The glucose-6-phosphate dehydrogenase from Trypanosoma cruzi: its role in the defense of the parasite against oxidative stress

The Trypanosoma cruzi glucose-6-phosphate dehydrogenase (G6PDH) is encoded by several genes located in three of the parasite chromosomes. All the sequences present two possible start codons, 111bp apart, also present in its Trypanosoma brucei counterpart. As the 37 residues comprised between the two candidate initiator methionines of T. brucei and T. cruzi G6PDHs constitute an unusual N-terminal extension only present in trypanosomatids, two forms of the T. cruzi G6PDH were expressed in Escherichia coli: a long one (Tc-G6PDH-L) translated from the first ATG codon, and a short one (Tc-G6PDH-S) translated from the second. Both were purified and their kinetic constants determined. The apparent K(m) for glucose-6-phosphate was 189.9, 98.4, and 288microM, for Tc-G6PDH-L, Tc-G6PDH-S and native Tc-G6PDH, respectively. The apparent K(m) for NADP was similar for both recombinant proteins. The Tc-G6PDH-L as well as the native enzyme, was inactivated by DTT while the Tc-G6PDH-S was unaffected by the reducing agent. This behavior could be related to the presence of two Cys groups in the N-terminal extension of the Tc-G6PDH-L similarly to the redox regulated G6PDHs from chloroplasts and cyanobacteria. This property, together with a remarkable induction (up to 46-fold) of the T. cruzi G6PDH in metacyclic trypomastigotes under oxidative stress conditions, suggests that the enzyme may play a prominent role in the defense mechanisms of the parasite against oxidative stress becoming an important target for chemotherapy. Western blots using antibodies against the N-terminal extension in Tc-G6PDH-L show that this form is expressed in the parasite.     

Dehydroepiandrosterone, glucose-6-phosphate dehydrogenase, and longevity

Dehydroepiandrosterone (DHEA) is an abundantly produced adrenal steroid whose biological role has never been clarified. DHEA is a potent uncompetitive inhibitor of mammalian glucose-6-phosphate dehydrogenase (G6PDH) and as a consequence lowers NADPH levels and reduces NADPH-dependent oxygen-free radical production. Overproduction of oxygen-free radicals, or oxidative stress, upregulates inflammation and cellular proliferation and is believed to play a critical role in the development of cancer, atherosclerosis, and Alzheimer's disease, as well as the basic aging process. Both in vitro and in vivo experimental studies strongly indicate that DHEA and related steroids inhibit inflammation and associated epithelial hyperplasia, carcinogenesis, and atherosclerosis, at least in part, through the inhibition of G6PDH and oxygen-free radical formation. Recent epidemiological findings in Sardinian males bearing the Mediterranean variant of G6PDH deficiency are consistent with the hypothesis that reduced G6PDH activity has a beneficial effect on age-related disease development and longevity. Clinical trials with DHEA are encumbered by the high oral doses required as well as the conversion of DHEA into active androgens. The use of less androgenic congeners as well as non-oral formulations may facilitate testing of this class of compounds.     

Purification, localisation and characterisation of glucose-6-phosphate dehydrogenase of Trypanosoma brucei

Cell-fractionation and digitonin titration of procyclic trypomastigotes of Trypanosoma brucei, revealed that almost half of the total NADP+ -dependent glucose-6-phosphate dehydrogenase (G6PDH) activity, the first enzyme of the pentose phosphate pathway (PPP), is associated with glycosomes. The specific activity of G6PDH in the purified organelles was increased 4-fold relative to a total cell extract and showed latency. Moreover, in the absence of detergents this activity was totally resistant to the action of trypsin. The cytosolic counterpart was neither latent, nor was it resistant to trypsin. Both cytosolic and glycosomal G6PDH activities behaved identically on phenyl-, CM-, heparin-, and Affigel-blue-Sepharose columns. Both isoenzymes had a subunit Mr of 62 000 and an isoelectric point of 6.85, while kinetic studies carried out on the partially purified G6PDH from both cell compartments did not reveal any differences. The purified enzyme had an apparent Km of 138 and 5.3 microM for glucose 6-phosphate (G6P), and for NADP+, respectively, and had a specific activity of 14 micromol. (min mg of protein)(-1). We conclude that while in procyclic stages of T. brucei G6PDH activity is present in two different cell compartments, i.e. the cytosol and the glycosomes, these two activities most likely represent one and the same isoenzyme.     

