Fatty acid ethyl ester (FAEE) associated acute pancreatitis: An ex-vivo study using human pancreatic acini

Abstract

Background & aim

Fatty acid ethyl esters (FAEEs), are produced by non-oxidative alcohol metabolism and can cause acinar cell damage and subsequent acute pancreatitis in rodent models. Even though experimental studies have elucidated the FAEE mediated early intra-acinar events, these mechanisms have not been well studied in humans. In the present study, we evaluate the early intra-acinar events and inflammatory response in human pancreatic acinar tissues and cells in an ex-vivo model.

Methods

Experiments were conducted using normal human pancreatic tissues exposed to FAEE. Sub-cellular fractionation was performed on tissue homogenates and trypsin and cathepsin B activities were estimated in these fractions. Acinar cell injury was evaluated by histology and immunohistochemistry. Cytokine release from exposed acinar cells was evaluated by performing Immuno-fluorescence. Serum was collected from patients with AP within the first 72 h of symptom onset for cytokine estimation using FACS.

Results

We observed significant trypsin activation and acinar cell injury in FAEE treated tissue. Cathepsin B was redistributed from lysosomal to zymogen compartment at 30 min of FAEE exposure. IHC results indicated the presence of apoptosis in pancreatic tissue at 1 & 2hrs of FAEE exposure. We also observed a time dependent increase in secretion of cytokines IL-6, IL-8, TNF-α from FAEE treated acinar tissue. There was also a significant elevation in plasma cytokines in patents with alcohol associated AP within 72 h of symptom onset.

Conclusion

Our data suggest that alcohol metabolites can cause acute acinar cell damage and subsequent cytokine release which could eventually culminant in SIRS.

Introduction

Acute pancreatitis (AP) is a necro-inflammatory condition of the pancreas with varying degrees of severity. Approximately 20—30% of affected individuals develop severe disease characterized by significant pancreatic necrosis and systemic inflammation, which may lead to organ failure and subsequent death [1,2]. Alcohol consumption is one of the major risk factors of AP. The exocrine pancreas metabolises alcohol by both oxidative and non-oxidative (OME and NOME) pathways and non-oxidative ethanol metabolism generates fatty acid ethyl esters (FAEEs) [2–5]. Several studies reported that FAEE causes more injury than ethanol itself. Prolonged administration of alcohol in animals has been found to increase digestive enzymes including trypsinogen, chymotrypsinogen, lipase as well as the lysosomal enzyme cathepsin B [6,7]. In experimental models of AP, FAEEs has been shown to produce several biochemical changes including increased intracellular calcium, oxidative stress, impairment in autophagy, colocalization of pancreatic digestive enzymes with lysosomes, alteration in the activation of transcriptional factors such as NF-kB and AP-1 [8–11]. Studies also reported that mitochondrial damage was central to this Ca²⁺ overload caused by loss of membrane potential and consequent decline in ATP, and acinar cell homeostasis [12,13]. However, data from rodent experimental models may not necessarily be extrapolated to human conditions because of the differences in the physiology and immune functions between humans and rodents.

We had previously reported bile acid-induced early acinar cell damage and inflammatory response in AP under in-vitro conditions, using human pancreatic acini. Our results demonstrated that during early time point of pancreatic injury, the acinar cell could secrete inflammatory cytokines such as tumour necrosis factoralpha (TNF-a) and interleukin-6 (IL6) which in turn could activate the circulating PBMCs that results in systemic inflammatory response syndrome (SIRS) [14,15]. Since insight into the pathogenesis of alcohol-induced AP in humans is still speculative, in the present study we aimed to investigate the early intra-acinar events in an in-vitro model of alcohol-induced human AP. Along with this we also conducted clinical acute pancreatitis studies, in which we collected blood samples from a series of patients who were admitted with clinical diagnosis of AP within 72hrs of symptom onset to understand the relation between early acinar injury with systemic inflammation. Identifying the mechanisms by which ethanol predisposes acinar cell damage could unravel several promising targets for therapeutic interventions. In this study, we conducted the experiments using human pancreatic tissues and acinar cells, and the pancreatic injury was induced using FAEEs.

Material and methods

Materials

Fatty acid ethyl esters, Collagenase IV from Clostridium histolyticum and Bovine serum albumin were obtained from Sigma-Aldrich (St. Louis, USA); Substrates Z-Arg-Arg-7AMC, BOC-Gln-Ala-Arg-7AMC hydrochloride were obtained from Sigma- Aldrich (Buchs, Switzerland), LDH cytotoxicity assay kit was obtained from G-Biosciences (St. Louis, USA). DMEM medium, Phosphate Buffered Saline were purchased from Himedia Laboratories (Mumbai, India). Amylase quantification kit was purchased from Accurex Biomedicals (Thane, India). Cytometric bead array (CBA) human inflammatory cytokines kit was purchased from BD Biosciences, USA. Polyclonal primary antibodies (anti-IL-6, anti-TNF-α, anti-caspase-3, anti-Annexin-V) were purchased from Abcam (Cambridge, UK), anti-amylase antibody was purchased from Santacruz (California, USA). Most of the chemicals used in the study were purchased either from Sigma-Aldrich (Germany, USA) or SDFCL Fine Chemicals limited (Mumbai, India).

