Pressure support ventilation attenuates ventilator-induced protein modifications in the diaphragm

This study was performed in accordance with the recommendations of the National Research Council's Guide for the Care and Use of Laboratory Animals [23]. This experiment was approved by the University of Clermont-Ferrand animal use committee. Forty-two adult male Sprague-Dawley rats (250 g) were individually housed and fed rat chow and water ad libitum and were maintained on a 12-hour light/dark photoperiod for 1 week before initiation of these experiments. Animals were randomly assigned to 6 or 18 hours of CMV or PSV with 21% O₂ (Figure 1). All surgical procedures were performed using aseptic techniques. After reaching a surgical plane of anesthesia (sodium pentobarbital, 50 mg/kg of body weight, intraperitoneal), animals were weighed and tracheostomized. The jugular vein was cannulated for the infusion of saline and sodium pentobarbital (5 mg/kg of body weight per hour). Body fluid homeostasis was maintained by administration of 2 mL/kg per hour intravenous electrolyte solution. The carotid artery was cannulated for measurement of arterial blood pressure, pH, and blood gas tensions (GEMpremier-3000 system; Instrumentation Laboratory, Lexington, MA, USA). Heart rate and electrical activity of the heart were monitored via a lead II electrocardiogram using needle electrodes placed subcutaneously. Throughout the ventilation period, animals received enteral nutrition (via a nasogastric tube) using the AIN-76 rodent diet with a nutrient composition of proteins, lipids, carbohydrates, and vitamins which provided an isocaloric diet (Research Diets, Inc., Brunswick, NJ, USA). Body temperature was monitored (rectal thermometer) and maintained at 37°C ± 1°C with a recirculating heating blanket. Continuing care during the experimental period included expressing the bladder, removing upper airway mucus, lubricating the eyes, rotating the animal, and passive movements of the limbs. Animals (both CMV and PSV) were regularly rotated to prevent atelectasis, to limit mechanical constraints, and to maintain ventilation/perfusion ratio homogeneity.

Immediately after inclusion, animals were mechanically ventilated using a volume-driven ventilator (Rodent Ventilator model 683; Harvard Apparatus, Holliston, MA, USA) for 6 hours (group 1) or 18 hours (group 2). The tidal volume was 10 mL/kg of body weight and the respiratory rate was 80 breaths per minute, with a fraction of inspired oxygen (FiO₂) of 21% but without positive end-expiratory pressure. These ventilatory conditions resulted in complete diaphragmatic inactivity and prevented noxious effects of a hypercapnia on the muscular contractile properties [2, 3, 24, 25]. At the end of the experimental period, each animal was weighed, and the costal diaphragm was rapidly dissected and frozen in liquid nitrogen. Samples were stored at -80°C until subsequent assay (except for samples in which protein synthesis and proteolysis were analyzed, which were treated as described below). At the same time, arterial blood was obtained for culture.

Animals were also anesthetized and mechanically ventilated for 6 hours (group 4) or 18 hours (group 5) as described above (model PSV ventilator DARHD01; IFMA, Aubière, France). The level of pressure support applied, determined during preliminary studies, allowed a minute volume of 200 ± 10 mL/minute (respiratory rate of 80 ± 10 breaths per minute and FiO₂ of 21%). The range of pressure support used was 5 to 7 cm H₂O. The ventilator had a pressure trigger. The expiratory trigger was fixed at 25% of peak inspiratory flow, and the maximum inspiratory time was set at 1 second. The ventilator did not have back-up ventilation. If the animal was not triggering, no pressure was released. Continuing care during the experiment was also applied as above. At the end of the experimental period, the costal diaphragm was rapidly removed, dissected, and frozen in liquid nitrogen. Samples were stored at -80°C.

Control animals (group 3) were free of intervention before inclusion (not mechanically ventilated). These animals were anesthetized and their diaphragms were rapidly dissected, frozen, and stored at -80°C until subsequent assay. Because of the biochemical constraints (variability of the solutions of Krebs-Henselheit), each day of experimentation required a control animal.

At the appropriate times (6 or 18 hours), the entire diaphragm, costal and crural, was removed, dissected, and weighed. All biochemical studies were conducted using the costal region of the diaphragm. Samples were rapidly frozen in liquid nitrogen and stored at -80°C until assay.

A two-way analysis of variance (StatView^®, version 5.0; SAS Institute Inc., Cary, NC, USA) with time (6 versus 18 hours) as one factor and modality (PSV versus CMV versus control) as the other factor was used. When appropriate, a post hoc protected least squares difference Fisher test was used. Values are mean ± standard deviation in the text and mean ± standard error of the mean in the tables and graphs. Statistical significance was defined a priori as a P value of less than 0.05.

