Granulocytosis and increased adhesion molecules after resistive loading of the diaphragm

Acute exercise increases the circulating pool of leukocytes, the magnitude of which is relative to the intensity and duration of the exercise 1, 2. The number of circulating nonspecific immune cells, including mobilisation of neutrophils from the marginated pools into the circulation, is mediated by the exercise intensity-dependent secretion of stress hormones, such as catecholamines 3, cortisol 4, 5 and growth hormone 3, 4. Catecholamine-mediated flushing of neutrophils from the marginated pools into the circulation occurs through the shear force of exercise-induced haemodynamics. Cortisol and various cytokines induce early release of neutrophils from the bone marrow 3, which is reflected by an increase in band cells after exhaustive exercise 3, 4.

Exertion-induced muscle injury can be initiated by unaccustomed exercise, resulting in structural disruption of muscle fibres during and post-exercise 5. This is followed by neutrophil infiltration and release of myocellular proteins, such as creatine kinase (CK) and myoglobin, into the circulation 4, 5. Marked systemic neutrophilia with a leftward shift (reflective of increased band cells) and enhanced capacity of neutrophils to release reactive oxygen species have been documented after endurance exercise and were correlated positively with subsequent efflux of CK 4, 6, 7. This suggests that after stressful exercise, mobilised and primed neutrophils may be important mediators of exertion-induced muscle damage 4. Furthermore, several investigators have recently implicated indices of acute inflammation to be a mechanism for Z-line disruption, myocellular protein release and delayed onset muscle soreness after eccentric exercise 8–10.

Local inflammation in injured muscle can include complement activation, upregulation of adhesion molecule expression on leukocytes and endothelium with subsequent migration, and infiltration of blood leukocytes into the affected tissue 9, 11. The selective recruitment of leukocytes into traumatised tissue in a defined sequence 11 involves the combined action of multiple cell adhesion molecules, including intracellular adhesion molecule (ICAM)-1 and vascular cell adhesion molecule (VCAM)-1, that are expressed on the endothelium and which bind to LFA-1 and VLA-4 expressed on leukocytes, respectively 12, 13.

Similar to exertion-induced injury of the limb muscles, marked diaphragm injury has been shown after 1.5 h of intensive inspiratory resistive loading (IRL) 14, 15. The IRL in this animal model is ∼60–70% of the maximal load, which might be similar to the load experienced by patients with chronic obstructive pulmonary disease (COPD) or asthma during an acute exacerbation. The role of adhesion molecules and the response of circulating and intramuscular populations of leukocytes have not yet been characterised in this model. The postulation presented here is that diaphragm injury induced by IRL increases circulating leukocytes, and upregulates ICAM-1 and VCAM-1 expression on endothelium, which promotes recruitment of neutrophils and macrophages/monocytes into the injured muscle. The purpose of this study was to quantify circulating leukocyte subsets, the expression of VCAM-1 and ICAM-1 in the diaphragm, infiltrating leukocyte subsets, and diaphragm injury after IRL.

METHODS

Ethical approval was provided by the University of British Columbia Committee on Animal Care (Canada). The study conformed to the animal care guidelines of the Canadian Council on Animal Care. Two groups of New Zealand White rabbits ranging 3.5–4.5 kg were studied: IRL (n = 8) and control (n = 7) groups.

RESULTS

DISCUSSION

The unique findings of this study are that IRL in rabbits induced an increase in circulating band cells, signifying bone marrow stimulation. Three days after the loading, increased expression of the adhesion molecules, ICAM-1 and VCAM-1, in diaphragm vessels was accompanied by increased macrophage and neutrophil recruitment into the injured muscle. This is the first study to show this sequence of inflammatory events after exertion-induced diaphragm injury, and it is speculated that these leukocytes contribute to the diaphragm muscle injury seen after resistive loading.

