Reprogramming of cholesterol sensing in epithelial cells supports pancreatic inflammation.

Our previous work demonstrates that ACLY supports inflammation-induced pancreatic carcinogenesis [ 13 ], but its role in the pathogenesis of pancreatitis remained unknown. To this end, we generated Mist1-Cre ER ;Acly fl/+ or Mist1-Cre ER ;Acly fl/fl mice that were administered tamoxifen to obtain ACLY-proficient and -deficient animals (acinar cell-restricted deletion; named ACLY WT and ACLY Δacini , respectively), which were then iteratively injected with cerulein (CER) to induce pancreatitis ( Figure 1A ). Figure 1 Genetic ablation of Acly alleviates pancreatitis through reduction of cholesterol accumulation. A. The upper panel illustrates experimental design: mice with indicated genotypes were administered tamoxifen for two consecutive days to elicit genetic ablation of ACLY. After one week, mice were then injected with cerulein. The bottom panel shows representative images of hematoxylyn and eosin staining of pancreas sections. Scale bar: 20 µm. B. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. C. Percentage of tissue-infiltrating cells was assessed in images from A (CER-treated animals). Each dot corresponds to one image analyzed. ∗ denotes statistical significance (p‹0.05; Student’s t-test). D. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. E. Quantification of trypsin activity in pancreata harvested from mice described in Figure 1A. F. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. G. Absolute levels of pancreatic cholesterol were quantified from tissue samples as in A. ∗ denotes statistical significance (p‹0.05; Student’s t-test). Figure 1 Genetic ablation of Acly alleviates pancreatitis through reduction of cholesterol accumulation. A. The upper panel illustrates experimental design: mice with indicated genotypes were administered tamoxifen for two consecutive days to elicit genetic ablation of ACLY. After one week, mice were then injected with cerulein. The bottom panel shows representative images of hematoxylyn and eosin staining of pancreas sections. Scale bar: 20 µm. B. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. C. Percentage of tissue-infiltrating cells was assessed in images from A (CER-treated animals). Each dot corresponds to one image analyzed. ∗ denotes statistical significance (p‹0.05; Student’s t-test). D. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. E. Quantification of trypsin activity in pancreata harvested from mice described in Figure 1A. F. Western blotting of whole-pancreas lysates for indicated proteins. Mouse genotypes and treatments are indicated and experimental replicates shown. G. Absolute levels of pancreatic cholesterol were quantified from tissue samples as in A. ∗ denotes statistical significance (p‹0.05; Student’s t-test). ACLY was efficiently ablated from tamoxifen-injected Mist1-Cre ER ;Acly fl/fl pancreata, with only minimal residual levels ( Figure 1B ). CER administration induced widespread inflammation, as expected: histological examination showed immune cell infiltration, dilation of interstitial space, acinar cell loss together with impaired autophagic flux ( Figure 1A ; Suppl. Fig. 1A ) in line with previous reports [ [22] , [23] , [24] , [25] , [26] ]. Overall, histological examination of pancreata revealed that ACLY Δacini mice showed fewer infiltrating cells ( Figure 1C ; Suppl. Fig. 1B ). We also observed more and wider intra-acinar vacuoles, although acinar cell morphology was minimally impacted. ACLY-deficient pancreata also exhibited reduced levels of lipidated LC3 and decreased buildup of ubiquitinated proteins ( Figure 1B ), which are indicative of an improved autophagic flux. Importantly, ACLY ablation markedly lowered CER-stimulated ER stress ( Figure 1D ), while the release of digestive enzymes was blunted ( Figure 1E-F ). Together these data indicate that several molecular/cellular markers of pancreatitis improve in mice lacking ACLY in acinar cells. ACLY produces acetyl-CoA, which is used for the de novo generation of lipids and for histone acetylation [ 13 ]. We speculated that acetyl-CoA production may support the pathogenesis of pancreatitis feeding these signaling/metabolic pathways. To test this hypothesis, we first interrogated histone acetylation to find that was only marginally impacted by ACLY ablation in acinar cells ( Suppl. Fig. 1C ). In contrast, we found elevated cholesterol levels upon induction of experimental pancreatitis in ACLY WT but not ACLY Δacini mice ( Figure 1G ). Triglycerides levels were not affected ( Suppl. Fig. 1D ). Interestingly, we observed that genetic ablation of Acly in purified acinar cells transcriptionally downregulates enzymes in the mevalonate pathway (MVP) ( Suppl. Fig. 1E ) suggesting that the absence of ACLY impairs cholesterol biosynthesis not exclusively through