SEARCH

SEARCH BY CITATION

Keywords:

  • actin;
  • AIP1;
  • AIP1 overexpression;
  • Arabidopsis;
  • cytoskeleton;
  • plant

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • • 
    Actin organization and dynamics are essential for cell division, growth and cytoplasmic streaming. Here we analyse the effects of the overexpression of Actin Interacting Protein 1 (AIP1) on Arabidopsis development.
  • • 
    Arabidopsis plants were transformed with an ethanol-inducible AIP1 construct and the characteristics of these plants were analysed after induction.
  • • 
    When AIP1 was increased to approx. 90% above wild-type values, root hair development and actin organization in all cell types examined were disrupted.
  • • 
    Our data demonstrate that AIP1 is a key regulator of actin organization and that its regulation is essential for normal plant cell morphogenesis.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The actin cytoskeleton is involved in the regulation of numerous cellular processes. It is a dynamic structure and the actin organization in a cell is determined by these dynamics. Although actin filaments can polymerize and depolymerize in the absence of actin-modulating proteins in vitro, F-actin dynamics within a cell is dependent on the cooperative activity of large numbers of actin-modulating proteins. One of the proteins that is involved in regulating the dynamics of the actin cytoskeleton is Actin Interacting Protein 1 (AIP1). AIP1 caps actin filaments weakly and enhances the actin-severing activity of the actin-modulating protein Actin Depolymerizing Factor (ADF). AIP1 is a conserved WD repeat protein that has been found in fungi, slime moulds, nematodes, frogs, mammals and plants (for review, see Ono, 2003). In Arabidopsis thaliana, two copies of AIP1 have been found (AIP1-1: At2g01330; and AIP1-2: At3g18060). Northern blotting shows that AIP1-1 is floral-specific and AIP1-2 is present throughout the plant (Allwood et al., 2002). Also, affymetrics data demonstrate pollen-specific expression of AIP1-1 and vegetative tissue expression of AIP1-2.

The function of AIP1 has been studied in a number of organisms, mainly through the analysis of gene knockouts. In the yeast Saccharomyces cerevisiae, AIP1 knockout does not cause a cell growth phenotype (Iida & Yahara, 1999; Rodal et al., 1999), although, together with ADF, AIP1 promotes turnover of tropomyosin-decorated actin filaments (Okada et al., 2006). However, an aip1 null mutation in the slime mould Dictyostelium causes reduction in cellular processes that require a dynamic actin cytoskeleton, such as growth rates, cytokinesis, fluid phase uptake, phagocytosis and cell motility (Konzok et al., 1999). In the nematode Caenorhabditis elegans, a null mutation in one of the two AIP1 genes causes defects in muscle contractile activity, which correlates with severe disorganization of actin filaments and accumulation of actin aggregates (Ono, 2001). Recently, we showed that inhibition of AIP1 in Arabidopsis leads to dramatic developmental defects and actin accumulation (Ketelaar et al., 2004). Although data are available on inhibition of AIP1 in a number of species, overexpression of AIP1 has only been studied in embryos of the frog Xenopus laevis. Microinjection of purified AIP1 in blastomeres arrests cleavage and subsequent development by inhibition of accumulation of cortical actin during cleavage furrow formation (Okada et al., 1999). Since AIP1 inhibition in Arabidopsis gives a severe developmental phenotype and causes severe actin disruption, the presence of functional AIP1 in plants may be more important than in other species. This prompted us to ask what would be the consequences of overexpressing AIP1 with regard to plant development and actin organization.

