Proteasome-Rich PaCS as an Oncofetal UPS Structure Handling Cytosolic Polyubiquitinated Proteins. In Vivo Occurrence, in Vitro Induction, and Biological Role.

In this article, we outline and discuss available information on the cellular site and mechanism of proteasome interaction with cytosolic polyubiquitinated proteins and heat-shock molecules. The particulate cytoplasmic structure (PaCS) formed by barrel-like particles, closely reproducing in vivo the high-resolution structure of 26S proteasome as isolated in vitro, has been detected in a variety of fetal and neoplastic cells, from living tissue or cultured cell lines. Specific trophic factors and interleukins were found to induce PaCS during in vitro differentiation of dendritic, natural killer (NK), or megakaryoblastic cells, apparently through activation of the MAPK-ERK pathway. Direct interaction of CagA bacterial oncoprotein with proteasome was shown inside the PaCSs of a Helicobacter pylori-infected gastric epithelium, a finding suggesting a role for PaCS in CagA-mediated gastric carcinogenesis. PaCS dissolution and autophagy were seen after withdrawal of inducing factors. PaCS-filled cell blebs and ectosomes were found in some cells and may represent a potential intercellular discharge and transport system of polyubiquitinated antigenic proteins. PaCS differs substantially from the inclusion bodies, sequestosomes, and aggresomes reported in proteinopathies like Huntington or Parkinson diseases, which usually lack PaCS. The latter seems more linked to conditions of increased cell proliferation/differentiation, implying an increased functional demand to the ubiquitin⁻proteasome system.

Russell body inducing threshold depends on the variable domain sequences of individual human IgG clones and the cellular protein homeostasis.

Russell bodies are intracellular aggregates of immunoglobulins. Although the mechanism of Russell body biogenesis has been extensively studied by using truncated mutant heavy chains, the importance of the variable domain sequences in this process and in immunoglobulin biosynthesis remains largely unknown. Using a panel of structurally and functionally normal human immunoglobulin Gs, we show that individual immunoglobulin G clones possess distinctive Russell body inducing propensities that can surface differently under normal and abnormal cellular conditions. Russell body inducing predisposition unique to each immunoglobulin G clone was corroborated by the intrinsic physicochemical properties encoded in the heavy chain variable domain/light chain variable domain sequence combinations that define each immunoglobulin G clone. While the sequence based intrinsic factors predispose certain immunoglobulin G clones to be more prone to induce Russell bodies, extrinsic factors such as stressful cell culture conditions also play roles in unmasking Russell body propensity from immunoglobulin G clones that are normally refractory to developing Russell bodies. By taking advantage of heterologous expression systems, we dissected the roles of individual subunit chains in Russell body formation and examined the effect of non-cognate subunit chain pair co-expression on Russell body forming propensity. The results suggest that the properties embedded in the variable domain of individual light chain clones and their compatibility with the partnering heavy chain variable domain sequences underscore the efficiency of immunoglobulin G biosynthesis, the threshold for Russell body induction, and the level of immunoglobulin G secretion. We propose that an interplay between the unique properties encoded in variable domain sequences and the state of protein homeostasis determines whether an immunoglobulin G expressing cell will develop the Russell body phenotype in a dynamic cellular setting.

Inhibition of nonsense-mediated mRNA decay (NMD) by a new chemical molecule reveals the dynamic of NMD factors in P-bodies.

In mammals, nonsense-mediated mRNA decay (NMD) is a quality-control mechanism that degrades mRNA harboring a premature termination codon to prevent the synthesis of truncated proteins. To gain insight into the NMD mechanism, we identified NMD inhibitor 1 (NMDI 1) as a small molecule inhibitor of the NMD pathway. We characterized the mode of action of this compound and demonstrated that it acts upstream of hUPF1. NMDI 1 induced the loss of interactions between hSMG5 and hUPF1 and the stabilization of hyperphosphorylated isoforms of hUPF1. Incubation of cells with NMDI 1 allowed us to demonstrate that NMD factors and mRNAs subject to NMD transit through processing bodies (P-bodies), as is the case in yeast. The results suggest a model in which mRNA and NMD factors are sequentially recruited to P-bodies.

Dark cell change of the cerebellar Purkinje cells induced by terbutaline under transient disruption of the blood-brain barrier in adult rats: morphological evaluation.

