J147 affects cognition and anxiety after surgery in Zucker rats

.


Introduction
Regardless of the type of surgery, patients may experience postsurgical complications.
In order to study the pathophysiology of POCD and subsequently investigate potential therapeutic interventions, we developed an animal model for POCD, closely mimicking the patient surgical processes [29][30][31].This model confirmed results from previous studies [19,[32][33][34], showing that surgery-induced local inflammatory responses could result in increased circulating inflammatory markers, reflected in the brain as neuroinflammation [29,30].Although anti-inflammatory interventions seem to be beneficial in prevention of POCD in experimental [23,[35][36][37] and clinical studies [38][39][40], none have reached clinical practice so far.In addition, recently we showed that treatment with the anti-inflammatory drug ibuprofen could improve cognitive outcome after surgery, but this effect seemed associated with increased rather than decreased neuroinflammation [30].Moreover, inhibition of peripheral inflammation may not be sufficient to recover cognitive impairment in POCD [41].Therefore, successful treatment for POCD should address the pathological consequences of neuroinflammation, combined with other routes of neuroprotection, to improve cognition outcomes.
In this regard, J147 could be a promising candidate as it was shown to be neuroprotective in the central nervous system in several common neurodegeneration models [42][43][44][45][46]. J147, initially developed as anti--Alzheimer's disease drug, displays general anti-dementia [45,47], anti-diabetic [43] and anti-depressive effects [48][49][50].J147 has a mechanism of action with broad effects, as it can improve cognitive performance and dendritic structure, promote cell survival, reduce inflammation and nerve cell death due to the loss of trophic support, and prevent oxidative stress and leakage of the BBB, [42,43,[45][46][47]51,52]. In addition, J147 is a high affinity molecular target for α-subunit of ATP synthase (ATP5A) in the mitochondria, regulating a variety of metabolic signaling pathways, such as oxidative phosphorylation (OXPHOS) [45,53].Moreover, it was shown to exert cerebral cytoprotective effects in an acute ischemic model [54] and is currently in Phase I clinical trials (NCT03838185).J147 may therefore provide a promising new therapeutic avenue in prevention of POCD.Our study in healthy young Wistar rats confirmed that J147 could prevent cognitive impairment following abdominal surgery [55].Although these results may be relevant in the POCD treatment research, these young healthy rats do not optimally represent the patients at risk for developing POCD.This patient group consists of mainly older patients with impaired pre-operative health, including underlying diseases such as hypertension, hyperglycemia, impaired kidney function and liver disease; all known risk factors for POCD [7,[56][57][58].In agreement with this, we previously showed that advanced age and presurgical comorbidities were reflected in more wide-spread cognitive decline and exaggerated neuroinflammation [6,59].However, our previous study on ibuprofen treatment indicated that effects on behavior and (neuro)inflammation in young rats do not necessarily predict effects in aged rats [30].Moreover, commercially obtained aged rats have grown older under pathogen free, optimal food and climate conditions, which is usually not the case for the patients at risk for POCD.Therefore, a model that better represents the frail older patient population undergoing surgery is needed.The Zucker rats may serve as good candidates, as they display several risk factors for POCD, including obesity, hypertension, hyperglycemia, low-grade (neuro) inflammation and mild cognitive dysfunction [60].Moreover, in our previous study with Zucker rats [61], J147 treatment showed positive effects on the early recovery after surgery, reflected in shorter time to regain activity and restored diurnality.
Aim of the present study was to study the development of POCD and explore therapeutic potential of J147 to prevent or alleviate POCD in Zucker rats, and elucidate the underlying mechanisms.The effects of both acute and chronic J147 treatment on neuroinflammation, cognitive function and metabolism were investigated after abdominal surgery.

Animals and housing
Since in patients no difference in prevalence of POCD was observed between men and women [62], and to facilitate comparison with our previous studies [29,55,61], in the present study only male rats were used.A total of 83 rats; 69 12-weeks old obese Zucker rats (Crl:ZUC (Orl)-lepri) and 14 12-weeks old Lean littermates, were obtained from Charles River Laboratories (Wilmington, Massachusetts, USA), and aged in our facility up to 25-30 weeks.Although the rats at 25-30 weeks of age may not represent aged rats, older rats could not be used as mortality steeply increases after this age [63].Until the start of the study, rats were group-housed with 2 or 3 rats per cage on reversed light:dark cycle 12:12 with the lights out at 09:00 am, under controlled conditions (temperature of 20+/-2 • and humidity of 50 % +/-10 %).The rats had food (Teklad 2018, Envigo) and water ad libitum.After surgery, rats were housed individually, in order to prevent damage to the permanent intravenous catheter.Approval of all experiments was given by the Centrale Commissie Dierproeven (CCD) of the Netherlands, and the local experimental animal committee (IvD) of the University Groningen, the Netherlands.All experiments were performed in accordance with relevant guidelines and regulations and the reporting in this manuscript follows the recommendations in the ARRIVE guidelines

Experimental protocol
The present study is a follow-up of our previous study on early postsurgical recovery [61].The experimental protocol is illustrated in Fig. 1.At the start of the experiment, one week before surgery, Zucker rats were randomly divided over four experimental groups: non-surgery control; surgery control; surgery + acute J147 and surgery + chronic J147 treatment.Non-surgery lean littermates served as control for the Zucker rat model.
To characterize the Zucker rat as potential high-risk model for developing POCD, presence of risk factors was evaluated by comparing non surgery control Zucker rats to their (non-surgery) lean littermates.Surgery Zucker rats were subjected to abdominal surgery according to Hovens et al. [29] and allowed to recover, and effects of acute and chronic treatment with J147 on postoperative cognitive (dys)function were investigated.Chronic J147 treatment, food plus J147, started one week before surgery and continued until the end of the experiment.Acute J147 treatment and its vehicle control were administered as intravenous injection at the start of surgery.Half of the Zucker control surgery group received a vehicle-only injection and served as the control group for the acute treatment; the other half remained on regular food (no J147) and served as control for the chronic treatment.After surgery, rats were housed individually in cages equipped with an activity tracker (Circadian Activity Monitor System, CAMS, by H.M. Cooper, J.A. Cooper, INSERM U846, Department of Chronobiology, Bron, France).Non-surgery control rats were transported to the surgery room, left there for about an hour, then were placed in individual cages before returning to the housing facility, together with the surgery animals.After surgery, body weight and food intake were obtained daily.At day 5, a subgroup of rats (control surgery and chronically J147 treated) were subjected to a glucose tolerance test.Day 7-11 post-surgery, rats were subjected to behavioral testing regarding exploratory behavior and cognitive performance.Day 14 rats were sacrificed.Blood samples collected from cardiac puncture and urine samples collected from the bladder, were stored at -80 • C for later analyses.Half of the brain was collected and processed for immunohistochemistry and the other half for molecular analyses.

