Hypoxia causes ischemic bowel necrosis in rats: The role of platelet-activating factor (PAF-acether)

GASTROENTEROLOGY 1990;99:979-999

Hypoxia Causes Ischemic Bowel Necrosis in Rats: The Role of Platelet-Activating Factor (PAF-Acether) MICHAEL

S. CAPLAN,

XIAO-MING

SUN,

and WE1 HSUEH

Department of Pathology, Children’s Memorial Hospital, and Department of Pediatrics, Evanston Hospital, Northwestern University Medical School, Chicago, Illinois

We have previously shown that injection of plateletactivating factor causes necrotizing enterocolitis in the rat and that platelet-activating factor is an endogenous mediator in lipopolysaccharide-induced bowel necrosis. Because hypoxia is a known predisposing factor for neonatal necrotizing enterocolitis, we investigated the effect of hypoxia on platelet-activating factor formation and intestinal necrosis. Young male Sprague-Dawley rats were made severely hypoxic by placing them in a 100% N, chamber for 2 minutes; moderate hypoxia was accomplished using 10% 0, for 15 or 30 minutes. To evaluate the role of plateletactivating factor on intestinal perfusion and injury, two platelet-activating factor antagonists, SRI 63-441 and WEB 2066, were injected 10 minutes before the hypoxic exposure. We found that plasma plateletactivating factor levels were significantly elevated after 2 minutes of severe hypoxia (13.6 f 2.9 ng/mL vs. control 2.1 f 0.6 ng/mL) and after 30 minutes of moderate hypoxia (41.1 + 11.7 ng/mL). This increase in platelet-activating factor level was not caused by decreased degradation, because neither plasma nor intestinal platelet-activating factor acetylhydrolase was decreased in the hypoxic rats. (Intestinal acetylhydrolase activity was actually increased.) Intestinal perfusion was markedly decreased at 30 minutes in hypoxic animals. In contrast, all platelet-activating factor antagonist-treated animals had normal intestinal perfusion. Histological examination of affected bowel from hypoxic animals showed early intestinal necrosis which was completely prevented by pretreatment with SRI 63-441 and WEB 2066. Because 30 minutes of hypoxia also resulted in metabolic acidosis, we further investigated if acidosis alone could induce platelet-activating factor release and bowel injury. We found that acidosis alone resulted in moderate increase of plasma platelet-activating factor but did not produce bowel injury. We conclude that platelet-activating factor plays a central role in

mediating hypoxia-induced intestinal necrosis. Acidosis may enhance the effect of hypoxia on plateletactivating factor production.

N

ecrotizing enterocolitis (NEC, ischemic bowel disease in neonates), is a potentially lethal disease affecting l%-7% of all premature infants (1,2). Despite the intensive study, its etiology remains unclear. Hypoxia has been associated with NEC in epidemiological surveys [1,2) and is implicated as an important factor in the pathogenesis of this disease (3). Previous investigators have devised animal models to study the effect of hypoxia on the development of ischemic bowel necrosis (3-6). It is suggested that hypoxia results in decreased intestinal perfusion which may lead to bowel necrosis. However, the mechanism of this decrease in perfusion is unknown. Plateletactivating factor [also called PAF, PAF-acether, or acetylglyceryletherphosphorylcholine) is an endogenous phospholipid with potent biological effects, including platelet and neutrophil aggregation and degranulation, systemic hypotension, pulmonary hypertension, and capillary leakage (7,8). We have previously reported that infusion of PAF into the mesenteric artery in the rat results in intestinal hypoperfusion and ischemic bowel necrosis (91. In addition, endotoxin causes PAF production and ischemic bowel necrosis, which can be prevented by pretreatment with PAF antagonists (10). These observations suggest that PAF is an endogenous mediator for ischemic bowel necrosis. Thus, both PAF and hypoxia seem

