Prenatal Exposure to Respiratory Syncytial Virus Alters Postnatal Immunity and Airway Smooth Muscle Contractility during Early-Life Reinfections

Maternal viral infections can have pathological effects on the developing fetus which last long after birth. Recently, maternal-fetal transmission of respiratory syncytial virus (RSV) was shown to cause postnatal airway hyperreactivity (AHR) during primary early-life reinfection; however, the influence of prenatal exposure to RSV on offspring airway immunity and smooth muscle contractility during recurrent postnatal reinfections remains unknown. Therefore, we sought to determine whether maternal RSV infection impairs specific aspects of cell-mediated offspring immunity during early-life reinfections and the mechanisms leading to AHR. Red fluorescent protein-expressing recombinant RSV (rrRSV) was inoculated into pregnant rat dams at midterm, followed by primary and secondary postnatal rrRSV inoculations of their offspring at early-life time points. Pups and weanlings were tested for specific lower airway leukocyte populations by flow cytometry; serum cytokine/chemokine concentrations by multiplex ELISA and neurotrophins concentrations by standard ELISA; and ex vivo lower airway smooth muscle (ASM) contraction by physiological tissue bath. Pups born to RSV-infected mothers displayed elevated total CD3+ T cells largely lacking CD4+ and CD8+ surface expression after both primary and secondary postnatal rrRSV infection. Cytokine/chemokine analyses revealed reduced IFN-γ, IL-2, IL-12, IL-17A, IL-18, and TNF-α, as well as elevated nerve growth factor (NGF) expression. Prenatal exposure to RSV also increased ASM reactivity and contractility during early-life rrRSV infection compared to non-exposed controls. We conclude that maternal RSV infection can predispose offspring to postnatal lower airways dysfunction by altering immunity development, NGF signaling, and ASM contraction during early-life RSV reinfections.


Introduction
Respiratory syncytial virus (RSV) is the leading cause of lower respiratory tract infection (LRTI) in children under 5 years of age worldwide and is hallmarked by potentially life-threatening bronchiolitis and pneumonia [1,2]. Furthermore, strong epidemiologic evidence associates infant RSV LRTI with increased risk of wheezing episodes and asthma later in life [3][4][5][6][7][8]. Despite this relationship, the exact mechanisms by which early life RSV LRTI predispose to chronic airway dysfunction remain poorly understood. Host immune responses and lower airway inflammation produced during RSV LRTI is clearly of great importance in clearing RSV infection and influence disease severity outcomes [9,10]. In particular, cytotoxic T lymphocytes are central in the control of active infection and viral clearance, which explains why immunocompromised individuals with deficient cell-mediated immunity experience more severe and prolonged RSV disease and shed the virus much longer [1,11].
Chronic airway dysfunction developing after early-life RSV LRTI results also from virusdriven modulation of local nerve growth factor (NGF) expression leading to increased neurotransmitters release and neuronal hyperreactivity [12][13][14]. Accordingly, increased NGF expression and cholinergic innervation were demonstrated within the lower airways of fetal rats exposed to RSV in utero [15], without significant change of the other key neurotrophin brain-derived neurotrophic factor (BDNF) [16]. The same study demonstrated the presence of a transplacental route of RSV transmission, the ability of this virus to infect fetal lower airway epithelium, and non-specific airway hyperreactivity (AHR) during postnatal RSV reinfection [15].
Among several aspects requiring additional investigation, the influence of maternal RSV infection on postnatal offspring immunity, neurotrophins expression, and mechanism of airway smooth muscle contractility during postnatal RSV LRTI remains largely unknown. Recently, vertical transmission of viral antigens was reported to impact postnatal immunity whereby macaques exposed to viral epitopes in utero displayed altered immunity after postnatal virus challenge [17]. Regarding RSV, the concept of maternal-to-fetal transmission during pregnancy is not unrealistic as evidenced by the documentation of multiple sites of extrapulmonary RSV infection [18][19][20][21][22][23][24][25]. Yet, the idea that a pregnant woman infected with RSV could potentially transmit the virus to her unborn child was unheard of until only recently and raises genuine concerns for potential implications in the pathogenesis of chronic airway diseases. Indeed a very recent document from the Advisory Committee on Immunization Practices of the Center for Disease Control and Prevention (CDC) has recommended the immunization of pregnant women to prevent maternal to infant transmission of the infection [26].
Our previous discovery of vertical RSV transmission led us to investigate whether in utero exposure to RSV influences specific aspects of cell-mediated host immunity and airway smooth muscle contractility during postnatal reinfections. We feel the results of this study demonstrate that maternal RSV infection conveys lasting effects on postnatal offspring immunity, which coincide with elevated NGF expression and airway smooth muscle contractility during recurrent early-life RSV LRTI.