Surgery
Abdominal surgery, with ischemia reperfusion, was performed as described in detail previously [29,30].Briefly, rats were anaesthetized using sevoflurane (±2,5 % in air/O 2 = 2/1) and received 0,01 mg/kg buprenorphine s.c. as analgesia.Rats were equipped with an indwelling jugular vein catheter to mimic insertion of a venous line in patients.At this moment acute J147 or vehicle was injected intravenously.Subsequently, abdominal surgery was performed, consisting of exteriorizing the intestines, and temporal clamping (30 min) of the upper mesenteric artery.Clamping and reperfusion were visually verified by absence or presence of pulsation in the mesenteric artery distal to the clamp site.After removal of the mesenteric artery clamp, the intestines were placed back and the abdominal wall and skin were closed separately by sutures.From induction to recovery, the procedure took 60-75 min.The animals were allowed to recover until they were fully awake, before returning to the housing facility.

Acute J147 treatment
Effects of acute J147 were studied after single i.v.injection of J147 at the start of surgery, in order to interfere with the acute inflammatory responses without hampering the subsequent wound healing processes.For acute treatment, crystalline J147 was dissolved in 15 % solutol/85 % saline (10 mg/ml J147) and kept at 37 • C until administration (1 ml/ kg).Vehicle treated rats received 1 ml/kg solvent (15 % solutol/85 % saline), at 37 • C. A fresh solution was prepared each day.Blood samples obtained in our previous study indicated that this procedure resulted in therapeutic plasma levels of J147 at the time of surgery [61].

Chronic J147 treatment
For chronic oral treatment, J147 was added to the food (Teklad 2018, Envigo) at a concentration of 833 mg/kg, aiming at approximately 30 mg/kg/day J147, based on an estimated food consumption of 21 g/ day and body weight of 750 g [64].Chronically treated rats received J147 food one week before surgery until the end of the experiment.J147 was provided by Abrexa Pharmaceuticals, Inc. Envigo prepared the food, Teklad 2018, with and without J147.J147 food was stored at -20 ⁰C (dark).Blood samples obtained in our previous study indicated therapeutic plasma levels of J147 at the time of surgery [61].At sacrifice, a separate blood sample collected from chronically treated rats (n = 10), and processed according to Oberman et al., [61], revealed a steady state plasma level of J147 of 19.8 ng/ml (QPS, the Netherlands).

Behavioral tests
Behavioral tests were performed in the dark (active) period of the rats, starting 7 days after surgery.This time course was chosen to match with the clinical setting of POCD diagnosed one month after surgery [65], referring to the shorter lifespan of rats.The lights went on at 9:00 p.m. and off at 9:00 a.m.Rats were allowed to eat the first hour of the dark period, as testing did not include the first hour and the last hour of the dark phase.General exploratory behavior and anxiety was tested in the Open field (OF) test; long-term spatial learning, spatial memory and cognitive flexibility in the Morris Water Maze (MWM); short-term spatial working memory in the novel location recognition test (NLR); the novel object recognition test (NOR) for short-term working memory on object recognition.

Open field
The Open Field (OF) test was performed to assess exploratory and anxiety-like behavior.The box consisted of a square (100 × 100 × 40 cm) divided into areas: the center area (60 × 60 cm), 4 side areas (20 × 60 cm) and 4 corner areas (20 × 20 cm).The rats were placed in the middle of the box and the trial started when the rat crossed one of the lines defining the center area.After five minutes the rat was gently removed from the box and returned to his cage.Before each rat, the box was cleaned with 70 % ethanol and allowed to dry.The movement as well as location of the rat were recorded as time in corner, walls and center, number of visits in the center and the distance walked, and analyzed with Ethovision (Noldus, the Netherlands).Exploratory behavior, as interest in environment [66], was measured as the distance the rat moved during the test, while depressive/anxiety-like behavior was obtained from the choice for the relative save area (corner and wall), as reciprocal of the percentage of time spent in center area and number of center visits.

Novel object and novel location recognition
The Novel Object (NOR) and Novel Location Recognition (NLR) were tested in random order in one procedure to assess spatial short-term memory and object memory, as described before [29].For this test, an open square box (50 × 50 × 40 cm) with 2 transparent and two grey walls was used.Objects used in the test were selected from a pilot study, investigating the preferences of rats for different objects.Rats were habituated to the box for 3 min, the day before the test.The test consisted of four phases; each phase took 3 min.Between each phase there was a break of 45 s in which the rat remained in the test box.The first phase was performed in an empty box.In the second phase, two identical objects were placed in both corners at the nearest end of the box.In this phase, the rat was allowed to explore both objects.After this exploration phase, either NOR test or NLR test was performed.Before the start of each phase, both objects were removed from the box and cleaned with 70 % ethanol and dried before use.During the NOR test one of the familiar objects was replaced by a new object and during the NLR test one of the familiar objects was relocated to the top of the box.After each test, the box was cleaned with 70 % ethanol and allowed to dry to remove smell cues before entering of the next animal.The tests were manually analyzed with Eline (University of Groningen).Baseline exploratory behavior was calculated by the time the rat spent exploring both of the objects at phase 2. Results from rats that showed no interest in either object, indicated by no exploration at either object, were omitted from analyses.Novel object preference and novel location preference were calculated as time spent on exploration of the novel or relocated object divided by time spent on exploration of both objects.

Morris water maze (MWM)
The Morris Water Maze consisted in total of six training sessions, two probe trials and three reversed training sessions.The tests assessed spatial learning, spatial memory and cognitive flexibility, respectively [29,67].The maze entailed a round pool (diameter 140 cm) with an invisible platform, submerged 1-2 cm below water level.The water had a temperature of 26 +/-1 • during the sessions.Visual cues surrounded the pool and divided it into four virtual quadrants.The platform was placed in the target quadrant during the training sessions, removed from the pool during the probe trials and relocated to the opposite quadrant during the reversed training sessions.The first day consisted of three training sessions.Each training session consisted of three trials for each rat during which the rats were placed in random order in each quadrant (except the target quadrant).The rats were allowed to find the platform within 60 s and left on it for 10 s.When the rat did not find the platform within 60 s, he was gently guided to the platform and left there for 10 s.After the 3 trials the rat was towel dried and placed back in his cage.The period between the training sessions of each rat was at least 2 h.The second day started with a probe trial to assess early spatial memory.During the probe trial, the rat was placed in the opposing quadrant of the target quadrant, while the platform was removed.The rat was allowed to swim for 60 s and was taken out of the water in the quadrant where the platform was supposed to be.The number of times the rat crossed the place of the platform and the time the rat spent in the target quadrant, were recorded.After the probe trial the day continued with three more training sessions starting one hour after the probe trial.At the third day the second probe trial was performed following the same procedure as with probe trial 1.After this probe trial three reversed training sessions were done consisting of three trials each session with the platform placed in the opposing quadrant to assess cognitive flexibility.For the MWM, the following parameters were calculated: spatial learning: per session the average latency to find the platform to create a learning curve; for the probe trials, the number of platform crossings and percentage in the target quadrant; memory consolidation: as the difference in latency to find the platform in the last learning session and in the first reversal session; cognitive flexibility: as the learning curve in the reversal trial.

