The neonatal susceptibility window for inhalant allergen sensitization in the atopically predisposed canine asthma model


Correspondence: Dr E. G. Barrett, Lovelace Respiratory Research Institute, 2425 Ridgecrest Dr SE, Albuquerque, NM 87108, USA. Email:

Senior author: Edward G. Barrett


Allergic asthma often begins in early life and, although many risk factors have been enumerated, the specific factors that initiate disease progression in an individual remain unclear. Using our dog model of early life allergen inhalation, we tested the hypothesis that the atopically biased neonatal immune system would exhibit tolerance to ragweed if allowed to mature normally before exposure or artificially through innate immune stimulation with early life exposure. Dogs were subjected to a series of inhalational ragweed exposures from 1 to 20 weeks old, with or without inhalation of a Toll-like receptor 4 (TLR4) agonist (CRX-527), or from 13 to 31 weeks old. Serum allergen-specific antibody response was assessed at 4, 8 and 20 weeks after the last sensitizing exposure. At 24 or 35 weeks old, airway hyper-responsiveness to methacholine and ragweed challenges and pulmonary inflammation by bronchoalveolar lavage were tested 1 and 4 days after ragweed challenge at 28 or 39 weeks old. Allergen-free immune maturation resulted in no airway hyper-responsiveness and very little ragweed-specific IgE relative to the control group, but eosinophilia developed upon ragweed challenge. TLR4 agonism yielded no airway hyper-responsiveness, but a strong airway neutrophilia developed upon ragweed challenge. Our data indicate that an atopic predisposition creates a critical window in which allergen exposure can lead to an asthmatic phenotype. Allergen-free immune maturation may lead to allergen tolerance. TLR4 agonism before early life allergen exposure may abrogate the development of allergen-specific bronchonconstriction, but allergen-specific pulmonary inflammation remains a strong concern.


In individuals at risk for allergic disease, such as allergic asthma, the immune system may be primed to develop an allergic phenotype very early in life; so preventive measures may need to occur in this early life period. An increased or prolonged perinatal bias toward a T helper type 2 (Th2) cytokine phenotype [i.e. increased interleukins (IL) 4, 5 and 13] may lead to abnormal reactions to otherwise innocuous antigens. The normal early immune response to environmental allergens is probably Th2-biased, as evidenced by allergen-specific immunoglobulin E (IgE, dependent on Th2 cytokines) in both children that later become atopic, as well as those who do not.[1] Atopic children maintain specific IgE responses to inhalant allergens beyond that of non-atopic children[1] and increased allergen-specific IgE in early life is strongly associated with allergic asthma development.[2] The specific genetic, epigenetic, and environmental factors that prolong the IgE response and initiate allergic disease progression remain unclear.

Exposure to allergen is a necessity for allergic sensitization, although there is some debate as to when sensitization begins,[3] and sensitization is a major risk factor in the development of chronic asthma.[2, 4-6] Hence, one mechanism of preventing sensitization may be avoidance, at least until the neonatal immune system has matured sufficiently to respond to allergens appropriately (i.e. develop tolerance to the allergens). A rodent model of maternal transmission of asthma risk has shown that allergic susceptibility (to ovalbumin) gradually declines into young adulthood in allergically predisposed offspring,[7] which suggests that the skewed response is reversible with immune maturation. Measures to reduce oral (food) and household allergens have been shown to decrease risk, but not prevent allergic asthma development in children.[8, 9] High-risk children raised with specific allergen control measures starting prenatally develop a specific IgE response with similar symptoms, but had better lung function at 3 years old compared with children raised without such control measures.[10] It is unknown if maternal (allergic mothers during gestation) and neonatal avoidance of seasonal aeroallergens such as ragweed pollen can directly result in allergic asthma prevention. Further, it is unknown if specific avoidance is even advisable. If the normal response is initially Th2 skewed early in life, but subsequently turns to tolerance, would allergen avoidance lead to a missed opportunity for development of tolerance and yield asthma upon allergen exposure later in life?[10]