Maternal rrRSV infection: experimental approach
To determine if maternal infection with rrRSV impacts the development of postnatal offspring immunity during early-life rrRSV infections, we bred Fischer-344 rats and confirmed pregnancy through vaginal swabbing to time gestation. Dams were inoculated intratracheally at mid-gestation (day E12) using recombinant RSV strain A2 expressing red fluorescent protein (rrRSV; 4 × 10 6 PFU; Fig 1A) or an equal volume of virus-free control medium. Term offspring were then inoculated with control medium or rrRSV (4 × 10 5 PFU) at postnatal day 10 (D10) only, or sequentially at D10 and D23 (5 × 10 5 PFU; Fig 1B).
In contrast to the changes in T cell subpopulations, we found only smaller, non-significant differences in B cell ( Fig 2D) and monocyte lineage ( Fig 2E) cell counts during primary postnatal rrRSV infection. CD4 + /CD8 + T cell ratios, confirmed the extent of CD8 + T cell suppression in RR pups compared to the CR pups (0.6 ± 0.05 vs. 0.19 ± 0.02 arbitrary units; P<0.0001; Fig 2F). Furthermore, lower airway CD8 + T cell counts were equivalent in RR and RC pups lacking postnatal rrRSV exposure, suggesting altered immune response to rrRSV in RR pups as a consequence of prenatal exposure.
Elevated CD3 + CD4 neg CD8 neg T cells persist during recurrent postnatal rrRSV LRTI Prior infection with RSV does not produce lasting immunity and demonstrates the virus' ability to evade host immune mechanisms. Therefore, we sought to assess whether recurrent postnatal rrRSV infections impact lower airway immunity. We again observed markedly elevated CD3 + T cells (P<0.01) in weanling rats from rrRSV-infected mothers during sequential postnatal rrRSV LRTI (RRR) compared to all other groups ( Fig 3A). Increased CD3 + T cell counts observed in RRR weanlings seem the result of maternal rrRSV infection, as CRR weanlings displayed comparatively less CD3 + T cells (3.7% ± 0.94% vs. 0.67% ± 0.16%; P<0.0001). CD3 + T cell counts observed in CRR weanlings after repeated postnatal rrRSV inoculation were equivalent to CD3 + T cell counts seen in the non-infected CCC weanlings. Unexpectedly, CD4 + and CD8 + T cell levels were similar in CRR and RRR weanlings (Fig 3B and 3C). This finding suggests that maternal rrRSV infection exerts a lasting influence on lower airway T cells counts in offspring during recurrent postnatal rrRSV LRTI, but this is not reflective of overall CD4 + and CD8 + subsets that instead tend to normalize after the primary infection.
Opposite to our findings in D15 pups, lower airway B cells were significantly reduced in CRR compared to RRR weanlings during a secondary postnatal rrRSV infection (0.95% ± 0.22% vs. 6.96% ± 0.36%; P<0.0001; Fig 3D), and total lower airway B cells in RRR weanlings were not significantly different from those of CCC and RCC weanlings. Similarly, lower airway monocyte lineage cells were significantly elevated in RRR compared to CRR weanlings at D28 (26.8% ± 4.89% vs. 10.42% ± 4.31%; P<0.05; Fig 3E). Also, unlike what we observed in D15 pups, the CD4 + /CD8 + T cells ratio was similar between RRR and CRR weanlings at D28 (1.64 ± 0.22 vs. 2.03 ± 0.23 arbitrary units; P = 0.246; Fig 3F). Collectively, these data suggest prenatal exposure to RSV fails to maintain lasting suppression of lower airway CD8 + T cell levels after repetitive postnatal RSV infections. However, CD8 + T cell activation status in RRR and CRR weanlings was not evaluated and remains important to characterize in future studies. The effect of maternal infection on lower airway cell counts in offspring during early-life rrRSV infections at D15 (RR vs. CR) and D28 (RRR vs. CRR) are summarized in Table 1.

