Inhibition of Intestinal Transit by Resuscitation-Induced Gut Edema is Reversed by L-NIL1,2

Journal of Surgical Research 129, 1–5 (2005) doi:10.1016/j.jss.2005.04.041

Inhibition of Intestinal Transit by Resuscitation-Induced Gut Edema is Reversed by L-NIL 1,2 S. D. Moore-Olufemi, M.D.,*,‡ H. Xue, M.D.,* S. J. Allen, M.D.,†,‡,§ F. A. Moore, M.D.,*,‡ R. H. Stewart, D.V.M., Ph.D.,§ G. A. Laine, Ph.D.,§ and C. S. Cox, Jr., M.D.,*,‡,§,3 *Department of Surgery, †Department of Anesthesiology, ‡Trauma Research Center, University of Texas–Houston Medical School, Houston, Texas; and §The Michael E. DeBakey Institute, Texas A&M University, College Station, Texas Submitted for publication March 9, 2005

transit and elevated iNOS protein expression. Pretreatment with L-NIL improved intestinal transit and decreased expression of iNOS protein without decreasing intestinal tissue water compared to edema animals. There was no difference in mucosal injury or MPO activity among groups. Conclusion. Gut edema delays intestinal transit via an iNOS-mediated mechanism. © 2005 Elsevier Inc. All rights

Background. Post-resuscitation gut edema and associated gut dysfunction is a common and significant clinical problem that occurs after traumatic injury and shock. We have shown previously that gut edema without ischemia/reperfusion injury delays intestinal transit [1]. We hypothesized that gut edema increases expression of inducible nitric oxide synthase (iNOS) protein, and that selective iNOS inhibition using L-NIL reverses the delayed intestinal transit associated with gut edema. Materials and methods. One hour prior to laparotomy, rats were pretreated with 10 mg/kg body weight of intraperitoneal L-NIL or saline vehicle and underwent 80 ml/kg body weight of 0.9% saline ⴙ superior mesenteric venous pressure elevation (Edema) or sham surgery (Sham). A duodenal catheter was placed to allow injection of a fluorescent dye for the measurement of intestinal transit. At 6 h, the small bowel was divided and the mean geometric center (MGC) of fluorescent dye was measured to determine transit. Ileum was harvested for histological assessment of mucosal injury, evaluation of iNOS protein expression by Western blotting, and MPO activity. Tissue water was determined using the wet-to-dry weight ratio to assess gut edema. Data are expressed as mean ⴞ SEM, n ⴝ 3– 6 and * ⴝ P <0.05 using ANOVA. Results. Gut edema, expressed as increased wet-todry ratio, was associated with decreased intestinal

reserved.

Key Words: ileus; resuscitation; edema; iNOS; L-NIL; trauma. INTRODUCTION

Ileus is a common problem that occurs after traumatic injury and shock, which can delay recovery and increase morbidity. Crystalloid-based fluid resuscitation after traumatic injury often results in gut edema, which can contribute to the abdominal compartment syndrome, requiring decompressive laparotomy [2]. However, even lesser degrees of post-resuscitation gut edema are associated with ileus, preventing the delivery of enteral nutrition leading to increased morbidity [3]. Acute alterations of Starling forces due to abdominal packing, increased capillary pressure, with venous outflow obstruction and crystalloid infusion (decreased plasma oncotic pressure) can cause gut edema. We have previously described a model of acute interstitial hydrostatic edema that mimics the causes of gut edema in damage-control surgical procedures (elevated abdominal venous pressures and large-volume crystalloid resuscitation), which, in the absence of gut ischemia/ reperfusion, resulted in decreased intestinal transit [1, 4]. Interestingly, two separate interventions that reduce gut edema after intestinal manipulation have

1

Supported by NIGMS Grants T32 GM 08792, P50 GM38529, K08 GM00675, and CDC Grant CCU-620069, and NHLBI Grant RO1 HL-36115. 2 Presented at the Association for Academic Surgery Meeting, November 11–13, 2004 in Houston, Texas. 3 To whom correspondence and reprint requests should be addressed at Department of Surgery, Division of Pediatric Surgery, UT–Houston Medical School, 6431 Fannin Street, Suite 5.246, Houston, Texas 77030. E-mail: [email protected].

1

0022-4804/05 $30.00 © 2005 Elsevier Inc. All rights reserved.

