Inflammatory diseases influence tissue metabolism, altering regulation of extracellular adenine nucleotides, with a resultant protective influence of adenosine. Ecto-5′-nucleotidase (CD73) is a central surface enzyme generating extracellular adenosine. Thus, we hypothesized that CD73 is protective in mucosal inflammation as modeled by trinitrobenzene sulfonate (TNBS) colitis. Initial studies revealed a >3-fold induction of CD73 mRNA levels after TNBS colitis. Additionally, the severity of colitis was increased, as determined by weight loss and colonic shortening, in cd73−/− mice relative to cd73+/+ controls. Likewise, enteral administration of the selective CD73 inhibitor α,β-methylene ADP to cd73+/+ mice resulted in a similar increase in severity of TNBS colitis. Gene array profiling of cytokine mRNA expression, verified by real-time PCR, revealed a >90% down-regulation of IFN-αA in cd73−/− mice and α,β-methylene ADP-treated cd73+/+ mice, compared with cd73+/+ mice. Exogenous administration of recombinant IFN-αA partially protected TNBS-treated cd73−/− mice. Cytokine profiling revealed similar increases in both IFN-γ and TNF-α mRNA in colitic animals, independent of genotype. However, IL-10 mRNA increased in wild-type mice on day 3 after TNBS administration, whereas cd73−/− mice mounted no IL-10 response. This IL-10 response was restored in the cd73−/− mice by exogenous IFN-αA. Further cytokine profiling revealed that this IL-10 induction is preceded by a transient IFN-αA induction on day 2 after TNBS exposure. Together, these studies indicate a critical regulatory role for CD73-modulated IFNαA in the acute inflammatory phase of TNBS colitis, thereby implicating IFN-αA as a protective element of adenosine signaling during mucosal inflammation.

Inflammatory diseases of the bowel result from a dysregulation of the interaction between immune response and barrier function with many investigational models involving disruption of specific elements of either of these processes. Specifically, the trinitrobenzene sulfonate (TNBS)4 model of murine colitis involves induction of both acute and chronic intestinal inflammation via the combination of barrier disruption with ethanol and Ag-mediated stimulation of an inflammatory response in response to TNBS (1). The resultant inflammatory response is characterized by IL-12-driven transmural infiltrative disease histologically similar to Crohn’s disease, but confined primarily to the colon (1). Previous work has demonstrated that a subset of adaptive responses to TNBS colitis are associated with an alteration in the expression of a panel of epithelial genes with established roles in both barrier function and immune regulation (2). One such gene is ecto-5′-nucleotidase (CD73), the gene encoding the 5′-ectonucleotidase, a critical enzyme in the breakdown of ATP to adenosine (Ado) (3).

There are several existing lines of evidence arguing for a role for CD73 in the regulation of the interaction between barrier and inflammation in the intestinal mucosa. It has been demonstrated that CD73 is extensively expressed on the luminal surface of colonic enterocytes (4). The role of CD73 in both endothelial (5, 6) and epithelial barrier function (7) has previously been demonstrated, and a further function in the regulation of neutrophil (PMN)-epithelial crosstalk has also been described (4, 5, 8). Generation of the cd73−/− mouse (9) has made it possible to further investigate the influence of global alterations in CD73 expression on inflammation and barrier. One resultant study has demonstrated the role of CD73 in the modulation of Ado-mediated PMN transendothelial migration (8). Specifically, this work revealed that CD73 activity is important for blocking/attenuating the posthypoxic exit of PMNs from the vasculature into both lung and colon (8). In light of its roles in the modulation of processes relevant to the onset of intestinal inflammation, such as barrier function and leukocyte trafficking, concomitant with its abundant expression on colonic epithelial cells we sought to further characterize CD73 activity in the context of murine TNBS-induced experimental colitis and to investigate its potential involvement in the regulation of other mediators of the innate immune response, such as cytokines.

Although a role for CD73 in cell signaling has been hypothesized (3), previous studies have focused on the direct effect of CD73 activity on downstream Ado production (7) and Ado receptor activation (5). Currently, there is little evidence connecting CD73 regulation itself to other downstream mediators of the immune response. In these studies, we use a gene array approach to define differences in inflammatory mediator expression profile in cd73−/− mice relative to cd73+/+ controls and determine that few differences exist with the intriguing exception of a significant reduction in the expression of the cytokine IFN-αA.

Type I IFNs have been of particular interest with respect to the regulation of inflammation with conflicting clinical and experimental evidence relating to their roles in intestinal inflammation. One of the more established therapeutic indications for IFN-α has taken advantage of its antiviral properties in the long-term treatment of viral hepatitides (10). However, case reports indicate that an unwelcome side effect of this therapy is the spontaneous development of an intestinal inflammatory condition which is similar to ulcerative colitis (11, 12, 13), but reversible on discontinuation of IFN therapy.

In contrast to these findings, type I IFNs have been found to be beneficial in treatment of viral-associated bowel inflammation such as herpes-associated colitis (14). Furthermore, in vitro comparison of virus-stimulated cytokine responses reveals decreased production of IFN-α by human intestinal monocytes isolated from patients with Crohn’s disease relative to controls (15). Subsequent clinical trials investigating potential benefits of type I IFN therapy in inflammatory bowel disease have yielded conflicting results with little benefit reported in Crohn’s disease (16), but more promising results in some studies looking at either IFN-β1 (17) or IFN-α2A (18, 19) in ulcerative colitis. Additional animal studies demonstrate a type I IFN-mediated anti-inflammatory effect in murine colitis models (20, 21). In view of these conflicting findings, the factors governing intestinal regulation of type I IFNs and their contributions to the regulation of mucosal inflammation remain an area of active interest.

In these studies, we demonstrate enhanced CD73 expression in TNBS colitis and further show an exacerbation of TNBS colitis in cd73−/− animals relative to wild-type controls. We then use gene array analysis to characterize global changes in cytokine expression associated with the cd73−/− genotype. Thus, we identify a significant decrease in colonic mucosal mRNA expression of the type I IFN in the CD73null phenotype. We subsequently determined that the enhancement of TNBS colitis is recapitulated by blockade of CD73 activity using the selective inhibitor α,β-methylene ADP (APCP). Furthermore, we demonstrate that inflammation, based on weight loss and colon length, can be attenuated back to the level of cd73+/+ controls in the CD73null animals via treatment with IFN-αA. However, IFN-αA treatment of cd73+/+ animals exposed to TNBS failed to provide the same protective effects in our model.

Further cytokine mRNA profiling indicates that whereas wild-type and cd73−/− animals have similar baseline expression of inflammatory cytokines, the cd73−/− animals are uniquely deficient in their ability to induce IL-10 mRNA production in response to TNBS exposure. Furthermore, this inability to mount an IL-10 response is rescued by exogenous IFN-αA administration in the cd73−/− mice. However, exogenous IFN-αA fails to significantly alter the IL-10 response of wild-type mice.

Given this finding that the IFN-αA-deficient cd73−/− mice are deficient in IL-10 mRNA induction, whereas the wild-type mice demonstrate a down-regulation of IFN-αA mRNA on day 3 after TNBS exposure, we sought to determine whether the intact IL-10 response might in fact be associated with an induction of IFN-αA earlier in the colitis course in the wild-type mice. Examination of the kinetics of the IFN-αA mRNA response in the wild-type mice demonstrated that the wild-type mice transiently but significantly induced IFN-αA on day 2 after TNBS colitis followed by a subsequent down-regulation of IFN-αA mRNA expression by day 3. These findings suggest a critical role for IFN-αA in the acute inflammatory phase of TNBS colitis, contributing to the induction of the anti-inflammatory cytokine IL-10, necessary for inflammatory resolution. The transient nature of this IFN-αA induction and its subsequent rapid down-regulation in the wild-type mice suggests the existence of a critical window for the IFN-αA response in the regulation of acute inflammation and may have serious implications for dosing and safety of IFN-αA as a therapeutic agent in ulcerative colitis.

Mice with homozygous deficiency in cd73 (cd73−/−) on the C57BL/6 SVJ strain were generated and characterized as previously described (9) and maintained by homozygous crosses. Control C57BL/6 mice (cd73+/+) were matched for age and gender in each experiment. All procedures involving animals were performed according to National Institutes of Health guidelines for use of live animals and were approved by the Institutional Animal Care and Use Committee at Brigham and Women’s Hospital.