Here we show that inducible AIP1 overexpression in Arabidopsis causes defects in actin organization that are different from the defects caused by AIP1 inhibition. Moreover, overexpression of AIP1 causes defects in root hair development. Our results show that not only is AIP1 essential during plant development, but that tight regulation of AIP1 expression is essential for proper actin organization and plant morphogenesis.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Cloning

The HA-AIP1 fusion was generated by amplifying full-length AIP1-1 (At2g01330). In the amino-terminal primer, an HA-tag sequence and a Gateway attB1 sequence was included (GGGGACAAGTTTGTACAAAAAAGCAGGCTCAGCCACCATGTACCCATACGATGTTCCAGATTACGCTATGGCGAAACTCCTCGAGACCTTTCC). The carboxy terminal primer contained a Gateway attB2 sequence (GGGGACCACTTTGTACAAGAAAGCTGGGTCTCACTGAGGTTCGATATGCCACAAAC). The resulting PCR product was cloned into pDONR207 (Invitrogen, Paisley, UK) and transferred to the p35S:alcR-RFA plasmid (described later).

The p35S:alcR plasmid was made Gateway-compatible by replacing the CAT gene in the pAlcA:CAT plasmid (Caddick et al., 1998) by the Gateway cloning cassette RFA. Subsequently, the alcA:RFA:nos cassettes were subcloned into the p35S:alcR (Caddick et al., 1998) cassette. The resulting p35S:alcR-RFA plasmid was used to introduce the HA-AIP1 cassette by a Gateway LR reaction.

The p35S:alcR-HA-AIP1 plasmid was transformed into Agrobacterium tumefaciens strain C58C3, and Arabidopsis Col-0 plants were transformed by floral dipping (Clough & Bent, 1998).

Plant culture, growth conditions and ethanol induction

Wild-type Arabidopsis Col-0 was used for all experiments and transformations. Ethanol-inducible AtAIP1-1 overexpression plants were selected by growing the T1 seeds on solid half-strength M & S medium, complemented with 50 µg ml−1 kanamycin. Kanamycin-resistant T2 seeds were used in all experiments. GFP, fused to the second actin binding domain of Arabidopsis fimbrin (GFP:FABD), serves as a live cell actin marker (Ketelaar et al., 2004). GFP:FABD was introduced into AtAIP1 overexpression lines by a cross with a Col-0 line, homozygous for a GFP:FABD insert. Seeds were potted in general-purpose compost and sand (4 : 1) and grown in a glasshouse. AIP1 overexpression was induced by watering the plants twice weekly with 0.5% ethanol. To observe the effects on root development, seeds were surface-sterilized by soaking in 70% ethanol for 1 min, followed by 10% bleach plus 0.05% Triton X-100 for 10 min. After sterilization, seeds were washed three times in sterile distilled water. Subsequently, a thin layer of solid medium (Wymer et al., 1997), covered by a 30 × 24 mm piece of biofoil (Vivascience via Merck, Poole, UK) was placed on a coverslip (50 × 24 mm). The solid medium that was not covered by biofoil was removed by making a straight cut with a sharp knife. Seeds were placed against the cut agar surface and allowed to germinate into the solid medium. 0.7% Plant agar (Duchefa, Haarlem, the Netherlands) was used to solidify the medium. The slides with seedlings were rested with the seeds towards the upper side on 1 ml pipette tips and contained in 70 mm Petridishes, wrapped with parafilm. Plants were cultured at 22°C in a long daylight regime (16 h light, 8 h dark) for 5 d. The AIP1 overexpression was induced by addition of 0.25% ethanol to the medium before placing the seeds.