2

JOURNAL OF SURGICAL RESEARCH: VOL. 129, NO. 1, NOVEMBER 2005

been shown to improve intestinal transit [5– 6]. Harada et al. used the diuretic ethacrynic acid in a rat model of intestinal manipulation that is known to cause ileus. Use of the diuretic decreased inflammation and improved intestinal transit. Similarly, Overhaus et al. used the same model but treated the animals with hypertonic saline. Hypertonic saline decreased inflammation and improved intestinal transit. While both interventions reduced the inflammatory response to intestinal manipulation, both interventions are known to reduce edema. Thus it is reasonable to consider that gut edema may be both an effect of injury and also an amplifier of the injury. Further, gut edema is a potential therapeutic target. Little is known about the cellular mechanisms by which gut edema affects intestinal transit. Increased expression of ileal inducible nitric oxide synthase (iNOS) and subsequent nitric oxide production serves to increase enterocyte and smooth muscle cGMP. This signaling pathway has been shown to inhibit smooth muscle contractility by inhibiting myosin light chain phosphorylation to prevent actin/myosin crossbridging [7– 8]. Previous work at our institution and others has demonstrated that iNOS expression is a critical mediator of ileus after hemorrhagic shock/ resuscitation, gut ischemia/reperfusion, or direct manipulation of the small intestine [9 –12]. Evidence points to resident macrophages and/or recruited leukocytes as the source of iNOS [11–12]. l-N6-(1iminoethyl)-lysine (L-NIL) is an l-arginine analogue that is a slow-binding, competitive iNOS inhibitor with a 30-fold selectivity for iNOS over nNOS or eNOS. We hypothesized that up-regulation of ileal iNOS protein plays a role in delayed intestinal transit associated with gut edema and selective iNOS inhibition using L-NIL would reverse gut edema-associated ileus. MATERIALS AND METHODS All procedures were approved by the University of Texas Animal Welfare Committee and were consistent with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Experimental groups. The four following experimental groups were studied: (1) Sham—rats underwent anesthesia, laparotomy, and instrumentation without mesenteric venous pressure elevation or resuscitation; (2) Sham ⫹ L-NIL—rats underwent the same protocol as group number 1 but with intraperitoneal L-NIL pretreatment at a dose of 10 mg/kg body weight; (3) Edema—rats underwent anesthesia, instrumentation, and mesenteric venous pressure elevation with crystalloid resuscitation (described below); (4) Edema ⫹ L-NIL—rats underwent the same protocol as group number 3 but with intraperitoneal L-NIL pretreatment at a dose of 10 mg/kg body weight. Acute gut edema preparation. Male Sprague Dawley rats, weighing between 300 and 350 g, were fasted 16 –18 h prior to surgery and given free access to water. Under general anesthesia with isoflurane, a midline laparotomy incision was made using aseptic technique. Each rat had a silastic catheter introduced into the mid-duodenum by needle puncture, advanced 1 cm distally, and fixed to the duodenal wall with a 6-O silk purse-string suture. The catheter was then passed through the musculature of the left abdominal wall and subcutaneous tissue toward

