|Home | About | Journals | Submit | Contact Us | Français|
Respiratory viral infections are the leading cause of acute illness during infancy and are closely linked to chronic inflammatory airway diseases later in life. However, the determinants of susceptibility to acute respiratory infections still need to be defined.
We investigated whether the individual variation in the antiviral response at birth determines the risk for acute respiratory illness in the first year of life.
We studied 82 children who were enrolled in a birth cohort study of inner-city children with at least one parent with allergy or asthma. We cultured cord blood monocytes and assessed interferon-γ (IFN-γ) and C-C chemokine ligand 5 (CCL5) mRNA production at 24 hours after inoculation with respiratory syncytial virus (RSV). We also monitored the frequency of acute respiratory illness at 3-month intervals and analyzed nasal lavage samples for respiratory viruses at the time of illness during the first year..
Respiratory infection was reported for 88% of subjects, and respiratory viruses were recovered in 74% of symptomatic children. We observed a wide range of antiviral responses in cord blood monocytes across the population. Furthermore, a decrease in production of IFN-γ (but not CCL5) mRNA in response to RSV infection of monocytes was associated with a significant increase in the frequency of upper respiratory tract infections (r = -0.42, p<0.001) and prevalence of ear and, sinus infections, pneumonias and respiratory hospitalizations.
Individual variations in the innate immune response to respiratory viruses are detectable even at birth, and these differences predict the susceptibility to acute respiratory illness during the first year of life.
Individual variations in the immune response to respiratory viruses are detectable at birth, and these differences predict the susceptibility to acute respiratory illness during first year of life.
Respiratory viral infections are a common cause of early childhood illness. Most of these infections are short-lived and self-limited, but some can be severe enough to require hospitalization. Indeed, respiratory viral infections are associated with 20% of all mortality in children under 5 years of age 1. In addition to the morbidity of the acute infection, respiratory viral infections with wheezing are strong indicators of subsequent asthma 2, 3. Therefore, predicting those infants at risk for respiratory infections is an important first step infections preventing acute and chronic respiratory disease. Previous studies have identified a variety of potential risk factors for lower respiratory viral infections during the first year of life. These factors include daycare attendance, number of siblings, small lung size, exposure to tobacco smoke, low birth weight, and premature birth 3-6. Infections due to respiratory syncytial virus (RSV) are particularly implicated in acute illness and chronic lung disease in the first two years of life 7. However, the majority of RSV infections in infants occur without any known risk factors 8. Thus, we still do not understand the precise mechanism for the wide variation in susceptibility to severe respiratory infections among children in these settings.
One possible explanation for the range of susceptibility to respiratory viral infection in early childhood is that there are definable variations in the antiviral response, such as a congenital deficiency in the innate immune response that can be detected even at the time of birth. A central ingredient of the innate immune response to respiratory viruses is the system for interferon (IFN) production and signaling 9. In that regard, a decrease in interferon-γ (IFN-γ) production from cord blood mononuclear cells (CBMCs) stimulated by phytohemagglutin or allergens has been associated with increased risk for acute respiratory illness during infancy 10, 11. Perhaps more relevant to viral infection, the lack of a detectable IFN-γ response to RSV in CBMCs was associated with decreased wheezing in the first year of life, but a detectable response was only found in a third of individuals so predictive power was limited 12. Therefore, in the present study, we developed alternative methods to determine whether the innate immune response of virus-infected CBMCs could predict the later development of respiratory illness. We used RSV to activate CBMCs based on the well-established association of RSV infection with subsequent childhood asthma 2, 13, 14. However, to monitor the innate immune response to RSV, we determined the induction of the genes encoding IFN-γ and a remarkably virus-responsive chemokine CCL5 based on sensitive and quantitative methods for mRNA detection 15. We also monitored IFN signal transduction by tracking the level of STAT1 activation in response to IFN-β stimulation. In both cases, we used cultured/adherent CBMCs to select for monocytes versus the mixed cell population that included T cells in previous studies. We assessed whether each of these immune endpoints could predict the development of respiratory illness during first year of life in a prospective birth cohort of children at high risk for asthma and allergic disease. The experimental matrix led to the unexpected finding of RSV-induced IFN-γ gene expression in monocytes as a predictor of subsequent respiratory viral illness.
