Introduction

Chapter 1

Introduction

1.1 International initiatives to explore the human genome

1.2 Infectious disease and evolution

1.3 Malaria as a selectionforce

1.3.1 Hemoglobinopathies and signatures of balancing

selection

1.3.2 Glucose 6-Phosphate dehydrogenase deficiency and

malaria resistance

1.3.3 Natural selection at the Duffy blood group locus

1.4 Host response to P.falciparum malaria

1.4.1 Host immune responses to parasite attack

1.4.2 Cytoadherance

1.5 Host genetic factors and P.falciparum malaria

1.5.1 Genes encoding adhesion molecules

1.5.2 Genes encoding immune regulatory molecules

1.5.3 Genes involved in the process of invasion and rosetting

1.5.4 Linkage and Genome wide association studies and

malaria

1.6 The diverse Indian population: a unique resource for

complex disease analysis

1.6.1 The Indian Genome Variation Consortium (IGVC)

1.7 Rationale

1. Introduction Ever since the rediscovery of Mendel's laws of heredity in the beginning of the

twentieth century, the quest to unfold the innumerable mysteries contained in the

human genome has come a long way. Scientific investigations catalyzed by

continuing technological advancement have enhanced our knowledge of the nature

and content of genetic information of the human genome. The human genome

contains information about human physiology, development, diseases as well as our

evolutionary past and continuing change. Inherited differences in DNA sequences

contribute to phenotypic variations, influencing a population's anthropological

characteristics, risk to diseases and response to environment and drugs. DNA

sequence variations are central for understanding population substructure, natural

selection as a result of disease pressure and genetic drift due to population

migration. Now, when the whole DNA sequence of humans is known to us, many

types of variations like STRs (Short Tandem Repeats or microsatellites), VNTRs

(Variable Number of Tandem Repeats), deletion/insertion mutations etc. have been

revealed. The most common type of genetic variation in the human genome is the

Single Nucleotide Polymorphism (SNP) which accounts for more that 80% of all

variation in the human genome. A SNP is a DNA point variation at a single base pair

position with frequency of more than 1% in a popUlation. It has been estimated that

the human DNA sequence carries a SNP every 1,000-2,000 nucleotides (Li et al., 1991). Over the years SNPs have proved to be extremely informative in disease gene

mapping and tracing evolutionary histories of populations. As SNPs are present in

sufficient density on the human genome, they also provide a tool for comprehensive

haplotype analyses. With high resolution SNP maps for some world populations now

available in public databases, a wide corridor has been opened for gaining insights

into the genetic basis of complex diseases and to answer basic questions regarding

human evolution.

1.1 International initiatives to explore the human genome

The most ambitious and challenging task ever undertaken for exploration of the

human genome is the one which deciphered the entire DNA sequence compiled in

23 pairs of chromosomes. The International Human Genome Project (HGP) was

formally initiated in 1990 as a multi-institutional international collaborative

program which aimed to sequence the 3 billion base pairs that make up the human

genome. The first draft sequence of human DNA was presented in February 2001

(HGP, 2001; Venter et ai., 2001). The results gave many fundamental i~sights

regarding our genome. It was estimated that there are about 25,000 genes encoding

proteins (previous estimation was much higher) and that only 1-1.5% of the genome

encodes proteins. The data also revealed new information about repetitive elements,

domain sharing and conservation, and gene duplication in the human genome. The

information provided by the HGP is organized and maintained in web portals like,

NCBI, Genbank, ENSEMBL, etc.

Deciphering of the human genome sequence has revolutionized our understanding

of the genetic mechanisms of complex diseases. Various approaches to identify the

causative variants for common and complex disorders have been carried out. Most

of the disease association approaches are focused on candidate gene variations and

their regions of linkage to a particular disorder by comparing a set of infected to

uninfected individuals. This 'direct' approach has already led to the discovery of

genetic risk factors for common disorders, like type-I diabetes, asthma, Alzheimer's

disease etc., but the applicability of such analysis is narrow and confined to a small

genomic region. Since a particular disease is a consequence of malfunctioning of

many genes, genome-wide variation screening can be more informative and

comprehensive. Genome-wide association studies can be made lucid if we have prior

knowledge of the set of sequence variants in the genome which could serve as

genetic markers to detect association between a particular genomic region and a

disease, immaterial of the functionality of the markers. The search for causative

variants can then be zeroed down to the genomic regions showing association with

the disease. The 'International HapMap Project' was initiated in late 2002 to

create a public, genome-wide database of common sequence variation in DNA

samples from popUlations with varying ancestry from parts of Africa (YRI: Yoruba),

Asia (JPT: Japanese and CHB: Han-Chinese) and Europe (CEU)(International

HapMap project, 2003). The phase I data release of the HapMap consortium in 2005

presented the database of more than a million SNPs across human genome

genotyped with 269 DNA samples. The data documented the distribution patterns of

SNPs, recombination hotspots, linkage disequilibrium and haplotype diversity

2

across chromosomes in the selected populations (International HapMap project,

2005; 2007). This information on sequence variation on the human genome can

serve as a guide for designing candidate gene, linkage-based and genome-wide

disease association studies.

Although the Human Genome Project gave a highly accurate sequence of nucleotide

bases across all chromosomes, we still have incomplete information about genomic

regions that encode transcripts for proteins and various RNAs as well as genomic

elements that regulate gene expression. To completely understand the human

genome and how genetic information is orchestrated into various biological

processes it is important to know the exact function of each genomic region. The

National Human Genome Research Institute (NHGRI) launched a public research

consortium named ENCODE, the Encyclopedia Of DNA Elements, in September

2003, to carry out a project to identify all functional elements in the human genome

sequence (ENCODE, 2004). The main aim of the ENCODE project is to present a

comprehensive catalog of all the structural and functional elements of the human

genome which is critical for understanding molecular mechanisms and information

flow through the 'central dogma'. In the recent release of data of the pilot phase, the

consortium has presented a detailed account of roughly 30Mb-nearly 1% of the

human genome (ENCODE project, 2007). Initial investigations have hinted that the

human genome is pervasively transcribed, such that the majority of its bases can be

found in primary transcripts, including non-protein-coding transcripts, and those

that extensively overlap one another. The results from the ENCODE project, when

complete, will provide the functional landscape of the human genome which will

have enormous applicability. The information can be channeled into future

researches dedicated to understand the intricate biological pathways operating in a

cell. The ENCODE data may also play a pivotal role in biomedical research for

understanding the biology of human health and disease.

1.21nfectious disease and evolution

Infectious diseases, as suggested by AE Garrod in his book "Inborn factors in

disease" (1931), have been a major selective force in human evolution and in

shaping our overall genetic makeup. More than fifty years ago, J.B.S. Haldane

3

published a speculative review which is now often cited as having inspired new

thinking on disease and evolution. The article states common knowledge of

infectious diseases and their possible potency as agents of natural selection and

also as accelerators of speciation (Haldane, 1949). Haldane's inferences largely came

from his observations in certain Mediterranean populations where there is an

extremely high gene frequency of thalassemia which he suggested might have come

under intense selection because it conferred a heterozygote advantage against

malaria.

From time to time modifications in the human genome may be selected so as to

combat pathogen attack. There are many examples in evolutionary history where

naturally occurring genetic defense mechanisms have evolved for resisting

infectious diseases. A number of red cell defects like Sickle cell trait, thalassemia,

duffy antigen variation and G6PD deficiency, which also confer resistance to the

malaria parasite, are classical examples of genetic footprints selected by the disease

pressure of malaria, one of the strongest known force of evolutionary selection on

human genome. Another finding is the introduction of a 32-base deletion in the

CCR-5 gene (gene for a chemokine receptor) which protects the individual from HIV

disease progression (Sampson et al., 1996; O'Brien et aI., 1997). Strong HLA-DR

associations have been found in cases of leishmaniasis (Cabrera et al., 1995),

onchocerciasis (Meyer et al., 1994), and filariasis (Yazdanbaksh et al., 1995). It has

been suggested, that the very high frequency of the common cystic fibrosis allele in

Northern Europeans might reflect selection against a major infectious disease,

possibly one of the diarrheal illnesses that swept across Europe in the past.

Similarly, genetic and epidemiological studies have provided some evidence that the

high frequency of Tay-Sachs disease in some Jewish populations reflects

heterozygote resistance to tuberculosis in some of the parts of Eastern Europe

(Weatherall, 1996).

While studying the co-evolution of species, it is important not only to consider the

genetic makeup of the host but also that of the pathogen. When any infectious

agent attacks the human host, it immediately encounters its defense machinery and

microenvironments. In a scenario where infectious agents have many evolutionary

4

advantages due to their short generation time, high fecundity and mutation rates,

the host immune mechanism needs to employ effective strategies to fight the

invading pathogen. The pathogen is under strong selection to constantly devise new

strategies for gaining access to the host while evading its defense system which is

often modulated and modified by the administration of antibiotics. Given the

evolutionary advantages of the pathogens and their ability to develop drug

resistance it may be speculated that pathogens may tend towards higher a rate of

virulence within the human host while ensuring better transmission from host to

host. This idea comes from an assumption that the high level of immunity in

vaccine-acquired host population can maintain more virulent forms of pathogens

which have evolved by immune-selection, than can naIve host populations. This

'battle of host and pathogen genomes' will continue where both species evolve

continuously to gain advantage over their rival and continue to coexist without a

clear winner.

