Prenatal T Cell Immunity to Wuchereria bancrofti and Its Effect ...

Prenatal T Cell Immunity to Wuchereria bancrofti and Its

Effect on Filarial Immunity and Infection Susceptibility

during Childhood

Indu Malhotra, Peter L. Mungai, Alex N. Wamachi, Daniel Tisch, John M. Kioko, John

H. Ouma, Eric Muchiri, James W. Kazura, Christopher L. King

Dow

nloaded from https://academ

ic.oup.com/jid/article/193/7/1005/2191556 by guest on 01 June 2022

http://akrurl.com/.2zohn

Neonatal Immunity and Childhood Susceptibility to Lymphatic Filariasis • JID 2006:193 (1 April) • 1005

M A J O R A R T I C L E

Prenatal T Cell Immunity to Wuchereria bancroftiand Its Effect on Filarial Immunity and InfectionSusceptibility during Childhood

Indu Malhotra,1 Peter L. Mungai,1 Alex N. Wamachi,3 Daniel Tisch,1 John M. Kioko,2 John H. Ouma,4

Eric Muchiri,2 James W. Kazura,1 and Christopher L. King1

1Center for Global Health and Diseases, Case Western Reserve University, Cleveland, Ohio; 2Division of Vector Borne Diseasesand 3Kenya Medical Research Institute, and 4Kenyatta University, Nairobi, Kenya

Background. Antenatal immune experience with Wuchereria bancrofti due to maternal filariasis may influencesusceptibility to infection. We tested the hypothesis that filarial-specific T cell responses at birth that are indicativeof in utero tolerance or sensitization affect the evolution of filarial-specific immunity and susceptibility to W.bancrofti infection during childhood.

Methods. A birth-cohort study of 159 Kenyan newborns was performed. Cord blood and peripheral bloodwere obtained annually to age 7 years and were assayed for filarial infection and filarial antigen–driven interferon(IFN)–g, interleukin (IL)–2, IL-5, and IL-13 production by lymphocytes.

Results. There was a 12.9-fold (95% confidence interval [CI], 2.5–107.2-fold) and a 4.8-fold (95% CI, 1.7–12.9-fold) increased risk of infection for immune-tolerant newborns (maternal infection present during gestation,with no filarial antigen–driven cord blood T cell response [ ]), compared with immune-sensitized (maternaln p 25infection present with cord blood T cell response [ ]) and unexposed (maternal infection absent [n p 24 n p

]) newborns. Cytokine responses developed at a later age in tolerant newborns, were characterized by impaired110IFN-g responses, and contrasted with those of filarial-sensitized newborns, who had sustained and elevated IL-5and IL-13 responses to age 7 years.

Conclusion. Prenatal immune experience, as determined by whether in utero priming to filarial antigen occurs,is a major determinant of childhood susceptibility to W. bancrofti infection.

More than 120 million residents of Africa, Latin America,

Asia, and the Pacific region are infected with the lym-

phatic filarial parasites Wuchereria bancrofti and Brugia

species, and 11 billion persons are at risk [1]. The dis-

tribution of infection and lymphatic disease attributable

to these worms varies widely across and within pop-

ulations in endemic areas, ranging from asymptomatic

uninfected individuals to persons with light to heavy

Received 25 August 2005; accepted 14 October 2005; electronically published22 February 2006.

Potential conflicts of interest: none reported.Presented in part: annual meeting of the International Centers for Tropical

Disease Research, Bethesda, Maryland, 14 May 2004.Financial support: National Institutes of Health, US Public Health Service (grants

AI36219 and AI33061).Reprints or correspondence: Dr. Christopher L. King, Center for Global Health

and Diseases, Case School of Medicine, Case Western Reserve University,Wolstein Research Building, 2103 Cornell Rd., Cleveland, OH 44106-7286([email protected]).

The Journal of Infectious Diseases 2006; 193:1005–13� 2006 by the Infectious Diseases Society of America. All rights reserved.0022-1899/2006/19307-0014$15.00

parasite burdens, with or without overt lymphatic path-

ological abnormalities [2]. These outcomes are not static;

they are dynamic in that increasing age and cumulative

exposure to mosquitoes bearing infective larvae correlate

with higher infection burden and a propensity to develop

lymphatic pathological abnormalities [3, 4]. Although

host genetic polymorphism and bacterial superinfection

may influence susceptibility to infection and disease [5–

7], prenatal filarial-specific immune tolerance or sensi-

tization associated with maternal infection during ges-

tation and adaptive T cell cytokine responses appear to

have a dominant effect. Cross-sectional surveys have

shown that children whose mothers were putatively

microfilaremic during gestation were more likely to be

microfilaremic, compared with children whose mothers

were amicrofilaremic during gestation [8–10]. Since

childhood microfilaremic status was independent of pa-

ternal infection, these associations have been interpreted

to indicate that the apparent increase in susceptibility

was due to neonatal tolerance conferred by maternal

Dow



1006 • JID 2006:193 (1 April) • Malhotra et al.

filariasis [9, 10]. Indirect evidence supporting the importance of

prenatal immune factors in determining the outcome of natural

exposure to infective larvae derives from several observations.

