Search

For several sesquiterpene lactones (STLs) found in Asteraceae plants, very interesting biomedical activities have been demonstrated. Chicory roots accumulate the guaianolide STLs 8-deoxylactucin, lactucin, and lactucopicrin predominantly in oxalated forms in the latex. In this work, a supercritical fluid extract fraction of chicory STLs containing 8-deoxylactucin and 11β,13-dihydro-8-deoxylactucin was shown to have anti-inflammatory activity in an inflamed intestinal mucosa model. To increase the accumulation of these two compounds in chicory taproots, the lactucin synthase that takes 8-deoxylactucin as the substrate for the regiospecific hydroxylation to generate lactucin needs to be inactivated. Three candidate cytochrome P450 enzymes of the CYP71 clan were identified in chicory. Their targeted inactivation using the CRISPR/Cas9 approach identified CYP71DD33 to have lactucin synthase activity. The analysis of the terpene profile of the taproots of plants with edits in CYP71DD33revealed a nearly complete elimination of the endogenous chicory STLs lactucin and lactucopicrin and their corresponding oxalates. Indeed, in the same lines, the interruption of biosynthesis resulted in a strong increase of 8-deoxylactucin and its derivatives. The enzyme activity of CYP71DD33 to convert 8-deoxylactucin to lactucin was additionally demonstrated in vitrousing yeast microsome assays. The identified chicory lactucin synthase gene is predominantly expressed in the chicory latex, indicating that the late steps in the STL biosynthesis take place in the latex. This study contributes to further elucidation of the STL pathway in chicory and shows that root chicory can be positioned as a crop from which different health products can be extracted.

Plants are gaining increasingly acceptance as a production platform for recombinant proteins. One reason for this is their ability to carry out posttranslational protein modifications in a similar if not identical way as mammalian cells. The capability of plants to carry out human-like complex glycosylation is well known. Moreover, the targeted manipulation of the plant N-glycosylation pathway allows the production of proteins carrying largely homogeneous, human-type oligosaccharides. These outstanding results have placed plants in a favourable position compared to other eukaryotic expression systems. This review provides a comprehensive summary of the N-glycosylation of plant-produced recombinant proteins, the possible impact of plant-specific N-glycans on the human immune system, and recent advances in engineering the plant N-glycosylation pathway towards the synthesis of (complex) human-type glycan structures, highlighting challenges and achievements in the application of these powerful technologies.

Background:Over the past 2 decades, molecular biology has shown that most major allergens exist in multiple isoforms. Very little is known about the relevance of allergen isoforms at the level of expressed protein (ie, actual allergen exposure). Objective:The aim of this study was to evaluate the applicability of state-of-the-art quadrupole time-of-flight mass spectrometry (Q-TOF MSMS) to the identification and quantification of allergen isoforms at the protein level. Methods:Q-TOF MSMS is a mass spectrometric approach for sequencing peptides and proteins. In our study it was applied to recombinant (r)Mal d 1, rBet v 1a and rBet v 1d, and natural (n)Mal d 1 from fruits of Malus domestica, cultivar Granny Smith. Results:Q-TOF MSMS allowed sequencing of about 70% of all amino acids in Mal d 1 and about 60% of those in Bet v 1. Mixing experiments with rBet v 1 isoforms and with rMal d 1 and nMal d 1 revealed that the technique allows identification of isoforms in mixtures down to a level of at least 5%. Recombinant Mal d 1 was identified as a Mal d 1a representative, whereas Granny Smith apples were shown to contain Mal d 1b-like allergen isoforms. In this apple cultivar hitherto unreported modifications of Mal d 1b were identified. Q-TOF MSMS allowed semiquantitative measurement of allergen at the femtomole to picomole level. Conclusion:Q-TOF MSMS is a powerful tool to find out whether an allergen isoform, detected at the cDNA level, is really expressed in quantities relevant for the allergic patient. (J Allergy Clin Immunol 2002;110:131-8.)