Comparative study of Leishmania mexicana and Trypanosoma brucei NAD-dependent glycerol-3-phosphate dehydrogenase

The NAD-dependent glycerol-3-phosphate dehydrogenases (G3PDH, EC 1.1.1.8) of Trypanosoma brucei and Leishmania mexicana are thought to have different roles in carbohydrate metabolism. Here the physicochemical and kinetic properties of natural G3PDH from T. brucei with the recombinant homologue of L. mexicana which share 63% positional identity are compared. Despite their supposed different functions in energy metabolism of the parasites the two G3PDHs have remarkably similar properties, including pH optima and K(m) value for dihydroxyacetone phosphate (DHAP) and NADH in the formation of glycerol 3-phosphate (G3P) and for NAD+ and G3P in the reverse reaction. Both enzymes are subject inhibition by dihydroxyacetone phosphate at concentrations above 0.2 mM and are inhibited by the trypanocidal drugs suramin and melarsen oxide at sub-micromolar concentrations.     

Dehydroepiandrosterone, epiandrosterone and synthetic derivatives inhibit Junin virus replication in vitro

In the present paper the in vitro antiviral activity of dehydroepiandrosterone (DHEA), epiandrosterone (EA) and 16 synthetic derivatives against Junin virus (JUNV) replication in Vero cells was studied. DHEA and EA caused a selective inhibition of the replication of JUNV and other members of the Arenaviridae family such as Pichinde virus and Tacaribe virus. The compounds were not virucidal to cell-free JUNV. The impairment of viral replication was not due to an inhibitory effect of the steroids on virus adsorption or internalization. An inhibitory effect of the compounds on JUNV protein synthesis and both intracellular and extracellular virus production was demonstrated. A partial inhibitory action on cell surface expression of JUNV glycoprotein G1 was also detected on DHEA- and EA-treated cultures. Like DHEA and EA, three compounds obtained from EA by chemical synthesis showed selectivity indexes higher than ribavirin, the only antiviral compound that has shown partial efficacy against arenavirus infections.     

Recombinant tumor necrosis factor alpha does not inhibit the growth of African trypanosomes in axenic cultures

Mice whose tumor necrosis factor alpha (TNF-alpha) genes were disrupted developed higher levels of parasitemia than wild-type mice following infection with Trypanosoma congolense IL1180 or T. brucei brucei GUTat3.1, confirming the results of earlier studies. To determine whether TNF-alpha directly affects the growth of these and other bloodstream forms of African trypanosomes, we studied the effects of recombinant mouse, human, and bovine TNF-alpha on the growth of two isolates of T. congolense, IL1180 and IL3338, and two isolates of T. brucei brucei, GUTat3.1 and ILTat1.1, under axenic culture conditions. The preparations of recombinant TNF-alpha used were biologically active as determined by their capacity to kill L929 cells. Of five recombinant TNF-alpha lots tested, one lot of mouse TNF-alpha inhibited the growth of both isolates of T. brucei brucei and one lot of bovine TNF-alpha inhibited the growth of T. brucei brucei ILTat1.1 but only at very high concentrations and without causing detectable killing of the parasites. The other lots of mouse recombinant TNF-alpha, as well as human TNF-alpha, did not affect the growth of any of the test trypanosomes even at maximal concentrations that could be attained in the culture systems (3,000 to 15,000 U of TNF-alpha/ml of medium). These results suggest that exogenously added recombinant TNF-alpha generally does not inhibit the growth of African trypanosomes under the culture conditions we used. The impact of TNF-alpha on trypanosome parasitemia may be indirect, at least with respect to the four strains of trypanosomes reported here.     