the back of the neck where it was exteriorized through an interscapular incision and fixed to the skin with 4-O silk suture. A rubber cap was used to seal the exposed end of the catheter. At the time of instrumentation, rats had a 4-O silk ligature placed around the superior mesenteric vein (SMV) and tied over a PE-10 silastic tube, which was removed after ligature was secured. This created acute mesenteric venous hypertension without occluding mesenteric venous blood flow. Post instrumentation, an external jugular venous catheter was placed, and 80 ml/kg body weight of 0.9% saline was administered in the Edema and Edema ⫹ L-NIL groups. Intestinal transit. Five and one-half hours after instrumentation, 0.1 ml of a 5 mm solution of nonabsorbable FITC-Dextran (FD4, MW ⫽ 9400, FITC content 0.008 mol/mol glucose; Sigma-Aldrich, St. Louis, MO) was injected into the catheter and flushed with 0.1 ml normal saline. Twenty-five minutes later, rats were anesthetized with an intraperitoneal injection of ketamine and the entire small intestine was removed and divided into 10 equal segments. The intraluminal contents were flushed with 3 ml of 5 mm Tris buffer (pH 10.3) to recover the FITC-Dextran. The FITC-Dextran concentration was measured using an optical scanner (STORM model 860; Amersham Biosciences, Piscataway, NJ) and expressed as a fraction of total tracer recovered and presented as the geometric center of distribution [1, 13]. At the conclusion of the procedure, the rats were euthanized. Intestinal tissue water. Intestinal samples were opened along the antimesenteric border and blotted dry. Wet weight was determined prior to placing tissue in an oven set to 60°C. Tissues were dried over 2–3 days to a constant dry weight and dry weights were measured and used to determine tissue water [(wet weight) – (dry weight)/dry weight]. Cytoplasmic and nuclear extract isolation and determination. Cytosolic and nuclear extracts were prepared using a modified method described by Deryckere and Gannon [14]. Frozen distal ileal segments (250 mg) were homogenized in 3 ml of buffer A [0.6% Nonidet P-40, 150 mmol/L NaCl, 10 mmol/L HEPES, pH 7.9, 1 mmol/L EDTA, pH 8.0, and 0.5 mmol/L phenyl methyl sulfonyl fluoride (PMSF)] containing 30 ␮l/ml of protease inhibitor cocktail (Sigma Chemical Co., St. Louis, MO). The samples were centrifuged at 2000 rpm for 1 min and incubated on ice for 20 min. After removal of the supernatant (cystolic fraction), the residual pellet was transferred to a microfuge tube containing buffer B (25% glycerol, 20 mmol/L HEPES, pH 7.9, 420 mmol/L NaCl, 1.2 mmol/L MgCl 2, 0.2 mmol/L EDTA, pH 8.0, 0.5 mmol/L DTT, and 0.5 mmol/L PMSF) and centrifuged at 12,000 rpm for 10 min. The cystolic extract was used to determine total protein and MPO concentrations. Total protein concentration was determined using the Bio-Rad microtiter plate assay protocol (Bio-Rad, Hercules, CA). iNOS protein expression. Total protein (200 ␮g) was loaded on a 7.5 or 10% precast SDS-PAGE gel (Bio-Rad, Hercules, CA) and electrophoresed in 1⫻ Tris buffer. The protein was transferred to nitrocellulose membranes and probed with anti-iNOS polyclonal (1: 1000 dilution; Cayman) antibody overnight. The blot was incubated with appropriate secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h. The secondary reaction was visualized with ECL (Amersham, Piscataway, NJ). iNOS densitometry was analyzed by image-analysis software (Optimas 6.1, Media Cybernetics, Silver Spring, MD) and expressed in arbitrary units. Histology. Ileal segments were stored in 10% formalin until processing. Tissue was embedded in paraffin blocks, sectioned in 5-␮m slices, placed on glass microscope slides, and stained with hematoxylin and eosin (H&E). Tissues were examined by light microscopy by a blinded pathologist and scored using a system described by Chiu et al. [15]: Grade 0 ⫽ normal mucosa; Grade 1 ⫽ subepithelial space developing at the tip of the villus; Grade 2 ⫽ lifting of the epithelial layer from the lamina propria and moderate extension of the subepithelial space; Grade 3 ⫽ some denuded tips of the villi and massive lifting of the epithelial layer; Grade 4 ⫽ dilated and exposed capillaries and denuded villi; and Grade 5 ⫽ hemorrhage, ulceration, and disintegrated lamina propria.

MOORE-OLUFEMI ET AL.: RESUSCITATION-INDUCED ILEUS

3

TABLE 1 Tissue Water Expressed as Wet:Dry Weight Ratios Sham

Sham/L-NIL

Edema

Edema/L-NIL

2.98 ⫾ 0.08*

3.32 ⫾ 0.05

4.28 ⫾ 1.24

3.81 ⫾ 0.21

*P ⬍ 0.05 compared to all other groups. Myeloperoxidase activity. Myeloperoxidase (MPO) enzyme activity was measured using the H 2O 2-dependent oxidation of 3,3=,5,5=tetramethylbenzidine (TMB) adapted from Suzuki et al. and indexed to the tissue wet weight [16]. MPO activity was assayed by adding 10 ␮l of supernatant from each sample and 100 ␮l of SureBlue TMB 1-Component Microwell Peroxidase (KPL, Gaithersburg, MD) in duplicate to a 96-well microplate. The reaction was terminated with 100 ␮l of 0.18M sulfuric acid and the change in absorbance at 450 nm was measured with a Kinetic Microplate Reader (Molecular Devices Corporation, Sunnyvale, CA) spectrophotometer. Statistical analysis. All data are expressed as mean ⫾ SEM using a commercial statistical software program (NCSS, Kaysville, UT). Statistical significance of differences among groups was determined by analysis of variance (ANOVA) followed by Duncan’s and Tukey–Kramer multiple comparison tests. A P value ⬍0.05 was considered significant.

