Because of the strain dependence on the development of respiratory disease following cigarette smoke (CS) exposure (Vlahos et al., 2006), male Balb/C mice were chosen for this experiment. 7-week-old male BALB/c mice were obtained from the Animal Resources Centre Pty. Ltd. (Perth, WA, Australia) and allowed to acclimatize for 1 week at the RMIT Animal Facility (Bundoora, VIC, Australia). Mice were housed in micro-isolator cages with an ambient temperature of 21°C under a 12:12 light/dark cycle (lights on at 7a.m. and off at 7p.m. AEST). Water and standard mouse chow were available ad libitum. Body weight was monitored 3 times a week. All experimental procedures and protocols were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes, and approved by the RMIT University Animal Ethics Committee (AEC 1533).

Mem71 (H3N1) is a strain of IAV with intermediate virulence, which is a result of genetic reassortant of A/Memphis/1/71 (H3N2) × A/Bellamy/42 (H1N1) (Reading et al., 1997; Gualano et al., 2008). The virus was grown in Madin-Darby canine kidney (MDCK) cells, which were maintained at 37°C/5% CO₂ in RPMI 1640 medium containing 2 mM L-glutamine, 2 mM sodium pyruvate, 30 μg/ml gentamycin, 100 I.U./ml penicillin, 100 μg/ml streptomycin and 10% v/v heat inactivated FCS (Thermo Fisher Scientific, Massachusetts, United States). For viral amplification, 90% confluent MDCK cells were washed twice in PBS and infected with egg grown virus at 0.01 plaque forming units/cell. After 1.5 h, the inoculum was removed, cells were washed twice with PBS, followed by the addition of a low protein, serum free medium (VP-SFM, Thermo Fisher Scientific, Massachusetts, United States) containing 0.5 μg/ml trypsin. Virus was harvested 48 h later, when there was extensive viral cytopathic effect. Flasks were shaken to loosen cells, the medium and cells were collected, briefly vortexed, subjected to a low-speed clearing spin and the clarified supernatant was collected and stored in aliquots at -80°C until required.

Mice were placed in 18L Perspex containers in a standard fume hood and exposed to smoke generated from three cigarettes (Winfield Red, Phillip Morris, VIC, Australia; the average particulate matter was 419 mg/m³) over a 1-h period. CS was generated in 60 ml tidal volumes over a 10 s period in order to mimic the inhalation rates observed in human smokers. Mice were exposed to CS in this fashion 3 times a day (15 min per cigarette), with a 2-h break between smoke sessions (9 cigarettes per day total). CS exposure occurred for 5 days a week for 8 weeks. Control animals (Sham) were also placed in 18L Perspex containers but were not exposed to CS. At the end of 8 weeks of CS exposure, mice were anesthetized with methoxyflurane (Medical Developments International, VIC, Australia) and intranasally inoculated with either sterile phosphate buffered saline (PBS; diluent) or influenza A virus (Mem71/H3N1 strain at 1 × 10^4.5 plaque forming units) dissolved in PBS. Following infection, CS exposure was ceased and mice were monitored periodically until the end of experiment for signs of distress (slowed respiration, reduced activity) and culled if symptoms persisted.

One day prior to termination, the locomotor activity of the mice was assessed in an open field arena. Locomotor activity was assessed between 8 a.m. and 1 p.m. to limit potential impacts of circadian rhythm. Mice were placed in the center of an empty 60 cm × 60 cm x 60 cm black plywood box and allowed to freely explore the environment for 8 min. The open field arena was cleaned with 70% ethanol prior to use and before each subsequent animal tested. The activity of the mice was filmed with a camcorder positioned above the arena (Panasonic HC-W585M, Osaka, Japan) and later scored on Ethovision XT software (v11.5; Noldus Information Technology, Wageningen, Netherlands). Briefly, a single zone was created in Ethovision that corresponded with the entire box. Mice were scored for total distance travelled within the zone during an 8-min period and trace maps were generated.

At the end of the protocol, mice were anesthetized with a mixture of ketamine (80 mg/kg)/xylazine (16 mg/kg) and incisions were made to expose the tibialis anterior (TA) muscle, with care taken to avoid cutting the fascia as previously described (Chan et al., 2021). Briefly, the distal tendon was separated from the surrounding fascia in order to free the tendon from the ankle joint. The mouse was transferred to a heated platform (37°C) attached to an in situ contractile apparatus (1300A Aurora 3-in-1 whole animal system; Aurora Scientific, Ontario, Canada) and the limb of interest was secured to the platform via a foot clamp and a needle inserted through the patellar tendon and into the platform. The separated tendon was tied and secured to an isometric force transducer via suture threads. Two fine electrodes were inserted into the muscle belly ∼3–5 mm apart and penetrating just below the surface of the muscle. The muscle was periodically bathed with warm saline to prevent drying.

