Root Hair Mutants
|
by: Sebastian Molnar
Introduction Root hairs are tubular outgrowths derived from root epidermal cells, and are involved in nutrient and water uptake, but may also be sites of entry for Rhizobium for the development of nitrogen fixing root nodules (Grierson et al. 1997; Schmidt et al. 1999). Root epidermal cells can have one of two fates: development into a trichoblast (or root-hair forming cell) or development into an atrichoblast (root-hairless cell). These two types of cells are organized on the root surface as longitudinal columns or cell files (Galway et al., 1994; Masucci et al., 1996). Root hair growth may be considered as two separate phases: initiation and elongation. At least two modes of root hair initiation have been recognized. One mode occurs by asymmetric division of the epidermal cell (in which the small cell generates a root hair and the large cell remains hairless), while in the second mode, cell fate is determined by a cell’s position relative to other cells (Galway et al., 1997). In Arabidopsis epidermal cells, the site of initiation (identified by a localized bulge in the cell wall) for a developing root hair normally occurs in the apical end, which suggests that trichoblasts are polarized (Schiefelbein and Somerville, 1990; Masucci and Schiefelbein, 1994). The elongation stage of root hair formation arises through tip growth. Studying root hairs is useful for understanding fundamental biological problems, such as cell polarity and positional-determinism. For example, tip growth is characteristic of root hairs, pollen tubes, and fungal hyphae (Heath, 1995). There is strong evidence for tip-high ion concentration gradients in these systems (Schiefelbein and Somerville, 1990; Schiefelbein et al., 1992; Heath, 1995), indicating that cell polarity is a widespread and important developmental requirement. As mentioned above, certain types of cells differentiate in response to positional cues, thus positional-determinism is essential for developmental programs (van den Berg et al., 1995). Since positional controls are known to exist in the root (i.e. trichoblast/atrichoblast determinism), and since single cells may be analyzed during morphogenesis, root hairs provide a simple model system that is easily accessible for answering many questions in developmental biology and cellular differentiation. Molecular genetic approaches have revealed several genes that are involved in root hair development. Some genes are required for root hair initiation, such as RHD1 (ROOT HAIR DEFECTIVE 1; Schiefelbein and Somerville, 1990), RHD6 (Masucci and Schiefelbein, 1994) and RHL1 (ROOTHAIRLESS1; Schneider et al., 1998). Other genes are responsible for elongation, such as RHD2, RHD3, RHD4 (Schiefelbein and Somerville, 1990), Tip1 (Schiefelbein et al., 1993), and COW1 (Grierson et al., 1997). Some genes are involved in determining cell fate, such as GL2 (Masucci et al., 1996), TTG (TRANSPARENT TESTA GLABROUS; Galway et al., 1994), and CPC (Wada et al., 1997). Many details of the mechanisms behind root and root hair development are as yet, unknown. However, through identifying the genes and spatial cues involved in the process, developmental programs may be better understood. In this paper, two aspects of root hair development will be examined. The first part will focus on root hair elongation, and in particular, with respect to a globally expressed gene, RHD3. The second part will deal with a genetic and structural system involved in determining root epidermal cell fate. RHD3 Through phenotypic screening of roots and root hairs in Arabidopsis mutants, Schiefelbein and Somerville (1990) designated RHD3 (ROOT HAIR DEFECTIVE 3) homozygous mutants as having short roots and short wavy root hairs (in comparison to wild type roots). The wavy appearance of RHD3 was suggested to be the result of an asymmetrical deposition of cell wall material at the root hair tip during growth (Schiefelbein and Somerville, 1990). The RHD3 plant itself is smaller in size than the wild type, indicating that RHD3 is not root hair specific. Also, RHD3 is not responsible for specifying cell type, as the position-dependent pattern of trichoblasts and atrichoblasts is unaffected in RHD3 (Wang et al. 1997). Analysis of different plant tissues in rhd3 mutants revealed that the gene is required for proper cell expansion (Wang et al., 1997; Galway et al., 1997). Thus, it was confirmed that RHD3 is not root hair specific, but rather, has a global effect on the plant with the consequence of reducing cell size. It has been suggested that RHD3 may participate in a hormonal response which influences cell expansion, since RHD3 plants were found to be insensitive various plant hormones such as IAA, ACC and 2,4-D (Wang et al. 1997). As mentioned above, it was suggested that the wavy appearance of RHD3 mutants may be due to an uneven deposition of cell wall material during elongation. Cell enlargement is accompanied by various cell wall modifications, and it serves to increase surface area of the cell (Wang et al. 1997). During normal tip growth, components of the cell wall are brought in by vesicle transport, however, in