Pre-exposure to mechanical ventilation and endotoxemia increases Pseudomonas aeruginosa growth in lung tissue during experimental porcine pneumonia.

BACKGROUND: Immune system suppression during critical care contributes to the risk of acquired bacterial infections with Pseudomonas (P.) aeruginosa. Repeated exposure to endotoxin can attenuate systemic inflammatory cytokine responses. Mechanical ventilation affects the systemic inflammatory response to various stimuli. AIM: To study the effect of pre-exposure to mechanical ventilation with and without endotoxin-induced systemic inflammation on P. aeruginosa growth and wet-to-dry weight measurements on lung tissue and plasma and bronchoalveolar lavage levels of tumor necrosis factor alpha, interleukins 6 and 10.
METHODS: Two groups of pigs were exposed to mechanical ventilation for 24 hours before bacterial inoculation and six h of experimental pneumonia (total experimental time 30 h): A30h+Etx (n = 6, endotoxin 0.063 μg x kg-1 x h-1) and B30h (n = 6, saline). A third group, C6h (n = 8), started the experiment at the bacterial inoculation unexposed to endotoxin or mechanical ventilation (total experimental time 6 h). Bacterial inoculation was performed by tracheal instillation of 1x1011 colony-forming units of P. aeruginosa. Bacterial cultures and wet-to-dry weight ratio analyses were done on lung tissue samples postmortem. Separate group comparisons were done between A30h+Etx vs.B30h (Inflammation) and B30h vs. C6h (Ventilation Time) during the bacterial phase of 6 h.
RESULTS: P. aeruginosa growth was highest in A30h+Etx, and lowest in C6h (Inflammation and Ventilation Time both p<0.05). Lung wet-to-dry weight ratios were highest in A30h+Etx and lowest in B30h (Inflammation p<0.01, Ventilation Time p<0.05). C6h had the highest TNF-α levels in plasma (Ventilation Time p<0.01). No differences in bronchoalveolar lavage variables between the groups were observed.
CONCLUSIONS: Mechanical ventilation and systemic inflammation before the onset of pneumonia increase the growth of P. aeruginosa in lung tissue. The attenuated growth of P. aeruginosa in the non-pre-exposed animals (C6h) was associated with a higher systemic TNF-α production elicited from the bacterial challenge.

Introduction

During critical illness, and especially during mechanical ventilation (MV), physical and immunological alterations occur that invite the high probability for infections with bacteria such as Pseudomonas (P.) aeruginosa. The attributable mortality from ventilator-associated pneumonia with P. aeruginosa is high. Moreover, treatment is increasingly difficult because of multidrug resistance and costs escalate from the prolonged length of stay [1]. The responsiveness and functionality of the individual immune system are central to the risk of developing an infection in general and to the onset of intensive care-related infections in particular [2]. Attempts have been made to describe the stage of reactivity of the immune system during sepsis based on clinical description or phenotype [3,4]. The ultimate goal is to tailor treatment to meet the clinical needs of patients based on the stage of immune reactivity. Concerning bacterial infections, any means to reduce inflammation-related organ damage while optimizing bacterial clearance is of utmost interest. We know from a previous experiment that twenty-four hours (h) of preceding MV combined with exposure to endotoxin attenuates the inflammatory cytokine response to a subsequent challenge of endotoxin, a phenomenon known as endotoxin tolerance (ET) [5]. Additionally, relatively small ventilatory interventions in healthy lungs, i.e. reduced tidal volumes from 10 to 6 mL x kg-1 and elevated positive end-expiratory pressure (PEEP) from 5 to 10 cmH2O, attenuate systemic and organ-specific inflammation and, most importantly, reduce P. aeruginosa burden in lung tissue in experimental pneumonia [6-8]. To our knowledge, scientific studies investigating weakened immune system responses to in vivo bacterial growth have not been performed in humans and are scarce in large animal models. As an experimental model animal, the size and anatomical features of the pig enable the use of machines and surveillance techniques relevant to intensive care. Additionally, the pig shares inflammatory traits with humans to the extent that it is suitable for pneumonia models [9]. The current experiment expands on the questions of inflammatory responsiveness and bacterial burden development after experimental intensive care. The underlying general hypothesis is that differences in inflammatory cytokine responsiveness, generated by any modality, correlate to differences in bacterial growth in lung tissue. Specifically, we hypothesized that two separate, clinically relevant entities, systemic inflammation and MV, would affect immune system reactivity and ultimately bacterial growth in experimental pneumonia. The primary aim was to investigate whether the bacterial growth in lung tissue six h after a bacterial challenge would be affected by a) a preceding inflammatory event, i.e. endotoxemia for 24 h and b) MV for 24 h. These two separate parts of the current experiment are referred to as Inflammation (comparing two 30 h groups) and Ventilation Time (comparing one 30 h and one 6 h group). Secondary aims were to investigate the development of lung injury and cytokine responses in plasma and bronchoalveolar lavage (BAL).

Materials and methods

The experiment contained three groups of healthy pigs of both sexes, between 9 and 12 weeks old and sexually immature (Fig 1). Two groups, A30h+Etx (n = 6) and B30h (n = 6), were studied for 30 h. The third group, C6h (n = 8), was derived from a previously published experiment and studied for 6 h [8]. Notably, the experiments were conducted simultaneously in time from the same animal batches. The animals were allocated to either experiment and experimental group by block randomization. The same methods and protocol were used for all groups. One animal in the 30-hour experiment and two in the six-hour experiment served as sham controls for a reference to normality and appreciation of the inflammatory model. The sham animals were not given endotoxin or a bacterial challenge but were in all other respects treated according to the protocol. A total of 24 animals were used in the production of the study. Additional information on methodology has been previously published [7] and a more detailed account is provided in the S2 File.

Fig 1

Experimental overview.

Experimental parts Inflammation A30h+Etx vs. B30h, Ventilation Time B30h vs. C6h, Etx (endotoxin), dotted line indicates preparatory surgery of approximately 1 h, Pseudomonas (P.) aeruginosa, CFU (colony forming unit), the bacterial phase is between 0–6 h, n (number of animals per group).

Ethics statement

The animals were handled as per the regulations of the Swedish Board of Agriculture. The Animal Research Ethics Board of Uppsala approved and issued the permit for the current experiment (Uppsala djurförsöksetiska nämnd, DNr C 250/11). The study was designed in compliance with the Minimum Quality Threshold in Pre-clinical Sepsis Studies (MQTiPSS) guidelines and reported in adherence to the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines [10]. The animals (Swedish farm pig) were acquired from a private source, Mångsbo Gård breeding facility, Uppsala, Sweden. In total 23 animals were used in the current experiment. The animals were allowed to eat and drink ad libitum up to 1 h before the start of the experiment. The pigs were sedated with tiletamin 3 milligrams (mg) x kilogram (kg)-1, zolazepam 3 mg x kg-1, and xylacin 2.2 mg x kg-1. Morphine 20 mg and ketamine 100 mg were given in an auricular vein. Anesthesia was maintained with pentobarbital 8 mg x kg-1 x h-1 and morphine 0.26 mg x kg-1 x h-1. To facilitate ventilator management and counteract shivering and coughing muscle relaxation was maintained with an infusion of rocuronium at an initial rate of 2 mg x kg-1 x h-1.Immediately after the experimental endpoint, the animals were euthanized by an intravenous injection of potassium chloride and mechanical ventilation was withdrawn.

Anesthesia and surgical procedure

Briefly, the anesthetized animals were tracheotomized and catheters were placed in a cervical artery, a central vein, the pulmonary artery, the portal vein by route of the splenic vein and the bladder.

Protocol

The P. aeruginosa pneumonia phase made up the last 6 h in all groups and initial ventilator settings were the same in all groups. During the first 24 h in groups A30h+Etx and B30h, the pigs were ventilated in a lateral position with 180° position changes every 6 h. During the last 6 h, the animals from all groups were in the supine position. During the preparatory surgery, a bolus of 20 mL x kg-1 of Ringer´s acetate was administered. Group A30h+Etx received an endotoxin infusion of 0.063 μg x kg-1 x h-1 for 24 h in an auricular vein (Escherichia (E.) coli: 0111:B4 (Sigma Chemical Co., St Louis, MO, USA)) while group B30h received an equivalent amount of saline 0.9%. After 24 h, an intra-tracheal inoculum of 1x1011 colony forming units (CFU) of P. aeruginosa type O3 was given. Group C6h received no pre-exposure to ventilation but started the experimental protocol after preparatory surgery at the time of bacterial inoculation.

Interventions

Initially, tidal volume was 10 mL x kg-1, PEEP 5 cm H2O, respiratory rate 25 x min-1 and inspired oxygen fraction (FiO2) 0.3. The respiration was adjusted to meet an arterial partial pressure of carbon dioxide (PaCO2) value from 4.5–6.5 kilo pascal (kPa) by an increment or decrement in the respiratory frequency of 10%. Predefined increments of FiO2 (0.3–0.6–0.8–1.0) were performed at arterial partial pressure of oxygen (PaO2) values below 10 kPa simultaneously with changes in predefined PEEP levels (5-8-10-14 cm H2O). If the plateau pressure were over 30 cm H2O, the tidal volume was reduced to 7 mL x kg-1 and the inspiratory to expiratory ratio was changed from 1:2 to 1:1. The alveolar recruitment maneuver (ARM) consisted of stepwise increments of PEEP (sequential 3 cm H2O increments for 5 seconds (s) each under control of systolic arterial blood pressure) until the peak pressure reached 35 cm H2O, followed by prolonged inspiration for 10 s. ARM was performed at the start of the protocol (-24 h) and after each change of position in the 30 h groups, as well as after bacterial inoculation at 0 h in all groups. During the first 90 minutes (min) of the experiment, norepinephrine was used in boluses of 40 μg if mean arterial pressure (MAP) equaled mean pulmonary arterial pressure (MPAP). A MAP value, regardless of MPAP, that was below 60 mmHg after 90 min was treated with a bolus of Ringer´s acetate 15 mL x kg-1, a 1 mL bolus of norepinephrine 20 μg x mL-1, followed by a norepinephrine infusion of the same concentration starting at 5 mL x h-1. At relapse of MAP below 60 mmHg, the infusion dose was doubled.

