAPI (c, f) at three time-points after nematode infection (14, 28 and 40 d after inoculation (DAI)). (a–c) Sections of galls in roots of wild-type A. thaliana Col-0 showing multiple giant cell nuclei. (d–f) Sections of galls in roots of the KRP4OE line. Arrows indicate the interconnected nuclei within the giant cells. Asterisk, giant cell; n, nematode. Bars, 50 μm.

Analysis of overexpressing KRP4 whole cleared galls, stained with PI, revealed that giant cell nuclei were not only elongated, as seen in sectioned galls, but were also interconnected. PI staining of nuclei enabled the 3D reconstruction of the distribution of nuclei and aberrant nuclear phenotypes within giant cells at different times after nematode inoculation (Fig. 3). During mitotic divisions, proper separation of sister chromatids seemed affected, suggestive of abnormal chromosome segregation (Fig. 3a–c). As mitotic divisions progress within giant cells, this effect on nuclei becomes more severe (Fig. 3d,e; Movies S5, S6) as seen in mature galls (Fig. 3f,f′; Movie S7).

Figure 3.

Ectopic expression of Kip-related protein 4 (KRP4) leads to the formation of aberrant nuclei in Arabidopsis giant cells induced by Meloidogyne incognita. 3D confocal projections of serial optical sections of whole-mount root samples stained with propidium iodide are shown. (a) Nuclei in anaphase in wild-type giant cells and (b) in the KRP4OE line. (c) Synchronous mitotic events in giant cells of the KRP4OE line. Mitotic figures are abnormally shaped, suggesting that chromosome separation might be affected (arrows). (d, d′) A 7-d-old gall showing interconnected nuclei within giant cells (arrows). (e) Fourteen-day-old and (f) 30-d-old galls showing grouped and interconnected nuclei within different giant cells (arrows), and in detail (f′). Asterisk, giant cell; n, nematode. Bars, (a, b, f′) 5 μm; (b–e) 20 μm.

To validate these observations, we performed in vivo observations using the GFP-KRP4 construct. Uninfected KRP4OE roots revealed a nuclear localization characterized by intensely fluorescing dots, exclusively in interphase cells (Fig. S3d). During mitosis, the GFP-KRP4 localized to chromosomes in prophase, metaphase, and early anaphase (Fig. S3e–h). As anaphase progressed, the fluorescence signal associated with the chromosomes became gradually diffused until it was no longer detected. After cytokinesis, accumulation of GFP-KRP4 was observed in nuclei of new daughter cells (Fig. S3i). Subsequently, in vivo localization of KRP4 during NFS development (7, 14, 28 and 40 DAI) revealed strong fluorescence in nuclei of giant cells and neighbouring cells (Fig. 4). This analysis suggested that the visibly interconnected nuclei were correlated with the presence of the KRP4 protein, exclusively within the nuclei of giant cells. This nuclear phenotype was not observed in nuclei of normal root cells as these underwent cytokinesis. This interconnected nuclear phenotype was specific to giant cell nuclei and was already observed at 7 DAI (Fig. 4a,a′). As giant cells developed, the nuclear defects became more pronounced as seen at 14 to 40 DAI (Fig. 4b–d, Movie S8). Irregularly shaped giant cell nuclei were connected by condensed chromatin stretches and formed separate clusters (Fig. 4b,d). In contrast, in the control line expressing GFP alone, all nuclei appeared to be individually separated (Fig. S4a,b).

Figure 4.

In vivo nuclear localization of GFP-KRP4 in Arabidopsis galls induced by Meloidogyne incognita. These images are 3D confocal projections. Left panel, GFP fluorescent images; right panel, overlays of GFP fluorescence with differential interference contrast. (a, a′) A 7-d-old gall showing abnormal nuclear division (red arrows) and interconnected nuclei within giant cells (white arrows). (b, b′) A 14-d-old gall showing interconnected nuclei within giant cells (arrows). (c, c′) A 30-d-old gall showing grouped and interconnected nuclei within giant cells (arrows). (d, d′) Isolated nuclei of 40-d-old giant cells showing bridges of interconnected nuclei. Asterisk, giant cell; n, nematode. Bars, (a–c) 20 μm; (d) 5 μm.