Aeroallergen avoidance is not always feasible and an alternative approach to allergic asthma prevention in high-risk individuals may be to artificially induce maturity in the neonatal immune system. A delayed onset of Th1 responsiveness appears to be normal in early childhood.[11-14] However, decreased Th1 responsiveness is not an intrinsic property of the immature immune system and can be overcome with appropriate maturational stimulation of the innate immune system,[13] as illustrated by a robust Th1 (interferon-γ) response in newborns vaccinated with mycobacterial antigens.[15] Microbial components and their analogues,[16] such as toll-like receptor 4 (TLR4) agonists, are well-known inducers of the innate immune system and in high-risk individuals may skew a pro-allergic, immature immune system toward a Th1 profile, providing protection from allergic sensitization. Some, but not all, epidemiological studies have indicated that endotoxin in the home may protect against atopic sensitization[17-19] and atopic asthma.[20] Further, lipopolysaccharide at relatively high doses before allergen sensitization (intranasal[21]) or after (intravenous[22]) has been shown to decrease non-specific airways hyper-responsiveness in rodent models of allergic asthma. However, it is unknown if a synthetic TLR4 agonist inhaled before allergen inhalation during sensitization would prevent allergen-specific bronchoconstriction, leading to tolerance and abrogating the asthmatic phenotype.

We hypothesized that an atopically biased, neonatal immune system would become tolerant to an aeroallergen if allowed to mature either normally, in the absence of allergen, or artificially through innate immune (TLR4) stimulation before allergen exposure. In rodent models, early life inhalation exposure to allergen appears to confer tolerance[23, 24] such that cofactors to the allergen are needed to develop allergic disease in early life.[25] Hence, we chose to test this hypothesis in our dog model of atopically predisposed asthma development[26] in which high IgE-producing beagle dogs are bred and their offspring are sensitized by early life inhalation of ragweed allergen, which coupled with an atopic predisposition is sufficient to develop an allergic asthma phenotype.[27]

Materials and methods


Beagles were born and maintained in the Lovelace Respiratory Research Institute (LRRI) colony in Albuquerque, NM. Dogs were housed in kennel buildings with access to indoor and outdoor runs. Dogs were allowed to nurse for up to 6 weeks with introduction of solid food at about 4 weeks of age. All animal procedures were carried out in accordance with protocols approved by the LRRI Institutional Animal Care and Use Committee.

Nine female beagle dogs used as breeders were previously sensitized to ragweed[26] and were known to have increased ragweed-specific IgE and IgG with increased airway hyper-responsiveness to inhaled ragweed and histamine and lung eosinophilia. Females were bred to one of two naive male beagle dogs. Offspring (a total of 53 puppies in 11 litters) were exposed by inhalation to either ragweed aerosol or filtered air based on the experimental timeline in Figs 1 and 2. Inhalation exposures were carried out as previously described[26] using a dynamic six-port canine exposure system that allows continuous, head-only (< 4 weeks of age) or muzzle only exposure. Offspring were exposed repeatedly from 5 to 20 min each time to aerosolized ragweed extract (short, Ambrosia artemisifolia; Greer Laboratories, Lenoir, NC). Aerosolization of the ragweed extract resulted in a mass median diameter particle of 1·18 ± 0·29 μm (week 1–4) and 1·22 ± 0·17 μm (week 8–20). The average total lung deposition of ragweed protein for the exposures was 157·92 ± 3·51 μg (week 1–4) and 436·93 ± 20·89 μg (week 8–20). Lung deposition was calculated as follows: [concentration (mg/m3) × minute ventilation (l per min) × exposure time (min) × deposition fraction], where minute ventilation was calculated based on body weight[28] and deposition fraction was estimated to be 10%.

Figure 1.

Exposure schedule and experimental timeline for sensitization starting early in life with (CRX/Air and CRX/RW) and without (Air and RW) CRX-527 exposures. RW, ragweed.

Figure 2.