The electrophoretic mobilities and activities of eleven enzymes of bloodstream and culture forms of Trypanosoma brucei compared

Eleven soluble enzymes in the supernatant of bloodstream Trypanosoma brucei were compared for electrophoretic mobility and activity with those of T. brucei cultures grown in 3 different media. All bands of each enzyme found in the bloodstream form were also present in the cultured material, but extra bands of malate dehydrogenase (MDH) (EC 1.1.1.37), aspartate aminotransferase (ASAT) (EC 2.6.1.1), and in 2 to 6 cultures of isocitrate dehydrogenase (ICD) (EC 1.1.1.42) were present in culture forms but not in bloodstream forms. An interfering enzyme, peculiar to cultured T. brucei, which reacted with 2-oxoglutarate and possibly a trace amount of ammonium ions, ran with the fast-moving ASAT bands. Threonine dehydrogenase activity, high in cultured trypanosomes irrespective of the medium used but low in bloodstream trypanosomes, was markedly lower in Trypanosoma evansi and a much passaged T. brucei 8/18. Glucosephosphate isomerase activity on the other hand was high in bloodstream and low in cultured trypanosomes. Glutamate dehydrogenase activity was too low to record reliably in bloodstream trypanosomes, but could be clearly detected in cultured forms. As the differences point to some changes in gene expression between the two forms, culture material is likely to replace trypanosomes from living animals for electrophoretic characterization only when considerable comparative work has been done.     

Inhibition of tumor development by dehydroepiandrosterone and related steroids

The naturally occurring adrenal steroid, dehydroepiandrosterone (DHEA), is a potent non-competitive inhibitor of mammalian glucose-6-phosphate dehydrogenase (G6PDH). Oral administration of DHEA to mice inhibits spontaneous breast cancer and chemically induced tumors of the lung and colon. Topical application of DHEA to mouse skin inhibits 7,12-dimethylbenz(a)anthracene (DMBA)-initiated and tetradecanoylphorbol-13-acetate (TPA)-promoted papillomas and DMBA-induced carcinomas at both the initiation and promotion phase. Evidence is presented that critical steps in the initiation process (mixed-function oxidase activation of a carcinogen) and promotion process (enhanced rates of cell proliferation and superoxide formation) all require NADPH and may be inhibited by DHEA and structural analogs as a result of a lowering of the NADPH cellular pool. Results obtained by others with fibroblasts and lymphocytes from individuals with the Mediterranean variant of G6PDH deficiency also indicate that a reduction in the NADPH cellular pool confers resistance to benzo(a)pyrene. Preliminary data suggest that food restriction may depress G6PDH levels and this may contribute to the tumor preventive effect of underfeeding.     

Evidence for kinetoplast and nuclear DNA homogeneity in Trypanosoma evansi isolates

The kinetoplast DNA minicircles from 13 stocks of trypanosomes designated as Trypanosoma evansi were digested with various restriction enzymes. We also examined the distribution of restriction site polymorphisms in the nuclear DNA of 9 of these stocks, using 7 different variable surface glycoprotein (VSG) and non-VSG probes. Restricted kinetoplast DNA (kDNA) fragments of some of these strains were cloned into M13 or PUC 18 vectors and sequenced. The restriction and sequence mapping showed that most of T. evansi isolates belonged to the A1 and A2 types of Borst and to two new closely related types A3 and A4. A notable exception was RoTat 4/1 derived from a Sudanese stock which was found to display a characteristic brucei-like minicircle heterogeneity. The T. evansi minicircles analysed are not only homogeneous in sequence but also the region similar to the conserved region in Trypanosoma brucei and Trypanosoma equiperdum is flanked on its 5' end by a palindromic repeat of part of the conserved region. The highly conserved sequence GGGCGGT which appears to correspond to the initiation of synthesis of one of the Okazaki fragments contains an additional G and is located as in T. brucei and T. equiperdum about 73 bp 5' from the ORI. The nuclear DNA analysis confirms the kDNA study in that all the T. evansi stocks are members of a very homogeneous group in terms of sequence divergence. Moreover, our analysis also confirms that T. evansi is more closely related to the West African T. b. brucei and T. b. gambiense than to other African trypanosomes.     

