Role of lipase-generated free fatty acids in converting mesenteric lymph from a noncytotoxic to a cytotoxic fluid

Abstract

Recent studies have shown that mesenteric lymph plays a very important role in the development of multiple-organ dysfunction syndrome under critical conditions. Great efforts have been made to identify the biologically active molecules in the lymph. We used a trauma-hemorrhagic shock (T/HS) model and the superior mesenteric artery occlusion (SMAO) model, representing a global and a localized intestinal ischemia-reperfusion insult, respectively, to investigate the role of free fatty acids (FFAs) in the cytotoxicity of mesenteric lymph in rats. Lymph was collected before, during, and after (post) shock or SMAO. The post-T/HS and SMAO lymph, but not the sham lymph, manifested cytotoxicity for human umbilical vein endothelial cells (HUVECs). HUVEC cytotoxicity was associated with increased FFAs, especially the FFA-to-protein ratio. Addition of albumin, especially delipidated albumin, reduced this cytotoxicity. Lipase treatment of trauma-sham shock (T/SS) lymph converted it from a noncytotoxic to a cytotoxic fluid, and its toxicity correlated with the FFA-to-protein ratio in a fashion similar to that of the T/HS lymph, further suggesting that FFAs were the key components leading to HUVEC cytotoxicity. Analysis of lymph by gas chromatography revealed that the main FFAs in the post-T/HS or lipase-treated T/SS lymph were palmitic, stearic, oleic, and linoleic acids. When added to the cell culture at levels comparable to those in T/HS lymph, all these FFAs were cytotoxic, with linoleic acid being the most potent. In conclusion, this study suggests that lipase-generated FFAs are the key components resulting in the cytotoxicity of T/HS and SMAO mesenteric lymph.

a body of work emerging over the last decade documenting that factors contained in intestinal (mesenteric) lymph can lead to systemic inflammation and organ injury has resulted in the generation of the gut lymph hypothesis of the multiple-organ dysfunction syndrome (MODS) (14). This hypothesis is based on the following major lines of evidence. First, ligation of the main lymph duct exiting the gut limits the development of MODS. Second, lymph from stressed animals has a wide range of tissue-injuring and proinflammatory activities when tested in vitro as well as when injected into naive animals (14). This work showing that factors carried in the intestinal lymph from the stressed gut significantly contribute to a systemic inflammatory state and MODS has been validated in a large number of acute conditions, including trauma-shock, burn injury, pancreatitis, isolated intestinal ischemia, and endotoxemia (2, 5, 11, 12, 14, 17, 23, 24, 27, 37). One common factor among all these insults is that they are associated with intestinal hypoperfusion and ischemia due to the shunting of blood from the splanchnic to the central circulation (14, 17). This intestinal ischemia-reperfusion insult results in release of biologically active molecules from the stressed gut into the mesenteric lymphatics, which in turn drain into the systemic circulation, where they exert their systemic effects. It is not surprising that a number of studies have been directed at identifying the exact factors in mesenteric lymph that are responsible for its in vitro and in vivo injury-inducing effects. These studies have ranged from isolation/purification approaches (10, 22, 33) to proteomics (16, 26, 29), as well as the direct measurement of putative factors, such as cytokines, bacterial products, and endogenous danger-signaling molecules or alarmins (9, 32). This work has documented that a number of likely putative lymph mediators, including endotoxin and cytokines, are not responsible for the physiological activities of shock lymph. Furthermore, while proteomic studies have documented differences between shock and sham-shock intestinal lymph, none of these studies has led to the identification of the exact factors in lymph contributing to its biological activity. In addition to the studies investigating protein factors, several studies have suggested that lipid factors are responsible for at least some of the biological activity of trauma-hemorrhagic shock (T/HS) lymph, including its cytotoxic and neutrophil-activating properties (10, 18, 33). Recent studies showing that T/HS lymph has increased lipoprotein lipase activity and that this increase in lipoprotein lipase activity appears to significantly contribute to the in vitro endothelial cytotoxicity of T/HS lymph support the involvement of lipid factors in the biological activity of T/HS lymph (13, 31). Thus this study tested the hypothesis that increases in specific free fatty acids (FFAs) are involved in the cytotoxicity of T/HS lymph. To test this hypothesis, two models associated with gut injury were used: 1) the total body T/HS ischemia-reperfusion model and 2) the isolated intestinal superior mesenteric artery occlusion (SMAO) ischemia-reperfusion model.

