Antiproliferative Activity of α-Tocopherol, γ-Tocopherol and Tocotrienols and Their Drug Interactions Evaluated Using Loewe and Chou–Talalay Models in HeLa and MCF-7 Cancer Cell Lines

Abstract

Background: Food rich in tocopherols (T) and tocotrienols (T3) are considered functional due to their ability to reduce oxidative stress and modulate anti-viability and pro-apoptotic pathways with anticancer potential; however, their efficacy differs between T and T3 and among isoforms (α and γ) likely due to differences in intracellular uptake and, consequently, in the activation of anticancer signaling pathways. To address these isoform-dependent differences, HeLa and MCF7 cancer cell lines were used to assess the antiproliferative activity of α-tocopherol (αT), γ-tocopherol (γT) and tocotrienols (Tocomin) as well as their pharmacological interactions according to Loewe and Chou–Talalay models. Methods: The tocol profile of the commercial mixture of T3 (Tocomin) was quantified by normal-phase HPLC. HeLa, MCF7, and ARPE-19 cells were cultured in DMEM supplemented with 10% FBS and exposed to αT, γT, or Tocomin (50–800 µg/mL; DMSO vehicle) for 48 h; viability was measured by the MTT assay and EC50 values were obtained from log(dose)–response fits (n = 3). Fixed-ratio (1:1) combinations were evaluated in HeLa and MCF7, and interactions were quantified using Loewe additivity and Chou–Talalay combination indices, supported by isobologram analysis. Results: Tocomin showed greater potency with αT and γT, and synergy with αT/γT; however, the combination of αT + γT showed antagonism in both cell lines. Conclusions: The higher potency of Tocomin and its synergistic interactions with αT or γT suggest that tocotrienol-rich mixtures may enhance the antiproliferative response, whereas combining αT and γT together may reduce efficacy under the tested conditions.

1. Introduction

Breast and cervical cancers remain major causes of cancer-related mortality in women, and dietary factors have been implicated as modifiable contributors to cancer risk and progression [1]. In this context, nutraceuticals such as vitamin E have attracted attention due to their antioxidant and anti-inflammatory properties and their potential to modulate tumor-related pathways [2]. Vitamin E comprises the tocol family, including tocopherols (T) and tocotrienols (T3), each present as α, β, γ, and δ isoforms; notably, α- and γ-tocopherol are predominant dietary isoforms [3]. While tocopherols have been widely studied for effects on growth and apoptosis-related signaling (kinase pathways and regulators such as BAX and p53), increasing evidence suggests that tocotrienols may exert stronger antiproliferative and pro-apoptotic activities in cancer models [4].

Importantly, the biological efficacy of tocols depends not only on intrinsic activity but also on isoform-dependent uptake, distribution, and metabolism. Tocols partition into lipid bilayers, and their cellular entry can be facilitated by lipid-transport mechanisms associated with membrane receptors and lipoproteins [5]. Systemically, preferential retention of αT is influenced by hepatic α-tocopherol transfer protein, whereas other isoforms and tocotrienols are more susceptible to CYP-dependent catabolism, potentially limiting intracellular accumulation and altering signaling outcomes [6]. These differences raise a key gap: commercial mixtures rich in tocotrienols and combinations of isoforms may yield non-additive effects, yet direct comparisons under the same experimental conditions and quantitative interaction analyses remain limited. We hypothesized that vitamin E isoforms differ in antiproliferative potency in HeLa and MCF-7 cells and that selected combinations exhibit non-additive interaction patterns. The research question addressed was as follows: Do α-tocopherol, γ-tocopherol, and tocotrienols differ in potency, and do their combinations deviate from additivity under a fixed-ratio design?