Abbreviations used in this paper: ABG, arterial blood gas; LP!S, lipopolyaaccharide; NEC, necmtizing enterocolitis; PAF, plateletactivating factor (PM-acetherk PRP, platelet-rich plasma; TNF, tumor necrosis factor; WBC, white blood cell. 0 1999 by the American Gastmantemlogical Association 9916.5065/9O/S3.00

980 CAPLAN ET AL.

implicated in the pathogenesis of NEC. The present study attempts to elucidate the relationship between PAF and hypoxia and investigates the effect of hypoxia on PAF production and metabolism and the role of PAF and hypoxia in the development of NEC. Materials and Methods Male 25-30-day-old Sprague-Dawley rats (Harlan Sprague-Dawley), weighing 55-90 g, were used for all experiments. Animals were anesthetized with pentobarbital (65 mg/kg IP), and polyethylene catheters were placed in the carotid artery for blood sampling and blood pressure recording and in the jugular vein for drug injection. “Acute” (severe] hypoxia was accomplished by placing the animal in a 100% N, chamber for 2 minutes, whereas “subacute” (moderate) hypoxia was effected using 10% O,, 15% CO,, and 75% N, for 15 or 30 minutes. Acidosis was produced after IV injection of ammonium chloride (NH&l, 300 mg/kg) in four divided doses during 15 minutes: the animals were studied at 30 minutes. At the end of each study period, blood was sampled for plasma PAF level, acetylhydrolase (the enzyme degrading PAF) activity, arterial blood gas (ABG) determination, and white blood cell (WBC) count and hematocrit value. At the end of 30 minutes of hypoxia and acidosis, the abdomen was opened and the intestine was exposed. The extent of intestinal perfusion was evaluated by IV injection of Evans blue dye (2 g/L, 0.5 mL). The Evans blue dye method is a qualitative technique that identifies the presence or absence of organ blood ilow by staining perfused tissue (ll,l2). Our previous studies have suggested that vasoconstriction of the microvasculature (13) and consequent ischemia (11) are the initial events which precede the subsequent event of intestinal necrosis. Because the dye was injected at the end of the experiment, presumably after vasoconstriction, the unstained areas were to represent areas of hypoperfusion and ischemia. Sections were taken from these areas, and tissue injury was subsequently confirmed by histological examinations. Our previous studies have shown that early, mild lesions often escape detection by the naked eye (11) and could only be identified by the dye injection method (131. Furthermore, there is an excellent correlation between good perfusion and normal histology, i.e., Evans blue dye perfused areas invariably show normal histology (11). However, sections from hypoperfused intestine may vary from normal histology to moderate necrosis. The percent of perfused bowel was assessed by measuring the length of normal intestine, dividing by the total intestinal length, and multiplying by 100. Representative intestinal sections were taken from hypoperfused areas and fixed in 10% formalin for histological examination, according to a previously described method (11). Briefly, tissue was embedded in paraffin, stained with H&E, and evaluated by light microscopy. Three sections were studied from hypoperfused areas of intestine from affected animals or from random areas of normally perfused intestine by a pathologist unaware of the code. Microscopic lesions were graded as follows: minimal, separation of the surface epithelial cells from lamina propria; mild, epithelial cell necrosis confined to the tips of the villi; and moderate, partial loss of intestinal villi.