First, permanent residents of a filariasis-endemic area of Indo-

nesia were found to have less-severe lymphatic pathological ab-

normalities, compared with recent migrants from nonendemic

areas who were first exposed to mosquito-borne infective larvae

as adults [11, 12]. Second, filarial-specific IgE and IgM, im-

munoglobulin isotypes that do not cross the placenta, have been

detected in cord blood from infants born to mothers with filarial

infection, suggesting in utero sensitization to filarial antigens [13–

16]. Third, Steel et al. [17] compared filarial antigen–driven cy-

tokine responses of Polynesian adolescents born to mothers who

were putatively microfilaremic with those of Polynesian adoles-

cents whose mothers were infection free. Although none of the

study participants was infected (mass treatment with antifilarial

drugs had been performed in the Cook Islands at least once since

the time of the participants’ birth 17–19 years earlier), the 10

individuals born to infection-free mothers had stronger filarial-

driven interleukin (IL)–2, IL-5, IL-10, and interferon (IFN)–g

responses than did the 11 individuals born to putatively micro-

filaremic mothers, suggesting that long-term immune tolerance

is conferred by maternal filarial infection during gestation. The

relationship between maternal filariasis and filarial-specific T cell

immunity in newborns has, however, not been examined in any

study. More importantly, it is not known whether or how ma-

ternally conferred immunity affects the evolution of parasite-

specific T cell immunity and susceptibility to infection during

childhood. We conducted a prospective cohort study of 159 Ken-

yan infants to address these gaps in knowledge.

SUBJECTS, MATERIALS, AND METHODS

Prospective cohort study. The institutional review boards of

the Kenya Medical Research Institute and University Hospitals

of Cleveland/Case Western Reserve University approved the

prospective study of infants whose mothers attended the an-

tenatal clinic and gave birth at Msambweni District Hospital

in Coast Province, Kenya. The community-specific prevalence

of W. bancrofti infection, on the basis of microfilaremia in the

area, has a range of 9.2%–21.0% [10]. Inclusion criteria were

uncomplicated singleton pregnancy, vaginal delivery, normal

birth weight, permanent residence, and intent to remain in the

area for the next 5 years. Exclusion criteria were complicated

pregnancies (e.g., pre-ecampsia and poorly controlled diabe-

tes), cesarean delivery, complicated vaginal delivery, and prema-

ture delivery (!36 weeks’ gestation). Systematic interventions

against filariasis did not occur during the time when the study

was conducted (1995–2002). In 2003, participants who were

found to be infected were given a single dose of diethylcar-

bamazine (6 mg/kg of body weight) plus 400 mg of albendazole,

under the auspices of a filariasis-elimination program initiated

by the Kenya Ministry of Health.

Diagnosis of filarial infection. The filarial-infection status

of pregnant women was determined before delivery, and the

status of their infants was determined 1 time per year up to

age 7 years on the basis of (1) IgG4 antibodies to a saline extract

of B. malayi adult male and female worms (BmA) [14], (2) W.

bancrofti–specific Og4C3 antigen in plasma (TropBioMed) [18],

and (3) microscopic counting of microfilaremia after Nuclepore

filtration of 1 mL of venous blood obtained between 2000 and

2400 h [19]. Maternal infection intensity was classified as light

when BmA-specific IgG4 antibody alone was positive, moderate

when BmA-specific IgG4 antibody plus Og4C3 antigen were

positive, and heavy when all 3 tests were positive. Of note,

pregnant women who were only BmA IgG4 positive could rep-

resent only continued exposure and may not have developed

a fully patent filarial infection. Children were considered to be

infected with W. bancrofti when Og4C3 levels exceeded 32 U/

mL (the cutoff value recommended by the manufacturer) for

2 consecutive years. Infants who did not participate in the first

follow-up examination at 1 year of age were excluded from

analysis. Children who participated in the first-year visit and

�1 subsequent annual follow-up visit were included. IgG and

IgE antibodies to B. malayi infective larvae were measured by

ELISA and were used to evaluate exposure to infective larvae

independent of infection status [20].

Diagnosis of nonfilarial infections. The presence of ma-

laria parasites in venous blood and placental intervillous blood

was determined by smear and real-time quantitative polymerase

chain reaction of the gene encoding the small-subunit ribo-

somal RNA [21]. Childhood infection with intestinal helminths

and urinary schistosomiasis was evaluated annually using the

Kato method for fecal specimens and Nuclepore filtration of

urine samples, respectively.

Antigens and mitogens. Parasite antigens were saline ex-

tracts of BmA and microfilariae [15]. Protein antigen concen-

trations (5 or 10 mg/mL BmA and 1 mg/mL microfilarial an-

tigen) were titrated to levels that failed to stimulate cytokine

production by mononuclear cells isolated from cord blood from

10 healthy US newborns and peripheral blood from 8 healthy

US adults. Phorbol 12-myristate 13-acetate (50 ng/mL; Cal-

biochem) and ionomycin (1 mg/mL; Calbiochem) were used as

mitogens in parallel cultures. Mycobacterial purified protein

derivative (PPD; Evans Medical) was used at a concentration

of 10 mg/mL.