Bacterial inoculum

The inoculation dose of P. aeruginosa, O-antigen serotyped to O3 by a slide agglutination test with commercial antisera (Bio-Rad Laboratories AB, Solna, Sweden) at the section for Clinical Microbiology and Infectious Medicine (Uppsala, Sweden), was a suspension of 1x1011 CFU dissolved in Lysogeny Broth according to Miller (VWR, Leuven, Belgium) at a total volume of 20 mL. The concentration of the suspension was estimated from an optical density reading and afterwards verified from overnight cultures. Before the start of the study protocol at 0 h, a suction catheter was inserted blindly into the tracheal tube until it reached mechanical resistance. After a BAL was secured for culture and laboratory analyses the bacterial suspension was injected blindly into the lungs via the same catheter.

Measurements and bacterial cultures

Physiological measurements and handling of laboratory samples followed earlier described sequences [7]. Six lung tissue samples were taken from each animal at the experimental endpoint. Three samples were from dorsal locations in the right upper, middle and lower lobes and three from the corresponding levels in the left lung (total of six samples). Bacterial cultures were performed after sequential dilution with saline 0.9% on Cystine-Lactose-Electrolyte-Deficient agar (BD Diagnostics, Stockholm, Sweden) plates kept at 37° C overnight. Six lung tissue samples of approximately 1 g were homogenized with 3 mL of saline 0.9% for 4 min with a Stomacher 80 Biomaster (Seward, Worthing, UK) before the single line sequential dilution. Six larger tissue samples of approximately 10 g were weighed before and after desiccation in 60° C for 12 h. BALs were performed directly before the start of the experiment at -24 h in the 30 h groups, in all groups at 0 h and before the experimental endpoint at 6 h. BAL was performed as a blind bronchial sampling with a suction catheter through the endotracheal tube with 20 mL of saline 0.9%. The BAL fluid was used for bacterial cultures after sequential dilution and cytokine measurements. Commercial porcine-specific sandwich enzyme-linked immunosorbent assay (ELISA) was used to determine tumor necrosis factor alpha (TNF-α), interleukin (IL) 6 and IL10 in plasma and BAL (only TNF-α and IL6), (DY690B(TNF-α) and DY686 (IL-6), R&D Systems, Minneapolis, MN, USA and KSC0102 (IL-10), Invitrogen, Camarillo, CA, USA). The lower detection limits in EDTA plasma were <230 pg × mL−1 for TNF-α, <60 pg × mL−1 for IL6 and <60 pg × mL−1 for IL10. All ELISAs had intra-assay coefficients of variation (CV) of <5% and a total CV of <10%. After enzymatic conversion of nitrate to nitrite by nitrate reductase, total nitrite concentration in urine and BAL at -24, 0 and 6 h was measured using the Parameter™ assay (SKGE001, R&D Systems, Minneapolis, MN, USA). The urine samples were diluted 1:5 before the assay according to the recommendations of the manufacturer. Plasma and BAL urea (reagent: 7D75-21) and albumin (reagent: 7D54-21, BCP method) were measured on a Mindray BS380 chemistry analyzer (Mindray, Shenzhen, China) with the reagents obtained from Abbott laboratories (Abbott Park, IL, USA). The total coefficient of variations (CV) were 2% at 10 mmol/L for the urea method and 1% at 30 g/L for the albumin method.

Statistics

The animals were allocated to groups of 6–8 animals by block randomization simultaneously with the previously published experiment [8]. Pro-inflammatory cytokine peaks during the first 24 h were anticipated in the two 30 h groups (A30h+Etx and B30h). These differences in relation the 6 h group (C6h)—with no pre-exposure before the bacterial challenge—constituted part of the constructed group separating inflammatory state characteristics. Therefore, comparative group statistics in the experimental parts Inflammation (A30h+Etx vs. B30h) and Ventilation Time (B30h vs. C6h) were based on data solely from the last 6 h of the experiment (the bacterial phase). No multigroup comparisons including all three groups were used in the experiment. A general linear model (GLM) was used for group comparisons in the lung tissue sample variables (i.e. bacterial growth and wet-to-dry weight ratio). Random effects were introduced into the model to account for the within-subject dependencies of the six simultaneous lung tissue samples from each individual, making the GLM a mixed model. Because the bacterial inoculum was delivered blindly to either the right or left lung, the tissue samples in each animal were statistically analyzed using three levels (cranial-middle-caudal) consisting of the right and left corresponding samples. Raw data graphical presentations of the lung tissue sample variables related to side and cranio-caudal distributions are presented in the supplementary files, and post hoc comparative statistics was done by ANOVAs (S1 and S2 Figs). Repeated measures were analyzed with analysis of variance (ANOVA) for repeated measures. Only the group factor is presented as a p-value in the results from either the GLM or ANOVA for repeated measures. Inoculated dose and bacterial counts in BAL were analyzed with Mann-Whitney U-test for each experimental part (Inflammation and Ventilation Time) based on non-normal distribution, but data was presented in a logarithmic form for coherence within the presentation. In analogy with earlier publications all cytokines were logarithmically transformed [6-8]. Urinary nitrite was analyzed with Mann-Whitney U-test for each experimental part based on the mean concentration value from three measurements during the bacterial phase. StatisticaTM (Statsoft, Tulsa, OK, version 13) was used for the statistical calculations and control of relevant assumptions. A p-value of < 0.05 was considered significant. A senior statistician approved the statistical design. Sham animals are only presented in the supplementary material as descriptive data. No power calculation was conducted for this specific experiment since we had no previous data on bacterial behavior in our models. Instead, we used the power calculation for the preceding inflammatory experiments [6-8]. It was based on a systemic TNF-alpha difference of 15% at 6 hours, an alpha error of 0,05, a power of 0,8, and an SD of 10%, which yielded six evaluable animals per group. The choice of 8 animals per group in the previously published day-based experiment [8] was based on this calculation while allowing for a slightly larger variability in the bacterial outcome variable. As we started with the day-based experiment we could appreciate the bacterial growth in lung tissue better. Based on this data we reduced the number of animals in the 30 h experiments, which were completed at the end of the experimental period, from eight to six to meet the 3R principle. In summary, we reduced the number of animals as we believed we could meet the required difference in the main outcome variable anyway.

Results

No animal died before the experimental endpoint at 6 h after the bacterial challenge. The animals had a mean weight of 25.2 ± 2 kg with no differences between the groups. The use of norepinephrine and fluid boluses in accordance with the protocol is described in Table 1. The larger dose of norepinephrine in group B30h could be traced to one animal that was circulatory unstable throughout the experiment. No differences were noted in the inoculation dose between any groups (Table 1). Regarding distribution within the lungs of bacterial growth and wet-to-dry ratios there was no difference between right and left sided samples. There was a significant difference in cranio-caudal direction regarding bacterial growth. Raw data of all samples in the experiment is presented in S1 and S2 Figs.

Table 1

Norepinephrine, fluid, inoculation dose.

Variable	Group	a) Pre-exposure	b) Bacterial phase	p
Norepinephrine	A30h+Etx	150(0/400)	200(0/600)
(μg)	B30h	100(0/600)	300(0/1200)	0.94
	C6h	-	0(0/40)	0.28
Fluid bolus	A30h+Etx	1	0
(n)	B30h	1	0	N/A
	C6h	-	0	N/A
Inoculation dose	A30h+Etx	-	11.0(10.9/11.1)
(log10 CFU)	B30h	-	10.8(10.8/11.1)	0.60
	C6h	-	10.9(10.7/11.0)	0.52

Norepinephrine and fluid boluses expressed in separated parts as a) dose during the first 24 h in A30+Etx and B30, and b) dose during the bacterial phase 0–6 h, (3/6 animals in A30+Etx and B30 each and 2/8 in C6 received norepinephrine), fluid boluses of Ringer´s acetate 15 mL x kg-1 expressed in absolute numbers, colony forming unit (CFU), inoculation dose given at 0 h, median(LQ/HQ), Mann-Whitney U-test on the total amounts during the experiment, p upper refers to A30h+Etx vs. B30h, lower to B30h vs. C6h, N/A (not applicable).

Experimental part Inflammation–A30h+Etx compares to B30h

Pseudomonas aeruginosa in lung tissue and bronchoalveolar lavages

The highest bacterial growth was in group A30h+Etx (Fig 2). No differences were detected in bacterial growth between the groups in BAL (Table 2).

Fig 2

Bacterial growth in lung tissue.

P. aeruginosa counts, log10 colony forming units per gram lung tissue. a) Dots represent raw data for each group, mean±SEM (line, brackets), general linear model, * denotes p<0.05, experimental parts Inflammation A30h+Etx vs. B30h, Ventilation Time B30h vs. C6h. b) Descriptive graphical presentation of distribution based on sample level (cranial, middle, caudal) for each group connected by a line, brackets SEM for each level.

Table 2

Pseudomonas aeruginosa and cytokines in bronchoalveolar lavages.

