Changes in IgE- and Antigen-dependent histamine-release in peripheral blood of Schistosoma mansoni-infected Ugandan fishermen after treatment with praziquantel

Background Parasite-specific IgE levels correlate with human resistance to reinfection with Schistosoma spp. after chemotherapy. Although the role of eosinophils in schistosomiasis has been the focus of a great deal of important research, the involvement of other Fcε receptor-bearing cells, such as mast cells and basophils, has not been investigated in relation to human immunity to schistosomes. Chemotherapy with praziquantel (PZQ) kills schistosomes living in an in vivo blood environment rich in IgE, eosinophils and basophils. This releases parasite Ags that have the potential to cross-link cell-bound IgE. However, systemic hypersensitivity reactions are not induced by treatment. Here, we describe the effects of schistosomiasis, and its treatment, on human basophil function by following changes in total cellular histamine and in vitro histamine-release induced by schistosome Ags or anti-IgE, in blood samples from infected Ugandan fishermen, who are continuously exposed to S. mansoni infection, before and 1-day and 21-days after PZQ treatment. Results There was a significant increase in the total cellular histamine in blood samples at 1-day post-treatment, followed by a very significant further increase by 21-days post-treatment. In vitro histamine-release induced by S. mansoni egg (SEA) or worm (SWA) Ags or anti-IgE antibody, was significantly reduced 1-day post-treatment. The degree of this reduction correlated with pre-treatment infection intensity. Twenty-1-days post-treatment, SEA-induced histamine-release was still significantly lower than at pretreatment. Histamine-release was not correlated to plasma concentrations of total or parasite-specific IgE, nor to specific IgG4 plasma concentrations. Conclusion The biology of human blood basophils is modulated by S. mansoni infection and praziquantel treatment. Infection intensity-dependent suppression of basophil histamine-release, histamine-dependent resistance to infection, and similarities with allergen desensitisation are discussed as possible explanations of these observations.

still significantly lower than at pretreatment. Histamine-release was not correlated to plasma concentrations of total or parasite-specific IgE, nor to specific IgG4 plasma concentrations.

Conclusion:
The biology of human blood basophils is modulated by S. mansoni infection and praziquantel treatment. Infection intensity-dependent suppression of basophil histamine-release, histamine-dependent resistance to infection, and similarities with allergen desensitisation are discussed as possible explanations of these observations.

Background
High levels of circulating IgE are characteristic of both parasitic helminth infections and hypersensitivity conditions such as asthma and allergy. IgE and other Th2 mediated responses have been shown to be important in immunity to helminth infections. In human populations living in schistosomiasis endemic areas, high levels of IL-4, IL-5 [1,2], eosinophilia [3] and parasite-specific IgE are associated with resistance to reinfection after chemotherapy [4][5][6]. In previous studies in Kenya, levels of IgE specific for the adult Schistosoma mansoni worm, when measured after PZQ treatment but before re-infection, negatively correlated with subsequent reinfection intensities [7]. Specific IgE responses against Ags present in the outer tegument of the adult worm were also significantly associated with resistance to reinfection after treatment [8]. Human IgE and eosinophils have been shown to combine in antibody-dependent, cellular cytotoxicity mechanisms (ADCC) to kill early schistosome larvae in vitro [9]. However, this mechanism may not be as effective in vivo as, on penetration of its vertebrate host, the parasite rapidly disguises its outer tegumental surface by absorbing host Ag [10] and also becomes innately refractory to ADCC killing [11]. The roles of other major Fcε receptor-bearing effector cells such as mast cells and basophils has yet to be defined in human immunity to schistosomiasis. In vitro basophil studies have suggested a secretagogue potential of some S. mansoni Ag [12,13] or of plasma factors from infected patients [13], but the relationship between S. mansoni infection and basophils, and its relationship with human susceptibility to infection/reinfection, is not known. The role of basophils in allergy is an active area of research. Interestingly, it is suggested that allergic diseases are less prevalent in areas that are endemic for helminth infections and, when they are present, the manifestations of these diseases are less severe in helminth-infected individuals [14]. Various immune regulatory processes have been put forward as candidate mechanisms for the control of the potentially adverse effects of IgE responses in connection to both potential hypersensitivity to helminth Ags themselves and allergy in general [15].
Chemotherapy to kill schistosome worms whilst they are living in an intravenous environment that is rich in IgE, eosinophils and basophils, would seem to have the potential to induce a systemic hypersensitivity reaction. Orally administered PZQ, the drug of choice, is rapidly absorbed into the blood, where it can be metabolised within 90 minute [16]. Within one hour of contact with PZQ, the outer tegument of the worm is severely disrupted [17]. This rapid disruption of the worm tegument would lead to the exposure of worm Ags, some known to be recognised by IgE [7], directly to the blood. Despite this, only a very few heavily infected older individuals have transient hypersensitivity responses, usually within a few hours of treatment, such as urticaria and oedema [18]. This suggests that some aspect(s) of infection or reactions between infection and host response to infection, circumvents the most potentially damaging effects of systemic interactions between specific-IgE, mass-released parasite Ags and immune effector cells such as mast cells, eosinophils and basophils.
Here we describe the effects of schistosomiasis and the intravenous killing of the parasite on basophil function by following the changes in total cellular histamine content and in vitro basophil histamine-release induced by schistosome Ags or anti-IgE Ab. The studies were carried out using washed blood from infected Ugandan fishermen, before and at 1-day and 21-days after they were treated with PZQ.

