Serum cytosolic β-glucosidase elevation and early ischemic injury to guinea pig small intestine

Serum cytosolic β-glucosidase elevation and early ischemic injury to guinea pig small intestine Sheri Morris, MD, William Hays, PhD, Miko Enomoto, BA, Robert Glew, PhD, Richard Feddersen, MD, Donald Fry, MD, and Don Morris, MD, Albuquerque, NM

Background. The lack of an early, sensitive marker for intestinal ischemia has led to delay in diagnosis and worsened outcome for patients with acute onset of this condition. Our preliminary studies revealed that guinea pig cytosolic β-glucosidase (CBG) is expressed predominantly in the small intestine, with lower levels in the liver and pancreas and undetectable levels in other organs. Cytosolic β-glucosidase was investigated as a serum marker of small intestinal ischemia in a guinea pig model. Methods. Guinea pigs underwent anesthesia, sham laparotomy, 30 minutes of mesenteric ischemia followed by 6 hours of reperfusion, 6 hours of sustained mesenteric ischemia, or closed-loop small bowel obstruction. Serum samples were assayed for CBG activity. At the conclusion of the ischemia/reperfusion experiments, small bowel samples were assayed for residual enzyme activity, and paraffin sections were graded for the severity of histologic injury. Results. Serum CBG activity rose rapidly after intestinal ischemia with and without reperfusion. Peak enzyme activities were elevated 23-fold for reperfused animals (P < .001) by 4 hours. For nonreperfused animals, peak serum CBG activities reached 29-fold above baseline and were significantly higher than the CBG activities of reperfused animals at 4 hours (P < .01) and at 6 hours (P < .05). Mucosal injury ranged from undetectable to moderate and corresponded in severity with both peak serum enzyme activity and decreased residual activity in the small bowel. In animals subjected to closed-loop obstruction, there was a mean increase of serum CBG of 9.2-fold from 4 to 6 hours after establishment of obstruction (P < .05). Conclusions. In the guinea pig model, CBG is a sensitive marker of ischemic injury caused by arterial occlusion or closed-loop obstruction of the small bowel. (Surgery 1999;125:202-10.) From the Departments of Surgery, Biochemistry, and Pathology, University of New Mexico School of Medicine, Albuquerque, NM

SMALL INTESTINAL ISCHEMIA has been recognized with increasing frequency during the past 25 years. Patients with this condition represent approximately 1 of every 1000 hospital admissions, and mortality rates remain at 40% to 70% in most medical centers.1-4 Acute mesenteric occlusion is the most common vascular cause of ischemic bowel disease.5,6 Intestinal ischemia is often implicated as a complicating factor in many diseases (eg, congestive heart failure, shock, sepsis, atherosclerotic vascular disease, inflammatory bowel disease) and results in a spectrum of injury ranging from completely reversible mucosal alterations to transmurSupported by dedicated health research funds of the University of New Mexico Health Sciences Center. Accepted for publication Sept 11, 1998. Reprint requests: Don Morris, MD, Department of Surgery, UNM Health Sciences Center, ACC 2nd Floor, 2211 Lomas Blvd NE, Albuquerque, NM 87131-5431. Copyright © 1999 by Mosby, Inc. 0039-6060/99/$8.00 + 0

202 SURGERY

11/56/95084

al infarction.1,7 Outcome may be improved by early diagnosis and treatment, with restoration of mucosal blood flow and prevention of progression of the injury toward infarction.8,9 Unfortunately, few diagnostic tests are specific for small intestinal ischemia. Physicians commonly use a combination of physical examination, radiographic studies, nonspecific laboratory tests, and open or endoscopic abdominal exploration to assess the clinical status of patients with suspected intestinal ischemia. At best, these tests are limited to the diagnosis of transmural infarcts and are insensitive to early damage.10 Tonometry and angiography are the most specific and sensitive tests for early acute bowel ischemia; however, both of these procedures are invasive and each is associated with particular complications and limitations.8,10-13 The lack of a sensitive marker for intestinal ischemia can lead to delayed intervention and a worsened outcome for patients.14 This problem underscores the need for a noninvasive test of high specificity and sensitivity for screening

Surgery Volume 125, Number 2

Morris et al 203

Fig 2. Serum CBG activity. Animals in group III (30minute ischemia, 6-hour reperfusion) and group IV (6hour ischemia, no reperfusion) were compared with control groups I (anesthesia) and II (sham laparotomy) (P < .001 for comparison of groups I through III at all time points after baseline). Time is plotted in hours after reperfusion; baseline activity is plotted on the ordinate axis. Activity is shown as mean ± SEM. For comparisons of groups III and IV, *** = P < .001; ** = P < .01; * = P < .05. Fig 1. Expression of CBG in guinea pig organs. Top, Specific activity of CBG in major organs. Representative data are shown. L, Liver; P, pancreas; S, stomach; D, duodenum; J, jejunum; I, ileum; C, cecum. Bottom, Western immunoblot of enriched guinea pig tissue extracts. Upper band in each lane corresponds to native CBG; lower band is proteolytic fragment of the enzyme that retains catalytic activity and is routinely seen in crude extracts of this enzyme. Samples loaded are as follows: lane 1, liver; lane 2, pancreas; lane 3, duodenum; lane 4, jejunum; lane 5, ileum.

