Similar T helper Th2-like cytokine mRNA expression in vernal keratoconjunctivitis regardless of atopic constitution

Abstract

Background: Many patients with vernal keratoconjunctivitis (VKC), a severe chronic allergic eye disease in children, exhibit IgE-sensitization, but about 40% of cases lack this immunologic trait. As a disease factor in VKC, IgE is thus not fully understood. The aim of this study was to investigate whether there are any differences in the conjunctival cytokine messenger (m)RNA pattern related to IgE-sensitization in children suffering from VKC.

Methods: Tissue samples were obtained from 16 symptomatic VKC subjects with sub-tarsal disease and six control subjects. Expression of mRNA for interleukin (IL)-4, IL-5, IL-13, and interferon (IFN)-γ was investigated with a semiquantitative reverse-transcriptase polymerase chain reaction (RT-PCR) technique. The presence of T cells, IgE⁺ cells, mast cells, and eosinophils was analyzed with immunohistochemical methods. Allergen-specific IgE antibodies were assessed in serum and with skin prick testing.

Results: Ten out of the 16 VKC subjects showed evidence of IgE-sensitization. No differences were detected for any tissue variable between VKC subjects with and without IgE-sensitization. Statistically significant increases over controls were found for both VKC groups with regard to all cell markers.

Conclusions: The amount of messenger RNA encoding cytokines and inflammatory cell markers in VKC did not correlate with IgE-sensitization. Tissue changes in all patient samples were characterized by a prevalence of T cells, eosinophils, mast cells and cell-bound IgE molecules. However, the role of cell-bound IgE molecules in VKC patients lacking IgE-sensitization remains to be determined.

According to the recently published EAACI position paper for nomenclature of allergic diseases, the rare entity of vernal keratoconjunctivitis (VKC), a chronic severe ocular disease with hypertrophic conjunctival changes occurring mainly in young boys, is classified as an IgE-mediated disorder (1). Indeed, findings in tissue samples from VKC, confirm the probable implication of mast cells, eosinophils and cell-bound IgE (2, 3). The atopic constitution, defined as presence of allergen-specific IgE antibodies, is further common among sufferers of VKC. However, many VKC patients do not exhibit evidence of IgE-sensitization. In a recent retrospective survey of 62 VKC subjects in the Stockholm area, only 60% had a definite diagnosis of IgE-mediated allergy (4). The role of IgE-sensitization in VKC is obscure for yet another reason, since many VKC patients do not present a disease course that varies with the exposure to the relevant sensitizing allergens (5). In addition, the most common ocular IgE-mediated disorder, seasonal allergic conjunctivitis due to pollen exposure, is not accompanied by gross papillary hypertrophy, the hallmark sign of VKC.

We have previously investigated the cellular composition of conjunctival specimens from children suffering from VKC (3). There were no apparent differences between the subjects with and those without IgE-sensitization, but the sample size was not large enough to allow a statistical analysis. Moreover, the study did not address the involvement of cytokines, now recognized as key regulators of allergic tissue responses. The cytokines interleukin (IL)-4 and IL-13 share many biological effects, and they are both instrumental in the immunoglobulin class switching to IgE production in B cells (6, 7). IL-5 promotes recruitment and survival of the eosinophilic granulocytes (8, 9). Although these cytokines may originate from various types of cells, an important source is a T helper cell subset, designated Th2 (10). The proallergic actions of Th2 secreted cytokines are thought to be inhibited by interferon (IFN)-γ, a cytokine that is mainly derived from another T helper cell subset, Th1 (11).

With the objective of further defining tissue inflammatory features in relation to IgE-sensitization, conjunctival biopsies from VKC subjects were examined for mRNA cytokine expression with a semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR) technique. In addition, immunocytochemical markers for mast cells, eosinophils, T cells, and IgE-bearing cells, all pivotal in the chronic allergic infiltrate, were analyzed.

Material and methods

Subjects and biopsies

Full-thickness papillary tissue was excised from the upper mid-tarsal area in 16 patients with present signs and symptoms of palpebral VKC. A majority, 69%, of the subjects showed corneal disease in the form of punctate erosions. They all had at least a two-year disease history. No topical or systemic treatment had been given for a minimum of one week preceding the operation, which was performed to alleviate symptoms. In addition, upper mid-tarsal conjunctival biopsies from six aged-matched children were obtained during squint surgery. These control subjects had no history of any allergic manifestation, ocular inflammation or previous eye surgery. All specimens were excised with a knife under general anesthesia. The tissue, having a minimum surface of 2 × 5 mm and a depth of 1 mm, was split and placed in 4% formaldehyde or snap frozen and stored at −70°C for further processing. The sampling procedure followed the tenet of the Declaration of Helsinki. Informed consent was obtained from the subjects and their parents. The project was approved by the local ethics committee.