Therefore, this study evaluated the antiproliferative activity of αT, γT, and a tocotrienol-rich commercial mixture (Tocomin), alone and in fixed-ratio combinations, in HeLa and MCF-7 cell lines. Interactions were characterized using Loewe additivity and the Chou–Talalay method. Loewe provides an additivity reference for agents assessed on the same endpoint (cell viability/proliferation), whereas Chou–Talalay quantifies interactions via the combination index (CI) across effect levels, classifying them as synergistic, additive, or antagonistic. Using both models on the same dataset strengthens interpretation and reduces model-dependent conclusions.

2. Materials and Methods

2.1. Samples and Reagents

This study used pure standards (αTocopherol, γTocopherol) from Sigma-Aldrich (St. Louis, MO, USA) and a commercial extract of tocopherols and tocotrienols (Tocomin) purchased from a commercial store (Tocomin SupraBio, Arlington Heights, IL, USA).

Dimethyl sulfoxide (DMSO), Dulbecco’s modified Eagle’s medium, fetal bovine serum (FBS), 3-(4,5-dimethylthiazol-2-yl) 2,5 diphenyltetrazolium bromide (MTT), sodium chloride (NaCl), sodium hydroxide (NaOH), trypan blue, trypsin-EDTA, sodium phosphate salts were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Non-essential amino acids, L-Glutamine (200 mM), penicillin–streptomycin (Gibco, Vienna, Austria) and isopropyl and isobutyl alcohol were obtained from JT-Baker (Avantar Performance Materials, Gliwice, Poland).

2.2. Quantification and Characterization of Commercial Extract of Tocopherols and Tocotrienols

The quantification of individual tocols in commercial extract Tocomin was performed by normal-phase HPLC (Perkin Elmer model 200, Waltham, MA, USA) according to the chromatographic conditions and methodology of Stevens-Barron et al. [7]. Tocomin and pure standards were mixed with 1 mL of hexane (HPLC grade), filtered with syringe membrane filters nylon (0.45 µM) and placed in HPLC vials wrapped in aluminum foil. The equipment for tocol identification and quantification consisted of HPLC equipment a LiChrosorb Si 60 (5 M, 25 0.4 cm) normal-phase column and a fluorescence detector at 285 nm excitation and 325 nm emission wavelengths and the experimental conditions were as follows: isocratic, the glow rate was 1.0 mL/mi, isopropanol:hexane (0.9:99.1 v/v) as mobile phase. The results were expressed in µg tocols/g of Tocomin. Representative chromatograms are provided in the Supplementary Materials (Supplementary Figure S1).

2.3. Cell Cultures

Human malignant cell lines MCF-7 (breast adenocarcinoma; ATCC HTB-22; RRID:CVCL_0031) and HeLa (cervical carcinoma; ATCC CCL-2; RRID:CVCL_0030), and the non-cancerous ARPE-19 cell line (retinal pigment epithelial cells; ATCC CRL-2302; RRID:CVCL_0145) were kindly provided to our laboratory by Dr. Salvador Enrique Meneses Sagrero (Department of Agriculture and Livestock, Universidad de Sonora, Hermosillo, Sonora, Mexico). According to the provider, these cell line stocks were originally obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).

Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS), non-essential amino acids, L-glutamine (from a 200 mM stock solution), penicillin (100 U/mL), and streptomycin (100 μg/mL). Cultures were maintained at 37 °C in a humidified atmosphere of 5% CO₂ and 95% air and grown in 25 cm² tissue-culture flasks.

