cDNA isolation, characterization, and protein intracellular localization of a katanin-like p60 subunit from Arabidopsis thaliana.

Katanin, a heterodimeric protein with ATP-dependent microtubule-severing activity, localizes to the centrosome in animal cells. Widespread occurrence is suspected as several species contain homologs to the katanin p60 subunit. Recently we isolated an Arabidopsis thaliana cDNA with significant identity to the p60 subunit of sea urchin katanin. Like p60, the encoded protein is a member of the AAA superfamily of ATPases, containing the Walker ATP binding consensus and the signature AAA minimal consensus sequences within a single larger AAA/CAD amino acid motif. Phylogenetic analysis placed the encoded protein in the AAA subfamily of cytoskeleton-interactive proteins, where it formed a strongly supported clade with 4 other members identified as katanin p60 subunits. The clone was named AtKSS (Arabidopsis thaliana katanin-like protein small subunit). Western blots, performed using a polyclonal antibody raised against recombinant AtKSS, revealed AtKSS is present in protein extracts of all Arabidopsis organs examined. To evaluate potential interactions between AtKSS and the cytoskeleton, the intracellular localization of AtKSS was correlated with that of tubulin. AtKSS was found in perinuclear regions during interphase, surrounding the spindle poles during mitosis, but was absent from the preprophase band and phragmoplast microtubule arrays. These data support the thesis that AtKSS is an Arabidopsis homolog of the p60 subunit of katanin. Its cell cycle-dependent distribution is consistent with microtubule-severing activity, but additional studies will better define its role.

Rapid degradation of auxin/indoleacetic acid proteins requires conserved amino acids of domain II and is proteasome dependent.

Auxin rapidly induces auxin/indoleacetic acid (Aux/IAA) transcription. The proteins encoded are short-lived nucleus-localized transcriptional regulators that share four conserved domains. In a transient assay measuring protein accumulation, an Aux/IAA 13-amino acid domain II consensus sequence was sufficient to target firefly luciferase (LUC) for low protein accumulation equivalent to that observed previously for full-length PSIAA6. Single amino acid substitutions in these 13 amino acids, corresponding to known auxin response mutants, resulted in a sixfold to 20-fold increase in protein accumulation. Naturally occurring variant amino acids had no effect. Residues identified as essential by single alanine substitutions were not sufficient when all flanking amino acids were alanine, indicating the importance of flanking regions. Using direct protein degradation measurements in transgenic Arabidopsis seedlings, full-length IAA1, PSIAA6, and the N-terminal 73 PSIAA6 amino acids targeted LUC for rapid degradation with 8-min half-lives. The C-terminal 109 amino acids did not affect LUC half-life. Smaller regions containing domain II also targeted LUC for rapid degradation, but the rates were not equivalent to those of the full-length protein. A single domain II substitution in the context of full-length PSIAA6 increased half-life 30-fold. Proteasome inhibitors affected Aux/IAA::LUC fusion protein accumulation, demonstrating the involvement of the proteasome.

Auxin modulates the degradation rate of Aux/IAA proteins.

Aux/IAA gene family members were first identified by their rapid transcriptional increase in response to auxin. Auxin/indole-3-acetic acid protein (Aux/IAA) luciferase (LUC) fusions expressed in Arabidopsis under control of a non-auxin-responsive promoter were used to monitor the effect of auxin on protein abundance independent of transcriptional regulation by auxin. After 2 hr in the presence of 1 microM exogenous dichlorophenoxyacetic acid (2,4D), a synthetic auxin, the levels of pea IAA6 (PSIAA6) and Arabidopsis IAA1 LUC activity were 35% and 67%, respectively, of mock-treated genetically identical seedlings, whereas the activity of LUC alone from equivalently treated seedlings remained unaltered. The steady-state level of an Aux/IAA fusion protein lacking domain II, one of the conserved domains found in all Aux/IAA proteins, was not reduced in the presence of auxin. Higher levels of exogenous auxin were required to affect the steady-state level of the PSIAA6LUC fusion with a point mutation in domain II. A 13-aa consensus sequence from domain II fused to LUC created an auxin-responsive fusion protein. The change in steady-state levels in response to auxin is extremely rapid, with a decrease in LUC activity detectable by 2 min after auxin application. Direct half-life measurements show that the decrease caused by exogenous auxin is due to the decrease in fusion protein half-life. These results suggest that auxin rapidly modulates the degradation rate of Aux/IAA proteins, with higher levels of auxin increasing the proteolytic rate of Aux/IAA family members.

