| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the Philippines, a study on the long-term impact of chemotherapy has provided some indications that acquired resistance to reinfection also occurs in S. japonicum infections.6 Using failure-time analysis on more than 5,000 individuals followed up over a period of eight years, children 7–13 years old were shown to get reinfected faster compared with older individuals (14–35 years old). This significant decrease in the risk of reinfection in the older age groups, which was significantly associated with prior infection, suggests that acquired resistance to reinfection with S. japonicum occurs in humans chronically exposed to this parasite.
The approach of comparing immune responses of two extreme groups of individuals, whether resistant or susceptible as distinguished by actual exposure, has been adopted by a number of investigators. Some of these responses have been correlated with resistance to reinfection. These include, among others, the development of a specific IgE response and increased levels of peripheral blood eosinophils.10 Further evidence for the role of IgE among resistant individuals was demonstrated in studies in Kenya,11 Brazil,12 and Zimbabwe.13 IgA has been implicated not only in eosinophil-mediated killing of parasites,14 but also in anti-fecundity functions15 in S. mansoni. In S. haematobium infection, the IgA response to soluble worm antigen preparation (SWAP) showed a negative correlation with intensity of infection.13
In contrast, susceptibility was correlated with the appearance of IgM and IgG2 antibodies2 and an overproduction of IgG4 antibody isotypes directed against carbohydrate epitopes of schistosome eggs antigens.16 These studies suggested a strong correlation with blocking of immunity in susceptible individuals, implicating downregulation of the immune system. Although fewer studies have been undertaken on S. japonicum, significant correlation of IgE responses to adult worms with age17 and with resistance18 have been reported, suggesting a similar pattern to that observed in S. mansoni and S. haematobium infections. Furthermore, a recent study showed that human IgA significantly targets the candidate vaccine molecule paramyosin,19 underscoring the potential for designing a paramyosin-based mucosal vaccine.20
Here we describe the antibody isotype patterns to native and recombinant parasite antigens resulting from S. japonicum infection among 155 individuals living in an endemic community in Leyte, The Philippines whose exposure, infection, and reinfection status had been observed and have been described in a three-year water contact study in a complementary paper.21 Specific antibody responses were correlated with each other and with several epidemiologic variables such as age, sex, and infection. This study was aimed primarily at evaluating antigen-specific antibody responses to native and recombinant antigens that are putative vaccine candidates against S. japonicum,35,36 In particular, their potential role, if any, in protection among putatively resistant individuals. In addition, antigen-specific antibody isotype responses were correlated with specific cellular immune responses the results of which are described in a complementary publication.22
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
All participants received yearly treatment regardless of parasitology results and therefore were assured of a zero infection baseline at the start of the water contact study. The data on water exposure and reinfection observations that were collected were used to describe the age- and sex-specific exposure and reinfection profiles of the study population at the start of the water contact study and after three years of follow-up treatment.
The water contact study site comprised approximately 216 hectares of land area. Thirty water observation sites were selected from a pilot ocular survey that identified potential human exposure sites. Local field workers were hired as observers for the water contact activities for the three year duration of the study. Water contact data collected included duration of contact in minutes, degree of contact to approximate percent skin surface exposure, activity types, and time of day.
Following a baseline parasitology diagnosis of a five-stool sample Kato-Katz (duplicate slides per stool sample = 10 slides) and treatment with praziquantel (50 mg/kg of body weight) given for all individuals enrolled in the water contact study (N = 412) regardless of the parasitology results, a 10-month post-treatment water contact observation (three cycles) was conducted. After every three cycles of water contact observations, parasitology diagnosis of five-stool Kato-Katz slides and treatment with praziquantel for all enrolled were again conducted. For the duration of the three-year water contact study, data from 10 observation cycles (water contact), four treatments, and four parasitology diagnoses of five-stool follow-ups were collected. Four types of water contact and exposure information were recorded; snail infection at the site of water contact (proportion of infected snails), the degree of water contact based on the "rule of nines" (percent of body surface), the duration of water contact (minutes), and the time of day of water contact. The time of day was scored using an arbitrary scale of 1 to 10. The time-of-day score was computed based on the diurnal pattern of cercarial shedding.