2.4. Proliferation Inhibitory Activity of Pure Compounds and Extracts

The cells were removed from the culture dish with trypsin-EDTA and seed into 96-well plates (100 µL/well). The individual pure compounds (αT, γT) and the commercial extract Tocomin were dissolved in DMSO to prepare a stock solution. Stock solutions of αT, γT, and Tocomin were prepared in 100% DMSO, and working solutions were obtained by dilution in complete culture medium immediately before use, where the final DMSO concentration was maintained constant at 0.1% (v/v) in all experimental conditions (vehicle control, single compounds, and combinations). For combination experiments, a fixed 1:1 ratio was applied on a mass basis (µg/mL), meaning equal µg/mL contributions of each compound at each combination dose level in the microplate (0, 50, 100, 200, 400 and 800 µg). The cell viability assay with the MTT reagent was performed according to Stevens-Barron et al. [7], and 50 µL of DMEM with stock solution was added and incubated for 48 h based on standard practice for viability/antiproliferative assays in adherent cancer cell lines to capture measurable growth inhibition [7]. After the incubating period, the cells were washed with 100 µL of phosphate-buffered solution (PBS), and then 100 mL of DMEM and 10 mL of MTT (5 µmg/mL) were added to each well and left to incubate for 4 h more. The formazan crystals were solubilized with isopropyl alcohol, and the absorbance of the samples was read at 570 nm in a UV-Vis microplate reader (BioRad Benchmark Plus, Hercules, CA, USA) with a reference wavelength of 630 nm. Cell survival was expressed as a percentage, where the optical density of cells treated with DMSO alone was considered 100% survival.

To establish baseline cytotoxicity and perform an initial evaluation of selectivity under the same exposure settings, ARPE-19 retinal pigment epithelial cells were used as a non-malignant reference model. HeLa and MCF-7, the main cancer models of interest for this investigation, were the only cells used in combination experiments and interaction analysis. As a result, the vitality of ARPE-19 cells was not assessed under combination treatments. Half-maximal effective concentrations (EC50) were estimated by nonlinear regression using a sigmoidal (four-parameter logistic) dose–response model rather than linear regression across the full dose range (% cells survival and the logarithm of the dose) [7].

Potency (EC50/EC50): Potency was defined as the concentration required to reduce viability to 50% (EC50). EC50 values were estimated from log(dose)–response curves by linear fitting in the corresponding range and reported as mean ± SD (n = 3). To compare potency between treatments, the relative change (fold-change) was calculated as the ratio between the EC50 of the comparator treatment and the EC50 of the treatment, such that values > 1 indicate greater potency of the tocotrienols (Equation (1)) [8].

$$Potency={\displaystyle \frac{EC50 TREATMENT}{EC50 COMPARATIVE TREATMENT}}$$

(1)

2.5. Experimental Design and Pharmacological Interaction Between Two Compounds

Drug interactions were evaluated using a fixed 1:1 (w/w) concentration ratio (µg/mL). Tocomin is a mixture and the isoforms differ in molecular weight. We used a mass-based ratio for standardization; evaluation of molar ratios and response-surface mapping will be addressed in follow-up studies. For each pair, the fixed-ratio mixture was prepared and tested as a serial dilution across multiple total concentrations (25 + 25, 50 + 50, 100 + 100, 200 + 200 and 400 + 400 µg/mL), enabling interaction assessment across several effect levels (Fa) specifically at the 1:1 ratio [9].

These experiments were carried out on two cell lines: HeLa and MCF7. The pure compounds were weighed and diluted in DMSO as a vehicle (µg/mL), according to the EC50 of each compound and Tocomin extract. The individual dose–response curve was measured using Emax model that describes the dose–response relationship using a saturation function according to the next equation (Equation (2)):

$$E(c){\displaystyle \frac{Emax}{EC50+C}}$$

(2)

The E(c) values are decisive in predicting the effect observed at the maximum concentration used in the trial, where Emax indicates the maximum possible effect and EC50 the concentration at which 50% of the effect is achieved [10].

Interaction analysis was conducted using the Loewe model based on the principle of additivity to evaluate interactions between αT + γT, αT + Tocomin, and γT + Tocomin, determining whether the combination of the two compounds produces a greater, lesser, or equal effect than the sum of the individual effects of each compound.

We calculated the effects of each molecule separately on cell viability at concentrations of 50, 100, 200, 300, 400 and 800 µg/mL. After the individual effects were obtained, we evaluated the combined effects of αT + γT, αT + Tocomin, and γT + Tocomin at the same ratio (1:1) with the following concentrations: 50 + 50, 100 + 100, 200 + 200, 400 + 400 µg/mL. Once the effects of each concentration on cell viability were calculated, a dose–response curve was constructed showing the relationship between the concentration of the molecule and the observed effect (alone and in combination).