GASTROENTEROLOGY Vol. 99, No. 4

To test the efficacy of PAF antagonists in preventing hypoxia-induced intestinal injury, SRI 63-441 [a generous gift from Dr. D. Handley, Sandoz Research Institute, East Hanover, NJ; 2.5 mg/kg IV] and WEB 2086 (a generous gift from Dr. H. Heuer, Boehringer Ingelheim, Mainz, Germany; 0.5 mg/kg IV] were given 10 minutes before the hypoxic challenge. SRI 63-441(14,15) and WEB 2086 (16) are structurally different PAF receptor antagonists. At the doses used, these drugs have no effects on systemic blood pressure or mesenteric blood flow in adult rats (our unpublished observation]. Blood was sampled from control, 2-minute, 15-minute, and 30-minute groups with hypoxia and acidosis for PAF determinations. Briefly, 0.5 mL blood was added to 0.05 mL citric acid to prevent clotting and inhibit acetylhydrolase activity. The sample was rapidly centrifuged in a microfuge, and 0.2 mL plasma was added to 4 mL of chloroform and methanol, 2:1 [vol/vol]. Next, 10’ cpm of [‘H]PAF (81 Ci/ mmole; Amersham Corporation, Arlington Heights, IL] were added to the sample for calculation of extraction efficiency, and the total lipids were extracted by Folch’s method (17). The extracted lipid was plated on a thin layer plate coated with silica gel G and developed in solvent system chloroform, methanol, and water, 65:35:6 (vol/vol/vol). The zone comigrating with standard PAF was scraped and eluted with chloroform, methanol, and water 1:2:0.8 [vol/vol/vol] and dried in N,. An aliquot was counted for calculation of recovery, and another aliquot was reconstituted in 0.5 mL phosphate-buffered saline with 5 mg/mL bovine serum albumin and assayed for PAF activity. The PAF was assayed using the rabbit platelet serotonin release method, according to previously described method [IO). Briefly, 1 &i of [3H]serotonin (10 Ci/mmol; Amersham, Arlington Heights, IL) was added to 1 mL of platelet-rich plasma (PRP) for 30 minutes. Five sets of tubes were prepared: (1) PRP (0.5 mL) and vehicle only, at 0°C with Triton X-100 [Sigma, St. Louis, MO] added at the end to lyse the platelets: (2) same as tube 1, except that no Triton-X was added: (3) PRP and sample, stirred for 5 minutes at room temperature; (4) PRP and vehicle, stirred for 5 minutes; (5) same as tube 3 except that 10 pg/mL SRI 63-119 was also added (to confirm that the active compound in the solution was PAF) (10). All tubes were then centrifuged, and the supernatants were counted. Serotonin release = (3 - 4)/(1 - 2) x 100%. Serotonin uptake = (1 - 2)/l x 100%. A standard curve using different concentrations (100 pg/mL to 1 ng/mL) of PAF was constructed with each assay, and the PAF concentration in the samples was determined by comparison with the standard curve. Data were considered reliable when the PAF concentration in the diluted sample was between 200-300 pg/mL and 700-800 pg/mL (in the straight portion of the curve]. Samples were reassayed using different dilutions when the concentration was beyond the range of these values. The intraassay variability was usually less than 10%. Acetylhydrolase is a PAF-specific enzyme that catalyzes the degradation of PAF to biologically inert lyso-PAF. Acetylhydrolase activity was assayed by using a radiolabeled substrate and by measuring reaction products. For plasma acetylhydrolase, IO5cpm of l-O-[SH]alkyl-2-acetyl-3phosphorylcholine, 80 pmol/L PAF, 0.1 mol/L HEPES buffer (pH 7.2), and 10 PL plasma (l:l-dilution with saline) in a final

HYPOXIA, PAF, AND TISSUE INJURY

October 1999

volume of 60 rL was incubated at 37°C for 30 minutes. To evaluate intestinal acetylhydrolase, control animals were compared with rats which were made hypoxic for 30 minutes. The entire small intestine was recovered, washed in ice cold saline, and homogenized in HEPES buffer (pH 7.2, 0.1 mol/L, 1:5 vol/vol) on ice. This sample was again diluted 1:5 with HEPES buffer, and a 20 rL aliquot was sampled for enzyme activity. Preliminary experiments were performed to determine the kinetics of acetylhydrolase activity: it was found that the reaction was linear at 30 minutes. The reaction was stopped with 50 CCL glacial acetic acid. Lipids were extracted with chloroform, methanol, and water (1:2:0.8],plated on silica gel G plates, and developed in chloroform/methanol/water, 65:35:6(vol/vol/vol). The zones that comigrated with PAF and lyso-PAF standards were scraped and counted in a scintillation counter. Acetylhydrolase activity was expressed as nanomoles of lyso-PAF formed per milliter per minute for plasma and micromoles per milligram per minute for intestine. The intraassay variability was usually below 5%. All values are expressed as mean + SEM. Means of more than two groups were compared by one-way analysis of variance (ANOVA) using the method of Scheffe to identify specific group differences. To compare two group means, a standard unpaired Student’s t test was used. P values less than 0.05 were considered significant. Results Acute severe hypoxia (mean ABG in pH/Pco,/ Po,/base excess: 7.27/57/18/- 2) occurred 2 minutes after exposure in 100% N, and resulted in an increase in plasma PAF levels from 2.1 + 0.8 ng/mL to 13.8 f 2.9 ng/mL (P < 0.05, Figure 1). Prolonged hypoxic

