Corresponding author: Mhan-Pyo Yang, DVM, PhD, Department of Veterinary Medicine, College of Veterinary Medicine, Chungbuk National University, 410 Sungbong-ro, Heungduk-gu, Cheongju, Chungbuk 361-763, Korea; e-mail: firstname.lastname@example.org
Background: The use of soybean oil-based lipid emulsion (SO-based LE) in parenteral nutrition has been reported to impair neutrophil functions in humans and rodents. As yet, little is understood about the effects of SO-based LE on canine immune responses.
Hypothesis: A short-term infusion with SO-based LE affects the phagocytic responses of canine peripheral blood polymorphonuclear neutrophilic leukocytes (PMNs).
Animals: Twenty-four healthy Beagle dogs.
Methods: Experimental study. Dogs were randomly assigned into groups of six and administered a 2-hour IV infusion with 0.9% NaCl solution or sufficient SO-based LE (INTRALIPOS 20%) to supply 40, 100, and 200% of the basal energy requirement (BER). PMN functions were determined after collecting blood samples before, immediately after, and 24 hours after the infusion.
Results: None of the treatments significantly affected the phagocytic capacity of PMNs or circulating leukocyte numbers. The infusion providing 200% of BERs significantly reduced PMN oxidative burst activity, filamentous actin polymerization, and Cdc42 Rho guanosine triphosphatase activity immediately after its delivery. However, these functions were restored to pre-infusion values 24 hours after the infusion. The lower calorie infusions did not have these effects.
Conclusions and Clinical Importance: These results suggest that short-term infusions with a supraphysiological dose of SO-based LE may decrease the immune functions of canine PMNs. However, more long-term studies will be needed to extrapolate the effect of SO-based LE with clinically relevant doses in a practical situation.
Neutrophils are key cellular components of the innate immune system and form the large majority of peripheral blood polymorphonuclear cells.1,2 They are dynamic, motile cells that have a unique capacity to phagocytize and thereby eliminate pathogens and cell debris.2 Several of their receptors recognize chemoattractant molecules, after which they migrate to the relevant sites.3 This brings neutrophils into contact with invading foreign particles and activates their phagocytic receptors, which then mediate their ingestion of the foreign particles.4 These responses involve the polymerization and rearrangement of cellular actin filaments.5,6 The ingested microbe is then killed by microbicidal enzymes and the oxidative burst caused by the formation of reactive oxygen species (ROS).2,7 Recently, the biological functions of neutrophils were found to be under the control of Ras-homologous (Rho) guanosine triphosphatases (GTPases), which are a subgroup of the Ras superfamily of low molecular weight (MW 20–30 kDa) GTP-binding proteins.2,8,9 Rho, Rac, and Cdc42 are the best characterized members of the Rho family GTPases.9 They localize at membranes and act as a molecular switch, cycling between an active GTP-bound and an inactive guanosine diphosphate (GDP)-bound state.10
Parenteral nutrition has been used for critically ill animals that are unable to take oral or enteral nutrition.11–13 Dogs receiving parenteral nutrition typically receive 40 to 60% of their energy requirements as lipids.14 The most commonly used nutrition lipids are in the form of a soybean oil-based lipid emulsion (SO-based LE) (also known as long-chain triacylglycerol lipid emulsion),15 which has a high ratio of n-6 to n-3 polyunsaturated fatty acids. Although lipid emulsions constitute the main source of fuel calories and fatty acids in parenteral nutrition formulations,13,15 they have been reported to have adverse effects on patient outcomes: it has been suggested that they impair immune defenses and alter inflammatory responses.15–17 However, although there are numerous reports about interactions between lipids and the phagocytic system, it remains unclear whether SO-based LE affects the phagocytic responses of neutrophils.18,19 Previous reports examining this issue have yielded conflicting observations. Thus, although several studies have suggested that in vivo treatment with SO-based LE can impair neutrophil functions in rats and humans,20–22 2 other studies have reported that parenteral nutrition with SO-based LE does not alter neutrophil phagocytosis, superoxide production, or chemotaxis,23,24 and yet another study found that it increased neutrophil chemotaxis in humans.25 It was recently postulated that these discrepancies may reflect differences in the amount of lipid emulsion or the experimental model or both that was used.18,19
To our knowledge, little is known regarding the effect of infusion with SO-based LE on the phagocytic responses of canine PMNs (peripheral blood polymorphonuclear neutrophilic leukocytes), and besides there have been no in vitro studies with canine leukocytes and SO-based LE. Thus, we sought to determine the effect of a short-term IV infusion with 3 doses of SO-based LE that provided 40, 100, and 200% of basal energy needs on canine PMN phagocytic capacity and oxidative burst activity (OBA). We also examined the effect of these treatments on filamentous polymeric actin (F-actin) and Cdc42 activation levels and on circulating leukocyte numbers.
