Ulinastatin ameliorates gastrointestinal injury sustained in a 2-hit porcine model of septic shock

a b s t r a c t

Background: Ulinastatin is protective against organ dysfunction in severe disease. We investigated the extent of gastrointestinal tract injury and the potential protective effect of ulinastatin in a 2-hit porcine model of septic shock.

Methods: Pigs were randomized to 4 groups, 3 septic shock groups (12 per group)–vancomycin (VAN), vanco- mycin + ulinastatin (VAN + ULI), and saline (SAL)–and a sham-operated group (n = 10). Septic shock was in- duced by 2 hits: acute lung injury and Staphylococcus aureus challenge. Four hours after the 2 hits, pigs in septic shock received a central venous injection of vancomycin, vancomycin + ulinastatin, or saline. Hemodynamic sta- tus and blood samples were obtained. Serum D-lactate, Diamine oxidase, and intestinal fatty acid-binding protein were determined, and gastrointestinal ATP enzyme activity was measured. Pathological and ultrastructural tests were performed.

Results: Gastrointestinal tract injury after septic shock was significant. Compared with the SAL and VAN groups, the VAN + ULI group had better hemodynamic parameters (improved mean arterial pressure and cardiac out- put) (Pb .05) and improved oxygen metabolism (oxygen delivery and consumption) (Pb .05). In VAN + ULI group, serum D-lactate, diamine oxidase, and intestinal fatty acid-binding protein were significantly reduced (Pb .05). Moreover, Na⁺-K⁺– and Ca²⁺-ATPase enzyme activity was significantly high (Pb .05). Pathological and ultrastructural changes showed that severe gastrointestinal injury was significantly ameliorated in the VAN + ULI group vs the SAL and VAN groups.

Conclusions: Gastrointestinal injury and abnormal Energy metabolism are remarkable following septic shock. Ulinastatin can improve energy metabolism and ameliorate injury to the gastroIntestinal mucosa in the early stage of septic shock.

Introduction

Sepsis is a leading cause of morbidity and mortality in critically ill patients, and its incidence is increasing annually worldwide [1] despite continuous improvements in treatment strategies, especially in antimicrobial therapy and supportive care. The pathogenesis of sepsis is complex and can often lead to immunologic dysfunction and Multiple organ dysfunction syndrome. Sepsis is therefore a major challenge for clinicians.

? Reprints: none to declare.

?? Sources of support: none to declare.

? Presentation: none to declare.

* Corresponding author at: Emergency Department of Beijing Chaoyang Hospital, Affil- iated to Capital Medical University, Chaoyang District, Beijing, China. Tel.: +86 13681392380; fax: +86 010 85231051.

E-mail address: lcscyyy@163.com (C. Li).

The gastrointestinal (GI) tract has a major barrier function [2], playing a pivotal role in protecting body from harmful intraluminal pathogens and large antigenic molecules. The GI tract is not only the main target for sepsis but also the vital organ that mediates sepsis. When a body goes into septic shock, ischemia, anoxia, and impairment of energy metabolism occur spontaneously [3]. The GI mucosa is vulner- able to hypoxia induced by severe sepsis or septic shock [4]. It is there- fore important to protect the integrity and normal function of the GI tract in preventing the development of sepsis.

Ulinastatin, a human urinary trypsin inhibitor, is a broad-spectrum protease inhibitor isolated and purified from human urine [5] that has a protective effect on many organs in various animal models [6,7]. The potential role of ulinastatin in treating sepsis deserves investigation. Re- cent experimental and clinical studies have indicated that ulinastatin ameliorates intestinal injury [8,9]. However, most of these experiments were performed on rodents, making it hard to obtain serial blood sam- ples and detailed hemodynamics. The aim of the present study was to evaluate the effect of ulinastatin on protecting the GI tract by developing

http://dx.doi.org/10.1016/j.ajem.2016.05.011

Fig. 1. Diagram of the experimental protocols. Pigs were randomized into 4 groups. All the pigs were studied for 24 hours. The pigs in the SHAM and SAL groups were infused with saline as a vehicle, the pigs in the VAN group received vancomycin, and the pigs in the VAN + ULI group were given both vancomycin and ULI. Blood sampling and data monitoring were performed at the times indicated.

a 2-hit sepsis model with acute lung injury and Staphylococcus aureus challenge in pigs.