Cytokine and antibody measurements. Mononuclear cells

were separated from heparin-anticoagulated cord blood from

newborns and peripheral blood from children by density-gra-

dient centrifugation on ficoll-hypaque, washed in RPMI 1640,

and suspended to a final concentration of cells/mL in62 � 10

Dow




Figure 1. Study design of newborn cohort followed for time to filarial infection

RPMI 1640 supplemented with 10% fetal calf serum, 4 mmol/

L l-glutamine, 25 mmol/L HEPES, and 80 mg/mL gentamicin

(BioWhittaker). Antigens or mitogens were added to duplicate

or triplicate cultures. Cultures with media alone served as con-

trols. Cells were incubated at 37�C in 5% CO2, and supernatants

were collected at 72 or 96 h for measurements of IL-2, IFN-

g, IL-5, and IL-13, which were performed as described else-

where [15]. The limits of detection were 38 pg/mL for IL-2,

40 pg/mL for IFN-g, 20 pg/mL for IL-5, and 32 pg/mL for IL-

13. We analyzed filarial antigen–driven cytokine responses only

when cord blood lymphocytes produced all 4 cytokines when

stimulated with mitogen. Cord blood from 7 infants produced

little or no cytokine production after mitogen stimulation, as

a consequence of poor lymphocyte viability ( ) or con-n p 4

tamination of cultures ( ) (e.g., spontaneous cytokine lev-n p 3

els were high and obscured any mitogen or antigen-induced

stimulation). None of these samples had detectable filarial an-

tigen–induced cytokine production and, thus, could not be in-

cluded in the analysis.

Statistical analysis. The Mannc-Whitney U test was used

for comparisons of cytokine responses by various groups of

children. The x2 test was used to compare the proportion of

positive responders among the groups. A log-rank test was used

to compare Kaplan-Meier survival curves for infection suscep-

tibility according to maternal infection status alone and new-

born filarial antigen–driven T cell responses combined with

maternal infection status. The proportional hazard assumption

was tested by assessing the log (�log)–estimated survival plot-

ted against the log of time, where the assumption would be

considered violated if plotted lines of exposure intersected.

Without violation of this assumption, a Cox proportional haz-

ard model was fit to compare the hazard across exposure status.

Finally, the independence of survival time to hazard at time t

in each exposure group was further tested by including time-

and log of time–dependent exposure covariates, using the Wald

statistic, where a significant time-dependent effect of exposure

would be considered to violate the proportional hazard as-

sumption. All analyses were performed using SAS (version 8.2;

SAS Institute).

RESULTS

An overview of the enrollment and follow-up of the participants

is presented in figure 1. A total of 193 mother-newborn pairs

were enrolled between 1995 and 1999; 83% of children in the

birth cohort ( ) remained in the study until follow-upn p 159

Dow




Table 1. Characteristics of the newborn cohort, grouped according to maternal infectionstatus and cord blood T cell cytokine response to filarial antigen.

Characteristic Sensitized Tolerant Unexposed

Maternal infection with filariasis Yes Yes NoNewborn T cell response to filarial antigen Yes No NoNo. 24 25 110Village of residence

I 4 (17) 7 (28) 15 (14)II 2 (8) 4 (16) 22 (20)III 5 (21) 2 (8) 8 (6)IV 8 (33) 7 (28) 35 (32)V 2 (8) 1 (4) 10 (9)VI 3 (13) 4 (16) 20 (18)

CoinfectionMalaria positive by PCR

Maternal placental blood 9 (38) 11 (44) 36 (33)Children during follow-up 19 (79) 21 (84) 94 (85)

Intestinal helminth during childhood 6 (25) 4 (16) 23 (21)Schistosome infection during childhood 0 (0) 1 (4) 9 (8)

NOTE. Data are no. (%) of values, unless otherwise noted. PCR, polymerase chain reaction.

ceased in 2002. Reasons for dropout and exclusion from anal-

ysis were death, emigration from the study area, and failure to

attend at least 2 follow-up visits. The mean duration of follow-

up was 4.3 years (range, 3.5–7.0 years).

Newborns were categorized according to the infection status

of their mothers and whether cord blood filarial antigen–driven

cytokine responses were detected (figure 1 and table 1). Of the

49 newborns whose mothers had indicators of active W. ban-

crofti infection, 24 had prima facie evidence of in utero sen-

sitization—that is, cord blood lymphocyte IL-5, IL-13, IFN-g,

and/or IL-2 responses to filarial antigen. The remaining 25

newborns of W. bancrofti–infected mothers were categorized as

tolerant—that is, cord blood lymphocyte antigen–driven cyto-

kine responses were not detected despite coexisting maternal

filarial infection. Cytokine responses by sensitized newborns in-

dicated a mixed type 2 (IL-5 and/or IL-13)/type 1 (IFN-g and/

or IL-2) immunity that did not correlate with the intensity of

maternal filarial infection (table 2).

We examined whether maternal infection intensity was as-

sociated with neonatal sensitization versus tolerance. Newborns

categorized as tolerant had mothers with higher infection bur-

dens, compared with those categorized as sensitized—that is,

the proportion of microfilaremic mothers was greater in the

former than in the latter group (13/25 vs. 7/24), whereas the

opposite relationship existed when maternal BmA-specific IgG4

alone was the indicator of maternal infection (4/25 vs. 9/24).

