Cortistatin, an antiinflammatory peptide withtherapeutic action in inflammatory bowel diseaseElena Gonzalez-Rey*, Nieves Varela*, Amir F. Sheibanie†, Alejo Chorny*, Doina Ganea†, and Mario Delgado*‡

*Institute of Parasitology and Biomedicine, Consejo Superior de Investigaciones Cientıficas, 18100 Granada, Spain; and †Department of Biological Sciences,Rutgers, The State University of New Jersey, Newark, NJ 07102

Edited by Arthur Weiss, University of California School of Medicine, San Francisco, CA, and approved January 11, 2006 (received for reviewOctober 17, 2005)

Cortistatin is a recently discovered cyclic neuropeptide related tosomatostatin that has emerged as a potential endogenous antiin-flammatory factor based on its production by, and binding to,immune cells. Crohn’s disease is a chronic debilitating diseasecharacterized by severe T helper 1 (Th1)-driven inflammation of thegastrointestinal tract. The aim of this study is to investigate thetherapeutic effect of cortistatin in a murine model of colitis.Cortistatin treatment significantly ameliorated the clinical andhistopathologic severity of the inflammatory colitis, abrogatingbody weight loss, diarrhea, and inflammation and increased thesurvival rate of the colitic mice. The therapeutic effect was asso-ciated with down-regulation of inflammatory and Th1-driven au-toimmune response, including the regulation of a wide spectrumof inflammatory mediators. In addition, a partial involvement ofregulatory IL-10-secreting T cells in this therapeutic effect wasdemonstrated. Importantly, cortistatin treatment was therapeuti-cally effective in established colitis and avoided the recurrence ofthe disease. This work identifies cortistatin as an antiinflammatoryfactor with the capacity to deactivate the intestinal inflammatoryresponse and restore mucosal immune tolerance at multiple levels.Consequently, cortistatin represents a multistep therapeutic ap-proach for the treatment of Crohn’s disease and other Th1-medi-ated inflammatory diseases.

autoimmunity � cytokines � inflammation � neuroimmunology

Inflammatory bowel disease (IBD) is a family of chronic, idio-pathic, relapsing, and tissue-destructive diseases characterized by

dysfunction of mucosal T cells, altered cytokine production, andcellular inflammation that ultimately leads to damage of the distalsmall intestine and the colonic mucosa (1). IBD is clinicallysubdivided into two phenotypes: Crohn’s disease (CD) and ulcer-ative colitis. CD is an incurable autoimmune disease with a prev-alence of 0.05% that leads to chronic inflammation resulting in arange of gastrointestinal and extraintestinal symptoms, includingabdominal pain, rectal bleeding, diarrhea, weight loss, skin and eyedisorders, and delayed growth and sexual maturation in children(2). Although its etiology remains unknown, there is circumstantialevidence to link CD to a failure of the mucosal immune system toattenuate the immune response to endogenous antigens (1). Severalanimal models of CD have been developed lately. Although manyof these models incompletely resemble the human disease (3, 4), thehapten-induced model of colonic inflammation, in which 2,4,6-trinitrobenzene sulfonic acid (TNBS) is delivered intrarectally,displays human CD-like clinical, histopathological, and immuno-logical features (4). In this model, intestinal inflammation resultsfrom a covalent binding of the haptenizing agent to autologous hostproteins with subsequent stimulation of a delayed-type hypersen-sitivity to TNBS-modified self-antigens. Similarly to human CD,TNBS-induced colitis is marked by an immune response to exag-gerated gut-associated lymphoid tissue development, which givesrise to a prolonged severe transmural inflamed intestinal mucosacharacterized by uncontrolled production of inflammatory cyto-kines and oligoclonal expansion and activation of CD4 T cellsspecifically associated with a T helper 1 (Th1) response (4).

Therapeutic agents currently used for CD are not entirelyeffective, are rather nonspecific, and have multiple adverse sideeffects (2). In most cases, surgical resection is the ultimate alter-native. Therefore, the present therapeutic strategy is to find drugsor agents that specifically modulate both components of the disease,i.e., the inflammatory and the Th1-driven responses.

Cortistatin is a recently discovered cyclic neuropeptide related tosomatostatin (5). Cortistatin binds to all five cloned somatostatinreceptors and shares many of the somatostatin pharmacologicaland functional properties, including the depression of neuronalactivity and inhibition of cell proliferation (6). However, cortistatinalso has many properties distinct from somatostatin, includinginduction of slow-wave sleep and reduction of locomotor activity(6). Cortistatin, but not somatostatin, has been detected in varioushuman immune cells, including lymphocytes, monocytes, macro-phages, and dendritic cells (7, 8). Some of the somatostatin immu-nomodulatory actions (9) with a predominant antiinflammatoryeffect might be shared by cortistatin. Also, because cortistatinexpression levels correlate with immune cell differentiation andactivation (7, 8), cortistatin might be a major endogenous regula-tory factor in the immune system. In addition to the binding tosomatostatin receptors in immune cells, cortistatin can also bind toother receptors, including the receptor for the growth hormonesecretagogue ghrelin (10), a hormone recently described as a potentantiinflammatory factor (11, 12). The aim of this study is toinvestigate the potential antiinflammatory action of cortistatin andits potential therapeutic use in the TNBS-induced model of colitis.

