Effects of subtotal pancreatectomy and long‐term glucose and lipid overload on insulin secretion and glucose homeostasis in minipigs

Abstract Introduction Nowadays, there are no strong diabetic pig models, yet they are required for various types of diabetes research. Using cutting‐edge techniques, we attempted to develop a type 2 diabetic minipig model in this study by combining a partial pancreatectomy (Px) with an energetic overload administered either orally or parenterally. Methods Different groups of minipigs, including Göttingen‐like (GL, n = 17) and Ossabaw (O, n = 4), were developed. Prior to and following each intervention, metabolic assessments were conducted. First, the metabolic responses of the Göttingen‐like (n = 3) and Ossabaw (n = 4) strains to a 2‐month High‐Fat, High‐Sucrose diet (HFHSD) were compared. Then, other groups of GL minipigs were established: with a single Px (n = 10), a Px combined with a 2‐month HFHSD (n = 6), and long‐term intraportal glucose and lipid infusions that were either preceded by a Px (n = 4) or not (n = 4). Results After the 2‐month HFHSD, there was no discernible change between the GL and O minipigs. The pancreatectomized group in GL minipigs showed a significantly lower Acute Insulin Response (AIR) (18.3 ± 10.0 IU/mL after Px vs. 34.9 ± 13.7 IU/mL before, p < .0005). In both long‐term intraportal infusion groups, an increase in the Insulinogenic (IGI) and Hepatic Insulin Resistance Indexes (HIRI) was found with a decrease in the AIR, especially in the pancreatectomized group (IGI: 4.2 ± 1.9 after vs. 1.5 ± 0.8 before, p < .05; HIRI (×10−5): 12.6 ± 7.9 after vs. 3.8 ± 4.3 before, p < .05; AIR: 24.4 ± 13.7 µIU/mL after vs. 43.9 ± 14.5 µIU/mL before, p < .005). Regardless of the group, there was no fasting hyperglycemia. Conclusions In this study, we used pancreatectomy followed by long‐term intraportal glucose and lipid infusions to develop an original minipig model with metabolic syndrome and early signs of glucose intolerance. We reaffirm the pig's usefulness as a preclinical model for the metabolic syndrome but without the fasting hyperglycemia that characterizes diabetes mellitus.


| INTRODUC TI ON
Type 2 diabetes, one of the major diseases of the twenty-first century, has an elevated prevalence (10.5% of the global population). 1 The World Health Organization (WHO) defines this disease as reaching a glycemia over 126 mg/dL after 8 h of fasting, twice validated, or over 200 mg/dL following an oral glucose tolerance test. 2 Currently, it is understood that a variety of pathophysiological processes plays a role in the onset of type 2 diabetes 1,3 : decrease of insulin secretion, insulin resistance as a result of an imbalanced intake of carbohydrates and lipids and "thrifty genotype". 4 Additionally, type 2 diabetes is highly heterogenous and can be divided into five novel subtypes: severe autoimmune diabetes (related to type 1 diabetes), severe insulin-deficient diabetes, severe insulin-resistant diabetes, mild obesity related diabetes and mild-age related diabetes. 5,6 Due to the disease's polymorphism, type 2 diabetes research needs the appropriate preclinical models in order to better understand pathophysiological pathways and create innovative, effective therapeutic approaches.
The most effective interventional treatment for type 2 diabetes in obese patients nowadays is metabolic surgery, 7-10 which enables an early diabetes remission independent of weight loss. [11][12][13][14] Current research on metabolic surgery is concentrated on understanding the link between the physiological changes that occur after intervention and the clinical benefit as well as on the development of new methods intended to improve the metabolic phenotype by reducing the associated complications. 15 However, using preclinical models in which the surgical procedure may be easily applied to humans is one of the issues with metabolic surgery research. Porcine models have more translational value than rat models, even if rats are more often employed as preclinical models of metabolic surgery. [16][17][18][19] Pigs are in fact omnivorous and are similar to humans concerning the morphology and the physiology of their gastrointestinal tract, pancreas, propension to obesity and sedentary, metabolic biomarker levels and drug pharmacokinetics, 20,21 making it a particularly suitable preclinical model for these kinds of studies.
However, the creation of type 2 diabetes itself is the main problem with the preclinical pig model. Current commercial pigs for meat production are the results of generations of selective breeding that targeted a phenotype able to store energy for later consumption by humans, likely making them protected against the deleterious effects of a "diabetogenic" environment. 22 But some minipig strains, such as the Göttingen one, on which the physiology of insulin secretion is similar to humans, 23,24 or the Ossabaw, recognized to be a natural model of metabolic syndrome, 25 were found to be more susceptible to metabolic issues. Because of this, various studies have attempted in the past to develop preclinical diabetic pig models from these strains. A number of strategies have been tried, including: a surgical strategy involving total or subtotal pancreatectomy, which results in a strong reduction in insulin secretion but has no effect on insulin sensitivity [26][27][28] ; a chemical strategy involving the use of beta-cell toxins, such as streptozotocin or alloxan, with variable results and significant hepato-and nephrotoxicity [29][30][31][32][33] ; dietary interventions with a High-Fat, High-Sucrose diet, resulting in an insulin-resistant and obesity-related phenotype but not type 2 diabetes [34][35][36][37][38] ; and genetic engineering, which can produce customized pig models [39][40][41] but is logistically challenging and may have unintended consequences. 42 In conclusion, no true type 2 diabetic pig with fasting hyperglycemia and insulin resistance conforming to type 2 diabetes definition has been created so far.
In the current study, we considered additional approaches involving the combination of existing methods: an oral energetic overload (a High-Fat, High-Sucrose diet) and a subtotal pancreatectomy were separately performed and then also combined, intended to outperform the pancreas's regulatory capabilities. A parenteral approach using long-term intraportal glucose and lipid infusions, combined or not with a prior subtotal pancreatectomy, was also attempted. If the parenteral nutrition has already been set up to induce metabolic disorders in a piglet model, 43,44 it was, to our best knowledge, never tested in the adult pig as an approach to induce type 2 diabetes.

