How proteins get into microbodies (peroxisomes, glyoxysomes, glycosomes)

Delineating transitions during the evolution of specialised peroxisomes: Glycosome formation in kinetoplastid and diplonemid protists

One peculiarity of protists belonging to classes Kinetoplastea and Diplonemea within the phylum Euglenozoa is compartmentalisation of most glycolytic enzymes within peroxisomes that are hence called glycosomes. This pathway is not sequestered in peroxisomes of the third Euglenozoan class, Euglenida. Previous analysis of well-studied kinetoplastids, the ‘TriTryps’ parasites Trypanosoma brucei, Trypanosoma cruzi and Leishmania spp., identified within glycosomes other metabolic processes usually not present in peroxisomes. In addition, trypanosomatid peroxins, i.e. proteins involved in biogenesis of these organelles, are divergent from human and yeast orthologues. In recent years, genomes, transcriptomes and proteomes for a variety of euglenozoans have become available. Here, we track the possible evolution of glycosomes by querying these databases, as well as the genome of Naegleria gruberi, a non-euglenozoan, which belongs to the same protist supergroup Discoba. We searched for orthologues of TriTryps proteins involved in glycosomal metabolism and biogenesis. Predicted cellular location(s) of each metabolic enzyme identified was inferred from presence or absence of peroxisomal-targeting signals. Combined with a survey of relevant literature, we refine extensively our previously postulated hypothesis about glycosome evolution. The data agree glycolysis was compartmentalised in a common ancestor of the kinetoplastids and diplonemids, yet additionally indicates most other processes found in glycosomes of extant trypanosomatids, but not in peroxisomes of other eukaryotes were either sequestered in this ancestor or shortly after separation of the two lineages. In contrast, peroxin divergence is evident in all euglenozoans. Following their gain of pathway complexity, subsequent evolution of peroxisome/glycosome function is complex. We hypothesize compartmentalisation in glycosomes of glycolytic enzymes, their cofactors and subsequently other metabolic enzymes provided selective advantage to kinetoplastids and diplonemids during their evolution in changing marine environments. We contend two specific properties derived from the ancestral peroxisomes were key: existence of nonselective pores for small solutes and the possibility of high turnover by pexophagy. Critically, such pores and pexophagy are characterised in extant trypanosomatids. Increasing amenability of free-living kinetoplastids and recently isolated diplonemids to experimental study means our hypothesis and interpretation of bioinformatic data are suited to experimental interrogation.

Recent insights into peroxisome biogenesis and associated diseases

Peroxisomes are single-membrane organelles present in eukaryotes. The functional importance of peroxisomes in humans is represented by peroxisome-deficient peroxisome biogenesis disorders (PBDs), including Zellweger syndrome. Defects in the genes that encode the 14 peroxins that are required for peroxisomal membrane assembly, matrix protein import and division have been identified in PBDs. A number of recent findings have advanced our understanding of the biology, physiology and consequences of functional defects in peroxisomes. In this Review, we discuss a cooperative cell defense mechanisms against oxidative stress that involves the localization of BAK (also known as BAK1) to peroxisomes, which alters peroxisomal membrane permeability, resulting in the export of catalase, a peroxisomal enzyme. Another important recent finding is the discovery of a nucleoside diphosphate kinase-like protein that has been shown to be essential for how the energy GTP is generated and provided for the fission of peroxisomes. With regard to PBDs, we newly identified a mild mutation, Pex26-F51L that causes only hearing loss. We will also discuss findings from a new PBD model mouse defective in Pex14, which manifested dysregulation of the BDNF-TrkB pathway, an essential signaling pathway in cerebellar morphogenesis. Here, we thus aim to provide a current view of peroxisome biogenesis and the molecular pathogenesis of PBDs.

Dissecting Leishmania infantum Energy Metabolism - A Systems Perspective

Leishmania infantum, causative agent of visceral leishmaniasis in humans, illustrates a complex lifecycle pertaining to two extreme environments, namely, the gut of the sandfly vector and human macrophages. Leishmania is capable of dynamically adapting and tactically switching between these critically hostile situations. The possible metabolic routes ventured by the parasite to achieve this exceptional adaptation to its varying environments are still poorly understood. In this study, we present an extensively reconstructed energy metabolism network of Leishmania infantum as an attempt to identify certain strategic metabolic routes preferred by the parasite to optimize its survival in such dynamic environments. The reconstructed network consists of 142 genes encoding for enzymes performing 237 reactions distributed across five distinct model compartments. We annotated the subcellular locations of different enzymes and their reactions on the basis of strong literature evidence and sequence-based detection of cellular localization signal within a protein sequence. To explore the diverse features of parasite metabolism the metabolic network was implemented and analyzed as a constraint-based model. Using a systems-based approach, we also put forth an extensive set of lethal reaction knockouts; some of which were validated using published data on Leishmania species. Performing a robustness analysis, the model was rigorously validated and tested for the secretion of overflow metabolites specific to Leishmania under varying extracellular oxygen uptake rate. Further, the fate of important non-essential amino acids in L. infantum metabolism was investigated. Stage-specific scenarios of L. infantum energy metabolism were incorporated in the model and key metabolic differences were outlined. Analysis of the model revealed the essentiality of glucose uptake, succinate fermentation, glutamate biosynthesis and an active TCA cycle as driving forces for parasite energy metabolism and its optimal growth. Finally, through our in silico knockout analysis, we could identify possible therapeutic targets that provide experimentally testable hypotheses.

Peroxisome-mitochondria interplay and disease

Peroxisomes and mitochondria are ubiquitous, highly dynamic organelles with an oxidative type of metabolism in eukaryotic cells. Over the years, substantial evidence has been provided that peroxisomes and mitochondria exhibit a close functional interplay which impacts on human health and development. The so-called "peroxisome-mitochondria connection" includes metabolic cooperation in the degradation of fatty acids, a redox-sensitive relationship, an overlap in key components of the membrane fission machineries and cooperation in anti-viral signalling and defence. Furthermore, combined peroxisome-mitochondria disorders with defects in organelle division have been revealed. In this review, we present the latest progress in the emerging field of peroxisomal and mitochondrial interplay in mammals with a particular emphasis on cooperative fatty acid β-oxidation, redox interplay, organelle dynamics, cooperation in anti-viral signalling and the resulting implications for disease.