Materials and Methods
The subjects of this study were twenty-four 3-year-old laboratory Beagles. The dogs weighed 9.03 ± 0.54 kg (mean ± SD) and all were healthy, as judged by a physical examination, indirect measurement of systolic blood pressure, examination of fecal specimens for the presence of parasites by use of a flotation technique, heartworm antigen testing, complete blood count analysis, serum biochemical analysis, urinalysis, ACTH response testing, and diagnostic imaging. All dogs were housed separately in cages with a 12-hour light: 12-hour dark cycle and were fed a commercial dieta and provided tap water. All experimental procedures were approved by the ethics committee of the Chungbuk National University.
The 24 dogs were randomly assigned to 4 treatment groups (6 dogs per treatment). The animals received a 2-hour IV infusion with a physiological saline (0.9% NaCl) solution (Treatment A) or admixtures of physiological saline with sufficient SO-based LEb to supply 40, 100, and 200% of the basal energy requirement (BER) (Treatments B, C, and D, respectively). The energy requirements of the dogs were calculated by using the following formula: BER = 30 × (body weight in kilograms) + 70.14,15 As shown in Table 1, the total infusion volumes of the 4 treatments differed slightly. Because maintenance fluid requirements were estimated to be 60 mL/kg/d,26 physiological saline was infused on its own in Treatment A whereas it was added at relevant quantities to the SO-based LE amounts used for Treatments B-D to increase the total fluid intake to 60 mL/kg/d. The Treatment B-D solutions were aseptically compounded in a 100-mL infusion setc with a burette just before administration. After a 24-hour fasting period, all treatments were infused for exactly 2 hours through an over-the-needle catheter inserted into a cephalic vein. The infusion rate was calculated from the maintenance fluid requirements as described (Table 1). Blood samples were collected by jugular venipuncture before (time 0), immediately after (time 2), and 24 hours after (time 24) the infusion.
Table 1. Details of the 4 infused treatments.
A (n = 6)
B (n = 6)
C (n = 6)
D (n = 6)
The infused total energy was calculated by the following formula: (the adjusted energy requirement ÷ 24 hours) × 2.
Data are expressed as means ± SD.
BER, basal energy requirement; SO-based LE, soybean oil-based lipid emulsion.
Blood samples drawn into a K2-EDTA bottle were analyzed with an automated CBC analyzerd and the total leukocyte, neutrophil, lymphocyte, monocyte, eosinophil, and basophil numbers were determined. The results were compared with previously established reference ranges.27
To evaluate PMN functions, the PMNs were isolated by density-gradient centrifugation immediately after collecting the blood samples. Briefly, heparinized blood samples were overlaid in a 1 : 1 ratio on a Histopaque solutione (specific gravity, 1.077 ± 0.001). After centrifugation at 400 ×g for 45 minutes at room temperature, the PMNs were collected from the upper layer of sedimented erythrocytes. To purify the PMNs, erythrocytes were allowed to sediment for 60 minutes in a phosphate-buffered saline (PBS) solution containing 1.5% dextranf (MW, 200,000). The floating cells were then gently collected and pelleted by centrifugation at 400 ×g for 5 minutes. The residual erythrocytes were lysed by a brief treatment with 0.83% NH4Cl in a tri(hydroxymethyl)-aminomethane-base buffer (pH 7.2) for 5 minutes. The purity of PMNs in the final cell suspension was found to be >96%, as determined by Wright-Giemsa staining analysis of a blood film obtained by use of cytocentrifugation. The resulting PMNs were resuspended in RPMI 1640 mediume supplemented with 2 mM l-glutamine, 0.02 mg gentamicin/mL, and 5% heat-inactivated fetal bovine serum.g