50

40 2 5 ; d

3o 20 10 0 A

B

C

0

E

Figure 1. Effect of hypoxia on PAP production. Plasma PAP level w= determined from young, anesthetized, and tracheotomized rata treated with: A, air; B, 100% N, for 2 minuteq C, 10% 0,15% CO,, and 75% N, for 15 minutes; D, 10% 0,, 15% CO,, and 75% N, for 30 minutes; or E, NH, Cl (309 mg/kg) for 30 minutes. Blood was collected at the end of the exposure period, and plasma was extracted, purified by TLC, and assayed for PAP activity as described in Materials and Methods. Data are expressed ae the mean * SEM. (See Table 1 for the number of animal8 in each group.) ?? P < 0.05 by one-way ANOVA using Scheffe for group differences.

981

Table 1. Effect of Hypoxia on Arterial Blood Gas Values Group Treatment A

B C

D E

Air 100% N, 10% 0, 10% 0, NH,Cl

Time of exposure

No. of animals

2 min 15 min 30 min 30 min

5 7 5 10 7

“Units for Pco, and PO, are millimeters milliequivalents per liter.

pH/ Pco,/Po,/base" (mean] 7.34/43/96/- 2 7.27/57/18/- 2 6.99/95/23/12 6.80/101/40/- 22 7.05/50/78/18

of mercury: base units are

exposure (10% 0,) at 15 and 30 minutes resulted in low PO, values, and worsening respiratory and metabolic acidosis (Table 1). After 15 minutes of hypoxia, the PAF level was 5.9 f 3.6 ng/mL (P = NS), but at 30 minutes the levels increased markedly to 41.1 f 11.7 ng/mL (P < 0.01, Figure 1). Acidosis for 30 minutes without hypoxia (7.05/50/78/- 16) resulted in increased plasma PAF levels of 28.5 f 9.1 ng/mL (see Figure I]. All animals (n = 14) exposed to hypoxia alone after 30 minutes showed various degrees of abnormal intestinal perfusion by Evans blue dye method (Figure 2). The mean percent of bowel perfused was 40% f 7% and ranged from 0%~95% (Figure 3). In contrast, all hypoxic animals (n = 5 for each group] pretreated with SRI 63-441 and WEB 2086 and animals treated with acidosis alone had normal intestinal perfusion. Histological examination of intestine from animals subjected to hypoxia alone found 6 of 14 with mild or moderate intestinal injury (Figure 4A, Table 2), and 3 of 14 with minimal injury (Table 2). Microscopic specimens from animals treated with acidosis alone and animals pretreated with WEB 2086 and SRI 63-441 pretreated animals showed complete prevention of hypoxia-induced bowel injury (Figure 4B): none of these animals showed any necrotic changes except for one rat in the SRI 63-441 group which showed focal minimal injury. Sections taken from heart, lung, liver, kidney, and spleen of hypoxic animals showed congestion in the lung as the only other abnormality. No necrosis was seen in any organ, except in the bowel. There were no differences in plasma acetylhydrolase activity between control or hypoxia groups at 2, 15, or 30 minutes (73.9 + 1.0, 72.9 f 2.4, 72.9 f 1.0, and 73.4 + 1.4 nmole . mL-’ . min-’ respectively; P = NS). In contrast, after 30 minutes of hypoxia, intestinal acetylhydrolase activity increased from 3.8 f 1.1 pmole - mg-’ . min-l to 5.4 f 0.6 pmole - mg-’ mine1 (P < .05, Figure 5). The WBC count decreased 30 minutes after hypoxia (2900 f 400 vs. control count of 6500 + 450, P < 0.01, Figure 6). Pretreatment with SRI 63-441 partially reversed the hypoxia-induced leukopenia (5100 f 300, P < 0.051, and WEB 2086 completely prevented leuko-