Activity
From the collected locomotion data, actograms were made, which were further analyzed regarding aspects of rhythmicity.Total activity was calculated per 24 h (12 h light and 12 h dark period) for postoperative day 13.Diurnality was calculated by subtracting the activity during the dark period from the activity during the light period, divided by total activity during 24 h.Diurnality assesses the shift in activity in the active (dark) period to the light (rest) period; lower values refer to more light activity for the nocturnal rats.All parameters were analyzed using ACTOVIEW for Excel 2010, programmed by C. Mulder, University of Groningen [68].

Sacrifice and tissue collection
Three days after behavioral testing, rats were anesthetized with pentobarbital (90 mg/kg).The heart was punctured and a blood sample was collected.Part of the blood sample was directly used for hematocrit determination, while the rest was centrifuged for 10 min at 2600 G, and plasma was collected and stored at -80 • C until further analysis.Urine samples were collected directly from the bladder.Samples were stored at -80 • C until further analysis.The rats were sacrificed by transcardiac perfusion with saline containing Heparin.Randomly regarding hemisphere, one half of the brain was fixated in paraformaldehyde (PFA, 4 %) and processed for immunohistochemical analyses; the other half was quickly frozen in liquid nitrogen and stored at -80 • for later molecular analyses.

Hematocrit
A small part of the blood was collected in a capillary tube with a plug at the end and centrifuged at 800 rpm for 8 min for the hematocrit determination.The red blood cell part were measured and haematocrit was calculated by (red blood cells/total blood)*100.

Plasma levels of inflammatory markers
Plasma levels of inflammatory markers (Il-1beta, IL-6, IL-10, TNFalpha and VGEF) were measured by multiplex (Bio-Techne, Luminex).However, since it appeared not possible to obtain relevant results with this multiplex, only IL-1beta (Invitrogen) and lipocalin 2 levels (Bioporto), reflecting the most relevant inflammatory markers [13,29,69], were obtained by separate ELISA kits.Analyses were performed according to manufacturer's instructions.

Plasma triglycerides
Zucker rats are characterized by high plasma triglycerides [70][71][72], therefore we also measured triglycerides at sacrifice.For the triglycerides measurements Roche/Hitachi GRO-PAP reagents were added to each sample (cat nr: 11730711216) according to manufacturer's instructions and absorbance was measured at 473 nm.

Glucose tolerance test
Since Zucker rats were reported to exert insulin resistance [41], effects on glucose tolerance [73] were obtained, at day 5 after surgery.Rats were fasted 4 h before the glucose tolerance test, from 9:00 until 13:00.For logistic reasons measurement were only performed in subgroups of control surgery and surgery with chronic J147 treatment rats.For that, glucose (500 mg/kg body weight) was administered via the jugular vein canula and blood samples (200 µl) were collected at 0, 5, 10, 15, and 30 min.Samples were split into 50 µl blood, stored at -20 • C for later measurement of glucose, and 150 µl, centrifuged at 2600 G for min and plasma stored at -20 • C until measurement of insulin levels.Glucose levels were measured using an autoanalyzer, facilitating a color shifting redox reaction with potassium ferricyanide (K 3 [Fe(CN) 6 ]) and glucose (C 6 H 12 O 6 ).The C 6 H 12 O 6 is diluted with 0,3 M NaCl, dialyzed with K 3 [Fe(CN) 6 ] and heated to 95 • C. The yellow color decreases as the glucose concentration in the sample increases.The absorption is measured at 420 nm.
The insulin levels in the plasma were measured using a rat Insulin RIA kit (Cat nr.#RI-13 K, Sigma Aldrich, Darmstadt, Germany).The analysis was executed following the manufacturer's protocol.The concentration of insulin was measured using a gamma counter which counts the radioactivity of radio-bound insulin in the solution.In addition to the time course of glucose and insulin levels, the ratio between these was calculated as measure for insulin sensitivity.

Proteinuria
As proteinuria had been reported in the Zucker rat model [74,75], and could be indicated a risk factor for POCD (Cavaliere et al., 2016), the effects of surgery and of J147 were measured as protein level in urine.Urinary protein levels were measured by Pyrogallol Red Method Microprotein kit (Euro Diagnostic Systems Pvt Ltd, India) according to manufacturer's instructions.

Tissue preparation
Brains were post-fixated in 4 % PFA in 0.1 M PB for 41-46 h at room temperature.After the post-fixation brains were washed out from 4 % PFA with 0.01 M PB during +/-48 h on a shaker, 100 rpm at room temperature.Brains were washed 8 times within the 48 h.After this, the brains were placed on 30 % sucrose in 0.01 M PB overnight at room temperature.When the brains had sunken, they are washed with MiliQ water and petted dry.Thereafter the brains were frozen in liquid nitrogen and stored at -80 • C until further analysis.
For immunohistochemical staining, brains were cut into 24 µm thick sections from the prefrontal cortex to the ventral hippocampus and stored free floating in 0,01 M PBS + 0,1 % Natrium Azide at 4 • C. Before staining, the free-floating sections were pretreated with 0,3 % H202 for 30 min.All sections were thoroughly rinsed with 0.01 M PBS between staining steps.Sections were transferred to glass slides and dehydrated through gradients of ethanol and xylol solutions.Labeling was analyzed blinded to the treatment.Slides were scanned at the department of Pathology, University Medical Center Groningen, the Netherlands.Image analysis software (Image-Pro-Plus 6.0) was used to analyze the photographs.

Microglia
To analyze neuroinflammation microglia were visualized [29,76].Sections were incubated for 3 days with 1:2500 rabbit-anti IBA-1 in 1 % BSA, 0,1 % TX at 4 • C. Followed by 2 h incubation with 1:500 goat-anti rabbit secondary antibody at room temperature.All sections were incubated for 1 h with avidin-biotin peroxidase complex (Vectastain ABCkit, Vector, Burlingame, USA) at room temperature.Labeling was visualized using a 0.075 mg/ml DAB solution activated with 0.1 % H2O2.All dilutions were made in 0.01 M PBS, except for the DAB that was made in MilliQ.Microglia activation was determined in the Hilus and CA1 region of the hippocampus.The number of microglia, the average total cell size and average cell body size were determined (400X magnification).Since microglia activation is characterized by an increased cell body size and retraction of the dendritic processes (Kreutzberg, 1996), the cell body to total cell size ratio was used as a measure of microglia activation, and calculated according to Hovens et al. [29,76].