Simultaneous Measurement of Phagocytic Capacity and OBA
Phagocytic capacity and OBA were evaluated simultaneously as described elsewhere.28 Briefly, the isolated PMNs were placed in 24-well plates at a density of 1 × 106 cells/mL and incubated for 2 hours at 37 °C in a 5% CO2-humidified atmosphere. For the final hour of culture, 20 μL of a carboxylate-modified polystyrene fluorescent microsphereh (size, 1.0 μm) suspension that had been adjusted to 1 × 109 beads/mL was added to each well. Fifteen minutes before the end of the culture period, 1 μM dihydrorhodamine 123e was added. The ROS-mediated conversion of the nonfluorescent dihydrorhodamine 123 into fluorescent rhodamine 123 was used to measure OBA29 and the phagocytic capacity was determined by estimating the number of PMNs containing phagocytosed fluorescent microspheres. This was done by subjecting the cells to flow cytometric analysis, as follows. The cultured cells were gently harvested, centrifuged at 400 ×g for 3 minutes at 4 °C, and washed 3 times with PBS solution containing 3 mM EDTA. The viability of the PMNs was >98% on the basis of their ability to exclude trypan blue dye. The cells were then resuspended in fixation bufferi in accordance with the manufacturer's instructions. All steps after beginning the cell culture were conducted in the dark. The cells were then analyzed by a multipurpose flow cytometerj and analysis softwarek with an argon laser set at 488 nm. Samples of 10,000 cells were assayed in triplicate. The FL1 channel was set to 505–545 nm to detect the green fluorescing rhodamine 123 and the FL3 channel was set to 630–660 nm to detect the red fluorescent microspheres. The cells were gated on the basis of forward and side light-scattering characteristics. Phagocytic capacity and OBA were expressed as percentages and mean fluorescence intensities (MFI, an arbitrary unit), respectively.
Determination of Cellular F-Actin Levels
To evaluate the effect of SO-based LE infusion on actin polymerization in the PMNs at the points of contact with the microspheres during phagocytosis, the total cellular F-actin levels were measured as described elsewhere.5,30 The isolated PMNs were placed in 24-well plates at a density of 1 × 106 cells/mL and incubated for 80 minutes at 37 °C in a 5% CO2-humidified atmosphere. For the final 20 minutes of culture, 20 μL of a carboxylate-modified polystyrene fluorescent microsphereh (size, 1.0 μm) suspension (1 × 109 beads/mL) was added. The cultured cells were then gently harvested, centrifuged at 400 ×g for 3 minutes at 4 °C, washed 3 times with PBS solution, and fixed with fixation bufferi at 4 °C according to the manufacturer's instructions. The fixed cells were washed 3 times with PBS solution and then stained in the dark for 15 minutes at 37 °C with 165 nM fluorescein isothiocyanate (FITC)-labeled phalloidine and 100 μg/mL lysophophatidylcholine.e The cells were washed and analyzed within 30 minutes by using a multipurpose flow cytometerj and analysis softwarek with the argon laser set at 488 nm. Samples of 10,000 cells each were assayed in triplicate. The FL1 channel was set to 505–530 nm to detect the green fluorescing FITC molecule. The F-actin levels were expressed as MFI. To assess changes in the cell shape and F-actin distribution, the fluorescence images of PMNs were visualized by confocal laser scanning microscopyl and analyzed with appropriate software.m The PMNs were counterstained with propidium iodidee nuclear stain to reveal the cell shape clearly.