Three putative α1→3/4‐fucosyltransferase (α1→3/4‐FucT) genes have been detected in the Arabidopsis thalianagenome. The products of two of these genes have been identified in vivo as core α1→3‐FucTs involved in N‐glycosylation. An orthologue of the third gene was isolated from a Beta vulgariscDNA library. The encoded enzyme efficiently fucosylates Galβ1→3GlcNAcβ1→3Galβ1→4Glc. Analysis of the product by 400 MHz 1H‐nuclear magnetic resonance spectroscopy showed that the product is α1→4‐fucosylated at the N‐acetylglucosamine residue. In vitro, the recombinant B. vulgarisα1→4‐FucT acts efficiently only on neutral type 1 chain‐based glycan structures. In plants the enzyme is expected to be involved in Lewisaformation on N‐linked glycans.

Three putative α1→3/4-fucosyltransferase (α1→3/4-FucT) genes have been detected in the Arabidopsis thalianagenome. The products of two of these genes have been identified in vivo as core α1→3-FucTs involved in N-glycosylation. An orthologue of the third gene was isolated from a Beta vulgariscDNA library. The encoded enzyme efficiently fucosylates Galβ1→3GlcNAcβ1→3Galβ1→4Glc. Analysis of the product by 400 MHz 1H-nuclear magnetic resonance spectroscopy showed that the product is α1→4-fucosylated at the N-acetylglucosamine residue. In vitro, the recombinant B. vulgarisα1→4-FucT acts efficiently only on neutral type 1 chain-based glycan structures. In plants the enzyme is expected to be involved in Lewis aformation on N-linked glycans.

Equistatin (EI) is a cysteine protease inhibitor that was isolated from the sea anemone Actinia equina.It belongs to a recently discovered group of thyroglobulin type-I domain inhibitors called thyropins. Since native EI is found only in low amounts in the body of sea anemone and expression of recombinant EI in Escherichia coliyielded only 1 mg/liter of protein, we used the Pichia pastorisexpression system to obtain higher yields. A cDNA encoding EI was inserted into pPIC9 vector and transformed into the P. pastoris,strain GS115. Clones expressing high levels of EI were selected from 48 transformants. Recombinant EI was produced in 2-liter shake flasks and recovered from the fermentation broth by affinity chromatography using CM-papain–Sepharose. SDS–PAGE and N-terminal sequence analysis revealed that EI was N-terminally intact and running at the expected molecular weight of 22 kDa. The equilibrium dissociation constants of EI with papain and bovine cathepsin D were determined and were found to be similar to the results for the native inhibitor. EI production was scaled up to a bench top fermentor with a 25 mg/liter yield of active EI.

A cDNA encoding a precursor of equistatin, a potent cysteine and aspartic proteinase inhibitor, was isolated from the sea anemone Actinia equina. The deduced amino acid sequence of a 199-amino-acid residue mature protein with 20 cysteine residues, forming three structurally similar thyroglobulin type-1 domains, is preceded by a typical eukaryotic signal peptide. The mature protein region and those coding for each of the domains were expressed in the periplasmic space of Escherichia coli, isolated, and characterized. The whole recombinant equistatin and its first domain, but not the second and third domains, inhibited the cysteine proteinase papain (Ki 0.60 nM) comparably to natural equistatin. Preliminary results on inhibition of cathepsin D, supported by structural comparison, show that the second domain is likely to be involved in activity against aspartic proteinases.

N-Acetylglucosaminyltransferase I (GlcNAcT-I, EC 2.4.1.101) is the enzyme which initiates the formation of complex N-linked glycans in eukaryotes by transforming GlcNAc to the oligo-mannosyl acceptor Man5GlcNAc2-Asn. The enzymatic activity and the structure that is synthesised by this enzyme are found in animals and plants but not in yeast. cDNAs encoding the enzyme have already been cloned from several mammals and the nematode Caenorhabditis elegans.In this article the cloning of an Arabidopsis thalianaGlcNAcT-I cDNA with homology to animal cDNAs is described. By expression of the plant cDNA in CHO Lec1 cells, a mammalian cell line deficient in GlcNAcT-I, it was shown that it encodes an active enzyme with the same enzymatic activity as the animal homologue. It has already been shown that a human GlcNAcT-I can complement an A. thalianamutant (cgl-1). Here it is shown that the reverse is also true, the plant glycosyltransferase is able to complement a mammalian mutant (Lec1) deficient in GlcNAcT-I.