Increases in total cellular histamine content of blood in S. mansoni infected individuals 1-day and 21-days after treatment
Washed blood, as defined in the Methods, was used to determine the total cellular histamine content of S. mansoni-infected individuals. By using total cellular histamine, rather than basophil number, we were able to compare the amount of histamine released with different stimuli with the amount of cellular histamine available in the blood at each of the times of sampling, irrespective of basophil number and degree of basophil activation. As most platelets are believed to be depleted during the washing steps and as basophils are the major contributor to blood cellular histamine, the most significant measured histamine was of basophil origin. Figure 1 shows that, between pre-treatment and 1-day post-treatment, there was a significant increase in total cellular histamine content of the blood of infected individuals (P = 0.019, n = 25), and a very significant increase by 21-days post-treat-ment (p = 0.001, n = 25). The greatest increase occurring between 1-day and 21-days post-treatment time points (p = 0.005, n = 25). Total cellular histamine content was also measured in the washed whole blood cells from eight individuals who had not been exposed to schistosomiasis at the same time as the 21-days post-treatment samples were processed. Values of median cellular histamine content for the infected study cohort at the three time-points and for the non-exposed group are summarised in Table  1. This shows that the median histamine content in the non-exposed group was close to the increased median histamine content found in the blood from the S. mansoniinfected group 21-days post-treatment. The non-exposed donors were not matched for age and sex with the infected cohort (and therefore are not considered as a control group), however, it is possible that the observed augmentation of total histamine after treatment may represent a return to a normal, uninfected, steady state rather than a treatment-related up-regulation. A classical interpretation of these observations would be that infected individuals have lowered basophil counts and after treatment their basophil counts return to levels comparable to those found in non-infected individuals. Alternatively, the low total cellular histamine content could be attributable to lowered histamine content per cell.