Variable	Group	-24 h	0 h	6 h	p
P. aeruginosa BAL	A30h+Etx	0.0(0.0/0.0)	0.0(0.0/0.0)	4.0(3.8/4.0)
(log10CFU x 100μL-1)	B30h	0.0(0.0/0.0)	0.0(0.0/0.0)	4.4(4.1/5.0)	0.24
	C6h		0.0(0.0/0.0)	3.8(3.3/4.6)	0.40
TNF-α BAL	A30h+Etx	1.6(1.0/2.3)	2.2(2.1/3.0)	3.5(2.6/3.5)
(log10ng x L-1)	B30h	1.5(1.2/1.7)	1.8(1.6/3.0)	3.4(2.7/3.5)	0.93
	C6h		1.8(1.7/1.9)	3.3(2.8/3.6)	0.75
IL6 BAL	A30h+Etx	2.8(1.7/2.9)	2.1(1.7/2.9)	2.3(1.7/2.9)
(log10ng x L-1)	B30h	2.1(2.0/2.6)	2.9(2.7/3.1)	2.8(2.3/3.1)	0.58
	C6h		2.6(1.7/2.9)	2.6(2.0/3.0)	0.65

Pseudomonas (P.) aeruginosa, bronchoalveolar lavage (BAL), colony forming unit (CFU), TNF-α (tumor necrosis factor alpha), IL6 (interleukin 6) at the start of the experiment -24 h (A30h+Etx and B30h), at the bacterial inoculation 0 h and at the end of the experiment 6 h, median(LQ/HQ), p upper refers to A30h+Etx vs. B30h, lower to B30h vs. C6h solely at 6h (no statistical differences between compared groups at -24 or 0 h presented in S1 Table), Mann-Whitney U test, N/A (not applicable).

Physiologic, hypoperfusion and lung injury variables

We found no differences between the groups in the ventilator variables; peak- (P peak), plateau- (P plateau), mean airway pressure (P mean), respiratory rate (RR) and tidal volume (VT); nor were there any differences in the circulatory variables: heart rate (HR), MPAP, pulmonary capillary wedge pressure (PCWP), MAP, cardiac index (CI) and lactate. Moreover, we found no differences in the arterial partial pressure of oxygen to the inspired oxygen fraction ratio (PaO2/FiO2) (Table 3). The wet-to-dry weight ratio was lower in B30h than in A30h+Etx (Fig 3). Additional analyzes of albumin in plasma and BAL were performed to calculate the alveolo-capillary permeability. However, too many albumin values were below detection limit in BAL to make the calculation. The data for albumin is presented in S1 Data.

Table 3

Physiologic variables, hypoperfusion and lung injury.

Variable	Group	-24 h	0 h	3 h	6 h	p
P peak	A30h+Etx	18±3	18±3	26±8	27±6
(cmH2O)	B30h	16±2	18±3	23±7	21±4	0.44
	C6h		18±2	20±2	21±2	0.44
P plateau	A30h+Etx	16±4	16±3	23±7	26±6
(cmH2O)	B30h	15±2	17±3	20±5	20±4	0.32
	C6h		17±2	20±2	21±2	0.27
P mean	A30h+Etx	9±2	9±1	11±3	13±4
(cmH2O)	B30h	9±2	9±1	10±2	10±2	0.35
	C6h		9±1	9±1	9±1	0.06
Respiratory rate	A30h+Etx	25±0	25±0	28±6	31±10
(breath x min-1)	B30h	25±0	25±0	26±1	26±2	0.27
	C6h		24±2	25±2	25±2	0.13
Tidal volume	A30h+Etx	239±54	210±37	185±46	182±50
(mL)	B30h	224±22	216±33	199±33	189±21	0.67
	C6h		248±17	251±21	245±19	<0.01*
HR	A30h+Etx	93±23	95±21	121±32	137±45
(beats x min-1)	B30h	93±19	94±32	121±27	126±21	0.82
	C6h		103±20	94±9	111±12	0.12
MPAP	A30h+Etx	18±3	18±3	25±7	26±4
(mmHg)	B30h	18±1	17±2	25±5	24±5	0.93
	C6h		22±4	29±14	34±14	0.31
PCWP	A30h+Etx	8±2	6±3	8±3	7±2
(mmHg)	B30h	8±2	5±2	7±2	8±1	0.7
	C6h		9±3	9±4	9±4	0.35
CI	A30h+Etx	3.1±1.0	3.2±0.8	3.9±1.3	4.2±1.7
(L x min-1 x m-2)	B30h	2.7±0.5	3.6±1.0	3.9±1.1	3.7±1.0	0.53
	C6h		3.2±0.7	2.5±0.3	2.4±0.8	<0.05*
MAP	A30h+Etx	88±14	81±5	99±8	97±22
(mmHg)	B30h	83±15	75±10	88±10	88±15	0.14
	C6h		93±22	79±15	75±20	<0.05*
Lactate	A30h+Etx	1.9±0.9	0.9±0.3	1.1±0.3	1.1±0.3
(mmol x L-1)	B30h	1.8±0.3	1.1±0.2	1.4±0.3	1.4±0.5	0.18
	C6h		1.8±0.4	1.6±0.7	1.5±0.6	<0.05*
PaO2/FiO2	A30h+Etx	450±44	348±127	244±133	213±125
	B30h	489±38	339±92	270±149	232±147	0.86
	C6h		446±43	378±56	343±61	<0.05*

Pressure (P), heart rate (HR), mean pulmonary arterial pressure (MPAP), pulmonary capillary wedge pressure (PCWP), cardiac index (CI), mean arterial pressure (MAP), arterial partial pressure of oxygen (PaO2), inspired oxygen fraction (FiO2), mean±SD, p upper refers to A30h+Etx vs. B30h, lower to B30h vs. C6h, ANOVA for repeated measures

* denotes p<0.05.

Fig 3

Wet-to-dry weight ratio of lung tissue.

Wet-to-dry ratios of lung tissue. a) Dots represent raw data for each group, mean±SEM (line, brackets), general linear model, experimental parts Inflammation A30h+Etx vs. B30h, Ventilation Time B30h vs. C6h, * denotes p<0.05, ** denotes p<0.001. b) Descriptive graphical presentation of distribution based on sample level (cranial, middle, caudal) for each group connected by a line, brackets SEM for each level.

Inflammatory variables in plasma, bronchoalveolar lavages and urine

The plasma levels of TNF-α displayed early peak values during the pre-exposure in A30h+Etx, but there were no differences to B30h in the last 6 h (Fig 4). IL6 showed biphasic peak values, during the pre-exposure and at the end of the experiment, that did not separate the groups (Table 4 and S3 Fig). No differences were seen in IL10, leukocytes, neutrophils, temperature or platelets between the groups (Table 4). Finally, we found no differences between the groups in TNF-α or IL6 levels in BAL at the end of the experiment (Table 2). Additional analyzes of urea in plasma and BAL were performed to calculate the dilution factor in the BAL samples. However, too many urea values were below detection limit in BAL to make the calculation. The data for urea is presented in S1 Data. At the bacterial inoculation the nitrite concentration in urine was reduced in both groups to around 20% of values at the start of the experiment at -24 h. The mean concentration during the bacterial phase of the experiment, 0 to 6 h, did not separate the groups (Fig 5).

Fig 4

TNF-α in plasma.

Mean±SEM, total group difference calculated between 0–6 h with ANOVA for repeated measures, experimental parts Inflammation A30h+Etx vs. B30h, Ventilation Time B30h vs. C6h, * denotes p<0.01 in Ventilation Time, axis scale changes at 0 hours.

Table 4

Plasma cytokines, inflammatory cells and temperature.

Variable	Group	-24 h	0 h	3 h	6 h	p
IL6	A30h+Etx	1.9±0.6	1.9±0.6	2.4±0.1	2.7±0.3
(log10 ngxL-1)	B30h	1.9±0.6	2.1±0.3	2.5±0.3	2.7±0.5	0.84
	C6h		1.1±0.3	2.4±0.3	3.1±0.3	0.63
IL10	A30h+Etx	1.0±0.9	0.4±0.6	0.3±0.6	0.9±0.6
(log10 ngxL-1)	B30h	0.3±0.7	0.6±0.7	0.3±0.6	1.0±1.0	0.99
	C6h		0.8±0.6	0.7±0.6	0.8±0.7	0.40
Leukocytes	A30h+Etx	13±6	15±6	16±7	23±11
(109 x L-1)	B30h	16±6	19±7	18±6	21±7	0.99
	C6h		14±5	15±6	18±7	0.38
Neutrophils	A30h+Etx	6±5	5±2	8±4	16±9
(109 x L-1)	B30h	8±5	10±8	11±6	15±6	0.61
	C6h		7±3	9±4	11±6	0.61
Temperature	A30h+Etx	38.1±0.4	38.8±2.1	38.8±2.4	39.1±2.1
(°C)	B30h	37.6±0.7	39.7±2.7	39.4±2.8	39.7±3.1	0.65
	C6h		38.4±0.9	38.6±0.9	39.1±1.0	0.86
Platelets	A30h+Etx	385±162	209±128	200±118	200±111
(109 x L-1)	B30h	323±60	218±101	199±91	205±91	0.99
	C6h		377±190	390±165	380±189	<0.05*

Interleukin (IL), mean±SD, p upper refers to A30h+Etx vs. B30h, lower to B30h vs. C6h, ANOVA for repeated measures

* denotes p<0.05.

Fig 5

Total nitrite in urine.