Tobacco and tomato plants were generated exhibiting insect resistance due to the introduction of modified cryIA(b) and cryIC genes of Bacillus thuringiensis. Limited modifications at selected regions of the coding sequences of both genes are sufficient to obtain resistance against Spodoptera exigua, Heliothis virescens and Manduca sexta. The criteria used to modify both genes demonstrate that the removal of sequence motifs potentially resulting in premature polyadenylation and transcript instability causes increased insect resistance. The expression of a cryIC-cryIA(b) fusion resulting in protection against S. exigua, H. virescens and M. sexta demonstrates the potential of expressing translational fusions, not only to broaden the insect resistance of transgenic plants, but also to simultaneously employ different gene classes in resistance management strategies.

To explore the feasibility of employing antibodies to obtain disease resistance against plant root pathogens, we have studied the expression of genes encoding antibodies in roots of transgenic plants. A model monoclonal antibody was used that binds to a fungal cutinase. Heavy and light chain cDNAs were amplified by PCR, fused to a signal sequence for secretion and cloned behind CaMV 35S and TR2' promoters in a single T-DNA. The chimeric genes were cloned both in tandem and in a divergent orientation. The roots of tobacco plants transformed with these constructs produced antibodies that were able to bind antigen in an ELISA. Immunoblotting showed assembly to a full-size antibody. In addition, a F(ab')2-like fragment was observed, which is probably formed by proteolytic processing. Both antibody species were properly targeted to the apoplast, but the full-size antibody was partially retained by the wall of suspension cells. The construct with divergent promoters showed a better performance than the construct with promoters in tandem. It directed the accumulation of functional antibodies to a maximum of 1.1% of total soluble protein, with half of the plants having levels higher than 0.35%. The high efficiency of this construct probably results from coordinated and balanced expression of light and heavy chain genes, as evidenced by RNA blot hybridization.

Short stretches of similar sequences have been detected in unrelated bacterial outer membrane proteins (Nikaido and Wu (1984) Proc. Natl. Acad. Sci. USA 81, 1048-1052). In the most pronounced similarity region, only a glycine residue is absolutely conserved. To investigate whether this glycine residue is essential for outer membrane incorporation, oligonucleotide-directed mutagenesis was applied to replace this residue, i.e. Gly-144, as well as two other Gly-residues in pore protein PhoE. Substitution of Gly-52 and Gly-258 by Ala And Val, respectively, did not influence outer membrane incorporation. However, the substitution of Gly-144 by Leu affected the efficiency of outer membrane incorporation. After in vitro synthesis this mutant protein was less efficiency precipitated with monoclonal antibodies that recognize coformational epitopes than wild-type PhoE, showing that the mutation interferes with correct folding of the protein.

The amino acid sequences necessary for entomocidal activity of the CryIA(b) protoxin of Bacillus thuringiensis were determined. Introduction of stop codons behind codons Arg601, Phe604 or Ala607 showed that amino acid residues C-terminal to Ala607 are not required for insecticidal activity and that activation by midgut proteases takes place distal to Ala607. The two shortest polypeptides, deleted for part of the highly conserved ß-strand, were prone to proteolytic degradation, explaining their lack of toxicity. Apparently, this ß-strand is essential for folding of the molecule into a stable conformation. Proteolytic activation at the N-terminus was investigated by removing the first 28 codons, resulting in a translation product extending from amino acid 29 to 607. This protein appeared to be toxic not only to susceptible insect larvae such as Manduca sexta and Heliothis virescens, but also to Escherichia coli cells. An additional mutant, encoding only amino acid residues 29–429, encompassing the complete putative pore forming domain, but lacking a large part of the receptor-binding domain, was similarly toxic to E. coli cells. This suggests a role for the N-terminal 28 amino acids in rendering the toxin inactive in Bacillus thuringiensis, and indicates that the cytolytic potential of the pore forming domain is only realized after proteolytic removal of these residues by proteases in the insect gut. In line with this hypothesis are results obtained with a mutant protein in which Arg28 at the cleavage site was replaced by Asp. This substitution prevented the protein from being cleaved by trypsin in vitro, and reduced its toxicity to M. sexta larvae.