halted anabolic carbon flux. This demonstrates that de novo cholesterol biosynthesis is boosted in CER-treated acinar cells. Enhanced channeling of carbon atoms toward the mevalonate pathway (MVP) likely requires both whole-cell metabolic rewiring and upregulation of MVP enzymes. We previously showed that human and murine pancreatitis elicits the elevation of epithelial cells’ cholesterol; we linked augmented levels to impaired trafficking in response to inflammation [ 5 ]. Because both statins administration and ACLY deletion improve the outcome of experimental pancreatitis ([ 5 ] and Figure 1 , respectively), we set out to understand the molecular underpinnings of cholesterol dyshomeostasis in pancreatitis. The response of epithelial cells to inflammation and their contribution to pathogenesis have been difficult to parse. Insight can now be gained by single cell-sequencing datasets that enable the molecular characterization of distinct cell types within a composite tissue. We analyzed multiple datasets [ 20 , [27] , [28] , [29] ]; while heterogeneous in terms of sample harvesting, all used cerulein to induce acute pancreatitis in mice and documented the efficient recovery of exocrine cells ( Suppl. Fig. 2A ). In all, we identified both canonical acini and metaplastic cells, which are exclusively present in cerulein-treated – but not saline-treated – mice, based on the expression of cell lineage markers. Leveraging cell lineage clustering, we could examine genes that were differentially-expressed specifically in acinar cells (and metaplastic cells) early after cerulein administration ( Suppl. Table 1 ). Gene ontology (GO) highlighted significant reprogramming of cellular metabolism, including changes in oxidative phosphorylation, mTORC1 signaling and cholesterol homeostasis ( Suppl. Fig. 2A ). These data align with our collective findings that organelle and cholesterol dyshomeostasis contributes to the pathogenesis of pancreatitis [ 30 ]. To specifically monitor the deregulation of cholesterol trafficking, primary acinar cells were isolated from either saline–or CER-injected mice and stained with filipin, a fluorescent macrolide that binds free cholesterol. We found that cerulein administration promotes re-distribution of cholesterol away from the plasma membrane and accumulation at lysosomes ( Figure 2A ). We next treated 266-6 acinoma cells with inflammatory cytokines and found that IL-6 induces compartmentalization of cholesterol within lysosomes ( Figure 2B ). TNFα elicits a milder effect in 266-6 cells; yet, augmented cholesterol content in Lamp1-positive organelles could be clearly observed ( Suppl. Fig. 2B ). Figure 2 Pancreatitis induces sequestration of cholesterol within lysosomes. A-B. Filipin staining to visualize cholesterol (gray). Lysosomes were immunostained with an anti-Lamp1 antibody (green). In A, WT mice were injected hourly (7x) with either saline or cerulein and acini purified after the last injection. Explants were spun down and immediately stained. Blown-outs magnify area highlighted by dashed, pink squares - in split channels. Scale bars: 100 µm. In B., 266-6 cells were starved and treated with IL-6 (0.08 ng/mL) for 24 hours. Scale bars: 50 µm. Quantification on the right. Asterisks denote statistical significance; ∗, p‹0.05, Student’s t-test. C. Western blotting of whole-pancreas lysates from indicated genotypes/treatments. Quantification on the right (densitometry of western blot bands). D. Western blotting of whole-cell lysates from primary acini treated as indicated for 30 minutes. EEY: eeyarerastin I (ERAD inhibitor); CCK: cholescystokynin. Figure 2 Pancreatitis induces sequestration of cholesterol within lysosomes. A-B. Filipin staining to visualize cholesterol (gray). Lysosomes were immunostained with an anti-Lamp1 antibody (green). In A, WT mice were injected hourly (7x) with either saline or cerulein and acini purified after the last injection. Explants were spun down and immediately stained. Blown-outs magnify area highlighted by dashed, pink squares - in split channels. Scale bars: 100 µm. In B., 266-6 cells were starved and treated with IL-6 (0.08 ng/mL) for 24 hours. Scale bars: 50 µm. Quantification on the right. Asterisks denote statistical significance; ∗, p‹0.05, Student’s t-test. C. Western blotting of whole-pancreas lysates from indicated genotypes/treatments. Quantification on the right (densitometry of western blot bands). D. Western blotting of whole-cell lysates from primary acini treated as indicated for 30 minutes. EEY: eeyarerastin I (ERAD inhibitor); CCK: cholescystokynin. While this can be partly ascribed to elevated number of lysosomal vesicles ( Suppl. Fig. 1A ; also reported in: [ 25 , 26 ]), we found that the lysosomal Niemann-Pick C-1 Protein (NPC1), a key cholesterol extruder, was markedly reduced upon induction of experimental pancreatitis ( Figure 2C ; in line with our previous findings: [ 5 ]). These data suggest the possibility that pancreatitis elicits the disproportionate accumulation of cholesterol at lysosomes in epithelial cells. We set out to characterize the mechanism of NPC1 downregulation. ACLY ablation rescues NPC1 protein levels in CER-treated mice ( Figure 2C ). In line, silencing of Acly upregulates NPC1 and decreases cholesterol sequestration at lysosomes in IL-6-treated 266-6 cells ( Suppl. Fig. 2C–D ). Together these suggest that cytoplasmic acetyl-CoA generation is critical for NPC1 expression and/or stability. Transcription levels of Npc1 in acinar cells were not reduced be CER administration ( Suppl. Fig. 2E ) leading to the hypothesis that NPC1 protein levels may be regulated post-translationally. In fact, endonuclease H digestion indicated that experimental pancreatitis disrupts NPC1 processing ( Suppl. Fig. 2F ) in line with the notion that inflammation causes ER stress and organelle dyshomeostasis ( Figure 1C , also in [ 16 , 30 ]). We thus tested the contribution of the endoplasmic reticulum-associated protein degradation (ERAD) pathway: addition of ERAD-specific inhibitor eeyarestin preserves NPC1 protein levels in primary acinar cells stimulated ex vivo with cholecystokinin (CCK) ( Figure 2D ). Collectively, these data demonstrate that pancreatitis deranges cholesterol trafficking, in part through misfolding and degradation of NPC1 leading to cholesterol accumulation within lysosomes/late endosomes. ACLY promotes NPC1 stability thus improving cholesterol homeostasis. To interrogate the patho-physiological role of lysosomal cholesterol accumulation, we pre-treated mice with either vehicle or the NPC inhibitor U18666A and next challenged those with CER. We observed that NPC1 targeting dramatically worsened tissue architecture at 2 days post CER administration. Specifically, a clear increase in metaplastic structures could be detected in U18666A-treated animals ( Figure 3A ; Suppl. Fig. 3A ). Treating primary acinar explants with U18666A also promoted acinar-to-ductal metaplasia ex vivo ( Figure 3B ), which was – in contrast – suppressed by blockade of cholesterol biosynthesis ( Suppl. Fig. 3B ). This aligns with our previous findings that cholesterol induces ADM through signaling mechanisms [ 31 ]. Figure 3 Lysosomal accumulation of cholesterol promotes ADM. A. Mice were treated with saline or U18666A (daily for 2 weeks; 10 mg/kg) before induction of pancreatitis (n=3, each treatment). Mice were sacrificed 2 days after CER administration. On the left, representative images of tissue histology (whole sections, stained with hematoxylin and eosin staining; magnifications of areas of interest). On the right, quantification of area of inflammation. Scale bar: 1 mm. B. Blinded morphological examination of acinar cell explants from WT mice treated with saline or U18666A (5 µg/mL). Duct-like structures were counted and plotted as percentage of total explants. ∗∗, p‹0.01, Student’s t-test. C. Changes in the expression of genes associated with mTORC1 signaling (left panel) or cholesterol homeostasis (right panel) upon induction of experimental pancreatitis, over 14-day span (from GSE235874 ). D-F. Western blotting to ascertain activation of canonical mTOR targets. In D, western blotting for ribosomal protein S6 phosphorylation. 266-6 murine acinoma cells were cultured in normal conditions or starved of cholesterol (MCD) and then stimulated with CCK (100 nM). In E, western blotting of pancreas lysates of mice treated as in A. In F, 266-6 cells were transfected with two different Npc1-targeting siRNAs or non-targeting control and harvested after 48 hours. G. Blinded morphological examination of acinar cell explants from WT mice (n=4) supplemented with cholesterol and treated with DMSO or Torin-1 for 48 hours. Duct-like structures were quantified as percentage of total explants, per optical fields. Dots show average value for each mouse explant. Asterisks denote statistical significance; ∗, p‹0.05, Student’s t-test. H. Working model describing the link between pancreatitis and cholesterol-dependent mTORC1 activation. Inflammation-exposed acinar cells fail to properly fold NPC1, which is degraded causing sequestration of cholesterol within lysosomes and activation of mTORC1 via the Ragulator complex. Figure 3 Lysosomal accumulation of cholesterol promotes ADM. A. Mice were treated with saline or U18666A (daily for 2 weeks; 10 mg/kg) before induction of pancreatitis (n=3, each treatment). Mice were sacrificed 2 days after CER administration. On the left, representative images of tissue histology (whole sections, stained with hematoxylin and eosin staining; magnifications of areas of interest). On the right, quantification of area of inflammation. Scale bar: 1 mm. B. Blinded morphological examination of acinar cell explants from WT mice treated with saline or U18666A (5 µg/mL). Duct-like structures were counted and plotted as percentage of total explants. ∗∗, p‹0.01, Student’s t-test. C. Changes in the expression of genes associated with mTORC1 signaling (left panel) or cholesterol homeostasis (right panel) upon induction of experimental pancreatitis, over 14-day span (from GSE235874 ). D-F. Western blotting to ascertain activation of canonical mTOR targets. In D, western blotting for ribosomal protein S6 phosphorylation. 