INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

RADIOLABELED CHEMOTACTIC peptides are promising tools for the imaging of inflammation and infection (4, 7). Chemotactic peptides are naturally released by bacteria and initiate leukocyte chemotaxis by binding to high-affinity receptors on the white blood cell membrane (21, 26). These receptors are present on polymorphonuclear neutrophils and monocytes. N-formyl-methionyl-leucyl-phenylalanine-lysine (fMLFK) is one of the most potent synthetic chemoattractants, is readily synthesized and radiolabeled, and has been shown to be an effective means for the rapid localization of focal sites of inflammation or infection (2). Specific binding of fMLFK to leukocyte receptors in vivo has not been definitely proved, but decreased uptake after application of an antagonist and high target-to-background ratios in inflamed tissue suggest leukocyte targeting (3, 8).

Previous studies evaluating the in vivo behavior and biodistribution of chemotactic peptides used models of bacterially or chemically induced muscular inflammation (2, 3, 6, 8, 24). These studies, however, did not document leukocyte infiltration histologically or otherwise and thus did not provide evidence that chemotactic peptide localization correlates with leukocyte infiltration. To address this issue, two well-established models of acute pancreatitis in rats have been applied in the present study. The first model produces mild pancreatitis, which is predominantly characterized by interstitial edema and increased capillary blood flow (14, 15). Leukocyte infiltration in the pancreas is limited. In contrast, the second model produces severe pancreatitis, characterized by large areas of necrosis, less pronounced edema, extensive leukocyte infiltration, and decreased microcirculatory perfusion and ischemia in the pancreas (12, 23).

Recently, our laboratory described the local and systemic distribution patterns of leukocytes in identical models of acute pancreatitis by using in vitro radiolabeled white blood cells (25). In the present study we evaluated the biodistribution of the nicotinyl hydrazine-derivatized chemotactic peptide analog ^99mTc-fMLFK-HYNIC in these two models and compared the results with data obtained by the technique established in the previous study. Because nonspecific inflammatory reactions, such as edema formation and alterations in tissue perfusion, may affect the tissue uptake of various radiopharmaceuticals, the combination of well-described specific and nonspecific inflammatory reactions in our models of mild and severe pancreatitis makes them appropriate for evaluating the in vivo behavior of chemotactic peptides.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and catheter placement. Experiments were performed in male Sprague-Dawley rats weighing 300-350 g. Care was provided in accordance with the procedures outlined in the Guide for Care and Use of Laboratory Animals [DHHS Publication No. (NIH) 85-23, revised 1985, Office of Science and Health Reports, Bethesda, MD 20892]. Animals were fasted overnight before the experiments but allowed free access to water.

Study design. Animals were randomly allocated to a control group or two groups of acute pancreatitis: mild edematous or severe necrotizing disease. Mild pancreatitis was induced by intravenous infusion of cerulein (5 µg · kg¹ · h¹; Takus, Farmitalia, Carlo Erba, Freiburg, Germany) for 6 h (15). Cerulein was reconstituted in normal saline and infused at 3 ml · kg¹ · h¹. Necrotizing pancreatitis was induced as follows and as described in detail elsewhere (23). The biliopancreatic duct was cannulated with a 24-gauge Teflon catheter (Critikon, Tampa, FL), and bile and pancreatic juice were drained by gravity for 5 min with the common hepatic duct clamped at the porta hepatis. Glycodeoxycholic acid (Sigma Chemical, St. Louis, MO) in glycylglycine-NaOH-buffered solution (pH 8.0, room temperature) was infused retrogradely into the biliopancreatic duct at a concentration of 10 mmol/l in a volume (1.2 ml/kg)-, time (10 min)-, and pressure (30 mmHg)-controlled fashion, followed by the intravenous infusion of cerulein over 6 h. Control animals received saline infusions alone.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of experimental protocol. A: cerulein or saline was infused for 6 h, and animals were killed when infusions were completed. For induction of severe pancreatitis, cerulein infusions were preceded by intraductal infusion of glycodeoxycholic acid (GDOC) for 10 min. B: 1 group of animals with severe pancreatitis received saline for 6 h after completion of cerulein infusions. All animals were injected with 99Tc-labeled chemotactic peptide formyl-methionyl-leucyl-phenylalanine-lysine (fMLFK) 3 h before organs were removed for radioactivity measurements.