TNBS colitis was induced with a modification of the technique of Morris et al. (22). Sensitization was induced by application of 1% TNBS (picrylsulfonic acid solution; Sigma-Aldrich) in 100% ethanol on day −7 to the shaved and abraded abdomen of each mouse. Colitis was induced on day 0 by intrarectal administration of 5 μl/g body weight of 2.5% TNBS solution in 50% ethanol after sedation of the animals using 14 μl/g body weight of 2.5% tribromoethanol solution in 1× PBS. TNBS was introduced using an epidural catheter (FlexTip Plus Epidural Catheter; Arrow International) which was inserted ∼6 cm per rectum. After TNBS administration, animals were suspended in a vertical position for at least 5 min to ensure colonic retention of the chemical. Control animals received a corresponding volume of 50% ethanol alone.

Only animals that showed an initial response to TNBS treatment, defined by at least a 5% weight loss relative to baseline weight, were included in the weight loss analysis for the TNBS treatment groups. (This resulted in the exclusion from analysis of five mice from the C57BL/6 TNBS group and two mice from the cd73−/− TNBS group.) Additionally, one cd73−/− animal in the TNBS treatment group died between days 2 and 3 before harvest and was thus excluded from our analyses.

In mice treated with the CD73 inhibitor, APCP (Sigma-Aldrich), APCP was prepared at a concentration of 500 μg/100 μl of PBS, and animals were treated by gavage with APCP 20 mg/kg/day for 4 days before harvest. In TNBS experiments, APCP treatment was initiated on the day −1, 1 day before the induction of colitis. Mice were then treated daily until harvest on day 3 (5 total doses). Control animals received corresponding volumes of PBS.

Total RNA derived from colonic scrapings (enriched in epithelial cells) was reverse-transcribed into cDNA and arrayed through the Massachusetts General Hospital DNA Core Facility (Boston, MA) using the PGA mouse v1.1 chip. For each experiment, three animals were arrayed in triplicate.

In the case of IFN-αA-treated mice, mouse IFN-αA (Serotec) was diluted in PBS to a concentration of 1 × 104 U/ml and sterile-filtered. A baseline weight was established for mice on day −1, the day before the induction of colitis and IFN-αA-treated mice received daily i.p. injections of IFN-αA until the morning of harvest with a goal dose of 40 U of IFN-αA per g of baseline body weight. Due to the range in baseline weights, animals were injected as follows. Animals weighing ≥22.5 g received 1000 U/day, animals with a baseline weight ranging from 17.5–22.4 g received 800 U/day, and animals weighing <17.5 g received 600 U/day. Control animals received daily injections of corresponding weight-adjusted volumes of sterile PBS i.p. (Two animals were excluded from the analysis of the IFN-αA-treated cd73−/− TNBS group. One animal was excluded due to an error in IFN-αA administration resulting in >110% of the intended dose and another due to frank peritonitis, attributed to injection-related bowel perforation, with abscess formation at the injection site noted at the time of harvest.)

As a further parameter of ongoing inflammation, colon length was determined by measurement of the distance from the most distal aspect of the cecum to the most terminal aspect of the rectum. Additionally, histological analysis was performed on whole sections of colon which were fixed in OCT compound (Sakura Finetek) at the time of harvest and frozen using liquid nitrogen; then sections were cut and processed for staining with H&E.

Histological examination was performed on three samples of the distal colon. Samples were fixed in 10% formalin before staining with H&E. All histological quantitation was performed in a blinded manner, using a previously described scoring system (23). The three independent parameters measured were severity of inflammation (0–3: none, slight, moderate, severe), extent of injury (0–3: none, mucosal, mucosal and submucosal, transmural), and crypt damage (0–4: none, basal 1 of 3 damaged, basal 2 of 3 damaged, only surface epithelium intake, entire crypt and epithelium lost). The score of each parameter was multiplied by a factor reflecting the percentage of tissue involvement (×1, 0–25%; ×2, 26–50%; ×3, 51–75%; ×4, 76–100%), and all numbers were summed. Maximum possible score was 40.

mRNA was isolated for transcriptional analysis from colonic mucosal scrapings homogenized in TRIzol (Invitrogen Life Technologies.) using a 550 sonic dismembrator (Fisher Scientific International) followed by phenol-chloroform extraction, as previously described (2). Reverse transcription was done using Iscript cDNA synthesis kit using the manufacturer’s instructions (Bio-Rad Laboratories).

The transcriptional profile of the cd73−/− animals relative to C57BL/6 mice was assessed from total RNA using quantitative genechip expression arrays (Affymetrix) as described previously (24). Results obtained by gene array analysis were subsequently verified by real-time PCR using iQ SYBR mix (iCycler; Bio-Rad), as described previously (2, 6). Primer sets were as follows: murine β-actin (sense primer, 5′-ctctccctcacgccatcctg-3′; antisense primer, 5′-tcacgcacgatttccctctcag-3′, 124 bp); murine IFN-αA (sense primer, 5′-ctgacccaggaagactacct-3′; antisense primer, 5′-ggctgaggaagacatggctct-3′, 138 bp); and murine CD73 (sense primer, 5′-ctggggcactctggttttga-3′; antisense primer, 5′-tccccgcaggcacttctttg-3′, 114 bp).

Murine inflammatory cytokine profiling was also performed with real-time PCR using iQ SYBR mix. For these experiments, the primer sets were as follows: murine β-actin (sense primer, 5′-ctaggcaccagggtgtgat-3′; antisense primer, 5′-tgccagatcttctccatgtc-3′, 148 bp); murine IFN- γ (sense primer, 5′-tcaagtggcatagatgtggaagaa-3′; antisense primer, tggctctgcaggattttcatg-3′, 92 bp); murine TNF α (sense primer, 5′-ccaccacgctcttctgtctac-3′; antisense primer, 5′-tgggctacaggcttgtcact-3′, 151 bp); and murine IL-10 (sense primer, 5′-atgctgcctgctcttactgactg-3′; antisense primer, 5′-cccaagtaacccttaaagtcctgc-3′, 216 bp).

Statistical analysis was performed using Student’s t test or ANOVA, as indicated.

Extracellular Ado signaling has been widely implicated in adaptive responses to inflammation (25). The liberation of extracellular Ado involves phosphohydrolysis of adenine nucleotide intermediates and is critically regulated by the terminal enzymatic step catalyzed by CD73 (26). Previous studies revealed that CD73 is most abundantly expressed in the colon (9) and that hypoxia-inducible factor 1-regulated CD73 may be protective during in mucosal inflammation (2, 7). We therefore sought to define the contribution of CD73 to colitic disease. Initially, we examined CD73 expression in cd73+/+ mice during the induction of colitis. As shown in Fig. 1,A, analysis of the mRNA from colonic mucosal scrapings revealed a 3.8- ± 1.1-fold induction in CD73 expression in day 3 TNBS-treated cd73+/+ C57BL/6 animals relative to ethanol-treated controls (∗, p < 0.025). Such findings are consistent with previous studies using TNBS colitis (2). Furthermore, as a confirmation of the null genotype, no CD73 mRNA was detected in mucosal scrapings from either ethanol- or TNBS-treated cd73−/− mice (Fig. 1 A).

FIGURE 1.

CD73 levels are induced in colitis, and deficiency of CD73 results in enhanced acute inflammation in TNBS colitis. A, Colonic mucosal scrapings from C57BL/6 cd73+/+ mice and cd73−/− mice sacrificed in the acute phase of TNBS colitis were assayed for CD73 mRNA levels by real-time PCR and compared with mRNA levels seen in ethanol (EtOH)-treated controls. Results are depicted as fold change relative to control ± SEM. ★, p < 0.025, Student’s t test, n = 4. B, The influence of the elimination of CD73 expression on changes in body weight following induction of TNBS colitis was then determined. Cd73−/− animals treated with TNBS (▵) experienced a faster weight loss following induction of colitis and failed to gain weight back compared with wild-type TNBS-treated controls (□) (p < 0.025, ANOVA). The percent baseline weight for ethanol-treated cd73+/+ (▪) and cd73−/− (▴) mice are shown for comparison. C, Colon length measurements were determined for the same animals (analyzed in A) after 3 days of colitis. Although both cd73+/+ and cd73−/− mice treated with TNBS had significant colon shortening relative to ethanol-treated controls. ★, p < 0.025, Student’s t test. TNBS-treated cd73−/− mice displayed a significant further decrease in colon length relative to cd73+/+ TNBS-treated mice. ★, p < 0.025, Student’s t test.

