Portal de Periódicos da CAPES

Autophagic Cell Death of Pancreatic Acinar Cells in Serine Protease Inhibitor Kazal Type 3—Deficient Mice

2005; Elsevier BV; Volume: 129; Issue: 2 Linguagem: Inglês

10.1016/j.gastro.2005.05.057

1528-0012

Masaki Ohmuraya, Masahiko Hirota, Masatake Araki, Noboru Mizushima, Makoto Matsui, Takuya Mizumoto, Kyoko Haruna, Shoen Kume, Motohiro Takeya, Minetaro Ogawa,

Autophagy in Disease and Therapy

Background & Aims: Serine protease inhibitor Kazal type 1 (SPINK1), which is structurally similar to epidermal growth factor, is thought to inhibit trypsin activity and to prevent pancreatitis. Point mutations in the SPINK1 gene seem to predispose humans to pancreatitis; however, the clinical significance of SPINK1 mutations remains controversial. This study aimed to elucidate the role of SPINK1. Methods: We generated Spink3-deficient (Spink3−/−) mice by gene targeting in mouse embryonic stem cells. Embryonic and neonatal pancreases were analyzed morphologically and molecularly. Specific probes were used to show the typical autophagy that occurs during acinar cell death. Results: In Spink3−/− mice, the pancreas developed normally up to 15.5 days after coitus. However, autophagic degeneration of acinar cells, but not ductal or islet cells, started from day 16.5 after coitus. Rapid onset of cell death occurred in the pancreas and duodenum within a few days after birth and resulted in death by 14.5 days after birth. There was limited inflammatory cell infiltration and no sign of apoptosis. At 7.5 days after birth, residual ductlike cells in the tubular complexes strongly expressed pancreatic duodenal homeodomain-containing protein 1, a marker of pancreatic stem cells, without any sign of acinar cell regeneration. Conclusions: The progressive disappearance of acinar cells in Spink3−/− mice was due to autophagic cell death and impaired regeneration. Thus, Spink3 has essential roles in the maintenance of integrity and regeneration of acinar cells. Background & Aims: Serine protease inhibitor Kazal type 1 (SPINK1), which is structurally similar to epidermal growth factor, is thought to inhibit trypsin activity and to prevent pancreatitis. Point mutations in the SPINK1 gene seem to predispose humans to pancreatitis; however, the clinical significance of SPINK1 mutations remains controversial. This study aimed to elucidate the role of SPINK1. Methods: We generated Spink3-deficient (Spink3−/−) mice by gene targeting in mouse embryonic stem cells. Embryonic and neonatal pancreases were analyzed morphologically and molecularly. Specific probes were used to show the typical autophagy that occurs during acinar cell death. Results: In Spink3−/− mice, the pancreas developed normally up to 15.5 days after coitus. However, autophagic degeneration of acinar cells, but not ductal or islet cells, started from day 16.5 after coitus. Rapid onset of cell death occurred in the pancreas and duodenum within a few days after birth and resulted in death by 14.5 days after birth. There was limited inflammatory cell infiltration and no sign of apoptosis. At 7.5 days after birth, residual ductlike cells in the tubular complexes strongly expressed pancreatic duodenal homeodomain-containing protein 1, a marker of pancreatic stem cells, without any sign of acinar cell regeneration. Conclusions: The progressive disappearance of acinar cells in Spink3−/− mice was due to autophagic cell death and impaired regeneration. Thus, Spink3 has essential roles in the maintenance of integrity and regeneration of acinar cells. Inappropriate activation of trypsinogen in the pancreas leads to pancreatitis. Once activated, trypsin is capable of activating many other digestive proenzymes in the pancreas and enhances autodigestion of the pancreas. Trypsin