Interactions of the COP9 signalosome with the E3 ubiquitin ligase SCFTIRI in mediating auxin response.

The COP9 signalosome is an evolutionary conserved multiprotein complex of unknown function that acts as a negative regulator of photomorphogenic seedling development in Arabidopsis. Here, we show that plants with reduced COP9 signalosome levels had decreased auxin response similar to loss-of-function mutants of the E3 ubiquitin ligase SCFTIR1. Furthermore, we found that the COP9 signalosome and SCFTIR1 interacted in vivo and that the COP9 signalosome was required for efficient degradation of PSIAA6, a candidate substrate of SCFTIR1. Thus, the COP9 signalosome may play an important role in mediating E3 ubiquitin ligase-mediated responses.

Protein degradation in signaling.

Recent studies have linked proteolysis by the ubiquitin/proteasome pathway to a variety of signaling pathways in higher plants. These links were uncovered by characterization of mutants altered in signaling or by targeted disruption of components of the proteolytic pathway. Significant advances have recently revealed connections between proteolysis and hormone responses, light perception, environmental adaptation, and floral development.

Ubiquitin-specific proteases from Arabidopsis thaliana: cloning of AtUBP5 and analysis of substrate specificity of AtUBP3, AtUBP4, and AtUBP5 using Escherichia coli in vivo and in vitro assays.

A cDNA for a new ubiquitin-specific protease (UBP), AtUBP5, was identified from Arabidopsis thaliana flower mRNA using an oligonucleotide made against the conserved UBP cysteine (Cys) box. The 924-amino-acid AtUBP5 contains the regions characteristic of all UBPs and has 35% identity and 53% similarity overall to a mammalian UBP (Unp), resulting from additional significant similarity outside these regions. AtUBP5 has 48% identity and 58% similarity overall to two uncharacterized Arabidopsis genomic sequences but is distinct outside the UBP conserved regions from two other previously published Arabidopsis UBPs, AtUBP3 and -4. Using in vivo Escherichia coli assays, which allow co-expression of GSTAtUBPs and substrates, we show that all three UBPs were active. AtUBP5 was active without 311 amino acids N-terminal to the active site cysteine, or without 233 nonconserved amino acids between the Cys and His boxes, or without both, indicating the core region was sufficient. In in vivo and in vitro assays, GSTAtUBP3, -4, and -5 exhibited preference for specific Ub-Ub linkages, suggesting accessibility and/or conformation is important and demonstrating that these enzymes cleave post-translationally. A chimeric UBP consisting of the AtUBP5 Cys box with AtUBP3 amino acids was active and exhibited AtUBP3 specificity, indicating that the modular nature of UBPs and specificity for cleavage sites is not determined by the Cys box.

Histidine-tagged ubiquitin substitutes for wild-type ubiquitin in Saccharomyces cerevisiae and facilitates isolation and identification of in vivo substrates of the ubiquitin pathway.

A general method for purification of any substrate of the ubiquitin pathway, the major eukaryotic proteolytic pathway, should utilize the common characteristic of covalent linkage of ubiquitin to substrate lysyl residues. The utility of a N-terminal histidine-tagged ubiquitin (HisUb) for in vivo conjugation and isolation of ubiquitinated proteins by metal chelation chromatography is conditioned by the requirement that HisUb conjugate to the same set of proteins as wild-type ubiquitin. Stringent in vivo tests with Saccharomyces cerevisiae strains expressing ubiquitins only from plasmids were performed to show that HisUb could substitute for wild-type ubiquitin. The utility of HisUb as a method for purification of proteins ubiquitinated in vivo was demonstrated by metal chelation chromatography of yeast extracts expressing HisUb and immunoblotting for Rpb1, the largest subunit of RNA polymerase II. A fraction of Rpb1 was present in the ubiquitinated form in vivo. The ability to use HisUb expression in transgenic organisms that retain expression of their endogenous ubiquitin genes was demonstrated through transgenic Arabidopsis thaliana expressing HisUb or its variant HisUbK48R. UbK48R is a version of ubiquitin capable of conjugation to proteins, but cannot serve as an attachment site for ubiquitin via the major in vivo interubiquitin linkage. Whereas transgenic plants expressing HisUb showed insignificant enrichment of ubiquitinated proteins, transgenic Arabidopsis lines expressing HisUbK48R gave a much better yield.