FIGURE 1.

CD73 levels are induced in colitis, and deficiency of CD73 results in enhanced acute inflammation in TNBS colitis. A, Colonic mucosal scrapings from C57BL/6 cd73+/+ mice and cd73−/− mice sacrificed in the acute phase of TNBS colitis were assayed for CD73 mRNA levels by real-time PCR and compared with mRNA levels seen in ethanol (EtOH)-treated controls. Results are depicted as fold change relative to control ± SEM. ★, p < 0.025, Student’s t test, n = 4. B, The influence of the elimination of CD73 expression on changes in body weight following induction of TNBS colitis was then determined. Cd73−/− animals treated with TNBS (▵) experienced a faster weight loss following induction of colitis and failed to gain weight back compared with wild-type TNBS-treated controls (□) (p < 0.025, ANOVA). The percent baseline weight for ethanol-treated cd73+/+ (▪) and cd73−/− (▴) mice are shown for comparison. C, Colon length measurements were determined for the same animals (analyzed in A) after 3 days of colitis. Although both cd73+/+ and cd73−/− mice treated with TNBS had significant colon shortening relative to ethanol-treated controls. ★, p < 0.025, Student’s t test. TNBS-treated cd73−/− mice displayed a significant further decrease in colon length relative to cd73+/+ TNBS-treated mice. ★, p < 0.025, Student’s t test.

Close modal

Our previous studies revealed that cd73−/− animals manifest no outward immunological defects when housed in specific pathogen-free conditions (9). Here, we compared the clinical course of TNBS-induced colitis in cd73+/+ and cd73−/− mice. Weight loss is a reliable method to assess TNBS colitis severity (1), and as shown in Fig. 1,B, cd73−/− mice exposed to TNBS lost weight more rapidly and failed to regain weight over the 5 days following induction of colitis (Fig. 1,B; p < 0.025 by ANOVA). Because of an increased incidence of death in the cd73−/− mice during the acute phase of colitis in our pilot experiments, mice were harvested after 3 days of colitis for all other experiments to optimize the chance for survival of the TNBS-treated cd73−/− mice. In these studies, cd73−/− animals sacrificed on day 3 after the induction of colitis also demonstrated a more severe decrease in colon length relative to genotype-specific, ethanol-treated controls, consistent with an increased inflammatory response (Fig. 1 C, p < 0.025). These studies reveal that cd73−/− animals are more susceptible to TNBS-induced colitis.

To rule out the possibility that increased susceptibility of cd73−/− mice to colitis could be attributed to a developmental etiology, we determined whether this phenotype could be mimicked by pharmacological blockade of CD73 using the selective inhibitor APCP. cd73+/+ C57BL/6 mice were treated with APCP (20 mg/kg) by gavage, a concentration we have previously shown to effectively inhibit CD73 in a systemic manner (7). As measured by HPLC, serum levels of APCP were 0.26 ± 0.11 mM in treated mice (data not shown). In a separate experiment, mice were subjected to TNBS colitis with or without APCP (20 mg/kg) administered by gavage beginning on day −1 and then daily (for a total of five doses) until harvest on day 3 following the induction of colitis. Similar to cd73−/− mice, wild-type APCP-treated mice exposed to TNBS lost weight more rapidly and failed to show recovery weight gain during the course of the experiment (Fig. 2; p < 0.05 by ANOVA), compared with mice receiving TNBS alone who, in this set of experiments, showed recovery weight gain back toward the level of ethanol-treated controls by day 3 (Fig. 2). Such findings confirm that altered CD73 activity likely explains our findings of increased susceptibility in cd73−/− animals.

FIGURE 2.

Pharmacological blockade of CD73 with APCP exacerbates TNBS colitis. To determine the influence of APCP treatment on acute inflammation in TNBS colitis, mice were treated daily with 20 mg/kg APCP per gavage, beginning on the day before induction of colitis and extending to the morning of harvest on day 3 of colitis. Results for daily weights on each day following the induction of colitis are depicted as percent baseline weight ± SEM. APCP-treated TNBS mice (▵) lost more weight and failed to regain weight as quickly as PBS-treated TNBS controls (□). p < 0.05 by ANOVA, n = 3. APCP-treated mice receiving ethanol alone (▴) did not show statistically significant differences in weight loss relative to PBS-treated mice receiving ethanol alone (▪).

FIGURE 2.

Pharmacological blockade of CD73 with APCP exacerbates TNBS colitis. To determine the influence of APCP treatment on acute inflammation in TNBS colitis, mice were treated daily with 20 mg/kg APCP per gavage, beginning on the day before induction of colitis and extending to the morning of harvest on day 3 of colitis. Results for daily weights on each day following the induction of colitis are depicted as percent baseline weight ± SEM. APCP-treated TNBS mice (▵) lost more weight and failed to regain weight as quickly as PBS-treated TNBS controls (□). p < 0.05 by ANOVA, n = 3. APCP-treated mice receiving ethanol alone (▴) did not show statistically significant differences in weight loss relative to PBS-treated mice receiving ethanol alone (▪).

Close modal

In an attempt to gain specific insight into potential mechanisms of increased colitis susceptibility of cd73−/− mice, we profiled the expression of colonic epithelial immunoregulators in noncolitic cd73−/− and cd73+/+ mice, under baseline conditions, using cDNA microarrays (n = 3 animals per group). This analysis was insightful inasmuch as it revealed that cd73−/− colonic epithelia (derived from mucosal scrapings enriched in epithelia) express nearly undetectable levels of IFN-α, specifically IFN-αA (also known as IFN-α3) (27). As shown in Table I, a comparative profile of a number of colitis-associated immunomodulators and their receptors, revealed significant changes only in IFN-αΑ, under baseline conditions. Indeed, whereas baseline mRNA expression of TNF-α, TNFR1, TNFR2, IFN-γR, IFN-αR, IL-6, IL-10, and TGFβ were not changed significantly in cd73−/− mice relative to controls, IFN-αA was repressed by as much as 91 ± 7% compared with cd73+/+ animals (p < 0.001), thereby suggesting a relatively selective reduction in IFN-αA expression in cd73−/− mice. This reduction in IFN-αA in cd73−/− mice was not confined to the intestinal tract, in that a reduction in IFN-αA mRNA expression in whole lung tissue of cd73−/− mice was also evident (92 ± 4% decrease compared with cd73+/+ animals; p < 0.001). Given these robust findings, we pursued the potential role of this alteration in baseline IFN-α expression in the setting of colonic inflammation.

Table I.

Colonic epithelial inflammatory gene expression in cd73+/+vs cd73−/− micea

Genecd73−/−:cd73+/+ Ratio
IFN-α-A 0.09 ± 0.01b 
IFN-γ-R2 0.76 ± 0.17 
IFN-γ-R1 0.81 ± 0.13 
IFN-α-R1 1.18 ± 0.28 
IFN-α-R2 1.20 ± 0.44 
TNF-α 0.87 ± 0.08 
TNF-α-R1 1.35 ± 0.37 
TNF-α-R2 0.74 ± 0.16 
IL-6 1.14 ± 0.39 
IL-10 0.93 ± 0.17 
TGF-β1 0.89 ± 0.24 
Genecd73−/−:cd73+/+ Ratio
IFN-α-A 0.09 ± 0.01b 
IFN-γ-R2 0.76 ± 0.17 
IFN-γ-R1 0.81 ± 0.13 
IFN-α-R1 1.18 ± 0.28 
IFN-α-R2 1.20 ± 0.44 
TNF-α 0.87 ± 0.08 
TNF-α-R1 1.35 ± 0.37 
TNF-α-R2 0.74 ± 0.16 
IL-6 1.14 ± 0.39 
IL-10 0.93 ± 0.17 
TGF-β1 0.89 ± 0.24 
a

Data represent ratio ± SD of microarray fluorescence derived from mRNA of colonic scrapings of cd73−/− (n = 3) and cd73+/+ mice (n = 3).

b

p < 0.001.

