Differential Cytokine and Antibody Responses to Adult and Larval Stages of Onchocerca volvulus Consistent with the Development of Concomitant Immunity

Abstract

The possibility of concomitant immunity and its potential mechanisms in Onchocerca volvulus infection were examined by analyzing cytokine and antibody responses to infective larval (third-stage larvae [L3] and molting L3 [mL3]), adult female worm (F-OvAg), and skin microfilaria (Smf) antigens in infected individuals in a region of hyperendemicity in Cameroon as a function of age. Peripheral blood mononuclear cell interleukin 5 (IL-5) responses to F-OvAg and Smf declined significantly with age (equivalent to years of exposure to O. volvulus). In contrast, IL-5 secretion in response to L3 and mL3 remained elevated with increasing age. Gamma interferon responses to L3, mL3, and F-OvAg were low or suppressed and unrelated to age, except for responses to Smf in older subjects. IL-10 levels were uniformly elevated, regardless of age, in response to L3, mL3, and F-OvAg but not to Smf, for which levels declined with age. A total of 49 to 60% of subjects had granulocyte-macrophage colony-stimulating factor responses to all O. volvulus antigens unrelated to age. Analysis of levels of stage-specific immunoglobulin G3 (IgG3) and IgE revealed a striking, age-dependent dissociation between antibody responses to larval antigens (L3 and a recombinant L3-specific protein, O. volvulus ALT-1) which were significantly increased or maintained with age and antibody responses to F-OvAg, which decreased. Levels of IgG1 to L3 and F-OvAg were elevated regardless of age, and levels of IgG4 increased significantly with age, although not to O. volvulus ALT-1, which may have unique L3-specific epitopes. Immunofluorescence staining of whole larvae showed that total anti-L3 immunoglobulin levels also increased with the age of the serum donor. The separate and distinct cytokine and antibody responses to adult and infective larval stages of O. volvulus which are age related are consistent with the acquisition of concomitant immunity in infected individuals.

The filarial parasite Onchocerca volvulus infects about 18 million people, and a further 100 million live in areas in which O. volvulus is endemic in Africa and Latin America. The resulting disease, onchocerciasis, is characterized by severe dermatitis and blindness (21). There is epidemiological evidence that acquired immunity against O. volvulus infection occurs in humans. For example, in regions of high endemicity, despite constant exposure to infected Simulium flies, 1 to 5% of the population exhibits no clinical manifestations of disease. These individuals are considered to be immune to infection and are referred to as putatively immune (PI) (15, 17, 46, 47). Furthermore, in chronically infected (INF) individuals, the number of skin microfilariae (mf) tends to level off between the ages of 20 and 40 years, suggesting that these individuals have developed a means of limiting acquired infections (11). It has been suggested that the means of limiting acquired infections is through concomitant immunity (36), whereby newly introduced infective-stage larvae (third-stage larvae [L3]) are eliminated while adult worms and mf are left unaffected. In individuals infected with lymphatic filariae, concomitant immunity was clearly defined (34) and was associated with stage-specific immune responses. Levels of antibodies against the infective Wuchereria bancrofti L3 increase with duration of exposure (7), and there are differences in the classes and isotypes of antibody responses to adult versus larval antigens of Brugia malayi (27). Because of the difficulty in obtaining O. volvulus L3, previous studies have been limited to adult or microfilarial extracts, and data on responses specific to the infective stages are lacking. Antigens from L3 and subsequent developmental stages of molting L3 (mL3) have been shown to be promising sources for protective antigens and targets for the control of filarial infections (8, 13, 30). It was previously shown that PI individuals express enhanced Th2 and Th1 responses to larval antigens (46); however, the mechanisms that may allow INF individuals to limit new infections are not known.

The results presented here represent the first comparative analysis of cellular and antibody responses in infected individuals of a broad range of ages to antigens of O. volvulus L3 and mL3 and stages associated with the establishment of patent infection, the adult female and skin mf.

MATERIALS AND METHODS

Study population.