266-6 murine acinoma cells were cultured in normal conditions or starved of cholesterol (MCD) and then stimulated with CCK (100 nM). In E, western blotting of pancreas lysates of mice treated as in A. In F, 266-6 cells were transfected with two different Npc1-targeting siRNAs or non-targeting control and harvested after 48 hours. G. Blinded morphological examination of acinar cell explants from WT mice (n=4) supplemented with cholesterol and treated with DMSO or Torin-1 for 48 hours. Duct-like structures were quantified as percentage of total explants, per optical fields. Dots show average value for each mouse explant. Asterisks denote statistical significance; ∗, p‹0.05, Student’s t-test. H. Working model describing the link between pancreatitis and cholesterol-dependent mTORC1 activation. Inflammation-exposed acinar cells fail to properly fold NPC1, which is degraded causing sequestration of cholesterol within lysosomes and activation of mTORC1 via the Ragulator complex. Time-resolved single-cell analyses show that cholesterol homeostasis and mTORC1 signaling are simultaneously deregulated and return to baseline levels within few days ( Figure 3C ; Suppl. Fig. 3C ), suggesting the intriguing possibility that the two events may be linked. The mechanistic Target of Rapamycin Complex 1 (mTORC1) senses nutrient availability at lysosomes and coordinate anabolic programs with cell growth. Cellular reprogramming in response to perturbations is also influenced by mTORC1 activity, which integrates multiple nutritional cues including cholesterol [ 14 , 15 ]. Indeed, cholesterol supplementation promotes activation of canonical mTORC1 targets in 266-6 cells ( Suppl. Fig. 3D ). We asked whether pancreatitis-associated cholesterol dyshomeostasis triggers mTORC1 signaling. We found that CER/CCK stimulation enhanced the phosphorylation of canonical mTOR target Ribosomal Protein S6, which was blunted in vitro by cholesterol deprivation using β-cyclodextrins ( Figure 3D ; Suppl. Fig. 3E–F ). Importantly, NPC1 genetic/pharmacological targeting elicits mTORC1 activation both in vivo and in vitro ( Figure 3E-F ). Together these observations indicate that pancreatitis determines rewiring of cholesterol trafficking in epithelial cells, leading to potentiated mTORC1 signaling and ADM. Elevated signaling through mTORC1 can be observed in proliferating acinar cells in both human and mouse pancreata ( Suppl. Fig. 3G ) [ 32 , 33 ]. In fact, proliferating acini are enriched in ADM markers [ 33 ], and mTORC1 inhibition using Torin1 suppresses cholesterol-induced ADM in isolated acini ( Figure 3G ).

All animal protocols were reviewed and approved by by the animal research committee of VA Greater Los Angeles Healthcare System in accordance with the US NIH Guide for the Care and Use of Laboratory Animals (National Academies Press, 2011). Acly f/f mice were previously described [ 13 ]. To generate Acly ΔACINI mice, Acly f/f mice were bred with Mist1-Cre ER transgenic mice (B6.129- Bhlha15 tm3(cre/ERT2)Skz /J; Jax strain: #029228) and injected subcutaneously with tamoxifen for 2 consecutive days, as previously described [ 20 ]. Genotyping was performed by PCR amplification of ear snips digested with proteinase K, using primers as in [ 13 , 20 ]. Effective recombination was tested by western blotting for ACLY on whole pancreatic tissue lysates; pancreatic expression of ACLY was the sole exclusion parameter used in our studies. Unless otherwise stated, all experimental mice were a mix of male and female. The numbers of animals used per experiment are stated in the figure legends. Trypsin activity was measured in tissue homogenates using a specific fluorogenic substrate, Boc-Gln-Ala-Arg- AMC, as described in [ 5 ]. Inflammatory cell infiltration was quantified on H&E-stained pancreatic tissue sections. At least 1000 acinar cells/mouse were analyzed on several high-power fields. Cerulein-induced (CER-AP) and l -arginine induced (Arg-AP) acute pancreatitis models were performed in 12-week-old C57BL/6 mice that received 7 hourly i.p. injections of 50 μg/kg cerulein or 3 hourly i.p. injections of 3.3 g/kg Arg; controls received physiologic saline injections. Mice were euthanized for analyses 7 h after the first cerulein or 24 h after the first Arg injection. NPC1 inhibitor, U18666A, was administered through daily i.p. injections/10 mg/kg) for 15 