Before examining the influence of this alteration in baseline IFN-αA expression on the response to inflammation, we verified the microarray findings of IFN-αA deficiency. As shown in Fig. 3, a real-time PCR screen of mRNA isolated from whole mouse colon (Fig. 3,B) comparing cd73−/− and cd73+/+ animals revealed a prominent deficiency of IFN-αA (with IFN-αA mRNA decreased by 77.6 ± 11.3%; p = 0.01). Furthermore, a similar decrease in IFN-αA expression (by 79.6 ± 7.3; p < 0.05) was observed in mRNA isolated from colonic mucosal scrapings (Fig. 3 C). Such findings confirm our microarray results and implicate a role for CD73 in regulation of IFN-αΑ. Furthermore, because mucosal scrapings are collected in such a way as to enrich for intestinal epithelial cells, these findings provide some support for the existence of a specific deficiency in epithelial IFN-αA mRNA production in the cd73−/− mice.

FIGURE 3.

Genetic or pharmacological inhibition of CD73 expression results in a specific decrease in IFN-αA mRNA expression resulting in altered IFN-αA regulation in response to TNBS colitis. A, Whole lung and colonic mucosal scrapings from cd73−/− animals and cd73+/+ mice were analyzed by gene array for relative mRNA expression of cytokines.p < 0.001, Student’s t test. This profound down-regulation of IFN-αA mRNA expression was confirmed by real-time PCR in both colonic mucosal scrapings (B, ★, p = 0.01, Student’s t test, n = 4–8 mice per group) and whole colon (C, ★, p < 0.05, Student’s t test, n = 7–8 mice per group). D, Mice were treated with 20 mg/kg amounts of the CD73 inhibitor APCP per gavage for 4 days; results are depicted as fold change in IFN-αA mRNA expression relative to that of PBS-gavaged controls. ★, p < 0.05, Student’s t test. E, Mucosal scrapings harvested from cd73−/− mice and cd73+/+ C57BL/6 mice, 3 days after induction of TNBS colitis, were analyzed for IFN-αA mRNA expression by real-time PCR, and the results are depicted as fold change relative to cd73+/+ controls. Exposure to TNBS resulted in a significant reduction in IFN-αA mRNA levels in cd73−/− mice relative to cd73+/+ controls. ★, p < 0.01, Student’s t test. In contrast, cd73−/− ethanol-treated control mice had a significantly lower baseline IFN-αA mRNA expression relative to cd73+/+ ethanol-treated controls. ★, p < 0.01 by Student’s t test). Additionally, cd73−/− TNBS-treated mice failed to show a significant change in IFN-αA mRNA expression relative to cd73−/− ethanol-treated controls (n = 6–8 mice per treatment group).

FIGURE 3.

Genetic or pharmacological inhibition of CD73 expression results in a specific decrease in IFN-αA mRNA expression resulting in altered IFN-αA regulation in response to TNBS colitis. A, Whole lung and colonic mucosal scrapings from cd73−/− animals and cd73+/+ mice were analyzed by gene array for relative mRNA expression of cytokines.p < 0.001, Student’s t test. This profound down-regulation of IFN-αA mRNA expression was confirmed by real-time PCR in both colonic mucosal scrapings (B, ★, p = 0.01, Student’s t test, n = 4–8 mice per group) and whole colon (C, ★, p < 0.05, Student’s t test, n = 7–8 mice per group). D, Mice were treated with 20 mg/kg amounts of the CD73 inhibitor APCP per gavage for 4 days; results are depicted as fold change in IFN-αA mRNA expression relative to that of PBS-gavaged controls. ★, p < 0.05, Student’s t test. E, Mucosal scrapings harvested from cd73−/− mice and cd73+/+ C57BL/6 mice, 3 days after induction of TNBS colitis, were analyzed for IFN-αA mRNA expression by real-time PCR, and the results are depicted as fold change relative to cd73+/+ controls. Exposure to TNBS resulted in a significant reduction in IFN-αA mRNA levels in cd73−/− mice relative to cd73+/+ controls. ★, p < 0.01, Student’s t test. In contrast, cd73−/− ethanol-treated control mice had a significantly lower baseline IFN-αA mRNA expression relative to cd73+/+ ethanol-treated controls. ★, p < 0.01 by Student’s t test). Additionally, cd73−/− TNBS-treated mice failed to show a significant change in IFN-αA mRNA expression relative to cd73−/− ethanol-treated controls (n = 6–8 mice per treatment group).

Close modal

The above studies in cd73−/− mice implicate CD73 as a limiting metabolic step for maintenance of colonic IFN-αA, in vivo. To rule out that such changes might be associated with a developmental defect in cd73−/− mice unrelated to the loss of CD73, we assessed colonic IFN-αA levels in cd73+/+ mice in response to APCP (20 mg/kg/day orally for 4 days). As shown in Fig. 3 D, epithelial IFN-αΑ levels in mice treated with APCP resulted in a 56.7 ± 8.2% decrease in (p < 0.05 compared with vehicle controls). This decrease in IFN-αA approximates that in cd73−/−, thereby strongly implicating CD73 in basal maintenance of IFN-α in colonic epithelial cells.

Although some evidence exists to suggest that monocytes from Crohn’s disease patients may express lower levels of virally stimulated IFN-αA (15), relative to controls, the influence of inflammatory bowel disease on IFN-αA expression remains unclear. We therefore analyzed the influence of TNBS colitis on IFN-αA mRNA expression (Fig. 3 E). Analysis of colonic mucosal scrapings from cd73+/+ mice exposed to TNBS revealed a significant reduction in IFN-αA expression (77.2 ± 14%; p < 0.01 compared with vehicle control). cd73−/− mice exposed to TNBS revealed no significant change in IFN-αA mRNA expression.

Katakura et al. (21) recently showed that TLR-9 agonists induce type I IFN. In the course of these studies, they showed that administration of the type I IFN IFN-β at a dose of 1000 U/mouse/day to SCID mice provides benefit in the dextran sulfate sodium murine model of experimental colitis (21). Because both cd73−/− mice and APCP-treated mice display both a more severe inflammatory phenotype in the acute phase of TNBS colitis and a significant decrease in IFN-αA mRNA expression, we sought to determine whether the severity of the colitic phenotype could be attenuated by rescue treatment with recombinant IFN-αA. For these purposes, cd73−/− and cd73+/+ mice were treated daily with i.p. injection of recombinant murine IFN-αA (40 U/g body weight), as described in Materials and Methods. This dose of IFN-αA was based both on our own pilot experiments (data not shown) and on previous work (21).

In these experiments, cd73+/+ mice demonstrated the expected weight loss pattern (Fig. 4 A) typically seen in TNBS colitis with an initial weight loss beginning on day I and then stabilizing over the 3 days of the experiment. cd73+/+ mice treated with both IFN-αA and TNBS showed no statistically significant difference in their weight curve relative to cd73+/+mice treated with TNBS alone. Although these findings are in conflict with those of Katakura et al., this difference may reflect inherent differences between the SCID mice used by Katakura et al. and the wild-type C57BL/6 mice used here. Alternatively, Katakura et al. used the dextran sulfate sodium colitis model rather than the TNBS model and stimulated their mice with IFN-β rather than IFN-αA. Although these two type I IFNs share the same receptor, there are a number of important biological differences between the various type I IFNs (28, 29), which could contribute to the differences in our results relative to those of Katakura et al.

FIGURE 4.

IFN-αA treatment rescues cd73−/− mice in TNBS colitis. cd73−/− mice treated with either 40 U/kg IFN-αA (IFN-αA/TNBS, ▵) or an equal volume of sterile PBS (TNBS, ○) daily, by i.p. injection beginning on the day before the induction of colitis, were then followed for weight change relative to similarly treated cd73+/+ TNBS mice (TNBS (•) and IFN-αA/TNBS (▴) and genotype-specific, ethanol-treated controls (cd73−/− (□), and cd73+/+ (▪), n = 13–20 mice per treatment group). Daily weights are depicted as mean percent of baseline weight ± SEM for cd73+/+ mice (A) and cd73−/− mice (B). Weights on day 3 following induction of colitis are depicted as mean percent of baseline weight ± SEM for each condition, as labeled (C). ★, p < 0.0001, ★★, p < 0.05). Results for mean colon length are depicted as mean colon length per kg baseline body weight ± SEM (D, ★, p < 0.0001; ★★, p < 0.05).

FIGURE 4.

