- Top of page
- Materials and methods
Helminth infections are among the most common parasitic diseases in tropical countries worldwide (Hotez et al. 2008). After decades of research, it is now well recognised that they exert a profound impact on the host's immune response and are able to survive in the blood system, intestine, body cavities, or host tissues for several years (Maizels et al. 2004; Bethony et al. 2006). Besides the modulation of the immune response through the parasites, helminth infections cause chronic disease in their hosts, leading to fatigue, iron deficiency anaemia, growth stunting, malnutrition and poor cognitive development (Bethony et al. 2006). Regions with highest prevalences coincide with underdeveloped regions with poor hygiene, inadequate water supply and sanitation and often a population that lives on less than 2 US dollars per day (Hotez et al. 2008).
In impoverished communities, undernutrition is common, such as those of poor areas of northern and north-eastern Minas Gerais State (POF 2004; Jardim-Botelho et al. 2008). A general state of nutritional deficiency is associated with impairment of cell-mediated immunity, phagocyte function, cytokine production, secretory antibody response and antibody affinity (Chandra 2002). In conjunction with helminth infections, this might aggravate the health status of individuals, as found for Schistosoma japonicum infection and patients with hepatic fibrosis (Coutinho et al. 2005). Conversely, malnutrition may influence the immune responses against helminth infections in endemic areas. Experimental infections in mice demonstrated that protein malnutrition reduces Th2 responses in the intestine of the host and increases nematode burden (Ing et al. 2000). Vice versa, malnutrition can influence the development of the parasite in its host. In experimental S. mansoni infections, the development of male and female schistosomes has been shown to be impaired in undernourished mice (Oliveira et al. 2003; Barros et al. 2009), and the scarcity of host micronutrients also affected nematode development and reduced parasitic burden in experimental models such as hamsters (Held et al. 2006) or mice (Bourgeois et al. 2007).
In the present longitudinal study, we evaluated systemic serum cytokine and chemokine markers for inflammation and Th1/Th2 responses in relation to helminth infection, parasite burden and/or nutritional parameters.
- Top of page
- Materials and methods
Demographic data of enrolled individuals are shown in Table 1. Due to a higher proportion of female participants and due to differences in mean age between the infection groups, statistical analyses are adjusted for sex and age. According to WHO classifications (WHO 2002), the mean intensity of infection in the different infection groups can be classified as light to moderate for hookworm and A. lumbricoides and as moderate for S. mansoni, with co-infected individuals having higher egg counts than their respective mono-infected counterparts (Table 1). Between the different groups mono- or co-infected with the respective helminth species, participants co-infected with the two intestinal helminths had significantly higher egg counts for hookworm and for A. lumbricoides when compared with their mono-infected counterparts (groups 2 and 3) (P = 0.009 and P < 0.001, respectively).
Table 1. Demographic data on 210 patients from Americaninhas, Minas Gerais, Brazil
|Group||Egg-negative||Mono-infection with either hookworm or Ascaris||Dual infection with both hookworm and Ascaris||Dual or triple infection with S. mansoni and hookworm and/or Ascaris||Mono-infection with S. mansoni|
|Numbers||n = 32||n = 28||n = 38||n = 90||n = 22|
To compare the nutritional status of the participants and its influence on the immune response, height-for-age and body mass scores (CDC HAZ, BMIC, and BMI) were evaluated in all of the individuals and compared between the different groups. In children and adolescents, stunting was related in 19.5% of participants. Furthermore, 5.9% presented with underweight and 21.9% of individuals had excessive weight. No statistically significant correlation was found between any of the inflammatory cytokines or chemokines tested, such as IL-1β, IL-6, TNF-α or CCL3 (MIP-1α) and HAZ scores before or after treatment (data not shown). When body mass was evaluated before treatment, CCL17 (TARC) serum concentrations were significantly higher in the group of eutrophic individuals (geometric mean 505 pg/ml) than in the groups of under- or overweight (291 and 349 pg/ml, respectively, P = 0.02 from Kruskal–Wallis test). In contrast, 12 months after treatment, CXCL10 (IP-10) concentrations were significantly higher in the group of individuals with overweight than the other groups (P < 0.01 for each comparison). The intensity of infection did not seem to negatively influence the participants' body mass, because egg counts for hookworm and A. lumbricoides were significantly higher in eutrophic individuals than in individuals with under- or overweight (P < 0.001; and P = 0.03, respectively).
