Final Progress Report

Proposal No. IBD-0110
Principal Investigator:   Scott E. Plevy, M.D.
Applicant Organization:  University of Pittsburgh (Pennsylvania, U.S.A.)
Project Title:  Clostridium perfringens as a novel therapeutic vehicle in inflammatory bowel disease
Period of Award:  August 1, 2004 - January 31, 2006

A. Summary of project aims:

The goal of this proposal is to create a novel biologic therapeutic targeted to Crohn’s disease using a normal constituent of the enteric bacterial flora, Clostridium perfringens. C. perfringens is best known as an important cause of self-limited food borne disease. In humans, vegetative C. perfringens sporulate in the small intestine. In the case of pathogenic organisms, spores lyse in the terminal ileum, releasing large amounts of C. perfringens enterotoxin (cpe). We have genetically engineered C. perfringens, replacing the plasmid-based cpe gene with HIV and SIV peptides. These peptides are delivered directly to the distal small bowel in rodents. In this proposal, we will genetically engineer C. perfringens to produce the anti-inflammatory cytokine IL-10. C. perfringens poses several significant advantages over other bacteriotherapeutic approaches: 1) molecules can be delivered specifically to the distal small bowel; 2) these molecules are produced at extremely high local concentrations and; 3) no significant immune responses occur against C. perfringens. If successful, these experiments will lead to non-human primate and human studies to evaluate C. perfringens as a novel therapeutic modality in Crohn’s disease.

Specific Aim 1. Construction of a recombinant C. perfringens that expresses bioactive IL-10.  We demonstrate that C. perfringens can be engineered to express HIV and SIV proteins.  We have subcloned the murine IL-10 gene into the C. perfringens enterotoxin (cpe) plasmid pJRC200, replacing the coding sequence for the cpe gene. We will measure IL-10 in bacterial lysates by Western blot and ELISA. We will determine whether bacterially produced IL-10 is bioactive by an IL-12 inhibition assay.

Specific Aim 2: Recombinant C. perfringens delivers murine IL-10 to the terminal ileum of mice. Delivery of spore contents to the murine intestine will be assessed using a recombinant strain of C. perfringens that expresses EGFP. Then, the IL-10 expressing bacteria will be administered to IL-10 deficient mice. IL-10 expression in regional mucosal tissue will be determined by ELISA.

Specific Aim 3. Therapeutic efficacy of recombinant C. perfringens expressing IL-10 in IL-10 deficient and SAMP1/YitFc mice.  We will assess recombinant C. perfringens expressing IL-10 as a potential therapeutic modality in two mouse models of IBD. We will also perform preliminary mechanistic studies to understand immunologic effects of this bacterial vector. Experiments will be performed to evaluate prevention of colitis, as well as treatment of established colitis by recombinant C. perfringens.

B. Accomplishments toward meeting those aims/ List of significant results:

Aim 1. Recombinant Clostridium perfringens (CP) that express interleukin-10 (IL-10) (IL-10 CP) has been successfully engineered. Sporulating IL-10 CP produce immunoreactive IL-10 detectable by ELISA and Western blot. In vitro and in vivo studies have been complicated by relative insolubility of the expressed protein leading to the conclusion that CP may in fact be most amenable to the delivery of short anti-inflammatory peptides. One such molecule, an IkB kinase inhibitor, NEMO binding domain peptide, linked to a cationic peptide transduction domain (PTD-NBD) has become a focus of recent studies (see aim 2).

To develop safer C. perfringens vectors for delivery of proteins to the mammalian GI tract, our collaborators have developed a genetically engineered C. perfringens with a deletion of the alpha toxin gene (Chen Y, et al. Construction of an Alpha Toxin Gene Knockout Mutant of Clostridium Perfringens Type A by Use of a Mobile Group II Intron. Appl Env Microbiol 2005; 11: 7542-7547). Additionally, studies have begun to assess the ability of nonpathogenic strains of clostridia to express plasmid-based genetically engineered proteins.