Exposure schedule and experimental timeline for delayed sensitization (Del/Air and Del/RW). RW, ragweed.

Experimental timeline

Dogs received a series of four weekly then four monthly air (Air) or ragweed (RW) exposures by inhalation starting at 6–8 days of age through to 20 weeks of age (Fig. 1, Air = 5, RW = 16) or from 13 through to 31 weeks of age (Fig. 2, Del/Air = 4, Del/RW = 10). Using this sensitization protocol, not all positive control (RW) dogs presented as ‘responders’ to the allergic sensitization protocol. To fit a clinically relevant model of allergic asthmatics, we separated dogs in the RW group who did not present an increase in pulmonary resistance of ≥ 150% for both the ragweed dose response and 5-min ragweed exposure [used to assess pulmonary inflammation by bronchoalveolar lavage (BAL)] into a ragweed non-responder group (RWNR, see Results; Airways response). A separate group of puppies was exposed by inhalation to a TLR4 agonist (CRX-527) three times before ragweed exposure (Fig. 1, CRX/Air = 6, CRX/RW = 12). Aerosolization of CRX-527 for an average exposure of 12 min resulted in a mass median diameter particle of 1·49 ± 0·06 μm and an average total lung deposition of 229·87 ± 8·62 μg. CRX-527 is a synthetic lipid, a mimetic that provides innate immune stimulation through TLR4 activation [provided as a gift from Corixa Corporation, Seattle, WA (aquired by GlaxoSmithKline)].

Serum samples were collected 4, 8 and 20 weeks after the last ragweed or mock sensitization and BAL fluid before and after exposure to ragweed by inhalation (~200–250 μg deposition) at week 28 or 39. In addition, airway hyper-responsiveness to specific and non-specific stimuli was measured at 24 or 35 weeks of age.

Pulmonary resistance

Airway reactivity was evaluated at 24 or 35 weeks of age by a routinely used method in our laboratory.[27] Dogs were anaesthetized and maintained on isoflurane during the procedure. Briefly, an oesophageal balloon catheter in the caudal thoracic oesophagus was used to estimate transpulmonary pressure and respiratory flow was measured using a pneumotachograph (Fleisch no. 1) connected to the endotracheal tube. Custom-designed software (LabView 5·1; National Instruments, Austin, TX) was used to facilitate integration of the flow signal to yield volume and to derive the total pulmonary resistance.

Once the dogs were in position, with pulmonary mechanical indices stabilized, airway responsiveness was assessed by obtaining dose–response curves for methacholine and ragweed aerosols. Methacholine was prepared at concentrations of 0·1, 0·3, 1, 3, 10 and 30 mg/ml in sterile water. Similarly, ragweed was prepared at concentrations of 0·1, 0·3, 1·0 and 3·0 mg/ml in sterile water. Aerosols were generated using a jet nebulizer (No. 950; Hospitak, Lindenhurst, NJ) that fed into the inspiratory leg of the ventilator circuit in close proximity to the end of the endotracheal tube. Five breaths of nebulized vehicle (water), methacholine or ragweed solution (standardized to an end-inspiratory pressure of 15 cmH2O) were allowed, and the response was assessed until peak resistance was reached. Pulmonary resistance values were allowed to recover within 10% of baseline values before subsequent challenge or until a plateau was reached. The response was evaluated by calculating the percentage change in peak resistance following methacholine or ragweed inhalation compared with the baseline.

Bronchoalveolar lavage

Bronchoalveolar lavage of left and right caudal lung lobes was performed 8 weeks after completion of the sensitization protocol corresponding to 28 (Fig. 1) or 39 (Fig. 2) weeks of age. BAL was performed by anaesthetizing dogs with isoflurane and a flexible fiberoptic bronchoscope (Pentax FB-15X, 4·8 mm outer diameter) was advanced into the lung and wedged in a segmental bronchus. Sample collection, processing and differential cell counts were performed as previously described.[27] Counts for right and left caudal lung lobes were averaged for the final count.

