7/29/2019 j.1365-3024.2007.00965.x

1/9

2007 Mahidol University 475Journal compilation 2007 Blackwell Publishing Ltd

Parasite Immunology, 2007, 29, 475 483 DOI: 10.1111/j.1365-3024.2007.00965.x

BlackwellPublishingLtdORIGINALARTICLESIgGsubclassesandIgE antibodyregulationinmalaria

Differential regulation of IgG subclasses and IgE antimalarial

antibody responses in complicated and uncomplicated

Plasmodium falciparum

malaria

P. TANGTEERAWATANA,

1

S. M. MONTGOMERY,

3,4,5

H. PERLMANN,

6

S. LOOAREESUWAN,

2

M. TROYE-BLOMBERG

6

& S. KHUSMITH

1

1

Department of Microbiology and Immunology, 2

Department of Clinical Tropical Medicine and Hospital for Tropical Diseases, Faculty ofTropical Medicine, Mahidol University, Bangkok, Thailand, 3

Clinical Epidemiology Unit, Department of Medicine, Karolinska UniversityHospital, Stockholm, Sweden, 4

Clinical Research Centre, rebro University Hospital, rebro, Sweden, 5

Department of Primary Care andSocial Medicine, Imperial College, London, UK, 6

Department of Immunology, Stockholm University, Stockholm, Sweden

SUMMARY

The aim of this study was to assess the immunoglobulin (Ig)-

subclass distribution of antimalarial antibody responses in 110

and 169 Thai patients with complicated and uncomplicated

Plasmodium falciparum

malaria, respectively. Antimalarial

plasma IgG subclasses and IgE antibody levels against a crude

malaria blood stages, and antigen preparation were determined

using enzyme-linked immunosorbent assay (ELISA). On

admission, the levels of anti-

P. falciparum

IgG1, IgG2 and

IgG3 were significantly lower in patients with complicated

malaria than uncomplicated malaria (IgG1, P

< 00001;

IgG2, P

< 00001; IgG3, P

< 00001). The levels of antimalarial

IgE were slightly lower, but not statistically significant(

P = 0389) in the complicated malaria. After adjustingall antibody levels and age, anti-

P. falciparum

IgG3 levels

remained significantly associated with complicated malaria.

None of the other antibody concentrations showed statisti-

cally significant associations with complicated malaria. The

anti-

P. falciparum

IgG3 levels were related to the IgG1 as well

as IgG2 levels. A correlation between anti-

P. falciparum

IgG2

and IgE was observed in the complicated malaria group, and

this may indicate their roles in the severity of disease. Our

data suggest that anti-

P. falciparum

IgG3 is associated with

a reduced risk of complicated malaria and that antimalarial

Ig-subclasses are differently regulated in patients with com-plicated and uncomplicated malaria.

Keywords

antibody,

complicated malaria,

IgE,

IgG subclasses,

P. falciparum

,

severity

INTRODUCTION

Anti-

Plasmodium falciparum

specific antibodies play a criticalrole in immune protection against the asexual blood stagesofP. falciparum

. Passive transfer of immunoglobulin (Ig) Gfrom adult Africans to Gambian children (1) or to adultThai patients (2) has been shown to protect against asexualblood stages of P. falciparum

infection. In areas wheremalaria is endemic, the cytophilic IgG1 and IgG3 subclassesare considered to protect against P. falciparum

, whereas IgG2and IgM are not. The latter Ig classes are even suggested toblock protective mechanisms (3). The correlation betweenisotype distribution and the clinical manifestation of malariais controversial. Anti-

P. falciparum

merozoite surfaceprotein (MSP) IgG3 antibodies have been shown to beassociated with either lower parasitaemias or lower risk ofclinical malaria (4,5). IgG1 antibodies to exoantigen areassociated with clinical protection, while IgG1 antibodiesto MSP2 sero group B are associated with increased riskof clinical infection (6,7). Likewise, IgG2 antibodies toexoantigen or crude schizont antigens have been shown tobe positively correlated to the number of malaria attacks or

clinical malaria (7,8), whereas high IgG2 and low IgG4antibody levels have been associated with resistance to

P. falciparum

in Burkina Faso (9). Thus, the cumulativeevidence of the role of different anti-

P. falciparum

Igsubclasses are contradictory and far from clear.

