Temperature-triggered in situ forming lipid mesophase gel for local treatment of ulcerative colitis

Ulcerative colitis is a chronic inflammatory bowel disease that strongly affects patient quality of life. Side effects of current therapies necessitate new treatment strategies that maximise the drug concentration at the site of inflammation, while minimizing systemic exposure. Capitalizing on the biocompatible and biodegradable structure of lipid mesophases, we present a temperature-triggered in situ forming lipid gel for topical treatment of colitis. We show that the gel is versatile and can host and release drugs of different polarities, including tofacitinib and tacrolimus, in a sustained manner. Further, we demonstrate its adherence to the colonic wall for at least 6 h, thus preventing leakage and improving drug bioavailability. Importantly, we find that loading known colitis treatment drugs into the temperature-triggered gel improves animal health in two mouse models of acute colitis. Overall, our temperature-triggered gel may prove beneficial in ameliorating colitis and decreasing adverse effects associated with systemic application of immunosuppressive treatments.


Reversibility of the transition.
Small-angle X-ray scattering (SAXS) was used to determine the lipid phase, and thus the reversibility of the transition. Source data are provided as a Source Data file.

Amplitude sweep experiments.
A stress-controlled rheometer (Modular Compact Rheometer MCR 72 from Anton Paar, Graz, Austria) was used in cone-plate geometry, 0.993° angle and 49.942 mm diameter. The temperature control was set either at 25 or 38 °C. An amplitude sweep was performed at 1 Hz between 0.002 and 100% strain to determine the linear viscoelastic regime (LVR), the yield and flow points.

Phase transition identification after in vivo applications.
Healthy animals were administered with 100 mL of TIF-Gel and either the excreted gel (with stool after 30 min) or the residual gel present in the colon after 6 h was collected and analyzed by SAXS (the animal was sacrificed, colon harvested and the residual gel washed 3x with PBS before analysis). As shown in Fig. S3, the Bragg reflections characteristic of L phase were present before administration at 25 °C, whereas the gel excreted with the stool showed the L Ia3d transition. Moreover, the lamellar phase absorbed heat and water during the experiment reaching a cubic (pn3m) phase as was already observed in the in vitro investigations.

Drug homogeneity into the gel structures.
Both drugs are dissolved into the gel matrix and they do not form crystals once incorporated into the lipidic gel (at least at the drug concentrations used in this study), as proven by the absence of reflections associated with a drug crystallization in the WAXS spectra at high q (see Figure S4, panel a). We carried out additional experiments to assess whether both drugs were homogeneously distributed into the gel matrix. To determine this, the gel (loaded with TAC or TOFA) was prepared as described in the manuscript and transferred into a 2 mL Eppendorf tube. The tube was centrifuged and kept at rest for 24 hours. Subsequently, the gel was divided into 3 different layers (Top, Middle, and Bottom), and the drug content evaluated in each. As shown in Figure S4 (panel b), each layer contains the same drug amount, confirming that TOFA and TAC were homogeneously distributed.

S6
Gels' phase identity with 10 % w/w amounts of drugs.
Measurements were performed on a Bruker AXS Micro, as described in the main text. MLO was used as the lipid component of the mesophases and mixed with weighed amounts of drugs (10% w/w) in sealed Pyrex tubes and alternatively centrifuging (10 minutes, 5000 g) several times at room temperature until a homogenous mixture was obtained. The mesophase was then equilibrated for 48 h at room temperature. were quantified. Mouse numbers are the same as in main text figure 3. No statistically significant differences in percentages were observed between the different treatment groups. Data were analyzed using one-way ANOVAs.
DCs, dendritic cells. Error bars are ± SEM. The gating strategy used is also depicted. Source data are provided as a Source Data file.

In vivo/ex vivo experiments to evaluate the adhesion of the TIF-Gel to the colon wall.
For in vivo adhesion testing, healthy animals (n= 11) received an enema of 100 l DiR (1,1'-dioctadecyl-3,3,3',3'tetramethylindotricarbocyanine iodide) loaded gel (DiR-TIF-Gel) under anesthesia as described in the experimental section. Animals were sacrificed after 30 minutes (

HPLC method: Tacrolimus
Tacrolimus (TAC) was detected by reverse-phase liquid chromatography using a Macherey-Nagel Nucleosil 100-5 C18 (4.0 x 250 mm; 5.0 µm particle size) column. The mobile phase consisted of methanol/water (80:20 v/v) + 0.1% trifluoroacetic acid at a flow rate of 1 mL/minutes, temperature 50 °C and UV detection at λ = 214 nm. An internal standard (ketoconazole, 20 µg/mL) was added to each sample to correct for inter-injection variation and UV detection at λ = 278 nm. Data were collected and analyzed using the software Chromeleon 7 (Thermo Fisher).

Stability Study of TAC and TOFA
The stability of the drugs -TOFA and TAC -was monitored over one month. At specific time points, an aliquot of the formulation was analyzed at the HPLC, and the content of the drug recorded. Data are expressed as relative percentage referred to day 0.

Dead volume of the syringe and cannula
To calculate the dead volume of the syringe 1 ml: Injekt®-F (Fine Dosage) Luer Solo (Luer Slip) (Braun) with the canula (size 20G, L × diam. 1.5 in. × 1.9 mm), the syringe was filled with different amount of formulation MLO + MilliQ water (84% lipid and 16% water) and the amount that came out of the syringe was recorded.
Supplementary figure 11. Calibration curve used to calculate the amount of formulation that has to be loaded into the syringe to obtain exactly 100 mg of formulation from the syringe with the cannula. Source data are provided as a Source Data file.

LC-MS/MS analysis
Samples, standards and QC were extracted by protein precipitation and analyzed by LC-MS/MS using the following method. For plasma samples, 10µL of plasma was mixed with 25 µL of precipitation solution (80:20 Acetonitrile: Methanol + 0.1 µM loperamide). The samples were centrifuged at 10000g for 10 minutes and 20µL of supernatant was diluted with 40µL of H2O+0.1%FA. The samples were centrifuged at 3400rpm for 10 minutes and 50µL of supernatant was diluted with 100µL of H2O+0.1%FA. All the samples were analyzed by LC-MS/MS (Shimadzu prominence HPLC coupled to an AB/SCIEX 4000 QTRAP) in positive MRM mode. The samples were separated on a Cortecs RP shield column (3x50mm 2.6u) using a fast gradient of 10 mM ammonium formate in water (A) and methanol (B). The gradient starts at 20%B and increases to 98%B in 2 minutes, hold for 0.5 minutes and equilibrates for 1.4 minutes. The MRM parameters were optimized for each analyte; the MRM transition 313. to 149.3 was selected for TOFA, 822.3 to 770.1 for TAC and 477.1 to 266.0 for loperamide (internal standard). The samples were quantified using a calibration curve prepared in matrix using the area ration of analyte to internal standard.