Investigating innate immune mechanisms in early-life development and outcomes of food allergy

The US Centers for Disease Control and Prevention has reported an approximately 50% increase in the prevalence of reported food allergies in children from 1997 to 2011,¹ with estimates of up to 6 million American children with food allergies at an economic cost of approximately $25 billion per year.^1,2 Notably, a good proportion of these subjects will have immunologic tolerance and resolution of food allergy.³ For example, food allergy resolution can be observed in 43% to 57% of children with milk allergy in early to late childhood (2–10 years)³: 47% to 50% of children with egg allergy in early to late childhood (2–9 years) and 22% of infants with peanut allergy by 4 years of age.^4,5 The underlying immunologic mechanisms leading to initial food sensitization, subsequent development of food allergies, and natural resolution versus persistent food allergy are poorly understood.

Studies of recent prospective birth cohorts suggest a role for altered early innate immunity in the development of CD4⁺ T_H2 responses and predisposition to allergic diseases.⁶ Functional characterization of the ontogeny of microbial pattern recognition responsiveness of PBMCs and CD4⁺ T cells over the first 5 years of life has revealed a link between the development of allergic disease in infants and an innate immune-induced PBMC hyperresponsiveness at birth, followed by an age-related retrograde in innate immune-induced cytokine responses by 5 years of life.⁶ These studies have led to the evolving concept that altered responses of innate inflammatory cytokines (IL-1β, IL-6, TNF-α, and IL-10) at critical times in T-cell development skew T-cell balance from CD4⁺ T_H1 responses to CD4⁺ T_H2 responses and allergy. The source of these cytokines, the relationship between cytokine production and PBMC populations, and what contribution the age-related altered trajectory of innate cytokine responses plays in persistent versus transient allergy outcomes remain unclear.

The study by Neeland et al⁷ published in this issue of the Journal describes an examination of early-life innate immune signatures in patients with and without egg allergy. The authors performed a longitudinal follow-up analysis of a population-based cohort of infants with challenge-confirmed egg allergy to determine the relationship between innate immune responses and (1) clinical phenotypes of egg allergy and (2) persistent or transient egg allergy outcomes. Examining a total of 54 patients from the Australian HealthNuts cohort, the authors assessed the PBMC immune profile, characterizing classical (CD14^high, CD16⁻), intermediate (CD14^high, CD16⁺), and nonclassical (CD14^low, CD16⁺) monocytes and dendritic cells (myeloid and plasmacytoid) and innate immune responses by multiplex cytometric bead array analysis at 1 year with subsequent follow-up and cross-sectional analysis of egg allergy status and the innate immune phenotype of PBMCs at 4 years of age.

The authors showed that the blood serum of 1-year-old infants with egg allergy had greater levels of classical and intermediate monocytes and myeloid dendritic cells at baseline than that of 1-year-old infants without egg allergy. All monocyte populations from infants with egg allergy expressed increased HLA-DR levels, suggesting activation; significantly increased baseline production of IL-8; and significantly reduced IL-12p70 and IL-10 production compared with that in monocyte populations from infants without egg allergy. Consistent with the concept of hyperresponsiveness of PBMCs to innate immune activation, the authors observed increased proinflammatory cytokine (IL-1β, IL-6, IL-8, IL-10, and TNF-α) production in LPS-induced, CD3-depleted PBMCs from infants with egg allergy compared with those without.

Stratifying infants with egg allergy into groups of those with persistent (n = 14) or transient (n = 22) allergy revealed that the PBMC monocyte subpopulation phenotype differed between infants with persistent and those with transient egg allergy at 1 year of life. Although the observed increased classical monocyte phenotype was maintained in both groups, the increased intermediate monocyte and myeloid dendritic cell phenotypes were lost in the transient egg allergy infant group. These findings suggest that at 1 year of age, children with a transient egg allergy phenotype have an intermediary PBMC innate immune signature distinct from that of children with persistent egg allergy.

PBMC immune profiles from 1 year of age were stratified by diagnosis at follow-up (2–4 years) into either persistent (n = 5) or transient (n = 6) egg allergy and compared with those of children without egg allergy (n = 13). These stratified analyses revealed that CD3-depleted PBMC baseline production of IL-1β and IL-8 was increased in patients with persistent egg allergy compared with that in patients with transient egg allergy, with only IL-8 production being increased in patients with persistent egg allergy compared to transient egg allergy. Analyses of innate immune–induced responses of PBMCs from infants with persistent and transient egg allergy revealed increased production of IL-1β and IL-6 compared with those seen in infants without egg allergy, with only IL-8 and TNF-α production being significantly different between infants with persistent and transient egg allergy. Notably, levels of IL-12p70 were greater in the absence of egg allergy than in the presence of persistent or transient egg allergy.

