Evaluation of Peripheral Blood Neutrophil Functions after an Oral Carbohydrate Overload in Obese and Insulin Resistant Horses

Background: Obesity and insulin resistance (IR) are conditions of increasing prevalence in populations of equids worldwide. The immune impairment described in metabolic dysfunction status in humans has been extensively reported with minimal data regarding horses. The objective of the study was to evaluate the effect of obesity as an isolated factor and in association with insulin resistance on apoptosis, phagocytosis and oxidative burst activity of neutrophils isolated from peripheral blood of lean and obese adult horses with or without insulin resistance, basally and after induction of hyperglycemia through an oral glucose test. Results: No differences in apoptosis were observed between experimental groups at any time point. Phagocytic capacity was signicantly diminished at baseline in the obese-IR group (P<0.05) but increased after stimulation of hyperglycemia (P=0.007). Basal reactive oxygen species production differed signicantly (P=0.0001) between the obese-insulin sensitive (IS) and lean-IS or obese-IR groups, and decreased signicantly after stimulation of hyperglycemia in the lean-IS and obese-IS groups (P<0.05). Conclusions: Results from this study showed that both metabolic status itself, and acute hyperglycemia, are factors that inuence PMNs functionality in horses, specically phagocytosis and oxidative burst. This indicates the need for close monitoring of immune function in horses with inammatory disease and concurrent obesity and insulin resistance.


Background
Endocrine disorders associated to obesity represent a worldwide health problem, with special importance in humans and companion animals [1,2]. In horses, the epidemic of obesity is a growing and concerning problem with rates of obesity as high as 51% reported [3,4]. Classically, as in other species, obesity in horses has been considered a risk factor for the development of insulin resistance (IR), which in conjunction with hyperinsulinemia is referred to as insulin dysregulation (ID) in equine populations [5].
Abnormal insulin metabolism is a key feature of endocrinopathic disorders and has been directly linked with the development of the life-threatening condition laminitis [6][7][8][9]. The close relation between obesity, IR and immune compromise is well recognized in humans and mice [10,11], while in horses this relation is not fully understood with fewer reports on the topic [12][13][14][15]. Although obesity and IR are closely linked, it must be acknowledged that these conditions can develop independently.
Obesity is not only a predictor of progression of later endocrine disease, but an altered metabolic state in itself [16][17][18]. The adipose tissue is an active endocrine organ, whose dysfunctional activity is mainly characterized by altered secretion of adipokines and increased expression of proin ammatory cytokines (principally tumor necrosis factor alpha (TNFα) and interleukin 1 (IL-1)) [19]. The activation of the immune system as consequence of metabolic dysfunction, leads to a characteristic low-grade chronic in ammation state reported in obese humans [20][21][22]. In horses, the state of chronic in ammation associated to endocrine dysfunction remains unclear, with inconsistent data reports linking body condition score (BCS) and secretion of proin ammatory cytokines [13,[23][24][25]. The complex interactions between the endocrine and innate immune system are still incompletely understood in equids [26][27][28].
Neutrophils (polymorphonuclear cells, PMNs) are part of the innate immune system and play a major role in host defense, as the rst cellular type to be recruited to ght in ammation and infection [29]. PMNs are the most abundant leukocyte cell type on the circulating blood of mammals, including horses [30]. In humans with poor metabolic control associated to metabolic conditions such as type 2 diabetes mellitus [31,32], high levels of circulating glucose and insulin are able to impair neutrophil functions, with hyperglycemia acting as an important downregulatory factor for PMNs activity [33,34]. In healthy humans, acute hyperglycemia has been associated with a decrease in neutrophil chemotaxis and phagocytosis [35], impaired reactive oxygen species (ROS) production [36] and inhibition of neutrophils extracellular traps (NETs) formation [37]. Although horses are known as good regulators of blood glucose levels, with pancreatic exhaustion occurring infrequently [19], limited data exists regarding neutrophil function in horses with energetic metabolism abnormalities. This study group recently reported results in obese non-insulin dysregulated horses [38], and only a few studies have reported altered PMNs function in hyperinsulinemic-obese horses [13] and aged horses suffering from pituitary pars intermedia dysfunction [14]. The effect of endocrine disturbances on PMNs functions in horses remains unknown, and more studies are needed to understand the close relation between the immune system and endocrine disorders. To the authors´ knowledge, minimal data is available regarding the assessment of neutrophil functionality of horses with metabolic disturbances, and no previous studies have been reported regarding the in vivo effect of acute induced hyperglycemia on equine neutrophil functions. Therefore, the purpose of the study is to evaluate apoptosis, phagocytosis and oxidative burst activity of neutrophils from peripheral blood, basally and after induction of acute hyperglycemia through an oral glucose test, in groups of lean and obese adult horses with or without insulin resistance.