Neurogenesis
To visualize newly formed neurons, sections were stained for double cortin (DCX) as described before [29].Sections were incubated for 3 days with 1:1000 goat-anti DCX.(Santa Cruz, Dallas, USA) in 3 % normal rabbit serum, 0,1 % TX at 4 • C, followed by incubation with 1:500 rabbit-anti goat secondary antibody (Jackson, Wet Grove, USA) for 2 h at room temperature.Sections were incubated for 2 h with avidin-biotin peroxidase complex (Vectastain ABCkit, Vector, Burlingame, USA) at room temperature.Labeling was visualized using a 0.075 mg/ml DAB solution enhanced with ammonium nickel sulfate solution and activated with 0.1 % H2O2.All dilutions were made in 0.01 M PBS, except for the DAB which was made in MilliQ.The area of DCX positive cells per length of the dentate gyrus of the hippocampus was considered a measure for neurogenesis.

Blood brain barrier protein
To visualize blood brain barrier (BBB) function, sections were stained for endothelial barrier antigen (EBA) [77,78].Sections were incubated overnight with 1:1500 mouse-anti EBA (BioLegend, San Diego, USA) in 3 % normal horse serum, 0.5 % Triton-X-100 and 0.01 M PBS followed by and incubation of 1 hour with 1:1500 goat-anti mouse (Jackson, Wet Grove, USA) in 3 % normal horse serum, 0.5 % Triton-X-100 and 0.01 M PBS.Thereafter, all sections were incubated for 1 hour with avidin-biotin peroxidase complex (Vectastain ABC kit, Vector, Burlingame, USA) at room temperature.Labeling was visualized using a 0.035 mg/ml DAB solution activated with 0.1 % H 2 O 2 .All dilutions were made in 0.01 M PBS, except for the DAB that was made in MilliQ.EBA staining was analyzed by Image J (Image J 1.53k, National Institutes of Health, Bethesda, Maryland, USA), as the coverage of EBA positive area in the whole hippocampus.

Blood brain barrier leakage
BBB leakage was obtained from presence of Immunoglobulin G (IgG), a protein which presence is normally restricted to plasma [79].To visualize IgG in the brain, sections were incubated overnight with 1:1000 donkey-anti rat antibody in 3 % BSA (Jackson ImmunoResearch Europe Ltd, Cambridgeshire, UK), followed by 1 h incubation with avidin-biotin peroxidase complex (Vectastain ABCkit, Vector, Burlingame, USA) at room temperature.Labeling was visualized using a 0.035 mg/ml DAB solution activated with 0.1 % H 2 O 2 .All dilutions were made in 0.01 M PBS, except for the DAB that was made in MilliQ.IgG staining was analyzed by Image J (Image J 1.53k, National Institutes of Health, Bethesda, Maryland, USA), as coverage was determined in the hippocampus.

Oxidative phosphorylation
As a measure for oxidative phosphorylation, expression of the 5 mitochondrial electron transport chain complexes was obtained [80], by Western blotting of hippocampus and prefrontal cortex tissue.Brain tissue of 6 randomly chosen rats of each group, was homogenized in ice-cold 50 mM tris HCL buffer at pH 7.4, and then the tissues were bead-beaded in a Tissue Lyser for 2 times 1 min.Thereafter, homogenates were sonicated for 2 times 2 s.Homogenates were centrifuged for 10 min at 1500 rpm and supernatant was collected and stored at -80 • C. Protein concentration was determined by a BCA assay (Thermo Scientific).Subsequently, western blotting was performed.Equal amounts of protein from each sample were loaded on a 10 % polyacrylamide SDS-gels.Coomassie Brilliant Blue staining was used to determine the total protein content of every sample.The ND6 subunit of complex I, the Ip subunit of complex II, the core protein 2 of complex III, subunit II of cytochrome c oxidase (complex IV) and the α subunit of the F1F0 ATP synthase (complex V) were measured using monoclonal antibodies (1:1000, MitoSciences, OR, USA).Proteins were visualized by Clar-ityWestern ECL Substrate (Bi0Rad #170-5061).All proteins were expressed as arbitrary units (AU) and corrected for total protein.The sum of the expression of the 5 proteins was used as an measure for total mitochondria [80].

Data analyses
Statistical analyses were performed using SPSS (IBM SPSS Statistics, Version 27, Armonk, NY).For this study we performed 4 separate statistical analyses to be able to answer the 4 sub questions regarding the Zucker strain, surgery and J147 treatment.Firstly, results of non-surgery Zucker rats and their (non-surgery) lean littermates were compared in order to evaluate the anticipated risk factors for POCD, related to this Zucker strain, by an independent sample t-test.Since half of the Zucker control surgery group received vehicle injection and served as control for acute treatment, while the other half served as food control for the chronic treatment, these groups were compared with an independent sample t-test to detect potential effects of vehicle treatment.Secondly, non-surgery and control surgery Zucker rats were compared by an independent sample t-test in order to examine effects of surgery.Thirdly, Zucker rat groups that underwent surgery, with and without active J147 treatment, were compared using one way ANOVA followed by post-hoc Dunnett's correction, in order to evaluate effects of treatment.Finally, Zucker rats were compared using one way Analysis of Variances (ANOVA) followed by post-hoc Dunnett's correction with Zucker nonsurgery as control group, to evaluated whether treated rats would return to values of non-surgery rats.Results that exceeded twice the standard deviation of its group were regarded as outlier and excluded from analyses of comparing means.Results from timed measurements were analyzed with ANOVA for repeated measures.Since the focus of this study was on POCD, correlations were made between behavior and other parameters, to obtain more insight in underlying mechanisms.For correlation analyses, lean rats were excluded.Correlations between parameters were obtained by a Pearson correlation.Data are presented as mean ± SEM.Differences were regarded statistically significant when p ≤ 0.05.Relevant trends were indicated by p ≤ 0.10.

General
From the 83 rats started with, 6 rats died spontaneously or had to be sacrificed prematurely before entering the study (1 lean; 5 Zucker rats), 2 rats died during the study and 1 had to be sacrificed shortly after surgery, because of losing too much body weight and showing signs of lethargy.In total 74 rats were subjected to the behavioral test.Since no significant effects of vehicle treatment were observed, control groups were pooled, resulting in the following experimental groups: 1. Nonsurgery Lean (n = 13); 2. Non-surgery Zucker (n = 14); 3. Zucker surgery control (n = 15), 4. Zucker surgery + acute J147 (n = 16) and 5. Zucker surgery + chronic J147 (n = 16).
General characterization of experimental groups is presented in Table 1.Obesity in the Zucker rats was confirmed by a significantly higher body weight compared to lean rats (t = -9584, p < 0,001).Surgery in Zucker rats significantly reduced body weight (t = -7497, p < 0,001), which was recovered 14 days later.