Changes in maximal histamine-release from the blood of S. mansoni-infected individuals 1-day after treatment
The maximal level of histamine released when washed whole blood cells were cultured in the presence of either anti-IgE Ab, or SEA or SWA schistosome Ag is shown in Table 2. It is expressed either as a percentage of total cellular histamine released, and thereafter called histaminereleasability, or as the absolute amount of released histamine in ng/ml. Histamine-releasability data reflect how much of the total cellular histamine present in a standard volume of blood can be released in response to either anti-IgE Ab or schistosome Ag. The absolute amount of released histamine reflects how much histamine a particular stimulant could release from cells in a standard volume of blood, whether this is a function of basophil numbers and/or their histamine-releasability state. Potential spontaneous release was subtracted from the final assay value during histamine measurement, thus only stimuli effects were measured. Figure 2 shows that before treatment, the basophils of most infected individuals were able to release histamine via an IgE-dependent pathway and after SEA or SWA stimulation. Basophils from individuals that had not been exposed to schistosomiasis showed no significant release upon stimulation with SEA or SWA used at the same concentrations under the same conditions (data not shown). Circulating specific and non-specific IgE is greatly elevated in chronically infected schistosomiasis patients [19,20] Increases in total cellular histamine content of blood from S. mansoni-infected individuals pre-treatment, 1-and 21-days post-treatment Figure 1 Increases in total cellular histamine content of blood from S. mansoni-infected individuals pre-treatment, 1-and 21-days post-treatment. The changes in total cellular histamine content (ng/ml) from washed blood of S. mansoni-infected individuals (n = 25) from pre-treatment levels to 1-day (black square) or 21-days (open square) after treatment with praziquantel. Any point that plots on the diagonal line is unchanged from the pre-treatment level. The differences in the levels of total cellular histamine were statistically significant between all time points and increased pre-treat-ment<1-day post-treatment<21-days post-treatment.  Median total cellular histamine content, expressed as ng/ml blood, from S. mansoni-infected individuals, and from non-infected individuals. Total cellular histamine content was measured in whole washed blood of each S. mansoni-infected individuals at pre-treatment, 1-day or 21days post-treatment, and in each non-infected individuals at the same time as 21-days post-treatment. Numbers in brackets are 25% and 75% percentiles, respectively. and this was verified in the present study. Median total IgE measured in the plasma of S. mansoni-infected individuals at pretreatment, 1-day and 21-days post-treatment was 587.5 UI (range 2603), 404.1 UI (range 2276.7) and 1616.5 UI (range 2850.4), respectively. The values are in the same range as previously reported in S. mansoniinfected patients [20,21] and are comparable to what has been reported in highly allergic individuals [22]. Noticeably, a rise in total IgE for a few months after the release of antigens in the blood has also been reported in allergic patients after desensitisation, but total IgE decreased in the longer term [23]. Our study did not include timepoints later than 21-days post-treatment but an elevation of serum total IgE has been reported 4 months after an anti-schistosomiasis treatment [21]. As might be expected, non-Ag-specific anti-IgE stimulation induces histaminerelease from the basophils from the study cohort. As shown in Figure 3, anti-SEA and anti-SWA IgE was detected in the plasma of the study participants at all time points and would, therefore, have been able to occupy basophil FcεR1.
At 1-day post-treatment, most individuals showed decreased histamine-releasability ( Figure 2). Histaminereleasability induced by SEA (p = 0.002, n = 21); SWA (p = 0.004, n = 23); and anti-IgE (p = 0.001, n = 23) dropped very significantly, as did the absolute amount of released histamine ( Table 2). This reduced ability to release cellular histamine, including after anti-IgE stimulation, suggests a general desensitisation, similar to that reported in the desensitization of individuals to specific allergens by either sub-optimal or super-optimal IgE receptor activation [24][25][26]. Praziquantel is a short acting drug that quickly causes the release of Ags from damaged worms and eggs. This may result in a rapid, in vivo basophil degranulation that desensitises the basophils to further Ag stimulation. Since total cellular histamine is low, the extent of such an early release would be limited, as shown in Table 2 by the absolute amount of released histamine lower at pre-treatment than at 21 days post-treatment, thus preventing the onset of hypersensitivity reactions after treatment. Additionally, any such early in vivo release of histamine may down-regulate further release through an autocrine mechanism involving H2 receptors [27].
A alternative explanation of this down-regulated histamine-releasability, in the presence of increased cellular histamine content, may be that after the worms and eggs destruction and subsequent release of parasite Ags by Praziquantel treatment, immature bone marrow basophils were freshly released into the blood. This mechanism has been suggested as the cause of increased basophil counts associated with reduced histaminerelease in allergic patients 5 days after treatment with rhG-CSF [28].

