Inflammatory and regulatory CCL and CXCL chemokine and cytokine cellular responses in patients with patent Mansonella perstans filariasis

Summary

Mansonella perstans (Mp) filariasis is present in large populations in sub‐Saharan Africa, and to what extent patent Mp infection modulates the expression of immunity in patients, notably their cellular cytokine and chemokine response profile, remains not well known. We studied the spontaneous and inducible cellular production of chemokines (C‐X‐C motif) ligand 9 (CXCL9) [monokine induced by interferon (IFN)‐γ (MIG)], CXCL‐10 [inducible protein (IP)‐10], chemokine (C‐C motif) ligand 24 (CCL24) (eotaxin‐2), CCL22 [macrophage‐derived chemokine (MDC)], CCL13 [monocyte chemotactic protein‐4 (MCP‐4)], CCL18 [pulmonary and activation‐regulated chemokine (PARC)], CCL17 [thymus‐ and activation‐regulated chemokine (TARC)] and interleukin (IL)‐27 in mansonelliasis patients (Mp‐PAT) and mansonelliasis‐free controls (CTRL). Freshly isolated peripheral mononuclear blood cells (PBMC) were stimulated with helminth, protozoan and bacterial antigens and mitogen [phytohaemagglutinin (PHA)]. PBMC from Mp‐PAT produced spontaneously (without antigen stimulation) significantly higher levels of eotaxin‐2, IL‐27, IL‐8, MCP‐4 and MDC than cells from CTRL, while IFN‐γ‐IP‐10 was lower in Mp‐PAT. Helminth antigens activated IL‐27 and MCP‐4 only in CTRL, while Ascaris antigen, Onchocerca antigen, Schistosoma antigen, Entamoeba antigen, Streptococcus antigen, Mycobacteria antigen and PHA stimulated MIG release in CTRL and Mp‐PAT. Notably, Entamoeba antigen and PHA strongly depressed (P < 0·0001) eotaxin‐2 (CCL24) production in both study groups. Multiple regression analyses disclosed in Mp‐PAT and CTRL dissimilar cellular chemokine and cytokine production levels being higher in Mp‐PAT for CCL24, IL‐27, IL‐8, MCP‐4, MDC and PARC (for all P < 0·0001), at baseline (P < 0·0001), in response to Entamoeba histolytica strain HM1 antigen (EhAg) (P < 0·0001), Onchocerca volvulus adult worm‐derived antigen (OvAg) (P = 0·005), PHA (P < 0·0001) and purified protein derivative (PPD) (P < 0·0001) stimulation. In Mp‐PAT with hookworm co‐infection, the cellular chemokine production of CXCL10 (IP‐10) was diminished. In summary, the chemokine and cytokine responses in Mp‐PAT were in general not depressed, PBMC from Mp‐PAT produced spontaneously and selectively inducible inflammatory and regulatory chemokines and cytokines at higher levels than CTRL and such diverse and distinctive reactivity supports that patent M. perstans infection will not polarize innate and adaptive cellular immune responsiveness in patients.

Introduction

Mansonelliasis is caused by four species of nematodes belonging to the genus Mansonella, i.e. M. perstans, M. streptocerca, M. ozzardi and M. rodhaini, three of which are endemic in Africa 1. M. perstans is considered to be the most frequent in sub‐Saharan Africa; more than 100 million people may be infected 1, 2. The transmission of M. perstans is through the bite of blood‐feeding Culicoides midges (Diptera: Ceratopogonidae). Adult M. perstans filariae are described in the connective tissue of serous body cavities and unsheathed microfilariae (Mf) circulate in peripheral blood 2. The longevity of adult M. perstans in humans is unknown, but Mf may persist for several months 2, 3. M. perstans is considered to be of little pathogenicity, and although often asymptomatic, infections may cause eosinophilia, subcutaneous swellings, aches, pains and skin rashes in a considerable proportion of patients 2. The lack of specificity of symptoms might be explained by a thoroughly modulated immune response which has developed and adapted to chronic infection and repeated reinfection in an endemic environment, but symptoms may also be caused by co‐infection with other filariae 2, 3.

The control of lymphatic filariasis is based on annual mass drug administration (MDA) of ivermectin together with albendazole, which will clear blood‐circulating microfilaria, and such intervention may also interrupt parasite transmission 4. Mansonelliasis is often co‐endemic with onchocerciasis 5, 6 and lymphatic filariasis 7, 8, 9, and the Onchocerciasis Control Programs in Africa (OCP, 1974–2002; APOC, 1995–2015) have largely controlled onchocerciasis the disease burden through the MDA of ivermectin 10, 11. With the implementation MDA of ivermectin, for decades large populations have become permanently negative for microfilaria (Mf) of O. volvulus, but ivermectin alone will not clear Mf of M. perstans and chronic mansonelliasis will persist 2, 5, 7, 12.

Ivermectin treatment will influence the parasite–host equilibrium and change the immune response profile in patients. In ivermectin‐treated onchocerciasis patients, T helper type 1 (Th1)‐type cytokines will reactivate while regulatory and Th2‐type‐promoting cytokines and chemokines lessen 13, 14, 15. Such changes may reflect decreasing eosinophil granulocyte activation against Mf 15, 16 and, in parallel, lower Plasmodium‐specific Th17 immune responses 17. De‐worming will alleviate the helminth‐induced cellular hyporesponsiveness; repeated anti‐helminth treatment results in significant increases in proinflammatory cytokine responses to Plasmodium falciparum antigens and mitogen with a significant decline in the expression of the inhibitory molecule cytotoxic T lymphocyte antigen (CTLA)‐4 on CD4⁺ T cells of treated individuals 18. To what extent persistent M. perstans infection may influence or bias cellular reactivity and the immune response profile in patients remains little known. We studied the in‐vitro cellular responsiveness of mononuclear peripheral blood cells from mansonelliasis patients to helminth, protozoan and bacterial antigen stimulation, and observed that significant proinflammatory chemokines and cytokines were produced.