The principal biologic parameters are summarized in Table 1. Blood gas/pH and cardiovascular homeostases were maintained constant in all animals during CMV and PSV. There were no significant differences in total body mass between groups and no group experienced a significant loss of body mass, indicating adequate hydration and nutrition during the experimental period (Table 2). All animals urinated and experienced intestinal transit during the experimental period. All blood cultures were negative for bacteria and none of the animals demonstrated sepsis signs.

Compared with control animals, diaphragmatic protein catabolism was significantly increased after 18 hours of CMV (33%, P = 0.0001) but not after 6 hours (Figure 2). There was a 36% increase in proteolysis between 6 and 18 hours of CMV (P = 0.0003). Compared with CMV, 6 and 18 hours of PSV showed no significant increase in proteolysis. Moreover, duration of PSV had no effect on total proteolysis evolution (4.18 ± 0.20 and 4.23 ± 0.12 nmol of tyrosine per milligram of protein per hour after 6 and 18 hours, respectively). Both chymotrypsin-like and tripeptydyl-peptidase 20S proteasome activities were increased after 18 hours of CMV (+50% versus controls and +45% versus CMV 6 hours). PSV did not increase 20S proteasome activities, regardless of the ventilation duration (6 or 18 hours).

Compared with control animals, CMV decreased diaphragmatic protein synthesis by 50% (P = 0.0012) after 6 hours and by 65% (P < 0.0001) after 18 hours of MV (Figure 3). The difference between 6 and 18 hours of CMV was 30%, which was not statistically significant. No variation of protein synthesis was observed during PSV. After 18 hours of MV, CMV showed a 94% reduction in protein synthesis compared with PSV (P = 0.0002).

Compared with control animals, protein oxidation, measured by myofibrillar protein carbonyl levels, was significantly increased after 18 hours of CMV (+63%, P < 0.001) and PSV (+82%, P < 0.0005) (Figure 4). Myofibrillar protein oxidation was not influenced by ventilator mode.

The anesthetic agent, sodium pentobarbital, could have affected the rate of muscle protein synthesis in the diaphragm. However, both MV and spontaneously breathing animals were anesthetized with sodium pentobarbital, so comparisons between groups are valid. Moreover, a previous study has reported that rats acutely anesthetized with sodium pentobarbital do not experience a significant decrease in protein synthesis in skeletal muscle [36]. Additionally, general anesthesia does not decrease protein synthesis in skeletal muscle in healthy humans undergoing abdominal surgery [37]. Collectively, these data indicate that protein synthesis is not altered by anesthesia per se. The influence of continued exposure of any given anesthetic agent (for example, 18 hours) would be difficult to separate from the reduced use during that state. However, the experiments reviewed above [36, 37] report normal rates of protein synthesis in limb-locomotor skeletal muscle during periods of time in which reduced use would not be expected to have an effect on protein synthesis. These reports [36, 37] indicate that anesthesia does not affect protein synthesis; therefore, the decreased rate of protein synthesis in the diaphragm during MV is attributable to MV, not to the anesthetic as previously reported by several authors [2, 4, 6, 11, 38].

Prolonged MV results in diaphragmatic atrophy and contractile dysfunction in animals. Evaluation of contractile diaphragmatic properties in PSV and CMV will have been clinically relevant. This study was not designed to respond to this question and we discuss only MV-induced diaphragmatic protein alterations. Further studies should focus on this point. Diaphragmatic contractions are avoided by CMV at a normal rate (80 cycles per minute). We have not tested this assessment but several authors have done so previously [4] and used this previously reviewed paper for a recent study [2, 11]. However, this does not exclude the possibility that the animals were triggering the ventilator during CMV in the present study. This is a real limitation of the manuscript.

In the present study, we simultaneously analyze the effects of MV on proteolysis, protein synthesis, and their kinetics. Consistent with earlier findings [2], our results confirm the increase in diaphragmatic proteolysis after 18 hours of CMV. Although diaphragmatic proteolytic injury has been implicated in the genesis of VIDD [7], less is known about modifications in diaphragmatic protein synthesis as a result of MV. Muscle atrophy can result from increased proteolysis [39], decreased protein synthesis [40], or both. Except for one recent study [11], none had considered the possibility that diaphragm atrophy associated with CMV could also result from decreased protein synthesis. We found both increased proteolysis and a time-dependent decrease in protein synthesis. Moreover, our results provide information about the probable kinetics of CMV-induced protein metabolism modifications. Indeed, the decrease in protein synthesis occurred extremely early (by the sixth hour of CMV), was worsened by the duration of MV, and preceded the increase in diaphragmatic proteolysis. It is interesting to note that, in the study of Shanely and colleagues [11], the results were obtained from the analysis of separate studies of in vitro proteolysis and in vivo protein synthesis. However, constant infusion of ¹³C-leucine, which is used in the analysis of in vivo protein synthesis, can modify an animal's protein profile by altering insulin release, on both the tissue and molecular levels [41], making interpretations between in vivo and in vitro models difficult. In addition, the nutritional profiles of animals can limit the interpretation. Indeed, some authors have compared the results obtained using fed [2] and unfed animals, implying a negative protein assessment [11, 41]. On the other hand, in vivo protein synthesis should be more relevant than in vitro proteolysis as used in our study. These methodological differences could explain some difference in the results.