and preliminary evaluation of patients suspected of or at risk for intestinal ischemia. Cytosolic β-glucosidase (CBG) is a 53-kd glycohydrolase that shows a high degree of amino acid sequence similarity with human intestinal lactasephlorhizin hydrolase.15 Studies in a guinea pig model suggest that this enzyme may function in the metabolism of dietary xenobiotic glycosides.16 CBG has been purified from human17 and guinea pig18 liver, as well as from other mammalian species.19 Chang et al20 reported that serum activity of several cytosolic and lysosomal glycohydrolases increased immediately on reperfusion after liver transplantation in swine. Cytosolic β-glucosidase level has been shown by Liu et al21 to rise in the peripheral circulation early in the course of porcine hepatic ischemia with reperfusion. The work by Liu et al21 suggested to us that CBG might be an early marker of cellular injury resulting from ischemia with reperfusion. In this study we show that high levels of CBG are expressed in guinea

pig small intestine, pancreas, and liver. We further demonstrate that CBG is an early, sensitive marker of small intestinal ischemia and ischemia/reperfusion injury in guinea pigs. MATERIAL AND METHODS Surgical procedure. Adult male Dunkun-Hartley guinea pigs (550 to 820 g) were purchased from Harlan (St Louis, Mo). The experiments were conducted under a protocol approved by the University of New Mexico Health Sciences Center Laboratory Animal Care and Use Committee. Guinea pigs were allowed free access to food and water before experimentation. Anesthesia was induced with ketamine (44 mg/kg intramuscularly), acepromazine (1 mg intramuscularly), and atropine (0.05 mg subcutaneously) and maintained with repeated doses of ketamine (14.5 mg/kg intramuscularly) and acepromazine (1 mg intramuscularly). Lidocaine (2%) was injected intradermally (1 to 3 mL) before skin incisions for local anesthesia.22 Animals were intubated with a 1.5-cm 14-gauge angiocatheter through tracheotomy, and the jugular vein was cannulated with polyethylene tubing (PE-10; Becton Dickinson, Parsippany, NJ). Animals were ventilated (UGOBasille model 7025 rodent ventilator; Stoelting) with a tidal volume of 7 mL/kg, respiratory rate of 60/min, and humidified 100% oxygen. Animals were monitored by electrocardiogram, venous

204 Morris et al

Surgery February 1999

Fig 4. Specific activity of CBG in jejunum. Specific activities were determined after death of each animal. Activity in jejunal samples was significantly lower in group IV than in groups I through III. ** = P < .01; * = P < .05.

Fig 3. Serum activities of LDH, ALT, and amylase. Enzyme activities were determined at baseline and at 2 and 4 hours after reperfusion in groups I, II, and III. A, Serum LDH; B, serum ALT; C, serum amylase. Activities are shown as mean ± SEM. For all comparisons, *** = P < .001; ** = P < .01; * = P < .05.

blood gas analysis, and rectal temperature. Each animal received hourly subcutaneous injections of 5 mL normal saline (NS, groups I through III) and intravenous supplementation every 15 minutes with 5 mL NS alternating with lactated Ringer’s solution (all groups). Blood samples were replaced with 3 volumes of NS in group I animals and 5 volumes of NS in groups II and III. The jugular venous catheter was flushed with heparin saline (75 units/mL, 0.2 mL/flush) after each blood sample or fluid administration. Thirty-eight animals were divided into 5 groups.

Group I (n = 8) served as anesthesia controls and underwent tracheotomy, jugular venous cannulation, and ventilatory support. Group II animals (n = 8) underwent these procedures and sham laparotomy and scheduled portal venous sampling. Animals in group III (n = 8) underwent complete arteriovenous small intestinal ischemia with reperfusion. Ischemia was achieved through occlusion of the proximal mesenteric vascular bundle with a small atraumatic vascular clamp. A second atraumatic clamp was applied across the proximal duodenum to ensure complete vascular occlusion. Both clamps were removed after 30 minutes of ischemia. Group IV animals (n = 7) were subjected to complete, irreversible mesenteric arterial occlusion. This was accomplished by placing microclips across the mesenteric artery just distal to a gastric branch, at the origin from the aorta. Venous outflow was not obstructed in group IV. Animals in group V (n = 7) were subjected to small bowel obstruction with intraluminal fluid distention. A segment of jejunum approximately 60 cm long was isolated with two ligatures. Before placement of the proximal ligature, a small enterotomy was performed and the lumen of the isolated segment was catheterized with PE-10 tubing. The catheter and enterotomy were secured with the same ligature used to isolate the proximal bowel segment. After isolation, the bowel segment was distended with approximately 300 mL 6% hetastarch (Hespan) in NS, and the catheter was clamped for the remainder of the experiment. Venous blood gas samples (0.2 mL) were obtained from the jugular vein on the same schedule in all groups. Peritoneal contents were replaced within the abdomen between manipulations. The skin was approximated with clamps and draped with a gauze sponge, which was frequently moist-