Materials and methods

Location of study, participants and examinations

This study was conducted in central Togo in West Africa, within the previously vector‐controlled area of the former OCP, where annually repeated mass drug administration (MDA) by community‐directed treatment with ivermectin (CDTI) has been applied since 1989. The mansonelliasis patients and endemic controls were from the Prefecture Tchaoudjo in the Central Region of Togo and permanent residents in rural villages. This investigation was authorized by the Ministry of Health in Togo (no. 0407/2007MS/CAB/DGS; no. 0060/2013/MS/CAB/DGS) and the Comité de Bioéthique pour la Recherche en Santé (no. 013/2015/CBRS). All participants gave their written informed consent, and for correct and complete understanding explanations were always given in the local language. At the time of sampling, participants were healthy male and female (non‐pregnant) individuals who resided permanently in the villages of Bouzalo (N09°06.134′; E001°02.588′) and Sagbadai (N09°03.819′; E001°04.473′). By means of mass drug administration through the community‐directed treatment with ivermectin (CDTI; 150 ug/kg), all participants received annually a single dose distributed by the National Onchocerciasis Control Program (NOCP) in Togo. Venous blood samples (18 ml), skin biopsies and stool and urine samples were collected from all participants, fresh stools were examined by microscopy for helminth and protozoa infections and the samples were then examined using the Kato–Katz methodology (helm‐TEST; Labmaster, Belo Horizonte, MG, Brazil). From each participant, 10 ml of urine were centrifuged and the sediment examined under a microscope for eggs of Schistosoma haematobium.

Microscopy examination for blood‐dwelling M. perstans microfilariae

Microfilariae of M. perstans were detected after Biocoll‐gradient centrifugation (Biochrom, Berlin, Germany) of 20 ml of whole blood samples (diluted 1 : 2 in RPMI) in the peripheral blood mononuclear cells (PBMC) fractions and polymorphonuclear cell pellets (PMNC)19. The PMNC pellets were resuspended in phosphate‐buffered saline (PBS) (part 1), mixed with an equal volume of 5% dextran 500 in PBS (part 2), and an equal volume of PBS (part 3) was added. Such resuspended PMNC were allowed to sediment at 1 g for 1 h at room temperature (RT); thereafter the supernatant was collected, centrifuged at 800 g for 15 min at RT and the pellets examined by microscopy for the presence of microfilaria of M. perstans. The isolated PBMC were dispensed in equal volumes into 24‐well cell culture plates, and after overnight incubation plate wells were examined under an inverted microscope for the presence of dwelling microfilariae of M. perstans.

Real‐time polymerase chain reaction (PCR) for the detection of M. perstans DNA in whole blood samples

For DNA extraction, whole venous blood samples (200 µl) were collected into microcentrifuge tubes and processed using the Qiagen DNA Investigator Kit (Qiagen, Hilden, Germany), according to the manufacturer's recommended protocol. After overnight proteinase K digestion at 56°C, three DNA elutions of 50 µl each were performed. The eluted DNA concentrations were determined and samples stored at −20°C before reverse transcription (RT)–PCR analysis. Extracted blood DNA concentrations ranged from 4 to 166 ng/µl. As previously applied 6 for the detection of M. perstans DNA, the real‐time quantitative PCR (qPCR) was carried out with the PCR cycler rotor gene RG 3000 (Corbett Research/Corbett Life Science, Qiagen NV, the Netherlands). The primer and probe sequences selection was carried out using the online software Primer3 (http://bioinfo.ut.ee/primer3-0.4.0/). The applied qPCR reaction was used to detect the M. perstans 18S and 5.8S ribosomal RNA gene, internal transcribed spacer 1 (GenBank: KJ631373). The qPCR primer pairs, probes and test conditions used for the detection of M. perstans (Mp) were: Mp‐primer forward 5′‐CTGCGGAAGGATCATTAA‐3′ (Tm 51.4°C); Mp‐primer reverse 5′‐TGCATGTTGCTAAATAAAAGTG‐3′ (Tm 52.8°C); and Mp‐probe 5′‐FAM‐CGAGCTTCCAAACAAATACATAATAAC‐TAM‐3′ (Tm 58.9°C) The RT–PCR conditions were 50°C/2 min (95°C/10 min, 95°C/15 s, 53°C/1 min) × 45 cycles.

Preparation of antigens

S. mansoni adult worms were isolated aseptically by perfusion from portal veins of infested mice; adult Ascaris lumbricoides were collected as expulsed worms from patients who received a 3‐day treatment of mebendazole, and the collected worms were washed extensively in sterile PBS (pH 7.6). Entamoeba histolytica trophozoites (axenic strain HB3) were a kind gift from Dr B. Walderich (Tübingen, Germany). Adult filarial worms of O. volvulus were isolated from nodules (onchocercomata), as described by Schulz‐Key et al. 1977 20. Adult A. lumbricoides, S. mansoni and O. volvulus were washed extensively in sterile PBS (pH 7.6), transferred into a Ten‐Broek tissue grinder and then homogenized on ice. Similarly, E. histolytica trophozoites were ground and homogenized. The homogenates were then sonicated twice (30% intensity, pulse 1 s) for 10 min on ice and centrifuged at 16000 g for 30 min at 4°C. The supernatants were sterile filtered (0·22 μm) and the protein concentration of each antigen was determined by the bicinchoninic acid (BCA) method (Pierce). The limulus amoebocyte lysate assay (E‐Toxate Kit; Sigma Aldrich, St Louis, MO, USA; ET0100) was used to detect endotoxin in the worm and protozoa antigen extracts. Endotoxin levels were at 0·25 EU/ml in the E. histolytica and S. mansoni extracts and 1·25 EU/ml in the A. lumbricoides and O. volvulus antigens. Purified protein derivative (PPD) from Mycobacterium tuberculosis was purchased from Behring (Marburg, Germany), and Streptolysin‐O (SL‐O) from Streptococcus pyogenes was obtained from Difco (Augsburg, Germany).

