Glucose-6-phosphate dehydrogenase-6-phosphogluconolactonase. A novel bifunctional enzyme in malaria parasites.

Plasmodium falciparum glucose 6-phosphate dehydrogenase (Pf Glc6PD), compared to other Glc6PDs has an additional 300 amino acids at the N-terminus. They are not related to Glc6PD but are similar to a family of proteins (devb) of unknown function, some of which are encoded next to Glc6PD in certain bacteria. The human devb homologue has recently been shown to have 6-phosphogluconolactonase (6PGL) activity. This suggests Pf Glc6PD may be a bifunctional enzyme, the evolution of which has involved the fusion of adjacent genes. Further functional analysis of Pf Glc6PD has been hampered because parts of the gene could not be cloned. We have isolated and sequenced the corresponding Plasmodium berghei gene and shown it encodes an enzyme (Pb Glc6PD) with the same structure as the P. falciparum enzyme. Pb Glc6PD is 950 amino acids long with significant sequence similarity in both the devb and Glc6PD domains with the P. falciparum enzyme. The P. berghei enzyme does not have an asparagine-rich segment between the N and C halves and it contains an insertion at the same point in the Glc6PD region as the P. falciparum enzyme but the insertion in the P. berghei is longer (110 versus 62 amino acids) and unrelated in sequence to the P. falciparum insertion. Though expression of this enzyme in bacteria produced largely insoluble protein, conditions were found where the full-length enzyme was produced in a soluble form which was purified via a histidine tag. We show that this enzyme has both Glc6PD and 6PGL activities. Thus the first two steps of the pentose phosphate pathway are catalysed by a single novel bifunctional enzyme in these parasites.

Identification of an intronic regulatory element in the human protein C (PROC) gene.

Regulatory DNA elements responsible for human protein C (PROC) gene expression have previously been identified in the upstream promoter region and first (untranslated) exon of the gene. Here we show that an additional sequence element located more than 500 bp downstream of the core promoter within intron 1 further enhances PROC promoter-driven reporter gene expression in human hepatoma cells. In common with core promoter constructs used in previous studies, the activity of this 3'-extended regulatory region is diminished by a naturally occurring promoter mutation. However, in contrast to constructs lacking intronic sequence, the promoter/intron regulatory region is repressed rather than activated by the transcription factor HNF-1. Using both conventional alignment procedures and complexity analysis to study the human and canine PROC sequences, we identified two conserved intronic regions, which were tested for their involvement in gene regulation. High-level gene expression from the intron-coupled promoter was dependent upon the integrity of a 142 bp sequence element, a duplicate copy of which is located in an upstream region of the PROC gene that possesses enhancer activity. These findings emphasise the potential importance of intragenic sequences for gene regulation and serve to illustrate that the results of PROC promoter/reporter gene experiments are critically dependent upon the sequence context. The identification of such intragenic elements is relevant to the analysis of human genetic disease since it will facilitate the detection and functional evaluation of regulatory mutations and polymorphisms.

Human hexose-6-phosphate dehydrogenase (glucose 1-dehydrogenase) encoded at 1p36: coding sequence and expression.

Using the published protein sequence from a rabbit microsomal glucose-6-phosphate dehydrogenase G6PD we have isolated and sequenced a cDNA clone coding for its human equivalent, which is also known as hexose-6-phosphate dehydrogenase (H6PD) and glucose dehydrogenase. The corresponding genomic sequence is in the databases enabling its localization to chromosome 1p36. The gene spans 37 kb and consists of 5 exons, the fifth of which codes for more than half of the 89 kDa protein. The first intron is a 10 kb insertion in the 5' untranslated sequence. The predicted mRNA has an exceptionally long (6.5 kb) 3' untranslated sequence. The predicted protein shows extensive homology with X-linked G6PD, suggesting the two genes share a common ancestor but no intron positions are conserved between the two genes suggesting the gene duplication was an ancient event. The C-terminal portion of the protein is not homologous with G6PD but shows limited homology with proteins of unknown function found throughout evolution and encoded next to G6PD in various micro-organisms. Intriguingly this C-terminal portion has some homology with the N-terminal sequence of Plasmodium falciparum G6PD.

Solution of the structure of tetrameric human glucose 6-phosphate dehydrogenase by molecular replacement.