activity is thought to be predominantly controlled by the serine protease inhibitor Kazal type 1 (SPINK1), which is also known as pancreatic secretory trypsin inhibitor. In the mouse, this homologous gene is designated as Spink3 (serine protease inhibitor Kazal type 3). SPINK1 is synthesized in the acinar cells of the pancreas and binds to trypsin to prevent further activation of pancreatic enzymes when trypsinogen is converted into trypsin. Thus, a lack of SPINK1 may result in the premature conversion of trypsinogen into active trypsin in acinar cells, thus leading to autodigestion of the exocrine pancreas by activated proteases. It is interesting to note that SPINK1 and epidermal growth factor (EGF) have structural similarities, including the number of amino acid residues and the presence of 3 intrachain disulfide bridges.1Yamamoto T. Nakamura Y. Nishide J. Emi M. Ogawa M. Mori T. Matsubara K. Molecular cloning and nucleotide sequence of human pancreatic secretory trypsin inhibitor (PSTI) cDNA.Biochem Biophys Res Commun. 1985; 132: 605-612Crossref PubMed Scopus (76) Google Scholar SPINK1 has been found to induce the proliferation of a variety of cell lines.2Ogawa M. Tsushima T. Ohba Y. Ogawa N. Tanaka S. Ishida M. Mori T. Stimulation of DNA synthesis in human fibroblasts by human pancreatic secretory trypsin inhibitor.Res Commun Chem Pathol Pharmacol. 1985; 50: 155-158PubMed Google Scholar, 3Niinobu T. Ogawa M. Murata A. Nishijima J. Mori T. Identification and characterization of receptors specific for human pancreatic secretory trypsin inhibitor.J Exp Med. 1990; 172: 1133-1142Crossref PubMed Scopus (27) Google Scholar Several mutations of the trypsinogen gene have been identified and are assumed to be pathogenic in patients with hereditary pancreatitis through the enhancement of intrapancreatic trypsin activity.4Whitcomb D.C. Gorry M.C. Preston R.A. Furey W. Sossenheimer M.J. Ulrich C.D. Martin S.P. Gates Jr, L.K. Amann S.T. Toskes P.P. et al.Hereditary pancreatitis is caused by a mutation in the cationic trypsinogen gene.Nat Genet. 1996; 14: 141-145Crossref PubMed Scopus (1297) Google Scholar, 5Gorry M.C. Gabbaizedeh D. Furey W. Gates Jr, L.K. Preston R.A. Aston C.E. Zhang Y. Ulrich C. Ehrlich G.D. Whitcomb D.C. Mutations in the cationic trypsinogen gene are associated with recurrent acute and chronic pancreatitis.Gastroenterology. 1997; 113: 1063-1068Abstract Full Text Full Text PDF PubMed Scopus (396) Google Scholar Although the mutations lead to an 80% likelihood of developing pancreatitis, they are not found in approximately 50% of patients. Mutations in the cystic fibrosis transmembrane conductance regulator are found in patients with chronic and idiopathic pancreatitis.6Sharer N. Schwarz M. Malone G. Howarth A. Painter J. Super M. Braganza J. Mutations of the cystic fibrosis gene in patients with chronic pancreatitis.N Engl J Med. 1998; 339: 645-652Crossref PubMed Scopus (786) Google Scholar, 7Cohn J.A. Friedman K.J. Noone P.G. Knowles M.R. Silverman L.M. Jowell P.S. Relation between mutations of the cystic fibrosis gene and idiopathic pancreatitis.N Engl J Med. 1998; 339: 653-658Crossref PubMed Scopus (792) Google Scholar However, additional gene mutations remain to be identified. Although the association of several mutations in the SPINK1 gene with familial and juvenile pancreatitis has been reported,8Witt H. Luck W. Hennies H.C. Classen M. Kage A. Lass U. Landt O. Becker M. Mutations in the gene encoding the serine protease inhibitor, Kazal type 1 are associated with chronic pancreatitis.Nat Genet. 2000; 25: 213-216Crossref PubMed Scopus (830) Google Scholar, 9Pfutzer R.H. Barmada M.M. Brunskill A.P. Finch R. Hart P.S. Neoptolemos J. Furey W.F. Whitcomb D.C. SPINK1/PSTI polymorphisms act as disease modifiers in familial and idiopathic chronic pancreatitis.Gastroenterology. 