Isolation of PBMC, cell culture experiments and determination of cytokine production

Heparinized venous blood was collected from mansonelliasis patients and endemic controls, and PBMC were isolated and cell culture experiments were conducted as described previously 14, 19. Briefly, PBMC were adjusted to 1 × 10⁷/ml in RPMI supplemented with 25 mM HEPES buffer, 100 U/ml penicillin and 100 mg/ml streptomycin, 0·25 mg/ml amphotericin B. Freshly isolated PBMC were cultured at a concentration of 2·5 × 10⁶ PBMC/ml in RPMI (as above) supplemented with 10% heat‐inactivated fetal calf serum (FCS) (Biochrom, Berlin, Germany) in the presence of either A. lumbricoides adult worm extract (AscAg; 5 μg/ml), E. histolytica strain HM1 antigen (EhAg; 10 μg/ml), O. volvulus adult worm‐derived antigen (OvAg, 35 µg/ml), M. tuberculosis PPD (100 µg/ml), phytohaemagglutinin (PHA) (1 : 100; Sigma, St Louis, MO, USA), S. mansoni adult worm extract (SmAg; 10 μg/ml) or Strep. pyogenes‐derived Streptolysin‐O (SL‐O, 1:50; Difco) in 5% CO₂ at 37°C and saturated humidity. Cell culture supernatants were collected after 48 h and stored below –20°C until further use. Cytokine secretion by stimulated PBMC was quantified by sandwich enzyme‐linked immunosorbent assay (ELISA) using cytokine‐ and chemokine–specific monoclonal and polyclonal antibodies, as recommended by the manufacturers. The detection limits of the cytokine and chemokine ELISAs (DuoSet; R&D Systems, Minneapolis, MN, USA) were at 50 pg/ml; all concentration values below that threshold were set to 0 pg/ml.

O. volvulus antigen‐specific ELISA

O. volvulus antigen‐specific (OvAg) immunoglobulin (Ig)G4 isotype reactivity was determined by ELISA as described by Mai et al. 14, 19. Briefly, microtitre plates (Corning 3690; Costar, Assay Plate) were coated with O. volvulus adult worm extract (OvAg 5 µg/ml) in PBS. Non‐specific binding capacity was blocked at room temperature (RT) for 2 h with PBS containing 5% fetal bovine serum (FBS). Samples and reference control sera were added to OvAg‐coated wells and incubated for 2 h at RT. After washing with PBS containing 0·05% Tween 20 (Sigma; P‐3563), horseradish peroxidase‐conjugated mouse anti‐human IgG4 monoclonal antibody (Invitrogen, Eugene, OR, USA) at a dilution of 1 : 500 was added for 2 h at RT. After washing as above, specific binding was visualized by addition of 3,3′,5,5′‐tetramethylbenzidine (TMB) substrate, reactions were stopped by addition of 0·5 M H₂SO₄ and the optical density was determined at 450 nm.

Data analysis

JMP software (versions 11.1.1; SAS Institute, Cary, NC, USA) was used for statistical analysis of data. Because of multiple comparisons, the level of significance was adjusted according to Bonferroni–Holm. For the cytokine and chemokine analyses, differences between groups were determined after logarithmic transformation to stabilize the variance of data [log (pg/ml + 1)]. The application of the Bonferroni–Holm adjustment resulted in an alpha level of α = 0·003. The data from the patient and control group were compared using Wilcoxon's rank sum test (Mann–Whitney U‐test). Multiple regression analysis was applied to analyse the chemokine and cytokine production in mansonelliasis patients and controls with and without hookworm co‐infection. The cytokine production was analysed with the predictors: study groups, cytokine, chemokine, antigen stimulations, mitogen stimulation, patient number × study groups (i.e. the random factor is the patient's number) and their corresponding interaction of degree 2. For post‐hoc testing and for comparison of the different groups with and without hookworm co‐infection, the Tukey–Kramer Test was applied. For multivariate analysis, epidemiological (gender, age) and immune parameters (cytokines, chemokines, antigen and mitogen stimulations) were added as covariates and compared with one another. Comparisons were made between mansonelliasis patients and controls with and without hookworm co‐infection.

Results

Study groups and patients' characteristics

The demographic, haematological and parasitology data for the study groups are shown in Table 1. All participants (n = 50) were treated annually with 150 µg/kg ivermectin for more than 15 years and were negative for microfilariae of O. volvulus at repeated skin biopsy examinations (Table 1). Antibody responses (IgG4) to O. volvulus antigen (OvAg) were similarly low in Mp‐PAT and in mansonelliasis‐free controls (Table 1). M. perstans microfilariae and M. perstans DNA were diagnosed in 37 participants (Mp‐PAT) and none in endemic controls (n = 13; CTRL). Hookworm larvae were detected in stool samples from Mp‐PAT (positive: n = 13; negative: n = 24) and also in endemic controls (positive: n = 3; negative: n = 10). Mp‐PAT presented with lower lymphocyte but higher eosinophil granulocyte counts (P = 0∙016) than CTRL.

Table 1.

