IJPS40001 Proof 1..13

PROOF u 40001 u IJPS u 169 u 2 u 2008 u EJM u CHECKED u 11/23/07

PEROXISOME AGGREGATION DURING CYTOKINESIS IN DIFFERENT ANGIOSPERM TAXA

David A. Collings1 and John D. I. Harper2

Plant Cell Biology Group, Research School of Biological Sciences, Australian National University, GPO Box 475, Canberra, Australian Capital Territory 2601, Australia; and E. H. Graham Centre for Agricultural Innovation, School of Agricultural and Veterinary Sciences, Charles Sturt University, Locked Bag 588, Wagga Wagga, New South Wales 2678, Australia

During cytokinesis in onion (Allium cepa L.) and leek (Allium porrum L.), all the peroxisomes present within the dividing cell aggregate in the phragmoplast adjacent to the developing cell plate. In order to understand the functional implications of this novel arrangement, especially in onion, which has hitherto been regarded as a model system in which to study plant cell division, we investigated how widespread the phenomenon was in selected monocots and dicots. During monocot cytokinesis, peroxisomes lacked aggregation in some taxa, notably grasses, and underwent partial aggregation in other taxa. However, complete aggregation of a cell’s entire complement of peroxisomes was restricted to the genus Allium. Although peroxisomal aggregation has been suggested to function in the formation of the cell plate, the distribution of partial and complete aggregation in monocots did not match known differences in primary cell wall biochemistry. No aggregation was seen during cytokinesis in dicots. Through quantification of peroxisome distribution in cells whose actin microfilaments were disrupted with either latrunculin or cytochalasin, we demonstrated that peroxisomal aggregation is a microfilament-dependent process in Allium and the closely related plant Tristagma, which shows only partial aggregation. We speculate whether analysis of the peroxisomal proteome might reveal novel function(s) for aggregated peroxisomes during cytokinesis.

Keywords: actin microﬁlaments, Allium, cell division, microtubules, peroxisomes, phragmoplast.

Introduction with lipid droplets that they metabolize, while leaf peroxisomes are closely appressed to the chloroplasts and mitochondria; The peroxisome is a small, ubiquitous eukaryote organelle together, these organelles form an integral part of the photores- that is bounded by a single membrane and is involved in many piratory pathway (Hayashi and Nishimura 2003). The mecha- metabolic processes, one of which is hydrogen peroxide metab- nisms that allow unspecialized peroxisomes to alternate between olism (Huang et al. 1983; Olsen and Harada 1995; Baker and stationary and rapidly moving states and allow specialized peroxi- Graham 2002). In plants, peroxisomes are remarkably plastic, somes to retain their necessary locations remain undiscovered. with their number, appearance, function, and location within We have observed that while interphase Allium cepa (onion) the cell dependent on cell type, stage of development, and inter- root tip and leek (Allium porrum L.) epidermal cells have ran- nal and external environmental cues (Huang et al. 1983). Apart domly distributed peroxisomes, in dividing cells, aggregated from the peroxisomes found in most tissues, generally referred peroxisomes associate with the cell plate. As observed in onion, to as unspecialized peroxisomes, specialized peroxisomes occur peroxisomes begin to aggregate in late anaphase, and through- in some tissues. These specialized peroxisomes include, in oil out cytokinesis, these aggregated peroxisomes form two dis- seeds, glyoxysomes that function in the metabolism of storage tinct layers immediately adjacent to either side of the outer lipids through the glyoxylate cycle. In leaves, specialized leaf edge of the developing cell plate. As with cytoplasmic peroxi- peroxisomes function in the recovery of fixed carbon through somal movement, this aggregation depends on microfilaments, the photorespiratory cycle (Olsen and Harada 1995). The spe- and agents that disrupt microfilament organization, such as cialization evident in peroxisomes leads to changes in their ap- cytochalasin and latrunculin, randomize peroxisomes during pearance and location in cells. In interphase cells, unspecialized division (Collings et al. 2003). We consider aggregation likely peroxisomes are randomly distributed, and they typically alternate to be functionally important, just as the nonrandom distribu- between being stationary and exhibiting rapid, microfilament- tions of the endoplasmic reticulum (ER) and Golgi apparatus Q1 dependent movements (Collings et al. 2002; Jedd and Chua during cytokinesis correspond to their functions in the forma- 2002; Mano et al. 2002; Mathur et al. 2002; reviewed by tion of the cell plate and new cell wall (Nebenfu¨hr et al. 1999; Muench and Mullen [2003]). Specialized peroxisomes, how- Dixit and Cyr 2002). We previously suggested two possible ever, are tethered at specific locations. Glyoxysomes associate functions that peroxisomes might perform during cytokinesis. The first is that the aggregated peroxisomes adjacent to the cell 1 Author for correspondence; e-mail: [email protected]. plate recycle membranes, which is necessary because the total 2 E-mail: [email protected]. surface area of the vesicles required to build the cell plate is Manuscript received May 2007; revised manuscript received September 2007. considerably greater than the surface area of the new plasma