In the analysis of the water contact data (unpublished data), four predictors of infection were found: snail infection, snail infection rate, frequency of exposure, and time of day of water contact. Based on a literature review, the a priori exposure index was formulated. This weighted index of potential exposure (WIPE) followed the functional formula:
 |
The WIPE (or exposure score) was computed as an aggregate value representing the exposure of an individual over a period of three years. The exposure values were categorized in quartiles giving the highest potential exposure on the 4th quartile exposure scores.
Egg output intensities remained very low throughout the duration of the study period. The average eggs per gram (EPG) of stool or egg output categories were therefore classified as low (1–19 EPG), moderate (20–99 EPG), and high (
100 EPG) egg counts instead of the World Health Organization categorization used for infection intensities.
Resistance was assumed for individuals with the following criteria:
1) individuals who initially had a moderate or high EPG and subsequently with a low EPG at least twice on follow-up despite having a high potential exposure value (belonging to the 3rd and 4th quartile of computed WIPE); 2) individuals who were initially infected and on follow-up became uninfected despite having a high exposure value; and 3) individuals who still had a low EPG at least twice on follow-up despite having a high exposure value.
Susceptibility was assumed for individuals with the following criteria: 1) individuals who initially had a moderate or high EPG and still had a moderate or high EPG at least twice on follow-up; 2) individuals who were initially uninfected and/or had a low level of infection and became moderately to highly infected at least twice on follow-up; and 3) individuals who became infected at least twice regardless of intensity of infection during the three-year follow-up with a minimal potential exposure value (belonging to the 1st and 2nd quartile of computed WIPE).
Human sera. Serum samples were collected from the 155 Macanip volunteer subjects. Sera were also collected from 17 healthy adults for comparative purposes. Nine were recruited from the Research Institute for Tropical Medicine (RITM) and eight were recruited from a nearby city and on interview had no history of travel to endemic municipalities. All procedures related to the study and the purpose of collecting blood were thoroughly explained to the subjects. They were told that this was totally voluntary and the subjects from the Macanip population were informed that if they refused to give blood it would not affect the standard care given to patients at the study site. All human adult participants or the parents or legal guardians of minors individuals were then asked to sign an informed consent form. The study was thoroughly reviewed and approved by the RITM Institutional and Ethical Review Board.
Approximately 10 mL of blood was drawn from the Macanip subjects and the healthy controls into plain VacutainerTM tubes and sera were separated by centrifugation using standard procedures. The sera were aliquoted and stored frozen until transported in cold packs back to the RITM for antibody isotype determinations.
Recombinant antigens. Five S. japonicum recombinant antigens were used: 97-kD full-length paramyosin (rPMY), 37-kD glyceraldehyde-3-phosphate dehydrogenase (rGAPDH), 14-kD fatty acid binding protein (rFABP), 22-kD tegumental membrane-associated antigen (rTEG), and 28-kD glutathi-one-S-transferase (rGST) were cloned, expressed, and purified at the Queensland Institute of Medical Research in Brisbane, Australia (PhD theses, Acosta, LP, University of the Phillipines, Manila, 1999). Briefly, recombinant antigen clones were cultured in Escherichi coli and isopropyl-ß-d-thiogalactopyranoside at a final concentration of 2 mM added to induce expression of proteins. Cells were collected by centrifugation, lysed using lysozyme (1 mg/mL), and cell debris was removed by centrifugation. Expressed proteins were purified in TALONTM (Clontech, Palo Alto, CA) resin and eluted in an imidazole gradient. Imidazole was removed using PD-10 desalting columns (Pharmacia-Biotech, Uppsala, Sweden), according to the manufacturer’s instructions. Glycerol (10% w/v) was added as a protectant and the recombinant proteins were aliquoted and stored at –80°C.