Demographics, haematological and parasitology data [median (min;max)] of the Mansonella perstans microfilariae (Mf)‐positive patients and M. perstans Mf‐negative endemic controls, their leucocyte counts and blood cell differential

	Mansonelliasis patients (n = 37)	Mansonelliasis‐free endemic controls (n = 13)
Age (min; max	49** (25; 75)	36 (21; 54)
Gender (F/M)	12/25	3/10
Haemoglobulin (g/dl)	16∙0 (10∙8; 23∙7)	15∙3 (11∙5; 17∙7)
Leucocytes (cells/μl)	5259 (3600; 8100)	5443 (4100; 8100)
Neutrophil granulocytes	34∙6% (27; 52)	35∙8% (27; 52)
Eosinophil granulocytes	1∙6% (0; 6)	0∙8% (0; 2)
Basophil granulocytes	0%	0%
Lymphocytes	61∙8% (46; 84)	61∙4% (48; 26)
Monocytes	2% (1; 4)	2% (2; 2)
M. perstans qPCR Ct‐value median (min; max)	Mp‐positive 36∙5 (29∙6; 42∙7)	Mp‐negative (none)
O. volvulus Mf/skin biospy	0	0
IgG4 OvAg‐ELISA OD median (min; max)	0∙232 (0; 0∙52)	0∙176 (0; 0∙398)
P. falciparum qPCR Ct‐value median (min; max)	Pf‐positive n = 26 33∙4 (24∙6; 38∙6)	Pf‐positive n = 6 37∙2 (34∙5; 39∙8)
Hookworm eggs/g stool median (min; max)	Positive n = 13 792 (144; 9264)	Positive n = 3 552 (72; 4728)
E. histolytica/E. dispar (cysts)	Positive n = 12	Positive n = 4

Wilcoxon rank sum test: patients (PAT) versus control (CTRL). qPCR = quantitative polymerase chain reaction; M/F = male/female; Ig = immunoglobulin; OD = optical density; ELISA = enzyme‐linked immunosorbent assay; Ct = cycle threshold.

^** P = 0∙016; P = 0∙001.

The spontaneous cellular production of chemokines and cytokine by PBMC from patients with patent M. perstans infection and mansonelliasis‐free controls

The spontaneous production, i.e. without antigen or mitogen stimulation, was elevated in mansonelliasis patients (Mp‐PATs) (Figs. 1 and 2, ‘1baseline’). PBMC from Mp‐PAT, compared to CTRL, spontaneously released significantly higher amounts of chemokine (C‐C motif) ligand 24 (CCL24)/eotaxin (P = 0·002), IL‐27 (P = 0∙002), IL‐8 (P = 0∙003), monocyte chemotactic protein‐4 (MCP4) (P = 0∙009) and macrophage‐derived chemokine (MDC) (P = 0∙028) (Figs. 1 and 2). In contrast, less IP‐10/chemokine (C‐X‐C motif) ligand 9 (CXCL9) CXCL9 (P = 0∙002) was secreted spontaneously by PBMC from Mp‐PAT than by cells from CTRL (Fig. 2, ‘1baseline’).

Figure 1.

The cellular production of chemokines eotaxin [chemokine (C‐C motif) ligand 24 (CCL24)], macrophage‐derived chemokine (MDC) (CCL22), thymus‐ and activation‐regulated chemokine (TARC) (CCL17) and the cytokine interleukin (IL)‐27 by peripheral blood mononuclear cells (PBMC) from mansonnelliasis patients (PAT, n = 37) and by PBMC from mansonelliasis‐free controls (CTRL, n = 13). Freshly isolated PBMC were either without stimulation with antigen or mitogen (baseline) or PBMC were stimulated with helminth antigens from Ascaris lumbricoides adult worm extract (AscAg; 5 μg/ml), Onchocerca volvulus adult worm‐antigen (OvAg, 35 µg/ml) or Schistosoma mansoni adult worm extract (SmAg; 10 µg/ml), or stimulated with Entamoeba histolytica strain HM1 antigen (EhAg; 10 μg/ml), Mycobacterium tuberculosis purified protein derivative (PPD, 100 µg/ml), the mitogen phytohaemagglutinin (PHA) (1 : 100, Sigma) or Streptococcus pyogenes‐derived Streptolysin‐O (SL‐O, 1 : 50). Cell culture supernatants were collected after 48 h and chemokine and cytokine secretion was quantified by specific enzyme‐linked immunosorbent assay (ELISA) (R&D Systems). The amounts of chemokines and cytokine released into cell culture supernatants are shown as box‐blots with the median and the 25% and 75% quartiles, the ×1·5 of the interquartile range and with all outliers as individual points. The cellular production (pg/ml) of chemokines and cytokine (i.e. without baseline subtraction) in CTRL and PAT was compared using Wilcoxon's rank sum test, and significantly different production levels between CTRL and PAT groups are indicated with P ≤ 0∙05. The level of significance was adjusted according to Bonferroni–Holm (α = 0·003). Significant differences between controls (CTRL) and patients (PAT) are indicated with P‐values and: CTRL > PAT = chemokine or cytokine production greater in controls than patients; CTRL < PAT = chemokine or cytokine production greater in patients than controls; n.s. = not significant.

Figure 2.