ResultsCortistatin Inhibits the Production of Inflammatory Mediators byActivated Macrophages in Vitro. To investigate the potential antiin-flammatory action of cortistatin, we evaluated first the effect ofcortistatin on the production of several inflammatory mediators byperitoneal macrophages. Cortistatin inhibited the production ofTNF-�, IL-6, IL-1�, IL-12, macrophage-inflammatory protein(MIP)-2, Rantes, and nitric oxide (NO) by activated macrophages(Fig. 5, which is published as supporting information on the PNASweb site). Interestingly, cortistatin showed a higher inhibitory effectthan the structurally related peptide somatostatin, and the soma-tostatin receptor-antagonist cyclosomatostatin partially reversedthe inhibitory effect of cortistatin, although it fully blocked theeffect of somatostatin (Fig. 5). These results suggest that cortistatincould exert its effects through both somatostatin receptor-dependent and -independent mechanisms. Indeed, a ghrelin-

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: CD, Crohn’s disease; IBD, inflammatory bowel disease; LPMC, lamina propriamononuclear cell; MIP, macrophage-inflammatory protein; MLN, mesenteric lymph node;MPO, myeloperoxidase; SAA, serum amyloid A; Th1, T helper 1; TNBS, 2,4,6-trinitrobenzenesulfonic acid; Treg, regulatory T cell.

‡To whom correspondence should be addressed at: Instituto de Parasitologia y Biomedi-cina, Consejo Superior de Investigaciones Cientıficas, Avenida Conocimiento, ParqueTecnologico de Ciencias de la Salud, Granada 18100, Spain. E-mail: mdelgado@ipb.csic.es.

© 2006 by The National Academy of Sciences of the USA

4228–4233 � PNAS � March 14, 2006 � vol. 103 � no. 11 www.pnas.org�cgi�doi�10.1073�pnas.0508997103

receptor antagonist partially blocked the inhibitory effect of cor-tistatin (Fig. 5).

Cortistatin Protects Against TNBS Colitis Development. We nextinvestigated the potential therapeutic action of cortistatin in theTNBS model of colitis. Mice subjected to intrarectal administrationof TNBS in 50% ethanol developed a severe illness characterizedby bloody diarrhea, rectal prolapse, pancolitis accompanied byextensive wasting syndrome, and a profound and sustained weightloss resulting in a mortality of 60% (Fig. 1 A–D). Mice treated withcortistatin 12 h after TNBS instillation rapidly recovered the lostbody weight, improved the wasting disease, and had a healthyappearance, with a survival of 90%, similar to control mice treatedwith 50% ethanol alone (Fig. 1 A–D). The therapeutic effect ofcortistatin was dose-dependent, showing maximal effects at dosesbetween 0.2 and 2 nmol (25–250 �g�kg) (Fig. 1B). Macroscopicexamination of colons obtained 3 and 7 days after colitis inductionshowed striking hyperemia, inflammation, and necrosis comparedwith control animals that only showed slight inflammation (Fig.1E). In contrast, the colons of cortistatin-treated mice showed nosigns of macroscopic inflammation (Fig. 1E). Histological exami-nation of the distal colon of mice given TNBS showed transmuralinflammation involving all layers of the bowel wall with a marked

increase in the thickness of the muscular layer, adherence tosurrounding tissues, patchy ulceration, epithelial cell loss, pro-nounced depletion of mucin-producing goblet cells, reduction of thedensity of the tubular glands, disseminated fibrosis, and focal lossof cripts (Fig. 1F). Inflammatory cell infiltrates consisted of mac-rophages, lymphocytes, and neutrophils in the lamina propria (Fig.1F), along with enlargements of lymphoid follicles in the colon(data not shown). Immunohistological analysis revealed CD4 Tcells, TNF-�-producing cells, and CD11b� cells, i.e., granulocytesand macrophages (Fig. 1G). Neutrophil infiltration correlated withincreased colonic myeloperoxidase (MPO) activity (Fig. 1H).When mice were treated with cortistatin, a striking improvement ofthese macroscopic and histological signs became apparent, with asignificant reduction in the inflammatory activity and neutrophilinfiltration (Fig. 1 F–H).