The expected effect (Ee) of the combination of two molecules (αT and γT, αT + Tocomin and γT + Tocomin) was calculated by adding the individual effects [11], according to the equation (Equation (3)):

Expected effect (Ee): EA + EB

(3)

where EA is the pharmacological effect of molecule A at the specific concentration and EB is the pharmacological effect of molecule B at the specific concentration (EC50). The values of the maximal effect by concentration in the experiment (400 µg/mL) were used to compare. In a 1:1 interaction study, if the effects are expected to be additive and simplicity is sought, the Loewe model is the appropriate one [12,13].

The calculation of the interaction index according to the Loewe (ICL) model [14] was performed using the following equation (Equation (4)):

$$ICL={\displaystyle \frac{CA}{{EC}_A}}+{\displaystyle \frac{CB}{{EC}_B}}$$

(4)

where CA and CB are the concentrations of drugs A and B in the mixture, and ECA and ECB are the EC50 values of drugs A and B used alone. An interaction index equal to 1 indicates additivity, whereas an ICL<1 indicates that the doses of A and B required to achieve a given effect in combination are lower than those expected under additivity and therefore indicate synergy (conversely ICL > 1 indicates antagonism). Based on the Loewe additivity analysis, we also calculated the Chou–Talalay combination index [15] using the following equation (Equation (5)):

$$ICC-T={\displaystyle \frac{d1}{Dx1}}+{\displaystyle \frac{d2}{Dx2}}$$

(5)

where d1 and d2 are the doses of compounds 1 and 2 in combination that achieve a given effect level (50% effect level) and Dx1 and Dx2 are the doses of each compound alone that produce the same effect level. Based on these results, an isobologram was constructed according to Loewe additivity method to visualize the expected additivity line and the observed interaction between compounds.

3. Results

3.1. Tocopherol and Tocotrienol Content from Tocomin Extract

Table 1. Quantification of tocols from Tocomin extract.

Tocol Isoform	Mg/g Oil	Mg per Capsule
αT	0.07 ± 0.00	17.84 ± 0.01
αT3	0.12 ± 0.01	23.35 ± 0.97
βT3	0.01 ± 0.00	1.198 ± 0.08
γT3	0.05 ± 0.00	10.34 ± 0.01
δT3	0.06 ± 0.00	11.84 ± 0.08
Total tocols	0.31 ± 0.01	64.57 ± 8.34

Values are expressed as milligrams per gram of oil and milligrams per capsule corresponding to a commercial presentation of Tocomin.

shows the profile of tocopherol isoforms in Tocomin, where the most abundant isoform was αT3 (0.12 mg/g of oil), followed by αT. No other tocopherol isoforms such as β, γ, and δ were detected; therefore, Tocomin can be considered an enriched sample of tocotrienols. Representative chromatograms (elution profiles), including overlays, are provided in Supplementary Figure S1.

3.2. Proliferation Inhibitory Activity from Pure Compounds Alone and Combined on HeLa and MCF7 Cancer Cells

3.2.1. Individual Effect of αT, γT and Tocomin on Cancer Cell Viability

Figure 1 shows micrographs of the cultures in both cell lines under the combined treatments of αT + Tocomin at 1:1 (400 + 400 µg/mL) and γT+Tocomin at 1:1 (400 + 400 µg/mL). Treatment duration (48 h) was selected based on standard practice for viability/antiproliferative assays in adherent cancer cell lines to capture measurable growth inhibition, and a dedicated time-course optimization was not conducted in the present study and should be addressed in future work.