The study was performed in the Kumba region, an area of hyperendemicity for onchocerciasis in southwest Cameroon. The individuals who consented to participate in the study were born or had resided for more than 10 years in villages around Kumba: Marumba I, Marumba II, Boa Bakundu, Bombanda, and Bombele. The individuals were screened and tested for the presence of mf in their skin snips and clinical symptoms of disease, such as dermatitis, nodules, and ocular lesions. None of the subjects had received ivermectin treatment prior to the study. Four skin snips were collected from each individual, and the average of the mf counts from the four snips was used in estimating the skin mf densities. During the screening process (46), we identified 168 individuals who were skin mf negative (mf−). To confirm their infectious status, their biopsy specimens were tested for the presence of a tandem repeat DNA specific for O. volvulus. The 150-mer DNA repeats were amplified by PCR and identified by Southern blotting with a specific internal O. volvulus probe (35). This confirmatory test resulted in a subgroup of 117 individuals who were mf− but PCR positive and thus considered infected but with an mf count of zero at the time of sample collection. Thirty two of the mf− PCR-positive individuals as well as 116 individuals having various skin mf densities were included in the present study group, which was comprised of 59 males and 89 females having skin mf counts ranging from 0 to 230 and a median age of 14 years (range, 3 to 62 years).

Antigens.

All parasite material was collected in our research facility at the Tropical Medicine Research Station, Kumba, Cameroon. Crude antigen extracts were prepared from different stages of O. volvulus, including L3 (obtained from infected black flies), mL3, adult female worms, and skin mf. mL3 was generated by incubating L3 in vitro in a 1:1 mixture of Iscove modified Dulbecco medium and NCTC-135-20% fetal calf serum-antibiotic-antimycotic solution (GIBCO BRL Life Technologies, Gaithersburg, Md.) for 3 days at 37°C. Larvae were collected after 1, 2, or 3 days in culture, washed in phosphate-buffered saline (PBS), and quick-frozen in N₂. Ultrastructural examination of such larvae by electron microscopy confirmed that the cultured larvae had started the molting process, as the separation between the cuticle of L3 and the newly synthesized cuticle of the fourth-stage larvae was evident in the cross sections. The mL3 antigen preparation was made from a pooled mixture of similar numbers of larvae that were collected on day 1, 2, or 3 of culturing.

Crude L3, mL3, adult female worm (F-OvAg), and skin microfilaria (Smf) antigens were prepared as described before (46). Briefly, the worms were ground to a powder by using a Bessman tissue pulverizer (Spectrum Lab Products, Houston, Tex.) and further disrupted by sonication before extraction in PBS containing 10 mM 3-[3-cholamidopropyl)-dimethylammonio]-2-hydroxy-1-propanesulfonate (Calbiochem, La Jolla, Calif.) and protease inhibitors (Sigma, St. Louis, Mo.; 2 mM phenylmethylsulfonyl fluoride, 0.2 mM Nα-p-tosyl-l-lysine chloromethyl ketone, 0.2 nM N-tosyl-l-phenylalanine chloromethyl ketone, 25 μg of leupeptin/ml, and 10 mM EDTA). The insoluble material was extracted twice in the same buffer for 12 h at 4°C. The pooled soluble extracts of each stage-specific preparation were then dialyzed against PBS, centrifuged at 4°C, and filter sterilized. All antigens tested negative for lipopolysaccharide (E-toxate assay; Sigma)

Lymphocyte stimulation.

Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation over Ficoll (Sigma). Proliferative assays with PBMCs (from 116 individuals) were done as described previously (14) by using, per 0.2-ml well, 2 × 10⁵ cells in RPMI 1640 medium containing 10% human AB serum, 25 mM HEPES, 2 mM l-glutamine, and 0.5 mg of gentamicin (Bio-Whittaker, Walkersville, Md.)/ml. For cytokine production, cells (from 73 individuals) were cultured at 2 × 10⁶/ml in RPMI 1640 medium containing 10% fetal calf serum, 25 mM HEPES, 2 mM l-glutamine, and 0.5 mg of gentamicin/ml. The cells were cultured for 2 or 5 days in the presence of the following final concentrations of antigen preparations: F-OvAg at 5 μg/ml, L3 at 0.5 μg/ml (equivalent to 50 L3 parasites/ml), mL3 at 0.36 μg/ml (equivalent to 50 mL3 parasites/ml), and Smf at 0.25 μg/ml (equivalent to 500 Smf parasites/ml). The antigens were used at concentrations determined to give optimal responses in infected individuals; similar antigen concentrations induced significant proliferation in infected individuals in comparison to control individuals (data not shown).

Proliferative responses to nonparasite antigens were obtained by using Streptolysin-O (SLO; 1:200; Difco, Detroit, Mich.) and pokeweed mitogen (1:200; Sigma). Cytokine responses to nonparasite antigens also were obtained by using SLO (1:100). In addition, PBMCs were cultured in the presence of a mitogenic stimulus of phorbol myristate acetate (50 ng/ml) plus ionomycin (1 μg/ml) (Calbiochem) (14). For proliferation, after 5 days cells were pulsed for 4 h with 1 μCi of [³H]thymidine (DuPont, Wilmington, Del.) before being harvested onto glass filters. The incorporation of thymidine was measured by liquid scintillation spectrophotometry. Data are expressed as the stimulation index, obtained by dividing the counts per minute (mean of triplicate cultures) for the antigen- or the mitogen-stimulated cultures by the mean counts per minute for the unstimulated control cultures. For cytokine production, the supernatants were harvested at day 5 (an optimal time point which we previously established) and stored at −70°C until analyzed for interleukin 5 (IL-5) (marker for Th2 phenotype), gamma interferon (IFN-γ) (marker for Th1 phenotype), IL-10, and granulocyte-macrophage colony-stimulating factor (GM-CSF) production. IL-5, IFN-γ, and GM-CSF were measured by using commercial sandwich enzyme-linked immunosorbent assay (ELISA) kits (R & D Systems, Minneapolis, Minn.), and IL-10 was measured by using an OptEIA kit according to the manufacturer's protocol (BD Pharmingen, San Diego, Calif.). Cytokine levels produced by 10⁶ PBMCs were expressed in picograms per milliliter, and the net antigen-specific production of a cytokine was calculated by subtracting the quantity of the cytokine produced by PBMCs cultured without antigen from that of the cytokine produced by PBMCs cultured with a specific antigen.

Antibody analysis.

Sera obtained from O. volvulus-infected individuals were analyzed for antibody to crude F-OvAg and L3 antigens by an established ELISA (26). In addition to crude extracts from different stages of the parasite, we included one recombinant protein, rOv-ALT-1, that has been well characterized in our laboratory. The rOv-ALT-1 transcript is an upregulated, L3 stage-specific transcript, and the protein is expressed only in L3 and during the transition from L3 to mL3 (26). Moreover, the recombinant protein induces significant protection in mice when used in combination with alum (1).

For immunoglobulin G (IgG) isotype ELISAs, antigens (F-OvAg, 5 μg/ml; L3, 2 μg/ml; rOv-ALT-1, 1 μg/ml; and glutathione S-transferase [GST], 1 μg/ml) were used to coat the wells of ELISA plates, and sera at a 1:200 dilution were reacted with the bound antigens. For the analysis of IgE levels, plates were coated with 10 μg of F-OvAg, rOv-ALT-1, or the control (GST)/ml and 5 μg of L3 extract/ml, and serum samples were preabsorbed with protein G-Sepharose (Pharmacia) before being used at a 1:20 dilution. For IgG subclass responses, the bound antibodies were detected by using a 1:1,000 dilution of monoclonal antibodies against different human subclasses (Hybridoma Reagent Laboratory, Kingsville, Md.). This step was followed by incubation with a 1:1,250 dilution of horseradish peroxidase-conjugated rabbit anti-mouse immunoglobulins (Kierkegaard & Perry Laboratories, Inc., Gaithersburg, Md.). IgE in the sera was detected by using an horseradish peroxidase-conjugated, ɛ-chain-specific anti-human IgE monoclonal antibody (Zymed, San Francisco, Calif.) at a 1:750 dilution. Tetramethylbenzidine (Sigma) was used as the substrate for all ELISAs, and the optical density (OD) was read at 450 nm. The OD values for rOv-ALT-1 are the net values after subtraction of the OD values for the control (GST).