Behavioral test
The open field test was used to examine general exploratory behavior and anxiety.In the distance moved in the OF test, as measure for exploratory behavior, Zucker rats moved significantly less than lean rats (t = 2313, p = 0,030), which was further exaggerated after surgery (Fig. 2B).No effects of treatment were observed.Whereas no differences in anxiety measures (time in center area) between lean and Zucker rats were found, surgery significantly reduced time in the center (t = 2415, p = 0,026) (Fig. 2A).Chronic-, but not acute J147 treatment significantly reversed this surgery effect (F = 4292, p = 0,048).Effects on time in the center area, were supported by effects on the number of center visits.
Short-term memory was tested in the NOR and NLR tasks (Fig. 2C  and D).No differences were observed between the groups in performance in the NOR (Fig. 2C); all rats performed above chance level (50 %).However, in the spatial memory associated NLR test, Zucker surgery rats tended to perform worse than the Zucker non-surgery rats (p < 0.1).Whereas acutely J147 treated rats even seemed to underperform compared to control surgery rats, chronic J147 treated animals showed a trend to improvement (p < 0.1) (Fig. 2D).
In the long-term spatial learning and memory test (Fig. 2E and F), in the MWM, a significant strain-related impaired performance in Zucker rats was indicated by longer latency in learning (t = 3346, p = 0,003) (Fig. 2E).Neither surgery, nor J147 treatment affected that.Whereas overall long-term spatial memory development was not different between groups (probe trial 2), Zucker rats seemed to develop their memory slower, since they do not seem to remember the position of the platform yet in probe trial 1, while they do in probe trial 2. A similar effect was observed in the control surgery group.Surprisingly, surgery rats that were acutely treated with J147 at the time of surgery, performed better two weeks later; these rats already knew the platform position in probe trial 1, and tended to outperform control surgery rats in probe trail 2 (p < 0,1).The chronically J147 treated rats only remembered where the platform was in probe trail 2. Although AUC of the learning curve in the reversal trial indicated less cognitive flexibility in Zucker rats compared to lean (AUC 73 ± 7 in Zucker versus 53 ± 8 s in lean rats; p = 0.074), no significant effects of surgery or treatment were observed in the reversal trial performance of Zucker rats.Moreover, no

Effects on metabolism 3.3.1. Activity
Fig. 3 shows the 24 h-activity, as well as the diurnality at postoperative day 13.Although Zucker rats had on average 25 % lower total daily activity, there was no significant difference between groups.Diurnality reflects the difference between light and dark activity of the night-active rats.Diurnality did not appear different between lean and Zucker rats, but was still affected by surgery after 13 days.Zucker control surgery rats had a lower diurnality score compared to Zucker non-surgery rats (t = -2175, p = 0,040), reflecting relatively more activity during the light period.In contrast to chronically treated rats, diurnality of acutely J147 treated Zucker rats did not statistically differ from Zucker non-surgery rats anymore.

Insulin sensitivity
Insulin sensitivity was measured with a glucose tolerance test.Plasma glucose levels increased similarly in Zucker surgery control rats and chronic J147 treated Zucker rats upon injection of glucose (500 mg/ kg body weight) (Fig. 4A).However, the insulin response was significantly lower in the chronic J147 treated rats compared to the Zucker surgery control group (F = 8350, p = 0,010) (Fig. 4B).As less insulin is made upon the same glucose response in Zucker chronic J147 treated rats, the ratio of insulin to glucose was significant lower in the Zucker chronic J147 treated rats (F = 12,801, p = 0,003), indicating lower insulin resistance after chronic J147 treatment (Fig. 4C).

Mitochondrial OXPHOS protein
As indication of mitochondrial function, expression of oxidative phosphorylation complexes (OXPHOS complex I -V proteins) was measured in the hippocampus and prefrontal cortex and data is displayed in Table 2 and Fig. 6.Although effects did not reach statistical significance, a consistent pattern is visible for all 5 complexes; Zucker rats displayed reduced expression of all proteins compared to lean, surgery increased levels back to lean values, without effect of treatment (Table 2).Accordingly, the sum of the proteins of complex I -V, as estimate for mitochondrial density, showed a similar pattern in the hippocampus, but a tendency for lower density in Zucker compared to lean and higher expression after surgery in the PFC (Fig. 6B and C, P < 0,1).

Proteinuria
Kidney function was examined by total protein concentration in urine on postoperative day 14.In all Zucker rat groups, a dichotomous response was visible, either protein levels appeared far above the detection level of the essay (800 mg/dl), or mostly below 500 mg/dl (Fig. 7A).Data for the low protein group is plotted in Fig. 7B.Whereas no difference was observed between lean and Zucker rats, surgery in Zucker rats significantly increased urinary protein level (t = -2775, p = 0,011), which was normalized by chronic (F = 5456, p = 0,003), but not by acute J147 treatment (Fig. 7B).When taken into account all Zucker rats (above and below 800 mg/dl) a significantly negative correlation between kidney weight and proteinuria was observed (r = -0.050,p = 0.050).A similar correlation was observed when only values below 800 mg/dl were included (r = -0.31,p = 0.056): the lower the kidney weight, the higher the proteinuria.

Effects on (neuro)inflammation 3.4.1. Plasma markers
Effects on systemic inflammation ware obtained from circulating levels of IL-1beta and NGAL.All Zucker rat groups had elevated plasma IL-1beta level at sacrifice, without effects of surgery or treatment (Fig. 8A).NGAL plasma levels were also elevated in the Zucker rats compared to lean littermates (t = -3919, p = 0,001) (Fig. 8B).However, in contrast to the plasma IL-1beta, NGAL was further elevated in the surgery groups 14 days after surgery.Acute J147 even further increased NGAL levels compared to surgery control rats (F = 5284, p = 0,023), with 6 rats showing values far above the detection level of the ELISA.In contrast, chronic J147 treated rats showed a slight decrease in plasma NGAL, being not significantly different from Zucker non-surgery rats anymore.

Neurogenesis and neuroinflammation
Neurogenesis was measured as the double cortin (DCX) positive area per length of the dentate gyrus (DG).No differences were found among the Zucker and lean groups in DCX.However, neurogenesis seemed to be lowered after surgery and partially preserved in the acute J147 treated group, however this did not reach statistical significance (Table 3).