Study site and approval

The studies were conducted at the Asian Institute of Gastroenterology, which is a high volume referral centre for pancreatic diseases. The study protocols were approved by the Asian Healthcare Foundation/Asian Institute of Gastroenterology Institutional Review Board (Reference no. AIG/AHF IRB 12/ 2011). Informed consent was obtained from all participants from whom tissue samples and clinical data were procured.

Human pancreatic specimen collection and processing

We procured normal human pancreatic tissues from patients who underwent Whipple’s surgery or distal pancreatectomy for indications other than chronic pancreatitis and pancreatic adenocarcinoma. We discarded samples that were non-viable and functionally impaired before the initiation of the experiments. We also excluded data from samples which initially appeared non-malignant but eventually showed malignancy on biopsy of the operative specimens. We conducted experiments in a stepwise manner. Firstly, we standardized our experiments in 3—5 sets followed by the actual experiments (n = 3) and finally validating the results in 2—3 independent sets. We presented the data from the experimental and validation sets, and for each experiments we used paired controls to avoid the interindividual variability as well as sampling errors.

Pancreatic tissues specimens were collected by cutting away from the transection margin of the resected pancreas and specimens were washed thrice with ice-cold oxygenated HEPES buffer (containing 127 mmol/L NaCl, 4.7 mmol/L KCl, 1.0 mmol/L Na2HPO4, 10 mmol/L HEPES, 1.06 mmol/L MgCl2, 1.28 mmol/L CaCl2,10 mmol/L D-glucose) and then transferred to the laboratory in fresh oxygenated ice-cold HEPES containing trypsin inhibitor and sodium pyruvate. After transportation to the lab, pancreatic specimens were washed with oxygenated Dulbecco’s modified Eagle’s medium containing 2% bovine serum albumin and specimens were processed for further experiments by removing the extracellular fat. Some tissues were minced into fragments of <0.5 mm2 without using collagenase. The minced fragments were exposed to noxious stimuli in complete DMEM medium with 100 μg/mL soybean trypsin inhibitor, in a cell culture incubator at 37 °C in an atmosphere of 5% CO₂. Remaining tissues were used for acinar clusters isolation. Briefly pancreatic tissue samples were injected with collagenase (200 U/mL) and minced into tiny bits. The minced pancreata were then incubated for 30—40 min in oxygenated HEPES buffer with collagenase at 37 °C in shaking water bath. For every 10 min of incubation, collagenase was drained from minced tissue and the suspension was pooled and filtered through a 140 mm nylon mesh. The filtrate was allowed to sediment in 4% BSA to get pure acini and finally acini were washed thrice and dispersed in fresh HEPES. Isolated acinar cells were then allowed to equilibrate for 5 min at 37 °C before conducting experiments.

Prior to conducting the experiments we look for morphological and functional integrity of the acini. For morphological examination H&E sections were made for each pancreatic specimens and analysed by pathologist who were blinded to the study groups.

To assess the functional viability, normal secretory response was quantified by checking the amylase activity against carbachol treatment with and without atropine pre-treatment. Briefly, freshly prepared pancreatic acinar fragments were treated with incremental doses (1 μM to 1 mM) of carbachol for 15 and 30 min. In a few experiments, the acinar tissues were pre-treated with 10 μM atropine for 20 min after which the media was replaced with fresh HEPES and treated with carbachol. Once the incubation was over, amylase was estimated in the media and expressed as percent of total. Experimental data was included for the samples showing morphological and functional activity. We have published the data on morphological integrity and functional activity earlier [14].

Induction of pancreatic acinar injury

Experiments were performed using pancreatic tissue sections under in-vitro conditions. 1–2 mm sized pancreatic fragments were exposed to different concentrations (10, 20, 50 μM) of commercially available Fatty acid ethyl esters (FAEE) (Cat. No.; 49454-U, Sigma Supelco, Bellefonte, USA) for different time periods (15mins, 30mins, 60mins,120mins, 3hrs, 6hrs, 18hrs, 24hrs). We used 49454-U - Fatty Acid Ethyl Esters (FAEES), containing C4-C24 even carbon saturated fatty acids at a concentration of 1000ng/ul and we used 850 mM ethanol as a vehicle to dissolve the components as described earlier [12]. Combination of ethanol with fatty acid ethyl esters mimicked an in vitro alcoholic environment. After induction using FAEE, tissue sections were immediately washed with PBS and used to perform biochemical and histological studies. To conduct the histological studies, FAEE treated and control pancreatic tissue fragments were fixed in 4% buffered formalin, following which blocks were prepared using these tissues. To perform H&E and immunohistochemistry staining, 140 μm tissue sections were prepared. For biochemical assays control and treated pancreatic tissue slices were homogenized in homogenization buffer and the resulting homogenates were used for experiments.

