Exp. Anim. 60(5), 433–444, 2011 —Review— Review Series: Frontiers of Model Animals for Human Diseases The Roles of Serine Protease Inhibitor Kazal Type 1 (SPINK1) in Pancreatic Diseases Masaki OHMURAYA1, 2) and Ken-ichi YAMAMURA2) 1)Priority Organization for Innovation and Excellence and 2)Institute of Resource Development and Analysis, Kumamoto University, Kumamoto 860-0811, Japan Abstract: Serine protease inhibitor Kazal type 1 (SPINK1) was originally identified as a trypsin inhibitor by Kazal et al. in 1948. SPINK1 is strongly elevated in pancreatitis and the elevation correlates with the severity of disease. In 2000, mutations in the SPINK1 gene were shown to be associated with chronic pancreatitis. Since then, there have been many reports on association between mutations in the SPINK1 genes and patients with pancreatitis. In 1982, SPINK1 was shown to be identical to tumor associated trypsin inhibitor (TATI). In addition, sequence similarities were detected between human epidermal growth factor (EGF) and human SPINK1 in 1983. Actually, SPINK1 was shown to stimulate growth of several cell lines including cancer cells in 1985. Recent clinical studies showed that high levels of SPINK1 protein in serum or urine were associated with adverse outcome in various cancer types. However, there was little evidence that showed in vivo function of SPINK1. Surprisingly, mice deficient in Spink3 (a mouse homologue gene of human SPINK1) showed excessive autophagy, but not pancreatitis in the exocrine pancreas, leading to autophagic cell death. We also demonstrated that SPINK1 acts as a growth factor through EGFR signaling. These data indicate that the role of the SPINK1 is not just as a trypsin inhibitor, but also as a growth factor as well as a negative regulator of autophagy. In this review, we summarize the roles of SPINK1/Spink3 in pancreatic diseases based on the data obtained from analyses using mouse models. Key words: autophagy, EGFR, growth factor, SPINK1/Spink3, trypsin inhibitor Introduction mutations in the SPINK1 gene are associated with chron- ic pancreatitis. Subsequently, exonic SPINK1 mutations, Serine protease inhibitor Kazal type 1 (SPINK1) was especially the N34S mutation, have been shown to be originally found as a trypsin inhibitor by Kazal et al. in associated with idiopathic and chronic pancreatitis [16, 1948 [25]. As SPINK1 was shown to be secreted from 66]. However, the incidence of this mutation is common, the pancreas and bind to trypsin to inhibit its activity, it occurring in 0.5–2.5% of the general population. Thus, was initially termed as pancreatic secretory trypsin in- it has been suggested that this mutation alone cannot hibitor (PSTI). In 2000, Witt et al. [75] showed that initiate the development of chronic pancreatitis and in- (Received 16 May 2011 / Accepted 20 June 2011) Address corresponding: K. Yamamura, Institute of Resource Development and Analysis, Kumamoto University, Kumamoto 860-0811, Japan 434 M. Ohmuraya AND K. Yamamura stead that this genetic variant acts as a disease modifier tant role for regulation of autophagy and autophagy is [59]. involved in the activation of trypsinogen. In this review, In 1982, Stenman et al. identified tumor-associated we will review the roles of SPINK1/Spink3 in the context trypsin inhibitor (TATI) in the urine of ovarian cancer of trypsin inhibitor, growth factor, and autophagy regu- patients [63]. TATI was promptly shown to be identical lator. to the SPINK1 based on the identical NH2-terminal amino acid sequence [21]. As tumors often expressed SPINK1/Spink3 as a Trypsin Inhibitor both TATI and trypsinogen, TATI was initially thought to have a similar function as in the pancreas, that is, Pancreatic digestive enzymes are stored as inactive protection of the tumor against the destructive activity precursors in pancreatic zymogen granules. Under nor- of trypsin within tumor cells. However, sequence simi- mal conditions, digestive enzyme activation is strictly larities were detected between human epidermal growth controlled to prevent autodigestion of the pancreas. factor (EGF) and human PSTI in 1983 [58], leading to However, in certain circumstances, excessive amounts the idea that SPINK1 has growth factor activity. Actu- of pancreatic trypsinogen are activated to trypsin, lead- ally, SPINK1 was shown to stimulate growth of several ing to activation of other zymogens and autodigestion cell lines including cancer cells [15, 40, 47]. In 1987, of the pancreas, pancreatitis. SPINK1 is synthesized in the human SPINK1 gene was identified as being ap- acinar cells of the pancreas and is thought to inhibit the proximately 7.5 kb consisting of 4 exons and was lo- trypsin activity in the pancreas. It possesses an active cated on chromosome 5 [20]. The gene