This study aimed to establish a cerebellar degeneration animal model and to characterize the dark cell change of Purkinje cells. We hypothesized that terbutaline, a ß2-adrenoceptor agonist, induces cerebellar degeneration not only in neonatal rats, but also in adult rats. Nine-week-old adult male Sprague-Dawley rats were anesthetized and infused with 25% mannitol via the left common carotid artery. Thirty seconds later, terbutaline was infused via the same artery. Dark-stained Purkinje cells were observed in the entire cerebellum on day 3. Prominent Bergmann glial cells accompanied by swelling of the glial processes were present, and were closely associated with the dark-stained Purkinje cells. These findings were found continuously throughout day 30. Ultrastructurally, dilated Golgi vesicles and/or endoplasmic reticulum and large lamella bodies were present in both severely changed and slightly changed Purkinje cells. Bergmann glial cells in the area of synaptic contacts of the severely changed Purkinje cells showed swelling. The Bergmann glial process in close contact with the slightly changed Purkinje cell dendrite in molecular layer showed slight swelling, and large lamella bodies in the dendrite were observed close to the dendritic spines. These findings may suggest that terbutaline induced a failure of Bergmann glial cell and resulted in dark cell degeneration of the Purkinje cells due to glutamate excitotoxicity.

The virion-packaged endoribonuclease of herpes simplex virus 1 cleaves mRNA in polyribosomes.

The virion host shutoff protein product of the U(L)41 gene of herpes simplex virus 1 is an endoribonuclease that selectively degrades mRNAs during the first hours after infection. Specifically, in contrast to the events in uninfected cells or cells infected with a mutant lacking the RNase, in wild-type virus-infected cells mRNA of housekeeping genes exemplified by GAPDH is degraded rapidly, whereas mRNAs containing AU elements are cleaved and the 5' cleavage product of these RNAs persists for many hours. We report that in wild-type virus-infected cells there was a rapid increase in the number and size of processing bodies (P-bodies). These P-bodies were also preset in cycloheximide (CHX)-treated cells but not in either treated or untreated uninfected cells or cells infected with the RNase minus mutant. Additional studies revealed that polyribosomes extracted from cytoplasm of wild-type virus-infected cells treated with CHX and displayed in sucrose gradients contained ribosome-loaded, truncated AU-rich mRNAs lacking the 3' UTR and poly(A) tails. The results suggest that the virion RNase is bound to polyribosomes by virtue of the reported association with translation machinery and cleaves the RNAs 5' to the AU elements. In contrast to the slow degradation of the of the residual 5' domain, the 3' UTR of the AU-rich mRNA and the GAPDH mRNA are rapidly degraded in wild-type virus-infected cells.

Formation of GW/P bodies as marker for microRNA-mediated regulation of innate immune signaling in THP-1 cells.

GW bodies (GWB or P bodies) are cytoplasmic foci thought to result from microRNA (miRNA) regulation of messenger RNA (mRNA) targets and subsequent mRNA degradation. The purpose of this study is to examine the effects of lipopolysaccharide (LPS) stimulation of human monocytes on GWB formation, miRNA induction, miRNA target regulation and downstream cytokine and chemokine expression. In response to LPS stimulation, the number of GWB consistently increased by twofold at 8 h after stimulation and this increase was abolished when the miRNA-effector proteins Rck/p54 or argonaute 2 were depleted. As the level of miR-146a increased from 19-fold up to 100-fold during LPS stimulation, the transfection of a miR-146a mimic into THP-1 cells was examined to determine whether miR-146a alone can induce similar changes in GWB. The results showed transfected miR-146a could produce a comparable increase in the number of GWB and this was accompanied by a reduction in major cytokines/chemokines induced by LPS. These data show that the increase in size and number of GWB may serve as a biomarker for miRNA-mediated gene regulation, and miR-146a has a significant role in the regulation of LPS-induced cytokine production in THP-1 cells.

TNF-alpha stimulation inhibits siRNA-mediated RNA interference through a mechanism involving poly-(A) tail stabilization.