Granulocytes increased across animals at 6 and 12 h, and band cell counts increased at 6 h post-IRL; however, the concentration of total circulating leukocytes across groups did not change over time. The increase in circulating neutrophils that accompanied the increased band cell counts implicates bone marrow release as the main source of immature circulating neutrophils in addition to mobilisation of marginated pools 20. In the presence of increased granulocyte and band cell counts, the constant counts of total circulating leukocytes after IRL were probably due to a fall in lymphocyte concentrations. This finding is consistent with previous findings after strenuous endurance exercise 21, 22. Physical exercise-induced transient redistribution of peripheral blood is characterised by a granulocytosis accompanied by lymphocytosis during exercise, followed by a more pronounced granulocytosis accompanied by lymphopenia after exercise 22. Wigernaes et al. 21 reported a continuous increase in neutrophils during active recovery with a fall in lymphocytes, which kept the total leukocytes constant at 120 min following a 60-min bout of uphill running at 83% of maximal oxygen uptake. In the present study, mononuclear cell counts tended to decrease in the IRL group at 6 h post-IRL compared with baseline, but the decrease did not reach a significant level, which could be due to the small change and small sample. Based on the means, variances and differences shown in the circulating mononuclear cells of the IRL group, a sample size of n = 45 would be required to achieve a power of 0.80 with a standardised difference of 0.83.

Neutrophil mobilisation after the control surgery and IRL appears to be a nonspecific stress response and is probably mediated by stress hormones, such as catecholamines 3, 23 and cortisol 4, 5. The significant increase in band cells after IRL, however, is a more specific injury rather than a stress response indicative of bone marrow release 23. Band cell release could be attributed to growth hormone and cytokines induced by the exertional stimulus 3, 4. Strenuous resistive breathing has been shown to induce increases in plasma interleukin (IL)-1β, IL-6 and tumour necrosis factor-α, which have been shown to be significantly blunted by antioxidants 24, 25. IL-6, IL-8 and granulocyte colony-stimulating factor may be associated with delayed-onset neutrophil mobilisation from the bone marrow reserve 1, 4. Previously, IL-6 and macrophage colony-stimulating factor responses were shown to be positively correlated with neutrophil mobilisation after exercise 20.

In the current study, it was found that resistive loading of the inspiratory muscles by small muscle mass exercise alters leukocyte trafficking. This is consistent with a recent report by Nemet et al. 26 who demonstrated that 10 min of low-intensity unilateral wrist flexion exercise can increase the concentration of inflammatory mediators, growth factors and circulating leukocytes in the control arm and not just in the exercising arm. They concluded that the increase in circulating leukocytes and inflammatory cytokines may have resulted from stimulation of the autonomic nervous system rather than the local working muscles.

In the present study, the magnitude of the increase in band cells after IRL was much less compared with that reported after marathon running 20. This is probably due to the much smaller muscle group recruited during IRL relative to running. Neutrophilia, and especially band neutrophil mobilisation was positively correlated to intensity of strenuous exercise as reflected by the exercise percentage maximal oxygen uptake, supporting the suggestion of Suzuki et al. 4, 20 that magnitude of the workload may be a determinant of the amount of neutrophil mobilisation from the bone marrow reserve. The present finding of a smaller circulating leukocyte redistribution after IRL was not surprising, because the small muscle mass of the inspiratory muscles does not increase exercise oxygen uptake to the same magnitude as marathon running. Although oxygen consumption was not measured in the current study, oxygen uptake during high and medium IRL in a dog model was 1.29- and 1.17-times the resting oxygen uptake, respectively 27. In contrast, whole body exercise at 60% of maximal oxygen uptake represents an estimated six-fold increase over resting oxygen uptake.

In the present study, the time course of no increase in circulating leukocytes and granulocytes during the exercise stimulus was somewhat different than reported data after other exercise stimuli. The primary reason why there was no early increase in circulating leukocytes during the IRL stimulus while the rabbits were anesthetised (2 h time point) was probably due to the inhibitory effects of anaesthetics on neutrophil function and kinetics. Thiopental, the anaesthetic used in the current study, decreases human neutrophil chemotaxis in a dose-dependent manner at clinically relevant concentrations 28. Previous work in the current authors' laboratory has shown that circulating leukocytes counts are lower in rabbits while anesthetised with a variety of different agents (unpublished data), which may be due in part to anaesthetic-induced decreases in stress and circulating catecholamines that result in increased margination of neutrophils in the lung capillaries. Although these anaesthetic effects have not been published, Boxer et al. 29 found that a β-antagonist, propranolol, will largely, but not completely, block the increase in adrenaline-mediated increase in circulating neutrophils.