Hookworm egg counts were negatively correlated with serum ferritin and IL-6 levels at baseline (ρ = −0.224, P = 0.003; ρ = −0.194, P = 0.005, respectively). Also, Ascaris epg were inversely correlated with serum TNF-α concentrations (ρ = −0.150, P = 0.031), whereas Schistosoma epg before treatment and after re-infection were positively correlated with serum IL-10 levels before treatment (ρ = 0.161, P = 0.02 and ρ = 0.220, P = 0.002). Both hookworm and S. mansoni epg were correlated with serum CCL11 (eotaxin-1) concentrations (ρ = 0.156, P = 0.024 and ρ = 0.201, P = 0.003, respectively), as were S. mansoni epg with CCL17 (TARC) (ρ = 0.259, P < 0.001). Higher baseline concentrations of IL-10 and CCL17 were associated with higher prevalence of S. mansoni, but not of the other parasites, at follow-up (odds ratio for IL-10 of 1.8 for each 10-fold increase, P = 0.004 and of 4.9 for each 10-fold increase in CCL17, P = 0.005 by logistic regression).
Table 2 shows an age- and sex-adjusted comparison of peripheral blood parameters between the patient groups. For serum ferritin values, individuals infected with intestinal helminths and/or co-infected with S. mansoni showed significantly lower serum ferritin values than egg-negative participants. Considering percentages of leucocyte subpopulations, individuals with intestinal helminths and individuals co-infected with S. mansoni presented with significantly higher percentages of eosinophil counts than egg-negative individuals or people with S. mansoni mono-infection. By contrast, the percentages of monocytes were highest in the latter groups. For both monocytes and lymphocytes, the lowest values were found in the two groups with multiple infections (Table 2).
Table 2. Baseline blood parameters of patients from Americaninhas either egg-negative or infected with hookworm, A. lumbricoides, and/or S. mansoni. Indicated are serum haemoglobin (g/dl) and serum ferritin values (μg/l), as well as cell counts for total leucocytes (mm3) and percentages of peripheral eosinophils, monocytes and lymphocytes. Analyses between the different infection groups were adjusted for age and sex
| ||Baseline geometric mean (95% confidence interval)|
| || n ||Egg-negative (n = 32)||Infection with either hookworm or Ascaris (n = 28)||Infection with both hookworm and Ascaris (n = 38)||Infection with S. mansoni and hookworm and/or Ascaris (n = 90)||Infection with S. mansoni(n = 22)||Kruskal–Wallis P valuea||Bootstrap P-values for pairwise comparisons of geometric meansa|
|Haemoglobin (g/dl)b||207|| || || || || ||0.40||–|
|Serum ferritin (μg/l)||174|| || || || || ||0.02||0.03† 0.001‡ 0.009§|
|Leucocytes (mm3)||207|| || || || || ||0.002||<0.001† <0.001‡ 0.01§|
|Eosinophils(%)||207|| || || || || ||<0.001||0.03† <0.001 ‡ <0.001§ <0.001¶ <0.001‖ <0.001#|
|Monocytesb(%)||207|| || || || || ||0.02||0.02† 0.01‡ 0.04§|
|Lymphocytes(%)||207|| || || || || ||0.03||0.002† 0.008‡|
Before and 12 months after treatment, we analysed the cytokines and chemokines that are indicators for inflammatory (IL-1β, IL-6, TNF-α, CCL-3), regulatory (IL-10), Th1-type (CXCL10), or Th2-type (IL-5, IL-13, CCL11, CCL17) responses during human infections. At pre-treatment, geometric mean values for IL-1β, IL-5, IL-6, IL-10, and for IL13 from the different groups were low and close to the respective detection limits of each assay, considering mostly less than 30% of participants in each group as responders (Table 3). A considerable percentage of the individuals' serum samples throughout the groups contained TNF-α, CCL3, and CXCL10, and CCL11 and CCL17 was detected in over 90% of samples. For CCL17, the group with co-infection of intestinal helminths and S. mansoni showed significantly higher concentrations than egg-negative individuals or persons with a mono-infection of intestinal nematodes (P < 0.01 for each) (Table 3). Similarly, CCL11 was also significantly elevated in the group co-infected with intestinal helminths and S. mansoni when compared with the egg-negative group (P < 0.05). Even though the highest CCL11 concentrations were detected in Schistosoma mansoni-co-infected individuals, the differences were not statistically different to individuals with any intestinal helminth infection. Overall, these two chemokines showed higher mean values in the infected groups than in egg-negative individuals (Table 3).