Immunoglobulin measurement

Whole blood was collected by jugular venepuncture at 24, 28 and 40 weeks of age or 35, 39 and 51 weeks of age, corresponding in both cases to 4, 8 and 20 weeks after the final exposure in the sensitization series. Serum was collected from clotted, centrifuged blood and frozen at −80° for subsequent evaluation of ragweed-specific IgE and IgG1 and IgG4. To measure ragweed-specific IgE and IgG1, 96-well plates were coated with 100 μl of 5 μg/ml ragweed in 0·05 m sodium bicarbonate buffer and incubated overnight at 4°. After washing and blocking with 1% bovine serum albumin (BSA) in 0·05 m Tris-buffered saline (TBS) for at least 1 hr at room temperature, 100 μl of serum samples at appropriate dilutions was added and plates were incubated at room temperature for 90 min. After washing, 100 μl of the appropriate biotin-conjugated goat anti-dog antibody at the manufacturer-recommended concentration (IgE at 1 : 5000, IgG1 at 1 : 15 000) in assay buffer (1% BSA in TBS + 0·05% Tween-20) was added to each well, plates were incubated for 75 min at room temperature and washed before adding 100 μl streptavidin–alkaline phosphatase conjugate at a dilution of 1 : 8000 in assay buffer. After 60 min of incubation with the conjugate at room temperature and washing, 100 μl para-nitrophenyl phosphate substrate (EMD Millipore Corporation, ES009, Billerica, MA) was added. Colour development was stopped by adding 50 μl of 1 m NaOH after 30–60 min and optical density was read at a wavelength of 405 nm on a microtitre plate reader. To measure ragweed-specific IgG4, 96-well plates were coated, blocked and washed as above followed by the addition of 100 μl of an antibody to IgG4 at a dilution of 1 : 10 in assay buffer (supernatant of an A5 cell line provided by Dr Michael Day, University of Bristol School of Veterinary Medicine, Bristol, UK). Plates were incubated for 60 min at room temperature and washed before adding 100 μl donkey anti-mouse IgG horseradish peroxidase conjugate (Jackson Laboratories, Westgrove, PA). After an additional 60 min of incubation at room temperature and washing, 100 μl of TMB substrate (BD Biosciences, San Diego, CA) was added. Colour development was stopped by adding 50 μl 1 m H2SO4 and optical density was read at a wavelength of 450 nm (reference 550 nm) after ~ 3 min. To normalize the data run in different assays, serum collected as a single sample from an allergic dog with high allergen-specific antibody titre was added to each plate and all data are expressed as a percentage of this control.

Statistical analysis

Two-way repeated measures analysis of variance (anova) was used in comparing antibody responses and cellular constituents found in BAL fluid. Repeated measures could not be used in comparing changes in pulmonary resistance among treatment groups because the dose–response was terminated when an individual had an increase in pulmonary resistance > 200% of baseline; hence, data were not available for each individual at every dose. A (non-repeated measures) two-way anova was used for comparing pulmonary resistance among treatment groups for both methacholine responses and ragweed-specific responses. Data for systemic immunity and airways response were log-transformed before statistical analyses to better fit the assumptions of homoscedasticity and normality for parametric testing. Data sets with potential outliers were subjected to Grubb's test to determine outliers. Systemic immunity data for one of the five dogs in the Air group was excluded based on outlier testing. All post hoc comparisons were performed with a Bonferroni post-test to compare each intervention with the positive control (RW) group. All statistical calculations were carried out using graphpad prism version 5.01 for Windows (GraphPad Software, San Diego, CA) at a significance level of ≤ 0·05.