Owing to the large impact that infectious diseases impose on human evolution and

population differentiation, characterization of nucleotide variability in genes that

play a role in resistance or susceptibility to infectious disease will be important for

understanding how selection shapes patterns of variability and linkage

disequilibrium (LD) in the human genome. Although most of the disease association

studies are aimed at finding a correlation between a DNA variant and the disease in

infected individuals, it is also necessary to examine patterns of genetic variation in

uninfected individuals to determine the impact that selection has on the population

as a whole. With the advent of dense maps of human genetic variations, it is now

possible to detect signatures of selection as a result of disease pressure and carry

out disease association studies at the genome-wide level. For better understanding

of co-evolution of the human host and pathogen, studies should be conducted

under a comparative paradigm. The availability of the complete genome sequence of

many pathogens has greatly accelerated our investigations to find footprints of

selection of genetic variants and the extent to which both host and the pathogen

have modulated and modified each others' genetic makeup throughout evolutionary

history.

1.3 Malaria as a selection force Malaria is the most common parasitic disease of the tropics caused by the sporozoa

of the genus Plasmodium, four species of which infect humans: P.falciparum, P.vivax, P.ovale, and P.malanae. Among these, P.falciparum causes the most severe

form of malaria which is the major cause of mortality and morbidity worldwide.

Every year millions of people suffer from the disease, among which Africa accounts

for the maximum number of deaths. During the course of evolution, malaria has

exerted a profound impact on the human genome. Below is a brief account of

certain genetic variants as examples of signatures of selection on the human

genome due to malaria disease pressure.

1.3.1 Hemoglobinopathies and signatures of balancing selection

The erythrocytic stage is the central and most important stage in the life cycle of the

Plasmodium parasite in the human host. The human genome, over the years, has

modified the genes related to RBC structure and function so as to confer resistance

to malaria. The hemoglobin structural defect, the 'sickle cell trait' was first studied

in relation to selection by malaria. Although over a hundred abnormal hemoglobins

have been identified, only those coding for HbS, HbC and HbE have reached

polymorphic frequencies of more than 10%. The hemoglobinopathies can be

classified in two categories; one class is of structural variants, which involve defects

in both Q and 13 chains and includes HbS, C and E. The other category, which

includes Q- and J3-thalassemias, are the variants which result due to reduced

production of normal forms of either Q or J3-components of hemoglobin. These red

cell defects have reached a state of balanced polymorphism in many parts of the

world which have a long history of malaria prevalence.

The sickle cell disease, which is caused by a single base substitution in the 13-globin chain (136: Glu-Nal) reSUlting in an unstable HbS variant, is widely

distributed throughout Africa, parts of Mediterranean, Middle East and central

India. Carriers of the HbS trait, the heterozygotes, are apparently healthy and are

protected against malaria. Thus, this hemoglobin defect is maintained in high

frequency in the form of balanced polymorphism in many malaria-endemic

populations (Allison, 1954). Various clinical studies and epidemiological data have

r ---shown that heterozygotes for HbS are about 90% protected against severe form of

malaria including cerebral malaria (Williams et ai., 200S) with 60% protection

against mortality observed in infants (Aidoo et ai., 2002). In India, the HbS is mostly

concentrated in Central India, in the states of Maharashtra, Andhra Pradesh,

Orissa and parts of Gujarat and Rajasthan (Sarnaik, 200S) with an average

frequency of 4.3% (Mohanty et ai., 2003). Orissa being hyperendemic for malaria,

alone shows -30% of HbS carrier frequency distribution among tribal and various

caste groups (Balgir, 200S). Although there is ample evidence supporting the

protective effect of HbS allele, the molecular mechanism by which it is mediated is

still unclear. Luzzato et ai., (1970) have suggested that sickling may provide a

mechanism for rapid clearance of infected RBCs by the spleen. Some in vitro studies

have also shown for both homozygote and heterozygotes of HbS that under low

oxygen levels both invasion and parasite growth are inhibited (Freidman et al., 1978; Paslov et ai., 1978). Carlson et al., (1994) have proposed that protection in

HbS carriers may be due to reduced rosetting of parasitized and unparasitized

RBCs. A recent study by Cholera et ai., (2008) has shown that binding of parasitized

AS erythrocytes (heterozygotes of HbS) to endothelial cells is significantly reduced

relative to the binding of parasitized normal erythrocytes and this reduced binding

correlates with the altered display of PfEMP-1 on RBC membrane. Another

hemoglobin variant that is implicated in malaria protection is HbC which is very

common in West Africa where it is shown to confer resistance to malaria in both

heterozygote and homozygote state (Roberts et al., 2004). A study conducted in

Mossi, Burkina Faso, showed upto 30% reduction in clinical malaria in HbSC

individuals and as much as 90% in homozygotes (Modiano et aI., 2001). Unlike HbS

and HbC, the results regarding the protective effect of HbE (~26: Glu-Lys) are less

conclusive. This may be because till now there is no convincing case-control data

and also HbE occurs generally in conjunction with the mild form of ~-thalassemia

which can be a confounding factor in case-control studies. Still, the HbE allele is

very common in malaria endemic regions of south-east Asia (Flatz et al., 1965).

Further support for the protective effect of HbE comes from the observation in Thai

patients where carriers of the HbE variant were protected from heavy parasite

burden and hence from the development of severe clinical malaria (Kesinee et al., 2002). In India, HbE is widely distributed in malaria-endemic regions of West

Bengal and the north-eastern states with a carrier rate of 23% in Assam (Deka et

al., 1988; Saha et al., 1990). Another genetic analysis with extended haplotypes

around HbE locus in south-east Asia has shown that the present frequency of the

HbE variant has been reached in less than 5000 years, a rate that can be achieved

only under some kind of selection pressure (Ohashi et al., 2004). Haplotype analysis

surrounding the Hb gene locus reveals that these mutations have arisen to their

present polymorphic status recently and independently in different populations

providing further evidence for selective pressure (Flint, 1993). There is evidence that

the HbS mutation appears to have occurred twice, once in Africa and once in India

or the Middle-East (Chebloune et al., 1988). It may be possible that although these

hemoglobin mutations have come under intense selection pressure due to malaria,

part of their present distribution has resulted from population migration, founder

effects and genetic drift.

Several lines of evidence suggest the protective role played by a- and (3-thalassemias

in malaria. The distribution of both a- and (3-thalassemias across the world

coincides with malaria prevalence (Livingstone, 1983). The a- and (3-thalassemias

are a consequence of deletions or point mutations in the non-coding portion of the

globin genes and cause inadequate synthesis of the a- and (3-globin chains

(Weatherall, 2001). Interestingly, each population with high incidence of

thalassemia carries a different set of mutations suggesting that this is not because

of population migration and the variation may have arisen locally and recently

subsequently expanding to high frequencies by selection. Many epidemiological

studies have demonstrated the direct relationship between thalassemia and malaria

incidence. By comparing previous malaria surveys with the known prevalence of (3-

thalassemia in Papua New Guinea (PNG), Hill et al., (1988) found that its frequency

co-varied with that of malaria. Another direct evidence for the protective effect of (3-

thalassemia against severe malaria comes from a study on Liberian children where

the carriers of (3-thalassemia showed nearly 50% reduction in risk of malaria

(Willcox et al., 1983). a-thalassemia is caused by deletion of one or more a globin

genes resulting in -a/aa or -a/-a genotype (a+ thalassemia). Studies in PNG and

Melanesia have shown that the a+-thalassemia has a genotype frequency

proportional to malaria incidence and as the disease becomes less prevalent with

increasing altitude in these regions, so does a+-thalassemia (Flint et al., 1986).

1.3.2 Glucose 6-Phosphate dehydrogenase deficiency and malaria resistance

Glucose 6-phosphate dehydrogenase (G6PD) is an important housekeeping enzyme

which catalyzes the first step of the pentose phosphate pathway which is the sole

source for the production of reducing capacity in the form of NADPH in

erythrocytes. NADPH generated by this pathway is used to reduce glutathione

which is used by the cells to neutralize free radicals produced during oxidative

stress. G6PD deficiency is the most common enzymopathy reported worldwide. At

the DNA level, a number of genetic variants have been discovered in many world

popUlations that may cause G6PD deficiency or reduced enzyme activity (Luzzato et al., 2001). Although G6PD deficiency results in a number of clinical disorders like

neonatal jaundice, hemolytic anemia and cardiovascular disorders (Beutler, 1994),

it is selectively maintained in some popUlations across the globe. Erythrocytes

deficient of G6PD cannot produce NADPH and reduced glutathione thus impairing

the cell's ability to combat oxidative damage caused by free radicals ultimately

resulting in hemolysis. The P.falciparum parasite, during its erythrocytic cycle,

produces free radicals as a result of metabolism which may cause hemolysis in the

absence of G6PD and hence the death of parasite. It may be due to this reason that

the G6PD deficiency is maintained as a balanced polymorphism in malaria hyper­

endemic regions worldwide. The G6PD gene is located on the telomeric region of the

long arm on the X-chromosome (Xq28). At the DNA level, more than a hundred

polymorphisms have been reported in the G6PD gene worldwide (Beutler, 1994) but

there are two very common variants which have been functionally characterized for

severely low enzyme activity. A single nucleotide change at two positions (202 and

376) in exon4 of the G6PD gene (the A- allele) results in only 12% enzyme activity

and is very common in sub-Saharan Africa (frequency 25%) but is present in

extremely low frequency in many other world popUlations (Beutler, 1994). The other

mutation causing low enzyme activity is the G6PD Med deficiency which results due

to a SNP in exon6. The Med deficiency is common in the Mediterranean (2-25%) and

in Kurdish Jews (70%). Ruwende et al., (1995) have showed that the A- deficiency

can reduce the risk of malarial infection by 46-55% in both heterozygous females

and hemizygous males. Another study by Tishkoff et al., (2002) has demonstrated

that the A- variant likely arose 3,840-11,760 years before present, consistent with

the estimated time for the spread of severe malaria (Livingstone, 1971). G6PD

deficiency has been widely studied in many populations across the world and in

general there is a correlation between its geographical distribution and history of

malaria prevalence. Although the exact molecular basis of G6PD deficiency is still

unclear, the A- allele seems to be the likely target as it is present in very high

frequency in most of Africa (Livingstone, 1985). In a study conducted on a set of

individuals from different ethnicity, it was found that the level of nucleotide

diversity at the G6PD locus fits the neutral model of molecular evolution but

strikingly there was little or no variation within A- and other deficiency alleles in

individuals of African descent (Saunders et al., 2002). The signature of selection at

the G6PD locus is also seen by the presence of unusually high levels of linkage

disequilibrium extending to a long genomic distance across the G6PD gene (Tishkoff

et al., 2001; Saunders et al., 2002; Sabeti et al., 2002).