MATERIALS AND METHODS

Animals.

Specific pathogen-free male Sprague-Dawley rats (Taconic Farms, Germantown, NY; 300–350 g body wt) were housed under barrier-sustained conditions and kept at 25°C with 12:12-h light-dark cycles. The rats had free access to water and chow (Teklad 22/5 Rodent Diet W-8640, Harlan Teklad, Madison, WI). All the experiments were approved by the New Jersey Medical School Animal Care Committee and were conducted in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the University of Medicine and Dentistry of New Jersey.

Collection of mesenteric lymph.

Animals were anesthetized by administration of pentobarbital sodium (50 mg/kg ip). A heating pad was applied as required to maintain body temperature at 37°C. After anesthesia, the animals underwent a 5-cm midline laparotomy, and a left medial visceral rotation was performed to expose the mesenteric lymph duct. A silicone catheter was inserted into the main efferent mesenteric lymphatic duct, secured in place with cyanomethacrylate glue (Super Glue, Duro), and exteriorized through a right flank stab wound. The laparotomy incision was closed in two layers with 3-0 silk. The mesenteric lymph was collected into a centrifuge tube on ice and centrifuged at 400 g for 15 min to remove the cells, and the supernatant was stored at −80°C.

T/HS or trauma-sham shock.

As previously described (11), a heparinized polyethylene (PE-50) catheter was introduced into the left femoral artery for blood pressure monitoring, and a 50-gauge silicone catheter was inserted into the right jugular vein for blood withdrawal and resuscitation. In the T/HS rats, hemorrhagic shock was induced by withdrawal of blood through the jugular vein catheter into a heparinized syringe (100 U/kg) until the mean arterial blood pressure reached 30 mmHg. The mean arterial blood pressure was maintained at 30–35 mmHg for 90 min by withdrawal or reinfusion of the shed blood (kept at 37°C) as required. At the end of the 90-min shock period, the T/HS animals were resuscitated by reinfusion of their shed blood. Lymph was collected separately before, during, and for 3 h after (post) shock. The trauma sham-shock (T/SS) animals underwent similar cannulation and laparotomy (trauma), but no blood was withdrawn, and the animals were not resuscitated.

SMAO.

After laparotomy, the superior mesenteric artery was separated near its origin with the aorta and reversibly occluded for 45 min using 4-0 suture. Immediate blanching of the small intestine and cecum verified that the blood supply to these intestinal segments had been completely shut off. Then the intestine was returned to the abdomen, which was temporarily closed with a hemostat. After 45 min of SMAO, the suture was removed, reperfusion of the intestinal segments was confirmed, and the abdominal incision was closed with sutures. Lymph was collected separately before, during, and for 3 h after SMAO.

Cytotoxicity of the lymph to the endothelial cells.

Human umbilical vein endothelial cells (HUVECs, Clonetics/BioWhittaker, Walkersville, MD) were grown at 37°C in 95% air-5% CO₂ using complete medium (EGM Bullet Kit, Clonetics/BioWhittaker), which was prepared by the addition of EGM SingleQuots, which contains 0.5 ml (10 ng/ml) of human recombinant EGF, 0.5 ml (1.0 mg/ml) of hydrocortisone, 0.5 ml (50 mg/ml) of gentamicin, 50 μg/ml amphotericin B (CC-4081), 2 ml (3 mg/ml) of bovine brain extract, and 10 ml of FBS to 500 ml of endothelial cell basal medium. The cells were harvested between passages 2 and 6, seeded at 2 × 10⁴ cells per well in 96-well plates, and grown to confluence over a 24-h period. The culture medium was removed and replaced by medium or medium containing lymph over a 5–50% (vol/vol) dose-response curve. After incubation for 3 or 18 h, the cytotoxicity of the lymph was determined by measurement of the viability of the cells using the mitochondrial tetrazolium [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)] test. Briefly, after certain times of incubation, the medium containing the lymph was removed by aspiration and replaced with 100 μl of endothelial cell basal medium (without phenol red) with 0.25 mg/ml MTT. After incubation at 37°C for 120 min, the cells were solubilized in DMSO (100 μl), and the extent of reduction of MTT to formazan within cells was quantitated at 570 nm using a microplate reader (Spectramax 250).