Nuclei within giant cells of the KRP4OE line remained connected until galls were fully grown, indicating that mitotic defects were caused by KRP4 overexpression. Giant cells in the KRP4OE line were much smaller than those in wild-type infected roots at the three time-points studied (Fig. S5a). As a consequence, the growth and development of the nematode were substantially delayed in the KRP4OE line (Fig. S5d,e). By contrast, in wild-type galls at 40 DAI, large egg masses were found associated with mature females (Fig. S5c). These results indicate that KRP4 overexpression negatively interferes with the nematode’s life cycle by hampering nematode development.


To monitor the distribution and behaviour of giant cell nuclei in nematode-infected Arabidopsis roots, 3D projections of PI-stained cleared gall samples were analysed and live-cell imaging of nuclear localized GFP-KRP4 protein was carried out using confocal microscopy.

Nuclear organization in galls induced by root-knot nematodes

A hallmark of giant cell development is the gradual increase in the number of nuclei followed by nuclear enlargement during giant feeding cell expansion. Cytological observations of giant cells during NFS development have been reported in different plant hosts (e.g. Jones & Northcote, 1972; Jones & Payne, 1978; Endo, 1987), and for several Arabidopsis mutants (e.g. Caillaud et al., 2008; Clément et al., 2009; Banora et al., 2011). The few reports on the distribution of nuclei in giant feeding cells relied on single sections or standard wide-field microscope analysis of flattened giant cells of infected host plants (e.g. Bird, 1961; Wiggers et al., 1990; Starr, 1993).

Herein, we show that optical sections, generated by standard confocal microscopy of WM galls, produce images revealing unprecedented information on the distribution of large nuclei within giant cells, from the stages of initial root infection by juvenile nematodes to mature galls containing reproducing females. Observations of mitotic events in giant cells of Arabidopsis confirmed that, within a single giant cell, amplification of nuclear material is highly synchronized, as observed in other plant hosts (e.g. Bird, 1961, 1973; Owens & Novotny-Specht, 1964; Huang & Maggenti, 1969; Wiggers et al., 1990).

KRP4 overexpression in galls disrupts nuclear division and affects nematode development

Histological observations of galls overexpressing the Arabidopsis cell cycle inhibitor KRP4 showed that, although nematodes succeeded in infecting the tissue and inducing the formation of feeding cells, nuclear morphology, as well as gall development, was severely affected. Cleared and PI-stained whole galls in the KRP4OE line revealed that nuclei presenting diverse shapes were interconnected, forming bridges, and were often clustered. These features became more severe as mitotic divisions occurred during gall development. Arabidopsis tissues deficient in the Microtubule-associated protein 65-3 show defects in karyokinesis and cytokinesis and subsequent nuclear connections (Caillaud et al., 2008). As a consequence, plant growth is stunted. Nuclear interconnections in the KRP4OE line were only observed in acytokinetic giant cells and were not observed in root meristematic and vascular cells of the KRP4OE line.