Airway response

After assessing responses to both ragweed-specific challenges, six of the 16 in the original positive control RW exposure group were determined to be non-responders in that they did not display a reproducible ragweed-specific response of ≥ 150% increase in pulmonary resistance at both 4 and 8 weeks following sensitization (RWNR; Fig. 3a,b). The RW group consisted of three males and seven females whereas the RWNR group consisted of three males and three females; three in the RWNR group were littermates to those in the RW group. Four weeks after the final sensitization, all groups had similar increases in non-specific pulmonary resistance in response to increasing doses of methacholine with no significant differences between treatment groups (data not shown). Pulmonary resistance in response to increasing doses of ragweed at 4 weeks after the last sensitization exposure was significantly increased at all doses only in the RW group, except the RW group response was not different from the RWNR group at 0·3 mg/ml ragweed. Values for all other groups remained near baseline pulmonary resistance levels for the range of ragweed concentrations used (Fig. 3a). At 8 weeks after sensitization, pulmonary resistance in response to ragweed was increased only in the RW group (< 0·0001; Fig. 3b).

Figure 3.

Allergen-free immune maturation before or Toll-like receptor 4 (TLR4) stimulation during inhalant allergen sensitization abolishes ragweed-specific airway reactivity. Four weeks after sensitization, the change in pulmonary resistance in the ragweed (RW) group was greater than all other treatment groups at each dose except for RW versus ragweed non-responder (RWNR) at the 0·3 mg/ml dose (a, > 0·05). Eight weeks after sensitization, the change in pulmonary resistance in the RW group was greater than all other treatment groups (b, *< 0·0001). Comparisons with RW in (a): a = at least < 0·05; b = RWNR > 0·05; others at least < 0·01; c = at least < 0·01; d = at least < 0·05.

Systemic immunity

Serum was obtained at 4, 8 and 20 weeks following the final sensitization, which corresponded to 35, 39 and 51 weeks of age in the delayed sensitization groups (Del/Air and Del/RW) and 24, 28 and 40 weeks of age in the other groups. Serum ragweed-specific IgE levels were significantly less in the Del/Air and Del/RW groups relative to the RW group at 4 weeks post-sensitization (Fig. 4a, < 0·001 and < 0·05, respectively) and significantly less at 20 weeks post-sensitization only in the Del/RW group (Fig. 4a, < 0·05).

Figure 4.

Allergen-free immune maturation before inhalant allergen sensitization affects the serum ragweed-specific immunoglobulin response to inhalant allergen sensitization. A delayed sensitization (matured immune system, Del/Air, Del/RW) results in less ragweed-specific IgE (a). The effect of treatment was statistically significant in each case (a, < 0·0001; b, < 0·0012; c, = 0·0001); however, post hoc testing could not define differences between ragweed (RW) and other exposure groups in (b) or (c).

Ragweed-specific serum IgG1 levels were readily apparent in the groups exposed to ragweed, but none were significantly distinguished from the RW group by post hoc testing (Fig. 4b). However, the effect of treatment was statistically significant with < 0·0001 by two-way anova.

Ragweed-specific serum IgG4 levels were readily apparent in the Del and CRX groups exposed to ragweed, but none were significantly distinguished from the RW group by post hoc testing (Fig. 4c). The effect of treatment was statistically significant with = 0·0001 by two-way anova.

Pulmonary inflammation

In BAL fluid obtained 8 weeks after sensitization, a neutrophilic response significantly greater than that of the RW group was found in the CRX/RW dogs 24 hr after ragweed challenge, which resolved by day 4 post-ragweed challenge (Fig. 5a, < 0·0001). Eosinophils were significantly increased 24 hr after ragweed challenge in BALF from the Del/Air and Del/RW groups above that of the RW group 1 day after ragweed challenge and resolved by day 4 (Fig. 5b, < 0·05). However, eosinophilia in response to the ragweed challenge was not apparent in the RW group.

Figure 5.

Allergen-free immune maturation before or Toll-like receptor 4 (TLR4) stimulation during inhalant allergen sensitization affects the pulmonary inflammatory response to allergen challenge at 8 weeks following sensitization. Inhaled CRX-527 during ragweed sensitization led to an increased neutrophilic response 1 day after ragweed challenge above that of ragweed alone (a, *< 0·0001). The ragweed exposed group (RW) had less eosinophilia than the Del/Air (b, #< 0·01) and Del/RW (b, ##< 0·05) treatment groups.