Animals
Seventeen adult horses, including eight geldings and nine non-pregnant mares were enrolled in the study.
Horses were privately owned or part of the Austral University of Chile veterinary teaching hospital research herd. An informed consent was signed before the performance of the study for privately owned horses. Animals did not participate in any other studies for at least a month prior to enrollment in this study. Breeds included eight Chilean horses and nine crossbreed horses. To ensure animals´ optimal health conditions, a complete clinical examination and blood analyses were performed before the onset of the study. Animals were selected and allocated into three groups based on their body condition score (BCS) and metabolic status (presence or absence of insulin resistance). BCS was assessed by two experienced observers, based on a 9-point scale [39]. To allocate horses in their correspondent group, all horses underwent an insulin tolerance test (ITT) as previously described before the performance of the study to determine their tissue insulin sensitivity or insulin resistance status [40]. The control lean insulin sensitive group (lean-IS) included seven horses with a regular BCS (4-6/9) and absence of insulin resistance. The second group comprised ve obese insulin sensitive (obese-IS) horses with an obese BCS (≥ 8/9) and absence of insulin resistance. The third and nal group, obese insulin resistant (obese-IR), included ve animals with an obese BCS (≥ 8/9) and presence of tissue insulin resistance. Characteristics data from the three groups are summarized in Table 1. The study was approved by the Bioethics Committee for the Use of Animals in Research of Austral University of Chile (Resolution N°3 08/2018). After experimentation, all animals resumed normal previous activities in their herds of origin. Table 1 Characteristics and endocrine data of the animal groups selected and enrolled in the study, lean-IS, obese-IS and obese-IR. Data is presented as mean ± standard deviation (SD). Body condition score data is presented as median (range). Within a row, values with different superscripts differs signi cantly (P<0.05). Lean-IS n=7; obese-IS n=5 and obese-IR n=5; insulin sensitive; IR, insulin resistance.

Study design
All horses underwent an infeed oral glucose test (infeed-OGT). The night before the test, horses were moved to stalls and fasted overnight. Free access to water was allowed. At 8:00 AM, horses were fed 1.0 g dextrose/kg, 0.15% of body weight alfalfa cubes and 200 g of wheat bran mixed with 200 mL of water, as previously described [41]. Horses were allowed a 15 min period to eat their meals, after which, if the meal wasn´t eaten, the test was discarded and repeated after a wash-out period of two days. Before the meal was offered to the horses, jugular blood samples were taken for the measurement of basal blood glucose, insulin and further laboratory analysis. After the meal was offered, serial blood samples were collected at 30, 60, 90 and 120 min for the analysis of blood glucose. At 120 min, blood samples were taken for the measurement of glucose, insulin and for laboratory analysis. Insulin was determined at baseline (time 0) and 120 minutes. Animals with a fasting insulin concentration > 20 µIU/mL, or ≥ 85 µIU/mL at 120 min were excluded from the study. Once the test was completed, animals were returned to paddocks. Tests were performed at the Austral University of Chile veterinary teaching hospital, with the exception of ve horses that were tested on eld, following the same protocol described above.