Recombinant human glucose 6-phosphate dehydrogenase (G6PD) has been crystallized and its structure solved by molecular replacement. Crystals of the natural mutant R459L grow under similar conditions in space groups P212121 and C2221 with eight or four 515-residue molecules in the asymmetric unit, respectively. A non-crystallographic 222 tetramer was found in the C2221 crystal form using a 4 A resolution data set and a dimer of the large beta + alpha domains of the Leuconostoc mesenteroides enzyme as a search model. This tetramer was the only successful search model for the P212121 crystal form using data to 3 A. Crystals of the deletion mutant DeltaG6PD grow in space group F222 with a monomer in the asymmetric unit; 2.5 A resolution data have been collected. Comparison of the packing of tetramers in the three space groups suggests that the N-terminal tail of the enzyme prevents crystallization with exact 222 molecular symmetry.

Amino acid substitutions at the dimer interface of human glucose-6-phosphate dehydrogenase that increase thermostability and reduce the stabilising effect of NADP.

Over 100 mutations of the G6PD gene have been documented. With the construction of the molecular model of glucose-6-phosphate dehydrogenase, based on the structure of the bacterial Leuconostoc mesenteroides glucose-6-phosphate dehydrogenase, it has been possible to superimpose these amino acid changes on to the structure of the glucose-6-phosphate dehydrogenase molecule. There are a large number of severe disease causing mutations at the dimer interface which usually cause decreased thermostability. We have used this knowledge to predict amino acid changes which would effect an increase in the stability of the dimer. The aspartic acid at residue 421 was chosen as it is a negatively charged residue at the centre of the dimer interface in an area rich in negatively charged residues. This residue was changed to a neutrally charged alanine or asparagine, or a positively charged lysine or arginine. The thermostability of the enzyme was increased when residue 421 was neutral (A or N) and increased further when positive (K or R). NADP is known to exert a concentration dependent stabilising effect on the glucose-6-phosphate dehydrogenase dimer. However the concentration-dependent stabilising effect of NADP was reduced in the residue-421 substitutions in a manner which was inversely proportional to charge change. These results suggest that changes at the dimer interface can also affect the distant (> 20 A) NADP-binding site, and vice versa; an attempt has been made to explain these interactions based on the molecular model of human glucose-6-phosphate dehydrogenase.

Functional analysis of an unusual length polymorphism in the human antithrombin III (AT3) gene promoter.

The prevalence of the alternative alleles of an unusual length polymorphism in the promoter of the human antithrombin III (AT3) gene was determined in a sample of 155 unrelated individuals from the Northern Irish population. The 108bp L allele and the 32bp S allele occurred at frequencies of 0.21 and 0.79 respectively. Some homology was noted between the L-specific sequence and the region immediately downstream. Residual homology was also evident between the L and S sequences, suggesting that the S allele was derived from the L allele during evolution by partial deletion followed by sequence divergence. The functional significance of the polymorphism was investigated by transient transfection of AT3 promoter/luciferase reporter gene constructs into two human hepatoma cell lines in vitro. The promoter strength of the L allele was found to be 1.6-fold higher than the S allele in HepG2 cells whereas in Hep3B cells, the strength of the S allele was 1.7-fold higher than that of the L allele. In order to evaluate the phenotypic consequences of the AT3 promoter polymorphism in vivo, plasma samples from the 155 control individuals were assayed for antithrombin III (ATIII) activity. Mean activities of the different promoter polymorphism genotypes (SS, LL, SL) were not significantly different. These results suggest that the AT3 promoter polymorphism does not contribute to the variation in plasma ATIII activity that occurs in the general population.

Disruption of a binding site for hepatocyte nuclear factor 1 in the protein C gene promoter is associated with hereditary thrombophilia.

A heterozygous T-->C transition was detected in the putative promoter region of the protein C (PROC) gene in a patient with type I protein C deficiency and a history of recurrent venous thrombosis. This mutation occurred 14 bp upstream of the transcription initiation site and within a sequence strongly homologous to the consensus binding site for the liver-enriched transcription factor, hepatocyte nuclear factor 1 (HNF-1). Transfection experiments demonstrated that a CAT reporter gene construct containing 626 bp of the putative PROC gene promoter was capable of driving CAT expression in HepG2 hepatoma cells. Levels of CAT expression from constructs bearing the mutation were found to be drastically reduced by comparison with the wild-type, consistent with the reduced plasma protein C antigen levels observed in the patient. Gel retardation and cotransfection experiments demonstrated that the mutation abolished both the binding and the transactivating ability of HNF-1 observed with the wild-type PROC gene promoter. Further, the ability of the mutation to disrupt HNF-1 binding appears to be a function not only of the nature of the nucleotide substitution and its position within the recognition sequence, but also of the relative affinity of the wild-type binding site for HNF-1. This analysis is therefore indicative of a vital role for HNF-1 in the expression of the PROC gene in vivo. Taken together with the identification of a human hepatoma cell line which contains HNF-1 but which does not express protein C, these findings are consistent with the view that HNF-1 is necessary although not sufficient for PROC gene expression in the liver.