The peroxin PEX14 of Neurospora crassa is essential for the biogenesis of both glyoxysomes and Woronin bodies

In the filamentous fungus Neurospora crassa, glyoxysomes and Woronin bodies coexist in the same cell. Because several glyoxysomal matrix proteins and also HEX1, the dominant protein of Woronin bodies, possess typical peroxisomal targeting signals, the question arises as to how protein targeting to these distinct yet related types of microbodies is achieved. Here we analyzed the function of the Neurospora ortholog of PEX14, an essential component of the peroxisomal import machinery. PEX14 interacted with both targeting signal receptors and was localized to glyoxysomes but was virtually absent from Woronin bodies. Nonetheless, a pex14Delta mutant not only failed to grow on fatty acids because of a defect in glyoxysomal beta-oxidation but also suffered from cytoplasmic bleeding, indicative of a defect in Woronin body-dependent septal pore plugging. Inspection of pex14Delta mutant hyphae by fluorescence and electron microscopy indeed revealed the absence of Woronin bodies. When these cells were subjected to subcellular fractionation, HEX1 was completely mislocalized to the cytosol. Expression of GFP-HEX1 in wild-type mycelia caused the staining of Woronin bodies and also of glyoxysomes in a targeting signal-dependent manner. Our data support the view that Woronin bodies emerge from glyoxysomes through import of HEX1 and subsequent fission.

The origin of eukaryotes: a reappraisal

Ever since the elucidation of the main structural and functional features of eukaryotic cells and subsequent discovery of the endosymbiotic origin of mitochondria and plastids, two opposing hypotheses have been proposed to account for the origin of eukaryotic cells. One hypothesis postulates that the main features of these cells, including their ability to capture food by endocytosis and to digest it intracellularly, were developed first, and later had a key role in the adoption of endosymbionts; the other proposes that the transformation was triggered by an interaction between two typical prokaryotic cells, one of which became the host and the other the endosymbiont. Re-examination of this question in the light of cell-biological and phylogenetic data leads to the conclusion that the first model is more likely to be the correct one.

A eukaryote without catalase-containing microbodies: Neurospora crassa exhibits a unique cellular distribution of its four catalases

Microbodies usually house catalase to decompose hydrogen peroxide generated within the organelle by the action of various oxidases. Here we have analyzed whether peroxisomes (i.e., catalase-containing microbodies) exist in Neurospora crassa. Three distinct catalase isoforms were identified by native catalase activity gels under various peroxisome-inducing conditions. Subcellular fractionation by density gradient centrifugation revealed that most of the spectrophotometrically measured activity was present in the light upper fractions, with an additional small peak coinciding with the peak fractions of HEX-1, the marker protein for Woronin bodies, a compartment related to the microbody family. However, neither in-gel assays nor monospecific antibodies generated against the three purified catalases detected the enzymes in any dense organellar fraction. Furthermore, staining of an N. crassa wild-type strain with 3,3'-diaminobenzidine and H(2)O(2) did not lead to catalase-dependent reaction products within microbodies. Nonetheless, N. crassa does possess a gene (cat-4) whose product is most similar to the peroxisomal type of monofunctional catalases. This novel protein indeed exhibited catalase activity, but was not localized to microbodies either. We conclude that N. crassa lacks catalase-containing peroxisomes, a characteristic that is probably restricted to a few filamentous fungi that produce little hydrogen peroxide within microbodies.

Peroxisome biogenesis: end of the debate

The long-standing and thorny issue of the origin of peroxisomes has at last been solved. New evidence demonstrates conclusively that the peroxisomal membrane originates from the endoplasmic reticulum. This process requires the two peroxins Pex3p and Pex19p leading to intermediate structures that then mature into functionally competent organelles.

Addition of oleic acid increases expression of recombinant human serum albumin by the AOX2 promoter in Pichia pastoris

The addition of several kinds of fatty acid to the culture medium of a recombinant human serum albumin (rHSA)-producing yeast, Pichia pastoris, resulted in increased expression levels of the product. Among the fatty acids tested, a small amount of oleic acid (0.01% (w/v)) doubled the rHSA production level in a shake-flask culture when measured by the reversed passive hemagglutination assay method. To elucidate this phenomenon, studies were conducted using deletion mutants from the AOX2 promoter region. Deletion mutants, designed for a detailed evaluation of the methanol regulation elements (AOX2-UAS, AOX2-URS1, and AOX2-URS2) did not respond to the addition of oleic acid. However, a deletion mutant that was not lacking an upstream region from the AOX2 promoter showed a response to oleic acid. The results implied the presence of an oleic acid-responsive element between nucleotides (nt) -1529 and -803, and it may lie between nt -1411 and -1403 in the AOX2 promoter of P. pastoris. The response to oleic acid was shown to function even when the level of rHSA expression was increased by a mutation in the AOX2 promoter. Therefore addition of oleic acid to the medium is likely to play an important role, in cooperation with gene manipulation, in achieving high expression levels of rHSA for the purpose of commercial production.

Evolution of energy metabolism and its compartmentation in Kinetoplastida

Kinetoplastida are protozoan organisms that probably diverged early in evolution from other eukaryotes. They are characterized by a number of unique features with respect to their energy and carbohydrate metabolism. These organisms possess peculiar peroxisomes, called glycosomes, which play a central role in this metabolism; the organelles harbour enzymes of several catabolic and anabolic routes, including major parts of the glycolytic and pentosephosphate pathways. The kinetoplastid mitochondrion is also unusual with regard to both its structural and functional properties.In this review, we describe the unique compartmentation of metabolism in Kinetoplastida and the metabolic properties resulting from this compartmentation. We discuss the evidence for our recently proposed hypothesis that a common ancestor of Kinetoplastida and Euglenida acquired a photosynthetic alga as an endosymbiont, contrary to the earlier notion that this event occurred at a later stage of evolution, in the Euglenida lineage alone. The endosymbiont was subsequently lost from the kinetoplastid lineage but, during that process, some of its pathways of energy and carbohydrate metabolism were sequestered in the kinetoplastid peroxisomes, which consequently became glycosomes. The evolution of the kinetoplastid glycosomes and the possible selective advantages of these organelles for Kinetoplastida are discussed. We propose that the possession of glycosomes provided metabolic flexibility that has been important for the organisms to adapt easily to changing environmental conditions. It is likely that metabolic flexibility has been an important selective advantage for many kinetoplastid species during their evolution into the highly successful parasites today found in many divergent taxonomic groups.Also addressed is the evolution of the kinetoplastid mitochondrion, from a supposedly pluripotent organelle, attributed to a single endosymbiotic event that resulted in all mitochondria and hydrogenosomes of extant eukaryotes. Furthermore, indications are presented that Kinetoplastida may have acquired other enzymes of energy and carbohydrate metabolism by various lateral gene transfer events different from those that involved the algal- and alpha-proteobacterial-like endosymbionts responsible for the respective formation of the glycosomes and mitochondria.