2000; 119: 615-623Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar, 10Chen J.M. Mercier B. Audrezet M.P. Raguenes O. Quere I. Ferec C. Mutations of the pancreatic secretory trypsin inhibitor (PSTI) gene in idiopathic chronic pancreatitis.Gastroenterology. 2001; 120: 1061-1064Abstract Full Text Full Text PDF PubMed Scopus (91) Google Scholar, 11Kuwata K. Hirota M. Sugita H. Kai M. Hayashi N. Nakamura M. Matsuura T. Adachi N. Nishimori I. Ogawa M. Genetic mutations in exons 3 and 4 of the pancreatic secretory trypsin inhibitor in patients with pancreatitis.J Gastroenterol. 2001; 36: 612-618Crossref PubMed Scopus (43) Google Scholar the clinical significance of these mutations remains controversial.12Shimosegawa T. Do point mutations in the PSTI (SPINK1) gene truly contribute to the pathogenesis of chronic pancreatitis?.J Gastroenterol. 2001; 36: 645-647Crossref PubMed Scopus (3) Google Scholar, 13Whitcomb D.C. How to think about SPINK and pancreatitis.Am J Gastroenterol. 2002; 97: 1085-1088Crossref PubMed Google Scholar, 14Threadgold J. Greenhalf W. Ellis I. Howes N. Lerch M.M. Simon P. Jansen J. Charnley R. Laugier R. Frulloni L. et al.The N34S mutation of SPINK1 (PSTI) is associated with a familial pattern of idiopathic chronic pancreatitis but does not cause the disease.Gut. 2002; 50: 675-681Crossref PubMed Scopus (174) Google Scholar, 15Chandak G.R. Idris M.M. Reddy D.N. Bhaskar S. Sriram P.V. Singh L. Mutations in the pancreatic secretory trypsin inhibitor gene (PSTI/SPINK1) rather than the cationic trypsinogen gene (PRSS1) are significantly associated with tropical calcific pancreatitis.J Med Genet. 2002; 39: 347-351Crossref PubMed Google Scholar, 16Drenth J.P. te Morsche R. Jansen J.B. Mutations in serine protease inhibitor Kazal type 1 are strongly associated with chronic pancreatitis.Gut. 2002; 50: 687-692Crossref PubMed Scopus (142) Google Scholar, 17Kuwata K. Hirota M. Nishimori I. Otsuki M. Ogawa M. Mutational analysis of the pancreatic secretory trypsin inhibitor gene in familial and juvenile pancreatitis in Japan.J Gastroenterol. 2003; 38: 365-370Crossref PubMed Scopus (25) Google Scholar To elucidate the role of Spink3, we produced Spink3-null mutant mice by gene targeting in mouse embryonic stem (ES) cells. Genomic DNA containing all the exons of the Spink3 gene was isolated from a C57BL/6-EMBL3 library (a gift from Dr Aizawa, Riken, Kobe, Japan) by hybridization with an 800—base pair genomic DNA probe containing intron 1. The targeting construction was produced in pBluescript II (Stratagene, La Jolla, CA) containing the phosphoglycerate kinase 1 promoter and the neomycin resistance gene (neo) cassette flanked by the mutated loxP sites lox71 and lox227218Araki K. Araki M. Yamamura K. Site-directed integration of the cre gene mediated by Cre recombinase using a combination of mutant lox sites.Nucleic Acids Res. 2002; 30: e103Crossref PubMed Scopus (109) Google Scholar and the diphtheria toxin A fragment with a polyadenylation cassette by using a polyoma enhancer/herpes simplex virus thymidine kinase promoter. The neo cassette was inserted in front of the initiation codon (Figure 1A). Electroporation with TT2 ES cells19Yagi T. Tokunaga T. Furuta Y. Nada S. Yoshida M. Tsukada T. Saga Y. Takeda N. Ikawa Y. Aizawa S. A novel ES cell line, TT2, with high germline-differentiating potency.Anal Biochem. 1993; 214: 70-76Crossref PubMed Scopus (415) Google Scholar (a gift from Dr Aizawa) and colony isolation were performed as previously described.20Araki K. Imaizumi T. Sekimoto T. Yoshinobu K. Yoshimuta J. Akizuki M. Miura K. Araki M. Yamamura K. Exchangeable gene trap using the Cre/mutated lox system.Cell Mol Biol (Noisy-le-Grand). 