Measurements of FFA, triglycerides, and protein in the lymph.

FFAs and triglycerides in the lymph and serum were measured using kits from Wako Pure Chemical Industries and Randox Laboratories, respectively. Protein in the lymph was measured using reagents and standards from Bio-Rad Laboratories. All assays were done according to the protocols provided by the manufacturers.

Assessment of the role of FFAs in the cytotoxicity of postshock lymph.

The role of FFAs in the cytotoxicity of postshock lymph was evaluated by comparing the cytotoxicity of postshock lymph with that of lipase-treated sham-shock lymph, where levels of FFAs similar to those of the postshock lymph specimens were produced. Equal amounts of lymph from five sham animals were pooled to represent sham lymph. Then pancreatic lipase (Sigma, porcine type IV-S, prepared in 5 mM CaCl₂) was added to aliquots of the pooled sham lymph to a final concentration of 0.1 mg (1,500 U)/ml to generate FFAs in the range comparable to postshock lymph. After 15–30 min of incubation at 37°C with pancreatic lipase, the reaction was stopped by the addition of orlistat, the pancreatic lipase inhibitor, at a final concentration of 2 μM. The doses of lipase and orlistat were selected on the basis of pilot experiments examining the dose response and time course of FFA generation of sham lymph by lipase, as well as the inhibitory efficiency of orlistat on lipase. In these pilot studies, HUVEC cytotoxicity was monitored. Then the cytotoxicity of the postshock lymph and lipase-treated sham lymph with different amounts of FFAs was tested on HUVECs, and the concentration that caused 50% cell death (LD₅₀) was used as the toxicity index. Then the role of FFAs in the cytotoxicity of postshock lymph was assessed by correlation and regression analyses of the characteristics of cytotoxicity of the postshock lymph and lipase-treated sham lymph.

Analysis of FFA in the lymph by gas chromatography.

Lipids in the lymph were extracted using the Folch method, and FFAs were separated from the other lipids by thin-layer chromatography, as described in our previous study (28). Then the FFAs were methylated by boron trifluoride in methanol, extracted with hexane, and analyzed by gas chromatography (25). The gas chromatograph (model GC-2014, Shimadzu Scientific Instruments, Columbia, MD) was equipped with an autoinjector (Shimadzu). The column (model DB-23, 123-2332) was 30 m long, with a 0.32-mm-ID bore and a film thickness of 0.25 μm (J & W Scientific, Folsom, CA). The column temperature of the gas chromatograph was maintained at 120°C for 1 min and then increased at 5°C/min from 120–240°C. The temperature of the injector and flame ionization detector was 250°C. A split (8:1) injection mode was used. The carrier gas was helium, with a column flow rate of 2.5 ml/min. Fatty acid was identified using retention times of authenticated fatty acid methyl ester standards (Matreya, Pleasant Gap, PA), and quantification was based on areas using Shimadzu Class VP 4.3 software.

Statistics.

Data were analyzed using SAS software (SAS Institute, Carey, NC). Difference between two groups was tested by Student's t-test. Differences among multiple groups were analyzed by one-way ANOVA with Tukey's post hoc multiple comparison test. P < 0.05 was considered statistically significant. The two regression lines based on the FFA and cytotoxicity of the postshock and lipase-treated sham lymph were compared by analysis of covariance. Values are means ± SE.

RESULTS

Post-T/HS and post-SMAO lymph has increased endothelial cell cytotoxicity.

A significant decrease in cellular viability was observed in HUVEC cultures incubated for 3 h with post-T/HS lymph at a final concentration of 5% (vol/vol) (Fig. 1A). In contrast to the post-T/HS lymph samples, neither T/SS lymph, nor pre-T/HS lymph, nor T/HS lymph collected during the 90-min T/HS period was cytotoxic. To determine whether lymph collected from rats subjected to SMAO was cytotoxic, viability of HUVEC cultures incubated with pre- or post-SMAO samples for 3 or 18 h was measured over a 5–50% (vol/vol) dose-response curve of lymph. The minimum concentration of post-SMAO lymph required to cause a significant decrease in HUVEC viability was 15% when the incubation period was 3 h (Fig. 1B). However, if the incubation period was prolonged to 18 h, a threshold level of cytotoxicity was observed, even at a lymph concentration of 5%, and the magnitude of cytotoxicity increased as the lymph concentration increased. Thus post-T/HS and post-SMAO lymph samples were cytotoxic for HUVECs, although the magnitude of cytotoxicity was less in the post-SMAO than post-T/HS lymph samples.