The G6PD deficiency information on various populations of India also shows

distinct patterns of distribution (Sukumar et al., 2002; 2004; Balgir, 2006). A

number of known and novel polymorphisms have been reported in Indian

populations. Among them the Med mutation is the commonest, while the A­

mutation is almost absent (Sukumar et al., 2002). A novel SNP, the 'G6PD-Orissa'

(Kaeda et al., 1995) was discovered in some tribal populations of Orissa where

malaria is hyper-endemic. Strikingly, the Med mutation, which is present in

considerable frequency in other parts of India, was found to be completely absent in

Orissa. Taken together, the G6PD deficiency data on India shows that the deficiency

trait is favored as the function of malarial prevalence within the population (Balgir,

2006) but the exact molecular basis of this is still unclear. More exhaustive studies

on the spread and distribution of individual mutations and LD patterns across the

G6PD locus may explain the molecular evolution of G6PD deficiency trait driven by

disease pressure.

1.3.3 Natural selection at the Duffy blood group locus

The Duffy blood group locus is located on chromosome 1 (1q22-23) and codes for

two co-dominant alleles FYA and FYB which cause a single amino acid change. The

product of the duffy gene, the Duffy blood group antigen, is expressed on the

erythrocyte membrane and serves as a receptor for proinflammatory chemokines.

Although the Duffy antigen might playa scavenging role to eliminate excess of toxic

chemokines produced during pathological states (Middleton et al., 1997) its main

role has been described as a receptor of the P.vivax parasite (Miller et al., 1975). A

single base pair mutation at -46 position in the FYB gene, which impairs its

promoter activity in RBCs by disrupting the GATA1 biding site, results in the FyB­

phenotype (Tournamille et al., 1995). The FyB- phenotype or the FYO allele is

incapable of binding the P. vivax parasite. The Duffy blood group locus (FY) shows

extreme patterns of geographical distribution and hence a likely target for

directional selection. The FYO allele is fixed in sub-Saharan Africa where it shows

highest FST values when compared with other neutral markers around its locus

(Hamblin et al., 2000). The absence of FY antigen might be the reason of complete

resistance to P. vivax infection in Africans (Livingstone et al., 1984). The positive

selection of the FYO allele in Africa is indicated by the finding that it is almost fixed

in most populations of Africa while it is nearly absent in popUlations outside Africa.

While G6PD deficiency is maintained as a balanced polymorphism in the

popUlation, the absence of Fy antigen is mostly present in the homozygous form as

this confers complete resistance to P. vivax infection. If positive selection due to

disease pressure, and not genetic drift, is alone responsible for the high FST values

at the FY locus, then the FST should diminish with distance from this locus. It was

shown by Hamblin et al., (2002) that in African populations, the maximum FST

values were at the Fy loci than those observed at neutral loci nearby, when

compared with Italian and Chinese samples which fit into the neutral model of

natural selection. Also, the level of genetic variation across both of the FYA and FYB

allele is surprisingly low in sub-Saharan Africa when compared to the Italian

population samples (Hamblin et al., 2000). The low degree of variations coupled

with high FST at the FY locus offers the best example of 'selective sweep' as a result

of P. vivax disease pressure making the African populations resistant to vivax

malaria. As most of the studies regarding FYO selection have concentrated on

Africa, little data is available on the distribution patterns of Fy alleles in other

P. vivax-endemic populations of the world. A recent report by Cavasini et al., (2007)

has shown that a low genotype frequency of the FYO is exhibited in P. vivax-endemic

areas of the Brazilian Amazon region. A study done on aboriginal tribes of Andaman

and Nicobar islands, India, showed very high frequency of the FYO genotype and

almost complete absence of P.vivax infection in that area (Das et al., 2005). Similar

findings have been reported in malaria endemic regions of PNG, where the

emergence of FYO allele suggests that the infection load by P. vivax is involved in the

selection of this erythrocyte polymorphism (Zimmerman et al., 1999). Where the

FYO allele is rare in popUlations outside Africa, the FYA allele shows the near­

fixation frequency in eastern Asia and the Pacific (Cavalli-Sforza et al., 1994). Unlike

FYO allele, the FYA allele has not been associated with any phenotype. A more

comprehensive study on the complex geographical distribution patterns of FY alleles

across globe along with the information on popUlation disease history and selective

forces acting on it may provide a detailed account of the directional selection of

these alleles in various world popUlations.

1.4 Host response to P./alciparum malaria

P.falciparum infection triggers host responses which are regulated by both innate

and adaptive arms of the immune system as well as specific interactions between

parasite and host molecules. In order to understand disease pathogenicity, it is

important to study host-parasite interactions and mechanisms involved in the

parasite infection process. Below is a brief account of the various host responses to

P.falciparum infection and the recent progress made in understanding the complex

nature of host-parasite interactions.

1.4.1 Host immune responses to parasite attack The innate immune response Innate immune mechanisms form the first line of defense of the host against any

antigenic attack. Majorly non-specific in nature, innate immunity refers to the basic

resistance to disease that a species possesses. The non-specific immune

mechanisms of the human host against P.falciparum are not clearly detailed but the

presence of low parasitemia following acute P.falciparum infection (Druille and

12

Perignon, 1994) suggests that innate immune responses check the blood stage

parasite growth. The main effectors for the development of this type of immunity are

natural killer (NK) cells, neutrophils and macrophages, among which NK cells are

the key players in the early innate immune response. NK cells are granulated

lymphocytes widespread throughout the body and are present in both lymphoid

organs and non-lymphoid peripheral tissues. NK cells have been implicated in viral

immunity and in defense against tumors. There is increasing evidence on the role of

NK cells as the chief effector cells during early onset of P.faiciparum malaria. It is

not clear how NK cells recognize parasitized erythrocytes (PEs) and to which

parasite ligand they bind. A recent report says that ICAM-1, a host surface receptor,

mediates the interaction between NK cells and PEs, however this interaction is

independent of the parasite ligand PrEMP1 (Baratin et ai., 2007). Following the

contact with PEs, the activated NK cells are the first to produce IFNy which is the

major cytokine to mediate cytotoxicity during the early phase of infection (Artavanis

et ai., 2002). Along with IFNy, NK cells also produce a number of other pro­

inflammatory cytokines and chemokines like TNF, IL-12, IL-18, CCL3, CCL4, CCL5

(RANTES) as elevated levels of these cytokines are detected in clinical malaria

samples (Kwaitkowski D, 1995; Lyke et ai., 2004; John et ai., 2006). It is now

believed that the NK cell mediated target cell destruction is influenced by a fine­

tuned interplay of this cytokine network. After the NK cells adhere to the PEs,

degranulation of perforin-containing granules occurs and the released perforin

damages the target cell. A number of host receptors have been recognized that are

involved in innate immune response to malarial parasite. Among them, several Toll

like receptors (TLRs), a central component of innate immunity, have been shown to

recognize a number of parasite ligand (Ishii et ai., 2005). TLR9 binds to hemozoin, a

heme polymer formed after the digestion of hemoglobin by the parasite (Coban et

ai., 2005; Parroche et al., 2006). Another parasite ligand, the glycosyl­

phosphatidylinositol (GPI) anchor, activates host cells primarily through TLR2

(Campos et ai., 2001). TLRs, after binding to parasitic ligands, initiate the innate

immune response by triggering the release of various cytokines. TLR4 stimulates

the production of certain pro-inflammatory cytokines whereas TLR1jTLR2

upregulates both pro- and anti-inflammatory cytokine release (McCall et ai., 2007). Besides TLR and NK cells mediated immune responses, a number of serum proteins

have been identified that play important role in innate defense mechanisms. The

mannose binding lectin (MBL2) acts as a pattern recognition receptor and binds

certain glycosylated parasite derived ligands (Klabunde et al., 2002). It is believed to

act as an opsonin for PEs and thus functions in their immune clearance (Garred et

al., 2003).