consecutive days (12 prior to induction of pancreatitis). Acinar cells were isolated as previously described (22). Briefly, pancreata from 6- to 8-week-old mice were collected upon sacrifice, washed twice in cold Hank's Balanced Salt Solution (HBSS, Biowest L0607) and sub-sequently minced. Tissue was then digested with 1 mg of collagenase P (Roche; #11215809103) in 5 mL HBSS at 37 °C for 30 min, occasionally inverting the tubes and interrupting the digestion every 10 min to mechanically disrupt the clogs by pipetting with progressively smaller pipette tubes. The reaction was stopped and tissue homogenate was washed twice with HBSS containing 5% calf serum (CS) and then filtered through a 500-μm mesh and a 100-μm cell strainer. The flow through was carefully laid onto an HBSS +30% CS solution and centrifuged. The pellet was resuspended in DMEM high glucose (Biowest, L0102) supplemented with 10% of FBS (Gibco, A5256701) (supplemented with indicated inhibitors) for suspension culture. Cells were plated in low-adhesion petri dishes, incubated at 37 °C for 24 h and then harvested, unless otherwise reported. For matrix-embedded culture, cells were resuspended in a 1:3 media:Matrigel (growth-factor reduced, Corning #356231) or a 1:2 media:collagen solution (purified Rat Collagen I, IBIDI #50201). Suspension (250 μL) was seeded onto a 48-well plate and incubated at 37 °C for solidification for at least 1 h. Upon Matrigel solidification, 500 μL of DMEM high glucose, containing 10% FBS, 0.1 mg/mL trypsin inhibitor (Sigma-Aldrich, T2011), and indicated inhibitors/supplements, was added to each well. Cells were monitored daily, and images were acquired at day 2 using a DMI6000B inverted light and fluorescent microscope. Upon collagen solidification, 500 μL of DMEM high glucose supplemented with 10% FBS and 0.1 mg/mL trypsin inhibitor were added to each well. The following day, the medium was replaced with DMEM high glucose supplemented with 10% FBS and 0.1 mg/mL trypsin inhibitor and recombinant EGF (100 ng/mL). Cells were monitored daily and images acquired at day 5 (unless otherwise reported in the figure legend) using a DMI6000B inverted light and fluorescence microscope. Cholesterol and triglycerides were measured as in [ 5 ] with Cholesterol and Triglyceride Assay kits, from Abcam (catalog# ab65390 and ab65336) following manufacturer's instructions. Fluorescence intensity (for cholesterol) and absorbance (for TAGs) were measured in a Tecan microplate reader. Each measurement was in duplicate, and the values were normalized per mg of total protein in the sample. Inhibitors and recombinant proteins used were as follows: Eeyarestatin (Cayman Chemicals; #10012609 – ERAD inhibitor), U18666A (Sigma-Aldrich; #U3633 – NPC1 inhibitor); Torin-1 (Sigma-Aldrich; #475991 – mTORC1 inhibitor), Atorvastatin (Cayman Chemicals; #10493 – HMGCR inhibitor), Zaragozic acid (Sigma-Aldrich; #Z2626 – SQS inhibitor), recombinant murine EGF (PeproTech; #215-09; 100 ng/mL); recombinant TGFa (R&D Systems; 100 ng/mL); recombinant murine IL6 (STEMCELL technologies #78052); bovine-extracted insulin (Sigma-Aldrich; #I5500; 50 ng/mL); cholesterol (Sigma-Aldrich; #S5442). Antibodies used for western blots include 4E-BP1 (53H11) (Cell Signaling, #9644), phospho-4E-BP1 (Thr37/46) (236B4) (Cell Signaling, #2855), ATP citrate lyase (ACLY) [EP704Y] (Abcam, #ab40793), amylase (Sigma-Aldrich, #A8273), BiP (Cell Signaling, #3177), CHOP (Cell Signaling, #5554), ERK (Cell Signaling, #9102), phospho-ERK (Cell Signaling, #5726), H3K9Ac (Cell Signaling, #9649), H3K27ac (Cell Signaling, #8173), H4K8ac (Cell Signaling, #2594), H4K12ac (Cell Signaling, #13944), H4K16ac (Cell Signaling, #13534), HO-1 (E6Z5G) (Cell Signaling, #82206), HSP90 (Santa Cruz Biotechnology, #sc-13119), LC3 (Cell Signaling, #2775), NPC1 (Abcam, 134113), NPC2 (ProteinTech, #19888-1-AP), PLIN2 (ProGen, #GP40), S6 Ribosomal Protein (54D2) (Cell Signaling, #2317), phospho-S6 Ribosomal Protein (Ser235/236) (Cell Signaling, #2211), p70-S6 kinase (Cell Signaling, #9202), phospho-p70 S6 kinase (Cell Signaling, #9234), SQSTM1/p62 (Cell Signaling, #5114), trypsinogen (Santa Cruz Biotechnology, #sc-137077), ubiquitin (Santa Cruz Biotechnology, #sc-8017). Antibodies used for IHC/immunofluorescence were as follows: CD68 (Abcam, #ab213363, 1:100), LAMP1 (AbCam, #ab208943, 1:2000), phospho-p70 S6 kinase (Cell Signaling, #9234, 1:100). For histologic evaluation, tissue samples were harvested as described [ 13 ]. Importantly, pancreata were laid on a planar surface and fixed with formalin overnight. Sectioning of paraffin-embedded tissues exposed the transverse axis and revealed whole organ morphology (4-μm sections; VIMM Histopathology Core). IHC was performed on paraffin-embedded sections. Tissue sections were dewaxed and rehydrated. Antigen retrieval was performed by boiling samples in citrate buffer