The MIN6 cell line was procured from National Centre for Cell Science (NCCS), Pune. Cell lines were maintained in Dulbecco's Modified Eagle's Medium (Sigma-Aldrich Chemicals Private Limited). Cells were passaged once weekly following detachment using trypsin/EDTA (Sigma-Aldrich Chemicals Private Limited)12. The cells were incubated at 37 ⁰C until confluent. Then, the cells were seeded at a concentration of 3.5 x 10⁵ cells/ml into a T125 cell culture flask until the flask became fully confluent. The trypan blue dye exclusion test was used to determine the number of viable cells present in cell suspension13. Cells treated with trypan blue were counted using a Countess® Automated Cell Counter.

Prior to the insulin release experiment, the cells were administered streptozotocin (STZ) to induce diabetes. The cells were incubated with a low dose of STZ (Sigma-Aldrich Chemicals Private Limited) (5 mM) for 1 hour14. The media used for incubation with STZ contained 2% FBS (Sigma-Aldrich Chemicals Private Limited) and then discarded off. After that, cells were incubated with KRBB (Sigma-Aldrich Chemicals Private Limited) for 1 hour, and the solution was discarded. STZ treatment was excluded from the positive control experiment.

Cell culture well plates were seeded with approximately 6.5 x 10⁴ cells per well in a total volume of 200 μl of media. The cells were seeded to allow two wells per treatment. The plates were incubated at 37 ⁰C under a 5% CO₂ atmosphere for 24 hours in DMEM media. The cells were treated with different molar concentrations of agomelatine (Sigma-Aldrich Chemicals Private Limited), thapsigargin (Sigma-Aldrich Chemicals Private Limited), and 2-APB (Sigma-Aldrich Chemicals Private Limited), singly as well as in combination, for 15 hours at 37 ⁰C. The treatments were categorized as follows:

Blank: containing media but no cells

Vehicle control 1: Ethanol (Sigma-Aldrich Chemicals Private Limited)

Vehicle control 2: DMSO (Sigma-Aldrich Chemicals Private Limited)

Positive control: Cells without streptozotocin treatment and cells with glipizide (Sigma-Aldrich Chemicals Private Limited) treatment

Group 1: Agomelatine

Group 2: Thapsigargin

Group 3: 2-APB

Group 4: Agomelatine + Thapsigargin

Group 5: Agomelatine + 2-APB

Group 6: Thapsigargin + 2-APB

Group 7 : Agomelatine + Thapsigargin + 2-APB

The combination ratios were kept at an equipotency ratio (i.e., ED₅₀ ratio) to ensure that each drug's contribution to the combination was approximately equal (Table 1)15. We plotted dose-response curves with the help of GraphPad Prism software16. ED₅₀ values were determined through dose-effect curves for each drug. The equipotent ratios were kept at 1:9 for agomelatine and thapsigargin, 1:90 for thapsigargin and 2-APB, 1:820 for agomelatine and 2-APB, and 0.9:9:82 for agomelatine, thapsigargin, and 2-APB together.

The MIN6 cells were treated with serial dilutions of each drug and drug combination described above. The media used for drug preparation was low glucose media with 2% FBS. Five dilutions (two-fold diluted) of each drug and combination plus controls were tested in two independent experiments with replicate samples. Chou and Talalay proposed ED₅₀ ratios as the basis for the experimental design17, 1819, 20.

Table 1

Single Drug
Drug	Concentrations
Agomelatine (nM)	200	100	50	25	12.5
Thapsigargin (μM)	4	2	1	0.5	0.25
2-APB (μM)	200	100	50	25	12.5
Two Drugs Combinations
Drug Combination 1	0.5 (EC50)	1 (EC50)	2 (EC50)	3 (EC50)	4 (EC50)
Agomelatine (nM)	22	43	87	130	173
Thapsigargin (μM)	0.2	0.4	0.8	1.2	1.5
Drug Combination 2
Thapsigargin (μM)	0.2	0.4	0.8	1.2	1.5
2-APB	18	37	74	111	148
Drug Combination 3
Agomelatine (nM)	22	43	87	130	173
2-APB	18	37	74	111	148
Three Drugs Combinations
Drug Combination 4	0.5 (EC50)	1 (EC50)	2 (EC50)	3 (EC50)	4 (EC50)
Agomelatine (nM)	22	43	87	130	173
Thapsigargin (μM)	0.2	0.4	0.8	1.2	1.5
2-APB (μM)	18	37	74	111	148

The different formulations of varying molarities of drugs and their combinations were administered, the cells incubated for 15 hours, and then the supernatant was collected for insulin quantification. The quantity of insulin was assessed by Rat/Mouse Insulin ELISA kit (Sigma-Aldrich Chemicals Private Limited), a Sandwich ELISA method. The activity of horseradish peroxidase in the presence of the substrate 3,3',5,5'-tetramethylbenzidine was used to quantify immobilized antibody-enzyme conjugates. After acidification of products, enzyme activity was evaluated spectrophotometrically by the increased absorbency at 450 nm, corrected from the absorbency at 590 nm.