For experiments with human pancreatic acini, freshly isolated human acini were resuspended in DMEM with 5% FBS and 0.025% soybean trypsin inhibitor and cells were seeded into six-well plate and allowed to equilibrate by incubating at 37 °C in a humidified 95% O₂/5% CO₂ atmosphere for 30mins. After incubation cells were treated with different concentrations of (10, 20, 50 μM) fatty acid ethyl esters for different time points (1hr, 2hrs) and control cells were kept untreated. After induction using FAEE, control and treated acinar cells were immediately washed with fresh PBS and used to perform biochemical and histological studies.

Subcellular fractionation

Pancreatic tissues were homogenized by using glass Dounce homogenizer (60 strokes per min) in homogenization buffer containing 250 mM sucrose, 20 mM HEPES-KOH (pH 7.0), 10 mM KCl, 1 mM EGTA, 2 mM MgCl2, 1 mM EDTA, 1 mM dithiothreitol. Pancreatic tissue homogenate was then processed through differential centrifugation to separate the zymogen and lysosome enriched fractions as described earlier [16]. Briefly, the homogenate was centrifuged at 1300 g to separate the zymogen enriched fraction. The resultant supernatant was re-centrifuged at 12000g for 13min and supernatant (lysosomal enriched fraction) was collected [17].

Measurement of Biochemical parameters

Pancreatic amylase activity was quantified in control and FAEE treated pancreatic tissue medium by using ERBA Chem-5 plus v2 auto analyser and was expressed as IU/ml. Trypsin activity was measured in whole pancreatic tissue and as well as in tissue subcellular fractions by using the substrate BOC-Gln-Ala-Arg-MCA (Sigma Aldrich, USA) in TAB buffer with 0.1%BSA(pH-8.1) according to the method described by Kawabata et al. [18]. Control and treated pancreatic tissue trypsin activity was measured (at a wavelength of Ex:350 nm and Em:460 nm) on Fluroskan Assent spectrofluorometer (Thermo Scientific, USA). Fluorometric values were analysed and expressed per mg of protein.

Cathepsin B activity was measured at 37 °C by using the substrate Z-Arg-Arg-7AMC (Sigma Aldrich, St. Louis, USA) in phosphate buffer (pH-6.5) containing 20 mM DTT according to Mcdonald & Ellis method [19]. Substrate hydrolysis by cathepsin B was measured fluorometrically with Ex at 355 nm and Em at 460 nm. Final values were expressed as total cathepsin B activity per mg of protein.

Histological examination of pancreatic injury

Acinar cell injury in control and treated pancreatic tissue fragments were identified and quantified by Hemotoxylin & Eosin staining. Briefly, 4 μm-thick serial sections were prepared from formalin-fixed paraffin-embedded (FFPE) tissue blocks and stained with Hema-toxylin/Eosin and examined by a senior research pathologist who was blinded to the study groups. Pancreatic tissue sections were examined for nuclear pyknosis, swollen and lightly stained cytoplasm and loss of membrane integrity, which were analysed in 5 randomly selected fields using an Olympus CX41 microscope system, as described previously [14].

Immunohistochemical studies

Immunohistochemistry was conducted on paraffin-embedded pancreatic tissue sections cut into 4–5 μm thin serial sections. Sections were deparaffinised and rehydrated in xylene and ethanol followed by antigen-retrieval with 10mM citrate buffer (pH-6.0 with 0.05%Tween 20). Peroxidase activity was blocked by 3% H2O2. Tissue sections were incubated at 4° C overnight with Rabbit polyclonal caspase-3 antibody (1:500, Abcam) which detects cleaved active caspase-3, and Annexin-V antibody (1:300, Abcam). After several steps of washing, the sections were incubated with goat polyclonal anti-rabbit secondary antibody (1:600) for 2 h at room temperature. Antibody staining was developed using the DAB detection kit (Invitrogen, USA), and counterstained by hematoxylin and examined under a light microscope. Caspase-3 and Annexin V positivity was quantified by calculating the mean percent of stain positive to total area in 5 randomly selected non-overlapping fields in each slide.

Immunofluorescence

IF was performed using human pancreatic tissue fragment sections and acinar cell blocks. Briefly, the sections were deparaffinised with xylene following which they were washed and rehydrated. Tissue sections were fixed in 100% chilled methanol. Blocking was performed by 4% fetal bovine serum. After blocking, sections were incubated with primary antibodies for amylase (1:300), TNF-a (1:500), and IL-6 (1:400). After overnight incubation at 4 °C, the tissue sections were stained with fluorescent-tagged secondary antibody (1:1000) at room temperature by avoiding light for 2hrs. Finally, sections were washed and mounted with mounting medium containing DAPI. Imaging was performed by capturing 5 randomly selected images from each section from each experimental set with the Olympus IX71 fluorescence microscope, and fluorescent images were captured using microscope and CARVII bioimager (BD Biosciences) using IP LAB software (BD Biosciences). Positive cells were counted by identifying the target positive cells in amylase positive cells from the total cell numbers and data expressed as mean (SEM).