Maternal rrRSV infection alters cytokine expression in offspring with early-life rrRSV LRTI
Serum samples were harvested from D15 pups during the peak of rrRSV infection and multiplex ELISA was used to measure the levels of 27 cytokines and chemokines. Compared to controls (CR), pups from rrRSV-infected mothers with early-life rrRSV reinfection (RR) displayed markedly reduced levels of IFN-γ (2.99 ± 1.61 pg/ml vs. 13.41 ± 4.31 pg/ml, P<0.05; Fig 4A), IL-2 (0.0 ± 0.0 pg/ml vs. 15.98 ± 2.07 pg/ml, P<0.0001; Fig 4B), IL-17A (12.75 ± 3.94 pg/ml vs. 39.58 ± 3.41 pg/ml, P<0.001; Fig 4C), and IL-18 (653.4 ± 144.7 pg/ml vs. 1,745 ± 611.2 pg/ml, P<0.01; Fig 4D). Remarkably, while IL-2 was amply detected in all other D15 groups, it was undetectable in all RR pups tested across multiple independent experiments and multiplex ELISA. Serum IL-2 was also reduced in RC compared CC pups (7.01 ± 3.51 pg/ml vs. 24.71 ± 6.39 pg/ml; P<0.05), suggesting maternal rrRSV infection conveys lasting modifications of early-life immune development, while the profound IL-2 suppression in RR pups indicated that competent activation of effector T cells in response to rrRSV antigens is not effective at generating proficient Th1 immunity as it has been reported in human studies [27,28]. Similarly, IL-18 data support the hypothesis that deficient IFN-γ production in RR pups was due to reduced macrophage activation during acute rrRSV LRTI, as alveolar macrophages have been shown to

cells; as well as (D) B cells and (E) monocyte lineage cells. (F)
The mean CD4 + /CD8 + T cell ratio is shown for each experimental group. Data for 3 experiments were combined and expressed as the mean ± SEM of !6 rats per group. Data are expressed as percentage of total CD45 + cells or total CD45 + / CD3 + cells. Two-tailed student's t test: *P<0.05, **P<0.01, ***P<0.001 compared to the RR group. express IFN-γ in response to IL-18 signaling [29] and we found no significant difference in lower airway monocyte lineage cells between CR and RR pups at D15.

Maternal rrRSV infection conveys lasting suppression of cytokines expression in offspring
Using serum obtained at D28 from sequentially rrRSV-inoculated weanlings, we utilized multiplex ELISA to assay levels of the same cytokines/chemokines array. We found that after    ml vs. 159.9 ± 11.15 pg/ml, P<0.02; Fig 5D), and TNF-α (68.85 ± 13.61 pg/ml vs. 101.7 ± 6.66 pg/ml, P<0.04; Fig 5E). Importantly, with the possible exception of IL-18 (429 ± 70.13 pg/ml vs. 627.4 ± 65.29 pg/ml, P = 0.055; Fig 5F), which showed a strong but statistically non-significant trend, the same cytokines found to be suppressed in RR pups at D15 remained significantly reduced in RRR weanlings during recurrent early-life rrRSV LRTI at D28. Systemic IL-12p70, IFN-γ, and TNF-α were significantly suppressed in RRR compared to CRR weanlings at D28, whereas lower airway CD4 + T and CD8 + T cell levels were shown to be equivalent. These findings support the concept that innate cytokine suppression in RRR weanlings results from impaired T eff cell activation and not reduced infiltration of T eff cells into the lower airways during recurrent early-life rrRSV LRTI. This hypothesis is also consistent with clinical reports of reduced CD8 + T cell activation, proliferation, and function during RSV LRTI leading to suppressed Th1 immunity hallmarked by deficient viral clearance, prolonged disease, and increased severity [30][31][32]. Additionally, CD3 + T cell counts in the lower airway of RRR weanlings were shown to be approximately 4.5-fold higher during recurrent postnatal rrRSV LRTI compared to CRR weanlings. Taken together, these data indicate maternal rrRSV infection conveys lasting effects on postnatal immune function possibly through the suppression of specific T-cell subsets during rrRSV reinfections.