Methods

This study was approved by the Animal Care and Use Committee of Beijing Chao-Yang Hospital, Capital Medical University (approval code: 113,266). All protocols were carried out in strict accordance with the guidelines for animal care and use established by the Capital Medical University Animal Care and Use Committee.

Animal preparation

Forty-six healthy domestic pigs aged 8-10 weeks, weighing 27 +- 2 kg and of either sex, were used. The animals were divided by sex on average and then were randomized to 4 groups according to the random number table: 3 septic shock groups (n = 12 per group)–vancomycin (VAN), vancomycin + ulinastatin (VAN + ULI), and saline (SAL)–and a sham-operated (SHAM) group (n = 10). The animals were fasted the night before the day of the experiment but had free access to water.

To induce anesthesia, propofol (1 mg/kg) was administered by intra- venous cannulation as a Bolus injection via a peripheral ear vein; pento- barbital (8 mg/[kg h]) was given intravenously to induce deep anesthesia, with saline infused intravenously at 10 mL/h as a vehicle.

All animals were intubated using a cuffed 6.5-mm endotracheal tube and maintained in a surgical plane for mechanical ventilation (Evita4; Drager, Lubeck, Germany) at a tidal volume of 8 mL/kg and a respiratory frequency of 12 breaths per minute with room air to maintain an end- tidal CO₂ concentration of 35-40 mm Hg. Vascular catheters were placed into the thoracic aorta, via the right femoral artery, for aortic pressure monitoring. A Swan-Ganz catheter (CCOmbo774HF75; Edwards Life Sciences, Irvine, CA) connected to a CCO monitor (Vigilance II, Edwards Life Sciences) was inserted into the left femoral vein and flow directed into the pulmonary artery to monitor the hemodynamic parameters. The entire study period lasted 24 hours. All animals were fluid resusci- tated, initially started on an infusion of 2 mL/[kg h] lactated Ringer‘s so- lution. The fluid rate was then adjusted depending on the filling pressures and hematocrit. Fluid resuscitation was limited to a maximum of 10 mL/[kg h] [10].

Bacterial preparation

Live S aureus was used to induce sepsis. S aureus (ATCC 25923) was incubated on a blood agar plate (BACT/Alert 3D; BioMerieux, Lyon, France). Colonies were suspended in brain heart infusion broth (BOXUN, Shanghai, China), modified for 18 hours at 37?C, and incubated in a shaking water bath with a speed of 40 rpm for ~20 hours at 37?C to reach the mid-log phase on the growth curve. The bacterial solution was then centrifuged and washed twice with normal saline. Finally, the

Table 1

Baseline characteristics among SHAM, SAL, VAN, and VAN + ULI groups

Characteristic	SHAM (n = 10)	SAL (n = 12)	VAN (n = 12)	VAN + ULI P^a (n = 12)
Body weight (kg)	27.43 +- 2.94	26.98 +- 2.28	27.22 +- 2.11	26.79 +- 2.41	.963
HR (bpm)	103.13 +- 5.61	101.50 +- 8.04	104.68 +- 6.23	102.33 +- 7.96	.834
Body temperature (?C)	38.45 +- 2.35	38.73 +- 3.17	38.54 +- 2.71	38.66 +- 2.88	.998
CO (L/min)	3.71 +- 0.42	3.65 +- 0.22	3.54 +- 0.34	3.43 +- 0.25	.342
MAP (mmHg)	108.21 +- 7.91	105.44 +- 8.24	100.32 +- 7.67	102.28 +- 7.52	.272

bpm = beats per minute; HR = heart rate.

^a PN .05 among the 4 groups (1-way ANOVA).

Table 2

Septic shock outcome 4 hours after 2 hits

of quality control, and these were measured after each round of smoke or air inhalation.

