J Mol Med (1998) 76:581–588 © Springer-Verlag 1998

REVIEW

&roles:Cyril Ruwende · Adrian Hill

Glucose-6-phosphate dehydrogenase deficiency and malaria

&misc:Received: 14 October 1996 / Accepted: 13 October 1997

&p.1:Abstract Glucose-6-phosphate dehydrogenase (G6PD) &kwd:Key words Glucose-6-phosphate dehydrogenase

is a cytoplasmic enzyme that is essential for a cell’s ca- deficiency · Severe malaria · Protection · Heterozygotes ·
pacity to withstand oxidant stress. G6PD deficiency is Hemizygotes
the commonest enzymopathy of humans, affecting over
400 million persons worldwide. The geographical corre- Abbreviations G6PD Glucose-6-phosphate
lation of its distribution with the historical endemicity of dehydrogenase&bdy:
malaria suggests that 66PD deficiency has risen in fre-
quency through natural selection by malaria. This is sup-
ported by data from in vitro studies that demonstrate im- G6PD
paired growth of P. falciparum parasites in G6PD-defi-
cient erythrocytes. Attempts to confirm that G6PD defi- Glucose-6-phosphate dehydrogenase G6PD is a cytoplas-
ciency is protective in field studies of malaria have yield- mic, so-called ‘house-keeping’ enzyme that catalyses the
ed conflicting results, but recent results from large case first and rate-limiting step of the hexose monophosphate
control studies conducted in East and West Africa pro- pathway for pentose phosphate synthesis. This pathway is
vide strong evidence that the most common African an important source of NADPH that is required for sever-
G6PD deficiency variant, G6PD A–, is associated with a al biosynthetic reactions and is essential for maintaining
significant reduction in the risk of severe malaria for adequate intracellular levels of reduced forms of glutathi-
both G6PD female heterozygotes and male hemizygotes. one and other sulphydryl groups. By preserving and re-
The effect of female homozygotes on severe malaria re- generating reduced forms of glutathione as well as pro-
mains unclear but can probably be assumed to be similar moting the stability of catalase, NADPH plays a major
to that of comparably deficient male hemizygotes. role in a cell’s ability to withstand oxidant stress. The
hexose monophosphate shunt catalysed by G6PD is also
CYRIL RUWENDE an important source of ribose which is essential for pro-
earned his doctorate from
Oxford University, UK. He is duction of nucleotide coenzymes, the replication of nu-
now a Resident in Medicine at cleic acids and, therefore, cell division [1]. The biological
Johns Hopkins University, and importance of G6PD is underlined by the fact that this en-
concentrates on genetic zyme has been detected in virtually every cell type in all
susceptibility to infection.
contemporary organisms [2].
ADRIAN HILL Along with the loci encoding colour blindness and fac-
also earned his medical degree tor VIII, the G6PD gene is located in the telomeric region
from Oxford University. His on the long arm (Xq28) of the X chromosome [3, 4]. The
interests relate to genetic
susceptibility to infectious G6PD gene consists of 13 exons and spans approximately
diseases and vaccine 18 kb [5]. The G6PD protein has a molecular weight of
development.&/fn-block: 59 kDa, and the active enzyme is composed variably of
two or four identical 515 amino acid subunits [6].
C. Ruwende (✉)
Department of Medicine, Johns Hopkins Medical Institutions,
1830 East Monument Street, Baltimore, MD 21205, USA Genetic polymorphism of G6PD
A. Hill
Wellcome Trust Centre for Human Genetics, The G6PD locus is probably the most polymorphic locus
University of Oxford, Windmill Road, Oxford OX 3 7BN, UK in humans, with over 300 allelic variants known, at least
582