Molecular and biochemical characterization of hexokinase from Trypanosoma cruzi

The Trypanosoma cruzi hexokinase gene has been cloned, sequenced, and expressed as an active enzyme in Escherichia coli. Sequence analysis revealed 67% identity with its counterpart in Trypanosoma brucei but low similarity with all other available hexokinase sequences including those of human. It contains an N-terminal peroxisome-targeting signal (PTS-2) and has a calculated basic isoelectric point (pI = 9.67), a feature often associated with glycosomal proteins. The polypeptide has a predicted mass of approximately 50 kDa similar to that of many non-vertebrate hexokinases and the vertebrate hexokinase isoenzyme IV. The natural enzyme was purified to homogeneity from T. cruzi epimastigotes and appeared to exist in several aggregation states, an apparent tetramer being the predominant form. Its kinetic properties were compared with those of the purified recombinant protein. Higher K(m) values for glucose and ATP were found for the (His)(6)-tag-containing recombinant hexokinase. However, removal of the tag produced an enzyme displaying similar values as the natural enzyme (K(m) for glucose = 43 and 60 microM for the natural and the recombinant protein, respectively). None of these enzymes presented activity with fructose. As reported previously for hexokinases from several trypanosomatids, no inhibition was exerted by glucose 6-phosphate (G6-P). In contrast, a mixed-type inhibition was observed with inorganic pyrophosphate (PPi, K(i) = 0.5mM).     

Role of expression site switching in the development of resistance to human Trypanosome Lytic Factor-1 in Trypanosoma brucei brucei

Human high-density lipoproteins (HDLs) play an important role in human innate immunity to infection by African trypanosomes with a minor subclass, Trypanosome Lytic Factor-1 (TLF-1), displaying highly selective cytotoxicity to the veterinary pathogen Trypanosoma brucei brucei but not against the human sleeping sickness pathogens Trypanosoma brucei gambiense or Trypanosoma brucei rhodesiense. T. b. rhodesiense has evolved the serum resistance associated protein (SRA) that binds and confers resistance to TLF-1 while T. b. gambiense lacks the gene for SRA indicating that these parasites have diverse mechanisms of resistance to TLF-1. Recently, we have shown that T. b. gambiense (group 1) resistance to TLF-1 correlated with the loss of the haptoglobin/hemoglobin receptor (HpHbR) expression, the protein responsible for high affinity binding and uptake of TLF-1. In the course of these studies we also examined TLF-1 resistant T. b. brucei cell lines, generated by long-term in vitro selection. We found that changes in TLF-1 susceptibility in T. b. brucei correlated with changes in variant surface glycoprotein (VSG) expression in addition to reduced TLF-1 binding and uptake. To determine whether the expressed VSG or expression site associated genes (ESAGs) contribute to TLF-1 resistance we prepared a TLF-1 resistant T. b. brucei with a selectable marker in a silent bloodstream expression site (BES). Drug treatment allowed rapid selection of trypanosomes that activated the tagged BES. These studies show that TLF-1 resistance in T. b. brucei is largely independent of the expressed VSG or ESAGs further supporting the central role of HpHbR expression in TLF-1 susceptibility in these cells.     

Telomere shortening and cell cycle arrest in Trypanosoma brucei expressing human telomeric repeat factor TRF1

Trypanosoma brucei has telomeres composed of 15 kb tracts of TTAGGG repeats that end in 3' overhangs and form t-loops. This structure is also present in mammalian cells and is thought to reflect the presence of telomere-binding proteins. The human TTAGGG repeat-binding factor TRF1 binds to telomeres and regulates their length. We attempted to interfere with the normal function of trypanosome telomeres by expressing human TRF1 in T. brucei. TRF1 localized to telomeres in cultured procyclic (midgut-stage) trypanosomes with great fidelity, but not in bloodstream-stage trypanosomes. Procyclic trypanosomes expressing high levels of TRF1 for extended periods of time exhibited shortening and increased size heterogeneity of their telomeres and the cell cycle was arrested in G1-S. These effects were not detected in cells expressing a TRF1 mutant incapable of binding to TTAGGG repeats. We argue that TRF1 displaces putative endogenous trypanosome telomere-binding proteins, not yet identified, and affects telomeres in ways that reflect its role as a negative regulator of telomere length in human cells.     