Representative micrographs are shown for selected concentrations to illustrate the observed phenotype, and the images at 25 µg/mL were not included because they were visually indistinguishable from the concentration of 50 µg/mL (no apparent morphological effect), and images at 800 µg/mL were not included because they reflected near-complete loss of viable cells, both consistent with the minimal and maximal-effect condition. Importantly, quantitative MTT dose–response analyses include all tested concentrations (25–800 µg/mL) as shown in Figure 2 and

Table 2. EC50 from pure tocopherols and a commercial mixture of tocols on HeLa and MCF7 cancer cells.

HeLa Cells
Viability (%)
Compound	EC50 µg/mL	25 µg/mL	50 µg/mL	100 µg/mL	200 µg/mL	400 µg/mL	800 µg/mL
αT	381.6 ± 2.58	94.6 ± 5.91	93.14 ± 5.93	75.24 ± 18.62	49.09 ± 4.63	32.18 ± 3.75	30.1 ± 3.70
γT	133.3 ± 2.21	85.1 ± 0.90	79.72 ± 9.31	61.10 ± 7.45	33.11 ± 4.42	18.90 ± 3.68	12.6 ± 2.10
Tocomin	104.17 ± 2.01	83.7 ± 3.60	78.57 ± 5.99	33.68 ± 5.19	31.40 ± 1.92	30.48 ± 5.18	27.8 ± 3.81
MCF7 cells
Viability (%)
Compound	EC50 µg/mL	25 µg/mL	50 µg/mL	100 µg/mL	200 µg/mL	400 µg/mL	800 µg/mL
αT	253.17 ± 2.40	98.04 ± 3.19	92.55 ± 6.40	90.95 ± 8.30	52.29 ± 16.6	35.10 ± 3.70	44.4 ± 10.3
γT	169.43 ± 2.22	98.80 ± 5.75	99.56 ± 6.00	77.45 ± 2.00	39.37 ± 2.70	14.28 ± 1.70	13.2 ± 1.10
Tocomin	79.92 ± 1.90	59.11 ± 2.90	86.47 ± 2.10	34.20 ± 2.80	31.72 ± 5.10	24.83 ± 3.30	23.2 ± 3.60

Values presented as µg/mL of alpha-tocopherol (αT), gamma-tocopherol (γT), and tocopherols extract (Tocomin) on cell viability, mean and standard deviation ± SD (n = 3).

. Micrographs for all tested concentrations, including 25 and 800 µg/mL, are provided in Supplementary Figure S2; the main figure shows selected concentrations for readability.

A notable effect is observed in the morphological changes that determine cell viability, such as loss of confluence and cell rounding, indicative of cell death [16]. However, in the ARPE cell line, no such cell death effects were observed; on the contrary, cell confluence was maintained across the concentrations of the molecules alone; ARPE cell viability was not evaluated in the combined treatments as in the HeLa and MCF7 cells.

Figure 2 shows the dosage response curves for αT, γT, and Tocomin in MCF7 (left) and HeLa (right) cells. In both lines, a dose-dependent decrease in viability was observed; Tocomin is the treatment with the greatest cytotoxic effect, followed by γT, while αT has the least effect in monotherapy. As a result, the Tocomin combinations increased the anti-viability activity, with the combination αT + Tocomin and γT + Tocomin, being more effective than the combination αT + γT, suggesting a potentiator effect attributed to Tocomin in both cell lines.

Tocomin showed higher efficacy in both cell lines (EC50 HeLa: 104.17 ± 2.01 µg/mL and EC50 MCF7: 79.92 ± 1.90 µg/mL), according to the EC50 values displayed in Table 2. This may indicate that Tocomin was 1.28 times more effective than γT (133.3 ± 2.21 µg/mL) and 3.66 times more potent than αT (381.6 ± 2.58 µg/mL) in HeLa. Tocomin was 2.12 times more potent than γT (169.43 ± 2.22 µg/mL) and 3.17 times more potent than αT (253.17 ± 2.40 µg/mL) in the MCF7 cell line. Tocomin consistently decreased viability to 33.68 ± 5.19 µg/mL (HeLa) and 34.20 ± 2.80 µg/mL (MCF7) at 100 µg/mL, while αT (75.24 ± 18.62 µg/mL and 90.95 ± 8.30 µg/mL) and γT (61.10 ± 7.45 µg/mL and 77.45 ± 2.00 µg/mL). Consequently, γT showed decreased residual viability even though Tocomin was the most effective at high dosages with a lower EC50. Therefore, although Tocomin was the most potent at high doses with a lower EC50, γT showed lower residual viability at 400 µg/mL (HeLa: 18.90 ± 3.68 µg/mL; MCF7: 14.28 ± 1.70 µg/mL), suggesting a greater maximum effect at high concentrations compared to Tocomin.