for 20 min and endogenous peroxidase was blunted by incubating samples with 3% H 2 O 2 for 10 min. Primary antibody was incubated overnight at 4 °C. Proteins in pancreatic tissue lysate were denatured by incubation at 80 °C for 5 min in the presence of 0.5% sodium dodecyl sulfate, and then treated with endonuclease H from Streptomyces plicatus (Glyko, Pro-Zyme, Hayward, CA) at 37 °C for 3 h according to the manufacturers’ protocols, and the samples were processed for immunoblotting as in [ 21 ]. Histones were purified using acid extraction, as previously described [ 13 ]. Briefly, acinar cells were resuspended in high-volume cold NIB buffer for 15 min. Nuclei were pelleted at 600 relative centrifugal force (rcf) for 5 min at 4 °C and washed twice using NIB without NP-40. The nuclear pellets were immediately resuspended in 0.4N H 2 SO 4 and rotated at 4 °C. After centrifugation, histones were precipitated from the supernatant by the addition of 20% tricholoracetic acid (TCA) for at least 1 h, followed by centrifugation. The pellet was washed once with acetone containing 0.1% HCl and finally with 100% acetone. Histone proteins were dried at room temperature and resuspended in water. Cells were seeded on gelatin-coated glass coverslips in 12-well plates at 30 000/50 000 cells per well and allowed to attach overnight. Cells were treated with specify compounds for 24h in DMEM high glucose (Biowest, L0102) supplemented with 1% of FBS (Gibco, A5256701) and then stained. Cells were first fixed with 4% paraformaldehyde (PFA) in PBS for 10 min at room temperature and then rinsed three times with PBS. Non-specific binding was blocked by incubating with 3% bovine serum albumin (BSA) in PBS (w/v) for 1 h at room temperature. Samples were then incubated overnight at 4 °C with the primary antibody diluted in blocking solution. Coverslips were rinsed three times with PBS and then labeled with fluorescently conjugated secondary antibodies (anti-Rabbit 488 Jackson ImmunoResearch 111-545-144, 1:500) and Filipin complex (0.5 mg/ml, Sigma–Aldrich F9765) in 3% BSA in PBS solution (w/v)) for 1 h 30 min at 37 °C, protected from light. Coverslips were rinsed three times with PBS, and two time with milliQ water and then mounted on glass slides using ProLong Glass mountant (Invitrogen, P36982 ). Cells were immediately imaged by ZEISS LSM900 inverted confocal microscope, at the magnification 63×. For staining of isolated acinar cells, immediately after extraction, cells were collected in serum-coated 1.5 mL tubes by centrifugation at 300× g for 5 min and washed twice with PBS. The staining protocol was performed as previously described for adherent cells, with the only modification being the final step where stained cells in suspension were transferred to 12-well plates containing glass coverslips and centrifuged at 4000g for 5 min to allow adhesion prior to mounting. For quantification of co-localization, 10–15 non-overlapping images were acquired from each coverslip. Raw, unprocessed images were imported into FIJI v.2.9.0/1.53t and converted to 8-bit images. Co-localization was quantified using the EzColocalization [ 42 ] plugin in ImageJ, and the Pearson's correlation coefficient (PCC) was used as the co-localization index. Publicly available single-cell RNA sequencing datasets of post-acute pancreatitis (AP) samples were mined in literature, and the GEO ( https://www.ncbi.nlm.nih.gov/geo/ ) accessions GSE235874 , GSE198183 and GSE207938 were selected given the presence of early post-AP timepoints in the available samples. For GSE235874 , pre-processed gene counts matrixes were downloaded from the accession page and used as input for the creation of Seurat objects in RStudio; for GSE198183 , cellranger output files were downloaded from the accession page and used as input for the creation of Seurat objects in RStudio; for GSE207938 , the h5ad object containing the analyzed dataset was downloaded from the accession page and converted in a Seurat object for subsequent analysis in RStudio. For each dataset, analysis and cluster annotation of single-cell derived data was conducted as described in the original articles using the CCA integration described in the Seurat package vignette; this includes the identification of acinar cells- and metaplastic cells-clusters. Dimensionality reduction plots were obtained using the DimPlot function from the Seurat package, while differentially expressed genes were calculated applying the FindMarkers function to joint acinar and ADM clusters over different timepoint. Subsequently, gene set enrichment analysis was performed using the Reactome collection in the EnrichR suite ( https://maayanlab.cloud/Enrichr/ ). To assess average expression of gene signatures contained in the MSig database ( https://www.gsea-msigdb.org/gsea/msigdb/index.jsp ), the AddModuleScore function from the Seurat package