rhd3 mutants vesicle distribution is different from that of the wild type. Secretory vesicles in rhd3 root hairs were found by Galway et al. (1997) to be present in higher concentrations at subapical regions, whereas in wild type root hairs, vesicles are localized primarily in the apical regions. Thus, asymmetrical vesicle distribution in RHD3 is likely to be responsible for the wavy phenotype. The major defect in RHD3 mutants is the reduced cell size. For example, rhd3 root hairs are less than one third of the volume of wild type hairs (Galway et al., 1997). Cell expansion in wild type plants occurs with an increase in vacuole size (e.g. through uptake of water) and a decrease in cytoplasm volume. In contrast, the relative amount of cytoplasm in rhd3 plants was shown by Galway et al. (1997) to be higher. What, then, is the role of the RHD3 gene product in the molecular mechanisms behind cell expansion? Wang et al. (1997) have cloned and sequenced the RHD3 gene, and found that it encodes an 89-kDa protein (802 amino acids) with putative GTP-binding motifs near the amino-terminal region. Three consensus sequence elements spaced at 40-80 amino acids were identified by Dever et al. (1987). These include GXXXXGK, DXXG, and NKXD. The first two elements are involved in binding with the phosphate moiety of GTP, while the third element is responsible for nucleotide specificity (McCormick, 1985; Dever et al 1987). The RHD3 gene was found to have the first two GTP-binding consensus sequences, but not the third. As indicated by Dever et al. (1987), the third consensus element (NKXD) is necessary for GTP binding: when NKXD is replaced by NKXW, both ITP (an inosine nucleotide) as well as GTP can be bound by the protein. Also, Clanton et al. (1986) have shown that when a tyrosine (116K) or a lysine (116Y) replaces the asparagine (116N) in NKXD of ras p21, GTP binding is abolished. Asparagine is a polar uncharged amino acid, as is tyrosine, but tyrosine is a much larger molecule (due to the benzene ring) than asparagine. This could explain why GTP is not bound with such an amino acid substitution -- i.e. the conformation of the site is not quite right for binding GTP. Lysine, on the other hand, is a basic amino acid, and its long flexible sidechain can form hydrogen bonds to the peptide carbonyl groups of aspartic acid and of histidine, therefore creating a hydrophobic pocket (la Cour et al., 1985). This may explain the inability of GTP-binding when the asparagine is substituted by a lysine. As indicated above, the NKXD element seems to be an essential component in GTP-binding (i.e. it is required for GTP specificity).This third consensus sequence is conserved in many GTP-binding proteins, however, in some cases (e.g. PEPCK and GTP:ATP phosphotransferase) the aspartic acid residue is replaced by tryptophan (Dever et al., 1987). Since RHD3 lacks this third element, the question is raised as to whether the gene really encodes a GTP-binding protein. Dever et al. (1987) have indicated that a protein with the first two elements (GXXXXGKS and DXXG) is likely to bind to phosphates, however, without the third element, specific GTP binding could not be predicted by sequence comparison. On the other hand, Dever et al. (1987) have noted that although their consensus sequence elements are based on several known GTP-binding proteins, some known GTP-binding proteins (e.g. alpha- and beta-tubulin) failed to match those criteria. One criterion in predicting GTP-binding is, not just the presence of the consensus sequences, but also the spacing of those sequences (Dever, et al. 1987). On average, the spacing between the first and second, and between the second and third sequences, ranges from 40 to 80 amino acids. In RHD3, the putative GTP-binding domains (GXXXXGKS and DXXG) are spaced at 42 amino acids (Wang et al. 1997). There are some GTP-binding proteins which do not strictly follow this spacing rule, such as GTP:AMP phosphotransferase, G protein, and transducin (Dever et al. 1987). Therefore, there may be several different classes of GTP-binding proteins that are possible, which do not follow the “standard rules”. In database searches, Wang et al. (1997) identified two new motifs in RHD3 and in related proteins from rice, yeast, and a protozoan (showing strong homology in the amino-terminal regions), but as yet, the function of these motifs have not been determined. The two novel motifs (FVIRD and NKDLDLP) are conserved in RHD3 and RHD3-like proteins, and are present within the first 300 amino acids from the amino-terminus (Wang et al. 1997). NKDLDLP bears a resemblance to the third GTP-binding consensus sequence, NKXD. Therefore, there is a possibility that this new consensus sequence (NKDLDLP) may be responsible for GTP specificity in RHD3 and RHD3-like proteins; or, it may be that FVIRD and NKDLDLP somehow interact with each other for GTP-binding. Since some GTP-binding proteins (e.g. GTPases) have been shown to participate in vesicle trafficking, and since RHD3-like proteins from different eukaryotic groups show significant homology in their amino-terminal regions, Wang et al. (1997) have postulated