The 110 newborns whose mothers did not have active W. ban-

crofti infection—that is, those negative for BmA-specific IgG4

antibody, filarial antigememia, and microfilaremia—were cat-

egorized as unexposed and, therefore, neither sensitized nor

tolerant. Cord blood lymphocytes from 106 (96%) of these

donors did not produce any cytokine in response to BmA or

microfilarial antigen. Three of the 4 newborns with filarial an-

tigen–driven responses had mothers with schistosomiasis ac-

companied by cord blood lymphocyte responses to schistosome

antigen. The 4 children whose lymphocytes responded to filarial

antigens but who did not have mothers who were infected with

or exposed to W. bancrofti were excluded from the category of

“sensitized infants” in subsequent analyses. There were no sig-

nificant differences among the 3 groups in maternal or child-

hood coinfections with malaria parasites, intestinal helminths,

or schistosomes (table 1).

To assess the impact of neonatal immunity on the evolution

of filarial T cell immunity during childhood, we evaluated cy-

tokine responses to BmA and microfilarial antigen annually.

The average duration of follow-up for the sensitized, tolerant,

and unexposed groups was similar (figure 1). The number of

children examined each year and the proportion in each group

with T cell cytokine responses to filarial antigen are shown in

table 3. Eighty-three percent of sensitized newborns (20/24)

had a recall response to filarial antigen at age 1 year, compared

with 12% (3/25) and 18% (20/110) in the tolerant and unex-

posed groups, respectively ( ). Filarial antigen–drivenP ! .001

cytokine responses persisted in the sensitized group (i.e., re-

sponses were observed for 75%–89% of children from age 2

through 7). In contrast, lymphocyte reactivity to filarial antigen

increased gradually with age in the other 2 groups (i.e., from

18%–66% at age 2–7 in the unexposed group). To examine the

specificity of these responses for filarial antigen, we performed

a similar analysis of PPD-driven IFN-g production (all partic-

ipants were given bacille Calmette-Guerin at birth, in accor-

dance with the national policy in Kenya). The proportion of

responders was similar in the 3 groups—in the sensitized group,

at age 1, 2, 3, 4, 5, 6, and 7 years, 87%, 91%, 94%, 79%, 83%,

Dow




Table 2. Filarial antigen–driven cytokine production by T cells of newborns with evi-dence of in utero sensitization to Wuchereria bancrofti.

Subject

Filarial antigen–drivencytokine production

by cord blood Maternal infection indicators

IL-5 (IL-13),pg/mL

IFN-g (IL-2),pg/mL

BmA-specific IgG4,U/mLa

Og4C3,U/mLb Microfilaria/mL

65 (47) (417) 225 796 152505 (233) 0 373 401 77339 (201) 0 1462 507 51208 (103) (56) 1239 367 29317 242 75 4492 337 7418 (188) 101 523 902 5631 45 (137) 1341 1121 228 56 2850 397 1110 0480 118 152 0 290 0553 129 (130) 728 288 035 31 183 271 210 0171 153 86 1500 191 0311 (45) 0 67 146 0179 239 0 487 96 053 55 (29) 632 80 064 0 (43) 398 0 0165 43 666 1010 0 0617 77 129 966 0 0354 139 936 988 0 0403 (143) 162 470 0 0313 53 36 318 0 0337 (69) (111) 262 0 0326 61 0 253 0 0356 0 350 232 0 0

NOTE. Net filarial antigen (BmA)–driven production of interferon (IFN)–g, interleukin (IL)–2, IL-13, andIL-5 (pg/mL) by cord blood lymphocytes was measured as described in Subjects, Materials, and Methods.Values in parentheses denote the level of IL-2 or IL-13 production, which was greater than that of IFN-gor IL-5 production for these subjects.

a A positive response was considered to occur at 1220 U/mL, a cutoff based on the mean plus 3 SDsof women residing in the Turkana region of Kenya, where lymphatic filariasis is not endemic.

b A circulating Og4C3 antigen level 132 U/mL was considered to be positive.

84%, and 92% of children, respectively, responded; in the tol-

erant group, at the same ages, 72%, 84%, 79%, 67%, 77%,

81%, and 73% of children, respectively, responded; and, in the

unexposed group, at the same ages, 77%, 81%, 83%, 78%, 79%,

81%, and 86% of children, respectively, responded ( ).P p .3

To evaluate the impact of prenatal immunity on cytokine re-

sponses during childhood, we compared filarial antigen–driven

IL-5 (type 2 cytokine) and IFN-g (type 1 cytokine) responses in

the 3 groups. IL-5 was dominant in the sensitized group, with

an age-related increase in production of this cytokine (figure 2,

upper panel). In contrast, the proportion of children with IL-5

responses in the tolerant and unexposed groups was lower at

each year of follow-up. The pattern of filarial antigen–driven

IFN-g responses (figure 2, lower panel) was different than that

of IL-5. First, production of this type 1 cytokine was minimal

at ages 1 year and 2 years in all 3 groups. Second, there was a

tendency for stronger responses beginning at age 3 years, espe-

cially among sensitized and unexposed newborns. Third, from

age 3 through 7 years, tolerant newborns had significantly weaker

IFN-g responses than did the other 2 groups. The pattern of IL-

13 and IL-2 production was similar to that of IL-5 and IFN-g

production, respectively. Microfilarial antigen–driven cytokine

responses were similar to those elicited by BmA (data not shown).

