Study populations
Participants were recruited based on responses to a questionnaire mailed three years (in 2007) after the outbreak to all persons with laboratory confirmed giardiasis during the outbreak [9]. Patients who reported chronic fatigue in this questionnaire were invited to participate in a thorough clinical evaluation and screening two years later (in 2009). Fifty-three individuals agreed to participate, and went through a clinical evaluation by specialists in internal medicine, psychiatry and neurology. They were evaluated for CFS or idiopathic chronic fatigue (ICF) according to the 1994 Fukuda criteria [12]. Those fulfilling the criteria, and had an onset of symptoms related to the Giardia infection, were categorized as PI-CSF or PI-ICF. Patients with sleep apnea syndrome, significant depression or anxiety disorders that could plausibly explain their fatigue were termed “fatigue other cause”. Individuals who had recovered well from the fatigue condition they had reported in the questionnaire two years previously were termed “fatigue recovered”. Five patients were excluded from this study after clinical evaluation (Fig. 1).
Two control groups were recruited; 22 individuals with normal fatigue score (=11) in the 2007 questionnaire (exposed, no-PI-fatigue group), and 10 healthy individuals not affected by the outbreak and without particular fatigue or abdominal symptoms (unexposed healthy controls) (Fig. 1). All participants were HIV negative and were not taking immunomodulatory medications or antibiotics.
Sampling and questionnaires
Participants were screened with a battery of routine blood tests and a magnetic resonance imaging (MRI) brain scan. Blood samples were taken between 08 am and 09 am after overnight fast and analyzed in parallel during the same period. Immunophenotyping results from these investigations have been reported previously [13]. Fecal samples were obtained and screened with microscopic examination and Giardia 18S PCR [14] of feces to rule out chronic giardiasis.
A total of 69 of the study participants underwent clinical characterization and were subject to one or more of the Giardia-specific immune response analyses presented in this paper. Some assays could not be done in all patients due to limited cell numbers (Fig. 1). Sampling and assays were stratified across groups and samples were blinded to laboratory personnel to avoid analytical bias. The severity of fatigue at the time of sampling was evaluated in all participants by the Fatigue Questionnaire [15], a validated set of 11 questions addressing different aspects of fatigue. Comorbid abdominal symptoms were recorded by the commonly used Rome II questionnaire [16].
Antigens
The Giardia antigens used in this study were made from culturing Giardia assemblage A strain WB-C6 (ATCC 50803) and assemblage B strain GS/M (ATCC 50581) in Diamond’s TYI-S-33 medium supplemented with bile as described previously[17]. Trophozoites were washed in PBS, treeze-thawed, then sonicated for 1 min at 20 W. After sentrifugation at 13000 x g, the supernatant containing soluble Giardia proteins was removed, and protein content was measured by the BCA protein assay kit (Pierce). Pilot testing showed that 10ug/mL of this mixed soluble Giardia antigens resulted in robust T cell responses and little background stimulation.
For the proliferation assay also a sterile filtered Candida albicans protein extract (403 skin prick test [Allergopharma]; 10 000 BU/mL), were added as an exploratory antigen and anti-CD3 + anti-CD28 as a control of T cell activation. Positive antigen controls were tuberculin purified protein derivate (PPD) (Statens Serum Institut) and Salmonella typhi LPS in all assays. Several concentrations were tested in pilot studies of all stimulation antigens and positive controls to optimize the assay.
PBMC acquisition and antigen-specifc immunity assays
Peripheral blood mononuclear cells (PBMC) were isolated from BD Vacutainer Na-citrate CPT tubes (BD, Franklin Lakes, NJ, USA) by density gradient separation. After harvesting, the PBMC were washed twice in PBS and were cultured in the presence or absence of investigational or control antigens at 37 °C in a humidified atmosphere of 5% CO2for six days in X-vivo 15 serum-free culture medium supplemented with l-glutamin, gentamicin, and phenol red (BioWhittaker).Proliferative responses were measured in triplicates of stimulated cultured cells by adding 3H-thymidine (Amersham International) after five days, and harvesting 18 h later. IIncorporated radioactivity was analysed by liquid scintillation counting in a β-counter. Proliferation in stimulated and unstimulated cultures was determined as median counts per minute (cpm) of each triplicate.
Activation of T cells was evaluated after six days in culture. 100 μL supernatants were carefully harvested from wells of unstimulated and stimulated PBMCs and frozen at −80 °C for later cytokine analysis. The cultured cells were then analysed by flow cytometric measurement of activated T cell subsets as published previously [17]. Briefly, after washing with PBS, the cultured cell suspensions (50 μL) were stained for 30 min in the dark with combination of 5 fluorescent dyes: CD3-ECD (Beckman Coulter), CD8a-FITC and CD4-PerCP/Cy5.5 and the activation markers CD26-PE and CD25-PE/Cy7 (BioLegend). After staining, cells were washed once, resuspended in PBS-paraformaldehyde solution (1%) and analyzed the same day using a Beckman Coulter Cytomics FC 500 MPL flow cytometer. In a typical acquisition 7×104 lymphocytes (min 2.3 ×104, max 1.7 ×105) were collected. The collected data were analyzed with FlowJo 7.6 software (Tree Star Inc, Ashland, OR, USA). Background responses in unstimulated cultures were adjusted for by subtracting these from responses in stimulated cultures.
Supernatants were kept frozen until analysis in three Bio-Plex assays (Bio-Rad Laboratories Inc., Hercules, CA, USA) for IFN-γ, TNF-α, IL-1β, IL-2, IL-4, IL-6, IL-9, IL-10, IL-13, IL-17A, IL-22, soluble (s) sCD40L, macrophage inflammatory protein 1 alpha (MIP-1α), MIP-1β, TGFβ1, TGFβ2, TGFβ3 and granulocyte macrophage colony-stimulation factor (GM-CSF) according to the manufacturer’s instructions. The cytokines TGFβ1, TGFβ2, TGFβ3 were analyzed in unstimulated and Giardia antigen stimulated cultures only. Observed concentrations (pg/ml) within the standard range were used for analysis. Values above the limit of quantitation were set to the highest value in the standard range, and values below this range were set to zero. For analytes where less than 50% of values were within range, further analyses were not done.
For both flow cytometric and cytokine assays, background responses in unstimulated cultures were subtracted from those in stimulated cultures for each participant. In some cases this gave negative values in the cytokine assay which were kept in the statistical analyses and interpreted as stimulant-induced decrease in production or increase in consumption of the relevant analyte.
Statistical analysis
Unless otherwise stated the data are presented as median (standard deviation (SD)). Chi-squared tests were used for categorical comparisons between groups. Linear regression analysis was used for correlation between response parameters as well as fatigue scores and sCD40L data. Comparison of T cell activation, proliferation and cytokine data between exposed and unexposed groups was done using the Mann Whitney U test. The significance level was set at p <0.05. To reduce false positive findings due to multiple comparisons within the exposed groups, we used the Kruskal-Wallis test across all exposed groups first, and further testing between the no PI-fatigue and PI-CFS/PI-ICF groups with Mann Whitney U test was only performed for variables with a p value less than 0.05. IBM SPSS Statistics version 23 (IBM Corp, Armonk, USA) was used for statistical analysis.