Total nitrite concentration (after conversion of nitrate to nitrite) in urine at -24 h in groups A30h+Etx (dark grey) and B30h (light grey), and in all groups during the bacterial phase 0, 3 and 6 h. Experimental part Inflammation A30h+Etx vs. B30h (p 0.30) and Ventilation Time B30h vs. C6h (white),* denotes p<0.05 in Ventilation Time. Mann-Whitney from mean concentration value in the bacterial phase 0–6 h. Median (HQ/LQ).

Experimental part Ventilation Time–B30h compares to C6h

The higher bacterial growth was in B30h (Fig 2). There were no differences between the groups in BAL (Table 2).

For the ventilator variables (P peak, P plateau, P mean, RR) there were no differences between the groups. VT was higher in C6h than in B30h. In the circulatory variables (HR, MPAP, PCWP) no differences were detected between the groups. C6h displayed lower MAP and CI and higher lactate than B30h. PaO2/FiO2 was higher in C6h than in B30h (Table 3). The wet-to-dry weight ratio was lower in B30h than in C6h (Fig 3).

Inflammatory variables in plasma and bronchoalveolar lavages

The TNF-α levels were higher in C6h than in B30h in plasma (Fig 4). No differences in IL6 (Table 4 and S3 Fig), IL10, leukocytes, neutrophils or temperature were seen between groups. Platelet levels were higher in C6h than in B30h (Table 4). At the end of the experiment, we observed no differences between the groups in TNF-α or IL6 in BAL (Table 2).

Sham animals

For the sham animals (30 h, n = 1 and 6 h, n = 2), all variables are presented in the supplementary material as descriptive statistics (S2 Table).

Discussion

Of the groups, the control group (C6h) had the lowest bacterial growth, the strongest pro-inflammatory cytokine response to the bacterial stimulus and intermediate edema development. The two separate parts of the experiment are discussed in sequence, although the rationale underlying the bacterial outcomes is overlapping.

A30h+Etx differed from B30h by higher bacterial counts in lung tissue and more evident edema formation, but there was no difference in systemic cytokine expression following the bacterial challenge. Group A30h+Etx was exposed to endotoxin for 24 h to model a state of general inflammation before the bacterial challenge. Both groups exhibited moderate IL6 peaks during the first 24 h, but the early TNF-α reaction in A30h+Etx was considerably more pronounced than in B30h. The difference in TNF-α response during the pre-exposure potentially affected the bacterial outcome in several ways. Endotoxin tolerance (ET) was originally described as a blunted fever response to a repeated endotoxemic challenge but has since proved to have multiple effects on the innate and the adaptive immune system toward a general state of reduced responsiveness [11]. The phenomenon is not limited to TNF-α expression but involves multiple cytokines to varying degrees. ET is believed to protect the host from excessive inflammatory reactions but also to bring an increased vulnerability to infections, indicating effects on the cellular defenses [12]. In a porcine experimental sepsis model using intra-venous P. aeruginosa the alveolar macrophages functionality was severely impaired [13]. In murine experiments ET was associated with increased survival in sepsis [14] and reduced P. aeruginosa clearance in lung tissue [15,16]. The relatively mild reactions in the pro-inflammatory cytokines, TNF-α and IL6, in A30h+Etx to the bacterial challenge were on these grounds anticipated in the present experiment, whereas the similar reaction in B30h was not. However, any non-infectious process that elicits a systemic inflammatory response, such as cardiac arrest or trauma, can potentially produce the same reduced cytokine response as endotoxin which could explain the mild cytokine reaction also in B30h. Summarily, the early TNF-α surge in A30h+Etx from endotoxin likely produced a wider array of effects on bacterial growth than could be differentiated from the cytokine responses to the bacterial challenge. Cytokines have pleiotropic effects. For example, the magnitude of pro-inflammation of IL6 or anti-inflammation of IL10 is dependent on the timing of the insult, the type of target cell and any pre-exposure to other cytokines or inflammatory events [17]. The timing of an inflammatory cytokine surge has a determining influence on the host’s handling of a P. aeruginosa challenge. IL4 given 24 h before a challenge in mice reduced both the TNF-α response and lethality. Conversely, simultaneous delivery has led to the opposite effects with enhanced TNF-α and lethality [14]. In another mouse experiment pre-treatment with anti-inflammatory IL10 reduced lung injury and lethality from intra-tracheal P. aeruginosa inoculation, indicating a possible role in bacterial outcome from the balance between pro- and anti-inflammatory cytokines at the time of bacterial challenge [18]. In the current experiment we did not find any significant differences between groups for IL10 or from post hoc comparisons during the bacterial phase as to the ratios of TNF-α/IL10 or IL6/IL10. Thus, other factors than cytokine levels or their respective ratios during the bacterial phase were more capable of exerting an effect on the bacterial outcome of P. aeruginosa.

Given no cardiac failure, edema in lung tissue is indicative of an inflammatory process and has long been associated with bacterial infection [19]. The presence of edema proved to have a negative effect on bacterial clearance in early experimental lung models and correlated with vulnerability to bacterial pneumonia [20]. TNF-α can initiate inflammatory edema in experimental settings [21] and sepsis and general inflammatory states affect the endothelial surfaces throughout the body [22]. Endotoxin in circulation results in shed glycocalyx from the endothelial luminal side of capillaries, thereby promoting fluid leak and edema formation. In human experimental settings this effect can be averted with a TNF-α blocking agent, indicating the direct effect of cytokine surges [23]. From this, we can hypothesize that the presence or escalation of edema during the bacterial phase in the last 6 h of our experiment resulted in impairing the immune system’s ability to reduce bacterial growth. There is a likely effect from endotoxin exposure on cellular defenses in the present experiment. The alveolar macrophage (AM) has been characterized as central to bacterial clearance and is susceptible to inflammatory influence on functionality [24,25]. In several lung damage models, from MV with large VT and zero PEEP to E. coli endotoxin, the resulting inflammation suppressed AM phagocytic and bactericidal action [26-28]. Even severe infections outside the lungs can affect bacterial clearance in the lungs [13]. Experimental E. coli peritonitis, releasing endotoxin and producing TNF-α in the liver, resulted in decreased alveolar neutrophil recruitment, phagocytosis and superoxide production in the lungs [15]. The same result was produced by intravenous endotoxin alone, highlighting the central role of cytokines in eliciting effects from different primary stimuli. The cellular defense functions of the AM are, to a variable degree, dependent on extra-cellular opsonization molecules, such as surfactant proteins from alveolar type II cells. Bacterial endotoxin is one of several agents that has multiple negative effects on surfactant production and function [29]. These changes, especially in surfactant protein-A and -D, lead to lower phagocytic and bactericidal ability from the AM [30]. Additionally, a condition of reduced surfactant production and function leads to atelectasis formation which, on its own, has a negative effect on bacterial clearance in porcine experimental settings [31]. Another point of influence of endotoxin is the complement system. A functional complement system is essential for an effective neutrophil function against Gram-negative bacteria, such as P. aeruginosa [32]. Endotoxin can deactivate complement by forming immune complexes with natural antibodies and so reduce functional levels [33]. Quantitative reductions in complement factor levels or functionality have, in experimental settings, severely hampered the antibacterial defenses of the lung [34]. Although these co-factors for effective cellular handling of bacteria, i.e. surfactant and complement, were not specifically addressed in this experiment, they remain likely contributors to the bacterial outcome.

As an inflammatory adjunctive measure and indication of inflammatory cell activity, we analyzed nitrate concentration in urine to proxy the total nitric oxide turnover during the experimental bacterial phase. Nitric oxide (NO) is an important inflammatory mediator molecule excreted from macrophages, neutrophils and endothelia upon inflammatory stimulus. The dominant part of NO production during an infection is attributed to inducible nitric oxide synthase (iNOS). Bursts of NO in response to inflammatory stimuli have multiple actions affecting direct bactericidal capacity, vascular permeability and scavenging of oxygen free radicals. The resulting effect from NO can differ depending on NO levels and timing during an infection and where NO exerts its dominant effect–in the mitochondrion of the macrophage or as part of the nitrous radical burst aimed at invading pathogens [35]. As the NO molecule is highly unstable measurements are practically directed toward stable metabolites such as nitrite, which is a significant source of NO in the system [36]. In the Inflammation experiment both groups had similarly suppressed nitrite levels in urine to around 20% of base line values after 24 h. The levels did not increase significantly during the bacterial phase in either group (Fig 5). A previous ovine endotoxin tolerance experiment has provided evidence of suppressed nitrite levels after endotoxin exposure and synchronously suppressed NOS activity harmonizing with the picture in the endotoxin exposed animals in group A30h+Etx in the current experiment [37]. The similar nitrite reduction in B30h was unexpected but speaks for a general suppression of inflammatory reactivity from unknown causes mentioned earlier.

Only small, non-significant differences were noted in inflammatory cells, hypoperfusion and physiologic variables. Most remarkable, given the large difference in edema formation, was the lack of difference in PaO2/FiO2. This result indicates a major edema development in A30h+Etx only after the bacterial challenge.

B30h differed from C6h by higher bacterial counts, lower edema formation and lower TNF-α expression from the bacterial challenge. Group B30h was exposed to 24 h of anesthesia and MV before the bacterial challenge as a way to evaluate the potential impact of time in basic intensive care treatment before the bacterial challenge. The 24 h in MV before the bacterial challenge in B30h resulted in differences in the inflammatory response, as well as the physiological state between the groups already at the time of bacterial challenge. These facts provide several plausible explanations for the bacterial outcome.