Cytoplasmic and peroxisomal catalases of the guinea pig liver: evidence for two distinct proteins

Catalase, a peroxisomal marker enzyme in the liver of most mammals, is found by immuno-electron microscopy in guinea pig (GP) hepatocytes not only in peroxisomes, but also in the cytoplasm (Beier et al. (1988) Eur. J. Cell Biol. 46, 129-135). We have been able to distinguish in GP liver homogenates between the cytosolic catalase and that part of the enzyme activity which is due to leakage of the enzyme from peroxisomes by adding 4% polyethylene glycol to the homogenization medium. This approach revealed that approximately 40% of the total catalase activity and almost all of alpha-hydroxy-acid oxidases are peroxisomal, while 60% of catalase is of genuine cytosolic origin. The cytosolic and peroxisomal catalases of guinea pig were purified to homogeneity and were analyzed by SDS-PAGE and isoelectric focussing. The cytosolic catalase exhibited a slightly higher Mr (approximately 1000) and a less acidic pI than the peroxisomal enzyme. Limited proteolysis and amino-acid analysis revealed also slight differences between the two molecular forms of catalase. Total RNA was isolated from guinea pig liver and translated in vitro by using a rabbit reticulocyte lysate system. Immunoprecipitation with an antibody against guinea pig catalase followed by high-resolution polyacrylamide gel electrophoresis revealed two polypeptide bands differing slightly in Mr. These observations suggest strongly, that cytoplasmic and peroxisomal catalases in guinea pig liver are two closely related but distinct proteins.

Cloning and characterization of the NAD-linked glycerol-3-phosphate dehydrogenases of Trypanosoma brucei brucei and Leishmania mexicana mexicana and expression of the trypanosome enzyme in Escherichia coli

A polyclonal antiserum raised against the purified glycosomal glycerol-3-phosphate dehydrogenase of Trypanosoma brucei brucei has been used to identify the corresponding cDNA clone in a T.b. brucei expression library. This cDNA was subsequently used to obtain genomic clones containing glycerol-3-phosphate dehydrogenase genes. Two tandemly arranged genes were detected in these clones. Characterization of one of the genes showed that it codes for a polypeptide of 353 amino acids, with a molecular mass of 37,651 Da and a calculated net charge of +8. Using the T.b. brucei gene as a probe, a corresponding glycerol-3-phosphate dehydrogenase gene was also identified in a genomic library of Leishmania mexicana mexicana. The L.m. mexicana gene codes for a polypeptide of 365 amino acids, with a molecular mass of 39,140 Da and a calculated net charge of +8. The amino-acid sequences of both polypeptides are 63% identical and carry a type-1 peroxisomal targeting signal (PTS1) SKM and -SKL at their respective C-termini. Moreover, the L.m. mexicana polypeptide also carries a short N-terminal extension reminiscent of a mitochondrial transit sequence. Subcellular localisation analysis showed that in L.m. mexicana the glycerol-3-phosphate dehydrogenase activity co-fractionated both with mitochondria and with glycosomes. This is not the case in T. brucei, where the enzyme is predominantly glycosomal. The two trypanosomatid sequences resemble their prokaryotic homologues (32-36%) more than their eukaryotic counterparts (25-31%) and carry typical prokaryotic signatures. The possible reason for this prokaryotic nature of a trypanosomatid glycerol-3-phosphate dehydrogenase is discussed.

Purification and characterization of phospho enol pyruvate carboxykinase from Trypanosoma brucei

ATP-dependent phospho enol pyruvate carboxykinase (EC 4.1.1.49; PEPCK, ATP) was purified from glycosomes of cultured procyclic Trypanosoma brucei to electrophoretic homogeneity. The purified enzyme exhibited a mean specific activity of 83 units mg-1, as measured in the carboxylation direction at 30 degrees C. A similar activity was obtained for the decarboxylation reaction. The enzyme was shown to be a homodimer in solution with a subunit molecular mass of 59 kDa. Amino acid sequence analysis suggested that the PEPCK (ATP) is identical to the trypanosomal protein p60, the sequence of which was previously predicted from the corresponding nucleotide sequence by other investigators. The basic nature of the enzyme was indicated by a high isoelectric point (pH 8.9). The enzyme was found to be strictly dependent on adenosine nucleotides for activity, as well as on the presence of Mn2+. Mg2+ was found to be ineffective as activator of the trypanosomal enzyme, but a combination of subsaturating (< or = 300 microM) concentrations of Mn2+ and high concentrations of Mg2+ caused a synergistic effect on the carboxylation activity, indicating a dual cation requirement. Mn2+ is necessary to activate the enzyme and Mn2+ or Mg2+ most likely forms the cation-nucleotide complex as the active form of the substrate. Relatively high (5 mM) levels of ATP were required to produce a significant inhibition of the carboxylation reaction. Quinolinic acid, a structural analogue of oxaloacetate, completely inhibited the decarboxylation reaction at a 1 mM concentration. The apparent Michaelis constants of the enzyme were 490 microM for PEP, 37 microM for oxaloacetate, 40 microM for ADP, 10.3 microM for ATP, 970 microM for Mn2+ and 26 mM for HCO3-. Endogenous substrate concentrations were found to be 327 nmol PEP, 1486 nmol ADP, 4200 nmol ATP and 11.5 nmol Mn2+ (ml cell volume)-1. Our kinetic data suggest that under physiological conditions PEPCK (ATP) in T. brucei is bidirectional and that its activity is regulated primarily by mass action. The physiological relevance of the enzyme in procyclic T. brucei is discussed.

