Deletion of the WD40 domain of ATG16L1 exacerbates acute pancreatitis, abolishes LAP-like non-canonical autophagy and slows trypsin degradation

Abstract

The WD40 domain (WDD) of ATG16L1 plays a pivotal role in non-canonical autophagy. This study examined the role of recently identified LAP-like non-canonical autophagy (LNCA) in acute pancreatitis. LNCA involves rapid single-membrane LC3 conjugation to endocytic vacuoles in pancreatic acinar cells. The rationale for this study was the previously observed presence of trypsin in the organelles undergoing LNCA; aberrant trypsin formation is an important factor in pancreatitis development. Here we report that the deletion of WDD (attained in ATG16L1[E230] mice) eliminated LNCA, aggravated caerulein-induced acute pancreatitis and suppressed the fast trypsin degradation observed in both a rapid caerulein-induced disease model and in caerulein-treated isolated pancreatic acinar cells. These experiments indicate that LNCA is a WDD-dependent mechanism and suggest that it plays not an activating but a protective role in acute pancreatitis. Furthermore, palmitoleic acid, another inducer of experimental acute pancreatitis, strongly inhibited LNCA, suggesting a novel mechanism of pancreatic lipotoxicity.Abbreviation: AMY: amylase; AP: acute pancreatitis; CASM: conjugation of Atg8 to single membranes; CCK: cholecystokinin; FAEE model: fatty acid and ethanol model; IL6: interleukin 6; LA: linoleic acid; LAP: LC3-associated phagocytosis; LMPO: lung myeloperoxidase; LNCA: LAP-like non-canonical autophagy; MAP1LC3/LC3: microtubule-associated protein 1 light chain 3; MPO: myeloperoxidase; PMPO: pancreatic myeloperoxidase; POA: palmitoleic acid; WDD: WD40 domain; WT: wild type.

Keywords: Amylase; LC3-associated phagocytosis; caerulein; cholecystokinin; endocytic vacuoles; palmitoleic acid.

Figure 1.

GFP-LC3 conjugation to endocytic vacuoles in pancreatic acinar cells isolated from GFP-LC3 mice and GFP-LC3 ATG16L1[E230] mice. Cells were incubated in physiological HEPES-buffered solution containing 100 pM CCK and Texas Red dextran (TRD) for 30 min at 35°C. GFP-LC3 ATG16L1[E230] is abbreviated to GFP-LC3 E230 on the graph. In this experiment we analyzed 425 vacuoles from 138 GFP-LC3 cells (79 GFP-LC3-conjugated vacuoles observed) and 596 vacuoles from 177 GFP-LC3 ATG16L1[E230] cells (only 2 GFP-LC3-conjugated vacuoles observed). In these experiments 5 GFP-LC3 mice and 6 GFP-LC3 ATG16L1[E230] mice were utilized for cell isolation. (A) the bars reveal the proportion of GFP-LC3-conjugated vacuoles (mean values ± standard error). (B) shows an example of a GFP-LC3-conjugated endocytic vacuole in an acinar cell isolated from a GFP-LC3 mouse and the lack of GFP-LC3 conjugation to an endocytic vacuole in an acinar cell isolated from a GFP-LC3 ATG16L1[E230] mouse. GFP fluorescence is shown in green color. Endocytic vacuoles were revealed by the fluorescence of endocytosed TRD (red color). TL on the left panels indicates transmitted light. Scale bars: 10 µm. Expanded fragments, containing vacuoles indicated by white arrows, are shown on the right panels.

Figure 2.

Severity of acute pancreatitis induced by a high dose of caerulein in wild-type and ATG16L1[E230] mice. The figure shows parameters characterizing the severity of acute pancreatitis (AP) in mice with deficient non-canonical autophagy (ATG16L1[E230] mice, abbreviated to E230 on the graph) and wild-type littermates (WT). Experimental AP was induced by 7 hourly intraperitoneal injections of caerulein (50 µg/kg); 12 ATG16L1[E230] mice and 13 WT mice were utilized in these experiments. Animals were humanely sacrificed 8 h after the first injection. Control experiments involved intraperitoneal injections of vehicle solution without caerulein (5 ATG16L1[E230] mice and 5 WT mice were used). Symbols above the bars illustrate the outcome of a Mann-Whitney test; symbol * indicates statistical significance (p < 0.05), ns indicates that the difference was not statistically significant. Specific p values were the following: serum AMY p = 0.011, pancreatic MPO (PMPO) p = 0.002, pancreatic trypsin p = 0.041, IL6 p = 0.37, lung MPO (LMPO) p = 0.64, total histopathology score p = 0.55 (information about the components of the histopathology score is summarized in Figure S2).