Metabolism of phospholipids and lysophospholipids by Trypanosoma brucei

African trypanosomes (Trypanosoma brucei brucei) rapidly metabolize exogenous 1-acyl-lysophospholipids by at least two routes: (1) hydrolysis by a phospholipase A1; (2) acylation by an acyl-CoA-dependent acyltransferase. In contrast to lysophospholipids, exogenous phospholipids are not rapidly metabolized by T. brucei. The acyltransferase (EC 2.3.1.23) converts exogenous 1-acyl lysophosphatidylcholine and exogenous acyl-CoA to phosphatidylcholine and CoA-SH. It is a membrane-bound enzyme and shows maximal activity within the first 2 min of exposure of trypanosomes to the exogenous substrates. The acyltransferase specificity for lysophospholipids is lysophosphatidylcholine greater than lysophosphatidylinositol greater than lysophosphatidylethanolamine greater than lysophosphatidate. Phosphatidylcholine enhances the enzyme activity towards lysophosphatidylethanolamine and lysophosphatidic acid. The preference for CoA acyl thioesters is oleoyl greater than palmitoyl greater than myristoyl greater than stearoyl greater than arachidonoyl, and this specificity distinguishes the protozoan enzyme from those of cells of mammalian hosts, which are specific for arachidonoyl-CoA. When the acyltransferase converts exogenous lysophosphatidylethanolamine to phosphatidylethanolamine, the latter is rapidly methylated to form dimethylphosphatidylethanolamine. There is also rapid hydrolysis of exogenous oleoyl-CoA by a thioester hydrolase in living trypanosomes, to yield free oleate and CoA-SH.     

The 6-phosphogluconate dehydrogenase from Trypanosoma cruzi: the absence of two inter-subunit salt bridges as a reason for enzyme instability

The third enzyme of the pentose phosphate pathway (PPP), 6-phosphogluconate dehydrogenase (6PGDH), is present in the four major stages of Trypanosoma cruzi, CL Brener clone. The enzyme was too unstable to be purified from epimastigote cell-free extracts. Two genes encoding 6PGDH were cloned and sequenced; the predicted amino acid sequences differ only in five non-essential residues. Since Southern blots suggested the presence of a single copy per haploid genome, the two genes found are probably alleles. One of these genes, encoding a protein with 78.6% identity with the Trypanosoma brucei 6PGDH, was expressed in Escherichia coli as an active recombinant enzyme, which was as unstable as the native 6PGDH. Modeling of the T. cruzi enzyme using the three-dimensional structure of the T. brucei 6PGDH as template suggested the lack of two out of five salt bridges proposed to strengthen subunit interactions in the active dimer. Restoring of these bridges by site-directed mutagenesis resulted in a more stable recombinant T. cruzi 6PGDH, which was used to determine the kinetic parameters. The K(m) value for 6-phosphogluconate (22.2+/-0.4 microM) was identical to the values reported for 6PGDHs from mammals, but the K(m) for NADP (5.9+/-0.2 microM) was significantly lower than the value reported for the human enzyme, and closer to that for the T. brucei enzyme. This suggests the possibility that inhibitors of the T. brucei 6PGDH, under development as potential drugs against African Trypanosomiasis, might also be successful for the chemotherapy of Chagas disease.     

Divalent metal requirements for catalysis and stability of the RNA triphosphatase from Trypanosoma cruzi

RNA triphosphatases act in the first step of the mRNA capping process, removing the gamma-phosphoryl group from the 5' end of nascent RNA. A metal-dependent catalysis is found in the enzymes from trypanosomes and several other lower eukaryotes. This contrasts with the cysteine-dependent activity of the corresponding enzymes of mammals, a difference that points to these enzymes as potential targets for drug design. This work describes the identification, expression, purification, enzyme kinetics, and the role of divalent metal in the ATPase activity of the RNA triphosphatase from Trypanosoma cruzi, the agent of Chagas' disease, and compares it with the previously characterized enzyme from Trypanosoma brucei. Sequence similarity of the T. cruzi enzyme with the RNA triphosphatase of Saccharomyces cerevisiae indicates that a tunnel domain containing the divalent metal forms its active site. Based on enzyme kinetics, circular dichroism, and intrinsic fluorescence analysis, a kinetic mechanism for the ATPase activity of the T. cruzi tunnel triphosphatase is proposed. A single metal is sufficient to interact with the enzyme through the formation of a productive MnATP-enzyme complex, while free ATP inhibits activity. Manganese is also required for the tunnel stability of the T. cruzi enzyme, while the T. brucei homologue remains stable in the absence of metal, as shown for other triphosphatases. These findings may be useful to devise specific triphosphatase inhibitors to the T. cruzi enzyme.     

