Volume 104, Issue 5 p. 729-739
Free Access

Respiratory disturbances in a mouse model of Parkinson's disease

Luiz M. Oliveira

Luiz M. Oliveira

Department of Pharmacology, Institute of Biomedical Science, University of São Paulo, São Paulo, SP, Brazil

Search for more papers by this author
Maria A. Oliveira

Maria A. Oliveira

Department of Pharmacology, Institute of Biomedical Science, University of São Paulo, São Paulo, SP, Brazil

Search for more papers by this author
Henrique T. Moriya

Henrique T. Moriya

Biomedical Engineering Laboratory, University of São Paulo, São Paulo, Brazil

Search for more papers by this author
Thiago S. Moreira

Thiago S. Moreira

Department of Physiology and Biophysics, Institute of Biomedical Science, University of São Paulo, São Paulo, SP, Brazil

Search for more papers by this author
Ana C. Takakura

Corresponding Author

Ana C. Takakura

Department of Pharmacology, Institute of Biomedical Science, University of São Paulo, São Paulo, SP, Brazil


Ana C. Takakura, Department of Pharmacology, Institute of Biomedical Science, University of São Paulo, 1524 Lineu Prestes Avenue, 05508-000 São Paulo, SP, Brazil.

Email: [email protected]

Search for more papers by this author
First published: 13 February 2019
Citations: 24

Edited by: Ken O'Halloran

Linked articles: This article is highlighted in a Viewpoint article by Lucking & O'Halloran. To read this paper, visit https://doi.org/10.1113/EP087653.


New Findings

  • What is the central question of this study?

    Clinical reports have described and suggested central and peripheral respiratory abnormalities in Parkinson's disease (PD) patients; however, these reports have never addressed the occurrence of these abnormalities in an animal model.

  • What is the main finding and its importance?

    A mouse model of PD has reduced neurokinin-1 receptor immunoreactivity in the pre-Bӧtzinger complex and Phox2b-expressing neurons in the retrotrapezoid nucleus. The PD mouse has impairments of respiratory frequency and the hypercapnic ventilatory response. Lung collagen deposition and ribcage stiffness appear in PD mice.

Parkinson's disease (PD) is a neurodegenerative motor disorder characterized by dopaminergic deficits in the brain. Parkinson's disease patients may experience shortness of breath, dyspnoea, breathing difficulties and pneumonia, which can be linked as a cause of morbidity and mortality of those patients. The aim of the present study was to clarify whether a mouse model of PD could develop central brainstem and lung respiratory abnormalities. Adult male C57BL/6 mice received bilateral injections of 6-hydroxydopamine (10 μg μl−1; 0.5 μl) or vehicle into the striatum. Ventilatory parameters were assessed in the 40 days after induction of PD, by whole-body plethysmography. In addition, measurements of respiratory input impedance (closed and opened thorax) were performed. 6-Hydroxydopamine reduced the number of tyrosine hydroxylase neurons in the substantia nigra pars compacta, the density of neurokinin-1 receptor immunoreactivity in the pre-Bӧtzinger complex and the number of Phox2b neurons in the retrotrapezoid nucleus. Physiological experiments revealed a reduction in resting respiratory frequency in PD animals, owing to an increase in expiratory time and a blunted hypercapnic ventilatory response. Measurements of respiratory input impedance showed that only PD animals with the thorax preserved had increased viscance, indicating that the ribcage could be stiff in this animal model of PD. Consistent with stiffened ribcage mechanics, abnormal collagen deposits in alveolar septa and airways were observed in PD animals. Our data showed that our mouse model of PD presented with neurodegeneration in respiratory brainstem centres and disruption of lung mechanical properties, suggesting that both central and peripheral deficiencies contribute to PD-related respiratory pathologies.


Breathing is a complex motor activity fundamental for life. The main function of the respiratory system is to provide oxygen to an organism and remove carbon dioxide originating from cellular metabolism. The ventilatory process is responsible for maintaining the plasma pH and body temperature values with​​in physiological ranges, also contributing to maintenance of the body's homeostasis (Dempsey et al., 2012).

Breathing problems are recognized as a cause of morbidity and mortality in patients with Parkinson's disease (PD), caused by neural damage in specific regions of the ventral medulla that are responsible for rhythm generation and chemoreception, such as the preBötzinger complex (preBötC) and retrotrapezoid nucleus (RTN), respectively (Baille et al., 2016; Benarroch, Schmeichel, Low, & Parisi, 2003; Owolabi, Nagoda, & Babashani, 2016; Seccombe et al., 2011; Torsney & Forsyth, 2017). Previous studies have also found that PD patients showed a reduction of pulmonary volumes and stiffening of respiratory muscles, indicating an impairment directly in the lung, ribcage and/or the muscles responsible for breathing (Baille et al., 2018; LeWitt et al., 2018; Owolabi et al., 2016; Seccombe et al., 2011). Additionally, clinical reports describe a predisposition to development of lung infections by PD patients, and to embolism and allergic reactions, compared with the general population (Hampson, Kieburtz, LeWitt, Leinonen, & Freed, 2017; Letsiou et al., 2017; Owolabi et al., 2016). Furthermore, areas in the brain responsible for the operation of swallowing and coughing movements can be downregulated in PD; therefore, lung infections in those patients could be associated with dysphagia (Hinkel et al., 2016; Schiffer & Kendall, 2018; van Hooren et al., 2016).