Group C6h reacted to the bacterial challenge with the full force of an unaffected innate immune system. The unaffected reaction led to an extensive inflammatory reaction with pro-inflammatory cytokines and edema development in lung tissue. Mortality in the early stages of sepsis is often related to profound inflammatory reactions, whereas later mortality relates to multiple organ failure or unresolved infections [38,39]. In C6h there was no previous inflammatory event that could induce ET and no previous cytokine surge that could affect macrophage function. Therefore, the higher TNF-α response in plasma to the bacterial challenge in C6h is an indication of the preserved functionality of the alveolar macrophages, as they are the most considerable source of cytokines in the inflammatory response from the lungs [40]. Additionally, urinary nitrate levels were significantly higher in the unexposed animals of C6h in comparison to the ventilated animals of B30h. The higher nitrite levels in C6h speaks for a higher NO activity from macrophages, neutrophils and endothelia and is a possible mechanistic explanation for both the lower bacterial levels and the greater edema development.

MV can produce inflammation if performed carelessly with cyclic atelectasis and overstretched alveoli, but in recent years the concept of protective ventilation has become the standard [41]. The term protective refers in the broader sense to the avoidance of inflicting iatrogenic harm. In our experiment the respiratory settings were moderate for PEEP and VT, and intermittent recruitments were done to avoid inflammation originating from the ventilator. However, during a general inflammatory stimulus supposedly safe ventilator settings can turn harmful and produce lung injury [42]. The development of lung injury is strongly correlated to immune system reactions with neutrophils in the central position [43]. Inflammatory edema follows the transmigration and activation of neutrophils [44]. The attenuated edema formation in group B30h suggests a relatively lower activation of neutrophils which, in turn, could explain the higher bacterial growth, supposedly associated with relatively lower bactericidal capacity. The edema formation correlates with the TNF-α responses at the bacterial challenge in the two groups. The present experiment was not a lung injury model, but B30h reached PaO2/FiO2 values bordering mild acute respiratory distress syndrome. However, PaO2/FiO2 decreased in B30h during the first 24 h despite intermittent recruitment maneuvers, and the lung parenchyma was significantly drier than in C6h at the end of the experiment. Most likely the lower PaO2/FiO2 in B30h reflected areas of atelectasis that existed at the time of bacterial challenge. Therefore, the anatomical conditions in the lungs differed between the groups from the start of the bacterial challenge. Another factor indicating atelectasis was that the VT was gradually reduced in B30h during the first 24 h and there was a significant difference between the groups throughout the bacterial phase. Atelectasis is a negative factor for bacterial clearance in porcine experimental settings. Thus, inversely, the relatively more open lung in C6h would be a favorable factor for reducing bacterial growth and the total bacterial load in lung tissue [31,45].

Our experiment has several limitations. Perhaps the most central factor for bacterial clearance was not evaluated in the experiment, i.e. the antibacterial capacity of the AM. Additionally, several possible influential co-factors discussed were not measured, namely surfactant, glycocalyx and complement. The experiment was not designed for these outcomes and would have needed a different laboratory set up but is in the plan for future experiments. Our experiment was arranged to resemble an intensive care setting with a minimal amount of inflammatory or potentially organ damaging insults to the research animals. With this configuration we actively chose to avoid repeated and large volume bronchoalveolar lavages after the bacterial inoculation for measurements and harvesting of alveolar macrophages and surfactant. Further, we actively decided against the use of a bronchoscope and chose the minimally invasive method of blind bronchial sampling common to clinical practice. Interventions that potentially could have yielded different results in the BAL variables would probably have influenced the bacterial growth in lung tissue as well as inflammatory variables in the lungs. The BAL variables did not separate the groups in this experiment. Our efforts to enhance the results by additional analyses of urea dilution failed because of too many values below detection in BAL. The same problem was found when analyzing albumin in plasma and BAL for alveolo-capillary permeability measurements as a marker of ling injury. We believe the method of blind bronchial sampling used was too unspecific to yield better data. Lastly, there were no microscopic tissue comparisons regarding lung damage. Based on the different localities of sampling, the relatively few animals per group and problems of representability based on inhomogeneous lungs it was not prioritized in the experiment given that it was deemed an underpowered measurement. The translational relevance of the current experiment lies in that it uses clinically similar intensive care conditions and relevant equipment to evaluate different aspects and factors of influence on bacterial growth in the lungs. The results underline the importance of inflammatory state, i.e. the reactivity of the immune system to a bacterial challenge, for the clinical manifestations yielded by an infection. Additionally, the ease by which we unknowingly can affect immune system reactivity by anesthesia and intensive care calls for further exploration on the cellular and clinical level.

Conclusions

Concomitant pre-exposure to MV and endotoxin increases the growth of P. aeruginosa in lung tissue from a subsequent bacterial challenge. Pre-exposure to only MV increases the bacterial growth similarly but to a lesser extent. In comparison, animals unexposed to MV or endotoxin show suppressed bacterial growth but at the cost of a more pronounced systemic inflammatory reaction with increased edema formation in lung tissue.

Distributions of Pseudomonas aeruginosa growth.

Raw data presentations of bacterial cultures in the experiment based on a) left-right side distribution within each group (post hoc one-way ANOVA between left (L) and right (R) within each group, all p>0.05), b) cranio-caudal distribution regardless of group (post hoc one-way multiple ANOVA p<0.05), mean indicated by line.

(TIF)

Click here for additional data file.

Distributions of wet-to-dry ratios.

Raw data presentations of all wet-to-dry measurements in the experiment based on a) left-right side distribution within each group (post hoc one-way ANOVA between left (L) and right (R) within each group, all p>0.05), b) cranio-caudal distribution regardless of group (post hoc ANOVA p 0.23), mean indicated by line.

(TIF)

Click here for additional data file.

IL6 in plasma.

Mean±SEM, total group difference calculated between 0–6 h with ANOVA for repeated measures. Experimental parts Inflammation A30h+Etx vs. B30h (group p 0.84, group*time 0.49), Ventilation Time B30h vs. C6h (group p 0.63, group*time p<0.05*), axis scale changes at 0 hours. Post hoc statistics (group*time) present differing dynamics in cytokine escalation in C6h from the bacterial challenge 0–6 h.

(TIF)

Click here for additional data file.

Pseudomonas aeruginosa and cytokines in bronchoalveolar lavages–complementary post hoc tests at– 24 and 0 h.

Pseudomonas (P.) aeruginosa, bronchoalveolar lavage (BAL), colony forming unit (CFU), TNF-α (tumor necrosis factor alpha), IL6 (interleukin 6) at the start of the experiment -24 h (A30h+Etx and B30h), at the bacterial inoculation 0 h and at the end of the experiment 6 h, median(LQ/HQ), p upper refers to A30h+Etx vs. B30h, lower to B30h vs. C6h.

(DOCX)

Click here for additional data file.

Physiologic and laboratory variables in the sham animals.

Sham 30h (S30h, n = 1), Sham 6h (S6h, n = 2), heart rate (HR), mean arterial pressure (MAP), mean pulmonary arterial pressure (MPAP), cardiac index (CI), pulmonary capillary wedge pressure (PCWP), pressure (P), wet-to-dry ratio (WD), tumor necrosis factor alpha (TNF-α), interleukin (IL). Sham 30 h (S30, n = 1) absolute value, sham 6 h (S6, n = 2), mean±SD, median (LQ/HQ) as presented in the main manuscript.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

Statement from statistician.

(PDF)

Click here for additional data file.

Supplementary materials and methods.

(DOCX)

Click here for additional data file.

27 Jul 2020

PONE-D-20-18632

Pre-exposure to mechanical ventilation and endotoxemia additively increases Pseudomonas aeruginosa growth in lung tissue during experimental porcine pneumonia

PLOS ONE

Dear Dr. Sperber,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Experimental work expanding our knowledge on lung physiology and pathology through studies into large animal models are highly appreciated. It is accepted that such studies are limited in the numbers of animals that can be included as well as in e.g. immunological parameters that can be analysed. However, as you may see from the detailed comments below, we would greatly appreciate if you could add some additional analyses from the BAL samples collected. Because you did not only collect BAL at the end of the experiment but also at -24h and 0h, respectively, TNF-alpha and IL-6 measurements at the other time points should be included. Such analyses should also comprise cell differentials to get an idea of the changes in immune cell composition in the alveolar space as well as additionally cytokines relevant at earlier time points of the pathologic processes in the lung. Data regarding BAL total protein should also be included. Further, the comments regarding the biostatistical analyses have to be considered. A detailed point-by-point response to the reviewers' comments will be mandatory.

Please submit your revised manuscript by Sep 06 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Heinz Fehrenbach

Academic Editor

PLOS ONE

Additional Editor Comments:

I have a few additional comments expanding on what reviewer #1 (see below) stated.

- In the statistics part you state "Comparative group statistics in the experimental parts Inflammation (A30h+Etx vs. B30h) and Ventilation Time (B30hvs. C6h) were based on data from the last 6 h of the experiment (the bacterial phase). No multigroup comparisons including all three groups were used in the experiment." However, you used group B30h twice in your statitics, which means that you apparently treated the setting as two independent experiments. This is clearly not the case. Although being no biostatistician, I think you have to adjust the level of significanse to the fact that you used group B30h twice.

- In the conclusion part of the abstract you state: "Mechanical ventilation and systemic inflammation before the onset of pneumonia additively increase the growth of P. aeruginosa in lung tissue." I do not think that you are able to demonstrate additivity of the effects with your setting and I also cannot find a discussion of this aspect in the manuscript. Because the abstract should contain only well supported statements, I suggest that you remove this sentence or at least consderably modify it to comply with your results.

- Please translate all terms in the supplemnetal material to English language.

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for including your ethics statement:

'The Animal Research Ethics Board of Uppsala issued the permit for the experiment (Uppsala djurförsöksetiska nämnd, DNr C 250/11).'

(a) Please amend your current ethics statement to include whether the Animal Research Ethics Board of Uppsala specifically approved your study.