Another anti-

P. falciparum

Ig class, which is increased inthe majority of individuals living in area of high endemicityis IgE. Anti-

P. falciparum

IgE antibodies have been shownto be significantly higher in patients with severe malaria

Correspondence

: Dr Srisin Khusmith, Department of Microbiologyand Immunology, Faculty of Tropical Medicine, MahidolUniversity, 420/6 Rajvithi Road, Bangkok 10400, Thailand(e-mail: tmskm@mahidol.ac.th).

Received: 9 March 2007

Accepted for publication: 11 July 2007

7/29/2019 j.1365-3024.2007.00965.x

2/9

476

2007 Mahidol UniversityJournal compilation 2007 Blackwell Publishing Ltd, Parasite Immunology

, 29

, 475483

P. Tangteerawatana et al.

Parasite Immunology

than in those with uncomplicated disease, suggesting a roleof IgE in the pathogenesis of malaria (1013). However, ina recent study conducted in Tanzania, anti-

P. falciparum

IgE levels in asymptomatic individuals were found to beassociated with a reduced risk for subsequent malariadisease (14). In addition, in a study comprising sympatric

ethnic groups, IgE levels were higher in the more malariaprotected tribe, the Fulani when compared to theirneighbours, the Dogon (15). Thus, these data suggest thatIgE can have both protective and pathogenic roles. Thereasons why people produce different anti-

P. falciparum

IgG subclasses are not fully understood. Factors such as theparasite strains, the levels of malaria transmission, or thehost genetic factors have been implicated in the regulationof the production of different anti-

P. falciparum

Ig classesand subclasses. Additional factors may include the parasite-antigen preparations used in the analysis.

To understand better the natural human defence

mechanism against the malaria parasite, we have investigatedanti-

P. falciparum

IgG subclasses and IgE antibody responsesdirected against a crude P. falciparum

extract in patientswith complicated or uncomplicated malaria. We evaluatedand described the isotype kinetic profile, and examined therelationship between the pattern of IgG subclasses andIgE with disease manifestation, comparing patients withcomplicated and uncomplicated forms of malaria.

MATERIALS AND METHODS

Malaria subjects

A total of 279 P. falciparum

malaria patients, 110 withcomplicated and 169 with uncomplicated malaria, wereenrolled in this study during the period between April 2002and May 2003. They had all been living in malaria endemicareas of the ThaiMyanmar border in the west and ThaiCambodia border in the east of Thailand. These areas areconsidered as having low and seasonal malaria transmission.The annual malaria incidence rates in 2001 were 26/1000population (16). Clinical manifestations of complicatedand uncomplicated malaria were defined according to WorldHealth Organization criteria (17). Cerebral malaria wasdefined as unrousable coma with positive asexual forms

ofP. falciparum

in blood smears, while other causes of comawere excluded. Severe malaria but not cerebral (noncerebralsevere malaria) was defined as patients exhibiting one ormore of the following signs: hyperparasitaemia (> 250 000parasites/mL); hypoglycaemia (glucose, < 22 nmol/L); severeanaemia (haematocrit, < 20% or haemoglobin, < 70 g/d);or increased serum level of creatinine (> 30 mg/dL).Uncomplicated malaria was characterized by a positiveblood smear and fever without other causes of infections

and no manifestations of severe malaria as describedabove. Patients diagnosed with malaria by symptoms andmicroscopic examination were transported for admission tothe Hospital for Tropical Diseases, Faculty of TropicalMedicine, Mahidol University, Bangkok where all the clinicalcharacteristics, haematological tests and biochemical tests,

for example, splenomegaly, G6PD-deficiency, Hb variantsblood groups, and so on,, were examined and recorded.Pregnant women or those who had been given antimalarialtreatment in the previous 2 weeks were excluded. The patientswith complicated malaria in the present study have all formsof severe malaria including cerebral malaria. The study hasbeen ethically approved by the Institute Review Boardof the Faculty of Tropical Medicine, Mahidol University,Bangkok, and fully informed consent was obtained from allthe patients.