The present study examined ICAM-1 and VCAM-1 expression and leukocyte subsets invasion in the diaphragm at 72 h post-IRL in this model, based on the previous time course studies that showed the greatest amount of muscle injury and inflammation 3 days after the exercise stimulus 30–34, including one study that showed a peaking of neutrophils and macrophages at 3 days 34. One previous IRL study has demonstrated that overt diaphragm injury, occurring at 72 h post-IRL, was accompanied by increased inflammatory cells in the diaphragm 14. The leukocyte–endothelial cell interaction is an early step in the cascade of events that leads to the development of the inflammatory response and cellular extravasation. It was proposed that the recruitment of granulocytes to the damaged diaphragm is supported by the widely accepted paradigm that inflammatory stimuli activate the endothelial cells of these vessels to express adhesion molecules and chemokines, which physically engage circulating leukocytes and promote their adhesion to these vessels. The adhesion molecules, ICAM-1 and VCAM-1, are normally expressed and upregulated during inflammation, or expressed only on activated endothelial cells in response to inflammatory stimuli 35.

The strong expression of ICAM-1 and VCAM-1 in the injured diaphragm supports the postulation that interaction between these molecules and leukocytes is a major factor in the migration of activated leukocytes into the inflamed diaphragm 36, 37. As expected, no VCAM-1 was expressed on the endothelium of vessels in the control group, but positive staining was observed following IRL.

The findings of increased expression of ICAM-1 and VCAM-1 in the diaphragm after IRL were similar to those reported after other types of exercise and in muscle diseases. Previous studies have shown that exercise can induce increases in soluble (s) ICAM-1 levels in the plasma 12, 38. Akimoto et al. 12 measured plasma concentrations of sICAM-1 before and after three types of exercise. The plasma concentration of sICAM-1, 1 day after a 42-km run or a downhill run, was significantly increased, whereas high-intensity bicycle ergometer exercise did not influence plasma concentrations of sICAM-1. It was concluded that plasma concentrations of sICAM-1 increase only after exercise associated with muscle damage 12. Immunohistopathological studies have shown the upregulation of ICAM-1 and VCAM-1 in endothelial cells and infiltrating cells in biopsies of muscle tissue from patients with polymyositis and dermatomyositis 39, 40, and with critical illness polyneuropathy and myopathy 18. Tissue expression of VCAM-1 is associated with muscle injury and pronounced inflammation in polymyositis 40. To the current authors' knowledge, no studies thus far have examined the role of adhesion molecules in the diaphragm after IRL.

Macrophages were the predominant inflammatory cell type in the diaphragm at 72-h post-IRL. Leukocytes, initially neutrophils and subsequently monocytes/macrophages, have been identified in areas with muscle injury after eccentric exercise 10. Neutrophils, the first inflammatory cell type to appear in injured muscle 11, 34, 41, have been suggested to have a role in both injury and regeneration of injured muscle. Although no direct in vivo evidence exists, neutrophils have been hypothesised to delay muscle regeneration by exacerbating the initial injury and/or by injuring myotubes through the release of free radicals and proteases 8. In addition, neutrophils could facilitate muscle regeneration by removing tissue debris from the injured area via phagocytosis 42 and by activating satellite cells 43. Macrophages, which increase in concentration 1–3 days after injury, are thought to contribute to muscle regeneration 11, but also may induce some injury. The beneficial contribution of macrophages to the events associated with muscle regeneration has been ascribed to their ability to phagocytose tissue debris 44 and to their capacity to cause myoblast proliferation in vitro 45. However, macrophages also have the ability to increase damage in muscle when there is an impaired capacity to generate nitric oxide, which serves to protect muscle membranes 46.