Native antigens. S. japonicum (Philippine strain) adult worms were recovered from the portal mesenteric vasculatures of laboratory infected rabbits by perfusion with heparinized saline as previously described.28 Soluble worm antigen preparation was prepared from adult worms washed in phosphate-buffered saline (PBS), homogenized in a glass tissue grinder, and ultra-centrifuged for two hours at 45,000 rpm. Soluble egg antigen (SEA) was likewise prepared from homogenized eggs collected and isolated from trypsin-digested infected rabbit livers. Homogenized eggs were likewise ultracentrifuged at 45,000 rpm and the supernatant was collected and sterilized. Total protein concentration for both SWAP and SEA was determined by the Lowry method using the Sigma (St. Louis, MO) protein assay kit (P-5656).
Antibody isotyping. Levels of antigen-specific binding of antibody isotypes IgG1, IgG2, IgG3, IgG4, IgA, IgE, and IgM were determined for the recombinant and native adult worm (SWAP) and egg (SEA) antigens. Briefly, 100 µL of recombinant or native antigens in carbonate buffer (pH 9.6) were coated onto 96-well, polyvinyl chloride, activated microplate wells (Titertek; ICN, Aurora, OH) and incubated overnight at 4°C. Antigen protein concentrations were optimized and the following final working concentrations were used for all samples and isotypes: SWAP at 20µg/mL, SEA at 5 µg/mL, rPMY at 1 µg/mL, rTEG at 10 µg/mL, rFABP at 2 µg/mL, rGAPDH at 10 µg/mL, and rGST28 at 2 µg/mL. Blank plates without antigen-coated wells were also processed for all serum samples and isotypes to check for non-specific binding. Phosphate-buffered saline, pH 7.2, alone was used in duplicate wells as reagent control for all plates processed. Appropriate dilutions of sera (1:40 for IgE, 1:100 for IgG2, IgG3, and IgA, 1:200 for IgG1 and IgG4, and 1:800 for IgM) (100 [UNKNOWN CHARACTER ‘‘]L/well) were dispensed in duplicate wells and incubated for one hour at 37°C or overnight at 4°C. After the plates were washed four times with wash buffer (PBS/0.05% Tween20, pH 7.2), dilutions of anti-human immunoglobulins in incubation buffer (PBS/0.3% bovine serum albumin/0.05% Tween 20, pH 7.2) were dispensed (100 µL/well) and incubated for one hour at 37°C. All antibodies (anti-human IgG1, IgG2, IgG3, IgG4, IgE, IgA, and IgM) were obtained from ICN Biomedical (Cleveland, OH). A third antibody, a biotinylated anti-mouse immunoglobulin in incubation buffer (Amersham Life Sciences, Buckinghamshire, United Kingdom) at a 1:3,000 dilution was added (100 µL/well) and the plates incubated for one hour at 37°C. After washing the plates six times, 100 µL of streptavidin/horseradish peroxidase conjugate (Amersham Life Sciences) (1:3,000 dilution) was added to the wells and the plates were incubated for 30 minutes at room temperature. The plates were then washed eight times with wash buffer and o-phenylene diamine dihydrochloride (OPD) substrate (Sigma; a Fast tablet yields a final concentration of 0.4 mg/mL of OPD, 0.4 mg/mL of urea hydrogen peroxide, and 0.05 M phosphate-citrate buffer) (200 µL/well) was added and the plates were incubated for 30 minutes in the dark for color development. Optical density (OD) values at 450 nm were measured in a Bio-Rad (Hercules, CA) microplate reader immediately after 30 minutes of color development or read at 492 nm after the reaction was stopped with 50 µL of 3 M H2SO4 per 200 µL of solution. Cut-off values for a positive result were taken as the OD value above the arithmetic mean absorbance (OD) ± 3 SD of all normal control samples per assay plate.
Statistical analysis.