Blood glucose and insulin analysis
Blood glucose was determined by a handheld glucometer validated for use in horses (OneTouch Ultra; LifeScan, Wayne, PA, USA) [42]. Hyperglycemia was diagnosed with blood glucose concentrations above 124 mg/dL [43]. Blood samples for insulin were collected in red top tubes and kept at room temperature to allow samples to clot for at least 1 hour. After this time, samples were kept in a cooler for transportation to the laboratory. All samples were centrifugated at 400 g for 10 minutes. Serum samples were harvested and transferred to new tubes and stored at -20 °C until analysis. Insulin concentration was determined using a chemiluminescence assay (ADVIA Centaur, Siemens Healthcare Diagnostics) as previously described [44].

Blood samples and neutrophil isolation
Peripheral blood samples (10 mL) obtained by jugular venipuncture, at baseline and 120 min after oral glucose administration, were stored in sterile tubes containing 1 mL of 3.8% w/v trisodium citrate as anticoagulant. Blood samples for neutrophil isolation were kept at room temperature for 2 hours until further processing in the laboratory. A discontinuous gradient (Percoll®GE Healthcare) was carried out following the protocol described by Borlone et al., (2017). Brie y, 85% density Percoll (4 mL) were placed in the bottom of a 15 mL propylene tube, followed by 70% Percoll (4 mL) and blood on top (4 mL). Tubes were centrifugated for 45 min at 1500 x g at room temperature for the obtention of two separately layers containing, mononuclear and PMNs cells respectively, for both times, baseline and 120 min. PMNs layer was aspirated and neutrophils were washed with phosphate-buffered saline solution (PBS 1X). Cell pellet was resuspended in 2 mL of PBS 1X, neutrophils were counted using a Neubauer chamber and immediately used.

Apoptosis Assay
In vitro neutrophil apoptosis was assessed at two-time points (baseline and 120 min) using the FITC annexin V apoptosis detection kit (BD Biosciences) according to the manufacturer's protocol. Neutrophils

Phagocytosis Assay
In vitro neutrophil phagocytic capacity was assessed at two-time points (baseline and 120 min) using a commercial phagocytosis kit (pHrodo™ Green E. Coli Bioparticles™, ThermoFisher Scienti c) according to the manufacturer's protocol. Brie y, 1 × 10 6 neutrophils were resuspended in PBS (100 µL) and incubated in cytometry tubes with 25 µL of pH-sensitive pHrodo™ (2 mg/mL) at 37 °C for 30 minutes. Samples incubated at 4 °C for 30 minutes were used as negative controls. After incubation, engulfment of pHrodo bioparticles by cells were immediately analyzed by ow cytometry (BD FACSCanto II cytometer, San Diego, CA, USA). Flow cytometry data was analyzed using FlowJo v10 software (FlowJo, LLC, USA).

Respiratory burst production
In vitro assessment of neutrophil ROS production at baseline and 120 min was determined using the luminol-dependent chemiluminescence method. In a white 96-well sterile plate, neutrophils resuspended in RPMI-1640 media (ThermoFisher Scienti c) were aliquoted at a concentration of 3 × 10 5

Statistical Analysis
Results are presented as least square mean (LSM) ± standard error (SE). Normality of the data was assessed using the Kolmogorov-Smirnov test. AUC values were obtained for ROS production curves and used for further analysis. A repeated measures linear mixed model was tted for the analysis of PMNs functions. Phagocytosis, apoptosis and AUC for ROS production were used as outcome variables, with time, group and the interaction between time and group as xed effects, including subjects as a random effect. Tukey's HSD post hoc was performed when signi cant differences were found. Statistical analyses were performed with the commercial software RStudio (RStudio, USA). Graphs were obtained using GraphPad Prism 5.00 for windows (GraphPad Software, La Jolla California, USA). Statistical signi cance was de ned at a p-value < 0.05.