Parkinson's disease is a prevalent, chronic and progressive neurodegenerative motor disorder, affecting 1% of the population over the age of 60 years (Blandini, 2013). Clinical investigations report that PD patients also present with non-motor symptoms, such as depression, sleep disturbances, impairment of the cardiovascular reflexes and breathing abnormalities (Blandini, Levandis, Bazzini, Nappi, & Armentero, 2007; Chaudhuri, Odin, Antonini, & Martinez-Martin, 2011; Lozano & Kalia, 2005; San Luciano et al., 2010). Moreover, PD is diagnosed based on the presence of one or more of the following motor symptoms: bradykinesia, postural instability and impaired coordination, body rigidity and tremor of the face and limbs (Seidel et al., 2015; Stefani et al., 2013). Several pieces of evidence suggest a massive degeneration in areas of the brainstem that are involved in control of breathing, mainly the preBötC and the RTN (Benarroch et al., 2003; Tuppy et al., 2015). Furthermore, it was recently proposed that neuronal loss could be linked to breathing abnormalities, such as bradypnoea at rest and impaired response to central chemoreflex activation (Oliveira, Tuppy, Moreira, & Takakura, 2017; Tuppy et al., 2015).

The purpose of this study was to determine whether it was possible to elicit breathing abnormalities in mouse model of PD, similar to those observed in a rat model of PD. We evaluated the following: (i) the expression of neurokinin-1 receptor (NK1r) in the preBötC and RTN and of the transcription factor Phox2b in the RTN in a mouse model of PD; (ii) the ventilation rate in alert PD-induced mice, by whole-body plethysmography; (iii) the respiratory mechanics in the closed and opened thorax of PD-induced mice; and (iv) the lung histoarchitecture in a mouse model of PD, compared with control animals.


2.1 Ethical approval

All experimental and surgical procedures conformed to the US National Institutes of Health and were approved by the Institutional Animal Care and Use Committee at the University of São Paulo (protocol no.: 58/2017). Breeding stock was acquired from The Jackson Laboratory (stock no. 000664; Sacramento, CA, USA), and pups were generated at Department of Pharmacology Vivarium at the University of São Paulo, Brazil.

2.2 Surgery and anaesthesia

Experiments were performed on 15 adult male C57BL/6 mice (weight 20–25 g, age 2 months; control, n = 7 and PD, n = 8). The animals were given free access to water and food and were housed in a temperature-controlled chamber at 24°C, with a 12 h–12 h light–dark cycle.

For selective chemical lesions, mice were fixed in a stereotaxic frame (Kopf model 1760). One injection per side of 6-OHDA (6-hydroxydopamine hydrochloride; Sigma, Saint Louis, MO, USA; 10 μg μl−1; 0.5 μl; bilaterally into the striatum) or vehicle (1 μg ascorbic acid in 1 μl of 0.9% saline) was made while the mice were under general anaesthesia induced by 5% isoflurane inhaled in 100% oxygen. We used the following coordinates in order to reach the region of the striatum: 0.5 mm rostral to bregma, 1.8–1.9 mm lateral to the mid-line and 3.5 mm below the surface of the skull. The injections were performed using pipettes with an external tip coupled to a Hamilton syringe. The 6-OHDA is taken into the neurons by reuptake by dopaminergic transporters in the striatum. Then, the neurotoxin is retrogradely transported and selectively kills dopaminergic neurons of the substantia nigra pars compacta (SNpc) by inhibition of mitochondrial complexes I and IV and formation of reactive oxygen species (Glinka, Tipton, & Youdim, 1996; Hernandes et al., 2013). The dose of 6-OHDA used in the present study was selected based on previous reports (Blandini, Armentero, & Martignoni, 2008; Fulceri et al., 2006; Lima, Oliveira, Botelho, Moreira, & Takakura, 2018; Oliveira et al., 2017; Schwarting & Huston, 1996). The 6-OHDA injections were performed bilaterally in the striatum to induce a retrograde neuronal nigrostriatal injury (Blum et al., 2001). For physiological experiments, the animals were recorded before and during 40 days after administration of 6-OHDA or vehicle into the striatum (Figure 1b). Surgery was performed using standard aseptic methods. After the surgery, the mice were treated with the antibiotic ampicillin (100 mg kg−1 i.m.) and the analgesic ketorolac (0.6 mg kg−1 s.c.). The toxin did not produce observable behavioural effects.