Screening for specific IgE antibody

Study and control subjects all underwent testing for IgE-sensitization. A positive diagnosis was defined as presence of IgE antibody to common inhalant or ingested allergens. A preliminary screening test was done with Phadiatop^®, Pharmacia CAP system, which determines IgE against any constituent of a panel of common aeroallergens. Serum IgE specific for individual inhalant allergens was further assessed with the Phadebas radioallergosorbent test^® or CAP RAST FEIA^® (Pharmacia). Values ≥ 0.35 kU/l were classified as positive. Food allergen sensitization was investigated either with skin prick testing (Soluprick^®, ALK, Copenhagen, Denmark), where a wheal diameter ≥ 3 mm was deemed positive in the presence of positive histamine and negative diluent controls, or with fx5^®, Pharmacia CAP system, RAST FEIA^®, a serum analysis of a range of common alimentary allergens. Total serum IgE was also analyzed with Phadebas IgE PRIST^® or Pharmacia CAP system IgE FEIA^®, normal reference range: 2–263 kU/l for individuals 4–20 years of age.

Immunohistochemical staining

Acetone fixed cryostat sections were stained with the three-step peroxidase-antiperoxidase (PAP) method as previously described (12). The sections were stained with anti-Leu-4 reacting with CD3, ‘pan T cells’ (Becton-Dickinson, San Jose, CA) and anti-IgE, clone C3102E8 (Southern Biotechnology, Birmingham, AL). Rabbit antimouse immunoglobulin (Dako, Copenhagen, Denmark) was used as the secondary antibody. Specificity tests included omission of the primary antibodies. Staining was not observed in these specimens. For five VKC subjects (involving three with and two without IgE-sensitization), the results of a previously performed alkaline phosphatase-antialkaline-phosphatase (APAAP) assay that had made use of the same primary antibodies were included (3), since the amount of frozen material was too limited to allow a PAP analysis.

Paraffin-embedded tissue was cut, deparaffinized, rehydrated and stained with the avidin-biotin complex (ABC) technique according to a previously employed protocol (3). The primary antibodies used were EG2 directed against eosinophil cationic protein and secreted forms of eosinophil protein X (Pharmacia & Upjohn) and AA1, which stains mast cell tryptase (Dako). With the AA1 antibody, pretreatment with pronase 0.05%, was performed. For one VKC subject, no material for formalin fixation was available.

The sections were viewed in a light microscope by one masked observer (P.M.). One representative section out of a set of three for each staining was analyzed. Cells with a positive coloration of the membrane or the cytoplasm according to the antibody staining property were counted manually at 500 × magnification. The total number of cells, regardless of epithelial or chorion localization, is given per 1 mm cross-section unit length.

RT-PCR for mRNA cytokine expression

Total RNA was extracted from approximately 50 sections, 6 µm thick, from each frozen specimen using RNAzol B^® (Biotecx Laboratories; Houston, TX). The extraction procedure followed the manufacturer's instructions save for the RNA precipitation that was done in the presence of glycogen (Boehringer Mannheim AB, Bromma, Sweden). The final RNA pellet was air-dried and resuspended in diethyl-pyrocarbonate-treated water. RNA quality was assessed and confirmed by electrophoresis on a 1% agarose gel (BRL, Gaithersburg, MD, USA).

First-strand cDNA was synthesized from total RNA with Omniscript^TM Reverse Transcriptase (Qiagen, KEBO, Stockholm, Sweden) according to the manufacturer's instructions in the presence of pd(T)₁₂₋₁₈ primer (Amersham Pharmacia Biotech, Uppsala, Sweden), and the RNase inhibitor RNasin (Promega, SDS, Falkenberg, Sweden) for 1 h at 37°C. The nucleotide sequences of the glyceraldehyde-3-phosphate dehydrogenase (G3PDH), IL-4, IL-5, and IFN-γ primers were identical to those provided by Clontech (Intermedica, Stockholm, Sweden). The IL-13 primers were described previously (13).

All biopsy samples to be analyzed for target cytokine gene expression were first investigated for expression of the ‘housekeeping’ gene G3PDH by competitive PCR. This procedure compensates for differences in the amounts and quality of the RNA extracted, and the efficiency of the reverse transcription in each sample. A competitive PCR experiment was performed in a final volume of 20 µl containing 2 µl of target cDNA (in appropriate dilutions determined in preliminary experiments) and 2 µl of internal standard cDNA (human pcr G3PDH MIMIC^™, Clontech). For each sample, three-fold dilution series of internal standard was run, 10⁻²−10⁻⁴ attomole/µl. PCR was conducted for 36 cycles with the following conditions: an initial denaturation step at 96°C for 1–2 min, denaturation at 94°C for 1 min, annealing at 60°C for 1 min, and extension at 72°C for 1 min, followed by a final extension step of 5 min at 72°C and cooling down to 4°C. The PCR products were separated on a 1.6% agarose gel (BRL). The ethidium bromide stained gel was photographed under UV light and analyzed with the Kodak Digital Science™ Electrophoresis Documentation and Analysis System 120 (Eastman Kodak Company, Rochester, NY). The data were imported into Microsoft™ Excel and analyzed as previously described (13). The samples were subsequently diluted accordingly in order to get a normalized concentration of G3PDH mRNA (0.0008 pM, 266 molecules) in each sample for the analyzes of cytokine mRNA expression.