1999; 45: 737-750PubMed Google Scholar Three targeted ES clones were obtained from among 360 G418-resistant clones. ES cells were aggregated with ICR molura as described.20Araki K. Imaizumi T. Sekimoto T. Yoshinobu K. Yoshimuta J. Akizuki M. Miura K. Araki M. Yamamura K. Exchangeable gene trap using the Cre/mutated lox system.Cell Mol Biol (Noisy-le-Grand). 1999; 45: 737-750PubMed Google Scholar Chimeras were mated with C57BL/6J mice, and germline transmission was obtained for all 3 lines. Noon on the day of vaginal plug detection was defined as 0.5 days after coitus. All experiments were performed in accordance with the Declaration of Helsinki and were approved by the Kumamoto University Ethics Committee for Animal Experiments. After overnight fasting, Spink3+/+ and Spink3+/− mice (5–7 weeks old and weighing 18–23 g) were given hourly intraperitoneal injections of saline as control (n = 5) or saline containing a supramaximal stimulating concentration of cerulein (50 μg/kg; n = 5) (Sigma-Aldrich Corp, Tokyo, Japan) for several hours (3–12 hours). One hour after the last injection, mice were killed, and the serum and pancreas were rapidly prepared for study. The serum was used for measurement of amylase activity. The pancreas was used for Western blot and trypsin assay analysis. DNA was digested with PstI and subjected to Southern blot hybridization. For genotyping of embryos and pups, polymerase chain reaction (PCR) analysis was performed with the following primers: neo cassette, 5′-AGAGGCTATTCGGCTATGAC-3′ and 5′-CACCATGATATTCGGCAAGC-3′; Spink3 exon 1, 5′-AGTTCTTCTGGCTTTTGCACCC-3′; and Spink3 intron 1, 5′-CTTTGCCACCACATCCCAAATG-3′. Total RNA was isolated from the trunk or intestine with Sepasol (Nacalai Tesque, Kyoto, Japan). For Northern blot analysis, 10 μg of RNA was applied to each lane and fractionated on a 1.4% agarose gel. Filter-bound RNA was sequentially hybridized with a digoxigenin-labeled RNA probe (Roche Molecular Biochemicals, Mannheim, Germany). Reverse-transcription PCR (RT-PCR) analysis was performed with the following primers: Spink3, 5′-AGTTCTTCTGGCTTTTGCACCC-3′ and 5′-CTCTTTTCCAGTCACCTTAGCT-3′; and trypsinogen, 5′-TGGCTTCCTAGAGGGAGGCAA-3′ and 5′-CACAGCCATAGCCCCAAGAGAC-3′. The pancreas was homogenized in lysate buffer (HEPES 50 mmol/L, pH 7.4, NaCl 150 mmol/L, Triton X-100 0.1%, glycerol 10%, NaF 1 mmol/L, sodium orthovanadate 2 mmol/L, ethylenediaminetetraacetic acid 1 mmol/L, and protease inhibitor cocktail [1:100 dilution; Sigma-Aldrich]). Extracts (20 μg of protein per lane) were applied to 20% (for Spink3 detection) or 16% (for microtubule-associated protein 1 light chain 3 [LC3] detection) polyacrylamide gel electrophoresis and transferred to an Immobilon polyvinylidene difluoride filter (Millipore, Billerica, MA). Primary antibodies to the following antigens (made in rabbit) were used at the indicated dilutions: pancreatic secretory trypsin inhibitor (Transgenic Inc, Kumamoto, Japan), 1:500; and LC3 (provided by Dr Tamotsu Yoshimori), 1:2000. An anti-rabbit immunoglobulin G antibody conjugated with horseradish peroxidase (Amersham Biosciences Corp, Piscataway, NJ) was used for detection. Quantification of the ratio of LC3-II to LC3-I was performed by the Densitograph software library version 4 (Atto, Tokyo, Japan). Unpaired Student t tests were used to calculate P values. For histological analysis, tissue was fixed overnight in 10% formalin, embedded in paraffin, sectioned, and stained with the H&E procedure. Immunohistochemistry was performed by using the following primary antibodies: rabbit anti-insulin antibody (diluted 1:200; Santa Cruz Biotechnology, Inc, Santa Cruz, CA); rabbit anti-glucagon antibody (Dako, Carpinteria, CA); goat anti-amylase antibody (diluted 1:200; Santa Cruz); rabbit anti-pancreatic duodenal