Fig. 1.

A: viability of human umbilical vein endothelial cells (HUVECs) before, during, and after (post) incubation in trauma-hemorrhagic shock (T/HS) and trauma-sham shock (T/SS) mesenteric lymph at a final concentration of 5% at 37°C for 3 h. Values are means ± SE (n = 5). *P < 0.05 vs. all other groups (by ANOVA). B: effect of 5–50% dose-response curve of pre- and post-superior mesenteric artery occlusion (SMAO) mesenteric lymph on viability of HUVECs after incubation at 37°C for 3 or 18 h. Values are means ± SE (n = 5). *P < 0.05 vs. pre-SMAO lymph samples at that concentration.

FFAs, but not protein or triglycerides, were significantly increased in the mesenteric lymph of rats after shock or SMAO.

To test the hypothesis that lymph cytotoxicity was associated with elevated lymph FFA levels, we measured FFA, protein, and triglyceride levels in lymph samples from rats subjected to T/HS or SMAO (Fig. 2). Protein and triglyceride levels were similar between the lymph samples collected before, during, and for 3 h after T/HS. Similarly, there was no difference in protein or triglyceride levels between the pre-SMAO and post-SMAO lymph samples. However, the FFA levels were significantly elevated in the post-T/HS and post-SMAO lymph samples, with a greater increase in the T/HS lymph samples (Fig. 2C). A time-course study of lymph FFA levels documented that the FFA levels were increased in the T/HS, as well as SMAO, lymph samples as early as 1 h after the end of the T/HS or SMAO period (Fig. 2D). However, consistent with the greater cytotoxic activity of the T/HS than SMAO lymph, the FFA levels continued to increase for the next 2 h and remained elevated during the entire 6-h collection period in the T/HS lymph samples, while they rapidly returned to baseline levels in the SMAO lymph samples.

Fig. 2.

Amounts of protein (A), triglycerides (B), and free fatty acids (FFA, C) in lymph of rats subjected to trauma-hemorrhagic shock (T/HS), trauma-sham shock (T/SS), or SMAO. Values are means ± SE (n = 5–8 lymph samples per group). Amounts of protein and triglycerides were similar among the groups. FFA levels in lymph of rats with T/HS or SMAO not sharing the same letters (a, b, c) are significantly different. D: dynamic changes of FFAs in mesenteric lymph of rats with T/HS or SMAO. Values are means ± SE. *P < 0.05 vs. pre-T/HS or SMAO (by ANOVA).

Cytotoxicity of mesenteric lymph is more closely related to the FFA-to-protein ratio than to the absolute FFA level.

The ability to accurately correlate T/HS lymph FFA levels and cytotoxic activity was confounded by the fact that the majority of the HUVECs were killed after 3 h of incubation with 5% T/HS lymph. Thus we determined the dose of T/HS lymph required to kill 50% of the HUVECs (LD₅₀) (Fig. 3) and correlated FFA, protein, and triglyceride levels of each T/HS lymph sample with its LD₅₀ value. A similar LD₅₀ value was determined for SMAO lymph samples. Since certain proteins, such as albumin, bind FFA and, by binding them, reduce their biological activity (22), we also evaluated the FFA-to-protein ratio. We observed a significant and similar inverse correlation between the FFA-to-protein ratio and the percentage of lymph required to cause an LD₅₀ cytotoxicity in the T/HS and SMAO lymph samples (Table 1).

Fig. 3.

Calculation of LD₅₀ (amount of lymph required to kill 50% of HUVECs) of T/HS lymph after a 3-h incubation with different concentrations of pooled T/HS lymph samples. Values are means ± SE; experiment was run twice.

Table 1.