IFN-αA treatment rescues cd73−/− mice in TNBS colitis. cd73−/− mice treated with either 40 U/kg IFN-αA (IFN-αA/TNBS, ▵) or an equal volume of sterile PBS (TNBS, ○) daily, by i.p. injection beginning on the day before the induction of colitis, were then followed for weight change relative to similarly treated cd73+/+ TNBS mice (TNBS (•) and IFN-αA/TNBS (▴) and genotype-specific, ethanol-treated controls (cd73−/− (□), and cd73+/+ (▪), n = 13–20 mice per treatment group). Daily weights are depicted as mean percent of baseline weight ± SEM for cd73+/+ mice (A) and cd73−/− mice (B). Weights on day 3 following induction of colitis are depicted as mean percent of baseline weight ± SEM for each condition, as labeled (C). ★, p < 0.0001, ★★, p < 0.05). Results for mean colon length are depicted as mean colon length per kg baseline body weight ± SEM (D, ★, p < 0.0001; ★★, p < 0.05).

Close modal

In contrast, cd73−/− mice treated with IFN-αA and TNBS showed decreased total weight loss relative to those with TNBS alone (Fig. 4,B; p < 0.005 by ANOVA). These differences in weight loss, expressed as percent baseline weight, were most marked on day 3 after induction of colitis (Fig. 4 C) with statistically significant weight loss seen in all TNBS treatment groups for both cd73+/+ and cd73−/− mice relative to genotype-specific, vehicle = treated controls. However, IFN-αA treatment resulted in a statistically significant decrease in relative weight lost with a mean percent baseline weight of 91.4 + 1.8% in the IFN-αA-treated cd73−/− animals compared with only 86.9 + 2.0% in the cd73−/− mice treated with TNBS alone (p < 0.05).

Colon length was again determined for each animal at the time of sacrifice as an additional indicator of colonic inflammation and edema. Because the male cd73−/− mice were markedly larger than age-matched wild-type controls, there were significant variations in mouse size within each experiment. To control for this, colon length measurements were normalized according to baseline body weight. Both TNBS-treated cd73+/+ and TNBS-treated cd73−/− mice had significantly shorter colons relative to ethanol-treated controls than did cd73+/+ and cd73−/− mice treated with both IFN-αA and TNBS. However, there was a significant reduction in colonic shortening in the cd73−/− mice treated with IFN-αA and TNBS relative to those receiving TNBS alone.

Histological staining of colonic sections demonstrated inflammatory cell infiltration into the intestinal epithelium in both wild-type and cd73−/− animals treated with TNBS (Fig. 5,A). Additionally, the tissue from IFN-αA/TNBS treated cd73−/− mice showed decreased inflammation and better preservation of crypts and villous architecture relative to those treated with TNBS alone. Histological scoring was performed on a subset of the same animals assayed for weight loss and colon length with three animals examined per treatment group (Fig. 5 B), demonstrating a trend toward an increase in total score for inflammation in the TNBS-treated cd73−/− animals relative to TNBS-treated wild-type animals, but this failed to achieve statistical significance over the number of sections scored. However, IFN-αA treatment, before and during TNBS colitis, resulted in significant reductions in scores for inflammation, extent of injury, crypt damage, and total composite score for both the wild-type and cd73−/− animals, relative to genotype-specific controls treated with TNBS alone.

FIGURE 5.

IFN-αA treatment attenuates inflammatory changes in the colonic mucosa of cd73−/− mice while stimulating proinflammatory changes in colonic mucosa of cd73+/+ C57BL/6 mice. A, Representative images from whole colon from each treatment group, as labeled, were processed for H&E staining and then analyzed by transmission bright-field microscopy. EtOH, Ethanol. B, Histological scoring was performed in a blinded manner as described. Scores for inflammation, injury extent, crypt damage, and total composite inflammation are depicted for TNBS-treated wild-type animals (with (white columns) and without (black columns) IFN-aA) and TNBS-treated cd73−/− animals (with (dark gray columns) and without (light gray columns) rIFN-αA). Results indicate statistically significant reductions in histological scores for all categories of inflammation in IFN-αA-treated animals relative to genotype-specific controls receiving TNBS alone. ★, p < 0.01.

FIGURE 5.

IFN-αA treatment attenuates inflammatory changes in the colonic mucosa of cd73−/− mice while stimulating proinflammatory changes in colonic mucosa of cd73+/+ C57BL/6 mice. A, Representative images from whole colon from each treatment group, as labeled, were processed for H&E staining and then analyzed by transmission bright-field microscopy. EtOH, Ethanol. B, Histological scoring was performed in a blinded manner as described. Scores for inflammation, injury extent, crypt damage, and total composite inflammation are depicted for TNBS-treated wild-type animals (with (white columns) and without (black columns) IFN-aA) and TNBS-treated cd73−/− animals (with (dark gray columns) and without (light gray columns) rIFN-αA). Results indicate statistically significant reductions in histological scores for all categories of inflammation in IFN-αA-treated animals relative to genotype-specific controls receiving TNBS alone. ★, p < 0.01.

Close modal

Although IFN-αA treatment, at least in part, rescued the cd73−/− colitis phenotype, administration of the same dose of IFN-αA per kg of body weight failed to diminish weight loss or colonic shortening in cd73+/+ mice exposed to TNBS colitis. Although these parameters of clinical disease were suggestive of a more severe inflammatory phenotype in the cd73+/+ mice treated with both IFN-αA and TNBS relative to those treated with TNBS alone, histological scoring of the wild-type TNBS mice treated with IFN-αA indicated decreased inflammation, injury, and crypt damage. It remains unclear why the observed discrepancy between histological scoring and clinical symptoms remain in cd73+/+ mice after IFN-αA administration. Although histological scoring of the tissue is but one parameter examined, it is possible that administration of exogenous IFN-αA to animals capable of baseline production could result in symptoms independent of colitis.

In an effort to further characterize the differential responses of the cd73−/− vs the wild-type mice to TNBS, we used real-time PCR to profile the intestinal mucosal production of key murine inflammatory cytokines following the induction of colitis. As expected, mRNA levels for IFN-γ (Fig. 6,A), TNF-α (Fig. 6,B), and IL-10 (Fig. 6,C) were significantly induced in the wild-type animals after TNBS exposure. Interestingly, although the cd73−/− animals showed inductions in IFN-γ and TNF-α mRNA that were similar to those of wild-type animals, the cd73−/− animals demonstrated a marked inability to up-regulate IL-10 mRNA after TNBS exposure. Furthermore, the ability to mount an IL-10 response to TNBS was restored in the cd73−/− animals treated with exogenous IFN-αA (Fig. 6 C). Thus, CD73-mediated IFN-αA may modulate colonic inflammation via regulation of IL-10 production.

FIGURE 6.

cd73−/− mice have an intact IFN-γ and TNF-α responses to TNBS but fail to induce IL-10 in the absence of IFN-αA. Colonic mucosal scrapings from wild-type (WT) or cd73−/− mice exposed to TNBS or TNBS and IFN-αA were analyzed by real-time PCR for relative expression of IFN-γ (A), TNF-α (B), and IL-10 (C). ★, p < 0.01; ★★, p < 0.001.

FIGURE 6.

cd73−/− mice have an intact IFN-γ and TNF-α responses to TNBS but fail to induce IL-10 in the absence of IFN-αA. Colonic mucosal scrapings from wild-type (WT) or cd73−/− mice exposed to TNBS or TNBS and IFN-αA were analyzed by real-time PCR for relative expression of IFN-γ (A), TNF-α (B), and IL-10 (C). ★, p < 0.01; ★★, p < 0.001.

Close modal

Although this finding was intriguing, the implication that the induction of murine IL-10 mRNA in response to inflammation requires IFN-αA appeared to be in conflict with the findings that wild-type mice demonstrate a down-regulation of IFN-αA message by day 3 of TNBS colitis (Fig. 3,E). To reconcile these two findings, we next examined the kinetics of both IFN-αA and IL-10 mRNA expression during the early course of TNBS colitis (Fig. 7,A). Intestinal mucosal scrapings from TNBS-treated wild-type C57BL/6 mice sacrificed on the indicated days were processed for RNA and assayed by real-time PCR for IFN-αA and IL-10 (Fig. 7 B) expression. Mucosal scrapings from TNBS-treated mice demonstrated a significant induction of IFN-αA relative to ethanol-treated controls (p < 0.05 by ANOVA) with levels increased by >3-fold on day 2 after induction of TNBS colitis (p < 0.025 by Student’s t test). Subsequent down-regulation of IFN-αA expression to baseline was observed by day 3 with further down-regulation relative to ethanol-treated controls evident by day 5. In contrast, TNBS-treated wild-type mice had a significant induction (p < 0.025 by ANOVA) of IL-10 mRNA overall with expression increased by >3-fold on day 3 (p < 0.025 by Student’s t test) and increasing to >5-fold relative to TNBS controls harvested on day 0 (p < 0.025 by Student’s t test). Although these results do not prove causality, they are temporally consistent with our findings that the IL-10 mRNA response to TNBS is restored by the administration of exogenous IFN-αA to our IFN-αA-deficient cd73−/− mice.