87 of which have reached polymorphic (i.e. >1%) fre- methionine substitution [11]. Although the nucleotide
quencies [7, 8]. These variants have been characterised 202 substitution accounts for at least 95% of the G6PD
biochemically based primarily on their differing residual A– molecular variants in Africa, in a minority of individ-
enzyme activities, electrophoretic mobility patterns and uals two other alternative second mutation sites have
also on their physicochemical (thermostability, chro- been identified at nucleotides 680 and 968 [11, 12].
matographic behaviour) and kinetic (Km for glucose-6
phosphate or NADPH, pH dependence and utilisation of
substrate analogues) properties [7]. G6PD deficiency
The G6PD variants are grouped into the following
classes depending on the degree of enzyme deficiency Frequency and distribution
and associated clinical symptoms:
G6PD deficiency is the commonest enzymopathy in man
– Class I: severely deficient associated with chronic non-
affecting over 400 million persons worldwide [13]. This
spherocytic anaemia
disorder, which is caused by a multitude of the different
– Class II: severely deficient, <10% residual enzyme ac-
structural allelic mutants of the G6PD gene referred to
tivity
above, is found mainly in the tropical and sub-tropical
– Class III: moderately deficient, 10–60% enzyme activity
regions of the world, with the highest rates, usually
– Class IV: near normal or normal enzyme activity,
5–30%, being found in Africa, Asia, the Middle East, the
60–150% enzyme activity
Mediterranean and Papua New Guinea [7]. Worldwide
– Class V: enzyme activity, >150%
the frequency figures range from 62% in Kurdish Jews to
To date, comparison of gene sequences encoding enzyme 0.1% in Japan and northern Europe [14].
variants to that of the normal G6PD B gene has led to the
identification of at least 34 different mutations [7]. These
mutations are widely spread throughout the gene, being Clinical features
found in all exons except exon 3 and exon 13. All but
one of these are point mutations associated with amino Clinical expression of G6PD deficiency is probably de-
acid substitutions [2]. The exception is G6PD Sunder- pendent on an interaction of the molecular properties of
land, which is due to a 3-bp deletion and results in the a given deficiency variant, exogenous factors and, possi-
loss of an isoleucine residue [9]. Interestingly, a substan- bly, additional genetic factors [7]. In unstressed normal
tial number of these variants are associated with variable cells G6PD activity is only 2% of total capacity [14], and
forms of G6PD enzyme deficiency. The absence in the therefore it is hardly surprising that most individuals
G6PD gene of larger deletions, or other mutations such with the more common class II and III G6PD deficiency
as nonsense mutations or frameshift mutations that variants are usually asymptomatic. Although there is no
would completely abolish the function of the protein, direct evidence to support this, it is likely that there is a
suggests that complete absence of the G6PD enzyme is correlation between the degree of enzyme deficiency and
incompatible with life [2]. the propensity to develop clinical symptoms.
In Africa G6PD is essentially a tri-allelic polymor- The most striking clinical syndrome associated with
phism (Table 1). G6PD B, the normal variant associated G6PD deficiency, acute haemolytic anaemia, occurs as a
with normal or 100% enzyme activity, is the commonest manifestation of this disorder on the mature red blood
allele, with frequencies of 60–80%. G6PD A which has cell. On account of its long non-nucleated life-span and
90% of the activity of G6PD B is the next commonest al- hence its impaired ability to generate adequate levels of
lele with frequency between 15–40%. The third allele NADPH and reduced forms of glutathione, the mature
which is the common deficiency allele in Africa is G6PD erythrocyte has a diminished reductive capacity to re-
A–. It is a class III variant with 12% enzyme activity, and spond to oxidant stress. Uncompensated oxidant stress in
it varies in frequency from 0% to 25% [7]. the erythrocyte leads to oxidation of haemoglobin to
G6PD A– is unique in that it contains two mutations. methaemoglobin, heinz body formation and membrane
The first at nucleotide 376, which on its own gives rise to damage [15]. In the extreme this leads to haemolysis
G6PD A [10], is an adenine to guanine substitution that while less severe oxident stress increases the deformabil-
results in an asparagine to aspartate amino acid substitu- ity of the erythrocyte and probably enhances the likeli-
tion while the second mutation, usually a guanine to ade- hood that the stressed cell will be removed from circula-
nine substitution at nucleotide 202, leads to a valine to tion by the reticuloendothelial system [16, 17].
Acute haemolytic anaemia is therefore the most fre-
quent clinical manifestation of G6PD deficiency. The
Table 1 G6PD alleles in Africa and their enzyme activities&/tbl.c:&
haemolysis is precipitated most commonly by infections
Alleles Class Enzyme activity Frequency but can also occur after the ingestion of drugs and food-
stuffs that contain oxidant components or in certain met-
G6PD B IV 100% 0.60–0.80 abolic conditions such as diabetic ketoacidosis [18].
G6PD A IV 80% 0.15–0.40 Agents with oxidant properties such as primaquine,
G6PD A– III 12% 0.00–0.25
sulphonamides, nitrofurantoin and several anti-inflam-
 583

