Gene Therapy (2008) 15, 1200–1209 & 2008 Macmillan Publishers Limited All rights reserved 0969-7128/08 $30.00 www.nature.com/gt ORIGINAL ARTICLE Oral IL-10 gene delivery in a microsphere-based formulation for local transfection and therapeutic efficacy in inflammatory bowel disease

MD Bhavsar and MM Amiji Department of Pharmaceutical Sciences, School of Pharmacy, Northeastern University, Boston, MA, USA

The objective of this study was to examine the potential of IL-1a, IL-1b and IL-12, as well as certain . The oral -10 (IL-10) gene therapy for the treatment therapeutic benefits of transfected IL-10 were further of inflammatory bowel disease (IBD). Nanoparticles-in-micro- demonstrated by an increase in body weight, favorable sphere oral system (NiMOS) was formulated with murine clinical activity score, restoration in colon length and weight, IL-10-expressing plasmid DNA in type-B gelatin nanoparticles, and suppression of inflammatory response as assessed by which were further encapsulated in poly(epsilon-caprolac- tissue histological analysis and myeloperoxidase activity. tone) microsphere matrix. Upon oral administration in an The results of this study provide highly encouraging evidence acute colitis model, IL-10 expression in the large intestine of oral gene delivery and transfection and potential utility in was measured by quantitative real-time PCR and ELISA. IBD therapy. The locally expressed IL-10 was able to suppress the Gene Therapy (2008) 15, 1200–1209; doi:10.1038/gt.2008.67; levels of proinflammatory , such as IFN-g, TNF-a, published online 17 April 2008

Keywords: nanoparticles-in-microsphere oral system; non-viral vector; oral gene therapy; IL-10 transfection; inflammatory bowel disease

Introduction IL-10 in suppression of inflammation in a trinitrobenze- nesulfonic acid (TNBS)-induced acute colitis animal Recent evidence supports the view that inflammatory model.6,9,10 Systemic administration of recombinant IL-10 bowel disease (IBD) is caused by genetically determined has been investigated in several small clinical trials with dysregulation of the mucosal immune response to limited success.11 Nakase et al.12 administered IL-10 luminal antigens derived from the intestinal micro- protein in gelatin microspheres to IL-10/(À/À) knockout flora.1–4 Activated T mediating inflamma- mice and showed some efficacy. Genetically engineered tion have been identified for the two phenotypes of IBD. Lactococcous lactis have also been developed for local Activated Th1 lymphocytes incline more toward the delivery of IL-10 protein in the intestinal lumen upon proinflammatory cytokines, which include tumor necro- oral administration.13 Local delivery of plasmid DNA sis factor-a (TNF-a), -g (IFN-g), interleukin-1 encoding for IL-10 has also been reported and much of (IL-1) and IL-12. The activated lymphocytes in ulcerative the effort has been made with the use of viral vectors, colitis exhibit increased levels of IL-5, IL-6 and trans- including replication-deficient adenoviruses.6,9,10,14 forming -b (TGF-b).5 Although viral vectors have been used successfully in Interleukin-10 plays a very important role in the preclinical studies, the potential for toxicity of viruses, immunological balance of the mucosal immune system. especially upon chronic administration, is a major It acts by inhibiting the production of proinflammatory concern when it comes to translation of this experimental cytokines from activated and, thus inhibits strategy into clinical reality.15,16 We and others17–19 have, antigen presentation and effector functions. IL-10 is therefore, focused on development of safe and effective also known to inhibit chemokines, including nonviral gene delivery systems. Among several exam- inflammatory protein-1a (MIP-1a) and eotaxin, along ples of nonviral vectors, polymeric gene delivery systems with inhibition in the expression of inducible nitric oxide have received significant attention.17–23 synthase and cyclooxygenase-2 in macrophages.6–8 Sev- In previous studies, we have described the formula- eral researchers have demonstrated the usefulness of tion development, characterization, optimization and preliminary in vivo transfection of reporter plasmids Correspondence: Professor MM Amiji, Department of Pharmaceu- upon oral administration of nanoparticles-in-micro- tical Sciences, School of Pharmacy, Northeastern University, Room sphere oral system (NiMOS) formulations in animal 110 Mugar Life Sciences Building, 360 Huntington Avenue, Boston, models.24,25 NiMOS is a biodegradable and biocompa- MA 02115-5005, USA. E-mail: [email protected] tible polymer-based microparticle system capable of Received 4 December 2007; revised 13 February 2008; accepted 19 plasmid DNA delivery and local transfection upon oral February 2008; published online 17 April 2008 administration. This delivery system is specifically Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1201 engineered for oral gene therapy, where the DNA- gelatin nanoparticles and in NiMOS. The plasmid DNA encapsulated gelatin nanoparticles can be released in encapsulation efficiency in gelatin nanoparticles was the intestine from poly(epsilon-caprolactone) matrix due greater than 93% at 0.5% (w/w) concentration. When to lipase-induced degradation. Once released, the nano- DNA-containing gelatin nanoparticles were further particles can be endocytosed by the cells of the intestinal encapsulated in NiMOS, the final DNA loading lumen to afford efficient transfection and transgene efficiency was found to be approximately 46% as expression. The solvent displacement technique em- determined by DNA extraction using dichloromethane ployed to formulate type B gelatin nanoparticles and protease digestion of gelatin nanoparticles to release encapsulating plasmid DNA has been previously opti- encapsulated DNA. mized and reported by our laboratory.26,27 To illustrate the utility of type B gelatin nanoparticles for systemic Local IL-10 transfection upon oral administration gene therapy, we have administered reporter and Assessment of plasmid composition and purity. The therapeutic (sFlt-1 expressing) plasmid DNA in poly pORF5-mIL-10 plasmid, purchased from InvivoGen (San (ethylene glycol)-modified nanoparticles for efficient and Diego, CA, USA) in transformed Escherichia coli, was sustained transgene expression and antiangiogenic ther- further amplified and purified from culture using the apeutic efficacy in solid tumor models.19,28–30 Plasmid Giga purification kit (Qiagen, Valencia, CA, To further evaluate the potential of NiMOS in oral USA). The plasmid composition and purity, as assessed therapeutic gene therapy, in this study, we show local by agarose gel electrophoresis, is shown in Figure 2a. transfection of IL-10 and therapeutic efficacy in TNBS- Lane 1 of the agarose gel is 2–10 kbp supercoiled ladder, induced acute colitis model developed in Balb/c mice. lane 2 is intact plasmid DNA (that is, pORF5-mIL-10), lane 3 is 1–10 kbp linear plasmid ladder, lane 4 is 100 kbp ladder and lane 5 is pORF5-mIL-10 plasmid DNA treated Results with endonucleases NcoI and NheI. The size of pORF5- IL-10-expressing plasmid DNA-encapsulated NiMOS mIL-10 is 3.7 kbp and the band seen in Figure 2a in lane 2 is between 3 and 5 kbp, suggesting that the plasmid Figure 1 shows the cross-sectional schematic of NiMOS DNA is pORF5-mIL-10. To further confirm the purity of and the scanning electron microscopy (SEM) images the isolated plasmid DNA, it was further incubated with obtained for gelatin nanoparticles. The SEM image restriction enzymes NcoI and NheI, which resulted in the confirmed the formation of small spherical type B gelatin formation of two fragment of plasmid DNA that were nanoparticles with encapsulated plasmid DNA expres- about 3.1 kbp and 570 bp. Lane 5 in Figure 2a shows two sing murine IL-10. The average particle diameter of the bands representing 3.1 kbp and the other smaller band resulting microspheres, as determined by Coulter analy- was between 500 and 600 bp, thus further confirming the sis using Multisizer 3 and by SEM, was in the range of composition of pORF5-mIL-10. 2–5 mm. Figure 1 also shows an inset of a NiMOS showing the spherical nature of NiMOS and smooth surface resulting from encapsulation of gelatin nanoparticles Local transfection assessment by quantitative within the poly(epsilon-caprolactone) matrix. PicoGreen RT-PCR. Quantitative real time (RT)-PCR analysis in dsDNA fluorescent assay was used to quantitate the the large intestine tissue was carried out to determine the capacity and efficiency of plasmid DNA encapsulation in IL-10 transgene expression at the mRNA level upon oral