Sample collection

Serial venous blood samples were collected in EDTA steriletubes before and after treatment at admission (Day 0), Day7 and Day 28. Plasma was collected, and heat inactivated at56

C for 30 min and stored at 20

C until use.

Microscopic observation

Thick and thin films were made for microscopic examinationafter standard Giemsa staining. Parasite densities weredetermined by calculating P. falciparum

parasites per

L ofblood, obtained from the number of parasites observedper 1000 erythrocytes in thin film or per 200 leucocytes inthick film.

Parasite extract preparation

Lysates of mature stages of Percoll-enriched P. falciparum

-infected erythrocytes of the laboratory strain F32 wereprepared and used as antigen for the detection of malariaantibodies as described (18).

Detection of anti-

P. falciparum

IgG subclasses and

IgE antibody levels

Levels of anti-

P. falciparum

IgG1, IgG2, IgG3 and IgEantibodies were measured by ELISA as described previously(13). Briefly, plates were coated overnight with 50

L/well of10

g/mL of P. falciparum

extract in sodium carbonatebuffer pH 96, and blocked with 05% w/v BSA in coatingbuffer for 3 h at 37

C. Samples were then added at differentdilutions, 1 : 400 for the determination of anti-

P. falciparum

IgG1; 1 : 100 for IgG2; 1 : 1600 for IgG3; and 1 : 100 for IgE.Samples were tested in duplicates. The sera were allowed to

7/29/2019 j.1365-3024.2007.00965.x

3/9

2007 Mahidol University

477

Journal compilation 2007 Blackwell Publishing Ltd, Parasite Immunology

, 29

, 475483

Volume 29, Number 9, September 2007 IgG subclasses and IgE antibody regulation in malaria

react for 1 h at room temperature for all determinations,with the exception of anti-

P. falciparum

IgE antibodies,which were incubated overnight. Anti-

P. falciparum

IgG1,IgG2, IgG3 or IgE were detected with biotinylated mousemonoclonal anti-human IgG1, 1 : 3000; IgG2, 1 : 3000;IgG3, 1 : 1000 (Pharmingen, USA), goat anti-human IgE,

1 : 8000 (Vector Laboratories, USA), and followed by ALP-conjugated streptavidin (diluted 1 : 2000) (Mabtech, Nacka,Sweden). The optical density values were read in a VmaxMicroplate Reader (Molecular Devices Corporation,Sunnyvale, CA, USA) at 405 nm wavelength. The concentra-tions were calculated from the standard curve obtained byincubating serial dilutions of either purified myelomaIgG1, IgG2 and IgG3 isotypes (Serotec, Oxford, UK) orhuman serum IgE (NIBSC, UK) well coated with goat anti-human IgG1 (Pharmingen, USA) for IgG1, goat anti-humanIgG (Jackson ImmunoResearch, Europe Ltd., Suffolk,UK) for IgG2, mouse anti-human IgG (Pharmingen, USA)

for IgG3 or goat anti-human IgE (Vector Laboratories,USA) for IgE and assayed as described above. Plasma IgGsubclasses and IgE optimum dilutions were tested beforethe experiments were performed. The optimum dilution foreach IgG subclass and IgE was then multiplied by eachIgG subclass and IgE concentrations obtained from thestandard curves.