Full-length, truncated, and mature forms of the CryIA(b) insecticidal crystal protein gene of Bacillus thuringiensis were engineered into the p10 locus of Autographa californica nuclear polyhedrosis virus (AcNPV). A signal sequence of Heliothis virescens juvenile hormone esterase was introduced at the N-terminus of these constructs to induce secretion. All recombinants, except those containing the mature toxin, produced high levels of CryIA(b) ICPs in insect cells. Thirty percent of the intracellular protoxin was N-glycosylated, suggesting that the protoxin was translocated across the ER membrane. Secretion into the medium, however, was limited. The production of the mature toxin was poor as a result of its cytotoxicity to insect cells. In a bioassay against second instar Spodoptera exigua larvae, using a recombinant expressing the Androctonus australis scorpion toxin gene in the same p10 locus as a positive control, the median survival time of AcNPV recombinants expressing the various B. thuringiensis CryIA(b) ICP constructs was not significant different from that of wild-type AcNPV. This suggests that production and/or secretion of B. thuringiensis (pro)toxins by AcNPV p10 recombinant viruses does not increase insecticidal activity since (i) the protoxins produced are inactive and not likely to be activated in vivo; (ii) secretion of the B. thuringiensis protoxins is poor; and (iii) production of the mature toxins results in cytotoxicity. Copyright 1995, 1999 Academic Press

Expression of single-chain antibody fragments (scFvs) in the plant cytosol is often cumbersome. It was unexpectedly shown that addition at the C-terminus of the ER retention signal KDEL resulted in significantly improved expression levels. In this report the cytosolic location of the scFv-CK was confirmed, excluding possible mistranslocation to other subcellular compartments. It was shown that expression of several other scFvs was also improved in tobacco protoplasts. In addition expression was improved in transgenic potato. Changing from KDEL to KDEI did not affect the enhanced protein expression level. Addition of the KDEL motif is a simple and straightforward tool to stabilize in planta cytosolic expression of many scFvs.

To obtain Bacillus thuringiensis crystal proteins with new properties and to identify the regions involved in insecticidal activity, we generated hybrid genes composed of cryIC and cryIE by in vivo recombination. Analysis of the hybrid proteins showed that domain III of CryIC is involved in the toxicity towards Spodoptera exigua and Mamestra brassicae. Transfer of this domain to CryIE, which is not active against these insects, resulted in a new protein with a broader activity. This hybrid protein binds to different receptors than CryIC, suggesting its use as an alternative for CryIC in resistance management programs.

PhoE protein of Escherichia coli K12 is an outer membrane protein which is supposed to span the membrane sixteen times. By creating a deletion which removes the last membrane-spanning fragment and studying the localization of the truncated PhoE, we show that this fragment is indispensable for trimerization and outer membrane localization. In addition, circumstantial evidence for the proposed topology model of the protein was obtained. An insertion mutation in a region supposed to be cell surface-exposed, interferes with the binding of a monoclonal antibody which recognizes a cell surface-exposed epitope of the protein.

Potato proteinase inhibitor II (PI2) is a serine proteinase inhibitor composed of two domains that are thought to bind independently to proteinases. To determine the activities of each domain separately, various inactive and active domain combinations were constructed by substituting amino acid residues in the active domains by alanines. These derivatives were expressed as soluble protein inEscherichia coli and exposed on M13 phage as fusions to gene 3 in a phagemid system for monovalent phage display. Inactivation of both active domains by Ala residues reduced binding of phage to trypsin and chymotrypsin by 95%. Ten times more phage were bound to proteinases by domain II compared to domain I, while a point mutation (Leu5 ? Arg) altered the binding specificity of domain I of PI2 phage from chymotrypsin to trypsin. The mutants were used to show that functional PI2 phage mixed with nonfunctional PI2 phage could be enriched 323 000-fold after three rounds of panning. Thus, these results open up the possibility to use phage display for the selection of engineered PI2 derivatives with improved binding characteristics towards digestive proteinases of plants pests.