Plasmid DNA (i.e., pORF5- mIL-10)-Containing Microsphere matrix composed Gelatin Nanoparticles of Poly(epsilon-caprolactone)

Particle Particle ~ 200 nm Cross Sectional View 2 -5 µm Size of NiMOS Size

DNA DNA Loading 93-98% Loading ~ 46 % Efficiency Efficiency

Figure 1 Schematic illustration showing the cross-sectional view of nanoparticles-in-microsphere oral system (NiMOS). On the left is the scanning electron microscopy (SEM) image of gelatin nanoparticles, which are less than 200 nm in diameter, and can physically encapsulate plasmid DNA at a loading efficiency of >93%. On the right is the SEM image of 2–5 mm NiMOS with the overall DNA encapsulation efficiency of B46%.

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1202 1 234 5

~3.13kbp

~3.7kbp ~ 4 Kbp ~ 3 Kbp 600 bp 570 bp

m IL-10 12500 m L32

10000

7500

5000

2500 mIL-10 (relative amounts) 0 Naïve Colitis Colitis Colitis No Colitis No Trtmnt Gel nps + pIL-10 NiMOS + pIL-10

200 ∗∗

150

100

50

IL-10 (pg/mg of total protein) 0 10 Naive

NiMOS + pIL-10 ** p<0.05 Colitis- No Trtmnt Gel nps + pIL- Figure 2 Murine interleukin (IL)-10 transgene expression using nanoparticles-in-microsphere oral system (NiMOS) in an acute colitis model. Qualitative agarose gel electrophoresis performed on pORF5-mIL-10 (B3.7 kbp) isolated from cultured Escherichia coli to show the purity and composition of the plasmid (a). Quantitative real time–PCR analysis performed on the large intestinal sections obtained from different control and treatment groups showing high mRNA transcript for IL-10 upon oral administration of the plasmid with NiMOS (b). Additionally, the levels of expressed IL-10 were also determined by ELISA (c). Each conscious animal received a 100 mg oral dose of pORF5-mIL-10 in gelatin nanoparticles or NiMOS. Mean±s.d. (n ¼ 4).

administration of the plasmid DNA in NiMOS. Figure 2b protein levels in the large intestine of the control and shows the results of quantitative RT-PCR analysis to treated animals as determined by ELISA. The group examine murine IL-10 mRNA transcript levels in the receiving gelatin nanoparticles encapsulating pORF5- large intestine of control and transfected animals. IL-10- mIL-10 as treatment showed similar levels of colonic expressing plasmid DNA when encapsulated in NiMOS IL-10 as seen for the saline group. Figure 2c also shows showed highest levels of mRNA expression. the results for colitis group administered with NiMOS formulation encapsulating plasmid DNA pORF5- mIL-10. We observed a significant difference (Po0.005) Measurement of transfected IL-10 levels by ELISA. in colonic IL-10 levels (B180 pg of IL-10 per mg of total Following oral administration of pORF5-mIL-10 at tissue protein) as for colitis group administered with 100 mg dose per mice in gelatin nanoparticles and in plasmid DNA in NiMOS, as compared (25 pg of IL-10 NiMOS to colitis-induced mice, the levels of murine per mg of total tissue protein) with the group receiving IL-10 expression was determined in the inflamed colon 4 saline and another group receiving gelatin nanoparticles days post-administration. Figure 2c shows murine IL-10 as treatment.