Surgery Volume 125, Number 2

Fig 5. Paraffin sections of intestine after ischemic injury. Left, Grade 1 changes in a group III animal. Degeneration and detachment of superficial epithelium are present. These changes are not readily distinguished from fixation and processing artifact and did not consistently distinguish group III from control groups. Right, Grade 3 changes in a group IV animal. Coagulation necrosis of the villi is present.

ened with saline to prevent desiccation. Tissue specimens were obtained from liver, small bowel, cecum, and pancreas after the final blood sampling. On completion of the experiment, animals were killed with pentobarbital (100 mg/kg) by cardiac puncture. Biochemical studies. Cytosolic β-glucosidase activity was determined using an end-point fluorometric assay modified from one described previously.23 Aliquots of serum (5 µL) were added to assay medium consisting of 0.2 mol/L Na citrate, pH 6.0, and 5 mmol/L 4-methylumbelliferyl-β−Dglucopyranoside (MUGlc) in a final volume of 0.1 mL. Reactions were incubated for 30 minutes at 37°C in a circulating water bath and terminated by the addition of 2.9 mL of 0.3 mol/L ammoniaglycine buffer, pH 10.0. Fluorescence of the 4methylumbelliferone product released by CBG was measured using a Turner model 450 fluorometer (Turner Instruments, Sequoia, Calif) with a band pass 360-nm excitation filter and a sharp cut 515nm emission filter. One unit of enzyme activity is defined as that cleaving 1 nanomole of MUGlc in 1 hour. Levels of lactate dehydrogenase (LDH), alanine aminotransferase (ALT), and amylase were assayed using an automated Kodak DT-60 Clinical Chemistry analyzer (Eastman Kodak, Rochester, NY). Serum samples were assayed batchwise for CBG activity immediately after death of each animal and then stored at –80°C for later assay of LDH, ALT, and amylase. The specific activity of CBG in solid organs was measured using the

Morris et al 205

Fig 6. Histologic severity scores vs peak serum β-glucosidase activity. Paraffin sections of small bowel from groups I through IV were scored for injury severity using the scoring system demonstrated in Fig 5. Animals were divided into 2 groups corresponding to severity scores of either 02 or 3-5, and their peak serum CBG activities were compared using a nonparametric Mann-Whitney test.

100,000g supernatant of organ samples that had been stored at –80°C until homogenization. For Western blot studies, the 100,000g supernatant was enriched for CBG activity by precipitation with 20% (w/v) PEG 4000, batch chromatography on diethylaminoethyl-cellulose, and concentration by ultrafiltration as described elsewhere.15 The enriched supernatant was dissolved in gel loading buffer at a final concentration of 2.5 µg/µL total protein and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Semi-dry electrophoretic transfer was performed using a Phastransfer apparatus (Pharmacia, Piscataway, NJ) and PVDF membranes (Bio-Rad, Hercules, Calif). Membranes were blocked for 2 hours with 5% (w/v) non-fat dry milk in phosphate-buffered saline solution with 0.5% (v/v) Triton X-100 and then incubated for 12 to 14 hours with 1:200 rabbit polyclonal antiserum raised against purified guinea pig liver CBG. Detection was by peroxidase staining using peroxidase-coupled goat anti-rabbit antiserum obtained from Sigma (Sigma Chemical Co, St Louis, Mo). Histologic analysis. Paraffin sections of small intestine, liver, and pancreas were obtained from animals from groups I through IV immediately after death. Each section was evaluated independently by a blinded observer according to a histologic scoring system previously used in an animal model of necrotizing enterocolitis.24 Samples of liver, small bowel, cecum, and pancreas were taken for histologic analysis and fixed overnight in formalin at 4° C. Paraffin sections were prepared and analyzed after staining with hematoxylin-eosin. A total of 5 animals each from groups I through III were available for analysis, along with all 7 animals