FIGURE 7.

Wild-type mice transiently induce IFN-αA on day 2 during TNBS colitis followed by a sustained induction of IL-10 beginning on day 3 after exposure to TNBS. Colonic mucosal scrapings from wild-type mice exposed to TNBS (closed bars) and sacrificed at varying time points after TNBS exposure, as indicated, were analyzed by real-time PCR for relative expression of IFN-αA (A) or IL-10 (B) relative to controls (open bars). Results are depicted as the mean ± SEM for three mice per condition, ★, p < 0.025 by Student’s t test.

FIGURE 7.

Wild-type mice transiently induce IFN-αA on day 2 during TNBS colitis followed by a sustained induction of IL-10 beginning on day 3 after exposure to TNBS. Colonic mucosal scrapings from wild-type mice exposed to TNBS (closed bars) and sacrificed at varying time points after TNBS exposure, as indicated, were analyzed by real-time PCR for relative expression of IFN-αA (A) or IL-10 (B) relative to controls (open bars). Results are depicted as the mean ± SEM for three mice per condition, ★, p < 0.025 by Student’s t test.

Close modal

In these studies, we have shown that the cd73−/− mice who are deficient in ATP/ADP-hydrolyzing activity and the generation of extracellular Ado are also deficient in IFN-αA production at baseline and lack the IL-10 response during hapten-based colitis. This IL-10 mRNA response on day 3 is restored in the cd73−/− mice in the presence of exogenous IFN-αA. The implications of these findings are summarized in the schematic depicted in Fig. 8. We propose that in wild-type C57/BL6 mice the exposure to TNBS in wild-type C57BL/6 mice triggers a CD73-dependent, transient induction of IFN-αA mRNA on day 2 followed by an induction of IL-10 mRNA first evident on day 3 and increasing further by day 5.

FIGURE 8.

TNBS-stimulated CD73-mediated induction of IFN-αA mRNA is temporally associated with a subsequent sustained induction of IL-10.

FIGURE 8.

TNBS-stimulated CD73-mediated induction of IFN-αA mRNA is temporally associated with a subsequent sustained induction of IL-10.

Close modal

Extracellular Ado signaling has been widely implicated as an endogenous protective mechanism for a number of inflammatory diseases (25). The rate-limiting surface enzyme for Ado liberation is CD73, and its regulation has been implicated in numerous inflammatory (30, 31) and hypoxic settings (2, 6, 9). The studies presented here indicate that CD73 activity inversely correlates with colitis severity. Furthermore, elimination of functional CD73 activity, either by genetic mutation in the cd73−/− mice or by pharmacological inhibition with APCP, results in an exacerbation of the acute inflammatory phase of TNBS colitis. Additional markers of inflammation, including colonic shortening and histological examination of colonic mucosal PMN infiltration, also support the finding of a more severe acute inflammatory response in the cd73−/− mice and therein implicate a protective role for Ado signaling in mucosal inflammation.

Our rationale for focusing on CD73 in colitis was severalfold. First, in a series of studies addressing the role of hypoxia-inducible factor in colitis, it was revealed that CD73 is a highly inducible gene during ongoing colitis. Indeed, TNBS colitis alone induced colonic epithelial CD73 mRNA by >5-fold (2). Moreover, as part of our study characterizing the phenotype of cd73−/− animals (9), analysis of CD73 enzyme activity (i.e., AMP-hydrolyzing activity inhibitable by APCP) in a broad range of tissues produced several important findings. First, there was a nearly 50-fold variation in CD73 enzyme activity in different tissues from cd73+/+ animals. Second, of the tissues surveyed, colon showed the highest level of enzyme activity, a somewhat surprising result given that a number of previous studies have suggested that the kidney likely carried the highest activity of any tissue (32). Instead, it was found that the colon expresses nearly twice as much activity as the kidney, with the rank order of tissue activity as follows: colon > kidney = brain > liver > lung > heart ≫ muscle (9). Furthermore, AMP-hydrolyzing activity that was not attributed to CD73 (i.e., nucleotidase and/or phosphatase activity that is not inhibited by APCP) also varied widely between individual tissues (up to 20-fold between different tissues). Why such seemingly disparate enzyme expression patterns exist is not clear at present. However, the data suggest tissue-specific patterns of extracellular nucleotide metabolism and differences in the relative contribution of CD73 to extracellular Ado generation among individual tissues.

Recent studies have detailed the role of Ado receptor signaling in multiple experimental models of colitis resulting from a variety of methods of induction of inflammation. Using T cell transfer approaches, Naganuma et al. (33) demonstrated that Ado A2A receptors (A2AR) appear to block the production of proinflammatory cytokines and that A2AR agonists attenuate T cell-mediated colitis. Additional work has revealed a protective role for A2AR in various other colitic models (34, 35). One recent study identified overexpression of Ado A2B receptor in experimental colitis (36), and two additional reports, one by Guzman et al. (37) using a rat colitis model and another by Mabley et al. (38) studying two independent mouse models of colitis, have indicated a protective role for the Ado A3 receptor. Taken together, these findings suggest a more universal role for Ado receptor signaling in the resolution of colitis, downstream of the method of induction of inflammation. Furthermore, our observation that genetic or pharmacological neutralization of the proximal step in Ado signaling (i.e., CD73) results in enhanced colitic disease is in general agreement with these findings.

A broad profiling of cd73−/− animals revealed that the loss of CD73 is closely linked to a profound reduction in the levels of IFN-αA, suggesting a relationship between CD73 activity and downstream regulation of IFN-αA expression. These findings are further supported by the demonstration that the pharmacological inhibition of CD73 activity also leads to down-regulation of IFN-αA mRNA expression. Furthermore, we show that IFN-αA replacement reduces severity of colitis associated with deletion of the CD73 gene and that levels of disease are restored to those seen in the wild-type TNBS-treated mice. At multiple levels, Katakura et al. (21) showed that the induction of type I IFN via TLR9 provides an innate protection during experimental colitic disease, and in the course of these studies, they also showed that administration of type I IFN provides benefit in experimental colitis. Likewise, type I IFN has been used with mixed success in human patients with inflammatory bowel disease, particularly ulcerative colitis (39).

The mechanisms underlying the clinical benefit of type I IFNs in colitis remain unclear. In the studies presented here, characterization of the kinetics of cytokine mRNA response in TNBS-treated wild-type mice demonstrates an early transient peak in IFN-αA mRNA seen on day 2 after TNBS exposure which is then down-regulated. This peak in IFN-αA is followed by an induction of IL-10 mRNA beginning on day 3 and continuing to increase in TNBS-treated wild-type mice relative to ethanol-treated controls on day 5. Further cytokine profiling indicates that the baseline deficiency of IFN-αA in the cd73−/− mice then leads to a specific inability to up-regulate expression of the immunomodulatory cytokine IL-10 in response to TNBS-induced colitis. This IL-10 response is rescued by exogenous administration of IFN-αA, restoring the animals to the TNBS phenotype of the wild-type mice.

Clear evidence supports a role for IL-10 in colonic inflammation with IL-10-deficient mice known to develop spontaneous chronic enterocolitis (40). Furthermore, IL-10 levels have been demonstrated to increase both in colonic tissue in experimental models of colitis (41) and in serum from human patients with active ulcerative colitis and in the recovery phase of both ulcerative colitis and Crohn’s disease (42). Here we demonstrate a role for CD73-mediated IFN-α in the production of IL-10 in response to colonic inflammation in the setting of TNBS colitis.