Statistical analysis

The data were analysed using

or S

V

computersoftware. Antibody levels are presented as geometric mean 95% CI. The MannWhitney U

-test was applied forcomparison between complicated and uncomplicated malariapatients. The relationship between IgG subclasses and IgEantibodies was examined using linear regression analysis.Multiple logistic regressions were used to investigate whetherthe levels of the IgG subclasses and IgE antibodies weresignificantly different between complicated and uncom-plicated malaria. The independent variables were anti

-

P. falciparum

IgG subclasses and IgE antibodies, and age.The levels of anti-

P. falciparum

IgG subclasses and IgEantibodies were grouped into equal fifths of their distribu-tions. Age was categorized into five groups (< 20, 2029,3039, 4049 and > 49 years). These measures were modelled

as a series of binary dummy variables. The highest fifthantibody level was used as the comparison group. Thus,univariate and multivariate analyses for categoricalvariables were performed. A P

-value less than 005 and 95%confidence intervals for odds ratios which do not cross 100were considered statistically significant.

To ensure that the results are not due to co-linearitybetween IgG1 and IgG2, two further adjusted logisticregression models had malaria severity as the dependent

variable. The first included measures of IgG1, IgG3and age, while the second included measures of IgG2, IgG3and age.

RESULTS

Characteristics of patients

The characteristics of 110 adults with complicated and 169adults with uncomplicated malaria are delineated in Table 1.No statistically significant difference was observed in theage distribution, sex ratio and nationality between thecomplicated and uncomplicated malaria groups. In addition,the blood group distribution, and percentages of individualswith haemoglobin variants and enzyme G6PD deficiencydid not differ between the complicated and uncomplicatedgroups. The mean parasite density was significantly higherin the complicated group compared to the uncomplicated

group (

P < 00001). When the highest temperatures werecompared between the uncomplicated and complicatedgroup, a significantly higher temperature was found in theuncomplicated group (

P = 0010). Individuals presentingwith complicated malaria were less likely to have hadprevious episodes of malaria, compared to those present-ing with uncomplicated disease (

P = 00004). Clinicalexamination revealed a significantly higher rate of spleenand liver enlargement in the patients with complicatedmalaria than those with uncomplicated malaria (

P = 0039and P

< 00001, for splenomegaly and hepatomegaly,respectively).

Association between anti-

P. falciparum

antibodies and

complicated malaria

Anti-

P. falciparum

serum antibody levels in patients weredetermined before and after treatment, that is, at Days 0, 7and 28. The geometric mean values of anti-

P. falciparum

IgG subclasses and IgE antibodies in both groups areillustrated in Figure 1. On Day 0 before treatment, thegeometric means of anti-

P. falciparum

IgG1, IgG2 and IgG3antibodies were lower in the patients with complicatedmalaria than in those with uncomplicated malaria(IgG1, P

= 00001; IgG2, P

< 00001; IgG3, P

< 00001). The

geometric mean of anti-

P. falciparum

IgE antibodywas slightly lower, although not statistically significant(

P = 0389) in the patients with complicated malaria thanuncomplicated malaria. On Days 7 and 28 after treatment,the geometric mean of anti-

P. falciparum

IgG1 and IgG3was still lower in patients with complicated malaria thanthose with uncomplicated malaria (IgG1, P

= 0001 andIgG3, P

= 0002 for Day 7; IgG1, P

= 00002 and IgG3,

P

= 00007 for Day 28). No significant differences were

7/29/2019 j.1365-3024.2007.00965.x

4/9

478

2007 Mahidol UniversityJournal compilation 2007 Blackwell Publishing Ltd, Parasite Immunology

, 29

, 475483

P. Tangteerawatana et al.

Parasite Immunology

observed for the geometric means of anti-

P. falciparum

IgG2 or IgE between patients with complicated comparedto those with uncomplicated malaria (IgG2, P

= 0882and IgE, P

= 0884 for Day 7; IgG2, P

= 0539 and IgE,

P

= 0318 for Day 28).The results of the logistic regression analysis comparing

associations of characteristics on Day 0 with risk ofcomplicated malaria are shown in Table 2. Univariateanalysis before adjustment showed that lower levels of anti-

P. falciparum

IgG3, IgG1 and IgG2 antibodies on Day 0were associated with a statistically significant increasedrisk of complicated malaria (OR > 100, P