homeodomain-containing protein 1 (Pdx1) antibody (diluted 1:500; Chemicon International, Inc, Temecula, CA); and monoclonal anti—proliferating cell nuclear antigen antibody (diluted 1:500; Novocastra Laboratories Ltd, Newcastle upon Tyne, UK). Reactivity of the primary antibodies with mouse antigens was confirmed. Primary antibodies were detected with a commercial biotin-streptavidin system (Vector Laboratories, Burlingame, CA). To evaluate the type of cells that synthesize DNA in the pancreas, mice were given a single intraperitoneal dose (100 mg/kg body weight) of bromodeoxyuridine (Sigma-Aldrich), a thymidine analogue, and killed 24 hours later. Sections of the pancreas were processed for paraffin embedding and immunostaining with an anti-bromodeoxyuridine antibody (diluted 1:20; Dako). For the detection of apoptosis, terminal deoxynucleotidyl transferase—mediated deoxyuridine triphosphate nick-end labeling (TUNEL) assay was performed by using an in situ apoptosis detection kit (Wako, Osaka, Japan). Pancreatic tissues were fixed with 2.5% glutaraldehyde and postfixed with 1% osmium tetroxide. After dehydration in a graded series of ethanol and propylene oxide, the samples were embedded in epoxy resin. Ultrathin sections were stained with uranyl acetate and lead citrate and then imaged with an H-7500 electron microscope (Hitachi, Tokyo, Japan). Pancreases from Spink3−/−-green fluorescent protein (GFP)-LC3 mice at 0.5 days after birth were dissected, fixed, and sectioned. GFP fluorescence was observed by using an IX81 fluorescence microscope (Olympus, Tokyo, Japan) equipped with an ORCA ER charge-coupled device camera (Hamamatsu Photonics, Hamamatsu, Japan). Trypsin activity of the pancreas was measured fluorometrically by using benzoyl-l-arginine p-nitroanilide as substrate according to previously described methods.21Kuwata K. Hirota M. Shimizu H. Nakae M. Nishihara S. Takimoto A. Mitsushima K. Kikuchi N. Endo K. Inoue M. et al.Functional analysis of recombinant pancreatic secretory trypsin inhibitor protein with amino-acid substitution.J Gastroenterol. 2002; 37: 928-934Crossref PubMed Scopus (76) Google Scholar The trypsin activity was corrected to the density of pancreas DNA. Unpaired Student t tests were used to calculate P values. P < .05 was considered to indicate a significant difference. Glucose was determined in the peripheral blood of Spink3+/+, Spink3+/−, or Spink3−/− mice at 0.5, 1.5, and 3.5 days after birth by use of commercial equipment (Glucocard) according to the supplier’s instructions (Arkray, Inc, Kyoto, Japan). The vector used for homologous recombination in ES cells to disrupt the Spink3 locus is shown in Figure 1A. Three targeted ES clones lacking Spink3 were identified by Southern blot analysis with a 5′ probe (Figure 1B) and were used to generate chimeric mice. Spink3+/− mice were healthy, fertile, and indistinguishable from their Spink3+/+ littermates, although the level of Spink3 expression was approximately half that of their Spink3+/+ littermates (Figure 1D and E). Male and female Spink3+/− mice were mated to produce Spink3−/− mice. Genotypes were determined by PCR analysis (Figure 1C). The ratio of living Spink3+/+, Spink3+/−, and Spink3−/− mice at 0.5 days after birth was 33:72:36 (n = 141), respectively. This matches the mendelian rate of 1:2:1. Thus, a Spink3 deficiency does not cause embryonic lethality. In homozygous newborn mice, no Spink3 messenger RNA (mRNA) or protein was detected (Figure 1D and E), thus indicating the production of a null allele of the Spink3 locus. At birth, Spink3−/− mice were indistinguishable macroscopically from their Spink3+/+ and Spink3+/− littermates and were fed milk. However, Spink3−/− mice did not gain weight (Figure 2A and B) and died by 14.5 days after birth (Figure 2C). They were severely dehydrated and had thin, cracking skin with very little fur. Spink3−/− mice at 18.5 days after coitus had a normal liver, gallbladder, spleen, stomach, duodenum, bile duct, and other viscera. Similarly, the pancreas appeared almost normal at 18.5 days after coitus (Figure 2D), thus suggesting normal embryonic development of the pancreas. However, the pancreas progressively disappeared after birth, and the duodenum became extended. The pancreatic bed was quite small and transparent, and the spleen was much smaller than that of wild-type mice at 3.5 days after birth (Figure 2E). To examine when the Spink3 and trypsinogen genes start to express, RT-PCR analyses were performed. Spink3 mRNAs and trypsinogen mRNAs were detected in embryos at 11.5 and at 15.5 days after coitus, respectively. Histological and immunohistochemical studies were performed to analyze the cause of the pancreatic deficiency. At 15.5 days after coitus, pancreases were normal, with finely branched ducts and distinct lumen in both Spink3+/+ and Spink3−/− embryos. At 16.5 days after coitus, acinar cells were clearly recognized in both Spink3+/+ and Spink3−/− embryos (Figure 3A), but Spink3−/− embryos showed mild vacuolization of acinar cells (Figure 3A). This suggests that vacuolization begins at around the time when zymogen granules appear. At 18.5 days after coitus, vacuolization of acinar cells became severe in Spink3−/− embryos (Figure 3B). At 0.5 days after birth, acinar cell degeneration was clearly evident in Spink3−/− mice (Figure 3C). At 1.5 days after birth, pancreatic lobules mainly comprised tubular complexes lined by flattened ductlike cells in a loose connective tissue stroma with few remaining acinar cells (Figure 3D). At 3.5 days after birth, degenerative changes became severe (Figure 3E), and only a few acinar cells were positive for amylase (Figure 3J). Until 3.5 days after birth, islets remained quite normal histologically, and the staining intensities for insulin (Figure 3A—E) and glucagon (data not shown) in Spink3−/− mice were similar to those in Spink3+/+ embryos. At 7.5 days after birth, rapid loss of acinar cells resulted in lobular contraction and led to tightly clustered tubular complexes associated with small numbers of residual acinar cells (Figure 3F). It is interesting to note that inflammatory cell infiltration was scarce in the pancreases of Spink3−/− mice. Vacuolization or degeneration of acinar cells was not observed in either Spink3+/+ or Spink3+/− mice at any stage (Figure 3A—F). These findings suggest that a lack of Spink3 activity induces massive selective acinar cell degeneration. In Spink3−/− mice at 7.5 days after birth, the ductlike cells in tubular complexes strongly expressed Pdx1 antigen, a marker of pancreatic stem cells (Figure 3L). However, they showed no mitotic figures in H&E staining (Figure 3F), no positive immunostaining with proliferating cell nuclear antigen (Figure 3L), and no bromodeoxyuridine incorporation (data not shown). Thus, there is no evidence of regeneration of acini in Spink3−/− mice during the period of our observations. Although spleens were small, histological sections showed a normal structure. The duodenum and small intestine of Spink3−/− mice looked normal up to 1.5 days after birth. However, at 3.5 days after birth, it was expanded, with the thin smooth muscle layer and villi mostly degenerated (Figure 4); the degeneration was probably caused by the Spink3 deficiency. In any case, atrophy of the pancreas by deficient exocrine function and degeneration of the duodenum and small intestine were likely to be the main causes of the severe retardation of growth and death of Spink3−/− mice. This is consistent with the data that mice lacking the gene encoding basic