1 2 INTERNATIONAL JOURNAL OF PLANT SCIENCES membrane (Otegui et al. 2001). Alternatively, were the harden- (FITC-conjugated sheep antirabbit IgG [Silenus-Amrad, Boronia, ing of the cell plate to involve peroxide-based cross-linking of Australia] and Cy-5-conjugated goat antimouse IgG [Jackson, pectins or extensins, then aggregated peroxisomes might scav- West Grove, PA] diluted at 1/100 and 1/200, respectively). Sec- enge peroxide diffusing into the cell, limiting oxidative damage tionswerethenwashedinPBS(threetimes,20mineach),stained (Collings et al. 2003). Evidence for both of these concepts is, with 0.1 mg/mL 49,6-diamidino-2-phenylindole (DAPI; Sigma) however, lacking. in PBS (5 min), and mounted in AF1 antifade agent (Citifluor, Any theory to explain the functions of aggregated peroxi- London). Immunolabeled root cells were imaged on a confocal somes during cell division must also account for differences in microscope (Leica SP2 system; Leica, Wetzlar, Germany) with a this phenomenon between species. We previously observed that 403 NA 1.4 oil immersion lens, with fourfold line averaging, and while aggregation occurs in various tissues in Allium, it does with optical series collected at 0.5- or 1.0-mm intervals. Using con- not occur in the roots of Arabidopsis (Collings et al. 2003). current excitation with both 488-nm argon and 633-nm helium/ Therefore, by surveying the distribution of the aggregation phe- neon lasers, fluorescence was collected with emission windows nomenon across diverse taxa, we might develop new insight set to 500–600 nm for FITC and 650–760 nm for Cy-5. into the functions performed by aggregated peroxisomes during For triple-labeling experiments, the same protocol was fol- cytokinesis. Our data indicate that while a partial form of ag- lowed. Primary antibodies for labeling the Golgi apparatus and gregation occurs during cell division in some monocot species, ER were mouse monoclonal anti-b-COP, a Golgi apparatus– highly ordered aggregation is limited to the genus Allium. resident protein (Couchy et al. 2003; clone M3A5 dilution 1/50, Sigma), and mouse monoclonal anti-Hsp70, an ER-resident protein (clone 1D9 dilution 1/200, StressGen, Victoria, BC) Material and Methods used in conjunction with rabbit anticatalase and rat monoclonal antitubulin (clone YOL 1/34 dilution 1/50, Sera-Lab, Crawley Q2 Plant Material Down, UK). Secondary antibodies used were Cy-5–conjugated Q3 Bulbs, seeds, and mature plants of a range of different spe- goat antirabbit IgG (Jackson), diluted 1/100, and cross-absorbed, cies were purchased from nurseries in Australia and from local species-specific Alexa 546–conjugated goat antimouse IgG and supermarkets. Seeds were germinated on moistened paper, while Alexa 488–conjugated goat antirat IgG (Molecular Probes, bulbs were allowed to sprout in loose potting mix. Roots were Eugene, OR), each diluted 1/200. Confocal settings were similar sampled when between 10 and 30 mm long. Mature plants to those used for double labeling, except that a 543-nm helium/ were used when seeds or bulbs were not available and were neon laser was also used, and emission windows were set at generally repotted in loose potting mix, with growing roots 500–530, 550–625, and 650–760 nm. To avoid bleed-through harvested when they became visible. from the Alexa 488 dye into the Alexa 546 image, images from Alexa 488 were collected sequentially from the concurrent imaging of Alexa 546 and Cy-5. Immunolabeling of Roots Excised root tips were fixed for 60 min in PME buffer (50 þ mM PIPES pH 7.0 K ,2mMMgSO4, 2 mM EGTA, and 0.1% Inhibitor Studies and Quantification of Triton X-100) containing 3.7% formaldehyde (diluted from Peroxisome Localization 37% formalin), then washed in PME (three times, 10 min each). For drug treatments, frozen stock solutions in DMSO of For specific experiments, the fixative was modified by the inclu- latrunculin B (MP Biomedicals) and cytochalasin D (Sigma) sion of 1.0% glutaraldehyde. Root tips were embedded in 3% were diluted to 2 and 5 mM, respectively, in distilled water, agar in PME buffer and cut into 150–200-mm-thick longitudi- and the DMSO concentration was adjusted to 1.0%. The api- nal sections with a Vibratome (Lancer Series 1000; Vibratome, cal 5–10 mm of root tips were exposed to these drug solu- St. Louis, MO). Sections were incubated in phosphate-buffered tions, or to DMSO controls, for 3–4 h before being fixed and saline (PBS; 131 mM NaCl, 5.1 mM Na2HPO4, and 1.56 mM processed for double labeling. Peroxisomal distributions were KH2PO4, pH 7.2) and processed in 96-well tissue culture plates determined from each image in optical series (1.0-mm inter- in which the well bases had been replaced with fine nylon vals) by manually counting peroxisomes associated with the gauze. For species where root tips were too small to section, phragmoplast or resident in the cytoplasm and by measuring root tip squashes were performed, following published methods cytoplasmic and phragmoplast areas with the program Meta- (Collings et al. 2003), by digesting cell walls in roots with morph (ver. 4.1.7, Universal Imaging, Downington, PA). Be- 1.0% (w/v) cellulysin Y6 and 0.1% (w/v) pectolyase Y23 (MP cause the root tip cells were not vacuolate, cytoplasmic areas Biomedicals, Seven Hills, Australia) in PME for 10 min. Roots were determined as cell areas minus the areas covered by nu- were then squashed between two multiwell slides (MP Biomed- clei and, when present, phragmoplasts. icals) coated with 0.1% polyethyleneimine (Sigma). Root sections or squashes were blocked in incubation buffer (PBS containing 1% BSA and 0.1% Triton X-100; 30 min). For double-labeling experiments, sections were incubated for 1–2 h Results at room temperature in primary antibodies (mouse anti-a-tubulin [clone B512, Sigma] and rabbit anticottonseed catalase [courtesy Peroxisomes, but Not Other Organelles, Aggregate of R. N. Trelease, Arizona State University], each diluted 1/1000 around the Cell Plate during Cytokinesis in Allium in incubation buffer). Material was washed in PBS (three times, Peroxisomes aggregated during cytokinesis in immunolabeled 20 min each) and labeled with secondary antibodies for 2 h sections of Allium roots (figs. 1, 2). Aggregation was observed in COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 3

Fig. 1 Peroxisomes aggregate at the cell plate during cytokinesis in Allium. In onion (Allium cepa) roots, immunolabeled microtubules (top row) and peroxisomes (bottom row) are shown in single, confocal optical sections with the root tip oriented downward. A, B, Peroxisomes were not aggregated in either interphase cells (int) or cells in preprophase (PPB). C, D, Aggregation began late in anaphase, when a single band of peroxisomes (D, arrow) developed at the middle of the mitotic spindle (sp). During early telophase (E, F), peroxisomes distributed into two bands (F, arrows) at the center of the barrel-shaped microtubule phragmoplast (phr), while in late anaphase (G, H), the microtubule phragmoplast expanded outward, and two rings of peroxisomes occurred immediately to the inside of the microtubules (H, arrows). Bar in H ¼ 10 mm for all images.