Degradation of Aux/IAA proteins is essential for normal auxin signalling.

The growth substance auxin mediates many cellular processes, including division, elongation and differentiation. PSIAA6 is a member of the Aux/IAA family of short-lived putative transcriptional regulators that share four conserved domains and whose mRNAs are rapidly induced in the presence of auxin. Here PSIAA6 was shown to serve as a dominant transferable degradation signal when present as a translational fusion with firefly luciferase (LUC), with an in vivo half-life of 13.5 min in transgenic Arabidopsis seedlings. In a transient assay system in tobacco protoplasts using steady-state differences as an indirect measure of protein half-life, LUC fusions with full-length PSIAA6 and IAA1, an Aux/IAA protein from Arabidopsis, resulted in protein accumulations that were 3.5 and 1. 0%, respectively, of that with LUC alone. An N-terminal region spanning conserved domain II of PSIAA6 containing amino acids 18-73 was shown to contain the necessary cis-acting element to confer low protein accumulation onto LUC, while a fusion protein with PSIAA6 amino acids 71-179 had only a slight effect. Single amino acid substitutions of PSIAA6 in conserved domain II, equivalent to those found in two alleles of axr3, a gene that encodes Aux/IAA protein IAA17, resulted in a greater than 50-fold increase in protein accumulation. Thus, the same mutations resulting in an altered auxin response phenotype increase Aux/IAA protein accumulation, providing a direct link between these two processes. In support of this model, transgenic plants engineered to over-express IAA17 have an axr3-like phenotype. Together, these data suggest that rapid degradation of Aux/IAA proteins is necessary for a normal auxin response.

Polypeptide tags, ubiquitous modifiers for plant protein regulation.

Evidence has emerged over the past few years that plants, like animals and fungi, employ a variety of polypeptides as tags to reversibly or irreversibly affect the function, structure, location, and/or turnover of numerous intracellular proteins. In plants, known polypeptide tags include ubiquitin, SUMO, RUB, and APG12, with the possibility of others. These modifiers are typically added post-translationally using individual sets of conjugase pathways that attach the polypeptides via an isopeptide bond to epsilon-lysyl amino group(s) in the targets. Often the tags can be removed subsequently by unique proteases that specifically cleave only the isopeptide bond. Examples also exist where the tag is added during translation upon fusion of the coding sequence of the tag with that of the target. Based on the number and diversity of targets, ubiquitin is the most influential modifier which mainly serves as a reusable signal for selective protein degradation by the 26S proteasome. In contrast, SUMO, RUB and APG12 become attached to a more limited number of targets and appear to have specialized functions, including roles in nuclear pore assembly/function, cell-cycle regulation, and lysosomal/vacuole trafficking, respectively. Based on their widespread occurrence in plants and their pervasive role in various biological processes, polypeptide tags likely play a prominent role in plant cell regulation.

Hematuria from arteriovesical fistula: unusual presentation of ruptured iliac artery aneurysm.

Iliac artery aneurysm rupture can be rapidly fatal if not diagnosed immediately. These aneurysms usually present in patients with other aneurysmal diseases of the aortoiliac arterial system. If not diagnosed and surgically repaired, iliac artery aneurysms can proceed to expand and ultimately rupture, usually presenting with back, flank, or abdominal pain and, possibly, signs of systemic shock. We present an unusual case report of a common iliac artery aneurysm rupture presenting as gross hematuria due to an arteriovesical fistula. Only three other cases of arteriovesical fistulae have been reported previously. Unlike the case presented, all three of these cases involved trauma or surgical instrumentation or manipulation of the bladder.

The rub family of ubiquitin-like proteins. Crystal structure of Arabidopsis rub1 and expression of multiple rubs in Arabidopsis.

Several proteins with significant identity to ubiquitin have been characterized recently. In contrast to ubiquitin's main role in targeting proteins for degradation, a described function of one family of ubiquitin-related proteins, the Rub family, is to serve as a stable post-translational modification of a complex involved in the G1-to-S cell cycle transition. Rub proteins have been found in animals, plants, and fungi and consist of 76 residues with 52-63% identity to ubiquitin. In this study three different RUB proteins within the plant Arabidopsis are identified; two differ by only 1 amino acid, while the third is only 77.6% identical to the other two. Genes encoding all three are expressed in multiple organs. In addition, we report the crystal structure of higher plant RUB1 at 1.7-A resolution to help elucidate the functional differences between Rub and ubiquitin. RUB1 contains a single globular domain with a flexible COOH-terminal extension. The overall RUB1 structure is very similar to ubiquitin. The majority of the amino acid differences between RUB1 and ubiquitin map to the surface. These changes alter the electrostatic surface potential in two regions and likely confer specificity between ubiquitin and RUB1 and their ubiquitin-activating enzyme (E1) or E1-like activating enzymes.