Longitudinal analyses of PBMC innate immune profiles in matched samples collected at 1 year of age and at follow-up in children with egg allergy revealed an age-dependent regression in classical and intermediate monocyte frequency in both patients with transient and those with persistent egg allergy and a significant increase in myeloid dendritic cell counts in infants with transient egg allergy but not those with persistent egg allergy. There were no differences in CD3-depleted PBMC innate immune–induced cytokine production between patients with persistent and those with transient egg allergy; however, at follow-up, baseline levels of CD3-depleted PBMC production of IL-6, IL-1β, and IL-8 were increased in infants with persistent versus those with transient egg allergy.

On the basis of these data sets, the authors concluded that in the first year of life, infants with persistent food allergy demonstrate a unique innate immune signature characterized by heightened circulating monocyte and dendritic cell levels and an exaggerated inflammatory cytokine profile at baseline and following innate immune activation.

The authors also revealed that children with persistent egg allergy demonstrated a higher rate of vitamin D deficiency at follow-up compared with children with transient egg allergy. Although serum vitamin D levels at follow-up negatively correlated with classical monocyte levels and positively correlated with myeloid dendritic cell counts in infants with transient egg allergy, these associations were not observed in patient with persistent egg allergy nor was there any relationship found between vitamin D levels and cytokine production at baseline or after innate immune activation. Given the known effects of vitamin D on monocytes, it would be of interest to examine whether vitamin D can directly modulate monocyte responsiveness to microbial stimulation and inflammatory cytokine production in infants with egg allergy.

The data described in the study by Neeland et al⁷ are consistent with previous reports of heightened PBMC innate immune responsiveness very early in life in patients with allergy relative to those without allergy. Limitations of this study include a lack of inclusion of cord blood analyses and that the PBMC analyses were performed on CD3-depleted PBMCs, which would eliminate any T cell–dependent amplification of monocyte responses. Furthermore, the individual contribution of the monocyte populations to the altered cytokine profile remains unclear. Given that the classical CD14^hiCD16⁻ monocytes consist of approximately 80% of total PBMC monocytes and are a prodigious source of cytokines, one would predict that the cytokines are derived from this population.⁸ However, intermediate CD14^hiCD16⁺ cells are also known to produce high levels of cytokines after Toll-like receptor 4 activation.⁹ Currently, it is unclear whether the differences observed in PBMC-derived innate immune cytokine profiles relate to phenotypic changes in monocyte populations or are simply reflective of altered numbers of monocyte subpopulations.

The strengths of the study include the longitudinal follow-up study design without any loss of subjects, combined assessment of PBMC subpopulations (monocytes and dendritic cells) in combination with baseline and innate immune–induced cytokine analyses, and being one of the few studies to investigate predisposition to food allergy based on innate immune signatures. The population was of a limited demographic, with no details on ethnicity or race as variables; therefore whether these findings can be generalized to other populations and races remains to be determined. Previous studies suggest that race, ethnicity, and sex can influence innate cytokine immune responses to vaccination, and therefore these variables might have an effect on interpretation of innate immune profiles.¹⁰

Another potential confounder is the influence of disease comorbidity. The authors reported that a subgroup analysis of subjects with and without eczema revealed that only those without eczema had differences in the innate monocyte cell populations; this finding raises the concern that innate immune profiles are likely influenced by comorbid disease, which could affect the reliability of analyzing such immune profiles without including comorbid diseases in phenotype analyses, particularly given that many patients with food allergy have or will have concomitant eczema, allergic rhinitis, and asthma. Future studies expanding on these initial observations defining innate immune signatures with subgroup analyses inclusive of different combinations of comorbid allergic diseases would be of interest. Furthermore, given the recent study revealing differences in allergic disease susceptibility in children based on environment (Amish vs Hutterite), inclusion of environmental exposure variables and analyses assessing environmental antigen–specific effects on innate immune signaling will be important.¹¹

Despite these limitations, this study does have important clinical implications suggesting 1) that immune profiling may be a possible first step in identifying infants at greater risk of persistent food allergy and 2) that development of immune profiling assay systems that could predict allergy outcomes at birth or at 1 year of age would improve the management of food allergy.

In summary, the study by Neeland et al⁷ further strengthens the evolving concept that an altered trajectory in myeloid innate inflammatory responses in early life can modify T-cell development, altering the balance of CD4⁺ T_H1 versus T_H2 responses and development of allergic disease (Fig 1); identifies potential immune profiles at birth or at 1 year of age that can predict egg allergy outcomes; and provides impetus for additional studies to assess the value and predictability of altered myeloid innate profiles in predicting natural resolution versus persistence of food allergy.

FIG 1.

Environmental exposures at a crucial time during immune programming (first year of life) directs dendritic cell (DC) and monocyte (Mono) phenotypes and CD4⁺ T-cell responses later in life. Development of maintenance of a DC and monocyte hyporesponsiveness phenotype in the first year of life reinforces T-cell tolerance and preservation of a nonallergic phenotype. Alteration of the trajectory of the myeloid innate immune phenotype during this period leads to exaggerated inflammatory cytokine production (IL-1β, IL-6, and IL-8) and shifts the balance of CD4⁺ T-cell programming toward a pro–type 2 response and development of allergic diseases, such as food allergy.