The cellular production of the chemokines chemokines (C‐X‐C motif) ligand 8 (CXCL8) [interleukin (IL‐8)], monokine induced by interferon (IFN)‐γ (MIG) (CXCL9), inducible protein (IP)‐10 (CXCL10), pulmonary and activation‐regulated chemokine (PARC) (CCL18) and (CCL13) monocyte chemotactic protein‐4 (MCP‐4) by peripheral blood mononuclear cells (PBMC) from mansonnelliasis patients (PAT, n = 37) and by PBMC from infection‐free endemic controls (CTRL, n = 13). Freshly isolated PBMC were without antigen and without mitogen stimulation (baseline). PBMC were stimulated with helminth antigens from Ascaris lumbricoides adult worm extract (AscAg; 5 μg/ml), Onchocerca volvulus adult worm‐antigen (OvAg, 35 µg/ml) or Schistosoma mansoni adult worm extract (SmAg; 10 µg/ml), or with Entamoeba histolytica strain HM1 antigen (EhAg; 10 μg/ml), Mycobacterium tuberculosis purified protein derivative (PPD, 100 µg/ml), the mitogen phytohaemagglutinin (PHA) (1: 100, Sigma) or Streptococcus pyogenes‐derived Streptolysin‐O (SL‐O, 1 : 50). Cell culture supernatants were collected after 48 h and chemokine and cytokine secretion was quantified by specific enzyme‐linked immunosorbent assay (ELISA) (R&D Systems). The amounts of chemokines and cytokines released into cell culture supernatants are shown as box‐plots with the median and the 25% and 75% quartiles, the ×1·5 of the interquartile range and with all outliers as individual points. The cellular production (in pg/ml) of chemokines and cytokine (i.e. without baseline subtraction) in CTRL and PAT was compared using Wilcoxon's rank sum test, and significant different differences between CTRL and PAT groups are indicated with P ≤ 0∙05. The level of significance was adjusted according to Bonferroni–Holm (α = 0·003). Significant differences between controls (CTRL) and patients (PAT) are indicated with P‐values and: CTRL > PAT = chemokine or cytokine production greater in controls than patients; CTRL < PAT = chemokine or cytokine production greater in patients than controls; n.s. = not significant; n.d. = not done.

The cellular production of Th2‐type chemokines CCL24 (eotaxin‐2), CCL22 (MDC) and CCL17 (TARC)

The spontaneous cellular production of CCL24 (eotaxin‐2) was significantly higher in Mp‐PAT than in endemic controls (P = 0·002) (Fig. 1a). Stimulation in vitro of PBMC from Mp‐PAT with helminth antigens (AscAg, OvAg, SmAg) did not induce production of CCL24 above baseline levels, but it is noteworthy that E. histolytica antigen (EhAg) and the mitogen (PHA) depressed CCL24 release in CTRLS and Mp‐PAT drastically (P < 0∙0001). The S. mansoni adult worm antigen (SmAg) depressed CCL24 in CTRLs (P < 0∙0001), but had no such effect in PAT (Fig. 1a). The chemokine CCL22 (MDC) was inducible above baseline only in PBMC from CTRLs and PAT in response to the mitogen PHA and antigen Streptolysin O (SLO) (Fig. 1b). In mansonelliasis patients the MDC production in responses to PHA, PPD and SLO was above the levels measured in cell cultures supernatants from PAT (Fig. 1b). The cellular production of CCL17/TARC (Fig. 1c) was similar in Mp‐PAT and CTRL; neither the helminth antigens (AscAg, OvAg), EhAg nor bacterial PPD and SLO activated a cellular production of CCL17/TARC above the spontaneous release (Fig. 1c). Following PBMC stimulation with SmAg, more CCL17/TARC was produced in CTRL than in Mp‐PAT (P = 0∙0004), while PHA induced higher levels of CCL17/TARC (P < 0∙0001) and CCL18/[pulmonary and activation‐regulated chemokine (PARC)] (P < 0∙0001) in Mp‐PAT than in CTRL.

The cellular production of regulatory IL‐27 and chemokine CCL18 (PARC)

The spontaneous release of IL‐27 was higher in Mp‐patients than CTRL (Fig. 1). In CTRL, the AscAg, EhAg, bacterial SLO and mitogen PHA‐activated PBMC responses remained below the IL‐27 amounts released by cells from mansonnelliasis patients (Fig. 1d). The cellular production of CCL18/PARC was similar in Mp‐PAT and CTRL and neither the helminth antigens (AscAg, OvAg), EhAg nor bacterial PPD and SLO activated a cellular production of CCL18/PARC above the spontaneous release (Fig. 2d).

The cellular production of proinflammatory chemokines CXCL8 (IL‐8), CXCL9 monokine induced by interferon (IFN)‐γ (MIG) and CXCL10 (IP‐10)

The amounts of CXCL8 (IL‐8) (Fig. 2a) produced by PBMC in response to mitogen PHA, bacteria‐derived SLO and PPD, extracts from Entamoeba‐Ag and helminth‐specific Ascaris‐Ag and Schistosoma‐Ag were higher in Mp‐PAT than in CTRL, but without significant differences. In Mp‐PAT the cellular release of CXCL8/IL‐8 in response to OvAg (13935 pg/ml) was more than twice as high as in CTRL (mean 6526 pg/ml) (Fig. 2a). The production of CXCL9 (monokine inducible by IFN‐γ; MIG) was inducible above baseline levels by helminth (AscAg, OvAg, SmAg), bacterial antigens (SLO, PPD) and the mitogen PHA (Fig. 2b). MIG responsiveness was similar in CTRL and Mp‐PAT and only PHA induced higher MIG responses in PAT than in CTRL (Fig. 2b). The chemokine CXCL10 (IP‐10) (Fig. 2c) was produced in higher amounts in CTRL than in Mp‐PAT when PBMC were stimulated with helminth antigens AscAg (P < 0·001), OvAg (P < 0·01) and SmAg (P < 0·006), and also in response to the EhAg (P < 0·0001). When PBMC were activated with the mitogen, PHA secretion of CXCL10 was higher in Mp‐PATs than CTRLs (P < 0·003), while the extracts from Strep. pyogenes (SLO) and M. tuberculosis (PPD) activated similar IP‐10/CXCL10 amounts in both groups.