STATA TM /Statistical software version 4.0 (Stata Corp., College Station, TX) was used for all data analysis. Relationships between the different antigen-specific antibody isotype responses, and between various cellular immune responses (see Acosta and others22), several epidemiologic parameters such as age, sex, infection intensities (at the start of the water contact study and after three-year follow-up), and exposure scores were initially examined by Spearman’s rank correlation coefficient test and/or the Pearson’s correlation coefficient test on log transformed data. Relationships of particular interest used the Mantel-Haenzel 
2 test for differences in frequencies and the Student’s t test for differences between means. Linear trends were evaluated by logistic regression analysis to allow for the effect of age, sex, antigen-specific antibody and cellular responses to predict a dichotomous classification of putative resistant and susceptible individuals. Cut-off values for all antibody titers were computed per microplate as average OD of all negative or normal controls ± 3 SD OD values above the computed cutoff value per microplate were considered positive reactions. Low and high responders were set based on estimated 50th percentile values. All statistical significance were set at the P < 0.05 level, unless indicated.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows the overall positivity rate and the mean OD values of antigen-specific antibody isotype responses. Due to the enormity of the data generated with absolute antibody levels by age groups and sex, we summarized the results of the antigen-specific antibody isotype patterns as percentage values. Absolute antibody levels can be found in the report by Acosta (PhD theses, Acosta, LP, University of the Phillipines, Manila, 1999).
|
50 years old) age groups. Similarly, IgG3, IgM, and IgE response rates induced by SWAP were also high, with positivity rates of more than 50% and consistent in most of the age groups. The highest IgE response to SWAP was evident in the 15–19-year-old age group, while the IgG3 response was highest in the 50-year-old age group. In contrast, percent IgG1, IgG2, and IgA positive response rates to SWAP were low, with an overall positivity rate less than 25%.
Antibody responses to SEA were similarly high for all isotypes, with the lowest response seen for IgA. In contrast to the low IgG2 response to SWAP (9.8% overall response), high IgG2 responses to SEA were observed across the different age groups with an overall 66% positivity rate. IgA responses to SEA appeared to be comparable for all age groups, but were significantly increased in the 50-year-old age group. As with SWAP, more than half of the youngest age group (5–9 years old) had IgM antibodies to SEA.
Antibody responses to the recombinant antigens tended to be lower than those observed for the native antigens. Nevertheless, age-dependent antibody responses, particularly for IgG4, IgG2, and IgG3, were induced by rPMY, rTEG, and rFABP. The IgA response to rPMY increased significantly commencing with the 15–19-year-old age group. Positivity rates for IgG1 and IgE responses to rPMY, rTEG, rFABP, and rGAPDH were consistently low across the different age groups, with the highest IgE response elicited by rPMY (as in SWAP/IgE, also highest in the 15–19-years-old age group). This was followed by rTEG, with an overall response rate of 15%. Likewise, rFABP and rGAPDH appeared to be poor targets of antibody responses, with only a few individuals showing positive IgG4 and IgA responses to these antigens. No IgE, IgA, or IgG2 responses to rGAPDH were measurable. Furthermore, no detectable specific antibody-binding to rGST28 was evident with any of the antibody isotypes tested by an enzyme-linked immunosorbent assay.
Specific antibody profiles for males versus females.
Differences in response rates and mean OD values between males and females were tested using the 2 test and Student’s t test on log transformed mean absolute OD values. The results indicated that more males had significantly higher IgG4 responses to SWAP and SEA, and marginally significant higher IgG4 response to GAPDH (P < 0.07). Although there were more males with IgG4 responses to rPMY and rTEG, these were not statistically significant. More females appeared to have higher IgE responses to rPMY and rTEG, and all those who were IgE positive in response to rFABP were females. Conversely, more males had high IgA responses to SEA, rPMY, and rFABP. However, these differences in specific IgE and IgA responses between males and females were not statistically significant.
Correlations between antigen-specific antibody responses with age, exposure, and reinfection intensities.
