Background

Positive end-expiratory airway pressure (PEEP) is a potent component of management for patients receiving mechanical ventilation (MV). For critically ill patients, PEEP improves gas exchange, increase end expiratory lung volume (EELV) and improves pulmonary homogeneity, improves clinical outcomes including mortality [1, 2]. However, recently, evidence has shown that the prolonged application of PEEP during mechanical ventilation may cause diaphragm remodeling, especially longitudinal muscle fiber atrophy [3]. PEEP may lead to the development of ventilation-induced diaphragm dysfunction (VIDD), making it difficult for patients to be weaned from MV.

Early studies had a particular focus on the effect of ventilator assistance (assisted or controlled ventilation) on diaphragm dysfunction. However, little attention has been given to the role of PEEP in this process. A variety of studies have shown that lower tidal volume (VT) and higher PEEP can play a lung protective role in mechanical ventilation [4,5,Western blot analysis

The diaphragm protein lysates were subjected to SDS‒PAGE gels for electrophoresis for 90 min and then transferred onto a polyvinylidene fluoride membrane (IPVH00010, Millipore, United States). The membrane was blocked and then incubated with primary antibodies against TGF-β1 (HA500496 1:1000, HuaAn Biotechnology, China), α-SMA (ER1003 1:1000, HuaAn Biotechnology, China), COL3 (HA720050 1:1000, HuaAn Biotechnology, China), COL1A1 (ET1609-68, 1:1000, HuaAn Biotechnology, China), COL6A1 (A9236, 1:1000, ABclonal, Woburn, MA, USA), and COL6A2 (A3796, 1:1000, ABclonal, Woburn, MA, USA) overnight at 4 ℃ and secondary antibodies at room temperature for 90 min and then incubated with ECL reagent. The final reported data of COL1A1, COL3, COL6A1, COL6A2, TGF-β1 and α-SMA were the band densities normalized to that of GAPDH.

RNA-seq analysis

RNA-seq was performed with 2 × 150 bp paired-end sequencing (PE150) on an Illumina Novaseq™ 6000 (LC-Bio Technology Co., Ltd., Hangzhou, China) following the protocol. Then, analysis of significant differences, GO enrichment and KEGG enrichment analyses were performed on the differentially expressed mRNAs.

Statistical analysis

Comparisons between of three groups were conducted by one-way analysis of variance (ANOVA) and nonparametric tests, and the values are presented as the means ± SDs. Statistical significance was considered when the p value was less than 0.05 by SPSS 19.0 statistical software.

Results

Physiological measurements and respiratory monitoring during model establishment

The initial body weights of the CON group, the MV group and the MV + PEEP group were 2.52 ± 0.09 kg, 2.56 ± 0.05 kg and 2.43 ± 0.02 kg, respectively. The respiratory parameters were monitored throughout (Fig. 1B–H and Additional file 1: Table S2), which indicated that the respiratory strategy we carried out provided sufficient ventilation support and maintained a good breathing pattern for rabbits. Moreover, the airway peak pressure (Ppeak) and mechanical power (MP) in the MV group were significantly lower than that in the MV + PEEP group (Ppeak: 11.52 ± 1.24 cmH2O and 15.88 ± 0.85 cmH2O, respectively, p < 0.001 MP: 0.66 ± 0.09 J/min and 0.99 ± 0.05 J/min, respectively, p < 0.001).

As shown in Table 1, none of the blood gas parameters at the end of the experiment were significantly different among groups. We found that the results of SBT between the MV group and the MV + PEEP group were not significantly different after mechanical ventilation (Table 2). After spontaneous breathing trial, the parameters of PETCO2 between the MV group and the MV + PEEP group were 29.50 ± 7.74 mmHg and 31.33 ± 4.59 mmHg, respectively.

Table 1 Body weight and blood gas analysis in three groups
Full size table
Table 2 SBT parameters in MV group and MV + PEEP group
Full size table

Diaphragm ultrasound can reveal the frequency of motion and the location of the diaphragm (Fig. 1A). When we applied 8 cm H2O PEEP to the diaphragm, the location of the diaphragm was altered and maintained in the new position under the persistent mechanical forces of PEEP. We detected the diaphragm excursion of these three groups before mechanical ventilation, after 48 h of mechanical ventilation and after SBT, as shown in Table 3. We found that mechanical ventilation with or without PEEP application had little influence on diaphragm excursion.