Western blotting

Western blotting was performed in proteins extracted from control and treated pancreatic tissue homogenates as described earlier [20,21]. Proteins were separated by using SDS-PAGE gels and electrophoretically transferred to the nitrocellulose membranes with the aid of IBlot system (Invitrogen). The membrane was blocked by incubating for 1hr in 5% (w/v) non-fat dry milk in Tris-buffered saline (TBS pH-7.5). The blots were then probed with primary antibody against to RIP3 in TBST (5% (w/v) non-fat dry milk with (0.05% (v/v) Tween 20). Membranes were washed with TBST three times and finally incubated with secondary antibody labelled with HRP for 1 h at room temperature. The blots were developed using enhanced chemiluminescence (ECL) detection kit (Invitrogen, USA). Expression of protein was quantified by calculating the intensities against beta-actin as internal control using band densitometry analysis.

Cytokines quantification

Cytokines (IL-6, IL-8, IL-10, IL1B, TNF-a, IL-12P70) released from injured acinar tissue, as well as in serum from patients with AP were quantified using BD FACS ARIA II system provided with BD cytometric bead array (CBA) human inflammatory cytokines kit (BD Biosciences, USA) according to manufacturer’s instructions.

Recruitment of AP patients and controls

To place our results from the experimental work into clinical perspective, we looked for circulating cytokine concentrations in 33 patients who were admitted with and without alcohol induced AP within the first 72 h of symptom onset, at Asian Institute of Gastroenterology Hospital. All the clinical parameters including patient demographics, presence of SIRS, the severity of AP (according to the Revised Atlanta Criteria), distribution of persistent organ failure, and mortality were recorded. We also collected blood samples from 10 blood donors that were used as controls for plasma cytokine assay. Collected blood samples from patients and controls were processed immediately wherein serum was separated and stored in -80 ° C for further FACS analysis.

Statistical Methods

A database was generated in Microsoft Excel for Mac (Ver. 14.6.9, Redmond, USA) and all statistical analyses were performed using SPSS (Ver. 20; Chicago, USA). Continuous variables were expressed as mean with a standard error of mean (SEM), while the categorical variables were represented as proportions (percentage). Distribution of continuous data was tested by the Shapiro-Wilk test. For comparing continuous variables between two groups, the student’s ‘t’ test (for normally distributed) or Mann-Whitney U test (for non-normally distributed) was applied. For comparison of cytokines in the controls and patient with AP (alcoholic and non-alcoholic groups), Kruskal-Wallis test was used with Dunn Bonferroni post hoc comparison between groups, as applicable. Cytokine values were log- transformed if the cytokine was grossly skewed, to maintain uniformity in interpretation. Comparison of categorical variables was done using the Fischer’s Exact or the chi-square tests. A two-tailed ‘p’ value of < 0.05 was considered statistical significance.he

Results

Procurement of human pancreatic samples

We used pancreatic tissue samples obtained from 47 patients across 4 years for experiments and discarded around 30 samples due to nonviability or functional impairment. The mean (SD) age of the patients was 52.2 (12.1) yrs and 34 (72.3%) were males. Thirty five (74.5%) patients underwent Whipples’s pancreaticoduodenectomy while the rest underwent distal pancreatectomy with or without splenectomy. The indications of surgery were ampullary adenoma with high grade dysplasia, indeterminate distal biliary stricture, cholangiocarcinoma, periampullary mass lesions, pancreatic NET, duodenal GIST, duodenal NET, IPMN, and symptomatic serous cystadenoma.

Human pancreatic acinar injury with alcohol metabolites

To standardise the optimal dose to induce pancreatic injury with alcoholic metabolites in-vitro, we initially conducted a doseresponse assay by treating the human pancreatic tissue fragments with increasing concentrations (10, 20, 50 μm) of FAEEs and identified the optimal dose that could induce pancreatic injury in in-vitro condition. As shown in Fig. 1, treatment of pancreatic tissue with 50 μm FAEE resulted in significant trypsin activity compared to the lower doses of FAEE. Furthermore, we confirmed the presence of pancreatic acinar cell injury by histological examination and showed more pancreatic damage compared to other concentrations. This indicated that FAEE could cause human pancreatic acinar injury, and 50 μm dose of FAEE could be used as optimal noxious stimuli to study the mechanisms in in-vitro setting of alcoholic pancreatitis.

Fig. 1.

a) Histogram showing trypsin activity as an indicator of acinar injury in response to different concentrations (10 μm, 20 μm, 50 μm) of FAEE. * indicates p = 0.02 and ** indicates p = 0.03 when compared to controls. (b) Time-dependent increase in amylase secretion into the medium after exposure ofhuman pancreatic acinar tissue to 50 μM FAEE for different time-points. * indicates p = 0.01 when compared to controls. (c, d) Histograms showing a time-dependent increase in trypsin activity (* indicates p = 0.03 and ** indicates p = 0.03 when compared to controls) and cathepsin B activity (* indicates p = 0.04 and ** indicates p = 0.04 when compared to controls) within the acinar tissue after exposure to 50 μM of FAEE for different time-points. Results are based on n = 6 experiments. Error bars indicate standard error of means (SEM).

As shown in Fig. 1 we observed increased amylase secretion into the medium from treated pancreatic tissue compared to controls after 30 min of FAEE exposure. We also observed a significant timedependent increase in the trypsin and cathepsin B activities in pancreatic tissue exposed to 50 μm FAEE.