The elevation in neutrophils and macrophages in the diaphragm in the present study is consistent with other observations after lengthening contractions of muscle 47, 48, electrically stimulated contractions 49 and exposure of muscle to toxin or poison 41, 42, but contrary to other recent reports 50, 51. Of these two differing reports, one described the extensor digitorum longus (EDL) muscle in response to in situ lengthening contractions 50 and the other investigated the diaphragm in response to resistive loading 51. The apparently discrepant results by Vassilakopoulos et al. 51 may actually be complementary to the present study and be explained by the different time courses examined. The study by Vassilakopoulos et al. 51 did not find increased myeloperoxidase activity in the diaphragm up to 6 h after resistive loading in the rat, which may indicate that neutrophils were recruited in the circulation, but did not enter the diaphragm until after 6 h. Lapointe et al. 50 concluded that neutrophils in the EDL muscle were not significantly elevated during the hours to days after overt injury caused by lengthening contractions. Alternative explanations for the lack of neutrophil influx in these other two studies 50, 51, in contrast to the current study, may be related to differences in species and/or lesser sensitivities in the techniques used to detect neutrophils. CD43, which was used to quantify neutrophils by Lapointe et al. 50, may be a less sensitive antibody for labelling neutrophils in skeletal muscle 48. Furthermore, when a histochemical technique for labelling neutrophils using diamino-benzadine, which illustrates the myeloperoxidase content of granulocytes, was piloted by the current authors, it was found to be less sensitive than using NP-5 to illustrate neutrophil defensin. The present study also differs from some reports of autoimmune muscle diseases, such as polymyositis, where the majority of the inflammatory cells are T-cells. This is not surprising considering the contrasting aetiologies 52.

The IRL sustained by rabbits in this study represents ∼60–70% of the maximal load; a previous report by the present group showed that the maximal diaphragmatic pressure of the rabbit is ∼65 cmH₂O during bilateral phrenic nerve stimulation 53. This load may appear high in terms of clinical relevance, but similar loads to the IRL model may occur during episodes of acute bronchospasm in asthmatics, acute exacerbations of COPD or during weaning trials in mechanically ventilated patients. Patients with stable severe COPD have increased diaphragmatic activation at rest, and 43% of the maximal electromyogram root-mean square 54, which increases to 81% during progressive exercise 55. A major clinical consequence of the inflammatory and injury response of the diaphragm after inspiratory resistive loading is the associated 35% reduction of its force production shown in earlier work from the current group 15. Thus, modulation of the inflammatory process during these conditions might facilitate prevention or minimise some of the potential diaphragm injury during these clinical scenarios.

Injury and inflammation in response to an exertional overload of the diaphragm may be part of the normal physiological response of adaptation and repair. In fact, inflammation may be essential to expedite optimal repair, as evidenced by a study in rodents that showed administration of nonsteroidal anti-inflammatories resulted in slower regeneration and a weaker myotendinous junction after a partial repair 56. If diaphragm injury overwhelms the reparative processes, however, injured myocytes may be replaced with connective tissue similar to repetitive strain models of limb muscle injury 57. Thus, in the event of an overwhelming injury response, carefully timed administration of anti-inflammatories that diminish the injury, followed by agents that increase muscle regeneration may improve clinical outcomes.

One limitation of this study is that an increased ICAM-1 and VCAM-1 expression associated with inflamed, injured diaphragm was observed, but cause and effect were not examined. Future studies are required to identify whether blocking these adhesion molecules would result in less diaphragm injury and/or force loss. A second limitation is the small number of animals in each group and that the IRL group showed a wide variance of VCAM-stained endothelial cells, including three animals who did not show positive VCAM-stained vessels. This variation in VCAM expression may reflect alternative triggers in addition to IRL such as hypoxaemia, or hypercapnia and/or individual differences in susceptibility to diaphragm injury due to factors such as inherent fibre-type characteristics and variable recruitment patterns of inspiratory muscles during IRL.

In conclusion, in this study, inspiratory resistive loading mobilised neutrophils into the circulation with increasing band cell counts, indicating bone marrow release as a major contributor for the increase in circulating neutrophils. Increased surface expression of intracellular adhesion molecule-1 and vascular cell adhesion molecule-1, and increased infiltration of neutrophils and macrophages/monocytes, are associated with inspiratory resistive loading-induced diaphragm injury and probably mediate muscle injury in this model.