982 CAPLAN ET AL.

GASTROENTEROLOGY Vol. 99, No. 4

Figure 2. Gross appearance of the opened abdomen following Evans blue dye injection from animals (A) subjected to 50 minutes of hypoxia and(B) pretreated with SRI 63-441 before hypoxic challenge.

penia (6900 f 880, P = NS). There were no significant differences in hematocrit values between control, hypoxia, SRI 63-441, or WEB 2086 treated animals (40.0% + 1.2%, 42.0% f l.O%, 39.8% + 2.1%, and 38.4% * 1.5‘70,respectively, Figure 5).

100 1

Discussion

200 tcmtml

Hypoxia

SRI

UEB

Figure 3. Hypoxia-induced hypoperfueion of small intestine and its prevention by PAF antagonists. Hypoxia was accomplished by placing the animal in 10% 0, 15% CO, and 75% N, for 30 minutes. Perfused areas are expressed as a percentage of the total length of the small intestine, as determined by Evans blue injection at the end of the experiment. All animals treated with SRI 63-441 (2.5 mg/kg) and WEB 2058 (0.5 n&kg), given before the hypoxia challenge, showed 100% perfusion, same as the control (sham-operated, air-breathing) animals. (See Table 2 for number of animals.) ?? P < 0.01 compared with control animals.

Necrotizing enterocolitis is the second leading cause of death in neonates; yet, its pathogenesis remains elusive. Clinically, the association of perinatal or neonatal hypoxia with NEC has been well established (1,2). In addition, several experimental models of hypoxia-induced bowel necrosis have been described. Hansbrough et al. reported a dog model that produced hypoxia using 10% 0, for 2 hours (41. Animals were allowed to recover, and 24 hours later, intestinal pathological specimens showed mild ischemic necrosis. Karna et al. worked with newborn piglets, accomplishing hypoxia with 11% 0,, and showed that hypoxia was more important than ischemia or acidosis in producing intestinal hypoperfusion (5). Barlow et al. developed a newborn rat model of hypoxia-induced intestinal necrosis (6). Placing the animals in a plastic bag for 3-5 minutes per day led to

October 1990

HYPOXIA, PAF, AND TISSUE INJURY

983

Figure 4. Microscopic apperuc

ante of the small intestine from animals (A) subjected to 30 minutes of hypoxia alone or (B) pretreated with SRI 6% 441 before hypoxia challenge.

apparent asphyxia and intestinal necrosis after 2-3 days in formula-fed animals. The mechanism of the hypoxia-induced intestinal injury in these models is unclear. Intestinal hypoperfusion has been demonstrated in some of these models (3,s); however, no specific mediator has been identified. Previous studies in our laboratory with the rat model showed a direct association between PAF and ischemic bowel necrosis. We have shown that PAF infused directly into the mesenteric artery produces intestinal hypoperfusion and pathological changes that mimick human NEC (9). In addition, injection of

lipopolysaccharide (LPS) causes increased intestinal levels of PAF and results in ischemic bowel necrosis (10). Finally, pretreating the rat with a PAF receptor antagonist prevents endotoxin-induced bowel necrosis (10). These observations indicate an important role of PAF in the development of NEC. In this report, we document that hypoxia leads to rapid release of PAF into the systemic circulation. In addition, after 30 minutes of moderate hypoxia, intestinal perfusion abnormalities occur with early changes of ischemic bowel injury. Because this bowel injury could be prevented by pretreatment with PAF recep-

