Molecular analysis of infant fecal microbiota in an Asian at-risk cohort-correlates with infant and childhood eczema.

BACKGROUND: Studies have suggested that selective microbial targets prevail in the fecal microbiota of infants with eczema. This study evaluated the composition of fecal microbiota of infants who developed eczema in the first 5 years of life and compared these with those of healthy controls.
FINDINGS: Children who developed eczema in the first 2 years, those with eczema at 5 years of age and healthy controls were selected from the placebo arm of a birth cohort of at-risk infants participating in a randomized double-blind trial on the protective effects of supplemental probiotics in early life on allergic outcomes. Molecular evaluation of fecal microbiota were conducted using Fluorescence In Situ Hybridization-Flow Cytometry (FISH-FC) for fecal samples collected. Longitudinal analysis of fecal microbiota composition at three days, one and three months and one year of life revealed higher abundance of Enterobacteriaceae [coefficient (B): 1.081, 95% CI: 0.229-1.933, adj p = 0.014] and Clostridium perfringens [coefficient (B): 0.521, 95% CI: 0.556-0.988, adj p = 0.03] in those who developed eczema in the first 2 years life. In those with eczema at 5 years of age, a lower abundance of Bifidobacterium was observed [coefficient (B): -27.635, 95% CI: -50.040 - -5.231, adj p = 0.018].
CONCLUSIONS: The differences in infant fecal microbiota observed in eczema subjects in this study support the notion that relative abundance of selective microbial targets may contribute to the subsequent development of eczema in childhood.

Findings

Background

Allergic diseases, such as eczema, are chronic inflammatory disorders with increasing global trends [1]. Intestinal microbiota play a role in the regulation of innate and adaptive immunity [2] and has been implicated in the development of allergy diseases [3]. The composition of fecal microbiota of infants and young children with eczema differ from healthy children [4]. This study aims to evaluate and monitor the composition, maturation, development of fecal microbiota at 4 times points of an at risk birth cohort in the first year of life, and compare these findings in relation to the development of eczema at 2 and 5 years of age with those of healthy controls.

Methods

Subjects with eczema and healthy controls were selected from the placebo arm of a birth cohort of at-risk infants participating in a randomized double-blind trial on the protective effects of supplemental probiotics in early life on allergic outcomes [5]. Subjects who developed eczema in the first 2 years (eczema, n = 28; healthy, n = 32), and those with eczema at 5 years of age (eczema, n = 15 [persistent from 2 years: n = 11; new cases: n = 4]; healthy, n = 19) were studied. Thirteen controls were excluded in the 5 year analysis because they developed allergen sensitization, rhinitis and wheeze at 5 years of age (Figure  1). Informed consent was obtained from all families. The study was approved by the hospital’s institutional ethical review board (Ref Code: 2006/00008). Stool samples were collected at 4 time points (3 days, 1, 3 and 12 months of age) as previously described [6] There were missing stool samples in 2% to 25% of subjects at different time points, but these were not different between cases and controls. Subjects with eczema in the first 2 years and at 5 years of age were subclassified into atopic (positive skin prick test to common allergens) eczema (2 yrs: n = 13; 5 yrs: n = 12) and non-atopic eczema (2 yrs: n = 15; 5 yrs: n = 3).

Figure 1

Flowchart of study subjects showing the progress of eczema development in the first 2 years and at 5 years of age. *Thirteen controls were excluded in the 5 year analysis because they developed allergen sensitization, rhinitis and wheeze at 5 years of age.

Molecular evaluation of fecal microbiota was conducted using/harnessing whole-cell-based detection approach based on Fluorescence In Situ Hybridization combined with Flow Cytometry (FISH-FC) for fecal samples collected at 3 days, 1, 3 and 12 months. 16S rRNA probes were selected to target and quantify the Eubacterium rectale-Clostridium coccoides group (Erec 482), Clostridium leptum subgroup (Clep 866 and the corresponding competitor probes), Bacteroides-Prevotella group (Bac 303), Bifidobacterium genus (Bif 164), Atopobium group (Ato 291), Lactobacilli-Enterococci group (Lab 158), Enterobacteriaceae family (Enter 1432), Clostridium perfringens (Cperf191), Clostridium difficile (Cdif198) and Escherichia coli (Eco 1531) as previously described [7].

Linear mixed model was used to evaluate the longitudinal differences (i.e. 4 time points) of bacterial targets with adjustments for gender, mode of delivery, breastfeeding up to 6 months and comorbidities (rhinitis and wheeze) [8].