Figure 3.

Severity of acute pancreatitis induced by a moderate dose of caerulein in wild-type and ATG16L1[E230] mice. The figure shows parameters characterizing the severity of acute pancreatitis (AP) in mice with deficient non-canonical autophagy (ATG16L1[E230] mice, abbreviated to E230 on the graph) and wild-type littermates (WT). Experimental AP was induced by 7 hourly intraperitoneal injections of caerulein (25 µg/kg); 12 ATG16L1[E230] mice and 18 WT mice were utilized in these experiments. Animals were humanely sacrificed 8 h after the first injection. Control experiments involved intraperitoneal injections of vehicle solution without caerulein (6 ATG16L1[E230] mice and 6 WT mice were used). Symbols above the bars illustrate the outcome of a Mann-Whitney test; symbol * indicates statistical significance (p < 0.05), ns indicates that the difference was not statistically significant. Specific p values were the following: serum AMY p = 0.047, pancreatic MPO (PMPO) p = 0.044, pancreatic trypsin p = 0.0002, IL6 p = 0.01, lung MPO (LMPO) p = 0.054, total histopathology score p = 0.65 (information about the components of the histopathology score is summarized in Figure S2).

Figure 4.

Deletion of WDD of ATG16L1 slows trypsin degradation and potentiates cell death. (A) rapid changes in the pancreatic trypsin levels in wild-type (WT) and ATG16L1[E230] (abbreviated to E230 on the graph) mice were induced by one intraperitoneal caerulein injection (50 µg/kg). Experiments were conducted on mice fed ad libitum. Trypsin levels were measured in pancreata obtained from untreated mice (i.e. before the caerulein injections, 0 h) as well as at 1 h, 2 h, 4 h and 6 h after the injection. Each time point on the graph shows the results of measurements (± standard error) obtained from 6 mice (normalized to the mean values from WT at 0 h for each of experiments). Symbol * indicates statistical significance (p < 0.05), ns indicates that the difference was not statistically significant. Trypsin levels at the 1 h time points were not significantly different between WT and E230 mice (p = 0.89). Note the substantial and statistically significant decline in the trypsin levels between 1 h and 2 h in WT mice (p = 0.005), and the absence of such decline in the WDD and LNCA-deficient E230 mice (p = 0.77). The significant difference in the trypsin levels between WT and ATG16L1[E230] developed at 2 h and disappeared at 6 h after the single injection. We have repeated critical early points of this experiment (0, 1 and 2 h) using mice fasted for 12 h (see Figure S3A). The rational for these additional experiments was to test putative interference from canonical autophagy that could be initiated in the fasted mice. Similar results were observed in experiments conducted on mice fed at libitum and fasted mice (Figure S3A). (B) apoptosis and necrosis of pancreatic acinar cells isolated from wild-type and ATG16L1[E230] mice. Apoptosis (right panel) and necrosis (left panel) were assessed at 14 h end point as the fluorescence of caspase 3/7 and propidium iodide, respectively. In these experiments 18 wild-type (WT) mice and 13 ATG16L1[E230] mice (E230) were utilized. The box plots (median line, box as first to third quartile, and whiskers as 1.5 times inter-quartile range) show the fluorescence measurements of the treatment groups divided by the fluorescence recorded in the corresponding vehicle control group. The p values for apoptosis were the following: 0.79 for CCK 10 nM, 0.98 for CCK 100 nM and 0.09 for cerulein (CER) 100 nM. The p values for necrosis were the following: 0.043 for CCK 10 nM, 0.054 for CCK 100 nM and 0.040 for cerulein (CER) 100 nM.

Figure 5.

