Article Text
Abstract
Background: A significant number of patients with chronic obstructive pulmonary disease (COPD) exhibit skeletal muscle wasting and decreased capillary area formation, which correlate with increased mortality.
Aim: To determine the molecular mechanisms mediating decreased capillary formation in COPD.
Methods: 24 patients with COPD and 12 matching controls were recruited. Patients with COPD were classified into mild, moderate and severe groups according to GOLD (global initiative for chronic obstructive lung disease) criteria. Biopsy specimens were obtained from the tibialis anterior muscle. Fibre typing and capillary formation, together with messenger RNA (mRNA) expression of hypoxia-inducible factors (HIF1α and HIF3α), vascular endothelial growth factors (VEGF-A, VEGF-B and VEGF-C isoforms) and von Hippel-Lindau (VHL) protein, were determined. VHL expression and localisation were further studied by immunohistochemistry.
Results: Skeletal muscle capillary formation decreased significantly with increasing disease severity. Compared with controls, a tendency to mRNA overexpression of HIF1α, HIF3α and VEGF isoforms was observed in mild and moderate COPD, which decreased at the severe stage. In contrast, skeletal muscle biopsy samples from patients with COPD exhibited significant overexpression of VHL at both the mRNA and protein level by immunohistochemistry. VHL protein was further determined to be localised to satellite cells.
Conclusions: Overexpression of VHL was identified in the skeletal muscle of patients with COPD. Increased VHL activity may have a negative effect on transduction of the hypoxic signal and may contribute to decreased capillarisation in skeletal muscles of patients with COPD.
Statistics from Altmetric.com
A significant number of patients with chronic obstructive pulmonary disease (COPD) develop extrapulmonary complications in the form of generalised weight loss and skeletal muscle wasting (collectively termed “pulmonary cachexia syndrome” (PCS)), which ultimately lead to aggravated dyspnoea, decreased exercise tolerance and increased sense of fatigue.1–3 PCS development has been shown to be independent of the degree of airflow obstruction and to correlate directly with increased mortality.2 The mechanisms leading to PCS development are still largely unknown.
Progressive pulmonary airflow limitation in COPD, in the absence of resting hypoxaemia, results in frequent transient periods of oxygen desaturation, which are considered a daily feature of the disease and can occur on performing such common tasks as walking, eating and washing.3–5 In addition, hypoxaemia, which necessitates oxygen therapy, is common during acute exacerbation of COPD and with disease progress. Peripheral skeletal muscles in patients with COPD not only exhibit wasting,6 but also changes attributed to the presence of hypoxia. These include: (1) a shift in muscle fibre type towards an increased proportion of the glycolytic fatigue-susceptible type II fibres relative to the aerobic less fatigue-susceptible type I fibres7 8; (2) altered mitochondrial enzyme activity7; (3) decreased and delayed skeletal muscle Vo2 uptake9; (4) significantly decreased muscle capillarisation including the number of capillaries/mm2 as well as the ratio of capillary/fibre.10 11
Tight regulation of systemic and intracellular O2 levels as well as the presence of an intact capillary network are essential for maintenance of normal cellular functions. The principal intracellular O2 level sensor is the widely expressed hypoxia-inducible factor (HIF), of which to date three isoforms (HIF1α, HIF2α and HIF3α) are known.12 13 Intracellular HIF concentrations are tightly regulated by the von Hippel-Lindau tumour-suppressor protein (pVHL), a product of the VHL gene and a component of the E3 ubiquitin ligase system.14 15 In response to hypoxia, intracellular HIF concentrations are stabilised and increase as the result of decreased degradation.16 HIF signal is then transduced to regulate the transcription of a multitude of gene responses16 including vascular endothelial growth factor (VEGF), which is central to the initiation of the angiogenic signal and neovascular formation.17 In skeletal muscles, reparative responses to damage and angiogenesis are undertaken by satellite cells occupying the stem cell niche on the periphery of skeletal muscle cells.18 Satellite cells have been shown to express HIF1α, VEGF and its receptors in human, cell and animal models.19–21
The aim of this study was to investigate our hypothesis that impaired coupling of the hypoxic–angiogenic signalling cascades mediates the decreased skeletal muscle capillarisation in patients with COPD.
METHODS
Subjects
The study was approved by the Uppsala University ethics board. Twenty-four patients with COPD (10 male and 14 female) were selected in a stable condition with no respiratory tract infections or exacerbation of their disease for at least 4 weeks before the study. All recruited patients were former smokers. Twelve (6 male and 6 female) healthy, non-smoking subjects matched for age and sex were recruited as controls. Exclusion criteria were known malignancy, cardiac failure, distal arteriopathy, recent surgery and severe endocrine, hepatic or renal disorders. Patients with COPD who were receiving a maintenance dose of oral steroids were not recruited to this study to avoid the known deleterious effects of steroid treatment on skeletal muscle morphology.