In Mp‐PAT and CTRL, multiple regression analyses disclosed dissimilar cellular chemokine and cytokine production levels, being higher in Mp‐PAT for CCL24, IL‐27, IL‐8, MCP4, MDC and PARC (P < 0∙0001 for all) at baseline (P < 0∙0001) in response to EhAg (P < 0∙0001), OvAg (P = 0∙005), PHA (P < 0∙0001) and PPD (P < 0∙0001) stimulation.

The cellular production of chemokines and IL‐27 in mansonelliasis patients and M. perstans‐negative controls co‐infected with hookworm

The cellular chemokine production of eotaxin (CCL24), MCP4 (CCL13), MDC (CCL22), MIG (CXCL9), PARC (CCL18), TARC (CCL17), CXCL8 (IL‐8) and the cytokine IL‐27 were similar in mansonelliasis patients and M. perstans‐negative controls co‐infected with hookworm (Table 2). Solely, the chemokine IP‐10 (CXCL19) was produced significantly less in hookworm‐positive mansonelliasis patients (Hook⁺Fil⁺) when their PBMC were stimulated with antigen extracts from A. lumbricoides (Tukey–Kramer test; P = 0∙0163), O. volvulus (P = 0∙0415) and E. histolytica antigen (P = 0∙0051). The CXCL10 (IP‐10) production levels were also lower in hookworm‐infected Mp‐PAT in response to S. mansoni (not significant) (Table 2).

Table 2.

The cellular production of the chemokine inducible by interferon (IFN)‐γ inducible protein 10 (IP‐10) (CXCL10) by peripheral blood mononuclear cells (PBMC) from mansonnelliasis (Mp⁺) patients co‐infected with hookworm (Hook⁺) and by PBMC from mansonelliasis‐free participants was compared. PBMC were without antigen (baseline) or with mitogen or antigen stimulation. In stool samples from mansonnelliasis (Mp⁺) patients hookworm larvae (Hook⁺) were detected (Hook⁺Mp⁺: n = 13) or not (Hook^–Mp⁺: n = 24) and also in mansonelliasis‐free (Mp^–) participants (Hook⁺Mp^–: n = 3; Hook^–Mp^–: n = 10). The mean cellular production in mansonelliasis patients without hookworm (Hook^–Mp⁺) and with hookworm co‐infection (Hook⁺Mp⁺) and in hookworm infected and mansonelliasis free (Hook⁺Mp^–) individuals and doubly negative (Hook^–Mp^–) controls were compared. The amounts of CXCL10 [inducible protein (IP)‐10] released into cell culture supernatants are shown as means (in pg/ml) with the 95% lower and 95% upper confidence intervals, and the chemokine CXCL10 (IP‐10) production was compared using the Tukey–Kramer test, and significant differences between study groups are indicated. Freshly isolated peripheral blood mononuclear cells PBMC (1 × 10⁶/ml in 500 µl) were stimulated with helminth (Helm) antigen extracts from Ascaris lumbricoides (AscAg; 5 μg/ml), Onchocerca volvulus (OvAg, 35 µg/ml), Schistosoma mansoni adult worm extract (SmAg; 10 µg/ml), or with protozoan (Protoz) antigen extract from Entamoeba histolytica strain HM1 antigen (EhAg; 10 μg/ml), or with bacteria (Bact) antigens from Mycobacterium tuberculosis purified protein derivative (PPD, 100 µg/ml) or Streptococcus pyogenes derived Streptolysin‐O (SLO, 1 : 50) or the mitogen (Mito) phytohaemagglutinin (PHA) (1: 100, Sigma). Cell culture supernatants were collected after 48 h and chemokine and cytokine secretion was quantified by specific ELISA (R&D Systems)

Antigen or mitogen stimulation	CXCL10 (IP‐10) pg/ml [(mean ) 95% lower CI; 95% upper CI)]				Tukey–Kramer test P‐value
	Study groups
	Hook–Mp–	Hook+Mp–	Hook–Mp+	Hook+Mp+
Baseline	454∙5 (425; 484)	450∙6 (401; 500)	374∙3 (279; 47)	378∙6 (324; 434)
	A	AB	AB	B	P = 0∙0184*
Bact‐PPD	2142∙8 (752; 3533)	925∙3 (0; 3138)	626∙7 (0; 500)	414 (0; 2888)
	A	A	A	A
Bact‐SLO	1334∙8 (304; 2366)	1623∙7 (5; 3243)	552∙7 (0; 3688)	2312∙6 (502; 4123)
	A	A	A	A
Helm‐AscAg	483 (464; 502)	451∙6 (420; 484)	406∙6 (345; 468)	418∙7 (382; 456)
	A	AB	AB	B	P = 0∙0163
Helm‐OvAg	566∙4 (513; 620)	543∙7 (455; 632)	490∙9 (319; 663)	413∙3 (314; 513)
	A	AB	AB	B	P = 0∙0415
Helm‐SmAg	1229∙7 (929; 1530)	1006∙6 (5535; 1478)	541∙7 (0; 1455)	484∙5 (0; 1012)
	A	A	A	A
Mito‐PHA	3183∙8 (2170; 4197)	2551∙1 (896; 4206)	2789 (0; 5993)	884∙5 (0; 2735)
	A	A	A	A
Protoz‐EhAg	446∙5 (430; 463)	433∙9 (407; 461)	392 (340; 444)	387∙5 (358; 418)
	A	AB	AB	B	P = 0∙0051

Study groups: Hook^–Mp^–: hookworm‐negative and Mansonella perstans‐negative; Hook⁺Mp^–: hookworm‐positive and M. perstans‐negative; Hook^–Mp⁺: hookworm‐negative and M. perstans‐positive; Hook⁺Mp⁺: hookworm‐positive and M. perstans‐positive; CI = confidence interval.

