Thirty-four adult female Lewis (LEW/SsNHsd) rats weighing 225–275 g (Harlan, Indianapolis, IN) were included in the study. Female Lewis rats were used due to the susceptibility of this gender and strain to persistent inflammation (Sternberg et al., 1989
; Tonelli et al., 2003
). In the experimental setup, group 1 (n
= 10) received VF injection of the LPS solution (from E. coli
serotype 0111:B4, Sigma-Aldrich Co., St. Louis, MO) to induce inflammation combined with trauma from needle penetration. LPS is a bacterial endotoxin associated with gram-negative bacteria, which produces a variety of physiological responses, including inflammatory and immune response modulation (Jacobs, 1981
). Group 2 (n
= 10) received VF injection of saline solution to model trauma from needle penetration alone; group 3 (n
= 10) did not receive any laryngeal manipulations and served as anesthetic control; and group 4 (n
= 4) received acute direct electrical stimulation of the iSLN to examine the excitability of the first-order neurons of the laryngeal afferent pathway within the nodose and jugular ganglia without changes in VF tissue integrity. All animals were maintained on a 12-h light/dark cycle and given ad libitum
access to food and water. All animals received humane care in compliance with the National Institute of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80-23, revised 1996). The study protocol was approved by the Animal Care and Use Committee of the National Institute of Neurological Disorders and Stroke, National Institutes of Health.
Each animal was initially anesthetized with a mixture of 3% isoflurane and 100% oxygen delivered via a ventilatory calibrated anesthetic pump to an induction chamber. The anesthetized animal was removed from the chamber, laid supine on the surgical carriage, and maintained on isoflurane through a head mask. An intravenous (IV) catheter was placed into the tail vein to maintain hydration, and a slow drip of saline with 0.1 ml of heparin solution was administered at a rate 3 ml/kg/h.
In groups 1, 2, and 3, isoflurane delivery was gradually reduced and replaced with 0.3–0.5 ml of propofol solution (10 mg/ml) to maintain anesthesia for 30 min. Group 1 and 2 animals were placed in the surgical chair with their mouth fixed in open position using a custom-designed laryngoscope to visualize the larynx (Inagi et al., 1998
). Using a transoral microlaryngoscopic approach (Zeiss, Germany), group 1 animals were injected with a sterile LPS solution (10 μg/5 μl) and group 2 animals received an equivalent volume of sterile saline injection (5 μl). Methylene blue crystals were added to both solutions prior to injection to mark the injection site. All injections were placed into the anterior or middle portion of the right VF using a 10 μl Hamilton syringe. Group 3 animals were maintained under propofol anesthesia for 30 min without any laryngeal manipulations. Animals in all groups were closely observed until full recovery and then returned to their cages in the animal care facility.
All animals in groups 1, 2, and 3 were allowed to survive for 72 h to capture the long-lasting VF tissue changes following inflammation and trauma. All animals were monitored by a veterinarian in the animal care facility during the survival period to detect any behavioral and/or physiological changes due to the experimental procedures. All animals' body weight and core body temperature were measured and recorded twice a day.
After the survival period, animals were again anesthetized with isoflurane as described above and then maintained on alpha-chloralose saline solution (0.7 mg/100 ml) at 8–20 μl/min via the tail IV over a 4-h period to control for unintended Fos expression due to handling of the animals and experimental manipulations. A circulating water blanket was placed underneath the animals to prevent hypothermia. In all animals, core body temperature using rectal temperature gauge, blood oxygenation level, heart rate, and respiratory rate were monitored continuously, and checks for the lack of a withdrawal response to a painful stimulus were conducted regularly at 30 min intervals. At the end of the 4-h period, the rats were deeply anesthetized with IV propofol solution (0.6 ml, 10 mg/ml) and perfused transcardially with phosphate buffer solution (PBS, pH 7.4) followed by 4% paraformaldehyde in 0.1 M PBS (pH 7.4) at the rate of 45 ml/min for 10 min.