matory agents are the most common drugs associated graphical correlation between the distribution of these
with haemolysis [18]. Fava beans (Vicia faba) commonly polymorphic deficiency variants with areas with histori-
ingested in the Mediterranean are the most well docu- cal endemicity of P. falciparum malaria suggests that
mented causative dietary agent and are associated with a disorder has risen in frequency through natural selection
well characterised condition, favism. Favism is associat- by malaria. The geographical distribution of G6PD defi-
ed with the severely deficient class II G6PD Mediterra- ciency can not be attributable solely to gene flow. In-
nean form and not the moderately deficient G6PD A– deed, the presence of many diverse G6PD variants that
form that is common in Africa. Although all victims of have arisen independently and reached polymorphic fre-
favism are G6PD deficient, not all (only 25%) G6PD- quencies in geographically disparate areas [7] further
deficient individuals develop favism after consumption supports the occurrence of natural selection of this disor-
of the fava beans [1, 7], suggesting that they may be oth- der. This hypothesis is further supported by the results of
er genetic or environmental factors involved in the patho- micromapping studies within relatively restricted geo-
genesis of this condition. graphical areas such as Kenya [24], Papua New Guinea
For class I G6PD deficiency variants the formed en- [25], Greece [26] and Sardinia [27] that have demon-
zyme is functionally so poor that the red cell life-span is strated a similarly remarkable geographical correlation
shortened even in the absence of stress, and hence class I between altitude and the distribution of G6PD deficiency
variants are associated with a chronic non-spherocytic with the lower altitude (<1000 m) areas, known to have
hemolytic anaemia [7] with affected individuals typically more intense malaria transmission, being clearly associ-
having mild to moderate anaemia and splenomegaly. The ated with higher frequencies for G6PD deficiency.
disadvantage of the chronic anaemia probably outweighs
any survival advantage afforded by these class I muta-
tions, and not surprisingly most of these mutations arise In vitro evidence
sporadically and are not usually propagated in popula-
tions [7]. Interestingly, most of the mutations that give Reports from early field studies that P. falciparum and P.
rise to these class I variants are clustered near the car- vivax parasites preferentially invade younger red blood
boxyl terminus of the G6PD protein [13]. cells that have relatively higher G6PD activity [28, 29],
Another serious clinical effect of G6PD deficiency is as well as the observation that in the presence of normal
icterus neonatorum or neonatal jaundice, which in severe and deficient erythrocytes malaria parasites preferential-
cases can lead to permanent neurological damage or ly develop in the normal cells [30], led investigators to
death. Increased red blood cell destruction accounts for propose that G6PD-deficient erythrocytes confer protec-
some of the hyperbilirubinaemia observed in this syn- tion against malaria by inhibiting erythrocyte invasion or
drome, but it is likely that severe enzyme deficiency in intracellular development of the malaria parasite [31,
the hepatocyte may impair the catabolism of bilirubin 32]. Since then there have been several independent
and thus also contribute to the development of jaundice studies in the literature reporting impaired growth of P.
[7]. falciparum in G6PD-deficient erythrocytes [33, 34], al-
though in some studies this was only observed when cul-
tures were subjected to oxidative stress [32]. Further-
X-chromosome activation and G6PD deficiency more, there are data that indicate that in heterozygous fe-
males, who as a consequence of variable X-chromosome
The G6PD gene is on the X chromosome and hence one inactivation have different proportions of normal and de-
of the two G6PD alleles present in females is subject to ficient cells, the degree of parasite growth inhibition is
inactivation. Variable X-chromosome inactivation means proportional to the percentage of deficient cells present
that expression of G6PD deficiency differs markedly [35].
among female heterozygotes as their red blood cell pop- Although there is growth inhibition in G6PD-deficient
ulations are variable mosaics of deficient and normal erythrocytes, it is now clear that after a few growth cy-
cells [19]. This phenomenon affects all somatic cells in cles the parasite can overcome the inhibition [36], and it
the body such that G6PD phenotypes have been success- had been suggested that the parasite achieved this by
fully used in the past to determine the clonal origins of producing its own G6PD enzyme [37, 38]. An ingenious
certain tumours and embryonal tissues in such female mechanism (based on the premise that expression of par-
G6PD heterozygotes [19–22]. asite G6PD enzyme is determined by G6PD genotype of
the host erythrocyte) was put forward [38] as a possible
mechanism to account for the results of a previous study
The G6PD deficiency and malaria hypothesis that had indicated that G6PD deficiency protection
against malaria was the sole prerogative of female het-
Epidemiological evidence erozygotes [39]. Hence in uniformly deficient red blood
cells such as those found in deficient hemizygous males
Although several different hypotheses have been ad- or deficient hemizygous females the parasite’s own in-
vanced to explain why G6PD deficiency has been select- duced G6PD enzyme would compensate for the lack of
ed for in different populations [1, 23]. The striking geo- the host’s enzyme. However, in female heterozygotes,
584