Correlation between protein, triglycerides, FFAs, and FFA-to- protein ratio with cytotoxicity of T/HS and SMAO lymph

	Correlation Coefficient
	T/HS lymph	SMAO lymph
Protein	0.226	0.350
Triglycerides	0.133	0.019
FFA	−0.645	−0.462
FFA-to-protein ratio	−0.891*	−0.868*

Values represent correlation between the amount of a specific component in the lymph and the percent concentration of lymph required to cause an LD₅₀ cytoxicity (n = 5–8).

T/HS, trauma-hemorrhagic shock; SMAO, superior mesenteric artery occlusion; FFA, free fatty acid.

^*P < 0.05.

Since the FFA-to-protein ratio correlated with cytotoxicity, we next tested the hypothesis that decreasing the FFA-to-protein ratio of post-T/HS lymph would reduce the cytotoxic activity of the lymph samples. We increased the protein content of the T/HS lymph samples by ∼50% by adding delipidated BSA (10 mg/ml) to the lymph samples and then measured cytotoxicity. Albumin was used, since albumin is the primary scavenger of FFA in vivo, and each albumin molecule can bind up to 13 FFA molecules (34). The addition of albumin, but not the control non-FFA-binding molecule ovalbumin, totally abrogated the cytotoxicity of the T/HS lymph samples after 3 or 18 h of incubation (Fig. 4). To further examine the importance of FFA binding by albumin, we also tested the notion that a lipid-containing albumin would be less effective than delipidated albumin in neutralizing postshock lymph cytotoxicity. In this experiment, we used BSA (Sigma) to which four FFAs were bound (2 molecules of oleic acid and 2 molecules of linoleic acid). While this fatty acid-bound albumin abrogated the cytotoxicity of postshock lymph after 3 h of incubation, its protective effect was lost when the incubation period of the lymph with the HUVECs was extended to 18 h (Fig. 4).

Fig. 4.

Effect of albumin (delipidated), fatty acid-bound albumin (FA-albumin), and ovalbumin (10 mg/ml) on cytotoxicity of T/HS lymph in HUVECs after 3 or 18 h of incubation. Values are means ± SE (n = 4–6 samples per group). *P < 0.001 vs. other groups.

Lipase-induced increases in FFA in sham-shock lymph resulted in cytotoxicity comparable to that in post-T/HS lymph, suggesting that FFAs are the key cytotoxic factors in T/HS lymph.

On the basis that increased FFAs in the T/HS lymph samples were mainly responsible for its cytotoxicity, we hypothesized that nontoxic sham-shock lymph would become cytotoxic if sufficient amounts of FFAs were released from chylomicrons in the lymph samples. As a first step in testing this hypothesis, we determined the effects of a dose response (Fig. 5A) and time course (Fig. 5B) of pancreatic lipase on FFA levels in sham-shock lymph specimens. We found that a 20-min incubation of sham-shock lymph with pancreatic lipase (0.1 mg/ml) generated FFA levels comparable to those in postshock lymph (Fig. 5B). However, since the presence of active lipase in the lymph would continue to generate more FFA after the lipase-containing lymph was added to the HUVEC culture and confound an accurate evaluation, we next determined the amount of the lipase inhibitor orlistat that would be required to inhibit lipase activity. A dose-response curve evaluating the effect of orlistat on lipase activity showed that 0.2 μM orlistat would be sufficient to inhibit the 0.1 mg/ml lipase activity (Fig. 5C). On the basis of these results, we tested the ability of lipase-treated sham-shock lymph to kill HUVECs. Orlistat was added to the lymph samples just before or 20 min after addition of the lipase. Addition of orlistat prior to the lipase blocked lipase activity and, hence, FFA generation, while addition of orlistat after 20 min of incubation with lipase prevented continued FFA generation. Neither lipase nor orlistat at the doses tested had cytotoxicity in the absence of sham-shock lymph (99 ± 5% HUVEC viability). Addition of lipase to sham-shock lymph resulted in these nontoxic sham-shock lymph samples becoming highly toxic, and this lipase-induced cytotoxicity was completely abrogated by pretreatment of the lymph samples with orlistat prior to addition of lipase (Fig. 5D).

Fig. 5.