that RHD3 may represent a new class of GTP-binding proteins required in the regulation of cell enlargement. At present, RHD3 has only been shown to be a GTP-binding protein through indirect means (e.g. through consensus sequence comparisons). Structural analysis (e.g. through x-ray crystallography) and modeling of the RHD3 gene product would confirm whether this protein binds GTP -- and if it does, how it binds -- and may reveal its physiological role in cell expansion. Wang et al. (1997) have indicated that hydropathy plots and secondary structure analysis show RHD3 to be largely hydrophilic and that there is a hydrophobic region near the carboxyl-terminus, however, they did not provide their analysis for this. Hydropathy profiles and transmembrane predictions of RHD3 are shown in the appendix. The hydropathy profile using SOSUI detected the presence of one transmembrane region, 23 amino acids long, between positions 687 and 709 (Appendix p1) (ref. #). On the other hand, a transmembrane prediction using TMpred revealed two possible models: a strongly preferred model containing four transmembrane helices, and an alternate model with three (Appendix p2) (ref. #). In both models, a region between the 155 to 131 position was detected as a potential transmembrane segment. A potential transmembrane region in the strongly preferred model between positions 731 to 750 was also detected. Both of these regions may be excluded as transmembrane segments, due to a low probability found using a similar method at TMHMM (Appendix p3) (ref. #). Therefore, it seems likely that there is at least one, and possibly two, transmembrane segments near the C-terminal region in RHD3, between positions 676 and 721. Also, using ScanProsite to search for functionality, identifies RHD3 as an ATP/GTP binding protein, however, only the first GTP-binding consensus sequence, GPQSSGKS, appears. On the other hand, a BLAST search (ref. #) reveals that RHD3 has a strong similarity to a putative GTP-binding protein, and some similarity to ORFs in yeast and in Entamoeba. These results suggest that RHD3 is a transmembrane GTP-binding protein, however, x-ray crystallography for structural analysis would be required for confirmation. As suggested by Wang et al. (1997), this type of GTP-binding protein may be widespread across the eukaryotes. A possible physiological role may be that the RHD3 protein is bound (by its transmembrane region) to the plasma membrane and binds GTP for signal transduction during cellular expansion. GL2 and TTG GLABRA-2 (GL2) has been shown to encode a protein with a homeodomain, and is involved in both trichome and root hair development (Di Cristina et al., 1996). Homeobox genes contain a 180 bp evolutionarily conserved sequence (i.e. a 60 amino acid motif, referred to as a homeodomain), and are involved in the development of phylogenetically distant organisms, such as animals, fungi, and plants (Chan et al., 1998). Homeodomains fold into a sequence-specific DNA-binding structure characterized by three alpha-helices separated by a loop and a turn. Several families of plant homeoboxes have been identified (for review, see Chan et al., 1998). The classification of GLABRA2 into one of these homeobox families has been under debate (Chan et al., 1998). Some have classified GL2 as an HD-Zip protein, whereas others have suggested that GL2 and GL2-like proteins have distinct characteristics, therefore warranting a class that is separate from HD-Zip (Chan et al., 1998). In any case, GL2 has been shown to dimerize for DNA-binding via a leucine zipper (Di Cristina et al. 1996), and it is responsible for controlling cell identity in a position-dependent manner during the differentiation of root epidermal cells (Masucci et al. 1996). How GL2 may be involved in the regulation of cell fate has been examined through observing the effects of GL2 mutants on development, and through elucidating the interactions of other proteins on GL2 expression. TRANSPARENT TESTA GLABROUS (TTG) is a protein that has been implicated in upregulating GL2 expression (Hung et al., 1998). In wild type plants, hairless cells develop when they are directly over cortical cells (therefore the epidermal cell contacts just one cortical cell), while root-hair cells develop when they are located above radial walls between cortical cells (and therefore contact more than one cortical cell) (Galway et al., 1994). In recessive ttg mutants (which exhibit a lack of trichomes, anthocyanins, and seed coat mucilage), this positional determinism is lost and root hair cells can develop in nearly all locations on the root epidermis (Galway et al. 1994). Thus, TTG was postulated as being required to maintain positional-dependent signalling for the differentiation of epidermal root cells into hairless cells. It was later demonstrated, through using ttg mutants and GL2:GUS fusions, that TTG is important in the regulation of GL2 expression (Hung et al. 1998). Low GL2 expression brought on by nonfunctional ttg mutants allows for inappropriate root-hair formation. Thus, TTG is responsible for maximal GL2 expression and inhibits the development of root hair formation