Table 3 Diaphragm excursion measured by ultrasound
Full size table

PEEP application during mechanical ventilation leads to extracellular matrix alteration and collagen deposition

According to the RNA-seq results, we identified a total of 665 differentially regulated genes among 29,076 genes between the MV group and the MV + PEEP group. Among these, the levels of 566 genes were significantly upregulated, and 99 were significantly downregulated (Fig. 2B). Interestingly, Gene Ontology (GO) analysis of differentially expressed genes showed that among the top 10 significantly enriched GO terms, 4 GO terms were associated with extracellular matrix and fibrosis (Fig. 2A). This finding indicated that the application of PEEP during mechanical ventilation may alter the extracellular matrix of the diaphragm. By further probing our data, we found some differentially regulated genes associated with collagen, as shown in Fig. 2B. Moreover, the mRNA expression levels of these significantly regulated genes were further verified to be increased in the MV + PEEP group compared to the MV group (COL 1A: 1.20 ± 0.26 and 3.68 ± 0.85, respectively, p = 0.002; COL 1A1: 1.00 ± 0.41 and 4.70 ± 1.20, respectively, p = 0.001; COL 1A2: 0.79 ± 0.36 and 2.78 ± 0.86, respectively, p = 0.008; COL 3A1: 1.56 ± 0.57 and 7.414 ± 1.60, respectively, p = 0.001; COL 5A3: 1.83 ± 0.97 and 4.40 ± 0.88, respectively, p = 0.008; COL 6A1: 1.29 ± 0.56 and 2.68 ± 0.12, respectively, p = 0.023; COL 6A2: 1.64 ± 0.86 and 4.53 ± 1.61, respectively, p = 0.014; COL 15A1: 1.28 ± 0.38 and 2.14 ± 0.53, respectively, p = 0.032; COL 16A1: 0.81 ± 0.12 and 2.44 ± 1.09, respectively, p = 0.026) (Fig. 2C–K). These results indicated that the application of PEEP during mechanical ventilation may lead to extracellular matrix alterations and collagen deposition.

Fig. 2
figure 2

Fibrosis participates in ventilation-induced diaphragm dysfunction. A Bubble chart of the top 10 categories for GO enrichment of the MV group and the PEEP group (n = 3). B Volcano plot of differential gene expression of the MV group and the PEEP group (n = 3). C–K The diaphragm of the two groups was subjected to qPCR to analyze different type of collagen expression. Values are represented as the mean ± SD (*p < 0.05, **p < 0.01, ***p < 0.001)

Full size image

PEEP application during mechanical ventilation contributes to diaphragm fibrosis

To further investigate the influence of extracellular matrix alteration and collagen deposition on the diaphragm, we carried out Masson’s trichrome staining to identify collagen fiber accumulation. Masson’s trichrome can stain collagen fibers of the diaphragm blue and muscle fibers red. As shown in Fig. 3A and B, the collagen fiber staining in the MV + PEEP group was stronger than that in the MV group (1.60 ± 0.37 and 4.45 ± 1.76, respectively, p = 0.001). Additionally, we found that other extracellular matrix and collagen-related mRNAs were upregulated in the MV + PEEP group compared to the MV group (FN: 1.13 ± 0.89 and 4.49 ± 3.28, respectively, p = 0.05; THBS1: 0.67 ± 0.32 and 2.76 ± 1.20, respectively, p = 0.006; THBS3: 1.12 ± 0.92 and 2.76 ± 1.68, respectively, p = 0.05) (Fig. 3C–E).

Fig. 3
figure 3

Application of PEEP during mechanical ventilation leads to diaphragm fibrosis. A, B Masson’s trichrome staining revealed fibrosis in the diaphragms of the MV + PEEP group. Scale bar = 50 μm. C–E Diaphragms of three groups were detected by qPCR to analyze fibrosis-related gene expression. Values are represented as the mean ± SD, n = 6 (*p < 0.05, **p < 0.01, ***p < 0.001)