who necessarily have mixed populations of deficient and The literature on previous G6PD deficiency/malaria
non-deficient erythrocytes, parasite adaptation would be field studies is full of conflicting reports [28–30, 39,
compromised, and thus the parasite growth and multipli- 43–47], summarised in Table 2. The most widely quoted
cation impaired by the parasites need to repeatedly field study was carried out by Luzzatto’s group in Nige-
switch on and off its own enzyme as it moved from defi- ria in the early 1970s [39]. The frequencies of the G6PD
cient to non-deficient host red blood cell. While confirm- A males and G6PD A–/B females were lower in those
ing the phenomenon of adaptation, subsequent studies children with malaria, suggesting that these genotypes
have found that the parasite G6PD levels do not appear are protective against the disease. Their observation that
to be affected by the host red cell genotype [40–42]. the mildly deficient G6PD A males and only G6PD A–/B
heterozygotes and not G6PDA–/A heterozygotes with
comparable G6PD activities, nor the more markedly de-
Field studies ficient hemizygous males and homozygous females, are
protected against malaria appears to indicate that the de-
It has long been recognised that definitive answers on the gree of enzyme deficiency per se is not involved the
malaria/G6PD hypothesis have to come from field stud- mechanism of protection against in G6PD deficiency.
ies. Such studies have needed to answer several impor- Unlike the sickle cell trait, whose protective effect in
tant questions. Firstly, is G6PD deficiency protective malaria has been relatively straightforward to prove in
against malaria, and if so, is it protective against uncom- vitro and field studies, the proposed protective role of
plicated mild malaria, severe malaria or both? Secondly, G6PD deficiency in malaria has proved difficult to veri-
if the disorder is protective, what is the extent of protec- fy. There are several reasons that may explain this. In
tion, and are all the different male and female deficiency contrast to sickle haemoglobin, G6PD follows a sex-
genotypes afforded similar protection? Answers to these linked rather than an autosomal inheritance pattern, and
questions would provide a basis for understanding of the males therefore have distinctly different genotypes from
mechanism of protection of this disorder against malaria females. In addition, there is considerable genetic hetero-
in addition to clarifying the evolutionary mechanisms re- geneity associated with G6PD deficiency, and in some
sponsible for the high prevalence of this genetic disorder populations more than one deficiency variant is present.
in most tropical and sub-tropical populations. More pertinently, it is known that fitness of the deficien-

Table 2 Summary of some field studies on G6PD deficiency and malaria&/tbl.c:&

Investigator n Comment Population

Studies supporting protective role of G6PD deficiency

Alison and Clyde [28] 532 Lower parasite rates and densities in deficient male East Africa
children; reduced levels similar to children with sickle
trait for both indices
Gilles et al. [43] 100 Reduced frequency of both deficient males and Nigeria
females in cases with very high parasite counts
(>100,000@mm3) compared to controls
Luzzatto et al. [30] – Greater rates of parasitisation of non-deficient Nigeria
erythrocytes in mixed erythrocyte populations of
female heterozygotes with acute malaria
Bienzle et al. [39] 702 Reduced frequency of female heterozygotes only in Nigeria
paediatric cases; reduced parasite densities in same
group
Butler [44] 277 Lower incidence of malaria in nonimmune deficient Vietnam
adult African-American males compared to their non-
deficient counterparts
Kar et al. [45] 708 Significantly reduced parasitisation rates in deficient India
males and females
Studies indicating no protective role of G6PD deficiency
Kruatrachue et al. [29] 203 Increased malaria in deficient males (1–3 years old) Thailand
compared to non-deficient males in same age group
Powell and Brewer [47] 16 No difference in parasitisation and parasitaemia levels USA
in experimentally infected deficient and non-deficient
nonimmune adult African-American male prison
inmates
Martin et al. [46] 68 No protection against cerebral malaria for any of the Nigeria
deficiency genotypes