RESULTS L-NIL

Does Not Prevent the Development of Gut Edema

As expected, the tissue water weight in Edema rats was ⬃70% greater than Sham animals (Table 1). However, pretreatment with L-NIL did not prevent the fluid accumulation in the intestine associated with venous hypertension and crystalloid resuscitation (Edema). L-NIL

FIG. 1. L-NIL improves delayed intestinal transit associated with resuscitation-induced gut edema. Rats were subjected to sham surgery with and without L-NIL pretreatment (Sham and Sham/LNIL), mesenteric venous pressure elevation, and crystalloid resuscitation with and without L-NIL pretreatment or vehicle (Edema, Edema/L-NIL). Intestinal transit of tracer along intestinal segments was then measured as detailed in Materials and Methods. The mean geometric center (MGC) of tracer is presented. *P ⬍ 0.05 for Edema versus Sham, Sham/L-NIL, and Edema/L-NIL; n ⫽ 5– 6. L-NIL

Does Not Affect Gut Myeloperoxidase Activity

There was no increase in MPO activity (Fig. 4) with gut edema compared to Sham or Sham ⫹ L-NIL, and there was no difference between Edema and Edema/LNIL animals.

Improves Delayed Intestinal Transit Associated with Gut Edema

Consistent with previous results [1, 9], the baseline mean geometric center (MGC) in Edema rats was ⬃45% lower than that of Sham animals, indicating delayed intestinal transit in rats with gut edema (Fig. 1). However, in rats with gut edema pretreated with L-NIL that underwent mesenteric pressure elevation, the MGC was comparable to that of Sham animals. There was no difference in intestinal transit in Sham rats pretreated with L-NIL. Gut Edema Increased iNOS Protein Expression

As depicted in Fig. 2, Edema was associated with increased expression of iNOS, which was blocked by L-NIL. Gut Edema Is Not Associated with Mucosal Injury

Representative hematoxylin and eosin stained sections of ileum from Sham, Sham/L-NIL, Edema, Edema/L-NIL, and Edema/Vehicle rats are depicted in Fig. 3. The ileum of Sham animals exhibited normal mucosal architecture with intact villi.

FIG. 2. (A & B) Resuscitation-induced gut edema is associated with increased ileal iNOS expression. Rats were subjected to sham surgery with and without L-NIL pretreatment (Sham and Sham/L-NIL), mesenteric venous pressure elevation, and crystalloid resuscitation with and without L-NIL pretreatment or vehicle (Edema, Edema/L-NIL). (A) Representative iNOS Western blot using 200 ␮g of cytoplasmic extracts. (B) iNOS densitometry using representative Western blot from (A) and expressed in arbitrary units. *P ⬍ 0.05 comparing Edema to all other groups.

4

JOURNAL OF SURGICAL RESEARCH: VOL. 129, NO. 1, NOVEMBER 2005

FIG. 3. (A & B) Resuscitation-induced gut edema does not cause mucosal injury. Rats were subjected to sham surgery with and without L-NIL pretreatment (Sham and Sham/L-NIL), mesenteric venous pressure elevation, and crystalloid resuscitation with and without L-NIL pretreatment or vehicle (Edema, Edema/L-NIL). (A) Formalinfixed, paraffin-embedded sections of ileum stained with hematoxylin and eosin. Magnification, ⫻20. n ⫽ 4 – 6 for each group. (B) Mucosal injury scores; there was no difference in mucosal injury among groups. n ⫽ 5– 6. (Color version of figure is available online.)

