3.3 Duct Obstruction Models
Many studies have demonstrated that obstruction of the pancreatic duct，either partial or complete，leads to morphological changes compatible with chronic pancreatitis.Proximal to the stenosis intraductal pressure increases and favors duct dilation with atrophy of acinar cells and replacement by fibrous tissue.In humans possible causes for a blockage of the duct are tumors（either adenocarcinoma or cystic or endocrine neoplasms）or more rarely cystic fibrosis that plugs the ductal lumen by its viscous mucin（Kloppel et al.2004）.Humans have one main pancreatic duct that enters the duodenum at the major duodenal papilla after joining the bile duct.Early in embryonic development two efferent pancreatic ducts，which origin form the ventral and dorsal part of the organ，fuse together to form one main duct.If this fusion fails to occur（in up to 9%of autopsy studies）two separate pancreatic ducts are preserved，a condition known as pancreas divisum（Brock et al.2013）.
3.3.1 Pancreatic Anatomy in Rodents
Currently the majority of pancreatic duct obstruction models are performed in rodents.However，the anatomy of the pancreas differs among species.Since most studies are carried out in mice or rats we want to give more detailed information regarding anatomy of the pancreas in these two species.Compared to the human pancreas in which gross anatomy shows a more compact or cohesive pattern the mouse pancreas is more loose as it consists of three separate lobes，the gastric，duodenal and the splenic lobe and is frequently interspersed with fatty，connective or lymphoid tissue（Treuting et al.2012）.Each lobe is drained by a separate pancreatic duct.Duct anatomy of the mouse pancreas shows variations that make ligation procedures more complex.Usually the ducts from the splenic and gastric lobe converge and will be completed by the duct from the duodenal lobe to form one common channel before opening into the common bile duct.In about 10%the splenic and the gastric duct join the duodenum separately with an accessory papilla（Watanabe et al.1995）.Like in mice the rat pancreas also consists of three lobes，two smaller ones（the biliary and the duodenal lobe）and a larger gastrosplenic lobe（Kara 2005）.The pancreatic duct system consists of two main ducts，the anterior and the posterior pancreatic duct.Both finally open into the biliopancreatic duct.In addition minor ducts of all three lobes are petering out separately to the biliopancreatic duct（Kara 2005）.
The size of pancreatic duct varies naturally between species.Magnetic resonance imaging（MRI）and autopsy studies revealed that the mouse pancreatic duct has an approximate diameter of 88 pm and a total tractable length of 5850 µm（Grippo et al.2011）.Caliber of the rat biliopancreatic duct is larger with around 1 mm and a length of ca.29 mm（Kara 2005）so that ligation procedures are technically easier.
Selective obstruction of one branch of the pancreatic duct and not of the whole biliopancreatic duct uncouples only one part of the pancreas that will be injured.For this reason induction of chronic pancreatitis can be locally controlled within the pancreas.Since the unobstructed part remains unaffected it can serve as an internal control（Kishi et al.2003；Sendler et al.2015）.Technique of（selective）ductal ligation forms a big advantage over other models of chronic pancreatitis as a“negative”control is created within the same animal ruling out confounding variables.
3.3.2 Ligation Models in Rodents
Ligation models belong to the best experimental models for mimicking chronic pancreatitis and are well established for mice and rats.In general first changes start within the first 7-10 days after the intervention with necrosis，an inflammatory cell infiltrate and first signs of acinar cell atrophy，ductal cell proliferation and initiation of fibrosis that is often visible with periductal and intralobular arrangement（Watanabe et al.1995；Sendler et al.2015；Churg and Richter 1971；Miyauchi et al.2007；Scoggins et al.2000；Yamamoto et al.2006）.Emergence of fibrosis is associated with activation of pancreatic stellate cells（PSCs）that transform from a quiescent to an active state by increased expression of alpha smooth muscle actin（aSMA）and deposition of extracellular matrix components（Apte et al.1999）.Later（≥14 days）changes include more extensive fibrosis with upregulation of collagen Ⅰ and Ⅲ production and a more perilobular distribution.In addition a fatty tissue replacement occurs（Watanabe et al.1995；Sendler et al.2015；Yamamoto et al.2006）.
