Glucose-Stimulated Insulin Secretion via Perfusion through the Mice Vasculature with an Intact Pancreas

Summary

Here, we provide details on performing glucose-stimulated insulin secretion (GSIS) via perfusion through the mice vasculature with an intact pancreas as an additional technique to assess the endocrine pancreatic function with reduction of possible confounders.

Abstract

Current methods for glucose-stimulated insulin secretion (GSIS) in rodents involve ex vivo perifusion or static incubation of pancreatic islets and pancreatic slices. Ex vivo perfusion and static incubation allow for multiple samples to be taken quickly with efficiency. However, these procedures subject the islets to trauma, intermittent hypoxia, and reduced paracrine insulin signaling, all of which can affect GSIS. These negative effects can be minimized via perfusion through the mice vasculature with an intact pancreas, which provides another comparison method for the different GSIS models. The backbone of this perfusion procedure involved i) vessel ligation to isolate the pancreas and ii) cannulation of the pancreatic vasculature. This model involved seven ligation points and two cannulations. The resultant perfused vasculature included the lower abdominal aorta, the celiac and superior mesenteric artery, the branching tributaries, and the portal vein.

The perfusate was introduced to the vasculature at ~2 mL/min via a peristaltic pump to PE-50 tubing that cannulates the lower abdominal aorta. The perfusates ran through the isolated vasculature and were collected through the portal vein via a cannulation with PE-30 tubing. Perfusate was collected over 3 min intervals in rounds of 30 min for each perfusate solution of 2.6 mM glucose, 16.8 mM glucose, and 2.6 mM glucose with 3-isobutyl-1-methylxanthine (IBMX) in Krebs-Ringer bicarbonate buffer (KRB, 16 mM HEPES and 0.1% BSA, pH 7.4). The IBMX solution was preceded by a normal 2.6 mM glucose solution as a wash cycle. Total time was ~1 h for the surgical procedure and 2 h for perfusate collection. Perfusates were frozen until an ELISA was performed to measure insulin secretion.

This method list hopes to provide a detailed procedural basis for insulin secretion measurement through pancreas perfusion through the vasculature, which will allow researchers to perform intact pancreas perfusion successfully in their laboratories.

Introduction

Quantifying and understanding beta-cell function has primarily been accomplished via ex vivo perifusion, static incubation, and, more recently, via pancreatic slices. While these procedures allow for the evaluation of isolated islets, the former two may put the islets through stressors that impact their ability to function. In a study by Rosenberg et al., canine islets isolated using standard methods showed increased beta cell apoptosis, which was evidenced by loss of the basement membrane, appearance of pyknotic nuclei, and apoptotic bodies¹. This study coincides with another report on the isolation of human pancreatic islets where morphologic changes depicting apoptosis were demonstrated soon after islet isolation^2,3. Increased beta-cell apoptosis led to reduced insulin release from these islets, supporting the concept that islet isolation predisposes them to damage and functional impairment.

In addition to apoptosis, the vasculature is lost with islet isolation, leading to impaired transportation of nutrients, metabolites, and hormones^4,5. This limitation has led to alternatives like transplantation of the islet cells into the eye to maintain some form of vasculature and innervation⁶. The change in vasculature architecture during islet isolation also creates a more compact islet, different from what is expected in vivo^7,8.

To add to the anatomical differences, the behavior of statically incubated islets differs from the isolated perfused pancreas. For statically incubated islets at low glucose, 77% of the total insulin released (3.31 ng out of 4.34 ng total) appeared to be present within the first 30 min and little in the succeeding time period. This is in stark contrast to the isolated perfused pancreas at low glucose concentrations, where there was virtually zero insulin release at all time points (<1.8 ng total)⁹. While further research must be done to fully elucidate this difference at low glucose concentrations, this difference in data may represent leakage from hypoxic dying cells, which was further examined via static incubation of monolayers of islets cells. Nevertheless, these dissimilarities between current methods and in vivo environments can cause impaired external validity (Table 1).