^99mTc-labeled chemotactic peptide. N-formyl-methionyl-leucyl-phenylalanine (fMLF) was purchased from Sigma Chemical. ^99mTcO₄ (⁹⁹Mo/^99mTc generator) and stannous glucohepatonate (Glucosan) were obtained from DuPont (Bilerica, MA). Hanks' balanced salt solution was purchased from GIBCO (Grand Island, NY). Dimethylformamide (DMF), trifluoroacetic acid (TFA), ether, petroleum ether, ethyl acetate, and p-cresol were obtained from Aldrich Chemical (St. Louis, MO). Inorganic salts were obtained from Fisher Scientific (St. Louis, MO). Instant thin-layer silica gel chromatographic strips were obtained from Gelman Laboratories (Ann Arbor, MI).

Peptide synthesis. fMLFK was synthesized and purified by standard solid-phase techniques (16), as previously described (18). The nicotinyl hydrazine-derivatized chemotactic peptide analog fMLFK-HYNIC was prepared as described below.

Radiolabeling with ^99mTc. A ^99mTc generator was eluted ~5 h after a previous elution to yield ~500 mCi of ^99mTcO₄. A typical elution contained ~3 nmol of ⁹⁹Tc; the ⁹⁹Tc-to-^99mTc ratio was ~1.5:1, and specific activity was >100,000 mCi/µmol. ^99mTc-glucohepatonate (^99mTc-Gluco) was used to provide the Tc(V) oxo species for radiolabeling the HYNIC-conjugated peptide (1, 7). ^99mTc-Gluco was prepared by adding ~300 mCi of ^99mTcO₄ in saline (~2.5 ml) to freeze-dried kits. The radiochemical purity of the product was >95% by instant thin-layer silica gel chromatography with acetone and saline as mobile phase solvents.

Myeloperoxidase activity. Excised pancreatic and pulmonary tissues were rinsed with saline, blotted dry, snap frozen in liquid nitrogen, and stored at 80°C until they were thawed for determination of myeloperoxidase activity by use of methods previously described with minor modifications (13, 25). Pancreatic or pulmonary tissue was homogenized in 0.1 M sodium phosphate buffer containing 0.5% hexadecyl trimethyl ammonium bromide and 5% soybean trypsin inhibitor (both from Sigma Chemical) before sonication and then directly frozen on dry ice. The specimens were freeze-thawed three times, and sonication was repeated after each cycle. Suspensions were then centrifuged at 20,000 g for 15 min, and the resulting supernatant was assayed. Myeloperoxidase activity was measured with a spectrophotometer (model UV-160, Shimazdu) at 470 nm by mixing an aliquot (25 µl) of the supernatant with 1.0 ml of 0.1 M sodium phosphate buffer (pH 7.0) containing 0.0016 ml of guaiacol (Sigma Chemical) and 0.0005% hydrogen peroxide (Sigma Chemical) as substrates. Myeloperoxidase activity is expressed as units per milligram wet weight of tissue.

Edema assessment. Pancreatic and pulmonary edema were evaluated by measuring the wet-to-dry weight ratio. Portions of the pancreatic tail and the left lung were removed after the animals were killed, trimmed of fat, and weighed. The water content was determined by calculating the wet-to-dry weight ratio from the initial weight (wet weight) and its weight after incubation at 160°C for 24 h (dry weight).

Histopathological evaluation. For histological evaluation the pancreas was fixed in 10% phosphate-buffered Formalin and embedded in paraffin. One coronal section was made in the plane of the flattened pancreas and stained with hematoxylin-eosin.

TAP measurement. TAP specifically reflects the conversion of trypsinogen to trypsin by cleavage of the TAP fragment (11). Blood samples (0.5 ml) for the measurement of TAP were collected in 0.2 mol/l EDTA to inactivate peptidases. After centrifugation of the samples (500 g, 10 min, 4°C) the remaining serum was coded and stored at 20°C until assayed.

