Structure-based design of a Cortistatin analogue with immunomodulatory activity in models of inflammatory bowel disease

Ulcerative colitis and Crohn’s disease are forms of inflammatory bowel disease whose incidence and prevalence are increasing worldwide. These diseases lead to chronic inflammation of the gastrointestinal tract as a result of an abnormal response of the immune system. Recent studies positioned Cortistatin, which shows low stability in plasma, as a candidate for IBD treatment. Here, using NMR structural information, we design five Cortistatin analogues adopting selected native Cortistatin conformations in solution. One of them, A5, preserves the anti-inflammatory and immunomodulatory activities of Cortistatin in vitro and in mouse models of the disease. Additionally, A5 displays an increased half-life in serum and a unique receptor binding profile, thereby overcoming the limitations of the native Cortistatin as a therapeutic agent. This study provides an efficient approach to the rational design of Cortistatin analogues and opens up new possibilities for the treatment of patients that fail to respond to other therapies.

b. Binding curves and IC50 values for Cortistatin analogues with respect to Somatostatin receptors 1-5 (SSTR1-SSTR5), using a Radio-ligand Binding Assay (Supplementary Table 1). All analogues are studied at identical concentrations and compared to the same SST-labeled reference. Data is represented as mean values (n=2), error bars represent the 95% Confidence Interval of the curve fitting. Effects of various concentrations of Cortistatin (CST) and analogue 5 (A5) on the production of inflammatory cytokines and nitric oxide by LPS-activated mouse peritoneal macrophages (a) and on cell proliferation and production of Th1-type cytokines by anti-CD3-activated mouse spleen cells (b). c,d. Synthesized analogues 2, 3 and 4 (A2, A3 and A4) did not significantly affect the inflammatory response of mouse Raw 264 macrophages (c) or T-cell activation of mouse spleen cells (d). Data are the mean ± SEM of three independent experiments (n=3), each performed as duplicates. Dots (in panels c,d) represent individual values of biologically independent cell cultures. Statistical significance between groups was assessed by paired, two-tailed Student's t-test. ***p<0.001, ****p<0.0001 vs. stimulated cells in the absence of peptides. Exact p-values are shown for p>0.001. Source data are provided as a Source Data file. Supplementary Fig. 3. Tolerogenic effect of analogue 5 but not of anti-TNFα antibodies on DSSinduced ulcerative colitis. Comparative therapeutic effects in a model of relapsing-remitting colitis by treatments with analogue 5 (A5) and anti-mouse TNFα antibody only during the first cycle of DSS exposure (see scheme for experimental design). Clinical signs were evaluated by disease activity indexes (scoring body weight loss, stool consistency and presence of fecal blood), survival rate and macroscopic signs of colon inflammation (damage score, length and weight of colon). Animals injected with saline instead of A5 were used as untreated colitic mice. n=8 mice/group. Data are mean ± SEM with dots representing individual values of biologically independent animals. Statistical differences between groups were calculated using two-tailed non-parametric Mann-Whitney test (for disease activity index and colon damage score), unpaired two-tailed Student's t-test (for colon length and weight) and Kaplan-Meier test (for survival). ***p<0.001; ****p<0.0001 versus untreated DSS-colitic mice (saline). Exact p-values are shown for p>0.001. NS, not significant. Source data are provided as a Source Data file. Administrations: s.c., subcutaneous; i.v., intravenous.
Supplementary Fig. 4. Protective and curative effects of analogue 5 on TNBS-induced acute colitis. Acute colitis was induced in mice by intrarectal TNBS and the animals were then treated with analogue 5 (A5) through different routes of administration following both protective (a) and curative (b) strategies as indicated in the schemes. Animals injected with saline instead of A5 were used as untreated colitic mice. Disease evolution and severity were monitored by survival and weight loss, colitis score and macroscopic colon damage score. n=8 mice/group. Data are mean ± SEM with dots representing individual values of biologically independent animals. Statistical differences between groups were calculated using two-tailed non-parametric Mann-Whitney test (for colon damage and colitis scores), unpaired two-tailed Student's t-test (for body weight) and Kaplan-Meier test (for survival). ***p<0.001; ****p<0.0001 versus untreated TNBS-colitic mice (saline). Exact p-values are shown for p>0.001. Source data are provided as a Source Data file. Administrations: s.c., subcutaneous; i.v., intravenous; p.o., oral.