Cdc42 Activation Assay and Western Blot Analysis
The Cdc42 activation levels in canine PMNs were determined by affinity precipitation by using a Cdc42 activation assay kitn in accordance with the manufacturer's protocol. Briefly, PMNs were isolated from blood samples that were collected immediately after the infusion. The PMNs were resuspended at a concentration of 1 × 107 cells/mL and incubated for 80 minutes at 37 °C in a 5% CO2-humidified atmosphere. The cells were left untreated (0 minutes) or exposed to 50 ng/mL phorbol myristate acetate (PMA) for the final 10 or 20 minutes of culture. The harvested PMNs were lysed in 500 μL of ice-cold Mg2+ lysis/wash (MLB) buffern supplemented with 5 μg/mL leupeptin, 5 μg/mL aprotinin, 1 mM sodium fluoride, and 1 mM sodium orthovanadate and clarified by centrifugation (22,000 ×g) for 10 minutes at 4 °C. The supernatants were then incubated for 45 minutes at 4 °C with gentle agitation with 5 μg of PAK-1 p21-binding domain (PBD)-agarose.n The PAK-1 PBD-agarose affinity precipitates were then washed 3 times with MLB buffer, eluted in 2 × Laemmli sample buffer for 5 minutes at 100 °C, electrophoresed on a 12% sodium dodecyl sulfate-polyacrylamide gel, and then transferred to a nitrocellulose membraneo with a Trans-Blot Cellp according to the supplier's instructions. Total cell lysates were analyzed as a loading control. Protein concentrations were determined by a Bio-Rad protein assayp using bovine serum albumin as the standard. The membranes were blocked for 30 minutes with PBS solution containing 3% nonfat dry milk and then incubated overnight with mouse monoclonal anti-human Cdc42 (mouse IgG1) antibodyn diluted to 1 μg/mL in freshly prepared PBS containing 3% nonfat dry milk. Mouse IgG1q served as the isotype control. Horseradish peroxidase-conjugated anti-mouse IgGn diluted to 1 : 3,000 in PBS solution containing 3% nonfat dry milk served as the secondary antibody. Immunoreactive proteins were visualized by the WEST-one Western Blot Detection Systemr according to the manufacturer's protocol. The signals were detected by using Chemi Doc EQp and analyzed by the Quantity One program (version 4.5.1).p The results were expressed as the ratio of GTP-Cdc42 to total-Cdc42 with the background subtracted.
All analyses were performed by statistical software.s Differences between the 4 different treatment groups were evaluated by means of a 1-way analysis of variance (ANOVA), which was followed by Dunnett's post hoc test. Two-group comparisons were performed by a 2-sample t-test. Normality tests (Kolmogorov-Smirnov) were performed to determine whether or not the results had a standard normal distribution. A P value of < .05 was considered significant. Results are shown as means±SD.
Effect of SO-Based LE Infusion on Circulating Leukocyte Numbers
To examine the effect of SO-based LE infusions on circulating leukocyte numbers, blood was taken before, immediately after, and 24 hours after the 2-hour infusion and complete blood counts were performed (Table 2). All cell counts of the circulating leukocytes numbers were within the reference ranges at all time points, and significant differences between the 3 SO-based LE treatment groups relative to the control group were not detected with regard to total leukocyte, neutrophil, lymphocyte, monocyte, eosinophil, and basophil numbers (P ranged from .832 to .188).
Table 2. Mean ± standard deviation of complete blood counts of samples collected from dogs infused with Treatments A-D for 2 hours.
Leukocyte Numbers (× 109/L)
No significant differences among the 4 treatments or within the treatments at the 3 points were noted, as determined by 1-way analysis of variance (ANOVA).
Effect of SO-Based LE Infusion on the Phagocytic Capacity and OBA of Canine PMNs
Relative to preinfusion values (time 0), none of the treatments had a significant effect on PMN phagocytic capacity when examined immediately (time 2) or 24 hours after the infusion (time 24) (Fig 1A; P ranged from .738 to .304). However, the PMNs from dogs who received 200% of their basal energy needs in their SO-based LE infusion (Treatment D) had significantly lower OBA values immediately after the infusion when compared with preinfusion values (P= .008) (Fig 1B), although these values returned to preinfusion levels 24 hours after infusion. The time 2 OBA values of the Treatment D group were also significantly lower than the time 2 OBA values of the Treatment A group (P= .004). In contrast, the time 2 and time 24 values of Treatments B (P= .806) and C (P= .071) had no significant effect on PMN OBA relative to preinfusion values.
Effect of SO-Based LE on F-Actin Levels in Canine PMNs
The PMN F-actin levels were significantly lower in the Treatment D dogs immediately after the infusion relative to the preinfusion levels (P= .026) but were restored to preinfusion values 24 hours after the infusion (Fig 2A). The time 2 values of Treatment D dogs were also significantly lower than the time 2 values of the Treatment A dogs (P= .013). However, the time 2 and time 24 values of the Treatment B and C dogs did not differ significantly relative to preinfusion values (P= .948 and .973, respectively). Confocal fluorescence imaging analysis confirmed that, relative to the time 2 PMNs from Treatment A dogs, the time 2 PMNs from the Treatment D dogs exhibited an overt decrease of actin polymerization at the points of microsphere contact (Fig 2B).