Fig. 4.
Outcomes of the glucose tolerance test for Zucker control surgery (n = 7-8) and Zucker surgery + chronic J147 treatment (n = 12-15).A) plasma glucose levels after injection of 500 mg/kg rat glucose, B) Plasma insulin levels after injection of 500 mg/kg rat glucose, and C) Ratio plasma insulin/glucose after injection of 500 mg/kg rat glucose.*: significant difference between indicated groups (p < 0.05).
Microglia activity, as measure for neuroinflammation, was obtained in the CA1 and Hilus area of the hippocampus.Results are shown in Fig. 9. Whereas no differences in microglia activity were observed in the Hilus area of the hippocampus (Fig. 9C), a significant increase in microglia activity was seen after surgery for all groups in the CA1 area (F = 8452, p < 0.001) (Fig. 9D).Neither acute-, nor chronic treatment with J147, altered this neuroinflammation parameter.Results are supported by the underlying parameters (Table 3).The surgery-associated increased microglia activity seemed mainly due to a lower dendrite area, with preserved cell body size.

Blood brain barrier
To study effects on BBB function, leakage of IgG into the brain (hippocampus) and blood brain barrier EBA staining were performed (Fig. 10).No significant difference in leakage was observed between Zucker rats and their lean controls.All surgery rats had a significantly higher leakage of IgG into the whole hippocampus (F = 11,605, p < 0,001) (Fig. 10C).EBA staining showed a tendency to lower values of after surgery.However, acute J147 treatment had a more pronounced protective effect on the EBA in de blood brain barrier (F = 2465, p = 0,046) (Fig. 10D).

Correlations
To unravel the potential mechanisms underlying the effects on behavior, regression analyses were performed.Markers of peripheral inflammation (IL1-beta and NGAL) appeared not correlated to any behavioral parameter.Similarly, no correlations were observed for proteinuria or insulin sensitivity with any behavioral parameter, and long-term spatial learning seemed not correlated to any measured parameter.Results of significant correlations are shown in Table 4. Whereas for short-term memory not many correlates were observed, long-term spatial memory seemed better with higher brain/hippocampal weight, and more neurogenesis.Interestingly, mitochondrial density appeared to provide a significant correlate for short-term memory (NOR), long-term memory, and cognitive flexibility.However, cognitive performance in Zucker rats did not seem to be correlated to (neuro)inflammatory responses.

General
The primary aim of this current study was to explore the Zucker rat model, presenting potential risk factors for post-operative dysfunction (POCD), and the effects of acute and chronic J147 treatment, regarding post-surgical cognitive function and (neuro)inflammation.Additionally, in order to elucidate on underlying mechanisms, effects on activity, daynight rhythm, anxiety/depressive-like behavior, neurogenesis, metabolism, and blood brain barrier function were obtained.Results showed that Zucker rats indeed expressed the anticipated risk factors for POCD development, including impaired cognitive performance, obesity, high plasma triglycerides, low-grade systemic inflammation and decreased neurogenesis, when compared to their lean littermates.However, major abdominal surgery in this model seemed to induce post-operative depression (POD) rather than POCD, with concomitantly significant (neuro)inflammation.Moreover, surgery impacted these rats by a disturbed day/night rhythm (diurnality), persistent weight loss, proteinuria and increased blood brain barrier (BBB) leakage.The lack of overt POCD in the Zucker rats limited the opportunity for therapeutic      improvement with J147.Nevertheless, whereas acute J147 positively affected long-term spatial memory, diurnality and blood brain barrier integrity, chronic J147 improved anxiety, short term spatial memory, glucose sensitivity and proteinuria.However, no significant effects of either J147 treatment were seen on (neuro)inflammation.Therefore, we conclude that the Zucker rat indeed may present signs of frailty, that are reported in patients, which resulted in POD, rather than POCD.Still, acute as well as chronic J147 treatment affected aspects of cognitive performance, without reducing (neuro)inflammation.Since acute and chronic J147 therapy seemed to have different beneficial effects, a combination may be worthwhile to investigate.Additionally, further optimization of doses and timing of treatment, as well as underlying mechanisms, would be subject to further investigation.

The Zucker rat as model for POCD
The Zucker rat model was used to better represent frail patients at risk for POCD.Most of our previous work on the effects of surgery was performed in male Wistar rats, in which surgery induced a transient (neuro)inflammatory response, that was associated with reduced cognitive performance [29].Known risk factors for POCD, such as older age [59], type of surgery [81] or previous infections [82], led to more extensive (neuro)inflammation and cognitive decline in these Wistar rats.However, neither risk factor had led to persistent post-operative depression-like behavior in healthy young rats or aged rats.Since, the risk factors were expressed in otherwise relative healthy young rats, while patients at risk for POCD are usually not healthy, the Zucker rat may provide a clinically more relevant model.The Zucker rat is an outbred strain with a genetic defect in the leptin receptor, resulting in hyperglycemia, obesity, hypertension and renal dysfunction [60].Moreover, these Zucker rats display low-grade (neuro)inflammation and cognitive dysfunction at a relatively young age (<6 months) [60].Results from this current study were in agreement with this literature, as Zucker rats displayed impaired cognitive performance (long-term learning and memory), obesity, high plasma triglycerides, low-grade systemic inflammation and decreased neurogenesis.As these factors

Table 3
Neurogenesis and microglial parameters for the different experimental groups (mean ± SEM); lean (n = 12-13), Zucker non-surgery (n = 13-14); Zucker control surgery (n = 13-15); Zucker surgery + acute J147 treatment (n = 14-15); Zucker surgery + chronic J147 treatment (n = 13-15).*: significant effect of strain, #: significant effect of surgery (p < 0.05).mimic risk factors for POCD in patients, we anticipated these Zucker rats to be vulnerable to developing POCD.Despite the potential predispositions, the Zucker rats seemed to develop only mild cognitive decline after surgery.In our previous rat studies, bodyweight loss appeared to be a good predictor for cognitive dysfunction and neuroinflammation in increased risk conditions [59].Zucker rats lost significant weight after surgery [61], but this did not exceed what we had observed in young healthy rats [29,55].However, in contrast to our previous studies (not published) body weight had not recovered at 14 days after surgery.This is reflected in the plasma triglycerides, as well as its source (liver weight), which were still significantly lower 14 days after surgery.In accordance with the mild body weight loss as predictor for cognitive performance, Zucker rats displayed only mild POCD; a slightly decreased short-term spatial memory in the NLR test, without effects on long-term learning and memory in the Morris Water Maze.Surprisingly, effects of surgery in the Zucker rats could rather be associated with post-operative depression (POD), indicated by lower exploration and more time in the save areas of the OF test.Lower exploration may reflect lower interest in the environment, as one of the central symptoms of depression (DSM-V), while spending more time in the save areas would suggest more anxiety [66].Moreover, similar to what we observed in one of our previous studies [69], exploration in the OF test appeared correlated to hippocampal microglia activity.Therefore, we concluded that surgery in the Zucker rat may induce POD, rather than POCD.Surgery is a major stressor for the body, and can trigger the release of hormones and neurotransmitters that affect mood and emotions.Moreover, chronic (neuro)inflammation as seen in our study may be involved in the development of depression [83,84].POD can also have a severe impact on patients.POD may include symptoms, such as persistent feelings of sadness, loss of interest or pleasure in activities, changes in appetite or weight, sleep disturbances, fatigue,