in cells located over cortical cells, but it is not required for the cell-type or spatial pattern of GL2 expression. It seems that TTG and GL2 encode general transcription factors for early cell type specification (Schiefelbein et al., 1997) with TTG as a positive regulator of GL2 (Hung et al. 1998). As mentioned earlier, root epidermal cells are one of two types: root-hair cells or root-hairless cells. These two types of cells are organized on the root surface as longitudinal columns or cell files, and epidermal cell fate is determined by a position-dependent mechanism (Galway et al., 1994; Masucci et al., 1996). Trichoblasts and atrichoblasts can be distinguished morphologically (e.g. there is a delay in vacuolation in root-hair forming cells which is not present in hairless cells) and biochemically (e.g. by toluidine blue staining) before becoming mature differentiated cells, therefore implying that cell fate is programmed in at an early stage in development (Galway et al., 1994). Masucci et al. (1996) have shown, using GUS reporter genes and in situ hybridization, that GL2 is preferentially expressed in differentiating hairless cells at an early stage. However, recessive gl2 mutations were found to not cause a complete cell type conversion from hairless to root-hair cells (i.e. two types of root hair cells were found in gl2 mutants: one type was identical to normal root-hair cells, while a second type had characteristics of both root-hair cells and hairless cells; Masucci et al. 1996). Thus, GL2 is a homeobox gene which controls a subset of genes for the inhibition of root hair formation, however, other genes may act earlier than GL2 to control other aspects of root epidermal cells (Masucci et al., 1996). It has been demonstrated that GL2 expression is regulated in a position-dependent mechanism for the control of cell-type patterning in both the root and hypocotyl (Hung et al., 1998). The patterning of stomata in the hypocotyl of Arabidopsis is influenced by ttg and gl2 mutations (i.e. the number of stomatal cells in the hypocotyl is increased; Hung et al., 1998). This position-dependent pattern is characterized through the differentiation of epidermal cells, into hairless cells in the root and nonstomatal cells in the hypocotyl, which are located outside of periclinal cortical cell walls, and is mainatined by GL2 and TTG (Hung et al., 1998). On the other hand, GL2 and TTG were found to have opposite effects for root-hair cells and trichomes. Thus, it has been suggested that TTG and GL2 act as positive regulators in trichome development (i.e. trichome formation is induced), and as negative regulators of root-hair and stomatal development (e.g. root hair formation is repressed, or alternately, hairless cell development is induced; Di Cristina et al., 1996; Hung et al., 1998). Conclusion. GL2 and TTG have been used to determine the relationships of other genes which have been identified as influencing root hair development (Wada et al., 1997; Wang et al., 1997; Schneider et al., 1998; Bernhardt and Tierney, 2000). For example, double mutants of rhd3 ttg and rhd3 gl2 were used by Wang et al. (1997) to determine whether RHD3 affects cell-type specification. Both double mutants generated additive effects (i.e. root hair cells appeared as short and wavy at all positions on the root epidermis), indicating that RHD3 acts independently of GL2 and TTG. As mentioned above, GL2 is preferentially expressed in differentiating hairless cells in wild type plants, and in rhd3 mutants, GL2:GUS reporter gene construct showed similar expression patterns to that in the wild type, confirming that RHD3 is independent from GL2 (Wang et al. 1997). In different study, a GUS construct with a root specific gene encoding a structural cell wall protein (AtPRP3) was found to have enhanced expression in gl2 and ttg recessive mutants, indicating that the AtPRP3 protein functions to contribute to root hair cell wall structure (Bernhardt and Tierney 2000). The general trend in using the methodologies such as those mentioned above, is the elucidation of developmental pathways which lead to cell polarity and positional-determinism, and ultimately to understanding cellular differentiation. Plant hormones, such as auxin and ethylene, play a major role in plant growth, however, the mechanisms of their action are poorly understood (Abel and Theologis, 1996). Some genes for root hair development, such as AtPRP, are regulated by auxin- and ethylene and therefore may be useful in analyzing hormonal pathways (Masucci and Schiefelbein, 1994; Bernhardt and Tierney, 2000). Signal transduction is a widespread -- for example, the intercellular signalling that must occur in position-dependent cellular differentiation -- and therefore, dissecting the developmental pathways which utilize such signalling would allow for greater understanding of the process. Homeobox genes, such as GL2, and their regulation, may remain evolutionarily conserved in diverse organisms. As such, studying a simple model system such as root hair development could provide insight into other, more complex systems.
References:
|
Get your own Free Home Page