&/tbl.:
 585
Fig. 1 Possible protective
mechanisms of G6PD deficien-
cy against severe malaria&ig.c:/f

cy phenotype is decreased significantly only under a lim- frequencies in children with the relatively rare condi-
ited number of specific circumstances and therefore on tion of complicated or severe malaria [49, 50] with those
average very little. Lastly, G6PD deficiency may interact of matched controls seems to offer a more sensitive
with other genetic factors and population specific non- measure of the effect of malaria resistance alleles [51,
malaria environmental factors, such as diet, to modify 52].
the net fitness of the carrier [48]. The largest field study on G6PD deficiency and ma-
In retrospect it would have been surprising if a signif- laria was carried out in two malaria endemic regions in
icant effect of G6PD deficiency had been clearly demon- East and West Africa and measured the frequencies of
strated in some of the major field studies carried out in the G6PD A–, G6PD A and G6PD B genotypes in over
the past. Invariably these clinical studies attempted to de- 2000 DNA samples collected from children under 10
fine the various genotypes from phenotypic measure- years [53]. This was the first and certainly the largest
ments of enzyme levels, electrophoretic mobility and cy- field study to use precise molecular techniques to un-
tochemical staining patterns [39, 43, 46]. This is difficult equivocally define G6PD genotypes in an investigation
because overlapping levels of enzyme activity are seen of the proposed protective effect of a G6PD deficiency
among genotypes, related partly to variable inactivation allele, such as G6PD A–, against malaria. There was no
of X chromosomes in female heterozygotes and partly to significant heterogeneity in odds ratios between the two
altered rates of erythrocyte turnover in acute malaria [7]. populations studied, and the results of both studies dem-
Furthermore, several of these studies were studying para- onstrated that the frequencies of both female heterozy-
site densities in cases of the more common clinical con- gotes and male hemizygotes are lower in the children
dition, mainly mild or uncomplicated malaria [39, 46]. with severe malaria than in controls. While female het-
Recently it has been found that comparison of genotype erozygotes were significantly protected against both se-
586