Molecular organization of hepatocyte peroxisomes

The matrix of peroxisomes has been considered to be homogeneous. However, a fine network of tubules is visible in electron micrographs at very high magnification. This substructure becomes more positive in a high-contrast photocopy and with an imaging-plate method. Clofibrate, bezafibrate, and aspirin increase peroxisomes. In proliferated peroxisomes, the density of matrix is low and the fine network is more visible. The effect of proliferators is more significant in males than in females. This sex difference may involve the action of estrogen, growth hormone, cytochrome P-450 and thyroxine. Mg-ATPase is localized on the limiting membrane of peroxisomes. Even on the membrane of irregular projections of proliferated peroxisomes, Mg-ATPase is evident cytochemically. Carnitine acetyltransferase is detectable in the matrix of proliferated peroxisomes. Withdrawal of proliferators results in a rapid decrease of peroxisomes. This may indicate the existence of peroxisome suppressors. Alternatively, dynamic transformation of vesicular to tubular types in peroxisome reticulum may occur. Such transformation has been described in lysosomes and mitochondria.

Ultrastructure of the main excretory duct epithelium of the female mouse submandibular gland with special reference to sexual dimorphism

The fine structure of the main excretory duct epithelium (MEDE) of female mouse submandibular gland was investigated by scanning and transmission electron microscopy and the results compared with the previously established structure of male mouse MEDE. A comparative analysis of the subepithelial capillaries of both sexes was also performed. In this pseudostratified epithelium, principal cell-types were observed: types-I, -II, -III and basal cells. This differed significantly from male MEDE, where type-II and -III are absent and type-I cells are the most numerous. The latter cell-type had abundant mitochondria, a few lipid-containing granules, lysosomes in the infra-nuclear cytoplasm and well-developed basal infoldings. These cells were also characterized by abundant glycogen granules throughout the cytoplasm, many profiles of strands of smooth endoplasmic reticulum in the apical region, and lysosomes in the infranuclear region. Type-II cells were the second most numerous. Their most characteristic features were the presence of tubular vesicles which appeared to be invaginated from the plasma membrane, RER, SER, free ribosomes, a few peroxisomes with nucleoids, and primary lysosomes in extremely light cytoplasm. They had many mitochondria throughout the cytoplasm, except in the apical region, a few lipid-containing granules and no basal infoldings. Type-III cells were very few and were characterized by well developed basal infoldings, abundant free ribosomes, RER, SER, vesicles containing moderately dense material, and many lipid-containing granules. They also had many mitochondria throughout the cytoplasm, except apically. Basal cells had a large nucleus and the cytoplasm had few organelles.(ABSTRACT TRUNCATED AT 250 WORDS)

Peroxisome assembly factor 1: nonsense mutation in a peroxisome-deficient Chinese hamster ovary cell mutant and deletion analysis

A cDNA encoding 35-kDa peroxisome assembly factor 1 (PAF-1), a peroxisomal integral membrane protein, was cloned from Chinese hamster ovary (CHO) cells and sequenced. The CHO PAF-1 comprised 304 amino acids, one residue shorter than rat or human PAF-1, and showed high homology to rat and human PAF-1: 90 and 86% at the nucleotide sequence level and 92 and 90% in amino acid sequence, respectively. PAF-1 from these three species contains a conserved cysteine-rich sequence at the C-terminal region which is exactly the same as that of a novel cysteine-rich RING finger motif family. PAF-1 cDNA from a peroxisome-deficient CHO cell mutant, Z65 (T. Tsukamoto, S. Yokota, and Y. Fujiki, J. Cell Biol. 110:651-660, 1990), contained a nonsense mutation at the codon for Trp-114, resulting in premature termination. Truncation in PAF-1 of either 19 amino acids from the N terminus or 92 residues from the C terminus maintained the peroxisome assembly-restoring activity when tested in both the Z65 mutant and the fibroblasts from a Zellweger patient. In contrast, deletion of 27 or 102 residues from the N or C terminus eliminated the activity. PAF-1 is encoded by free polysomal RNA, consistent with a general rule for biogenesis of peroxisomal proteins, including membrane polypeptides, implying the posttranslational transport and integration of PAF-1 into peroxisomal membrane.

Peroxisome assembly factor 1: nonsense mutation in a peroxisome-deficient Chinese hamster ovary cell mutant and deletion analysis

A cDNA encoding 35-kDa peroxisome assembly factor 1 (PAF-1), a peroxisomal integral membrane protein, was cloned from Chinese hamster ovary (CHO) cells and sequenced. The CHO PAF-1 comprised 304 amino acids, one residue shorter than rat or human PAF-1, and showed high homology to rat and human PAF-1: 90 and 86% at the nucleotide sequence level and 92 and 90% in amino acid sequence, respectively. PAF-1 from these three species contains a conserved cysteine-rich sequence at the C-terminal region which is exactly the same as that of a novel cysteine-rich RING finger motif family. PAF-1 cDNA from a peroxisome-deficient CHO cell mutant, Z65 (T. Tsukamoto, S. Yokota, and Y. Fujiki, J. Cell Biol. 110:651-660, 1990), contained a nonsense mutation at the codon for Trp-114, resulting in premature termination. Truncation in PAF-1 of either 19 amino acids from the N terminus or 92 residues from the C terminus maintained the peroxisome assembly-restoring activity when tested in both the Z65 mutant and the fibroblasts from a Zellweger patient. In contrast, deletion of 27 or 102 residues from the N or C terminus eliminated the activity. PAF-1 is encoded by free polysomal RNA, consistent with a general rule for biogenesis of peroxisomal proteins, including membrane polypeptides, implying the posttranslational transport and integration of PAF-1 into peroxisomal membrane.

Urate oxidase is imported into peroxisomes recognizing the C-terminal SKL motif of proteins

Rat liver urate oxidase synthesized from cDNA through coupled transcription and translation was incubated at 26 degrees C for 60 min with purified peroxisomes from rat liver. Urate oxidase was efficiently imported into the peroxisomes, as determined by resistance to externally added proteinase K. The amount of imported urate oxidase increased with time and the import was temperature dependent. A synthetic peptide composed of the C-terminal 10 amino acid residues of acyl-CoA oxidase (the C-terminal tripeptide is Ser-Lys-Leu) inhibited the import of urate oxidase, whereas other peptides, in which the C-terminal Ser-Lys-Leu (SKL) sequence was deleted or mutated, were not effective. Two mutant urate oxidase proteins in which the C-terminal Ser-Arg-Leu (SRL) sequence was deleted or mutated to Ser-Glu-Leu (SEL) were not imported into peroxisomes. With substitution of a lysine residue for arginine in the SRL tripeptide at the C-terminus the import activity was retained. These results show that urate oxidase is important into peroxisomes via a common pathway with acyl-CoA oxidase, and that the C-terminal SRL sequence functions as a peroxisomal-targeting signal.