Group 4 animals were intubated using an endotracheal tube and maintained under isoflurane anesthesia during exposure of the iSLN. The rats were placed into a stereotaxic frame in the supine position, and the trachea was exposed from the cricoid cartilage to the sternal notch using blunt dissection through a midline incision at the level of the hyoid bone. The endotracheal tube was replaced with a tracheal cannula to maintain the animal under isoflurane anesthesia and to allow the larynx to move freely during stimulation of the iSLN. Bipolar hooked wire electrodes were placed into the thyroarytenoid muscle bilaterally to confirm an electromyographic (EMG) response to stimulation of the iSLN. The right iSLN was exposed and a bipolar stimulating electrode was positioned over the nerve at the level of its entry into the larynx. A grounding electrode was placed in the right thigh of the animal. Isoflurane delivery was then gradually reduced and replaced with alpha-chloralose solution via tail IV (1.4 mg/100 ml) with an initial bolus of 0.2–0.4 ml to maintain anesthetic effect at 8–20 μl/min over a 4-h period (the total dose of 670–6.0 mg/kg). The ventilator connected to the tracheal cannula was used to maintain regular breathing. The rats were kept under quiet conditions with low lighting for 3 h before electrical stimulation of the iSLN was performed. This allowed for the control of c-fos expression due to handling of the animal and surgical procedures, which were not relevant to the experimental conditions. During this 3-h period, saline-moistened gauze was placed over the surgical site to maintain tissue hydration, and the stimulating electrode was immersed in warm (37°C) mineral oil. Electrical stimulation of the right iSLN was performed using a 100–500 μA stimulus with a pulse width of 0.2 ms and frequency of 0.5 Hz over a period of 45 min. Recordings of vital signs and the EMG of the thyroarytenoid muscle confirmed elicitation of the laryngeal reflex without affecting cardiac or respiratory function. Following the electrical stimulation of iSLN, the animals were maintained under quiet conditions for 20 min. The animals then received an IV injection of heparin solution (0.1 ml, 5000 units/1 mL) followed by barbiturate overdose and bilateral pneumothoraces. Following euthanasia, the animals were perfused with PBS (pH 7.4) followed by 4% paraformaldehyde in 0.1 M PBS (pH 7.4) at the rate of 45 ml/min.
Following perfusion, the brains, larynges, and bilateral portions of the vagal nerve containing the nodose and jugular ganglia were harvested for tissue processing. The brains and larynges were post-fixed overnight in paraformaldehyde and cryoprotected in 30% sucrose solution at 4°C for 3 days. The nodose and jugular ganglia were first blocked in a 20% gelatin solution to preserve tissue entity and then fixed in 4% paraformaldehyde solution overnight before cryoprotection in 30% sucrose solution at 4°C for 2 days. The brainstem was separated from the spinal cord through the transverse section at the level of the obex and from the forebrain at the level of the rostral periaqueductal gray (PAG), serially sectioned in the stereotaxic frontal plane at 40 μm on a freezing microtome (Zeiss, Germany), and collected into separate wells with 0.1M PBS (pH 7.4) for further immunostaining for Fos protein, pro-inflammatory cytokine IL-1β, microglia-specific ionized calcium binding adaptor molecule 1 (Iba-1), and control Nissl stain.
Brain sections were washed twice in PBS and incubated in 0.1% H2O2 for 30 min to inhibit endogenous peroxidase activity. Following another three washes in PBS, sections were blocked in 0.4% PBS/Triton X-100 (PBS-T) containing 1.5% normal goat serum for Fos immunostain, 1.5% normal rabbit serum for IL-1β stain, and 1.5% normal goat serum and 1% bovine serum albumin (BSA) for Iba-1 stain, respectively, for 1 h at room temperature. Sections were then incubated with rabbit anti-Fos polyclonal antibody (Ab-5, dilution 1:5000; Calbiochem, San Diego, CA), anti-rat IL-1β antibody (dilution 1:1000; R&D Systems, Minneapolis, MN), and rabbit anti-Iba1 antibody (dilution 1:5000; Wako Chemicals, Richmond, VA), respectively, in PBS-T overnight at 4°C. After three washes in PBS, the sections were incubated in biotinylated anti-rabbit IgG antibody for Fos and Iba-1 immunostains (dilution 1:200, Vector Laboratories, Burlingame, CA), respectively, and in biotinylated anti-goat IgG antibody for IL-1β immunostain (dilution 1:200, Vector Laboratories, Burlingame, CA) for 1 h at room temperature. For the control staining, only normal rabbit IgG was used. Sections were further incubated in Vectastain Elite ABC-kit (Vector Laboratories, Burlingame, CA) for 1 h at room temperature and in the DAB substrate kit solution (Vector Laboratories, Burlingame, CA) for 5 min at room temperature to visualize the bound antibody as a brown reaction product. Sections were then washed and mounted onto chrome-alum-gelatin coated slides. After drying overnight, the control sections were counterstained with cresyl violet. All slides were then dehydrated through an ascending alcohol series and xylene and cover slipped.
