CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4:754 –759
A Phase I Trial With Transgenic Bacteria Expressing Interleukin-10 in Crohn’s Disease HENRI BRAAT,* PIETER ROTTIERS,‡ DANIEL W. HOMMES,§ NATHALIE HUYGHEBAERT,储 ERIK REMAUT,‡ JEAN–PAUL REMON,储 SANDER J. H. VAN DEVENTER,* SABINE NEIRYNCK,‡,¶ MAIKEL P. PEPPELENBOSCH,** and LOTHAR STEIDLER‡,¶ *Department of Experimental Internal Medicine, Academic Medical Centre, Amsterdam, The Netherlands; ‡Department for Molecular Biomedical Research, Ghent University, and Flanders Interuniversity Institute for Biotechnology (VIB), Gent, Belgium; §Department of Gastroenterology, Academic Medical Centre, Amsterdam, The Netherlands; 储Laboratory of Pharmaceutical Technology, Ghent University, Gent, Belgium; ¶Alimentary Pharmabiotic Centre, University College Cork, Cork, Ireland; and **Department of Cell Biology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
Background & Aims: The use of living, genetically modified bacteria is an effective approach for topical delivery of immunomodulatory proteins. This strategy circumvents systemic side effects and allows long-term treatment of chronic diseases. However, treatment of patients with a living, genetically modified bacterium raises questions about the safety for human subjects per se and the biologic containment of the transgene. Methods: We treated Crohn’s disease patients with genetically modified Lactococcus lactis (LL-Thy12) in which the thymidylate synthase gene was replaced with a synthetic sequence encoding mature human interleukin10. Ten patients were included in a placebo-uncontrolled trial. Patients were assessed daily for the presence of potential adverse effects by direct questioning and assessment of disease activity. We evaluated the presence and kinetics of LL-Thy12 release in the stool of patients by conventional culturing and quantitative polymerase chain reaction of LL-Thy12 gene sequences. Results: Treatment with LL-Thy12 was safe because only minor adverse events were present, and a decrease in disease activity was observed. Moreover, fecally recovered LLThy12 bacteria were dependent on thymidine for growth and interleukin-10 production, indicating that the containment strategy was effective. Conclusions: Here we show that the use of genetically modified bacteria for mucosal delivery of proteins is a feasible strategy in human beings. This novel strategy avoids systemic side effects and is biologically contained; therefore it is suitable as maintenance treatment for chronic intestinal disease.
rohn’s disease is a chronic transmural granulomatous inflammation of the gastrointestinal tract, with an increasing incidence in Western society.1 Current therapy is mainly focused on remission induction, but safe and effective strategies for remission maintenance are generally considered to be the challenge of contemporary
C
research.2 In healthy individuals, tight regulation of the mucosal immune system prevents excessive inflammatory responses toward normal intestinal bacteria. The crucial role of interleukin-10 (IL-10) in this process is highlighted by experiments in IL-10 – deficient mice. These animals develop a chronic bowel disease resembling Crohn’s disease in human beings, which is in part caused by a loss of suppression of the mucosal immune response toward the normal intestinal bacterial flora.3 Unfortunately, systemic IL-10 treatment of Crohn’s disease patients is not very effective in inducing clinical remission and is associated with considerable side effects,4,5 which are partly due to the fact that systemic IL-10 induces the proinflammatory cytokine interferon-␥.6 However, studies in experimental models suggest that topical treatment with IL-10 is effective for preventing disease. Intragastric administration of a recombinant Lactococcus lactis strain, secreting murine IL-10, prevented onset of colitis in IL-10 knockout mice, caused a 50% reduction of the inflammation in dextran sulfate sodium-induced chronic colitis,7 and prevented colonic inflammation in trinitrobenzene sulfonic acid– colitis (unpublished data). Importantly, unmodified L lactis does not have beneficial effects on colonic inflammation in these models. Live, genetically modified bacteria might be a suitable way for topical delivery of recombinant proteins in the human gastrointestinal tract, but such a strategy has never been applied and gives rise to important questions regarding clinical and environmental safety as well as technological feasibility. To address these issues we conAbbreviations used in this paper: CFU, colony-forming units; IL-10, interleukin-10; LL-Thy12, Lactococcus lactis– human interleukin-10; Q-PCR, quantitative polymerase chain reaction © 2006 by the American Gastroenterological Association Institute 1542-3565/06/$32.00 doi:10.1016/j.cgh.2006.03.028
June 2006
ducted a clinical trial with a genetically modified L lactis. The parent bacterium is a food-grade bacterium and generally regarded as safe according to the United States Food and Drug Administration. Moreover, it is not a normal gut resident, and it does not colonize the human intestinal tract.8,9 The thyAgene of L lactis was replaced with a synthetic sequence encoding mature human IL10, fused at its N-terminus to a lactococcal secretion signal. This method provides a biologically contained bacterium (LL-Thy12) that is thymidine- or thyminedependent and unable to survive in the environment without continuous supplementation of either one of both compounds.10 In this study, we show that treatment of human beings with a genetically modified bacterium secreting a human protein is clinically safe and biologically contained and is therefore a realistic option.