984 CAPLAN ET AL.

GASTROENTEROLOGY

Vol. 99, No. 4

9000

Table 2. Hypoxia-Induced intestinal Injury and Its Prevention by Platelet-Activating Factor Antagonists

a000 -

Intestinal injury”

Treatment

No. of animals

None

Minimal

Mild

Air Hypoxia (30 min) SRI 83-441 + hypoxia WEB 2088 + hypoxia NH&l

5 14 5 5 7

5 5 4 5 7

0 3 1 0 0

0 5 0 0 0

Moderate 0 1 0 0 0

“Intestinal injury was graded by microscopic examination. Sections were taken from the nonperfnsed areas of the small intestine: pathological changes of the most severe lesions are presented in the table. (In acidotic animals and animals pretreated with PAF antagonists in which no hypoperfused bowel was observed, random sections were taken for histological examination). Microscopic lesions were graded by one of the authors who was unaware of the codes. Minimal, separation of the surface epithelial cells from lamina propria; mild, epithelial cell necrosis confined to the tips of villi; moderate, partial loss of intestinal villi. None of the animals showed necrosis beyond mucosal layer. tor antagonists before the hypoxic insult, it is concluded that the endogenously released PAF plays a central role in mediating hypoxia-induced intestinal necrosis. Because 30 minutes of hypoxia also resulted in metabolic acidosis, we also investigated the role of acidosis itself in inducing PAF production and intestinal injury. We found that acidosis alone produced elevation of circulating, plasma PAF with no resulting intestinal injury. Thus, although both hypoxia and acidosis lead to endogenous PAF production, only the former causes bowel injury. A possible explanation for this difference is that hypoxia is known to be injurious to all cells, whereas moderate acidosis (pH

1001

,8

?Plana ?

Control

2

Hypoxia

Figure 5. Acetylhydrolase activity ln plasma (open bars) and intestine (hatched bars) of control (air-breathing) and hypoxic (at 30 minutes) rats. AcetyIhydrolase activity was measured using radiolabeled PAF and measuring reaction products, lyso-PAF. (See Materials and Methods.) (See Table 1 for number of animals.) *P .= 8.08 using two-tailed Student’s t test.

8mtrol

Hypoxia

581

uE8

Figure 6. Effect of hypoxia on peripheral WBC count (A) and hematocrlt (B) and the preventive effect of PAF antagonists. (See Table 1 for number of animals.) *P < 8.85 compared with controls.

7.05 in our experiment] may not be as cytotoxic; also, the injurious effects of hypoxia and PAF are additive. Moreover, hypoxia causes increased activity in the sympathetic system, which induces splanchnic vasoconstriction (18); we have previously shown that PAF also induced norepinephrine release in the gut (19). Thus, the effects of hypoxia and PAF on splanchnic vasoconstriction may also be additive. In clinical situations, hypoxia is usually accompanied by acidosis, and it is possible that metabolic acidosis enhances PAF production and may also aggravate tissue injury. Platelet-activating factor was released into the systemic circulation after 2 minutes of severe hypoxia. To our knowledge, no other studies have documented such a rapid increase of PAF in response to hypoxia. Prevost et al. (20) have shown that after 4 hours of hypoxia, PAF increased in rat bronchoalveolar lavage. It is well known that PAF can increase acutely in response to other stimulants. Chang et al. have shown that 5 minufes after LPS injection, plasma PAF increases twofold, and by 20 minutes after LPS, PAF increased ninefold (21). The rate of PAF elevation seems to depend on the severity and duration of the stimulus. In our study, moderate hypoxia of 15 minutes duration was not sufficient to increase the plasma PAF