We further performed H&E staining in FAEE treated pancreatic acinar tissue. FAEE exposed pancreatic tissue sections showed a time-dependent increase in the acinar injury (Fig. 2a). As shown in Fig. 2b, there were significant differences in the areas of injury between the control and treated tissue after 2 and 4hrs of exposure, which progressed further by increasing the time of exposure. H&E staining along with biochemical parameters confirmed that FAEE can cause acinar damage and can be used as noxious stimuli to induce in-vitro acinar cell injury.

Fig. 2.

a) Representative H&E micrographs (10× magnification) of pancreatic tissue slices showing injury after exposure to 50 μm FAEE for 1, 2 and 4hrs. b) Quantitative data of tissue injury from n = 4 experiments indicate a statistically significant difference between FAEE treated and control tissue at 2 and 4 h respectively. * and ** indicates p values of 0.009 and 0.03 respectively, suggesting statistically significant difference when compared to the respective controls. Error bars indicate standard error of mean (SEM).

Cathepsin B redistribution in FAEE treated human pancreatic tissue

Previous experimental acute pancreatitis studies conducted in rodent models had demonstrated that zymogen and lysosomal contents could co-localize and cathepsin B in co-localized organelles could activate the trypsinogen to trypsin [18]. We evaluated for the co-localization of zymogens and lysosomes in in-vitro experimental human acute pancreatitis setting, and observed redistribution of cathepsin B from a lysosome enriched fraction to zymogen enriched fraction. As shown in Fig. 3a, in FAEE treated pancreatic tissue cathepsin B was significantly higher in the zymogen enriched fraction with a corresponding reduction in the lysosome enriched fraction. This was further validated by checking the ratios of cathepsin B in the zymogen and lysosomal fractions (zymogen: lysosome ratio) which was maximum at 30mins after the exposure of FAEE (Fig. 3b). These observations confirmed that alcohol metabolites could cause a redistribution of cathepsin B in human pancreatic acinar cell.

Fig. 3.

a) Bar diagram representing the redistribution of lysosomal cathepsin B from lysosomal to zymogen compartment after stimulation with 50 μM FAEE for different time points. Each bar indicates the Cathepsin B activity per mg of protein in the zymogen (black bar) and lysosomal (grey bar) enriched fractions. b) Bar diagram representing the zymogen to lysosome ratio of cathepsin B after treatment with 50 mM FAEE for different time points. Cathepsin B activity in the zymogen fraction was significantly higher after 30 min of FAEE induction.

* in (a) indicates statistically significant difference between zymogen and lysosome enriched fraction. * in (b) indicates statistically significant difference of the Zymogen:Lysosome fraction compared to untreated sample, while

** indicates statistically significant difference compared to 15mins and 60mins sample.

Data are based on n = 5 experiments.

Identification type of human pancreatic acinar injury in response to FAEE

Experimental AP studies reported that severity of pancreatitis directly correlated with the type of acinar cell injury (apoptosis/necrosis) [19]. To identify the mechanism of early acinar injury in in-vitro experimental human alcoholic pancreatitis, we performed IHC for apoptosis markers such as caspase-3 and Annexin-V, in FAEE treated pancreatic tissue sections. As shown in Fig. 4a (upper panel), we observed an incremental expression of caspase-3 in pancreatic tissue sections treated with FAEE for 1and 2hrs. To confirm the presence of apoptosis in the FAEE exposed human acinar tissue, we also evaluated for the expression of the early apoptotic marker, Annexin-V. FAEE treated tissues showed acinar cells with intense Annexin-V positive staining which was incremental from 1hr to 2 h of exposure. Fig. 4b and c shows the quantitative data of the expression of caspase-3 and Annexin-V respectively. These observations indicate that FAEE could induce apoptosis in human pancreata, early in the pathogenesis of AP.

Fig. 4.

a) Representative IHC pictures (40 x) showing active caspase-3 and Annexin-V expression in human pancreatic tissue treated with 50 μm FAEE at for 1 and 2hrs. b) and c) Histograms representing the quantitation of caspase-3 and Annexin-V in FAEE treated pancreatic tissue compared to control in each slide from n = 3 experiments. * and ** indicates p = 0.001 each in (b) and p < 0.0001 each in (c) based on 3 individual experiments. Error bars indicate standard error of mean (SEM).

Secretion of cytokines from FAEE treated human pancreatic tissue

Clinical studies have shown that acute pancreatitis is associated with increased circulating levels of inflammatory cytokines such as TNF-α, IL-1β, IL-6. These mediators are believed to participate in pathophysiology and disease progression; and responsible for the development of SIRS and early organ failure. However, the earliest source of these cytokines following acinar injury have been speculative. In our previous study, we demonstrated that in-vitro exposure of pancreatic tissue to TLCS could cause secretion of IL-8 and IL-6 from the pancreatic acinar cells incrementally. As shown in Fig. 5a, stimulation of pancreatic tissue with FAEE also showed secretion of IL-6 and IL-8 into the medium in a time-dependent incremental manner. Table 1 depicts the concentration of the cytokines IL-6, IL-1b and IL-8 detected in the media of FAEE treated tissues in a time-dependent manner. Cytokine secretion was maximum at 24hrs after induction of AP. We further conducted IF studies in isolated acinar cells to confirm whether acinar cells are the primary source for these secreted cytokines. As shown in Fig. 5b–d, human pancreatic acini treated with FAEE showed significantly higher intra acinar expression of IL-6 (p = 0.007) and TNF-a (p = 0.01) compared to control acini, indicating that alcohol metabolites can activate acinar cells to release cytokines.