Methods Patient Selection and Treatment Regime Patients were screened for eligibility 1 week before enrollment, once informed consent had been obtained. Eligible patients had Crohn’s disease, confirmed by routine endoscopic and histologic examination, for at least 3 months’ duration and a score on the CDAI ranging from 220 – 450. The CDAI incorporates 8 variables related to the disease activity: the number of liquid or very soft stools, the severity of abdominal pain or cramping, general well-being, the presence of extraintestinal manifestations, abdominal mass, use of antidiarrheal drugs or opioids, hematocrit, and body weight. These items yield composite scores ranging from 0 to approximately 600; higher scores indicate greater disease activity. Colonoscopy was performed at baseline to confirm diagnosis.11 The endoscopic score obtained during colonoscopy ranges from 0 –3 and is generally used at the Academic Medical Center to indicate disease activity. Zero indicates no inflammation, 1 indicates the presence of edema and aphthous ulceration; 2 indicates the presence of large separated ulcerations, and 3 indicates the presence of only limited islets of normal mucosa. Patients receiving the following medications were eligible: 5-aminosalicylates and corticosteroids (at least 8 weeks, with stable dosage regimen for at least 2 weeks), methotrexate, azathioprine, and 6-mercaptopurine (at least 8 weeks, with stable dosage regimen for at least 10 weeks). All previous Crohn’s disease medications were stopped at least 4 weeks before the study (8 weeks for infliximab). Patients were excluded if they exhibited significant comorbidity, an ileostoma, used antibiotics, or had a positive culture for enteral pathogens. Patients were allowed to take Crohn’s disease–related medication, but no change in medication dosage was allowed during the experiment. One day before the treatment period, all eligible patients underwent physical and routine laboratory examination and were hospitalized on a governmental approved ward for gene therapy in the Academic Medical Center. Patients received 10
IL-10 SECRETING L LACTIS IN CD
755
capsules with 1 ⫻ 1010 colony-forming units (CFU) of LLThy12 twice daily for 7 days. During the study period, patients received 4 g of cholate acid binder twice daily (Questran; Zambon, Amersfoort, The Netherlands) and 40 mg of proton pump inhibitor once daily (Pantozol; Altana Pharma BV, Hoofddorp, The Netherlands) to improve LL-Thy12 viability.
Specific Primers for Thy12 and Real-Time Polymerase Chain Reaction LL-Thy12-specific primer sets targeting the synthetic IL-10 gene and L lactis genus-specific primers designed on the basis of 16S rDNA gene sequence were used to quantify fecal LL-Thy12. The synthetic IL-10 gene primers were designed by using the lactococcal usp45 secretion signal (UsphIL10_forward: 5=-CAGCCCCGTTGTCAGGTGTTTAC-3=) and the synthetic, L lactis codon optimized hIL-10 gene8 (Usp-hIL10_reverse: 5=-TCACGAAGCATGTTTGGCAAGTTAC-3=). The 16S rDNA primers specific for L lactis were 16S L lactis_forward: 5=-CGTAGGTCCCGAGCGTTGTC-3= and 16S L lactis_reverse: 5=-GGTTGAGCCACTGCCTTTTACAC-3=.