Protein organization in the matrix of leaf peroxisomes. A multi-enzyme complex involved in photorespiratory metabolism

This report is an investigation on how the compartmentation of peroxisomal metabolism, involved in the photorespiratory cycle, is accomplished. With isolated peroxisomes from spinach leaves the conversion of serine to glycerate, as coupled to the conversion of glycolate to glycine, was measured. Not only with intact but also with osmotically shocked peroxisomes, which had retained the aggregated state of the peroxisomal matrix but lost the integrity of the boundary membrane, the rates of glycerate synthesis were as high as required for the photorespiratory cycle in vivo. With both intact and shocked peroxisomes the intermediates glyoxylate and hydroxypyruvate did not equilibrate with the medium. It appears from these results that the apparent compartmentation of peroxisomal metabolism is not due to the function of the boundary membrane but to the organization of peroxisomal enzymes in multi-enzyme complexes. When glycolate was added to peroxisomes without transamination partners, glyoxylate was released from the peroxisomes while the peroxisomal matrix partially disintegrated. With solubilized peroxisomes a partial reconstitution of functional enzyme complexes was achieved by the addition of poly(ethylene glycol). The function of the apparently very stable peroxisomal multi-enzyme complexes in protecting the cells from the toxic intermediates H2O2 and glyoxylate is discussed.

Biogenesis of peroxisomes: isolation and characterization of two distinct peroxisomal populations from normal and regenerating rat liver

According to Poole et al. (1970, J. Cell Biol. 45:408-415), newly synthesized peroxisomal proteins are incorporated uniformly into peroxisomes (PO) of different size classes, suggesting that rat hepatic PO form a homogeneous population. There is however increasing cytochemical and biochemical evidence that PO in rat liver are heterogenous, undergoing significant modulations in shape and size in process of PO morphogenesis (Yamamoto and Fahimi, 1987. J. Cell Biol. 105:713-722). In the present study, the kinetics of incorporation of newly synthesized proteins into distinct PO-subpopulations have been studied using short-term in vivo labeling (5-90 min). Two distinct "heavy" and "light" crude PO fractions were prepared by differential pelleting from normal and regenerating liver, and highly purified PO were subsequently isolated by density-dependent metrizamide gradient centrifugation according to Völkl and Fahimi (1985. Eur. J. Biochem. 149:257-265). The peroxisomal fractions banded at 1.20 and 1.24 g x cm-3. They differed in their mean diameters and form-factors and particularly in respect to the activity of beta-oxidation enzymes which was higher in the "light" PO. Whereas the "light" PO exhibited a single immunoreactive band with the antibody to the 70-kD peroxisomal membrane protein the "heavy" PO contained an additional (68 kD) band. In pulse-labeling experiments "light" PO showed clearly a higher initial rate of incorporation than the "heavy" PO. The relative specific activity in the "heavy" PO fraction, however increased progressively reaching that of "light" PO by 90 min. These observations provide evidence for the existence of different PO populations in rat liver which differ in their morphological and biochemical properties as well as in their rates of incorporation of new proteins.

A knowledge base for predicting protein localization sites in eukaryotic cells

To automate examination of massive amounts of sequence data for biological function, it is important to computerize interpretation based on empirical knowledge of sequence-function relationships. For this purpose, we have been constructing a knowledge base by organizing various experimental and computational observations as a collection of if-then rules. Here we report an expert system, which utilizes this knowledge base, for predicting localization sites of proteins only from the information on the amino acid sequence and the source origin. We collected data for 401 eukaryotic proteins with known localization sites (subcellular and extracellular) and divided them into training data and testing data. Fourteen localization sites were distinguished for animal cells and 17 for plant cells. When sorting signals were not well characterized experimentally, various sequence features were computationally derived from the training data. It was found that 66% of the training data and 59% of the testing data were correctly predicted by our expert system. This artificial intelligence approach is powerful and flexible enough to be used in genome analyses.

Antibodies directed against a yeast carboxyl-terminal peroxisomal targeting signal specifically recognize peroxisomal proteins from various yeasts

The carboxyl-terminal tripeptide Ala-Lys-Ile is essential for targeting Candida tropicalis trifunctional enzyme (hydratase-dehydrogenase-epimerase) to peroxisomes of both Candida albicans and Saccharomyces cerevisiae (Aitchison,J.D., Murray, W.W. and Rachubinski, R. A. (1991).J. Biol. Chem. 266, 23197-23203). We investigated the possibility that this tripeptide may act as a general peroxisomal targeting signal (PTS) for other proteins in the yeasts C. tropicalis, C. albicans, Yarrowia lipolytica and S. cerevisiae, and in rat liver. Anti-AKI antibodies raised against the carboxyl-terminal 12 amino acids of trifunctional enzyme were used to search for this PTS in proteins of these yeasts and of rat liver. The anti-AKI antibodies reacted exclusively with multiple peroxisomal proteins from the yeasts C. tropicalis, C. albicans and Y. lipolytica. There was a weak reaction of the antibodies with one peroxisomal protein from S. cerevisiae and no reaction with peroxisomal proteins from rat liver. Antibodies directed against a synthetic peptide containing a carboxyl-terminal Ser-Lys-Leu PTS (Gould, S. J., Krisans, S., Keller, G.-A. and Subramani, S. (1990). J. Cell Biol. 110,27-34) reacted with multiple peroxisomal proteins of rat liver and with peroxisomal proteins of yeast distinct from those identified with anti-AKI antibodies. These results provide evidence that several peroxisomal proteins of different yeasts contain a PTS antigenically similar to that of C. tropicalis trifunctional enzyme and that this signal is absent from peroxisomal proteins from at least one mammalian system, rat liver.

Is the cytosolic catalase induced by peroxisome proliferators in mouse liver on its way to the peroxisomes?

Dietary treatment of male C57B1/6 mice with clofibrate, nafenopin or WY-14.643 resulted in a modest (at most 2-fold) increase in the total catalase activity in the whole homogenate and mitochondrial fraction prepared from the livers of these animals. On the other hand, the catalase activity recovered in the cytosolic fraction was increased 12- to 18-fold, i.e. 30-35% of the total catalase activity in the hepatic homogenate was present in the high-speed supernatant fraction after treatment with these peroxisome proliferators. A study of the time course of the changes in peroxisomal and cytosolic catalase activities demonstrated that the peroxisomal activity both increased upon initiation of exposure and decreased after termination of treatment several days after the increase and decrease, respectively, in the corresponding cytosolic activity. This finding suggests that the cytosolic catalase may be on its way to incorporation into peroxisomes.

