Lung volume dependence of respiratory function in rodent models of diabetes mellitus

Uncontrolled diabetes mellitus (DM) can result in long-term hyperglycemia that activates various molecular pathways [1], leading to inflammation [2], endothelial [3, 4], smooth muscle cell dysfunction [5], and protein destruction [6]. These mechanisms involve the formation of advanced glycation end products (AGEs) [7], which cause deterioration to the structure and function of matrix proteins and affect matrix–matrix and matrix–cell interactions [8].

The pathophysiological consequences of DM may also damage the pulmonary endothelium and bronchial smooth muscle cells, so the lungs can be among the organs most affected by DM [9‐12]. Endothelial dysfunction contributes to the remodeling of the airways and lung parenchyma related to oxidative stress and the overproduction of reactive oxygen species [13‐15]. These processes manifest as the sustained contraction of bronchial smooth muscle cells [13, 14, 16, 17], thickening of the alveolar walls [10, 18], and changes to the elastin–collagen matrix in the lung parenchyma [10, 19, 20]. Hyperglycemia also affects type II pneumocyte cells, resulting in decreased surfactant biosynthesis and secretion [21, 22].

Conflicting results have been reported for the effects of DM on lung function. Some studies have demonstrated declines in spirometry outcomes [23‐26], whereas others found no detectable change [27‐30]. Possible explanations for this discrepancy include the dependence of spirometry results on the effort and cooperation of patients [31, 32], and substantial differences between study populations, such as the type and severity of DM, age, and comorbidities. Furthermore, forced expiratory parameters have limited sensitivity to detecting detrimental changes in lung tissue and peripheral airway mechanics [9, 27, 30], the aspects of the lung affected primarily by DM [30, 33]. Thus, it remains unclear how DM affects airway function and the viscoelastic mechanics of respiratory tissue, including its dissipative and elastic properties. In particular, it is unknown whether DM influences the changes in mechanical parameters that occur with changes in lung volume.

This study aimed to characterize changes in the airway and respiratory tissue viscoelastic parameters to clarify the pulmonary consequences of hyperglycemia. Therefore, we applied a measurement technique that allowed the exact characterization of these changes in well-established animal models of type 1 and type 2 DM. We hypothesized that DM would result in the development of compromised airway function and changes in airway responsiveness resulting from the sustained contractile response of the bronchial smooth muscle, as well as deterioration to both the dissipative and elastic components of respiratory tissue viscoelasticity. We also assessed the contribution to the DM-induced changes in lung function of the loss of lung volume, gas exchange abnormalities, proliferation, and fiber–fiber interaction by evaluating changes in the functional residual capacity (FRC), the arterial partial pressure of oxygen to the inspired oxygen fraction (PaO₂/FiO₂), intrapulmonary shunt (Qs/Qt), and lung histology.

Body weights, serum glucose levels, and the initial lung function parameters are summarized in Table 1. At the end of the 12-week treatment period, the mean body weight of the rats in the DM1 group was significantly lower than those in the DM2 and control groups. Significant elevations in blood glucose level were observed in the rats of the DM1 and DM2 groups compared to the control animals. The FRC measurements showed a significant reduction in static lung volume in the DM1 group rats; however, when the FRC values were normalized to body weight, they were significantly higher for the DM1 group rats than for those in the other two groups. Specific airway and respiratory tissue parameters were significantly higher in the DM1 and DM2 groups than in the control group.

Figure 2 shows how the airway and viscoelastic parameters of the respiratory tissues varied with the PEEP for the three groups. Raw was significantly greater in both groups of diabetic animals than in the control group at all PEEP levels (p < 0.001). Two-way ANOVAs showed significant interactions between group and PEEP level for the parameters G (p < 0.001) and H (p < 0.001), indicating that the treatment affected these parameters’ dependence on PEEP. The dissipative properties of the respiratory tissues reflected by G were significantly compromised in the DM1 group when PEEP = 0 cmH₂O. The respiratory tissue stiffness was elevated in both the DM1 (p < 0.001) and the DM2 (p = 0.039) groups at all PEEP levels, with more pronounced differences at low lung volumes. The dissociated PEEP dependences of G and H resulted in changes in η that varied between the groups, with significant decreases observed in the DM1 group for all three levels of PEEP (p < 0.001) and in the DM2 group when the PEEP was 3 or 6 cmH₂O (p = 0.014) compared to the control group.