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1203 Evaluation of therapeutic efficacy of IL-10 in acute reported in Figure 3. Figures 3a and b shows the results colitis model obtained for IL-1a and IL-1b. During colitis the levels of Therapeutic efficacy studies were performed on various both these proinflammatory cytokines were raised to above À1 treatment groups. In this study, we looked at many 150 pg mg of total protein. A significant decrease to about À1 different parameters ranging from expression of inflam- 20 and 10 pg mg of total protein for IL-1a and IL-1b, matory cytokines and chemokines, body weight changes, respectively, was observed for the colitis group receiving macroscopic clinical activity score, as well as histological NiMOS treatment. The colitis group receiving plasmid analysis of large intestine and tissue myeloperoxidase DNA gelatin nanoparticles did not show any significant (MPO) activity. decrease in the levels of both of these cytokines, where the values remained above 100 pg mgÀ1 for both. We also observed a decrease in the levels of other proinflammatory Proinflammatory and profiling. The cytokines, such as IFN-g,TNF-a and IL-12 in the NiMOS- expression profile of proinflammatory cytokines and treated group, as seen in Figures 3c–e, respectively. chemokines upon successful murine IL-10 transfection are In contrast, the group that received DNA encapsulated in

IL-1α IL-1β 200 200

150 150

100 100 protein) protein) (pg/mg of total (pg/mg of total

β β α α 50 50 IL-1 IL-1

0 0

No Colitis

Naive- No Colitis Naive- Colitis- No Trtmnt Colitis- No Trtmnt

Colitis- Gel nps + pIL-10Colitis- NiMOS + pIL-10 Colitis- Gel nps + pIL-10Colitis- NiMOS + pIL-10

IFN-γ TNF-α 150 200

150 100

100 protein (pg/mg of total protein)

(pg/mg of total γ γ 50 α 50 IFN- TNF- 0 0

IL-10

No Colitis l nps + p Naive- No Colitis Naive- Colitis- No Trtmnt Colitis- No Trtmnt

Colitis- Gel nps + pIL-10Colitis- NiMOS + pIL-10 Colitis- Ge Colitis- NiMOS + pIL-10

IL-12 25 2000 20 1750 1500 15 1250 MCP-1 MIP-1α 1000 RANTES

protein) 10 750

of total protein) 500

5 Chemokine (pg/mg IL-12 (pg/mg of total 250 0 0

IL-10

Naive- No Colitis Naive- No Colitis Colitis- No Trtmnt Colitis- No Trtmnt

Colitis- Gel nps + pIL-10 Colitis- NiMOS + pIL-10 Colitis- Gel nps + p Colitis- NiMOS + pIL-10 Figure 3 Reduction in the levels of proinflammatory cytokines and chemokines upon delivery of murine interleukin (IL)-10-expressing plasmid DNA administered in nanoparticles-in-microsphere oral system (NiMOS). The measured proinflammatory factors included IL-1a (a), IL-1b (b), interferon (IFN)-g (c), IL-12 (d) and (TNF)-a (e) and proinflammatory chemokines (monocyte chemotactic protein (MCP)-1, monocyte inflammatory protein (MIP)-1a and Regulated on Activation, Normal T Expressed and Secreted (RANTES)) (f). Each conscious animal received a 100 mg oral dose of pORF5-mIL-10 in gelatin nanoparticles or NiMOS. Mean±s.d. (n ¼ 4).

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1204 gelatin nanoparticles did not show any decrease in the of MCP-1 and MIP-1a as determined by ELISA were levels of these proinflammatory cytokines. This is due to about 80 and 9 pg mgÀ1 of total protein, respectively. the fact that gelatin nanoparticles were not able to transfect These levels were increased to 1600 and 850 pg mgÀ1 for thecellsatthesiteofinflammationcausingnochangeinthe MCP-1 and MIP-1a, respectively, for the colitis group levels of IFN-g and subsequently in levels of TNF-a and receiving saline and gelatin nanoparticles. Upon admin- IL-12. Upon induction of colitis the values of TNF-a and istration of NiMOS formulation, these levels were IL-12 were raised to about 180 and 20 pg mgÀ1 of total brought back to 550 and 50 pg mgÀ1 of total protein for protein, respetively. Whereas the normal values for these MCP-1 and MIP-1a, respectively. We did not observe any cytokinesasdeterminedbyELISAwerefoundtobe change in the levels of Regulated on Activation, Normal 8pgmgÀ1 of total protein for IL-12 and 15 pg mgÀ1 of total T Expressed and Secreted (RANTES) in the colitis or protein for TNF-a. Similar levels were observed for colitis upon treatment with IL-10-expressing plasmid DNA in group receiving with gelatin nanoparticles as treatment. any of the formulations. We observed a significant decrease in the levels of these cytokines for colitis group receiving NiMOS to about Body weight changes and clinical activity score. To 6.5 pg mgÀ1 for IL-12 and 14 pg mgÀ1 for TNF-a. further confirm the therapeutic benefits that can be Figure 3f shows the expression levels of different achieved with IL-10 oral gene therapy, change in body proinflammatory chemokines. We observed a decrease in weight of colitis-induced mice after different treatments the levels of two chemokines, monocyte chemotactic was examined. Figures 4a and b shows the percent protein 1 (MCP-1) and MIP-1a, upon increase in the IL-10 change in body weight of mice upon acute colitis levels at the site of inflammation, Figure 3f. The baseline induction and after administration of various treatments.