All participants underwent spirometry with reversibility testing. Patients with COPD were classified into mild, moderate and severe groups according to GOLD (global initiative for chronic obstructive lung disease) criteria.22 Table 1 summarises the anthropometric characteristics of all subjects recruited to the study.
Height and length were measured, and body mass index was calculated as weight (kg)/length2 (cm). Fat free mass was determined using dual x-ray absorptiometry in a fasting state, and fat free mass index was calculated as fat free mass/length2. Arterial blood gases were determined in all recruited groups by arterial puncture of the radial artery.
Muscle biopsies
Samples were obtained from the bulk of the tibialis anterior muscle of the dominant leg under local anaesthesia using a semi-open needle biopsy as previously described.23 The samples were frozen in isopentane cooled to its freezing point in liquid nitrogen, and stored at −80°C until analyses were performed.
Immunohistochemistry
Serial transverse sections, 5 μm thick, were cut at −22°C using a Leica CM1850 cryostat attached to positively charged glass slides (Superfrost; Menzel Gläser, Braunschweig, Germany). Muscle fibre composition was determined by immunohistochemical staining using the monoclonal antibodies N2.261 and A4.951 (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, Iowa, USA) at dilutions of 1:100.24 With the N2.261 antibody, type IIa fibres are strongly stained, type IIx fibres are unstained, and type I and type IIxa fibres are slightly stained. With the A4.951 antibody, type I fibres are strongly stained, whereas type IIa, IIxa and type IIx fibres are unstained. Type I–IIa fibres are stained with both antibodies. Capillaries were identified using monoclonal antibodies for CD31 (MO823; Dako, Glostrup, Denmark; dilution 1:20), which recognises PECAM-1, a transmembranous glycoprotein that is strongly expressed in vascular endothelial cells.25 VHL was identified using monoclonal antibodies (GTX11189; GeneTex, Inc, San Antonio, Texas, USA) at dilutions of 1:25 and 1:50. The indirect two-step dextran-polymer technique (En Vision system; DakoCytomation, Glostrup, Denmark) was used with diaminobenzidine. Sections stained for CD31 and diaminobenzidine as chromogen were counterstained with eosin. Before application of the primary antibodies, the sections were incubated in buffer containing bovine serum albumin for 5 min at dilutions of 1:50 and 1:400. The number of VHL-immunoreactive cells were quantitatively scored. From each case, the number of VHL-positive cells were counted from six randomly chosen, non-overlapping, microscope fields at an objective magnification of ×40. Only cells containing visible nuclei were included.
The control stains entailed: (i) omission of the primary antibody; (ii) replacement of the first layer of antibody by non-immune serum diluted 1:10 and by the diluent alone; (iii) preincubation (24 h) of primary antibody with the relevant antigen (10 nmol/ml diluted antibody solution), respectively, before application to the sections.
Capillary morphology
Analyses were performed using a light microscope (Nikon Eclipse E400, Japan) connected to a computerised image system (SPOT Insight; Diagnostic Instrument, Sterling Heights, Michigan, USA). The cross-sectional area corresponding to a mean of 401 fibres for each subject was photographed at a magnitude of ×4. The photographed image was used to determine fibre type composition for the muscle sample. The fibres were designated type I, type IIa, type IIx, type IIx-a or type I–IIa (IIc). Four to ten randomly selected areas, corresponding to a mean of 80 fibres, were photographed at a magnitude of ×20. Fibre area and fibre perimeter for type I and type IIa fibres were determined from these photographs. For each subject, at least 30 fibres of each type were analysed. To make sure that a cross-sectional section of the muscle fibre was achieved for determination of the perimeter, the obliquity of fibre sectioning was assessed using the form factor that represents: (4p×fibre area)/(fibre perimeter)2.25 There were no differences between the form factors measured between the groups (control group, 0.80 (0.03); mild group, 0.78 (0.04); moderate group, 0.78 (0.03); severe group, 0.79 (0.02)). Capillaries were analysed from photographs with a magnitude of ×20 of four to ten randomly selected cross-sectional areas, corresponding to a mean of 80 fibres for each subject. The parameter measured for capillaries was the number of capillaries around a single fibre (CAF) and the CAF relative to the fibre area (CAFA).
RNA isolation
Muscle biopsy specimens were frozen in liquid nitrogen and disrupted in a Mikro-Dismembrator II (B Braun, Melsungen, Germany). The RNeasy Fibrous Tissue Mini Kit (Qiagen, Valencia, California, USA) was used to isolate total RNA according to the manufacturer’s instructions. The integrity and quantity of the isolated RNA was evaluated with the Agilent 2100 bioanalyser using the RNA 6000 Nano Assay Kit (Agilent Technologies, Santa Clara, California, USA). All isolated RNA was stored at −80°C until use.