Supplementary Fig. 5. Tolerogenic effect of analogue 5 but not of treatments of reference on TNBS-induced chronic colitis.
Chronic colitis was induced in mice by intrarectal injections of increasing doses of TNBS once a week and the animals were then treated with analogue 5 (A5), antimouse TNFα antibody or mesalazine only during the first week as indicated in the scheme. Animals injected with saline instead of A5 were used as untreated colitic mice. Disease evolution and severity were monitored by survival and weight loss, colitis score and macroscopic colon damage score. n=8 mice/group. Data are mean ± SEM with dots representing individual values of biologically independent animals. Statistical differences between untreated (saline) and treated groups were calculated using two-tailed non-parametric Mann-Whitney test (for colon damage and colitis scores), unpaired two-tailed Student's t-test (for body weight) and Kaplan-Meier test (for survival). Exact p-values are shown. Source data are provided as a Source Data file. Administrations: s.c., subcutaneous; i.v., intravenous; p.o., oral.
Supplementary Fig. 6. Treatment with analogue 5 regulates Th1-and Th17-induced inflammatory responses and triggers regulatory T cells in TNBS-induced colitis. a. Mice (n=6 mice/group) with TNBS-induced acute colitis were treated with saline or analogue 5 (A5) as indicated in the scheme. Mice injected with 50% ethanol were used as basal controls. b. Cell proliferation and cytokine production by mesenteric lymph node (MLN) cells isolated from all experimental groups in basal conditions (stimuli: -) or in response to T-cell stimulation (PMA+ConA, stimuli: +). c. Levels of IgG autoantibodies for TNBS in sera. d. Cell proliferation and cytokine production by MLN cells isolated from untreated colitic mice in basal conditions (unstimulated) or in response to ex vivo stimulation (PMA+ConA) in the absence (none) or presence of A5. e. Flow cytometry analysis of regulatory T cells in MLNs isolated from all experimental groups (results show the percentage of CD25 + FoxP3 + cells in gated CD4 + T cell population). Data are mean ± SEM with dots representing individual values of biologically independent animals. Statistical differences between groups were calculated using unpaired two-tailed Student's t-test. ***p<0.001; ****p<0.0001 versus untreated TNBS-induced colitic mice (saline) or versus untreated stimulated-MLN cells (none, panel d). Exact p-values are shown for p>0.001. Source data are provided as a Source Data file. f. Gating strategy to identify CD25+FoxP3+ regulatory T cells (Treg) in the CD4+ population of mesenteric lymph nodes isolated from colitic mice shown in Fig. 6c and Supplementary Fig. 6e. g. Gating strategy to identify the percentage of CD4+ lymphocytes producing Th1, Th17 and Treg cytokines in activated mesenteric lymph node cell cultures shown in Fig. 6b. Staining with 7-aminoactinomycin D (7-AAD) was used to exclude dead cells in all analyses. Supplementary Fig. 7. Structural properties of analogue 5, (A5). a. Amide/proton alpha (above) and amide/aromatic (below) regions of the analogue 5 ( 1 H-NOESY experiment), with peak assignments displayed. b. Comparison of the chemical shifts (full 1D spectrum of analogue 5) are shown as a dilution series of spectra recorded at 1, 0.5, 0.25, 0.125 and 0.062 mM of a peptide solution in 90%H2O/10%D2O. No differences are observed between the different datasets with the exception of the signal to noise ratio that is affected by the dilution. c. Comparison of Phi, Psi and Chi1 values of A5 to Octreotide and four SST analogues described in the literature. These SST analogues were selected for the comparison, due to their specific preferences towards different SSTRs. d. Pharmacophores described in the literature for SST analogues and for Cortistatin and A5 described in this work. The pharmacophores are depicted as distances between key residues (gamma Carbons). Distances are represented as ranges to include the distance dispersion in the ensemble of conformations.

Supplementary Tables
Supplementary Table 1 Binding curves and IC50 values for Cortistatin analogues with respect to Somatostatin receptors 1-5 (SSTR1-SSTR5), using a Radio-ligand Binding Assay (Supplementary Figure 1b). Cells were incubated in HEPES buffer pH 7.4 with the new analogues of Cortistatin at 10 different concentrations in a range of 0.1 nM to 10 µM for 2-4 h and 125 I-Tyr11 SST14 was used as radio-ligand and SST14 as cold ligand. Membranes were filtered and washed 3 times and the filters were counted to determine [ 125 I]-SST-14 specifically bound (performed at Eurofins Panlabs, Inc.). The radioactivity obtained in the absence of SST14 is considered as total binding and that obtained in the presence of 1 µM of SST14 is considered as nonspecific binding. Specific binding was considered as the difference between total and nonspecific. Biochemical assay results are presented as the percent inhibition of specific binding (% inhibition) as a mean of duplicates (n=2), and error bars represent the 95% Confidence Interval of the curve fitting. The reliability of the assay was assessed by using SST-14 reference standards that were run as an integral part of each binding assay with the five (SSTR1-SSTR5). IC50 values were determined by a non-linear, least squares regression analysis using Prism 9 (GraphPad Software, USA), using the Dose-response inhibition (three-parameters) model. Mouse peritoneal macrophages were cultured in medium (unstimulated) or stimulated with LPS (1 μg/ml) in the absence or presence of Analogue 5 (100 nM). Mouse spleen cells were cultured in medium alone (unstimulated) or stimulated with anti-CD3 antibodies (2 μg/ml) in the absence or presence of analogue 5 (100 nM). After 24 hours (macrophages) or 48 hours (spleen cells) of culture, cell viability and apoptosis were determined. The percentage of cell viability relative to the initiation of cell culture was determined using AlamarBlue assay (Thermo Fisher Scientific). The percentage of apoptotic cells was determined by flow cytometric analysis of Annexin V positive 7-amino-actinomycin D negative cell population (R&D Systems). Data are the mean ± SEM of 5 independent cultures performed in duplicates.

Supplementary Table 4. Oligos used in the determination of gene expression by real-time PCR
forward reverse Total RNA was isolated from colon sections obtained at day 10 after TNBS injection following the manufacturer's protocol (Tripure, Roche). Precipitated RNA was treated with DNase 1 (Sigma) before reverse transcription (RevertAid First Strand cDNA Synthesis Kit, ThermoFisher Scientific). SYBER green quantitative PCR (SensiFast Sybr No-Rox mix, Bioline) was performed as described in the Methods Section.

Supplementary Notes
Supplementary Note 1: NMR assignments, spectra and structure calculation statistics.