To determine whether neonatal immunity associated with

maternal filariasis influenced infection susceptibility, we com-

pared the risk of acquiring W. bancrofti infection by offspring

of infected versus uninfected mothers and of newborns cate-

gorized as sensitized, tolerant, or unexposed. The 49 children

whose mothers had filariasis included a higher proportion who

became infected between ages 1 year and 7 years, compared

with the 110 children of mothers without filariasis, but these

values were not significantly different ( [1 df];2x p 2.55 P p

; hazard ratio for children of infected vs. uninfected women,.10

0.941 [95% confidence interval {CI}, 0.495–1.813]) (figure 3,

Dow




Table 3. Filarial antigen–driven cytokine responses at ages 1 year to 7 years, in children classified at birthas sensitized, tolerant, or unexposed (no active maternal filarial infection during gestation).

Age,years

Sensitized Tolerant Unexposed

Totalexamined

With filarialantigen–driven

cytokine responseTotal

examined


cytokine responseTotal

examined


cytokine response

1 24 20 (83) 25 3 (12) 110 20 (18)2 21 18 (86) 21 2 (10) 106 24 (23)3 22 18 (82) 22 2 (9) 102 37 (36)4 20 16 (80) 20 3 (15) 88 36 (41)5 18 16 (89) 20 5 (25) 80 41 (51)6 16 14 (88) 17 6 (35) 75 35 (47)7 16 12 (75) 13 7 (47) 62 41 (66)

NOTE. Data are no. (%) of children.

upper panel). Newborns with evidence of filarial-specific im-

mune tolerance had a 12.9-fold greater risk of infection, com-

pared with those with evidence of sensitization ( [12x p 9.43

df] ; ; 95% CI, 2.5–107.2-fold). The tolerant group wasP ! .001

also 4.8 times more likely to become infected, compared with

newborns born to uninfected women ( [1 df];2x p 11.88 P !

; 95% CI, 1.7–12.9-fold). There was no difference in the.001

risk of infection between the sensitized and the nonsensitized

unexposed groups ( [1 df]; ) (figure 3, lower2x p 0.81 P p .33

panel).

To examine whether differences in infection susceptibility

were related to exposure to mosquito-borne infective larvae,

we measured IgG and IgE antibodies to infective larvae annual-

ly [20]. A positive response indicating exposure to mosquito-

borne filarial larvae was scored when the ELISA optical density

was greater than the mean plus 3 SDs of control serum samples

from children residing in the Turkana region in Kenya. Intes-

tinal helminth infections and echinococcosis are endemic, but

lymphatic filariasis is absent, in Turkana. No differences among

the groups were observed. The proportions of children who

were IgG and/or IgE antibody positive at ages 1 year through

7 years were 8%, 10%, 14%, 25%, 45%, 47%, and 62%, re-

spectively, in the tolerant group; 4%, 10%, 14%, 25%, 33%,

38%, and 44%, respectively, in the sensitized group; and 3%,

5%, 10%, 13%, 19%, 28%, and 41%, respectively, in the unex-

posed group ( ). Clustering of infection-positive childrenP 1 .1

according to village or household of residence was not observed

(table 1).

DISCUSSION

Our results show that in utero acquisition of T cell immunity

to filarial antigens has a major impact on susceptibility to in-

fection with W. bancrofti during childhood. Prenatal filarial-

specific immune tolerance as a consequence of active maternal

filariasis increased the risk of infection during the first 7 years

after birth 12.9-fold, compared with that in newborns with

evidence of in utero sensitization, and 4.8-fold, compared with

that in newborns who were neither tolerant nor sensitized,

because their mothers lacked active W. bancrofti infection. Con-

versely, newborns with prenatal immune sensitization marked

by filarial-responsive T cells at birth maintained this immunity

throughout childhood and had reduced susceptibility to infec-

tion, suggesting that prenatal lymphocyte priming confers par-

tial protection against infective larvae or the establishment of

adult worms.

Susceptibility to filarial infection may be affected by factors in

addition to the antenatal immune mechanisms examined here.

First, heterogeneity in exposure to infective larvae could account

for differences in infection outcome. It is not possible to directly

measure the cumulative exposure of an individual to mosquitoes

bearing infective larvae. We used the presence of IgE and IgG

antibodies to infective larvae as a surrogate measure of exposure.

The proportion of children in each group who developed such

antibody responses over time was similar. Second, coinfection

with malaria or intestinal worms could induce bias in T cell

cytokine responses and thereby affect susceptibility to W. bancrofti

infection. Malaria, geohelminth, and schistosome infection rates

were similar among the 3 groups. Third, genetic differences may

affect susceptibility to lymphatic filariasis [22–24], but, to our

knowledge, there are no data demonstrating the contribution of

specific genes to infection susceptibility in humans.