Details are in the caption following the image
Characterization of Parkinson's disease (PD) mouse model and degeneration of important areas involved in respiratory control. (a) Experimental time course. (b) Whole-body plethysmography scheme. (c) Respiratory mechanics scheme. (d–f) Photomicrographs showing a reduction of catecholaminergic neurons in the SNpc (d) and no changes in the C1 region (e) and LC (f) of PD mice compared with control animals. (g–i) Mean of catecholaminergic neurons in SNpc (g), C1 region (h) and LC (i) in control or PD-induced animals. *Different from vehicle (one-way ANOVA followed by Newman–Keuls multiple comparisons test, P < 0.05). Scale bars: = 250 μm (d1 and e1; applies to d–f). Abbreviations: LC, locus coeruleus; SNpc, substantia nigra pars compacta; 4V, fourth ventricle; CPu, Caudate/Putamen

2.3 Ventilation

Measurements of respiratory rate (fR; in breaths per minute), tidal volume (VT; in microlitres per gram), expiratory time (tE; in milliseconds) and inspiratory time (tI; in milliseconds) were recorded by whole-body plethysmography (Buxco, Sharon, Connecticut, CT, USA; Figure 1b). Ventilation (urn:x-wiley:09580670:media:eph12452:eph12452-math-0001; in millilitres per minute per gram) was calculated as the product of fR and VT. Forty millilitres animal chambers were ventilated continuously with a mixture of 79% nitrogen and 21% oxygen (unless otherwise required by the protocol) at a rate of 500 ml min−1. Volume calibration was performed during each ventilation measurement throughout the course of the experiments by injecting a known volume of air (1 ml) inside the chamber. All experiments were performed at room temperature (21–23°C). Unanaethetized mice were allowed ≥2 h to acclimate to the chamber environment at normocapnia (21% O2, 79% N2 and <0.5% CO2) before measurements of baseline were taken. Hypercapnia was induced by titrating CO2 into the respiratory mixture up to a level of 7% (21% O2, 72% N2). Hypoxia was induced by continuously flushing with a mixture of 79% N2 and 21% O2 for 20 min. Recordings were taken at 1 min intervals in baseline conditions and 20 min after exposure to hypercapnia or hypoxia.

2.4 Measurement of airway responsiveness and respiratory mechanics

Animals were anaesthetized with an i.p. injection of ketamine (120 mg kg−1; Syntec, Brazil) and xylazine (12 mg kg−1; Ceva, Brazil). The trachea was cannulated with an 18-gauge metallic cannula (12.7 mm length), and the jugular vein was cannulated for later injection of acetyl-β-methyl-choline chloride (methacholine, MCh; Sigma-Aldrich, St Louis, MO, USA), a muscarinic agonist. Mice were artificially ventilated with a positive end-expiratory pressure of 3 cmH2O, at a frequency of 150 breaths min−1 and a tidal volume of 10 mL kg−1, using a mouse ventilator (flexiVent; SCIREQ, Montreal, Quebec, Canada; Figure 1c). Neuromuscular block was created by i.p. injection of pancuronium bromide (1 mg kg−1; Cristália, Brazil). After 7 min, a 6 s recruitment manoeuvre (pressure ramp until 30 cmH2O over 3 s, followed by a 3 s end-inspiratory pause) was performed twice during the default ventilation. A 3 s multifrequency (1–20.5 Hz) volume perturbation was applied to the lungs for the measurement of respiratory input impedance. From this, we obtained the parameters of airway resistance (RAW), tissue viscance (G) and tissue elastance (H). These parameters were first measured after the injection of 20 μl (10 g body weight) −1 of PBS, the diluent for MCh, via the jugular vein, and MCh 300 μg kg−1 in a volume of 20 μl (10 g body weight) −1. Next, the thorax was opened, and the same doses of PBS and MCh were injected again. From the data collected, the mean points after the PBS injection and the peak response after the injection of MCh 300 μg kg−1 were selected.

2.5 Brain histology, cell counting, imaging and data analysis

After the assessment of respiratory mechanics, the animals were deeply anaesthetized with 60 mg kg−1 pentobarbital i.p. and injected with heparin (50 units, transcardially). Finally, they were perfused through the ascending aorta and pulmonary artery with 50 ml of PBS (pH 7.4), followed by 4% phosphate-buffered (0.1 m; pH 7.4; 100 ml) paraformaldehyde (Electron Microscopy Sciences, Fort Washington, PA, USA). The brain was then removed and stored in the perfusion fixative for 24–48 h at 4°C. A series of coronal sections (40 μm) were stored in cryoprotectant solution at −20°C (20% glycerol plus 30% ethylene glycol in 50 ml phosphate buffer, pH 7.4) for later histological processing. All histochemical procedures were done using free-floating sections, in accordance with previously described protocols (Stornetta et al., 2006; Takakura et al., 2006, 2008).