Changes in the pathogenicity of relapsed diminazene aceturate (DA)-resistant Trypanosoma brucei brucei as the trypanosomes are transmitted from DA-treated hosts to another set of animals

This study investigated the changes in pathogenicity of relapsed diminazene aceturate (DA)-resistant Trypanosoma brucei brucei (TBB) as the trypanosomes are transmitted from DA-treated hosts to another set of animals. The Federe strain of T. brucei brucei which was known to be DA resistant was used to infect two groups of rats, one of which was treated at the peak of infection. There was temporary clearance of the trypanosomes from blood of the treated group and subsequent relapse of the infection. The relapsed trypanosomes were used to infect a group of rats (group A1), and the pathogenicity of the relapsed T. brucei brucei in this rat group was compared to that of a group infected with primary T. brucei brucei stock that was not treated with DA (group B1). Results showed that when compared with the B1 rats, the A1 group had a slower onset of parasitaemia, significantly lower (p < 0.01) level of parasitaemia all through the study, significantly longer (p < 0.01) time period to reach peak parasitaemia, significantly longer (p < 0.01) post-infection survival time and significantly higher (p < 0.05) packed cell volume from day 6 post-infection. Further sequential transmission of this same T. brucei brucei infection to another group of rats (A2 and B2) produced the same effects on the above parameters used to assess pathogenicity. It was concluded that beyond temporarily clearing trypanosomes from blood of treated animals, treatment with DA reduces the pathogenicity of DA-resistant T. brucei brucei, and this reduced pathogenicity is carried on as the T. brucei brucei is further sequentially transmitted to another set of animals.     

Kinetoplast DNA from Trypanosoma vivax and T. congolense

We have analysed kinetoplast DNA (kDNA) of the African trypanosomes Trypanosoma vivax and T. congolense. The maxi-circles from these organisms resemble those of T. brucei in size, but only to a limited extent in sequence as judged from restriction enzyme digests and DNA X DNA hybridization. The kDNA networks of T. vivax have three distinguishing features: they contain the highest maxi-circle concentration of any kDNA (at least twice that of T. brucei); they contain the smallest mini-circles (465 bp) yet found thus far and the width of the kDNA nucleoid in thin sections is correspondingly small (55 nm against 91 nm for T. brucei); they contain a substantial fraction of mini-circle dimers.     

Early course of infection in susceptible and resistant strains of mice, using [3H]uridine-labeled Trypanosoma brucei subsp. brucei.

A radioisotopic technique utilizing [3H]uridine was developed to determine the tissue distribution of intravenously inoculated Trypanosoma brucei spp. brucei in susceptible (Swiss-Webster) and resistant (deer) strains of mice. The reliability of the technique was tested by using unincorporated [3H]uridine and heat-killed (labeled) trypanosomes in parallel experiments. The inoculations with viable (labeled) organisms showed that during the initial 24 h T. brucei subsp. brucei was cleared from the bloodstream of deer mice to a significantly greater extent than in Swiss-Webster mice. In addition, significantly more radioactivity was found in the liver, spleen, and kidneys of the deer mice. Autoradiographs of trypanosomes in selected tissues supported the distribution observed with labeled organisms. All of the above experiments involved the highly virulent, monomorphic bloodstream form of T. brucei subsp. brucei. Similar experiments with less virulent tissue culture-adapted trypanosomes showed that these organisms were readily cleared from the bloodstream, even in highly susceptible Swiss-Webster mice. The results suggest that avoidance of phagocytosis may be an important virulence factor of T. brucei subsp. brucei and contributes to the variation observed in species and strain susceptible to trypanosomiasis.     

Glucose-6-phosphate dehydrogenase activity of the malaria parasite Plasmodium falciparum

Schizonts of Plasmodium falciparum, grown either in normal or glucose-6-phosphate dehydrogenase (G6PDH) deficient human red cells, contain an electrophoretically slow-moving form of G6PDH. The slow mobility of the G6PDH in non-dissociating polyacrylamide gels is due to its large size (Mr ca. 450,000) rather than to its charge. The activity of this enzyme was less than 10% of normal red cell G6PDH. These characteristics of the parasite-associated G6PDH were unaltered when parasites were grown in red cells from a G6PDH A+B+ heterozygote or following the introduction of a heterologous G6PDH into resealed ghosts. Differential absorption of the parasite-associated and red cell G6PDHs was demonstrated with antisera containing antibodies to red cell G6PDH. These studies show that a novel form of G6PDH is associated with P. falciparum in normal red cells without the requirements for induction by one or several cycles of multiplication in G6PDH deficient red cells.