In vivo import of firefly luciferase into the glycosomes of Trypanosoma brucei and mutational analysis of the C-terminal targeting signal

The compartmentalization of glycolytic enzymes into specialized organelles, the glycosomes, allows the bloodstream form of Trypanosoma brucei to rely solely on glycolysis for its energy production. The biogenesis of glycosomes in these parasites has been studied intensively as a potential target for chemotherapy. We have adapted the recently developed methods for stable transformation of T. brucei to the in vivo analysis of glycosomal protein import. Firefly luciferase, a peroxisomal protein in the lantern of the insect, was expressed in stable transformants of the procyclic form of T. brucei, where it was found to accumulate inside the glycosomes. Mutational analysis of the peroxisomal targeting signal serine-lysine-leucine (SKL) located at the C-terminus of luciferase showed that replacement of the serine residue (Serine548) with a small neutral amino acid (A, C, G, H, N, P, T) still resulted in an import efficiency of 50-100% of the wild-type luciferase. Lysine549 could be substituted with an amino acid capable of hydrogen bonding (H, M, N, Q, R, S), whereas the C-terminal leucine550 could be replaced with a subset of hydrophobic amino acids (I, M, Y). Thus, a peroxisome-like C-terminal SKL-dependent targeting mechanism may function in T. brucei to import luciferase into the glycosomes. However, a few significant differences exist between the glycosomal targeting signals identified here and the tripeptide sequences that direct proteins to mammalian or yeast peroxisomes.

Suramin prevents import of acyl-CoA oxidase into rat liver peroxisomes

The inhibitory effect of suramin on the import of [35S]acyl-CoA oxidase into purified rat liver peroxisomes was investigated in vitro. The import of acyl-CoA oxidase was inhibited completely by 10 microM suramin, whilst the latency of catalase remained unchanged. The important value decreased 60% by pretreatment of peroxisomes with 10 microM suramin, but it did not decrease by pretreatment of translation products. Polysulfonate compounds which have two clusters of negative charges, such as Cibacron blue F3GA and Trypan blue, as well as suramin, inhibited the import, whilst mono- and disulfonate compounds did not.

Peroxisome assembly mutations in humans: structural heterogeneity in Zellweger syndrome

Empty membrane ghosts of peroxisomes were found in fibroblasts from a patient with Zellweger's syndrome, a genetic disease of humans (Santos et al: Science 239:1536-1538, 1988). Import of soluble matrix proteins into the organelle was defective. We have now studied fibroblasts from seven patients representing five complementation groups of the syndrome (defined by complementation for peroxisome enzyme function). We find that empty peroxisome ghosts are present in all seven cell samples. Three patients, representing three complementation groups, give the same membrane pattern by immunofluorescence: few large ghosts. Three other patients, representing two complementation groups, give a second pattern: many large ghosts. The seventh patient's pattern is distinct. Thus, all seven of these patients exhibit Peroxisome IMport (PIM) mutations. Since membrane assembly occurs in these cells, the results indicate that biogenesis of organelle content and membrane proteins proceed by different mechanisms. Growth and division of the empty peroxisomal membrane must occur, but are modified by the mutations (ghost size and abundance vary). Cell fusion and immunofluorescence analyses of peroxisome size and catalase packaging formally demonstrate genetic complementation groups for peroxisome assembly in Zellweger syndrome.

Mutations in the 70K peroxisomal membrane protein gene in Zellweger syndrome

The peroxisomal membrane protein, with a relative molecular mass of 70,000 (M(r) 70K) (PMP70), is an important component of peroxisomal membranes and an ATP-binding cassette protein. To investigate its possible involvement in Zellweger syndrome (ZS), an inborn error of peroxisome assembly, we cloned and sequenced cDNAs for human PMP70 and mapped the gene to chromosome 1. Amongst 32 probands with ZS or related disorders, we found two mutant PMP70 alleles in single ZS probands from the same complementation group. One allele has a donor splice site mutation and the second a missense mutation. Our results suggest that PMP70 plays an important role in peroxisome biogenesis and that mutations in PMP70 may be responsible for a subset of ZS patients.

PAS3, a Saccharomyces cerevisiae gene encoding a peroxisomal integral membrane protein essential for peroxisome biogenesis

Saccharomyces cerevisiae pas3-mutants are described which conform the pas-phenotype recently reported for the peroxisomal assembly mutants pas1-1 and pas2 (Erdmann, R., M. Veenhuis, D. Mertens, and W.-H Kunau, 1989, Proc. Natl. Acad. Sci. USA. 86:5419-5423). The isolation of pas3-mutants enabled us to clone the PAS3 gene by functional complementation. DNA sequence analysis revealed a 50.6-kD protein with at least one domain of sufficient length and hydrophobicity to span a lipid bilayer. To verify these predictions antibodies were raised against a truncated portion of the PAS3 coding region overexpressed in E. coli. Pas3p was identified as a 48 kD peroxisomal integral membrane protein. It is shown that a lack of this protein causes the peroxisome-deficient phenotype and the cytosolic mislocalization of peroxisomal matrix enzymes. Based on protease digestion experiments Pas3p is discussed to be anchored in the peroxisomal membrane by its amino-terminus while the bulk of the molecule is exposed to the cytosol. These findings are consistent with the possibility that Pas3p is one component of the peroxisomal import machinery.

Evolutionary conservation of a microbody targeting signal that targets proteins to peroxisomes, glyoxysomes, and glycosomes

Peroxisomes, glyoxysomes, glycosomes, and hydrogenosomes have each been classified as microbodies, i.e., subcellular organelles with an electron-dense matrix that is bound by a single membrane. We investigated whether these organelles might share a common evolutionary origin by asking if targeting signals used for translocation of proteins into these microbodies are related. A peroxisomal targeting signal (PTS) consisting of the COOH-terminal tripeptide serine-lysine-leucine-COOH has been identified in a number of peroxisomal proteins (Gould, S.J., G.-A. Keller, N. Hosken, J. Wilkinson, and S. Subramani. 1989. J. Cell Biol. 108:1657-1664). Antibodies raised to a peptide ending in this sequence (SKL-COOH) recognize a number of peroxisomal proteins. Immunocryoelectron microscopy experiments using this anti-SKL antibody revealed the presence of proteins containing the PTS within glyoxysomes of cells from Pichia pastoris, germinating castor bean seeds, and Neurospora crassa, as well as within the glycosomes of Trypanosoma brucei. Western blot analysis of purified organelle fractions revealed the presence of many proteins containing this PTS in both glyoxysomes and glycosomes. These results indicate that at least one of the signals, and therefore the mechanism, for protein translocation into peroxisomes, glyoxysomes, and glycosomes has been conserved, lending support to a common evolutionary origin for these microbodies. Hydrogenosomes, the fourth type of microbody, did not contain proteins that cross-reacted with the anti-PTS antibody, suggesting that this organelle is unrelated to microbodies.