Airway resistance (left) and its changes relative to the baseline (right) during the MCh provocation tests are shown in Fig. 3. Significant elevations in the basal Raw values were observed in the DM1 (p < 0.001) and DM2 (p < 0.05) groups. These differences remained at the lower doses of MCh, whereas the absolute value of Raw became significantly lower in the DM1 rats at the highest dose of MCh (p < 0.05). Due to the significantly higher baseline Raw values in the diabetic rats, the differences in the dose-response curves to MCh between the protocol groups is more obvious when the responses are expressed as relative changes to baseline (right). In the control and DM2 groups, elevations in Raw were statistically significant from the 8 μg/kg/min MCh dose (p < 0.05), whereas in the DM1 group this increase was only observed after 16 μg/kg/min (p < 0.05). At the highest MCh dose, the MCh-induced relative increases in Raw were significantly lower in both DM1 and DM2 groups compared to control (p = 0.001). The characteristic shifts in the dose–response curves showed that the PD₅₀ for MCh was significantly higher for the DM1 group (25.3 ± 20 μg/kg/min) than for the control group (8.9 ± 3.3 μg/kg/min, p = 0.001) and the DM2 group (12.7 ± 7.4 μg/kg/min, p = 0.026).

The parameters reflecting the gas exchange properties of the lungs are summarized in Fig. 4. The rats with healthy lungs in the control group exhibited moderate values of intrapulmonary shunt (< 9.3%) and physiological PaO₂/FiO₂ (> 438 mmHg); these indices exhibited systematic improvements with increasing PEEP to 3 (p < 0.02) and 6 cmH₂O (p < 0.05). In the DM2 group, there was a near-significant tendency for Qs/Qt to be reduced compared to the control group at a PEEP of 6 cmH₂O (p = 0.065), whereas PaO₂/FiO₂ was reduced at a PEEP of 3 cmH₂O (p = 0.05) and tended to diminish at a PEEP of 6 cmH₂O (p = 0.069). In the DM1 group at all three PEEP levels, there were significant increases in the intrapulmonary shunt (p < 0.001) associated with markedly compromised PaO₂/FiO₂ (p < 0.001). In addition, changes in the PEEP dependence of the gas exchange parameters were observed in the DM1 group, with no monotonous improvements in Qs/Qt and PaO₂/FiO₂ with increasing PEEP.

Figure 5 shows the expression of collagen in the three groups. The percentage area of collagen was significantly higher in both DM groups than in the control group (p < 0.001). The collagen content of the lung parenchyma was greater in the DM1 group than in the DM2 group (p < 0.001).

Figure 6 shows the relationship between the percentage area of collagen obtained by lung histology and the viscoelastic dissipative and elastic mechanical parameters. There were significant correlations between the histology and both tissue mechanical indices, G and H, with correlation coefficients of 0.67 (p < 0.001) and 0.63 (p < 0.001), respectively.

The characterization of the pulmonary effects of DM in this study revealed detrimental changes in airway function, viscoelastic tissue mechanics, gas exchange, and collagen expression in the lung, with more severe manifestations in the rat model of type 1 DM than in that for type 2 DM. The increase in the basal airway tone with DM was associated with compromised dissipative and elastic tissue mechanics. The model of type 1 DM also showed diminished airway responsiveness to an exogenous cholinergic stimulus. These adverse mechanical and functional changes were accompanied by an increased intrapulmonary shunt and impaired PaO₂/FiO₂. Increasing the lung volume had a beneficial effect on the lung mechanics in both diabetic groups, whereas it had no benefit on gas exchange.

