Cytokine Levels and Profiles in Children Related to Sickle Cell Disease and Asthma Status

Introduction

J amaican children with sickle cell disease (SCD) are more likely to have asthma than Jamaican ethnically matched controls (Knight-Madden and others 2005). The pathophysiology underlying this association is unclear. Asthma is associated with a pro-inflammatory T helper cell (Th)-2 cytokine profile with elaboration of more interleukin (IL)-4 stimulated blood mononuclear cells (Park and others 1996; Magnan and others 2000). In addition, there is more messenger ribonucleic acid for IL-4 in bronchial biopsies from asthmatics than controls (Humbert and others 1996). IL-4 plays several roles in atopic inflammation, including the up-regulation of vascular cell adhesion molecule (VCAM)-1 expression on vascular endothelial cells, increasing IgE secretion by B lymphocytes, inducing the hypersecretion of mucus, and recruitment of cells into areas of allergic inflammation (Steinke and Borish 2001). Up-regulation of VCAM-1 could be very significant in patients with SCD, as VCAM-1 is responsible for the binding of sickle erythrocytes to the vascular endothelium, leading to vascular injury (Pathare and others 2003). Indeed, increased levels of VCAM-1 are associated with increased morbidity in patients with SCD.

Studies of cytokine levels in patients with SCD have yielded conflicting data. Some studies have reported higher IL-4 levels in SCD patients. Measurable serum levels of IL-4 were found in 13% of SCD patients, but only 2% of controls (Taylor and others 1997). One study found that IL-4 levels were higher in SCD patients at steady state than during a veno-occlusive crisis (Pathare and others 2004), but another study found the opposite (Musa and others 2010). Raghupathy and others (2000) reported higher IL-4 levels in patients with SCD than in controls and detectable levels of IL-4 were found in a larger proportion of patients with SCD than controls (69% versus 33%). IL-2 levels and IFN-γ levels were similar in patients with SCD and controls hence Th-2 cytokine patterns were observed more frequently in patients with SCD than in controls (Raghupathy and others 2000). In one study, IL-4 was undetectable in 21 SCD patients at steady state (Bourantas and others 1998), whereas in another lower levels of IL-4 were reported in SCD patients compared with healthy controls (Rautonen and others 1992). Asthma status was not reported in any of the above studies (Rautonen and others 1992; Taylor and others 1997; Bourantas and others 1998; Raghupathy and others 2000; Pathare and others 2004; Musa and others 2010) and a possible explanation for the different results found in the various studies could be the proportion of patients with concurrent asthma in the different locations. Indeed, we found that 47.5% of children with SCD in Jamaica had asthma (Knight-Madden and others 2005) but only 9.4% of those in the United Kingdom (Sylvester and others 2007). Further among the UK children, asthma was not more common in SCD children than in controls (Sylvester and others 2007). The aim of this study was to test the hypothesis that cytokine levels and profiles of SCD children reflected their asthma status. To determine whether findings in Jamaica were due to local factors or more generalizable to other populations with SCD, cytokine profiles were also tested in a sample of children in the United Kingdom. Control participants were required as there are no published data regarding IL-4 levels or cytokine profiles in healthy Jamaicans. As a consequence, cytokine profiles were examined in SCD children and ethnically matched controls living in Jamaica or London and related to asthma status.

Materials and Methods

A random sample of 80 children with hemoglobin (Hb) SS disease aged 5 to 10 years was recruited from the Sickle Cell Unit, Jamaica (Knight-Madden and others 2005). Each child was matched to a control subject by age, address, and ethnic group. In London, 53 with HbSS disease were recruited from 2 specialist clinics and 11 age and ethnic-matched controls were also recruited (Sylvester and others 2007). The study was approved by the University Hospital of the West Indies/University of the West Indies/Faculty of Medical Sciences Ethical Committee and the King's College Hospital Research Ethics Committee. Parents gave informed written consent for their child to take part.