3.2.2. Pharmacological Interactions from Combined Compounds on HeLa and MCF7 Cancer Cells

Table 3. Maximum possible effect (Emax), median effective concentration (EC50) and observed effect (E©) of compounds alone and combined on HeLa and MCF7 cancer cells.

HeLa
	Emax	EC50	E©
αT	67.82	213.47	0.318
γT	81.1	133.3	0.608
Tocomin	69.52	104.2	0.667
αT + γT	84.15	395.46	0.21
αT + Tocomin	76.26	51.05	1.49
γT + Tocomin	79.46	73.99	1.07
MCF7
	Emax	EC50	E©
αT	64.9	235.17	0.276
γT	85.72	169.43	0.506
Tocomin	75.17	79.92	0.941
αT + γT	81.24	476.36	0.17
αT + Tocomin	83.92	43.06	1.94
γT + Tocomin	69.73	99.05	0.70

shows the maximum efficacy parameters (Emax), which indicate how much the compound can bring the cell to a state of death or non-viability, which is seen as the maximum achievable cytotoxicity, where HeLa γT showed the highest maximum efficacy (Emax 81. 1) and a higher potency than αT (EC50 133 vs. 213.5), while Tocomin was the most potent (EC50 104.2) with a comparable Emax of 84.15. On the other hand, in MCF7 monotherapy, the pattern remained the same, with αT being the least potent (EC50 235.17; Emax 64.9), while γT was more effective (Emax 85.72) with intermediate potency (EC50 169.43). Tocomin was also the most potent with an EC50 of 79.92 and an Emax of 75.17.

The observed effect (Ec) is essential for comparing the strength of the actual effect under a reference condition. The results observed in the combinations, αT and γT, in the HeLa cell, increased the maximum efficacy (Emax 84.15) with a marked loss of potency (EC50 395.46) and a lower observed effect (Ec 0.21), consistent with competitive behavior. In contrast, αT and Tocomin extremely reduced the EC50 (51.05), increasing the observed effect (Ec 1.49), with a supra-additive phenomenon being observed under the experimental conditions. In γT and Tocomin, potency also improved compared to the individual compounds. In the MCF7 line, the combination of αT and γT also showed a loss of potency (EC50 476.36) and a low observed effect (Ec 0.17), unlike the combination of αT and Tocomin, where the combination was more effective with an Emax of 83.92, an EC50 of 43.06, and an Ec of 1.94. It was different in HeLa, where γT and Tocomin did not improve performance in MCF7 (lower Emax 69.73 and Ec 0.70), which does suggest dependence on the cellular environment.

Table 4. Expect effect of combination, combination effect, and interaction index from aT, gT and tocomin combined on viability of HeLa and MCF7 cells.