Naive - No Colitis Naive- No Colitis Colitis- No Trtmnt Colitis- No Trtmnt Colitis- NiMOS + pIL-10 110 Colitis- Gel nps + pIL-10 Colitis- Gel nps + pIL-10 Colitis- NiMOS + pIL-10 12.5 ∗∗

100 ∗∗ 10.0

∗∗ 90 ∗∗ 7.5

5.0 80 % Original Body weight

Colon Length (cm) 2.5 70

012345678910 0.0 Time (days)

∗∗ 4 ∗∗ 0.3

3

2 0.2

1

Clinical Activity Score 0.1 Colon Weight (g) 0 012345678910 ** p< 0.05 Time (days) 0.0 Figure 4 Changes in body weight, clinical activity score, and the lengths and weights of colonic tissue upon oral administration of murine interleukin (IL)-10-expressing plasmid DNA in nanoparticles-in-microsphere oral system (NiMOS). The body weight change was used as a marker of therapeutic efficacy achieved with locally expressed IL-10 over the course of 8 days (a). The clinical activity scores in control and treatment animals as measured using an aggregate of body weight changes, rectal bleeding and stool consistency (b). Additionally, the colon length (c) and colon weights (d) were also measured. Each conscious animal received a 100 mg oral dose of pORF5-mIL-10 in gelatin nanoparticles or NiMOS. Mean±s.d. (n ¼ 4).

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1205 Figure 4a shows change in body weight over the whole Lastly, the colon length and colon weight were the period of study, and Figure 4b shows the change in body other two parameters that were observed and are weight after day 4, which corresponds to administration reported in Figures 4d and e, respectively. We observed of the second dose of TNBS. NiMOS formulation showed a significant decrease in colon length and colon weight a considerable increase in transfection at the site of upon colitis induction, which was also observed for inflammation, which resulted in significant gain in body gelatin nanoparticle treatment group. For the NiMOS weight of mice starting from 2 days post-administration treatment group, we observed an increase in the length of the formulation, and complete restoration of body and weight of the colon back to normal, which was a weight was observed 4 days post-administration. The significant gain when compared to colitis and gelatin group receiving saline and plasmid DNA in gelatin nanoparticles group. The results of this study showed a nanoparticles, on the other hand, showed significant loss reduction in colon length during colitis to about 6 cm in body weight upon colitis induction and formulation and the weight was reduced to about 200 mg. The colitis administration. The loss in body weight observed in case group treated with NiMOS showed a significant of DNA-encapsulated gelatin nanoparticles was as high (Po0.01) increase in both colon length (9 cm) and weight as 25% of the original body weight after 4 days post- back (B300 mg) to baseline level. administration. Figure 4c shows the results of clinical activity scored Tissue histological evaluation and MPO activity. To according to the criteria mentioned before. The clinical further demonstrate the efficacy of locally expressed activity score of the disease for NiMOS group returned IL-10 in suppressing inflammation in an acute colitis back to normal, less than 1 on day 4 post-administration model, the large intestinal tissue cryosections were of the formulation, whereas high activity levels, greater stained with hematoxylin and eosin and observed with than 3 similar to colitis group were observed for group a bright-field microscope. Figure 5a shows the changes administered with gelatin nanoparticles. occurring at the cellular level upon colitis induction and

Colitis - Gel nps/pIL-10 Colitis - NiMOS/pIL-10 Naïve - No Colitis Colitis - No Trtmnt Trtmnt Trtmnt

Higher magnification

40 ∗∗

30

20 ** p<0.05

10 MPO activity/g of tissue

∗∗

0 Naïve - No Colitis Colitis - No Trtmnt Colitis - Gel Colitis - nps/pIL-10 Trtmnt NiMOS/pIL-10 Trtmnt Figure 5 The therapeutic benefits of oral interleukin (IL)-10 gene therapy as determined by tissue histology and myeloperoxidase (MPO) activity upon oral administration of murine IL-10-expressing plasmid DNA in nanoparticles-in-microsphere oral system (NiMOS). Images of hematoxylin and eosin stained colon tissue cryosections (a) and the tissue MPO activity (b). Each conscious animal received a 100 mg oral dose of pORF5-mIL-10 in gelatin nanoparticles or NiMOS. Mean±s.d. (n ¼ 4).