To explore other novel means of direct mucosal delivery of the IL-10 gene to mouse models of IBD, we have begun a collaboration under the auspices of this project with Dr. Jude Samulsky at the University of North Carolina at Chapel Hill. Dr. Samulski has pioneered the technology to develop tissue specific adeno-associated virus (AAV) serotypes. The goal of this collaboration is to create a novel biologic gene therapeutic targeted to inflammatory bowel disease (IBD) using an AAV specifically selected to transduce cells of the gastrointestinal (GI) tract. Although viral gene delivery approaches to treat IBD have been explored in animal models since the 1990’s, prevailing strategies have significant limitations, including low transduction efficiency of the intestinal mucosa and a broad host range of viral infectivity which limits the ability to specifically target GI immune responses. Therefore, we will screen and validate recombinant AAV hybrid serotypes that selectively and efficiently transduce normal and inflamed GI mucosa. Although focused on IBD, this project, if successful, may have global implications for gastrointestinal biology: Mutant AAV may be generated that specifically transduce different locations in the GI tract (ie – ileum versus colon), different cell types (enterocyte subpopulations, intestinal epithelial stem cells, mucosal dendritic cells), or have a predilection for inflamed versus in tact mucosa. Therefore, AAV vectors generated in this pilot project could have implications not only for IBD, but for mucosal vaccination strategies, colon cancer, intestinal epithelial growth and development, and intestinal failure, to name a few.  As a proof of concept, the therapeutic efficacy of recombinant AAV expressing IL-10 will be tested in IL-10 deficient (-/-) mice. This work has been recently funded by a pilot and feasibility grant through the NIH supported Center for Gastrointestinal Biology and Disease at UNC-Chapel Hill (P30 DK34987).

Aim 2. We demonstrated delivery of fluorescent label and an SIV p27 protein expressed in CP to mouse ileum and jejunum. However, no IL-10 was detectable with the IL-10 CP vector when delivered by oral gavage to wild type and IL-10-/- mice, and no therapeutic efficacy was demonstrated in a pilot experiment of 5 IL-10-/- mice. We speculate that IL-10 formed an insoluble precipitate in vitro and in vivo thus interfering with tissue accessibility and biologic activity. Therefore, further efforts were shifted to identify small immunologically active peptides that may have efficacy in IBD models and may be expressed in CP without the noted solubility concerns. To this end, we first determined efficacy and mechanism of a novel cell permeable peptide NF-kB inhibitor (PTD-NBD) in murine colitis. The NF-κB family of transcription factors is a central regulator of chronic inflammation. The phosphorylation of IκB proteins by a specific IκB kinase (IKK) is a key step in NF-κB activation.  IKK is a complex of two catalytic subunits, IKKα and IKKβ, and a regulatory subunit, “NF-κB essential modulator” (NEMO). An N-terminal region of NEMO associates with a hexapeptide sequence within the C-terminus of both IKKα and IKKβ, named the NEMO binding domain (NBD). A cell permeable peptide containing a cationic protein transduction doman (PTD) coupled to NBD (PTD-NBD) disrupts the association of NEMO with IKKs in vitro and blocks NF-κB activation in vivo. As a therapeutic modality, PTD-NBD has the theoretic advantage of selectively inhibiting activated NF-kB, and the potential for local and systemic administration. To determine transduction efficiency, murine bone-marrow (BM) derived macrophages were incubated with two PTDs (8K and HIV TAT).  NF-kB activity was evaluated using an NF-kB luciferase reporter plasmid transfected into RAW264.7 cells.  IL-10 deficient (-/-) mice at 10 weeks of age were injected IP with PTD-NBD (1, 5, or 10 mg/kg and vehicle) daily for two weeks. Colitis severity was scored by a pathologist blinded to treatment group. TNF and IL-12 p40 production were determined by ELISA in supernatants from LPS (100 ng/ml) plus IFN-g (10 ng/ml) activated splenocytes and intestinal explant cultures. The PTD peptides transduce BM-derived macrophages with close to100% efficiency.  In RAW264.7 cells, PTD-NBD dose-dependently inhibits LPS-induced NF-kB activity. In vivo, IL-10-/- mice treated IP with PTD-NBD (n=4) demonstrate amelioration of colitis (colitis score 1.6±0.4 vs. 5.4±0.9) compared to IL-10-/- mice treated with vehicle (n=4). Intestinal explants of treated mice express less TNF (70.5±43.4 pg/ml vs. 219±34.3 pg/ml) and IL-12 p40 than controls.  Decreased LPS plus IFN-γ activated IL-12 p40 is demonstrated in splenocytes cultured ex vivo from PTD-NBD treated IL-10-/- mice. In conclusion, a cell permeable peptide inhibitor of IKK, PTD-NBD, ameliorates colitis in IL-10-/- mice.  Mechanistically, mucosal explants and splenocytes cultured ex vivo for 24 hours demonstrate decreased cytokine expression, suggesting that inhibition is not related to the continued presence of the short-lived PTD-NBD peptide.  These results demonstrate that inhibitors of IKK, in particular a PTD-NBD peptide, could be therapeutic in the treatment of colitis.  Further work will develop this peptide as a local therapy, with and without the CP vector for delivery. This work was presented at a research symposium at DDW 2006 with the BMRP acknowledged as the supporting organization. Furthermore, this work was recently funded by the NIH through the STTR (R41) mechanism in collaboration with the small biotechnology company that owns patents on the NBD peptide, Theralogics Inc.