Results Study Design The Medical Ethical Commission of the Academic Medical Center and the Dutch Administration of Public Health, Environment and Nature approved a limited clinical trail with LL-Thy12 in patients under physical containment (http://www.cogem.net). This study aimed at evaluating clinical as well as environmental safety of LL-Thy12. In total, 10 patients with moderate to severe Crohn’s disease (see Supplemental Table 1 at www. cghjournal.org) were included in the study. All patients were admitted to an isolated ward during the study period, and all stool material was collected and decontaminated by chemical treatment except for samples investigated for the presence of LL-Thy12. Patients took 10 enteric-coated (Eudragit FS 30 D and LD30D-55; Rhom & Haas GmbH, Darmstad, Germany) capsules containing about 1010 CFU of lyophilized LL-Thy1212 twice daily for 7 consecutive days. Despite the Eudragit coating, gastric acid still negatively influences survival of LL-Thy12; thus patients were co-treated with a proton pump inhibitor.12 Clinical Safety During treatment, patients were assessed daily for the presence of potential adverse effects by direct questioning and, when necessary, additional testing. The same procedure was performed at days 8, 14, and 28 (Table 1). One patient was withdrawn from the study on the second day because of persistent vomiting and sub-
756
BRAAT ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 6
Table 1. Adverse Events in LL-Thy12–Treated Patients n
Patient no.
Adverse events Stomach pain Abdominal pain Headache Common cold Flatulence Vomiting Small bowel obstruction
1 1 1 1 3 1 1
10 4 10 10 4, 9, and 10 3 5
Serious adverse events Exacerbation during LL-Thy12 Exacerbation after withdrawal LL-Thy12
0 4
1, 2, 4, and 6
NOTE. Patient no., see Supplemental Table 1 at http://www.cgh journal.org.
sequent noncompliance with the dose regimen; he was finally treated with infliximab. No other study withdrawals or serious adverse events occurred during the supplementation period; adverse events were generally minor (flatulence) and either might be related to the administration of large amounts of metabolically active L lactis or common to patients with Crohn’s disease (Table 1). To determine possible disease progression we used the
CDAI13 and measured serum levels of C-reactive protein at days ⫺1, 8, 14, and 28 (Figure 1). Patients had a mean decrease in CDAI of 71.7 after 1 week of treatment, the decrease in soft stool being a prominent factor in this, and a concomitant decrease in C-reactive protein levels (Figure 1). Clinical benefit was observed in 8 of 10 patients; 5 patients went into complete clinical remission (CDAI, ⬍150), and 3 patients exhibited a clinical response (decrease in CDAI, ⬎70). One patient was excluded from the study, and 1 patient only exhibited a partial response. Four patients developed a significant relapse with an increase in CDAI of more than 100 points after discontinuation of LL-Thy12 treatment. Azathioprine and 6-mercaptopurine have a delayed onset of efficacy of several months. It is thus possible that patients within the trial could have started on their azathioprine/6-mercaptopurine 8 weeks before entry. In this regard the clinical effects documented may have been due to azathioprine/6-mercaptopurine rather than the L lactis (Supplemental Table 1), although patients with or without azathioprine/6-mercaptopurine did not do significantly better than patients without this medication. In conclusion, treatment with LL-Thy12 was
Figure 1. Clinical scores of patients after LL-Thy12 administration. (A) CDAI and C-reactive protein levels in patients at days ⫺1, 8, 14, and 28. (B) Represents the mean CDAI and C-reactive protein (⫾ standard error of the mean) of 10 patients during treatment and follow-up. The black bar indicates the treatment period.