was employed to assign a z-score to each cell, and then violin plot were obtained using the VlnPlot function from the same package. For immunoblotting, cells were lysed in lysis buffer (50 mM Tris–HCl pH7.4, 300 mM NaCl, 1 mM EDTA, 1 % NP-40), supplemented with phosphatase inhibitor cocktail (Sigma–Aldrich, P0044) and protease inhibitor cocktail (Sigma–Aldrich, P8340). Pancreatic tissues were collected, immediately snap-frozen in liquid nitrogen, and stored at −80 °C until use. For protein extraction, frozen tissues were pulverized under liquid nitrogen using a pre-chilled mortar and pestle to obtain a fine powder. The powdered tissue was then resuspended in lysis buffer containing 50 mM tris, 150 mM NaCl, 10 mM MgCl2, 0.5 mM dithiothreitol, 1 mM EDTA, and 10% (w/v) glycerol supplemented with 1% (w/v) Triton, 1% (w/v) SDS, supplemented with phosphatase inhibitor cocktail and protease inhibitor cocktail. Protein lysates were quantified by Bradford assay, resuspend in 5x sample buffer (GenScript, MB01015) and resolved by 4–12% SDS–PAGE gels (Millipore, MP41G) and transferred to nitrocellulose membranes (Biorad, 1620112). Membranes were blocked in TBS-T 5 % milk for 30 minutes at room temperature. Incubation with the appropriate dilution of primary antibody was carried out at 4 °C overnight, followed by fluorophore-conjugated secondary antibodies (AlexaFluor 790 donkey anti-rabbit IgG, Jackson Immunolaboratories, 711-655-152 or AlexaFluor 488 Goat Anti-Mouse IgG, 115-545-003) or horseradish peroxidase conjugated secondary antibodies (Goat anti-rabbit HRP, PI-1001-1) for 1 h at room temperature. Signals were visualized using Invitrogen iBright Imaging Systems. Data are presented as the means of experimental replicates with their respective standard deviations (SD), unless otherwise indicated. Student's two-tailed t tests (two-sample equal variance, two-tailed distribution) were used for analyses, unless otherwise indicated. Significance was defined as follows: ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

Here we used ACLY-deficient mice as a tool to uncover a metabolic-signaling conduit that contributes to the pathogenesis of pancreatitis. In those animals, targeting de novo cholesterol synthesis ameliorates several hallmarks of pancreatitis (in line with our prior observation using statins [ 5 ]; also in: [ 23 ]), including endoplasmic reticulum (ER) stress. We demonstrated that pancreatitis-associated ER stress leads to ERAD-dependent degradation of the lysosomal-resident cholesterol transporter, NPC1. Decreased levels of NPC1 cause alterations in cholesterol trafficking and its accumulation within lysosomes [ 34 ], where is sensed by the Ragulator complex and promotes mTORC1 activation [ 14 , 15 ]. In acinar cells this induces acinar-to-ductal metaplasia, a common response of exocrine cells to inflammation [ 3 ]. At the same time, cholesterol is kept astray from the ER leading to enhanced compensatory biosynthesis. The beneficial effects of ACLY deletion are thought to be a consequence of reduced ER stress, lysosomal cholesterol accumulation and dampened mTORC1 signaling ( Figure 3H ). This study elucidates several novel aspects with significant ramifications. First, we defined a physiological role for cholesterol-mTORC1 signaling axis. The ability of mTOR complex 1 to sense lysosomal cholesterol has been shown in models where NPC1 function was ablated, a condition associated to the fatal neurological disorder Niemann-Pick disease type C. Recently, progressive accumulation of cholesterol within lysosomes has been shown in cellular models of senescence and linked to aberrant mTORC1 activation [ 35 ]. We show that this metabolic/signaling conduit is responsive to physiological stimuli (e.g.: CCK/cerulein; inflammatory cytokines). Excessive mTORC1 signaling unlocks acinar cell plasticity, which contributes to the pathogenesis of pancreatitis and play multifaced roles in tissue regeneration. This has obvious translational ramifications, as circulating cholesterol levels are highly variable but can be modulated; understanding how that impacts the onset of pancreatitis may lead to improved dietary management. In terms of basic biology, our findings delineate the signaling cascade regulated by cholesterol trafficking and add to the definition of cholesterol as signaling metabolite [ 36 ]. Second, we showed that elevation of intrapancreatic cholesterol is mediated by de novo synthesis in acinar cells. While extra-hepatic tissues can synthetize cholesterol, including beta cells and the small intestine [ 37 , 38 ], this has not been documented in non-neoplastic exocrine pancreas. Third, we expand the characterization of metabolic reprogramming in cell metaplasia [ 39 ]. It will be interesting to assess whether the intertwine between metabolite compartmentalization, nutrient sensing (e.g.: mTORC1 