Changes in maximal histamine release from the blood of S. mansoni-infected individuals 21-days after treatment
At 21-days post-treatment, SEA or anti-IgE stimulated histamine-releasability by the blood of most patients was still significantly lower than it was before treatment (Figure 2; SEA: p = 0.017, anti-IgE: p = 0.059). Thus, it appeared that the blood basophils were desensitized 1day post-treatment and that, in relation to anti-IgE and SEA at least, this desensitised state was still evident 21days after treatment. However, the absolute amount of released histamine was significantly higher at 21-days post-treatment compared with pre-treatment levels with SEA (p = 0.009), SWA (p = 0.000) or anti-IgE Ab (p = 0.008) stimulation ( Figure 2 and Table 2). At 21-days post-treatment, histamine-releasability was not higher than pre-treatment % histamine-releasability, suggesting   (Fig 2a), SWA (Fig 2b) and anti-IgE (Fig 2c). Any point that plots on the diagonal line is unchanged from the pre-treatment level. The differences between the levels of histamine-releasability between pretreatment and 1-day post-treatment were statistically significant for all stimuli. The differences between the levels of histamine-releasability between pre-treatment and 21-days post-treatment were statistically significant and nearly significant for SEA-stimulated and anti-IgE-stimulated histamine release, respectively. that the observed higher absolute amount of released histamine was the consequence of the up-regulation of total cellular histamine content.
At 21-days post-treatment there would have been few, if any, parasites remaining in the blood, and little or no parasite Ag in the circulation [29]. For example, at this timepoint CAA, a diagnostic schistosome gut-associated circulating Ag, could not be detected in the plasma of 98% of the treated individuals (data not shown). Thus, basophils present at 21-days post-treatment would not have been subject to in vivo Ag challenge prior to in vitro culture. In these circumstances the basophils would have reacted normally to in vitro Ag stimulation via surface receptor bound Abs, including IgE. However, the continued reduction in histamine-releasability 21-days post-treatment, compared with the infected pre-treatment state, suggests the presence of a, yet to be identified, infection-associated priming factor that is removed or becomes ineffective after treatment.
We examined the possibility that, alternatively, an Ag-specific desensitisation, comparable to the immuno-therapeutic desensitisation of allergic patients, could have occurred. A role for IL-10 and TNFα of T-cell origin has been postulated in basophil desensitisation after wasp venom immunotherapy [30], and these cytokines were detectable in the plasma at all time-points. We found however no correlation between the plasma concentration of these cytokines and histamine-releasability after treatment in this study cohort (data not shown). The mechanism down-regulating histamine-releasability at 21 days post-treatment is therefore likely to be different from the one involved in therapeutic immuno-desensitisation.

Comparison of SEA-, SWA-and anti-IgE-induced histamine-releasability
We assumed that anti-IgE would induce the highest possible IgE-dependent histamine-releasability and compared it to that of SEA and SWA, to assess the secretagogue potential of the parasite Ag. As shown in Figure 2, both SWA and SEA induced significantly higher histaminereleasability than anti-IgE Ab at all study time-points: SEA Vs anti-IgE pre-treatment p = 0.002 (n = 32); 1-day posttreatment p = 0.002 (n = 29), 21-days post-treatment p = 0.004 (n-25); SWA Vs anti-IgE pre-treatment p = 0.004 (n = 32), 1-day post-treatment p = 0.002 (n = 29), 21-days post-treatment p = 0.000 (n = 25). There was no statistically significant difference between histamine-releasability induced by SEA and that induced by SWA at pretreatment and at 1 day post-treatment, but at 21 days posttreatment, SWA-induced histamine-releasability was significantly higher than SEA-induced histamine-releasability (p = 0.005, N = 25). Parasite factors, such as a S. mansoni analogue to human translationally controlled tumour protein (TCTP), may induce an additional IgEindependent histamine release, or enhanced non Ag-specific, IgE-dependent release [13,31]. These non-classical pathways could therefore be involved in the additional histamine-release. However we didn't detect histaminerelease from basophils from non-infected individuals after SEA or SWA stimulation under the same conditions. This enhanced histamine-releasability could then appear to be specifically induced by the experience of infection. We however have to consider the fact that our methodology may have been less sensitive than those used in the description of these non-classical pathways and that low levels of histamine-release in non-infected individuals may not have been detected. Thus, it is possible that parasite factors may have directly induced additional or enhanced histamine-releasability.

Post-treatment changes in in vitro basophil sensitivity to anti-IgE and parasite Ags
The sensitivity of blood cells to parasite Ag or anti-IgE stimulation was assessed by determining the lowest of the 9 concentrations of Ag or of the 3 concentrations of anti-IgE used in the assay, that was capable of triggering a significant histamine-release. The lower the concentration required, the higher the blood cell sensitivity. When the changes in sensitivity between consecutive time-points were compared, significant or highly significant negative associations were found between the changes from pre-treatment to 1-day post-treatment and the changes from 1-day to 21-days post-treatment (r = -0.636, p = 0.001 for anti-IgE; r = -0.568, p = 0.003 for SEA; r = -0.465, p = 0.019 for SWA; n = 25, for all conditions). Thus, the individuals whose basophil sensitivity decreased from pre-treatment to 1day post-treatment tended to be the same individuals whose basophil sensitivity had increased by 21-days posttreatment.
Whereas histamine-releasability shows the potential strength of histamine-release, sensitivity reflects the Ag concentration required to induce degranulation. In a mouse experimental model, sensitivity, contrarily to histamine-releasability, was shown to be correlated to affinity of the IgE for the Ag [32]. This reduced sensitivity to Ag or anti-IgE Ab could reflect qualitative rather than quantitative changes in plasma antibodies. The parallel changes in histamine-releasability and in basophil sensitivity show that reduced basophil sensitivity could be partly responsible for the post-treatment reduction in histamine-releasability in blood culture after treatment.