| Animals and housing
The study included a total of 21 healthy 1-year-old minipigs: 4 Ossabaw minipigs (DTU, Lyngby, Denmark) and 17 Göttingen-like preclinical model for the metabolic syndrome but without the fasting hyperglycemia that characterizes diabetes mellitus.

K E Y W O R D S
energetic overload, hyperglycemia, minipig model, pancreatectomy, type 2 diabetes (Pannier, Wylder, France), weighing respectively 48.2 ± 1.9 kg and 31.7 ± 11.0 kg. Our local minipig strain, called Göttingen-like, was created more than 30 years ago as a consequence of an initial crossing with the Göttingen strain. To limit the metabolic differences related to the female hormonal cycle, only males were included. At the start of the protocol, animals were either surgically castrated or delivered castrated in the animal facility. All animals were housed and enriched in individual boxes in conventional conditions. Water was provided ad libitum and 400 g of standard food (Swine Engrais-F S25/T, Uneal Cooperative) was given twice a day. The composition of the standard diet was detailed in Table 1. The light/dark cycle was 12 h of light and 12 h of darkness with a temperature between 19 and 24°C. Pigs benefited from a 15-day acclimatization period.

| Study design
In this study, we wanted to create a preclinical pig model of type 2 diabetes that corresponds to the WHO definition. 2 This led to the combination of a partial pancreatectomy, a surgical method of insulin deprivation, with methods of energetic overload, via oral or intraportal administration. In order to determine which strain of minipigs was most suited for our strategy, we first assessed how both the Göttingen-like strain (n = 3) and the Ossabaw one (n = 4) responded to a 2-month High-Fat, High-Sucrose diet (HFHSD). We decided thereafter to discard the Ossabaw strain because Göttingen-like showed a phenotype closer to our expectations. Following this choice, we combined a subtotal pancreatectomy with 2 months of HFHSD in Göttingen-like minipigs (n = 6). Finally, we explored a different strategy by infusing intraportal glucose and lipids for 3 weeks as a parenteral energetic overload in two groups of Göttingen-like minipigs: in Group 1 (n = 4), no subtotal pancreatectomy was initially performed; in Group 2 (n = 4), a subtotal pancreatectomy was performed prior to the energetic overload. The impact of the subtotal pancreatectomy on glucose metabolism in the Göttingen-like strain was simultaneously examined (n = 10), including the animals subjected to the HFHSD following the pancreatectomy (n = 6) and those of Group 2 (n = 4). Figure 1 displays the general design of this research.  The external jugular vein was exposed in the neck region after skin and muscle incision. After venotomy, the catheter (Hickman® 9.6F Single-Lumen CV Catheter, Bard Access System, USA) was inserted and linked to the vein with two ligatures (Vicryl® Bobine 2/0, Ethicon, France). It was tunnelized via the subcutaneous tissue from the incision zone to the dorsal area of the neck. Muscular and cutaneous layers were then closed by simple overlock (respectively Polysorb® 2/0, Medtronic, France and Mersilene® 1, Ethicon, France). This catheter remained throughout the duration of the procedure and was kept operational by administering 5 mL of physiological serum that had been heparinized (1 mL heparin at 5000 IU/mL for 250 mL NaCl 0.9%) after each usage or every 2 days if it was not.