This study used well-established models to induce type 1 and 2 DM [34‐39]. Type 1 DM was induced by a single high dose of STZ to cause the destruction of the pancreas, whereas type 2 DM was induced by a low dose of STZ to cause diffuse degeneration of the pancreatic cells, combined with a high-fat diet. As a result of these treatments, the blood glucose levels in both DM groups were markedly higher after 12 weeks than those of the control group. The body weight of the type 1 DM group rats was reduced, which can be explained by insulin deficiency [47, 48]. As with other chronic complications of DM, the pulmonary complications would be expected to manifest in the late stage of the condition [12]. For this reason, a 12-week period was allowed for the animals to develop lung dysfunction. This long period with markedly elevated serum glucose levels allowed the development of adverse pulmonary symptoms without a fatal outcome.

Forced oscillatory assessment of airway mechanics showed that DM caused a deterioration of basal Raw, with more severe changes occurring in the type 1 DM model rats (Fig. 2 PEEP0). As these changes were also apparent in the specific airway resistance values (Table 1) of both DM groups, the differences in lung size cannot solely be accounted for by the compromised airway mechanics. Instead, these pathological changes in the airways may be explained by the decreased vagal tone [49], excessive mucus production [50], low-grade chronic inflammation [51], activation of inflammatory pathways [52], or bronchial smooth muscle cell proliferation [5, 53]. The present study’s finding of increased airway tone in the DM rat models is in qualitative agreement with the results of previous studies that assessed airway resistance in DM [27, 54], providing supporting evidence for the potential development of chronic airway obstruction in patients with DM.

The effects of DM on the mechanical properties of the lung tissue has been studied in an ex-vivo experiment, and the results were limited to assess the elastic behavior of the lung parenchyma [55] However, dissipative properties of the lung tissue play an important role in determining the physiological lung mechanics and the changes of this component is characteristic in various lung diseases [56]. In the present study using an in-vivo setting, significant deteriorations were observed in the viscoelastic mechanical parameters of the respiratory tissues in both groups of diabetic rats (Fig. 2), and this adverse change affected both the dissipative and elastic properties. These differences compared to the control group remained after normalization of the values to the FRC (Table 1), indicating that intrinsic changes to the dissipative and elastic properties of the respiratory tissues can be anticipated. Collagen is the main determinant of overall lung tissue viscoelasticity [57, 58]. In the present study, the volume of collagen increased in the DM models, in agreement with results reported for patients with DM [59] (Fig. 5). There were significant correlations between the histological and mechanical findings (Fig. 6), which suggested that the overexpression of collagen may be a primary cause of the compromised tissue damping and elastance observed in DM. The underlying pathophysiological mechanisms responsible for this extracellular matrix remodeling may be related to the activation of pathological pathways that contribute to the formation of AGEs, which also stimulate the production of extracellular matrix components, including collagens, thereby affecting the interactions of the extracellular matrix [1].

Increasing numbers of patients with DM require surgical procedures that involve mechanical ventilation, which has increased the burden on healthcare providers [60, 61]. There is, therefore, a need for awareness that adequate mechanical ventilation should be provided for these patients. Open lung strategies require the application of an appropriate PEEP; however, the impact of different levels of PEEP on respiratory function in DM has not been clarified. Our findings demonstrated that the between-group differences in the airway resistance and in the tissue mechanical parameters representing viscoelastic dissipation and elastance disappeared at a moderately elevated PEEP (6 cmH₂O; Fig. 2). The excessive PEEP dependence of the respiratory mechanical parameters was consistent with the diminished surfactant function reported previously in models of DM [21, 22]. It suggests that applying PEEP can have a beneficial effect on respiratory mechanics in this metabolic disease. The mechanical improvements were reflected in the decreased intrapulmonary shunt and increased PaO₂/FiO₂ in the DM rats when a PEEP of 3 cmH₂O (Fig. 4) was applied. Nevertheless, substantial differences remained even with elevated PEEP in both groups of DM rats; this can be attributed to the persistent alveolar-capillary barrier damage observed in DM [26, 59, 62]. The type-II pneumocyte and surfactant layer damage [21, 22] along with the low-grade inflammation [26] observed in DM can contribute to the alveolo-capillary dysfunction. Noticeably, there was recurrent deterioration of PaO₂/FiO₂ and Qs/Qt in the type 1 DM rats at a PEEP of 6 cmH₂O. This group would be expected to develop the well-established adverse pulmonary vascular consequences of hyperglycemia [63]. Accordingly, formation of AGEs as detailed above contributes to a proliferation of pulmonary endothelial and vascular smooth muscle cells and thicker basal lamina [64], which may have resulted in the intra-acinar and alveolar arterioles becoming prone to collapse when the PEEP exerted an external mechanical load on the capillary network.