Western blotting and quantification of AIP1 expression

Tissues were frozen in liquid nitrogen and ground in a mortar with a pestle. The frozen material was directly added to boiling SDS-PAGE loading buffer and boiled for 3 min. After boiling, samples were kept on ice and spun in a tabletop centrifuge for 1 min at maximum speed. Samples were fractionated on one-dimensional gels, and blotted onto nitrocellulose membranes as described by Hussey et al. (1988). The membrane was probed with polyclonal mouse anti-AtAIP1-1 antiserum (Allwood et al., 2002) diluted 1 : 200, or monoclonal mouse anti-α-tubulin clone DM1A (Sigma, St Louis MO, USA) diluted 1 : 1000 in TBS-T (10 mm Tris, pH 7.4, 140 mm NaCl, and 0.05% Tween 20) buffer including 1% (w/v) fat-free dried milk. Anti-mouse secondary antibody conjugated to horseradish peroxidase (Amersham Pharmacia, Piscataway, NJ, USA) was used 1 : 5000 diluted in TBS-T buffer with 1% (v/w) fat-free dried milk. Protein bands cross-reacting with anti-AtAIP1-1 or α-tubulin were visualized using an chemoluminescence detection kit (Amersham Pharmacia). The average intensity of bands was measured using Photoshop 7.0 (Adobe Systems, Inc., San Jose, CA, USA). The intensity value of the tubulin lane in the control was used as a reference (100) and all intensity values were expressed as a proportion of the control. The standardized AIP1 amount was divided by the standardized tubulin amount of the same treatment. The value of the control was set to 100% and the other values were adjusted similarly. Unsaturated images were used for all calculations.

Imaging

A Zeiss LSM510 META system was used for confocal imaging. Green fluorescent protein (GFP) fluorescence was visualized using the Argon-ion laser at 4% of the maximum power, combined with main dichroic beamsplitter HFT 488 nm and a BP 505–530 nm emission filter.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

AIP1 protein quantity

Arabidopsis plants harbouring ethanol-inducible, HA-tagged AIP1 overexpression constructs were generated. We constructed an N-terminal fusion of an HA-tag to AIP1-1 and transformed Arabidopsis Col-0 with this construct. In two independent transgenic lines, we tested for HA-AIP1 expression 3 d after induction with 0.5% ethanol by probing a protein gel blot with an anti-HA antibody. Both lines harboured a protein of approx. 67 kDa, which correlates with the mass of the chimeric HA-AIP1 protein (Fig. 1). We tested the amount of AIP1 overexpression by comparing the AIP1 expression between the different lines on a protein gel blot, using polyclonal antiserum raised against Arabidopsis AIP1-1, which detects both Arabidopsis AIP1-1 and AIP1-2 (Fig. 1) (Allwood et al., 2002). The intensity values were corrected for the protein amounts of tubulin as a loading control. The AIP1 amounts in the two HA-AIP1 lines were, respectively, 89 and 47% up-regulated compared with wild-type Col-0 (Fig. 2).

image

Figure 1. Western blot of Arabidopsis uninduced and induced HA-AIP1 lines, probed with anti HA-tag, AIP1 and α-tubulin antibodies.

Download figure to PowerPoint

image

Figure 2. Quantification of AIP1 expression in the two overexpressing lines, compared with an uninduced line of Arabidopsis. The HA-AIP1-overexpressing lines show an increase in the total amount of AIP1 of 89 and 47%.

Download figure to PowerPoint

Root hair growth is disturbed when AIP1 expression is increased

We tested the plants overexpressing AIP1 (the HA-AIP1 lines) for changes in shape or development by two different methods. We grew plants on soil until a rosette had formed. Subsequently AIP1 overexpression was induced by watering the plants twice weekly with 25 ml of water with 0.5% ethanol per plant. This method was designed to analyse aerial organs and is similar to the method we previously used to analyse AIP1 RNAi lines (Ketelaar et al., 2004). Alternatively, we placed seeds on sandwiches of solid medium (Wymer et al., 1997) between a cover slip and biofoil (Ketelaar et al., 2004), containing 0.25% ethanol, and allowed them to germinate in the presence of ethanol. This method is suitable to study defects in root and root hair development. The optimal concentrations of ethanol had been determined in previous work (Ketelaar et al., 2004). Both lines showed similar responses to HA-AIP1 induction and the results from HA-AIP1 line 1 are shown. We did not observe any significant differences between the AIP1-overexpressing plants and the Col-0 wild-type plants grown on soil and after induction with ethanol. Not only did the stature of the plants not differ, but the shapes and sizes of individual cell types, such as trichomes, leaf epidermal cells and stomata, also did not differ significantly (not shown). The root length and width of plants grown in the sandwiches were also unchanged. However, root hairs appeared short and slightly swollen, with occasional branches when compared with wild-type root hairs (Fig. 3a,b). The average root hair width was increased from 11.62 (± 0.59 µm SD) in uninduced HA-AIP1 control root hairs to 19.10 (± 1.48 µm SD) in the tubes and 26.87 (± 4.18 µm SD) in swollen tips of root hairs that developed on HA-AIP1-overexpressing roots. The increases in width of both the root hair tubes and the tips are significantly different compared with uninduced root hairs (Kolmogorov–Smirnov test). The root hair length was reduced from 403.4 (± 39.8 µm SD) in uninduced HA-AIP1 control root hairs to 52.0 (± 21.9 µm SD) in HA-AIP1 overexpressing roots that were growing in the presence of 0.25% ethanol. This difference was also statistically relevant, as assessed using the Kolmogorov–Smirnov test. We did not detect any changes in root hair density, which indicates that all root epidermal cells that are designated to form a root hair do so.