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1206 treatment of inflammation with IL-10 gene therapy. As to protect the payload from the gastrointestinal barriers, can be seen, the naive colon shows very good mucosal such a high level of transfection can be a result of architecture including the presence of goblet cells. In exposure of NiMOS to the underlying cells of the contrast, the large intestinal colitis-induced tissue gastrointestinal tract in case of colitis. showed classical features of inflammation, including The therapeutic benefits that can be obtained with thickening of luminal wall, loss of mucosal architecture high local levels of murine IL-10 in colitis-induced mice and cellular infiltration. These features were also are shown in Figure 3. As reported for Crohn’s disease, observed in the group administered with pORF5-mIL- an upregulation in the expression of Th-1-derived 10 in gelatin nanoparticles confirming earlier observation cytokines, such as IL-1, IL-12, IFN-g and TNF-a, was of lack of IL-10 transfection and absence of any anti- observed in our TNBS-induced murine colitis model. inflammatory effect with gelatin nanoparticles alone. In Crohn’s disease is characterized by an increase in contrast, colitis-bearing animals treated with plasmid activation of transcription factor nuclear factor-kB with DNA in NiMOS showed a significant decrease in downstream effects on cytokine production, such as inflammation with the histology of colon almost restor- TNF-a and immune regulation.31,32 The NiMOS treat- ing back to that the native state in such a short-time ment showed a 10-fold decrease in local levels of TNF-a. period. We also observed similar reduction in the levels of other Furthermore, we determined the MPO activity levels proinflammatory Th-1-derived cytokines, such as IL-12, as a marker of neutrophil infiltration and inflammation. IL-a, IL-1b and IFN-g. IL-10 has been known for its role As shown in Figure 5b, there was a significant increase in as a cytokine synthesis inhibitor and is also known to MPO activity levels, about 30 MPO activity per gram of modulate expression levels of Th-1-derived cytokines.4,6 tissue, were observed upon induction of acute colitis in Along with reduction in levels of proinflammatory Balb/c mice. The MPO activity upon treatment with cytokines, there was also significant reduction in certain transfected IL-10 using NiMOS formulations was sig- chemokines, such as MCP-1and MIP-1a. A direct action nificantly reduced (Po0.005) to about 2 MPO activity per of IL-10 on these chemokines has not been reported, but gram of tissue and was comparable to the enzyme these chemokines are known to modulate inflammatory activity in the normal noncolitis colon. In contrast, the conditions by regulating immune cell infiltration at the tissue MPO activity levels were significantly higher in disease site. To this effect, systemic administration of colitis group treated with pORF5-mIL-10 encapsulated MIP-1a has been reported to exacerbate IBD in mouse in gelatin nanoparticles group. This result supports our models.33 Reduction in the levels of Th-1-derived previous findings obtained from histological analysis cytokines, such as IFN-g, IL-1, IL-12 and TNF-a, due to of the large intestine specimens, whereby high levels the presence of IL-10 could have resulted in the cessation of cellular infiltration and inflammatory activity was of ongoing inflammatory process at the disease site. observed in control and gelatin nanoparticle-treated In contrast, higher levels of inflammatory cytokines animals. and chemokines, comparable with colitis group receiving saline treatment, were observed in animals treated with DNA encapsulated in gelatin nanoparticles. This could Discussion be due to the poor stability of gelatin nanoparticles in protease-rich gastrointestinal fluid leading to premature In this study, we describe the use of NiMOS for delivery release and degradation of the DNA. We did not observe of therapeutic plasmid DNA encoding murine IL-10 in any changes in RANTES levels for any of the treatment TNBS-induced acute colitis model developed in Balb/c groups. Previously, high levels of RANTES have been mice (Figure 2). The RT-PCR analysis results showed reported in chronic murine models of TNBS-induced very high levels of transcription of transgene (murine colitis.34 As we used an acute TNBS-induced colitis IL-10), as determined by the mRNA levels, for the colitis model, such an increase in RANTES levels were not group administered with NiMOS treatment. The high observed during the course of this study. levels seen for this group can be attributed to the murine The reduction in the local burden of proinflammatory IL-10 transgene, as such high levels were not observed cytokines and chemokines caused by excess IL-10 upon for other treatment groups, including gelatin nanoparti- administration of NiMOS treatment resulted in reversal cles. Some what high levels of IL-10 mRNA, compared to of experimental colitis, which was reflected in the body noncolitis group, was observed for colitis groups with weight changes, macroscopic disease activity score and saline and gelatin nanoparticles as treatment but this colon morphology (Figure 4). Upon induction of colitis could be a compensatory response due to induced colitis. using a hydroalcoholic TNBS solution, the animals start The ELISA results for IL-10 were in good correlation with losing body weight and this can be used successfully to the RT-PCR data. A high level of murine IL-10 protein determine the therapeutic benefits that can be obtained was observed for NiMOS treatment group, which would from various treatments.35 Among all the treatment have generated from high local levels of murine IL-10 groups, only the animals in colitis group administered mRNA levels observed during the RT-PCR analysis. We with NiMOS formulation showed weight gain back to observed relatively higher levels of IL-10 in the colitis the original 4 days after the loss of more than 10% of group administered with saline alone as compared to their original body weight upon administration of naive animals. This increase in IL-10 level has been second dose of TNBS The NiMOS treatment not only suggested to be compensatory IL-10, which is expressed reversed the loss in body weight, but also caused a in response to the induced colitis.6 The ELISA results reduction in the disease activity score. The disease state along with the RT-PCR results clearly demonstrate the also caused a decrease in colon length and colon weight high transfection ability of NiMOS formulation upon in mice and these values reversed back to baseline upon oral administration. In addition to the ability of NiMOS administration of NiMOS treatment to colitis mice. The