Results

The demographic characteristics and data showing the relative abundance of fecal bacterial groups for children who developed eczema by 2 and at 5 years of age and their healthy controls are summarized in Tables  1, 2 and 3 respectively. Longitudinal monitoring of the dynamics of intestinal bacterial colonization over 4 time points (3 days, one month, three months and one year of life) in the first year of life showed higher relative abundance of Enterobacteriaceae [coefficient (B): 1.081, 95% CI: 0.229-1.933, adj p = 0.014] in children with eczema by 2 years of age, and Clostridium perfringens [coefficient (B): 0.521, 95% CI: 0.556-0.988, adj p = 0.03] in eczema n = 28) by 2 years of age compared to healthy controls (n = 32) (Table  2 and Figure  2a).

Table 1

Demographic characteristics of children with eczema by 2 years and at 5 years of age and their healthy controls

	By 2 years of age			At 5 years of age
	Eczema (n = 28)	Healthy control (n = 32)	p value	Eczema (n = 15)	Healthy control (n = 19)	p value
Gender (%)
Male	14 (50)	15 (46.9)	0.809	6 (40)	9 (47.4)	0.667
Female	14 (50)	17 (53.1)		9 (60)	10 (52.6)
Mode of delivery (%)
Lower segment caesarean section	7 (25)	9 (28.1)	0.785	4 (26.7)	4 (21.1)	1
Vaginal delivery	21 (75)	23 (71.9)		11 (73.3)	15 (78.9)
Feeding history from birth to month 6 (%)
Breastfeeding and formula feeding	21 (75)	26 (81.3)	0.558	11 (73.3)	16 (84.2)	0.672
Total formula feeding	7 (25)	6 (18.7)		4 (26.7)	3 (15.8)

Table 2

Relative abundance of fecal bacterial groups for children with eczema by 2 years of age and their healthy controls

		Healthy		Eczema by 2 years of age
Time points	Bacterial group	n	Mean (SD)	n	Mean (SD)
3 day	E.rectale - C. coccoides	21	0.27 (1.139)	19	ND
	Clostridium leptum	21	3.08 (9.155)	19	ND
	Bacteriodes-Prevotella	21	5.56 (11.654)	19	6.88 (19.386)
	Bifidobacterium	21	29.35 (36.872)	19	15.88 (30.175)
	Atopobium	21	0.44 (0.801)	19	0.86 (3.555)
	Lactobacilli–Enterococci	21	3.22 (8.631)	19	0.49 (1.370)
	Enterobacteriaceae	21	26.22 (35.55)	19	41.64 (37.819)
	Clostridium perfringens	12	2.05 (7.093)	16	2.99 (8.137)
	Clostridium difficile	12	ND	16	0.66 (2.365)
	E.coli	11	14.17 (29.862)	16	34.98 (40.363)
1 month	E.rectale–C. coccoides	23	0.29 (0.589)	19	6.65 (18.510)
	Clostridium leptum	23	ND	19	2.47 (9.339)
	Bacteriodes-Prevotella	23	1.65 (3.309)	19	6.01 (16.408)
	Bifidobacterium	23	41.27 (38.514)	19	28.38 (33.945)
	Atopobium	23	6.52 (12.635)	19	2.32 (6.41)
	Lactobacilli–Enterococci	23	2.74 (7.265)	19	7.98 (15.762)
	Enterobacteriaceae	23	15.63 (23.726)	19	20.1 (23.837)
	Clostridium perfringens	15	0.19 (0.751)	18	0.55 (1.482)
	Clostridium difficile	15	ND	18	ND
	E.coli	15	9.01 (18.952)	18	3.62 (4.631)
3 month	E.rectale–C. coccoides	28	5.45 (10.551)	24	7.43 (13.389)
	Clostridium leptum	28	0.46 (1.590)	24	0.01 (0.059)
	Bacteriodes-Prevotella	28	3.53 (7.845)	24	3.32 (5.126)
	Bifidobacterium	28	55.86 (32.344)	24	54.21 (33.827)
	Atopobium	28	10.19 (16.06)	24	3.72 (7.037)
	Lactobacilli–Enterococci	28	2.77 (3.121)	24	2.62 (3.404)
	Enterobacteriaceae	28	3.8 (7.222)	24	6.51 (10.024)
	Clostridium perfringens	17	0.72 (1.547)	20	2.26 (6.246)
	Clostridium difficile	17	ND	20	ND
	E.coli	17	1.16 (2.725)	20	4.26 (10.987)
1 year	E.rectale–C. coccoides	27	30.40 (18.445)	24	31.25 (16.0.42)
	Clostridium leptum	27	4.22 (7.022)	24	1.33 (3.047)
	Bacteriodes–Prevotella	27	10.05 (10.082)	24	4.175 (6.049)
	Bifidobacterium	27	28.88 (23.227)	24	27.92 (25.196)
	Atopobium	27	5.91 (8.639)	24	5.05 (9.337)
	Lactobacilli–Enterococci	27	1.57 (3.437)	24	1.45 (4.99)
	Enterobacteriaceae	27	0.10 (0.3)	24	1.65 (3.162)
	Clostridium perfringens	18	ND	24	0.30 (1.457)
	Clostridium difficile	18	ND	22	0.07 (0.328)
	E.coli	18	0.05 (0.129)	23	0.83 (2.22)