In addition to causing hypoxia, both ex vivo perifusion and static incubation involve culturing and separating the islets from the pancreas into tissue culture dishes to keep them in a suspended state. During this process, collagenase is typically perfused through the pancreatic duct to allow for pancreatic tissue dissociation. Collagenase binds to collagen fibrils and unravels them with the help of neutral proteases, gelatinases, and endogenous pancreatic proteases. Together, these proteases continue to digest until the islets are released from their extracellular matrix and acinar-cell attachments. Balancing this mixture of proteases yields varying results in both the purity and functionality of the isolated islets¹⁰. The efficacy of collagenase also relies on animal characteristics like strain, age, weight, and diabetes status. The use of these proteases in conjunction with animal-specific traits can harm pancreatic islets during the isolation process. When administering collagenase, these proteases also activate endogenous pancreatic proteolytic enzymes that aid the digestion of the basement membrane. However, controlling endogenous pancreatic enzyme activation is not yet possible; thus, this may play a role in islet harm during the cultivation process^11,12,13. One report from Edmonton showed endogenous protease activity affecting islet isolation yield once reaching a sufficient enzyme activity level for isolation^14,15.

Current infusion techniques of collagenase (intraductal) have been shown to also penetrate the islet interior, leading to possible islet fragmentation and, thus, low islet yields¹⁶. In a study by Balamurugan et al., collagenase had been seen to localize in islets and acinar tissue within infused mouse pancreas, leading to intense inflammation and activation of apoptotic pathways¹⁷. Islets that did survive the isolation phase showed reduced ability to release insulin as well.

In addition, other factors like mechanical stress during islet preparation and hypoxia-induced necrosis during the culture all play a role in islet functioning¹⁸. These traits can influence the viability of the islet when isolating the pancreas and, therefore, the resultant function.

Apart from islet isolation and evaluation via ex vivo perifusion and static incubation, islets in living pancreatic slices are another ex vivo method used to measure endocrine cell responses¹⁹. In this system, the chemical stress from islet isolation procedures is avoided. Additionally, the islet microenvironment remains intact, allowing for the study of islet physiology closer to its in vivo state²⁰. Similar to ex vivo perifusion and static incubation, this is a method that will continue to play an important role in acquiring knowledge of the pancreas. However, some limitations include mechanical stress during slice generation, as well as lacking an overall holistic view of an intact pancreas for pathology and physiology. The slice represents a particular region within the pancreas, which decreases its generalizability due to the decreased sample size and limited view²¹.

This islet isolation and other methods are accompanied by various points of stressors that may be imposed on the islets. Mechanical stress, hypoxia, protease overactivity, and limited view all play a role in the resultant islet yield and overall external validity. These factors can be reduced through perfusion of the pancreas via the mice vasculature, which minimizes all of the above stressors while also allowing for examination of the effects of other secretagogues and pharmacological agents present in the body due to maintaining cellular integrity²². Prior studies have shown the impact of islet blood flow on plasma insulin levels²³; in turn, vasoactive molecules like angiotensin peptides and sympathetic agonists, which change islet blood flow, would alter islet function and insulin release^24,25. Pancreas perfusion would allow further assessment of how these and other different molecules may affect endocrine cell response. This system hopes to allow for further conceptual advancement of our understanding of the endocrine pancreas biology in the presence of cross-talk in its natural microenvironment with an intact vasculature and innervation. Additionally, perfusion of the intact pancreas in this protocol allows for isolation of the organ, reducing potential confounders from other organs like the liver, which clears insulin from circulation, when compared to perfusion of the whole mouse. Perfusion of the whole mouse would allow for reduced surgical technique; however, it would still require the collection of the perfusate through a vessel after the pancreas and prior to reaching the liver -- likely the portal vein. Portal vein collection would continue to necessitate cannulation. While whole mouse perfusion would provide the benefit of reduced surgical steps, there would be limitations due to the confounders created by organs like the kidneys and lungs producing angiotensin peptides. However, this would be a novel approach that could prove useful when examining the effects of vasoactive molecules produced from other organs on pancreatic islets, which is the primary priority.

Whole pancreas perfusion has been successfully performed in the past by others^7,22,26. In this updated protocol, the use of mice poses an increased technical challenge due to their smaller anatomy compared to rats. However, this procedure successfully demonstrates the feasibility of this protocol by demonstrating a well-characterized insulin secretory profile from the intact pancreas of 3 independent mice.

Protocol

All research performed on rodents is in compliance with Joslin Diabetes Research Center guidelines (approval number 22-01).
NOTE: This constitutes a non-survival procedure with a primary and secondary method of euthanasia as requested by the IACUC committee.