product consists site (a lysine residue at position 41 and isoleucine residue of 79 amino acids including a 23 amino acid signal pep- at position 42, corresponding to positions 18 and 19 of tide with three intramolecular disulfide bridges (Cys9- the mature peptide) that serves as a specific target sub- Cys33, Cys16-Cys35, and Cys24-Cys56) [3, 20]. strate for trypsin. Trypsin and SPINK1 produce a com- SPINK1, with the molecular weight of 6.2 kD, is se- plex with a covalent bond between the catalytic serine creted by the acinar cells of the exocrine pancreas into residue of the enzyme and the lysine residue of SPINK1. the pancreatic juice. Later on, SPINK1 was shown to SPINK1 is considered to be the first-line safeguard sys- compete with EGF in the binding to Swiss 3T3 fibroblast tem that can inactivate about 20% of total trypsin activ- cells to almost the same extent, suggesting that SPINK1 ity if trypsinogen is accidentally converted to trypsin in binds to the EGF receptor (EGFR) [14]. In conjunction acinar cells. ��������������������������������������������Acute pancreatitis (AP) is a severe, debili- with the growth factor activity of SPINK1, roles of tating and sometimes fatal inflammatory disease. Its SPINK1 in cancer were suggested by many investiga- incidence has escalated dramatically in recent years. tions, such as a relationship with poor survival, invasion, However, no specific therapy exists.��������������������� Importantly, chron- or recurrence (see reference 54 for a review). However, ic pancreatitis is a risk factor for pancreatic ductal ad- detailed study has not been performed until recently. enocarcinoma (PDA) [34, 36]. This was clearly demon- During the past few years, we studied extensively the strated in cases of hereditary chronic pancreatitis and function of human SPINK1 and mouse Spink3 (mouse incidence of pancreatic cancer in these patients increased homologue of human SPINK1 is termed as Spink3) using by 53 times more than that in the control [35]. To prevent especially knockout mouse models. Through studies pancreatic cancer, it is important to elucidate the mo- using Spink3 knockout mice, we clearly demonstrated lecular mechanisms of pancreatitis development. How- that Spink3 acts a trypsin inhibitor in vivo. In vitro stud- ever, the causes of pancreatitis are so heterogeneous, and ies using several cancer cell lines revealed that the both genetic and environmental factors are involved in SPINK1 can stimulate growth of cancer cell lines through the onset of pancreatitis. Thus, one of the best approach- the EGF receptor and its downstream signaling pathway, es is to produce and analyze a mouse model for human MAPK/ERK. Furthermore, through the studies using hereditary pancreatitis. various genetically engineered mice, we found a com- Hereditary chronic pancreatitis (HCP) is a very rare pletely new function of Spink3. Spink3 plays an impor- form of early onset chronic pancreatitis. With the excep- SPINK1 in PaNCrEatic diseases 435 tion of the young age at diagnosis and a slower progres- Although the trypsin inhibitory activity of SPINK1 sion, the clinical course, morphological features, and was demonstrated by in vitro studies, there has been no laboratory findings of HCP do not differ from those of evidence demonstrating trypsin inhibitory activity in patients with alcoholic chronic pancreatitis. So far, gene vivo. If the Spink3 acts as a trypsin inhibitor in vivo, mutations of cationic trypsinogen (PRSS1), anionic pancreatitis will develop in the absence of Spink3. To trypsinogen (PRSS2), SPINK1, cystic fibrosis transmem- analyze this, we generated the Spink3-deficient mice by brane conductance regulator (CFTR), chymotrypsinogen gene targeting [50]. In this experiment, we used the C (CTRC), and calcium-sensing receptor (CASR) have exchangeable gene targeting method, which was a mod- been identified as being associated with HCP (72). In ified form of the exchangeable gene trapping method [1, the case of the PRSS1 gene, pancreatitis can be caused 2]. We produced a null allele in the mouse Spink3 locus by various molecular mechanisms. These include the using a targeting vector that contained a neomycin re- increase of stability by mutation in the autolysis site [5, sistance gene flanked by lox71 and lox2272 (Fig. 1A). 8, 73], increase of autocatalytic conversion of trypsino- Using a replacement vector that carries the human mu- gen to active trypsin by mutation modifying the “sub- tant SPINK1 gene flanked by lox66 and loxP, we can strate” property of trypsinogen [12, 65], disturbance of produce the humanized allele carrying the human mutant intracellular transport resulting in colocalization of SPINK1 gene in the mouse Spink3 locus. Thus, we can trypsinogen with lysosomal enzyme, cathepsin