Fig. 5.

a) FACS images representing cytokine in acinar induction media. Proinflammatory cytokines (IL-6, IL-8) concentration were at baseline in control and these cytokines concentrations were increased along with time and showing maximum increase at 18–24hrs of treatment with 50 μM FAEE. b) Representative immunofluorescence (IF) images with scale bar of 50 μm indicating IL-6 and TNF-a expression in isolated human acinar cells treatment with 50 μm FAEE for 1hr. Green fluorescence in the second column indicates amylase expression, while red fluorescence in the third column indicates IL-6 and TNF-a respectively, Yellow fluorescence in the last column as indicated by white arrows are merged images of IL-6, and TNF-a with amylase implying the location of IL-6 and TNF- a within the acini. Positive IL-6 and TNF-a positive cells were counted in amylase positive cells against the number of nuclei (DAPI stain) in at least 5 randomly selected images in each slide from each experiment (n = 3), using the Olympus IX71 fluorescence microscope, and all the fluorescent images were captured using the CARVII bioimager (BD Biosciences). Results were expressed as number of positive cells per 100 nuclei. c) and d) Histograms showing the quantitative representation of IL-6 and TNF-a positivity respectively, which was significantly higher than that of controls (p = 0.007 and 0.01 respectively) in isolated acini. Error bars indicate SEM.

Table 1. Time dependent concentration of cytokines secreted into the media after induction of FAEE-induced pancreatic acinar injury.

Time (in hours) after induction of acinar injury with FAEE	Mean (SEM) cytokine concentration in the media (in pg/ml)
	IL-8	IL-1β	IL-6
Untreated	7.5 (5.8)	5.0 (2.0)	1.5 (0.4)
3	68.6 (20.7)	8.0 (2.2)	120.0 (87.77)
6	181.7 (94.7)	14.3 (5.1)	209.9 (159.6)
18	1995.3 (1038.7)	4.6 (0.8)	596.1 (356.0)
24	9644.2 (885.6)	18.4 (9.6)	4987.9 (2805.1)
36	124.8 (58.2)	2.9 (1.5)	42.3 (22.5)

Cytokine expression in patients with acute pancreatitis

To place our experimental data into clinical perspective, we quantified the circulating cytokine concentration in patients with and without alcohol induced AP patients who were admitted within 72hrs of symptom onset. Table 2 shows the demography and clinical characteristics of study patients. Out of the total of 33 recruited patients, 14 patients (5 with MAP and 9 with SAP) were alcohol induced. Among the non-alcohol induced patients, the risk factor of AP were gall stones (n = 10), hypertriglyceridemia (n = 1), medications (n = 1) and idiopathic (n = 9). Overall mortality was 27.3%, out of which 60% (3/5) of alcohol induced and 50% (2/4) of the nonalcohol induced AP patients died within the first week of illness.

Table 2. Clinical characteristics of patients with acute pancreatitis admitted within 72hrs of symptom onset.

Clinical parameters	Alcohol-induced AP (n = 14)	Non-alcohol related AP (n = 19)
Age (in yrs) [Median, IQR]	37.5 (29.0-41.3)	38.0 (22.0-54.0)
Male gender (n; %)	13 (92.9)	15 (78.9)
Disease severity       MAP (n; %)	0 (0)	5 (26.3)
MSAP (n; %)	5 (35.7)	5 (26.3)
SAP (n; %)	9 (64.3)	9 (47.4)
BUN (mg/dl) at admission [Median, IQR]	38.0 (27.0-52.8)	40.0 (20.0-65.0)
Creatinine (mg/dl) at admission [Median, IQR]	1.8 (1.0-2.6)	1.0 (0.7-2.5)
SIRS score at admission [Median, IQR]	3.0 (2.0-3.0)	2.0 (2.0-3.0)
BISAP score at admission [Median, IQR]	3.0 (3.0-3.25)	3.0 (3.0-4.0)
APACHE II score at admission [Median, IQR]	9.0 (7.3-19.8)	8.0 (5.0-18.0)
Renal failure (n; %)	8 (57.4)	8 (42.1)
Respiratory failure (n; %)	3 (21.4)	1 (5.3)
Multiorgan failure [Median, IQR]	3 (21.4)	1 (5.3)
Need for ICU (n; %)	11 (78.6)	10 (52.6)
Total ICU days [Median, IQR]	8.0 (6.0-9.0)	7.5 (5.0-14.5)
Total hospital days [Median, IQR]	11.5 (6.3-19.3)	9.0 (5.0-18.0)
Infected necrosis [Median, IQR]	9 (64.3)	7 (36.8)
Need for intervention (n; %)	3 (21.4)	3 (15.8)
In-hospital death (n; %)	5 (35.7)	4(21.1)
First week mortality (n; %)	3 (21.4)	2 (10.5)
Day of mortality from onset [Median, IQR]	4.0 (1.5-14.0)	5.5 (2.3-6.3)