Table 4
Correlations between indicated behavioural tests and other (neuro) inflammatory and metabolic parameters in Zucker rats.difficulty concentrating, feelings of worthlessness or guilt, and thoughts of death or suicide [2], and may even occur in up to 60 % of patients depending on type of surgery (Tegegne et al.,2022; [2]).However, as POD was not the main focus of this study, more research is needed towards surgery induced depressive-like behavior in Zucker rats, including more behavioral tests and a better reflection in brain areas involved, such as the basolateral amygdala.Zucker rats indeed developed an inflammatory response after surgery [61], which persisted at least until 14 days after surgery.However, IL-1beta, as cytokine with predictive value for POCD [13], was not affected by surgery, which would be consistent with the lack of overt POCD in the Zucker rats.Peripheral inflammation was indicated by higher plasma NGAL levels, which we previously suggested as a good predictor for long-term cognitive performance after surgery [69].This was reflected in neuroinflammation, as higher microglia activity in the hippocampal CA1 area.In line with the higher inflammatory response of the Zucker rats, increased blood brain barrier (BBB) leakage was observed, reflected in higher IgG and lower EBA in the hippocampus.These results indicated that surgery negatively affects the integrity of the BBB, which corresponds with literature showing that BBB dysfunction is linked to postoperative cognitive impairments [26,28].
Although surgery reduced activity as well as diurnality up to 5 days after surgery [61], diurnality was still affected 14 days after surgery, resulting in more day activity in the normally night-active rats.Clinical studies showed that surgery could induce disruptions in circadian rhythm, causing sleep disturbances and changes in the balance of day-time and nighttime activity [85][86][87][88][89].However, this relationship between activity and circadian sleep/wake rhythm is bi-directional, as less activity can also affect the circadian rhythm, and may be an indicator for worse postoperative outcome, such as poor recovery and cognitive problems [90,91].
Finally, the Zucker rat model is based on a disturbed leptin metabolism, as a result from a receptor deficiency leading to high levels of circulating leptin [92].Although major knowledge regarding the relationship between leptin and surgery comes from bariatric, rather than general surgery, there is evidence that leptin signaling plays a role in post-operative dysfunction.In multiple studies, higher circulating leptin levels haven been associated with reduced cognitive risk [93][94][95][96][97][98][99][100][101].Moreover, preoperatively high leptin levels were associated with lower development of delirium in patients after surgery [102] and administration of leptin can alleviate symptoms of AD in preclinical studies [103].Moreover, stimulation of leptin production was also shown to improve POCD symptoms in wild type mice, but not in leptin receptor-deficient (db/db) mice [104].
All in all, although Zucker rats may display clinically known risk factors for POCD, these rats did not develop substantial cognitive decline after surgery, but showed signs of POD, which coincided with neuroinflammation.Therefore, Zucker rats may not be the better representable model for POCD that we had anticipated.

Effects of J147
J147 is a promising new drug with proposed anti-aging and antidementia properties [42][43][44][45][46].In the present study, the effects of J147 after acute administration at the start of surgery, and chronic administration in food, starting one week before surgery were investigated in Zucker rats.Both administration routes resulted in plasma concentrations well within the therapeutic range of 3-70 ng/ml [105], at the time of surgery [61].Chronic treated J147 rats showed a stable blood plasma concentration of 11.9 ± 5,1 ng/ml at the time of surgery as well as at the end of the experiment (19.8 ng/ml).Acutely administered J147 yielded blood plasma levels of almost 450 ng/ml at 1 hour after injection, and declined to 50 ng/ml at 24 h [61].Our previous studies showed that J147 protected against long-term surgery associated cognitive decline in healthy young Wister rats [55], and improved early post-surgical recovery in Zucker rats [61].As POCD can affect different brain areas [11], different cognitive tests were included.Whereas the NLR represents short-term spatial memory and the MWM reflects long-term spatial learning, memory and cognitive flexibility, which are mainly associated with hippocampal (CA1) function [106], the short-term NOR rather is PFC function-associated [107].Since the Zucker rats did not seem to present major cognitive decline after surgery, there was little opportunity for improvement by J147.Nevertheless, there was a tendency for improvement of short-term spatial memory after chronic treatment, and a tendency for improvement-of long-term spatial memory after acute treatment.These positive effects of J147 on cognition are in agreement with previous studies that also showed these effects in other dementia models [45][46][47]108].However, our previous study in young healthy rats showed that chronic-but not acute, J147 improved both short-term and long-term spatial memory [55].Interestingly, the opposite short-term/long-term cognitive effects have been observed before in our own lab in an Alzheimer model [109] and other labs [110].Sanderson and Bannerman suggested that the opposite effects on cognitive performance can result from competitive processes that contribute to short-term and long-term memory [110].Therefore, it could be possible that under risk conditions a reduction in processes underlying short-term memory can lead to an increase in the processes underlying long-term memory.
The OF test was used to evaluate effects on mood.Exploration as estimate for interest in environment was obtained from walking distance.Although Zucker rats after surgery showed similar overall activity as non-surgery rats, distance in the OF test was significantly reduced.This was not affected by J147.In contrast, reduced time in the center area was significantly increased by chronic, but not by acute J147 treatment, suggesting an anxiolytic effect of J147.This is in general agreement with studies from our lab [55] and others [49,50].Moreover, these studies stress the notion that anxiety and cognition are closely associated and interacting processes [111][112][113][114], sharing brain structures, such as the amygdala and hippocampus [115,116].
Furthermore, the surgery-induced shift from primarily dark active to more light active behavior seemed to be shifted back by acute J147 treatment.In our previous study on effects of J147 on early recovery after surgery, the same pattern was seen [61].Shifts and disturbances in circadian rhythm could cause cognitive problems after surgery [117,118].To our knowledge there is no literature on the effect of J147 on the circadian rhythm yet.However, as J147 shows similarities in bio-activity to curcumin [48,119], J147 may act through the same mechanisms.Curcumin can alter the molecular circadian clock, by inhibiting inflammation through NF-κB, regulating circadian rhythms [120][121][122][123] and the immune system [124].Although based on circulating cytokine levels and microglia activity, no anti-inflammatory effects of J147 could be observed in our models of postoperative recovery ( [55,61] and present study), it cannot be excluded that J147 has acted through other inflammatory pathways to restore the circadian clock disturbances after surgery.