| Subtotal pancreatectomy
By reclining the stomach cranially and the intestinal system caudally, the pancreas was made accessible. From the tail to the head, the dissection was carried out (splenic lobe). In the retroportal region, the connecting lobe was similarly dissected and largely removed. Before section and extraction, ligatures between the splenic and the duodenal lobes were performed, and TA B L E 1 Composition (in %) of the standard diet and the High-Fat High-Sucrose (HFHS) diet given to the minipigs.

| Intraportal glucose and lipid infusions
Over the period of 3 weeks, the intraportal catheter was used to administer lipid and glucose infusions twice daily for 2 h. infusion's glycemia above 500 mg/dL. Each infusion was preceded by a 500 mg/kg bolus of 50% glucose solution to raise blood glucose levels to more than 500 mg/dL within 1 min. All the infusions were performed in an awake animal. On EDTA and heparinised tubes, blood samples were obtained before the MMT was administered (t = 0 min) and at various time intervals afterwards (t = 15, t = 30, t = 60, t = 90, t = 120, and t = 180 min).

| Intravenous glucose tolerance test (IVGTT)
Following an overnight fast, a 50-% glucose solution (G50®, B. Braun, France) was intravenously administered into the CVC at a dose of 500 mg/kg. On EDTA tubes, blood samples were taken in the awake animal before (t = 0 min) and following the administration of glucose at t = 1, t = 3, t = 5, t = 10, t = 15, and t = 30 min.
Plasma was collected from each tube, centrifuged at 4000 rpm for 10 min at 4°C, and then stored at −80°C until analyses.

| Biological analyses
The amperometric glucose oxidase method was used to measure the level of glucose in blood (glucometer Accu-Chek Performa®, Roche, France, or Nova Biomedical StatStrip Xpress®, DSI, USA). A DXI Access Immunoassay System (Beckman Coulter) with an assay range between 0.3 and 300 μIU/mL was used to measure the plasma insulin concentrations, as previously mentioned. 47 Plasma lipid profile (total cholesterol, LDL, HDL and triglycerides) was assessed using an Abbott Architect C4000® clinical chemistry analyser.

| Calculations and statistical analyses
For data analysis, GraphPad Prism v8® software was employed. For curves, the results were expressed as mean ± SEM, and for histograms, as mean ± SD. Depending on the situation, paired or unpaired Student's t-tests were used to analyse the variables. A Two-Way ANOVA and Sidak post-hoc tests were used to compare blood glucose and plasma insulin levels during the MMT and IVGTT between the different strains of minipigs or between baseline and after diabetogenic interventions. For each comparison of blood glucose or insulin evolution during metabolic test, the effect of time of the metabolic test (called "time") and strain (Göttingen-like or Ossabaw, called "strain") or diabetogenic intervention ("HFHSD", "pancreatectomy", or "infusions") was systematically assessed. The presence of interaction between "time" and "strain" or "time" and "intervention" was also evaluated. The calculation of Insulinogenic Index was performed to evaluate the postprandial early insulin secretion as for glycemia ( Figure 2C) and insulin ( Figure 2D) did not change between the two steps. After HFHSD, Ossabaw minipigs showed a trend of lower postprandial blood glucose levels ( Figure 2E) accompanied by a trend of higher insulin peak secretion ( Figure 2F).
However, during the IVGTT, there were no discernible changes between the glucose decline ( Figure 2G) and corresponding insulin concentrations ( Figure 2H). Ossabaw minipigs' fasting blood glucose appeared lower than Göttingen-like ones at baseline, and it was significantly lower after HFHSD than those of Göttingenlike (70.0 ± 3.4 after vs. 79.0 ± 4.6 mg/dL before; p < .05) ( Figure 2I). p < .05). These findings indicated that Ossabaw minipigs had a better early insulin response than Göttingen-like minipigs. We thus decided to proceed with our strategy using the Göttingen-like strain.