Correlations using Pearson’s product moment coefficients in associating antigen-specific antibody responses with age, exposure, and reinfection intensities were computed on log-transformed data (Table 2). IgG4 and IgG2 antibody responses to SWAP, rPMY, and rTEG, and IgG3 responses to rPMY and rTEG appeared to correlate well with age. The IgA response to rPMY and rFABP and the IgE response to rFABP also showed a significant correlation with age. Exposure appeared to be significantly associated with IgG4 responses to SWAP, SEA, and rPMY and with IgG2 and IgG3 responses to SWAP and rPMY.
|
Table 3 shows the overall odds ratio (OR) correlating antigen-specific antibody responses with resistance or susceptibility. When we controlled for the effect of age and sex in stratified Mantel-Haenzel 2 analysis, IgG4 responses to SWAP showed a significant positive association with resistance among males but not in females and to rPMY among males in the older age groups (20 years old). Likewise, IgA and IgE responses to SWAP predicted a probability of resistance, but only in the younger age groups. Conversely, IgE responses to rPMY appeared to predict susceptibility among males in the older age groups (20 years old). While the IgG4 response to rPMY appeared to predict susceptibility in the younger age group, IgM responses to SWAP and SEA consistently predicted susceptibility.
|
. Negative associations were seen for IL-10 and SWAP-induced IgA and IgM antibody responses. IgE reactivity to rPMY antigen was also negatively associated with rPMY-induced IL-10 production. This implies a cross-regulating function for IL-10 in modulating Th2-type antibody responses.
Table 4 shows the results of logistic regression analysis on the dichotomous assumption of predicting resistance, allowing for the effect of age, sex, and cellular and antibody responses to specific antigens to predict probability of becoming resistant or susceptible to infection. The results suggest that age, sex, IFN- response to SWAP and rPMY, and IgA and IgM responses to SWAP are important predictors of susceptibility or resistance in this population. The age effect was significant only when taken as a binomial category (5–19 and 20 years old) and not as a continuous variable.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A weak but significant positive correlation was observed with age for IgG4 responses to SWAP (r = 0.1881, P < 0.005), which was found to be stronger for the recombinant antigens rPMY (r = 0.2844, P < 0.005) and rTEG (r = 0.3368, P < 0.005). IgG2 and IgA responses to rPMY and rTEG also showed some degree of age-dependent responses. In contrast with studies on S. mansoni and S. haematobium, we were unable to show any significant association with age and IgE responses to any of the antigens, but we were able to see some degree of association with these responses with resistance. This implies that in S. japonicum, resistance, particularly IgE responses, does not necessarily develop slowly with age. We have therefore shown a somewhat different picture of antibody profiles in the S. japonicum-endemic situation compared with the other schistosome species. The data complement similar antibody isotype studies in China on schistosomiasis japonica.29,30
In this study, degree of exposure was associated with antibody levels, as seen for the strong correlation between exposure and recombinant antigen (rPMY), implicating the need for specific antigens triggering immune memory. Sex differences in terms of antibody responses were also observed. Males showed significantly higher IgG4 responses to crude worm (SWAP) and egg antigens. While IgE responses to rPMY, rTEG, and rFABP were higher among females, IgA responses to SEA, rPMY, and rFABP tended to be higher among males. These differences in antibody responses between males and females have also been observed in other studies. Recently, several reports have pointed to sex31,32 and other intrinsic factors such as host genetics33 that can impact immunity in schistosomiasis and these need further investigation.
As indicated earlier, previous work on S. mansoni,1–4,11 S. haematobium,5–10 and S. japonicum17–19 has correlated resistance with high levels of IgE and/or IgA, while IgG4 and IgM may block IgE pathways. In this study, we did not see any clear association between IgE and IgA responses and resistance using simple analysis. However, after controlling for the effects of age and/or sex in a stratified analysis we conclude the following: 1) Negative associations were found for IgG4 response to rPMY among age groups 5–19 years old, thus predicting susceptibility in this age group (N = 62, P < 0.05). 2) There was a positive association of the IgE response to SWAP among females who were less than 20 years of age (N = 26, P < 0.01), thus predicting resistance. 3) There was a positive association of the IgA response to SWAP among those who were less than 20 years of age (N = 39, P < 0.05), thus predicting resistance in both males and females. 4) There was a negative association of IgM responses to SWAP and SEA (N = 144, P < 0.05), thus predicting susceptibility. 5) There was a contrasting result for SWAP response to IgG4 among males where a positive association was observed (N = 82, P < 0.05) (strong association for those in the older age groups) predicting resistance. 6) There was a negative association of the IgE response to rPMY in the older age groups (20 years old), thus predicting susceptibility (N = 82, P < 0.05).