The only know function for CD73-mediated AMP hydrolysis is in the formation of adenosine (26). In the colon, previous studies have revealed that CD73 is expressed most prominently as a glycosylphosphatidylinositol-linked membrane protein on apical (luminal) surface of colonic epithelia (4). In the current studies, we have not precisely localized the source of IFN-αA in the colon. A number of limitations in protein detection have hindered our abilities to localize IFN-αA within tissue, and epithelial isolations are sufficiently contaminated to disallow specific conclusions. Thus, although we presume, based on our data obtained from mucosal scrapings, that IFN-αA is derived primarily from colonic epithelia, it is possible that other sources exist. Indeed, although type I IFNs can be made by most any nucleated cell (43), plasmacytoid dendritic cells (pDC) are a major source of IFN-α (44). However, past studies in the virology literature reveal that mature pDCs down-regulate their production of IFN-α via an A2AR-mediated mechanism (45), making them less likely to be directly responsible for a CD73-dependent deficiency in IFN-α. Nonetheless, a more indirect role for pDCs related to Ado-mediated alterations in chemotaxis and recruitment to epithelial tissues cannot be ruled out. Although one report has indicated that CD73 mRNA is expressed in monocyte-derived dendritic cells (46), essentially nothing is known about the function of CD73 on dendritic cells. Thus far, difficulties in obtaining pDC in adequate numbers for evaluation have hindered our attempts at characterizing CD73 function in these cells.

These studies further demonstrate conflicting results with respect to the influence of IFN-αA on the evolution of TNBS-associated acute inflammation in wild-type CD57/BL6 mice with total histological scoring supporting a beneficial effect, whereas other parameters indicate little or no improvement in the presence of exogenous IFN-αA. Our findings addressing exogenous administration of recombinant IFN-α support the recent clinical investigation of the importance of appropriate levels of IFN-α in the regulation of colonic inflammation.

Taken together, these findings indicate that CD73 functions in an endogenous protective mechanism in the acute phase of TNBS colitis through an IFN-αA-dependent manner. The lack of down-regulation of IFN-αA in response to colitis in the cd73−/− mice may indicate that IFN-αA production is already maximally repressed in these mice. Alternatively, the fact that IFN-αA is induced then rapidly down-regulated in the wild-type mice after induction of colitis may indicate that IFN-αA is involved in down-regulating its own production, possibly in a CD73-dependent manner. This hypothesis is supported by studies revealing CD73-dependent IFN-α regulation of Ado signaling in human endothelial cells (47). Further studies will be necessary to determine the molecular mechanisms underlying these effects as well as to determine whether a critical window for IFN-αA dosing exists and whether we might consider evaluating CD73 activity in patients who benefit from IFN-α therapy.

We acknowledge Dr. Linda Thompson for providing breeding stocks of cd73−/− mice, without which this work would not be possible.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by National Institutes of Health Grants DK50189, DE016191, HL60569, and DK62007 and by a grant from the Crohn’s and Colitis Foundation of America. S.P.C. holds the Kern Family Professorship of Medicine at the University of Colorado Health Sciences Center.

4

Abbreviations used in this paper: TNBS, trinitrobenzene sulfonate; CD73, ecto-5′-nucleotidase; Ado, adenosine; APCP, α,β-methylene ADP; PMN, polymorphonuclear cell or neutrophil; A2AR, adenosine A2A receptor; A2BR, adenosine A2B receptor; pDC, plasmacytoid dendritic cells.