vere malaria (46%) and mild malaria (41%), male hemi- malaria resistance afforded to female heterozygotes is
zygotes were only significantly protected against severe little different from that of male hemizygotes, the overall
malaria (58%). Unfortunately, female G6PD A– homo- fitness of female heterozygotes is likely to be greater
zygotes were too rare to allow measurement of the sus- than that of hemizygotes and homozygotes. This situa-
ceptibility of this genotype to severe malaria. A possible tion constitutes a balanced polymorphism [57] and
protective effect of the G6PD A genotype, with 85% of perhaps explains the observed rarity of populations in
normal enzyme activity, was sought by comparing the which frequencies of G6PD deficiency are in excess of
frequencies of this genotype between the clinical groups 50% [7].
in which an effect was perhaps most likely to be detect- Most studies addressing the malaria G6PD question
able, i.e. males with severe malaria to control males. have involved the more common P. falciparum malaria,
This did not demonstrate any significant differences in and hence there is a paucity data on the role of G6PD de-
the frequency of this genotype between these groups. ficiency in non-falciparum malaria. One study in Nigeria
Heterozygosity for haemoglobin S was strongly protec- reported a lower than expected frequency of female het-
tive against both severe and mild malaria in both areas, erozygotes amongst 33 girls with P. malariae associated
but no clear evidence of interaction between G6PD and nephrotic syndrome [1]. Another study by Kar et al. [45]
haemoglobin S genotypes was observed, although the in northern India reported protection in female heterozy-
power of the study to detect this was low. gotes and male hemizygotes against both P. falciparum
The data from this African case-control study strong- and P. vivax malaria.
ly suggest that the G6PD A– allele is associated with There are still several issues that have not been ad-
substantial resistance to severe malaria in hemizygous dressed on the malaria G6PD hypothesis. Data on the
males as well as in heterozygous females. Its results con- role of enzyme-deficient female homozygotes as well as
cur with in vitro studies showing impaired growth of P. that of the numerous other known G6PD deficiency vari-
falciparum in enzyme-deficient erythrocytes [33] and ants on severe malaria are still needed. It will also be in-
suggest that the degree of enzyme deficiency is central to teresting to pursue these current data with in vitro work
the protective mechanism of G6PD deficiency against to assess the possible effects of G6PD deficiency on im-
malaria. There are several mechanisms which might ex- portant host-parasite phenomena such as sequestration,
plain this phenomenon at the molecular level (Fig. 1). rosetting and macrophage release of key cytokines such
The most likely attributes protection to reduced multipli- as tumor necrosis factor, interleukin-4, and interferon-γ
cation in deficient erythrocytes, probably as a result of in response to parasite infection. Furthermore elucidation
the intracellular accumulation of toxic oxidized sub- of the precise biochemical pathways with which oxida-
stances such as disulphide glutathione and haemozoin tive stress in deficient erythrocytes interferes with para-
[17, 34, 54]. However, there is some evidence that infect- site growth may provide additional impetus for efforts to
ed deficient erythrocytes are more susceptible to haemo- modify existing or design novel chemotherapeutic and
lysis as a result of increased methaemoglobin and release possibly immunotherapeutic agents that more effectively
of ferriheme, a known cytolytic agent [55], and that in target essential enzymes and cofactors in the parasite
addition, they are more readily phagocytosed by cells of growth process within the infected erythrocyte.
the reticuloendothelial system as a result of the erythro-
cyte changes (i.e. methaemoglobin and heinz body for-
mation, membrane damage) that are associated with the
References
oxidant stress imposed by parasite infection [17].
The degree of protection associated with G6PD defi- 1. Sodiende O (1992) Glucose-6-phosphate dehydrogenase defi-
ciency reported by Ruwende et al. [51, 56] is less than ciency. Ballieres Clin Hamatol 5:367–382
that afforded to carriers of sickle haemoglobin but equal 2. Vulliamy T, Mason P, Luzzatto L (1992) The molecular basis
of glucose-6-phosphate dehydrogenase deficiency. Trends
to or greater than that associated with some HLA vari- Genet 8:138–143
ants and thalassaemias. The African A– variant has a 3. Pai GS, Sprenkle JA, Do TT, Marena CE, Migeon BR (1980)
higher level of enzyme activity than the most prevalent Localisation of loci for hypoxanthine phosphoribosyltransfer-
Mediterranean and some Asian types of G6PD deficien- ase and glucose-6-phosphate dehydrogenase and biochemical