| Reduction of acute insulin response after subtotal pancreatectomy in Göttingen-like minipigs
We assessed how a subtotal pancreatectomy affected glucose metabolism ( Figure 3). After the surgical procedure, there was no rise in fasting blood glucose ( Figure 3A). Mixed Meal Tests did not reveal any appreciable changes in blood glucose ( Figure 3B) or insulin levels ( Figure 3C). As a result, there was no change in the Insulinogenic Index ( Figure 3D). When compared to before the intervention, the speed at which the glucose levels declined during the IVGTT following pancreatectomy was slower (blood glucose levels of respectively 175.1 ± 12.4 mg/dL after vs. 109.4 ± 13.1 mg/dL before at 30 min, p < .05) ( Figure 3E). A significant interaction between "time" and "pancreatectomy" was thus reported (p < .05). Plasma insulin levels during IVGTT were significantly lower after pancreatectomy than at the baseline and especially at 3 and 5 min (respectively 24 ± 3.2 μIU/ mL after vs. 45 ± 5.1 μIU/mL before and 21 ± 2.8 μIU/mL after vs. 43 ± 6.1 μIU/mL, p < .05) ( Figure 3F). A significant interaction between "time" and "pancreatectomy" was thus noticed (p < .0001). As a result, following pancreatectomy, the Acute Insulin Response was significantly decreased (18.3 ± 10.0 μIU/mL after vs. 34.9 ± 13.7 μIU/ mL before, p < .0005) ( Figure 3G). Finally, there was no significant change reported in fasting plasma lipid profile after subtotal pancreatectomy ( Figure 3H).

| No significant change in glucose metabolism following the combination of a subtotal pancreatectomy with a 2-month HFHSD in Göttingen-like minipigs
The metabolic phenotypic changes following a subtotal pancreatectomy and 2 months of HFHSD as an oral energy overload were then examined (Figure 4). Following the protocol, animals gained weight (26.3 ± 5.9 kg after, compared to 21.3 ± 3.6 kg before, p < .05).
Following this approach, no rise in fasting blood glucose ( Figure 4A) was observed. With a more pronounced peak at 30 min and a faster return to baseline following the procedure, postprandial blood glucose dynamics were different from before, even if not significantly ( Figure 4B). Although there was a trend to higher postprandial insulin levels ( Figure 4C), the Insulinogenic Index did not significantly change ( Figure 4D). Additionally, the Hepatic Insulin Resistance Index modestly but not significantly increased ( Figure 4E). The IVGTT revealed no significant changes in glucose tolerance ( Figure 4F), insulin levels ( Figure 4G), or Acute Insulin Response ( Figure 4H). Finally, the levels of fasting plasma lipids were globally increased after intervention