Humoral immunity and malaria It has been observed that in areas where malaria is hyper-endemic, adults acquire

relative resistance to malaria in comparison to children in whom acute infections

develop. Although the mechanisms underlying this naturally acquired immunity are

incompletely understood, there is strong evidence to suggest that antibodies against

parasitic proteins playa protective role during falciparum malaria (McGregor et al.,

1988). Elevated levels of serum IgM, IgE and IgG isotypes were found in patients

following parasite infection (McGregor and Wilson, 1988; Derowitz, 1989; Brasseur

et al., 1990). Passive transfer of IgG from immune serum from West African donors

to East African (McGregor et al., 1963) and Thai patients (Bouharoun-Tayoun et al., 1990) substantially decreased the parasite load, indicating that some of the

important antigens inducing protective responses are shared by P.falciparum parasites worldwide regardless of geographical origin. Various studies have been

carried out to assess the distribution of antibody isotypes with protective immunity,

although no clear pattern has emerged. In protected individuals, cytophilic

antibodies of IgG 1 and IgG3 isotype predominate, while in unprotected subjects,

especially children, either noncytophilic IgG2 and IgM classes or overall low levels of

antimalarial antibodies prevail (Bouharoun-Tayoun, 1992; Philips, 1994; Santhou

et al., 1997). Generally, several fold increase of IgG3 isotype levels occurs in

response to falciparum infection and is associated with a reduced risk of

complicated malaria (Tangteerawatana et al., 2007). Elevated concentrations of IgG2

antibodies may be associated with decreased risk to severe malaria as it has also

been shown that individuals carrying a protective functional allelic variant of the

Fcy receptor IIA (H131) binds the non cytophilic isotype IgG2 more efficiently than

the cytophilic isotypes (Nasr et aI., 2007). All these findings suggest that

antimalarial Ig-subclasses are differently regulated in patients.

Of major importance for the development of antibody mediated humoral immunity

to the blood stage of P.falciparum are parasite antigens expressed on the surface of

infected RBCs. The predominant antigens involved are members of highly variant

families, e.g. var gene product PfEMP1, expressed on parasitized erythrocytes, is the

major blood stage antigen and shows extremely high degree of antigenic variation.

The PfEMPl molecule contains binding sites for many host receptors expressed on

vascular endothelium and is the major parasite ligand to mediate cytoadherance.

The high antigenic variability of PfEMPl and its interaction with many host

receptors helps in parasite survival in the blood stream and also prevents parasites

from being destroyed by the spleen. Other than var gene products, there are many

multi-gene families which encode parasite antigens on the RBCs, e.g rif genes,

stevor and surfin gene families. Other candidate parasite antigens inducing

protective antibody response are polymorphic merozoite surface proteins (MSPs)

which have role to play during erythrocyte invasion and there is accumulating

evidence that a vaccine incorporating a portion of the conserved C-terminal of MSP3

was safe, immunogenic and induced antibodies with a strong anti-parasite effect

(Audran et al., 2005). The trypsin-resistant Variant Surface Antigen (VSA) expressed

on RBC membrane is also important for the development of blood stage immunity

(Neilsen et ai., 2002) and has been shown to induce female-specific antibody

response during pregnancy-associated malaria (Sharling et al., 2004).

The various mechanisms involved in antibody-mediated protection are still not very

clear but cytophilic antibodies (IgG 1, IgG3) may exert anti-parasitic effects by ADCC

(antibody dependent cell-mediated cytotoxicity) or ADCI (antibody dependent

cellular inhibition). Using mononuclear cells from Gambian children, Brown et al., (1986) showed that IgM is more effective than IgG in mediating ADCC. Antibodies

capable of binding to the parasitized RBC surface could not only prevent

sequestration to capillary endothelium but also enhance spleen clearance of these

cells (Udheinya et al., 1981; Jensen et al., 1989). Malaria infections in both human

and experimental animals are also associated with high production of IgE and anti­

Ig,E antibodies (Perlmann et al., 1994). Elevated IgE levels induce the production of

TNF and nitric oxide (NO) from mononuclear cells (Perlmann et al., 1995).

15

Overproduction of TNF is considered to be a major pathogenic mechanism

responsible for fever and tissue lesions in P. Jaiciparum malaria.

Cell-mediated immune mechanisms against malaria Cell-mediated immune responses during malarial infection, involving T lymphocytes

and antigen presenting cells (APCs) protect the host against both pre-eythrocytic

and erythrocytic parasite stages. T lymphocytes are produced in the bone marrow

and mature in thymus. Their cell surface receptors recognize antigens which are

processed and presented by APCs in association with MHC (major histocompatibility

complex) molecules. There are two major populations of T-cells, the CD4+ T-cells (T

helper cells) and CD8+ T-cells (cytotoxic T-cells). Of these T-cell populations, the

CD4+ cells are essential for immune protection against blood stages in both murine

and human malaria (Weindanz and Long, 1988; Troye-Blomberg et ai., 1994), whereas the CD8+ cells are shown to be involved in various effector mechanisms in

pre-erythrocytic stages of P.faiciparum infection (Schofield et ai., 1987; Sano et ai., 2001). The CD4+ T-cells playa central role in the cellular immune response against

P.faiciparum and are essential for both acquisition and regulation of malaria

immunity. Experiments on both rodents and humans with peripheral blood have

suggested the existence of malaria-specific but functionally distinct subsets of T­

helper (Th) cells (Troye-Blomberg et al., 1988; 1994). The two major subsets of Th

cells are Th1 and Th2 cells which are classified on the basis of different profiles of

cytokines they secrete. The Th1 and Th2 cells cross-regulate each other's

differentiation and activity via production of cytokines. The pro-inflammatory

cytokines produced by activated Th 1 cells like TNF, interferon-y (IFNy), IL-6 may be

protective and enhance parasite killing by NK cells and macrophages. Enhanced

plasma levels of TNF, IL-6 and IL-1 have also been shown to correlate with severe

P.faiciparum malaria (Kern et ai., 1989; Grau et ai., 1989; Kwaitkowski et ai., 1990). IL-12 produced by macrophages and by primed Th1 cells contributes to protective

immunity by inducing the production of IFNy by NK cells which helps in parasite

killing by cytotoxicity (Crutcher et ai., 1995; Doolan et ai., 1999). It has been also

shown that the proinflammatory cytokine IL18 plays an important role in the IL12-

mediated regulation of IFNy and activation of Th1 and NK cells (Dinarello et ai., 1999). The Th1 immune responses are regulated by the Th2 cytokines which check

excessive inflammation and tissue damage caused due to Th 1 cytokines and may

assist in preventing disease progression. The anti-inflammatory Th2 cytokine ILI0

has been shown to down regulate TNF production thereby preventing excessive

tissue damage mediated by NK cells during falciparum malaria (Ho et al., 1995).

Another Th2 cytokine, IL4 inhibits IFNy production (Snapper et al., 1987) and

correlates negatively with the parasite load in P.falciparum infected individuals

(Mshana et al., 1991). Many of the recent studies on P.falciparum malaria

emphasize on the balance between Thl and Th2 cytokines during stages of disease

severity. Thus, higher ILI0jTNF ratio in children from malaria holoendemic regions

of western Kenya suffering from mild malaria may prevent development of severe

disease progression by controlling excessive inflammatory activities of TNF (Caroline

et al., 1998). Also the anti-malarial plasma IgE levels in Tanzanian patients

correlated with an increased ratio of IL4jIFNy producing cells suggesting a shift in

the balance between Th 1 and Th2 cells in naturally P.falciparum primed individuals

(Elghazali et al., 1997).

The second class of T-Iymphocytes, the CD8+ T-cells play important role in pre­

erythrocytic stages of the malarial parasite where they have been shown to destroy

infected hepatocytes (Hoffman et al., 1989). lt is however not clear exactly how

CD8+ T-cells are activated and initiate their cytotoxic activity. A recent report in

rodent malaria has shown that CD8+ T-cells are primed in skin draining lymph

nodes in response to the circumsporozoite protein epitope of P.yoelli (Sumana et al .• 2007). The presence of CD8+ T-cells in in vitro cultures of peripheral blood of

mononuclear cells can inhibit parasite antigen induced proliferation and cytokine

secretion by CD4+ T-cells (Mshana et al., 1990) suggesting a role of CD8+ T-cells in

mediating immunosuppression during acute infection.

1.4.2 Cytoadherance

P.falciparum malaria distinguishes itself from other forms of malaria due to its

unique capability to adhere to the capillary and postcapillary venular endothelium

by a process called cytoadherance. Due to the sequestration of parasitized

erythrocytes (PEs), a number of alterations take place in the microcirculatory

environment resulting in multiple organ dysfunction and manifestation of more

severe forms of malaria, like cerebral malaria. The key players in the process of

cytoadherance are specific parasite ligands and host receptor molecules generally

called 'adhesion molecules'. These adhesion molecules are surface receptors that

are involved in cell-cell interactions. They are central to the recruitment and

trafficking of immune cells and generation of an active immune response. While

most of the adhesion molecules are constitutively expressed on the cell surface,

expression of some is upregulated by cytokines produced in response to parasite

attack. A number of host receptors and parasitic ligands that take part in

sequestration of PEs have been identified and are discussed below:

Endothelial receptors

CD36

CD36 or platelet glycoprotein IV, is a class B, 88kDa cell surface scavenger protein,

largely found on monocytes, RBCs, platelets and endothelial cells. Its natural

ligands are collagen (Tandon et al., 1989), thrombospondin (TSP) (Asch et aI., 1987)

and both natural and oxidized forms of high-density, low-density and very-low­

density lipoproteins (Calvo et ai., 1998). The CD36 molecule (Fig.!.l) has two

transmembrane domains, two short cytoplasmic domains and a large extracellular

domain which is heavily N-glycosylated (Oquendo et ai., 1989; Vega et al., 1991). It

was first reported by Barnwell et ai., (1989) that CD36 glycoprotein is the main cell

receptor for ligands on PEs when they showed that affinity-purified CD36 molecule

specifically binds PEs under static flow conditions. A monoclonal antibody to CD36,

OKM5, inhibits and reverses the cytoadherence of PEs to a number of target cells

(Talle et ai., 1983; Barnwell et aI., 1985) in vitro, including dermal microvascular r----------+. Hydrophilic Domain

Extracellular

Membrane

COOH Intracellular

Fig. 1.1: Schematic representation of the CD36 molecule.