Our findings corroborate a proposed critical window for allergen sensitization[29] through the use of a dog model of atopically predisposed asthma development. When dogs that are predisposed to an allergic asthma phenotype were reared without gestational or neonatal exposure to ragweed allergen through their first 13 weeks of life (Del/RW group), they exhibited what is presumed to be a mature immune response to a ragweed sensitization protocol (i.e. became tolerant of inhaled ragweed), as evidenced by a lack of bronchoconstriction with inhaled ragweed challenge (Fig. 3b). They displayed a measurable serum ragweed-specific IgG response (Fig. 4b,c) without an apparent ragweed-specific IgE response (Fig. 4a) at this later time period after sensitization. On the other hand, dogs that were exposed to ragweed by inhalation beginning at their first week of life exhibited an early asthmatic phenotype primarily characterized by increased pulmonary resistance after allergen exposure, as well as increased serum ragweed-specific IgE relative to the delayed sensitization group (Fig. 4a). Further, innate immune stimulation before ragweed sensitization induced a lack of bronchoconstriction to inhaled ragweed allergen (Fig. 3b), but subsequent ragweed exposure resulted in a profound inflammatory response as evidenced by a neutrophilic influx following ragweed challenge (Fig. 5a).

An atopic predisposition imparted by the mother may create the neonatal window in which allergen exposure will lead to an asthmatic phenotype. We previously evaluated the effect of parental allergic status on asthmatic phenotype development in the offspring[27] and found that the offspring from non-allergic parents sensitized on the same schedule as in the current study did not develop an asthmatic phenotype, whereas offspring of allergic mothers did develop an asthmatic phenotype. In the current study, offspring of allergic mothers allowed to mature to 13 weeks of age before inhaled allergen exposure did not develop allergen-specific increases in pulmonary resistance or show non-specific airway hyper-reactivity. In agreement with our data, a mouse model of the maternal transmission of asthma risk showed offspring of asthmatic mothers (but not of normal mothers) had mild airway inflammation and a lack of non-specific airway hyper-responsiveness if allowed to mature to 10 weeks of age before sensitization.[7] Also, maternal atopy plus early allergen exposure (mite allergen in this case) in children has been suggested to increase the risk of childhood asthma.[30] Taken together, the data indicate that the maternally contributed atopic predisposition creates a sensitization window in early life in which aeroallergen exposure may lead to an asthmatic phenotype that can be abrogated by gestational and neonatal aeroallergen avoidance until the atopically predisposed immune system is more mature.