Aim 3. Experiments not performed due to technical difficulties as stated above encountered in aim 2 and changes in direction of project.

C. Lay summary of the progress report:

The human IBDs, CD and UC, affect over one million Americans. There are many pressing, unmet medical needs to develop safer and more effective treatments for these lifelong, debilitating illnesses.  The bacteria that normally reside in all of our intestines are known to be important in the initiation and perpetuation of inflammatory bowel disease (IBD).  In fact, most researchers believe that IBD may be caused by an abnormal inflammatory reaction against bacteria that in most people constantly exist inside the intestine and do not cause any harm.  Therefore, an exciting area to develop new treatments for IBD is through the characterization or creation of bacteria that may suppress inflammation in the intestine.  One approach has been to genetically engineer harmless intestinal bacteria to express anti-inflammatory proteins.  In this proposal, we genetically engineered a normal resident of the intestinal bacterial population, Clostridium perfringens (C. perfringens) to express the anti-inflammatory protein interleukin-10 (IL-10).  C. perfringens is a potentially novel way to treat Crohn’s disease (CD), because it can deliver this anti-inflammatory protein directly to the terminal ileum, which is the area of the intestine most frequently affected in CD.  In addition, by delivering this protein directly to the intestine rather than through the bloodstream, as is the case with many new therapies, this may be a safe treatment, without side effects related to suppressing the immune system throughout the entire body.  Although the bacteria was successfully created, the expressed anti-inflammatory protein, due to its large size was not soluble, and therefore did not work to treat murine IBD in one mouse model. However, our search for other anti-inflammatory peptides that are smaller and therefore more amenable to delivery in this bacteria led to exciting results. We had tested a new treatment in mouse models of IBD that can potentially be administered by mouth and works by turning off a protein, NF-κB, which is a “master switch” that turns on inflammation in the intestine. We have demonstrated that this short peptide ameliorates chronic IBD in one mouse model and we have obtained funding from the NIH to continue these studies. Additionally, we have received a small pilot and feasibility grant from the NIH to look at novel viruses as a way to deliver IL-10 to the intestine in IBD. The viral approach may have advantages over bacteria in that a virus will infect a cell and then the cell itself will express IL-10. Finally, we are pursuing further studies with our collaborators at the University of Pittsburgh who have now developed new strains of bacteria that may be safer for delivery of proteins to the intestine.