The differences in cellular immune responses to filarial an-

tigens observed here suggest that several mechanisms may con-

tribute to the control of susceptibility to infection. With respect

to children who were relatively resistant to infection, prenatal

priming generated a population of filarial-specific lymphocytes

that persisted from birth through age 7 years. The dominant

cytokine was IL-5 (IL-13 followed a similar pattern but was

detected in a lower proportion of participants). These children

produced little filarial antigen–driven IFN-g and IL-2, even

though their T cells at birth produced both type 2 (IL-5 and/

or IL-13) and type 1 (IFN-g and/or IL-2) cytokines. This result

Dow




Figure 2. Age-related filarial antigen–driven interleukin (IL)–5 and in-terferon (IFN)–g production by peripheral blood mononuclear cells fromchildren with evidence of in utero sensitization, tolerance, or neither(unexposed). Peak levels of IL-5 (upper panel) and IFN-g (lower panel)are shown for cells cultured with filarial antigen (BmA) at 48 or62 � 1096 h. Bars represent net geometric mean cytokine production (cell cultureswith antigen minus media alone) � SE. Asterisks indicate significantdifferences ( , Mann-Whitney U test) between the sensitized groupP ! .05and either the tolerant group or the unexposed group (upper panel) andbetween the tolerant group and either the sensitized group or the unex-posed group (lower panel).

Figure 3. Kaplan-Meier time-to-infection survival curves for childrenof filarial-infected mothers (solid line) vs. children of uninfected mothers(dashed line) (A) and children with evidence of in utero sensitization(dotted line), in utero tolerance (solid line), or neither (unexposed; dashedline) (B). Time to infection for the 2 groups in panel A did not differ( [1 df ]; hazard ratio, 0.941 [95% confidence interval, 0.495–2x p 2.551.813]). The difference in time to infection among the 3 groups in panelB was significant ( ), as was the comparison between childrenP ! .0001with evidence of in utero sensitization and those with in utero tolerance( ) and between children with in utero tolerance and those whoP ! .001were unexposed ( ). There was not a significant difference be-P ! .001tween children with evidence of in utero sensitization and those whowere unexposed in utero ( ).P 1 .10

suggests that filarial-specific lymphocytes with a predominant

type 2 phenotype are more likely to survive into childhood. This

phenomenon has also been observed following fetal exposure

to allergens [25]. The persistence of filarial antigen–driven IL-5

among putatively resistant children implicates a protective role

of this cytokine. This conclusion is supported by observations

in mouse models of filariasis. Mice deficient in IL-5 were more

susceptible to filarial infection than were wild-type controls [26,

27], and IL-5–transgenic mice were more resistant than were

normal mice [28].

The lack of filarial antigen–specific responses in some children

of filarial-infected women represents either immune tolerance or

a lack of fetal exposure to filarial antigens. We believe that im-

mune tolerance is the most likely explanation, because mothers

of tolerant newborns had higher infection burdens than did

mothers whose newborns had evidence of lymphocyte priming.

Additionally, the proportion of children in the tolerant group

with recall responses and the amount of IFN-g produced were

less than those in the other groups, indicating a reduced fre-

quency of antigen-reactive cells that may arise from clonal de-

letion or T cell anergy [29]. Alternatively, regulatory T cells pro-

ducing IL-10 and tumor growth factor–b may be expanded [30–

32]. In either case, impaired IFN-g responses correlated with

increased susceptibility to infection during childhood (figure 2).

Strong filarial antigen–driven IFN-g production is characteristic

of putatively immune, “endemic normal” subjects and has been

suggested to be linked with protection against filariasis [33, 34].

Dow




Humoral immunity to infective-stage larvae has been impli-

cated in protection in animal models of filarias [35, 36]. Prelim-

inary data indicate that children with evidence of in utero sen-

sitization have increased antilarval IgE antibodies, compared with

IgG4 antibodies, whereas the reverse relationship exists in chil-

dren who become tolerant in utero. A similar relationship be-

tween larval-specific IgE and IgG4 has been suggested to correlate

with protection against human lymphatic filariasis [37]. Elevated

antiparasite IgE:IgG4 ratios have been associated with acquired

resistance in human schistosomiasis [38, 39].

The mechanisms by which the fetus is exposed to filarial an-

tigen in utero and the factors determining whether such exposure

results in priming or tolerance remain poorly understood. Mi-

crofilariae rarely cross the placenta [40, 41], and one report sug-

gested that filarial antigen is found in 10% of cord blood samples

from newborns born to mothers with filariasis [42]. However,

we have not detected filarial antigen in cord blood, using the

Og4C3 or other antigen-capture assays [43]. Factors altering the

integrity of the placenta (such as damage resulting from placental

malaria, which is coendemic in the area) may affect the time

and efficiency of passage of filarial antigen from the maternal to

the fetal circulation during gestation [44].

A practical implication of the present study relates to the

potential impact that mass treatment programs aimed at elim-

inating transmission of W. bancrofti [45] may have on prenatal

immune priming and tolerance. On the one hand, reduction

of filarial infection burdens in women of childbearing age may

diminish prenatal immune tolerance and, thus, decrease the

risk of infection during childhood. On the other hand, the risk

of developing lymphatic pathological abnormalities might in-

crease if transmission is not stopped and if exposure to infective

larvae is delayed to later in life. With the recent commencement

of a mass treatment program in Kenya, continued monitoring

of both the current and a newly established birth cohort may

help to address some of these questions.