206 Morris et al

Fig 7. Serum CBG activity in closed-loop obstruction. Animals in group V (n = 7) were subjected to closed-loop obstruction; peak serum activity (occurring from 4 to 6 hours in all animals) is plotted in comparison with baseline CBG activities. Median elevation over baseline was 9.2fold (range, 2.9- to 130-fold).

from group IV. Injury was assessed by a blinded observer according to the following protocol: 0, no diagnostic changes; 1, changes of uncertain significance, including vacuolization or detachment of surface epithelium; 2, patchy partial-thickness mucosal necrosis; 3, confluent partial-thickness mucosal necrosis; 4, full-thickness mucosal necrosis; 5, bowel wall (muscularis propria) necrosis. Statistical analysis. Differences in serum enzyme activities between groups I through III were analyzed by nonparametric Kruskall-Wallis one-way analysis of variance. Separate analyses were performed at each time point. Serum CBG activities in groups III and IV were compared at each time point using a Mann-Whitney test. Analysis of group V data was performed using a Wilcoxon signed rank test for paired data. Statistical significance for all tests was defined as P < .05. RESULTS CBG activity in normal tissue. A survey of extracts of various guinea pig tissues revealed that CBG activity was greatest in the small bowel, pancreas, and liver, with much lower levels in the stomach and colon (Fig 1, A). The identity of CBG was confirmed by Western immunoblot analysis (Fig 1, B). Immuno–cross-reactive protein in the small bowel and pancreas was identical in molecular weight to that expressed in the liver. The intensity of staining, as determined by densitometric scanning, corresponded with the specific activity of CBG in each tissue. There was negligible enzyme activity and no immuno–cross-reactive material in the other organs tested (brain, lung, heart, kidney, spleen, adrenal, and skeletal muscle)(data not shown). Serum CBG activity with ischemic injury. The

Surgery February 1999

baseline serum CBG activity was less than 8.0 U/mL in the portal and systemic samples in all animals. The mean serum CBG activity was not significantly different between the 2 control groups (groups I and II) and remained less than 15 U/mL throughout the experiment. There was an immediate and statistically significant increase in the serum activity of CBG on reperfusion after 30 minutes of small intestinal ischemia (group III)(Fig 2). Enzyme activity peaked after 2 hours and was elevated 23-fold over baseline activity. The increases in CBG activity at each time point were statistically significant when compared with the controls. With sustained mesenteric arterial ischemia to 85% of the small intestine (group IV, no reperfusion), serum CBG activity increased within 1 hour of ischemia and rose to an average of 29-fold over baseline at 4 hours. This elevation was significantly greater than group III (Fig 2). In group III, both the artery and the venous outflow were occluded for 30 minutes, whereas in group IV the venous outflow was specifically preserved. The activity of LDH increased in group III on reperfusion and was significantly greater than the activity in control animals (Fig 3, A). LDH activity was also significantly elevated at 4 hours in the sham laparotomy control group (group II). The activity of ALT in group III increased after reperfusion and was significantly greater than the activity in the control groups (Fig 3, B). Serum amylase levels in group III were not significantly different from controls throughout the experiment (Fig 3, C). CBG activity in liver and jejunum after ischemia. The specific activity of CBG in the jejunum and liver was measured in tissue samples taken immediately after the death of each animal. Cytosolic βglucosidase activity in jejunal samples from animals in group IV (nonreperfused) was significantly lower than in samples from groups I through III (Fig 4). Conversely, there were no significant differences in the activity of CBG in liver samples from all 4 groups (data not shown). Histology of the small intestine, liver, and pancreas. Paraffin sections of small bowel obtained from groups I through III revealed injury ranging from nondiagnostic changes to mild mucosal injury, which was characterized by vacuolization or detachment of surface epithelium (Fig 5). Sections of small intestine from group IV after 6 hours of arterial ischemia exhibited moderate mucosal injury, which was manifested as patchy to partialthickness mucosal necrosis, coagulation necrosis of the villi, involvement of neighboring villi, and, in some animals, full-thickness mucosal necrosis. Injury was graded numerically in terms of severity.