| Alterations of insulin secretion pattern and insulin resistance after long-term intraportal glucose and lipid infusions in Göttingenlike minipigs
In two groups of minipigs, one without prior pancreatectomy and the other following subtotal pancreatectomy, we infused long-term intraportal glucose and lipid ( Figure 5). Animals of each group gained a lit- 24.7 ± 5.1 μIU/mL before, and 55.3 ± 10.0 μIU/mL after infusions at 30 min vs. 37.1 ± 7.6 μIU/mL before; not significant) and a significant interaction between "time" and "infusions" was discovered (p < .005) ( Figure 5B). Blood glucose levels decreased during IVGTT more slowly than they did before protocol (159.0 ± 14.2 mg/dL after infusions vs. 70.4 ± 28.6 mg/dL before at 30 min; not significant) ( Figure 5C) and insulin levels globally decreased, with an exception at 30 min ( Figure 5D).
Following procedure, postprandial blood glucose levels in Group 2 fell globally ( Figure 5E), similar to Group 1 and a significant interaction between "time" and "intervention" was observed (p < .05). Plasma insulin levels rose for the first 30 min (76.7 ± 10.6 μIU/mL after protocol at 15 min vs. 25.5 ± 5.1 μIU/mL before, and 75.4 ± 19.0 μIU/mL after protocol at 30 min vs. 31.0 ± 6.5 μIU/mL before; not significant) and a significant interaction between "time" and "intervention" was reported (p < .0001) ( Figure 5F). IVGTT findings after protocol revealed a slower lowering of blood glucose ( Figure 5G) and especially lower insulin levels with a significant intervention observed between "time" and "intervention" (p < .05) ( Figure 5H). Finally, whether or not a subtotal pancreatectomy had been performed prior to the intraportal glucose and lipid infusion, no increase in fasting blood glucose was observed ( Figure 5I).
5.3 ± 2.3 before; not significant, and 12.6 ± 7.9 after vs. 3.8 ± 4.3 before; p < .05, respectively for Groups 1 and 2) ( Figure 5K). Additionally, both groups' Acute Insulin Responses reduced (28.7 ± 7.5 µIU/mL after vs. 38.6 ± 13.3 µIU/mL before; not significant, and 24.4 ± 13.7 µIU/mL after vs. 43.9 ± 14.5 µIU/mL before; p < .005 before, respectively for The results in the Ossabaw strain were unexpected. The HFHSD induced a response that was highly comparable to that of the Göttingen-like strain, with an early insulin secretion that appeared to be even more effective than the Göttingen one in the baseline state. However, Ossabaw minipigs have a reputation for being the strain that is most susceptible to metabolic syndrome. 25,52 In fact, they developed, in the "Ossabaw Georgia island" where they come from, a "thrifty genotype" that enabled them to easily store energy from low-nutritive substrates because of the severe selection pressure imposed by the dry climate of the Ossabaw island. Thus, it is claimed that Ossabaw minipigs serve as a natural model for reproducing the symptoms of type 2 diabetes and the metabolic syndrome, similar to those populations that are predisposed to these diseases naturally. 3 However, given that no fasting hyperglycemia could be generated only after a diet in previous research, [53][54][55] it would appear that the expression of their metabolic syndrome would be more focused on lipidic dysregulations than disorders of glucose metabolism. 56,57 Fasting lipid levels of Ossabaw minipigs were much greater than those of the Göttingen-like strain in our study, particularly in terms of cholesterol, which makes this strain well-suited for the investigation of hypercholesterolemia illnesses 58 but not for studies of diabetes. We continued the combination protocol, which included a subtotal pancreatectomy, followed by five more months of HFHSD, in two minipigs of this strain. These two minipigs showed no signs of metabolic change (data not shown), demonstrating that this strain is not susceptible to develop type 2 diabetes.
The decision to perform a pancreatectomy was made considering the highly variable and toxic effects of streptozotocin 59 and alloxan. 60 Additionally, a surgical pancreatic mass excision is easier to control than one caused by toxic chemicals, 42,46 which is why this way of generating an insulin deficiency was chosen. The partial pancreatectomy's subsequent impact on glucose metabolism was unexpectedly modest, with the only discernible change being a reduction in the acute insulin response, which is the first phase of insulin secretion. We also observed that following pancreatectomy, insulin release reached a plateau. Nevertheless, there was no change in insulin secretion throughout the oral glucose challenge. As previously   Conceptualization (lead); funding acquisition (equal); investigation (equal); methodology (lead); project administration (supporting); resources (lead); supervision (lead); writing -review and editing (lead).

ACK N OWLED G M ENTS
The authors would like to extend their sincere gratitude to the

FU N D I N G I N FO R M ATI O N
The authors warmly thank the European Genomic Institute for Diabetes (EGID) for the funding and its precious contribution for animal disposal and housing, the Institut National de la Santé et de la Recherche Médicale (Inserm) for the stipend "Poste d'Accueil" and the Agence Nationale de Recherche (ANR).

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.