SHAM (n = 10)

SAL

(n = 12)

VAN

(n = 12)

VAN + ULI (n = 12)

Live S aureus (3 x 10¹¹CFU) was instilled into the lung lobes (via an endotracheal tube) after ALI. Septic shock was induced within 4 hours in

Survival 10 5 7 11

CO (L/min) 3.71 +- 0.28 1.90 +- 0.19^a 2.31 +- 0.20^a 2.14 +- 0.25^a

MAP (mmHg) 115.31 +- 5.22 53.75 +- 5.81^a 59.21 +- 6.13^a 51.33 +- 6.02^a ^a Pb .05 vs SHAM (1-way ANOVA).

bacteria were resuspended in sterile saline. The number of bacteria in solution was adjusted to 3 x 10¹¹colony-forming units (CFU)/mL.

Experimental protocols

Septic shock was induced by a 2-hit model, with ALI as the first hit followed by S aureus instillation as the second hit. As described previ- ously [10], ALI was induced by smoke inhalation. In short, all animals ex- cept the Sham group were insufflated with 4 x 12 breaths of cotton smoke at b 40?C, which was generated and delivered by a modified bee smoker filled with 40 g of burning cotton-toweling material. The SHAM group received 4 x 12 breaths of room air via a bee smoker. A sig- nal extraction pulse CO-oximeter (Rad-57; Masimo Corp, Irvine, CA) monitored the arterial carboxyhemoglobin concentrations as a method

all animals with the 2 hits, which showed a decrease in mean arterial pressure (MAP) and cardiac output (CO) by hemodynamic monitoring. The drugs were delivered in a randomized manner by using the sealed envelope method, which was blinded to the investigators. Apart from those in the SHAM group, the pigs received a central venous injection of vancomycin (20 mg/kg) (Lilly Technology Center, Indianapolis, IN), vancomycin (20 mg/kg) + ulinastatin (30,000 U/kg) (Techpool Bio- Pharma Co Ltd, Guangzhou, China), or saline.

Hemodynamic parameters were continuously recorded. Blood sam- ples were obtained at 0, 4, 8, 12, 16, 20, and 24 hours. Finally, the anes- thetized animals were euthanized by infusing a lethal dose of potassium chloride. The GI tissues were dissected, and the samples were taken from either the ileum, for intestinal tissue, or the arcus minor ventriculi (for the gastric tissue) to measure the activities of both Na⁺-K⁺– and Ca²⁺-ATPase and for histopathology (Fig. 1).

Measurements

Hemodynamic data, including CO and MAP, were continuously record- ed. Oxygen metabolism parameters, including oxygen delivery (1.34 x Hb

Fig. 2. Changes in hemodynamics and oxygen metabolism variables. A, CO. B, MAP. C, Lactate . D, Oxygen delivery (DO₂). E, Oxygen consumption (VO₂). The values are reported as mean +- SD. ^aPb .05 vs SHAM; ^bPb .05 vs SAL; ^cPb .05 vs VAN (1-way MRM ANOVA).

Fig. 3. Concentrations of D-LA (A), DAO (B), and I-FABP (C) in serum after septic shock. ^aPb

.05 vs SHAM; ^bPb .05 vs SAL; ^cPb .05 vs VAN (1-way MRM ANOVA). D-LA = D-lactate.

x SaO2x CO+ 0.003 x PaO2), oxygen consumption (CaO2- CvO2x CO), and serum lactate, were calculated by blood-gas analyses (GEM Premier 3000 Blood gas analyzer; Instrumentation Laboratory, Lexington, MA).

Table 3

Changes in Na⁺-K⁺– and Ca²⁺-ATPase activity in gastric and intestinal tissues

Items Intestinal tissue Gastric tissue Na⁺-K⁺-ATPase (umol/g prot)

SHAM 213.66 +- 28.67 317.71 +- 38.56

SAL 151.25 +- 20.34^a 89.42 +- 22.05^a,c

VAN 158.58 +- 26.15^a 189.74 +- 33.45^a,b VAN + ULI 180.62 +- 20.47^a,b,c 253.38 +- 25.78^a,b,c

Ca²⁺-ATPase (umol/g prot)

Venous blood samples were taken at 0, 4, 8, 12, 16, 20, and 24 hours.

D-Lactate, diamine oxidase (DAO), and intestinal fatty acid-binding pro-

tein (I-FABP) were determined using enzyme-linked immunosorbent assay kits (Beijing Fangchengjiahong Technology Co Ltd, Beijing, China). Specific spectrophotometric kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) were used to determine Na⁺-K⁺– and Ca²⁺– ATPase activity in GI tissues. Using standard formulas, enzyme activity was determined by measuring the optical density of the inorganic phos-

phate decomposed from ATP by the tissue protein.