We analyzed cord blood samples from 82 newborns that were enrolled in the Urban Environment and Childhood Asthma (URECA) study. This group represents a subset of the 178 children enrolled at the St Louis site, which in turn was a subset of the total number of children enrolled at the Baltimore, Boston, and New York city sites between February 2005 and March 2007 as described previously 16-18. Subjects were required to have at least one parent with allergic rhinitis, eczema and/or asthma and to reside in an area with >20% of the residents below the poverty level as well as being born at ≥34 weeks gestation. At the St Louis site, a small number of children without an allergic parent (n = 5) were also recruited for comparison. After enrollment, all subjects were monitored for any episodes of acute respiratory illness over the next year along with quarterly assessments of respiratory (and non-respiratory) illness and wheezing by questionnaire. Nasal lavage samples were obtained whenever a caregiver reported an acute respiratory illness and at the time of a 1-year follow-up visit. The Washington University Human Research Protection Office approved the study protocol.
Cord blood samples were collected in the delivery room and CBMCs were isolated by density gradient centrifugation using Accuspin tubes (Sigma) within 16 hours of collection as described previously 17, 19. When sufficient amounts of sample were available (i.e., in 82 of 178 subjects), the cells were resuspended in RPMI medium with 10% FBS, 2mM L-glutamine, and 1 mM non-essential amino acids to a final concentration of 8 × 105 per ml and plated in 4-well Labtek chambers (500 μl per well) for real-time PCR assay) and in 2-well Labtek chambers (1 ml per well) for STAT1 activation assay as described below. Nonadherent cells were removed after 24 hours, and adherent cells were cultured for 5 days with media changes on days 1, 2 and 4 and removal of additional nonadherent cells. At the end of the cell culture period, the adherent cells were >95% positive for CD68 immunostaining as a marker of monocytic lineage and therefore designated as cord blood monocyte cultures. The approach avoided purification methods (such as magnetic-bead selection or FACS) that modify the cell membrane or cell culture methods (such as growth factor supplementation) that promote full differentiation and polarization and thereby aimed to obtain cells of the monocyte lineage that were representative of naïve lung tissue monocytes and macrophages (the target for respiratory viral infection in vivo).
On culture day 5, cord blood monocytes were infected with RSV (A2 strain) at MOI 7.5 or an equivalent amount of UV-inactivated RSV (RSV-UV). Cellular RNA was isolated immediately and 24 hours after inoculation using the RNAeasy mini kit (QIAGEN, Valencia, CA) and transcribed to cDNA using the high capacity cDNA reverse transcription kit (Applied Biosystems). Single target quantitative real-time PCR was used to monitor IFN-γ and CCL5 mRNA and RSV RNA levels. For IFN-γ and CCL5 mRNA, primers were obtained from Applied Biosystems (Hs00174575_A1 and Hs00174143_A1). For RSV RNA, primers 5312F (5’-TCCCTACGGTTGTGATCGATAGA-3’) and 5396R (5’-TGATGGGAAGTAGTAGTGTAAAGTTGGT-3’) and probe 5349T (5’-AGGTAATACAGCCAAATC -3’) targeting the viral L gene were based on the sequence of RSV strain CRD2 (GenBank accession number DQ340570). For GAPDH mRNA, primers 50F (5’-CAGCCGAGCCACATCCCTCAGACACCAT-3’) and 125R (5’-CTTTACCAGAGTTAAAAGCAGCCCTGGTGACCA-3’) and probe 88T (5’-AGGTCGGAGTCAACCGATTTGGTCGTATTG-3’) were used. Plasmids encoding CCL5 and IFN-γ (OriGene, Rockville, MD) and a portion of the RSV L-gene (nt 5400-7016) and GAPDH gene sequence (GenBank accession number NM_002046) were used to generate RNA standards. RT-PCR was performed using TaqMan real-time PCR Master Mix with 5 μL of sample cDNA in accordance with the manufacturer's protocol (Applied Biosystems). All data for gene copy number was normalized to GAPDH level.
In a subset of cord blood samples (n = 63) with an adequate number of cells, we also assessed IFN signal transduction by monitoring the level of STAT1 phosphorylation in response to IFN stimulation. The corresponding CBMCs were processed as described above and serum starved on day 4 of culture. On culture day 5, the cells were incubated with IFN-β (100 U/ml) for 30 minutes. Cell lysate was harvested after treatment with cell lysis buffer (Cell Signaling, MA). The level of total STAT1 was determined using ELISA (Invitrogen), and phosphorylated STAT1 (Tyr701) was determined using a sandwich ELISA (PathScan Phospho-Stat1, Cell signaling, MA).
Nasal lavage samples were obtained during acute respiratory illnesses during the first year of life and at the one-year follow-up visit. For illness samples, a respiratory symptom scorecard was completed as described previously 10, 18. When the score indicated a moderate to severe respiratory illness, the site staff obtained a nasal lavage sample within 48 hours. All nasal lavage samples were processed for identification of nine respiratory viruses using a PCR-based assay as described previously 18.