The TA was stimulated twice with a sustained contraction at 100 Hz with a 2-min rest interval in between, in order to settle the muscle and tighten the knots. Optimal muscle length (L₀) was determined by increasing the length of the muscle with a micromanipulator. The muscle was stimulated with twitch contractions (single action potential allowing contraction then relaxation) every 30 s until a repeatable maximum peak twitch force was obtained. L₀ was measured using precision digital calipers as the distance between the distal myotendinous junction and the insertion of the TA at the base of the knee. A force frequency test was performed which involved contractions induced at increasing frequencies (10, 20, 30, 50, 80, 100, 150, 200, 250, and 300 Hz) at 2-min intervals. Forces were converted to a digital signal and recorded by DYNAMIC MUSCLE ANALYSIS 611ATM (Aurora Scientific).

Mice were euthanized with sodium pentobarbitone (240 mg/kg; Virbac Australia, NSW, Australia) via intraperitoneal injection. Bronchoalveolar lavage fluid (BALF) was collected by flushing out the lungs with 400 µL ice-cold PBS followed by three aliquots of 300 µL ice-cold PBS, generating ∼1 ml of BALF per mouse. The total number of viable cells in the BALF was measured by diluting 50 µL of BALF with 50 µL of an acridine orange stain (Invitrogen, VIC, Australia). Cells were counted on a standard Neubauer hemocytometer under fluorescent light with an Olympus BX53 microscope (Olympus, Tokyo, Japan). To differentiate between the cell populations in the BALF, 50,000 cells were centrifuged using ∼50–200 µL of BALF at 400rpm for 10 min on a Shandon Cytospin 3 (Thermo Fisher Scientific, Massachusetts, United States). Dried cytospots were fixed with Shandon^TM Kwik-Diff^TM fixative (Thermo Fisher Scientific, Massachusetts, United States) and stained with Hemacolor Rapid Red and Blue dye (Merck, Australia). Cells were identified and differentiated into macrophages, lymphocytes, and neutrophils according to standard morphological criteria, with a minimum of 500 cells counted per slide. Hind limb muscles (quadriceps, TA, calf muscles) were dissected from each mouse, weighed, snap frozen in liquid nitrogen, and stored at −80°C until required. Muscles for morphological analysis were frozen in Tissue-Tek O.C.T compound (ProSciTech, QLD, Australia) under liquid nitrogen cooled isopentane, then stored at -80°C until required.

Frozen O.C.T embedded muscles were sectioned in a cryostat (Leica Biosystems, United States) for immunofluorescent staining of slow skeletal muscle myosin heavy chain (SM1; Santa Cruz Biotechnology Inc., TA, United States) and laminin (Novus Biologicals, Colorado, United States), as previously described (Chan et al., 2020) with slight modifications. Briefly, sections were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer for 10 min at room temperature. Slides were then rinsed with 1X PBS +0.1% Tween® 20 (PBST) and incubated with blocking buffer (5% fetal bovine serum, 0.5% Triton-X, 0.01% sodium azide, 1X PBS) for 2 h at room temperature. Fluorophore conjugated antibodies for SM1 and laminin were added to sections and allowed to incubate overnight at 4°C. Myofiber cross-sectional area was determined by staining sections for laminin. Muscle fiber typing was determined by staining muscle sections for SM1. Following overnight incubation, sections were washed with 1X PBST and slides were mounted with Fluoromount-G™ Mounting Medium, with DAPI (Thermo Fisher Scientific, Massachusetts, United States) and cover slipped.

Slides were viewed and imaged using an Olympus Slide Scanner (VS120-S5, Olympus, United States) and analyzed using cellSens™ Life Science Imaging Software (Olympus, United States). Briefly, user defined boundaries consistent across each sample (size and location) were created and the cross-sectional area of fibers within boundaries was measured using the cellSens^TM measuring tool. Fiber types were identified based on the staining color and oxidative myofibers were manually counted and expressed as a percentage of all fibers within boundaries.