Despite the progress made in understanding plant CDK inhibitors of the ICK/KRP family, their specific functions and physiological effects in plants remain to be elucidated. To date, detailed functional analyses of the KRP1 and KRP2 Arabidopsis genes have been carried out (Wang et al., 2000; De Veylder et al., 2001; Schnittger et al., 2003; Zhou et al., 2003; Verkest et al., 2005; Weinl et al., 2005; Jakoby et al., 2006; Roeder et al., 2010; Sanz et al., 2011), whereas for the other ICK/KRP family members little is known (Bemis & Torii, 2007; Kim et al., 2008; Liu et al., 2008; Anzola et al., 2010). Strong overexpression of KRPs (KRP1, KRP2, KRP6 and KRP7) in Arabidopsis seedlings affects cell proliferation and consequently plant growth. However, different Arabidopsis ICK/KRP members may have distinct roles, as their expression patterns vary within different plant tissues (de Almeida Engler et al., 2009 in http://www.psb.ugent.be/ishi/) and also during the cell cycle (Menges et al., 2005). KRP4 transcripts seem to be preferentially expressed in mitotically dividing cells such as root and shoot meristems (Ormenese et al., 2004; Marjanac et al., 2009). KRP4 promoter activity or transcripts are not detected in vascular root tissue of Arabidopsis (http://www.psb.ugent.be/ishi/, Sample ID:152103) or in NFSs induced within the vascular cylinder (P. Vieira et al. and de Almeida et al., unpublished data). At any event, KRP4 overexpression in NFSs revealed that giant cell nuclei were unable to divide properly. Multiple defects in giant cell nuclei, such as irregular shape, size and interconnections, were observed during our microscopic observations. In wild-type giant cells, nuclei tend to group, appearing to be interconnected, in sectioned galls. Here our PI analyses showed that nuclei were unconnected within mature giant cells of Arabidopsis. However, we cannot exclude the possibility that occasionally abnormal division of genetic material might occur (Wiggers et al., 1990), leading to the formation of aberrant nuclei. The occurrence of cumulative defects in nuclei within a single giant cell overexpressing KRP4 suggests the occurrence of abnormal chromosome segregation leading to an abnormal nuclear phenotype. Therefore, ectopic KRP4 expression might cause a mitotic instability, possibly hampering the proper separation of sister chromatids. KRP4 degradation, during sister chromatid separation, appears to be a prerequisite for normal mitotic cell division (Boruc et al., 2010 and our data). Therefore, we propose that the continuous misexpression of KRP4 in giant cells will drive a cumulative negative effect on mitotic events, resulting in clustered and interconnected nuclei.

Despite the effects of KRP4 overexpression on mitotic events in giant cells, resulting in an aberrant nuclear morphology, these feeding cells could be induced and maintained during nematode parasitism. The competence of the parasitic nematode to alter normal cellular processes might allow mitosis to progress even under abnormal conditions. Nevertheless, under ectopic KRP4 expression, giant cells often contained little cytoplasm, and expansion was hampered. Giant cells constitute the main source of nutrient supply for nematode development and most females associated with these galls were small and did not produce eggs.

The precise functions of KRP4 in plant cells are still unknown. Co-localization of KRP4 with chromosomes during mitosis in normal cells suggests its regulatory role during cell division: for example, mediating the association of CDKA;1 with chromosomes during mitosis (Boruc et al., 2010). Also, Boruc et al. (2010) suggested that KRP4 binds to the CDK/cyclin/CKS (CDK subunit) complex to modulate its activity in a concentration-dependent manner.

Concluding remarks

Despite numerous studies on plant–nematode interactions, morphological and cellular studies have relied mainly on the study of single histological sections of infected roots. Herein, we present a simple procedure that allows the 3D imaging of nuclei in giant feeding cells induced by parasitic root-knot nematodes. This methodology provides new insights into the organization and morphology of complex organelles, such as nuclei, in large biological samples, such as nematode-induced galls. This approach is complementary to histological and in vivo studies. The WM procedure applied here also allowed us to easily identify the distinct nematode life cycle stages associated with each gall. The identification of the precise developmental stage of nematodes is difficult and often impossible when solely based on standard tissue sectioning. Our method allowed us to study the nuclear structure of a cell cycle mutant compared with the wild type and to follow the parasite stage simultaneously. Our data provide insights into KRP4 function and validate functional predictions for this cell cycle inhibitor. We present further evidence that KRP4 is probably involved in chromatid separation during chromosome segregation (Boruc et al., 2010), also illustrating that the WM method proposed, combined with other studies, can help to reveal gene function.

Several strategies are currently used to target nematode parasitism genes as well as the genes involved in nematode metabolism, via RNAi constructs, to inhibit gall formation or nematode development (e.g. Bakhetia et al., 2005; Huang et al., 2006). The WM method presented here has the potential to be a very useful tool for readily acquiring insights into the cellular and nuclear changes occurring in complex multicellular biological specimens, such as nematode-induced galls.


We especially thank Eugenia Russinova for providing the KRP4OE line. We also thank Vanesa Sanchez for help during our initial experiments, and Nathalie Marteu for the production of biological material and Erin Matticola for language correction. P.V. was supported by a doctoral scholarship from Fundação para a Ciência e para a Tecnologia, Portugal (SFHR\BD\41339\2007).