TaqMan PCR
For gene expression analysis, 0.5 μg total RNA was diluted in RNase-free water to 20 μl for the synthesis of complementary DNA by reverse transcription using SuperScript II (Invitrogen, Carlsbad, California, USA) as described by the manufacturer. The final complementary DNA product was stored at −20°C until use. Genes included for expression analysis are HIF1α (Hs00936368_m1), HIF3α (Hs00541709_m1), VEGF (Hs00900054_m1), VEGFB/VEGFC (Hs 00) and VHL (Hs00184451_m1). All of the genes were designed and premixed by Applied Biosystems (Foster City, California, USA) provided as Assay-on-demand. Cyclophilin was used as the housekeeping gene for normalisation. Details for primers and probes for cyclophilin are: cyclophilin forward primer, 5′-TGCTGGACCCAACACAAATG-3′; cyclophilin reverse primer, 5′-TGCCATCCAACCACTCAGTC-3′; cyclophilin probe, 5′-TTCCCAGTTTTTCATCTGCACTGCCA-3′. The probes were labelled with FAM as reporter dye and TAMRA as quencher dye. ABI Prism 7700 Sequence Detection System (Applied Biosystems) was used for TaqMan PCR amplifications. During amplification, thermal cycler conditions were as follows: each sample was analysed twice in duplicate for 2 min at 50°C, 10 min at 95°C, 15 s at 95°C and 1 min at 60°C. The PCR amplification was correlated against a standard curve.
Statistical analysis
Capillary immunohistochemistry data are reported as median and range. Kruskal–Wallis one-way analysis of variance was performed, with group as the categorical value. When a significant difference was found, an all-pairwise comparison was performed to compare median values. Statistical analyses were performed using Statistix 8 (Analytic Software, Tallahassee, Florida, USA).
mRNA and quantitative analysis of VHL immunohistochemistry were analysed by regression analysis to estimate differences between groups, expressed as β coefficients with 95% CI, and determine statistically significant differences. Normality of data distribution was checked using the Shapiro–Wilk test. Significance was determined at p = 0.05.
RESULTS
Capillary area and morphology
Significant differences in CAF in both type I (p<0.006) and type IIa fibres (p<0.002) were observed, with the lowest CAF value in the groups with moderate and severe COPD (table 2). When adjusted for fibre area, CAFA did not differ significantly for either fibre type. A step-down decrease in skeletal muscle capillarisation was observed with increasing disease severity (fig 1).
Gene expression in COPD and control groups
A consistent trend towards overexpression of HIF1α (Fig 2A), HIF3α (Fig 2B), VEGF-A (Fig 3A), VEGF-B (Fig 3B) and VEGF-C (Fig 3C) was observed in mild and moderate COPD compared with controls. Statistical significance was achieved for VEGF-A, VEGF-B and HIF3 in patients with moderate COPD compared with controls (table 3). HIF2 mRNA was also analysed, but did not reach confident detection levels in our hands.
There was significant overexpression of VHL mRNA in mild COPD compared with controls (β = 11.9, 95% CI 3.3 to 20.6, p = 0.008; table 3 and fig 4). In addition, the mean level of VHL mRNA expression exhibited non-significant overexpression in moderate COPD versus controls (β = 7.1, 95% CI −0.8 to 15.1, p = 0.077) and similar levels to controls in severe COPD (β = −1.6, 95% CI −11.2 to 8.1, p = 0.741; table 3). Using specific pVHL antibodies and immunohistochemistry, a protein overexpression pattern that mirrored the pattern of mRNA overexpression was also observed (fig 5). Hence, the number of cells displaying pVHL immunoreactivity were significantly increased in patients with mild (20.2 (2.0) cells per microscope field; β = 10.8, 95% CI 8.9 to 12.7, p = 0.001; fig 5B) and moderate (14.5 (0.6); β = 5.1, 95% CI 3.1 to 7.1, p = 0.001; fig 5C) COPD compared with controls (9.4 (0.5); fig 5A). In patients with severe disease, the number of immunoreactive cells was lower than in the controls (7.5 (1.7); fig 5D). Expression of pVHL immunoreactivity was localised to the spindle-shaped satellite cells lying immediately beneath the fibre sarcolemma.