A: generation of FFAs in pooled sham lymph by different doses of pancreatic lipase. Lipase was prepared in 5 mM CaCl₂ and added at 1:20 (vol/vol) to pooled sham lymph to final concentrations of 0–0.4 mg/ml and incubated for 20 min at 37°C, and FFAs in lymph were determined by chemical assay. Values are means ± SE of 2 experiments. B: time course of FFA generation by pancreatic lipase. Lipase was added to pooled sham lymph to a final concentration of 0.1 mg/ml. At 0–40 min, an aliquot of the lymph was obtained, and FFAs were determined by chemical assay. Values are means ± SE of 2 experiments. C: inhibition of lipase by different doses of orlistat. Orlistat (dissolved in DMSO) was added at 1:20 (vol/vol) to aliquots of pooled sham lymph to a final concentration of 0.002–200 mM. Samples were mixed, lipase was added to the lymph to a final concentration of 0.1 mg/ml, and samples were incubated at 37°C. By the end of the 30-min incubation, FFAs in lymph were measured by chemical assay. Values are means ± SE of 2 experiments. D: effect of T/SS lymph with or without lipase or lipase + orlistat on viability of HUVECs after 3 h of incubation. Addition of lipase resulted in the nontoxic T/SS lymph samples becoming toxic, and this toxicity-inducing ability of lipase was totally abrogated by pretreatment with orlistat. Values are means ± SE of 3 experiments. *P < 0.05 vs. other groups.

Using different incubation times of the sham-shock lymph samples with lipase followed by orlistat treatment, we generated sham-shock lymph samples containing 8.5–17 mM FFA. The increased FFA levels generated in the T/SS lipase-treated lymph samples were similar to those in the actual T/HS lymph samples (Fig. 6A), while the levels of triglycerides, cholesterol, and phospholipids were not affected by lipase treatment and were similar between the groups (data not shown). When the LD₅₀ values of these sham lymph samples with different amounts of FFA were tested on HUVECs, a significant correlation (r² = −0.83, P = 0.032) between the FFA-to-protein ratio and cytotoxicity was observed (Fig. 6B). Importantly, when the LD₅₀ curves of the lipase-treated sham-shock lymph specimens were compared with those of the actual post-T/HS lymph specimens, they were virtually identical (Fig. 6B). This observation supports the concept that increased FFAs are largely responsible for the cytotoxicity in postshock lymph.

Fig. 6.

A: lipase treatment of T/SS lymph samples results in an increase in total FFA similar to that of T/HS lymph. Values are means ± SE (n = 6 lymph samples per group). *P < 0.05 vs. T/SS. B: relationship between cytotoxicity [as expressed by lymph concentration to cause 50% HUVEC cell death (LD₅₀)] and FFA-to-protein ratio in T/HS lymph (●, solid line) and lipase-treated sham-shock lymph (○, dashed line) samples (n = 5 lymph samples per group).

Identification of FFAs responsible for T/HS lymph cytotoxicity.

The lipid composition of post-T/HS, T/SS, and lipase-treated T/SS lymph samples was analyzed by gas chromatography. The concentrations of triglycerides, cholesterol, and phospholipids were similar between the post-T/HS, T/SS, and lipase-treated T/SS lymph samples (data not shown). However, four fatty acid species [palmitic (16.0), stearic (18.0), oleic (18.1), and linoleic (18.2)] were present at higher concentrations in the T/HS and lipase-treated T/SS than in the T/SS lymph samples (Fig. 7).

Fig. 7.

Concentration of different FFAs in T/HS, T/SS, and lipase-treated T/SS lymph samples. Values are means ± SE (n = 5 lymph samples per group). *P < 0.05 vs. T/SS.

To directly test whether the increase in any of these four FFAs could account for the cytotoxicity of post-T/HS lymph, each of these FFAs was added to the medium at a concentration equivalent to 10% T/HS lymph; then HUVEC cultures were incubated with the FFA-supplemented medium. All four of these FFAs had cytotoxic activity, but the greatest activity was exerted by linoleic acid (Fig. 8A). To further investigate the potential cytotoxicity of these FFAs, we supplemented T/SS lymph specimens with the FFAs. Cytotoxicity of the FFAs was less when the individual FFAs were added to the complex biological T/SS lymph fluid than when they were added to the medium (Fig. 8B). However, when combinations of FFAs that contained linoleic acid were added to the T/SS lymph specimens, cytotoxicity reached that of the T/HS lymph samples (Fig. 8B).