^*Groups that are not linked by the same letter differ significantly.

Discussion

The impact of ivermectin treatment on M. perstans

The present study was conducted in central Togo, where the mass drug administration (MDA) of ivermectin implemented by the OCP (1974–2002) and APOC (1995–2015) will progressively eliminate onchocerciasis and may reduce O. volvulus parasite transmission 21, 22, 23, 24, but the drug will not affect M. perstans and mansonelliasis will persist. The villages from whence patients originated are situated in mosaic forest savanna, and manonelliasis prevalence was recently detected with 7% in Sagbadai and 19% in Bouzalo 6, and annual MDA with ivermectin has been implemented there for more than 2 decades 25. Similarly observed in the Guinea and mosaic forest savanna zones in Cameroon, M. perstans has not changed after > 8 years of MDA treatments, but prevalence and infection intensities decrease in communities within the deciduous equatorial rainforest, suggesting that ivermectin has a partial effect on M. perstans 26. In savanna‐type western Burkina Faso, M. perstans did not respond during a 14‐year period to bi‐annual ivermectin treatments 7. Both studies used thick blood smears for M. perstans diagnosis, which may have missed low‐level changes of Mf counts in the savanna‐type ecological zones. To what extent persistent M. perstans infection and repeated ivermectin treatments may modulate, activate or counteract patients' cellular immune responsiveness remains unclear. In ivermectin‐treated onchocerciasis patients the expression of immunity will change; with parasite clearance, depressed cellular responsiveness of both types 1 and 2 will reactivate to O. volvulus and to bystander antigens, while parasite‐specific antibody responses will lessen gradually 13, 14, 15, 16, 17, 18. Repeated treatments with ivermectin may have changed or even reactivated cellular responsiveness, as observed in onchocerciasis, but these changes may not suffice to eliminate patent M. perstans infection. In the present study we observed that, with patent M. perstans infection, the spontaneous and antigen‐inducible cellular production levels of several types 1 and 2 proinflammatory and regulatory chemokines and cytokines were significantly higher in Mp‐PAT than in CTRL. Such mixed and non‐polarized responses may account for the lack of grave immune‐mediated pathology with mansonelliasis, and may represent a balance between a tolerable parasite load, causing little vascular and lymphatic damage and an immune adaptation which, to some extent, may limit parasite numbers without causing severe disease manifestation.

Th2‐type chemokine cellular production of eotaxin, TARC, MDC and CXCL8 (IL‐8)

In mansonelliasis patients, the prominent cellular release of the eosinophil‐ and neutrophil‐activating chemokines eotaxin‐2 (CCL24) (Fig. 1a) and CXCL8 (IL‐8) (Fig. 2a) may enhance the capacity of granulocytes for activation and chemotaxis. Although CXCL8 (IL‐8) and eotaxin (CCL24) persisted at elevated levels, blood‐circulating microfilariae of M. perstans were not eliminated by granulocyte‐mediated destruction. Immobilization and microfilariae destruction in infested tissues may trigger inflammatory responses, as observed in sowda‐type onchocerciasis 27, in lymphatic filariasis patients following treatment with diethylcarbamazine (DEC) 28, and serious adverse events have occurred in ivermectin‐treated onchocerciasis patients heavily co‐infected with Loa loa 29, 30. Cellular killing of microfilariae of O. volvulus was granulocyte‐mediated and serum‐dependent 31, and antigen extracts of O. volvulus which contain the symbiotic bacteria Wolbachia elicited strong macrophage and neutrophil activation and chemotaxis via induction of CXCL8 (IL‐8) 32. All study participants were repeatedly treated with ivermectin, and it remains unknown to what extent repeated ivermectin treatments will modify or modulate in mansonelliasis patients their filaria‐specific immune responses, and whether such changes may enhance their resistance or susceptibility to co‐infections with protozoa, bacterial and viral pathogens. In onchocerciasis patients, the Th2‐type chemokines CCL22 (MDC) and CCL17 (TARC) increased temporarily and shortly after primary ivermectin treatment, and then diminished significantly with the reduced parasite load 33. Enhanced levels of CCL17 (TARC) and CCL22 (MDC) may support elimination of Mf, but both were not inducible by helminth antigens (AscAg, OvAg, SmAg), neither in Mp‐PAT nor in CTRL. In the present study, only PHA and bacterial PPD and SLO enhanced their production in Mp‐PAT, and such selective inducible cellular release of MDC and TARC chemokines indicates functional Th2‐type cell recruitment and tissue‐specific migration of lymphocytes in Mp‐PAT 34.

Th1‐type and proinflammatory chemokine cellular production of MCP‐4, MIG and IP‐10

The monocyte chemoattractant protein 4 [CCL13 (MCP‐4)] is found in many chronic inflammatory diseases, and displays anti‐microbial activity against Gram‐negative bacteria 35. In the present work, the cellular release of CCL13 (MCP‐4) following stimulation with mitogen and bacteria‐ or helminth‐derived antigens was significantly higher in Mp‐PAT than in CTRLs. The ‘promiscuous’ binding of CCL13 (MCP‐4) to several chemokine receptors, i.e. CCR1, CCR2 and CCR3, may enhance cytokine secretion, activate further effecter cells and then facilitate the cell‐mediated clearance of microfilaria of M. perstans.