Fig. 6 demonstrates the concentration of cytokines among patients with alcohol and non-alcohol induced AP compared to healthy controls. Cytokine analysis revealed four cytokines (IL-6, IL- 8, IL-10, TNF-α) to be elevated significantly in all the AP patients compared to controls. The most predominant cytokine that was elevated in these patients was IL-6, followed by IL-8, TNF- α and IL- 10. Even though IL-6 and IL-8 were higher than controls in both alcohol and non-alcohol induced AP, TNF- α was significantly higher than controls only in the patients with alcohol induced AP (Fig. 6 c). Marked elevation of IL-6 was observed in patients with alcohol induced AP who had SIRS and organ failure (data not shown).

Fig. 6.

Box and whisker plots showing circulating cytokine (a) IL-6, (b) IL-8,(c) IL-10, (d) TNF-a concentration in healthy controls and patients with and without alcoholic AP.

Sample size of patients included for the circulating cytokine assay was 33 (14 alcohol induced AP while 19 non-alcoholic AP).

Discussion

Experimental AP studies involving rodent models have reported that ethanol metabolites can induce pancreatic digestive enzyme secretion and plays an important role in causing pancreatic injury [22]. Several studies proposed varied theories to explain the deleterious effects of alcohol causing AP, such as release of free fatty acids by FAEE, increased intra cellular Ca²⁺, mitochondrial dysfunction, increased oxidative stress, shift of acinar cell apoptosis to necrosis and association fat [23–25]. Besides this, it has also been shown that the ethanol exposure resulted in the activation of inflammatory transcription factors such as NF-kB, AP-1 and other inflammatory molecules thereby resulting in increased trypsin release [26]. FAEEs are non-oxidative metabolites of alcohol that are synthesized maximally in the pancreas and liver after alcohol consumption [27]. The common FAEEs include the saturated esters of ethyl palmitate (16:0) and ethyl stearate (18:0), and unsaturated esters of ethyl oleate (16:1), ethyl palmitoleate (18:1), ethyl linoleate (18:2), and ethyl arachidonate (20:4). Of these, ethyl palmitate (16:0) is the most abundant and maintains high stability [28]. However, the early mechanism contributing to pancreatic injury and pancreatic injury mediated systemic inflammation (SIRS) in human alcoholic acute pancreatitis is not well studied. We, therefore, focused to study the mechanism of early pancreatic injury in alcoholic AP.

To develop in-vitro alcoholic pancreatitis model, we procured normal pancreatic tissues and induced pancreatic injury with FAEE as described in the methods. It has been reported that FAEEs, products of non-oxidative ethanol metabolites, could induce pancreatic injury both in-vivo [28] and in-vitro [29]. We demonstrated the presence of pancreatic injury with FAEE as evidenced by an elevation of trypsin and cathepsin in pancreatic tissue induction medium. We observed early re-distribution of cathepsin B from lysosome enriched to zymogen enriched compartment after FAEE induced pancreatic injury. This is similar to previous studies reported in rodent models of AP [30]. The redistributed cathepsin B in acinar cells could be the possible factor which activates trypsinogen to active trypsin thereby leading to acinar injury. While the precise mechanism by which FAEEs induce pancreatic injury is still speculative, several mechanisms have been implicated, including mitochondrial dysfunction by concentration-dependent uncoupling of oxidative phosphorylation, alteration of membrane permeability and increased fragility of lysosomal membranes [31–33]. The latter mechanisms aligns to our results that suggested co-localization, i.e. redistribution of cathepsin B from lysosomal to the zymogen compartment. Interestingly, a recent study has clearly shown that even though FAEEs can cause more severe pancreatic injury to the pancreatic acinar cells, they in fact cause milder damage compared to their parent unsaturated fatty acids in equivalent concentrations [34]. This could explain our observations of mild early injury to the human pancreatic acinar cells in the form of apoptosis.

In our previous study, we reported the presence of early autophagy in bile acid-induced AP. However, we did not observe autophagy in FAEE induced AP. Subsequently, we attempted to evaluate other forms of acinar cell injury in the FAEE treated pancreatic tissues. We observed apoptosis in the FAEE treated human pancreatic acini as evidenced by the markers Annexin-V and caspase-3. Surprisingly, this observation was contrary to those reported in experimental animal models where chronic ethanol exposure was shown to inhibit apoptosis and accelerated necrosis [35]. However, these findings in experimental animal models were observed after chronic ethanol exposure, as opposed to acute alcohol exposure in our experimental setting. In our current studies, we observed substantial tissue injury at 4hrs after induction on H&E staining that could represent acinar necrosis. This could have been either as a result of natural time-dependent autolysis or transformation from apoptosis to necrosis [35]. In order to address this issue, we did evaluate for the expression RIP3, a marker of necrotic cell injury, at different time-points after induction of AP by performing Western blot. However, our results were inconclusive and therefore we opted not to include the ambiguous data. Nevertheless, the objective of our study was to evaluate the early mechanisms of alcohol related AP in humans, and, given the findings, alcohol induced human AP is characterized by pancreatic acinar cell apoptosis early in the disease progression. Whether there could be subsequent transformation to acinar cell necrosis at later time points needs to be investigated further.