Pathologic and ultrastructural analysis

Specimens of GI tissues were formalin fixed, cut into 4-um sections, and stained with hematoxylin and eosin (HE) (original magni- fication, x 100) for pathologic observation under a light microscope (CX41; Olympus, Tokyo, Japan). The ultrastructure of GI tissue was observed using transmission electron microscopy (JEM-1230; Jeol, Tokyo, Japan). The results were interpreted blindly by 2 indepen- dent observers.

Statistical analysis

Statistical analysis was performed using SPSS 17.0 (Chicago, IL). Data are expressed as the mean +- standard deviation. One-way analysis of variance (ANOVA) was used to compare the differences among the groups. The continuous variables were fixed to normal distributions, and equal variances were analyzed using the Kolmogorov-Smirnov and homogeneity of variance tests. One-way MRM ANOVA was used to determine the differences between each time point in each group. Comparisons of survival among groups were made using Pearson ?² test. Statistical differences were considered significant at Pb .05.

Results
1. Outcomes

The baseline characteristics of the 4 groups are shown in Table 1. No significant changes in body weight, heart rate, body temperature, CO, and MAP were detected (PN .05).

Septic shock was defined as MAP less than 60 mmHg, which is ac- cording to the criteria of the “Surviving Sepsis Campaign: International Guidelines for Management of Severe Sepsis and Septic Shock” (2012) [11]. As shown in Table 2, septic shock was induced within 4 hours in all the animals that received the 2 hits. Compared with those in the SHAM group, CO and MAP were significantly lower in the SAL, VAN, and VAN + ULI groups according to hemodynamic monitoring (Pb

.05). However, there were no significant differences among the 3 poste- rior groups (PN .05).

In the VAN + ULI group, only 1 piglet died at 24 hours after the 2 hits. A significant difference in survival to the end of the 24-hour exper- iment period among the groups was demonstrated using Pearson ?² test (P= .005).

Hemodynamics and oxygen metabolism

After septic shock, CO and MAP were significantly lower in the SAL, VAN, and VAN + ULI groups compared with the SHAM group (Pb .05). CO and MAP were significantly higher in the VAN + ULI group at 16, 20, and 24 hours compared with the SAL and VAN group (Fig. 2A and B). Throughout the study, serum lactate concentrations were signifi-

SHAM 291.09 +- 39.45

SAL 142.93 +- 40.14^a,c

VAN 184.26 +- 45.85^a,b

293.12 +- 57.35

90.13 +- 27.45^a,c

166.51 +- 36.9^a,b

cantly increased compared with baseline after septic shock in the SAL, VAN, and VAN + ULI groups (Pb .05). However, the lactate concentra-

VAN + ULI 251.47 +- 36.17^a,b,c 207.08 +- 54.05^a,b,c

^a Pb .05 vs SHAM.

^b Pb .05 vs SAL.

^c Pb .05 vs VAN (1-way ANOVA).

tions were lower in the VAN + ULI group than those in the SAL and VAN groups at 8 and 12 hours (Fig. 2C). Oxygen delivery and consump- tion were significantly higher in the VAN + ULI group than those in the SAL and VAN groups at 20 and 24 hours (Fig. 2D and E).

Fig. 4. Pathological changes in the intestinal structure observed by light microscopy in 4 groups (HE staining x100). A, SHAM group. B, SAL group. C, VAN group. D, VAN + ULI group.

Fig. 5. Pathological changes in the gastric tissue structure observed by light microscopy in 4 groups (HE original magnification x100). A, SHAM group. B, SAL group. C, VAN group. D, VAN + ULI group.

Fig. 6. The ultrastructural changes of the intestinal tissue observed by electron microscopy in the 4 groups (original magnification x10,000). A, SHAM group. B, SAL group. C, VAN group. D, VAN + ULI group.