The control of mRNA stability is a complex biological process that involves numerous factors, including microRNA (miRNA) and short interfering RNA (siRNA). Here, we show that short interfering RNA (siRNA) and microRNA share some similarities in their response to cellular stress. miR16 expedites the degradation of mRNAs containing AU-rich elements (ARE) in their 3' untranslated region (UTR). si20 is an siRNA designed to target a non-ARE sequence in the TNF 3'UTR. We found that both si20 and miR16/ARE-mediated degradation of mRNAs can be inhibited by stimulating cells with different stresses. By analyzing TNF-alpha stimulation-mediated stabilization of si20- and miR16-targeted mRNA, we show that this stabilization is not caused by modifying si20 and miR16 loading into Ago2 complexes, or mRNA targeting to Ago2, but by inhibiting mRNA deadenylation. This is the first report showing that a specific siRNA-mediated mRNA degradation can be regulated by inflammatory stimuli, and that deadenylation is involved in this siRNA-mediated mRNA decay.

Neuropathological effect of carbamate molluscicides on the land snail, Eobania vermiculata.

The present study was designed to investigate the neuropathological effect of the two carbamate pesticides: methomyl and methiocarb on the neurons of the buccal ganglia in the land snail Eobania vermiculata using topical application and baiting technique. Their in vivo effects on acetylcholinesterase (AChE, EC 3.1.1.7) activity were also investigated. Sublethal dose and concentration (1/4 LD(50) and 1/4 LC(50)) of both pesticides were used, and the experiment lasted for 14 days. Histopathological and ultrastuctural alterations in the buccal ganglia were more obvious after the baiting technique treatment than after the topical application method, and methomyl was found to be more toxic than methiocarb. These alterations included shrinkage of the perikarya of neurons, increased cytoplasmic basophilia, and extreme indentation of the plasma membrane. In addition, the nuclei appeared karyolitic, eccentric, and highly shrunken with an irregular nuclear envelope. The most outstanding symptom observed after topical application of methiocarb was a highly vacuolated cytoplasm with a peripheral increase in electron density associated with dense accumulations of free ribosomes. On the other hand, an increased number of lysosomes and autophagosomes were observed after topical application of methomyl. Mitochondrial damage, increased number of lipid droplets, and myelin figures were frequently observed in ganglia treated with either methomyl or methiocarb. Moreover, it was noticed that both compounds induced reductions in AChE activity. However, methomyl exhibited more potency in reducing AChE activity than methiocarb.

Accelerated proteasomal activity induced by Pb2+, Ga3+, or Cu2+ exposure does not induce degradation of alpha-synuclein.

The involvement of environmental heavy metals in Parkinson's disease (PD) has been suggested by epidemiologic studies; however, the mechanism of this effect is unknown. PD is characterized by the aggregation of alpha-synuclein in Lewy bodies. We previously showed that Pb2+ accelerates proteasomal activity. Therefore, we examined the effect of Pb2+, Ga3+, and Cu2+ on alpha-synuclein in human SH-SY5Y cells. The heavy metals induced an increase in heme-oxygenase-1 levels without significant cell death or ROS generation. The metals inhibited ALA-dehydratase, which is the inhibitory subunit of the proteasome, thereby accelerating proteasomal activity and decreasing protein levels of CDK-1 and PBGD. However, alpha-synuclein protein levels increased after exposure to metals, similar to the effect obtained with the proteasome inhibitor, hemin, suggesting that alpha-synuclein is inaccessible to proteasomal degradation. Indeed, electron microscopy revealed the formation of aggresomes in Pb2+- or hemin-treated cells. Thus, although heavy metals enhance proteasomal activity, alpha-synuclein is protected from degradation, and its protein levels and aggregation are increased.

The Ca2+-binding protein ALG-2 is recruited to endoplasmic reticulum exit sites by Sec31A and stabilizes the localization of Sec31A.

The formation of transport vesicles that bud from endoplasmic reticulum (ER) exit sites is dependent on the COPII coat made up of three components: the small GTPase Sar1, the Sec23/24 complex, and the Sec13/31 complex. Here, we provide evidence that apoptosis-linked gene 2 (ALG-2), a Ca(2+)-binding protein of unknown function, regulates the COPII function at ER exit sites in mammalian cells. ALG-2 bound to the Pro-rich region of Sec31A, a ubiquitously expressed mammalian orthologue of yeast Sec31, in a Ca(2+)-dependent manner and colocalized with Sec31A at ER exit sites. A Ca(2+) binding-deficient ALG-2 mutant, which did not bind Sec31A, lost the ability to localize to ER exit sites. Overexpression of the Pro-rich region of Sec31A or RNA interference-mediated Sec31A depletion also abolished the ALG-2 localization at these sites. In contrast, depletion of ALG-2 substantially reduced the level of Sec31A associated with the membrane at ER exit sites. Finally, treatment with a cell-permeable Ca(2+) chelator caused the mislocalization of ALG-2, which was accompanied by a reduced level of Sec31A at ER exit sites. We conclude that ALG-2 is recruited to ER exit sites via Ca(2+)-dependent interaction with Sec31A and in turn stabilizes the localization of Sec31A at these sites.