Using an immunofluorescence technique, Phox2b immunoreactivity was counted after detection using a rabbit anti-Phox2b antibody (1:800; gift of J. F. Brunet, Ecole Normale Superieure, Paris, France). Sections were incubated for 24 h at room temperature (22–24°C) and diluted in PBS containing 10% normal horse serum (26050; Jackson ImmunoResearch Laboratories) and 0.3% Triton X-100. Sections were then rinsed in phosphate buffer and incubated for 2 h in Alexa 594 donkey anti-rabbit antibody (711-585-152; 1:200; Jackson ImmunoResearch Laboratories). Using the immunoperoxidase technique, tyrosine hydroxylase (TH) was counted after detection using a mouse anti-TH antibody (MAB318; 1:1000; Millipore, Temecula, CA, USA) and NK1r immunoreactivity (NK1r-ir) was densitometrically assessed after detection using a mouse anti-NK1r antibody (39-6100; 1:10,000; Invitrogen). Sections were incubated for 24 h at room temperature and diluted in PBS containing 10% normal horse serum (26050; Jackson ImmunoResearch Laboratories) and 0.3% Triton X-100. After three rinses, the sections were transferred to the appropriate affinity-purified biotinylated secondary antibody, donkey anti-mouse biotin (715-065-151; 1:500; Jackson ImmunoResearch Laboratories) for TH and NK1, diluted in PBS containing 1% normal donkey serum and 0.3% Triton X-100, incubated for 24 h at room temperature, rinsed again and exposed to Extravidin (Sigma-Aldrich; 1:1500) for 4 h at room temperature. Peroxidase reactions were visualized using the glucose oxidase procedure and 3,3′-diaminobenzidine (DAB) tetrahydrochloride as chromogen, and the enzymatic reaction was initiated with β-d-glucose.

All the sections were mounted onto slides in rostrocaudal sequential order, dried, and covered with DPX (Aldrich, Milwaukee, WI, USA). Coverslips were affixed with nail polish.

A conventional multifunction microscope (Nikon, Japan) was used to image sections and perform subsequent analysis (Barna, Takakura, & Moreira, 2012; Silva, Takakura, & Moreira, 2016; Stornetta et al., 2006; Takakura, Barna, Cruz, Colombari, & Moreira, 2014). A one-in-four series of 30-μm-thick brain sections were used per animal, causing each section analysed to be 120 μm apart. The sections were counted bilaterally, and the numbers reported in the Results section correspond exactly to the counts of one-in-four sections in a series. Photographs were taken with a colour camera (DS-Fi3; Nikon, Japan). An image analysis software package (ImageJ, v.1.41; US National Institutes of Health, Bethesda, MD, USA) was used for cell counting, and a technical illustration software package (Canvas, v.9.0; ACD Systems, Miami, FL, USA) was used for line drawings, assembly of figures and labelling. The neuroanatomical nomenclature used was defined by Paxinos and Franklin (2012).

2.6 Lung histology, cell counting, imaging and data analysis

Lung tissue was placed in paraformaldehyde, then embedded in paraffin. The tissue samples were cut using a microtome (5 μm thickness) and were stained with Haematoxylin and Eosin (H&E), using the Picrosirius and Orcein methods. Lung morphology from 10 non-coincident areas (H&E staining) were imaged at a magnification of ×200. To determine the pulmonary structure, we applied the point-counting technique, using an integrating eyepiece with a coherent system, composed of a 100-point grid consisting of 50 lines of known length. Collagen and elastic fibre samples were analysed after Picrosirius polarization and Orcein stain methods, respectively. Densitometric analysis was used to quantify collagen and elastic fibres in the parenchyma, airway and blood vessels. Images were captured using a colour camera (DS-Fi3; Nikon, Japan). An image analysis software package (ImageJ, v.1.41) was used for cell counting, along with a technical illustration software package (Canvas, v.9.0) for creating line drawings.

2.7 Statistics

Data are reported as means ± SD. Statistical analysis was performed using SigmaStat v.11.0 (Jandel Corporation, Point Richmond, CA, USA). All the data passed the normality test (Shapiro–Wilk test). One-way ANOVA followed by the Student–Newman–Keuls multiple comparisons test or two-way ANOVA followed by Bonferroni post hoc test were used as appropriate. For each analysis, P < 0.05 was considered statistically significant.