image

Figure 3. Root hairs are thicker, swollen and shorter in Arabidopsis HA-AIP1-expressing lines compared with wild-type plants. (a) A root fragment of a wild-type plant with long root hairs of more or less the same diameter. (b) A root fragment of an HA-AIP1-overexpressing plant. Root hairs are short, thick and swollen, but the root hair density and root cell lengths are not affected. Bars, 100 µm.

Download figure to PowerPoint

The actin configuration in AIP1-overexpressing lines

We tested for defects in actin organization in the HA-AIP1-1 lines that overexpress AIP1. To visualize the actin cytoskeleton in live plants, we crossed the AIP1 overexpression lines with plants expressing GFP, fused to the second actin binding domain of Arabidopsis fimbrin (GFP-FABD) (Ketelaar et al., 2004). First, we focused on the actin organization in root hairs, since these developed abnormally. We were not able to image growing root hairs of the AIP1-overexpressing lines; therefore we focused on fully grown root hairs. The actin cytoskeleton in fully grown wild-type root hairs consists of several actin cables that run through the root hair tube more or less longitudinally and loop through the tip (Fig. 4a). In the root hairs of HA-AIP1-overexpressing lines, the number of actin bundles was larger and the degree of alignment with the longitudinal axis of the root hair seemed somewhat reduced (Fig. 4b–f).

image

Figure 4. The actin cytoskeleton in root hairs of Arabidopsis HA-AIP1-expressing lines consists of many thick bundles of F-actin. (a) The actin cytoskeleton of a fully grown wild-type root hair. (b–f) The actin cytoskeleton in several differently shaped root hairs of an HA-AIP1-expressing plant. Bars, 15 µm.

Download figure to PowerPoint

Although there is no detectable phenotype in aerial organs, we analysed the actin organization in stem epidermal cells. Owing to their shape and organization, these cells are ideal for detecting changes in actin organization. In wild-type cells, bundles of actin filaments run through the cortical cytoplasm, transvacuolar strands and perinuclear cytoplasm. Most of these bundles are oriented more or less longitudinally to the cell axis (Fig. 5a). In cells of induced HA-AIP1-overexpressing lines, the bundles of actin filaments appear finer and their orientation has changed from more or less longitudinal to more transverse, with a larger variation in orientation (Fig. 5b). We measured the angle of actin cables to the long axis of cells and found that the angle in HA-AIP1-overexpressing lines was higher than in uninduced HA-AIP1 control lines (Fig. 6).

image

Figure 5. The actin organization in stem epidermal cells is altered in the Arabidopsis HA-AIP1-overexpressing lines. (a) The stem epidermis of a control plant; (b) the epidermis of an induced HA-AIP1 plant. The organization of the actin is modified in that bundles appear thinner and the orientation is changed. Bars, 50 µm.