Serum D-lactate, DAO, and I-FABP

Baseline D-lactate, DAO, and I-FABP serum levels did not differ among the 4 groups (Fig. 3). Septic shock resulted in the significant ele- vation in serum D-lactate, DAO, and I-FABP levels relative to the SHAM group (Pb .05). D-Lactate and DAO remained elevated after septic shock, whereas I-FABP peaked at 4 hours and then began to decrease. Compared with the SAL and VAN group, expression of D-lactate and DAO in the VAN + ULI group was significantly reduced at 16 hours, whereas the expression of I-FABP in the VAN + ULI group was signifi- cantly reduced at 12 hours.

ATPase activity

There was significantly greater Na⁺-K⁺– and Ca²⁺-ATPase activity in both gastric and intestinal tissues in the VAN + ULI group compared with the SAL or VAN groups (Pb .05; Table 3).

Histopathology

Slices of GI tissue were observed under a light microscope (Figs. 4-7). The SAL group displayed high levels of inflammatory cell infiltration, vascular congestion and expansion (Fig. 4B, black arrow), interruptions and erosion in the intestinal villous epithelium, as well as disordered ar- rangement of the cells of the lamina (Fig. 4B, white arrow). In the VAN group, the structure of the intestinal villi was nearly intact, but many in- filtrating inflammatory cells were present (Fig. 4C, black arrows). In the SHAM and VAN + ULI groups, the gastric mucosa and intestinal villi were intact (Fig. 4A and D).

The SHAM group displayed structural integrity of the gastric mucosa (Fig. 5A). On the other hand, the gastric mucosa in the SAL group was damaged, and the cells were disordered (Fig. 5B, black arrows). In the VAN group, the cell arrangement was slightly disordered, but there was visible infiltration with inflammatory cells (Fig. 5C, black arrow). The mucosal cell arrangement and morphology were near normal in the VAN + ULI and SHAM groups (Fig. 5D).

Electron microscopy showed that intestinal mitochondrial injury was severe in the SAL group, which showed characteristic cavity changes (Fig. 6B, black arrow) and indistinct swelling (Fig. 6B, white arrow). In the VAN group, parts of the intestinal microvilli were defective (Fig. 6C, white arrow), and the structure of the intestinal mitochondria was not distinct, with hazy cristae (Fig. 6C, black arrow). The VAN + ULI group had signifi- cantly attenuated mitochondrial injury compared with the SAL group, maintaining mitochondrial structure and integrity (Fig. 6D, white arrow).

The gastric glandular structure from each group was relatively normal, especially in SHAM group and VAN + ULI group, there were no significant change in cell size and no obvious nuclear condensation (Fig. 7A and D). The SAL group had severe injury, displaying both nuclear deformation and shrinkage in parietal cells (Fig. 7B, white arrow). Mitochondrial swelling and cracked cristae were observed in the VAN group (Fig. 7C, white arrow). Parietal cells in the VAN + ULI group were almost as normal as those in the SHAM group. The VAN + ULI group had little intracellular dam- age in the mitochondrial architecture at 24 hours after septic shock.

Discussion

In the past, septic shock was considered to be difficult to replicate in animal models [12]. Systemic inflammatory response syndrome or

Fig. 7. The ultrastructural changes of the gastric tissue observed by electron microscopy in the 4 groups (original magnification x10,000). A, SHAM group. B, SAL group. C, VAN group. D, VAN + ULI group.

multiple organ dysfunction syndrome induced by sepsis is often associ- ated with a priming insult and usually at least 1 or more comorbidities. Eissner et al [13] tried a 2-hit model with hemorrhagic shock and Pseudomonas aeruginosa challenge to induce sepsis. According to the 2-hit theory, this enhanced inflammatory response leads to further tissue damage and organ dysfunction, which seems similar to clinical sepsis [4].

To study the GI tract and make the results more credible, we injured the respiratory system as the first hit to make the abdomen completely free of interference. Moreover, recent clinical studies have demonstrat- ed that sepsis caused by Gram-positive bacteria has been increasing and that these are responsible for 50% of cases of severe sepsis or septic shock in hospital intensive care units [14]. S aureus is the leading cause of sepsis among gram-positive bacteria [15].

In the present study, inhalation of S aureus was used as the second hit to induce septic shock. Both CO and MAP decreased significantly in all animals except those in the SHAM group, and lactate concentration was markedly raised within 4 hours after the 2 hits. This indicated pro- nounced circulatory changes and deterioration of Gas exchange, which demonstrated successful induction of septic shock.