3.1 Neural dysfunction in a mouse model of PD

The initial series of experiments were designed to identify any neuroanatomical differences between this experimental mouse model of PD, compared with control animals. To evaluate the effects of toxin injection, TH immunoreactivity (TH+) was examined within the SNpc in control or PD-induced mice. We found that 6-OHDA reduced the number of TH+ neurons in the SNpc 40 days after the toxin injection (60.3 ± 19.3 TH+ neurons versus control, 198.4 ± 36.5 TH+ neurons; P < 0.001; one-way ANOVA), as represented in Figure 1d and g. To assess the specificity of toxin injection, TH+ was also detected in the C1 and locus coeruleus (LC) regions in control and PD mice. However, our results showed that the number of neurons in each region did not experience notable change in PD mice, when compared with the control animals (Figure 1e, f, h and i).

In the brainstems of the adult PD-induced or control mice, at the rostr-caudal level of the preBötC (bregma −6.74 mm) ventral to the semi-compact nucleus ambiguus (scAmb), the immunoreactivity for NK1 receptors was reduced in PD animals (28.4 ± 6.4 versus control, 41.1 ± 2.8% expression; P = 0.02; one-way ANOVA, n = 4; Figure 2a2–d2). No changes were observed in the density of NK1 receptors in the BötC, rostral and caudal ventral respiratory groups (Figure 2a1–d1, a3–d3 and a4–d4). In other sections, Phox2b was detected in the rostrocaudal axis of the RTN (bregma −5.77 to −7.00 mm), which indicated a significant reduction in PD animals (37.8 ± 9.4 versus control, 61.3 ± 7.1 Phox2b+ neurons; P = 0.009; one-way ANOVA, n = 4; Figure 3a–c and f). At the same RTN bregma level and axis, we evaluated the NK1r expression and, as represented in Figure 3d, e and g, the NK1r expression in PD animals was reduced (34.4 ± 6.3 versus control, 56.8 ± 9.9% expression; P = 0.009; one-way ANOVA, n = 4).

Details are in the caption following the image
Neuroanatomical impairment in PD mice. (a) Schematic representation of the brainstem areas analysed (bregma level −6.64 to −8.00 mm). (b–d) Representative photomicrographs from NK1r-ir in control (b) and PD animals (c) and mean of NK1r-ir (d) in the Bötzinger complex (1), pre-Bötzinger complex (2), rostral ventral respiratory group (rVRG; 3) and caudal ventral respiratory group (cVRG; 4). *Different from vehicle (one-way ANOVA followed by Newman–Keuls multiple comparisons test, P < 0.05). Scale bar: 80 μm (b4; applies to b and c). Abbreviations: Amb, nucleus ambiguus; AP, area postrema; cAmb, compact nucleus ambiguus; cc, central canal; Cu, cuneate nucleus; DMV, dorsal motor nucleus of vagus; Gr, gracile nucleus; IO, inferior olives; Li, lineares; py, pyramidal tract; scAmb, semi-compact nucleus ambiguus; Sp5, spinal trigeminal tract; vsc, ventral spinocerebellar tract; XII, hypoglossal nucleus; CPu, Caudate/Putamen
Details are in the caption following the image
Reduction of Phox2b neurons and neurokin-1 receptor immunoreactivity (NK1r-ir) in the retrotrapezoid nucleus (RTN) in PD mice. (a) Schematic representation of the RTN area analysed (bregma level −6.36 mm). (b, c) Representative photomicrographs from Phox2b immunoreactivity (Phox2b-ir) in control (b) and PD animals (c) in the RTN. (d, e) Representative photomicrographs from NK1r-ir in control (d) and PD animals (e) in the RTN. (f, g) Number of Phox2b neurons (f) and NK1r-ir (g) in the RTN of control or PD mice. *Different from vehicle (one-way ANOVA followed by Newman–Keuls multiple comparisons test, P < 0.05). Scale bars: 500 μm (a), 80 μm (c; applies to b and c), 20 μm (b1; applies to b1–c1), 80 μm (e; applies to d and e) and 20 μm (e1; applies to d1 and e1). Abbreviations: py, pyramidal tract; Sp5, spinal trigeminal tract; VII, facial nucleus; VS, ventral surface; 4V, 4th ventricle; CPu, Caudate/Putamen

Physiological and anatomical experiments were conducted on the same animals. The PD and control mice were monitored for changes in breathing at rest and during activation of the central or peripheral chemoreceptors (hypercapnia, 7% CO2 and hypoxia, 8% O2), using the whole-body plethysmograph respiratory method. We observed that PD-induced mice presented a reduction of resting fR (156.3 ± 9.7 breaths min−1 versus control, 176.1 ± 9.8 breaths min−1; P = 0.001; two-way ANOVA, n = 7) and urn:x-wiley:09580670:media:eph12452:eph12452-math-0002 (0.53 ± 0.08 ml min−1 versus control, 0.65 ± 0.10 ml min−1; P = 0.018; two-way ANOVA, n = 7) and an increase of expiratory time (0.36 ± 0.05 ms versus control, 0.26 ± 0.04 ms; P = 0.002; two-way ANOVA; Figure 4a–f). No changes were identified in resting VT and tI (Figure 4c and e). Additionally, we observed a higher fR irregularity in PD-induced mice (0.11 ± 0.06 breaths min−1 versus control, 0.030 ± 0.03 breaths min−1; P = 0.005; two-way ANOVA; Figure 4g).