onion (A. cepa), leek (A. porrum), and several other species, in- adjacent to the cell plate (fig. 2C, arrow), ER remained present cluding two further crops (spring onions, A. fistulosum, and gar- throughout the cytoplasm, including in reticulate cortical ar- lic, A. sativum) and three ornamentals (A. cowanii, A. moly,and rays (arrowheads). Similarly, the Golgi apparatus accumulated A. sphaerocephalum). These species all showed similar patterns in the division plane but also remained present throughout the of peroxisome aggregation during cytokinesis (table 1 summa- cytoplasm (fig. 2F, arrow). rizes data for all species tested). In onion roots, the randomly organized peroxisomes of interphase and preprophase cells (fig. 1A,1B) gathered late in Peroxisomes Do Not Aggregate during Cytokinesis anaphase at the center of the division plane into a single band in Dicots and Some Monocots (fig. 1C,1D, arrow). In early telophase, this band of peroxi- In contrast to Allium, none of the other species that we inves- somes resolved into a pair of bands on either side of the form- tigated showed peroxisomes that were tightly aggregated during cell plate (fig. 1E,1F, arrows). As telophase proceeded, the ing cytokinesis (figs. 3, 4; table 1). Of the dicots that we have microtubule phragmoplast expanded outward to the edge of tested, Arabidopsis, cucumber (fig. 3A,3B), and anemone (fig. Q4 the cell (fig. 1G), and the peroxisomes arranged into two par- 3C,3D) all retained randomly organized peroxisomes through- allel rings immediately to the inside of the microtubules and out cell division. Similarly, numerous monocot species also had on either side of the outer edge of the developing cell plate completely random peroxisomes during cytokinesis, including (fig. 1H, arrows). Aggregation involved essentially all peroxi- maize (fig. 3E,3F), rice and other grasses (not shown), and somes in a cell and occurred in all telophase cells whether the Freesia (fig. 3G,3H; table 1). plane of cell division was transverse (fig. 1F,1H) or oblique (fig. 1D). Further, aggregation occurred in all cell divisions in Allium, including cells within the root proper (fig. 1C–1H), Many Monocots Show Partial Aggregation of the root cap (data not shown), and the leaf epidermal cells, Peroxisomes during Cytokinesis and during stomatal pore formation in the leaf (data not Between the two extremes of Allium and the dicots exam- Q5 shown). Furthermore, roughly even numbers of peroxisomes ined, we found that root cells from a range of monocot species Q6 were inherited by each daughter cell (fig. 1F,1H). showed partial aggregation of peroxisomes during cytokinesis, Triple-labeling experiments were used to compare how per- adjacent to the late phragmoplast (fig. 4; table 1). As with Al- oxisomal aggregation in leek roots contrasted with the distribu- lium, this partial aggregation occurs only in late anaphase and tion of other organelles (fig. 2). Peroxisomes aggregated during telophase. Telophase Tradescantia cells had an increased den- cytokinesis, but this aggregation was different from that of two sity of peroxisomes adjacent to the phragmoplast and cell plate Q7 other organelles that accumulated at the division plane. Al- (fig. 4B, arrow), while in telophase Arum cells, peroxisomes as- though the ER accumulated in the division plane and, notably, sociated with the late phragmoplast were often brighter than 4 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 2 Peroxisome aggregation during cytokinesis is different to the localization of other organelles. Single optical sections through leek (Allium porrum) roots show that different organelles have distinct aggregation patterns during cytokinesis. Microtubules (A, D) and peroxisomes (B, E) are shown triple-labeled with either endoplasmic reticulum (ER, C) or the Golgi apparatus (F). The ER accumulated at the cell plate (C, arrow) but also occurred throughout the cytoplasm, including in a reticulate array in the cell periphery (arrowheads). Discrete Golgi vesicles occurred throughout the cytoplasm, but the buildup of punctate labeling within the division plane indicated Golgi-derived vesicles (F, arrow). Bar in F ¼ 10 mm for all images. other peroxisomes within the cell or in neighboring cells (fig. 4D, occurred during cytokinesis in Brodiaea (fig. 4O,4P)orinthere- arrowhead). This partial-aggregation effect was, however, differ- lated genus Triteleia (table 1). ent from the aggregation effect in Allium for several reasons. Un- like complete aggregation, which began in early telophase and formed a double ring of all the cell’s peroxisomes on the inside of Spatial Analysis of Aggregated and Partially the late microtubule phragmoplast, partial aggregation was char- Aggregated Peroxisomes acterized by development late in telophase of an incomplete Peroxisomal aggregation in Allium depends on the interaction (hence partial) localization of peroxisomes to the outer edge of between myosin and microfilaments, in that the actin-disrupting the phragmoplast. Further, levels of aggregation sometimes varied drugs cytochalasin D and latrunculin B and the myosin inhibitor markedly between different cells in the one plant. BDM all block the aggregation of peroxisomes. While inhibi- The partial aggregation of peroxisomes during cytokinesis tor studies suggested that a similar microfilament-dependent occurred in numerous monocot families (table 1) that are taxo- localization occurred during partial aggregation (fig. 5), because nomically divergent. Yet within these families, species exist this process is much less pronounced than the full aggregation where no aggregation was apparent. For example, while some found in Allium, we conducted a numerical analysis that con- members of the Amaryllidaceae, including Amaryllis (fig. 4E, firmed microfilament dependency (table 2). Q8 4F)andCrinum (data not shown), showed a strong pattern of We first characterized the effects of 2-mM latrunculin B and partial aggregation during cytokinesis, Agapanthus did not (fig. 5-mM cytochalasin D on aggregation in Allium. Both cytochala- 4G,4H). Even within the Alliaceae, the complete aggregation sin and latrunculin greatly reduced the levels of peroxisome of all peroxisomes during cytokinesis occurred only in the aggregation (fig. 5C–5F), compared to those found during cyto- genus Allium. Related genera, including Tristagma, Tulbaghia, kinesis in controls (fig. 5A,5B). When quantified, these drugs Q9 Nothoscordum (fig. 4I–4N), and Leucocoryne (data not shown), caused no changes in the relative density of peroxisomes found showed only partial peroxisomal aggregation, while in Ipheion, in interphase cells but caused a significant reduction in the asym- no aggregation was apparent (data not shown; table 1). We also metrical distribution of peroxisomes during cytokinesis. Instead tested several genera traditionally placed within the Alliaceae of peroxisomes being 40-fold more commonly associated with that molecular evidence now places within the Themidaceae. the phragmoplast than with the remaining cytoplasm of the cell, Consistent with this reclassification, no peroxisomal aggregation this ratio drops to less than twofold (table 2). In Tristagma,an COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 5

Table 1 Taxonomic Distribution of Peroxisomal Aggregation during Cytokinesis Aggregated Aggregated Common peroxisomes Common peroxisomes Family, species name in roots Family, species name in roots Monocots: Gladiolus sp. Partial Agavaceae: Poaceae: Dracaena draco Dragon tree No Avena sativa Oats No Yucca filamentosa No Oryza sativa Rice No Alliaceae: Zea mays Maize No Allium cepa Onion Yesa,b,c,d Themidaceae: Allium cowanii Yesd Brodiaea elegans No Allium fistulosum Spring onion Yesd Triteleia ixioides No Allium moly Yesd Xanthorrhoeaceae: Allium porrum Leek Yesa,b,c,d Xanthorrhoea preissii Grass tree No Allium sativum Garlic Yesd Zingiberaceae: Allium sphaerocephalum Yesd Zingiber officinalis Ginger No Ipheion peregrinans Nod Dicots: Leucocoryne sp. Partiald Apiaceae: Nothoscordum dialystemon Partiale Coriandrum sativum Coriander Noa Tristagma uniflora Partiald Brassicaceae: Tulbaghia violacea Partiald Arabidopsis thaliana No Amaryllidaceae: Cucurbitaceae: Agapanthus sp. No Cucurbita pepo Pumpkin No Amaryllis belladonna Partial Fabaceae: Crinum alba Partial Pisum sativum Pea No Crinum pedunculatum Partiala,b Linaceae: Haemanthus katherinae Blood lily Partial Linum usitasitissumum Flax Noa Araceae: Malvaceae: Arum italicum Partial Alcea rosea Noa Colocasia esculenta Taro Partial Proteaceae: Commelinaceae: Banksia burdettii No Tradescantia virginiana Spiderwort Partial Ranunculaceae: Hyacinthaceae: Anemone coronaria No Lachenalia sp. No Scrophulariaceae: Ornithogalum arabicum No Antirrhinum majus Snapdragon Noa Iridaceae: Solanaceae: Freesia refracta No Nicotiana tabacum Tobacco Noa,f Liliaceae: Umbelliferaceae: Lilium longiflorum Partial Daucus carota Carrot Noa a Observed in root tip squashes. b Data from Collings et al. (2003). c Also demonstrated in leaves. d Alliaceous chemistry (onion smell) present. e Alliaceous chemistry absent. f Reported in tobacco BY-2 cells (Collings et al. 2003) but not replicated in this study.