Surgery Volume 125, Number 2

Areas of severe injury were seen in group IV animals only. The histologic severity scores corresponded with the serum CBG activities for each animal; animals with the most severe histologic scores (scores 3 through 5) had significantly higher peak serum enzyme activities (Fig 6). Sections of small intestine from groups I and II, as well as liver and pancreas sections from groups I through IV, did not show pathologic changes. Serum CBG activity with closed-loop obstruction. The effects of small bowel obstruction were modeled in group V (n = 7). A 60-cm segment of jejunum was isolated with ligatures and distended with approximately 300 ml 6% (w/v) hetastarch solution. Before the establishment of the obstruction, the baseline CBG activity was 4.8 ± 1.0 U/mL. In contrast to the animals in groups III and IV, there was a gradual increase in serum CBG in group V. Serum CBG activity rose to a mean of 9.2fold above baseline (44 ± 11 U/mL, P < .02) at 4 to 6 hours after establishment of obstruction (Fig 7). The increase in serum CBG activity ranged from 2.9- to 26-fold in group V animals when comparing peak and baseline serum CBG activities. DISCUSSION The principal finding of this study is that CBG is a sensitive marker of both catastrophic and noncatastrophic small bowel injury. There was a rapid rise in serum CBG activity in animals subjected to both sustained ischemia and ischemia with reperfusion. The degree of CBG elevation correlated well with the extent of injury, as determined from histologic sections. In animals subjected to small bowel obstruction, there was a more gradual increase in serum CBG activity, which was consistent with the less acute nature of the injury. Furthermore, there appeared to be no significant contribution to serum CBG activity from either liver or pancreas, indicating the specificity of CBG for intestinal injury in this experimental model. The diagnosis of small intestinal ischemia remains an elusive goal and one that is generally made only at laparotomy and after significant intestinal necrosis has occurred. The significant delay usually associated with the diagnosis of intestinal ischemia ultimately prolongs the period of ischemia and worsens intestinal injury. A sensitive screening laboratory test that would facilitate early diagnosis and treatment could significantly decrease the 70% mortality rate associated with small intestinal ischemia.8,9,14 Our data demonstrate the sensitivity of CBG to early mucosal injury. The rapid increase in enzyme activity on reperfusion after 30 minutes of intestinal ischemia (group

Morris et al 207

III) was highly significant when compared with controls. Although CBG activity peaked between 2 and 4 hours after ischemia/reperfusion in this group, CBG elevations after sustained ischemia (group IV) were significant as early as 1 hour after arterial occlusion, and peak elevations occurred at 4 hours. Serum CBG elevations were significantly higher at 4 and 6 hours in group IV as compared with those in group III, and these data corresponded with the severity of tissue injury scored on histologic sections from these animals. Intestines from animals that experienced 30 minutes of small intestinal ischemia followed by reperfusion showed evidence ranging from no injury to mild tissue injury, whereas tissue obtained from animals subjected to 6 hours of intestinal ischemia showed moderate to severe injury. Higher CBG activity at completion of the experiment was associated with more severe histologic changes. These results demonstrate the sensitivity of CBG for detecting not only the presence of early ischemic injury, but the severity of this injury as well. Mesenteric ischemia as modeled in groups III and IV is a catastrophic but relatively infrequent occurrence. A more commonly encountered situation involves small bowel obstruction caused by adhesions, tumor, hernia, etc. Ischemic injury to the mucosa in small bowel obstruction derives from occlusion of the vascular pedicle supplying the obstructed segment (strangulation) and from luminal distention compressing the submucosal vasculature.25 In group V we modeled the effects of luminal distention by infusing a large volume of osmotically active solution (6% hetastarch) to rapidly achieve a state that clinically may require much longer to develop. In comparison to groups III and IV, the animals in group V were subjected to a much less severe injury, given that the extrinsic vascular supply to the affected bowel segment was not compromised. Predictably, the resulting elevations in serum CBG followed a more indolent time course, but they were still significantly elevated relative to baseline (Fig 7). In the guinea pig, CBG is expressed in the small bowel, liver, and pancreas. Therefore, although the experimental manipulations were performed exclusively on the small bowel, it was necessary to consider the possible release of CBG from either the liver or the pancreas as a result of secondary injury to these organs. In preliminary experiments we found that combined occlusion of both the superior mesenteric artery and the superior mesenteric vein blocked elevations in serum CBG despite obvious infarction of the bowel (data not shown). Consequently, venous return was specifically pre-