HeLa
	Expected effect of combination (EA + EB) (1:1) 400 µg/mL	Combination effect (1:1) 400 µg/mL	Interaction index according to the Loewe model (ICL)		Interaction index according to the Chou–Talay model (ICC-T)
αT + γT	51.07	52.5	2.4	Antagonism	2.4	Antagonism
αT + Tocomin	62.65	26.07	0.36	Sinergy	0.52	Sinergy
γT + Tocomin	49.36	23.23	0.52	Sinergy	0.54	Sinergy
MCF7
	Expected effect of combination (EA + EB) (1:1) 400 µg/mL	Combination effect (1:1) 400 µg/mL	Interaction index according to the Loewe model (ICL)		Interaction index according to the Chou–Talay model (ICC-T)
αT + γT	53.99	72.82	2.41	Antagonism	4.83	Antagonism
αT + Tocomin	59.03	23.65	0.36	Sinergy	0.72	Sinergy
γT + Tocomin	31.46	30.47	0.91	Sinergy	1.82	Sinergy

Values expressed in µg/mL, the expected effect of the combination and the real effect of the combination are compared at the same concentration, and the interaction index has been established with the Loewe and Chou–Talay model. Values >1 indicate antagonism, =1 additivity and <1 indicate synergism.

shows two levels of evidence of interaction for the 1:1 condition. First, the direct comparison of Expected vs. Observed indicates whether the mixture produces lower viability (stronger effect) or higher viability (weaker effect) than predicted. Second, the Loewe (ICL) and Chou–Talalay (ICC-T) models formalize this interaction (CI < 1 synergy, CI ≈ 1 additivity, CI > 1 antagonism). In HeLa, αT + γT showed a slightly less potent effect than expected in terms of viability (Expected 51.07 vs. Observed 52.5), and both indices indicated antagonism (ICL = 2.4; ICC-T = 2.4). In contrast, αT + Tocomin and γT+Tocomin reduced viability more than expected and showed synergy for both models (αT + Tocomin: ICL = 0.36; ICC-T = 0.52; γT+Tocomin: ICL = 0.52; ICC-T = 0.54). In MCF7, αT + γT showed more marked antagonism than in HeLa (ICL = 2.41; ICC-T = 4.83). αT + Tocomin showed synergy (ICL = 0.36; ICC-T = 0.72), although less intense according to Chou–Talalay than in HeLa. For γT + Tocomin, Loewe suggested behaviors close to additivity (ICL = 0.91), while Chou–Talalay indicated antagonism (ICC-T = 1.82), evidencing a discrepancy between models. This discrepancy may occur because Loewe and Chou–Talalay are based on different assumptions, and Chou–Talalay’s CI depends heavily on the shape of the dose–response curves (slope m) and the effect level (fa) [17,18]. Overall, Tocomin tends to enhance the effect in HeLa; in MCF7, the pattern is more dependent on the model and suggests less robust interaction for γT + Tocomin.

Figure 3 shows the isobologram, where the behavior between lines can be observed. The combination of isoforms αT + γT showed antagonism in both lines, stronger in MCF7. The response observed in cell viability and its possible anticancer effect reflects not only intrinsic potency but also the effect of effective intracellular exposure, determined by the uptake of isoforms in the membrane, partitioning, and transport associated with serum proteins. The observations in the combination of αT + γT could indicate competition between isoforms for delivery processes via vesicular trafficking routes, which reduces the effective fraction of each compound in membranes and organelles, manifesting as antagonism, mainly in MCF7. However, combinations with Tocomin (rich in tocotrienols) could benefit from more efficient cellular delivery in the presence of serum albumin and/or complementary metabolic targets, generating a greater than expected reduction in viability: synergy. Interaction outcomes may be ratio-dependent; therefore, the synergy/antagonism patterns reported here should be interpreted as specific to the 1:1 fixed-ratio design and exposure conditions. Confirmation using additional ratios or full dose–matrix response-surface designs is required to generalize these interaction patterns.

4. Discussion

In this study, MCF-7 and HeLa were selected as well-established in vitro models of breast and cervical cancer, respectively, allowing direct comparison with previous studies on antiproliferative agents and drug interactions. ARPE-19, a non-tumor epithelial cell line, was included as an initial reference to evaluating selectivity under comparable experimental conditions and as evidence that the concentrations used in the study were not cytotoxic to healthy cells and did not interfere with the antiproliferative results of the combination. We recognize that normal counterparts with compatible tissue (MCF-10A and normal cervical epithelial models) would further reinforce translational relevance and are planned for future work.