Engineering in vivo instability of firefly luciferase and Escherichia coli beta-glucuronidase in higher plants using recognition elements from the ubiquitin pathway.

The ubiquitin pathway targets proteins for degradation through the post-translational covalent attachment of the 76 amino acid protein ubiquitin to epsilon-amino lysyl groups on substrate proteins. Two instability determinants recognized by the ubiquitin pathway in Saccharomyces cerevisiae have been identified. One is described by the N-end rule and requires specific destabilizing residues at the substrate protein N-termini along with a proximal lysyl residue for ubiquitin conjugation. The second is a linear uncleavable N-terminal ubiquitin moiety. The ability of these two determinants to function in higher plants was investigated in tobacco protoplast transient transfection assays using DNA encoding variants of well characterized reporter enzymes as substrates: firefly luciferase that is localized to peroxisomes (pxLUC), a cytosolic version of LUC (cLUC), and Escherichia coli beta-glucuronidase (GUS). cLUC with phenylalanine encoded at its mature N-terminus was 10-fold less abundant than cLUC with methionine at its mature N-terminus. GUS with phenylalanine encoded at its mature N-terminus was 3-fold less abundant than GUS with methionine at its mature N-terminus. The presence of a uncleavable N-terminal ubiquitin fusion resulted in 50-fold lower protein accumulation of cLUC, but had no effect on GUS. Both instability determinants had a much larger effect on cLUC than on pxLUC, suggesting that these degradation signals are either unrecognized or poorly recognized in the peroxisomes.

The ubiquitin-related protein RUB1 and auxin response in Arabidopsis.

The AXR1 (auxin-resistant) protein, which has features of the ubiquitin-activating enzyme E1, is required for normal response to the plant hormone auxin in Arabidopsis thaliana. ECR1 functions together with AXR1 to activate members of the RUB/NEDD8 family of ubiquitin-related proteins. Extracts from mutant seedlings lacking AXR1 did not promote formation of the RUB-ECR1 thiolester, indicating that AXR1 is the major activity in this tissue. AXR1 was localized primarily to the nucleus of dividing and elongating cells, suggesting that the targets of RUB modification are nuclear. These results indicate that auxin response depends on RUB modification of one or more nuclear proteins.

Modification of yeast Cdc53p by the ubiquitin-related protein rub1p affects function of the SCFCdc4 complex.

The RUB1/NEDD-8 family of ubiquitin-related genes is widely represented among eukaryotes. Here we report that Cdc53p in Saccharomyces cerevisiae, a member of the Cullin family of proteins, is stably modified by the covalent attachment of a single Rub1p molecule. Two genes have been identified that are required for Rub1p conjugation to Cdc53p. The first gene, designated ENR2, encodes a protein with sequence similarity to the amino-terminal half of the ubiquitin-activating enzyme. By analogy with Aos1p, we infer that Enr2p functions in a bipartite Rub1p-activating enzyme. The second gene is SKP1, shown previously to be required for some ubiquitin-conjugation events. A deletion allele of ENR2 is lethal with temperature-sensitive alleles of cdc34 and enhances the phenotypes of cdc4, cdc53, and skp1, strongly implying that Rub1p conjugation to Cdc53p is required for optimal assembly or function of the E3 complex SCFCdc4. Consistent with this model, both enr2delta and an allele of Cdc53p that is not Rub1p modified, render cells sensitive to alterations in the levels of Cdc4p, Cdc34p, and Cdc53p.

A model for the evolution of polyubiquitin genes from the study of Arabidopsis thaliana ecotypes.