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1207 significant decrease in disease activity for the NiMOS poly(vinyl alcohol) by repeated centrifugation, and treatment group can only be attributed to the increased freeze-drying to obtained free-flowing powder. local levels of IL-10 upon treatment. The IL-10 levels were seen to reduce the burden of proinflammatory Development of TNBS-induced acute colitis model cytokines in the colon and led to a reduction in the in Balb/c mice process of stimulation of the immune system maintain- The animal studies described below were performed ing the disease state. according to the experimental protocol approved by the Lastly, we studied tissue histology along with deter- Institutional Animal Care and Use Committee at North- mination of tissue cellular infiltration using MPO activity eastern University (Boston, MA, USA). Female Balb/c (Figure 5). The hematoxylin and eosin were used to stain mice, 8–10 weeks old, weighing approximately 20 g, the tissue cryosections of colon obtained from different were purchased from Charles River Laboratories treatment groups. The microscopic evaluation of non- (Wilmington, MA, USA). Before the model development, colitis tissue section revealed a normal mucosal archi- all mice were anesthetized by intraperitoneal injection tecture along with presence of goblet cells. Upon of ketamine-xylazine. Following anesthesia, 2.5 mg of induction of colitis, the composure of colon was TNBS dissolved in 100 ml of 40% (v/v) ethanol solution compromised accompanied with loss in the protective was administered rectally to the anesthetized mice twice epithelial layer and heavy infiltration of immune cells. on day 0 (at the start of the experiment) and at day 4 The lack in ability of gelatin nanoparticles to successfully using a 1.2 mm polyethylene catheter. Acute colitis deliver the transgene to the cells of the colon inflicted development was confirmed by measurements of with colitis resulted in no change in the ongoing weight loss, stool consistency, rectal bleeding and tissue inflammatory process occurring in the colon. For the histology. NiMOS treatment group the colon architecture was reversed back to normal. The high local levels of IL-10 Oral administration of IL-10 plasmid encapsulated were not only able to reduce the levels of inflammatory cytokines and chemokines and cellular infiltration, but in NiMOS this in turn led to proliferation of normal tissue cells When acute colitis was confirmed in Balb/c mice after causing a reversal in the disease state. The reduction in administration of the two TNBS doses, the animals were cellular infiltration for NiMOS group was further fasted for 24 h and divided into control and treatment demonstrated by MPO activity which was significantly groups. In the treatment group, conscious animals lower, almost back to normal levels, where as it remained received NiMOS encapsulated with pORF5-mIL-10 at a very high for colitis group treated with saline and dose of 100 mg by oral gavage. Additionally, gelatin gelatin nanoparticles. This study further indicates that nanoparticles encapsulated with pORF5-mIL-10 was IL-10 plays a crucial role in inflammation and can also administered to another group at a dose of 100 mg and result in reduction of inflammation by acting as an served as a control. Control animals did not receive any immunomodulatory cytokine. IBD in humans is a treatment. Changes in the body weight was measured chronic disease involving cycles of remission and relapse periodically. In addition, after 4 days post-administration upon treatment with many different kinds of drugs of the formulations, the the animals were killed by CO2 and steroidal molecules. Gene therapy has the potential inhalation and the large intestine was surgically excised to provide prolonged remission to patients suffering for evaluation of IL-10 expression and therapeutic from this chronic disease mediated by the interplay efficacy. between pro- and anti-inflammatory cytokines at the disease site. Evaluation of in vivo IL-10 transgene expression The expression of murine IL-10 upon oral administration of the plasmid in gelatin nanoparticles and NiMOS Materials and methods formulations in colitis-induced Balb/c mice was evalu- ated by quantitative RT-PCR and by measuring the Preparation of plasmid DNA-encapsulated NiMOS protein levels using Quantikine sandwich ELISA (R&D Nanoparticles-in-microsphere oral system were prepared Systems, Minneapolis, MN, USA). using murine IL-10-expressing plasmid DNA (pORF5- mIL-10)-encapsulated type-B gelatin nanoparticles in Quantitative RT-PCR. Quantitative RT-PCR was per- poly(epsilon-caprolactone) matrix to form 2–5 mm sized formed on homogenized large intestinal tissue to microspheres as previously described.24,25 Briefly, 10 mg determine the levels of mRNA transcript according of plasmid DNA-encapsulated gelatin nanoparticles to the published procedure.36 The sequences of the primers were suspended in distilled deionized water. This used were as follows: (1) murine IL-10 1: CCAAGCCTTA suspension was added to 2.0 ml of dichloromethane TCGGAAATGA, (2) murine IL-10 2: TCTCACCCAGGG containing 0.5% (w/v) poly(epsilon-caprolactone) and AATTCAAA, (3) murine L32 1: AGCAACAAGAAAAC homogenized at room temperature. The formed primary CAAGCACAT and (4) murine L32 2: TTGACATTGTG emulsion was added to an external aqueous phase GACCAGGAACT. The results obtained from the RT-PCR containing 0.1% (w/v) poly(vinyl alcohol) and further reaction were then analyzed by comparative Ct analysis homogenized to obtain a stable secondary emulsion. The for determination of relative murine IL-10 cDNA in the formed double emulsion was stirred to evaporate the samples. The PCR reaction product was then analyzed dichloromethane followed by purification by centrifuga- on a 1.8% agarose gel. After running the gel for 30 min at tion and finally lyophilized. The microparticle formula- 110 V, the image was recorded under UV light and tion was obtained by allowing dichloromethane to processed to visually confirm the formation of PCR evaporate at room temperature, washing off the excess product.