Cecropin B is a small antibacterial peptide from the giant silkmothHyalophora cecropia. To reveal the potential of this peptide for engineering bacterial disease resistance into crops, several cecropin B gene constructs were made either for expression in the cytosol or for secretion. All constructs were cloned in a plant expression vector and introduced in tobacco viaAgrobacterium tumefaciens. A cDNA-derived cecropin B gene construct lacking the amino-terminal signal peptide was poorly expressed in transgenic plants at the mRNA level, whereas plants harbouring a full-length cDNA-derived construct containing the insect signal peptide, showed increased cecropin B-mRNA levels. Highest expression was found in plants harbouring a construct with a plant-gene-derived signal peptide. In none of the transgenic plants could the cecropin B peptide be detected. This is most likely caused by breakdown of the peptide by plant endogenous proteases, since a chemically synthesized cecropin B peptide was degraded within seconds in various plant cell extracts. This degradation could be prevented by the addition of specific protease inhibitors and by boiling the extract prior to adding the peptide. In addition, anionic detergents, in contrast to cationic, zwitter-ionic or non-ionic detergents, could prevent this degradation. Nevertheless, transgenic tobacco plants were evaluated for resistance toPseudomonas solanacearum, the causal agent of bacterial wilt of many crops, andP. syringae pv.tabaci, the causal agent of bacterial wildfire, which are highly susceptible to cecropin Bin vitro. No resistance was found. These experiments indicate that introduction and expression of cecropin B genes in tobacco does not result in detectable cecropin B protein levels and resistance to bacterial infections, most likely due to degradation of the protein by endogenous proteases.

The growth hormone gene of the rainbow trout (tGH-II) was expressed in leaves of transgenic tobacco plants and seeds of transgenicArabidopsis plants using tissue-specific promoters. Although in the leaves and the seeds comparble amounts of tGH-II mRNA could be detected, the protein could only be identified in the tobacco leaves. Passage of the hormone into the secretory pathway, mediated by the signal sequence of the extracellular tobacco PRI-b (pathogenesis-related) protein, resulted in correct disulphide bridge formation and (partial) glycosylation of the hormone. In contrast, cytoplasmic expression resulted in misfolding and partial breakdown of the protein. The data demonstrate that synthesis, folding and glycosylation of heterologous proteins in plants is dependent both on subcellular location as well as on the tissue or cell type in which the protein is expressed.

A better understanding of the mode of action ofBacillus thuringiensisdelta-endotoxins is needed to develop strategies which may prevent or slow down selection for resistance. We studied the effect of Cry1C on several different cultured insect cell lines by means of toxicity assays, ligand blotting, and toxin binding studies. A clear difference in sensitivity toward Cry1C between the insect cell lines was observed.Spodoptera frugiperdacell line Sf9 was most sensitive, whereasSpodoptera exiguacell lines SeUCR and SelZD2109 showed intermediate sensitivity.Mamestra brassicae(Mb0503) andDrosophila melanogaster(Dm1) cells were the least sensitive as compared to Sf9 cells. Ligand blot analysis of SDS–PAGE size-separated proteins showed that Cry1C specifically binds to a 40-kDa protein in Sf9, SeUCR, and SelZD2109 cells. Cry1Ab does not bind to this protein. The Cry1C-binding protein was not observed in Mb0503 and Dm1 cells, suggesting that the presence of the 40-kDa Cry1C-binding protein is correlated with sensitivity toward Cry1C.

Several models for the transport of proteins across membranes predict a role for lipids. If these models are correct, then alterations in lipid metabolism may affect protein export and vice versa. We are investigating this possibility by studying Escherichia coli K-12 mutants with defects in protein export or phospholipid metabolism. A temperature-sensitive secA mutant, which is defective in protein export at 42°C, exhibited severe pleiotropic effects on membrane biogenesis. Incubation of this strain at 42°C resulted in the appearance of intracytoplasmic membranes, in alterations in lipopolysaccharide structure and in decreased cardiolipin and C18:1 fatty acid content. On the other hand, a pgsA mutant which is defective in the synthesis of acidic phospholipids, exhibited a protein export defect when studied in vivo or in vitro. These results are in agreement with a postulated role of membrane lipids in protein export.