Details are in the caption following the image
Breathing abnormalities in PD mice. (a) Representative respiratory recordings from control and PD mice in room air (before and 40 days after 6-OHDA or vehicle injections) and during hypercapnia (7% CO2) and hypoxia (8% O2) at 40 days after 6-OHDA or vehicle injections. (b–g) Changes in respiratory frequency (fR; b), tidal volume (VT; c), ventilation (urn:x-wiley:09580670:media:eph12452:eph12452-math-0003; d), inspiratory time (tI; e), expiratory time (tE; f) and the fR irregularity score at room air (g). (h–m) Changes in fR (h), VT (i) and urn:x-wiley:09580670:media:eph12452:eph12452-math-0004 (j) during hypercapnia and in fR (k), VT (l) and urn:x-wiley:09580670:media:eph12452:eph12452-math-0005 (m) during hypoxia in control and PD mice. *Different from vehicle (two-way ANOVA followed by Bonferroni post hoc test, P < 0.05)

The results of our experiments corroborated the findings of previous studies; the hypercapnic ventilatory response in PD animals was the only response that experienced impairment. Figure 4h shows that tachypnoeic response to CO2 was reduced in PD animals (369.6 ± 19.8 breaths min−1 versus control, 418.2 ± 30.0 breaths min−1; P = 0.006; two-way ANOVA; Figure 4h), whereas Figure 4k–m shows that the hypoxic ventilatory response remained consistent between the PD and control animals.

3.2 Peripheral breathing dysfunction in a mouse model of PD

To evaluate the lung properties, assessment of respiratory mechanics was performed 40 days after induction of PD, both before and after opening of the thorax. The baseline data showed increased tissue viscance (G) in PD-induced animals, when the samples were collected before opening the thorax (4.9 ± 0.9 cmH2O ml−1 versus control, 4.1 ± 0.6 cmH2O ml−1; P = 0.046; one-way ANOVA; Figure 5c). Once the thorax had been opened, the tissue viscance was reduced in both the PD and the control groups (Figure 5c). Additionally, we observed an increase of airway resistance (RAW) in the PD-induced group (0.33 ± 0.04 cmH2O s ml−1 versus control, 0.27 ± 0.2 cmH2O s ml−1; P = 0.01; one-way ANOVA; Figure 5a), only before the thorax had been opened. For PD animals, we observed an increase of tissue elastance (H) in the closed (31.3 ± 5.5 cmH2O ml−1 versus control, 25.4 ± 1.6 cmH2O ml−1; P = 0.019; one-way ANOVA) and opened thorax configurations (30.8 ± 5.4 cmH2O ml−1 versus control, 24.0 ± 3.7 cmH2O ml−1; P = 0.017; one-way ANOVA; Figure 5b), compared with control animals.

Details are in the caption following the image
Evidence of dysfunction of lung properties and architecture in PD mice. (a–c) Changes in airway resistance (RAW; a), tissue elastance (H; b) and tissue viscance (G; c) while the thorax had been preserved or after thoracotomy in control and PD mice. *Different from vehicle, different from closed thorax (Vehicle); different from closed thorax (6-OHDA) (one-way ANOVA followed by Newman–Keuls multiple comparisons test, P < 0.05). (d–f) Methacholine (MCh) challenges evaluated with the thorax opened for airway resistance (d), tissue elastance (e) and tissue viscance (f) during i.v. infusion of saline and methacholine in control and PD mice. *Different from vehicle (two-way ANOVA followed by Bonferroni post hoc test, P < 0.05). (g–n) Representative photomicrographs from the lung parenchyma of control and PD mice stained with Haematoxylin and Eosin (; g, j), Picrosirius (k, l) and Orcein (m, n). Scale bars: 100 μm (g; applies to g and k), 25 μm (g1; applies to g1 and k1), 100 μm (i and j; applies to i, j, m and n) and 25 μm (i1 and j1; applies to i1, j1, m1 and n1). Abbreviation: BV, blood vessels; CPu, Caudate/Putamen

The MCh challenge is illustrated in Figure 5d–f and shows that only those animals in the opened thorax configuration responded to MCh (300 μg kg−1, i.v.). We did not observe any changes in the results of the MCh challenge between our model of PD compared with the control animals.