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1208 Murine IL-10 ELISA. Sandwich ELISA kit from R&D with murine IL-10-expressing plasmid DNA delivered Systems was used to determine the levels of expressed orally in NiMOS. Gelatin nanoparticles, on the other IL-10 in the large intestine, following transfection with hand, were not as effective in oral DNA delivery control and NiMOS formulations.37 Following substrate probably due to rapid degradation in the protease- and addition, the optical density was measured at 450 nm. nuclease-rich gastrointestinal environment. The locally The IL-10 concentrations that were obtained from the transfected IL-10 was very effective in reducing the levels ELISA experiments were then normalized to the total of proinflammatory cytokines and chemokines, such as protein content and the values are expressed as picogram MCP-1 and MIP-1a, which are secreted by activated of murine IL-10 expressed per mg of total protein. leukocytes and other tissue cells for recruitment and proliferation of macrophages and at the site Therapeutic efficacy of locally expressed IL-10 in acute of inflammation. The therapeutic benefits of transfected colitis model IL-10 were further demonstrated by gain in body weight Inflammatory cytokine and chemokine profiling. The and a significant decrease in the clinical activity score. changes in pro- and anti-inflammatory cytokine concen- Tissue histology also showed complete lack of cell trations upon treatment of IL-10-expressing plasmid in infiltration and recovery of mucosal architecture in the the large intestine was evaluated by using Q-Plex Mouse treatment group. These results are highly encouraging Cytokine Array ELISA kit purchased from Bio-Legend for the potential development of oral gene therapy using (San Diego, CA, USA). The cytokine assays were NiMOS, especially for the treatment of chronic diseases performed according to the suggested protocol. The such as IBD. luminescence intensity obtained for each cytokine was used to calculate the concentration of cytokines in the tissue samples using a calibration curve. Acknowledgements We are grateful to David Nyugen in Professor Robert Macroscopic efficacy evaluations. Following oral Langer’s laboratory at MIT (Cambridge, MA, USA) for administration of various control and test formulations, the use the Coulter particle size analysis instrument. the macroscopic factors were reported every day until Dr Takeshi Sano and Dr Alan Jerusalmi from the Beth the end of the study and the colitis activity was Israel Deaconess Medical Center (Boston, MA, USA) are quantified with clinical score by assessing the weight acknowledged for providing technical assistance with loss, stool consistency and rectal bleeding as previously the development of murine acute colitis model. described.38 Along with the above mentioned para- meters, the macroscopic efficacy evaluation also con- sisted colon length (in cm) and colon weight (in grams) References measurements. 1 Blumberg RS, Saubermann LJ, Strober W. Animal models of Microscopic efficacy evaluations. Microscopic criteria mucosal inflammation and their relation to human inflammatory for damage and inflammation was assessed by light bowel disease. Curr Opin Immunol 1999; 11: 648–656. microscopy of hematoxylin and eosin stained histological 2 Blumberg RS, Strober W. Prospects for research in inflammatory sections obtained from tissue samples taken from a bowel disease. JAMA 2001; 285: 643–647. region of the inflamed colon immediately adjacent to the 3 Isaacs KL, Lewis JD, Sandborn WJ, Sands BE, Targan SR. State of gross macroscopic damage. After staining, the sections the art: IBD therapy and clinical trials in IBD. Inflamm Bowel Dis were imaged using bright-field microscopy and the 2005; 11 (Suppl 1): S3–S12. mucosal architectural change, cellular infiltration, 4 Korzenik JR, Podolsky DK. Evolving knowledge and therapy external muscle thickening, presence of crypt abscess, of inflammatory bowel disease. Nat Rev Drug Discov 2006; 5: goblet cell depletion, signs of edema, surface epithelial 197–209. cell hyperplasia and signs of epithelial regeneration were 5 Fichtner-Feigl S, Fuss IJ, Preiss JC, Strober W, Kitani A. assessed to qualitatively determine the degree of Treatment of murine Th1- and Th2-mediated inflammatory bowel disease with NF-kappa B decoy oligonucleotides. J Clin mucosal damage and repair in control and treated 14,39 Invest 2005; 115: 3057–3071. groups, respectively. 6 Li MC, He SH. IL-10 and its related cytokines for treatment of inflammatory bowel disease. World J Gastroenterol 2004; 10: Determination of tissue MPO activity. The tissue MPO 620–625. activity is a reliable index of inflammation based on the 7 Prieto J, Herraiz M, Sangro B, Qian C, Mazzolini G, Melero I et al. 40 infiltration of activated neutrophils. The large intestinal The promise of gene therapy in gastrointestinal and liver specimen were minced in 1.0 ml of hexadecyltrimethy- diseases. Gut 2003; 52 (Suppl 2): ii49–ii54. lammonium bromide buffer (0.5% in 50 mM phosphate 8 Sato Y, Takahashi S, Kinouchi Y, Shiraki M, Endo K, Matsumura buffer) on ice and homogenized. The change in absor- Y et al. IL-10 deficiency leads to somatic mutations in a model of bance upon addition of o-dianisidine hydrochloride and IBD. Carcinogenesis 2006; 27: 1068–1073. hydrogen peroxide was measured at 460 nm for 3 min 9 Lindsay J, Van Montfrans C, Brennan F, Van Deventer S, and the values were calculated using a reported equation Drillenburg P, Hodgson H et al. IL-10 gene therapy prevents and normalized to a gram of tissue.41 TNBS-induced colitis. Gene Therapy 2002; 9: 1715–1721. 10 Lindsay JO, Sandison A, Cohen P, Brennan FM, Hodgson HJ. IL-10 gene therapy is therapeutic for dextran sodium sulfate- Conclusions induced murine colitis. Dig Dis Sci 2004; 49: 1327–1334. 11 Herfarth H, Scholmerich J. IL-10 therapy in Crohn’s disease: The results of study demonstrated successful transfec- at the crossroads. Treatment of Crohn’s disease with the anti- tion at the inflammatory site in the gastrointestinal tract inflammatory cytokine . Gut 2002; 50: 146–147.