18

endothelial cells and C32 melanoma cells, as well as purified CD36 immobilized on

plastic. Almost all natural isolates of P.faiciparum adhere to CD36 (Hasler et ai., 1990) suggesting its importance in pathogenicity during malarial infection. The

molecular basis of CD36 interaction with PEs is not fully understood. However,

certain regions in CD36 have been implicated in cytoadherance, all lying within an

immunodominant and hydrophobic extracellular region of the molecule (Baruch,

1999). As soon as the parasite enters the human body, the

monocytes/macrophages being the essential effectors of innate immunity are

actively involved in phagocytocis of PEs and mostly use CD36 to directly mediate

the process (Peiser et ai., 2002). It has been shown that exposure of nonopsonized

PEs to Fc receptor-blocked monocytes resulted in significant PE phagocytosis,

accompanied by intense clustering of CD36 around PEs and this phagocytosis was

blocked by 70-80% by pretreatment of monocytes with antibodies directed to CD36

(McGilvrey et ai., 2000). CD36 may also playa role in regulating the dendritic cell

(DC) function. CD36-binding PEs inhibit the maturation and function of DCs,

resulting in low expression of MHC class II molecules, down regulation of IL12 and

IFNy and enhanced production of IL10 cytokines (Urban et ai., 1999). Intriguingly,

CD36 is also involved in the recognition and ingestion of apoptotic cells, which

intum modulate DC function in response to inflammatory stimuli in a similar

manner (Urban, 2000). By interacting with DCs through CD36, P. Jaiciparum could

be mimicking the uptake of apoptotic cells to moderate the host immune response

to infection and thereby delay adaptive immunity thus giving the pathogen a

survival advantage. At the same time, reduced secretion of IL-12 and IFNy from DCs

could limit excessive inflammatory responses associated with adverse clinical

outcomes during severe P.falciparum malaria. Taken together, these findings

indicate the contribution of CD36 to the initial innate immune response and also as

a mediator in the development of adaptive immunity apart from being a host

molecule mediating cytoadesion of PEs.

ICAM-l

The Intracellular Adhesion Molecule (ICAM-1) is a 76-115kDa, immunoglobin

superfamily surface glycoprotein. It has five extracellular Ig like domains,

hydrophobic transmembrane domain and a short cytoplasmic tail (Chapter 4, Fig.

19

4.13}. The endodomain of the ICAM-1 molecule interacts with a-actin and ezrin

(cytoplasmic linker) and thereby participates in a number of signaling cascades.

ICAM-1 is widely expressed on venular endothelium of many cell types (epithelial,

endothelial, monocytic, fibroblast, B-and T-cell lines) including endothelial surface

of Blood Brain Barrier (BBB). It is a natural receptor for lymphocyte function­

associated antigen-1 (LFA-1 or CDlla/CD18 integrin) and due to this interaction,

ICAM-1 is involved in the firm adhesion and extravasation of leukocytes to the

endothelium (Marlin and Springer, 1987). Many studies have highlighted the role of

ICAM-1 in adhesion and migration ofT lymphocytes across BBB (Wong et ai., 1999).

During many pathological conditions, the integrity of BBB is disrupted resulting in

intense infiltration of T lymphocytes and progression of the disease. Upregulation of

ICAM-1 on brain vasculature and increased trafficking of T-cells during brain

inflammation leads to severe disease consequences like cerebral malaria. ICAM-1

was first identified by Berendt et ai. (1989) as the endothelial cell receptor for PEs

that mediate cytoadherance during P.faiciparum malaria. Molecular studies have

shown that PEs bind to the first immunoglobulin domain of ICAM-1 (Berendt et ai., 1992; Ockenhouse et ai., 1992). McCormick et ai., (1997) showed that HDMEC

(human dermal microvascular endothelial cells), which constitutively express CD36,

mediate high levels of PE adhesion in the absence of ICAM-1 and adhesion is

enhanced after the induction of ICAM-1, indicating substantial synergy between

these two receptors. The expression of ICAM-1 is upregulated during many

infectious diseases including severe P.falciparum malaria (Berendt et al., 1989). The

observations that higher levels of pro-inflammatory cytokines, TNF, ILl, IFNy are

associated with severe falciparum malaria (Kwaitkowski D, 1995; Awandare et ai., 2006) and that these cytokines also increase the expression of ICAM-1 (Rothlein et ai., 1988) suggests the role ofICAM-1 in severe malaria pathogenicity. ICAM-1 is the

potential host receptor involved in the adhesion of PEs in brain vasculature

(Newbold et ai., 1997). Histopathological examination of brain tissues of the patients

who have died of cerebral malaria showed high expression of ICAM-1 in brain

endothelial cells and co-localization with sequestered PEs, indicating the

involvement of ICAM-1 in cerebral malaria (Turner et ai., 1994). The detailed

mechanism involving ICAM-1 and cythoadherance during falciparum malaria is still

unknown. However, ICAM-1 may playa crucial role in various immune processes

triggered during malaria.

PECAMl (CD31)

The platelet endothelial adhesion molecule-1 (PECAM1jCD31) is a 130kDa

glycoprotein of the immunoglobin superfamily and is constitutively expressed on

human platelets, intracellular junctions of resting endothelial cells and on

circulating monocytes, granulocytes and certain T-cell subsets (Muller etal., 1989; Newman et al., 1990; Albelda et al., 1990). The PECAM1 molecule is composed of

six Ig-like extracellular domains, a short transmembrane domain and a long

cytoplasmic tail having multiple sites for phophorylation, lipid modifications and

other post-translational modifications (Newman et al., 1990) (Chapter 4, Fig. 4.14).

PECAM 1 has been shown to be involved in interendothelial adhesion and leukocyte­

endothelial interactions, particularly during transmigration. The adhesion process

pertaining to PECAM 1 is complex because PECAM 1 is capable of mediating both

homophilic and multiple heterophilic adhesive interactions (Newton et al., 1997).

Treutiger et al., (1997) identified that P.falciparum infected RBCs adhere to PECAM1

on vascular endothelium and also to recombinant PECAM 1 adsorbed on plastic.

Certain cytokines like TNF and IFNy have been shown to alter PECAM 1 adhesion to

PEs (Stewart et al., 1996; Treutiger et al., 1997; Bujan et al., 1999). A recent study

has shown that PECAM-1 and FcyRIla are colocalized on the platelet membrane and

PECAM-1 down-regulates FcyRIIa-mediated platelet responses (Thai et al., 2003).

Taking together, all these observations suggest that PECAM-1 is a virulence­

associated endothelial receptor of P. Jalciparum-infected RBCs.

VCAMl

The vascular cell adhesion molecule-1 (VCAM1) is a cell surface glycoprotein and a

member of the Ig-superfamily. Expression of VCAM 1 is inducible on vascular

endothelium during pathological conditions. However, it is constitutively expressed

on some non-vascular cells like DCs of skin and monocytes (Rice et al., 1990).

VCAM1 mediates cell-cell interaction via its natural ligand VLA4 (Very late antigen)

integrin which is expressed on monocytes, lymphocytes and some granulocytes

(Elices et al., 1990; Taichmen et al., 1991). Ockenhouse et al., (1992) showed that

(lb'i3~L ~-; byq V*,V]

21

VCAM 1 serves as a cell surface receptor for PEs on activated endothelial cells.

VCAMI expression can also be induced on BBB (Myron et al., 1991) where, together

with ICAM-l, it interacts with the junction protein ezrin and mediates leukocyte

adhesion and transmigration (Olga et al., 2002). Thus VCAMI may playa role in

manifestation of cerebral malaria.

Thrombospondin Thrombospondin (TSP), a 450kDa, trimeric extracellular matrix glycoprotein which

binds to CD36, a/f3 integrins, collagen, plasminogen etc. (Chen et al., 2000), acts as

a molecular facilitator to direct the clustering of receptors, growth factors and

matrix components, to specialized domains for adhesion and signal transduction.

TSP was the first molecule to be identified as a PE receptor. TSP being the natural

receptor for CD36, binding of PE to CD36 and TSP is highly correlated. Anti CD36

antibody OKM5 inhibits TSP-mediated binding to PEs. TSP also serves as a receptor

for the modified band-3 protein (Lucas et al., 1998) which is an adhesin on PE

surface. A specific RGD motif present on the type 3 repeats (T3) of TSP has been

shown to be the recognition site for modified band-3 protein on PEs (Eda et al.,

1999). Subsequent investigations showed that, although TSP may contribute to

cytoadherance, it is not sufficient to mediate the process by itself.