A potential factor in allergic asthma development is a reduced or delayed maturation of the immune system at the time of allergen exposure. It appears that the normal post-parturient immune system is immature in both innate and adaptive immune cell responsiveness, as illustrated by a deficiency in Th1-type responsiveness relative to adults,[11-13] and atopic predisposition may augment this deficiency, leading to an asthmatic phenotype. For example, in children that progress to allergic disease, TLR-mediated innate immune responses (TLR4 included) were increased at birth but subsequently fell below those of non-allergic children by 1–2 years of age.[31, 32] Even though the neonatal immune system is immature, TLR agonism has been shown to stimulate innate immune cells such as cord blood mononuclear cells, inducing a robust Th1 type response.[12, 33, 34] Hence, a synthetic TLR agonist plus allergen would theoretically circumvent the pro-asthmatic response to aeroallergen exposure. A murine study of allergen sensitization after very early life exposure to inhaled lipopolysaccharide and/or allergen indicates that exposure to the combination is required for allergen tolerance (as opposed to lipopolysaccharide inhalation alone) at least 3 days before beginning sensitization, which led to an increased asthmatic phenotype.[35] In our model, we pre-treated immediately before sensitization exposures with CRX-527, a synthetic lipid a mimetic that does not require CD14 for TLR4 ligand activity.[36] The pre-treatment abolished the allergen-specific increase in pulmonary resistance. However, the response to allergen does not appear to be one of tolerance. Even though CRX-527 prevented allergen-specific bronchoconstriction, its use caused a ragweed-specific airway neutrophilia and increased serum ragweed-specific IgE similar to the RW group, but no airways hyper-responsiveness. Neutrophils are common in asthma exacerbations and may play a primary role in the initiation of asthma symptoms, as well as being the primary inflammatory cell in the neutrophilic asthma phenotype.[37] It is unknown if TLR-mediated innate immune function is hyper-responsive at birth in our model as it is in children,[31, 32] which if stimulated in the presence of allergen, as in our study, may produce a chronic allergen-specific inflammatory response. Another concern in the use of TLR4 agonists is appropriate dosage because low-dose TLR4 agonism (lipopolysaccharide) appears to modify the response such that a Th2 response dominates, whereas at higher doses, a Th1 response dominates in a murine model.[21] A recent study with addition of lipopolysaccharide to allergen in a mouse model of airway sensitization has confirmed that low-dose lipopolysaccharide contributes to allergic sensitization and that a neutrophilic response contributes to airways hyper-responsiveness.[38] In a previous study, we have shown that a TLR9 agonist (CpG) given in the same manner as the CRX-527 in the current study can abrogate both the increased pulmonary resistance and the inflammatory response resulting from allergen exposure in this model.[39] Hence, TLR agonists hold promise as a preventative therapy for allergic asthma, but the dosage, selectivity of TLR agonism and potential manifestation of inflammatory and allergic disease due to their use early in life require further studies.

Dogs that were treated with CRX-527 by inhalation immediately before the first, fourth and fifth of eight ragweed exposures did not exhibit non-specific or ragweed-specific bronchoconstriction. Interestingly, CRX-527-exposed dogs had a noticeable increase in serum ragweed-specific IgG4. IgG4 is often increased along with IgE in allergic individuals and it is suggested that IgG4 may be able to block an allergic response to allergen (reviewed in Aalberse et al.[40]). Our data support this possibility in dogs as the CRX-527 treated group did not have increased pulmonary resistance following ragweed challenge (Fig. 3b) whereas they did exhibit an increase in serum ragweed-specific IgE similar to the RW group (Fig. 4a), and also had a noticeable increase in serum ragweed-specific IgG4 (Fig. 4c). Although it is possible that IgG4 can directly block an IgE-mediated hypersensitivity response, it may also be that IgG4 is secondarily up-regulated in parallel to the induced immune response that is more directly responsible for the lack of bronchoconstrictive response to allergen, e.g. regulatory T-cell maturation and IL-10/IL-21 production inhibiting the allergic response while driving IgG4 production.[40] Alternatively, CRX-527-treated dogs may be exhibiting a ‘modified Th2 response’[41] in which the combination of CRX-527 and ragweed has stimulated a stronger Th2 response (above that of ragweed only), primarily manifested by B cells that have switched to IgG4 and IgG1 producer and memory cells in which the IgG4-switched B cells may be induced to switch to IgE-producing cells upon further allergen exposure. An unfortunate consequence of this hypothesis is the production of allergen-specific memory cells (IgG4 switched) that have the potential to switch to allergen-specific IgE plasma cells which, coupled with the allergen-specific neutrophilia seen in the CRX/RW group (Fig. 5a), argues against the use of CRX-527 as an early life intervention for asthma prevention.