(neuro)inflammation
As previous studies indicated that POCD is highly associated with peripheral inflammation [13,23,29] and neuroinflammation [29,55,61]) and literature indicated that J147 has anti-inflammatory properties [42,47], different (neuro)inflammatory markers were included in the current study.Although in the current study, J147 treatment had beneficial effects, this did not seem attributable to inhibition of inflammation, as neither IL-1beta or NGAL levels indicated reduced peripheral inflammation, nor microglia activity suggested lower neuroinflammation.This is in agreement with our previous studies where we also did not see an anti-inflammatory effect of J147 [55] and [61].The discrepancy with other studies could arise due to timing of collection of the materials.Neuroinflammation after surgery in rats has a distinct time course, which is elevated 1 week after surgery, subsides in the second week and becomes declined at the third week [29].Moreover, it could well be that J147 shifted the time course of the inflammatory response, rather than, or in combination with, altering the magnitude.

Blood-brain barrier function
Previous research suggests that disruption in the BBB is a factor in development of POCD [26,28].BBB integrity in the hippocampus was reduced by surgery, as indicated by significantly increased leakage (IgG) and a tendency to reduced barrier protein expression (EBA).Although acute J147 treatment seemed to even further increase plasma NGAL levels, BBB function seemed restored towards higher barrier protein expression, without effect on IgG leakage.From literature, J147 is known to have neuroprotective properties, as it can preserve BBB homeostasis and vascular function [42].However, we did not see a reduction in leakage of IgG in the hippocampus, which could be due to the timing.We measured BBB leakage 14 days after surgery, therefore J147 may have improved BBB integrity, but remaining IgG was not yet cleared from the brain.

Metabolism and proteinuria
In this current study, J147 played a role in multiple metabolic processes, as it improved insulin sensitivity and altered plasma triglycerides.Moreover, J147 lowered surgery-induced proteinuria.More specifically, chronically J147 treated animals displayed higher insulin sensitivity after surgery, even after a short period (2 weeks) on J147.Previous study also confirmed a positive effect of J147 on glucose metabolism [43].Moreover, chronically J147 treated animals had lower proteinuria, suggesting J147 also has a positive effect on kidney functioning.Although J147 is not extensively studied in kidney failure, Kepchia et al. showed that J147 improved physiological markers for brain and kidney function, suggesting a protective role of J147 in kidney failure [125].Another possible action mechanism for J147 is trough regulation of the mitochondrial oxidative phosphorylation (OXPHOS) system.Literature showed that J147 could also affect the mitochondrial Alfa-F1 subunit of ATP Synthase (ATP5A) of the OXPHOS system.This results in an increase in intracellular calcium, leading to activation of the AMPK/mTOR pathway, which is a canonical longevity mechanism [45].To investigate whether J147 also affects expression of the OXPHOS proteins in Zucker rats that were subjected to abdominal surgery, we measured OXPHOS protein levels in the hippocampus and prefrontal cortex (PFC) of the brain.J174 treatment had no effect on total and individual expression levels of OXPHOS complexes, both in PFC and hippocampus.Surprisingly, OXPHOS levels in the hippocampus was significantly correlated with long-term memory and cognitive flexibility in the MWM test, while OXPHOS in the PFC was correlated with all spatial learning and memory test outcomes, as well as exploration in the OF.
All in all, J147 may improve aspects of metabolism, possibly through its impact on mitochondrial OXPHOS regulation.However, further research is necessary to fully understand the mechanisms and potential therapeutic applications of these findings.

Limitations and future perspectives
Zucker rats, expressing signs of frailty, and potentially providing a clinically relevant model for POCD research, displayed only marginal cognitive decline after surgery, hence leaving little opportunity for potential improvement with J147 therapy.
One of the most obvious risk factors in the Zucker rat model was obesity.Although obesity is considered as a risk factor for development of POCD, it could also induce beneficial effects in surgery.This contradiction is called the obesity paradox [126,127].Composition of food, malnutrition and inflammation status are suggested to be major players in the difference between healthy and non-healthy obesity [128].An unhealthy diet as cause for obesity is suggested to lead to more postoperative cognitive problems [129].Therefore, obesity in our Zucker rat model, resulting from eating more of the healthy rat chow, may not be the risk factor for POCD, indicated in patients.
To our knowledge there is no difference for males and female regarding prevalence of POCD [62].Therefore, in order to facilitate comparison with our previous studies [29,55,61], in the present study only male rats were used.Still, it may be worthwhile to compare male and female rats, in a follow-up study.
In addition, POCD could be less prominent in our Zucker rats due to their low-grade inflammation status.Low-grade inflammation before surgery could also present a protective effect against the effects of surgery; pre-conditioning [130,131].Low-grade inflammation primes the immune system, and could improving cognitive functioning after surgery [132].
Age is most prominent risk factor for POCD in patients.Although the Zucker rats were middle aged (25-30 weeks old), since spontaneous mortality steeply increases after middle age [63], older Zucker rats could not be used, as that could potentially bias our results.Moreover, as we used outbred rats, on the one hand this may better represent the patient population, but on the other hand induces more variation in the data, hampering strong statistically significant outcomes.
In hindsight, the interaction of the elevated leptin levels, characteristic for Zucker rats, and POCD development may have been more extensively explored.
Finally, although the dose, time and route of J147 administration were chosen carefully, regarding the time course of the (neuro)inflammatory response [29], chronic and acute J147 treatment gave different results.Optimization of timing could provide more effective treatment and may help to better understand the development of post-operative complications and the action mechanism of J147.

Conclusion
Although the Zucker rat model indeed displayed risk factors for POCD, surgery seemed to induce POD, rather than POCD in these rats.Treatment with J147 showed positive effects on behavioral and metabolic parameters, but the mixed effect of acute and chronic treatment may suggest a combination for optimal treatment.Exploring underlying mechanisms revealed no anti-(neuro)inflammatory effect of J147.Potential effects through altered metabolism need further investigation.

Fig. 2 .
Fig. 2. Behavioural test for the different experimental groups (mean ± SEM); lean (n = 11-13), Zucker non-surgery (n = 12-14); Zucker control surgery (n = 13-15); Zucker surgery + acute J147 treatment (n = 14-16); Zucker surgery + chronic J147 treatment (n = 14-16).A) Time spend in center in the Open Field test, B) distance moved per second in the Open Field test, C) Novel object recognition test, D) Novel location recognition test, E) Learning curve of the Morris Water Maze, including the AUC as inlay, and F) probe 1 and 2 of the Morris Water Maze test.*: significant difference between indicated groups (p < 0.05).^: significant different van chance level (50 % NOR and NLR and 15seconds in the MWM probe trails; dashed lines).