Short stretches of similar sequences have been detected in unrelated bacterial outer membrane proteins (Nikaido and Wu (1984) Proc. Natl. Acad. Sci. USA 81, 1048‐1052). In the most pronounced similarity region, only a glycine residue is absolutely conserved. To investigate whether this glycine residue is essential for outer membrane incorporation, oligonucleotide‐directed mutagenesis was applied to replace this residue, i.e. Gly‐144, as well as two other Gly‐residues in pore protein PhoE. Substitution of Gly‐52 and Gly‐258 by Ala And Val, respectively, did not influence outer membrane incorporation. However, the substitution of Gly‐144 by Leu affected the efficiency of outer membrane incorporation. After in vitro synthesis this mutant protein was less efficiency precipitated with monoclonal antibodies that recognize coformational epitopes than wild‐type PhoE, showing that the mutation interferes with correct folding of the protein.

Expression of single‐chain antibody fragments (scFvs) in the plant cytosol is often cumbersome. It was unexpectedly shown that addition at the C‐terminus of the ER retention signal KDEL resulted in significantly improved expression levels. In this report the cytosolic location of the scFv‐CK was confirmed, excluding possible mistranslocation to other subcellular compartments. It was shown that expression of several other scFvs was also improved in tobacco protoplasts. In addition expression was improved in transgenic potato. Changing from KDEL to KDEI did not affect the enhanced protein expression level. Addition of the KDEL motif is a simple and straightforward tool to stabilize in planta cytosolic expression of many scFvs.

To study the effects of small insertions on the biogenesis and functioning of outer membrane pore protein PhoE of Escherichia coli K12, oligonucleotides were inserted at five different sites in the phoE gene. The proteins encoded by the mutant alleles all appeared to be incorporated into the outer membrane since cells producing these proteins bound either phages specific for the PhoE protein or monoclonal antibodies or both. However, one mutant protein was apparently not incorporated normally since expression of this protein was lethal and the protein did not function as a pore. Amino acid residues 75 and 280 of the PhoE protein appear to be exposed on the cell surface because insertions at these sites interfere with the binding of PhoE-specific phages or monoclonal antibodies to whole cells, respectively.

The effects of subcellular localization on single-chain antibody (scFv) expression levels in transgenic tobacco was evaluated using an scFv construct of a model antibody possessing different targeting signals. For translocation into the secretory pathway a secretory signal sequence preceded the scFv gene (scFv-S). For cytosolic expression the scFv antibody gene lacked such a signal sequence (scFv-C). Also, both constructs were provided with the endoplasmic reticulum (ER) retention signal KDEL (scFv-SK and scFv-CK, respectively). The expression of the different scFv constructs in transgenic tobacco plants was controlled by a CaMV 35S promoter with double enhancer. The scFv-S and scFv-SK antibody genes reached expression levels of 0.01% and 1% of the total soluble protein, respectively. Surprisingly, scFv-CK transformants showed considerable expression of up to 0.2% whereas scFv-C transformants did not show any accumulation of the scFv antibody. The differences in protein expression levels could not be explained by the steady-state levels of the mRNAs. Transient expression assays with leaf protoplasts confirmed these expression levels observed in transgenic plants, although the expression level of the scFv-S construct was higher. Furthermore, these assays showed that both the secretory signal and the ER retention signal were recognized in the plant cells. The scFv-CK protein was located intracellularly, presumably in the cytosol. The increase in scFv protein stability in the presence of the KDEL retention signal is discussed.