helix-loop-helix protein p48 died soon after birth because of a complete absence of exocrine pancreatic tissue.22Krapp A. Knofler M. Ledermann B. Burki K. Berney C. Zoerkler N. Hagenbuchle O. Wellauer P.K. The bHLH protein PTF1-p48 is essential for the formation of the exocrine and the correct spatial organization of the endocrine pancreas.Genes Dev. 1998; 12: 3752-3763Crossref PubMed Scopus (430) Google Scholar Apoptosis did not seem to be involved in this process, because there was no increase in the number of apoptotic cells detected by TUNEL assay (data not shown). We also performed transmission electron microscopic analysis of the pancreas. At 18.5 days after coitus, the cytoplasm of acinar cells in Spink3+/+ mice was filled with zymogen granules, but many vacuoles were observed in Spink3−/− mice (Figure 5A). At 0.5 days after birth, Spink3−/− acinar cells showed extensive vacuolization, but only a small amount of vacuolization was observed in Spink3+/+ mice (Figure 5B). At a higher magnification, some vacuoles were found to contain cellular organelles, thus indicating that they were autophagic vacuoles (Figure 5C). Nuclear changes and apoptotic body formation were not observed (Figure 5D). Endocrine (arrowhead in Figure 5E) and ductal epithelial (arrow in Figure 5E and F) cells showed normal structures. At 1.5 days after birth, both the number and size of acinar cells were markedly reduced in Spink3−/− mice (Figure 5F), and those remaining contained various types of vacuoles compressing the nuclei and resulting in cell death (Figure 5G). Occasionally, a huge vacuole containing various digested organelles was observed (Figure 5H). Some of the degenerated acinar cells were phagocytosed by macrophages (Figure 5I). These morphological findings showed that the massive amount of acinar cell death in Spink3−/− mice was associated neither with substantial infiltration of inflammatory cells, as found in necrosis, nor with apoptosis, the best characterized form of programmed cell death. The appearance of numerous cytoplasmic vacuoles before nuclear alteration indicates that acinar cell death in Spink3−/− mice is similar to an autophagic cell death, a form of major physiological cell death.23Clarke P.G. Developmental cell death morphological diversity and multiple mechanisms.Anat Embryol (Berl). 1990; 181: 195-213Crossref PubMed Scopus (1521) Google Scholar, 24Lockshin R.A. Zakeri Z. Programmed cell death and apoptosis origins of the theory.Nat Rev Mol Cell Biol. 2001; 2: 545-550Crossref PubMed Scopus (268) Google Scholar, 25Leist M. Jaattela M. Four deaths and a funeral from caspases to alternative mechanisms.Nat Rev Mol Cell Biol. 2001; 2: 589-598Crossref PubMed Scopus (1365) Google Scholar To confirm the presence of autophagosomes, we analyzed the expression of LC3, an autophagosome-associated protein.26Kabeya Y. Mizushima N. Ueno T. Yamamoto A. Kirisako T. Noda T. Kominami E. Ohsumi Y. Yoshimori T. LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing.EMBO J. 2000; 19: 5720-5728Crossref PubMed Scopus (5305) Google Scholar There are 2 forms of LC3 protein: LC3-I and LC3-II. Cytosolic LC3-I protein (18 kilodaltons) is converted into LC3-II (16 kilodaltons) and is associated with the autophagosome membrane. The amount of LC3-II is thus correlated with the extent of autophagosome formation. As shown in Figure 6A, Spink3−/− pancreases possessed more LC3-II than those of Spink3+/+, as in mice with cerulein-induced pancreatitis (Figure 6C). Densitometric analysis showed that the ratio of LC3-II to LC3-I was 0.56 (SD, 0.24) and 2.23 (SD, 0.10) in Spink3+/+ and Spink3−/− mice, respectively. The ratio of LC3-II to LC3-I was 0.70 (SD, 0.46) and 2.06 (SD, 0.57) in control mice and mice with cerulein-induced pancreatitis, respectively. Thus, the