Table 3. Pre-treatment serum cytokine and chemokine concentrations (pg/ml) in egg-negative individuals and patients infected with hookworm, A. lumbricoides, and/or S. mansoni. Indicated are geometric mean values with confidence intervals in brackets and percentages of responders. Numbers in brackets (n) indicate the numbers of volunteers in each group. For statistical analyses, a Kruskal–Wallis P-value is indicated for comparison within all the groups, and a bootstrap P-value is assigned for pairwise comparisons
| ||Baseline geometric mean pg/ml (95% confidence interval) and per cent above detection threshold by baseline infection groupa|
| || n ||Egg-negative (n = 32)||Infection with either hookworm or Ascaris (n = 28)||Infection with both hookworm and Ascaris (n = 38)||Infection with S. mansoni and hookworm and/or Ascaris (n = 90)||Infection with S. mansoni (n = 22)||Kruskal–Wallis P-value||Bootstrap P-values for pairwise comparisons of geometric means|
|IL-1β||210|| || || || || ||0.74||–|
|IL-5||208|| || || || || ||0.29||–|
|IL-6||210|| || || || || ||0.34||–|
|IL-10||208|| || || || || ||0.05||–|
|IL-13||204|| || || || || ||0.67||–|
|TNF-α||208|| || || || || ||0.54||–|
|CXCL10/IP-10||210|| || || || || ||0.47||–|
|CCL17/TARC||208|| || || || || ||0.01||0.002† 0.004‡|
|CCL3/MIP-1α||210|| || || || || ||0.30||–|
|CCL11/Eotaxin-1||210|| || || || || ||0.04||0.015†|
Table 4 shows the mean cytokine and chemokine concentrations in the different infection groups 12 months after each of the helminth species was cleared in all infected individuals by chemotherapy. 107 participants were newly infected with intestinal helminths and/or S. mansoni. Cytokine and chemokine geometric mean values for IL-1β, IL-5, IL-6, IL-10, IL-13 and TNF-α remained low 12 months after treatment, with negligible cytokine concentrations in sera of most individuals (Table 4). Mean CXCL10 concentrations were highest in participants multiply infected with intestinal helminths and S. mansoni, whereas in the other groups, the values dropped to lower levels compared with pre-treatment concentrations. Similar to pre-treatment values, CCL17 concentrations were significantly higher in mono- and in co-infections with S. mansoni. Also, CCL3 and CCL11 concentrations were highest in the group co-infected with intestinal helminths and S. mansoni. For CCL3, these differences were significant (Table 4). The conclusions and differences in cytokine and chemokine patterns between groups before and after treatment are not affected if calculations are adjusted for sex and age.
Table 4. Post-treatment serum cytokine and chemokine concentrations (pg/ml) in egg-negative individuals and patients infected with hookworm, A. lumbricoides, and/or S. mansoni. Indicated are geometric mean values with confidence intervals in brackets and percentages of responders. Numbers in brackets (n) indicate the numbers of volunteers in each group. For statistical analyses, a Kruskal–Wallis P-value is indicated for comparison within all the groups and a bootstrap P-value is assigned for pairwise comparisons
| ||Follow-up geometric mean pg/ml (95% confidence interval) and per cent above detection threshold by follow-up infection groupa|
| || n ||Egg-negative (n = 98)||Mono-infection with either hookworm or Ascaris (n = 44)||Dual infection with both hookworm and Ascaris (n = 36)||Dual or triple infection with S. mansoni and hookworm and/or Ascaris (n = 16)||Mono-infection with S. mansoni (n = 11)||Kruskal–Wallis P-value||Bootstrap P-values for pairwise comparisons of geometric means|
|IL-1β||202|| || || || || ||0.16||–|
|IL-5||201|| || || || || ||0.74||–|
|IL-6||202|| || || || || ||0.44||–|
|IL-10||202|| || || || || ||0.12||–|
|IL-13||202|| || || || || ||0.78||–|
|TNF-α||202|| || || || || ||0.54||–|
|CXCL10/IP-10||202|| || || || || ||0.03||†0.002|
| ||202|| || || || || ||0.01||0.002† 0.002‡ 0.010§ 0.035¶ 0.003‖|
| ||202|| || || || || ||0.03||<0.001† 0.004‡ <0.001§ 0.042¶|
|CCL11/Eotaxin-1||202|| || || || || ||0.19||–|
When changes in cytokine/chemokine patterns before and after treatment were investigated, we did not find any significant differences between individuals who cleared helminth infections and individuals who got re-infected with any of the parasite species under investigation (data not shown). If only S. mansoni infections were considered, we found a higher increase in CCL3 (MIP-1α) secretions in schistosomiasis patients than in participants who were not re-infected after treatment (P = 0.06). On the other hand, the decrease in CCL17 (TARC) concentrations was higher in those individuals who did not get re-infected with S. mansoni than in with participants re-infected with S. mansoni, where the CCL17 levels remained nearly unchanged. For the change in CCL17 secretion, these differences were significant between the re-infection group and the group that remained egg-negative 12 months after treatment. (P = 0.04) (Table 5). As serum concentrations for IL-1β, IL-5, IL-6, IL-10, IL-13 and TNF-α were low or lower than the detection limit, the fold changes were not shown in Table 5.