and ileus. Using the aforementioned model of acute hydrostatic gut edema, we have shown a reduction in ileal smooth muscle myosin light chain phosphorylation and a reduction in ileal myosin light chain kinase protein expression, correlating with a reduction in ex vivo ileal contractility [17]. Myosin light chain phosphorylation is required for actin/myosin cross-bridging during muscle contraction. NO/cGMP inhibits MLC phosphatase, which phosphorylates MLC, implying that this pathway is directly involved in impairing ileal smooth muscle contractility. This mechanistic link is in contrast to the PMN-mediated injury hypothesis whereby degranulating PMNs that have sequestered in the injured ileum inhibit contractility due to cellular injury. A common pathophysiologic link between ischemia/reperfusion and surgical manipulation of the intestine is thought to be recruitment of macrophages and neutrophils into the muscularis layer, decreasing the contractile properties of the muscle via injury [10 – 12]. Given that ileal MPO does not increase, it is unlikely that this mechanism of injury occurs with acute hydrostatic gut edema. Further evidence against PMN-mediated injury occurring with hydrostatic gut edema is the lack of histological evidence of cellular damage. An unexpected finding of this experiment was that L-NIL decreased iNOS protein expression. L-NIL is an l-arginine analogue thought to selectively downregulate iNOS enzymatic activity via competitive inhibition. However, our data are consistent with findings in other models that have used L-NIL. Specifically, McCartney-Francis et al. showed an unexpected decrease in iNOS protein expression in a model of joint inflammation after pretreatment with L-NIL; their findings were confirmed using QT PCR to demonstrate a reduction in iNOS message as well [18]. Recent work suggests that iNOS inhibition serves to decrease iNOS protein synthesis via a NO/cGMP/PKG pathway [19].

DISCUSSION

Our study showed that gut edema increased ileal iNOS protein expression and ileal iNOS inhibition with L-NIL improved the delayed intestinal transit associated with resuscitation-induced gut edema. As with previous studies, gut edema did not affect ileal MPO activity or cause histological mucosal injury. Prior studies have emphasized the role of ischemia/ reperfusion, hemorrhagic shock, and surgical manipulation in promoting injury-induced ileus [10 –12]. These models have implicated iNOS protein expression as a pivotal mediator of ileus. The source of ileal iNOS is thought to be from both ileal enterocytes as well as resident macrophages. The mechanism by which the iNOS/NO pathway affects gut motility has not been determined. However, initial work implicates the iNOS/NO/cGMP pathway as the critical mechanistic link between reduced ileal smooth muscle contractility

FIG. 4. L-NIL does not affect MPO activity. Rats were subjected to sham surgery with and without L-NIL pretreatment (Sham and Sham/L-NIL), mesenteric venous pressure elevation, and crystalloid resuscitation with and without L-NIL pretreatment or vehicle (Edema, Edema/L-NIL). There was no difference in MPO activity among groups. n ⫽ 5– 6.

MOORE-OLUFEMI ET AL.: RESUSCITATION-INDUCED ILEUS

Specifically, NO production as a result of increased iNOS activates protein G kinase. Protein G kinase is a serine-threonine kinase that can potentially modulate phosphorylation/activation of AP-1. Since AP-1 is a known transcription factor of iNOS, inhibition of NO production by any means, including competitive inhibition of iNOS using L-NIL, could serve to decrease iNOS transcription. Another possible explanation is that L-NIL down-regulates upstream signaling that leads to suppression of iNOS. We have recently shown that L-NIL decreases the DNA binding activity of one nuclear transcription factor for iNOS, STAT-3 (unpublished data), and STAT-3 DNA binding is associated with iNOS gene expression in a number of models [20]. Alternatively, L-NIL may decrease iNOS expression through another, unrelated pathway as a component of non-iNOS-specific effects. Further data that examine the effects of gut edema ⫹ L-NIL on iNOS mRNA could help define a mechanism by which L-NIL affects ileal iNOS protein. Is there biological plausibility for the observed phenomenon of ileal interstitial edema causing iNOS expression? Acute interstitial edema rapidly expands the interstitial matrix, presumably creating both a radial and a shear-type stress on the overlying enterocytes and smooth muscle. In vascular biology, shear stress applied to cultured smooth muscles induces iNOS expression in the smooth muscle cells [21]. Other alternative explanations include acute paracellular barrier dysfunction resulting in the initiation of a non-PMNmediated inflammatory cascade that results in iNOS expression. While we have shown that gut edema without injury increases paracellular permeability, it seems less likely that a proinflammatory cascade would be initiated without the chemotaxis and leukosequestration of PMNs in the ileum [1]. The data in this experiment are important because gut edema is a reasonable potential therapeutic target when designing resuscitation protocols. Strategies to limit gut edema using hypertonic/hyperoncotic fluids and/or early abdominal decompression to decrease mesenteric venous pressure may help to prevent resuscitation-induced ileus. Further elucidation of the mechanisms of gut edema-associated ileus could allow pharmacologic treatment of ileus if alternative resuscitation strategies were unsuccessful. In summary, our data show that the ileus associated with resuscitationinduced gut edema is prevented by selective iNOS inhibition. REFERENCES Moore-Olufemi, S. D., Xue, H., Attuwaybi, B. O., et al. Resuscitation-induced gut edema and intestinal dysfunction. J. Trauma 58: 264, 2005. 2. Balogh, Z., McKinley, B. A., Cox, C. S., Jr., et al. Abdominal

3.