HYPOXIA,

October 1990

level, whereas 30 minutes of moderate hypoxia resulted in a marked elevation. The main source of PAF is unclear. It is possible that other organs such as lung also contribute to the production of PAF in addition to the intestine. Acetylhydrolase, the PAF-specific degradating enzyme, is present in many tissues, including plasma and intestine (22). We have shown in this report that the plasma acetylhydrolase activity does not change in response to hypoxia. Interestingly, the intestinal acetylhydrolase activity increases after 30 minutes of hypoxia. This response may tend to protect the animal from increased local PAF production, a phenomenon that we demonstrated in our previous model (10). The acute rise in PAF must therefore be a result of increased PAF production and not inhibition of degradation Previous models have associated hypoxia with decreased intestinal perfusion. We have shown that hypoxia leads to abnormal intestinal perfusion and subsequent necrosis as a response to increased production of PAF. However, compared with our previous models of bowel necrosis induced by PAF (91, LPS (91, or tumor necrosis factor (231, the hypoxia-induced necrosis is much milder and is usually limited to the mucosa. This observation may carry some clinical relevance. It is well known that a severe perinatal hypoxic episode alone usually does not cause NEC. A second injury, i.e., enteral feeding, which results in an increase of intestinal flora and endotoxin, is usually required to initiate the development of NEC (1). We hypothesize that the present model of hypoxiainduced bowel injury mimics the initial hypoxic insult observed clinically in neonates. As a result, the mucosal barrier is impaired. The breakdown of the mucosal barrier facilitates the subsequent entry of bacteria and bacterial endotoxin after enteral feeding or infection, which initiates the process of NEC. Once bacteria and endotoxin enter the systemic circulation, shock and other systemic changes, which aggravate the bowel injury, develop. Moreover, PAF is a known “priming” agent for leukocytes (23).Our previous studies have also shown that PAF synergized with LPS (91 and tumor necrosis factor (TNF)-a (24)to induce bowel injury. Thus, it is plausible to propose that after the initial hypoxic insult and PAF production, a small amount of LPS could trigger the process of tissue injury. Our previous models of experimental NEC probably resemble the later, systemic stage of NEC. We have shown that LPS (9), PAF (9), and TNF-a (24) each play a role in the development of experimental NEC. Furthermore, secondary mediators, such as leukotrienes (11,13), prostaglandins (11,13), norepinephrine (191, and toxic oxygen radicals (251, some of which are probably released from inflammatory cells (e.g., phagocytes and mast cells) in the bowel, may also be

PAF, AND TISSUE

INJURY

985

involved in the pathophysiological mechanism of the development of NEC. Regardless of the model we used, pretreatment with PAF antagonists invariably prevented the development of NEC (10,12,24). These observations suggest a potential usefulness of PAF antagonists in preventing ischemic bowel necrosis in clinical medicine.

References 1. Walsh MC, Kliegman, RM. Necrotizing enterocolitis: treatment based on staging criteria. Pediatr Clin North Am 1986;33:179201. 2. Ryder RW, Shelton JD, Guinan ME. Necrotizing enterocolitis: a prospective multicenter investigation. Am J Epidemiol1980;112: 113-123. 3. Touloukian

4.

5.

6.