Human bronchial epithelial cell line, BEAS-2B (CRL-9609; American Type Culture Collection, United States), were grown as monolayers in a humidified incubator with 5% CO₂ at 37°C in LHC-9 medium (Thermo Fisher Scientific, Massachusetts, United States) supplemented with 50 μg/ml gentamycin and 10% fetal calf serum (Thermo Fisher Scientific, Massachusetts, United States), with medium changes every 2–3 days. C2C12 murine myoblasts (CRL-1772, American Type Culture Collection) were cultured in high glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Thermo Fisher Scientific, Massachusetts, United States) supplemented with 1% penicillin/streptomycin (100 units/mL penicillin and 100 μg/ml streptomycin; Thermo Fisher Scientific, Massachusetts, United States) and 10% fetal bovine serum (FBS; Thermo Fisher Scientific, Massachusetts, United States). Cells were cultured in a T-75 flask and were passaged at 70–80% confluence. The flasks were kept in a humidified incubator at 37°C with 5% CO₂. To induce differentiation, confluent monolayers of C2C12 myoblasts were switched to differentiation medium (DM) consist of high-glucose DMEM supplemented with 1% penicillin/streptomycin and 2% horse serum (Thermo Fisher Scientific, United States) and the DM was changed daily. All experiments were performed on day 6 when majority of myoblasts have transformed into mature myotubes.

Cigarette smoke extract (CSE) was generated as previously described (Chan et al., 2021) with slight modifications. Briefly, one cigarette (Winfield Red, Phillip Morris International, Australia) was bubbled through a 25 ml of pre-warmed experimental media at a rate of 3 ml/s to produce 100% CSE stock solution. The stock solution was sterile filtered and diluted to obtain desirable concentration for experimentation. For the CM study, BEAS-2B cells were stimulated with either polyinosinic:polycytidylic acid (poly I:C, 4 μg ml⁻¹), 25% CSE, or the combination of the two for 6 h. At the end of the stimulation, the media from the respective conditions were harvested and centrifuged at 400 g, 5 min at 4°C. The supernatants representing the CM were transferred onto differentiated C2C12 myotubes for 18 h.

Total RNA was extracted from muscle tissue and cells using the RNeasy® Mini Kit (Qiagen, Germany) according to manufacturer instructions. RNA was reverse transcribed to cDNA using the High-Capacity RNA-to-cDNA Kit (Thermo Fisher Scientific, Massachusetts, United States) according to manufacturer instructions and real time PCR reactions were performed on the QuantStudio 7™ (Applied Biosciences, United States) using mouse specific TaqMan ® Gene Expression Assays: Igf-eb (AIKALFT), Mstn (Mm01254559_m1), Cybb (Mm01287743_m1), Gpx1 (Mm00656767_g1), Il-6 (Mm00443258_m1), and Tnfα (Mm00446190_m1). All reactions were performed in triplicate and data obtained was normalized against glyceraldehyde 3-phosphate dehydrogenase (GAPDH), used as the reference gene, prior to analysis using the delta-delta Ct method.

Tissue samples were homogenized in 500 µL of ice cold RIPA lysis buffer [150 mM NaCl, 1% Triton X-100, 0.5% Na-deoxycholate, 0.1% sodium dodecyl phosphate, 50 mM Tris (pH 8)] supplemented with β-mercaptoethanol (1%) and protease and phosphatase inhibitor cocktail (1%, Cell Signaling Technology, United States). The samples were incubated on ice for 30 min before being centrifuged for 10 min at 14,000 g at 4°C and supernatant was collected for immediate use or stored at −80°C. Protein concentrations were quantified through the use of a commercially available colorimetric bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Massachusetts, United States), following manufacturer instructions. For western blots, protein samples were denatured in SDS loading buffer (4x Laemmli buffer; 50 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol, 1% β-mercaptoethanol, 12.5 mM EDTA, 0.02% bromophenol blue). Western blot analyses were performed on 12% SDS PAGE gels to assess key signalling molecules including phospho-/pan- STAT1, phospho-/pan- JNK, GAPDH, BiP, phospho-/pan-eIF2α, phospho-/pan-p70S6 kinase, phospho-/pan- S6rp, phospho-/pan- 4E-BP1, tubulin (Cell Signaling Technology, United States), MAFbx and MURF-1 (Abcam, United States). For oxyblots, protein samples were derivatized and stabilized using the OxyBlot Protein Oxidation Detection kit (Merck, Massachusetts, United States) for immunoblot detection of carbonyl groups, following manufacturer instructions. Membranes were developed using chemiluminescent substrates and imaged using the ChemiDoc system (BioRad Laboratories Inc., United States). Quantitative densitometry analysis of bands of interest was performed using Image Lab software (Ver. 6.0, Bio-Rad Laboratories Inc.).

All data are presented as mean +SEM unless otherwise stated; n represents the number of mice. Multivariate analysis was performed via two-way ANOVA, followed by Tukey post hoc test where significant differences were found. All statistical analyses were performed using PRISM 9^TM (GraphPad Software, CA, United States). In all cases, statistical significance was assumed when p < 0.05.