(b) Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to " ext-link-type="uri" xlink:type="simple">http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research."

3. At this time, we request that you  please report additional details in your Methods section regarding animal care, as per our editorial guidelines:

(a) Please state the source and number of mice used in the study

(b) Please provide details of animal welfare (e.g., shelter, food, water, environmental enrichment)

(c) Please provide the name and dosage of the specific anaesthetic agent used in your study

(d) Please include the method of euthanasia

Thank you for your attention to these requests.

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: N/A

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Study by Sperber et al.

The study reports on the Pseudomonas aeruginosa growth of lung tissues after pre-exposure to mechanical ventilation with or without endotoxin-induced systemic inflammation in large animal (pig) models.

From interpretation point of view, there were three groups and two comparable parts in current study:

Group #1-A30h+Etx: endotoxin(+) + mechanical ventilation(+) + bacterial inoculation(+)

Group #2-B30h: endotoxin(-) + mechanical ventilation(+) + bacterial inoculation(+)

Group #3-C6h: endotoxin(-) + mechanical ventilation(-) + bacterial inoculation(+)

Comparable part #1-inflammation: A30h+Etx vs B30h

Comparable part #2-ventilation time: B30h vs C6h

The authors sampled lung tissues of three different regions (cranial, middle, and caudal) from both right and left lung of the same animal based on the heterogeneity of lung injury in large animals. Bacterial growth (colony forming unit) and lung edema (wet-to-dry ratio) from lung tissues are assessed as well as cytokines (TNF-α, IL 6 and 10) in bronchoalveolar lavage (BAL) and plasma to measure lung and systemic inflammation. The authors found significant bacterial growth of lung tissues in animals pre-exposed to mechanical ventilation and additionally to systemic inflammation for 24 hours. Lung edema was severe in A30h+Etx, and mild in B30h. The authors also found higher plasma TNF-α in C6h when compared to B30h. No differences of bacterial growth and cytokine levels in BAL between the groups were observed. This is an observational study on an important clinical topic performed in large animals. The methodology in this study is complete, and the statistical analysis appears appropriate. However, no mechanisms were investigated. Also, the causal relationship between each observation (bacterial growth, lung edema and systemic inflammation) seems not clear and needs to be improved.

Major comments:

1) From my understanding that the cytokines TNF-α and IL6 are the early responded factors during inflammation, it is anticipated that these cytokines would decrease or return to the normal levels later, as the results presented in current study. Thus, it's not clear why the authors measured TNF-α and IL6 at the end of current studies (the time periods were approximately 30 hours). Also, the higher levels of TNF-α in C6h might be due to the time frame after bacterial infection, in which time period TNF-α increased. Could the authors clarify these?

2) In terms of cytokines TNF-α and IL6 mainly produced by macrophages, these results in current studies might reflect the functional role for circulatory macrophages. How about the other immune cells, such as neutrophils, as they also play important roles in bacterial infection and clearance? It would be better if the authors could provide some evidence (such as other cytokines) to indicate the functional activity for other immune cells.

3) Due to the infection of P. aeruginosa in lung, I am curious about the local lung injury after pneumonia. In terms of lung damage, the authors measured oxygen index (PaO2/FiO2), wet-to-dry ratio, and cytokines in BAL. There was difference only in wet-to-dry ratio in each comparison. Since the lung tissues and BAL were harvested in current study, could the authors provide more measurements for lung damage, such as protein concentration in BAL, histological lung injury staining, cytokines and surfactants in lung tissues? These results would not only help support local lung injury after bacterial infection, but also might provide some potential explanation for bacterial growth.

4) In discussion of experimental part: A30h+Etx vs B30h, the authors discussed the potential effect of TNF-α peak and lung edema existed in A30h+Etx on impairing the immune system’s ability, which subsequently reduced bacterial growth. The results did not show difference in cytokines of BAL and plasma in A30h+Etx vs B30h at the end of study. Could the authors explain this? It would be more strength if the authors could provide more other evidences for the compromise immunity.

5) Page 26/Line 427-428: "Therefore, the higher TNF-α response to the bacterial challenge in C6h is an indication of the preserved functionality of the alveolar macrophages, ...". I think there would be "the higher plasma TNF-α response...". Could the authors improve the statement? How to explain the results of no difference of TNF-α in BAL between B30h and C6h, based on the preserved functionality of the alveolar macrophages in C6h?

6) Page26/Ling 438-441: "The attenuated edema formation in group B30 suggests a relatively lower activation of neutrophils which, in turn, could explain the higher bacterial growth, supposedly associated with relatively lower bactericidal capacity." In the discussion, the reason for the attenuated edema in B30h was attributed to the relatively lower activation of neutrophils. Could the authors provide the evidence for the lower activity of neutrophils in lung? This would enhance the strength of this statement and support the finding of higher bacterial growth.

Minor comments:

7) Page 9/Line 41 in Abstract: "A third group, C6h (n=8), ...". Please improve the description of the group C6h.

8) Page 9/Line 46-49: the results of abstract. Given there were two comparisons: A30h+Etx vs B30h and B30h vs C6h, it's not clear what comparison the P-values in results indicated. Could the authors describe these most important findings using numerical results (and statistical significance)?

9) Page 13/Line 140-142: "The alveolar recruitment maneuver (ARM) consisted of stepwise increments of PEEP until the peak pressure reached 35 cm H2O, followed by prolonged inspiration for 10 seconds (s)." Could the authors provide the details for the stepwise increments of PEEP (#-# cmH2O)?

10) Page 16/Line 216-220, the authors reported no difference of bacterial growth and wet-to-dry ratios between right and left sided lung. Could the authors improve the presentation of raw data from right and left side (supplementary Figure 1a and 2a) in each group, but not in all groups?

11) The P-value presented in Supplementary Figure 3 was a little bit confused based on only two comparable parts in this study. Could the authors clearly show the P values for each effect: inflammation, ventilation time and their interaction, in the figure legend of in the plot?

12) Page 21/Line 311: "The highest bacterial growth was in B30h (Figure 2)." Based on the comparison between B30h and C6h, perhaps it's better to present the data as: "The higher bacterial growth was in B30h...".

13) Page 26/Line 438-439: "The attenuated edema formation in group B30 suggests a relatively lower activation of neutrophils..." Please correct the group name "B30h".

14) Page 25/Line 411-412: "This result indicates a major edema development in

B30h+Etx only after the bacterial challenge." Here I think the group should be A30h+Etx.

Reviewer #2: The topic of this article is very interesting and relevant. I have just a few comments to make:

1) Although the hypothesis and research questions are well defined, were there any kind of sample size calculation

2) Please report data as mean and SD not as mean and sEM, especially since you have unequal group sizes.

3) I suggest tol elaborate a bit more the clinical / translational meaning of your findings.

4) Please comment why you did the inoculation in a blind way. Given the species it would have been easy to deliver everything in a very standardized way e.g. using fiberoptic bronchoscopy.

5) I suggest to explain also in this papers method section why you did not perform a sample size calculation. In my opinion crossreferencing is not sufficient in this case.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

8 Sep 2020

PONE-D-20-18632

Pre-exposure to mechanical ventilation and endotoxemia additively increases Pseudomonas aeruginosa growth in lung tissue during experimental porcine pneumonia

PLOS ONE

However, as you may see from the detailed comments below, we would greatly appreciate if you could add some additional analyses from the BAL samples collected. Because you did not only collect BAL at the end of the experiment but also at -24h and 0h, respectively, TNF-alpha and IL-6 measurements at the other time points should be included. Such analyses should also comprise cell differentials to get an idea of the changes in immune cell composition in the alveolar space as well as additionally cytokines relevant at earlier time points of the pathologic processes in the lung. Data regarding BAL total protein should also be included.

Further, the comments regarding the biostatistical analyses have to be considered. A detailed point-by-point response to the reviewers' comments will be mandatory.

Please see further for responses to these comments.

Additional Editor Comments:

I have a few additional comments expanding on what reviewer #1 (see below) stated.

- In the statistics part you state "Comparative group statistics in the experimental parts Inflammation (A30h+Etx vs. B30h) and Ventilation Time (B30hvs. C6h) were based on data from the last 6 h of the experiment (the bacterial phase). No multigroup comparisons including all three groups were used in the experiment." However, you used group B30h twice in your statistics, which means that you apparently treated the setting as two independent experiments. This is clearly not the case. Although being no biostatistician, I think you have to adjust the level of significance to the fact that you used group B30h twice.

We do agree with the Editor that the questions regarding the statistical design are very central, and therefore, we have consulted an external senior statistician to critically revise our original approach and to comment. Please see original answer in supplementary file Statement from statistician.

In conclusion, the statistical review concluded that the objection from the editor could be divided into two parts. The two parts of the statistical concerns are commented below;

1. Are the separate experiments independent since they share one group in common? Yes, the three individual groups should be considered independent and the two experiments can therefore be considered separate from a statistical viewpoint.

2. Is a correction of the level of significance warranted? Correction is not warranted on the basis of the statistical status of the two separate experiments. However, correction could be warranted if the number of statistical analyses is high. As the investigation (both parts) is exploratory in nature with variables that biologically influence each other, no correction is deemed necessary.

- In the conclusion part of the abstract you state: "Mechanical ventilation and systemic inflammation before the onset of pneumonia additively increase the growth of P. aeruginosa in lung tissue." I do not think that you are able to demonstrate additivity of the effects with your setting and I also cannot find a discussion of this aspect in the manuscript. Because the abstract should contain only well supported statements, I suggest that you remove this sentence or at least consderably modify it to comply with your results.