The chemokines CXCL9 (MIG) and CXCL10 (IP‐10) bind to their receptor CXCR3, activate and recruit T cells, eosinophils, monocytes and natural killer (NK) cells to inflamed tissues, and as such may also contribute to tissue damage 36. In both Mp‐PAT and CTRL, helminth extracts (AscAg, OvAg, SmAg) as well as protozoan and bacterial antigens (PPD, SLO) induced the production of CXCL9 (MIG), always being higher in Mp‐PAT. Such responsiveness indicated that helminth parasites will not only activate proinflammatory Th2‐type chemokines. CXCL10 (IP‐10) and its receptor CXCR3 contribute to the pathogenesis of chronic inflammatory arthritis 36, and CXCL10 (IP‐10) and CXCL9 (MIG) were strongly augmented in children with severe malaria 37, 38. The lower secretion of CXCL10 (IP‐10) in Mp‐PAT in response to AscAg and EhAg supported that IP‐10 mediated recruitment of inflammatory cells and induction of inflammatory cytokines was lower in Mp‐PAT.

Regulatory and anti‐inflammatory cytokine and chemokine production of IL‐27 and PARC

Regulatory T cells (T_reg) are potent suppressors of the adaptive immune response, with the ability to guide monocyte differentiation towards alternatively activated macrophages (AAM) 39. Chronic human filarial infection is associated with increased levels of immune regulatory cytokines produced by T_reg and the presence of monocytes characterized by an AAM phenotype expressing genes encoding the alternative activation markers resistin, MRC1, CCL18 (PARC) and MGL 40. A feature of AAM is increased production of CCL18 (PARC) 40, and as observed in the present study, PBMC from Mp‐PAT did not differ from CTRL in their capacity to produce CCL18 (PARC) in response to bacteria‐ or helminth‐specific antigens, suggesting that cellular production of CCL18 (PARC) was equilibrated in both study groups.

The spontaneous IL‐27 production by PBMC was higher in Mp‐PAT than in CTRL, and IL‐27 responses above baseline were inducible only in CTRL. The properties of IL‐27 to modulate inflammation, both by promoting IL‐10 as well as by antagonizing Th‐17 responses, will limit infection‐induced pathology 41. In children, the levels of regulatory IL‐27 rose with an increasing number of parasite infections 42, but with life‐threatening severe malaria the plasma levels of IL‐27 were reduced while proinflammatory chemokines were markedly high 37, 38. The elevated proinflammatory Th2‐type eotaxin‐2 (CCL24) and regulatory IL‐27 cytokine persisted in Mp‐PAT, but to what extent such a response profile may control, facilitate or prevent patent M. perstans infection, or limit pathogenesis, remains unanswered.

The mixed expression of chemokine and cytokine cellular responses with mansonelliasis

In patients with patent M. perstans infection, we found spontaneously elevated and distinctively inducible proinflammatory CXCL8 (IL‐8), Th2‐type eotaxin‐2 (CCL24), monocyte‐derived chemokine MDC and regulatory IL‐27 responses in Mp‐PAT, while the cellular production levels of the Th1‐type chemokine CXCL10 (IP‐10: inducible by IFN‐γ) were depressed. Such a response profile suggests that the blood‐circulating Mf of M. perstans may activate neutrophil and eosinophil granulocyte‐mediated defence mechanisms, and in parallel M. perstans will stimulate regulatory cytokine responses which dampen aberrant inflammation. A similar response profile was observed in mansonelliasis patients in Cameroon; their inflammatory type IL‐17A and T helper types 1 and 2 cytokines IFN‐γ, IL‐10 and IL‐13 were enhanced upon restimulation with M. perstans antigen extract. The monokine inflammatory protein 1β (MIP‐1β) was also measured at significantly higher concentrations, while the serum levels of CXCL8 (IL‐8) and CCL5 [regulated upon activation normal T cell expresssed and secreted (RANTES)] were lower in patients than in M. perstans‐negative individuals 43. In Mali, patients with W. bancrofti and/or M. perstans infections, the plasma levels of IL‐10 were elevated but the chemokine IP‐10 concentration low and, concurrently, bystander malaria antigen‐induced cellular production of IP‐10 diminished while IL‐10 was high 44. In those infected with either W. bancrofti or M. perstans, the frequencies of malaria‐specific Th1 and Th17 T cells were dramatically reduced, and such a response profile may alter specific T cell responses to concomitant parasite infections 45. In hookworm co‐infected Mp‐PAT, the cellular production of CXCL10 (IP‐10) was diminished in response to helminth AscAg, OvAg and SmAg and Entamoeba (Em) antigen extracts, and such lessened Th1‐type CXCL10 (IP‐10) production in filariasis patients could attenuate inflammatory immune responses associated with, e.g. severe malaria 38, 46.

Conclusions

In patients with patent M. perstans infection the innate cellular chemokine and cytokine production was at higher levels than in mansonelliasis‐free endemic controls, and cellular responses of both types 1 and 2 were selectively inducible by helminth‐, bacteria‐ and protozoa‐specific antigens and mitogen stimulation. The observed mixed Th1‐, Th2‐ and regulatory‐type cellular reactivity supports that M. perstans will not broadly suppress innate and adaptive immunity in patients. Such a non‐polarized cytokine and chemokine response profile with patent infection may facilitate M. perstans persistence and account for the lack of severe vascular and lymphatic immune‐mediated pathology, but this immune adaptation may alter resistance and susceptibility to other protozoan and metazoan parasites which are co‐endemic in the study area.

Author contributions

B. W., K. K., R. G. G., P. S. V. and P. T. S. conceived, designed and performed the experiments. M. B., R. G. G. and P. T. S. recruited and examined patients. B. W., K. K., V. A., P. W. P., P. T. S. and C. K' analyzed the data and wrote the manuscript. All authors read and approved the final manuscript.

Disclosure

The authors have no conflicts of interest to declare.