Blood samples were obtained when the children were clinically well and had had no acute symptoms for at least 2 weeks. The blood was spun to yield serum samples. All serum samples were stored at −70°C and thawed at the time of assay. Undiluted samples were assayed in the same laboratory using the same methods. Circulating levels of interferon γ (IFN-γ) and IL-4 were measured by a sandwich enzyme-linked immunosorbent assay (ELISA) using sensitive validated assays (reagents from R&D Systems) as previously described (Hussain and others 1996, 1998). One hundred microliters of monoclonal anti-cytokine antibody, diluted in 100 mM bicarbonate buffer (pH 9.5) to a concentration of 5 μg/mL, was coated onto each well of ELISA plates (Maxisorp immuno-plates; Nunc) overnight at 4°C. Plates were washed 3 times in phosphate-buffered saline (PBS) supplemented with 0.1% bovine serum albumin (BSA; Sigma Chemical Co.) and 0.01% Tween 20 (Sigma), and nonspecific binding sites were blocked by incubation with dilution buffer (PBS with 1% BSA) for 1 h at room temperature. One hundred microliters of standard solutions or undiluted serum was added for 1 h at room temperature with constant shaking using a Dynex IS89 Shaker (Dynex Technologies). After washing 100 μL of biotinylated monoclonal anti-IL-4 and IFN-γ was added and diluted 1/200 in dilution buffer supplemented with 2% mouse serum and incubated for 1 h at room temperature with constant shaking. After washing, 100 mL of horseradish peroxidase-conjugated streptavidin (Dako) was added, diluted 1 in 2,000, and incubated for 1 h at room temperature while constantly shaken. The reaction was developed using 100 mL of substrate [0.4 mg/mL o-phenylenediamine in citrate phosphate buffer (pH 5.0) containing 4 mL/mL of 3% hydrogen peroxide] and terminated with 50 mL of 4N H₂SO₄. The optical density of the samples was read in a Tecan Sunrise (Tecan) automated plate reader at 490 nm, and cytokine levels were read from standard curves constructed using recombinant cytokine (R&D Systems). The sensitivity of the assays was 10 pg/mL for IL-4 and 25 pg/mL for IFN-γ. The reproducibility of the assay has been previously reported (Hussain and others 1996, 1998).

Results

Jamaican children with SCD when compared with Jamaican controls were significantly more likely to be diagnosed with asthma (P=0.001), more likely to be IL-4 positive (P<0.001), and more likely to be classified as having a Th-2 pattern (<0.001, Table 1). British children with SCD, when compared with the British controls, were less likely to have been diagnosed with asthma (P=0.04) and less likely to be classified as having a Th-2 pattern (P=0.006, Table 1). Regression analysis demonstrated that among Jamaican children, SCD status, but not asthma status, acute chest syndrome (ACS) history, or gender, was predictive of IL-4 positivity and Th-2 status (P<0.001). In the British children none of those variables were significant predictors of IL-4 positivity or Th status.

Discussion

We have demonstrated that although IL-4 positivity, Th-2 status, and asthma were associated with SCD genotype, the relationship was opposite in the Jamaican compared with the British children. In the Jamaican children, SCD was associated with higher risk of asthma and a Th-2 pattern, whereas the reverse was true in the children living in the United Kingdom. Th-2 pattern has previously been reported in adults (Raghupathy and others 2000) and children (Taylor and others 1997) with SCD. The underlying mechanism is unclear. SCD is a chronic inflammatory disease marked by acute inflammatory episodes (Hebbel and others 2004; Murphy and Knight-Madden 2006). The chronic inflammatory vasculopathy is associated with an increased number of circulating endothelial cells (Strijbos and others 2009), which have been shown to be associated with increased levels of IL-4 (Kowal-Vern and others 2005). We hypothesize that an increased number of circulating endothelial cells may have accounted for the Th-2 pattern observed in this study, but further work in which VCAM-1 and circulating endothelial cells are measured is necessary to test that hypothesis. Some authors have reported decreased Th-2 cytokines in individuals with SCD. In one study, IL-4 was undetectable in 21 SCD patients at steady state (Bourantas and others 1998), whereas in another lower levels of IL-4 were reported in SCD patients compared with healthy controls (Rautonen and others 1992). Our hypothesis that differences in IL-4 levels and Th-2 status between SCD populations were due to differences in asthma status has not been supported by the data.