Polyubiquitin genes encode the highly conserved 76-amino acid protein ubiquitin that is covalently attached to substrate proteins targeting most for degradation. Polyubiquitin genes are characterized by the presence of tandem repeats of the 228 bp that encode a ubiquitin monomer. Five polyubiquitin genes UBQ3, UBQ4, UBQ10, UBQ11, and UBQ14, previously isolated from Arabidopsis thaliana ecotype Columbia [10] encode identical mature ubiquitin proteins, but differ in synonymous substitutions, nature of amino acids terminating the open reading frame, and in the number of ubiquitin repeats. The presence of these five genes in nine other Arabidopsis ecotypes was verified by polymerase chain reaction (PCR). Size differences in UBQ3 and UBQ11 amplified products from several ecotypes were observed, suggesting that alleles differ in ubiquitin repeat number. DNA sequence of UBQ11 alleles from each size class (ecotypes Be-0, Ler. and Rld-0) verified that PCR product size differences resulted from changes in the number of ubiquitin repeats. Nucleotide sequence between two UBQ11 alleles containing the same number of repeats was identical. Transcript size differences for UBQ3 and UBQ11 mRNAs between ecotypes Columbia and Landsberg indicated that repeat number changes did not inactivate these genes. Nucleotide sequence comparisons between UBQ11 repeats from different ecotypes suggest that first repeats are related to each other and last repeats are related to each other. We hypothesize that changes in UBQ11 ubiquitin repeat number occurred via the contraction and/or expansion of specific internal repeats or portions thereof by misalignment of alleles and recombination, most likely via unequal crossing-over events.

AtUBP3 and AtUBP4 are two closely related Arabidopsis thaliana ubiquitin-specific proteases present in the nucleus.

The ubiquitin-specific proteases (UBPs) are a class of enzymes vital to the ubiquitin pathway. These enzymes cleave ubiquitin at its C-terminus from two types of substrates containing (i) ubiquitin in an alpha-amino linkage, as found in the primary ubiquitin translation products, polyubiquitin and ubiquitin-ribosomal fusion proteins, or (ii) ubiquitin in an epsilon-amino linkage, as found in multiubiquitin chains either unattached or conjugated to cellular proteins. We have isolated cDNAs for two Arabidopsis thaliana genes, AtUBP3 and AtUBP4, which encode UBPs that are 93% identical. These two cDNAs represent the only two members of this subgroup and encode the smallest UBPs described to date in any organism. Using in vivo assays in Escherichia coli that allow the coexpression of a UBP with a putative substrate, we have shown that AtUBP3 and AtUBP4 can specifically deubiquitinate the artificial substrate Ub-X-beta-gal but cannot act upon the natural alpha-amino-linked ubiquitin fusions Arabidopsis Ub-CEP52 and Arabidopsis polyubiquitin. Affinity-purified antibody prepared against AtUBP3 expressed in E. coli recognizes both AtUBP3 and AtUBP4. AtUBP3 and/or AtUBP4 are present in all Arabidopsis organs examined and at multiple developmental stages. Subcellular localization studies show that AtUBP3 and/or AtUBP4 are present in nuclear extracts. Possible physiological roles for these UBPs are discussed.

Independent modulation of Arabidopsis thaliana polyubiquitin mRNAs in different organs and in response to environmental changes.

The highly conserved protein ubiquitin is encoded by five polyubiquitin genes in Arabidopsis thaliana ecotype Columbia that have been divided into two subtypes, the UBQ3/UBQ4 subtype and the UBQ10/UBQ11/UBQ14 subtype. Northern analysis using gene-specific oligonucleotides as hybridization probes and enzyme activity measurements from transgenic plants expressing beta-glucuronidase (GUS) under the control of individual polyubiquitin 5' flanking regions were used to determine the development and environmental regulation of polyubiquitin transcription and mRNA accumulation. Polyubiquitin mRNA levels within and between subtypes were independently modulated. UBQ3 mRNA levels were three-fold higher than UBQ4 mRNA levels in vegetative organs, but only two-thirds of the UBQ4 mRNA levels in flowers. UBQ3 mRNA was modulated by dark/light treatments, while mRNAs from UBQ and all members of the other subtype were unaffected. Similarly, within the UBQ10/UBQ11/UBQ14 subtype, UBQ11/UBQ14 mRNAs were modulated differently in seedlings after a two-hour heat-shock treatment. Among all the polyubiquitin genes, UBQ10 mRNA level was the most constant in all organs and environmental conditions examined. Transgenic plants transformed with a UBQ10 5' flanking region::GUS gene contained higher levels of GUS activity than transgenic plants expressing GUS under the control of UBQ3 5' flanking regions. In conclusion, the relative abundance of different Arabidopsis polyubiquitin mRNAs, even those produced from highly similar genes within a subtype, appears to be modulated independently in response to developmental and environmental cues.