We next investigated whether cortistatin would be effectiveduring the later phases of the disease with colitis fully established.Administration of cortistatin on 3 consecutive days starting 6 daysafter onset of disease rapidly reversed the lost body weight (Fig. 2A).In addition, we examined whether cortistatin was able to preventdisease recurrence. TNBS-treated mice reexposed on day 9 to asecond dose of TNBS rapidly died (100% mortality) due to severecolitis and body weight loss. Mice receiving an unique dose of

Fig. 1. Treatment with cortistatin protects against TNBS colitis development. Colitis was induced by intracolonic administration of TNBS (3 mg per mouse) in50% ethanol. Mice were treated i.p. with cortistatin (2 nmol per mouse or different doses in B) 12 h after TNBS injection. Mice treated with 50% ethanol wereused as controls. (A–D) Clinical evolution and severity was monitored by body weight changes (A and B), colitis score (C), and survival (D). (E) Macroscopic-damagescore was determined at 3 days after TNBS administration. (F) Histopathologic analysis was performed in hematoxylin�eosin-stained sections of colons at day3 of disease. (Magnification, �200.) (G) Inflammatory infiltrates in the colons (day 3) were phenotypically characterized by immunostaining against CD11b(monocyte�macrophage and neutrophils), TNF-�-producing cells, or CD4 T cells. (Magnification, �200.) (H) Colonic MPO activity was determined in the acutephase of the disease (day 3). n � 12–18 mice per group; *, P � 0.001 versus TNBS-treated mice.

Gonzalez-Rey et al. PNAS � March 14, 2006 � vol. 103 � no. 11 � 4229

MED

ICA

LSC

IEN

CES

cortistatin 12 h after the initial colitis induction survived and did notsuffer disease recurrence after a second administration of TNBS(Fig. 2B). The therapeutic effect of cortistatin was confirmed in SJLmice, a murine strain more susceptible for TNBS-induced colitis(Fig. 2B).

Cortistatin was more efficient at ameliorating the body weightloss and colitis than its two structurally related peptides, somatosta-tin and octreotide, and both somatostatin- and ghrelin-receptorantagonists partially reversed cortistatin’s effects (Fig. 6, which ispublished as supporting information on the PNAS web site).

Treatment with Cortistatin Reduces Systemic and Mucosal Inflamma-tory Responses in Mice with TNBS-Induced Colitis. We next evaluatedthe effect of cortistatin on the production of inflammatory medi-ators that are mechanistically linked to TNBS-induced colitis.Cortistatin dramatically reduced protein and mRNA expression ofinflammatory cytokines (TNF-�, IFN-�, IL-6, IL-1�, IL-1�, IL-12,IL-18, IL-17, IL-15, and macrophage migration inhibitory factor),chemokines (Rantes, MIP-1�, MIP-1�, MIP-3�, monocyte che-moattractant proteins 1 and 3, IP-10, and MIP-2) and chemokine

receptors (CCR-1, CCR-2, CCR-3, CCR-5, and CCR-7) in themucosa of colitic mice (Fig. 3). In addition, colons of cortistatin-treated mice showed increased levels of the antiinflammatorycytokine IL-10 and chemokine receptors CCR-4 and CCR-8 (Fig.3). The decrease in inflammatory mediators could be a conse-quence of the diminished infiltration of inflammatory cells in thecolonic mucosa in the cortistatin-treated colitic mice. However,lamina propria mononuclear cells (LPMC) isolated from cortista-tin-treated mice produced lower levels of proinflammatory factors(TNF-�, IL-6, and MIP-2) upon in vitro culture in comparison withTNBS mice (Fig. 3C). These findings suggest that, in addition to thereduction in inflammatory infiltration, cortistatin administrationdeactivates the inflammatory response in the colonic mucosa. Thebroad antiinflammatory activity of cortistatin in the colon wasaccompanied by down-regulation of the systemic inflammatoryresponse implicated in colonic inflammation (Fig. 3D). Cortistatindecreased the TNBS-induced serum levels of the proinflammatorycytokines TNF-�, IL-1�, IL-6, and MIP-2, and of serum amyloid A(SAA), a hepatic acute phase protein involved in tissue damage ininflammatory conditions.