activation) and cholesterol biosynthesis contributes to inflammation-promoted carcinogenesis. Of note, mutant KRAS rewires cell metabolism to unleash acinar cell plasticity and initiate carcinogenesis [ 12 ]. Fourth, our observations indicate that cellular compartments of inflammatory acinar cells communicate to affect cell plasticity. In fact, ER stress leads to misfolding of NPC1 to impact the endo-lysosomal machinery. This identifies a mechanistic link between organelle dyshomeostasis and signal transduction that contributes to disease pathogenesis. Our study has few limitations. We do not show direct evidence in human samples/cells; this is mainly due to the lack of acute pancreatitis biopsies and the challenge to culture human acinar cells (or the lack of immortalized human cell lines). In addition, the inter-organelle communication we describe does not fully account for changes in calcium dynamics, which may impact NPC1 folding. Our work does not dissect the exact sensing route for lysosomal cholesterol, which can be sensed by either SLC38A9 (via GATOR) or by LYCHOS (via the Ragulator complex), although single cell studies from normal pancreata indicate that only the former is expressed at meaningful levels in mouse and human acinar cells [ 32 , 33 , 40 ]. Finally, tissue dynamics elicited by cholesterol-mTORC1 signaling in the context of pancreatitis remains to be elucidated. We demonstrated that elevated cholesterol availability promotes acinar cells metaplasia, while targeting de novo synthesis ameliorates markers of inflammation. We might conclude that metaplastic cells are pro-inflammatory, in line with the notion that epithelial-stromal crosstalk is critical for the onset of pancreatitis [ 41 ]. Notwithstanding, the study identifies novel ad targetable mechanisms involved in the onset of pancreatitis. It also links statins- and diet-modulated cholesterol availability to physiological mTORC1 activation and unveils the significance of metabolic compartmentalization in exocrine cells.

Pancreatitis is an inflammatory disorder of the pancreas that constitutes a common cause of hospitalization and commands careful clinical management. Patients develop significant complications in 20–25% of cases, including permanent tissue damage [ 1 , 2 ]. Yet, our understanding of disease pathogenesis is incomplete and there is a lack of effective treatments. Several experimental models of pancreatitis can be applied to rodents. Administration of supra-physiological doses of cholecystokinin-analog cerulein is the most used [ 3 , 4 ]. These models have been instrumental to assess the molecular events that unfold in epithelial cells in response to inflammation [ 5 , 6 ]. However, how metabolic reprogramming supports the development of pancreatitis and the ensuing tissue regeneration is poorly investigated. Pancreatitis is characterized by organelle dyshomeostasis in acinar cells that includes altered mitochondrial function and dynamics, endoplasmic reticulum (ER) stress and impaired lysosomal/autophagic machinery [ [7] , [8] , [9] ]. In acinar cells, these events lead to either cellular collapse or adaptive responses, like cell cycle re-entry or acinar-to-ductal metaplasia (ADM) [ 3 , 10 , 11 ]. The physiological role of the latter is unclear, but is associated with tissue regeneration post-injury and can be hijacked by oncogenic KRAS to initiate ductal adenocarcinoma from acini [ 12 ]. We previously reported that genetic ablation of ATP-Citrate Lyase (ACLY) in pancreatic epithelial cells suppresses oncogene-induced ADM, reducing nucleo-cytoplasmic availability of acetyl-CoA to restrain global levels of histone acetylation and cholesterol biosynthesis [ 13 ]. We did not examine how ACLY deletion affects the development of pancreatitis. We reasoned it could disrupt cholesterol sensing to impact pathogenesis. Cholesterol levels fluctuate in the cell and are monitored carefully through multiple mechanisms. Being at the crossroad of cholesterol trafficking makes lysosomes the ideal recess for its sensing. Indeed, two lysosomal resident proteins – SLC38A9 and LYCHOS – can bind cholesterol and signal augmented levels to the mechanistic Target of Rapamycin Complex −1 (mTORC1) [ 14 , 15 ]. Activation of both mTORC1 and mTORC2 has been documented in experimental models of acute pancreatitis [ [16] , [17] , [18] , [19] ], where they promote ADM [ 19 ]. Here we show that inflammatory acinar cells fail to correctly fold the lysosomal cholesterol transporter Niemann-Pick disease type C protein 1 (NPC1), leading to sequestration of cholesterol within lysosomes and hyperactivation of mTORC1. We also show that suppressing NPC1 function enhances ADM to worsen the outcome of experimental pancreatitis. Our findings demonstrate the physiological role of lysosomal cholesterol sensing.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.