IFA.

The antibody response to the surface of O. volvulus L3 and skin mf was measured by an indirect immunofluorescence assay (IFA) with cryopreserved larvae and mf collected in Kumba, Cameroon. Serum samples were randomly selected from individuals of a broad range of ages (5 to 45 years). The total level of binding was determined initially for 20 individuals and later against L3 only for an additional 14 individuals. Scoring the intensity of the fluorescence was done in a semiquantitative manner as described by Kurniawan-Atmadja et al. (27) and Helmy et al. (22). Briefly, 25 to 50 live L3 parasites or 100 mf were incubated with sera at a 1:20 dilution for 1 h in PBS, followed by washing and incubation with a 1:15 dilution of fluorescein isothiocyanate-conjugated rabbit anti-human immunoglobulins (Dako; F200). After additional washings, the larvae were mounted for viewing under epifluorescence illumination (Zeiss Axioskop) at a magnification of ×10.

Statistical analysis.

Spearman's rank correlation test was used to test the significance of the correlation between the age or the skin mf counts of the individuals and their antibody titers or the cytokine quantities produced by their PBMCs (expressed as the correlation coefficient, r). Fisher's exact test was used for comparison of the response rate (responders versus nonresponders) in the different study groups. Cytokine production was considered positive when the individual had a net production above 10 pg/ml. Comparisons between the skin mf counts in the different age groups were made by using the nonparametric Mann-Whitney U test. A P value of <0.05 was considered significant.

RESULTS

Skin mf counts and age.

The distribution of skin mf counts with age in our study group as a whole (n = 148) is shown in Fig. 1. In this particular cohort, a leveling off of median skin mf counts occurs at about 11 to 15 years of age. Skin mf counts in the age group of 3- to 5-year-old individuals were significantly lower (P < 0.001) than those in all other age groups. Skin mf counts in the age group of 6- to 10-year-old individuals were also significantly lower than those in age groups of 16- to 30-year-old individuals and older individuals (P < 0.05).

FIG. 1.

Average skin mf counts per individual in the patient cohort (n = 148) arranged according to age range (years). Median values are indicated by horizontal bars. An asterisk indicates a P value of <0.001 for comparisons with all other age groups; a double asterisk indicates a P value of 0.032 for comparisons with individuals 16 to 30 years old and older. P values were determined by the Mann-Whitney U test.

Lymphocyte proliferation in response to L3 antigen increases with age.

We compared the proliferative responses of PBMCs from infected individuals (n = 116) to infective larval (L3) and adult female worm (F-OvAg) antigens as a function of age (Fig. 2). Proliferative responses to L3 antigen increased significantly with age (r = 0.205; P = 0.02). In contrast, there was a downward, although not statistically significant, trend with age in response to F-OvAg antigen.

FIG. 2.

Correlation analysis of proliferative responses of PBMCs in response to O. volvulus L3 (top) and F-OvAg (bottom) antigens and the age (years) of the PBMC donor (n = 116). Correlations were tested by Spearman's rank correlation test. The r and P values are given for significant correlations. SI, stimulation index.

Cytokine responses vary with parasite stage and age of the individual.

When we analyzed cytokine responses of PBMCs to stage-specific O. volvulus antigens in the group as a whole, regardless of age, Th2-type responses were dominant (Table 1). Significantly more individuals generated IL-5 than IFN-γ to all four antigens. IL-10 responses were also much more frequent than IFN-γ responses.

TABLE 1.