FGFR3 intracellular mutations induce tyrosine phosphorylation in the Golgi and defective glycosylation.

Mutations of the Fibroblast Growth Factor Receptor 3 (FGFR3) gene have been implicated in a series of skeletal dysplasias including hypochondroplasia, achondroplasia and thanatophoric dysplasia. The severity of these diseases ranges from mild dwarfism to severe dwarfism and to perinatal lethality, respectively. Although it is considered that the mutations give rise to constitutively active receptors, it remains unclear how the different mutations are functionally linked to the severity of the different pathologies. By examining various FGFR3 mutations in a HEK cell culture model, including the uncharacterized X807R mutation, it was found that only the mutations affecting the intracellular domain, induced premature receptor phosphorylation and inhibited receptor glycosylation, suggesting that premature receptor tyrosine phosphorylation of the native receptor inhibits its glycosylation. Moreover, these mutations appeared to be associated with elevated receptor signaling in the Golgi apparatus. In conclusion, although pathological severity could not be correlated with a single factor arising from FGFR3 mutations, these results suggest that intracellular domain mutations define a distinct means by which mutated FGFR3 could disrupt bone development.

Inhibition on LS-174T cell growth and activity of telomerase in vitro and in vivo by arsenic trioxide.

Arsenic trioxide (As2O3) shows a significant therapeutic effect upon acute promyelocytic leukemia (APL) and can induce the apoptosis of NB4 cells, which attracts scholars' great attention. Especially, the therapeutic effect on solid carcinoma has been paid more close attention to. The present study is to evaluate the effect of As2O3 on human colorectal carcinoma cells (LS-174T cell) and the activity of telomerase in vitro and in vivo. This research made use of the electron microscope, polymerase chain reaction-enzyme-linked immunosorbent assay (PCR-ELISA), fluorescence-activated cell sorter (FACS), MTT in vitro and in vivo (LS-174T xenograft model of nude mice). With the increasing concentration of As2O3, the ratio of living cells to dead cells decreased significantly, and the IC50 value was 5.23micromol/L; cells of the experimental groups endured a series of morphological changes similar to the features of apoptosis. Apoptosis curve of FACS pictures appeared after 24h, and the cells showed apoptosis in a time-dependent manner; As2O3 can inhibit the activity of telomerase of the cell extraction, obviously, in a concentration-dependent and time-dependent manner after 24h. As to the inhibition impact of As2O3 on the xenograft model of nude mice in the two indexes, tumor volume and weight, there was a significant difference between As2O3 and the control group; there was no difference between As2O3 and the fluorouracil (5-FU) group; in the group of peritoneal injections of As2O3, the cancer cells connected loosely with each other, nucleus changed markedly, and heterochromatin concentrated under the nucleus membrane. From the in vitro and in vivo experiment, we can see that As2O3 inhibited LS-174T cell growth mainly by inducing cell apoptosis, partly by the inhibition of telomerase activity.

Aflatoxin treatment brings about generation of multinucleate giant spermatids (symplasts) through opening of cytoplasmic bridges: light and transmission electron microscopic study in Swiss mouse.

Aflatoxins are dietary mycotoxins, which are a health hazard. Sub-symptomatic exposure to aflatoxins is known to produce male reproductive toxic effects with several manifestations. With a view to find if aflatoxins would produce multinucleate giant cells or symplasts in the seminiferous epithelium, we treated male Swiss mice with aflatoxin B(1) for 35 days and subjected the testis to light and transmission electron microscopic analysis. We found abundant symplastic spermatids in the seminiferous epithelium of treated mice. The origin of these cells was traced to opening of cytoplamic bridges. Due to widening of cytoplasmic bridge, the cytoplasm of spermatid(s) in a clone entered a cytoplasm-rich spermatid, followed by the nucleus/nuclei. Subsequently, the bridge(s) collapsed resulting in spherical symplasts. The study, in addition to revealing yet another manifestation of aflatoxin-induced disruption of spermatogenesis, also provides first direct evidence for opening of cytoplasmic bridges as the mechanism underlying origin of spermatid symplasts.