cy, and the associated disorder is milder [7]. Therefore if evidence of X chromosome expression from studies of a hu-
man X-autosome translocation. Proc Natl Acad Sci USA 77:
the degree of enzyme deficiency is important for the 2810–2813
mechanism of protection, it is likely that protection 4. Keats B (1983) Genetic mapping: X chromosome. Hum Genet
against malaria conferred by G6PD deficiency is at least 64:28
as great in many non-African populations with variants 5. Chen EY, Cheng A, Lee A, Kuang W-J, Hillier L, Green P,
Sclessinger D, Ciccodicola A, D’Urso M (1991) Sequence of
associated with greater G6PD enzyme deficiency. human glucose-6-phosphate dehydrogenase cloned in plasm-
The main clinical complications associated with ids and a yeast artificial chromosome. Genomics 10:792–
G6PD deficiency today are acute haemolysis and neona- 800
tal jaundice [7]. Individuals with more severe enzyme 6. Persico MG, Viglietto G, Martino G, Toniolo D, Paonessa G,
Moscatelli C, Dono R, Vulliamy TJ, Luzzatto L, D’Urso M
deficiency such as male hemizygotes and deficient fe- (1986) Isolation of human cDNA clones: Primary structure of
male homozygotes appear more susceptible to all of protein and unusual 5′ non-coding region. Nucleic Acids Res
these manifestations. Since the estimate of the degree of 14:2511
 587
7. Luzzatto L, Mehta A (1989) Human erythrocyte glucose-6- 29. Kruatrachue M, Charoenlarp P, Chongsophajaisiddhi T, Hari-
phosphate dehydrogenase deficiency. In: Scriver CR, Baudet nasuta C (1962) Erythrocyte glucose-6-phosphate dehydroge-
AL, Sly WS, Valle D (eds) The metabolic basis of inherited nase and malaria in Thailand. Lancet II:1183–1186
disease, 6th edn. McGraw-Hill, New York, pp 2237-2265 30. Luzzatto L, Usanga E, Reddy S (1969) Glucose-6-phosphate
8. Luzzatto L, Battistuzzi G (1985) Glucose-6-phosphate dehy- dehydrogenase deficient red cells: Resistance to infection by
drogenase. Adv Med Genet 14:217–329 malarial parasites. Science 164:839–841
9. MacDonald D, Town M, Mason P, Vulliamy TJ, Luzzatto L, 31. Luzzatto L (1979) Genetics of red cells and susceptibility to
Goff DK (1991) Deficiency in red blood cells. Nature 351: malaria. Blood 54:961–976
115 32. Friedman MJ (1979) Oxidant damage mediates variant red cell
10. Vulliamy T, D’Urso M, Battistuzzi G, Estrada M, Foulkes NS, resistance to malaria. Nature 280:245–247
Martini G, Calabro V, Poggi V, Giordano R, Town M, Luzzatto 33. Roth EF Jnr, Raventos-Suarez C, Rinaldi A, Nagel RL (1983)
L, Perisco MG (1988) Diverse point mutations in the human Glucose-6-phosphate dehydrogenase deficiency inhibits in vit-
glucose-6-phosphate dehydrogenase deficiency gene cause en- ro growth of Plasmodium falciparum malaria. Proc Natl Acad
zyme deficiency and mild or severe hemolytic anemia. Proc Sci USA. 80:298–299
Natl Acad Sci USA 85:5171–5175 34. Miller J, Golenser J, Kullgren B, Spira DT (1984) Plasmodium
11. Hirono A, Beutler E (1988) Molecular cloning and nucleotide falciparum: thiol status and growth in normal and deficient hu-
sequence of cDNA for human glucose-6-phosphate dehydro- man erythrocytes. Exp Parasitol 57:239–247
genase deficiency. Proc Natl Acad Sci USA 85:3951–3954 35. Roth EF Jnr, Raventos-Suarez C, Rinaldi A, Nagel RL (1983)
12. Beutler E, Kuhl W, Vives-Corrons JL, Prchal JT (1989) Mole- The effect of X chromosome inactivation on the inhibi-
cular heterogeneity of glucose-6-phosphate dehydrogenase de- tion of Plasmodium falciparum malaria growth. by glucose-6-
ficiency A–. Blood 74:2550–2555 phosphate dehydrogenase deficient red cells. Blood 62:866–
13. Beutler E (1991) Glucose-6-phosphate dehydrogenase defi- 868
ciency. N Engl J Med 324:169–174 36. Roth EF Jnr, Schulman S (1988) The adaptation of Plasmodi-
14. WHO Working Group (1989) Glucose-6-phosphate dehydro- um falciparum to oxidative stress in G6PD deficient human
genase deficiency. Bull WHO. 67:601–611 erythrocytes. Br J Haematol 70:363–367
15. Beutler E (1983) Glucose-6-phosphate dehydrogenase defi- 37. Hempelmann E, Wilson RJM (1981) Detection of glucose-6-
ciency. In: Stanbury JB, Wyngaarden J, Frerickson DS, Gold- phosphate dehydrogenase in malarial parasites. Mol Biochem
stein J (eds) The metabolic basis of inherited disease, 5th edn. Parasitol 2:197–204
McGraw-Hill, New York, pp 1629–1653 38. Usanga EA, Luzzatto L (1985) Adaptation of Plasmodium fal-
16. Arese P, Mannazzu I, Turrini F, Galiano S, Gaetani GF (1986) ciparum to G6PD-deficient host red cells by production of
Etiological aspects of favism. In: Yoshida A, Beutler E (eds) parasite-coded enzyme. Nature 313:793–795
Glucose-6-phosphate dehydrogenase. Academic, Orlando, pp 39. Bienzle U, Ayeni O, Lucas AO, Luzzatto L (1972) Glucose-6-
45–75 phosphate dehydrogenase and malaria. Greater resistance of
17. Turrini F, Schwarzer E, Arese P (1993) The involvement of female heterozygotes for enzyme deficiency and of males with