It is a valid point. Even if the two entities of mechanical ventilation and endotoxemia yield increasingly higher bacterial growth when added, compared to the unexposed animals, we have not proved it to be additively so. We have taken the word “additively” out of the title and the abstract.

- Please translate all terms in the supplemental material to English language.

The Swedish words in the Supplementary data file have been translated into English.

2. Thank you for including your ethics statement:

'The Animal Research Ethics Board of Uppsala issued the permit for the experiment (Uppsala djurförsöksetiska nämnd, DNr C 250/11).'

(a) Please amend your current ethics statement to include whether the Animal Research Ethics Board of Uppsala specifically approved your study.

We have corrected the wording to approved and issued the permit for the current experiment (Ethical statement).

(b) Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

At this time, we request that you please report additional details in your Methods section regarding animal care, as per our editorial guidelines:

(a) Please state the source and number of mice used in the study

The animals (Swedish farm pig) were acquired from a private source, Mångsbo Gård breeding facility, Uppsala, Sweden. In total 23 animals were used in the current experiment. (Added to the Ethics statement.)

(b) Please provide details of animal welfare (e.g., shelter, food, water, environmental enrichment)

The animals were allowed to eat and drink ad libitum up to 1 h before the start of the experiment.

(c) Please provide the name and dosage of the specific anaesthetic agent used in your study

The pigs were sedated with tiletamin 3 milligrams (mg) x kilogram (kg)-1, zolazepam 3 mg x kg-1, and xylacin 2.2 mg x kg-1. Morphine 20 mg and ketamine 100 mg were given in an auricular vein. Anesthesia was maintained with pentobarbital 8 mg x kg-1 x h-1 and morphine 0.26 mg x kg-1 x h-1. To facilitate ventilator management and counteract shivering and coughing muscle relaxation was maintained with an infusion of rocuronium at an initial rate of 2 mg x kg-1 x h-1.

(Added to the Ethics statement.)

(d) Please include the method of euthanasia

Immediately after the experimental endpoint, the animals were euthanized by an intravenous injection of potassium chloride and mechanical ventilation was withdrawn.

(Added to the Ethics statement.)

Comments to the Author

5. Review Comments to the Author

Reviewer #1: Study by Sperber et al.

The study reports on the Pseudomonas aeruginosa growth of lung tissues after pre-exposure to mechanical ventilation with or without endotoxin-induced systemic inflammation in large animal (pig) models.

From interpretation point of view, there were three groups and two comparable parts in current study:

Group #1-A30h+Etx: endotoxin(+) + mechanical ventilation(+) + bacterial inoculation(+)

Group #2-B30h: endotoxin(-) + mechanical ventilation(+) + bacterial inoculation(+)

Group #3-C6h: endotoxin(-) + mechanical ventilation(-) + bacterial inoculation(+)

Comparable part #1-inflammation: A30h+Etx vs B30h

Comparable part #2-ventilation time: B30h vs C6h

The authors sampled lung tissues of three different regions (cranial, middle, and caudal) from both right and left lung of the same animal based on the heterogeneity of lung injury in large animals. Bacterial growth (colony forming unit) and lung edema (wet-to-dry ratio) from lung tissues are assessed as well as cytokines (TNF-α, IL 6 and 10) in bronchoalveolar lavage (BAL) and plasma to measure lung and systemic inflammation. The authors found significant bacterial growth of lung tissues in animals pre-exposed to mechanical ventilation and additionally to systemic inflammation for 24 hours. Lung edema was severe in A30h+Etx, and mild in B30h. The authors also found higher plasma TNF-α in C6h when compared to B30h. No differences of bacterial growth and cytokine levels in BAL between the groups were observed. This is an observational study on an important clinical topic performed in large animals. The methodology in this study is complete, and the statistical analysis appears appropriate. However, no mechanisms were investigated. Also, the causal relationship between each observation (bacterial growth, lung edema and systemic inflammation) seems not clear and needs to be improved.

Major comments:

1) From my understanding that the cytokines TNF-α and IL6 are the early responded factors during inflammation, it is anticipated that these cytokines would decrease or return to the normal levels later, as the results presented in current study. Thus, it's not clear why the authors measured TNF-α and IL6 at the end of current studies (the time periods were approximately 30 hours). Also, the higher levels of TNF-α in C6h might be due to the time frame after bacterial infection, in which time period TNF-α increased. Could the authors clarify these?

Thank you for a valuable comment. The TNF-α and IL6 responses are indeed early responses to inflammatory stimuli and will naturally wane to lower levels after an initial rise. However, the main point of the current experiment (two separate parts) was to relate bacterial growth in lung tissue to the inflammatory response during the bacterial phase which constituted the last six hours in all three groups. The inflammatory responsiveness at the time of an infection is clinically relevant for the patients’ ability to fight pathogens and for the effects of general inflammation, i.e. sepsis. Although these differences in clinical phenotype are known, they are hard to model in studies because of the problems of keeping experimental animals under long term intensive care before experimental interventions. Our experiment is an effort in this direction.

We constructed a way to investigate the influence of inflammatory responsiveness on bacterial growth. It was our hypothesis that prior inflammatory activation as seen from endotoxin infusion would entail a lower response in TNF-α and IL6 when given a secondary inflammatory challenge in the form of bacteria. Further we hypothesized that the secondary inflammatory response would affect the bacterial growth. It was not anticipated that the same response would be seen in the animals that did not receive endotoxin during 24 h. The discussion about the phenomenon of endotoxin tolerance, what sets it in play and its influence on bacterial growth is directed at this experimental layout. The cytokine response in the 6 h group, that did not have any prior inflammatory insult except preparatory surgery, represent the unaffected inflammatory system. Indeed, the cytokine response is higher in the 6 h group and entails both lower bacterial levels and higher edema formation in the lungs. We strongly believe that the time frame has influence on the cytokine responses. The investigation of the influence of time in mechanical ventilation was the intention of the experimental part named Ventilation Time, in which we compared one 30 h group with the 6 h group.

To clarify we have added a sentence in the Statistics section (p 10, line 216).

2) In terms of cytokines TNF-α and IL6 mainly produced by macrophages, these results in current studies might reflect the functional role for circulatory macrophages. How about the other immune cells, such as neutrophils, as they also play important roles in bacterial infection and clearance? It would be better if the authors could provide some evidence (such as other cytokines) to indicate the functional activity for other immune cells.

Thank you for a valid comment. Macrophages are described as the main source of inflammatory cytokines in the initial response to an invading pathogen. Circulating leukocytes, as opposed to tissue macrophages, have proven to not be a major source of cytokines in human sepsis as described by Gille-Johnson et al 2012. In the lungs, the alveolar macrophage is abundant in numbers and a very likely major contributor to systemic cytokines in the current experiment. Additional sources of cytokines are recruited neutrophils and endothelia. These cells all have in common an ability to produce nitric oxide (NO) by inducible nitrous oxide synthase. The NO molecule has many disparate effects in an infection, not least to act bactericidal and produce capillary leakage. To target immune cell function better we made additional analyzes of nitrite in urine as a proxy for total NO turnover. The NO molecule is highly unstable and short lived why this stable byproduct is more practical to analyze. The results show that both 30 h groups had suppressed levels of nitrite to around 20% of baseline, and that the 6 h group had significantly higher levels during the whole bacterial phase. We interpret this as a mechanistic explanation of our findings. Exemplified by the 6 h group – higher nitrite indicates higher immune cell activity which entails lower bacterial levels and higher edema development in the Ventilation Time experiment. We have made additions to Results, added a new Figure (5) and comments in the discussion. (p 20, line 437)

3) Due to the infection of P. aeruginosa in lung, I am curious about the local lung injury after pneumonia. In terms of lung damage, the authors measured oxygen index (PaO2/FiO2), wet-to-dry ratio, and cytokines in BAL. There was difference only in wet-to-dry ratio in each comparison. Since the lung tissues and BAL were harvested in current study, could the authors provide more measurements for lung damage, such as protein concentration in BAL, histological lung injury staining, cytokines and surfactants in lung tissues? These results would not only help support local lung injury after bacterial infection, but also might provide some potential explanation for bacterial growth.

Thank you for a valuable comment. During the planning of the experiment we considered alternative BAL methods. However, since the experiment entailed pulmonary outcomes, and the development of edema and P/F ratios were relevant physiological outcome variables we decided to use as little intervention as possible within the lungs. Our choice of a minimal procedure, as is common in clinical practice, was therefore chosen deliberately for this experiment. Repeated BALs or higher volume BAL would have affected the model for our primary outcomes negatively, although the BAL itself may have had higher quality.

As an additional effort to improve the BAL results, we tried to calculate the dilution factor of the BAL sample to normalize each sample and make more effective comparisons. The basis for this approach was that different samples may have had different amount of saline added to the alveolar lining fluid and standardization could impact the results. The dilution factor could potentially be used on both cytokines and bacterial counts in BAL. We analyzed urea in plasma and BAL, a method described for the purpose, as urea supposedly freely flows to the alveolar lining fluid yielding the same concentration as in plasma. Regrettably, most urea values were below detection limit in the BAL fluid effectively making this approach not viable. Our conclusion is that the dilution from saline in the BAL fluid was substantial but not quantifiable in the current experiment.

Regarding lung injury and protein counts as suggested from the editor and reviewer, we have made additional analyzes of albumin in plasma and BAL to evaluate the alveolo-capillary permeability. We chose to use an albumin rather than a total protein assay as the alveolo-capillary leakage would be influenced by the size of the protein studied. A total protein assay would detect proteins of different molecular sizes while an albumin method only measures albumin with a molecular weight of 67 kDa. Another problem with total protein methods is that the methods do not measure all proteins equally. E.g. 1 mg bovine serum albumin gives approximately twice the absorbance as 1 mg IgG with an assay utilizing a Coomassie reagent. Like urea most albumin samples were below detection limit in BAL fluid and the variable could not be produced. As a total protein analysis would suffer from the same problem of dilution as the other variables in BAL fluid it was decided a non-viable endeavor and not pursued further.