Notably，exocrine insufficiency is hardly seen in the ligation models unless a high fat diet is fed.On the one hand this is caused by the incomplete obstruction of the pancreatic ductal system so that one part of the pancreas remains unaffected and keeps up secretory enzyme production.Secondly very early after ligation a subgroup of pancreatic cells，intermediate cells，start a transdifferentiation and display both exocrine and endocrine phenotypes as indicated by co-expression of insulin and amylase（Bertelli and Bendayan 1997）.These cells are detectable within the islets of Langerhans and within exocrine tissue and indicate a high degree of plasticity of the organ within the first days after the insult to maintain exocrine and endocrine function.
The lack of endocrine insufficiency might be also explained by an increased islet beta-cell proliferation in duct-ligated rats.The Beta cell population nearly doubled within the first week after the ligation and small islets and islet-cell clusters developed（Wang et al.1995）.Interestingly，glucose transporter type 2（GLUT-2），the major glucose transporter isoform，was expressed in ductal cells，besides its primary location on insulin secreting beta cells.Obviously ligationstimulated ductal cells reach a metaplastic state and acquire properties of endocrine cells with higher glucose sensitivity（Wang et al.1995）.(https://www.daowen.com)
3.3.3 Variations of Classical Ligation Procedures
A continuous pancreatic duct hypertension was induced by double-ligation of the bilio-pancreatic duct at its proximal and distal end in rats.Both pancreatic juice and bile were selective collected and diverted into the duodenum.Two weeks after the procedure mucoprotein concentration of the pancreatic juice was increased despite a decrease of digestive enzyme contents indicating that non-enzymatic protein secretion starts on sustained pancreatic duct hypertension（Yamamoto et al.2006）.Previous studies in humans already showed that non-enzymatic protein secretion leads to enhanced viscosity of the pancreatic juice and formation of intraluminal protein plugs（Harada et al.1981）.First histomorphologic changes were observed earlier，i.e.at already 7 days，with intralobular and perilobular fibrosis and a peak of acinar cell apoptosis.
Possibly the observation time needs to be much longer for studying effects on exocrine function.In a model established by Isaksson and collaborators the rat pancreatic duct was obstructed with a glue-like substance consisting of either acrylate or prolamine and animals were monitored for 5 months in total.Exocrine insufficiency was present as evidenced by reduced secretory enzyme levels whereas endocrine function was not impaired（Isaksson et al.1983）.These results also tell us that the investigation of long-term effects on the pancreas is technically possible when using duct occlusion models but an observation time of months will be required.
3.3.4 Ligation Models in Non-rodents
There are fewer studies dealing with mammals whose pancreatic anatomy looks more similar to the human one.In dogs ligation procedures were studied that led to typical morphological characteristics of chronic pancreatitis.These changes developed within 3-6 months.Pancreatic excretory capacity was diminished，too（Tanaka et al.1988，1998）.Compared to rats connective tissue replacement was stronger，however acinar atrophy was less marked（Churg and Richter 1971）.The degree of pancreatic damage was much more distinct，when ligation was performed in conjunction with chronic ischemia or ethanol administration to these dogs（Tanaka et al.1988）.
3.3.5 Limiting Factors of Ligation Models
Comparison of ligation models is limited as the location of ligation differs among studies.In addition other stimuli were applied to animals such as caerulein（Sendler et al.2015），ischemia（Tanaka et al.1988）or ethanol feeding（Tanaka et al.1988）that affect severity of the disease.Knowledge of the exact location of pancreatic duct ligation is of high importance when comparing the extent of changes of chronic pancreatitis.
Ligation procedures are rather invasive procedure and demand surgical skills and exact knowledge on the anatomy of abdominal organs.Moreover technical equipment has to be available for doing the operation.These factors set limitations on general application of duct-ligations models.A learning curve when doing the procedure is inevitable and has to be taken into account when analyzing the results.