Differences between the cytokine levels in the children living in Jamaica and in the United Kingdom were not due to technical issues, as the samples were stored under similar conditions and the cytokines assayed in the same laboratory. The differences may be due to differences in the environment to which the children were exposed. Worm infestation is more frequent in tropical climates, where children play outside all year round. In Jamaican primary school children, the prevalence of Trichuris trichiura was reported to range from 42% to 47% and that of Ascaris lumbricoides from 15% to 37% (Wong and others 1994). In addition, Le Souef and others (2000) have postulated that in tropical environments, owing to the need for defense against helminths, there is a predisposition toward genes related to Th-2 immune responses. The tropical climate in Jamaica may promote more exposure to particular allergens such as cockroach (Montealegre and others 2004; Knight-Madden and others 2006). A further explanation for our results is that immunization exposures are also different. Basic immunizations covered by government programs in Jamaica do not include the Hemophilus influenzae booster at 18 months, hepatitis, influenza, Prevnar, or chicken pox vaccines. These may be more commonly given to British children. There are few data regarding the immune responses to vaccines in healthy children (DeStefano and others 2002).

We found that IL-4 levels were increased in Jamaican children with SCD who had been well for at least 2 weeks. These findings were similar to those of Raghupathy and others (2000) examined children living in Kuwait and who also found higher levels of IL-4 in SCD children than controls. They report similar levels of IL-4 positivity to our findings with SCD (54.9% versus 69%) and controls (22.1 versus 33.3%). Musa and others (2010), working in Nigeria and reporting on an older SCD group (22.4±6.3 years), found that IL-4 levels were normal in patients who had been in steady state for 2 to 3 months, but were elevated in those who recently had a vaso-occlusive crisis. The apparent difference between our results and those of Musa and others (2010) may be due to a different definition of steady state and the different ages of the patients investigated. It remains to be determined how much of the elevation in IL-4 is associated with a vaso-occlusive crisis. We report that IFN-γ levels were more frequently measurable in children with SCD than in controls, which partially agrees with that of Musa and others (2010), as they reported elevated levels of another Th-1 cytokine (IL-2), in SCD patients and not in controls. Raghupathy and others, however, found no difference in IFN-γ levels between SCD children and controls. These differences, as before, may have been explained by environmental factors that impact significantly on cytokine profiles and as a result influence the predisposition to complications of SCD. The incidence of ACS was higher in the Jamaican children than the UK children in our study and they had different cytokine profiles. Further, there may be genetic differences regarding Th status. For example, among tribes in Africa, the Fulani are more likely to have increased expression of TH2-related genes (IL-4 and GATA3) compared with sympatric Mossi from Burkina Faso (Torcia and others 2008).

The children were recruited into a study to assess the effect of asthma on SCD and this may have influenced who agreed to take part. Indeed, the asthma prevalence of the controls in the United Kingdom was 36.4%, higher than the 24.3% and 22.9% noted in previous studies (McCann and others 2002; Pearce and others 2007). The control population in the United Kingdom, however, was small, so caution must be exercised in interpreting this difference. In Jamaica, the asthma prevalence of 21.3% in the controls was similar to the 20.8% reported in Jamaican school children (Nichols and Longsworth 1995). A limited number of cytokines were assayed, but they were the cytokines that typify the Th-1 and Th-2 responses.

In conclusion, these data suggest that cytokine profiles, in particular IL-4 levels, may be different in children in developed and developing countries and also in children with SCD compared with ethnic-matched controls. It would be interesting to explore whether these differences are due to genetic or environmental factors.