Descriptive data were expressed as percentages, means ± standard deviations (SD), or medians with interquartile ranges (IQR) for non-normally distributed data. To test differences between specific groups, chi-squared or Fisher's exact tests were used to compare categorical variables, while unpaired t-tests were used to compare continuous variables. Appropriate log-transformations were made to the data to yield an approximately normal distribution. For non-normally distributed data, the Wilcoxon-rank-sum tests (Mann-Whitney-U tests) were used to compare groups. Each measurement was standardized as ratio over control to minimize variability in day-to-day experiments as follows: IFN-γ mRNA response to RSV = IFN-γ mRNA copies with RSV/IFN-γ mRNA copies without RSV, CCL5 mRNA response to RSV = CCL5 mRNA copies with RSV infection/CCL5 mRNA copies with no RSV infection, and STAT1 phosphorylation with IFN-β = (P-STAT1 with IFN-β / Total STAT1 with IFN-β) / (P-STAT-1 no stimulation / Total STAT1 no stimulation). RSV level was expressed as RSV RNA copy number with infection minus the value for no infection. To test for associations between variables, Spearman correlations were calculated. We examined the possibility of confounding by the following variables: gender, breastfeeding, maternal smoking during pregnancy and the first year of the child's life, an overall sum of bedroom allergen exposure (Mus m, Bla g, Fel d), and exposure to endotoxin (recombinant factor-C assay) and ergosterol. We evaluated whether any of these variables were associated with the outcomes under analysis, and we found only that maternal smoking during the first year of life was related to the pneumonia outcome and the sum of bedroom allergen exposures was related to sinus infections. However, in logistic models controlling for these variables, the odds ratios changed by less than 5% when adjusted, so we present unadjusted relationships here. All statistical tests were two-tailed, and p-values of less than 0.05 were considered statistically significant. Statistical procedures were conducted using both SAS 9.2 and R 2.12.2 software.
We processed all cord blood samples that contained an adequate number of cells, representing 82 of the total of 178 children who were enrolled at the St Louis site of the URECA cohort. Among the 82 newborns, 85% of the babies were African-Americans, the mean age of the mother at the time of delivery was 23.7 years of age, and 63% of the infants had at least one parent with asthma (Table 1). Subjects were reported to have an average of 4.2 upper respiratory tract infections (colds), 1.3 wheezing illnesses, and an all-causes hospitalization rate of 15% during the first year of life. The basic demographics and first year outcomes for this group of 82 children were not significantly different from the remaining group of 96 children that were not part of the present analysis (Supplementary Table 1).
We collected nasal lavage samples during symptomatic respiratory episodes (n = 38) and at routine 1-year follow-up visits (n = 64). Respiratory viruses were detected in 74% of the symptomatic episodes and in 48% of the routine visits (Table 2). Human rhinovirus was found with the highest and RSV with the next highest frequency in symptomatic episodes, but both could also be detected asymptomatic individuals. Together, the findings indicate that the majority of respiratory illnesses during the first year of life are associated with detectable levels of respiratory viral pathogens, but asymptomatic infants also have a high rate of apparent viral carriage.
The distribution of values for RSV-induced IFN-γ and CCL5 mRNA, RSV RNA, and STAT1 activation showed a broad range among the subject population suggesting a degree of individual variation among subjects for each of these responses (Table 3). This finding further suggested that the differences in responses might translate to variable degrees of protection against viral infection.
We next investigated whether individual variation in antiviral response was associated with the development of respiratory infections during the first year of life. We found that IFN-γ responses to RSV in cord blood monocyte cultures correlated inversely with the number of upper respiratory tract infections during the first year of life (r = -0.42, p<0.001) (Figure 1). There was no correlation between IFN-γ response and available measures that might associate with subject atopy, i.e., number of parents with asthma, number of parents with hayfever, and presence of parental asthma at the time of initial screening. The relationship to respiratory tract infections was selective for the IFN-γ response, since we found no association between CCL5 response, RSV titer, or STAT1 activation with the frequency of upper respiratory tract infections (r = -0.17, p = 0.17; r = -0.01, p = 0.94; and r = -0.02, p = 0.91, respectively). We found no association of IFN-γ response, CCL5 response, RSV titer, or STAT1 activation with the frequency of wheezing episodes in the first year of life (r = -0.05, p = 0.71; r = -0.12, p = 0.32; r = -0.04, p = 0.75; and r = 0.12, p = 0.36, respectively).