Glucose-responsive and oleic acid-responsive elements in the gene encoding the peroxisomal trifunctional enzyme of Candida tropicalis

We have investigated the regulation of expression of the gene (HDE), encoding the peroxisomal trifunctional enzyme hydratase-dehydrogenase-epimerase (HDE), of the diploid yeast Candida tropicalis. Heterologous expression in Saccharomyces cerevisiae of constructs containing deletions in the upstream region of the HDE gene has allowed for determination of regions responsible for the control of expression of the HDE gene. Expression was monitored by immunoblot analysis of yeast lysates with anti-HDE serum. Regions have been identified that are responsible for both repression by glucose and induction by oleic acid. A glucose-responsive region lies between nucleotides (nt) -526 and -393. An oleic acid-responsive region lies between nt -393 and -341. An additional region controlling derepression by nonfermentable carbon sources is located downstream from nt -341. Comparison of the nt sequences of these regions to upstream regions of other oleic acid-responsive genes of C. tropicalis has identified possible consensus nt sequences for glucose- and oleic acid-responsive upstream elements in these genes. The regulation of the HDE gene in S. cerevisiae closely resembles that found in C. tropicalis, suggesting that similar mechanisms of transcriptional control operate in both yeasts.

Regulation of transcription of the gene coding for peroxisomal 3-oxoacyl-CoA thiolase of Saccharomyces cerevisiae

Transferring Saccharomyces cerevisiae cells from glucose- to oleate-containing growth media results in a significant increase in the number and volume of peroxisomes. To investigate this proliferation process we studied the transcriptional regulation of the gene coding for peroxisomal 3-oxoacyl-CoA thiolase (EC 2.3.1.16) in response to the switch in carbon source. Expression was proved to be repressed during growth on glucose, derepressed during growth on glycerol, and induced during growth on oleate as the sole carbon source. By deletion and mutational analysis of sequences upstream of this gene, we have identified a region which is involved in the regulation of transcription. It is contained within a 52-base-pair sequence, UAST52 (upstream activation sequence thiolase 52), located between 203 and 151 nucleotides upstream of the translational initiation codon. This sequence proved to be required for repression, derepression and induction of transcription, and was able to activate transcription from the truncated version of the heterologous iso-1-cytochrome-c (CYC1) promoter in a similar way as in the wild-type promoter context. Sequence comparison revealed that the UAST52 contained a sequence motif ('beta-oxidation box') that is very similar to sequences located in the 5'-upstream regions of the genes coding for two other beta-oxidation enzymes of S. cerevisiae: the peroxisomal acyl-CoA oxidase and the peroxisomal trifunctional beta-oxidation enzyme of S. cerevisiae. Mutational analysis of the 'beta-oxidation box' indicates that this sequence motif acts as a UAS in vivo. Sequence comparison also revealed that just upstream of the 'beta-oxidation box', between positions -213 and -201, a potential binding site occurred for the yeast multifunctional autonomously replicating sequence binding factor ABF1. Gel-retardation-competition experiments indicate that ABF1 binds specifically to this sequence.

Studies on the intracellular distributions of soluble epoxide hydrolase and of catalase by digitonin-permeabilization of hepatocytes isolated from control and clofibrate-treated mice

Digitonin permeabilization of hepatocytes from control and clofibrate-treated (0.5% by mass, 10 days) male C57bl/6 mice was used to study the intracellular distributions of soluble ('cytosolic') epoxide hydrolase and of catalase. The following conclusions were drawn. (1) About 60% of the total soluble epoxide hydrolase activity in control mouse hepatocytes is situated in the cytosol. (2) The rest is not mitochondrial, but probably peroxisomal. (3) Of the total catalase activity in control mouse hepatocytes, 5-10% is found in the cytosol. (4) Treatment of mice with clofibrate increases the total hepatocyte activity of soluble epoxide hydrolase 4-fold, but does not influence the relative distribution of this enzyme between cytosol and peroxisomes. (5) The total catalase activity is increased 3.5-fold by clofibrate treatment and 15-35% of this activity is shifted from the peroxisomes to the cytosol.

Recognition of a peroxisomal tripeptide entry signal by the glycosomes of Trypanosoma brucei

Trypanosomes compartmentalise the first 9 enzymes of glycolysis and glycerol metabolism in peroxisome-like microbodies known as glycosomes. The identity of the sequences that direct proteins into the glycosome has until now been uncertain. We show here that the peroxisomal tripeptide entry signal is sufficient to cause association of a bacterial enzyme with the glycosomes of Trypanosoma brucei. However, it works less efficiently than the C-terminal 21 amino acids of trypanosome glycosomal phosphoglycerate kinase.

Protein transport and compartmentation in yeast

Many newly synthesized proteins must be translocated across one or more membranes to reach their destination in the individual organelles or membrane systems. Translocation, mostly requiring an energy source, a signal on the protein itself, loose conformation of the protein and the presence of cytosolic and/or membrane receptor-like proteins, is often accompanied by covalent modifications of transported proteins. In this review I discuss these aspects of protein transport via the classical secretory pathway and/or special translocation mechanisms in the unicellular eukaryotic organism Saccharomyces cerevisiae.

Acyl-CoA oxidase, peroxisomal thiolase and dihydroxyacetone phosphate acyltransferase: aberrant subcellular localization in Zellweger syndrome

We have studied the presence and subcellular localization of peroxisomal 3-oxoacylcoenzyme A thiolase, acylcoenzyme A oxidase and acyl-CoA: dihydroxyacetonephosphate acyltransferase (DHAPAT) in fibroblasts from control subjects and patients with an inherited deficiency of peroxisomes (Zellweger syndrome), using immunofluorescence spectroscopy and density gradient centrifugation techniques. The results show that Zellweger cells contain unprocessed thiolase and unprocessed acyl-CoA oxidase which are associated with structures containing a peroxisomal integral membrane protein of 69 kDa and having a density much lower than that of normal peroxisomes. The residual DHAPAT activity present in Zellweger cells is also contained in these structures. We conclude that these structures represent defectively assembled peroxisomes which may still be capable of importing some peroxisomal proteins.