Frequencies of individuals with positive cytokine responses to O. volvulus antigens

Antigen	% of individuals having the following cytokine response to O. volvulus crude antigena:
	IL-5	IFN-γ	IL-10	GM-CSF
L3	81.6 (71)b	20.0 (60)	79.0 (55)b	48.8 (54)
mL3	78.3 (60)b	21.3 (60)	72.7 (44)b	51.1 (43)
F-OvAg	90.4 (73)b	42.8 (63)	90.9 (55)b	56.5 (56)
Smf	53.3 (45)b	26.6 (45)	70.5 (34)b	60.0 (45)

^aTotal numbers of individuals in each experiment are given in parentheses. A positive response was taken as a cytokine concentration of >10 pg/ml above the concentration in untreated control wells.

^bThe difference between IL-5 or IL-10 and IFN-γ responses was statistically significant at a P value of <0.05, as determined by Fisher's exact test.

However, correlation analysis between the amounts of IL-5 and the age of the individuals revealed a negative association between F-OvAg-stimulated IL-5 release and age (r = −0.389; P = 0.0007) and between Smf-stimulated IL-5 release and age (r = −0.560; P < 0.0001) (Fig. 3). A similar inverse relationship was also observed for the IL-5 levels and skin mf densities of the individuals (for F-OvAg, the r value was −0.438 and the P value was 0.0001; for Smf, the r value was −0.581 and the P value was <0.0001). Thus, in response to F-OvAg and Smf, IL-5 levels decrease as skin mf densities and ages of the individuals increase. In marked contrast, the quantities of IL-5 released from PBMCs stimulated with L3 and mL3 remained elevated with increasing age (Fig. 3) or skin mf densities (data not shown) of the individuals.

FIG. 3.

Correlation analysis between amounts of IL-5, IFN-γ, IL-10, and GM-CSF generated by PBMCs in response to O. volvulus infective larval (L3 and mL3) and adult (F-OvAg) antigens and Smf antigen and the age (years)of the PBMC donor. Correlations were tested by Spearman's rank correlation test. The r and P values are given for significant correlations. The number (n) of individuals in each analysis is indicated.

IFN-γ levels in the group as a whole were low or suppressed (negative values) for all the antigens tested, except for responses to Smf in individuals over the age of 15 years. There was no correlation between IFN-γ concentrations and age or skin mf densities in response to F-OvAg, L3, and mL3. However, a positive correlation between IFN-γ levels and age was observed in response to Smf (r = 0.300; P = 0.046) (Fig. 3), independent of the skin mf counts of the donors.

Since IL-10 is a potential mediator of T-cell hyporesponsiveness to adult and larval antigens in chronic onchocerciasis, we measured IL-10 secretion by PBMCs in response to F-OvAg, L3, mL3, and Smf in our cohort. More individuals produced IL-10 than IFN-γ in response to all of the O. volvulus antigens (Table 1), and there was a particularly strong IL-10 response to F-OvAg (Fig. 3). Although IL-10 secretion was elevated in response to the first three antigens, it was not statistically related to age or skin mf densities. In contrast, the IL-10 response to Smf declined significantly with age (r = −0.336; P = 0.055). There was an apparent inverse relationship between IL-10 and IFN-γ production in response to Smf; however, it was not statistically significant.

A total of 48.8 to 60% of the group as a whole secreted GM-CSF when stimulated with the O. volvulus antigens (Table 1), although the amounts of this cytokine were unrelated to the age of the PBMC donor (Fig. 3) or skin mf count (data not shown). However, in the individuals who responded, the levels of GM-CSF were positively correlated with those of IL-5 (r = 0.408; P = 0.002) after L3 antigen stimulation. Peripheral blood eosinophil or neutrophil counts were not significantly correlated with GM-CSF and IL-5 secretion in the latter individuals (data not shown).

PBMCs from the group as a whole produced similarly high levels of IL-5, IFN-γ, IL-10, and GM-CSF in response to either a mitogenic stimulus (phorbol myristate acetate plus ionomycin) or SLO, a nonparasitic antigen, regardless of age. These levels were uniformly higher than the upper limit of detection (≥500 pg/ml).