Full size image

We next investigated the alteration of different collagen types. Sirius red staining is another standard method for evaluating collagen fibers in tissues. The complex of collagen and Sirius red is much more birefringent than that of other proteins; thus, it appears brighter than the other tissues with polarized microscopy [14]. Moreover, collagen type I fibers can appear yellowish-orange to red, while collagen type III fibers appear green to yellowish-green with polarized microscopy [15]. We found that regardless of the total collagen fibers detected by the bright field microscope or the collagen type I and III fibers visualized by the polarized microscope, the MV + PEEP group presented more positive staining than the MV group (bright field microscope: 3.33 ± 0.72 and 7.35 ± 0.75, respectively, p = 0.001; polarized microscope: 1.80 ± 0.32 and 3.42 ± 0.28, respectively, p = 0.001) (Fig. 4A–C). Similar results were found by western blotting: the protein expression levels of collagen types 1A1, III, 6A1 and 6A2 were significantly increased in the MV + PEEP group compared to the MV group (COL 1A1: 1.18 ± 0.16 and 1.88 ± 0.20, respectively, p = 0.05; COL III: 1.75 ± 0.52 and 3.76 ± 0.07, respectively, p = 0.001; COL 6A1: 1.70 ± 0.21 and 4.93 ± 0.80, respectively, p = 0.001; COL 6A2: 1.24 ± 0.06 and 2.06 ± 0.34, respectively, p = 0.04) (Fig. 4D–H). We also found that compared to that of the MV group, the positive immunofluorescence of COL III in the MV + PEEP group was stronger (1.13 ± 0.37 and 2.73 ± 0.83, respectively, p = 0.003) (Fig. 4I, J). Taken together, our present study found that the application of PEEP during mechanical ventilation contributes to collagen fiber accumulation, indicating fibrosis in the diaphragm.

Fig. 4
figure 4

Application of PEEP during mechanical ventilation leads to collagen deposition in the diaphragm. A–C Sirius red staining with bright field and polarized microscopy revealed increased collagen fibers in the diaphragms of the MV + PEEP group. Scale bar = 50 μm. D–H Diaphragm protein lysates of the three groups were detected by Western blotting, and the expression of different collagen types was quantified. I, J Diaphragmatic tissue sections were subjected to immunofluorescence staining to analyze the expression of collagen III. Red represents collagen III; green represents WGA; blue represents Hoechst. Scale bar = 50 μm. Values are represented as the mean ± SD. n = 6 (*p < 0.05, **p < 0.01, ***p < 0.001)

Full size image

PEEP application during mechanical ventilation induced fibrosis in the diaphragm is associating with TGFβ-1 upregulation

TGFβ-1 is known as one of the most effective fibrogenic factors, playing an important role in the expression of collagen fibers. We found that the expression of TGFβ-1 at both the mRNA and protein levels was upregulated in the MV + PEEP group (mRNA: 1.35 ± 0.50 and 2.42 ± 0.77, respectively, p = 0.03; protein: 1.20 ± 0.23 and 2.30 ± 0.65, respectively, p = 0.04) (Fig. 5A–C). Additionally, we found that the expression of alpha smooth muscle actin (α-SMA) in the MV + PEEP group was significantly upregulated (1.06 ± 0.46 and 2.15 ± 0.37, respectively, p = 0.03) (Fig. 5A, D). In summary, the upregulation of TGFβ-1 in mechanical ventilation with PEEP application may be the potential mechanism of fibrosis in the MV + PEEP group.

Fig. 5
figure 5

PEEP application during mechanical ventilation induced fibrosis in the diaphragm is associating with TGFβ-1 and α-SMA upregulation. A–D Diaphragm protein lysates of the three groups were detected by Western blotting, and the expression of TGFβ-1 and α-SMA were quantified. Diaphragms of the three groups were detected by qPCR to analyze fibrosis-related gene expression. Values are represented as the mean ± SD. n = 6 (*p < 0.05)

Full size image

Discussion

PEEP is a common and important method in acute hypoxic respiratory failure. Most acute respiratory distress syndrome (ARDS) patients who receive mechanical ventilation will receive PEEP of 5 to 12 cm H2O in combination with lung-protective ventilation modes to ameliorate oxygenation and prevent atelectasis [16, 17]. However, it has been demonstrated that after 18 h of MV with 2.5 cm H2O of PEEP, diaphragm fibers adapt to the altered shape by absorbing serially linked sarcomeres, which is termed longitudinal atrophy [8]. Recent research proved that MV with 10 cm H2O of PEEP for 12 h worsens diaphragm atrophy induced by a ventilator in rats by inducing oxidative stress [7]. However, the influence of PEEP application is still controversial. Sassoon et al. found that 48 h of MV with 8 cm H2O of PEEP did not exacerbate diaphragm dysfunction in rabbits [18]. The potential explanation is that the profound effects of CMV on the diaphragm made the additional influence of PEEP undetectable. It has even been reported that high levels of PEEP application preserved diaphragm contractility in a rat ARDS model.