7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

RJ, Posch JN, Spencer RR. The pathogenesis of ischemic gastroenterocolitis of the neonate: selective gut mucosal ischemia in asphyxiated neonatal piglets. J Pediatr Surg 1972;7:194-205. Hansbrough F, Priebe CJ Jr, Falterman KW, Bornside GH, Welsh RA. Pathogenesis of early necrotizing enterocolitis in the hypoxic neonatal dog. Am J Surg 1983;145:189-175. Karna P, Senagore A, Chou CC. Comparison of the effect of asphyxia, hypoxia, and acidosis on intestinal blood flow and 0, uptake in newborn piglets. Pediatr Res 198629929-932. Barlow B, Santulli TV, Heird WC, Pitt J, Blanc WA, Schullinger JN. An experimental study of acute neonatal enterocolitis-the importance of breast milk. J Pediatr Surg 1974;9:587-594. Benveniste J. Paf-acether, an ether phospholipid with biological activity. In: Karnovsky ML, Leaf A, Bolis LC, eds. Biological membranes: aberrations in membrane structure and function. New York: Liss 198873-85. Hanahan DJ. Platelet activating factor: a biologically active phosphoglyceride. Ann Rev Biochem 1986;35:483-509. Gonzalez-Crussi F, Hsueh W. Experimental model of ischemic bowel necrosis: the role of platelet activating factor and endotoxin. Am J Path01 1983;112:127-135. Hsueh W, Gonzalez-Crussi F, Arroyave JL. Platelet activating factor is an endogenous mediator for bowel necrosis in endotoxemia. FASEB J 1987;1:403-405. Hsueh W, Gonzalez-Crussi F, Arroyave JL. Platelet-activating factor-induced ischemic bowel necrosis-an investigation of secondary mediators in its pathogenesis. Am J Path01 1986122: 231-239. Hsueh W, Gonzalez-Crussi F, Arroyave F, Anderson JL, Lee ML, Houlihan WJ. Platelet activating factor-induced ischemic bowel necrosis: the effect of PAF antagonists. Eur J Pharmacol 1986:123:79-83. Hsueh W, Gonzalez-Quasi F, Arroyave JL. Release of leukotriene C, by isolated, perfused rat small intestine in response to platelet-activating factor. J Clin Invest 1986;78:108-114. Handley DA, Tomesch JC, Saunders RN. Inhibition of PAFinduced systemic responses in the rat, guinea pig, dog and primate by the receptor antagonist SRI 63-441. Thromb Haemost 1986;56:40-44. Handley DA, Van Valen RG, Tomesch JC, Melden MK, Jaffe JM, Ballard FH, Saunders RN. Biological properties of the antagonist SRI 63-441 in the PAF and endotoxin models of hypotension in rat and dog. Immunopharmacology 1987;13:125132. Casals-Stenzel J, Muacevic G, Weber KH. Pharmacological actions of WEB 2086, a new specific antagonist to platelet activating factor. J Pharm Exp Ther 226;241:974-981.

986 CAPLAN ET AL.

17. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957;228:497-509. 18. Hoka S, Bosnjak ZJ. Arimura H, Kampine JP. Regional venous outflow, blood volume, and sympathetic nerve activity during severe hypoxia. Am J Physiol1989;256:H162-170. 19. Hsueh W, Gonzalez-Crussi F, Arroyave JL. Sequential release of leukotrienes and norepinephrine in rat bowel after plateletactivating factor: a mechanistic study of platelet-activating factor-induced bowel necrosis. Gastroenterology 1988;94:14121418. 20. Prevost MC, Cariven C, Simon MF. Chap H. Douste-Blazy L. Platelet activating factor (PAF-acether) is released into rat pulmonary alveolar fluid as a consequence of hypoxia. Biochem Biophys Res Comm 1984;119:58-63. 21. Chang SW, Feddersen CO, Henson PM, Voelkel NF. Platelet activating factor mediates hemodynamic changes and lung injury in endotoxin-treated rats. J Clin Invest 1987;79:1498-1509. 22. Farr RS, Wardlow ML, Cox CP, Meng KE, Greene DE. Human

GASTROENTEROLOGY

Vol. 99. No. 4

serum acid-labile factor is a acylhydrolase that inactivates platelet activating factor. Federation Proc 1983;42:3120-3122. 23. Vercellotti GM, Yin HQ, Gustafson KS, Nelson RD. Jacob HS. Platelet-activating factor primes neutrophil responses to agonists: role in promoting neutrophil-mediated endothelial damage. Blood 1988;71:1100-1107. 24. Sun XM, Hsueh W. Bowel necrosis induced by tumor necrosis factor in rats is mediated by platelet-activating factor. J Clin Invest 1988;81:1328-1331. 25. Cueva JP, Hsueh W. Role of oxygen derived free radicals in platelet activating factor induced bowel necrosis. Gut 1988;29: 1207-1212.

Received December 21,1989. Accepted April 9.1990. Send requests for reprints to: Wei Hsueh, M.D., Ph.D., Department of Pathology, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, Illinois 60614. This work was supported by NIH Grant No. DK34574.