International Journal of Colorectal Disease

, Volume 21, Issue 2, pp 114–120

Decreased levels of interleukin-12p40 in the serum of patients with Whipple’s disease


  • A. Kalt
    • Department of DermatologyThe University of the Saarland
  • T. Schneider
    • Department of Internal Medicine ICharité, Campus Benjamin Franklin Hospital
  • S. Ring
    • Department of GastroenterologyDeutsche Klinik für Diagnostik
  • J. Hoffmann
    • Department of Internal Medicine ICharité, Campus Benjamin Franklin Hospital
  • M. Zeitz
    • Department of Internal Medicine ICharité, Campus Benjamin Franklin Hospital
  • A. Stallmach
    • Department of GastroenterologyMarienhospital, Katholische Kliniken Essen-Nordwest
  • D. H. Persing
    • Department of Clinical MicrobiologyMayo Clinic
    • Department of Internal MedicineSt. Josef Hospital
Original Article

DOI: 10.1007/s00384-005-0778-6

Cite this article as:
Kalt, A., Schneider, T., Ring, S. et al. Int J Colorectal Dis (2006) 21: 114. doi:10.1007/s00384-005-0778-6



An impaired production of interleukin (IL)-12 and T cell interferon-γ (IFN-γ) of in vitro stimulated monocytes has been discussed as a pathogenic factor in Whipple’s disease (WD). It is unclear whether this defect of cellular immunity is translated to the humoral immune system and to serum correlates.


We analyzed the serum of 40 patients with Whipple’s disease in various degrees of disease activity by sandwich enzyme-linked immunosorbent assay for differences in cytokine and cell adhesion molecule concentrations compared with age- and sex-matched controls.


We observed a highly significant reduction of IL-12p40 levels (patients, 0.18±0.05 ng/ml (mean±SEM); controls, 3.19±0.39 ng/ml; p<0.01) in all stages of disease activity, whereas the concentration of IL-12p70 was comparable with controls. Furthermore, we observed a slight decrease in tumour necrosis factor α (TNF-α) concentrations in the serum of patients (patients, 6.36±0.90 pg/ml; controls, 10.5±1.23 pg/ml; p<0,05). The levels of other cytokines such as IFN-γ, IL-2, IL-13 and transforming growth factor β, as well as soluble cell adhesion molecules lymphocyte function-associated antigen 3 and intercellular adhesion molecule 1, were not significantly different compared with controls. Levels of immunoglobulin G2 (IgG2) measured in the serum of WD patients were below normal in 24 of 29 patients and were even below the 95% confidence interval in 10 patients.


Our data demonstrate a persistent defect of the cellular immune response with decreased serum concentrations of IL-12p40 and TNF-α and decreased IgG2 levels in a large group of WD patients. These data support as in vivo finding the results obtained in previous investigations with stimulated monocytes/lymphocytes. The isolated decrease in IL-12p40 may hint at possible defects in the IL-12/IFN-γ promoter system.


Whipple’s diseaseImmunologyCytokinesImmunoglobulin



analysis of variance


enzyme-linked immunosorbent assay


intercellular adhesion molecule


interferon consensus sequence binding protein


interferon gamma


immunoglobulin G2




lymphocyte function-associated antigen




standard deviation


transforming growth factor beta


tumor necrosis factor alpha


Whipple’s disease


Whipple’s disease (WD) is a rare systemic disorder first described by George Hoyt Whipple in 1907 [1, 2]. Caucasian men are approximately eight times more often afflicted than women, and the disorder becomes clinically manifest mostly in middle age [35]. Frequent symptoms of this chronic relapsing disease are oligoarthropathies, followed by diarrhoea and severe weight loss [38]. Nearly every organ system can be affected.

The diagnosis of WD is often made by examination of small-bowel biopsies or biopsies of other organs showing the pathognomonic periodic acid–Schiff (PAS) positive inclusions [9, 10] that represent extracellular bacteria or bacteria within macrophages [11]. In addition, other diagnostic tests, including electron microscopy of specimens and polymerase chain reaction (PCR) [10, 12], can be used for diagnosis. Recently, it has become possible to culture the causative organism, Tropheryma whipplei, in human fibroblasts [1315] and to generate specific antibodies that in future could become useful in the diagnosis of WD [13, 14, 16, 17].

A coincidence of bacterial infection by T. whipplei [18], with alterations of cellular immune defence and defects of the macrophage system, has been discussed in the pathogenesis of WD. Of all patients, 28–44% have reported to be HLA-B27 positive [3, 19, 20].

The low prevalence of WD in the population combined with the probably ubiquitous occurrence of T. whipplei suggest a low pathogenic nature of the causative organism [21, 22].