Distinct O. volvulus stage-specific antibody isotype responses develop with age.

Analysis of antigen-specific IgGl, IgG3, IgG4, and IgE in sera from O. volvulus-infected individuals was performed with F-OvAg and L3 antigens as well as with rOv-ALT-1, which has been well characterized in our laboratory (26) and which induces significant protection in mice when used in combination with alum (1). Correlation analysis between the ELISA values (OD values) for each isotype and serum donor age was performed, and the results are shown in Fig. 4.

FIG. 4.

Correlation analysis between OD values in IgG1, IgG3, IgG4, and IgE ELISAs with O. volvulus L3, F-OvAg, and rOv-ALT-1 antigens and the age (years) of the serum donor. Correlations were tested by Spearman's rank correlation test. The r and P values are given for significant correlations. The number (n) of individuals in each analysis is indicated.

Although the IgG1 responses to F-OvAg and L3 antigens were not correlated with age, they were elevated in all age groups. The IgG3 response to crude L3 antigen was, however, positively correlated with age (r = 0.250; P = 0.022), but that to F-OvAg antigen was not. IgG4 levels increased significantly with age in response to both L3 antigen (r = 0.250; P = 0.020) and F-OvAg antigen (r = 0.298; P = 0.021). With the exception of the IgG4 response to L3 (r = 0.224; P = 0.044; n = 81), there were no significant correlations between levels of IgG antibodies and skin mf counts. IgG2 analysis was excluded in this study because this isotype predominantly recognizes cross-reacting phospholipids and carbohydrates (3, 12). Interestingly, when the antibody response against rOv-ALT-1 was analyzed, significant increases with age were seen for the IgG1 (r = 0.256; P = 0.048) and IgG3 (r = 0.279; P = 0.031) levels, whereas the IgG4 levels in response to rOv-ALT-1 were almost uniformly low, regardless of age.

The level of IgE antibodies to L3 antigen was significantly upregulated with age (r = 0.225; P = 0.021) and was elevated in the majority of samples in response to rOv-ALT-1. Conversely, the IgE level in response to F-OvAg antigen tended to decline with age and was negatively correlated with skin mf densities (r = − 0.336; P = 0.009).

Surface antibody reactivity with L3 increases with age.

The antibody response to the surface of L3 was measured by indirect immunofluorescence with live L3 parasites harvested from infected black flies or live Smf parasites purified from human skin snips and cryopreserved in Kumba, Cameroon. Analysis of serum total immunoglobulin reactivity with the L3 or Smf surface by immunofluorescence staining for 20 individuals (5 to 45 years old) indicated that significantly more O. volvulus-infected individuals had anti-L3 surface antibodies (75%) than had anti-Smf surface reactivity (35%). The difference was determined by Fisher's exact test (P = 0.02). Further analysis of serum total immunoglobulin reactivity with the L3 surface (Fig. 5) with age for 34 individuals (5 to 60 years old) showed an enhanced intensity of L3 surface recognition with increasing age of the serum donors (r = 0.37; P = 0.02).

FIG. 5.

Total antibody responses to the surface of intact O. volvulus L3, as assessed by IFA, versus age (years) of the serum donor (n = 34). The intensity of fluorescence was scored from 0 to 3. Correlation was tested by Spearman's rank correlation test. The r and P values are given for significant correlations.

DISCUSSION

The significant finding of this study is that separate and distinct cytokine and antibody responses to adult and infective larval stages of O. volvulus are a function of duration of exposure to infection and are consistent with a state of concomitant immunity in infected individuals. The existence of parasite infective-stage-specific immunity has been well documented in lymphatic filariasis (7, 27, 34) and schistosomiasis (4), and epidemiological data support the concept of concomitant immunity in these infections. In bancroftian filariasis, protective immune responses develop over a number of years, regardless of microfilaremic status (4). In onchocerciasis, the fact that parasite infection as measured by mf counts in the skin reaches a plateau after 20 to 40 years (11) and, in the present cohort, after 10 to 15 years of exposure (Fig. 1) indicates the possibility of acquired concomitant immunity to further infection. However, the potential mechanisms of such resistance to the newly invading infective larvae have not been studied to date. Because skin mf counts are not directly related to the burden of O. volvulus infection and fluctuate considerably with time in the same individual and because the status of concomitant immunity is necessarily acquired with years of exposure, equating to age in our study group, irrespective of skin mf status, all of our immunological analyses are correlated with age.