Insulin and glucose response to in-feed OGT
All animals tolerated the dextrose powder in the meal during the in-feed OGT and no adverse effects were observed in the horses. LSM ± SE for blood glucose concentrations at baseline showed no signi cant differences (P > 0.05) between the lean insulin sensitive (lean-IS) (92.1 ± 13.9), obese insulin sensitive (obese-IS) (91.6 ± 16.4) and obese insulin resistance (obese-IR) (111.4 ± 16.4) groups (Fig. 1A). All baseline glucose measurements were within the reference range established for the species, and hyperglycemia was not detected. At 30 min, a signi cant difference (P < 0.05) was observed between the lean-IS (112.9 ± 13.9) and obese-IR (166.6 ± 16.4) groups, but no signi cant differences were observed (P > 0.05) when the control group (lean-IS) was compared to the obese-IS animals (117.4 ± 16.4) (Fig. 1A).

Discussion
The present study aimed to evaluate changes in peripheral blood neutrophil apoptosis, phagocytosis and ROS production, in response to acute hyperglycemia, in horses with obesity and insulin resistance. Our results provide evidence that obesity per se can modulate neutrophil oxidative burst independently of insulin sensitivity. Moreover, obesity in association to IR impairs peripheral blood neutrophil phagocytosis but not ROS production. Induced hyperglycemia increased phagocytosis in the obese-IR group and diminished ROS production in all study groups, although this decrease was not signi cant for the obese-IR group. To the authors´ knowledge this is the rst report regarding the effect of in vivo hyperglycemia, obesity and IR, on these aspects of neutrophil functionality.
The respiratory burst is an important oxygen-dependent function of neutrophils, highly involved in their bactericidal capacity [46]. In the present study, a marked increase in basal ROS production was observed in the obese-IS group when compared to both, the lean-IS and obese-IR groups. These results are consistent with those reported by Brotfain et al. (2015), who reported overproduction of ROS in morbidly obese non-diabetic human patients. Conversely, obese-IR horses showed a minimal increase in ROS production, with results similar to those observed in the lean-IS group. This was an interesting and unexpected result. In humans, IR is a condition characterized by high ROS production and is a key feature of type 2 diabetes mellitus [47]. Studies conducted in subjects with type 2 diabetes have widely documented this marked increase in ROS production [48,49]. Furthermore, high levels of ROS have been linked to endothelial damage in humans suffering from metabolic disorders [50]. Likewise, endothelial dysfunction associated with endocrinophatic laminitis has been reported in horses with metabolic disturbances such as metabolic syndrome [51]. In horses few studies have explored the relation between metabolic disorders and neutrophil respiratory burst [13,14]. Holbrook et al., (2012) observed a marked increase in ROS production in hyperinsulinemic obese horses when compared to a lean group. Although none of the animals tested in our study were hyperinsulinemic, the contradictory results obtained between Holbrook's and the present study requires further research. On the other hand, in our study acute hyperglycemia induced a signi cant reduction of neutrophil ROS production in both insulin sensitive groups, but no signi cant decrease was observed in the obese-IR group. In healthy humans, induced acute hyperglycemia assessed through different clamp techniques failed to in uence neutrophils oxidative burst response [52,53]. However, in diabetic patients, hyperglycemia is described as an important priming factor of neutrophils responsible for their excessive response under stimulation, including an exaggerated production of ROS [54,55]. Furthermore, hyperglycemia stimulates the activation of the polyol pathway, an alternative metabolic route, in which glucose at high concentrations is converted to sorbitol by the enzyme aldose reductase [31]. In diabetic patients the activation of this metabolic pathway has been linked with increased neutrophils ROS production, which levels were restored after the use of an aldose reductase inhibitor [56]. Although pancreatic exhaustion and consequent poor glycemic control are uncommon in horses [19], cases of type 2 diabetes have been reported [57] and an activation of this alternative pathway cannot be discarded. To date, no data is available regarding the effect of hyperglycemia and alternative metabolic routes activation on neutrophil functions of horses with metabolic disorders. Thus, these results suggest that metabolic status of horses and hyperglycemia affect the responsiveness of blood neutrophils, with an exacerbated oxidative burst response observed in the obese-IS group, which could play a detrimental role in horses´ health.