Effect of SO-Based LE on Cdc42 Activity in Canine PMNs
Western blot analysis of the canine PMNs taken immediately after infusion with Treatments A and D revealed that Cdc42 protein was expressed by the canine PMNs (Fig 3). Cdc42 activation was assessed by examining the response of the PMNs to 0, 10, or 20 minutes of culture with 50 ng/mL PMA. The Treatment A PMNs responded significantly to PMA treatment, because they had higher levels of the GTP-bound form of Cdc42 when they had been cultured with PMA for 20 minutes than when they had not been cultured with PMA (P= .021). In contrast, the Treatment D PMNs did not respond significantly to PMA treatment (P= .241). Moreover, compared with the Treatment A PMNs, the Treatment D PMNs showed significantly lower Cdc42 activation after being exposed to PMA for 0 (P= .033), 10 (P= .017), and 20 (P= .006) minutes.
Here we showed that a single 2-hour infusion at a clinically relevant infusion rate with SO-based LE that provided 40, 100, or 200% of energy requirements did not alter circulating leukocyte numbers in dogs. This contradicts a previous study showing that infusion with SO-based LE elevated the granulocyte counts of healthy humans.20 However, we observed that the treatment providing 200% of BERs (a supraphysiological dose) inhibited the OBA of canine PMNs during phagocytosis, although this activity had recovered 24 hours after the infusion.
Actin is a major cytoplasmic component of neutrophils and exists in 2 physical states, globular monomeric (G)- and F-actin.30,31 Upon activation during phagocytosis, G-actin converts into F-actin and localizes at the cell cortex.31,32 It is well known that actin polymerization plays an essential role in a variety of neutrophil functions, including chemotaxis, phagocytosis, and ROS production.5,33–35 Moreover, abnormal actin dynamics in neutrophils have been linked to increased susceptibility to infection by humans36 and may be associated with myelodysplastic syndromes.37 It has also been demonstrated that many of the respiratory burst components are associated with the cytoskeleton,38 and that an intact actin cytoskeleton is required for prolonged OBA during neutrophil phagocytosis.39 Given these links between actin dynamics and OBA and other neutrophil functions, we next examined the effect of SO-based LE infusion on actin polymerization in canine PMNs. For this, we used phalloidin, which binds much more tightly to F-actin than to G-actin.5 Because phalloidin binds to F-actin at a 1 : 1 ratio, the amount of phalloidin bound is equivalent to the amount of F-actin present.5 We found that the short-term infusion with the high dose of SO-based LE remarkably reduced the total cellular F-actin levels in the canine PMNs at the points of contact with the microspheres during phagocytosis. Those results enabled us to speculate a possibility that the attenuation of OBA by the SO-based LE may be correlated with an actin-dependent mechanism in canine PMNs. However, the lower doses of SO-based LE did not have this effect. It appears that canine PMN functions will most likely only be attenuated by short-term IV infusion with SO-based LE when the infusion dose per hour is far beyond a clinically relevant dose. This observation is in agreement with the suggestion of several authors18,19 that some of the discrepancies between previous studies may be because of differences in the amounts of lipid emulsions that were infused.
We recently showed that the oxidative burst of naive canine PMNs is triggered only when phagocytosis is stimulated by microspheres.28 This suggests that naive canine PMNs revealing decreased ROS production may also exhibit impaired phagocytosis.28 Unexpectedly, however, even though we used the same microspheres and found that Treatment D significantly suppressed canine PMN OBA, neither Treatment D nor the lower calorie treatments suppressed PMN phagocytic capacity (although Treatment D reduced the mean phagocytic capacity at time 2 relative to preinfusion levels or the effect of Treatment A at time 2, this effect was not statistically significant). This indicates that the inhibitory effect of a supraphysiologic dose of SO-based LE on ROS production does not necessarily mean an equivalent reduction in phagocytic capacity. However, given that we used only a single and relatively short infusion period, and phagocytic capacity was examined only by measuring microsphere uptake, additional experiments assessing the effect of a supraphysiologic dose of SO-based LE on the phagocytic capacity of canine PMNs are needed. Meanwhile, here we used the unopsonized microspheres for stimulation of phagocytic responses. The achievement of maximum phagocytic responses may be directly influenced by whether the microspheres are opsonized or not.1,4,30 Consequently, if we employed an opsonized foreign particle as a stimulant, there is also a possibility that this would lead a difference in our interpretation of the effect of SO-based LE on phagocytic capacity and OBA of canine PMNs.