The three-drug combination was carried out17, and data were analyzed using the CompuSyn computer software package. The drugs and their combined data were analyzed to determine whether there were significant difference sbetween treatments using analysis of variance (ANOVA). Post hoc analysis was done using Dunnett's test to compare the treatment groups with controls, in order to identify the treatments with the strongest effects. One-way ANOVA and Dunnett's test was carried out using Minitab.

Table 2

S.No	Exposure Group and insulin level	Concentrations
1	Agomelatine (nM)	200	100	50	25	12.5
	Insulin level (ng/mL)	28.857	29.704	28.914	28.920	28.879
2	Thapsigargin (μM)	4	2	1	0.5	0.25
	Insulin level (ng/mL)	27.868	28.339	27.829	26.432	25.525
3	2-APB (μM)	200	100	50	25	12.5
	Insulin level (ng/mL)	25.761	26.691	26.113	26.167	25.821

Table 3

S.No.	Drugs combinations (μM)	CI values
1.	Agomelatine + Thapsigargin
(i)	1.673	0.08704
(ii)	1.33	0.04140
(iii)	0.887	0.03978
(iv)	0.443	0.01943
(v)	0.222	0.00974
2.	Thapsigargin + 2-APB
(i)	149.5	0.05615
(ii)	112.2	0.17927
(iii)	74.8	0.10518
(iv)	37.4	0.20425
(v)	18.2	0.02882
3.	Agomelatine + 2-APB
(i)	148.173	2.57705
(ii)	111.13	0.04467
(iii)	74.087	0.05587
(iv)	37.043	7.02E-5
(v)	18.022	3.83E-5
4.	Agomelatine + Thapsigargin + 2-APB
(i)	149.673	1.20476
(ii)	112.33	3.73327
(iii)	74.887	6.62564
(iv)	37.443	0.67536
(v)	18.222	0.64944

Standard curves for insulin absorbance were constructed using calibration samples of known concentration, and the R-squared values of the curves were used as a data analysis tool to correlate insulin level and absorbance. The cells were exposed to ethanol at 200 μM and DMSO at 4 μM as vehicle controls to check for any effect of these on insulin excretion from the cells. These solutions were found to have no effect on insulin secretion. DMSO containing the compound of interest at a known concentration was diluted with media and added to wells. After 15 hours, the media was collected for insulin content analysis.

In order to determine synergism or antagonism between drugs, it is essential to know the "potency" and the "shape" of the dose-effect curve for each drug. The ED₅₀ of each medication was calculated in the previous sections, and the constant combination ratio experiment was conducted at an equipotency ratio [e.g., ED₅₀¹/ED₅₀² ratio] to ensure that the contributions of each drug's effects to the combination were nearly equal. The dose limits were defined from the single drug test to encompass concentrations below and above the ED₅₀ values of each compound.

Figure 1

Figure 2

Figure 3

The fa–CI plots of binary and ternary combinations for insulin ELISA are shown in Figure 1. The CI value versus fa (fraction of insulin secretion with respect to the control) is shown as fa-CI plot. The CI values for the different drugs at different dose levels are also shown in Table 3. Agomelatine and thapsigargin showed synergism across nearly the whole range of dose levels. However, in terms of insulin secretion, thapsigargin and 2-APB showed a stronger synergism than the above group. The combination of agomelatine and 2-APB at a high dose of 148.173 μM concentration showed antagonism, while at a low dose, it appeared to be synergistic. The ternary combination also induced dual antagonistic and synergistic effects, being antagonistic at 149.673 μM, 112.33 μM, and 74.887 μM, but synergistic at dose levels below 37 μM.

The interaction of two and three medication combinations through quantitative values of CI at f_a = 0.9 shows that agomelatine and 2-APB yielded synergistic effects, but agomelatine + thapsigargin and thapsigargin +2-APB were particularly antagonistic to one another in regards to insulin secretion by MIN6 cells. The concept of the polygonogram comes not from mathematical derivations but rather from its practical utility17. To some extent, the polygonogram can also be used to project (or predict) the outcomes of drug combinations even when the tests have not yet been completed21 (Figure 2).