Regarding phagocytosis, the obese-IR group showed an important decrease in the phagocytic rate of neutrophils compared to the insulin-sensitive groups, regardless of their body condition score. Consistent with our results, a decreased PMNs phagocytic activity has been reported in humans and rats with type 2 diabetes [58][59][60]. Furthermore, results from the insulin sensitive groups (lean-IS and obese-IS) are consistent with those described by Brotfain et al., (2015), where no differences in neutrophil phagocytosis were found between lean and non-diabetic morbidly obese subjects. Few studies in horses have aimed to determinate the effect of metabolic diseases on neutrophil phagocytosis [13,14]. In a study conducted in obese-hyperinsulinemic horses, no differences were reported in the phagocytosis of bacteria when compared to a lean-control group [13]. Although our results are not consistent with that previous report, our study population differed in that none of the animals were hyperinsulinemic, but rather, insulin resistant, which is a more prevalent condition in clinical patient populations. Interestingly, hyperglycemia in the obese-IR group appears to increase the phagocytic rate of neutrophils. This nding differs from available in vitro data, in which hyperglycemia has been reported to impair and decrease phagocytosis of human PMNs [33,61]. On the other hand, apoptosis is an important neutrophil function for the maintenance of tissue homeostasis [62]. Alterations in the process of neutrophil programmed cell death has been associated with chronic in ammation, tissue injury and some diseases in humans [63]. In the present study, apoptotic capacity of neutrophils appears to not be in uenced by either, hyperglycemia or the metabolic status of horses. Although a slightly decrease in the apoptotic rate was observed in the two obese groups, no signi cative differences were reached. Thus far, little information is known regarding the relation between metabolic disorders and neutrophil apoptosis in horses. Only one study reported an increased expression of apoptotic related proteins in adipose and hepatic tissue of horses with equine metabolic syndrome [64]. However, other reports have studied the effects of different diseases, such as asthma [65], colitis and lipopolysaccharide infusion [66,67], upon neutrophil apoptosis. Interestingly, delayed neutrophil apoptotic rates have been reported in the studies mentioned above. Diverse studies in humans and mice had demonstrated that both, a dysfunctional metabolic status like type 2 diabetes mellitus and hyperglycemia are able to decrease the rate of neutrophil apoptosis [33,68,69]. Our study do not support similar ndings in the equine species, but it must be acknowledged that only spontaneous neutrophil apoptosis was assessed, in absence of stimuli reagents. Stimuli are able to enhance the prolonged lifespan of neutrophils, and therefore differences are easier to observe when metabolic disturbances are present [69].
The low number of horses tested per group is the main limitation of the present study. A larger study population will be able to increase the power of the test and provide more robust evidence. Furthermore, the effect of hyperglycemia over neutrophils functions was determined at a single time point after the carbohydrate overdose. In vitro experiments with different glucose concentrations, and in vivo trials with persistent hyperglycemia, could provide more information on the effect of glucose on neutrophil functions. Additionally, only one laboratory technique was performed for the assessment of each neutrophil function in this study. Future experimental studies testing different techniques and stimuli, and including other neutrophil functions such as chemotaxis, adherence capacity and NETosis should be explored. Moreover, future studies should include insulin dysregulated animals, as this group was not included in the present report.

Conclusion
In conclusion, our results address the effect of obesity as an isolated factor and in association with IR on blood neutrophil functionality in adult horses, indicating that phagocytosis and oxidative burst functions can be altered by metabolic status itself and by acute hyperglycemia in horses. Furthermore, our results support the fact that endocrinology is a dynamic science, and that the extrapolation of results from other species to the equine population should be done cautiously.