In-vivo analysis of morphogenesis in plants.

Although many molecular regulators of morphogenesis have been identified in plants, it remains largely unknown how the molecular networks influence local cell shape and how cell growth, form, and position are coordinated during tissue and organ formations. So far, analyses of gene function in morphogenesis have mainly focused on the qualitative analysis of phenotypes, often providing limited mechanistic insight into how particular factors act. For this reason, there has been a growing interest in mathematical and computational models to formalize and test hypotheses. These require much more rigorous, quantitative approaches; in parallel, new quantitative and correlative imaging pipelines have been developed to study morphogenesis. Here, we describe a number of such methods, focusing on live imaging.

Molecular aspects of microtubule dynamics in plants.

Microtubules are highly dynamic structures that play a major role in a wide range of processes, including cell morphogenesis, cell division, intracellular transport and signaling. The recent identification in plants of proteins involved in microtubule organization has begun to reveal how cytoskeleton dynamics are controlled.

Cellular basis of shoot apical meristem development.

Shoot apical meristems are composed of proliferating, embryonic type cells, that generate tissues and organs throughout the life of the plant. This review covers the cell biology of the higher plant shoot apical meristem (SAM). The first section describes the molecular basis of plant cell growth and division. The genetic mechanisms, that operate in meristem function and the identification of several key regulators of meristem behavior are described in the second section, and intercellular communication and coordination of cellular behavior in the third part. Finally, we discuss some recent results that indicate interaction between the cellular regulators, such as the cell cycle control genes and developmental regulators.

KNAT2: evidence for a link between knotted-like genes and carpel development.

The KNAT2 (for KNOTTED-like from Arabidopsis thaliana 2) homeobox gene is expressed in the vegetative apical meristem. It is also active during flower development, suggesting a function in the structuring of flowers. To investigate its role, we used a DEXAMETHASONE (DEX)-inducible system to generate transgenic plants that overexpressed a fusion of KNAT2 with the hormone binding domain of the glucocorticoid receptor. DEX-induced plants were similar to plants overexpressing the closely related KNAT1 gene, indicating overlapping functions, although we observed differences as well. In particular, KNAT2-GR activation induced ectopic carpel features. First, KNAT2 induced the homeotic conversion of nucellus into carpel-like structures. Second, KNAT2 induced stigmatic papillae on rosette leaves in the ap2-5 background. Third, ectopic expression of the carpel identity gene AGAMOUS (AG) was observed in carpels and ovules. Interestingly, the homeotic conversion was not dependent on AG activity, because it was maintained in the ag-1 ap2-5 double mutant. Therefore, our data indicate that KNAT2 also must activate other carpel regulators. Together, these results suggest that KNAT2 plays a role in carpel development.

Developmental control of cell division patterns in the shoot apex.

The shoot apical meristem is a group of rapidly dividing cells that generate all aerial parts of the plant. It is a highly organised structure, which can be divided into functionally distinct domains, characterised by specific proliferation rates of the individual cells. Genetic studies have enabled the identification of regulators of meristem function. These factors are involved in the formation and maintenance of the meristem, as well as in the formation of the primordia. Somehow, they must also govern cell proliferation rates within the shoot apex. Possible links between meristem regulators and the cell cycle machinery will be discussed. In order to analyse the role of cell proliferation in development, cell cycle gene expression has been perturbed using transgenic approaches and mutation. The effect of these alterations on growth and development at the shoot apex will be presented. Together, these studies give a first insight into the regulatory networks controlling the cell cycle and into the significance of cell proliferation in plant development.

PIN-FORMED 1 regulates cell fate at the periphery of the shoot apical meristem.

The process of organ positioning has been addressed, using the pin-formed 1 (pin1) mutant as a tool. PIN1 is a transmembrane protein involved in auxin transport in Arabidopsis. Loss of function severely affects organ initiation, and pin1 mutants are characterised by an inflorescence meristem that does not initiate any flowers, resulting in the formation of a naked inflorescence stem. This phenotype, combined with the proposed role of PIN1 in hormone transport, makes the mutant an ideal tool to study organ formation and phyllotaxis, and here we present a detailed analysis of the molecular modifications at the shoot apex caused by the mutation. We show that meristem structure and function are not severely affected in the mutant. Major alterations, however, are observed at the periphery of the pin1 meristem, where organ initiation should occur. Although two very early markers of organ initiation, LEAFY and AINTEGUMENTA, are expressed at the periphery of the mutant meristem, the cells are not recruited into distinct primordia. Instead a ring-like domain expressing those primordium specific genes is observed around the meristem. This ring-like domain also expresses a boundary marker, CUP-SHAPED COTYLEDON 2, involved in organ separation, showing that the zone at the meristem periphery has a hybrid identity. This implies that PIN1 is not only involved in organ outgrowth, but that it is also necessary for organ separation and positioning. A model is presented in which PIN1 and the local distribution of auxin control phyllotaxis.