June 2006
IL-10 SECRETING L LACTIS IN CD
757
Figure 2. Biologic containment of LL-Thy12. (A) Quantification of LL-Thy12 bacterial DNA was performed by using an iCycler Real-Time PCR detection System (BioRad). DNA melting curves were used to monitor product specificities. Detection was based on fluorescence resonance energy transfer, with a SYBR Green 490 fluorophore. Thermal cycling conditions were as follows: 95°C for 10 min, 40 cycles of 95°C for 15 s, and 60°C for 1 min. The CT value was the cycle at which a statistically significant increase in fluorescence intensity was first detected in association with a logarithmic increase in PCR product. The detection system constructed a standard curve by plotting the CT value against each dilution of the known standard and used this to determine the quantitative value for test samples from the CT value detected. The known standard was DNA extracted from serial diluted LL-Thy12 in feces (obtained from patients before treatment), from which real-time PCR detection was linear for cell counts ranging from 109–104 LL-Thy12 cells per g feces. All real-time PCR reactions were performed in duplicate. (B) Quantification of LL-Thy12 during and after the treatment period. To evaluate the kinetics of LL-Thy12 release, the daily amount of LL-Thy12 genome equivalents in the fecal samples was divided by the total amount of LL-Thy12 genome equivalents detected during the entire study period. Depicted is the mean ⫾ standard error of the mean of all patients. (C) Analysis of LL-Thy12–specific sequences per g feces as estimated by using specific primer sets that target the synthetic IL-10 gene used in this study or L lactis species–specific primers designed on the basis of 16S rDNA gene sequence (□). The data represented here were obtained from 1 patient.
clinically well-tolerated and did not induce systemic and/or long-term side effects. The lack of a control group, as is intrinsic to a phase I study, does not allow conclusions regarding the clinical efficacy, but experimental data suggest that the combination of L lactis MG1363, the parental strain of live LL-Thy12, and the transgene are necessary for effectivity. Furthermore, L lactis is not known to have beneficial effects on Crohn’s disease in human beings and showed no effect on animal enterocolitis.14 It is uncertain what properties of LL-Thy12 are responsible for the synergistic activity observed in animal models, or whether epitope mimicry15 contributes to the potential beneficial effects observed in this trial. Biologic Containment We evaluated the presence and kinetics of LL-Thy12 release in the stool of patients during and after the treatment period. Because of unexpected high background in the patients involved, the method for detection of specific LLThy12 colonies, as applied earlier in pigs10 and validated before this study on feces from healthy individuals, could not be used. As an alternative we extracted bacterial DNA from stool samples and determined the number of LLThy12 genome equivalents by quantitative polymerase chain reaction (Q-PCR) of the synthetic IL-10 gene (a LL-Thy12–specific DNA sequence). The method was linear for cell counts ranging from 104–109 LL-Thy12 cells per g stool (Figure 2A). We used the QIAamp DNA Stool Kit (QIAGEN, Hilden, Germany) to extract bacterial DNA from fecal samples, prepared as described.10 Analysis of
daily stool samples showed that a maximum amount of total, live, and dead LL-Thy12 was recovered on day 4 (33.9% of the total amount detected during the entire treatment period) (Figure 2B). Two days after termination of the therapy, LL-Thy12–specific DNA sequences could no longer be detected by using this methodology, indicating that less than 104 cells per g stool were present (Figure 2B). The maximal percentage of viable, ie, IL-10 –secreting, LL-Thy12 was determined from fresh stool samples, diluted, and plated on thymidine-containing solid agar GM17 plates.10 On average, total bacterial count was 1 ⫻ 1010 CFU per g stool. In several experiments, colony lifting followed by immunoprobing of 1000 –2000 of these individual bacterial colonies showed that none secreted IL-10. Q-PCR data (Figure 2B) had shown that on the day of maximal recovery, the total number of live and dead LL-Thy12 was 16.4 ⫻ 107 cells per g stool. Therefore, 100% viability would yield an average of 16 IL-10 –secreting colonies per 1000 CFU on the thymidine plates. Because none were detected, we can deduce that on the day of maximal recovery, less than 6.25% of all LL-Thy12 were still alive. We also determined viable LL-Thy12 count in daily stool samples of a patient. To this end, bacterial isolates were kept at 4°C for 30 days in M9-azide medium, which substantially enriches for L lactis.10 Plating on thymidine-containing GM17 agar plates and subsequent PCR analysis for the presence of LL-Thy12–specific DNA sequences in 100 individual colonies revealed that only a minority of these represented viable LL-Thy12
758
BRAAT ET AL
(10.4%, 10.5%, and 13% at days 2, 4, and 6, respectively). Growth of these LL-Thy12 isolates was strictly dependent on the addition of thymidine, and IL-10 production was exclusively associated with the LL-Thy12 genotype. In addition, daily stools of a patient with a typical LL-Thy12 release pattern was analyzed by Q-PCR with either LL-Thy-12–specific primers or L lactis 16S rDNA-specific primers. We observed a strict concomitance between the values obtained with both sets of primers (Figure 2C).