Our data showed that PSV limits MV-induced increases in proteolysis and decreases in protein synthesis. Moreover, in contrast to CMV, modifications in protein metabolism were not affected by PSV duration. Because of differences in proteolysis/protein synthesis ratios, we hypothesized that PSV allows the maintenance of protein turnover. In addition, because CMV decreased protein synthesis, it is likely that CMV decreases or completely inhibits protein turnover. These differences in modification of metabolism may be due to differences in the type of diaphragmatic muscle damage caused by CMV and PSV. Indeed, as for peripheral skeletal muscle models, during PSV the diaphragm is subjected to exercise type activity through an increase in respiratory activity (versus CMV) [42‐44]. This exercise would protect the diaphragm from modifications related to muscular inactivity caused by CMV. During CMV, there is a complete absence of neural activation and mechanical activity in the diaphragm [4, 45], which undergoes passive shortening during mechanical expansion of the lungs [46, 47]. This trauma has been implicated in the genesis of VIDD [2, 11], in particular during sarcomere injury [48, 49] and during decreased force-generating capacity of the diaphragm [7, 50]. There has been little determination of the types of proteins implicated in CMV-induced metabolic damage. CMV has been shown to decrease the rate of mixed muscle protein synthesis by 30% and to decrease the rate of myosin heavy chain protein synthesis by 65% [11]. Although our study was not designed to analyze the type of proteins involved in the reduction of protein synthesis, it shed new light on the changes in protein synthesis associated with the conservation of diaphragm activity. Further experiments are necessary to determine the specific proteins implicated in the increased protein turnover observed with PSV. Our results also confirm that the 20S proteasome is involved in MV-induced proteolytic damage [2, 10]. CMV increases 20S proteasome activity in parallel with the increase in diaphragmatic proteolysis. After 18 hours of CMV, we observed an increase in the activity of extralysosomal TPPII, which degrades peptides generated by the proteasome. Similarly, 72 hours of CMV increased the level of MAF-box mRNA, which encodes an E3 ligase implicated in the ubiquitination of proteins targeted for degradation via the proteasome [38]. Together, these findings indicate the importance of the ubiquitin-proteasome pathway in CMV-induced diaphragmatic muscle damage and in overall regulation of muscle proteolysis [51] (as well as the importance of this enzymatic system within the skeletal muscle proteolytic machinery [52, 53]).

Little is currently known concerning the triggers or molecular signals of MV-induced protein metabolism modifications and muscle atrophy [51, 54]. Oxidative injury is induced by MV, and increased protein oxidation and lipid peroxidation were found to be associated with CMV [2, 55]. Oxidative stress occurs within a few hours after the start of CMV [9, 56] and may play a central role in the pathogenesis of CMV-induced diaphragmatic atrophy [7]. Oxidized proteins are associated with increased proteolysis, which generates muscle atrophy and dysfunction [57, 58]. Because PSV does not increase proteolysis (contrary to CMV) or decrease protein synthesis, it is likely that PSV causes less oxidative injury. Our results confirm that CMV is associated with diaphragmatic oxidative stress as indicated by an increase in protein myofibrillar oxidation. The increase in protein carbonyl levels parallels the increase in 20S proteasome activity, which specializes in degrading proteins oxidized by reactive oxygen species [7, 59]. Thus, oxidized proteins may generate an increase in 20S proteasome activity. Contrary to our hypothesis, we observed a similar oxidation of myofibrillar protein with PSV. Thus, even if MV causes oxidative stress, our findings support the hypothesis that protein oxidation probably does not trigger the diaphragmatic proteolytic damage generated by CMV and its associated diaphragmatic dysfunction. Nevertheless, an overproduction of free radicals may constitute the molecular signal of CMV-increased proteolysis, either in mitochondria (as suggested by an increase in manganese-superoxide dismutase activity [9]) or via other metabolic pathways (such as that involving xanthine oxidase [12]). There is also the possibility that other diaphragmatic regulating factors (such as apoptosis) might be involved [60].