The degradation of these bacteria by macrophages may be impaired, as intact intracellular bacteria have been found in bowel specimen [23]. Complement receptor 3, a molecule involved in phagocytosis and antigen processing and in mediating interferon-γ (IFN-γ)-induced killing of intracellular pathogens, has a reduced expression in WD patients [24]. Other investigations with stimulated peripheral and mucosal monocytes/lymphocytes of WD patients suggested an impairment of cytokine production, i.e. defects of monocyte interleukin (IL)-12 production, a reduced T-cell IFN-γ production and an increase in Th2-related cytokines (IL-4) [25, 26]. These data go along with the observation that growth of T. whipplei in macrophages was achieved by deactivation of phagocytes with IL-4 and IL-10 [27].

Because a reduced production of IL-12 in vitro has been reported, we examined IL-12 levels in vivo by analyzing serum samples. During inflammatory immune responses, the interaction by immune competent cells is arranged by cell adhesion molecules. In many chronic inflammatory states, elevated levels of soluble intercellular adhesion molecule 1 (sICAM-1) and lymphocyte function-associated antigen (LFA)-3, both occurring at the surface of macrophages and interacting with ligands on T cells, can be found.

Previous studies in a small patient group have also suggested reduced immunoglobulin G2 (IgG2) serum levels [25] (interpreted secondary to the Th1 deficiency [28, 29]). Therefore, we intended to investigate the serum concentration of cytokines and cell adhesion molecules that may translate immune defects in vivo and may serve as screening purposes.

Patients and methods

Patients and controls

Our study includes 40 patients with Whipple’s disease in various degrees of disease activity, compared to 30 healthy, age- and sex-matched controls. The mean age of WD patients was 64 years and 4 months, compared to 63 years and 9 months in controls. The female to male was 1:7 in WD patients and 1:9 in controls.

Blood samples of patients and controls were taken from July 1997 to January 1999. Serum samples were kept frozen at −20°C for a maximum of 3 years until usage; repeated defrosting was avoided.

The serum of 25 patients in group 3 (as described in the succeeding sections) was kindly made available by David H. Persing, M.D., Department of Clinical Microbiology, Mayo Clinic, Rochester, MN, USA.

To distinguish between transient and persistent defects in the immune system, the 40 WD patients were stratified into three predefined groups according to their disease activity (clinical signs and histology), as shown in Table 1.
Table 1

Patient groups

Group 1

Patients in acute disease stage → clinical symptoms, histology positive (3 cases)

Group 2

Patients with subclinical disease on antibiotics → clinical or histological still positive (5 cases)

Group 3

Patients in remission → without clinical or histological signs of disease activity (32 cases)

Distribution of patients into three groups according to their disease activity by clinical signs and histology

IL-12p40 and IgG2 levels were determined first. The other tests were done with the rest of the available sera. The correct number of analyzed patients is mentioned in the legends of the figures.


Serum samples of patients and controls were analyzed by sandwich enzyme-linked immunosorbent assay (ELISA) using paired antibodies to determine levels of cytokines and soluble cell adhesion molecules. Our investigations included the following parameters: IL-12p40, IL-13, IFN-γ, tumour necrosis factor α (TNF-α), transforming growth factor beta (TGF-β) (Pharmingen, Hamburg, Germany), IL-2, IL-12p70 (R&D Systems, Wiesbaden, Germany), soluble LFA-3 (sLFA-3) (as described in Hoffmann et al. [30]) and sICAM-1 (Southern Biotechnology Associates, Inc., Birmingham, AL, USA). Immunoglobulin G2 (IgG2) was measured by ELISA in the routine laboratory of the National Institutes of Health, Bethesda, MD, USA (Chief: T.A. Fleisher, M.D.); the normal ranges are from 282 to 576 mg/dl (mean 429 mg/dl).

Statistical methods

Evaluation of data was done by using the SPSS programme (SPSS Inc., Version 10.0.5, Chicago, IL). Data four times above the standard deviation were excluded. The Kolmogorov–Smirnow test was used to prove normal distribution. Significance of differences between patient groups and controls was determined by (1) analysis of variance (ANOVA) and the Scheffé procedure for the post hoc comparison (more than two groups) or (2) the Student’s t test (two groups).

If necessary, other non-parametric tests were used for subgroup analysis as indicated: (1) Kruskal–Wallis test (more than two groups) or (2) Mann–Whitney’s U test (two groups). Post hoc correction was done by Bonferroni–Holm test.

Differences were considered significant if the p value was <0.05.


In this study we analyzed serum samples from a total of 40 WD patients with different stages of the disease and from 30 healthy persons for in vivo differences with regard to cytokines from Th1 cells (IL-2 and IFN-γ), Th2 cells (IL-13 and TGF-β) and macrophages (IL-12 and TNF-α), as well as to soluble ICAM-1 and LFA-3 and to IgG2.