Fig. 8.

A: effect of palmitic (P), stearic (S), oleic (O), and linoleic (L) acids added to culture medium at a concentration equal to that in 10% T/HS lymph on viability of HUVECS after incubation at 37°C for 18 h. B: cytotoxic effects of FFA alone or in combination when added to T/SS lymph samples in sufficient amounts to raise their levels to that observed in 10% T/HS lymph samples. Viability of HUVECS was measured after incubation at 37°C for 18 h. Values are means ± SE; samples were run 3 separate times. Values not sharing the same letter (a, b, c) are significantly different from each other.

DISCUSSION

The current study supports the role of lipid factors, especially certain unsaturated FFAs, in T/HS and SMAO lymph as putative endothelial cytotoxic factors. This conclusion is based on three major lines of evidence: 1) cytotoxic T/HS and SMAO lymph samples had elevated FFA levels, and the FFA-to-protein ratio of individual T/HS lymph samples directly correlated with cytotoxicity; 2) addition of lipase to the noncytotoxic T/SS lymph samples resulted in these nontoxic lymph samples becoming cytotoxic; and 3) medium or T/SS lymph specimens supplemented with the specific FFAs found to be increased in the T/HS and lipase-treated T/SS samples were able to kill HUVECs. Furthermore, it appeared that the FFA-to-protein ratio was more important than the absolute FFA levels. This notion was supported by studies showing that the FFA-to-protein ratio in the T/HS and SMAO lymph samples correlated better with cytotoxicity than did the absolute FFA levels and that the addition of albumin to the cytotoxic T/HS lymph samples abrogated its cytotoxicity. This protective effect of albumin is consistent with the previously described physiological role of albumin as a lipid scavenger, to which most plasma FFAs are bound, as well as with studies indicating that FFAs bound to albumin have reduced biological activity (15, 34). Thus, although it appears that increased levels of certain FFAs are involved in the cytotoxic activity of T/HS, SMAO, and lipase-treated T/SS lymph samples, certain caveats must be considered: 1) since albumin scavenges molecules other than FFAs, part of albumin's protective effect may be related to its scavenging of non-FFA molecules; and 2) we have not directly measured the amount of FFA bound to albumin in the lymph samples and, thus, do not know if the degree of albumin saturation with FFA differs between the T/SS and T/HS lymph specimens.

The mechanisms by which T/HS and SMAO lead to increased FFA levels in intestinal lymph remain to be fully defined but appear to involve an increase in lymph levels of lipase activity, since our previous T/HS studies document that lymph lipase activity is increased (7, 31). Several factors could contribute to this increase in lipase activity in the T/HS and SMAO lymph specimens: 1) a stress-induced increase in hepatic and lipoprotein lipase production, 2) absorption of pancreatic lipase across the injured gut mucosa, and 3) the lipase-inducing effects of heparin (7, 31). This increase in lymph lipase activity would in turn result in the liberation of FFAs from chylomicrons in the lymph and, thereby, increase lymph's cytotoxic capacity.

The observation that FFAs can be cytotoxic under certain circumstances has been reported previously (7, 8, 35). Chung et al. (7) reported that lipolytically released FFAs from the plasma of normal volunteers fed a fatty meal are cytotoxic. They also found that treating hypertriglyceridemic serum in vitro with purified lipoprotein lipase resulted in partitioning of FFAs onto lipoproteins and that these FFA-enriched lipoprotein fractions were cytotoxic for mouse peritoneal macrophages (8). However, the liberated FFAs that bound to albumin did not manifest cytotoxicity (8). This observation further highlights the importance of albumin as a lipid scavenger and the influence of the level of albumin and its lipid-binding capacity, as well as the absolute FFA levels, on FFA-induced cytoxicity. Penn et al. (30) found that lipid cytotoxic factors were produced in the intestine of rats subjected to a severe model of intestinal ischemia.