We also investigated whether there was an association of antiviral responses with the occurrence of infections at other respiratory and non-respiratory sites (i.e., sinus, ear, croup, stomach) as well as any association with the number of reported hospitalizations for respiratory illness. We found that IFN-γ responses to RSV in cord blood monocyte cultures were inversely related to the frequency of ear infections (p = 0.01), sinus infections (p = 0.04), pneumonias (p = 0.02) and respiratory hospitalizations (p = 0.05) (Figure 2). We found no differences in IFN-γ responses between those who did and did not have croup, “stomach flu”, or unexplained fevers (Figure 2 and data not shown). Together, these findings reinforce the association of a decreased IFN-γ response to RSV with the development of increased respiratory virus infections in the first year of life.
In this study, we provide evidence that a decreased antiviral IFN response at the time of birth is selectively associated with an increase in acute respiratory infections in first year of life among infants at high risk for the development of asthma and allergic disease. In support of this relationship between antiviral response and respiratory infection, we show that: 1) RSV-driven induction of IFN-γ mRNA production in cord blood monocytes is variable among infants at birth; (2) decreased levels of RSV-induced IFN-γ mRNA in cord blood monocytes are associated with a significant increase in the frequency of upper respiratory tract infections as well as prevalence of ear infections, sinus infections, pneumonias and respiratory illnesses requiring hospitalization; 3) levels of RSV-induced CCL5 mRNA expression (another highly inducible antiviral system) and IFN-driven activation of STAT1 (the downstream target of the IFN receptors) are not associated with this phenotype for subsequent illness; 4) symptomatic respiratory illnesses were frequently associated with detectable levels of respiratory viral pathogens; and 5) levels of RSV-induced IFN-γ mRNA were not linked to other types of infections such as croup or stomach flu due to other types of pathogens. Together, these findings provide for a close relationship between RSV-driven IFN-γ mRNA production and the development of respiratory viral illness and in turn suggest the possibility that a decrease in this type of response might lead to an increase in this type of illness.
The present findings offer distinct insights from those obtained previously. For example, others found that phytohemagglutinin- and allergen-stimulated IFN-γ production in CBMCs was inversely correlated with frequency of respiratory viral infection in first year of life 10, 11. However, these studies likely measured the responsiveness of T cells since nonadherent cells were not eliminated and T cell mitogen and antigen were used for stimulation. Moreover, the effect of viral infection itself was not assessed in these studies. Other studies examined the capacity of PBMCs to produce IFN-γ during RSV illness in children but in this case, cells were activated with PMA-ionomycin and cross-linking antibodies to T cell co-stimulatory receptors and the response was localized to CD3+ T cells 20. Similarly, others again made no attempt to purify cells, then stimulated cells with T cell mitogen, and did not separately quantify IFN-γ mRNA levels 21. The same approach was taken in previous reports of an association between decreased IFN-γ production from CBMCs and an increased risk of allergic sensitization and recurrent wheezing 12, 22, 23. Here again, this may reflect the focus on T cell production of IFN-γ and the proposed role of Th1 versus Th2 cytokines in the development of atopy and asthma.
After these studies were done, we have come to better recognize the critical role of the innate immune system in controlling viral infection and postviral asthma 9. In particular, the IFN and the monocyte-macrophage systems are required for protective immunity against respiratory viruses such as RSV 24-26, and these same systems are capable of driving postviral asthma independent of the adaptive immune system at least in experimental models 27-29. Therefore, the present approach was designed to directly assess the innate immune response to viral infection and was done so using the relevant cell type (purified monocytes) and stimulus (RSV infection) as well as more specific and sensitive methods (real-time quantitative PCR) than applied previously. The upshot is the first evidence that the monocyte IFN-γ gene also serves as a marker and might even participate in the susceptibility to infection during infancy. This unexpected finding provides a new lead for control of the antiviral response, since the previous view was that IFN-γ gene expression was silenced in the monocyte lineage and was only active in lymphoid cells 30.