Table 5. Fold change in serum cytokine/chemokine levels in patients who remained egg-negative and patients who got re-infected with S. mansoni 12 months after treatment
| ||Geometric mean fold change from baseline to follow-up (95% confidence interval)||Kruskal–Wallis P-valuea|
| || n ||S. mansoni positive at baseline but egg-negative at follow-up (n = 83)||S. mansoni egg-positive at baseline and re-infected at follow-up (n = 26)|| |
- Top of page
- Materials and methods
In the present study, we followed the serum cytokine and chemokine profile of a cohort of individuals from an area highly endemic for intestinal helminths and for schistosomiasis before and 12 months after treatment. The impact of nutritional parameters, infection with different helminth species and intensity of infection was investigated in terms of altered immune responses, for example, markers for inflammation, regulatory IL-10 production and Th1 and Th2 responses. Before treatment, we could not associate any altered nutritional parameters with increased inflammatory responses. Also, the highest infection intensities were found in eutrophic participants. Most importantly, infection with intestinal helminths and/or S. mansoni was associated with higher CCL11 and CCL17 concentrations, rather than up-regulation of inflammatory cytokines. Both chemokines were found to be serological markers for multiple helminth infections, especially in combination with a S. mansoni infection.
In agreement with a former epidemiological study in this area (Fleming et al. 2006), our emphasis was laid upon the main, most prevalent helminth species and on individuals either mono- or co-infected with them, for example, N. americanus, A. lumbricoides and S. mansoni. Similar to the epidemiological study (Fleming et al. 2006), the randomly selected participants in the immunological study had higher egg counts when more than one helminth species was detected. Such synergistic effects between soil-transmitted helminths and schistosomes have also been reported by others (Chamone et al. 1990; Howard et al. 2002; Raso et al. 2006; De Cássia et al. 2007).
It is well recognised that malnutrition is the commonest cause of immunodeficiency and can aggravate various kinds of infectious diseases (Chandra 2002). However, in our participants we were not able to discover any association between abnormal nutritional parameters and increased inflammatory serum cytokines or chemokines. This might be due to the random selection of individuals and the limited total number of participants with altered nutritional parameters in this cohort. In contrary, the significantly elevated serum CCL17 concentrations in eutrophic individuals before treatment might merely reflect a state of immunocompetence and a response to the presence of multiple helminth infections throughout the population. The significantly elevated CXCL10 concentrations in serum from persons with overweight at 12 months after treatment may be attributed to the action and contribution of adipose tissue to the production of IFN-γ and inflammatory cytokines (Marti et al. 2001; Conway & Rene 2004). CXCL10 secretion is under genetic control of IFN-γ and is up-regulated in Th1-type inflammatory processes (Saha et al. 2010) and is thought to be a good marker for interferon-dominated responses (Dixon et al. 2000). Possibly, the higher CXCL10 serum concentrations in individuals with overweight became apparent due to the reduction in worm burden and diminished, helminth-induced immunomodulation.
Hookworm and Ascaris intensity of infection did neither correlate with under- nor with overweight. In contrary, highest intensities of infection were found in eutrophic individuals. This might be explained by the selection of a restricted number of persons for this immunological study, because in a former epidemiological study, stunting was significantly associated with A. lumbricoides and low body mass with hookworm infection in a subgroup of children and adults, respectively (Jardim-Botelho et al. 2008). Other studies with school children also found an association between stunting and high intensities of A. lumbricoides or S. mansoni infections (Assis et al. 2004; Payne et al. 2007, respectively).
