Lysophosphatidylcholines Enriched with cis and trans Palmitoleic Acid Regulate Insulin Secretion via GPR119 Receptor

Among lipids, lysophosphatidylcholines (LPCs) with various fatty acyl chains have been identified as potential agonists of G protein-coupled receptors (GPCRs). Recently, targeting GPCRs has been switched to diabetes and obesity. Concomitantly, our last findings indicate the insulin secretagogue properties of cis and trans palmitoleic acid (16:1, n-7) resulting from GPCR activation, however, associated with different signaling pathways. We here report the synthesis of LPCs bearing two geometrical isomers of palmitoleic acids and investigation of their impact on human pancreatic β cells viability, insulin secretion, and activation of four GPCRs previously demonstrated to be targeted by free fatty acids and LPCs. Moreover, molecular modeling was exploited to investigate the probable binding sites of tested ligands and calculate their affinity toward GPR40, GPR55, GPR119, and GPR120 receptors.


Molecular modeling
To keep the continuity of the methodology, all molecular docking, conventional molecular dynamics and trajectory postprocessing was performed using GNINA 1.0 1 , AMBER software 2 with a protocol described previously 3 , with the exception that ligands RESP charge fitting at a total charge of 0 site-oriented molecular docking performed after obtaining initial poses from blind docking and MMGBSA calculations with topology and trajectory stripped from the membrane after the calculations for increased accuracy.The structure of GPR40 and GPR119 receptors were downloaded from the protein data base (ID: 4PHU, 7XZ5 respectively) Receptors GPR55 and GPR120 were folded by AlphaFold2 4 .Then receptors were embedded into the membrane using CHARMM-GUI 5 .Visualization of proteinligand complexes was performed with the use of PyMOL software with the help of PLIP interaction profiler 6 .
Stability of structures during simulations was validated by a radius of gyration (Rg), root-mean square deviation (RMSD) of Cα atoms of residues employed in secondary structure formation, solvent-accessible surface area of ligand (SASA) as well as its RMSD throughout the simulations.

In silico mutagenesis
Models obtained after molecular dynamics was paste into ICM-Pro (ver.3.9) software (MolsoftLLC) where implemented module to in silico mutagenesis was used to calculate binding energy of receptor pocket.Subsequently, all amino acids involved in ligand -protein interaction was replaced with alanine.Then new binding energy (ΔGbind mut) was calculated and compared to wild type protein binding energy (ΔGbind wt).The difference in the binding free energies was calculated with formula: ΔΔGbind = ΔGbind mut -Δgbind wt .
Mutation of Phe157 and Trp238 in both analyzed models caused the biggest change in binding energy.It leads to the conclusion that these two amino acid residues are the most important for ligand binding in the case of LPCs.Qian et al. also indicated that Trp238 is the most important residue but Phe157 was not highlighted. 7Probably Phe157 is necessary in ligand stabilization.Also, Phe241 mutation caused reduction of binding energy in both models because of its nonpolar side chain which stabilizes ligand inside the binding pocket.For both models, Glu261 and Trp265 have significant importance in ligand-protein interaction.Moreover, the mutation of amino acids located at the entrance to the binding pocket in the GPR119-LPC(16:1(9E)) model lowers binding energy (Fig S9).This observation leads to the conclusion that all these mentioned amino acids play a crucial role in ligand binding.Data represent the means ± SD from at least 3 simulations.

Calcium flux measurements
To perform intracellular Ca 2+ measurements ([Ca 2+ ] i ), EndoC-βH1 cells were seeded onto coated 96-well plates (4 x 10 4 cells per well) in 100 µl of culture medium and incubated for 24 h under optimum conditions.On the following day, the culture medium was substituted with Ca5 buffer containing 2 mM or 20 mM glucose and based on the given protocol from Screen QuestTM Fluo-8 No Wash Calcium Assay Kit, [Ca 2+ ] i was measured.Additionally, to examine plausible membrane permeability caused by perused compounds, propidium iodide (PI) was added (final concentration of 1 µg/ml) just before running the experiment in the microplate reader.Changes in fluorescence measurements for calcium flux (Ex/Em 490/520 nm) and PI intercalation (Ex/Em 535/617 nm) were monitored upon the addition of LPCs.Fluorescence measurements were corrected with reference to background fluorescence.

Glucose stimulated insulin secretion (GSIS)
EndoC-βH1 cells were seeded onto a coated 24-well plate at the density of 2.5 × 10

siRNA transfection preceding GSIS
To find out if GPR119 gene silencing affects GSIS, EndoC-βH1 cells were seeded onto a coated 24-well plate at the density of 1 × 10 5 cells/well in 1 ml of culture medium and incubated at 37 °C for 24 hours.Then the medium was substituted with a medium containing siRNA (25 nM) and transfection reagent (0,4%), without antibiotics according to the transfection protocol provided by the producer.After 48 hours GSIS procedure proceeded as described above.
Figure S11.Insulin secretion by EndoC-βH1 cells after 48 hours of transfection with siRNA directed to silence GPR119 expression.

cAMP synthesis measurements
Analysis of stimulated cAMP synthesis was performed using a Cyclic AMP ELISA Kit based on the supplier's protocol.EndoC-βH1 cells were seeded onto 6-well plates (2 x 10 6 cells per well) and cultivated for 48 h under optimum conditions.Subsequently, the standard culture medium was replaced with Ca5 buffer with 2 mM glucose and incubated for 1 h under standard conditions.Next, the buffer was changed to a fresh one containing 20 mM glucose.Fatty acids, with a final concentration of 10 µM, and IBMX, with a final concentration of 1mM, were also added.Cells were again incubated for 30 minutes under the same conditions.Accumulated cAMP concentration was measured in cell lysates and recalculated with respect to the protein content of cells, which was previously measured using Bradford Protein assay.

Safety statement
No unexpected or unusually high safety hazards were encountered.

Fig S1 .
Fig S1.Root mean square deviation of Cα atoms contributing to secondary structure formation.Small shifts are observable during the displacement of α-helices not embedded inside the membrane.

Fig S5 .
Fig S5.Structure of GPR119 from different angle with binding pocket indication (yellow arrow).A -presentation of a biding pocket cavity from the extracellular side.B -side view of GPR119 with the end of biding pocket.Carrangement of seven transmembrane alpha helices in GPR119.D -extracellular view of GPR119 with the direction of binding pocket entrance.

Fig
Fig S9.Binding energy changes in mutants where amino acids located in receptor pocket are replaced with alanine.
Cells were incubated for 30 minutes at 37 °C.200 µl of sample was collected and mixed with BSA at the final concentration of 0.1%.The buffer was then substituted with a new portion of Ca5 buffer supplemented with 20 mM glucose and tested compounds and/or antagonists were added at the same concentrations.Cells were incubated for 30 minutes once again.200 µl sample was collected again and mixed with BSA.Further on, the adherent cells were lysed with the addition of 0.1 M HCl (100 µl/well).Collected samples were used to evaluate insulin content via competitive enzyme-linked immunosorbent (ELISA) assay and the amount of secreted insulin was normalized to the protein content of the respective cell lysates measured by Bradford Protein assay.