Identification of mutations associated with peroxisome-to-mitochondrion mistargeting of alanine/glyoxylate aminotransferase in primary hyperoxaluria type 1

We have previously shown that in some patients with primary hyperoxaluria type 1 (PH1), disease is associated with mistargeting of the normally peroxisomal enzyme alanine/glyoxylate aminotransferase (AGT) to mitochondria (Danpure, C.J., P.J. Cooper, P.J. Wise, and P.R. Jennings. J. Cell Biol. 108:1345-1352). We have synthesized, amplified, cloned, and sequenced AGT cDNA from a PH1 patient with mitochondrial AGT (mAGT). This identified three point mutations that cause amino acid substitutions in the predicted AGT protein sequence. Using PCR and allele-specific oligonucleotide hybridization, a range of PH1 patients and controls were screened for these mutations. This revealed that all eight PH1 patients with mAGT carried at least one allele with the same three mutations. Two were homozygous for this allele and six were heterozygous. In at least three of the heterozygotes, it appeared that only the mutant allele was expressed. All three mutations were absent from PH1 patients lacking mAGT. One mutation encoding a Gly----Arg substitution at residue 170 was not found in any of the control individuals. However, the other two mutations, encoding Pro----Leu and Ile----Met substitutions at residues 11 and 340, respectively, cosegregated in the normal population at an allelic frequency of 5-10%. In an individual homozygous for this allele (substitutions at residues 11 and 340) only a small proportion of AGT appeared to be rerouted to mitochondria. It is suggested that the substitution at residue 11 generates an amphiphilic alpha-helix with characteristics similar to recognized mitochondrial targeting sequences, the full functional expression of which is dependent upon coexpression of the substitution at residue 170, which may induce defective peroxisomal import.

Structure/function relationships in methylotrophic yeasts

This symposium marks the 15th anniversary of the discovery of microbodies in methylotrophic yeasts. In the intervening years much has been learned about the structure, function and biogenesis of these organelles and these advances are described. As our endeavours continued, unexpected results have confused commonly held views. This was for instance the case when microbody-minus mutants of yeasts became available which showed that some microbody matrix enzymes may be functional when present in the cytosol while others are not. At the molecular level, our understanding of structure/function relationships is also expanding. Examples are structural elements which relate to protein topogenesis and function of enzymes in different cell compartments. Other, perhaps more unusual, adaptations have also been encountered; some involve protein-protein interactions or even modified cofactors which possibly have helped methylotrophic yeasts to establish and/or maintain themselves in natural ecosystems.

Cloning and nucleotide sequence of cDNA encoding human liver serine-pyruvate aminotransferase

Cloned cDNAs for human liver serine-pyruvate aminotransferase (Ser-PyrAT) were obtained by screening of a human liver cDNA library with a fragment of cDNA for rat mitochondrial Ser-PyrAT as a probe. Two clones were isolated from 50,000 transformants. Both clones contained approximately 1.5 kb cDNA inserts and were shown to almost completely overlap each other on restriction enzyme mapping and DNA sequencing. The nucleotide sequence of the mRNA coding for human liver Ser-PyrAT was determined from those of the cDNA clones. The mRNA comprises at least 1487 nucleotides, and encodes a polypeptide consisting of 392 amino acid residues with a molecular mass of 43,039 Da. The amino acid composition determined on acid hydrolysis of the purified enzyme showed good agreement with that deduced from the nucleotide sequence of the cDNA. In vitro translation of the mRNA derived from one of the isolated clones, pHspt12, as well as that of mRNA extracted from human liver, yielded a product of 43 kDa which reacted with rabbit anti-(rat mitochondrial Ser-PyrAT) serum. Comparison of the deduced amino acid sequences of human Ser-PyrAT and the mature form of rat mitochondrial Ser-PyrAT revealed 79.3% identity. Although human Ser-PyrAT appears to be synthesized as the mature size, the 5'-noncoding region of human Ser-PyrAT mRNA contains a nucleotide sequence which would encode, if translated, an amino acid sequence similar to that of the N-terminal extension peptide of the precursor for rat mitochondrial Ser-PyrAT.

Characterization of an in vitro assay for import of 3-phosphoglycerate kinase into the glycosomes of Trypanosoma brucei

Glycosomes are microbody organelles found in kinetoplastida, where they serve to compartmentalize the enzymes of the glycolytic pathway. In order to identify the mechanism by which these enzymes are targeted to the glycosome, we have modified the in vitro import assay developed by Dovey et al. (Proc. Natl. Acad. Sci. USA 85:2598-2602, 1988). This assay measures the uptake of in vitro-translated Trypanosoma brucei glycosomal 3-phosphoglycerate kinase (gPGK) by purified glycosomes. Up to 50% of the total 35S-gPGK in the glycosomal fraction was resistant to extraction by 3 M urea or treatment with proteinase K (500 micrograms/ml). The glycosome-associated 35S-gPGK could be chemically cross-linked to the endogenous glycosomal proteins to form a sodium dodecyl sulfate-resistant complex, suggesting that it is close to the intraglycosomal protein matrix. Deoxycholate solubilized the glycosome and thereby rendered the glycosome-associated 35S-gPGK fully susceptible to proteinase K. However, the glycosome-associated 35S-gPGK was not digested by proteinase K in the presence of Triton X-100, which cannot dissolve the glycosomal protein core. The 35S-gPGK synthesized in vitro was able to bind directly to protein cores, where it became resistant to urea extraction and proteinase K digestion. However, the 35S-gPGK-protein core complex exhibited a much higher density than the 35S-gPGK-glycosome complex and was readily separable in sucrose gradients. Thus, in our in vitro import assay, the 35S-gPGK appeared to associate with intact glycosomes, possibly reflecting import of protein into the organelle. Complete denaturation of the 35S-gPGK in 8 M urea prior to the assay enhanced the efficiency of its association with glycosomes. Native gPGK did not compete with the association of in vitro-translated gPGK unless it was denatured. The assay exhibited time and temperature dependence, but it did not require externally added ATP and was not inhibited by the nonhydrolyzable analogs adenosine-5'-(beta,gamma-imido)-triphosphate and gamma-S-ATP. However, the presence of 20 to 30 microM ATP inside the glycosome may fulfill the requirement for protein import.