onionlike plant from the Alliaceae that shows only partial aggre- The complete, microfilament-dependent aggregation of all of a gation of peroxisomes during cytokinesis, we observed a similar cell’s peroxisomes to the division plane adjacent to the forming microfilament-dependent localization of peroxisomes during cy- cell plate was limited to the genus Allium. This aggregation pat- tokinesis. While partial aggregation was apparent in control roots tern was also restricted to peroxisomes, as even in Allium,the (fig. 5G,5H), both latrunculin and cytochalasin reduced aggre- ER and Golgi apparatus showed only partial aggregation at the gation (fig. 5I–5L). When quantified, the ratio of peroxisomes division plane. However, a diverse range of monocot species ex- associated with the phragmoplast to those associated with other hibited partial aggregation of peroxisomes during cytokinesis. cytoplasm dropped significantly, from above four to below two, Allium root tips have been a classical system in which to study Q10 again confirming microfilament dependency (table 2). cell division, including at the light-microscope level (e.g., Kellicott 1904), in some of the earliest electron microscopy studies of plants Discussion (e.g., Rozsa and Wyckoff 1951; Porter and Machado 1960), and for the first immunofluorescence demonstration of microtubules Our data demonstrate that the behavior of peroxisomes dur- in cell wall–containing plant cells (Wick et al. 1981). Despite the ing cytokinesis varies markedly among higher plant species. fact that Allium is a model system for cell division, our results 6 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 3 Peroxisomes do not aggregate during cytokinesis in dividing cells of the dicots tested or in some monocot roots. Microtubules (top row) and peroxisomes (bottom row) are shown in single representative confocal optical sections of telophase cells in the dicots Curcurbita pepo (cucumber; A, B) and Anemone coronaria (anemone; C, D) and the monocots Zea mays (maize; E, F) and Freesia odorata (freesia; G, H). Peroxisome aggregation did not occur with the microtubule phragmoplast (phr), the mitotic spindle (sp), or preprophase bands (PPB). Bar in H ¼ 10 mm for all images. demonstrate that, with regard to peroxisome distribution, it has ing of the cell plate through peroxide-based cross-linking of highly unusual cell divisions and that significant differences exist pectins or extensins (Collings et al. 2003). between it and other model systems of monocot cell division, The cell biology data provided in this study give no direct evi- such as Haemanthus and Tradescantia, which show partial ag- dence for specific roles of peroxisomes during plant cell division, gregation, and the grasses and dicots, such as Arabidopsis,which although the taxonomic distribution of partial aggregation— lack aggregation. limited to some monocots, with complete aggregation found only in Allium—may be significant. However, our evidence would seem to rule out three possible explanations for aggregation. Why Do Peroxisomes Aggregate during Cytokinesis? 1. That full aggregation is limited to Allium might suggest What is the function, then, if any, of aggregated peroxisomes some unexplained link between aggregation and the distinctive during cytokinesis in Allium, and do peroxisomes perform any sulfurous aroma, or alliaceous chemistry, present in plants from functions when partially aggregated in other monocot cell divi- this genus. However, no correlation exists between the presence sions? By analogy with other organelles, we suggest that perox- or absence of alliaceous chemistry in the Alliaceae and the pres- isomal aggregation likely has some function in monocots. Both ence of full or partial peroxisomal aggregation (table 1). the accumulation of ER in the mitotic spindle and phragmo- 2. It would seem unlikely that peroxisomal aggregation oc- plast of dividing plant cells (Gupton et al. 2006 and references curs to ensure the equal partitioning of peroxisomes into the therein) and the accumulation of the Golgi apparatus into the daughter cells, as plastids and mitochondria segregate during ‘‘Golgi band’’ in the plane of cell division during metaphase division without aggregating. Although few studies have di- and adjacent to the cell plate during cytokinesis (Nebenfu¨hr rectly looked at plastid or mitochondrial distribution during et al. 2000; Dixit and Cyr 2002) match the roles these organ- cell division, there have been no indications that they show any elles play in the deposition of the cell plate and new cell wall. type of aggregation. In tobacco cells, while the Golgi apparatus Indeed, it has been observed that the Golgi apparatus can be re- aggregates at the division plane, plastids and mitochondria are cruited to locations where it is needed (Nebenfu¨hr et al. 1999, excluded (Kawazu et al. 1995). Further, electron micrographs 2000). We previously suggested that aggregated peroxisomes of onion root tips, in which peroxisomes can be seen aggregat- might function during cytokinesis, either in the recycling of ex- ing, fail to show changes in either plastid or mitochondrial dis- cess membrane no longer required by the cell or in the harden- tributions (Porter and Machado 1960; Hanzely and Vigil 1975; COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 7

Fig. 4 Many monocot species show partial aggregation of peroxisomes during cytokinesis. Microtubules and peroxisomes are shown in single confocal optical sections of telophase cells of Tradescantia virginiana (A, B), Arum italicum (C, D), Amaryllis belladonna (E, F), Agapanthus sp. (G, H), Tristagma uniﬂora (I, J), Tulbaghia violacea (K, L), Nothoscordum dialystemon (M, N), and Brodiaea elegans (O, P). Varying levels and forms of peroxisomal aggregation were observed during cytokinesis. Peroxisomes associated with the outer edge of the phragmoplast (phr)in Tradescantiaand Amaryllis (B, F, arrows), whereas phragmoplast-associated peroxisomes in Arum appeared more brightly labeled than neighboring cytoplasmic peroxisomes (D, arrowheads). Within the Alliaceae, complete aggregation of peroxisomes during cytokinesis was limited to Allium. Partial aggregation occurred strongly in Tristagma (I)andNothoscordum (N, arrows) but only weakly in Tulbaghia (L, arrow). By contrast, Brodiaea (O), from the Themidaceae, lacked peroxisomal aggregation during cytokinesis. Bar in P ¼ 10 mm for all images.