Other receptors Several other host receptors have been shown to interact with PEs. P-selectin,

secreted by activated platelets and endothelial cells, was shown to bind PEs under

flow conditions in a Ca++ dependent process through its lectin domain which

interacts with a sialic acid residue on the parasite ligand (Ho et al., 1998). Studies

with P.berghei infected mice have indicated that P-selectin is important for the

development of malarial pathogenesis but may not required for leukocyte adhesion

in the brain (Chang et al., 2003). Elevated plasma levels of soluble P-selectin were

reported in an African population of a malaria-endemic region (Facer et al., 1994).

These results raise the possibility that P-selectin may have an important beneficial

anti-inflammatory function during malaria infection. Another receptor, E-selectin,

whose expression on endothelial cells is induced by cytokines IL-l and TNF (Pober

et al., 1987), has also been shown to bind many parasite isolates (Rachanee et al.,

22

1996). Elevated levels of soluble E-selectin have been shown to correlate with

disease severity (Peyron et ai., 1994). One of the most serious forms of malaria

develops in pregnant women when late stages of PEs frequently sequester to the

placenta, a condition associated with low birth weight of the offspring and a major

cause of neonatal morbidity and mortality in areas where malaria is endemic.

Experimental evidence has shown that parasites obtained directly from human

placentas adhere to chondroitin sulfate A (CSA) expressed on syncytiotrophoblasts

that line the placental intervillous space (Fried and Duffy, 1996). CSA is a

glycosaminoglycan carrying a sulfate group and is present in the placenta in its low

sulfated form and optimally supports PE binding since the CSA levels are

significantly increased in the P.falciparum infected placenta (Achur et al., 2000; 2007). The risk of getting placental malaria is more in primigravid women as

compared to multigravidas and this is because antibodies against CSA-binding

parasites are produced during acquired humoral response to the parasite attack.

These anti-adhesion antibodies limit the accumulation of parasites in the placenta,

making women and the fetus resistant to placental malaria in subsequent

pregnancies.

Parasite cytoadherant ligands PjEMPl

Plasmodium Jaiciparum erythrocyte membrane protein-1 (PfEMP1) is the key

virulence factor of the parasite that is involved in rosetting and sequestration to

uninfected erythrocytes and endothelial cells (EC) of the host. The protein is

encoded by a large family of var genes (Su et ai., 1995; Smith et al., 2000) which are

predominantly localized in subtelomeric regions of all 14 chromosomes and show

extremely high degree of antigenic variation (Su et al., 1995). The extracellular

domain contains two binding motifs, Duffy binding-like (DBL) and cysteine-rich

interdomain region (CIDR) having a diverse binding potential that depends on their

primary sequence. This domain serves as a ligand for a myriad host receptors. The

CIDR1 domain of PfEMP1 is believed to be involved in adhesion to the host CD36

molecule forming a stable and long lasting anchor for PEs to bind to the endothelial

cells. Parasite entry in brain is mediated by cytoadherance between PfEMP1 and

ICAM-1 through the complex DBLfj and C2 PfEMP1 domains (Smith et al., 2000).

23

Besides sequestering to the EC, PEs also adhere to uninfected RBCs, a process

called rosetting. Complement receptor-1 (CR1), present on RBC membrane, plays an

important role in the formation of rosettes by interacting with the DBLla domains

of PrEMPl (Cooke et al., 2001). The considerable antigenic variation undergone by

PrEMPl contributes to the immune evasion and confers survival advantage to the

parasite.

Clag

It has been observed that subtelomeric deletions at chromosome 9 of P.Jalciparum

are associated with loss of binding of PEs to C32 me1enoma cells (Dey et al., 1993).

This chromosomal region was termed as clag9 and contains 9 exons which encode

for a 220 kDa protein distinct from PrEMPl (Holt et al., 1998). Clag9 has been

implicated in binding of PEs to the host endothelial receptor CD36 and targeted

gene disruption of clag9 in the stably cytoadherent P.Jalciparum line 3D7 abolishes

binding to melanoma cells or CD36 (Trenholme, 1999). The exact role played by

clag9 during cytoadhesion is still unknown. It is also hypothesized that clag9 may

be involved in the folding, transport or chaperoning ofPfEMPl (Trenholme, 1999). A

recent study has shown that clag9 is initially trafficked to the rhoptries of

P.Jalciparum (Gardiner, 2004). While the exact function of most rhoptry proteins is

not known, they are thought to be involved in invasion, formation of the

parasitophorus vacuole membrane and modification of the host cell membrane

during invasion (Sam-Yellowe et al., 1998; Douki et al., 2003). Thus it is also

possible that clag9 is involved in trafficking of other parasite proteins involved in

cytoadhesion or in the remodelling of the host red cell so that these proteins can be

trafficked to the correct location.

Sequestrin Sequestrin is a 270kDa protein present on PE surface and was shown to bind CD36

when it was immunoprecipitated by an anti-idiotypic antibody to anti-CD36 (MAb,

OKM8) (Ockenhouse et al., 1991). This binding was competitively inhibited by

soluble CD36. Sequestrin is a candidate malaria vaccine antigen, and anti-Id

antibodies that recognize this molecule may be useful for passive immunotherapy of

cerebral and severe falciparum malaria.

24

Modified band3 protein The modified band 3 protein or MB3P, a 65kDa protein, was immunoprecipated by

antibodies specific for the cytoplasmic domain and the N-terminal side of the

membrane-spanning region of human band 3 protein, but was not recognized by an

antibody specific to the C-terminal side of membrane spanning region (Crandall et ai., 1991). It was found that this 65kDa protein is truncated and covalently modified

band 3 protein (the anion transporter) of human erythrocytes. It contains the first

540 amino acids of human band 3. Thrombospondin is the specific host receptor for

MB3P (Lucas et ai., 1998). Based on peptide sequence of human band 3, synthetic

peptides HPLQKTY and YVKRVK were able to block PE adhesion in dose-dependent

manner (Lucas et ai., 1998).

Rifins and Surfins Rifins belong to the repetitive interspersed family of genes called as 'rif genes which

are generally found in close association with the var gene family present in clusters

at the telomeres of P.faiciparum chromosomes. A recent study reveals different

subcellular localization patterns for rifin variants which belong to two distinct

subgroups designated as A- and B-type rifins. While A-type rifins were found to be

associated with the parasite and transported to the surface of infected erythrocytes

via Maurer's clefts, B-type rifins appeared to be mostly retained inside the parasite

(Petter et al., 2007). Although their function remains unknown, immunity to rifin

proteins is associated with a stable immune response over time and with rapid

clearance of parasites from circulation (Abdel-Latif et ai., 2002; 2003). Furthermore,

rifin-specific IgG2 subclass antibodies are predominant in cerebral malaria patients

and may have important functions in malaria pathology. Another repetitive

interspersed family of genes called 'Surfins' have been recently identified. These are

located within or close to the subtelomeric region of the parasite chromosomes

(Winter et ai., 2005). Surfin proteins show structural and sequence similarities with

exported surface-exposed proteins. Surfins were found to co-transport with ?rEMP1

and rifin to the erythrocyte surface and also accumulated in the parasitophorous

vacuole. In released merozoites, surfins are present in an amorphous cap at the

parasite apex, indicating that they may be involved in the invasion of erythrocytes

(Winter et ai., 2005).

25

1.S Host genetic/actors and P./alciparum malaria

Malaria being the major killer of human populations is also the strongest known

force for evolutionary selection in the recent past. With accumulating evidence, it is

now apparent that the genetic variability due to selection by malaria is not confined

only to polymorphisms involving erythrocytes. During a relatively short time that

human beings have been exposed to malaria, there has been a remarkable change

in their genetic makeup as part of their adaptation to this major killer. A number of

studies carried out during the last few years have reported genetic associations with

susceptibility and resistance to P.falciparum malaria involving genes of immune

system, inflammation, cytoadhesion etc. Below is a brief account of recent advances

in the analysis of human genetic variations that may have resulted from repeated

exposure of populations to malaria.

1.5.1 Genes encoding adhesion molecules

As explained above, sequestration of the parasite to the host capillary endothelium

marks the most unique and fatal event of the host-pathogen interaction during

falciparum malaria. Molecular analysis of the mechanisms underlying

cytoadherance has produced a new class of candidate genes for malaria. Nucleotide

variations in the genes encoding adhesion molecules have shown differential

patterns of susceptibility/resistance to malaria at the population level. Fernandez­

Reyes et al., (1997) identified a mutation in the codon 29 of the [CAM1 gene in the

Kilifi region of Kenya. Homozygotes for this mutation were found to be more

frequent in patients suffering from cerebral malaria than in controls. However, there

exists an inverse correlation between this polymorphism and falciparum malaria in

Gabonese children suffering from the disease (Jurgen et al., 1999). In another study

conducted on Gambian malaria patients, no correlation was found between the

ICAM-1 codon 29 mutation (ICAM-lkilifi) and disease severity (Bellamy et al., 1998).

This mutation is almost absent in Thai and European populations but present at

very high frequency in most of the African populations and African-Americans living

in north America (Jurgen et al., 1999). The high allele frequency of this [CAM1 SNP

in African populations and their differential associations to falciparum malaria

suggests that there may be other selection pressures acting on the populations

apart from malaria.