1. Beginning

1. Administer anesthesia.
   1. Inject fasted, normoglycemic mice (Flox/Cre 53-74 week-old M/F used in experiments) with ketamine/xylazine mix (overdose was used as primary method of euthanasia: ketamine 100-120 mg/kg and xylazine 10 mg/kg; total volume of mixture injected during trials ~0.08-0.14 mL) administered intraperitoneally via 28 G needle.
      NOTE: Use of normoglycemic, fasted mice is important to reduce effects of ketamine; however, an anesthetic can be substituted.
   2. Ensure the mouse is properly anesthetized by pinching the posterior foot with forceps to check for reflex.
      NOTE: Lack of reflex is denoted by lack of movement of the pinched leg.
   3. Administer sterile ophthalmic ointment to eyes to prevent dryness while under anesthesia and prevent corneal ulcers.
2. Secure the subject.
   1. Ensure the mouse is secure and will stay in place during the procedure, which this experiment completes via magnets attached to rubber bands that connect to the limbs with the magnets attached to a magnetic board.
3. Clean the incision site.
   1. Ensure that the mouse is sprayed down with 70% ethanol to prevent contamination from the mouse's hair,
      NOTE: Shaving the mouse would allow for less chance of contamination from the fur.
4. Perform incision.
   1. Using forceps/tweezers, pinch 2-3 cm from the anus and cut upwards while piercing the peritoneum with curved scissors.
   2. Create a diagonal incision from the midline to the ribs and lateral portion of the mouse.
   3. Create an identical incision on the opposite side.
      NOTE: The goal is to create adequate sight of the abdominal cavity from near the rectum to the diaphragm

2. Ligation

1. Ligate the gastric artery.
   1. Move intestines to the animal's right with blunt forceps/fingers. Lift the stomach to locate the gastric artery and vein.
      NOTE: Gastric artery and vein identified through visualization of vessels running from the lower esophagus to the upper stomach.
   2. Take a 7-0 suture (no needle) and place it under the gastric artery using blunt forceps in preparation for ligation.
   3. Create a surgeon's knot and ensure adequate ligation via a firm knot.
      NOTE: The type of knot is not important; prevention of blood flow can be determined via a change of color of the ligated organ to a paler appearance or removal of the organ and monitoring for blood expulsion from the ligation point. Ligation of the gastric artery ensures no influence from gastric contents. This also reduces leakage to the gastric vessels and directs perfusate through to the celiac artery.
2. Ligate left renal vessels.
   1. Repeat steps 2.1.1-2.1.3 for the left renal vessels.
      1. Identify the left renal vessels by visualizing bright red and dark purple vessels connecting to the left kidney. This step prevents leakage to the renal vessels and redirects perfusate to the celiac artery.
3. Ligate the upper abdominal aorta.
   1. Repeat steps 2.1.1-2.1.3.
      1. Identify the upper abdominal aorta by visualizing the midline bright red vessel with branching vessels of the celiac artery, superior mesenteric artery, and left renal vessels. Perform ligation superior to the celiac artery to reduce leakage through the upper abdominal aorta and ensure flow to the celiac artery and eventually, the pancreas.
4. Ligate right renal vessels.
   1. Move intestines to the animal's left. Pinch the tissue/fat surrounding the right kidney as leverage to identify the right renal vessels.
   2. Repeat steps 2.1.1-2.1.3.
      1. Identify right renal vessels by visualizing purple/dark red vessels (both right renal veins with an artery lying underneath that is not visible) connected to the right kidney with connection to both inferior vena cava and abdominal aorta. This step parallels step 2.2 with the left renal vessels, allowing for further isolation of the pancreas; however, be wary of the proximity of this vessel to the inferior vena cava.
5. Ligate the abdominal portion of the inferior vena cava.
   1. Identify the abdominal portion of the inferior vena cava through visualization of a larger purple vessel with a branching connection to the right renal vein.
   2. Repeat steps 2.1.1-2.1.3.
      NOTE: This step reduces potential confounders from the blood going through to the portal vein, where the perfusate will be collected.
6. Ligate the lower abdominal aorta, before bifurcation (Figure 1)
   1. Repeat steps 2.1.1-2.1.3.
   2. After ligating just before the bifurcation, place a suture thread under the aorta 0.5 cm or so above the first knot. Use this suture thread to tie the cannulation into place.
   3. Use the space between the suture thread and ligated suture to insert the tube for cannulation. This ligation ensures perfusate runs up through the lower abdominal aorta towards the celiac artery.
7. Ligate the portal vein (Figure 2)
   1. Move intestines to the animal's left to allow for proper visualization of the portal vein.
   2. Create an incision into the diaphragm and ribs to create more space, as needed, to view the portal vein and for appropriate traction for ligation and cannulation.
   3. Ligate the vessel as superior as possible (closer to the liver) to ensure space for cannulation later.
   4. Place a second suture inferior to the first but above the pancreas. Step 2.7.2 will cause the subject to pass away. Additionally, this step is important to allow for perfusate to be collected and prevent potential backflow superior to the cannulation point.