Some of the data presented here are compatible with human studies on S. mansoni and S. haematobium. However, contrasting results with these studies were seen between the IgG4 response to SWAP, which appears to be associated with resistance, and between the IgE response to rPMY, which was associated with susceptibility. It appears in this S. japonicum-exposed community that resistance may be associated with mechanisms directed towards a Th1 response. Unlike other studies, where younger age groups elicited higher IgG4 responses, in this population, there were higher IgG4 responses in the older age groups. A recent study has provided proof that IgG4 could act as an inhibitory cytokine for IgE synthesis, which was observed to be differentially regulated by IL-10.34 Conversely, IgE appeared to be associated with susceptibility in older individuals. This also supports our hypothesis that in this population a Th1 response predominates, where IFN- could strongly mediate protection, of particular interest being the significant Th1-type immune responses elicited by rPMY, a well recognized vaccine candidate against schistosomiasis.20,35,36
When we tested the current antibody data and cytokine data presented in our complementary cellular study22 in a logistic analysis to allow for the effect of other variables in predicting resistance or susceptibility, several remained predictors of resistance: IFN- response to SWAP, IgA response to SWAP, IFN- response to rPMY, and IFN- response to rTEG. The IgM response to SWAP was consistently associated with susceptibility. Although this analysis strongly points to the importance of IFN- in protection, there were some indications that antibodies, particularly IgG4, IgE, and IgA responses to SWAP (an antigenic preparation containing numerous potentially protective epitopes), which are mediated by Th2 type responses, are also involved in inducing protection in humans.
Received May 28, 2003. Accepted for publication August 8, 2003.
Acknowledgments: The authors thank Fe Aligui, Genevive Hernandez, and Ernesto Abeto for expert assistance, and Dr. Li Yuesheng for his comments on drafts of the manuscript.
Financial support: This study was supported by the UNDP/World Bank/World Health Organization Special Program for Research and Training in Tropical Diseases and the National Health and Medical Research Council of Australia.
Authors’ addresses: Luz P. Acosta, Gemiliano dL. Aligui, and Remigio M. Olveda, Research Institute for Tropical Medicine, Alabang, Muntinlupa City, The Philippines, Telephone: 632-807-26-28, Fax 632-842-22-45, E-mails:lacosta{at}ritm.gov.ph, galigui{at}ritm.gov.ph, and remio{at}ritm.gov.ph. Donald P. McManus, Molecular Parasitology Laboratory, The Queensland Institute of Medical Research, 300 Herston Road, Brisbane, Queensland 4006, Australia, Telephone: 61-7-3362-0401, Fax 61-7-3362-0104, E-mail:donM{at}qimr.edu.au. Wil-fred U. Tiu, Department of Parasitology, College of Public Health, University of the Philippines, 624 Pedro Gil Street, Ermita, Manila, The Philippines, Telephone: 63-2-596808, Fax: 63-2-5211394, E-mail:cospeed{at}pacific.net.ph.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Jiz, J. F. Friedman, T. Leenstra, B. Jarilla, A. Pablo, G. Langdon, S. Pond-Tor, H.-W. Wu, D. Manalo, R. Olveda, et al. Immunoglobulin E (IgE) Responses to Paramyosin Predict Resistance to Reinfection with Schistosoma japonicum and Are Attenuated by IgG4 Infect. Immun., May 1, 2009; 77(5): 2051 - 2058. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. NARA, K. IIZUMI, H. OHMAE, O. S. SY, S. TSUBOTA, Y. INABA, A. TSUBOUCHI, M. TANABE, S. KOJIMA, and T. AOKI ANTIBODY ISOTYPE RESPONSES TO PARAMYOSIN, A VACCINE CANDIDATE FOR SCHISTOSOMIASIS, AND THEIR CORRELATIONS WITH RESISTANCE AND FIBROSIS IN PATIENTS INFECTED WITH SCHISTOSOMA JAPONICUM IN LEYTE, THE PHILIPPINES Am J Trop Med Hyg, February 1, 2007; 76(2): 384 - 391. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