26

Another important adhesion molecule that has been studied extensively in

association to falciparum malaria is CD36. Aitman et al. (2000) have showed that

two mutations, T1264G in exon10 and G1439C in exon12, which encode truncated

CD36 proteins and are the molecular basis of CD36 deficiency, were found in high

frequency in African patients suffering from severe malaria. A contrasting report by

Pain et al. (2001) has demonstrated that the same T1264G mutation in

heterozygous state is associated with protection from severe malaria in children

from the Kilifi area of Kenya. A variation screening of the CD36 gene done in Thai

malaria patients has shown that the frequencies of two SNPs in the upstream

promoter of the gene at positions -53 and -14 were significantly decreased

compared to controls (Omi et al., 2003). A repeat polymorphism, (TG}t2 in intron 3

was also found to be associated with reduced risk to cerebral malaria in the same

study (Omi et al., 2003). It was further shown that intron 3(TG) 12 is involved in the

nonproduction of the variant CD36 transcript that lacks exons 4 and 5. Since exon

5 of the gene is known to encode the ligand-binding domain for P. !alciparum-infected erythrocytes, intron3 (TG}t2 may be responsible for protection from cerebral

malaria in Thailand (Omi et al., 2003).

The adhesion molecule PECAM-1 has also been studied in relation to falciparum

malaria. Homozygotes of two non-synonymous SNPs, L125V and S563N, were

higher in patients suffering from cerebral malaria in Thailand, suggesting that these

genotypes are one of the risk factors for cerebral malaria in Thailand (Kikuchi et al., 2001). There are many other host adhesion molecules which are involved in parasite

sequestration during acute falciparum malaria. Variation screening of the genes

may lead to the discovery of certain new polymorphisms that can be correlated to

malaria disease manifestation.

1.5.2 Genes encoding immune regulatory molecules

Genetic polymorphisms in genes encoding immune regulatory molecules may have

varying impact on the manifestation of infectious diseases including malaria. Over

the last ten years, many studies have established correlation of genetic variants in

immune regulatory molecules with malaria. Tumor necrosis factor (TNF), the major

mediator of innate and humoral immune response during malaria, is the most

27

extensively studied cytokine in relation to the effect of its genetic variants on

falciparum malaria. Many promoter variants of the TNF gene exhibit differential

associations to malaria and TNF production in different populations. McGuire et ai., (1994) found that Gambian children who were homozygous for a polymorphism at -

308 nt position in the TNF enhancer, had a seven-fold increased risk of dying from

malaria. A study conducted on Kenyan children, also showed that the -308

polymorphism was strongly correlated with pre-term delivery and infant mortality

(Aidoo et ai., 2001). Wilson et ai., (1997) have showed that the -308 SNP results in

increased TNF production although there are conflicting reports regarding the

functional significance and disease association for this SNP. In Malian children and

Thai adults, there was no correlation of the -308 polymorphism with either TNF

production or malaria severity (Canbantous et ai., 2006; Ohashi et ai., 2001).

Likewise, another enhancer SNP at -238 position has been shown to correlate with

severe malarial anemia in Gambia (McGuire et ai., 1999) and high parasitemia in

Burkina Faso (Flori et al., 2005). However, this SNP showed no correlation with

malaria in Malian children (Cabantous et al., 2006). Other TNFpromoter variants at

positions -1031, -857 and -376 have been associated to cerebral malaria in Thai,

Myanmar and Gambian patients respectively (Hananantachai et ai., 2007; Ubalee et

ai., 2001; Knight et ai., 1999). The fact that particular variants of the same gene are

associated with different disease manifestations of severe malaria in different

populations is intriguing and emphasizes the need for better understanding of their

functional significance and of specific mechanisms involved in the role of these

SNPs in the TNF transcription regulation.

Many polymorphisms in genes encoding interleukins and their correlation to

malaria have been documented. A VNTR polymorphism in intron 3 of the IL4 gene

has been associated with cerebral malaria in Ghanaian children (Gyan et ai., 2004).

The promoter SNP at position -590 in the IL4 gene, which is believed to increase IL-

4 transcription (Borish et ai., 1994), correlates with enhanced P. Jaiciparum-specific IgE levels during falciparum malaria in West Africa (Verra et ai., 2004). However, no

correlation was established between this SNP and malaria disease severity. An

insertion/deletion polymorphism in the IL12B cytokine promoter has been

correlated to increased mortality in Tanzanian children having cerebral malaria but

28

not in Kenyan children with severe malaria (Morahan et al., 2002). Furthermore,

homozygotes for this polymorphism had decreased production of nitric oxide, which

is in part regulated by IL1213 (Morahan et al., 2002). IL-12 is important in

generation of Th1 immune response and thus various !L12 gene polymorphisms

and their effect on the regulation of IL-12 production may influence the outcome of

malaria infection. Polymorphisms in many other cytokines have been shown to

correlate with falciparum malaria in various African and south-Asian populations.

For example, carriers of a SNP in exon 5 of the !L1B gene show high parasitemia

(Gyan et al., 2002) in nonsevere malaria patients suggesting that this SNP may play

a possible role in clinical outcome of non-severe malaria. A CA repeat polymorphism

and a SNP at position -183 in the interferon-y (IFNG) promoter have been associated

with protection from cerebral malaria in Malian children (Cabantous et ai., 2005).

Genetic polymorphisms in various host molecules playing important roles in innate

immune responses during malaria have also been documented. Low plasma levels

of mannose binding lectin (MBL) and two SNPs in codons 54 and 57 have been

shown to correla~e with susceptibility to severe malaria in Gabonese children,

suggesting that deficient innate immune responses, in the form of low MBL levels,

may be a risk factor for malaria severity (Luty et ai., 1998). Genetic variants of toll

like receptors have also been examined in relation to falciparum malaria. The TLR4

Asp/Gly, codon299 and the TLR9 T-1486C polymorphisms increased the risk of low

birth weight in infants and increased the risk of maternal anemia in P.falciparum infected pregnant women in Ghana (Frank et al., 2006). These findings suggest that

TLR4 and TLR9 may playa role in the manifestation of malaria during pregnancy.

Various polymorphisms in the genes encoding cytokine receptors have been

associated to falciparum malaria. Two SNPs at nucleotide position 17470 and codon

168 in the interferon alpha receptor (IFNAR1) was associated to reduced risk to

cerebral malaria in Gambia (Aucan et al., 2003) suggesting a role of type 1 interferon

pathway in resistance to cerebral malaria. Another SNP in the IFNGRl gene has also

been implicated in falciparum malaria. In a case-control study conducted on

Gambian children, it was found that heterozygotes for the -56 SNP in the promoter

of IFNGRl gene, were protected against cerebral malaria and also the death

resulting from it (Koch et al., 2002). Another immune regulatory molecule that has

29

been implicated in malaria susceptibility is the human IgG receptor FcyRIIa (CD32)

which is an important link between the cellular and humoral arms of the immune

system. It is a low affinity receptor for immunoglobulin subtypes IgG 1-4 and also

binds C-reactive protein (CRP) with high affinity. A polymorphism in exon 4 of

FCGR2A (Arg/His, codon131, G/A) alters its function in vitro; the product of the G

allele (131R) has preferential affinity for IgG1 and IgG3 while the A allele (131H)

product binds efficiently to IgG2 while retaining its affinity for IgG 1 and IgG3

(Warmerdam et ai., 1991). The high-affinity IgG2 binding 131H allele has been

correlated with susceptibility to severe P.faiciparum malaria in Africa, particularly in

children in Kenya and Gambia (Shi et ai., 2001; Cooke et ai., 2003) but failed to

show significant independent association with severity of malaria in Thai adults

(Omi et ai., 2002). A recent report (Nasr et ai., 2007) has implicated the 131H allele

in protection from severe malaria in Sudan. Ethnic differences in the distribution of

allelic variants of many immune regulatory genes and differences in predisposition

to P.falciparum malaria at the population level suggests that in areas where malaria

transmission is high, there may exist strong biases in allele frequencies of various

polymorphisms. New variants from many other immune regulatory genes should be

investigated for better understanding of the patterns of susceptibility or resistance

of a popUlation to the disease.

1.5.3 Genes involved in erythrocyte invasion and rosetting

P.falciparum invasion of the host erythrocytes is an important event of the asexual

cycle of the parasite in the human host and is central to the pathology of the

disease. During P.falciparum malaria, the circulating merozoites invade RBCs by

two independent pathways involving two different ligands and host receptors. The

invasion pathway which involves the erythrocyte membrane protein Band3

(SLC4A1) is a sialic acid independent pathway and a 27 base-pair deletion mutation

in SLC4Al gene leads to a condition known as 'South-east Asian Ovalocytosis'

(SAO). The mutant is lethal in the homozygous state but seems to confer malaria

protection when heterozygous (Genton et ai., 1995; Allen et ai., 1999). SAO is quite

common in Melanesia with prevalence up to 35% in parts of PNG (Mgone et al., 1996). The exact molecular mechanism involving protection with malaria and SAO

is unclear, but it has been suggested that the Band3 deletion mutation may result

30

in structural changes in the RBC membrane that in some way interferes with the

merozoite invasion of RBCs. Another SNP in the promoter (-T512C) of the Band3

gene has been shown to correlate with severe malarial anemia and fatality in

Ghanaian children (Kalckreuth et al., 2006). In the same study it was shown that

this promoter polymorphism also results in higher transcriptional activity and is

also associated with higher metabolic acidosis in cerebral malaria patients. The

other invasion pathway which is sialic acid dependent, uses Glycophorin as the

host receptor for parasitized RBCs. Deletion of exon3 in the Glycophorin C (G¥Pq

gene resulting in serologic altered phenotype of Gerbich (Ge) blood group system,

called as Ge negativity, has been found in high frequency in many parts of PNG

where malaria is hyperendemic (Booth et aI., 1972; Serjeantson et al., 1994; Patel et

al., 2004). Maier et al. (2002) have shown that the P.falciparum ligand EBA140 does

not bind GYPC in Ge negative erythrocytes, nor can P. Jalciparum invade such cells.