208 Morris et al

served in groups I, II, IV, and V. The absence of an increase in serum amylase level, together with the absence of pathologic changes in the pancreas, indicated that significant pancreatic injury was unlikely. However, the elevation in ALT level seen in group III animals suggested the possibility of liver injury being responsible for some of the serum CBG elevation. Several reports describe hepatic injury and dysfunction after intestinal ischemia/reperfusion injury.26-28 Current data suggest that this may be due to increased neutrophil adhesion in liver sinusoids, with subsequent inflammatory damage.29 Therefore some release of CBG from the liver is probable given the magnitude of the injury incurred by the animals in groups III and IV. The practical question is really the magnitude of the enzyme increase and the major organ responsible for this increase. In our studies there was considerable depletion of CBG in the small bowel but not in the liver (Fig 4). Thus, in the context of this experimental model, our data are consistent with the small bowel being the predominant source of serum CBG activity. The question arises as to whether the increases in serum CBG activity could reflect increased expression, especially in the liver. After ischemia/reperfusion injury, the expression of heat shock proteins is increased in the liver.30 In rat kidney subjected to ischemia/reperfusion injury, increases in the mRNA for selected heat shock proteins are increased as much as 8- to 10-fold within 6 hours after reperfusion, with increases in the respective protein being detectable by 24 hours after reperfusion.31 In contrast, hepatic ischemia/reperfusion injury causes a dramatic decrease in the synthesis of both acute phase proteins and secretory proteins in general, which are evident within 1 hour after reperfusion.32 These data suggest that, although heat shock proteins can be rapidly induced after ischemia/reperfusion injury, hepatic protein synthesis in general is decreased. In the case of CBG, we are unaware of any data that might suggest that this enzyme is either a heat shock protein or an acute phase reactant. Furthermore, we observed increases greater than 20-fold over baseline serum CBG activity in less than 4 hours after reperfusion. Increased translation sufficient to account for an increase of this magnitude seems improbable. Perhaps most significantly, there is no signal sequence in the mRNA for CBG; it is an intracellular enzyme and is not localized to any membranes of the secretory pathway.15 Therefore the only way for CBG to reach the circulation is for it to be released during the process of membrane damage. The distribution of CBG in human tissues

Surgery February 1999

appears to be very similar to what we have described in the guinea pig. Previous reports have documented the presence of this enzyme in the mucosa of human small bowel,33,34 which we have recently confirmed using samples of normal human small intestine obtained from the margins of surgical resection (unpublished observations). Newmark et al35 demonstrated much higher activity of this enzyme in human small intestine than in the colon, whereas the absence of the enzyme from human kidney36 is consistent with our findings in the guinea pig (Fig 1). In preliminary studies we have documented several-fold elevations in serum CBG activity of 3 patients who were found at laparotomy to have ischemic bowel (unpublished observation). Collectively, these data suggest that CBG may be useful as a marker of ischemic injury to human intestine. Many serum markers have been evaluated for their use in screening patients for intestinal ischemia, including creatine kinase and aspartate transaminase,37 alkaline phosphatase,38 and LDH.39 These enzymes have proved unsuitable as markers for intestinal ischemia because of insufficient sensitivity or specificity.37-39 However, recent work with intestinal fatty acid binding protein (I-FABP) indicates that this protein has promise as a marker of bowel ischemia. The study by Gollin et al40 indicated that I-FABP is a sensitive and early marker for intestinal ischemia and ischemia/reperfusion injury in the rat model. A subsequent report by Kanda et al41 demonstrated a 100% sensitivity of elevated I-FABP for intestinal infarction and a somewhat lower sensitivity for intestinal strangulation. Patients with abdominal emergencies not involving infarction or strangulation had significantly lower levels of serum I-FABP. Future studies comparing IFABP and CBG in different models of intestinal injury could be very informative. The use of a simple biochemical test that is specific for small intestinal injury would permit rapid identification of patients with questionable clinical findings and encourage the earlier use of more invasive diagnostic or therapeutic modalities. In our study, CBG assays were performed immediately on completion of the experiments. Accurate results were obtained within 30 minutes, which would allow for real-time clinical management decisions to be made on the basis of the test results. Organ specificity could be obtained clinically by comparing serum ALT and amylase activities with CBG activity. A lack of a significant increase in either ALT or amylase activity would suggest intestinal rather than liver or pancreatic injury as a source for the CBG. Future developments in the molecular