The amount of biopsy material from most patients did not allow competitive PCR for the cytokines. The PCR conditions for IL-4, IL-5, IL-13, and IFN-γ were the same as for competitive PCR for G3PDH except that the annealing temperature was 55°C and the number of cycles was 38–40. The mRNA expression was evaluated on an ordinal scale from 0 to 3 (0: no signal detected; 1: weak signal; 2: moderate signal; 3: strong signal).

Statistics

The Kruskal-Wallis ANOVA followed by intergroup comparisons with the two-sided Mann–Whitney test were used to analyze variable differences between VKC subjects with and without IgE-sensitization and controls. The Kendall rank correlation coefficient was used to explore associations between certain parameters. A probability (P) level of < 0.05 was considered significant.

Results

In the disease group, 10 subjects with and six without IgE-sensitization were identified (Table 1). Of the six control subjects, none displayed serologic evidence of allergen-specific IgE antibodies. The comparison of tissue inflammatory variables between those with and those without IgE-sensitization in the VKC populations, the main investigational objective, yielded no statistically significant differences (Figs 1, 2). Statistically significant increases in the number of CD3⁺, IgE⁺, AA1⁺, and EG2⁺ cells were found for both immunologic subgroups of VKC in comparison with controls (Fig. 1). Whereas CD3⁺, IgE⁺ and AA1⁺ cells were evenly distributed throughout the epithelium and substantia propria, EG2⁺ cells tended to be localized in the epithelium and in the apical parts of the stroma in VKC tissue. The control specimens differed in this respect, since AA1⁺ cells were mainly found in the stroma and EG2⁺ cells were absent.

Table 1.  VKC patient and control characteristics

	n	Male/ female	Age (years)		Total serum-IgE† (kU/L)
			Median	Range	Median	Range
* IgE-sensitization = presence of IgE antibodies to common inhalant allergens and/or to ingested allergens as assessed in serum or by skin prick testing. † Normal reference ranges: 2–263 kU/l for 4–20-year-old individuals.
VKC with IgE-sensitization*	10	10/0	10.5	(7–17)	300	(9–> 1000)
VKC without IgE-sensitization	6	6/0	8.5	(6–11)	105	(13–270)
Controls	6	3/3	8.0	(7–12)	33	(< 2–> 1000)

Figure 1. Enumeration of cells positive for cell markers CD3 (A), IgE (B), AA1 (C), and EG2 (D) in conjunctival tissue. Note the different scales on the y-axis. For one VKC subject with IgE-sensitization no AA1 or EG2 staining was done. Subjects with IgE-sensitization are shown with open symbols, and those without IgE-sensitization are shown with closed symbols. P-levels < 0.05 are marked with *.CD3⁺: VKC with IgE-sensitization vs controls: P = 0.0003; VKC without IgE-sensitization vs controls: P = 0.004. IgE⁺: VKC with IgE-sensitization vs controls: P = 0.0003; VKC without IgE-sensitization vs controls: P = 0.002. AA1⁺: VKC with IgE-sensitization vs controls: P = 0.0004; VKC without IgE-sensitization vs controls: P = 0.002. EG2⁺: VKC with IgE-sensitization vs controls: P = 0.0004; VKC without IgE-sensitization vs controls: P = 0.002.

Figure 2. Cytokine mRNA expression in conjunctival biopsies. The samples were analyzed by semiquantitative RT-PCR with specific primers for IL-5 (A), IL-13 (B), and IFN-γ (C) after normalization for the housekeeping gene G3PDH. The relative signal strength was expressed on an ordinal scale from 0 to 3. Subjects with IgE-sensitization are shown with open symbols, and those without IgE-sensitization with closed symbols. P-levels < 0.05 are shown with *. IL-5: VKC with IgE-sensitization vs controls: P = 0.007; VKC without IgE-sensitization vs controls: P = 0.18. IL-13: VKC with IgE-sensitization vs controls: P = 0.01; VKC without IgE-sensitization vs controls: P = 0.13.

Expression of mRNA for the cytokines IL-5 and IL-13 was significantly more pronounced in the VKC subjects with IgE-sensitization vs controls. Similar trends were found between VKC subjects without IgE-sensitization and controls but they failed to reach significance (Fig. 2). Signals of mRNA encoding IL-4 were only identified in three VKC samples, out of which two were associated with IgE-sensitization. The level of mRNA for IFN-γ was sparsely and similarly distributed in all study groups (Fig. 2).