ND, not detected.

Table 3

Relative abundance of fecal bacterial groups for children with eczema at 5 years of age and their healthy controls

		Healthy		Eczema at 5 years of age
Time points	Bacterial group	n	Mean (SD)	n	Mean (SD)
3 day	E.rectale–C. coccoides	10	ND	9	ND
	Clostridium leptum	10	2.7 (8.526)	9	3.74 (11.217)
	Bacteriodes-Prevotella	10	10.32 (15.729)	9	7.49 (19.590)
	Bifidobacterium	10	21.65 (33.2)	9	24.45 (34.194)
	Atopobium	10	0.21 (0.439)	9	0.12 (0.229)
	Lactobacilli–Enterococci	10	5.48 (12.272)	9	0.64 (1.307)
	Enterobacteriaceae	10	47.73 (41.03)	9	30.89 (39.958)
	Clostridium perfringens	7	ND	5	ND
	Clostridium difficile	7	ND	5	ND
	E.coli	7	21.87 (35.99)	5	20.36 (39.873)
1 month	E.rectale–C. coccoides	13	0.32 (0.670)	10	0.74 (1.687)
	Clostridium leptum	13	ND	10	4.06 (12.829)
	Bacteriodes-Prevotella	13	2.92 (4.008)	10	6.56 (20.824)
	Bifidobacterium	13	48.29 (36.270)	10	21.28 (28.332)
	Atopobium	13	5.15 (9.403)	10	6.01 (12.463)
	Lactobacilli–Enterococci	13	1.91 (3.987)	10	5.77 (18.246)
	Enterobacteriaceae	13	19.54 (28.826)	10	18.19 (20.289)
	Clostridium perfringens	10	0.29 (0.92)	6	ND
	Clostridium difficile	10	ND	6	ND
	E.coli	10	11.62 (22.528)	6	5.27 (7.354)
3 month	E.rectale–C. coccoides	16	4.83 (10.187)	12	7.67 (14.804)
	Clostridium leptum	16	0.17 (0.688)	12	0.2 (0.678)
	Bacteriodes-Prevotella	16	2.83 (5.904)	12	6.42 (9.913)
	Bifidobacterium	16	55.21 (33.0)	12	42.557 (35.960)
	Atopobium	16	12.73 (19.887)	12	5.96 (9.282)
	Lactobacilli–Enterococci	16	2.79 (3.085)	12	3.07 (3.726)
	Enterobacteriaceae	16	4.28 (6.634)	12	10.94 (14.169)
	Clostridium perfringens	12	0.98 (1.791)	7	6.50 (9.55)
	Clostridium difficile	12	ND	7	ND
	E.coli	12	1.64 (3.152)	7	6.83 (13.998)
1 year	E.rectale–C. coccoides	15	28.20 (19.361)	12	29.78 (15.534)
	Clostridium leptum	15	3.54 (7.102)	12	1.84 (2.359)
	Bacteriodes-Prevotella	15	9.22 (10.450)	12	4.86 (8.232)
	Bifidobacterium	15	37.97 (23.939)	12	18.88 (24.0)
	Atopobium	15	6.45 (9.575)	12	4.40 (5.199)
	Lactobacilli–Enterococci	15	1.59 (3.724)	12	3.29 (6.987)
	Enterobacteriaceae	15	0.04 (0.152)	12	1.59 (3.133)
	Clostridium perfringens	12	ND	9	ND
	Clostridium difficile	12	ND	7	0.22 (0.582)
	E.coli	12	0.07 (0.155)	8	1.78 (3.521)

ND, not detected.