Acknowledgments

We appreciate the assistance of Adams Omollo, Kephar Otieno, andElton K. Mzungu, for technical help, and of Grace Watutu, for data entry.We are grateful to the maternity nurses and the women and children inKenya who participated in this study.

References

1. Michael E, Bundy DA, Grenfell BT. Re-assessing the global prevalenceand distribution of lymphatic filariasis. Parasitology 1996; 112(Pt 4):409–28.

2. Ottesen EA. The Wellcome Trust Lecture. Infection and disease inlymphatic filariasis: an immunological perspective. Parasitology 1992;104(Suppl):S71–9.

3. Kazura JW, Bockarie M, Alexander N, et al. Transmission intensity andits relationship to infection and disease due to Wuchereria bancrofti inPapua New Guinea. J Infect Dis 1997; 176:242–6.

4. Tobian AA, Tarongka N, Baisor M, Bockarie M, Kazura JW, King CL.Sensitivity and specificity of ultrasound detection and risk factors forfilarial-associated hydroceles. Am J Trop Med Hyg 2003; 68:638–42.

5. Choi EH, Nutman TB, Chanock SJ. Genetic variation in immune func-tion and susceptibility to human filariasis. Expert Rev Mol Diagn2003; 3:367–74.

6. Dreyer G, Noroes J, Figueredo-Silva J, Piessens WF. Pathogenesis oflymphatic disease in bancroftian filariasis: a clinical perspective. Par-asitol Today 2000; 16:544–8.

7. Nutman TB, Kumaraswami V. Regulation of the immune response inlymphatic filariasis: perspectives on acute and chronic infection withWuchereria bancrofti in South India. Parasite Immunol 2001; 23:389–99.

8. Alexander ND, Kazura JW, Bockarie MJ, et al. Parental infection con-founded with local infection intensity as risk factors for childhood mi-crofilaraemia in bancroftian filariasis. Trans R Soc Trop Med Hyg 1998;92:23–4.

9. Lammie PJ, Hitch WL, Walker Allen EM, Hightower W, Eberhard ML.Maternal filarial infection as risk factor for infection in children. Lancet1991; 337:1005–6.

10. Malhotra I, Ouma JH, Wamachi A, et al. Influence of maternal filariasison childhood infection and immunity to Wuchereria bancrofti in Kenya.Infect Immun 2003; 71:5231–7.

11. Partono F, Pribadi PW, Soewarta A. Epidemiological and clinical featuresof Brugia timori in a newly established village: Karakuak, West Flores,Indonesia. Am J Trop Med Hyg 1978; 27:910–5.

12. Partono F, Purnomo. Clinical features of timorian filariasis amongimmigrants to an endemic area in West Flores, Indonesia. SoutheastAsian J Trop Med Public Health 1978; 9:338–43.

13. Dissanayake S, de Silva LV, Ismail MM. IgM antibody to filarial antigensin human cord blood: possibility of transplacental infection. Trans RSoc Trop Med Hyg 1980; 74:542–4.

14. King CL, Malhotra I, Mungai P, et al. B cell sensitization to helminthicinfection develops in utero in humans. J Immunol 1998; 160:3578–84.

15. Malhotra I, Ouma J, Wamachi A, et al. In utero exposure to helminthand mycobacterial antigens generates cytokine responses similar to thatobserved in adults. J Clin Invest 1997; 99:1759–66.

16. Weil GJ, Hussain R, Kumaraswami V, Tripathy SP, Phillips KS, OttesenEA. Prenatal allergic sensitization to helminth antigens in offspring ofparasite-infected mothers. J Clin Invest 1983; 71:1124–9.

17. Steel C, Guinea A, McCarthy JS, Ottesen EA. Long-term effect of prenatalexposure to maternal microfilaraemia on immune responsiveness to fi-larial parasite antigens. Lancet 1994; 343:890–3.

18. More SJ, Copeman DB. A highly specific and sensitive monoclonalantibody-based ELISA for the detection of circulating antigen in ban-croftian filariasis. Trop Med Parasitol 1990; 41:403–6.

19. Eberhard ML, Lammie PJ. Laboratory diagnosis of filariasis. Clin LabMed 1991; 11:977–1010.

20. Helmy H, Weil GJ, Faris R, et al. Human antibody responses to Wucher-eria bancrofti infective larvae. Parasite Immunol 2000; 22:89–96.

21. Malhotra I, Mungai P, Muchiri E, et al. Distinct Th1- and Th2-typeprenatal cytokine responses to Plasmodium falciparum erythrocyte in-vasion ligands. Infect Immun 2005; 73:3462–70.

22. Terhell AJ, Houwing-Duistermaat JJ, Ruiterman Y, Haarbrink M, AbadiK, Yazdanbakhsh M. Clustering of Brugia malayi infection in a com-munity in South-Sulawesi, Indonesia. Parasitology 2000; 120:23–9.