Gibberellin and ethylene control endoreduplication levels in the Arabidopsis thaliana hypocotyl.

We have previously shown that endoreduplication levels in hypocotyls of Arabidopsis thaliana (L.) Heynh. are under negative control of phytochromes. In this study, the hormonal regulation of this process was analysed using a collection of A. thaliana mutants. The results show that two hormones in particular, gibberellin (GA) and ethylene, play distinct roles. Hypocotyl cells of the GA-deficient mutant ga1-11 grown in the dark did not elongate and showed a greatly reduced endoreduplication. Normal endoreduplication could be restored by supplying 10(-9) M of the gibberellin GA(4+7), whereas the restoration of normal cell growth required 100-fold higher concentrations. The GA-insensitive mutant gai showed reduced cell elongation but normal ploidy levels. We conclude that (i) GA(4+7) has a global positive effect on endoreduplication and (ii) that endoreduplication is more sensitive to GA(4+7) than cell elongation. Ethylene had a completely different effect. It induced an extra round of endoreduplication both in light- and dark-grown seedlings and acted mainly on discrete steps rather than having a global effect on endoreduplication. The genes EIN2 and CTR1, components of the ethylene signal transduction pathway were both involved in this process.

A chromosomal paracentric inversion associated with T-DNA integration in Arabidopsis.

T-DNA integration in the nuclear plant genome may lead to rearrangements of the plant target site. Here we present evidence for a chromosomal inversion of 26 cM bordered by two T-DNAs in direct orientation, which is linked to the mgoun2 mutation. The integration sites of the T-DNAs map at positions 80 and 106 of chromosome I and we show that each T-DNA is bordered by plant sequences from positions 80 and 106, respectively. Although the T-DNAs are physically distant, they are genetically closely linked. In addition, three markers located on the chromosome segment between the two T-DNA integration sites show no recombination with the mgo2 mutation. We show that the inversion cannot be a consequence of a recombination event between the two T-DNAs, but that the integration of the T-DNAs and the inversion were two temporally linked events. T-DNA integration mechanisms that could have led to this inversion are discussed.

Cellular parameters of the shoot apical meristem in Arabidopsis.

The shoot apical meristem (SAM) is a small group of dividing cells that generate all of the aerial parts of the plant. With the goal of providing a framework for the analysis of Arabidopsis meristems at the cellular level, we performed a detailed morphometric study of actively growing inflorescence apices of the Landsberg erecta and Wassilewskija ecotypes. For this purpose, cell size, spatial distribution of mitotic cells, and the mitotic index were determined in a series of optical sections made with a confocal laser scanning microscope. The results allowed us to identify zones within the inflorescence SAM with different cell proliferation rates. In particular, we were able to define a central area that was four to six cells wide and had a low mitotic index. We used this technique to compare the meristem of the wild type with the enlarged meristems of two mutants, clavata3-1 (clv3-1) and mgoun2 (mgo2). One of the proposed functions of the CLV genes is to limit cell division rates in the center of the meristem. Our data allowed us to reject this hypothesis, because the mitotic index was reduced in the inflorescence meristem of the clv3-1 mutant. We also observed a large zone of slowly dividing cells in meristems of clv3-1 seedlings. This zone was not detectable in the wild type. These results suggest that the central area is increased in size in the mutant meristem, which is in line with the hypothesis that the CLV3 gene is necessary for the transition of cells from the central to the peripheral zone. Genetic and microscopic analyses suggest that mgo2 is impaired in the production of primordia, and we previously proposed that the increased size of the mgo2 meristem could be due to an accumulation of cells at the periphery. Our morphometric analysis showed that mgo2 meristems, in contrast to those of clv3-1, have an enlarged periphery with high cell proliferation rates. This confirms that clv3-1 and mgo2 lead to meristem overgrowth by affecting different aspects of meristem function.

Phytochrome controls the number of endoreduplication cycles in the Arabidopsis thaliana hypocotyl.