In that regard, we note that IFN-γ production is generally attributed to Natural Killer (NK) cells, NKT cells, and T cells whereas monocytes and macrophages are solely a target of IFN-γ action. Indeed, studies of atopic disease in infancy often focus exclusively on T cell production of IFN-γ 31. This circumstance is also likely due to the lower levels of IFN-γ produced by monocytes and macrophages under conditions used in previous studies. Here we are able to measure IFN-γ production by using a sensitive assay for the corresponding mRNA. Whether this level of IFN production has functional consequences is uncertain, but its utility as a biomarker for susceptibility to viral infection proved quite useful. In contrast, it appears that the monocyte-macrophage lineage is critical for host defense against respiratory viruses (including RSV). In particular, lung macrophages are charged with clearance of infected cells without dying themselves, and this protection derives from an anti-apoptotic survival function of the chemokine CCL5 25. If this function is lost (e.g., in mice that are CCL5 deficient), the host is more susceptible to respiratory viral infection. Other work suggests that CCL5 is also needed to direct dendritic cell traffic in the face of viral infection 32. Each of these observations are consistent with those linking CCL5 promoter gene polymorphisms to susceptibility to severe RSV bronchiolitis 33. Nonetheless, we did not find that RSV induction of CCL5 gene expression was significantly associated with respiratory viral infection rates in the first year of life. It is still possible, however, that CCL5 production at the level of the lung macrophage would be predictive of susceptibility to infection, especially given the heterogeneity of monocyte-macrophage populations in the circulation and the lung.
Our study was conducted in a population selected for high prevalence of atopic disease. However, we do not expect that IFN-γ response or viral susceptibility is attributable to atopy. Indeed, we found no association of the IFN-γ response with the available measures associated with atopy in our subject group, since full evaluation of atopic status in our subjects was not yet performed at 1 year of age. Similarly, others found no association between lower respiratory illness in the first year of life and the occurrence of parental atopy or subject eczema 11. These findings suggest that viral susceptibility can be independent of atopy, and certainly there is evidence that this can be the case in experimental models of respiratory viral infection 25. However, it will require a non-atopic cohort and/or follow-up of the present cohort to formally test the relationship between virus-induced IFN-γ production in monocytes, viral susceptibility, and atopic status in humans.
The molecular mechanism for RSV induction of IFN-γ gene expression still needs to be defined. Thus, the pathogen recognition receptor (PRR) system is responsible for mediating viral induction of the various forms of IFN-α, IFN-β, and IFN-λ, and RSV is remarkably effective in blocking the induction (and signaling) of these IFN species 34, 35. By contrast, the PRR system of Toll-like receptors (TLRs) and RIG-I-like receptors (RLRs) does not appear to regulate IFN-γ gene expression or signaling. Instead, IFN-γ gene expression is subject to a distinct type of positive and negative regulation at pre-transcriptional, transcriptional, and post-transcriptional levels at least in the case of lymphoid (NK, NKT, and T) cells 30. However, these regulatory mechanisms have not been studied in monocytes or in response to RSV in any cell type. We could not define this regulatory mechanism with such limited human samples, but our work should open this new field of research.
RSV is the most common cause of serious respiratory illness during infancy, and severe RSV-induced bronchiolitis is linked to subsequent wheezing illness/asthma. Further, paramyxoviral infection is established as a high-fidelity experimental model of asthma. Moreover, RSV and related paramyxoviruses are easily detected in monocytes and macrophages in the lung. Hence, we chose RSV for our study of newborns and their monocytes, unaware of course that the subsequent results would predict respiratory illness associated with rhinovirus. With such a severe limitation in cell sample size, we were unable to test multiple viruses, but an analysis of the response to other types of viruses (including rhinovirus) and viral strains (including other RSV types) would be a valuable goal in the future.
In sum, we report that congenital variations in the innate immune response may predict the susceptibility to acute respiratory illness during the first year of life. Our effort uncovered evidence that the IFN-γ monocyte response to virus rather than T cell response to mitogen/allergen might be linked to the development of viral infections and eventually postviral asthma. We were able to define this relationship despite a relatively small sample size. Sample availability for complex immunologic analysis served to limit the number of URECA participants that could be studied. However, the demographic characteristics of our cohort were no different from the overall group, suggesting that we studied a representative sample of subjects. In addition, our study included mostly African-American children, so the result may not be generalized to children with other racial backgrounds. Interestingly, others have recently found that ISGs (e.g., PYHIN1) are also linked to the development of asthma, particularly in individuals of African descent 36. Our findings are also consistent with observations of decreased IFN-α production in response to RSV in PBMCs from older children and adults with allergic asthma 37, 38. Thus, these early events in infancy may carry over to a similar deficit in antiviral defense in later life. Together, the findings suggest that a full analysis of IFN production pathways may provide key insights into the susceptibility to respiratory viral infection and subsequent chronic obstructive lung diseases such as asthma.
Decreased IFN-γ response to RSV in cord blood monocytes was predictive of increased upper respiratory tract infections during the first year of life in an inner city cohort at high-risk for allergy and asthma.
Supported by grants from the National Institutes of Health (National Institute of Allergy and Infectious Diseases and National Heart, Lung, and Blood Institute).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.