The blocked nodose and jugular ganglia were serially sectioned at 20 μm on a freezing microtome (Zeiss, Germany). The sections were processed for Fos immunostain as described above.
The larynges obtained from the animals with LPS and vehicle injections were sectioned at 20 μm and stained with Meyer's hematoxylin and eosin (H&E) for verification of the injection site and evaluation of VF changes due to experimental procedures.
Data analysis was conducted blinded without prior knowledge of animal group identity. For each of Fos, IL-1β and Iba-1 immunostains, brainstem sections 100 μm apart from each other were examined bilaterally and quantified from the rostral PAG to the caudal medulla in each animal. The brainstem sections were reconstructed using the corresponding Nissl-stained control sections and the stereotactic atlas of the rat brain (Paxinos and Watson, 1998
) on an image analysis system (Neurolucida, MicroBrightField, Colchester, VT). Examined brainstem regions included the second-order sensory nuclei of laryngeal afferents, that is the NTS, spinal trigeminal nucleus (Sp5), and intermediate/parvicellular reticular formation (IRF/PCRF) (Altschuler et al., 1989
; Patrickson et al., 1991
; Travers and Norgren, 1995
; Hayakawa et al., 2001
); the laryngeal motor nucleus, that is the NA (Gacek, 1975
; Yoshida et al., 1982
); the higher-order nucleus of laryngeal control, that is the PAG (Ambalavanar et al., 1999
); and the non-specific nuclei of the laryngeal control, that is the area postrema (AP) and locus coeruleus (LC) (Ambalavanar et al., 2004
). Within the NTS, we examined its medial (NTSm), interstitial (NTSi), and intermediate (NTSim) subnuclei; within the PAG, we examined its lateral (LPAG), dorsomedial (DMPAG), dorsolateral (DLPAG) and ventrolateral (VLPAG) subnuclei. The mean number of immunopositive cells was computed in each of the examined regions on the right (ipsilateral to injection) and left (contralateral to injection) sides of the brainstem (Neurolucida, MicroBrightField, Colchester, VT).
The nodose and jugular ganglia, the sites of the first-order neurons of laryngeal afferents, were examined in vehicle-, LPS-treated, and iSLN-stimulated animals to determine the effects of VF trauma, inflammation, and acute iSLN electrical stimulation, respectively, on FLI in these neurons.
The H&E-stained laryngeal tissue from the LPS- and vehicle-treated animals was examined to quantify the extent and severity of VF changes. The latter was assessed by computing the ratio between the total area of VF damage (in μm2
) and the total area of the injected VF (in μm2
), which resulted from the use of methylene blue, on manually outlined laryngeal sections (Neurolucida, MicroBrightField, Colchester, VT). In addition, a scoring system from 0 (none or normal) to 3 (severe or diffuse) (Tsai et al., 2006
) was used to estimate epithelial damage and severity of changes in the VF tissue induced by local inflammation and/or trauma.
Because some animals had no response in the same region (marked as zero at quantitative analysis), the data were transformed by adding a constant of one to each cell for each region for all animals before statistical analysis. As the Shapiro–Wilk test found that data were not normally distributed (W = 0.811 p = 0.019 in the LPS group; W = 0.723 p = 0.002 in the vehicle group; W = 0.841 p = 0.045 in the anesthetic group), we used non-parametric tests to assess the statistical differences between the groups while accounting for the variance differences in FLI. We conducted a priori Kruskal–Wallis non-parametric analysis to estimate the overall group differences between the LPS-, vehicle-injected, and anesthetic groups, including all examined brainstem structures, at p = 0.05. If the group effect was statistically significant on any structure, those structures were further examined using separate post hoc Mann–Whitney U-tests to determine the significance of differences in Fos expression between the groups. Because only two brainstem structures (IRF/PCRF and NTSi) were significantly different on the initial Kruskal–Wallis test, the post-hoc Mann–Whitney tests were conducted on these structures only with the significance level adjusted to 0.025 to correct for multiple comparisons.
The relationships between FLI in the brain structures that showed differences between the groups (i.e., IRF/PCRF and NTSi) and the severity of VF changes in LPS- and vehicle-treated animals were assessed using Spearman's rank order correlation coefficients at a priori R = 0.6 and p = 0.05.
Because only spurious IL-1β expression was found in few animals across all experimental groups (2 LPS-treated, 3 vehicle-treated, and 1 anesthetic control) and no positive Iba-1 immunoreactivity was identified in either group, the effects of IL-1β and Iba-1 expressions were not quantified for statistical analyses.