In pups first inoculated with rrRSV on D10-but not again on D23-all cytokines and chemokines analyzed via multiplex ELISA were not significantly different across treatment groups at D28. While IL-2 and IFN-γ continued to display non-equivalent levels in vertically infected weanlings compared to controls, these differences were not significant (S1 Fig). Of particular importance for the proposed link between RSV and asthma, IL-5 was markedly reduced in RRR compared to CRR weanlings during recurrent postnatal rrRSV LRTI (360.6 ± 55.77 pg/ ml vs. 488.1 ± 17.92 pg/ml; P<0.05). We also found that IL-4 (90.46 ± 23.62 pg/ml vs. 145.7 ± 16 pg/ml; P = 0.0671) and IL-13 (84.3 ± 23.84 pg/ml vs. 138.7 ± 13.42 pg/ml; P = 0.0585) were lower in RRR compared to CRR weanlings, but these differences were not significant. Therefore, the differences in cytokine and chemokine expression observed between CRR and RRR weanlings at D28 seemingly diminish in the absence of secondary early-life rrRSV infection.
Prenatal exposure to rrRSV is associated with increased postnatal NGF expression As we have previously shown elevated neurotrophin expression in vertically infected fetal rats [15], we sought to investigate what influence maternal rrRSV infection may have on neurotrophin expression in offspring with primary early-life rrRSV LRTI. Thus, we measured NGF and BDNF concentration in serum collected from pups at D15. We found higher NGF concentration in RR compared to CR pups at D15 (4,599 ± 254.9 pg/ml vs. 2,395 ± 236.9 pg/ml; P<0.0001; Fig 6A). These data suggest a pathogenic link between elevated NGF and increased response to nerve stimulation previously observed in vertically infected pups during early-life rrRSV LRTI [15].
We next sought to find additional endpoints to capture intrinsic ASM dysfunction. To do so, we identified and optimized a novel parameter which we termed "tissue recovery time" (TRT). To measure TRT, we performed ex vivo electrical field stimulation (EFS) on tracheal ring specimens using a 10-s pulse of electrical frequency at 30 Hz and recorded the length of time required for a specimen to reach its initial baseline resting tension. Our EFS single-frequency TRT data closely parallel and effectively support methacholine challenge findings ( Fig  7B). Specifically, we observed a 35% increase of TRT in RRR compared to CRR weanlings (2.61 ± 0.5 min vs. 1.69 ± 0.46 min; P = 0.1130), which is equivalent to an average increase of 54.75 s in contraction time above resting baseline. This finding suggests that weanling rats from RSV-infected mothers require a greater amount of time (almost 1 full minute) to elapse before their lower airway smooth muscle returns to its resting state during secondary early-life rrRSV infection. Although the mechanism remains unclear, taken together our tissue bath data suggest that the lower airway smooth muscle of prenatally exposed rats is primed for both stronger and longer contractions in response to a triggering stimulus well after birth.