Statistical analysis. Values are means ± SE. ANOVA was used to show an overall difference between groups, Student's t-test to make pairwise comparisons of normal distributed parameters, and regression analysis to evaluate the correlation between ^99mTc radioactivity and myeloperoxidase activity in pancreatic tissue. P < 0.05 was considered to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Absolute uptake of ^99mTc in the pancreas (percent injected dose/g) increased in proportion to the severity of pancreatitis (Fig. 2A). Uptake was increased in edematous (P = 0.04) and necrotizing (P < 0.001) pancreatitis compared with controls, with significantly higher levels in necrotizing than in edematous pancreatitis (P = 0.001). To assess whether the pancreatic uptake was affected by intravascularly retained ^99mTc-fMLFK-HYNIC, pancreas-to-blood ratios of ^99mTc activity were calculated. Pancreas-to-blood ratios were significantly increased in severe pancreatitis (P < 0.001 vs. control, P = 0.002 vs. cerulein) but showed only a trend in mild pancreatitis (Table 1). Pancreas-to-blood ratios were similar in animals with severe pancreatitis killed at 6 and at 12 h (Table 1).

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Chemotactic peptide uptake [A; expressed as percent injected dose (ID)/g] and myeloperoxidase activity (B) in pancreas. Uptake of chemotactic peptide 99mTc-fMLFK-HYNIC and myeloperoxidase activity levels in pancreatic tissue were increased in mild (cerulein) and severe (GDOC) pancreatitis (* P = 0.04;  P < 0.001 vs. control), with higher levels in severe than in mild disease ( P  0.03).

Myeloperoxidase activity in the pancreas demonstrated increasing leukocyte sequestration with worsening of pancreatitis (Fig. 2B). Although there was a moderate increase in cerulein pancreatitis (P < 0.001 vs. control), a major increase of leukocyte sequestration was found in necrotizing pancreatitis (P < 0.001 vs. control, P = 0.03 vs. cerulein). Myeloperoxidase levels in the pancreas of animals with severe pancreatitis that were killed at 12 h showed no significant difference from levels in animals killed at 6 h (18.7 ± 1.4 and 15.9 ± 1.6, respectively).

Pancreatitis-associated inflammatory response in the lungs, as reflected by pulmonary myeloperoxidase activity, was present in severe pancreatitis (P < 0.001 vs. control, P = 0.03 vs. cerulein) and to a lesser degree in mild pancreatitis (P = 0.02 vs. control; Fig. 3A). Absolute uptake of ^99mTc-labeled chemotactic peptide in the lungs showed similar patterns, with increased levels in animals with severe pancreatitis (P = 0.04 vs. control, P = 0.03 vs. cerulein; Fig. 3B). Lung-to-blood ratios of ^99mTc activity were only slightly higher in severe pancreatitis than in mild pancreatitis or controls (Table 1).

View larger version (35K): [in this window] [in a new window]   Fig. 3.   Myeloperoxidase activity (A) and chemotactic peptide uptake (B) in lung. Similar to myeloperoxidase activity in pancreatic tissue, myeloperoxidase activity in lung tissue was increased in mild (cerulein) and severe (GDOC) pancreatitis (* P = 0.02;  P < 0.001 vs. control). Levels were higher in severe than in mild disease ( P = 0.03). Uptake of chemotactic peptide in lung showed increased levels only in severe pancreatitis (* P = 0.04 vs. control;  P = 0.03 vs. cerulein).

No significant differences were found in chemotactic peptide uptake in muscle or muscle-to-blood ratio between the two groups of pancreatitis and controls (Table 1), indicating that the uptake of chemotactic peptide in pancreas and lungs is organ specific and not due to nonspecific processes. Slightly higher levels of uptake of ^99mTc in blood were found in animals with mild and severe pancreatitis than in controls (controls: 0.36 ± 0.07% injected dose/ml; cerulein: 0.41 ± 0.04% injected dose/ml; glycodeoxycholic acid: 0.50 ± 0.05% injected dose/ml; differences not statistically significant).

Figure 4 assesses the correlation between tissue uptake of chemotactic peptide and leukocyte sequestration and demonstrates a highly significant positive correlation (r = 0.7395, P < 0.001) between ^99mTc-labeled fMLFK uptake and myeloperoxidase activity in pancreatic tissue.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Correlation between chemotactic peptide uptake and myeloperoxidase activity: regression analysis between uptake of chemotactic peptide and leukocyte sequestration as assessed by myeloperoxidase activity in pancreatic tissue. Solid line, curve of best fit; dashed lines, 95% confidence limits.