In the same group of animals (n = 4), we quantified morphological parameters in PD and control mice. As shown in Table 1, we did not find any changes in histoarchitecture when comparing PD versus control animals. However, we did notice an increase in collagen fibre deposits in the alveolar septa and in the airways of the PD-induced animals (Figure 5k–n), compared with the control animals (Figure 5g–j; Table 1).

Table 1. Lung histology
Contituent (%) Sham 6-OHDA P value
Normal area 90.0 ± 3.5 90.3 ± 3.6 0.962
Alveolar collapse 7.5 ± 0.6 7.3 ± 0.3 0.781
Collagen fibres in alveolar septa 12.5 ± 1.4 18.3 ± 1.5* 0.032
Elastic fibres in alveolar septa 11.8 ± 2.0 13.5 ± 1.5 0.517
Collagen fibres in airway 14.5 ± 1.5 21.8 ± 1.2* 0.011
Elastic fibres in airway 11.3 ± 1.7 12.3 ± 1.8 0.711
Collagen fibres in blood vessels 19.8 ± 1.7 22.0 ± 1.9 0.408
Elastic fibres in blood vessels 40.3 ± 2.2 39.3 ± 2.8 0.790
  • Percentage of normal area, alveolar collapse and elastic and collagen fibres in alveolar septa, airways and blood vessels. Values are means ± SEM. All values were computed in 10 random, non-coincident fields per mouse. *Different from vehicle (Student's unpaired t test, P < 0.05; n = 4). Abbreviation: 6-OHDA, 6-hydroxydopamine.


We believe that these results provide confirmation of breathing disturbances at rest and during central chemoreflex challenge in our mouse model of PD. Furthermore, our findings show, for the first time, that PD-induced mice experience lung dysfunctions and histological changes. Taking all these findings into consideration, we believe these results provide significant insight into respiratory abnormalities experienced in this animal model of PD, and it is our hope that these findings can be leveraged to help with clinical assistance of patients with PD.

4.1 Neural control of breathing in animals with induced PD

Parkinson's disease is the second most common neurodegenerative disease, after Alzheimer's disease, affecting ∼2–4% of the general population between the ages of 60 and 80 years (Blandini, 2013). Parkinson's disease is characterized by the chronic and progressive death of dopaminergic neurons in regions of the brain responsible for fine motor adjustments (Lees, Hardy, & Revesz, 2009; Seidel et al., 2015). Clinically, PD patients can develop breathing abnormalities, in addition to non-motor symptoms, such as sleep disorders, constipation, autonomic failure, depression and cognitive dysfunction (Bernal-Pacheco, Limotai, Go, & Fernandez, 2012; Chaudhuri et al., 2011; Raggi, Bella, Pennisi, Neri, & Ferri, 2013).

Breathing in mammals is controlled by the integrated network of excitatory and inhibitory neurons in the ventral portion of the medulla, known as the preBötC. These neurons are responsible for the generation of breathing (eupnoea), gasps and sighs (Janczewski, Tashima, Hsu, Cui, & Feldman, 2013; Kam et al., 2013; Zanella et al., 2014). In previous studies performed on rats, a reduction of NK1 receptor expression was observed in the preBötC, with a strong correlation with the reduction of the resting fR in PD animals (Benarroch et al., 2003; Tuppy et al., 2015). From the data collected in our mouse model of PD, we also observed a reduction in the density of NK1 receptors in the preBötC and a reduction in resting fR.

Additionally, data from our experiments identified another important cluster of neurons, located within the ventral medullary surface (the so-called retrotrapezoid nucleus) and involved in central chemoreception, that was also affected in PD-induced rats and mice. This region is composed of glutamatergic neurons (that do not express tyrosine hydroxylase and choline acetyltransferase) that express the transcription factor Phox2b (Guyenet & Bayliss, 2015; Ruffault et al., 2015; Silva, Tanabe, Moreira, & Takakura, 2016; Takakura et al., 2006, 2014). In rat models of PD, the impaired respiratory hypercapnic response observed also showed a strong correlation with the reduction of Phox2b+ neurons in the RTN (Fernandes-Junior, Carvalho, Moreira, & Takakura, 2018).

Previous studies have shown that in the rodent model of PD induced by 6-OHDA, there were no anatomical changes of catecholaminergic neurons in the LC (Farrand et al., 2017; Ostock, Lindenbach, Goldenberg, Kampton, & Bishop, 2014; Shin et al., 2014; Tuppy et al., 2015), which was confirmed by the present study. Previous findings also demonstrated in a rat model of PD that central chemoreception is initially reduced, but then recovers during the course of PD (Oliveira et al., 2017). One explanation for that recovery is the fact that the LC is intact and responsible for the activation of breathing during hypercapnia (Oliveira et al., 2017).