Cortistatin Suppresses Th1 Cytokine Response and Stimulates IL-10Production in TNBS-Induced Colitis. Although macrophages andneutrophils are the major sources of inflammatory mediators, CD4T cells play a key role in the initiation and perpetuation of CD byproducing IFN-�, a potent inducer of the inflammatory response(4). In fact, CD and TNBS-induced colitis are considered Th1-typecell-mediated autoimmune diseases (4, 13, 14). Therefore, wedetermined the effect of the cortistatin treatment of TNBS-inducedcolitis on the ability of LPMC and draining mesenteric lymph node(MLN) cells to produce IFN-� and proliferate in vitro. LPMC andMLN cells obtained from TNBS-treated mice proliferate more andproduced significantly more IFN-� than ethanol-treated mice,and in vitro activation of these cells caused further cell expansionand increased amounts of IFN-� (Fig. 4 A and B). In contrast,LPMC and MLN cells isolated from cortistatin-treated coliticanimals proliferate less and produce significantly lower amounts ofIFN-� than do cells isolated from TNBS-treated mice, even afterpotent T cell stimulation (Fig. 4 A and B). In addition, theproduction of the regulatory cytokine IL-10 was significantly in-creased in LPMC and MLN cells obtained from cortistatin-treatedmice, specially upon in vitro activation; the Th2-type cytokine IL-4was not significantly affected (Fig. 4B). Thus, cortistatin decreasedTh1 cytokine production in vivo and abrogated the responsivenessof LPMC and MLN cells to subsequent in vitro stimulation. Thedecreased IFN-� production and proliferation of LPMC and MLNcells is specific to cells residing in the lamina propria environmentor draining MLN, because splenocytes from cortistatin-treated anduntreated TNBS mice equally proliferate and produce IFN-� uponstimulation (data not shown). Given that the decrease in IFN-�production induced by cortistatin treatment could be a conse-quence of either down-regulation of IFN-� release or inhibition ofTh1 cell differentiation and that the production of IL-10 could bedue to macrophages and CD4 T cells, we determined the intracel-lular expression of these cytokines by flow cytometry in sorted CD4T cells. Cortistatin significantly decreased the number of IL-2�IFN-�-producing Th1 cells and increased the number of IL-10-producing CD4 T cells in LPMC and MLN (Fig. 4B). Thus,cortistatin administration to colitic mice regulates the generation�differentiation of autoreactive�inflammatory Th1 cells and, pre-sumably, regulatory IL-10-secreting T cells.

DiscussionThe present study reports cortistatin as a peptide with potentantiinflammatory actions in vivo. Our data demonstrate that cor-tistatin provides a highly effective treatment for TNBS-inducedcolitis, a murine experimental model of CD. A single injection ofcortistatin at the onset of the disease ameliorated the clinical and

Fig. 2. Cortistatin treatment abrogates established colitis and reduces dis-ease recurrence. (A) Established colitis. Colitis was induced by intracolonicadministration of TNBS (1.5 mg per mouse) at days 0 and 6. Mice were treateddaily for 3 consecutive days with cortistatin (2 nmol per mouse) starting 6 daysafter TNBS administration (arrow). Disease progression was assessed by bodyweight loss. n � 8 mice per group. (B) Disease recurrence. Colitis was inducedin BALB�c and SJL mice by intracolonic administration of TNBS (3 mg permouse) at days 0 and 9 (arrows). Mice were treated i.p. with cortistatin (2 nmolper mouse) 12 h after TNBS injection. Controls were given with a secondinjection of ethanol on day 9. Disease progression was assessed by bodyweight loss and survival percentage. The numbers in parentheses representdaily mortality percentages after the second TNBS infusion. n � 8–10 mice pergroup.

4230 � www.pnas.org�cgi�doi�10.1073�pnas.0508997103 Gonzalez-Rey et al.

histopathologic severity of the wasting disease, abrogating bodyweight loss, diarrhea, and intestinal inflammation and reduced thehigh mortality caused by this syndrome. From a therapeutic pointof view, it is extremely important to take into account the ability ofdelayed administration of cortistatin to ameliorate ongoing diseaseand that an initial treatment with cortistatin prevented recurrenceof the disease given a second dose of TNBS, which fulfills anessential prerequisite for an anticolitic agent, because treatment isstarted after the onset of CD patients.

There are several potential mechanisms by which cortistatintherapy can modulate the effector phase of TNBS colitis. CD andTNBS-induced colitis are characterized by transmural inflamma-tion of the colon because of an IL-12-driven, Th1 cell-mediatedresponse to TNBS-haptenated colonic proteins and�or crossreac-tive luminal antigens (4). During the effector phase of bowelinflammation, both innate and acquired immune responses overlap,and multiple inflammatory mediators are involved. Cortistatinstrongly reduced mucosal inflammation by down-regulating theproduction of a wide panel of mediators involved in the local andsystemic inflammatory response. Chemokines are responsible forthe mucosal infiltration and activation of various leukocyte popu-lations, which contribute to colitis development (15). The fact thatcortistatin treatment reduced the production of a plethora ofchemokines could partially explain the absence of inflammatoryinfiltrates in the colonic mucosa of cortistatin-treated mice, beingespecially relevant for chemokines, such as MIP-2 (chemotactic forneutrophils), IP-10 (for Th1 cells), and Rantes�MIP-1� (for mac-rophages and T cells), all of which are involved in CD pathogenesis