1
Neurath, M., I. Fuss, W. Strober.
2000
. TNBS-colitis.
Int. Rev. Immunol.
19
:
51
-62.
2
Karhausen, J., G. T. Furuta, J. E. Tomaszewski, R. S. Johnson, S. P. Colgan, V. H. Haase.
2004
. Epithelial hypoxia-inducible factor-1 is protective in murine experimental colitis.
J. Clin. Invest.
114
:
1098
-1106.
3
Zimmerman, H..
1992
. 5′-Nucleotidase: molecular structure and functional aspects.
Biochem. J.
285
:
345
-365.
4
Strohmeier, G. R., W. I. Lencer, T. W. Patapoff, L. F. Thompson, S. L. Carlson, S. J. Moe, R. J. Mrsny, J. L. Madara.
1997
. Surface expression, polarization, and functional significance of CD73 in human intestinal epithelia.
J. Clin. Invest.
99
:
2588
-2601.
5
Lennon, P. F., C. T. Taylor, G. L. Stahl, S. P. Colgan.
1998
. Neutrophil-derived 5′-adenosine monophosphate promotes endothelial barrier function via CD73-mediated conversion to adenosine and endothelial A2B receptor activation.
J. Exp. Med.
188
:
1433
-1443.
6
Eltzschig, H. K., J. C. Ibla, G. T. Furuta, M. O. Leonard, K. A. Jacobson, K. Enjyoji, S. C. Robson, S. P. Colgan.
2003
. Coordinated adenine nucleotide phosphohydrolysis and nucleoside signaling in posthypoxic endothelium: role of ectonucleotidases and adenosine A2B receptors.
J. Exp. Med.
198
:
783
-796.
7
Synnestvedt, K., G. T. Furuta, K. M. Comerford, N. Louis, J. Karhausen, H. K. Eltzschig, K. R. Hansen, L. F. Thompson, S. P. Colgan.
2002
. Ecto-5′-nucleotidase (CD73) regulation by hypoxia-inducible factor-1 mediates permeability changes in intestinal epithelia.
J. Clin. Invest.
110
:
993
-1002.
8
Eltzschig, H., L. Thompson, J. Karhausen, R. Cotta, J. Ibla, S. Robson, S. Colgan.
2004
. Endogenous adenosine produced during hypoxia attenuates neutrophil accumulation: coordination by extracellular nucleotide metabolism.
Blood
104
:
3986
-3992.
9
Thompson, L., H. Eltzschig, J. Ibla, C. V. D. Wiele, R. Resta, J. Morote-Garcia, S. Colgan.
2004
. Crucial role for ecto-5′-nucleotidase (CD73) in vascular leakage during hypoxia.
J. Exp. Med.
200
:
1395
-1405.
10
Manns, M. P., H. Wedemeyer, M. Cornberg.
2006
. Treating viral hepatitis C: efficacy, side effects, and complications.
Gut
55
:
1350
-1359.
11
Wenner, W. J., Jr, D. A. Piccoli.
1997
. Colitis associated with α interferon?.
J. Clin. Gastroenterol.
25
:
398
-399.
12
Mavrogiannis, C., I. S. Papanikolaou, I. S. Elefsiniotis, D. I. Psilopoulos, A. Karameris, G. Karvountzis.
2001
. Ulcerative colitis associated with interferon treatment for chronic hepatitis C.
J. Hepatol.
34
:
964
-965.
13
Sprenger, R., M. Sagmeister, F. Offner.
2005
. Acute ulcerative colitis during successful interferon/ribavirin treatment for chronic hepatitis.
Gut
54
:
438
-439.
14
Ruther, U., C. Nunnensiek, H. A. Muller, H. Bader, U. May, P. Jipp.
1998
. Interferon α (IFN α 2a) therapy for herpes virus-associated inflammatory bowel disease (ulcerative colitis and Crohn’s disease).
Hepatogastroenterology
45
:
691
-699.
15
Capobianchi, M. R., S. Fais, F. Mercuri, M. Boirivant, F. Dianzani, F. Pallone.
1992
. Interferon-α (IFN-α) production by human intestinal mononuclear cells: response to virus in control subjects and in Crohn’s disease.
Gut
33
:
897
-901.
16
Gasche, C., W. Reinisch, H. Vogelsang, R. Potzi, E. Markis, M. Micksche, H. P. Wirth, A. Gangl, H. Lochs.
1995
. Prospective evaluation of interferon-α in treatment of chronic active Crohn’s disease.
Dig. Dis. Sci.
40
:
800
-804.
17
Nikolaus, S., P. Rutgeerts, R. Fedorak, A. H. Steinhart, G. E. Wild, D. Theuer, J. Mohrle, S. Schreiber.
2003
. Interferon β-1a in ulcerative colitis: a placebo controlled, randomised, dose escalating study.
Gut
52
:
1286
-1290.
18
Sumer, N., M. Palabiyikoglu.
1995
. Induction of remission by interferon-α in patients with chronic active ulcerative colitis.
Eur. J. Gastroenterol. Hepatol.
7
:
597
-602.
19
Madsen, S. M., P. Schlichting, B. Davidsen, O. H. Nielsen, B. Federspiel, P. Riis, P. Munkholm.
2001
. An open-labeled, randomized study comparing systemic interferon-α-2A and prednisolone enemas in the treatment of left-sided ulcerative colitis.
Am. J. Gastroenterol.
96
:
1807
-1815.
20
Uguralp, S., B. Mizrak, A. B. Karabulut, N. Gurbuz, M. Demircan.
2004
. Interferon-α reduces the development of experimental necrotizing enterocolitis.
Cytokine
25
:
204
-211.
21
Katakura, K., J. Lee, D. Rachmilewitz, G. Li, L. Eckmann, E. Raz.
2005
. Toll-like receptor 9-induced type I IFN protects mice from experimental colitis.
J. Clin. Invest.
115
:
695
-702.
22
Morris, G. P., P. L. Beck, M. S. Herridge, W. T. Depew, M. R. Szewczuk, J. L. Wallace.
1989
. Hapten-induced model of chronic inflammation and ulceration in the rat colon.
Gastroenterology
96
:
795
-803.
23
Dieleman, L. A., M. J. Palmen, H. Akol, E. Bloemena, A. S. Pena, S. G. Meuwissen, E. P. Van Rees.
1998
. Chronic experimental colitis induced by dextran sulphate sodium (DSS) is characterized by Th1 and Th2 cytokines.
Clin. Exp. Immunol.
114
:
385
-391.
24
Furuta, G. T., J. R. Turner, C. T. Taylor, R. M. Hershberg, K. M. Comerford, S. Narravula, D. K. Podolsky, S. P. Colgan.
2001
. Hypoxia-inducible factor 1-dependent induction of intestinal trefoil factor protects barrier function during hypoxia.
J. Exp. Med.
193
:
1027
-1034.
25
Hasko, G., B. N. Cronstein.
2004
. Adenosine: an endogenous regulator of innate immunity.
Trends Immunol.
25
:
33
-39.
26
Hunsucker, S. A., B. S. Mitchell, J. Spychala.
2005
. The 5′-nucleotidases as regulators of nucleotide and drug metabolism.
Pharmacol. Ther.
107
:
1
-30.
27
van Pesch, V., H. Lanaya, J. C. Renauld, T. Michiels.
2004
. Characterization of the murine α interferon gene family.
J. Virol.
78
:
8219
-8228.
28
Kalie, E., D. A. Jaitin, R. Abramovich, G. Schreiber.
2007
. An interferon α2 mutant optimized by phage display for IFNAR1 binding confers specifically enhanced antitumor activities.
J. Biol. Chem.
282
:
11602
-11611.
29
Brideau-Anderson, A. D., X. Huang, S. C. Sun, T. T. Chen, D. Stark, I. J. Sas, L. Zadik, G. N. Dawes, D. R. Guptill, R. McCord, et al
2007
. Directed evolution of gene-shuffled IFN-α molecules with activity profiles tailored for treatment of chronic viral diseases.
Proc. Natl. Acad. Sci. USA
104
:
8269
-8274.
30
Zernecke, A., K. Bidzhekov, B. Ozuyaman, L. Fraemohs, E. A. Liehn, J. M. Luscher-Firzlaff, B. Luscher, J. Schrader, C. Weber.
2006
. CD73/ecto-5′-nucleotidase protects against vascular inflammation and neointima formation.
Circulation
113
:
2120
-2127.
31
Volmer, J. B., L. F. Thompson, M. R. Blackburn.
2006
. Ecto-5′-nucleotidase (CD73)-mediated adenosine production is tissue protective in a model of bleomycin-induced lung injury.
J. Immunol.
176
:
4449
-4458.
32
Zimmermann, H., N. Braun.
1999
. Ecto-nucleotidases–molecular structures, catalytic properties, and functional roles in the nervous system.
Prog. Brain Res.
120
:
371
-385.
33
Naganuma, M., E. B. Wiznerowicz, C. M. Lappas, J. Linden, M. T. Worthington, P. B. Ernst.
2006
. Cutting edge: critical role for A2A adenosine receptors in the T cell-mediated regulation of colitis.
J. Immunol.
177
:
2765
-2769.
34
Cavalcante, I. C., M. V. Castro, A. R. Barreto, G. W. Sullivan, M. Vale, P. R. Almeida, J. Linden, J. M. Rieger, F. Q. Cunha, R. L. Guerrant, et al
2006
. Effect of novel A2A adenosine receptor agonist ATL 313 on Clostridium difficile toxin A-induced murine ileal enteritis.
Infect. Immun.
74
:
2606
-2612.
35
Odashima, M., G. Bamias, J. Rivera-Nieves, J. Linden, C. C. Nast, C. A. Moskaluk, M. Marini, K. Sugawara, K. Kozaiwa, M. Otaka, et al
2005
. Activation of A2A adenosine receptor attenuates intestinal inflammation in animal models of inflammatory bowel disease.
Gastroenterology
129
:
26
-33.
36
Kolachala, V., V. Asamoah, L. Wang, T. S. Obertone, T. R. Ziegler, D. Merlin, S. V. Sitaraman.
2005
. TNF-α upregulates adenosine 2b (A2b) receptor expression and signaling in intestinal epithelial cells: a basis for A2bR overexpression in colitis.
Cell Mol. Life Sci.
62
:
2647
-2657.
37
Guzman, J., J. G. Yu, Z. Suntres, A. Bozarov, H. Cooke, N. Javed, H. Auer, J. Palatini, H. H. Hassanain, A. J. Cardounel, et al
2006
. ADOA3R as a therapeutic target in experimental colitis: proof by validated high-density oligonucleotide microarray analysis.
Inflamm. Bowel Dis.
12
:
766
-789.
38
Mabley, J., F. Soriano, P. Pacher, G. Hasko, A. Marton, R. Wallace, A. Salzman, C. Szabo.
2003
. The adenosine A3 receptor agonist, N6-(3-iodobenzyl)-adenosine-5′-N-methyluronamide, is protective in two murine models of colitis.
Eur. J. Pharmacol.
466
:
323
-329.
39
Tilg, H., A. Kaser.
2004
. Type I interferons and their therapeutic role in Th2-regulated inflammatory disorders.
Expert Opin. Biol. Ther.
4
:
469
-481.
40
Kuhn, R., J. Lohler, D. Rennick, K. Rajewsky, W. Muller.
1993
. Interleukin-10-deficient mice develop chronic enterocolitis.
Cell
75
:
263
-274.
41
Tomoyose, M., K. Mitsuyama, H. Ishida, A. Toyonaga, K. Tanikawa.
1998
. Role of interleukin-10 in a murine model of dextran sulfate sodium-induced colitis.
Scand. J. Gastroenterol.
33
:
435
-440.
42
Mitsuyama, K., N. Tomiyasu, K. Takaki, J. Masuda, H. Yamasaki, K. Kuwaki, T. Takeda, S. Kitazaki, O. Tsuruta, M. Sata.
2006
. Interleukin-10 in the pathophysiology of inflammatory bowel disease: increased serum concentrations during the recovery phase.
Mediat. Inflamm.
2006
:
26875
43
Doly, J., A. Civas, S. Navarro, G. Uze.
1998
. Type I interferons: expression and signalization.
Cell Mol. Life Sci.
54
:
1109
-1121.
44
Siegal, F. P., N. Kadowaki, M. Shodell, P. A. Fitzgerald-Bocarsly, K. Shah, S. Ho, S. Antonenko, Y. J. Liu.
1999
. The nature of the principal type 1 interferon-producing cells in human blood.
Science
284
:
1835
-1837.
45
Schnurr, M., T. Toy, A. Shin, G. Hartmann, S. Rothenfusser, J. Soellner, I. D. Davis, J. Cebon, E. Maraskovsky.
2004
. Role of adenosine receptors in regulating chemotaxis and cytokine production of plasmacytoid dendritic cells.
Blood
103
:
1391
-1397.
46
Berchtold, S., A. L. Ogilvie, C. Bogdan, P. Muhl-Zurbes, A. Ogilvie, G. Schuler, A. Steinkasserer.
1999
. Human monocyte derived dendritic cells express functional P2X and P2Y receptors as well as ecto-nucleotidases.
FEBS Lett.
458
:
424
-428.
47
Niemela, J., T. Henttinen, G. G. Yegutkin, L. Airas, A. M. Kujari, P. Rajala, S. Jalkanen.
2004
. IFN-α induced adenosine production on the endothelium: a mechanism mediated by CD73 (ecto-5′-nucleotidase) up-regulation.
J. Immunol.
172
:
1646
-1653.