As expected, several blood parameters were altered in individuals infected with STH or S. mansoni. Notably, when compared with egg-negative participants, serum ferritin concentrations and peripheral blood eosinophil counts were significantly altered in mono-infections with STH or co-infections with intestinal helminths and/or schistosomes, while altered to a lesser extent in S. mansoni mono-infections.
As no significantly elevated levels of inflammatory serum cytokines (IL-1β, IL-6, TNF-α) were detected in our egg-positive individuals at pre-treatment, our main attention was drawn to markers for a Th2-type or regulatory immune response. At pre-treatment, the most abundant chemokines detected in more than 90% of all individuals were CCL11 (eotaxin-1) and CCL17 (TARC). Serum concentrations of both of them seem to be linked to the presence of a S. mansoni infection, because both chemokines significantly correlated with S. mansoni egg counts and highest mean concentrations resulted in schistosomiasis patients, either mono- or co-infected with this parasite. The role of type-2-dominant CCL11 is mostly attributed to the activation and attraction of eosinophils (Teixeira et al. 1997) and most probably reflects the chronic exposure to and infection with different helminth species. For experimental S. mansoni infections in mice, it was shown that CCL11 was highly up-regulated during egg-antigen-induced granuloma formation in the lungs (Qiu et al. 2001). Production of CCL17 can be regulated by Th1- as well as by Th2-type cytokines (Qiu et al. 2001). However, As in our study, serum CCL17 concentrations are a marker for helminth and especially for S. mansoni infection, it is more likely that in this case, increased CCL17 concentrations were induced by Th2-type immune responses, as proposed by Zlotnik and Yoshie (2000). Similarly to our results, CCL17 secretion was elevated in supernatants of antigen-stimulated PBMCs from individuals multiply infected with helminths and protozoan parasites (Soboslay et al. 2006; Hamm et al. 2009). In another human trematode infection, increased CCL11 and CCL17 concentrations were also detected in pleural effusion samples from individuals infected with Paragonimus westermani (Matsumoto et al. 2002). Serum concentrations of inflammatory cytokines, such as IL-1β and IL-6, were low and only up to 20% of participants in each group were considered as responders. Even though percentages of responders were similar to Coutinho et al. (2006), there were no significant differences in our groups and IL-6 concentrations did not correlate with S. mansoni egg counts, but such a state of general inflammation might only be present in patients with a high intensity of infection or even in cases were hepatic fibrosis has already developed, as described for S. japonicum (Coutinho et al. 2005, 2006).
Twelve months after treatment, significantly elevated CCL17 concentrations persisted in S. mansoni mono- and co-infections. However, elevated CCL17 values are now accompanied by increased CCL3 and CXCL10 concentrations, especially in co-infected individuals. Because these two type-1-dominant chemokines are thought to be moderately up-regulated during early granuloma formation (Qiu et al. 2001), they might indicate an acute state of re-infection in our participants, rather than a marker for severe disease, as described in hepatosplenic patients (Souza et al. 2005). When compared before and after treatment, re-infection in these individuals was associated with a trend to increased CCL3 serum concentrations and maintained CCL17 concentration, with the latter being significantly higher than in individuals who remained egg-negative after treatment. In human schistosomiasis, it is well known that IL-10 plays a crucial role in the down-modulation of Th1-type immune responses (Sher et al. 1991), reduces immune-mediated pathology (Hoffmann et al. 2000; Hesse et al. 2004), and might also facilitate parasite survival. The higher percentages of responders for IL-10 in participants with S. mansoni and the correlations between serum IL-10 and S. mansoni egg counts, also described by others (Reimert et al. 2006), indicate a general state of immunomodulation in these individuals to diminish egg-induced pathology (Hoffmann et al. 2000). In human schistosomiasis, it was shown that eosinophils are a common source of IL-10, and these cells may therefore contribute to the modulation of the immune response (Kayaba et al. 2001), which, in our field study, might also have a crucial impact on intestinal helminth co-infections in terms of parasite survival, and reproduction.
In summary, our participants infected with STH and/or S. mansoni did not reveal any changes in inflammatory cytokines due to altered nutritional parameters or infection intensities, which might be due to in part to the reduced number of individuals in this cohort study and mostly to low-to-moderate infection intensities. In contrary, the presence of STH and/or S. mansoni is closely associated with higher type-2 serum chemokine concentrations, such as CCL11 and CCL17. The driving force for such altered chemokine responses seemed to be a S. mansoni infection, rather than hookworm or A. lumbricoides infections, as S. mansoni-co-infected individuals presented with elevated serum levels for these chemokines.