23. Romia SA, el-Ganayni GA, Makhlouf LM, Handousa AE. HLA antigensand blood groups in bancroftian filariasis. J Egypt Soc Parasitol 1988;18:211–20.

24. Choi EH, Zimmerman PA, Foster CB, et al. Genetic polymorphismsin molecules of innate immunity and susceptibility to infection withWuchereria bancrofti in South India. Genes Immun 2001; 2:248–53.

25. Prescott SL, Macaubas C, Smallacombe T, Holt BJ, Sly PD, Holt PG.Development of allergen-specific T-cell memory in atopic and normalchildren. Lancet 1999; 353:196–200.

26. Le Goff L, Loke P, Ali HF, Taylor DW, Allen JE. Interleukin-5 is essentialfor vaccine-mediated immunity but not innate resistance to a filarialparasite. Infect Immun 2000; 68:2513–7.

Dow




27. Volkmann L, Bain O, Saeftel M, et al. Murine filariasis: interleukin 4and interleukin 5 lead to containment of different worm developmentalstages. Med Microbiol Immunol (Berl) 2003; 192:23–31.

28. Martin C, Al-Qaoud KM, Ungeheuer MN, et al. IL-5 is essential forvaccine-induced protection and for resolution of primary infection inmurine filariasis. Med Microbiol Immunol (Berl) 2000; 189:67–74.

29. King CL, Kumaraswami V, Poindexter RW, et al. Immunologic tolerancein lymphatic filariasis: diminished parasite-specific T and B lymphocyteprecursor frequency in the microfilaremic state. J Clin Invest 1992; 89:1403–10.

30. Babu S, Kumaraswami V, Nutman TB. Transcriptional control of im-paired Th1 responses in patent lymphatic filariasis by T-box expressedin T cells and suppressor of cytokine signaling genes. Infect Immun2005; 73:3394–401.

31. Gillan V, Devaney E. Regulatory T cells modulate Th2 responses in-duced by Brugia pahangi third-stage larvae. Infect Immun 2005; 73:4034–42.

32. Taylor MD, LeGoff L, Harris A, Malone E, Allen JE, Maizels RM.Removal of regulatory T cell activity reverses hyporesponsivenessand leads to filarial parasite clearance in vivo. J Immunol 2005; 174:4924–33.

33. Dimock KA, Eberhard ML, Lammie PJ. Th1-like antifilarial immuneresponses predominate in antigen-negative persons. Infect Immun1996; 64:2962–7.

34. Ravindran B, Satapathy AK, Sahoo PK, Mohanty MC. Protective im-munity in human lymphatic filariasis: problems and prospects. MedMicrobiol Immunol (Berl) 2003; 192:41–6.

35. Paciorkowski N, Shultz LD, Rajan TV. Primed peritoneal B lymphocytesare sufficient to transfer protection against Brugia pahangi infection inmice. Infect Immun 2003; 71:1370–8.

36. Partono F, Oemijati S. The association of clinical filariasis and Wucher-

eria bancrofti infections in Jakarta. Southeast Asian J Trop Med PublicHealth 1978; 9:260–3.

37. Nielsen NO, Bloch P, Simonsen PE. Lymphatic filariasis-specific im-mune responses in relation to lymphoedema grade and infection status.II. Humoral responses. Trans R Soc Trop Med Hyg 2002; 96:453–8.

38. Demeure CE, Rihet P, Abel L, Ouattara M, Bourgois A, Dessein AJ.Resistance to Schistosoma mansoni in humans: influence of the IgE/IgG4 balance and IgG2 in immunity to reinfection after chemothera-py. J Infect Dis 1993; 168:1000–8.

39. Hagan P, Blumenthal UJ, Dunn D, Simpson AJ, Wilkins HA. HumanIgE, IgG4 and resistance to reinfection with Schistosoma haematobium.Nature 1991; 349:243–5.

40. Campello TR, Ferreira RS, Pires ML, et al. A study of placentas fromWuchereria bancrofti microfilaraemic and amicrofilaraemic mothers. JTrop Med Hyg 1993; 96:251–5.

41. Eberhard ML, Hitch WL, McNeeley DF, Lammie PJ. Transplacental trans-mission of Wuchereria bancrofti in Haitian women. J Parasitol 1993; 79:62–6.

42. Hitch WL, Eberhard ML, Lammie PJ. Investigation of the influenceof maternal infection with Wuchereria bancrofti on the humoral andcellular responses of neonates to filarial antigens. Ann Trop Med Par-asitol 1997; 91:461–9.

43. Ramzy RM, Gad AM, Faris R, Weil GJ. Evaluation of a monoclonal-antibody based antigen assay for diagnosis of Wuchereria bancrofti in-fection in Egypt. Am J Trop Med Hyg 1991; 44:691–5.

44. Crocker IP, Tanner OM, Myers JE, Bulmer JN, Walraven G, Baker PN.Syncytiotrophoblast degradation and the pathophysiology of the ma-laria-infected placenta. Placenta 2004; 25:273–82.

45. Ottesen EA. The global programme to eliminate lymphatic filariasis.Trop Med Int Health 2000; 5:591–4.

Dow



Prenatal T Cell Immunity to Wuchereria bancrofti and Its Effect ...

Documents