The lack of relationship between anti-IgE-and Ag-induced histamine-releasability, and plasma total IgE, specific IgE and specific IgG4
Plasma levels of total IgE, anti-SEA and anti-SWA IgE and IgG4 were measured at the 3 study time points. Histamine-releasability was not associated with level of total IgE at any time-point. Indeed, despite a statistically significant increase in total IgE between pre-treatment and 21days post-treatment (p = 0.024, N= 24), anti-IgE stimulated histamine-releasability was lower in the blood of most patients at 21-days post-treatment. Histaminereleasability was associated with parasite-specific plasma IgE levels only for SEA-induced histamine-release with anti-SEA IgE at 1-day post-treatment (r = 0.513, p = 0.012, n = 23), when histamine-releasability was at its lowest. There was no statistically significant change in plasma anti-SEA IgE between pre-treatment and 1-day post-treatment, as shown in Figure 3, while there was a very significant decrease in SEA-induced histamine-releasability. No significant correlation between IgG4 levels and SEA or SWA Ag-stimulated histamine-releasability was found.
Thus it would appear that post-treatment changes in the level of histamine-releasability in vitro were regulated either at the cellular level or by serum factors affecting the in vivo sensitisation, priming or neutralisation of the basophils or of other surface-bound factors, other than ELISA-detectable parasite-specific IgE or IgG4. A similar lack of relationship between plasma Ab and histaminerelease has been reported from studies of cord-blood basophils passively-sensitized with plasma from S. mansoni infected adults [18] and RBL-2H3 cells transfected with FcεRI sensitized with plasma from allergic patients [33].

The influence of pre-treatment infection intensity on posttreatment anti-IgE-and Ag-induced histamine releasability
The intensity of S. mansoni infection pre-treatment, expressed in eggs per gram of faeces, was compared with the changes in histamine-releasability pre-and post-treatment. Figure 4 shows that at 1-day post-treatment, the time when histamine-releasability was lowest, the pretreatment intensity of infection correlated negatively with SEA-, SWA-and anti-IgE-stimulated % histamine-release. Thus, the greater the intensity of S. mansoni-infection pretreatment, the lower the releasability of blood cell histamine at 1-day post-treatment. This clearly demonstrates that the decline in histamine-releasability seen after treatment was dependent on the S. mansoni-infection rather than on the direct action of the drug itself. As the degree of 1-day post-treatment suppression of histaminereleasability was dependent on the intensity of infection, but is triggered by treatment, this suggests that suppression of % histamine-release is related to the amount of Ag that is released in vivo by the drug-induced disruption of different numbers of infecting parasites. This desensitized state persists until 21-days post-treatment in relation to SEA ( Figure 2). Interestingly, a negative association between level of Ag exposure and subsequent reactivity has been reported in bee-keepers allergic to bee venom, in whom the number of stings was negatively correlated to skin sensitivity and who showed less side reaction to venom desensitisation than other bee venom allergic patients who have been less exposed to bee stings before treatment [34].
Although the relationship between infection and low post-treatment histamine-releasability could result from an infection intensity-dependent down-regulation mechanism there is a particularly interesting alternative explanation. It has been suggested from studies in the murine model of schistosomiasis that parasite induced histaminerelease may regulate the intensity of schistosome infection by triggering inflammation reactions that prevent superinfection [35]. In relation to the present study, it is possible that a high histamine-releasability response to parasite Ag might be protective against infection/re-infection. If this were true, the association between low infection intensity and high histamine-releasability would result in a restriction of parasite numbers by an IgE-dependent histamine-release, not intense infection suppressing histamine-release. With regard to this hypothesis, it would be interesting to test if the magnitude of in vitro histaminereleasability in response to treatment was predictive of the subsequent resistance or susceptibility of individuals to re-infection with S. mansoni.