CS exposure suppressed body weight gain (∼2 g lower than Sham) which was maintained throughout the course of the experiment (Figure 1A). Following IAV infection, Sham mice experienced a ∼9% reduction in body weight compared to the non-infected Sham mice. Under CS condition, IAV infection also caused a similar reduction in body weight. CS exposure or IAV alone produced a robust increase in total cellularity within the BALF (Figure 1B). This was attributed to an increase in macrophage, neutrophil, and lymphocyte counts (at day 3). This was further augmented by the combination of CS exposure and IAV infection (Figures 1B–E) recapitulating that of AECOPD in human patients (Bozinovski et al., 2012; Wang et al., 2019). Similar to our previous observations (Oostwoud et al., 2016), the IAV infection-driven exacerbation was characterized by a peak in airway inflammation at day 3 that was resolved by day 10, marking the resolution phase of infection.

CS exposure caused a global reduction in mass of the hind limb muscles (Figures 2A–E), which was sustained throughout the course of the experiment. This was characterized by a 13 and 30% decrease in mass of the TA (Figure 2A) and plantaris (Figure 2C) respectively, which are predominated by fast-twitch fibers. Similar trends were observed in the slow-twitch fiber predominant soleus with a 22% decrease observed (Figure 2B), along with a 17% loss in both the mixed-fiber predominant gastrocnemius and quadriceps (Figures 2D,E). Despite the exacerbated airway inflammation, the combination of CS and IAV infection (at day 3) did not result in any further loss of muscle mass. By day 10, most of the muscle loss driven by the combination of CS exposure and IAV infection returned to Sham level (Figures 2A–E), suggesting viral exacerbation is unlikely to have additive effects on CS-induced muscle loss.

In line with the loss of TA muscle mass, CS exposure also caused a ∼10% reduction in mean myofiber cross−sectional area (CSA), which persisted throughout the course of experiment (Figure 3A). Despite the seemingly unaffected muscle mass, IAV infection also resulted in a similar reduction in mean myofiber CSA, in the Sham group. However, the suppressive effects of IAV infection on mean myofiber CSA did not appear to be alleviated on day 10, unlike that of muscle mass (Figure 3A). Furthermore, immunostaining detected a significant reduction in the proportion of oxidative muscle fibers following CS exposure, suggesting the presence of a slow-to-fast fiber type shift, like that seen in people with stable COPD (Gosker et al., 2007). In contrast, IAV infection did not significantly impact fiber type composition, irrespective of CS exposure or time after infection (Figure 3B), suggesting viral exacerbation is unlikely to contribute to fiber type transformation of limb muscle.

To investigate the impacts of viral exacerbation on limb muscle function, we performed detailed functional analysis on day 3, when airway inflammation was at its peak (Figures 1B–E). Our correlational analyses revealed that the loss of TA mass is directly correlated with the peak increase (i.e., day 3) in BALF neutrophilia (Figure 4A), which is a common characteristic of both stable COPD and AECOPD (Gompertz et al., 2001; Hoenderdos and Condliffe, 2013). Although a similar correlative trend was observed on day 10, the R₂ value was much lower (Figure 4B), suggesting the main effects of viral exacerbation on limb muscle are most likely to occur on day 3. Our muscle function analysis found that either IAV infection or CS exposure caused a significant reduction in contractile function of the TA muscle, which was further deteriorated when IAV infection was combined with CS exposure (Figure 4C). In line with this, the specific force analysis demonstrated a 41 and 36% decrease in contractile function by IAV infection and CS exposure respectively when compared to Sham animals, while a 52% decrease was observed when IAV infection was combined with CS exposure (Figure 4D), suggesting the manifestation of muscle weakness. As muscle weakness is closely linked to reduced physical activity levels (Degens and Alway, 2006; Passey et al., 2016), we assessed the locomotor activity of the mice using an open field arena. Despite its prominent effect on muscle strength, IAV infection had no apparent effect on total distance travelled, irrespective of CS exposure (Figure 4E), hence physical inactivity is unlikely to contribute to the observed muscle weakness. These findings suggest that viral exacerbation primarily impacts muscle strength rather than mass, and this is not compounded by changes in locomotor activity.