Morris et al 209

Surgery Volume 125, Number 2

biology of this enzyme could lead to organ-specific CBG assays. In conclusion, we have shown that CBG is an early marker of ischemic intestinal injury in the guinea pig model. Future studies are needed to further explore the behavior of this enzyme in models of bowel obstruction, as well as in different pathologic situations such as short segment ischemia and recurrent episodes of ischemia with reperfusion. In addition, mechanisms of CBG release should be investigated, as well as the metabolism and clearance of this enzyme. We thank Dr Mark Eichenger and Riley Nelson for assistance with blood gas monitoring and ventilation equipment and Dr David Johnston for helpful suggestions and discussions. REFERENCES 1. Boley SJ, Brandt LJ, Veith FJ. Ischemic disorders of the intestines. Curr Probl Surg 1978;15:1-85. 2. Kairaluoma MI, Karkola P, Heikkinen E, Huttunen R, Mokka RE, Larmi TK. Mesenteric infarction. Am J Surg 1977;133:188-93. 3. Ottinger LW. The surgical management of acute occlusion of the superior mesenteric artery. Ann Surg 1978;188:721-31. 4. Pierce GE, Brockenbrough ED. The spectrum of mesenteric infarction. Am J Surg 1970;119:233-9. 5. Hildebrand HD, Zieler RE. Mesenteric vascular disease. Am J Surg 1980;139:188-92. 6. Smith JS, Patterson LT. Acute mesenteric infarction. Am Surg 1976;42:562-7. 7. Mitsudo S, Brandt LJ. Pathology of intestinal ischemia. Surg Clin North Am 1992;72:43-63. 8. Boley SJ, Sprayregan S, Siegelman SS, Veith FJ. Initial results from an aggressive roentgenological and surgical approach to acute mesenteric ischemia. Surgery 1977;82:848-55. 9. Kaleya RN, Sammartano RJ, Boley SJ. Aggressive approach to acute mesenteric ischemia. Surg Clin North Am 1992;72:157-82. 10. Kurland B, Brandt LJ, Delaney HM. Diagnostic tests for intestinal ischemia. Surg Clin North Am 1992;72:85-105. 11. Boley SJ, Brandt LJ, Veith FJ, et al. A new provocative test for chronic mesenteric ischemia. Am J Gastroenterol 1991;86:888-91. 12. Bergofsky E. Determination of tissue O2 tensions by hollow visceral tonometers: effect of breathing enriched O2 mixtures. J Clin Invest 1964;43:193-200. 13. Bakal CW, Sprayregen S, Wolf EL. Radiology in intestinal ischemia: angiographic diagnosis and management. Surg Clin North Am 1992;72:125-41. 14. Heys SD, Brittenden J, Crofts TJ. Acute mesenteric ischemia: the continuing difficulty in early diagnosis. Postgrad Med J 1993;69:48-51. 15. Hays WS, Jenison SA, Yamada T, Pastuzyn A, Glew RH. Primary structure of the cytosolic β-glucosidase of guinea pig liver. Biochem J 1996;319:829-37. 16. Gopalan V, Pastuszyn A, Galey WR, Glew RH. Exolytic hydrolysis of toxic plant glucosides by guinea pig liver cytosolic β-glucosidase. J Biol Chem 1991;267:14027-32. 17. Daniels LB, Coyle PJ, Chiao YB, Glew RH, Labow RS. Purification and characterization of a cytosolic broad-speci-

18.

19.

20.

21.

22.

23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33. 34. 35.

ficity β-glucosidase from human liver. J Biol Chem 1981;256:13004-13. LaMarco KL, Glew RH. Hydrolysis of a naturally occurring β-glucoside by a broad-specificity β-glucosidase from liver. Biochem J 1986;237:469-76. Glew RH, Gopalan V, Forsyth GW, VanderJagt DJ. The mammalian cytosolic broad-specificity β-glucosidase. In: Esen A, editor. β-Glucosidases: biochemistry and molecular biology. Washington: American Chemical Society; 1993. p 83-112. Chang CT, Imai K, Chang YC, Hayashi T, Kohno H, Nagasue N, Nakamura T. Plasma lysosomal enzymes after liver transplantation in the pig. Enzyme 1991;45:145-54. Liu W, Schob O, Pugmire JE, Jackson DE, Zucker KA, Fry DE, Glew RH. Glycohydrolases as markers of hepatic ischemia-reperfusion injury and recovery. Hepatology 1996;24:157-62. Shucard DW, Andrew M, Beauford C. A safe and fast-acting surgical anesthetic for use in the guinea pig. J Appl Physiol 1975;38:538-9. Peters SP, Lee RE, Glew RH. Differentiation of β-glucocerebrosidase from β-glucosidase in human tissues using sodium taurocholate. Arch Biochem Biophys 1975;175:569-82. Musemeche CA, Baker J, Feddersen RM. A model of intestinal ischemia in the neonatal rat utilizing superior mesenteric artery occlusion and intraluminal PAF. J Surg Res 1995;58:724-7. Bass KN, Jones B, Bulkley GB. Current management of small-bowel obstruction. Adv Surg 1997;31:1-34. Poggetti RS, Moore FA, Moore EE, Koieke K, Banerjee A. Simultaneous liver and lung injury following gut ischemia is mediated by xanthine oxidase. J Trauma 1992;32:723-8. Poggetti RS, Moore EE, Moore FA, Koieke K, Banerjee A. Gut ischemia/reperfusion-induced liver dysfunction occurs despite sustained oxygen consumption. J Surg Res 1992;52:432-6. Turnage RH, Bagnasco J, Burger KS, Guice KS, Oldham KT, Hinshaw DB. Hepatocellular oxidant stress following intestinal ischemia/reperfusion injury. J Surg Res 1991;51:467-71. Horie Y, Wolf R, Miyasaka M, Anderson DC, Granger DN. Leukocyte adhesion and hepatic microvacular responses to intestinal ischemia/reperfusion in rats. Gastroenterology 1996;111:666-73. Aoe T, Inaba H, Kon S, Imai M, Aono M, Mizuguchi T, et al. Heat shock protein 70 messenger RNA reflects the severity of ischemia/reperfusion injury in the perfused rat liver. Crit Care Med 1997;25:324-9. Maines MD, Mayer RD, Ewing JF, McCoubrey WK. Induction of kidney heme oxygenase-1 (HSP32) mRNA and protein by ischemia/reperfusion: possible role of heme as both promoter of tissue damage and regulator of HSP32. J Pharmacol Exp Ther 1993;264:457-62. von Allmen D, Li S, Hasselgren P-O, Fischer JE. Effect of ischemia on protein synthesis in the septic liver. Surg Gynecol Obstet 1991;172:441-8. Asp N-G, Dahlqvist A, Koldovsky O. Human small intestinal β-galactosidases. Biochem J 1969;114:351-9. Gray GM, Santiago NA. Intestinal β-galactosidases. J Clin Invest 1969;48:716-28. Newmark J, Brady RO, Grimley PM, Gal AE, Waller SG, Thistlethwaite JR. Amygdalin (Laetrile) and prunasin β-glucosidases: distribution in germ-free rat and in human tumor tissue. Proc Natl Acad Sci USA 1981;78:6513-6.