Early life exposure to inhaled ragweed led to a pro-asthmatic response with ragweed-specific bronchoconstriction and serum IgE response; however, other asthmatic characteristics were not present, such as eosinophilia (Fig. 2b) and non-specific airway hyper-responsiveness (data not shown) and a significant proportion of the ragweed exposed offspring of allergic mothers did not present an asthmatic phenotype (six of 16 determined to be non-responders, RWNR group). Eosinophils are present in the baseline and day 4 samples (Fig. 5b), but we cannot be sure whether an inflammatory state of the day 1 samples led to degranulation, confounding our ability to distinguish these cells from others. This may indicate that the eosinophils were activated but not heavily recruited into the lung at this age. The lack of these characteristics in our model may indicate the immaturity or early stage of the asthmatic phenotype, reflect the variability in an inhalation-only sensitization model, or represent an overall insufficient allergic stimulus. A previous study in our laboratory has shown that a more robust asthmatic phenotype can occur in the offspring of allergic mothers subjected to the same inhalation-only sensitization protocol.[27] With this inherent variability in the model, an adjuvant may be required to provide a more distinct separation of normal versus asthmatic phenotypes. Using alum in systemic administration of allergen has proven to generate a robust asthmatic phenotype in dogs sensitized in our laboratory with combined systemic (+ alum) and inhaled allergen,[26] and we currently see at least 80% presenting an asthmatic phenotype using this adjuvanted sensitization. ‘Suboptimal’ sensitization protocols used in murine studies of maternal transmission of asthma risk also rely on one intraperitoneal allergen/alum administration in addition to inhaled allergen to generate a phenotype distinguishable across treatments.[25] Further (unrelated) studies with some of the dogs from the current study indicate that an asthmatic progression of eosinophilia, airway hyper-responsiveness and high IgE response does occur (unpublished observations) later in life; however, prospective assessment of asthmatic phenotype development into adulthood of these dogs was outside the scope of this study. Interestingly, for future studies this model presents a unique opportunity to examine critical early life factors in the development of allergic asthma. With our finding that only ~ 63% of atopically predisposed neonates responded with an asthmatic phenotype (with individuals of the same litters in both responder and non-responder groups), we have an opportunity to critically evaluate the early life environmental factors (allergen dose, maternal interaction/breastfeeding, air pollution, smoke, etc.), as well as gender and individual immune maturation that contribute to the responder phenotype in a genetically homogeneous study group (similar to ‘twin’ studies in human research), which can further be compared to non-atopically predisposed offspring.

The appearance of ragweed-specific immunoglobulins in the non-ragweed (Air) exposed groups confounds the interpretation of the data. This may indicate some level of background contamination in the assay procedure or alternative ragweed exposure (from ragweed-exposed littermates); however, the ragweed-specific airway response in only the RW group lends confidence to the findings in comparing serum IgE in the RW and Del/RW groups. We do not have formal data at our facility on ambient ragweed concurrent with the time-frame of this study. However, local municipal reports indicated (from 2004 to 2009) an average ragweed season of 25 days with an average daily pollen count of 5·9 grains/m3 of air sampled. As a gross approximation, assuming a maximum minute ventilation of 2·12 l/min (20-week-old dog) for 25 days, total ventilation may be ~77 m3, which would provide exposure to a total of ~454 pollen grains presented to the upper airway. By our approximation (from the Greer pollen product RM56, Greer Laboratories, Lenoir NC) this is ~0·556 μg allergen in a pollen season presented to the upper airway, which is substantially less than the four weekly exposures of ~157 μg each and four monthly exposures of ~436 μg each that is estimated to reach the lung during sensitizing exposures. Although we cannot exclude a small ambient exposure, we believe it to be a very low amount compared with our sensitization exposures that are expressed with respect to lung deposition, which only resulted in ~ 63% of dogs responding with an allergen-specific airway response.

In the present study, we have shown that the maturity of the immune system relative to timing of exposure to inhaled allergen is important in the development of allergen-specific bronchoconstriction in a dog model of atopically predisposed asthma development. Also, strict aeroallergen avoidance in early life allows immune maturation, enabling an appropriate response to subsequent aeroallergen exposure. Further, the use of innate immune stimulants may prove useful in circumventing allergy development as in allergen vaccines for high-risk neonatal populations. However, long-term effects of TLR agonists on immune regulation and potential disease manifestation need to be further investigated.


The authors wish to thank Dr Julie Wilder for critically evaluating the manuscript before submission. This work was performed with support from National Institutes of Health grant R01AI061787.


The authors declare that they have no competing interests.