Previous studies from our laboratory (Chan et al., 2021) and others (Barreiro et al., 2010; Fermoselle et al., 2012) have demonstrated the involvement of oxidative stress in CS-induced muscle dysfunction, however, its role in IAV infection remains unclear. An oxyblot assay on the TA muscle detected a 60% increase in protein carbonylation following IAV infection in Sham mice, and this was not further elevated when combined with CS exposure (Figure 5A). At the molecular level, IAV infection elicited a 15-fold increase in Cybb [NADPH oxidase 2 (Nox2)] and a 3-fold increase in glutathione peroxidase 1 (Gpx-1) mRNA expression. However, this response appeared to be completely blunted when IAV infection was combined with CS exposure (Table 1). IAV infection also subdued myogenic signaling, characterized by a downregulation in mRNA expression of myogenin (Myog; 60% reduction), a positive regulator of myogenesis, and concomitant upregulation of myostatin (Mstn; 10-fold), which inhibits myogenesis, irrespective of CS exposure (Table 1). In line with the induction of oxidative stress and blunted myogenic signaling, the protein abundance of binding immunoglobulin protein (BiP) and muscle RING-finger protein-1 (MuRF-1) were markedly elevated by IAV infection, irrespective of CS exposure (Figure 5B). BiP is a cellular redox sensor and a master regulator of protein folding (Wang et al., 2014), while MuRF-1 is a downstream effector of myostatin that promotes ubiquitin-mediated protein degradation in skeletal muscle (Desgeorges et al., 2017). These findings imply that IAV infection does not further exacerbate the oxidative burden in the muscle or increase the disruption to myogenic signaling incurred by CS exposure.

To ascertain the contributory role of airway inflammation in the disruption of myogenic signaling, we treated bronchial epithelial cells (BEAS−2B) with poly I:C, a well-established viral mimetic used to exacerbate inflammation in lung injury models (Vella et al., 2020), in the presence or absence of CSE, to recapitulate a viral exacerbation. Poly I:C stimulated the phosphorylation of signal transducer and activator of transcription 1 (STAT1; Figure 6A) and resulted in a 20-fold increase in the expression of interleukin-6 (Il-6; Figure 6B). In contrast, CSE elicited rapid phosphorylation of c-Jun N-terminal kinase (JNK) which was accompanied by an increased expression of tumor necrosis factor α (Tnfα: Figures 6A,B). Although cell viability was seemingly unaltered (Supplementary Figure S1A), co-exposure to poly I:C and CSE resulted in a marked increase in interferon β (Ifnβ; 2.4-fold), Il-6 (54-fold) and Tnfα (∼12,000-fold) mRNA expression, suggesting an amplified toll-like receptor (TLR)-3-driven inflammatory response in these bronchial epithelial cells.

To determine whether the TLR3-driven inflammatory response of the lung cells may potentiate myogenic disruption by CSE, we transferred the culture medium of the BEAS−2B cells (i.e., CM) onto C2C12 myotubes for a time-course experiment, as outlined in Figure 7. Myotubes treated acutely (4-h exposure), with CM from poly I:C-treated cells displayed no observable changes in key signal transducers of protein folding [BiP, eukaryotic initiation factor 2α (eIF2α)], protein synthesis [p70 S6 Kinase, S6 ribosomal protein, eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1)] or in muscle-specific F-box (MAFbx); a negative regulator of muscle mass (Passey et al., 2016). Despite not having an apparent effect on the structural integrity of the myotubes (Supplementary Figure S1B), 24-h exposure to the poly I:C CM resulted in a marked increase in BiP and eIF2α phosphorylation (Figures 7A–C), suppression of S6 ribosomal protein phosphorylation (Figures 7A,E), and elevation of MAFbx abundance (Figures 7A,G), suggesting a shift in myogenic balance. CM from CSE-exposed cells had distinct impacts on the myogenic signaling, characterized by an even greater phosphorylation of eIF2α without an increase in BiP protein expression (Figure 7B), increased p70 S6 Kinase phosphorylation (Figure 7D), diminished 4E-BP1 phosphorylation (Figure 7F) and a concomitant increase in MAFbx protein expression (Figure 7G). CM from the poly I:C + CSE-exposed lung cells increased BiP and MAFbx abundance, like the poly I:C alone experiments, and furthermore, significantly suppressed the phosphorylation of S6 ribosomal protein.

We next performed qPCR on the CM-exposed myotubes for markers of myogenesis and inflammation. Unlike CSE, direct stimulation of C2C12 myotubes with poly I:C had no significant impact on myogenic genes and Il6 expression (Figure 7H). Similar to what was seen in our in vivo IAV infection model (Table 1), poly I:C CM caused a ∼50% reduction in Myog mRNA expression, without any significant changes observed in the gene expression of either Fbx32, which encodes for the MAFbx protein, or Il6 (Figure 7H). Importantly, CM from poly I:C + CSE did not further increase these changes, suggesting the TLR3-driven inflammatory response of the lung cells is unlikely to have an addictive effect on myogenic disruption driven by CSE.