(15). In addition to the regulation of cell recruitment to the laminapropria during colitis, cortistatin regulates the inflammatory cellactivation and cytokine production. Thus, cortistatin down-regulated the production of the proinflammatory�cytotoxic cyto-kines TNF-�, IFN-�, IL-6, IL-1�, IL-12, IL-15, IL-17, IL-18, andmacrophage migration inhibitory factor by mucosal immune cellsand increased the levels of the antiinflammatory cytokine IL-10.The decrease in inflammatory mediators could be the consequenceof a diminished infiltration of inflammatory cells in the colonicmucosa in the cortistatin-treated TNBS mice. However, the factthat LPMC isolated from cortistatin-treated mice produced lowerlevels of proinflammatory factors upon in vitro activation arguesagainst this hypothesis, which suggests that, in addition to thereduction in inflammatory infiltration, cortistatin administrationdeactivates the inflammatory response. In this study, we alsodemonstrated that cortistatin acts as a macrophage-deactivatingfactor by down-regulating the production of a wide board ofinflammatory mediators. Therefore, it is plausible that deactivationof resident and infiltrating mucosal macrophages is a major mech-anism involved in the antiinflammatory action of cortistatin in IBD.The local antiinflammatory action of cortistatin was also evidentsystemically. Of special consideration is the cortistatin reduction ofSAA, an acute phase protein used for clinical monitoring of CD(16), which has been associated with tissue damage in severalinflammatory conditions (17), although its precise role in intestinalinflammation is not known.

TNBS-induced colitis also is a Th1-mediated disease, requiring Tcell activation as the central initiating event that subsequently leads

Fig. 3. Treatment with cortistatin re-duces systemic and mucosal inflammatoryresponses in mice with TNBS-induced coli-tis. Colitis was induced by intracolonic ad-ministration of TNBS. Mice were treatedi.p. with cortistatin (2 nmol per mouse) 12 hafter TNBS injection. Mice treated with eth-anol alone were used as controls. At thepeak of the disease (day 3), serum was col-lected, protein extracts and total RNA wereobtained from colons, and LPMC were iso-lated. (A) Cytokine�chemokine contents inprotein extracts were determined by ELISA.n � 5–6 mice per group. (B) Gene expres-sion of several inflammatory�autoimmunemediators was determined by microarray.Results are representative of two separateexperiments (n � 4–5 mice per group). MIF,macrophage migration inhibitory factor;MCP, monocyte chemoattractant protein.(C) Isolated LPMC were cultured in vitro for24 h, and the cytokine�chemokine con-tents in the supernatants were determinedby ELISA. n � 5–6 mice per group. (D) Cy-tokine�chemokine and SAA contents weredetermined in sera by ELISA. n � 5–6 miceper group; *, P � 0.001 versus TNBS-treatedmice.

Gonzalez-Rey et al. PNAS � March 14, 2006 � vol. 103 � no. 11 � 4231

MED

ICA

LSC

IEN

CES

to macrophage recruitment and activation (4). The bias toward Th1cytokines (mainly IFN-� and TNF-�) is crucial in the establishmentof chronic inflammation. Our results demonstrate that the expres-sion of the Th1-type cytokines IFN-� and TNF-� in the colon isdown-regulated in response to cortistatin after the induction ofTNBS colitis. It appears that the inhibition of the Th1 response iscaused by a direct action on LPMC and draining MLN cells,because LPMC and MLN cells obtained from cortistatin-treatedanimals are refractory to Th1 cell stimulation. In contrast to IFN-�and TNF-�, cortistatin increased the production of IL-10 in vivo.The effect of cotistatin on IFN-� and IL-10 production by T cellsseems to be independent of the action of the peptide on macro-phages, because cortistatin directly affects T cells in the absence ofantigen-presenting cells (our unpublished data). However, the factthat cortistatin increased IL-10 but not IL-4 production in CD4 Tcells from LPMC and MLN argues against a shift toward Th2responses. IL-10 has been recognized recently as a signaturecytokine for a subset of CD4 T cells that exert regulatory functions(18). Active suppression by IL-10-secreting regulatory T cells (Treg)plays a key role in the control of self-antigen-reactive T cells and theinduction of peripheral tolerance in vivo (18). In fact, deletion ordysfunction of these suppressive cells result in the appearance ofmultiple autoimmune inflammatory disorders, especially in theintestine (19–21). Our data suggest that cortistatin induces thegeneration�activation of IL-10-secreting Treg cells. This findingcorrelates with the fact that cortistatin increased the colonicexpression of CCR-4 and CCR-8, two chemokine receptors ex-pressed in Treg cells (22, 23). Cortistatin may favor the recruitmentof Treg to the inflamed mucosa. However, we did not observed asignificant increase in the expression of ligands for CCR4 andCCR8 (CCL17 and CCL1) in the colon after cortistatin treatment.Although the mechanism is still unknown, preliminary experimentsindicate that cortistatin induces the differentiation of tolerogenicdendritic cells with capacity to generate IL-10-secreting Treg cells(our unpublished results).