Discussion We conducted the first human trial with a genetically engineered, therapeutic bacterium, and the results obtained indicate that such a strategy can be both safe for the patient as well as biologically contained. Such therapy encompasses important advances compared with current systemic treatment. The clinical parameters obtained, including the assessment of adverse events and scoring for disease activity, showed that the topical application of recombinant proteins by L lactis–mediated delivery is indeed safe for the patient, and that systemic side effects can be circumvented by such an approach. This absence of major side effects allows long-term treatment, as is necessary for the management of chronic disease. The containment strategy, to prevent uncontrolled environmental spreading of the transgene, is guaranteed at 3 different levels. First, the parent organism, L lactis subspecies MG1363, is a non-pathogenic, food-derived bacterium that was cured of all resident plasmids and was shown to be deficient of conjugative transposition.16,17 Second, the organism is thyA-deficient and unable to replicate in an environment lacking thymine or thymidine. Finally, the transgene is stably incorporated in the bacterial chromosome, preventing lateral dissemination of the transgene and obviating the need for antibiotic selection markers. Despite continued daily intake of 2 ⫻ 1011 CFU LL-Thy12 during the treatment period, a decrease of LL-Thy12 CFU in feces was detected, indicating that thymine less death was highly functional, and prominent LL-Thy12 DNA degradation occurred during transit. This observation might be explained by an increased transition time during the course of the treatment period. Indeed, extrapolated weekly soft stool frequency decreased from 44.0 ⫾ 7.5 to 20.5 ⫾ 4.3 on days ⫺1 and 8, respectively. The strict concomitance in bacterial count (Q-PCR) detected with a species-specific probe for L lactis 16s RNA genes and with a LL-Thy12– specific probe for the hIL-10 transgene, the thymidine growth dependency of LL-Thy12 and the restriction of
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 6
IL-10 production to LL-Thy12 suggest that transition of the bacterium through the human gastrointestinal tract yields genetically intact LL-Thy12. Hence, the results obtained demonstrate the efficacy of the chosen biologic containment strategy. Systemic IL-10 has been disappointing in the treatment of Crohn’s disease.4,5 Thus, mode of IL-10 delivery might be important for the effects observed in the present study. In agreement, we have preliminary evidence that bacterial context of IL-10 presentation is important because dendritic cells matured in the presence of IL-10 were still capable of sustaining a mixed lymphocyte reaction, whereas live bacteria producing a comparable amount of IL-10 did not induce T-cell proliferation in such a reaction (H.B., unpublished observations). Thus, the strong clinical effect observed in this study might well be due to the combination of a lactic acid bacterium and local IL-10 production. In conclusion, the results obtained suggest that the chosen therapeutic strategy is safe and biologically contained. This observation allows a regulatory approved design of an out-of-hospital placebo-controlled trial to test the clinical effect of LL-Thy12. Bacterial-based topical delivery of biologically active proteins is a novel and highly promising avenue for combating mucosal disease. This study demonstrates, as has been alluded to in a number of recent comments,18 –20 that this approach is a realistic option in human medicine. Our results clearly hold promise for the safe application of live genetic modified L lactis as an efficient therapeutic tool in human beings with chronic intestinal inflammation.
Supplementary Data Note: to access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org.