Vaughn et al. 1996; Collings et al. 2003; K. Vaughn, personal phase are unlikely to be mitochondria, as described, but are communication). Examination of micrographs from an earlier likely to be peroxisomes (Rozsa and Wyckoff 1951). Systems electron microscopy study of onion root tip cell divisions sug- do exist, however, where that partitioning of plastids and mito- gests that the structures aggregated near the cell plate in ana- chondria is regulated. When quantiﬁed in dividing tobacco 8 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 5 Peroxisomal aggregation during cytokinesis is microﬁlament dependent. Microtubules and peroxisomes are shown in maximum projections of confocal optical series covering a depth of ca. 2.5 mm for telophase cells from roots of Allium (A–F) and Tristagma, an onionlike plant from the Alliaceae (G–L). Control treatments (A, B, G, H) were compared to 3–4-h treatments with either 2-mM latrunculin B (C, D, I, J)or 5-mM cytochalasin D (E, F, K, L). In control cells of Allium, the aggregation of peroxisomes (B, arrows) with the microtubule phragmoplast (A, phr) broke down after both latrunculin (D) and cytochalasin treatments (F), so that instead of being associated with the phragmoplast (location indicated with asterisks), peroxisomes were found throughout the cells. In Tristagma, the partial aggregation of peroxisomes during cytokinesis (H, arrow) is also blocked by latrunculin (J) and cytochalasin (L), such that peroxisomes remained randomly positioned in the cytoplasm. Bar in L ¼ 10 mm for all images.

culture cells, the nonaggregated ER, plastids, and mitochondria there are major differences in the composition of the primary partition equally into the daughter cells, in a process requiring cell wall. The primary cross-linking glycan in dicots and many microfilaments (Sheahan et al. 2004). monocots is xyloglucan, whereas the remaining commelinoid 3. Partial aggregation of peroxisomes occurs in numerous monocots use glucuronoarabinoxylan as their main cross-linking monocot families that are taxonomically divergent (fig. 6) glucan (Buchanan et al. 2000). However, partial aggregation Q12 Q11 (Dahlgren 1989; Tamura et al. 2004). Among flowering plants, occurs in both commelinoid monocots, such as Tradescantia COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 9