Figure 5 shows a gamma camera image of pancreata from control animals and animals with pancreatitis (Fig. 5A) and corresponding histopathological sections (Figs. 5, B-D). Although in the control pancreas grains were predominantly localized to the spleen, there were relatively more grains over the pancreas in mild and severe pancreatitis. The amount of grains localized to the pancreas was larger in severe than in mild disease. No major grain localization was seen over the duodenal loop in either group. Histological sections showed regular pancreatic morphology in controls (Fig. 5B). Interstitial edema and moderate leukocyte infiltration were present in mild pancreatitis (Fig. 5C), whereas marked leukocyte infiltration and acinar cell necrosis were characteristic findings in severe pancreatitis (Fig. 5D).

View larger version (152K): [in this window] [in a new window]   Fig. 5.   Gamma camera imaging of pancreas (A) and corresponding histology (B-D). Because rat pancreas is a thin organ obscured anteriorly by liver, where chemotactic peptide uptake is high, pancreas of controls and animals with mild edematous (cerulein) and severe necrotizing (GDOC) pancreatitis was excised en bloc with adjacent spleen and duodenal loop (A). Histopathological evaluation of correlating pancreatic tissue showed regular pancreatic morphology in controls (B). Mild pancreatitis was characterized by interstitial edema and moderate leukocyte infiltration (C). In severe pancreatitis, marked leukocyte infiltration and extended areas of acinar necrosis were present (D).

Edema formation, as quantitated by wet-to-dry weight ratio, was found in pancreata of all animals with pancreatitis (P < 0.001 vs. control; Table 2). Edema was significantly less in necrotizing pancreatitis than in edematous pancreatitis (P = 0.002), confirming published results (25). No significant differences were found in pulmonary edema between controls and animals with mild or severe pancreatitis killed at 6 h, but significant pulmonary edema was found in animals with necrotizing pancreatitis killed at 12 h (P = 0.001 vs. control).

The concentration of TAP in plasma, reflecting trypsinogen activation to active trypsin, was increased in all animals with acute pancreatitis compared with controls (Table 2). TAP concentrations were higher in necrotizing than in edematous pancreatitis (P < 0.001), confirming the severity of pancreatitis.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Radiolabeled chemotactic peptides have been proposed as localizing agents for inflammation and infection (8, 24), but there is only indirect evidence that they are retained in inflamed tissue by high-affinity binding to fMLF receptors on the membranes of leukocytes (polymorphonuclear neutrophils and mononuclear phagocytes). Previous studies have shown that the chemotactic peptide ^99mTc-fMLFK-HYNIC localizes with high target-to-background ratios in bacterially or chemically induced muscle inflammation (3, 6, 24). Coinjection of a cold peptide (moderate antagonist) (3) or injection of a control peptide with low receptor-binding affinity (24) resulted in significantly lower target-to-background ratios, suggesting that the retention of the high-affinity peptide fMLFK is due to receptor-specific binding at the inflammatory site. However, receptor specificity and leukocyte binding have not been completely established, and significant amounts of peptide accumulation may be due to nonspecific interactions or leakage into the inflammatory site coincident with edema formation.

In the present study we have used two well-established experimental models of acute pancreatitis to evaluate the characteristics of ^99mTc-labeled chemotactic peptide fMLFK in localizing inflammatory sites. The first model is one of mild edematous pancreatitis, characterized by hyperemia, pronounced edema, hyperamylasemia, and moderate trypsinogen activation (14, 15, 19), findings confirmed in this study. In contrast, the second model results in necrotizing pancreatitis with markedly impaired pancreatic microcirculatory perfusion, significantly less tissue edema, more trypsinogen activation, and concomitant development of lung injury, reflecting a systemic inflammatory response (10, 12, 22). Pancreatic leukocyte infiltration occurs in both models, but to a much greater degree in severe than in mild pancreatitis. Pulmonary leukocyte sequestration reminiscent of acute respiratory distress syndrome-related lung injury (9) is also greater in necrotizing than in edematous pancreatitis (10). A recent study from our laboratory confirmed this leukocyte distribution pattern by using in vitro radiolabeled white blood cells (25).