Certain correlations were sought after separately in the two disease subgroups. There was a significant relationship between IL-5 mRNA expression and EG2⁺ cells in samples from VKC patients with IgE-sensitization (P = 0.02, n = 9). In addition, mRNA for IL-13 and IL-5 were significantly associated in VKC subjects with (P = 0.02, n = 10) and without IgE-sensitization (P = 0.02, n = 6). There was no relationship between total serum-IgE and IgE-bearing cells in the tissue. Nor was there any correlation between IL-13 and IgE-bearing cells or CD3⁺ cells and any cytokine mRNA expression. No reciprocity was found between mRNA for IFN-γ and the other cytokines.

Discussion

The fact that IgE-sensitization is not invariably a disease factor in VKC has long been recognized but has been afforded only limited attention in scientific work. The aim of the present investigation was to study whether there were any divergent patterns in mRNA transcripts for Th1 and Th2 cytokines or in the cellular infiltrate related to IgE-sensitization in VKC patients. The vast body of evidence indicating a reciprocal role for Th1 and Th2 related cytokines in ocular IgE-mediated disease (14–19) justifies this research approach.

In the present investigation, disease was associated with an increase of mRNA encoding Th2-like cytokines with no clear-cut differences between the two immunologic VKC subgroups. Our results are in line with those from asthma studies comparing individuals with and without IgE-sensitization (20, 21). Moreover, in two pathology studies of VKC, mention was made that Th2 reactivity was also a feature in subjects lacking IgE-sensitization, corroborating our observation (14, 18). It is noteworthy that we observed significantly more mRNA signals for IL-13 and IL-5 only in the VKC population exhibiting IgE-sensitization as compared with the controls. Whether this difference reflects a real discrepancy between the two immunologic subgroups of VKC or whether this may be explained by the relatively few samples of VKC subjects without IgE-sensitization included in our investigation is at present unclear. Nevertheless, IL-13 and IL-5 gene expression were significantly associated with one another in both disease groups. Other investigators have demonstrated a closely related secretion of these cytokines after allergen stimulation in T cell lines (22) or in bronchial explants from asthmatics with IgE-sensitization (23). The significant association between mRNA signals for IL-5 and IL-13 established also in the VKC group lacking IgE-sensitization we find remarkable, but the stimulating factors responsible for the enhanced expression of these transcripts in patients with no proven allergen sensitivity remains elusive at present.

We found it interesting that IL-13 mRNA was the most markedly expressed cytokine transcript in biopsies from patients with VKC, similar to what has been found in an ex vivo bronchial challenge study (23). IL-13 is secreted for a longer time than IL-4 after Th2 stimulation and has been theorized to be particularly important in persistent IgE production as well as in eosinophil recruitment (21). However, we found no correlation between IL-13 transcripts and IgE⁺ cells or EG2⁺ cells. In contrast, mRNA expression for IL-5 was significantly associated with EG2⁺ cells in the disease subgroup with IgE-sensitization. This seemingly highlights the function of IL-5 in increasing eosinophil lifespan (9), and confirms findings from a nasal allergen provocation study (24). We found IL-4 mRNA signals in only three out of 16 patient samples. Previous tissue cytokine investigations concerning vernal disease have indeed implicated IL-4 but, contrary to the present study, no semiquantitative comparisons between different cytokine transcript signals were made in these studies (14, 16–18). Finally, no conclusions could be drawn regarding the putative capacity of IFN-γ to counteract Th2 driven inflammation, since IFN-γ mRNA expression was scarce, equally distributed, and not reciprocal relative to other cytokines in any group.

Our failure to establish any differences in the expression of immunocytochemical markers between the two VKC cohorts parallels data from previous studies focusing on similar mechanistic issues in VKC (3, 25) and asthma (26, 27). Of particular interest are cell-bound IgE antibodies because allergen cross-linking of IgE on mast cells is the crucial initiating event in IgE mediated allergic inflammation. Many of the IgE⁺ cells in the VKC samples had a morphology suggestive of mast cells. However the pathogenic role of IgE⁺ mast cells in VKC subjects lacking IgE sensitization remains to be determined.

Acknowledgments

We thank Anne Svensson, Margareta Oskarsson, and Berit Spångberg for their excellent technical assistance. This investigation was supported by grant 16X-7924 from the Swedish Medical Research Council, by grants from the Swedish Council for Work Life Research, the Swedish Foundation for Health Care Sciences and Allergy Research, the Swedish Association against Asthma and Allergy, the Hesselman's Foundation, Consul Th C Bergh's Foundation, Karolinska Institutet, and the Foundation for Eye Research in Stockholm, Sweden.