Tryptamine induces cell death with ultrastructural features of autophagy in neurons and glia: Possible relevance for neurodegenerative disorders.

Tryptamine derivatives are a family of biogenic amines that have been suggested to be modulators of brain function at physiological concentrations. However, pharmacological concentrations of these amines display amphetamine-like properties, and they seem to play a role in brain disorders. Amphetamines induce autophagy in nerve cells, and this type of cell death has also been involved in neurodegenerative diseases. In the present work, we clearly demonstrate for the very first time that high concentrations of tryptamine (0.1-1 mM) induce autophagy in HT22 and SK-N-SH nerve cell lines and in primary cultures of astrocytes, glial cells being less sensitive than neurons. Ultrastructural cell morphology shows all of the typical hallmarks of autophagy. There is no nuclear chromatin condensation, endoplasmic reticulum and mitochondria are swollen, and a great number of double-membraned autophagosomes and residual bodies can be shown in the cytoplasm. Autophagosomes and residual bodies contain mitochondria, membranes, and vesicles and remain unabridged until the cell membrane is disrupted and the cell dies. The same results have been found when cells were incubated with high concentrations of 5-methoxytryptamine (0.1-1 mM). Our results establish a possible link between the role of tryptamine derivatives in brain disorders and the presence of autophagic cell death in these kinds of disorders.

Analysis of eosinophilic round bodies formed after injection of enamel matrix derivative into the backs of rats.

BACKGROUND: Enamel matrix derivative (EMD) is used in dental clinics for the regeneration of alveolar bone. Its effects have not yet been clarified, although it induces eosinophilic round bodies (ERBs) and cartilage formation at the injection site. The objective of this experiment was to examine the histopathologic and biochemical properties of ERBs formed after EMD injection. METHODS: The backs of Sprague-Dawley rats injected with various concentrations of EMD were examined histopathologically. For biochemical examinations, ERBs were microdissected out from the sections. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), matrix-assisted laser-desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS), and database analysis of ERBs were carried out. RESULTS: The histopathological findings were consistent with a foreign body reaction. Numerous ERBs were observed 7 days after injection of 30.0 mg/ml EMD. Histopathologically, ERBs did not contain polysaccharide, amyloid, or hemosiderin. The cells surrounding ERBs were not macrophages or vascular endothelial cells. SDS-PAGE of the microdissected ERBs revealed an intense band at around the 40-kDa region. MALDI-TOF MS showed that the spectrum for ERBs has only a single strong ion intensity. Analysis of the amino acid sequence revealed that the ERBs were composed of various molecular fragments, which all contained an identical seven amino acid sequence. In addition, these peptides are a component of amelogenin. CONCLUSIONS: A high concentration of EMD induces ERBs that consist of a 40-kDa protein which includes a constituent part of amelogenin. The ERBs (or remaining EMD) might promote mesenchymal cell differentiation into hard tissue-forming cells around the EMD injection site.

Subcellular distribution of di-(2-ethylhexyl)phthalate in rat testis.

Subcellular distribution of di-(2-ethylhexyl)phthalate (DEHP) in the testis was studied by single oral administration of [3,4,5,6-(3)H]-phthalic acid di-(2-ethylhexyl) ester (DEHP-3H) or phthalic acid di-(2-ethyl[1-(3)H]hexyl) ester (3H-DEHP) to 8-week-old male rats. Autoradiographs and electron microscopic autoradiographs were prepared from the testis, liver and kidney at 6 and 24 hr after administration and distribution of radioactive materials in the tissues were observed. In the autoradiographic specimen at 6 hr after administration of DEHP 3H-labeled at phthalic acid moiety (DEHP-3H), many grains were observed in the testis, mainly at the basal area of seminiferous tubules at the stages IX to I of the spermatogenic cycle. Electron microscopic autoradiographs taken at the same time revealed that localization of grains were in the smooth-surfaced endoplasmic reticulum and mitochondria of Sertoli cells. A few grains were also present at the Golgi apparatus and lysosome of Sertoli cells, and at the interfaces between the Sertoli cells or between Sertoli cells and spermatocytes, and in the cytoplasm of spermatocytes. Autoradiographs of the liver revealed grains in the centrilobular hepatocytes, localized at mitochondria, rough-surfaced endoplasmic reticulum and peroxisomes. In the kidney, the radioactivity was localized at the brush border of the tubular cells in the pars recta of proximal tubules. In the 24-hr specimen, the grain density in the seminiferous tubules obviously decreased. On the other hand, by autoradiography with DEHP 3H-labeled at the alcohol (3H-DEHP), only a few grains were observed in autoradiographs of the testes at 6 hr after administration. No grains were noted in autoradiographs of the liver and kidney with 3H-DEHP. The results showed that the phthalic acid ester was splitted rapidly in the body and only the phthalic acid moiety distributed into the cells.