A consistent finding in chronic onchocerciasis is proliferative hyporesponsiveness of lymphocytes to adult antigens (10, 41). Our results are consistent with these earlier studies but additionally reveal an age-related increase in PBMC responsiveness to larval antigens (Fig. 2). Infective larval antigens are believed to be the targets of protective immunity in PI individuals (46) and in animal models of O. volvulus (1) and other (8, 13) filarial infections.

Chronic filarial infections are typified by Th2 responses (31, 34), and this fact was confirmed by the domination of IL-5 over IFN-γ responses to crude parasite extracts in the present study. With the exception of the response to the Smf antigen, IFN-γ secretion was either absent or suppressed by O. volvulus antigens. In previous studies, IFN-γ responses to F-OvAg were either low (3) or downregulated with age (44). The elevated IFN-γ response to the Smf antigen with age may be part of the response to increasing numbers of mf with age and the immune response to mf during chronic infection. High levels of IFN-γ production by splenocytes from B. pahangi-infected mice were recently reported following stimulation by homologous microfilarial antigens (38). In B. malayi infection in mice, adult worms stimulate a strong Th2 cytokine response (IL-4), whereas mf drive a Th1-biased response (IFN-γ) (29).

In the present study, IL-5 responses to the larval antigens were strikingly different from those stimulated by extracts from parasite stages involved in establishing patent infection, i.e., the female adult worms (F-OvAg) and skin microfilariae (Smf), which declined with increasing age and skin mf counts. This differential antilarval IL-5 response could contribute to protective immunity to new O. volvulus infection. Although not significantly correlated with age, in individuals (approximately 50%) who secreted GM-CSF to L3 antigens, the amounts of this cytokine were positively associated with IL-5 generation. The reason for the lack of GM-CSF responses in individuals who had nevertheless generated IL-5 to L3 antigens is presently unknown. However, elevated antilarval (L3 and mL3) IL-5 and GM-CSF responses were previously observed in PI individuals compared toINF individuals (46). This combination of cytokines would be particularly effective in attracting and activating cells from the granulocyte-macrophage lineage (neutrophils, eosinophils, and basophils) (19) in response to the migrating larvae in the skin. We found no correlation of O. volvulus antigen-specific IL-5 or GM-CSF with circulating eosinophil counts, a finding which may have been due to the prevalence of other eosinophilia-inducing helminth infections. Alternatively, newly recruited eosinophils may localize rapidly in tissues at the site of cytokine secretion.

IL-10 has been implicated in T-cell hyporeponsiveness and suppressed Th1 responses to lymphatic filarial antigens (32, 39) and O. volvulus antigens (5, 10). In the present study, IL-10 levels were high in most cases, especially in response to F-OvAg; however, in initial studies (data not shown), we found that neutralization of IL-10 had no effect on IFN-γ (or IL-5) production by PBMCs from INF individuals in response to F-OvAg. Thus, IL-10 is unlikely to be the only factor responsible for the downregulation of T-cell responses in filariasis; parasite-derived immunoregulatory molecules may also be important (33).

Antibody responses in onchocerciasis were analyzed in earlier studies with adult worm extracts (2, 6, 37, 43) or recombinant O. volvulus antigens with or without adult worm extracts (3, 18, 20, 40, 42, 45). However, with the exception of one recent study (16), none analyzed responses to larva-specific antigens as a function of duration of exposure to O. volvulus. In the present study, we show, for the first time, that differential specific IgG1, IgG3, and IgE responses to larval antigens, crude L3 extract, rOv-ALT-1, and F-OvAg antigen develop over years of exposure to the parasite. Our antilarval antibody findings are at variance with a recently published study which reported extremely low levels of IgG1, IgG3, and IgG4 and almost undetectable levels of IgE to L3 extract in comparison with F-OvAg (16). This variance may reflect the different larval antigen extraction techniques used, which might have resulted in different recoveries of cuticular antigens.