Table 2 et al. 2005) and with the localization of green fluorescent protein (GFP) fusions of the tail domain from this myosin to per- Q13 Peroxisomal Densities during Cytokinesis Show That oxisomes (Riesen and Hanson 2007). The relationship between Aggregation and Partial Aggregation Are Microfilament-Dependent Processes organelles and the cytoskeleton is, however, complicated, and any statement that a particular organelle depends solely on the Allium Tristagma microfilament cytoskeleton requires caution. Further, peroxi- Control: somal aggregation requires two discrete processes, the transport Interphase: n 13 12 Cytoplasm (px/pL) 73 6 5666 8 Cytokinesis: n 65 Cytoplasm (px/pL) 6.4 6 1.8 24 6 7 Phragmoplast (px/pL) 198 6 20 105 6 32 Ratio 41 6 8 4.3 6 .6 Latrunculin B (2 mM): Interphase: n 713 Cytoplasm (px/pL) 90 6 6546 4 Cytokinesis: n 44 Cytoplasm (px/pL) 47 6 6a 32 6 7 Phragmoplast (px/pL) 69 6 2a 50 6 10 Ratio 1.5 6 .2a 1.6 6 .1 Cytochalasin D (5 mM): Interphase: n 93 Cytoplasm (px/pL) 71 6 7516 4 Cytokinesis: n 42 Cytoplasm (px/pL) 49 6 5a 29 6 2 Phragmoplast (px/pL) 70 6 8a 35 6 16 Ratio 1.5 6 .2a 1.2 6 .6a Note. n ¼ number of replicate cells. px=pL ¼ peroxisomes present per picoliter; this is equivalent to the number of peroxisomes present in a cube with 10-mm sides. The ratio is calculated for indi- vidual cytokinetic cells as the ratio of peroxisomal density associated with the phragmoplast divided by that in the remaining cytoplasm. a Drug treatment values significantly different from drug-free controls (P < 0:05, t-test). and Arum (fig. 4A,4B), and noncommelinoid monocots, so these variations in monocot cell wall biochemistry do not match the distribution of peroxisomal aggregation (fig. 6). Differences in the primary wall are, therefore, not likely to be the cause of aggregation. This need not mean that aggregated peroxisomes do not play a role in cell wall formation. The observed differ- Fig. 6 Peroxisomal aggregation is widespread among monocots, ences in monocot cell walls relate to the primary cell wall, with aggregation occurring in both commelinoid and noncommelinoid whereas peroxisomal aggregation adjacent to the phragmo- monocots. This image, based on work of Dahlgren (1989) and Buchanan plast and cell plate might relate to biochemical differences in et al. (2000), shows orders (with names in boxes) whose size relates to cell plate and middle lamella formation. the number of species and whose position reflects phylogenetic affinity. The 34 monocot species tested belonged to 12 different families (names listed in parentheses) from four different orders. Orders in which The Cytoskeleton and Peroxisomes partial aggregation was observed in some or all species are indicated in Quantification of peroxisomal distributions after drug treat- dark gray, whereas two orders in which aggregation was never ob- ments demonstrated that the full aggregation of peroxisomes served (Poales and Zingiberales) are shown in light gray. Untested orders are shown in white. Among monocot species, primary-wall biochem- in Allium, along with the partial aggregation of peroxisomes istry is an important distinguishing feature between the commelinoid in Tristagma, a genus from the Alliaceae and a close relative monocots (lower half of the figure), which use glucuronoarabinoxylan of Allium, are processes dependent on actin microfilaments as their main cross-linking glucan, and the noncommelinoid mono- (Collings et al. 2003; this study). These observations are con- cots, which, along with dicots, use xyloglucan. There was, however, no sistent with the immunolocalization of a specific myosin XI apparent correlation between peroxisomal aggregation and primary- motor from the surface of Arabidopsis peroxisomes (Hashimoto wall biochemistry. 10 INTERNATIONAL JOURNAL OF PLANT SCIENCES of peroxisome to the cell plate and their tethering there. These study) and that aggregated peroxisomes in Allium colocalize processes may rely on different cytoskeletal components. with microfilaments and not microtubules (Collings et al. Cytoskeleton-based motility of peroxisomes in interphase 2003). Our observations strongly suggest that microfilaments cells depends on microfilaments. The motility of peroxisomes function in both transport to the cell plate and tethering but (Collings et al. 2002; Jedd and Chua 2002; Mano et al. 2002; do not discriminate between the two processes. Further, the Mathur et al. 2002; reviewed by Muench and Mullen [2003]), data do not rule out a role for microtubules in either transport mitochondria (van Gestel et al. 2002), plastids (Kandasamy or tethering, although such a role is unlikely. Proof of a solely Q14 and Meagher 1999), subcortical ER (Boevink et al. 1998 and microfilament-dependent process will require careful analysis references therein), and the Golgi apparatus (Boevink et al. of the effects of microfilament and microtubule disruption, but 1998; Nebenfu¨ hr et al. 1999) are inhibited by microfilament- such experiments are not easy. Immunofluorescent labeling disrupting drugs but not by drugs that depolymerize microtu- after microfilament disruption, as used in this study, failed to bules. Microtubule-based trafficking of organelles has not been distinguish between roles for microfilaments in motility and demonstrated in interphase higher-plant cells, although kinesins tethering and cannot be applied after microtubule depolymer- (microtubule-based motors) associate with mitochondria (Ni ization because the loss of the mitotic spindle blocks cell divi- et al. 2005; Romagnoli et al. 2007) and vesicles (Lu et al. sion at metaphase, before peroxisomal aggregation occurs. 2005). Microtubule-based motility is, however, likely during Instead, living cells must be analyzed. We developed a system cell division, where the organization of the ER during mitosis in which aggregated peroxisomes were visible in the epidermal and cytokinesis relies on the integrity of the microtubule cyto- cells of whole Allium leaves transiently expressing peroxisome- skeleton (Gupton et al. 2006), where kinesins associate with targeted GFP (Collings et al. 2003). We have not, however, been vesicles in cytokinesis (Lee et al. 2001; Vanstraelen et al. 2006) able to use this to investigate time courses through division af- and where microfilament disruption does not inhibit accumula- ter microfilament and microtubule disruption because dividing tion of vesicles in the cell plate. cells cannot be recognized before peroxisomal aggregation. In- Cytoskeletal tethering of organelles, so that they remain sta- stead, we plan to investigate these time courses using cultures tionary at a specific location in the cell while surrounded by of stably transformed Allium (see Eady et al. 2003) expressing streaming cytoplasm, may be as important as cytoskeleton-based peroxisome-targeted GFP to distinguish the roles of microtu- organelle motility. While the subcortical ER shows continual bules and microfilaments in peroxisomal aggregation. motility, organelles, including peroxisomes, mitochondria, and the Golgi apparatus, show what has been described as ‘‘stop- Outlook and-go’’ motion characterized by periods of rapid motion and periods where the organelle is stationary, tethered within the It remains possible that analysis of the mechanism and pro- streaming cytoplasm (Nebenfu¨hr et al. 1999). Microtubules may teins involved in full aggregation in Allium and partial aggrega- function in this tethering. In living tobacco cells, nonmotile mito- tion in some of its close relatives may reveal important chondria often align parallel to the microtubule cytoskeleton, information concerning organelle positioning. Although full ge- and this organization is disrupted by microtubule depolymeriza- nome sequences are not available for any plant that shows par- tion (van Gestel et al. 2002). In tobacco, mitochondrial fusion tial or full aggregation (rice, maize, poplar, and Arabidopsis all during cell dedifferentiation is also microtubule dependent (Shea- lack peroxisomal aggregation), it is significant that a large data- han et al. 2005), and mitochondrial kinesins decrease mitochon- base of ca. 11,000 Allium cDNA sequences is available (Kuhl drial motility in tobacco pollen tubes (Romagnoli et al. 2007). et al. 2004). Furthermore, cell division in Allium roots can be Microtubule-based positioning of mitochondria also occurs in synchronized with hydroxyurea (Clain and Brulfert 1980; Do- Arabidopsis (Logan et al. 2003). Microtubules also function in lezˇel et al. 1999). As nonspecialized peroxisomes can be puri- the positioning of plastids in various plants (Kwok and Hanson fied from various tissues, including bean roots (Theimer and 2003; Chuong et al. 2006), while microtubule depolymerization Heidinger 1974) and Arabidopsis (Fukao et al. 2002, 2003), it increases the number of tobacco cells that show motile Golgi ap- is tempting to speculate whether Allium and Tristagma roots paratus, suggesting that the microtubules act to limit streaming might provide a convenient system in which to study the cyto- (Nebenfu¨hr et al. 1999). Similar suggestions have also been skeletal and biochemical control mechanisms that regulate per- made for peroxisomes. Numerous peroxisomal matrix proteins oxisome positioning and partitioning. Analysis of differences in have been purified based on their affinities for either microtu- the peroxisomal proteome (Fukao et al. 2002, 2003) between bules or tubulin (Chuong et al. 2002, 2004; Harper et al. 2002), interphase and cytokinetic cells of Allium and Tristagma and and it has been proposed that briefly tethering peroxisomes to comparisons to nonaggregating monocots, such as rice and microtubules may facilitate protein uptake into the peroxisome maize, might reveal proteins implicated in the functioning of (Muench and Mullen 2003). No evidence, however, has been aggregated peroxisomes during cytokinesis. provided for this. The switching mechanisms that turn organelle motility and tethering on and off, and whether they are the same for different organelles, have not been determined. Acknowledgments Understanding the roles of the cytoskeleton in peroxisomal aggregation during cell division is, therefore, complicated by the D. A. Collings acknowledges the receipt of an Australian Re- existence of both cytoskeleton-based motility and cytoskeleton- search Council Research Fellowship and Discovery Grant, based tethering. Using microfilament antagonists, we have shown DP0208806, and professors Jim Pratley and Deirdre Lemerle that microfilaments are necessary for the aggregation of per- for an Adjunct Lectureship at Charles Sturt University. J. D. I. oxisomes in Allium and Tristagma (Collings et al. 2003; this Harper thanks professors Adrienne Hardham and Richard COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 11

Williamson from the Plant Cell Biology Group (Australian Na- and Dr. Rosemary White (the Commonwealth Scientiﬁc and In- tional University [ANU]) for a visiting fellowship. We acknowl- dustrial Research Organisation [CSIRO] Division of Plant Indus- edge support from the E. H. Graham Centre for Agricultural try Microscopy Unit) for assistance with confocal microscopy. Innovation and a Charles Sturt University competitive grant We also thank two anonymous reviewers for their helpful and and thank Dr. Daryl Webb (ANU Electron Microscopy Unit) constructive comments.

