Filaggrin gene polymorphisms in Indian children with atopic dermatitis: A cross-sectional multicentre study

Plain Language Summary

Atopic dermatitis is a chronic, itchy, relapsing inflammatory skin disorder seen more commonly in children. Filaggrin is a protein which plays an important role in maintaining the normal barrier function of the skin. Defects in the filaggrin gene can cause reduced or absent production of filaggrin, contributing to the development of atopic dermatitis. This study was conducted on 75 children at two different hospitals in India, to identify the filaggrin gene variants and assess their role in atopic dermatitis. There were 20 different filaggrin gene variants identified in 26 children in our study. Among them, 16 filaggrin gene variants were unique to our study population while four had been reported previously in Asian and European populations. These filaggrin gene variants in our study were significantly associated with an early age of onset of the atopic dermatitis and increased serum levels of the IgE class of antibodies.

Introduction

Atopic dermatitis is a chronic, progressive, relapsing inflammatory skin disorder characterised by varying degrees of pruritus and inflammation seen in 20% of children and 1–3% of adults.¹ The prevalence of atopic dermatitis in India is around 0.5–7.2 per cent.² Atopic dermatitis is the result of a complex interplay of several environmental, dietary, behavioural and immunological factors in genetically predisposed individuals.² Filaggrin (filament aggregating protein, FLG) gene is an important component of skin barrier function and plays an important role in epidermal differentiation. Epidermal barrier and FLG null variants are thought to play a key role in the pathogenesis of atopic dermatitis, as well as in allergic rhinitis and asthma.³ Null variation is a change in the gene that results in non-transcription of mRNA or translation of protein or both. Null variants can be due to frameshifts or nonsense substitutions. FLG gene studies have been carried out in many geographical populations, which include European and Asian populations (Japanese, Chinese, Korean, Taiwan and Singaporean). Patients with FLG variants have been reported to have more persistent and severe disease, a higher incidence of herpes virus infections, association with ichthyosis vulgaris, palmar hyperlinearity, keratosis pilaris, asthma and a greater risk of multiple allergies than patients with atopic dermatitis without FLG variants.⁴ FLG gene variants may predict early onset, severe disease, worse prognosis, persistence of the disease in adulthood, allergen sensitisation and development of IgE-mediated food allergies.⁵

Several cohort studies have revealed that only 14–56% of patients harbour FLG gene null variants as a predisposing factor.^{6 , 7} The profilaggrin/FLG gene resides on chromosome 1q21.3 within the epidermal differentiation complex and consists of three exons. Exon 3 is extremely large (>12 kb) and sequencing is challenging as it encodes most of the profilaggrin polypeptide with almost complete homologous 10, 11 and 12 repeats.⁵ Null variants in exon 3 result in complete loss of the expressed protein. Every race is likely to have a unique set of FLG variants which could predict the severity of the disease.^{4 , 8} Most of the studies have analysed the more common loss of function variants, which could have resulted in under-reporting of other significant loss of function variants seen in different ethnic populations. Recent technology like next-generation sequencing can analyse the entire coding region of FLG to identify unreported loss of function variants.³

A recent Indian study has shown a prevalence of variant R501X to be 2.2% in children concomitantly associated with asthma and atopic dermatitis.⁹ Handa et al. have reported a prevalence of 33.7% of FLG variants (S2889X, 2282del4, R501X and Q2417X) associated with hand eczema from the Indian population.¹⁰ FLG variants could predict the outcome of atopic dermatitis which would help clinicians to counsel parents about the likely severity of atopic dermatitis. We aimed to study the FLG gene polymorphisms in Indian children with atopic dermatitis and do a genotype-phenotype correlation.

Materials and methods

Results

Clinical characteristics of study children

We recruited 80 unrelated children diagnosed with atopic dermatitis in the study. Five children failed the DNA quality control (insufficient DNA sample), hence we analysed 75 children [Figure 1]. Among these 75 children, 50 were boys and 25 were girls aged 4 months to 18 years (mean age, 5.2 ± 3.9 years). The clinical characteristics of children with and without FLG variants are presented in Table 1. The mean age of onset of the disease was 2.7 years (10 days to 12 years). Around 93.4% of children presented with atopic dermatitis in the early age group (birth to 8 years). Seventy children (93.3%) had moderate to severe disease in our study. In the present cohort, palmar hyperlinearity was seen in 42.7% and xerosis in 76% of children. Serum IgE levels could be done in only 55 children due to logistical reasons.