3. Cannulation

1. Cannulate the lower abdominal aorta.
   1. Create a 0.5-1.5 mm incision with scissors into the lower abdominal aorta between the bottom knot and superior suture.
   2. Using the lowest knot as leverage, insert the tube into the aorta via the nick until it reaches the more superior suture; take care not to tear.
   3. Tie the more superior suture into a knot to lock the tube into place.
      NOTE: This step allows for the perfusate to enter the enclosed system. With the ligations in place, the perfusate should travel from the entry point in the lower abdominal aorta through to the celiac artery, then the pancreatic vessels, ending with the hepatic portal vein where the perfusate can be collected.
2. Cannulate the portal vein.
   1. Create a 0.5-1 mm incision into the portal vein just after the portal vein leaves the pancreas.
   2. Holding the top knot as leverage with blunt forceps in one hand, insert the tube into the portal vein using forceps to grip the tube with the other hand.
   3. Use the second suture to tie the tube into place, if possible.
      NOTE: This step is the end point of the enclosed system where the perfusate should exit with the impact of the pancreas on the fluid present.

4. Collection

1. Perform perfusion.
   1. Before placing the lower abdominal aorta tubing, ensure the perfusion system has begun with the perfusate flowing at the desired speed (~2 mL/min).
      NOTE: The perfusate should drain from the portal vein cannulation and into a collection tube. A polymerase chain reaction (PCR) tube can be used for easier storage. The perfusate solutions used here are 2.6 mM glucose with 1.2 mM 3-isobutyl-1-methylxanthine (IBMX), 16.8 mM glucose, and 2.6 mM glucose with 3-isobutyl-1-methylxanthine (IBMX) in Krebs-Ringer bicarbonate buffer (KRB, 16 mM HEPES and 0.1% BSA, pH 7.4). Perfusate solution began with a low-glucose solution to high-glucose, back to low glucose, and ended with an IBX solution for control.
2. Collect the perfusate.
   1. Collect the perfusate over 30-min intervals into PCR tubes (replaced tubes every 3 min).
   2. After each 30-minute interval, switch the solution.
   3. Perform an enzyme-linked immunosorbent assay (ELISA) with the samples. This can be done at a later point if stored appropriately.

Results

The results after ELISA should demonstrate insulin secretion from the mouse pancreas to an extent dependent on perfusate glucose levels. Because of the novelty of this procedure in mice, analyzing the outcome and its success depends upon comparison with multiple other trials of a perfused mice pancreas. A biphasic insulin response to high glucose concentration is considered physiological, as assessed in human subjects and with islet perifusion methods. The results presented are similar to...

Discussion

The protocol outlined above aims to create an enclosed system composed of the intact pancreas, two inlets (celiac and superior mesenteric artery), and one outlet (hepatic portal vein). All other vessels are ligated to prevent leakage when perfusate is flown through, as well as to ensure isolation of the pancreas itself with its microenvironment. The ligation steps in this protocol also allow for the flow of the perfusate through the pancreas without travel to other organs that may potentially confound the results, like t...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Fiber membrane oxygenator	Harvard Apparatus	73-3757	Allows oxygen to enter perfusate
Heating pad	Braintree scientific	39OM
Ketamine/xylazine	Patterson Veterinary	78908598
Magnetic board	Braintree scientific	39OM
Magnets with rubber bands	Braintree scientific	39OM
Microscopy	N/A	N/A	No specific company
Oxygen tank	N/A	N/A	No specific company
PCR tubes	ThermoFisher/Invitrogen	AM12225	Used for perfusate collection
PE-30 tube	Braintree scientific	PE30
PE-50 tube	Braintree scientific	PE50
Peristaltic pump	N/A	N/A	No specific company
Surgical tools	N/A	N/A	No specific company
Suture	N/A	N/A	No specific company