Although we have observed a reduction in the number of neurons or receptors in crucial regions involved in the neural control of breathing in this mouse model of PD, the underlying mechanisms are still unknown. It is possible that SNpc dopaminergic neurons project indirectly to ventral respiratory column (VRC) through the periaqueductal grey matter, as already demonstrated in rats (Lima et al., 2018). Other possibilities are the presence of a dysfunctional blood–brain barrier, inflammation or an increase of oxidative stress in this model that could lead to death of brainstem neurons or loss of their markers. However, further studies are necessary to clarify these mechanisms.

The neuroanatomical and physiological changes observed in both the mouse model and the rat model of PD were very similar overall. The primary difference observed between the two models was the timetable of when the changes first become noticeable. In rats, the respiratory changes started 40 days after the induction of PD, whereas in mice we began to notice respiratory impairment 10 days after the injection of 6-OHDA into the striatum.

4.2 Impairment of lung properties in PD mice

The lungs are one of the most sensitive organs in the body; however, few studies have described the negative changes and impact on the lungs during the development of PD (Baille et al., 2018; Seccombe et al., 2011; Torsney & Forsyth, 2017).

Parkinson's disease patients can experience an impairment in the tone, contractility and coordination of the thoracic musculature, affecting both respiratory mechanics and pulmonary functions (Baille et al., 2018; Owolabi et al., 2016; Reyes, Ziman, & Nosaka, 2013). Therefore, studies that evaluate the respiratory mechanics in a model of PD might lead to greater understanding of the effects of the disease. We applied the constant phase model to analyse the respiratory mechanics in proximal and distal portions of respiratory system (Hantos, Adamicza, Govaerts, & Daróczy, 1992).

In our PD model, we verified an increase in G, a parameter related to the energy dissipation in the distal portion of the respiratory system (Hantos et al., 1992). Recent studies have provided evidence of changes in the lung volumes experienced by PD patients, such as airflow limitation and ventilatory muscle weakness. In one of the recent studies, by Baille et al. (2018), it was observed that PD patients exhibited inspiratory muscle weakness during the early stages of the disease. Thus, the mechanical changes observed in our study might also be correlated with stiffening of the lung and ribcage, which have also been observed in these previous studies. Therefore, we evaluated the respiratory mechanics without the presence of ribcage effects in order to gain a better understanding of the changes in the respiratory mechanics of our PD model. In the closed thorax, we observed that not only G, but also RAW and the tissue elastance (H) increased in PD. However, in the open thorax, only H was increased in PD. The data collected suggest, for the first time, that stiffening in the ribcage could have an effect on the distension and relaxation that occurs during a normal breath in PD animals (Baille et al., 2016; Hampson et al., 2017). Additionally, we verified abnormal collagen fibre deposition in alveolar septa and in airways in PD-induced animals. These findings suggest that PD-induced animals might experience stiffening of both the ribcage and lung tissue. It seems that disturbances in alveolar septa and airways were not sufficient to induce changes in respiratory mechanics, because the increase in RAW and G disappeared when the thorax was opened.

Parkinson's disease patients present with muscle stiffening and rigidity during the late stages of the disease (Baille et al., 2018; LeWitt et al., 2018; Owolabi et al., 2016; Seccombe et al., 2011). However, no physiological or molecular mechanisms have been described yet. In our experiments, we could also observe such changes in an experimental model of PD. If that is the case, new approaches should be performed to investigate the effects promoted by neurodegeneration at SNpc or brainstem levels and correlate them with the organ dysfunction.

There are several possible explanations for the correlation of PD and the high risk of lung diseases. Our experiments showed similar responsiveness to methacholine challenge between the groups. Furthermore, several patterns of abnormalities have been demonstrated for symptoms such as restrictive pulmonary dysfunction, obstructive airflow disease and upper airway obstruction. We believe that future studies are needed to investigate any connection between neural degeneration and peripheral distress.


Our research and evidence from previous studies suggest that neurological pathologies could affect peripheral tissues in several ways. Previous studies have already demonstrated that during the development of PD, the network responsible for rhythm generation could become impaired. In addition, based on our findings, we hypothesize that during the development of PD, the lungs and the ribcage can become stiff, which could contribute to impaired inspiration and expiration during breathing.


We thank Dr L. V. Rossoni (University of São Paulo) for providing the microscope apparatus, Dr. R. Soncini (Federal University of Alfenas) for staining support and Dr W. T. Lima for the mechanical ventilation apparatus. We also thank Liza Severs and Brandon Knapp for carefully reviewing and revising the manuscript.


    L.M.O., T.S.M. and A.C.T. designed the experiments. L.M.O., M.A.O. and H.T.M. collected and analysed the data. All the authors wrote the manuscript, approved the final version and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.


      None declared.