Accumulation of heme oxygenase-1 (HSP32) in Xenopus laevis A6 kidney epithelial cells treated with sodium arsenite, cadmium chloride or proteasomal inhibitors.

The present study examined the effect of sodium arsenite, cadmium chloride, heat shock and the proteasomal inhibitors MG132, withaferin A and celastrol on heme oxygenase-1 (HO-1; also known as HSP32) accumulation in Xenopus laevis A6 kidney epithelial cells. Immunoblot analysis revealed that HO-1 accumulation was not induced by heat shock but was enhanced by sodium arsenite and cadmium chloride in a dose- and time-dependent fashion. Immunocytochemistry revealed that these metals induced HO-1 accumulation in a granular pattern primarily in the cytoplasm. Additionally, in 20% of the cells arsenite induced the formation of large HO-1-containing perinuclear structures. In cells recovering from sodium arsenite or cadmium chloride treatment, HO-1 accumulation initially increased to a maximum at 12h followed by a 50% reduction at 48 h. This initial increase in HO-1 levels was likely the result of new synthesis as it was inhibited by cycloheximide. Interestingly, treatment of cells with a mild heat shock enhanced HO-1 accumulation induced by low concentrations of sodium arsenite and cadmium chloride. Finally, we determined that HO-1 accumulation was induced in A6 cells by the proteasomal inhibitors, MG132, withaferin A and celastrol. An examination of heavy metal and proteasomal inhibitor-induced HO-1 accumulation in amphibians is of importance given the presence of toxic heavy metals in aquatic habitats.

PaCS is a novel cytoplasmic structure containing functional proteasome and inducible by cytokines/trophic factors.

A variety of ubiquitinated protein-containing cytoplasmic structures has been reported, from aggresomes to aggresome-like induced structures/sequestosomes or particle-rich cytoplasmic structures (PaCSs) that we recently observed in some human diseases. Nevertheless, the morphological and cytochemical patterns of the different structures remain largely unknown thus jeopardizing their univocal identification. Here, we show that PaCSs resulted from proteasome and polyubiquitinated protein accumulation into well-demarcated, membrane-free, cytoskeleton-poor areas enriched in glycogen and glycosaminoglycans. A major requirement for PaCS detection by either electron or confocal microscopy was the addition of osmium to aldehyde fixatives. However, by analyzing living cells, we found that proteasome chymotrypsin-like activity concentrated in well-defined cytoplasmic structures identified as PaCSs by ultrastructural morphology and immunocytochemistry of the same cells. PaCSs differed ultrastructurally and cytochemically from sequestosomes which may coexist with PaCSs. In human dendritic or natural killer cells, PaCSs were induced in vitro by cytokines/trophic factors during differentiation/activation from blood progenitors. Our results provide evidence that PaCS is indeed a novel distinctive cytoplasmic structure which may play a critical role in the ubiquitin-proteasome system response to immune, infectious or proneoplastic stimuli.

Mechanism of microtubule array expansion in the cytokinetic phragmoplast.

In land plants, the cell plate partitions the daughter cells at cytokinesis. The cell plate initially forms between daughter nuclei and expands centrifugally until reaching the plasma membrane. The centrifugal development of the cell plate is driven by the centrifugal expansion of the phragmoplast microtubule array, but the molecular mechanism underlying this expansion is unknown. Here, we show that the phragmoplast array comprises stable microtubule bundles and dynamic microtubules. We find that the dynamic microtubules are nucleated by γ-tubulin on stable bundles. The dynamic microtubules elongate at the plus ends and form new bundles preferentially at the leading edge of the phragmoplast. At the same time, they are moved away from the cell plate, maintaining a restricted distribution of minus ends. We propose that cycles of attachment of γ-tubulin complexes onto the microtubule bundles, microtubule nucleation and bundling, accompanied by minus-end-directed motility, drive the centrifugal development of the phragmoplast.