The improvement in wasting disease seen in mice treated withcortistatin (�90%) compares favorably to that achieved with othertherapies, such as blocking of IL-12, TNF-�, IL-6 or IFN-�, ortreatment with prednisolone or mesalamine (4, 13, 24–27), some ofthem widely used in treating IBD patients. The capacity of cor-tistatin to regulate a wide spectrum of inflammatory mediators, inaddition to suppression of Th1-type responses and potential gen-eration of Treg cells, might offer a therapeutic advantage overneutralizing Abs directed against a single mediator.

Cortistatin shares many structural and functional properties withsomatostatin. The amino acid sequences of these cyclic peptides, thegene structures, the partial coexpression, and the activation ofcommon receptors and signaling pathways suggest that the twopeptides exert similar functions. Although somatostatin-deficientmice do not display an overt phenotype (28), the lack of increasedcortistatin expression in these mice argues against a compensatoryrole of cortistatin. The distinct functions of cortistatin in thenervous system (5, 6) and the present work support this hypothesis.In fact, cortistatin is significantly more efficient in preventing colitisdevelopment than somatostatin or its agonist octreotide. Thesuperior potency of cortistatin in reducing inflammation may residein its capacity to activate different receptors and transductionpathways. Whereas somatostatin and octreotide only bind to so-matostatin receptors, cortistatin also can activate other receptors,including the receptor of ghrelin, a orexigenic hormone recentlyidentified as a potent antiinflammatory factor (10–12). We havefound that ghrelin also protects against TNBS-induced colitis byinhibiting the production of inflammatory cytokines and chemo-kines by macrophages by down-regulating the production of Th1cytokines by CD4 T cells and by increasing the production of IL-10by macrophages and T lymphocytes (29). Therefore, the possibilityexists that cortistatin is exerting its therapeutic effect on IBD, atleast partially, through the ghrelin receptors. Indeed, cortistatineffects on inflammation and colitis were partially reversed bysomatostatin- and ghrelin-receptor antagonists. Finally, the partic-ipation of cortistatin-specific receptors, not yet identified, cannot beruled out.

Of physiological relevance is the observation that the expressionof cortistatin is increased in activated inflammatory cells (7, 8).Although the levels of cortistatin have not been yet measured in CDpatients, it is attractive to speculate that the body responds to anexacerbated inflammatory response by increasing the peripheralproduction of endogenous antiinflammatory factors, including cor-tistatin. In our study, the animals did not exhibit side effects,probably because a short period of treatment with the peptide isenough to get a significant disease remission without recurrence,and plasma cortistatin levels only slightly increased after cortistatintreatment (32 versus 62 pg�ml). Extending the use of cortistatin tothe human system, however, will depend on the peptide dosage andthe expression of somatostatin�cortistatin�ghrelin receptors inhuman immune cells, because species-related differences in expres-sion have been found (7–9). In addition, based in its somatostatin-like structure, cortistatin should be unstable and susceptible to

Fig. 4. Cortistatin suppresses Th1 cytokine response and stimulates IL-10 production in TNBS-induced colitis. Colitis was induced by intracolonic administrationof TNBS, and mice were treated i.p. with cortistatin (2 nmol per mouse) 12 h after TNBS injection. Mice treated with ethanol alone were used as controls. MLNand LPMC were isolated at the peak of the disease (day 3), and cultured with medium alone (unstimulated) or with 12-myristate 13-acetate plus ConA(stimulated). (A) Proliferation was determined after 4 days of culture. The cytokine contents in the supernatants were determined by ELISA after 48 h of culture.(B) Stimulated lymphocytes were analyzed for CD4, and intracellular cytokine expression was analyzed by flow cytometry. Double staining for IFN-��IL-2 orIL-4�IL-10 expression was performed in gated CD4 T cells. The number of IFN-�- and IL-10-expressing T cells relative to 104 CD4 T cells was determined (bar graphs).Data represent values from two independent experiments (n � 5–6 mice per group per experiment). *, P � 0.001 versus TNBS-treated mice.

4232 � www.pnas.org�cgi�doi�10.1073�pnas.0508997103 Gonzalez-Rey et al.

degradation, making the peptide not very useful for clinical appli-cation in theory. However, the in vivo effects showed by cortistatinin this model are very convincing. Therefore, some of the fragmentresultants of the degradation of the peptide could still retainbiological functions, as previously described in other systems (30).In fact, no increased intact cortistatin levels were detected in colonsof cortistatin-treated mice.

In summary, this work identifies cortistatin as an immunomodu-latory factor with the capacity to deactivate the inflammatoryresponse in vivo at multiple levels and provides a powerful rationalefor the assessment of the efficacy of cortistatin as a therapeuticapproach to the treatment of CD.