4.

5.

6.

7.

8.

9.

10.

11. 12.

13.

14.

15. 16.

17.

18.

19.

20.

1.

21.

5

compartment syndrome: the cause or effect of post-injury multiple organ failure. Shock 20: 483, 2003. Balogh, Z., McKinley, B. A., Cocanour, C. S., et al. Supranormal trauma resuscitation causes more cases of abdominal compartment syndrome. Arch. Surg. 138: 637, 2003. Moore-Olufemi, S. D., Xue, H., Allen, S. J., et al. Effects of primary and secondary intra-abdominal hypertension on mesenteric lymphatic resistance and flow: Implications for the abdominal compartment syndrome. Shock (In press). Harada, T., Cruz, R. J., Delude, R. L., et al. Ethacrinic acid ameliorates ileus induce by bowel manipulation in mice. Crit. Care Med. 32: A93, 2004. Overhaus, M., Behrendt, F. I., Doering, J. G., et al. Hypertonic saline inhibits rat post-operative ileus. J. Am. Coll. Surg. (Surg. Forum) 199: S23, 2004. Attuwaybi, B., Kozar, R. A., Gates, K. S., et al. Hypertonic saline prevents inflammation, injury, and impaired intestinal transit after gut ischemia/reperfusion by inducing heme oxygenase 1 enzyme. J. Trauma 56: 749, 2004. Lincoln, T. M., Dey, N., and Sellak, H. cGMP-dependent protein kinase signaling mechanisms in smooth muscle: from the regulation of tone to gene expression. J. Appl. Physiol. 91: 1421, 2001. Nishimura, N., and van Breemen, C. Direct regulation of smooth muscle contractile elements by second messengers. Biochem. Biophys. Res. Commun. 163: 929, 1989. Hassoun, H. T., Weisbrodt, N. W., Mercer, D. W., et al. Inducible nitric oxide synthase mediates gut ischemia/reperfusioninduced ileus only after severe insults. J. Surg. Res. 97: 150, 2001. Turler, A., Moore, B. A., Pezzone, M. A., et al. Colonic postoperative inflammatory ileus in the rat. Ann. Surg. 236: 56, 2002. Schwarz, N. T., Kalff, J. C., Turler, A., et al. Selective jejunal manipulation causes postoperative pan-enteric inflammation and dysmotility. Gastroenterology 126: 159, 2004. Miller, M. S., Galligan, J. J., and Burks, T. F. Accurate measurement of intestinal transit in the rat. J. Pharmacol. Methods 6: 211, 1981. Deryckere, F., and Gannon, F. A one hour mini preparation technique for extraction of DNA-binding proteins from animal tissues. Biotechniques 16: 405, 1994. Chiu, C., McArdle, A. H., Brown, R., et al. Intestinal mucosal lesion in low-flow states. Arch. Surg. 101: 478, 1970. Suzuki, K., Ota, H., Sasgawa, S., et al. Assay method for myeloperoxidase in human polymorphonuclear leukocytes. Anal. Biochem. 132: 345, 1983. Uray, K., Allen, S. J., Laine, G. A., et al. Intestinal edema reduces intestinal motility through decreased myosin light chain phosphorylation. FASEB J. 19: A744, 2005. Yu, Z., and Kone, B. C. The STAT3 DNA-binding domain mediates interaction with NF-kappaB p65 and inducible nitric oxide synthase transrepression in mesangial cells. J. Am. Soc. Nephrol. 3: 585, 2004. Gertzberg, N., Clements, R., Jaspers I., et al. Tumor necrosis factor-alpha-induced activating protein-1 activity is modulated by nitric-oxide mediated protein kinase G activation. Am. J. Respir. Cell Mol. Biol. 22: 105, 2000. McCartney-Francis, N. L., Song, X., Mizel, D. E., et al. Selective inhibition of inducible nitric oxide synthase exacerbates erosive joint disease. J. Immunol. 166: 2734, 2001. Gosgnach, W., Messika-Zeitoun, D., Gonzalez, W., et al. Shear stress induces iNOS expression in cultured smooth muscle cells: role of oxidative stress. Am. J. Physiol. 279: C1880, 2000.