The present study demonstrates that pancreatic uptake of the chemotactic peptide ^99mTc-fMLFK-HYNIC quantitatively correlates with leukocyte infiltration in acute pancreatitis. Leukocyte infiltration was quantified by measurement of myeloperoxidase activity, an enzyme stored in the azurophilic granules of polymorphonuclear leukocytes (13), and confirmed by histology. With leukocyte infiltration being more pronounced in severe than in mild pancreatitis, our data provide strong evidence that chemotactic peptides can be used to differentiate between different severities of this disease. This is also the first study showing that chemotactic peptides localize at an acute intra-abdominal inflammatory site. However, and more importantly, significantly increased pancreas-to-blood ratios in severe pancreatitis indicate high target-to-background ratios necessary for using chemotactic peptides in external imaging.

Other microcirculatory phenomena that may influence the trafficking of chemotactic peptides to leukocytes, such as alterations in endothelial permeability or capillary perfusion, may affect the uptake or concentration of chemotactic peptides at sites of inflammation. In our models, pancreatic edema formation was more prominent in edematous than in severe pancreatitis (25). Nonetheless, there was greater uptake of chemotactic peptide in the pancreas in severe than in mild pancreatitis, correlating with the greater myeloperoxidase activity. The finding confirms that leukocyte accumulation is more important than plasma leakage in chemotactic peptide localization (24). Although the study does not directly address whether alterations of tissue perfusion at the inflammatory site may affect chemotactic peptide uptake, the two models of pancreatitis used in this study have been well characterized previously using intravital microscopy and diffuse reflectance spectroscopy (12, 14). In mild edematous pancreatitis, pancreatic blood flow is temporarily increased, whereas capillary perfusion is markedly reduced and even shut down in severe pancreatitis. Inasmuch as these perfusion abnormalities are discordant with the relative uptake of chemotactic peptide observed in these experiments, we infer that they cannot account for the findings. The combined weight of evidence and each of its elements are consistent with a specific quantitating link between chemotactic peptide concentration and leukocyte sequestration.

The results obtained by using chemotactic peptides to localize pancreatic inflammation are similar to those of a previous study using in vitro radiolabeled white blood cells (25). Such labeled white blood cells accumulate in an inflammatory locus only after replacement in the circulation and subsequent migration to the extravascular destination. As a consequence, radiolabeled white blood cells required injection before induction of pancreatitis in the previous study. Because chemotactic peptides are very small molecules that can diffuse rapidly through tissues, they will bind to circulating leukocytes and to extravascular cells that have already localized at a site of inflammation (2, 7). In the present study, chemotactic peptides were injected well after the induction of the disease and still localized successfully in the pancreas and the lungs. Myeloperoxidase levels had reached a plateau by 6 h after induction of severe pancreatitis and remained stable for the next 6 h, suggesting that leukocyte migration was at its maximum in this time frame; chemotactic peptide levels mirrored the finding.

Unlike in vitro radiolabeled white blood cells, chemotactic peptides can be readily synthesized and radiolabeled, and the technique does not involve the potential hazards associated with handling of blood and reinjection of labeled cells. Additionally, special care must be taken to avoid damage and activation of the cells in the processing of in vitro radiolabeled leukocytes. Cell damage and activation induced by in vitro procedures lead to initial entrapment of labeled cells in the lungs after their injection, with their subsequent delayed release into the systemic circulation (20).

Chemotactic peptides have been investigated principally for imaging local inflammatory processes but not associated remote organ injury. The systemic inflammatory response to severe local inflammation may produce other sites of leukocyte accumulation, as illustrated in the lungs in our model of necrotizing pancreatitis. Chemotactic peptide uptake and myeloperoxidase activity in the lungs were significantly increased in severe pancreatitis. Uptake of chemotactic peptide in the lungs was not caused by the development of vascular leakage, inasmuch as no significant pulmonary edema formation was detected in any group at 6 h. The absence of increased chemotactic peptide uptake into muscle in animals with severe acute pancreatitis indicates that the pulmonary uptake is organ specific and not due to generalized systemic leukocyte sequestration.

This study thus demonstrates that the uptake of ^99mTc-labeled chemotactic peptide fMLFK correlates quantitatively with leukocyte sequestration at specific inflammatory sites and is unaffected by nonspecific microcirculatory events. Their localization in acute experimental pancreatitis and associated lung injury suggests that chemotactic peptides have potential use in demonstrating and perhaps assessing the degree of injury in intra-abdominal inflammatory processes and distant organs. Future studies are needed to validate the application of chemotactic peptides for external imaging in human pancreatitis and other intra-abdominal inflammatory processes.