Effect of palmitoleic acid on LAP-like non-canonical autophagy and cell death of pancreatic acinar cells isolated from wild-type and ATG16L1[E230] mice. (A) effect of palmitoleic acid on GFP-LC3 conjugation to endocytic vacuoles in pancreatic acinar cells. Cells were isolated from GFP-LC3 transgenic mice and incubated in physiological HEPES-buffered solution containing Texas red dextran (TRD) at 35°C. Symbol * indicates statistically significant difference from the values obtained in experiments with 10 nM CCK (without POA). The bars reveal the proportion of GFP-LC3-conjugated vacuoles (mean values ± standard error). Application of 10 nM CCK for 30 min resulted in the generation of endocytic vacuoles and GFP-LC3 conjugation to the endocytic vacuoles (405 vacuoles, 101 cells were analyzed, 5 mice were used for cell isolation in these experiments). In all experiments shown in this figure the extracellular solution contained 1% ethanol (utilized as a vehicle for POA). In separate experiments we found that this ethanol concentration had no effect on the number of vacuoles (p = 0.97, not shown) and the proportion of GFP-LC3-conjugated vacuoles (p = 0.92, not shown) produced by 10 nM of CCK; 665 vacuoles from 177 cells (derived from 9 mice) treated with 1% ethanol (added to the extracellular solution described in the section “isolation of pancreatic acinar cells”) and 708 vacuoles from 163 cells (derived from 6 mice) maintained in the ethanol-free extracellular solution, were utilized in these experiments. Notably, we did not observe GFP-LC3-conjugated endocytic vacuoles in cells treated with 200 µm palmitoleic acid (POA) (133 vacuoles, 61 cells were analyzed, 5 mice were used for cell isolation). Furthermore, addition of 200 µm POA to the solution containing 10 nM CCK completely inhibited GFP-LC3 conjugation to endocytic vacuoles (in this experiment both CCK and POA were applied simultaneously for 30 min (177 vacuoles, 71 cells were analyzed, 5 mice were used for cell isolation). We next conducted an experiment involving application of 10 nM CCK followed by the application of POA after a delay of 10 min (in the continuous presence of 10 nM CCK, POA was present for 20 min). In this configuration POA also strongly inhibited GFP-LC3 conjugation to the endocytic vacuoles (p = 0.008; 440 vacuoles; 130 cells were analyzed; 6 mice were used for cell isolation). Effects of lower concentrations of POA are shown on the figure S6. (B) effect of palmitoleic acid on apoptosis and necrosis of pancreatic acinar cells isolated from wild-type and ATG16L1[E230] mice. In these experiments we utilized 100 µm of POA. Apoptosis (right panel) and necrosis (left panel) were assessed at 14 h end point as the fluorescence of caspase 3/7 and propidium iodide, respectively. In these experiments 16 wild-type (WT) mice and 11 ATG16L1[E230] mice (E230) were utilized. The box plots (median line, box as first to third quartile, and whiskers as 1.5 times inter-quartile range) show the fluorescence measurements of the treatment groups divided by the fluorescence recorded in the corresponding vehicle control group; ns indicates that the difference was not statistically significant. The p value for apoptosis was 0.27. The p value for necrosis was 0.51.

Figure 6.

Severity of acute pancreatitis induced by palmitoleic acid and ethanol in wild-type and ATG16L1[E230] mice. The figure shows parameters characterizing the severity of acute pancreatitis in WDD and LNCA-deficient ATG16L1[E230] mice (abbreviated to E230 on the graph) and wild-type littermates (WT). Experimental acute pancreatitis was induced by two hourly intraperitoneal injections of 1.35 g/kg ethanol and 150 mg/kg palmitoleic acid (POA) this type of experimental pancreatitis is abbreviated as the FAEE model; 11 ATG16L1[E230] mice and 11 WT mice were utilized in these experiments. Animals were humanely sacrificed 24 h after the first injection. Control experiments involved intraperitoneal injections of vehicle solution (5 ATG16L1[E230] mice and 5 WT mice were used). Symbols above the bars illustrate the outcome of a Mann-Whitney test; symbol ns indicates that the difference was not statistically significant (p > 0.05). We did not observe significant difference in any of the measured parameters. Specific p values were the following: for serum AMY p = 0.09, for pancreatic MPO (PMPO) p = 0.19, for pancreatic trypsin p = 0.55, for IL6 p = 0.89, for lung MPO (LMPO) p = 0.99, for total histopathology score p = 0.44; p values for the components of the total histopathology score were the following: for edema p = 0.99 (not shown), for leukocyte infiltration p = 0.61 (not shown) and for necrosis p = 0.5 (not shown).