Gene Therapy Oral IL-10 gene therapy for IBD MD Bhavsar and MM Amiji 1209 12 Nakase H, Okazaki K, Tabata Y, Ozeki M, Watanabe N, Ohana M 28 Kaul G, Amiji M. Cellular interactions and in vitro DNA et al. New cytokine delivery system using gelatin microspheres transfection studies with poly(ethylene glycol)-modified gelatin containing interleukin-10 for experimental inflammatory bowel nanoparticles. J Pharm Sci 2005; 94: 184–198. disease. J Pharmacol Exp Ther 2002; 301: 59–65. 29 Kommareddy S, Amiji M. Poly(ethylene glycol)-modified 13 Steidler L, Hans W, Schotte L, Neirynck S, Obermeier F, Falk W thiolated gelatin nanoparticles for glutathione-responsive in- et al. Treatment of murine colitis by secreting tracellular DNA delivery. Nanomedicine 2007; 3: 32–42. interleukin-10. Science 2000; 289: 1352–1355. 30 Kommareddy S, Amiji M. Antiangiogenic gene therapy with 14 Barbara G, Xing Z, Hogaboam CM, Gauldie J, Collins SM. systemically administered sFlt-1 plasmid DNA in engineered Interleukin 10 gene transfer prevents experimental colitis in rats. gelatin-based nanovectors. Cancer Gene Ther 2007; 14: 488–498. Gut 2000; 46: 344–349. 31 Ardizzone S, Bianchi Porro G. Inflammatory bowel disease: new 15 Lecollinet S, Gavard F, Havenga MJ, Spiller OB, Lemckert A, insights into pathogenesis and treatment. J Intern Med 2002; 252: Goudsmit J et al. Improved gene delivery to intestinal mucosa by 475–496. adenoviral vectors bearing subgroup B and d fibers. J Virol 2006; 32 Gaya DR, Russell RK, Nimmo ER, Satsangi J. New genes in 80: 2747–2759. inflammatory bowel disease: lessons for complex disease? Lancet 16 Romano G, Pacilio C, Giordano A. Gene transfer technology in 2006; 367: 1271–1284. therapy: current applications and future goals. Stem Cells 1999; 33 Pender SL, Chance V, Whiting CV, Buckley M, Edwards M, 17: 191–202. Pettipher R et al. Systemic administration of the chemokine 17 Dang JM, Leong KW. Natural polymers for gene delivery and inflammatory protein 1alpha exacerbates inflam- tissue engineering. Adv Drug Deliv Rev 2006; 58: 487–499. matory bowel disease in a mouse model. Gut 2005; 54: 1114–1120. 18 Dubensky Jr TW, Liu MA, Ulmer JB. Delivery systems for gene- 34 Ajuebor MN, Hogaboam CM, Kunkel SL, Proudfoot AE, Wallace based vaccines. Mol Med 2000; 6: 723–732. JL. The chemokine RANTES is a crucial mediator of the 19 Kaul G, Amiji M. Tumor-targeted gene delivery using poly progression from acute to chronic colitis in the rat. J Immunol (ethylene glycol)-modified gelatin nanoparticles: in vitro and 2001; 166: 552–558. in vivo studies. Pharm Res 2005; 22: 951–961. 35 Lindsay JO, Ciesielski CJ, Scheinin T, Hodgson HJ, Brennan FM. 20 Lee KY, Kwon IC, Kim YH, Jo WH, Jeong SY. Preparation of The prevention and treatment of murine colitis using gene chitosan self-aggregates as a gene delivery system. J Control therapy with adenoviral vectors encoding IL-10. J Immunol 2001; Release 1998; 51: 213–220. 166: 7625–7633. 21 Luo D, Saltzman WM. Synthetic DNA delivery systems. Nat 36 Egger B, Bajaj-Elliott M, MacDonald TT, Inglin R, Eysselein VE, Biotechnol 2000; 18: 33–37. Buchler MW. Characterisation of acute murine dextran sodium 22 Mansouri S, Lavigne P, Corsi K, Benderdour M, Beaumont E, sulphate colitis: cytokine profile and dose dependency. Digestion Fernandes JC. Chitosan-DNA nanoparticles as non-viral vectors 2000; 62: 240–248. in gene therapy: strategies to improve transfection efficacy. Eur J 37 Gasche C, Bakos S, Dejaco C, Tillinger W, Zakeri S, Reinisch W. Pharm Biopharm 2004; 57: 1–8. IL-10 secretion and sensitivity in normal human intestine and 23 Martien R, Loretz B, Schnurch AB. Oral gene delivery: design of inflammatory bowel disease. J Clin Immunol 2000; 20: 362–370. polymeric carrier systems shielding toward intestinal enzymatic 38 van Montfrans C, Bennink RJ, de Bruin K, de Jonge W, Verberne attack. Biopolymers 2006; 83: 327–336. HJ, Ten Kate FJ et al. In vivo evaluation of 111In-labeled T- 24 Bhavsar M, Amiji M. Gastrointestinal distribution and in vivo homing in experimental colitis. J Nucl Med 2004; 45: transfection studies with nanoparticles-in-microsphere oral 1759–1765. system (NiMOS). J Control Release 2007; 119: 339–348. 39 Melgar S, Karlsson A, Michae¨lsson E. Acute colitis induced by 25 Bhavsar MD, Tiwari SB, Amiji MM. Formulation optimiza- dextran sulfate sodium progresses to chronicity in C57BL/6 but tion for the nanoparticles-in-microsphere hybrid oral delivery not in BALB/c mice: correlation between symptoms and system using factorial design. J Control Release 2006; 110: inflammation. Am J Physiol Gastrointest Liver Physiol 2005; 288: 422–430. G1328–G1338. 26 Kaul G, Amiji M. Long-circulating poly(ethylene glycol)- 40 Talero E, Sanchez-Fidalgo S, Ramon Calvo J, Motilva V. Galanin modified gelatin nanoparticles for intracellular delivery. Pharm in the trinitrobenzene sulfonic acid rat model of experimental Res 2002; 19: 1061–1067. colitis. Int Immunopharmacol 2006; 6: 1404–1412. 27 Kommareddy S, Amiji M. Preparation and evaluation of 41 Jordan JE, Zhao ZQ, Sato H, Taft S, Vinten-Johansen J. Adenosine thiol-modified gelatin nanoparticles for intracellular DNA A2 receptor activation attenuates reperfusion injury by inhibiting delivery in response to glutathione. Bioconjug Chem 2005; 16: neutrophil accumulation, superoxide generation and coronary 1423–1432. endothelial adherence. J Pharmacol Exp Ther 1997; 280: 301–309.

Gene Therapy