Conclusions
The present work describes changes in basophils biology and the modulation of basophil function, induced by the treatment of S. mansoni infection. Some of these changes may be a return to steady non-infected state. S. mansoni treatment-induced human basophil immune modulation is associated with pre-treatment infection intensity. The mechanisms involved in post-treatment basophil desensitisation at 1-day post-treatment may have similarities to those that induce desensitisation to allergens with immunotherapy. Analysis of histamine-release from naïve The relationship between S. mansoni pre-treatment infection intensity and the histamine-releasability from the blood of infected individuals, pre-and post-treatment, after stimulation with anti-IgE or schistosome Ag basophils passively sensitised with the plasma from the same individuals should provide data on the plasma factors involved in the regulation of the basophil response. It is hoped that such studies, coupled to the monitoring of reinfection after treatment, will provide detailed information about the role of basophils in IgE mediated immunological protection against reinfection after treatment and mechanisms by which the potential adverse effects of IgEmediated immune effector mechanisms are down-regulated. Adult men were, through their occupation, the part of the population the most exposed to infection. The select study group comprised males aged between 18 and 45 years old (mean age 34.5 years), having resided in Bugoigo for at least three years and consenting to participate. The selection was made after parasitological examination of three stool samples per individual, with two Kato thick smears per sample using 50 mg of faeces per slide [36]. All the selected individuals had detectable S. mansoni eggs, but those with over 70 eggs per slide were excluded to reduce the number of outliers. The mean pre-treatment egg count for the selected cohort was 282 (range = 983) eggs per gram of faeces.

Non-infected group
A group of 8 healthy volunteers, members of the sample collection team, was bled at the same time as the third bleed of the study group. This control group comprised 5 Africans (2 females and 3 males) and 3 Europeans (1 female and 2 males).

Blood collection and PZQ treatment
Informed consent was obtained from all those who participated in this study, in line with the National guidelines of the Ugandan Ministry of Health, whose ethical review committees approved all the protocols used, and the US Department of Health and Human Services. Thirty mlblood samples taken by venipuncture in heparinised syringes (10 U/ml heparin Na salt, Sigma, UK) were collected from the 33 participants before they received a single dose of 40 mg/kg body weight of PZQ. The participants were asked to come back to donate blood samples a second time (1-day post-treatment) exactly 24 hours after having been treated. A third sample was taken 21-days post-treatment from the 28 participants who came back to donate blood. Only 3 ml of the collected blood was used for histamine assays. The rest of the blood was used for other assays that were parts of the same main study. At the completion of this study the whole Bugoigo community was treated with PZQ.
Antigen coating of microtitre plates used for histamine release Glass fibre microtitre plates from Ref Lab, Denmark were coated with 25 µl of the following substances: Histamine (50 ng/ml) added to four wells; α-IgE from DAKO, Denmark, used in three concentrations (1:200 (7 µg/ml), 1:1000, 1:2000), each concentration was added in two wells; SEA and SWA, used in nine concentrations with the dilution factor 3.5 and each concentration was added into two wells. The highest concentration of SEA and SWA was 50 µg/ml. All dilutions were made in distilled water containing 5% glycerol. Subsequently the allergen coated microtitre plates were dried for 6 hours at 37°C, and thereafter the plates were packed and sealed until use in Bugoigo. Preliminary experiments using blood from 3 S. mansoni-infected patients showed that histamine-release performed in Ag-coated plates (stored for 3 months at 20°C) varied less than 5% from histamine release induced by freshly prepared Ag. Control experiments using blood from 5 non-infected individuals showed no histamine-release to SEA and SWA (data not shown).

Histamine release assay
Histamine-release was performed using the glass fibre assay described elsewhere [37], which allows the capture of released histamine, irrespective of basophil source, after stimulation. The assay was however simplified for direct release of histamine by peripheral blood without passive sensitisation. In brief 3 ml of each blood sample was washed twice in PIPES buffer (Ref Lab, Denmark) at room temperature to eliminate platelets and plasma factors not already cell-bound. The samples were reconstituted to the initial volume with PIPES buffer and thereafter referred to as "washed blood", with the addition of IL-3 (5 ng/ml washed blood). Substances in the Ag-coated plates were dissolved by adding 25 µl PIPES buffer to each well prior to addition of 25 µl washed blood per well. Histamine was released and subsequently bound to glass fibres in the microtitre plate. Thereafter the plate was washed with distilled water. The plates were stored in the dark at room temperature during the 21-days study period in Bugoigo and subsequently shipped to the Ref Lab for histamine analysis.

Histamine analysis
Histamine was measured by spectrofluorometry as described elsewhere [37]. Spontaneous release is automat-was part, DWD set up the larger study of which this work was part and wrote the manuscript.