< 005), and noassociation with complicated malaria was observed for theanti-

P. falciparum

IgE levels. To exclude the possibility ofconfounding by age and co-linearity between anti-

P. falciparum

IgG1, IgG2, IgG3 and IgE antibodies, multiple simultaneousadjustment was performed for these factors as describedin the statistical analysis section. After adjustment, onlylower levels of anti-

P. falciparum

IgG3 antibodies remainedsignificantly associated with a greater risk of complicatedmalaria (OR > 100, P < 005), whereas no significant

association of anti-P. falciparum IgG1, IgG2 and IgEantibodies with the risk of succumbing to complicatedmalaria was observed.

The analysis performed for Day 0 levels was repeated forsubsequent days, that is, Days 7 and 28, and the negativeassociation of IgG3 levels in patients with complicatedmalaria was consistently observed and independently fromthe other Ig levels and age, although the magnitude of theassociation was reduced (data not shown).

Table 1 Characteristics of patients with complicated and uncomplicated malaria

Characteristics Complicated malaria (110)a Uncomplicated malaria (169)a P-valueb

Agec (range) 250 (1467) 240 (1365) 0849Sex (male/female) 75/35 133/36 0137Nationalityd

Burmese 17 (155) 24 (142) 0614Karen 18 (164) 27 (160)Mon 37 (336) 72 (426)Thai 38 (345) 45 (266)Laos 0 1 (06)

Blood groupsd

A 21 (191) 27 (160) 0556AB 6 (55) 7 (41)B 34 (309) 58 (343)O 47 (427) 66 (391)NR 2 (81) 11 (65)

Haemoglobin variantsd,e

Hb variants 14 (127) 29 (172) 0825Normal 95 (864) 133 (787)NR 1 (09) 7 (41)

G6PD deficiencyd

Deficiency 13 (118) 14 (83) 0597Normal 96 (873) 153 (905)NR 1 (09) 2 (12)

Parasitaemia (range)f 117 322 (331 336 900) 5641 (17238 800) < 00001Highest fever in Cg 381 1005 384 0981 0010Previous malaria episodesd

Yes 30 (273) 93 (550) < 00004No 78 (709) 76 (45)NR 2 (18)

Splenomegalyh

Positive 18 (164) 3 (81) 0039Negative 92 (834) 166 (982)

Hepatomegalyh

Positive 66 (60) 32 (189) < 00001Negative 44 (40) 137 (811)

aNumber of patients; bMannWhitney U-test used for the analysis; cMedian; dPercentages (in parentheses); eThe Haemoglobin (Hb) variantsinclude HbE, HbF; fGeometric mean/L of blood; gMean SD; hPercentages of individuals with enlarged spleens and/or liver(in parentheses). NR, No report.

7/29/2019 j.1365-3024.2007.00965.x

5/9

2007 Mahidol University 479Journal compilation 2007 Blackwell Publishing Ltd, Parasite Immunology, 29, 475483

Volume 29, Number 9, September 2007 IgG subclasses and IgE antibody regulation in malaria

The lack of association of IgG1 and IgG2 with malariaseverity in the adjusted model was not due exclusively toco-linearity between these two antibodies. When adjustedonly for age and IgG3, the association of IgG1 with malariaseverity was not statistically significant. Compared with thehighest level, the other levels from low to high had oddsratios (and 95% CI) of: 048 (016142); 077 (029202);085 (034213); and 047 (018123). Similarly, there was

no statistically significant association of IgG2 with malariaseverity after adjustment for age and IgG3, producingodds ratios of: 073 (025219); 105 (038285); 147 (058372); and 078 (030203).

Anti-P. falciparum IgG1, IgG2, IgG3 and IgE relationship

At admission, 110 patients with complicated and 169with uncomplicated malaria had detectable levels of anti-

P. falciparum IgG subclasses and IgE antibodies. The levelsof the different malaria-specific IgG subclasses and IgEvaried greatly between individuals. The correlation betweenthe specific IgG1 and IgG3 antibody levels (r = 0295, P