The Escherichia coli single-stranded DNA binding protein (SSB) is implicated in DNA replication, recombination and repair1. On the chromosome, the ssb gene is located adjacent to the excision repair gene uvrA, but the two genes are transcribed in opposite directions2. uvrA has been shown to be part of the E. coli SOS system by introducing Mud(Ap, lac) insertions distal to the regulatory region of the gene in the chromosome3. Recent investigations suggest that SSB is also involved in the SOS response1,4,5. However, because the SSB protein is essential to the cell, the inducibility of the ssb gene cannot be investigated by the insertion method. Therefore, we used plas-mids harbouring the regulatory region of ssb fused to the galK structural gene, while leaving an intact ssb gene in the chromosome. We show here that expression of the ssb gene is dependent on two promoters of which one is damage inducible. Evidence is presented that the divergently transcribed ssb and uvrA genes are controlled by a common LexA binding site.

Nucleotide-sugar transporters (NSTs) form a family of structurally related transmembrane proteins that transport nucleotide-sugars from the cytoplasm to the endoplasmic reticulum and Golgi lumen. In these organelles, activated sugars are substrates for various glycosyltransferases involved in oligo- and polysaccharide biosynthesis. The Arabidopsis thaliana genome contains more than 40 members of this transporter gene family, of which only a few are functionally characterized. In this study, two Arabidopsis UDP-galactose transporter cDNAs (UDP-GalT1 and UDP-GalT2) are isolated by expression cloning using a Chinese hamster ovary cell line (CHO-Lec8) deficient in UDP-galactose transport. The isolated genes show only 21% identity to each other and very limited sequence identity with human and yeast UDP-galactose transporters and other NSTs. Despite this low overall identity, the two proteins clearly belong to the same gene family. Besides complementing Lec8 cells, the two NSTs are shown to transport exclusively UDP-galactose by an in vitro NST assay. The most homologous proteins with known function are plant transporters that locate in the inner chloroplast membrane and transport triose-phosphate, phosphoenolpyruvate, glucose-6-phosphate, and xylulose 5-phosphate. Also, the latter proteins are members of the same family, which therefore has been named the NST/triose-phosphate transporter family.

Crystal proteins of Bacillus thuringiensis are known for their insecticidal specificity. This specificity is, to a large extent, determined by the interaction of the proteins with high-affinity binding sites on the epithelial membrane of the midgut of sensitive insects. In particular, domain II of the three domains of the toxic moiety has been implicated in specificity. To determine which sequences of the protein are involved in binding, loops of domain II which terminate in the molecular apex of CryIA(b) were replaced by the corresponding regions of CryIE, a protein with different binding characteristics and insect specificity. In contrast to expression of the wild-type genes, expression of the mutant alleles in Escherichia coli resulted in the formation of biologically inactive, insoluble aggregates. Although these aggregates could be solubilized in vitro using urea, in contrast to the wild-type CryIA(b), the mutant proteins did not correctly refold as is shown by their increased protease sensitivity and lack of biological activity. The results indicate that engineering CryI proteins, based on the CryIIIA structure, is likely to prove difficult, particularly since the conformation of CryIIIA and CryI proteins might differ in domain II.

The regulation of the ssb gene of E.coli has been studied. We reported earlier that the SOS box of the neighbouring uvrA gene also controls the transcription of the ssb gene. Detailed analysis of the upstream region of ssb by S1 mapping reveals the existence of three in vivo functional promoters of which the most upstream one (PI) is inducible by DNA damage. Measurement of galactokinase synthesis using galK fusion plasmids indicates that the uninduced level of transcription from the PI promoter is low. Ssb multicopy plasmids lacking the PI promoter still complement the UV sensitivity of an Ssb mutant. The role of the three promoters in the regulation of the level of Ssb protein in the cell, is discussed.

Introduction of positively charged amino acids immediately downstream of the signal sequence in prokaryotic precursor proteins is known to affect the export process. However, it is not clear whether different positively charged amino acids affect the export process similarly. To investigate this, the glutamate at position +2 of outer membrane protein PhoE was substituted by arginine, lysine of histidine. Pulse-chase experiments revealed that the Lys and Arg residues at position +2 caused a reduced processing rate, and that the effect was markedly more severe in the case of the Arg residue. Trypsin accessibility experiments revealed that the accumulated precursors were present in the cytoplasm. Since the degree of the inhibitory effect corresponded to the pKa of the different positively charged amino acids, this suggests that the positively charged residues must be deprotonated during the secretory process.