Discussion
In vivo rodent models of vertical RSV transmission have shown that maternal-to-fetal transfer of this virus is possible and predisposes to airway hyperreactivity during a primary early-life infection with the same pathogen [15]. However, whether maternal RSV infection has consequences for the development of the offspring immune system and how it responds during a postnatal reinfection has not been characterized until now. Our results demonstrate that maternal RSV infection alters postnatal offspring immunity, resulting in lower airway lymphocyte profiles that are significantly different from that of a RSV-naïve host during a first postnatal RSV LRTI. Specifically, we found large CD3 + T cell elevations in the context of reduced CD4 + T, CD8 + T, and B cell populations, which remind of the findings reported by Welliver and colleagues in children succumbing to severe RSV disease [33]. One of the limitations of the present study, however, is lack of in depth analysis of RSV-specific T cell responses and immunologic memory development in our vertically infected rat pups, which we plan to examine in future studies.
As infants are infected with RSV during an early development window where host responses to the virus are incomplete and fail to prevent reinfection [1,[34][35][36], we studied whether maternal RSV infection affects lower airway immune cell counts during repeated postnatal RSV LRTI and observed sustained CD3 + T cell elevation in the lungs of weanling rats from infected mothers during a second early-life reinfection. Similar to that seen in our primary infection experiments, the CD3 + T cell population observed in weanlings from infected mothers consisted primarily of CD3 + CD4and CD3 + CD8 -T cells that were not present in controls from non-infected mothers. This finding suggests that maternal RSV infection facilitates the maintenance of phenotypically different CD3 + T cell subsets within the lower airways of offspring with recurrent early-life RSV LRTI, which might interfere with CD4 + T and CD8 + T cell activation and function. While remaining important to investigate, the influence Another finding with potential clinical implications is that offspring from RSV-infected mothers displayed elevated B cell and monocyte lineage cell counts within their lower airways during repetitive postnatal RSV LRTI. Thus, the ineffective immunoglobulin production observed in several human studies [35][36][37][38][39] might arise from defective maturation of competent B cells and/or monocytes. Additionally, the combination of low cytokine levels and high B cell counts in RRR rats may be reflective of what has been reported in clinical studies by Mejias et al. [40] on whole blood gene expression, showing that specific T and B cell genes involved in both innate and adaptive immunity are consistently under-expressed in infants with the most severe RSV disease.
Our data also demonstrate that alterations of host anti-RSV immunity persist across consecutive early-life reinfections, suggesting that postnatal antiviral responses are programmed in utero through modification of the developing immune system. A possible explanation is that in our experiments the fetus is exposed to RSV in its preimmune phase, i.e., before the establishment of immunological self-tolerance is complete, and it is therefore conceivable that viral antigens might be recognized as self during an early-life reinfection. In support of this hypothesis, the prolonged IL-2 suppression in vertically infected rats during postnatal RSV LRTI suggests that those pups may be failing to recognize the virus as pathogenic and non-self, since IL-2 is produced physiologically as a result of naïve Th0-to-Th1 phenotype switching [27].
Our data are also well aligned with clinical reports of reduced IFN-γ levels correlating with greater disease severity in infants with RSV LRTI who fail to mount competent T-cell mediated immune responses [41,42]. Taken together, these observations suggest that vertical RSV transmission might predispose to impaired CD4 + and CD8 + T cell-mediated immunity during early-life RSV LRTI. It will be necessary to thoroughly examine and confirm this hypothesis with future studies in order to determine whether the reduced host immunity we are observing is due to functional tolerance mediated by a currently unknown CD3 + T cell subset capable of inhibiting CD4 + and CD8 + T cell activation, or it is rather a consequence of transferred maternal immunity.
We have shown previously that vertical RSV transmission leads to NGF overexpression in fetuses extracted from infected dams, as well as in the lungs of offspring [15]. Accordingly, pharmacological blockade of TrkA, the cognate receptor for NGF, reduced AHR in vertically infected pups during early-life RSV reinfection [15]. In the present study, we observed nearly 2-fold higher NGF levels in pups from RSV infected-mothers compared to controls noninfected in utero. The translational relevance of these data derives from clinical evidence that early-life RSV infection is linked to chronic airway dysfunction in humans [3][4][5][6][7][8]43], which is mediated at least in part by NGF [14,15,44]. Furthermore, increased NGF/TrkA axis expression in tracheal aspirates from mechanically ventilated infants with naturally occurring RSV LRTI suggests that disease severity closely parallels local concentrations of this neurotrophin [14]. As a potential practical application of these studies, NGF has been proposed as an effective biomarker to predict infant RSV LRTI severity [45].
Lastly, we found that dysfunctional host immunity and elevated NGF levels are associated with stronger and prolonged ASM contraction in vertically infected rats. Prenatal RSV exposure resulted in postnatal AHR in response to both ex vivo methacholine challenge and EFS, suggesting that RSV interferes with the development of both pre-and post-synaptic neuronal function. Furthermore, ASM of weanling rats from RSV-infected mothers required an appreciably greater amount of time to return to its baseline tone following contraction. Notably, vertical RSV transmission conveyed a lasting influence on ASM function observed across multiple, temporally spaced early-life RSV LRTI. The significance of these findings stems again from the known epidemiologic association between early-life RSV LRTI and the development of recurrent wheezing and chronic asthma well after the acute phase of infection has resolved [15,46]. It is also important to note that direct stimulation of lower airway smooth muscle from pups with recurrent postnatal RSV LRTI produced significantly lower reactivity to methacholine in the absence of in utero exposure. This is important as it implies that prenatal RSV exposure is required for the development of postnatal AHR, whereas being infected one or more times in early-life may not be sufficient.
In conclusion, this study provides evidence that rats born to RSV infected mothers have altered immune responses, amplified neurotrophic signaling, and increased contractility of their lower airways smooth muscle during postnatal reinfections with the same virus. The most significant effects measured after primary early life infection were a sharp increase in CD3 + CD4 neg CD8 neg T cells combined with virtual suppression of the production of key Th1-type cytokines like IFN-γ and IL-2; a large increase in the expression of the prototypical neurotrophin NGF; and increased pre-and post-synaptic reactivity of the ASM combined with intrinsic hyper-contractility delaying its return to resting tone. These changes persistedalbeit to a lesser degree-after secondary reinfection and might provide a plausible explanation to the development of chronic airway dysfunction in a subpopulation of children with history of severe RSV infections in infancy.