Figure 2

Longitudinal comparison of the fecal microbiota of children with and without eczema. Linear mixed model analysis of fecal microbiota of children with eczema in (a) the first 2 years of life and (b) at 5 years of age, compared to healthy controls. Only microbial targets with statistically significant difference between eczema and controls are shown.

Sub-group analysis of atopic eczema (n = 13) and non-atopic eczema (n = 15) cases with onset in first 2 years of life yielded similar results, with higher abundance of Enterobacteriaceae compared to controls [atopic eczema: coefficient (B): 0.949, 95% CI: 0.214-1.683, adj p = 0.013; non atopic eczema: coefficient (B): 1.119, 95% CI: 0.136-2.102, adj p = 0.027], and Clostridium perfringens for non-atopic eczema only [coefficient (B): 0.713, 95% CI: 0.124-1.303, adj p = 0.022]. This analysis was also carried out with controls (n = 19) who were healthy all the way till 5 years of age (that is did not develop eczema after the age of 2 years), and similar findings were observed (data not shown).

For atopic eczema at 5 years old, a lower relative abundance of Bifidobacterium was found in cases (n = 15) compared to healthy controls (n = 19) [coefficient (B): -27.635, 95% CI: -50.040 - -5.231, adj p = 0.018] (Table  3 and Figure  2b).

Discussion

In this study significant differences in fecal microbiota signatures analyzed at 4 times points in the first year of life were observed in infants who developed eczema in the first 5 years of life. These data suggests that specific microbial signatures appearing in early life might be predictive of eczema. Previous studies have also compared infant microbiota at more than one time point in infancy [3,9]. Sjögren et al. reported lower prevalence of lactobacilli group I and Bifidobacterium adolescentis at 1 week after birth for atopic dermatitis compared to non-allergic controls, but no significant differences were found for the other time points i.e. 1 month and 2 months [3]. In another study, reduced microbial diversity at 1 month and 12 months were associated with the presence of serum specific IgE, while allergic rhinitis group at school age showed less microbial diversity at 1 month [9]. Instead of making comparisons between single time points, our study utilized linear mixed model analysis to enabled comparisons over 4 time points in the first year of life. Our findings indicate that a consistent pattern in profile of fecal microbiota over the first year was associated with development of childhood eczema.

Our findings support the single time point cross sectional studies, where increased numbers of bacterium from Clostridium cluster IV and XIVa have been reported. These clusters of Clostridium are normally found in abundance in the adult intestine [10,11]. Although we did not find any difference for Clostridium cluster IV and XIVa in our study but we did observe an increase in Clostridium perfringens in children with eczema. This findings was supported by a recent study where Clostridium cluster I was associated with higher risk of developing atopic dermatitis [12].

An increased abundance of Enterobacteriaceae has also been observed in atopic subjects [13]. We have also previously shown by 16S rRNA pyrosequencing in this same cohort that fecal Enterobacteriaceae was more abundant at the age of 1 month in eczema compared to match controls [14]. Higher abundance of Enterobacteriaceae has been reported to be associated with intestinal inflammation (dextran sodium sulfate induced colitis), suggesting that Enterobacteriaceae may be responsible for promoting immune dysregulation [15].

The reduced abundance of Bifidobacterium in eczema compared to healthy controls has been observed in several other studies [3,7], and is associated with an increased risk of atopic dermatitis [16]. We did not observe reduced abundance of Bifidobacterium for eczema in the first 2 years of life compared to healthy controls. However, in subjects whose onset of eczema occurred by the age of 2 years (n = 28), we found that in those with eczema persistent till 5 years (n = 11) had lower abundances of Bifidobacterium compared with those with transient eczema (resolved before 5 years) (n = 17) as well as healthy controls, but these differences did reach statistical significance (p = 0.099) (data not shown). These observed differences, occurring in early life, in the maturation and development of the infant gut microbiota between subjects with atopic dermatitis and healthy controls suggests that the gut microbiota may influence immune system development [17] and contribute to disease severity [2].

In conclusion, this prospective study which profiled the dynamics of intestinal bacterial colonization over infancy supports the notion that relative abundance of selective microbial targets may contribute to the subsequent development of eczema in childhood.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

GCY performed the experiments, data analysis and statistical analysis. EXLL drafted the manuscript. GCY and BWL helped to revise the manuscript. QSL participated in collation of clinical data. MA, LPCS and BWL participated in the study design and helped in coordination of sample and clinical data collection. All authors read and approved the final manuscript.