Close

Figure 1

Flow chart of the study

Export to PPT

Table 1 Clinical characteristics of Indian children with and without FLG variants

Clinical features	AD with FLG variants (n = 26)	AD without FLG variants (n = 49)	Total (n = 75)	P
Age at enrolment (years)
Mean ± SD	4.57 ± 3.97	5.61 ± 3.88	5.25 ± 3.92	0.277
Range	0.4–13	0.4–18	0.4–18
Median (IQR)	3.2 (1.8, 6)	5 (2.45, 8.13)	4 (2, 8)
Gender
Boys	16 (61.5%)	34 (69.4%)	50 (66.7%)	0.608
Girls	10 (38.4%)	15 (30.6%)	25 (33.3%)
Age of onset
Infantile (birth–<1 year)	17 (65.4%)	18 (36.7%)	35 (46.7%)	0.041
Early childhood (1–8 years)	7 (26.9%)	28 (57.1%)	35 (46.7%)
Late childhood (8–18 years)	2 (7.7%)	3 (6.1%)	5 (6.7%)
Family history	10 (38.5%)	22 (44.9%)	32 (42.7%)	0.632
Allergic disease association (allergic rhinitis and/or asthma)	10 (38.5%)	24 (49.0%)	34 (45.3%)	0.468
SCORAD
Mild (<15)	1 (3.8%)	4 (8.2%)	5 (6.7%)	0.653
Moderate (15–40)	16 (61.5%)	35 (71.4%)	51 (68%)	0.440
Severe (>40)	9 (34.6%)	10 (24.4%)	19 (25.3%)	0.264
Pityriasis alba	6 (23.1%)	14 (28.6%)	20 (26.7%)	0.785
Palmar hyperlinearity	12 (46.2%)	20 (40.8%)	32 (42.7%)	0.807
Xerosis	21 (80.8%)	36 (73.5%)	57 (76%)	0.577
Cheilitis	3 (11.5%)	12 (24.5%)	15 (20%)	0.234

P value <0.05 was considered significant. SD, Standard deviation; IQR, Interquartile range; AD, Atopic dermatitis; FLG, Filaggrin gene; SCORAD: SCORing Atopic Dermatitis

Discussion

There are more than 500 stop-gain variants in FLG described in the genome aggregation database.¹⁶ The most common FLG variants shared among Japanese, Koreans, Taiwanese, Chinese, Singaporeans and Asians are R501X and c.3321delA. In the European population, the most prevalent mutations are R501X and c.2282del4. The three most common FLG variants R501X, c.2282del4 and E2422X are shared in both European and Asian populations.⁸ Meng et al. found c.3321delA to be associated with xerosis, ichthyosis vulgaris, palmar hyperlinearity, keratosis pilaris, white dermographism and disease severity but not associated with early age of onset.¹⁷ Previous studies found that FLG loss of function variants were not commonly seen in persons of African ancestry, which could be due to inappropriate genotyping techniques. Recent studies with massively parallel sequencing have shown uncommon FLG variants to be associated with persistent atopic dermatitis in African ancestry as compared to European ancestry. However, the frequency of FLG loss of function mutations in African ancestry is lower. Copy number variation has been associated as a risk factor for atopic dermatitis and this has been documented in African-American pediatric patients with moderate to severe atopic dermatitis.¹⁸

About 20 FLG loss of function variants in 26 children (6 frameshift, 13 nonsense and 1 splice site variant) were seen in our study.