Literature Cited

Baker A, IA Graham 2002 Plant peroxisomes: biochemistry, cell Hashimoto K, H Igarashi, S Mano, N Nishimura, T Shimmen, E biology and biotechnological applications. Kluwer, Dordrecht. Yokota 2005 Peroxisomal localization of a myosin XI isoform in Boevink P, K Oparka, S Santa Cruz, B Martin, A Betteridge, C Arabidopsis thaliana. Plant Cell Physiol 45:782–789. Hawes 1998 Stacks on tracks: the plant Golgi apparatus trafﬁcs on Hayashi M, M Nishimura 2003 Entering a new era of research on an actin/ER network. Plant J 15:441–447. plant peroxisomes. Curr Opin Plant Biol 6:577–582. Buchanan BB, W Gruissem, RL Jones 2000 Biochemistry and molecular Huang AHC, RN Trelease, TS Moore Jr 1983 Plant peroxisomes. biology of plants. American Society of Plant Physiologists, Rockville, Academic Press, New York. MD. Jedd G, N-H Chua 2002 Visualization of peroxisomes in living plant Chuong SDX, VR Franceschi, GE Edwards 2006 The cytoskeleton cells reveals acto-myosin-dependent cytoplasmic streaming and

maintains organelle partitioning required for single-cell C4 photo- peroxisome budding. Plant Cell Physiol 43:384–392. synthesis in Chenopodiaceae species. Plant Cell 18:2207–2223. Kandasamy MK, RB Meagher 1999 Actin-organelle interaction: Chuong SDX, AG Good, GJ Taylor, MC Freeman, GBG Moorhead, association with chloroplast in Arabidopsis leaf mesophyll cells. DG Muench 2004 Large-scale identiﬁcation of tubulin binding Cell Motil Cytoskelet 44:110–118. proteins provides insight on subcellular trafﬁcking, metabolic chan- Kawazu T, S Kawano, T Kuroiwa 1995 Distribution of the Golgi neling, and signaling in plant cells. Mol Cell Proteomics 3:970–983. apparatus in the mitosis of cultured tobacco cells as revealed by

Chuong SDX, RT Mullen, DG Muench 2002 Identification of a rice DIOC6 fluorescence microscopy. Protoplasma 186:183–192. RNA- and microtubule-binding protein as the multifunctional protein Kellicott WE 1904 The daily periodicity of cell-division and of (MFP), a peroxisomal enzyme involved in the b-oxidation of fatty elongation in the root of Allium. Bull Torrey Bot Club 31:529–550. acids. J Biol Chem 277:2419–2429. Kuhl JC, F Cheung, Q Yuan, W Martin, Y Zewdie, J McCallum, A Clain E, A Brulfert 1980 Hydroxyurea-induced mitotic synchroniza- Catanach, et al 2004 A unique set of 11,008 of onion expressed se- tion in Allium sativum root meristems. Planta 150:26–31. quence tags reveals expressed sequence and genomic differences Collings DA, JDI Harper, J Marc, RL Overall, RT Mullen 2002 Life between the monocot orders Asparagales and Poales. Plant Cell 16: in the fast lane: actin-based motility of plant peroxisomes. Can J Bot 114–125. 80:430–441. Kwok EY, MR Hanson 2003 Microfilaments and microtubules Collings DA, JDI Harper, KC Vaughn 2003 The association of perox- control the morphology and movement of non-green plastids and isomes with the developing cell plate in dividing onion root cells stromules in Nicotiana tabacum. Plant J 35:16–26. depends on actin microfilaments and myosin. Planta 218:204–216. Lee Y-RJ, HM Giang, B Liu 2001 A novel plant kinesin-related protein Couchy I, S Bolte, M-T Crosnier, S Brown, B Satiat-Jeunemaitre 2003 specifically associates with the phragmoplast organelles. Plant Cell Identification and localization of a b-COP-like protein involved in the 13:2427–2439. morphodynamics of the plant Golgi apparatus. J Exp Bot 54:2053–2063. Logan DC, I Scott, AK Tobin 2003 The genetic control of plant Dahlgren G 1989 An updated angiosperm classification. Bot J Linn mitochondrial morphology and dynamics. Plant J 36:500–509. Soc 100:197–203. Lu L, Y-RJ Lee, R Pan, JN Maloof, B Liu 2005 An internal motor Dixit R, RJ Cyr 2002 Golgi secretion is not required for marking the kinesin is associated with the Golgi apparatus and plays a role in preprophase band site in cultured tobacco cells. Plant J 29:99–108. trichome morphogenesis in Arabidopsis. Mol Biol Cell 16:811–823. Dolezˇel J, J Cˇ ´ıhalı´kova´, J Weiserova´, S Lucretti 1999 Cell cycle Mano S, C Nakamori, M Hayashi, A Kato, M Kondo, M Nishi- synchronization in plant root meristems. Methods Cell Sci 21:95–107. mura 2002 Distribution and characterization of peroxisomes in Eady C, S Davis, J Farrant, J Reader, F Kenel 2003 Agrobacterium Arabidopsis by visualization with GFP: dynamic morphology and tumefaciens–mediated transformation and regeneration of herbicide actin-dependent movement. Plant Cell Physiol 43:331–341. resistant onion (Allium cepa) plants. Ann Appl Biol 142:213–217. Mathur J, N Mathur, M Hu¨ lskamp 2002 Simultaneous visualization Fukao Y, M Hayashi, I Hara-Nishimura, M Nishimura 2003 Novel of peroxisomes and cytoskeletal elements reveals actin and not glyoxysomal protein kinase, GPK1, identified by proteomic analysis microtubule-based peroxisome motility in plants. Plant Physiol 128: of glyoxysomes in etiolated cotyledons of Arabidopsis thaliana. 1031–1045. Plant Cell Physiol 44:1002–1012. Muench DG, RT Mullen 2003 Peroxisome dynamics in plant cells: a Fukao Y, M Hayashi, M Nishimura 2002 Proteomic analysis of leaf role for the cytoskeleton. Plant Sci 164:307–315. peroxisomal proteins in greening cotyledons of Arabidopsis Nebenfu¨ hr A, JA Frohlick, LA Staehelin 2000 Redistribution of thaliana. Plant Cell Physiol 43:689–696. Golgi stacks and other organelles during mitosis and cytokinesis in Gupton SL, DA Collings, NS Allen 2006 Endoplasmic reticulum plant cells. Plant Physiol 124:135–151. targeted GFP reveals ER organization in tobacco NT-1 cells during Nebenfu¨ hr A, LA Gallagher, TG Dunahy, JA Frohlick, AM Mazurkie- cell division. Plant Physiol Biochem 44:95–105. wicz, JB Meehl, LA Staehelin 1999 Stop-and-go movements of plant Hanzely L, EL Vigil 1975 Fine structural and cytochemical analysis Golgi stacks are mediated by the acto-myosin system. Plant Physiol of phragmosomes (microbodies) during cytokinesis in Allium root 121:1127–1141. tip cells. Protoplasma 86:269–277. Ni CZ, HQ Wang, T Xu, Z Qu, GQ Liu 2005 AtKP1, a kinesin-like Harper JDI, ND Weerakoon, JC Gardiner, LM Blackman, J protein, mainly localizes to mitochondria in Arabidopsis thaliana. Marc 2002 A 75-kDa plant protein isolated by tubulin-affinity chro- Cell Res 15:725–733. matography is a peroxisomal matrix enzyme. Can J Bot 80:1018– Olsen LJ, JJ Harada 1995 Peroxisomes and their assembly in higher 1027. plants. Annu Rev Plant Physiol Plant Mol Biol 46:123–146. 12 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Otegui MS, DN Mastronarde, B-H Kang, SY Bednarek, LA inheritance of chloroplasts, mitochondria and endoplasmic reticu- Staehelin 2001 Three-dimensional analysis of syncitial-type cell lum in dividing protoplasts. Plant J 37:379–390. plates during endosperm cellularization visualized by high resolu- Tamura MN, J Yamashita, S Fuse, M Haraguchi 2004 Molecular tion electron tomography. Plant Cell 13:2033–2051. phylogeny of monocotyledons inferred from combined analysis of Porter KR, RD Machado 1960 Studies on the endoplasmic reticu- plastid matK and rbcL gene sequences. J Plant Res 117:109–120. lum. IV. Its form and distribution during mitosis in cells of onion Theimer RR, P Heidinger 1974 Control of particulate urate oxidase root tips. J Biophys Biochem Cytol 7:167–180. activity in bean roots by external nitrogen supply. Z Pflanzenphysiol Riesen D, MR Hanson 2007 Association of six YFP-myosin XI-tail 73:360–370. fusions with mobile plant cell organelles. BMC Plant Biol 7:6. van Gestel K, RH Ko¨ hler, J-P Verbelen 2002 Plant mitochondria Romagnoli S, G Cai, C Faleri, E Yokota, T Shimmen, M Cresti 2007 move on F-actin, but their positioning in the cortical cytoplasm de- Microtubule- and actin-filament-dependent motors are distributed on pends on both F-actin and microtubules. J Exp Bot 53:659–667. pollen tube mitochondria and contribute differently to their move- Vanstraelen M, D van Damme, R de Rycke, E Mylle, D Inze´,D ment. Plant Cell Physiol 48:345–361. Geelen 2006 Cell cycle-dependent targeting of a kinesin at the Rozsa G, RWG Wyckoff 1951 The electron microscopy of onion plasma membrane demarcates the division site in plant cells. Curr root tip cells. Exp Cell Res 2:630–641. Biol 16:308–314. Sheahan MB, DW McCurdy, RJ Rose 2005 Mitochondria as a con- Vaughn KC, JC Hoffman, MG Hahn, LA Staehelin 1996 The nected population: ensuring continuity of the mitochondrial genome herbicide dichlobenil disrupts cell plate formation: immunogold during plant cell dedifferentiation through massive mitochondrial characterization. Protoplasma 194:117–132. fusion. Plant J 44:744–755. Wick SM, RW Seagull, M Osborn, K Weber, BES Gunning 1981 Im- Sheahan MB, RJ Rose, DW McCurdy 2004 Organelle inheritance in munofluorescence microscopy of organized microtubule arrays in plant cell division: the actin cytoskeleton is required for unbiased structurally stabilized meristematic plant cells. J Cell Biol 89:685–690. COLLINGS & HARPER—PEROXISOMAL AGGREGATION DURING CYTOKINESIS IN ANGIOSPERMS 13