Download figure to PowerPoint

image

Figure 6. The average absolute angle of actin bundles to the long axis of stem epidermal cells is increased in the Arabidopsis HA-AIP1 lines. Error bars indicate SD.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We have shown that AIP1 overexpression leads to changes in plant development and actin configuration. We have used an ethanol-inducible system to avoid problems with adaptation of the plant to higher quantities of AIP1. The phenotypes of the two transgenic lines overexpressing HA-AIP1 resembled each other, even though total AIP1 expression values were somewhat different. In ethanol-induced AIP1 RNAi lines, the actin cytoskeleton in shoot epidermal cells is dramatically altered. A thick mass of actin lies in the middle of the cell and no individual bundles can be found (Ketelaar et al., 2004). AIP1 overexpression does not cause this actin accumulation, but it does cause a different type of disorganization of the actin cytoskeleton. The more or less longitudinal bundles of actin disappear and are replaced by finer actin cables that are more transversely oriented. This fits well with the in vitro data that are available on Arabidopsis AIP1 (Allwood et al., 2002). Actin turnover increases so that there is less polymerized actin. For this reason, actin may not organize properly, and shorter, thinner actin cables appear. It is surprising that a dramatically altered actin organization does not lead to changes in expansion rates or expansion direction. This could indicate that changes in actin organization only occur after the cells have stopped expanding, or that cell growth is not affected by the altered actin organization. It is not possible to see in which cells the HA-AIP1 overexpression construct is switched on after induction with ethanol. In ethanol-inducible AIP1 RNAi lines, this is not a problem, since RNAi spreads systemically throughout the plant (Meins et al., 2005). In HA-AIP1 overexpression lines, the HA-AIP1 fusion protein does not spread systemically, and will only appear in cells that do express the fusion protein. From fluorescent protein expression studies, we know that the ethanol-inducible construct predominantly expresses in fully grown cells, with the exception of root hair forming root epidermal cells (T. Ketelaar & P. J. Hussey, unpublished). This suggests that most cell types are fully grown by the time HA-AIP1 overexpression is induced. This also suggests that the HA-IAP1 expression in induced cells may be higher than the increase that we observed on protein blots, since in some cell types, HA-AIP1 overexpression may not be induced.

HA-AIP1 overexpression induces formation of short, swollen root hairs

Can the dramatic morphological changes in root hairs by HA-AIP1 overexpression be attributed to the specific mode of expansion of root hairs? Root hairs are tip-growing cells. Tip-growing cells expand very rapidly over a very small area of their surface, so that a tubular structure is formed. The only plant cells that expand by tip growth are root hairs and pollen tubes. Since the construct we have used does not express in pollen, only root hairs display a phenotype. It has been demonstrated that root hair growth is extremely sensitive to actin depolymerization (Miller et al., 1999; Ketelaar et al., 2003); root hair growth is inhibited by concentrations of 1 µm cytochalasin D (CytD) and the diameter of the tip of growing root hairs increases at submicromolar concentrations of CytD. Cells that expand over a larger cell surface area are less sensitive to actin-depolymerizing drugs, as root growth continues at the concentrations of CytD already described (unpublished results). Therefore, limited changes in actin dynamics are more likely to cause a root hair phenotype than a phenotype in other cell types. The observed phenotype in root hairs fits well with data from the literature. AIP1 enhances the activity of ADF (Allwood et al., 2002), which leads to increased actin depolymerization. The swollen root hairs resemble the swollen tips of root hairs during CytD-mediated actin depolymerization (Ketelaar et al., 2003). Plants expressing reduced amounts of AIP1 do form fewer root hairs, where growth rate and final length are reduced. However, the diameter of these root hairs is not dramatically altered (Ketelaar et al., 2004). This may indicate that a minimal amount of actin dynamics is required for proper root hair elongation, whereas there is a minimal amount of F-actin required for focused root hair expansion.

Although the effects of AIP1 overexpression on plant development are less dramatic than the effects of AIP1 inhibition, actin organization and root hair development are disrupted. This indicates that tight regulation of AIP1 expression is essential for proper plant development.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was funded by grants from the BBSRC & EU HPRN-CT-2002–00265 ‘TIPNET’.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References