Regarding tissue samples we did only structurally harvest samples that were used for cultures and wet-to-dry weight measurements. We initially considered to save samples for later tissue analysis but decided against it as we were unsure about where the samples should be harvested to get representability since the lungs where highly inhomogeneous. Therefore, we don’t have samples to analyze that would benefit the experiment.

Summarily, the method of BAL did not yield data of sufficient quality to make conclusions in the current experiment. Efforts to better the results in our experiment by additional analyses regrettably did not deliver usable data. The additional analyzes, urea and albumin, are referred to in the main manuscript Results section but data is only presented in the Supplementary data file as they were not fit for tabular presentation. An updated comment on the BAL method is added to the Discussion (p 23, line 514).

4) In discussion of experimental part: A30h+Etx vs B30h, the authors discussed the potential effect of TNF-α peak and lung edema existed in A30h+Etx on impairing the immune system’s ability, which subsequently reduced bacterial growth. The results did not show difference in cytokines of BAL and plasma in A30h+Etx vs B30h at the end of study. Could the authors explain this? It would be more strength if the authors could provide more other evidences for the compromise immunity.

We believe the BAL method was inadequate to produce better quality results. Our efforts to improve the quality of the BAL samples by additional analyses of urea regrettably failed (see answer to comment 3).

The analysis of nitrate in urine addresses the question of additional data regarding compromised immune function (see answer to comment 2).

5) Page 26/Line 427-428: "Therefore, the higher TNF-α response to the bacterial challenge in C6h is an indication of the preserved functionality of the alveolar macrophages, ...". I think there would be "the higher plasma TNF-α response...". Could the authors improve the statement? How to explain the results of no difference of TNF-α in BAL between B30h and C6h, based on the preserved functionality of the alveolar macrophages in C6h?

The sentence is corrected accordingly. The lack of difference in BAL cytokines is addressed in the previous comment. The preserved functionality of alveolar macrophages (and other main cytokine producers such as neutrophils and endothelia) in the 6 h group is depicted in higher levels of pro-inflammatory cytokines in plasma and in higher levels of urinary nitrate (see answer to comment 2).

6) Page26/Ling 438-441: "The attenuated edema formation in group B30 suggests a relatively lower activation of neutrophils which, in turn, could explain the higher bacterial growth, supposedly associated with relatively lower bactericidal capacity." In the discussion, the reason for the attenuated edema in B30h was attributed to the relatively lower activation of neutrophils. Could the authors provide the evidence for the lower activity of neutrophils in lung? This would enhance the strength of this statement and support the finding of higher bacterial growth.

The additional results from nitrite analyses in urine comprises the total function of main producers of NO from iNOS. We cannot separate the part attributable to neutrophils. B30h had approximately 20% of the nitrite level of C6h at the time of bacterial challenge, and the difference was significant during the bacterial phase. We have no further possibility to analyze neutrophil function, regrettably.

Minor comments:

7) Page 9/Line 41 in Abstract: "A third group, C6h (n=8), ...". Please improve the description of the group C6h.

The description is corrected into “A third group, C6h (n=8), started the experiment at the bacterial inoculation unexposed to endotoxin or mechanical ventilation (total experimental time 6 h)”.

8) Page 9/Line 46-49: the results of abstract. Given there were two comparisons: A30h+Etx vs B30h and B30h vs C6h, it's not clear what comparison the P-values in results indicated. Could the authors describe these most important findings using numerical results (and statistical significance)?

The abstract is updated and corrected for greater clarity.

9) Page 13/Line 140-142: "The alveolar recruitment maneuver (ARM) consisted of stepwise increments of PEEP until the peak pressure reached 35 cm H2O, followed by prolonged inspiration for 10 seconds (s)." Could the authors provide the details for the stepwise increments of PEEP (#-# cmH2O)?

The sentence has been corrected into “The alveolar recruitment maneuver (ARM) consisted of stepwise increments of PEEP (sequential 3 cm H2O increments for 5 seconds (s) each under control of systolic arterial blood pressure) until the peak pressure reached 35 cm H2O, followed by prolonged inspiration for 10 s.

10) Page 16/Line 216-220, the authors reported no difference of bacterial growth and wet-to-dry ratios between right and left sided lung. Could the authors improve the presentation of raw data from right and left side (supplementary Figure 1a and 2a) in each group, but not in all groups?

The Supplemental Figures 1a and 2a are updated accordingly and the captions are corrected.

11) The P-value presented in Supplementary Figure 3 was a little bit confused based on only two comparable parts in this study. Could the authors clearly show the P values for each effect: inflammation, ventilation time and their interaction, in the figure legend of in the plot?

The primary statistics is presented in Table 4 Plasma cytokines, inflammatory cells and temperature and describes only p-values for the group factor (Inflammation p 0.84, Ventilation Time p 0.63) and not the interaction of group*time. No differences between groups are present based on the group factor only. Group*time can indicate a different dynamic between two comparable groups even if the total sum of values is similar – such as the case with IL6. As this dynamic is of some interest especially in C6h and it is mentioned briefly in the manuscript, we chose to add a presentation as a supplementary file. The caption is updated with additional information.

12) Page 21/Line 311: "The highest bacterial growth was in B30h (Figure 2)." Based on the comparison between B30h and C6h, perhaps it's better to present the data as: "The higher bacterial growth was in B30h...".

The point is valid, and the sentence is changed accordingly.

13) Page 26/Line 438-439: "The attenuated edema formation in group B30 suggests a relatively lower activation of neutrophils..." Please correct the group name "B30h".

Corrected accordingly.

14) Page 25/Line 411-412: "This result indicates a major edema development in

B30h+Etx only after the bacterial challenge." Here I think the group should be A30h+Etx.

Corrected accordingly.

Reviewer #2: The topic of this article is very interesting and relevant. I have just a few comments to make:

1) Although the hypothesis and research questions are well defined, were there any kind of sample size calculation

No power calculation was conducted for this specific experiment since we had no previous data on bacterial behavior in our models. Instead, we used the power calculation for the preceding inflammatory experiments (referred to as 6-8 in the main manuscript). It was based on a systemic TNF-alpha difference of 15% at 6 hours, an alpha error of 0,05, a power of 0,8, and an SD of 10%, which yielded six evaluable animals per group. The choice of 8 animals per group in the previously published day-based experiment (8) was based on this calculation while allowing for a slightly larger variability in the bacterial outcome variable.

As we started with the daybased (8) experiment we could appreciate the bacterial growth in lung tissue better. Based on this data we reduced the number of animals in the 30 h experiments, which were completed at the end of the experimental period, from eight to six to meet the 3R principle. In summary, we reduced the number of animals as we believed we could meet the required difference in the main outcome variable anyway.

The above sequence is added to the end of the Statistics section in the manuscript (p 11, line 243).

2) Please report data as mean and SD not as mean and sEM, especially since you have unequal group sizes.

All tabular data in the manuscript is presented as mean±SD or median(LQ/HQ) as appropriate. In the graphical presentations we find that mean±SEM (66.6% confidence interval) better presents the message intended graphically. One figure, nitrate in urine, is presented as a non-parametric box plot with median(LQ/HQ). The rational for tabular and graphical presentations have been the standard in our previous publications. Importantly, the comparative statistics is not influenced by these choices of presentation.

3) I suggest to elaborate a bit more the clinical / translational meaning of your findings.

We have made additions to the sequence in the manuscript.

The translational relevance of the current experiment lies in that it uses clinically similar intensive care conditions and relevant equipment to evaluate different aspects and factors of influence on bacterial growth in the lungs. The results underline the importance of inflammatory state, i.e. the reactivity of the immune system to a bacterial challenge, for the clinical manifestations yielded by an infection. Additionally, the ease by which we in clinical care unknowingly can affect immune system reactivity by anesthesia and intensive care calls for further exploration on the cellular level. (p 23, line 529)

4) Please comment why you did the inoculation in a blind way. Given the species it would have been easy to deliver everything in a very standardized way e.g. using fiberoptic bronchoscopy.

We agree with the view that a more thorough BAL method potentially could have given different results in the BAL variables. They were not the primary outcome variables in the current experiment and we actively decided to use a minimally invasive method so as to affect the lungs as little as possible. This is partly addressed in a previous comment from reviewer 1 (comment 3), and a sentence is added in the discussion on limitations (p 23, line 516) .

5) I suggest to explain also in this papers method section why you did not perform a sample size calculation. In my opinion crossreferencing is not sufficient in this case.

The explanation for group sizes is added to the statistics section, see response to comment 1.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

2 Oct 2020

Pre-exposure to mechanical ventilation and endotoxemia increases Pseudomonas aeruginosa growth in lung tissue during experimental porcine pneumonia

PONE-D-20-18632R1

Dear Dr. Sperber,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Heinz Fehrenbach

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

All my specific comments were adequately addressed.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for addressing all my concerns to great detail. Authors also performed additional work. The manuscript has significantly improved. I have no additional questions.

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

15 Oct 2020

PONE-D-20-18632R1

Pre-exposure to mechanical ventilation and endotoxemia increases Pseudomonas aeruginosa growth in lung tissue during experimental porcine pneumonia

Dear Dr. Sperber:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Dr. Heinz Fehrenbach

Academic Editor

PLOS ONE