210 Morris et al

Surgery February 1999

36. Freese A, Brady RO, Gal AE. A β-glucosidase in feline kidney that hydrolyzes amygdalin (Laetrile). Arch Biochem Biophys 1980;201:363-8. 37. DeToma G, Marazano D, Salvatore P, Cerza F, De Cesare E, Giacovazzo M, et al. Enzymatic and metabolic changes in the peripheral serum after superior mesenteric artery ligation in dogs. Ital J Surg Sci 1983;13:269-73. 38. Barnett S, Davidson E, Bradley E. Intestinal alkaline phosphatase and base deficit in mesenteric occlusion. J Surg Res 1976;20:243-6.

39. Calman C, Hershey F, Skaggs J. Serum lactate dehydrogenase in the diagnosis of the acute surgical abdomen. Surgery 1958;44:43-51. 40. Gollin G, Marks C, Marks WH. Intestinal fatty acid-binding protein in serum and urine reflects early ischemic injury to the small bowel. Surgery 1993;113:545-51. 41. Kanda T, Fuji H, Tani T, Murakami H, Suda T, Sakai Y, et al. Intestinal fatty acid-binding protein is a useful marker for mesenteric infarction in humans. Gastroenterology 1996; 110:339-43.

Acknowledgment We would like to thank the reviewers listed below who contributed their time recently to review manuscripts for Surgery. These individuals, as well as members of the Editorial Board, commit their time and careful consideration to ensure that articles in Surgery reflect the highest standards of scholarship and relevance. Andrew L. Warshaw Michael G. Sarr Editors in Chief Arbeit, Jeffery M. University of California Ascher, Nancy L. University of California, San Francisco Barber, W. Henry University of Mississippi Medical Center Baxter, B. Timothy University of Nebraska Medical Center Becker, James M. Boston Medical Center Behrns, Kevin E. University of North Carolina School of Medicine Bell, Richard H. Veterans Administration Puget Sound Health Care System Birkmeyer, John D. Department of Veterans Affairs Medical Center Bockman, Dale E. Medical College of Georgia Bonjer, H. Jaap University Hospital Rotterdam Dykzigt Brooks, David C. Brigham and Women’s Hospital Buchler, M. W. University of Bern–Inselpital Burdick, James F. Johns Hopkins Hospital Callery, Mark University of Massachusetts Medical Center Cherry, Kenneth Mayo Clinic Choti, Michael A. Johns Hopkins Hospital Clagett, G. Patrick University of Texas, Southwestern Medical School Clark, Orlo H. University of California, San Francisco, Mount Zion Medical Center

Coran, Arnold University of Michigan Medical Center Cryer, H. Gill University of California, Los Angeles, Medical Center Deitch, Edwin A. New Jersey Medical School Dellinger, E. Patchen University of Washington Medical Center Delmonico, Francis L. Massachusetts General Hospital Evans, Douglas B. MD Anderson Cancer Center Evans, Roger W. Mayo Clinic Evers, B. Mark University of Texas Medical Branch Fan, S. T. Queen Mary Hospital, University of Hong Kong Ferguson, Charles M. Massachusetts General Hospital Fernandez del Castillo, Carlos Massachusetts General Hospital Filler, Robert University of Toronto, Hospital for Sick Children Fink, Aaron S. Atlanta Veterans Administration Medical Center Fink, Mitchell P. Beth Israel Deaconess Medical Center Fleshman, James Washington University Medical Center Foitzik, Thomas Freie Universitat Berlin Friedmann, Theodore University of California, San Diego Gadacz, Thomas R. Medical College of Georgia Galandiak, Susan University of Louisville continued on pg. 242