amount of LC3-II increased in Spink3−/− mice, thus suggesting that autophagic activity is appreciably promoted in the Spink3−/− pancreas. Furthermore, we used GFP-LC3 mice, in which the expression and localization of LC3 can be monitored by the detection of GFP fluorescence.27Mizushima N. Yamamoto A. Matsui M. Yoshimori T. Ohsumi Y. In vivo analysis of autophagy in response to nutrient starvation using transgenic mice expressing a fluorescent autophagosome marker.Mol Biol Cell. 2004; 15: 1101-1111Crossref PubMed Scopus (1867) Google Scholar In Spink3+/−-GFP-LC3 mice at 0.5 days after birth, only a few dots of fluorescence were detected in the cytoplasm of acinar cells (Figure 6B, left). In contrast, many fluorescent dots were observed in the cytoplasm of acinar cells of Spink3−/−-GFP-LC3 mice at 0.5 days after birth (Figure 6B, right). Some dots can be recognized as ringlike structures, which may represent the attachment of GFP-LC3 protein to the autophagosome membrane. The number of GFP-LC3 dots was fewer than that of vacuoles observed by electron microscopy. We believe that this difference can be explained by the fact that LC3 gradually dissociates from the autophagic vacuoles after fusion with lysosomes. These results suggest extensive autophagosome formation in Spink3−/− acinar cells. Serum amylases were increased in Spink3−/− mice at 0.5 days after coitus as in mice with cerulein-induced pancreatitis (Figure 6D). This suggests that the last moment of cell death is accompanied by disruption of the membrane, thus releasing various cellular constituents, including amylase. In Spink3+/+ mice, the serum amylase level after 3, 6, and 9 injections was 14,080, 17,860, and 33,800 IU/L, respectively (n = 5). In Spink3+/− mice, the serum amylase level after 3, 6, and 9 injections was 12,000, 20,500, and 33,620 IU/L, respectively (n = 5). These results suggest that there is no difference in sensitivity to cerulein-induced pancreatitis. In addition, there was no significant difference in trypsin activities after 9 injections in Spink3+/+ mice (0.25 unit [Optical Density (OD)/DNA concentration (μg/μL)] SD = 0.03; n = 5) or Spink3+/− mice (0.23 SD = 0.02; n = 5), although these activities were increased compared with those in mice with saline injection (0.19 SD = 0.03; n = 5). The trypsin activity in the pancreases of Spink3−/− mice (0.55 and 1.1 at 0.5 and 1.5 days after birth, respectively; n = 3) was almost equal to that in wild-type mice (0.68 and 1.02; n = 3) and hetero-type mice (0.59 and 0.94; n = 3), thus suggesting that trypsin is not significantly activated in Spink3−/− mice. There were no significant differences in blood glucose levels of Spink3+/+, Spink3+/−, or Spink3−/− mice. We think that initially SPINK1 binds to trypsin to prevent activation of pancreatic enzymes and that a lack of SPINK1 may result in the premature conversion of trypsinogen into active trypsin within acinar cells, thus leading to autodigestion of the exocrine pancreas by activated proteases. However, in Spink3−/− mice, autophagic degeneration of acinar cells appeared at 16.5 days after coitus, and the rapid onset of cell death occurred in the pancreas and duodenum within a few days after birth, thus resulting in death by 14.5 days after birth. Thus, Spink3 has essential roles for maintenance of the integrity and regeneration of acinar cells during the perinatal stage. Spink3−/− mice died by 14.5 days after birth. Blood glucose levels were approximately the same in Spink3+/+, Spink3+/−, and Spink3−/− mice at 3.5 days after birth, when approximately 50% of Spink3−/− mice had already died. Thus, postnatal death may be due to exocrine dysfunction of the pancreas caused by the loss of acinar cells, but not of endocrine cells, and to malabsorption