The maintenance or upregulation of cytophilic (IgG1, IgG3, and IgE) and complement-fixing (IgG3) antibody responses distinct for larval antigens, as was found in this study, could be a mechanism by which concomitant immunity is maintained, thereby limiting the overall parasite load. Elevated IgG3 to the L3-specific S1 O. volvulus recombinant protecting antigen has been associated with the PI state in areas in which O. volvulus is endemic (9). Importantly, the IFA results corroborated our antibody ELISA findings and demonstrated a similar age-related increase in antibody binding, in this case to the L3 surface. Only a minority of the individuals had anti-skin mf reactivity. Subsequent analysis of a few serum samples with a high level of total binding to the L3 surface revealed that this response was mostly associated with IgG1, IgG3, IgE, and/or IgM antibodies (data not shown). Antilarval stage-specific immunity associated with possible concomitant immunity has been also described for B. malayi filariasis; INF individuals expressed antibodies to the surface of L3 but not to mf (27). In accordance with the idea that concomitant immunity develops with age, “immune” adults possessed anti-L3 surface antibody while “nonimmune” children did not during lymphatic filariasis.

Elevated levels of IgG4 antibodies are produced during chronic filarial infections, and coexpression with IgE antibodies may be beneficial, as the former isotype can block IgE-mediated allergic responses to mf and thus limit skin pathology (4, 23). Both L3- and F-OvAg-specific IgG4 responses increased with age in our study, but high IgE and low IgG4 levels to rOv-ALT-1 may enhance its vaccine potential. Shared antigens in crude extracts of L3 and F-OvAg may stimulate IgG4 responses, whereas larva-specific rOv-ALT-1 may lack these epitopes. rOv-ALT-1 is potentially important in the process of molting from L3 to L4 (26) and, when used in alum to vaccinate mice, was recently shown to induce significant reductions in viable and molting L3 parasites in diffusion chambers (1). Our present results would support its potential as a target antigen for the control of human O. volvulus infection.

In conclusion, our findings indicate separate and distinct cellular and antibody responses to O. volvulus infective larvae compared with those stimulated by adult female worms and their offspring (mf). These differential antilarval responses develop over years of exposure, consistent with the acquisition of a state of concomitant immunity. Antilarval Th2 and cytophilic antibody responses are maintained or elevated with duration of exposure to infection, while those to adult female worm and mf stages decline or are suppressed. The antilarval responses bear many of the hallmarks required for efficient antibody-dependent cellular cytotoxicity reactions—eosinophil- and neutrophil-stimulating cytokines and elevated cytophilic antibodies—which could effectively control new infection. In diffusion chambers, protection against O. volvulus L3 in mice previously immunized with irradiated L3 was dependent on Th2 responses (25, 28), and larval killing was coincident with maximal levels of IL-5, IgE, and eosinophils in the chambers (28). Further support for potential protective antibody-dependent cellular cytotoxicity mechanisms have been provided by PI and INF serum antibody-dependent killing of O. volvulus L3 in vitro by human neutrophils (24). Now that this phenomenon has been identified, selection of suitable patients for further in vitro studies of our recombinant larval O. volvulus antigen vaccine candidates will be greatly facilitated. At present, all of the cloned O. volvulus antigens found to be protective in the diffusion chamber model for mice are similarly recognized by sera from PI and INF individuals (1). Although effector mechanisms against infective L3 in the PI and concomitantly immune states may be very similar, the target antigens may be different. Detailed analysis of antigen recognition in older individuals with high Th2 and cytophilic antibody responses to L3 or mL3 antigens could yield promising new vaccine candidates.