Among the documented FLG loss of function variants in our study, 20% of the variants were previously reported in European and Asian populations and 80% were unique variants.^{16 , 17 , 19 , 20} India is home to more than 2000 ethnic populations, whose genetic origins are complex. The present Indian population is genetically heterogeneous with an admixture of ancestral north Indians comprising the majority and having roots in the Middle East, Central Asia and Europe and ancestral south Indians that have no relation to any population outside India.²¹ Further complexity is added by the relatively recent immigration of populations such as Siddis, Muslims and Jews.²² Ours being a two-centre study, had both these sets of populations (East Indians and South Indians) and thus reported variants of Asia and European as well as a high number of unique variants. Chauhan et al. studied the prevalence of R501X mutation in Indian children with allergic diseases. Their study was a case-control study study enrolling 90 children with allergic diseases (asthma, rhinitis and atopic dermatitis). The mutant R501X was seen in 3.3% of children with asthma and 2.2% in children with asthma concomitantly with eczema. The limitation of their study was that screening of only one FLG null variant was done.⁹

Several studies have shown the strong association of FLG null mutation with different atopic dermatitis phenotypes.¹⁶ FLG null variants in our study were significantly associated with early age of onset (infantile) and elevated serum IgE levels, which was concordant with several earlier studies.^{23 , 24 , 25} FLG variants have a significant effect on the age of onset of atopic dermatitis and can lead to early onset (less than 1 year) and persistence of the disease into adulthood.^{23 , 24} Although previous studies have shown a significant association between the FLG null variants with moderate to severe atopic dermatitis (odds ratio 2.03–13.4), our study did not show a statistically significant correlation.^{4 , 11 , 26 , 27} FLG mutation is a strong genetic predisposing factor for atopic dermatitis, but other factors like unidentified candidate genes, copy number variations and epigenetic factors may also play an important role in the pathogenesis of atopic dermatitis, which could explain the non-correlation of the severity of atopic dermatitis with FLG variants in our study, as well as the fact that SCORAD is variable at different points of time.¹⁶ Interestingly Park et al. in a study on Korean patients with atopic dermatitis did not observe any significant association between FLG null variants and clinical features of atopic dermatitis.¹⁸

Ota et al. studied the effect of FLG loss of function mutation on the severity of skin lesions and skin barrier function in the Japanese population. Eight (14.5%) patients had a loss of function mutations in the FLG gene in their study. Stratum corneum was collected from three different sites (extremities, neck and trunk) using the tape-stripping method. The amount of FLG protein and total amino acid in the stratum corneum was measured in both mutation carriers and non-carriers. FLG abnormalities had little effect on the severity of dermatitis, FLG protein and total amino acid content in the stratum corneum in the lesional skin. Their study suggested that the activation of Th2 dominant inflammatory cells along with FLG abnormalities plays a role in suppressing the production of FLG in lesional skin.²⁸

Previous studies have shown FLG mutation to be associated with elevated levels of IgE in Japanese, Koreans, Singaporeans, Chinese and European populations.^{4 , 29} In our study too, serum IgE levels were significantly associated with FLG null variants. We were unable to find any association between FLG variants with allergic disease association, family history, palmar hyperlinearity, xerosis, cheilitis and keratosis pilaris and this was similar to other studies.^{4 , 7 , 26}

We found no significant associations in severity or clinical manifestations of atopic dermatitis between children carrying one single variant and children with two or more variants. This could again be explained by the multifactorial nature of the disease. Though FLG mutation is not an independent risk factor for the development of different atopic dermatitis phenotypes, studies on FLG variants may help identify some infants at risk of developing atopic dermatitis, guiding early interventions to delay the onset and decrease the severity of atopic dermatitis.

Limitations

The main limitation of our study is its relatively small sample size. As a result, there could have been false negative or positive results due to low statistical power. The small sample size in our study also did not permit a statistical correlation between specific null variants and atopic dermatitis phenotypes. Larger population-based studies are required to substantiate the associations between FLG mutations and atopic dermatitis phenotypes in the Indian population. We could not look for copy number variation in the present study due to logistic reasons.

Conclusion

Our study documented a high prevalence of FLG null variants in Indian children. The FLG variants identified in our study included ones unique to this population and also those that overlapped with Asian and European populations. The presence of more than one FLG mutations in an individual did not correlate significantly with the severity of AD. We found a significant association of FLG variants with an early age of onset and elevated serum IgE levels.