This provides compelling evidence that Ge negativity may have arisen in Melanesian

(PNG) populations through natural selection by severe malaria.

The P.falciparum parasite is also involved in a phenomenon known as 'rosetting'

where uninfected erythrocytes cluster around a parasitized erythrocyte. Rowe et al., (1997) have identified Complement Receptor-1 (CR1) as the host molecule which

mediates rosetting with the parasite ligand PfEMP1 on PEs on host erythrocytes.

The CR1 molecule is primarily involved in clearance of immune complexes and

control of complement activation. During falciparum malaria, the CR1 molecule on

uninfected RBCs binds to parasitized RBCs to form clumps or rosettes which

contribute to severe malaria pathogenesis. It may be speCUlated that erythrocytes

with low CR1 surface expression will show reduced rosetting and hence confer

protection from severe malaria. On the other hand, high CR1 levels may be

beneficial in rapid clearance of opsonized parasitized RBCs to the spleen. A HindIII­

RFLP (restriction fragment length polymorphism) in the intron27 of the CR1 gene

has been correlated to the number of CR1 molecules on the RBC membrane (Wilson

et al., 1986). This polymorphism and low CR1 expression have been correlated to

severe falciparum malaria in Thailand (Nagayasu et al., 2001). A contrasting result

has been seen in Gambia where there exists no association between the HindIII­

RFLP and falciparum malaria (Bellamy et al., 1998; Zimmerman et al., 2003). The

31

correlation of the HindIII-RFLP with low CRI expression was not found in a study

conducted on Malian adults (Rowe et al., 2002) and in PNG the low CRI levels

correlated with another SNP in exon 22 (A3650G) (Cockburn et aI., 2004).

1.5.4 Linkage and genome wide association studies and malaria The various genes discussed above, shown to affect the susceptibility or resistance

to malaria have been identified by what is known as the 'candidate gene approach'

in which genetic variations are studied in singularity and their association with the

disease is established in a case-control format. This segregation approach has

helped in understanding host-genetic factors during malaria. However, a more

comprehensive approach is required to address certain issues like the extent to

which these genes influence the outcome of malarial infection when taken together,

identification of new genes that may affect malaria disease manifestation, and

existence of linkage disequilibrium across genomic regions controlling blood

infection levels. New research tools in which whole genomes can be scanned for

genetic variations affecting disease phenotype and are mapped and identified for

various association analysis including linkage patterns and their transmission

across generations have been developed. Initial studies done on the Gambian twins

and families in Burkina Faso have established a linkage of polymorphism between

genes of MHC (major histocompatibility complex) region on chromosome 6 (6p23)

with mild malaria (Jepson et al., 1997; Flori et al., 2002). In a sib-pair linkage

analysis done on malaria endemic Burkina Faso recruits using microsattelite

markers, an association was observed between 5q31-33 chromosomal region and

parasitaemia (Rihet et al., 1998; Flori et al., 2003). A similar trend of linkage was

seen among families of south Cameroon (Garcia et al., 1998). The 5q31-33 genomic

region contains genes which code for various interleukins and other molecules of

immunological importance like growth factors and receptors etc. The association of

this region with P.faiciparum blood infection levels indicates the likelihood of this

locus in control of parasitaemia. The validation of these results in many other

malaria-endemic popUlations may explain more comprehensively the critical role

played by this locus in malaria disease progression. Recently, linkage mapping and

SNP genotyping with a linked region on chromosome 21 (21 q22.11) conducted on

Gambian, Kenyan and Vietnamese patients have shown correlation of certain SNPs

32

with susceptibility to severe malaria (Aucan et al., 2003; Khor et al., 2007). This

locus contains receptors for various interferon and interleukins suggesting the

importance of this locus in the immune regulation during malaria. An autosome­

wide linkage analysis for P.falciparum infection intensity and mild clinical malaria

among Ghanaian children carried out by typing 10,000 SNPs has identified certain

chromosomal loci that are positively associated with a set of selected malaria related

phenotypes (Timmann et al., 2007). The study has found prominent signal of

linkage on chromosome 10 (10p15.3-14) correlated to frequency of malaria fever

episodes, along with a signal on chromosome 1 (lp36) with association with both

parasite prevalence and anemia. Interestingly, this study failed to find any signal of

association with malaria related phenotype at chromosome 5 (5q31-33) and to the

MHC region (6q23) which has been previously linked to P.falciparum malaria.

1.6 The diverse Indian population: a unique resource for complex disease

analysis

The Indian subcontinent is a conglomeration of the cultural, linguistic and genetic

diversity of its inhabitants. Though the exact time of entry of modern humans

(Homo sapiens sapiens) in India remains uncertain, it has been suggested that

modern man reached the north-western periphery of Indian subcontinent around

70,000 ybp during the middle Paleolithic period and then spread to many parts of

the country (Singh, 2002). The demographic expansion and population admixture of

ethnic groups due to waves of migration resulted in extensive social, cultural,

biological and linguistic diversity among the people of India. Linguistically, Indians

can be categorized as speakers of the Indo-European, Dravidian, Austro-Asiatic and

Tibeto-Burman language families. The Indian popUlation comprising of more than a

billion people, consists of distinct religious communities with hierarchical castes

and sub-castes and isolated tribal groups. Strict social rules that govern mating

patterns as well as geographical isolation make Indian populations highly

endogamous thus providing a unique template for dissecting complex disorders and

mapping underlying genes. Since candidate gene polymorphism-disease association

analyses are highly influenced by population genetic substructure, it is important to

assess the extent of population stratification and genetic differentiation among

endogamous groups in context of genetic variations in candidate disease genes.

33

1.6.1 The Indian Genome Variation Consortium (IGVC)

SNPs are the most common type of genetic variation in the human genome, and

have been extensively used for disease association and linkage disequilibrium

studies. There are a number of public databases that have SNP information on

many world populations, e.g. dbSNP (Sherry et al., 2001), HapMap (The

International HapMap Consortium, 2003), Celera (Kerlavage et al., 2002) etc.

However, the Indian population, which is one-sixth of the world popUlation, is not

represented in any of these databases. Given the uniqueness of the Indian

population in terms of its ethnic diversity and largely endogamaous nature,

information on its genetic variability will provide enormous insights for

understanding popUlation substructure and allow complex disease mapping. With

this objective, six national laboratories affiliated to the Council of Scientific and

Industrial Research (CSIR) initiated a network program on predictive medicine using

repeats and single nucleotide polymorphisms in the year of 2002 (IGVC, 2005). The

project aimed to provide a variation database of SNPs and repeats from over a

thousand genes on individuals belonging to various castes, tribes and religious

groups that make up the diverse Indian population. The genes selected for variation

screening were relevant candidates for complex diseases (e.g. cardiovascular,

neurological disorders) and infectious diseases as well as genes relevant to

pharmacogenomics. It was designed as a large-scale, comprehensive genetic study

of the Indian popUlation with wide-reaching implications. This lead to the formation

of a platform, the Indian Genome Variation Consortium (IGVC), which also included

the making of a portal, the Indian Genome Variation Database (IGVdb). In its first

phase, the project generated data on genetic variability of the Indian popUlation and

opened up immense possibilities for carrying out downstream studies and analysis

(IGVC, 2008). Our laboratory was a part of this consortium, and analyzed some of

the data generated in the project which was primarily focused on genetic variability

in genes that have a role to play during P.falciparum malaria.

1.7 Rationale

The incidence of P.falciparum malaria in India is high with -0.9 million cases

reported annually in the last few years (www.nvbdcp.gov.in). Several regions of the

country are 'high risk' for P.falciparum with the infection accounting for more than

34

+80% of all malaria cases in certain areas (Singh et al., 2003; Sharma et al., 2004).

Studies directed towards understanding host-genetic interactions during

P.faiciparum malaria have largely been carried out on African, Thai and PNG

populations with very few reports from India. The studies done on Indian

populations have concentrated on red cell defects and G6PD deficiency and their

association with malaria endemicity (Mohanty et al., 2003; Balgir, 2006). To

investigate the P.falciparum susceptibility/resistance profiles of diverse Indian

populations in the context of host genetic factors and degree of disease endemicity,

it is important to first study the distribution of variations (SNPs) in a range of host

genes related to malaria. The present study is focused on the analysis of genetic

variability, at the pan-India level, in specific gene loci that playa significant role

during P.faiciparum infection in humans. Association of selected SNPs with disease

severity would then be investigated in a case-control format with ethnically-matched

patients and controls drawn from a P.falciparum endemic and a non-endemic region

of India. The objectives of the study are:

• To discover and validate novel/ reported SNPs in the Indian population in

genes encoding immune regulatory and adhesion molecules that play a role

in P.falciparum malaria pathogenesis.

• To investigate functional correlation between identified SNPs and levels of

corresponding immune regulatory molecules.

• Apply the pan-India SNP data for disease association studies with malaria

patients drawn from a P.falciparum-endemic and a non-endemic region of

India.

35