MethodsInduction of Colitis and Study Design. Colitis was induced in 6- to8-week-old BALB�c and SJL mice (The Jackson Laboratory) asdescribed in ref. 13. Briefly, 3 mg of TNBS (Sigma) in 50% ethanol(to break the intestinal epithelial barrier) was administered into thecolon via a catheter. Control mice received 50% ethanol alone.Animals were treated i.p. with medium or with different concen-trations (0.05–2.0 nmol per mouse; 6–250 �g�kg) of cortistatin 1–29(American Peptide, Sunnyvale, CA), somatostatin, or octreotide(Sigma) 12 h after TNBS instillation. In some experiments, cyclo-somatostatin or [D-Lys-3]-growth hormone releasing peptide 6(Lys-GHRP-6; Sigma) was injected i.p. (250 �g per mouse) withcortistatin or octreotide. To study the therapeutic effect of thedelayed administration of cortistatin on established colitis, cortista-tin (2 nmol per mouse) was injected i.p. for 3 consecutive daysstarting 6 days after TNBS administration. To study the effect ondisease recurrence, 1.5 mg of TNBS was administered at days 0 and9, and cortistatin (2 nmol per mouse) was injected i.p. 12 h after thefirst TNBS infusion. Animals were monitored daily for appearanceof diarrhea, loss of body weight, and survival. Some animals werekilled at the peak of the disease (day 3), blood samples werecollected by cardiac puncture, and a segment of the colon wasexcised for macroscopic damage evaluation and weighed. Tissuesegments were immediately frozen in liquid nitrogen for histolog-ical and immunohistological studies, protein extraction and cyto-kine determination, MPO activity measurement, total RNA ex-traction, and LPMC isolation (see Supporting Methods, which ispublished as supporting information on the PNAS web site).

Macroscopic and Microscopic Damage Evaluation. Colons were ex-amined under a dissecting microscope and graded for macroscopiclesions on a scale from 0 to 10 based on criteria reflectinginflammation, such as hyperemia, thickening of the bowel, and theextent of ulceration (31). Scores for stool consistency and rectalbleeding were assessed according to the procedures published inref. 32. For histopathologic analysis, colons were fixed, sectioned,and stained with hematoxylin�eosin, and inflammation was gradedfrom 0 to 4 as follows in a blinded fashion: 0, no signs ofinflammation; 1, low leukocyte infiltration; 2, moderate leukocyteinfiltration; 3, high leukocyte infiltration, moderate fibrosis, highvascular density, thickening of the colon wall, moderate goblet cellloss, and focal loss of crypts; 4, transmural infiltrations, massive lossof goblet cell, extensive fibrosis, and diffuse loss of crypts. Forimmunohistological analysis, sections were stained with 5 �g�mlFITC- and phycoerythrin-labeled anti-CD4, anti-CD11b, or anti-TNF-� mAbs and examined by fluorescent microscopy.

Macrophage Cultures. Peritoneal exudate cells obtained fromBALB�c mice were incubated in complete medium at 106 cellsper milliliter, and, after 2 h at 37°C, nonadherent cells wereremoved. Macrophage monolayers were incubated with com-plete medium in the absence (unstimulated) or presence of 1�g�ml LPS from Escherichia coli serotype 055:B5 (Sigma) anddifferent concentrations of cortistatin. Supernatants were col-lected at different times, and cytokine levels were determined.The amount of NO formed was estimated from the accumulationof the stable NO metabolite nitrite by the Griess assay.

Data Analysis. All values are expressed as means � SD. Thedifferences between groups were analyzed by Mann–Whitney Utest and, if appropriate, by Kruskal–Wallis ANOVA test. Sur-vival curves were analyzed by the Kaplan–Meyer log-rank test.Changes in body weight were compared by use of the Wilcoxonmatched-pair signed-rank test.

This work was supported by Spanish Ministry of Health Grant PI04�0674, a Ramon Areces Foundation grant (to M.D.), and NationalInstitutes of Health Grants AI52306 and AI47325 (to D.G.).

1. Fiocchi, C. (1998) Gastroenterology 115, 182–205.2. Hanauer, S. B. & Present, D. H. (2003) Rev. Gastroenterol. Disord. 3, 81–92.3. Elson, C. O., Sartor, R. B., Tennyson, G. S. & Riddell, R. H. (1995) Gastroenterology

109, 1344–1367.4. Strober, W., Fuss, I. J. & Blumberg, R. S. (2002) Annu. Rev. Immunol. 20,

495–549.5. de Lecea, L. Criado, J. R., Prospero-Garcia, O., Gautvik, K. M., Schweitzer, P.,

Danielson, P. E., Dunlop, C. L., Siggins, G. R., Henriksen, S. J. & Sutcliffe, J. G.(1996) Nature 381, 242–245.