A majority of the cells in the Arabidopsis hypocotyl undergo endoreduplication. The number of endocycles in this organ is partially controlled by light. Up to two cycles occur in light-grown hypocotyls, whereas in the dark about 30% of the cells go through a third cycle. Is the inhibition of the third endocycle in the light an indirect result of the reduced cell size in the light-grown hypocotyl, or is it under independent light control? To address this question, the authors examined the temporal and spacial patterns of endoreduplication in light- or dark-grown plants and report here on the following observations: (i) during germination two endocycles take place prior to any significant cell expansion; (ii) in the dark the third cycle is completed very early during cell growth; and (iii) a mutation that dramatically reduces cell size does not interfere with the third endocycle. The authors then used mutants to study the way light controls the third endocycle and found that the third endocycle is completely suppressed in far red light through the action of phytochrome A and, to a lesser extent, in red light by phytochrome B. Furthermore, no 16C nuclei were observed in dark-grown constitutive photomorphogenic 1 seedlings. And, finally the hypocotyl of the cryptochrome mutant, hy4, grown in blue light was about three times longer than that of the wild-type without a significant difference in ploidy levels. Together, the results support the view that the inhibition of the third endocycle in light-grown hypocotyls is not the consequence of a simple feed-back mechanism coupling the number of cycles to the cell volume, but an integral part of the phytochrome-controlled photomorphogenic program.

MGOUN1 and MGOUN2: two genes required for primordium initiation at the shoot apical and floral meristems in Arabidopsis thaliana.

We report two new recessive mutations in Arabidopsis, mgoun1 and mgoun2 which cause a reduction in the number of leaves and floral organs, larger meristems and fasciation of the inflorescence stem. Although meristem structure is affected in the mutants, we provide evidence that its overall organisation is normal, as shown by the expression patterns of two meristem markers. Microscopical analyses suggest that both mutations affect organ primordia production. mgo1 strongly inhibits leaf production in a weak allele of shoot meristemless, stm-2. In addition, mgo1 and 2 severely reduce the ability of the fasciata1 and 2 mutants to initiate organs, although meristem formation per se was not inhibited. The strong allele, stm-5, is epistatic to mgo1, showing that the presence of meristematic cells is essential for MGO1 function. These results suggest a role for the MGO genes in primordia initiation although a more general role in meristem function can not be excluded. We describe a form of fasciation which is radically different from that described for clavata, which is thought to have an increased size of the meristem centre. Instead of one enlarged central meristem mgo1 and 2 show a continuous fragmentation of the shoot apex into multiple meristems, which leads to the formation of many extra branches. The phenotype of mgo1 clv3 and mgo2 clv3 double mutants suggest that the MGO and CLV genes are involved in different events. In conclusion, our results reveal two new components of the regulatory network controlling meristem function and primordia formation. A model for MGO genes is discussed.

Endoreduplication and development: rule without dividing?

Endoreduplication, a strategy to amplify nuclear DNA without cell division, is very common but poorly understood in plants. Recent findings in Drosophila provide a first picture of the molecular mechanism, which appears to be conserved between plants and animals. In Arabidopsis, the study of trichomes, leaf epidermis and hypocotyl cells sheds new light on the developmental regulation of this process, and its relation to cell expansion.

Cellular basis of hypocotyl growth in Arabidopsis thaliana.

The Arabidopsis thaliana hypocotyl is widely used to study the effects of light and plant growth factors on cell elongation. To provide a framework for the molecular-genetic analysis of cell elongation in this organ, here we describe, at the cellular level, its morphology and growth and identify a number of characteristic, developmental differences between light-grown and dark-grown hypocotyls. First, in the light epidermal cells show a characteristic differentiation that is not observed in the dark. Second, elongation growth of this organ does not involve significant cortical or epidermal cell divisions. However, endoreduplication occurs, as revealed by the presence of 4C and 8C nuclei. In addition, 16C nuclei were found specifically in dark-grown seedlings. Third, in the dark epidermal cells elongate along a steep, acropetal spatial and temporal gradient along the hypocotyl. In contrast, in the light all epidermal cells elongated continuously during the entire growth period. These morphological and physiological differences, in combination with previously reported genetic data (T. Desnos, V. Orbovic, C. Bellini, J. Kronenberger, M. Caboche, J. Traas, H. Höfte [1996] Development 122: 683-693), illustrate that light does not simply inhibit hypocotyl growth in a cell-autonomous fashion, but that the observed growth response to light is a part of an integrated developmental change throughout the elongating organ.

Plant CDC2 is not only targeted to the pre-prophase band, but also co-localizes with the spindle, phragmoplast, and chromosomes.