hemozoin toxicity in depression of cellular immunity. Parasitol non-deficiency variant. Lancet I:107–110
Today 9:297–300 40. Roth EF Jr, Schulman S (1988) The adaptation of Plasmodium
18. Greene L S (1993) G6PD Deficiency as protection against fal- falciparum to oxidative stress in G6PD deficient human eryth-
ciparum malaria: an epidemiological critique of population rocytes. Br J Haematol 70:363–367
and experimental studies. Yearbook Phys Anthropol 36:153– 41. Ling IT, Wilson RJM (1988) Glucose-6-phosphate dehydroge-
178 nase activity of the malaria parasite Plasmodium falciparum.
19. Beutler E, Yeh M, Fairbanks VF (1962) The normal human fe- Mol Biochem Parasitol 31:47–56
male as a mosaic of X chromosome activity. Studies using the 42. Kurdi-Haidar B, Luzzatto L (1990) Expression and character-
gene for G-6-PD deficiency as a marker. Proc Natl Acad Sci ization of glucose-6-phosphate dehydrogenase of Plasmodium
USA 48:9 falciparum. Mol Biochem Parasitol 41:83–92
20. Linder D, Gartler SM Glucose-6-phosphate dehydrogenase 43. Gilles HM, Fletcher KA, Hendrickse RG, Linder R, Reddy S,
mosaicism: utilization as cell marker in the study of leiomyo- Allan N (1967) Glucose-6-phosphate dehydrogenase deficien-
mas. Science 150:67–69 cy, sickling and malaria in African children in south western
21. Beutler E (1967) Value of genetic variants of glucose-6-phos- Nigeria. Lancet I:138–140
phate dehydrogenase in tracing the origins of malignant tu- 44. Butler T (1973) G6PD deficiency and malaria in black Ameri-
mours. N Engl J Med 276:389–39 cans in Vietnam. Military Med 138:153–155
22. Fialkow PJ (1976) Clonal origin of human tumours. Biochim 45. Kar S, Seth S, Seth PK (1992) Prevalence of malaria in Ao
Biophys 458:283–321 Nagas and its association with G6PD and HbE. Human Biolo-
23. Lisa A, Astolfi P, Degioanni A, De Pasquale C, Zei G (1994) gy 64:187–197
Differential fertility as a mechanism maintaining balanced 46. Martin SK, Miller LH, Alling D, Okoye VC, Esan GJF, Os-
polymorphism in Sardinia. Hum Biol 66:683–698 unkoya BO, Deane M (1979) Severe malaria and glucose-6-
24. Alison AC (1960) Glucose-6-phosphate dehydrogenase defi- phosphate dehydrogenase deficiency: reappraisal of the malar-
ciency in red blood cells of East Africans. Nature 186:531– ia G-6-PD hypothesis. Lancet I:524–526
532 47. Powell RD, Brewster GJ (1965) Glucose-6-phosphate dehy-
25. Yenchitsomanus P, Sumers KM, Board PG, Bhatia KK, Jones drogenase deficiency and falciparum malaria. Am J Trop Med
GL, Johnson K, Nurse GT (1986) Alpha thalassaemia in Papua Hyg 14:358–362
New Guinea. Hum Genet 74:432 48. Martin SK (1994) The malaria/G6PD hypothesis revisited.
26. Stamatoyannopoulos G, Panayotopoulos A, Motulsky AG Parasitol Today 10:251–252
(1966) The distribution of glucose-6-phosphate dehydrogenase 49. Greenwood BM, Marsh K, Snow RW (1991) Why do some
deficiency in Greece. Am J Hum Genet 18:296 African children develop severe malaria. Parasitol Today 7:
27. Siniscalco M, Bernini L, Filippi G, Latte B, Meera Khan P, 277–281
Piomelli S, Rattazzi M (1966) Population genetics of haemo- 50. Marsh K (1992) Severe malaria in African children. Parasitol-
globin variants, thalassaemia and glucose-6-phosphate dehy- ogy 104: S53-S59
drogenase deficiency, with particular reference to the malaria 51. Hill AVS, Allsopp CEM, Kwiatkowski D, Antsey NM, Tum-
hypothesis. Bull WHO 34:379–393 wasi P, Rowe PA, Bennet S, Brewster D, McMichael AJ,
28. Alison AC, Clyde DF (1961) Malaria in African children with Greenwood BM (1991) Common West African HLA Antigens
deficient erythrocyte glucose-6-phosphate dehydrogenase. are associated with protection from severe malaria. Nature
BMJ 1:1346–1349 352:495–500
588
52. McGuire W, Hill AVS, Allsopp CEM, Greenwood BM, of Plasmodium falciparum in glucose-6-phosphate dehydroge-
Kwiatkowski D (1994) Variation in the TNF-α promoter re- nase deficient erythrocytes is enhanced by oxidants and crisis
gion associated with susceptibility to cerebral malaria. Nature form factor. Trop Med Parasitol 39:272–276
371:508–511 55. Janney SK, Joist JH, Fitch CD (1986) Excess release of ferri-
53. Ruwende C, Khoo SC, Snow RW, Yates SNR, Kwiatkowski D, heme in G6PD-deficient erythrocytes: possible cause of hae-
Gupta A, Warn P, Allsopp CEM, Gilbert SC, Peschu N, New- molysis and resistance to malaria. Blood 67:331–333
bold CI, Greenwood BM, Marsh K, Hill AVS (1995) Natural 56. Hill AVS (1992) Malaria resistance genes: a natural selection.
protection of hemi- and heterozygotes for G6PD deficiency in Trans R Soc Trop Med Hyg 86:225–226, 232
Africa by resistance to severe malaria. Nature 376:246–249 57. Motulsky AG (1960) Metabolic polymorphisms and the role of
54. Golenser J, Miller J, Spira DT, Kosower NS, Vande waa JA, infectious diseases in human evolution. Hum Biol 32:28–62
Jensen JB (1988) Inhibition of intraerythrocytic development