The Rho GTPase Cdc42 has recently received special attention because it was shown to tightly regulate the innate immune responses of neutrophils as well as macrophages, including their actin polymerization, chemotaxis, and ROS production.40–42 For example, it was reported recently that impairment of Cdc42 activation in human alveolar macrophages may contribute to their reduced mannose receptor-mediated phagocytosis of unopsonized Pneumocystis organisms.43 It has also been shown that Cdc42 can regulate oxygen radical generation in mouse endothelial cells by reorganizing actin filaments, and that this increases cell migration.35 Although these key regulatory functions of Rho GTPase Cdc42 were elucidated by studies on laboratory rodents and human cells, it is thought that they probably also occur in virtually all mammals. These observations led us to examine whether the short-term inhibitory effects of the supraphysiological SO-based LE dose on canine PMN F-actin levels and OBA are accompanied by reduced Cdc42 activity. Indeed, Treatment D reduced Cdc42 activation in canine PMNs. This was true regardless of whether or not the cells were stimulated with PMA, which directly activates physiologically relevant signaling pathways, including the Cdc42 GTPase-mediated pathway.2,43
In clinical medicine, nutritional therapy, including parenteral nutrition, has been suggested to be important for immunoregulation, inflammation, neoplasia, and vascular diseases.44,45 Fatty acids may modulate the immune system by regulating arachidonate metabolism and eicosanoid production, modifying membrane fluidity, and regulating gene expression.46–48 It has been suggested that lipid emulsion-induced impairments of leukocyte migration and phagocytosis may be associated with lipid-induced changes of the cell membrane.18,20 This indicates that the fatty acid composition of phagocytes may directly influence their functional responses such as actin polymerization and OBA after their signaling molecules are triggered. Consequently, it may be that the ability of high-dose SO-based LE to suppress canine PMN functions may correlate with changed membrane fluidity that arises from rapid Cdc42-dependent interactions between the lipid emulsion and PMNs. To clarify the relationship between Cdc42 activity and the effects of SO-based LE, further research on other Rho-GTPase members (Rac and Rho) will be necessary.
In summary, it appears that a short-term infusion with a supraphysiologic dose of SO-based LE decreases canine PMN production of ROS and F-actin polymerization in a Cdc42-mediated manner, although the high infusion rates of SO-based LE would be impossible to achieve under clinical conditions. On the one hand, short-term infusions with clinically relevant doses of SO-based LE may not significantly inhibit canine PMN functions, whereas on the other hand clinicians will need to interpret our results with caution because our study has certain limitations.
aProPlan, Nestle Purina PetCare Korea Ltd, Seoul, Korea
bINTRALIPOS 20% INJ, Fresenius Kabi Korea Ltd, Seoul, Korea
cJ.M.S.(K) Medical Supply Co Ltd, Seoul, Korea
dCELL-DYN 3700, Abbott Diagnostics, Abbott Park, IL
eSigma-Aldrich Inc, St Louis, MO
fWako Pure Chemical Industries Ltd, Osaka, Japan
gInvitrogen Co, Grand Island, NY
hTransFluoSpheres, Molecular Probes Inc, Eugene, OR
iBD Cytofix, BD Biosciences, San Jose, CA
jFACSCalibur system, Becton Dickinson Immunocytometry Systems, San Jose, CA
kCELLQuest, version 3.3, Becton Dickinson Immunocytometry Systems
Authorship Contribution: Ji-Houn Kang designed research, performed research, analyzed data, and wrote this manuscript.
Mhan-Pyo Yang designed research, analyzed data, wrote this manuscript, was involved in drafting this manuscript, and had an opportunity to approve subsequent revisions.
Grants: This work was supported by a Korea Research Foundation Grant funded by the Korean Government (MOEHRD) (KRF-2007-313-E00540), a graduate fellowship provided by second Brain Korea 21 projects, and a research grant from the Chungbuk National University.