6. Spier, A. D. & de Lecea, L. (2000) Brain Res. Rev. 33, 228–241.7. Dalm, V. A. (2003) Am. J. Physiol. 285, E344–E353.8. Dalm, V. A. (2003) J. Clin. Endocrinol. Metab. 88, 270–276.9. Krantic, S. (2000) Peptides 21, 1941–1964.

10. Deghenghi, R., Papotti, M., Ghigo, E. and Muccioli, G. (2001) J. Endocrinol. Invest.24, RC1–RC3.

11. Dixit, V. D., Schaffer, E. M., Pyle, R. S., Collins, G. D., Sakthivel, S. K., Palaniappan,R., Lillard, J. W., Jr., & Taub, D. D. (2004) J. Clin. Invest. 114, 57–66.

12. Granado, M., Priego, T., Martin, A. I. Villanua, M. A. & Lopez-Calderon, A. (2005)Am. J. Physiol. 288, E486–E492.

13. Neurath, M. F., Fuss, I., Kelsall, B. L., Stuber, E. & Strober, W. (1995) J. Exp. Med.182, 1281–1290.

14. Neurath, M. F., Finotto, S. & Glimcher, L. H. (2002) Nat. Med. 8, 567–573.15. McCormack, G., Moriaty, D., O’Donoghue, D. P., McCormick, P. A., Sheahan, D.

& Baird, A. W. (2001) Inflammation Res. 50, 491–495.16. Niederau, C., Backmerhoff, F., Schumacher, B. & Niederau, C. (1997) Hepatogas-

troenterology 44, 90–107.17. Uhlar, C. M. & Whitehead, A. S. (1999) Eur. J. Biochem. 265, 501–523.18. Thompson, C. & Powrie, F. (2004) Curr. Opin. Pharmacol. 4, 408–414.

19. Singh, B., Read, S., Asseman, C., Malmstrom, V., Mottet, C., Stephens, L. A.,Stepankova, R., Tlaskalova, H. & Powrie, F. (2001) Immunol. Rev. 182, 190–200.

20. Groux, H., O’Garra, A., Bigler, M., Rouleau, M., Antonenko, S., de Vries, J. E. &Roncarolo, M. G. (1997) Nature 389, 737–742.

21. Asseman, C., Mauze, S., Leach, M. W., Coffman, R. L. & Powrie, F. (1999) J. Exp.Med. 190, 995–1004.

22. Iellem, A., Mariani, M., Lang, R., Recalde, H., Panina-Bordignon, P., Sinigaglia, F.& D’Ambrosio, D. (2001) J. Exp. Med. 194, 847–853.

23. Wang, L., Wells, A. D., Dorf, M. E., Ozkaynak, E. & Hancock, W. W. (2005) J. Exp.Med. 201, 1037–1044.

24. Stuber, E., Strober, W. & Neurath, M. (1996) J. Exp. Med. 183, 693–698.25. Amstrong, A. M., Foulkes, R., Jennings, G., Gannon, D., Kirk, S. J. & Gardiner, K. R.

(2001) Br. J. Surg. 88, 235–240.26. Fiorucci, S., Mencarelli, A., Palazzetti, B., Sprague, A. G., Distrutti, E., Morelli, A.,

Novobrantseva, T. I., Cirino, G., Koteliansky, V. E. & de Fougerolles, A. R. (2002)Immunity 17, 769–780.

27. Atreya, R., Mudter, J., Finotto, S., Mullberg, J., Jostock, T., Wirtz, S., Schutz, M.,Bartsch, B., Holtmann, M., Becker, C., et al. (2000) Nat. Med. 6, 183–588.

28. Ramirez, J. L., Mouchantaf, R., Kumar, U., Otero-Corchon, V., Rubinstein, M., Low,M. J. & Patel, Y. C. (2002) Mol. Endocrinol. 16, 1951–1963.

29. Gonzalez-Rey, E., Chorny, A. & Delgado, M. Gastroenterology, in press.30. Criado, J. R., Li, H., Jiang, X., Spina M., Huitron-Resendiz, S., Liapakis, G., Calbet,

M., Siehler, S., Henriksen, S. J., Koob, G., et al. (1999) J. Neurosci. Res. 56, 611–619.31. Fiorucci, S., Mencarelli, A., Palazzetti, B., Distrutti, E., Vergnolle, N., Hollenberg,

M. D., Wallace, J. L., Morelli, A. & Cirino, G. (2001) Proc. Natl. Acad. Sci. USA 98,13936–13941.

32. Kihara, N., De la Fuente, S. G., Fujino, K., Takahashi, T., Pappas, T. N. & Mantyh,C. R. (2003) Gut 52, 713–719.

Gonzalez-Rey et al. PNAS � March 14, 2006 � vol. 103 � no. 11 � 4233

MED

ICA

LSC

IEN

CES