Ethics statement
All procedures utilized in this study adhered to the NIH Guide for the Care and Use of Laboratory Animals and were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of the Lerner Research Institute at Cleveland Clinic.

Animals
Ten week old, pathogen-free Fischer 344 (F-344) rats (Envigo RMS, Indianapolis, IN) were housed under sterile conditions within a rat ABSL-2 facility using polycarbonate isolation cages mounted on racks providing positive individual ventilation with class-100 air at a rate of one cage change/minute (Lab Products, Seaford, DE). All cages and components including bedding, water hydra-packs, and chow were autoclaved before use and handled inside class-100 laminar flow hoods. Dams were pregnancy timed and were dosed with virus-free medium or inoculated with rrRSV at E12, as previously described [47].

rrRSV preparation
Recombinant RSV-A 2 expressing the red fluorescent protein (RFP) gene (rrRSV) was kindly provided by Dr. Mark Peeples (Nationwide Children's Hospital, Columbus, OH) and Dr. Peter Collins (National Institutes of Health, Bethesda, MD) [48]. Expression of viable RFP requires successful full-length RSV replication and the rrRSV strain construct (BN1) used in these experiments is described elsewhere [48,49]. Stock rrRSV was propagated using HEp-2 cells (ATCC CCL-23; American Type Culture Collection, Manassas, VA) in 1X DMEM with 10% fetal bovine serum. HEp-2 cells at 70% confluence were inoculated, harvested and titrated as described previously [15]. To obtain virus-free inoculum, HEp-2 cells were identically cultured and harvested.

rrRSV inoculation
Dams received rrRSV suspension (4.0 × 10 6 PFU) or an equal volume of virus-free medium by intratracheal (i.t.) instillation at E12 [15], and their litters were delivered at full term. At D10, pups received rrRSV (4.0 × 10 5 PFU i.t.) or an equal volume of virus-free medium. This established 4 groups-CC, CR, RC, and RR-where 'C' stands for control and 'R' stands for rrRSV, with the left letter referring to the prenatal inoculation status and the right letter referring to the postnatal inoculation status. Terminal experiments were conducted 5 days post-inoculation (D15). For postnatal reinfection at D23, weanlings previously inoculated at D10 received another dose of rrRSV (5.0 × 10 5 PFU) or an equal volume of virus-free medium. This established a distinct set of 4 additional groups-CCC, CRR, RCC, and RRR-with terminal experiments performed 5 days later (D28).

Bronchoalveolar lavage
Rats were intubated using an 18-gauge blunted cannula inserted below the cricoid cartilage before 1.5 ml of chilled lavage buffer (1X PBS, 0.5% BSA, 0.5 mM EDTA, 25 mM HEPES) was squirted and retrieved. This procedure was repeated for a total of 2 independent lavages. Cells were stained for viability with Live/Dead Stain (Life Technologies, Carlsbad, CA), and then stained using FITC anti-CD45. Next, cells were stained with the following cocktail of secondary conjugated anti-rat monoclonal antibodies diluted per manufacturers' instructions: V450 anti-CD4, APC anti-CD3, PE-Cy5 anti-CD45RA, PE-Cy7 anti-CD11b/c, and PE anti-CD8α (all from BD Biosciences, San Jose, CA). The choice to use CD45RA as a marker for B cells is a unique advantage of working with a rat model where mature B cells can be easily distinguished from other leukocytes on the basis of their expression of CD45 isoforms. Flow cytometry was performed with FACSAria (BD Biosciences) and data were analyzed using FlowJo software (TreeStar, Ashland, OR).
In both assays, three cartilage ring-wide tracheal segments were suspended between stainless steel supports in tissue baths (Radnoti, LLC; Monrovia, CA). Passive tension was set at 0.3 gm during a tissue equilibration phase of 1 h. Isometric force changes were recorded using PowerLab Systems software (ADInstruments, Colorado Springs, CO). After reference contraction in response to KCl (80 mM), buffer was replaced and samples returned to resting tension.
To obtain cumulative dose-response curves, increasing concentrations of methacholine (Sigma-Aldrich, St. Louis, MO) were added to the bath solution and contraction force expressed in grams was normalized to KCl. EFS frequency-response curves were generated by increasing frequency incrementally from 0.1 to 30 Hz, using a submaximal voltage of 32 V, 10-ms pulse duration, and 11-s square wave trains. We also developed a strategy to quantify intrinsic muscle contractility by measuring the time required for specimens to return to resting tension after peak contraction during the 30-Hz EFS pulse interval (tissue recovery time, TRT).

Statistics
All in vivo data are expressed as mean ± SEM. Two-tailed, unpaired Student's t-test analysis was performed to determine statistical significance using Prism software (GraphPad, La Jolla, CA). P-value of <0.05 was considered statistically significant.