Decrease of serum IL-12p40 in WD patients, but normal levels of IL-12p70

Previous investigations with stimulated peripheral blood mononuclear cells (PBMC) and lamina propria lymphocytes (LPL) have shown a significantly decreased IL-12p40 production in vitro [25, 26].

As shown in Fig. 1a,b, we observed significantly reduced serum levels of IL-12p40 in vivo in all patients as well as in subgroup analysis of all three groups of patients as compared with controls. The average concentration of IL-12p40 of all patients was 0.18±0.05 ng/ml as compared to 3.19±0.39 ng/ml (p<0.001) in controls. In contrast, the levels of IL-12p70 in WD patients (Fig. 1c,d) were within normal range (all patients, 0.38±0.10 pg/ml; controls, 0.31±0.06 pg/ml).
Fig. 1

ad Serum IL-12p40 or IL-12p70 concentration as determined by ELISA. a, c Data of all WD patients and of healthy controls (box plot analysis). b, d Data (mean±SEM) of the three subgroups and of controls (no box plot analysis available because of the small size of patient groups). Although the numbers in the subgroups 1 and 2 are too small, we formally did a statistic of the data (b and d). IL-12p40: 35 WD patients (group 1, n=2; group 2, n=4; group 3, n=29 patients), 28 controls; statistical method, ANOVA. a Significantly reduced concentration of IL-12p40 in patients with Whipple’s disease (0.18±0.05 ng/ml) compared to controls (3.19±0.39 ng/ml). b No significant differences between the three groups of patients depending on their disease stage (group 1, 0.40±0.10 ng/ml; group 2, 0.06±0.02 ng/ml, group 3, 0.18±0.06 ng/ml). IL-12p40 concentrations in all subgroups are reduced to a similar degree compared to controls. IL-12p70: 15 WD patients (group 1, n=3; group 2, n=5; group 3, n=7 patients), 24 controls; statistical method, Kruskal–Wallis test. c No significant differences of IL-12p70 concentration between all WD patients (0.38±0.10 pg/ml) and healthy controls (0.31±0.06 pg/ml). d No significant differences between the three groups of patients depending on their disease stage (group 1, 0.37±0.21 pg/ml; group 2, 0.43±0.17 pg/ml; group 3, 0.35±0.19 pg/ml). Similar IL-12p70 concentrations in all subgroups compared to controls

Decrease of TNF-α in WD patients

Comparison of TNF-α levels (Fig. 2a,b) showed a reduced concentration of this cytokine in the serum of all patients (n=14) in contrast to control subjects (all patients, 6.36±0.90 pg/ml; controls, 10.5±1.23 pg/ml; p<0.05). In previous studies in a small group of WD patients, the level of TNF-α after stimulation of monocytes was not different from that of controls [26].
Fig. 2

Serum TNF-α concentration as determined by ELISA. Data including mean±SEM of 14 WD patients as a whole and three subgroups (group 1, n=3; group 2, n=4; group 3, n=7 patients), as well as 30 healthy controls (statistical method, ANOVA). Although the numbers in the subgroups 1 and 2 are too small, we formally did a statistic of the data (b). a Significantly reduced TNF-α concentration in the serum of WD patients (6.36±0.90 pg/ml) compared to controls (10.5±1.23 pg/ml). b No significant differences between the three groups of patients depending on their disease stage (group 1, 5±0 pg/ml; group 2, 8±3 pg/ml; group 3, 6±0.72 pg/ml)

Reduced serum levels of IgG2 in WD patients

A small previous study suggested reduced IgG2 serum levels among WD patients (interpreted secondary to the Th1 deficiency) [25].

As shown in Fig. 3, we observed significantly less IgG2 in the serum of WD patients than in the serum samples of controls. The mean of IgG2 was 292±28 mg/dl within the patient group, compared to 429 mg/dl (range, 282–576 mg/dl) in the reference group. In 83% of the levels (24 of 29), IgG2 measured in the serum of WD patients were below normal; 10 of 29 levels were even below the 95% confidence interval.
Fig. 3

Serum IgG2 concentration as determined by ELISA. Figures show data including mean±SD of 29 WD patients compared to a reference population (statistical method, Student’s t test). Reduced levels of IgG2 subclass in the serum of WD patients. The mean of IgG2 was 292±28 mg/dl within the patient group, compared to 429±147 mg/dl (95% confidence interval) in the reference population. In 83% (24 of 29 levels), IgG2 measured in the serum of WD patients were below normal; 10 of 29 levels were even below the 95% confidence-interval