Following the identification of acinar cell injury with FAEE, we studied for the competence of acinar tissues to produce inflammatory cytokines upon treatment with FAEE. In our previous study, we reported that bile acid could induce the secretion of IL-6 and IL- 8 from the acinar cell. Similar to our previous report, in this study also we observed a time-dependent increase in cytokines secretion (especially IL-6 and IL8) in response FAEE treatment and it was maximum at 18–24hrs after treatment with FAEE. Interestingly, the concentrations of these cytokines were much higher in FAEE treated acinar tissue compared to TLCS treated tissue which we reported earlier. Since we used pancreatic tissue sections in these experiments, it might have been possible that resident immune cells could have contributed to cytokine release. Therefore we performed IF studies to evaluate the localization of the cytokines using freshly isolated acinar cells. We observed presence of TNF-a and IL-6 in FAEE treated acinar cells which confirmed that the cytokines were indeed secreted by the acinar cells. However, even though there was secretion of TNF-α by the acinar cells, we did not observe a parallel elevation in its concentration in the medium. One of the reasons for this observation could be the short half-life of TNF-α. On the contrary, a more plausible and important reason could be the binding of the secreted TNF-alpha to the membrane receptors that precluded their release into the medium. We had demonstrated distribution of TNF-alpha along the basolateral membrane of pancreatic acinar cells compared to interstitial distribution of IL-6 earlier in our studies on biliary acute pancreatitis [14].

In our previous study on bile acid-induced pancreatic injury, we demonstrated that the pancreatic acini secreted cytokine could active circulating PBMCs which could, in turn, mount a systemic inflammatory response syndrome. Observations from the current study make it plausible that in alcohol induced AP also acinar cell injury mediates SIRS in a similar fashion. To place this speculation into a clinical perspective, we evaluated circulating cytokines in patients who were admitted with AP within 72hrs of onset of symptoms. Similar to our observations in biliary AP, we observed a high concentration of IL-6 and IL-8 in patients systemic circulation. This suggests that irrespective of the initial trigger, acinar cell injury is followed by acinar cytokine secretion within the pancreatic interstitial space, which eventually lead to the progression of this local cytokine response towards a systemic inflammatory response (SIRS). Whether the initial type of acinar injury and cytokine secretion are independent or interrelated events in human AP, need to be evaluated further.

Our study had a few limitations. Firstly, we did not perform any experiments to prove that trypsin activation was initiated within the co-localized organelles. However, from previous evidence that human cathepsin B could activate human trypsinogen to active trypsin by cleaving trypsinogen activation peptide [26] and earlier studies from our lab and others on experimental confirmation of co-localization of zymogen and lysosome convinces us that trypsinogen activation was initiated within the co-localized organelles itself. Secondly, we could not obtain confirmatory evidence regarding the presence or absence of another form of acinar cell death, such as necrosis. Further confirmatory studies will have to be conducted to address this issue. Thirdly, the sample size of patients recruited in this study was low. The study site is a tertiary care referral centre for pancreatic diseases. Therefore the majority of the patients that we cater to are the ones with complication who present late in the disease course. This is also the reason why the proportion of patients with moderately severe and severe disease, and mortality rate were higher than what is seen in the general population. Nevertheless, to the best of our knowledge, this study is the first to evaluate the early pathogenesis of AP using treatment of human pancreatic acini with alcohol metabolites (FAEEs). Our model mimics a real-world AP scenario and therefore the results could be extrapolated to clinical AP.

In conclusion, we have demonstrated that in alcohol induced AP in humans, there is early co-localization of zymogen and lysosome resulting in trypsin activation, apoptosis of pancreatic acinar cells, and parallel secretion of pro-inflammatory cytokines. Based on our results, we propose the sequence of events depicted in Fig. 7 that could lead to induction of alcoholic acute pancreatitis in humans. Further confirmation and reinforcement of these sequence of events would identify putative targets for the pharmacotherapy of this currently incurable disease.

Fig. 7.

Schematic representation of the sequence of early mechanistic events in alcohol metabolite (FAEE) induced human acinar injury and progression of systemic inflammatory response. This figure represents the injury within the first 2 h, when apoptosis is predominant.

There could be a possibility of transformation apoptosis to necrosis in the subsequent hours that needs to be evaluated further.

Abbreviations

AP

Acute pancreatitis

SIRS

Systemic inflammatory response syndrome

FAEE

Fatty acid ethyl esters

IHC

Immunohistochemistry

PBS

Phosphate buffered saline

DMEM

Dulbecco’s modified Eagle medium

IF

Immunofluorescence

FFPE

Formalin fixed paraffin embedded

FACS

Fluorescence activated cell-sorting

TNF-α

Tumour necrosis factor-alpha

IL-6

interleukin-6