In our study, Na⁺-K⁺– and Ca²⁺-ATPase activity in the GI tissue of the pigs in the SAL group was significantly lower than that in the GI tis- sue of the pigs in the SHAM group, which demonstrates that energy metabolic dysfunction occurred after septic shock. This was consistent with the histopathologic and ultrastructural changes. After septic shock, there were different degrees of mucosal damage in GI tissues, and many infiltrating inflammatory cells were found. Meanwhile, in the SAL group, the gastric mitochondria were swollen with fragmented cristae, and the structure of the intestinal mitochondria was blurred

with hazy cristae. Moreover, all the indices (serum DAO, D-lactate, and I-FABP) reflecting GI tract injury [16-18] were significantly increased when septic shock was induced, indicating that GI mucosal ischemia and injury occurred early after septic shock.

Ulinastatin, a multifunctional Kunitz-type serine protease inhibitor, protects many organs from severe diseases. It is used widely to treat pancreatitis, shock, and disseminated intravascular coagulation. Al- though ulinastatin can enhance the immune response and adjust in- flammatory reactions, it cannot kill bacteria directly [9]. Concomitant antibiotic treatment must therefore be given.

In the present study, the VAN + ULI group had better hemodynam- ics (MAP and CO) and improved oxygen metabolism (oxygen delivery and consumption). Serum DAO, D-lactate, and I-FABP in the VAN + ULI group were lower than those in the SAL and VAN groups during the subsequent experimental period. This was similar to the results from a previous, parallel-controlled clinical study [9]. In addition, the ac- tivity of Na⁺-K⁺– and Ca²⁺-ATPase in the GI tissue was significantly high in the VAN + ULI group, indicating that ulinastatin improves ener- gy metabolism and maintains GI tract function under ischemic condi- tions following septic shock.

Furthermore, observation of histopathologic and ultrastructural changes showed that severe GI tissue injury was significantly ameliorat- ed in the VAN + ULI group compared with that in the SAL and VAN groups. This suggests that ulinastatin improves cellular metabolism fol- lowing septic shock by reducing energy requirements during the time at which oxygen supply is limited.

In a systematic review and meta-analysis of 29 randomized con- trolled trials [19], ulinastatin decreased intensive care unit mortality; improved oxygenation; decreased the duration of hospital stay; and

significantly improved inflammatory markers and, to a lesser extent, organ dysfunction.

The protective effect of ulinastatin on vital organs may be mediated in several ways.

Ulinastatin improves Ischemia-reperfusion injury in multiple or- gans by suppressing excessive generation of superoxide anion radicals in blood, oxidative stress, early inflammation, and endo- thelial injury [20]. In the present study, ulinastatin elevated MAP and CO after septic shock and improved oxygen, energy, and cell metabolism.
Ulinastatin inhibits inflammation by suppressing infiltration of neutrophils and release of elastase and Inflammatory mediators from neutrophils. This was confirmed by our results.
Ulinastatin helps to protect mucosal integrity. Sepsis leads to in- testinal mucosal injury, Bacterial translocation, and further aggra- vation of intestinal and remote organ injury [21]. Li et al [22] reported that ulinastatin decreased intestinal mucosal apoptosis and inhibited bacterial translocation. In the present study, ulinastatin significantly inhibited the increase in serum DAO, D- lactate, and I-FABP, reflecting the integrity and severity of GI mu- cosal injury, so as to interrupt the vicious cycle of multiple organ dysfunction syndrome caused by sepsis.

Our study had some limitations that should be noted. First, we used apparently healthy young pigs, but most patients experiencing sepsis or septic shock are elderly with chronic underlying diseases. Second, the end point of the study was 24 hours after septic shock, and the long- term effect of the drug should be evaluated in further studies.

Conclusion

In conclusion, our exploratory experiments demonstrated that GI in- jury and abnormal energy metabolism are significant following septic shock. Ulinastatin can improve energy metabolism and ameliorate inju- ry to the GI mucosa in the early stage of septic shock.

Acknowledgements

The authors would like to thank Yingying Fang and Liangxing Zhao for their technical assistance.

References

use of sequential organ failure assessment parameters in an awake porcine model of severe Staphylococcus aureus sepsis. APMIS 2012;120:909-21.

AJEM