DISCUSSION
The results of this study provide the first evidence of VHL overexpression at both the mRNA and protein level in skeletal muscles of patients with COPD. Our data show identical patterns of expression of VHL mRNA and pVHL by immunohistochemistry. Thus, significant VHL overexpression was observed in mild and moderate COPD, with the highest concentrations found in mild COPD. VHL expression in severe COPD was not significantly different from controls. Increased pVHL expression was then further determined by immunohistochemistry to be localised to satellite cells in the stem cell niche.18 Satellite cells have been shown to be not only responsible for myogenesis, but also to actively induce angiogenesis by expressing26 and secreting27 VEGF. mRNA expression of the various signalling molecules included in this study, as well as pVHL by immunohistochemistry, showed increases in the mild and moderate stages of COPD and decreases in severe COPD. The mechanisms mediating the latter response are unclear. Hypoxia-induced responses have, however, been shown to be attenuated in the presence of chronic, prolonged and severe hypoxia,28 29 such as would occur in severe stages of COPD with its prolonged course and markedly decreased pulmonary functions.
Take-home messages
This study shows overexpression of von Hippel-Lindau (VHL) at both the mRNA and protein level in skeletal muscles of patients with chronic obstructive pulmonary disease.
Increased VHL activity may have a negative effect on tissue capillarisation by adversely affecting transduction of the hypoxic signal towards capillary formation.
It is unlikely that overexpression of VHL would lead to complete blockade of HIF signalling to VEGF, as this would lead to markedly decreased capillary formation, ischaemia and probably muscle necrosis, something that has not previously been observed in COPD. It is thus more likely that VEGF concentrations are maintained at levels that are inadequate to maintain normal capillary formation, hence the gradual decrease in tissue capillarisation with increasing disease formation, hence the gradual decrease in tissue capillarisation with increasing disease severity. The molecular mechanisms leading to enhanced VHL expression in skeletal muscle from patients with COPD are unclear. Our findings raise the interesting possibility that VHL activity may have increased as a secondary response to enhanced HIF activity. Indeed, it has previously been suggested that HIF and pVHL constitute a closed signalling loop, in which pVHL acts as a signalling check-point for HIF transcription.15 Furthermore, hypoxia has been shown to directly upregulate the endogenous concentrations of pVHL independently of its activity.30 However a primary cause cannot be excluded. This raises the intriguing possibility that some people may have increased VHL activity, which would adversely affect their response to a hypoxic insult such as would occur if they smoked leading to development of COPD. Alternatively, smoking may induce molecular changes in the VHL gene, which is already known to be mutation-prone.14 All of these alternatives need to be addressed in future investigations.
In this study, we opted to investigate the tibialis anterior muscle rather than the widely investigated vastus lateralis muscle. The rationales for this choice were: firstly, we wanted to investigate whether COPD changes shown in vastus lateralis also occur in another muscle type; secondly, there is easier access to this muscle by the semi-open needle biopsy method we used. Our findings show similar changes in both muscle types including fibre-type shift from type I to type II and decreased capillarisation. This further confirmed the systemic nature of COPD pathology and that the morphological changes induced by hypoxia are more likely to be a general occurrence in all skeletal muscles.
A drawback of this investigation is the small number of patients included in each disease group. However, the total number of patients included in the study is similar to, or even higher than, those included in previously published reports. Furthermore, in contrast with most published reports on COPD, we avoided pooling material from different patient groups, allowing us to study the three disease stages separately. Nevertheless, the findings of this report need to be confirmed and studied in a larger patient population, probably concentrating on the mild and moderate stages of COPD.
In conclusion, our results identify VHL as a potential new player in the aetiopathology of COPD. In this scenario (fig 6), VHL overexpression may have a negative effect on the transduction of the HIF signal to VEGF, leading to a decrease in tissue capillarisation, development of PCS, and, ultimately, increased risk of mortality.2 The findings of this study present the possibility of viewing COPD aetiopathology and its peripheral skeletal muscle complications in a different light by considering the crucial interplay between the hypoxic and angiogenic signalling cascades. In this context, VHL overexpression may potentially represent a diagnostic marker for development of skeletal muscle dysfunction at the early mild stage of COPD. Furthermore, VHL overexpression may represent a future target for therapeutic intervention strategies that aim to restore normal VHL function, thus preventing development of PCS and its adverse and mortal consequences in patients with COPD.
Acknowledgments
This project was supported by funds from the Research Committee, Örebro University Hospital. We are grateful to Anders Magnuson, Örebro University Hospital, for skilful statistical assistance, Dr Tommie Olofsson, Department of Pathology, Uppsala University Hospital, for expert technical assistance, and Dr Ian Jones, Department of Clinical Chemistry, Örebro University Hospital, for kindly reviewing the manuscript. We are also grateful for the assistance of Brita Lindstrand, Irene Eriksson and Margareta Larsson, Department of Respiratory Medicine, and Lisbeth Lindvall, Department of Pathology, Örebro University Hospital, for expert nursing and laboratory technical work.
REFERENCES
Footnotes
Competing interests: None.
Ethics approval: Obtained.
Patient consent: Obtained.