In our study, we adopted PEEP of 8 cm H2O which has been confirmed that 8 cm H2O PEEP will not cause pulmonary overdistension or circulatory dysfunction [18]; In addition, the ventilation mode of volume assist/control preserves the spontaneous breathing effort of the experimental rabbits, which is closer to clinical application and protects the diaphragm function to a certain extent [19]. However, this may contribute to the patient-ventilator asynchrony which may also have impact on diaphragm. We made attempts to explore the role of patient-ventilator asynchrony on diaphragm dysfunction as shown in Additional file 1: Table S3. Based on current data, patient-ventilator asynchrony index between the MV group and the MV + PEEP group is not significant different. Interestingly, we found that mechanical ventilation with 8 cm H2O PEEP application led to fibrosis in the diaphragm in healthy rabbits by upregulating the expression of TGF-β1, which may be a potential cause of aggravating diaphragm dysfunction.

Fibrosis refers to the accumulation of ECM. Although ECM accounts for 10% of skeletal muscle mass and plays a major role in force transmission, maintenance, and muscle fiber repair [20], the abnormal accumulation of ECM, especially collagens, impairs muscle function and regeneration after injury [12, 21, 22]. Brass et al. demonstrated that high-fat diet (HFD) feeding promotes the role of thrombospondin 1 (THBS1) in obesity-related respiratory dysfunction by increasing FAP-mediated fibrogenesis and promoting fibrotic remodeling of the diaphragm [23]. A study of Duchenne muscular dystrophy (DMD) identified changes in the ECM structure and mechanics of the diaphragm during disease progression, and the role of collagen tissue in diaphragm function should be further investigated [11]. A previous study showed that a diaphragmatic injury model induced by high tidal volume ventilation in mice results in diaphragm dysfunction, which is associated with the activation of fibrosis-relevant proteins such as type I and III procollagen and TGF-β1. Mechanical stretch was considered to be the reason for this high tidal volume ventilation-induced fibrosis in the diaphragm [24]. Similarly, in our study, we found that PEEP application led to collagen deposition and fibrosis in a mechanical ventilated diaphragm. This finding may be due to the altered location of the diaphragm that we detected with diaphragm ultrasound (Fig. 1A). During the process of 48 h of mechanical ventilation, the diaphragm suffered an 8 cm H2O mechanical force. Persistent mechanical stretching might account for the fibrotic activation observed in our present research.

As an important regulator of ECM accumulation, TGF-β1 is considered to be a key driver in the development of fibrosis [25]. Several studies have demonstrated that TGF- β1 signaling is critical in the progression of several diseases, such as pulmonary fibrosis, cardiac fibrosis and liver fibrosis [26,27,28]. In the diaphragm of an MDX mouse model, TGF-β1 levels were significantly upregulated, accompanied by an increase in nonmuscle tissue [29]. Baptiste et al. found that a TGF-β pathway inhibitor alleviates diaphragmatic contraction dysfunction induced by sepsis [30]. In addition, TGF-β, especially TGF-β1, and its downstream signaling pathway are potent regulators of α-SMA gene expression during damage repair [31, 32]. Cumulative α-SMA expression is considered to be a classic hallmarks of myofibroblast differentiation, representing mature myofibroblasts [33]. Persistent activation of myofibroblasts leads to accumulation and contraction of collagenous ECM and eventually contributes to the development of fibrosis. In our present study, we found that 48 h of mechanical ventilation with 8 cm H2O PEEP application can significantly increase the expression of TGF-β1 accompanied by α-SMA in the diaphragms of healthy rabbits. This study may provide a potential therapeutic target to guide clinical practices preventing PEEP application-induced diaphragm dysfunction.

There were several limitations in this study. First, although fibrosis leads to skeletal muscle weakness, compared to collagen deposition, measurements of muscle contractile properties may better describe diaphragm weakness. Due to the limitation of the experimental conditions, direct measurements of muscle contractile properties were absent. Thus, we carried out SBT test to indirectly reflect the function of diaphragms between the MV group and the MV + PEEP group. However, the negative result of SBT test (as shown in Table 2) didn’t imply that fibrosis has no effect on diaphragmatic function, because the accuracy of SBT test on representing diaphragm function may be not enough to reflect the effect of diaphragm fibrosis on diaphragm function. In fact, the SBT test is not a perfect predictor of respiratory function. In clinical practice, there are also cases where SBT test passed but extubation failed.

Conclusions

Upon comparing the diaphragm between mechanically ventilated rabbits with/without PEEP application, we demonstrated that PEEP application for 48 h in mechanically ventilated rabbits will cause collagen deposition and fibrosis in the diaphragm. Moreover, activation of the TGF-β1 signaling pathway and myofibroblast differentiation may be the potential mechanism of this diaphragmatic fibrosis. These findings might provide novel therapeutic targets for PEEP application-induced diaphragm dysfunction.