References 1. Bouma G, Strober W. The immunological and genetic basis of inflammatory bowel disease. Nat Rev Immunol 2003;3:521–533. 2. Egan LJ, Sandborn WJ. Advances in the treatment of Crohn’s disease. Gastroenterology 2004;126:1574 –1581. 3. Kuhn R, Lohler J, Rennick D, et al. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 1993;75:263–274. 4. Fedorak RN, Gangl A, Elson CO, et al. Recombinant human interleukin 10 in the treatment of patients with mild to moderately active Crohn’s disease: the Interleukin 10 Inflammatory Bowel Disease Cooperative Study Group. Gastroenterology 2000;119:1473–1482. 5. Targan SR, Hanauer SB, van Deventer SJ, et al. A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn’s disease: Crohn’s Disease cA2 Study Group. N Engl J Med 1997;337:1029 –1035. 6. Tilg H, van Montfrans C, van den Ende A, et al. Treatment of Crohn’s disease with recombinant human interleukin 10 induces the proinflammatory cytokine interferon gamma. Gut 2002;50:191–195.
June 2006
7. Steidler L, Hans W, Schotte L, et al. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science 2000; 289:1352–1355. 8. Gruzza M, Fons M, Ouriet MF, et al. Study of gene transfer in vitro and in the digestive tract of gnotobiotic mice from Lactococcus lactis strains to various strains belonging to human intestinal flora. Microb Releases 1994;2:183–189. 9. Klijn N, Weerkamp AH, de Vos WM. Genetic marking of Lactococcus lactis shows its survival in the human gastrointestinal tract. Appl Environ Microbiol 1995;61:2771–2774. 10. Steidler L, Neirynck S, Huyghebaert N, et al. Biological containment of genetically modified Lactococcus lactis for intestinal delivery of human interleukin 10. Nat Biotechnol 2003;21:785–789. 11. Hommes DW, van Deventer SJ. Endoscopy in inflammatory bowel diseases. Gastroenterology 2004;126:1561–1573. 12. Huyghebaert N, Vermeire A, Neirynck S, et al. Development of an enteric-coated formulation containing freeze-dried, viable recombinant Lactococcus lactis for the ileal mucosal delivery of human interleukin-10. Eur J Pharm Biopharm 2005;60:349 –359. 13. Best WR, Becktel JM, Singleton JW, et al. Development of a Crohn’s disease activity index: National Cooperative Crohn’s Disease Study. Gastroenterology 1976;70:439 – 444. 14. Vandenbroucke K, Hans W, Van Huysse J, et al. Active delivery of trefoil factors by genetically modified Lactococcus lactis prevents and heals acute colitis in mice. Gastroenterology 2004;127:502–513. 15. Watanabe T, Kitani A, Murray PJ, et al. NOD2 is a negative regulator of Toll-like receptor 2-mediated T helper type 1 responses. Nat Immunol 2004;5:800 – 808. 16. Gasson MJ. Plasmid complements of Streptococcus lactis NCDO
View publication stats
IL-10 SECRETING L LACTIS IN CD
17.
18. 19.
20.
759
712 and other lactic streptococci after protoplast-induced curing. J Bacteriol 1983;154:1–9. Bringel F, Van Alstine GL, Scott JR. A host factor absent from Lactococcus lactis subspecies lactis MG1363 is required for conjugative transposition. Mol Microbiol 1991;5:2983–2993. Syvanen M. Churning out safer microbes for drug delivery. Nat Biotechnol 2003;21:758 –759. Bradbury J. Beyond pills and jabs: researchers develop new ways to get drugs to the right place at the right time. Lancet 2003; 362:1984 –1985. Baker M. Better living through microbes. Nat Biotechnol 2005; 23:645– 647.
Address requests for reprints to: Maikel P. Peppelenbosch, MD, Department of Cell Biology, University Medical Center Groningen, University of Groningen, A. Deusinglaan 1, 9713 AV, Groningen, The Netherlands. e-mail:
[email protected]; fax: ⴙ31-50-363 2512. H.B. was supported by the Broad Medical Research Program. D.W.H. is Clinical Fellow of The Netherlands Organization for Health Research and Development. The Science Foundation Ireland (SFI/01/F.1/B036) partly supported L.S. and S.N. M.P. is supported by the “Maag Lever Darm Stichting.” P.R is supported by the Flanders Interuniversity Institute for Biotechnology. E.R., N.H., and J.P.R. are supported by the Research Fund of Ghent University (GOA project no. 12050700). The authors thank I. Bruggeman for technical assistance, S. Cusack for reading the manuscript, and Dr Z. Zelinkova and all nurses on F5 Zuid for their enthusiastic contribution.