A polyclonal antiserum against the p34cdc2 homologue of Arabidopsis thaliana, CDC2aAt, was used in parallel with a polyclonal antiserum against the PSTAIRE motif to study the subcellular localization of CDC2 during the cell cycle of isolated root tip cells of Medicago sativa. During interphase, CDC2 was located in the nucleus and in the cytoplasm. The cytoplasmic localization persisted during the complete cell cycle, whereas the nuclear signal disappeared at nuclear envelope breakdown. At the beginning of anaphase, the anti-CDC2aAt antibody transiently co-localized with condensed chromosomes. The chromosomal co-localization disappeared as anaphase continued and remained excluded from the separated chromosomes until cytokinesis, when CDC2 re-located to the newly forming nuclei. We also observed a co-localization of CDC2 with three microtubular structures, the pre-prophase band, the spindle, and the phragmoplast.

Developmental and cell cycle regulation of alfalfa nucMs1, a plant homolog of the yeast Nsr1 and mammalian nucleolin.

We report here the isolation and characterization of the nucMs1 alfalfa cDNA, whose predicted amino acid sequence structurally resembles the yeast Nsr1 protein and animal nucleolins. These proteins consist of an N-terminal acidic domain, centrally located RNA recognition motifs (RRMs), and a C-terminal glycine- and arginine-rich domain. In comparison with animal nucleolins that contain four RRMs, NucMs1 more closely resembles the yeast Nsr1 protein, which contains only two RRMs. A NucMs1 C-terminal peptide antibody specifically recognized a 95-kD nucleolar protein in alfalfa cells that changed its localization in a cell cycle-dependent manner. The nucMs1 transcript and p95nucMs1 protein levels correlated with cell proliferation, and nucMs1 gene expression was found to be induced in the G1 phase upon mitogenic stimulation of G0-arrested leaf cells. In situ hybridization analysis of different alfalfa organs during various developmental stages showed that nucMs1 gene expression is highest in root meristematic cells, but it is also found in other meristematic cells of the plant body. nucMs1 expression is tightly linked to cell proliferation but does not depend on a particular cell cycle phase. No nucMs1 expression was observed in cells that had exited the cell cycle and were undergoing differentiation or polar growth, indicating that nucMs1 may not be necessary for processes other than cell proliferation.

Procuste1 mutants identify two distinct genetic pathways controlling hypocotyl cell elongation, respectively in dark- and light-grown Arabidopsis seedlings.

Plant morphogenesis is dependent on a tight control of cell division and expansion. Cell elongation during post-embryonic hypocotyl growth is under the control of a light-regulated developmental switch. Light is generally believed to exert its effects on hypocotyl elongation through a phytochrome-and blue-light receptor-mediated inhibitory action on a so far unknown cell elongation mechanism. We describe here a new class of allelic mutants in Arabidopsis, at the locus PROCUSTE1 (prc1-1 to -4), which have a hypocotyl elongation defect specifically associated with the dark-grown development program. Normal hypocotyl elongation is restored in plants grown in white, blue or red light. In agreement with this, the constitutive photomorphogenic mutation cop1-6, which induces a de-etiolated phenotype in the dark, is epistatic to prc1-2 for the hypocotyl phenotype. Epistasis analyses in red and blue light respectively, indicate that phytochrome B but not the blue light receptor HY4, is required for the switch from PRC1-dependent to PRC1-independent elongation. The conditional hypocotyl growth defect is associated with a deformation of the hypocotyl surface due to an uncontrolled swelling of epidermal, cortical or endodermal cells, suggesting a defect in the structure of the expanding cell wall. A similar phenotype was observed in elongating roots, which was however, independent of the light conditions. The aerial part of mature mutant plants grown in the light was indistinguishable from the wild type. prc1 mutants provide a means of distinguishing, for the first time, two genetic pathways regulating hypocotyl cell elongation respectively in dark- and light-grown seedlings, whereby light not only inhibits hypocotyl growth, but also activates a PRC1-independent cell elongation program.

The higher plant Arabidopsis thaliana encodes a functional CDC48 homologue which is highly expressed in dividing and expanding cells.

We have identified an Arabidopsis thaliana CDC48 gene which, unlike the putative mammalian homologue vasolin-containing protein (VCP), functionally complements Saccharomyces cerevisiae cdc48 mutants. CDC48 is an essential gene in S. cerevisiae and genetic studies suggest a role in spindle pole body separation. Biochemical studies link VCP function to membrane trafficking and signal transduction. We have described the AtCDC48 expression pattern in a multicellular eukaryote; the zones of cell division, expansion and differentiation are physically separated in higher plants, thus allowing the analysis of in situ expression patterns with respect to the state of cell proliferation. AtCDC48 is highly expressed in the proliferating cells of the vegetative shoot, root, floral inflorescence and flowers, and in rapidly growing cells. AtCDC48 mRNA and the encoded protein are up-regulated in the developing microspores and ovules. AtCDC48 expression is down-regulated in most differentiated cell types. AtCDC48p was primarily localized to the nucleus and, during cytokinesis, to the phragmoplast, a site where membrane vesicles are targeted in the deposition of new cell wall materials. This study shows that the essential cell division function of CDC48 has been conserved by, at least, some multicellular eukaryotes and suggests that in higher plants, CDC48 functions in cell division and growth processes.