Queries

Q1 The introduction of the abbreviation ‘‘ER’’ has been relocated to the first use of the ‘‘parent’’ term. Q2 N.B. Queries 3–14 were e-mailed to you during the manuscript editing stage, but unfortunately we did not receive your reply before the article had to be sent out for typesetting. If you have already responded to these queries, please ignore the following. If not, please do answer them now. Q3 Were all three antibodies ‘‘used in conjunction with rabbit anticatalase...,’’ or just mouse monoclonal anti-Hsp70? Q4 In the sentence beginning ‘‘As telophase proceeded,’’ can ‘‘arranged’’ be changed to ‘‘gathered,’’ ‘‘arranged themselves,’’ ‘‘resolved,’’ or something similar? This revision is recommended because ‘‘arranged’’ normally has an object. Q5 In the sentence beginning ‘‘Further, aggregation occurred,’’ can ‘‘data not shown’’ be changed to ‘‘not shown’’ or ‘‘not illustrated’’? This revision is recommended because the root cap and the leaf are contrasted with the root proper, which is illustrated in figure 1 but for which no data per se is offered. Similar changes are recommended in similar contexts throughout the article. Q6 In the last sentence of the paragraph, does ‘‘roughly even numbers’’ mean ‘‘roughly equal numbers’’? This interpretation is suggested by item 2 in the second paragraph of ‘‘Why Do Peroxisomes Aggregate during Cytokinesis’’? Q7 In the first sentence of the figure 2 legend, the passage ‘‘different to the localization’’ has been changed to ‘‘different from the localization’’ on the basis of the sentence beginning ‘‘Peroxisomes aggregated’’ in text. Is this the correct interpretation, or did you intend to write ‘‘different due to the localization’’? If neither, please clarify this passage. Q8 Table 2 has been inverted because in its original form it would have required a page of its own. Because of this reconfigu- ration, some of the footnotes have been consolidated into a table note. In the table itself: (a) It has been assumed that the former footnote a (‘‘Number of replicate cells’’) is an explanation of ‘‘n ¼ ’’ in all cases. Is this correct? If not what do the ‘‘n ¼ ’’ counts represent? (b) It has also been assumed that the unit px/pL applies to all ‘‘Cytoplasm’’ and ‘‘Phragmo- plast’’ entries. Is this correct? Q9 The second sentence of this paragraph has been recast to avoid comparing ‘‘aggregated peroxisomes’’ themselves with ‘‘peroxisome aggregation.’’ Q10 The last sentence of the paragraph has been revised to describe the comparison more precisely. Does the revised sentence express your intended meaning accurately? Q11 In the figure 6 legend, the sixth sentence has been recast to read ‘‘Among monocot species, primary-wall biochemistry is an important distinguishing feature between the commelinoid monocots (lower half of the figure), which use glucuronoarabinoxylan....’’ This revision was made primarily to place ‘‘between’’ closer to ‘‘distinguishing.’’ Does the revised sentence express your intended meaning accurately? Q12 Are ‘‘glycan’’ and ‘‘glucan’’ interchangeable? Q13 Is ‘‘green fluorescent protein’’ correct for GFP? Q14 In the sentence beginning ‘‘Further, the data do not rule out,’’ the words ‘‘although unlikely’’ have been placed at the end of the sentence and expanded into ‘‘although such a role is unlikely,’’ so as not to interrupt ‘‘role for microtubules....’’ Change OK?