Comparable levels of IFN-γ, IL-2, IL-13, TGF-β, sLFA-3 and sICAM-1

The evaluation of other cytokines or cell adhesion molecules did not demonstrate further significant differences. The levels of IFN-γ and IL-2 (both Th1-related cytokines), IL-13 and TGF-β (produced by Th2 cells) and of soluble LFA-3 and ICAM-1 (surface molecules of macrophages) were within normal ranges as compared to controls (Table 2).
Table 2

Comparison of serum cytokine and cell adhesion molecule levels in WD patients and controls


Group 1

Group 2

Group 3

All patients


IFN-γ (pg/ml)













IL-2 (pg/ml)













IL-13 (pg/ml)













ICAM-1 (ng/ml)



















LFA-3 (ng/ml)



















TGF-β (ng/ml)



















Data of IFN-γ, IL-2, IL-13, ICAM-1, LFA-3 and TGF-β: no significant differences between patients and controls

*Below detection limit

Lack of association of the immune defect to disease stage point to a persistent immune defect

In our set of data we were not able to reveal significant differences depending on the disease stage. In untreated patients with active disease (group 1), in treated patients with positive histology (group 2) and in patients without any evidence of disease (group 3), we observed similar results.


Alterations of the host immune defence have been discussed in the pathogenesis of WD. Our data demonstrate a persistent defect of cellular immunity in vivo independent of the disease stage, with decreased serum concentrations of IL-12p40, TNF-α and IgG2 in a large group of WD patients. These alterations indicate a basic defect of cell interaction and stimulation not due to activity of the disease, presence of bacteria or modulation by therapy.

Previous studies with stimulated monocytes and lymphocytes have shown a reduced IL-12p40 concentration in vitro [26]. In patients with reduced p40 levels (n=35), the level of p70 was essentially normal (n=15). In addition, the studies revealed reduced serum levels of TNF-α, whereas the ability of stimulated monocytes to produce TNF-α has been shown to be normal [26]. The reduced serum levels of TNF-α are more likely an unspecific finding and not a major factor in the pathogenesis of the disease. As another difference from in vitro studies, we found no decrease in IFN-γ levels or alterations of Th2 response in the serum. Overall, we interpreted these differences due to the investigation of different compartments.

Reduced levels of IL-12p40 point to an impairment of the interaction between macrophages and Th1 cells, which might be responsible for the impaired degradation of T. whipplei. Consequently, the activation of Th1 cells and the switch of specialized B lymphocytes to IgG2-producing B cells may be reduced. The dependence of IgG2 production on stimulation by IL-12-activated Th1 cells has been reported before [28].

A reduced level of IL-12p40 leads to a reduced activation of Th1 cells and therefore to a reduced production of IFN-γ [3133]. The production of IL-12p40 by macrophages depends on stimulation with bacterial lipopolysaccharides (LPS) or proteoglycans and interaction by cell adhesion molecules and IFN-γ [29]. Although it remains unclear whether the primary defect is the IL-12 or IFN-γ production, a defective IL-12p40 production seems to be more probable. IL-12p40 can be induced by LPS alone, whereas IFN-γ acts secondarily as a potentiator of IL-12p40 induction by providing a positive feedback mechanism and does not induce IL-12 by itself [3436].

The isolated decrease in IL-12p40 in all stages of WD, compared to normal levels of IL-12p70, may point to defects of the IL-12/IFN-γ promoter system. Recent investigations revealed the necessity of a specific protein in macrophages for production of the subunit p40. This protein—interferon consensus sequence binding protein (ICSBP)—is activated by stimulation of the IFN-γ-receptor (by IFN-γ or bacterial LPS) and regulates IL-12p40 transcription. Macrophages of mice lacking ICSBP are not able to produce selectively IL-12p40 after stimulation [36]. Furthermore, it is well known from earlier investigations that persons with mutations in the IL-12 gene or in the IL-12-receptor-β1-chain gene, which lead to a defective IL-12 response, suffer from infections with intracellular pathogens such as Mycobacterium avium-intracellulare [37, 38]. It seems possible that similar but more subtle defects facilitate the infection with T. whipplei and its clinical manifestation.

Previous findings of an increased IL-4 production may be rather secondary to the lack of inhibiting Th1 cytokines but may act as supporting factor, as they may deactivate macrophages [27].

Overall, the decrease in IL-12 might be one factor leading to impaired secondary immune function, decreased killing of the bacterium and finally to manifestation of the disease. Although it cannot formally be excluded that reduced IL-12 levels are a consequence of the disease, the data in patients in complete remission argue against it.

In further investigations, it will be interesting to determine why IL-12p40 induction and production are impaired in WD monocytes.


We would like to acknowledge Prof. Dr. Feurle, Neuwied, Germany, for the provision of serum samples of some patients of group 1 (n=3), which have already been included in a therapy study, Studie zur Initialtherapie des Morbus Whipple (SIMW) of WD.

Copyright information

© Springer-Verlag 2005