A novel Arabidopsis type 1 protein phosphatase is highly expressed in male and female tissues and functionally complements a conditional cell cycle mutant of Aspergillus.

Type 1 protein phosphatases are very highly conserved throughout eukaryotes where they regulate a number of key metabolic and morphogenetic processes. A cDNA, AtPP1bg, representing a new member of the type 1 protein phosphatase gene family in Arabidopsis has been isolated on the basis of hybridization with the Aspergillus bimG protein phosphatase gene. The AtPP1bg gene potentially encodes a 37 kDa protein very closely related to PP1 but with divergent N- and C-termini. The predicted amino acid sequence shows 71% identity to the ORF of the bimG gene. When expressed in Aspergillus under the alcA promoter, this phosphatase complements the temperature-sensitive bimG11 mutation allowing nearly normal vegetative growth at 37 degrees C (but not at 42 degrees C). Notably, the plant PP1 does not support morphogenesis (conidiation) at 37 degrees C. This may indicate that conidophore formation has particular phosphatase requirement(s) which the plant PP1 cannot supply. The pattern of expression of the AtPP1bg transcript has been studied during development of the plant. In situ hybridization of Arabidopsis with antisense probes shows that this phosphatase gene is expressed at a low level throughout the plant, but is strongly upregulated within developing flowers, especially in the tapetum, the developing and mature pollen and in the ovaries. This implies that the AtPP1bg either has a specialized role in the formation of these organs, or that there is an increased requirement for protein phosphatase 1 at these stages. It was found that the level of AtPP1bg transcript, as judged by the relative intensity of staining in different cells within the floral meristems, did not vary during the cell cycle.

An actin network is present in the cytoplasm throughout the cell cycle of carrot cells and associates with the dividing nucleus.

We have studied the F-actin network in cycling suspension culture cells of carrot (Daucus carota L.) using rhodaminyl lysine phallotoxin (RLP). In addition to conventional fixation with formaldehyde, we have used two different nonfixation methods before adding RLP: extracting cells in a stabilizing buffer; inducing transient pores in the plasma membrane with pulses of direct current (electroporation). These alternative methods for introducing RLP revealed additional features of the actin network not seen in aldehyde-fixed cells. The three-dimensional organization of this network in nonflattened cells was demonstrated by projecting stereopairs derived from through-focal series of computer-enhanced images. F-actin is present in interphase cells in four interconnected configurations: a meshwork surrounding the nucleus; thick cables in transvacuolar strands and deep in the cytoplasm; a finer network of bundles within the cortical cytoplasm; even finer filaments that run in ordered transverse array around the cell periphery. The actin network is organized differently during division but it does not disappear as do the cortical microtubules. RLP stains a central filamentous cortical band as the chromatin begins to condense (preprophase); it stains the mitotic spindle (as recently shown by Seagull et al. [Seagull, R. W., M. Falconer, and C. A. Weerdenburg, 1987, J. Cell Biol., 104:995-1004] for aldehyde fixed suspension cells) and the cytokinetic apparatus (as shown by Clayton, L., and C. W. Lloyd, 1985, Exp. Cell Res., 156:231-238). However, it is now shown that an additional network of F-actin persists in the cytoplasm throughout division associating in turn with the preprophase band, the mitotic spindle, and the cytokinetic phragmoplast.

Gold labeling of microtubules in cleaved whole mounts of cortical root cells.

A method is described for localizing microtubules using gold-labeled antibodies in combination with anti-tubulin. Cortex cells of Equisetum hyemale are broken open while still in buffer, after initially being attached to poly-L-lysine-coated grids. Thus, the cytoplasm becomes accessible to the antibodies. After application of the antibodies, the cleaved cells are post-fixed, stained, dehydrated, and critical point-dried. Different fixation procedures are compared: fixation in paraformaldehyde, in glutaraldehyde, and in glutaraldehyde followed by a sodium borohydride reduction step. All three methods result in good labeling of the microtubules, with low backgrounds. However, organization of the cytoplasm is best preserved in cells fixed in glutaraldehyde without sodium borohydride treatment. The method is highly suitable for studying the membrane-bound cytoskeleton because detergent extraction and/or embedding are avoided.