An important physiological feature of airways is the adaptation of their caliber in response to exogenous stimuli. The results for the type 1 DM group demonstrated diminished airway responsiveness to MCh, which indicated that the adaptation ability of the bronchomotor tone had been severely compromised. Pathophysiological mechanisms that may have been involved in the decreased airway responsiveness include compromised vagal tone development due to autonomic diabetic neuropathy [49, 65], smooth muscle cell dysfunction [66], and/or epithelial damage [13]. Hyperinsulinemia, insulin resistance and hyperglycemia can lead to hyperproliferation and phenotype changing of the airway and bronchial smooth muscle cells and induce tracheal wall thickening [5, 53]. The smooth muscle cell damage might occur due to various factors, for instance disturbances in the nitric oxide synthesis [13], TGF-β and Rho-associated protein kinase pathway activation [16, 67, 68]. Nonetheless, conflicting results have been reported for the effects of DM on airway responsiveness. The findings of the present study are consistent with earlier findings of a reduction in the bronchial response to cholinergic stimuli in DM [13, 29, 69]. Previous reports of no change in airway responsiveness [28, 66, 70] or the development of airway hyperresponsiveness [16, 17, 71] may be explained by the insensitivity of the methods to assess airway function [66], the lack of airway innervation [16, 17, 70], or the involvement of confounding factors, such as, smoking, genetic differences and/or phenotype, and the duration of DM [28, 71]. Furthermore, in previous clinical studies, diabetes was often associated with comorbidities affecting the respiratory system, which may cause variation in the manifested pulmonary effects, and can contribute to the discordant results in the literature.

Several methodological aspects of the present study warrant consideration. As well-established models were adapted in the present study to produce the most important features of DM, including hyperglycemia and insulin deficiency, verification of the effectiveness of the STZ treatments was limited to evaluating serum glucose levels; a detailed characterization of the resulting metabolic profile would go beyond the focus of the present investigation. The models of type 1 and type 2 DM differed only in the dose of STZ and the diet regimen. The high dose of STZ administered to the rats in the type 1 DM group was likely to destroy most of the pancreatic β-cells, whereas less severe β-cell destruction in the type 2 DM group was complemented by a high-fat diet to induce a combination of impaired insulin production and peripheral insulin resistance [35, 48]. The difference in body weight between the groups suggested marked differences in their metabolic status; however, the respiratory consequences were analogous in the two DM groups, with symptoms differing only in severity. This study was designed to focus primarily on the changes in the mechanical properties of the respiratory system in diabetes. However, there are plethora of diagnostic tools to characterize pulmonary dysfunction both at on organ and at cellular levels. Analyses of bronchoalveolar lavage fluid to assess altered surfactant function, or quantitative collagen and smooth muscle cell assays by using immunohistochemistry methods would give further insights into the underlying mechanisms; these methods may be subjects of forthcoming investigations.

In summary, distinct measurements of airway and respiratory tissue viscoelastic parameters in models of type 1 and type 2 DM showed evidence of detrimental changes in both compartments. The decline in airway function was reflected in elevated airway resistance and abnormal adaptation of the airways to exogenous constrictor stimuli. Lung tissue remodeling was manifested in compromised viscoelastic tissue mechanics, which affected dissipative and elastic properties, and the concurrent overexpression of collagen fibers in the extracellular matrix. These detrimental mechanical defects were overcome by applying PEEP, which demonstrated alveolar collapsibility in DM. However, even with the application of PEEP, gas exchange parameters remained compromised in the models of type 1 and type 2 DM, even deteriorated further in the type 1 DM model, indicating alveolo-capillary dysfunction. These findings may contribute to an improved understanding of the pulmonary consequences of DM and hold promise for the advancement of therapeutic interventions. PEEP titration during mechanical ventilation guided by respiratory mechanical parameters might be beneficial for this increasingly prevalent patient population.