Numerous possible biologically active components in T/HS or SMAO lymph could contribute to its cytotoxicity: 1) proinflammatory molecules released from inflammatory cells (e.g., neutrophils, macrophages, lymphocytes, and mast cells) or activated plasma factors reaching the interstitium of the gut, 2) endogenous danger-signaling molecules or alarmins (3, 20) released from damaged epithelial and other parenchymal cells, 3) alarmins produced by the degradation of interstitial or cellular molecules by digestive enzymes crossing the injured mucosal barrier, 4) bacteria or bacterial products, or even 5) dietary products. Studies to date have excluded many of these factors, including bacteria, bacterial products, cytokines, certain danger-associated molecular pattern molecules, activated complement, and others, as being responsible for the cytotoxic activities of T/HS lymph (1, 9, 32). Thus, while certain unsaturated FFAs have been implicated as major contributors to the endothelial toxicity of T/HS and SMAO lymph, it remains feasible that T/HS lymph contains non-FFA cytotoxic factors. This possibility is supported by studies showing that modified albumin species in T/HS lymph have cytotoxic activity (22). Furthermore, T/HS lymph from Arixtra-treated rats, in which neither T/HS lymph lipase activity nor FFA levels were increased, retained cytotoxic activity, although the magnitude of cytotoxic activity was greatly reduced compared with heparin-treated rats (13). In considering the question of possible non-FFA cytotoxic factors in T/HS lymph, valuable information is gained from a comparison of the two regression lines between the lipase-treated T/SS and T/HS lymph samples, illustrating the relationship between the FFA-to-protein ratio and the LD₅₀ of the lymph samples (Fig. 6). Since the only action of lipase would be to increase lymph FFA levels, the fact that the slope and regression lines of the T/HS and lipase-treated T/SS lymph samples are very similar suggests that the increase in FFA is a major factor in the cytotoxicity of T/HS lymph. Thus, although T/HS lymph may contain non-FFA cytotoxic factors, it appears that the post-T/HS increase in FFA is an important factor in the in vitro cytotoxicity of T/HS lymph.

The mechanism by which these unsaturated fatty acids killed the HUVECs was not directly investigated in this study. However, Choi et al. (6) found that palmitic acid can induce apoptosis of HUVECs and that the mechanism involves the dephosphorylation and activation of glycogen synthase kinase-3β. Similarly, Hufnagel et al. (21) found that treating lipoprotein with lipoprotein lipase liberated FFAs, especially linoleic and oleic acids, which induced the apoptosis of HUVECs by activating serine/threonine protein phosphatase type 2C and dephosphorylation of Bad. Others have found that treating postprandial plasma triglyceride-rich lipoprotein (VLDL and chylomicrons) with lipoprotein lipase results in increased levels of saturated and unsaturated FFAs, most pronouncedly linoleic acid and its hydroxylated metabolites, which are capable of inducing TNF-α and reactive oxygen species production in human aortic endothelial cells (36). In this context, Chung et al. (7) found that the lipolytic remnant products of triglyceride-rich lipoproteins produced after a meal rich in polyunsaturated fat are more injurious to arterial wall cells than those produced after a meal rich in saturated fat.

In summary, our studies have identified increased lymph FFA levels as significantly contributing to the in vitro cytotoxic effects of lymph from rats subjected to acute insults associated with an intestinal ischemia-reperfusion injury. However, this notion that lipase-generated FFAs may have deleterious effects does not seem to be limited to acute injury models. Studies have shown that the postprandial generation of proinflammatory and cytotoxic FFAs by lipoprotein lipase appears to be involved in a variety of inflammatory and endothelial cell disease-based processes, including atherosclerosis (19, 38). If we extrapolate our preclinical results with human studies documenting the biological effects of lipoprotein lipases on plasma lipid species, it is not unreasonable to question whether a patient's plasma lipid profile might be a factor in the development of acute injury- and/or stress-induced systemic inflammation or MODS. In support of this speculation is a report that increased serum levels of 18-carbon unsaturated FFAs were a predictor of the acute respiratory distress syndrome in at-risk intensive care unit patients (4).

GRANTS

This study is supported by National Institutes of Health Grants GM-59841 and T32 069330.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

X.Q., P.T., and E.A.D. are responsible for conception and design of the research; X.Q., W.D., S.M.S., S.U.S., D.C.P., T.R., and R.J.J. performed the experiments; X.Q. and E.A.D. analyzed the data; X.Q. and E.A.D. interpreted the results of the experiments; X.Q. and E.A.D. prepared the figures; X.Q., P.T., and E.A.D. drafted the manuscript.