Our interaction analysis was performed using a single fixed 1:1 concentration ratio, implemented as serial dilutions across total doses. Although this design enables interaction assessment at multiple effect levels (Fa), interaction outcomes may vary with the mixture proportion. Therefore, the reported synergy/antagonism should not be generalized beyond the tested ratio. Future studies will evaluate multiple ratios and/or full dose–matrix response-surface experiments to confirm robustness across mixture proportions.

While we observed isoform-dependent differences in antiproliferative potency and non-additive interaction patterns under the tested conditions, we did not directly measure molecular endpoints (protein or gene expression, transporter activity, or pathway flux).

Proposed mechanistic interpretations in this discussion should be considered hypothesis-driven and grounded in the prior literature rather than experimentally demonstrated in the present study. Our dataset is limited to in vitro viability outcomes and pharmacological interaction modeling (Loewe additivity and Chou–Talalay analyses) and does not include direct mechanistic validation (uptake/accumulation assays, ROS measurements, apoptosis markers, or pathway-level gene/protein expression). Therefore, references to potential involvement of membrane-associated effects or signaling pathways (NF-κB, mevalonate-related signaling, or redox buffering) are provided to contextualize the observed interaction patterns and to generate testable hypotheses for future studies, rather than to establish causality.

The relationship with isoforms and competitive cellular uptake lies in a lower IC50 for Tocomin, which has a mixture rich in tocotrienols. This would be consistent with more efficient intracellular delivery, producing a shift to the left of the dose–response curve (greater potency). In addition, it has been suggested that lipid receptors and transporters such as CD36 and SR-BI may be involved in the mass transport of vitamin E across the membrane. Overexpression of CD36 increases the absorption of alpha and gamma-tocopherol in cells, indicating that both isoforms may have common entry routes and therefore compete when they are located at the same time [19,20]. Under this explanation, combinations of isoforms can alter the effective intracellular proportion, either through saturation or transporter competition, thereby modifying the anti-viability effect without necessarily changing the intrinsic activity of the compound.

The differences in EC50 partially reflect in our results’ differences in intracellular accumulation determined by affinity and competition of uptake pathways [6], which would manifest itself in the reflected potency. In media with fetal bovine serum, bovine albumin is dominant and may skew bioavailability. There is direct evidence that bovine albumin increases cellular uptake of tocotrienols and decreases that of tocopherols, with differences between isoforms [21], Fetal bovine serum acts as a carrier and changes the route of entry. Some of the tocotrienols can enter from lipoproteins via receptors such as SR-BI, if in the medium the tocotrienols captured by albumin could decrease the flow through the lipoprotein route to SR-BI, affecting tocopherols more than tocotrienols if the latter are delivered well from the complex with albumin as proposed by Nakatomi [21], which coincides with the results obtained where Tocomin presented the lowest EC50.

In both cell lines, the tocotrienol-rich extract (Tocomin) was more potent than αT and γT. In addition to their antioxidant function, tocotrienols have a distinctive metabolic mechanism, which acts on the mevalonate cholesterol axis through post-transcriptional suppression of HMG-CoA reductase, involving growth signals and prenylation processes [22]. Signaling and apoptosis tend to differ between cells, and in the case of the MCF7 line, it has been reported that tocotrienols modulate pathways associated with proliferation and apoptosis [23,24]. On the other hand, the study of Nakatomi [21] declares that albumin favors the uptake of tocotrienols over tocopherols, which would imply that the same dose in the cell depicts more tocotrienols into the intracellular medium, which translates into a lower EC50 in Tocomin.

These results indicate that the equal γT + Tocomin combination may exhibit different performance depending on the cellular environment. In Hela cells, γT + Tocomin showed consistent behavior with an enhanced effect (lower EC50: 73.99) and a relatively high observed effect (E© = 1.07), indicating that the combination does not limit the response. In contrast, in MCF-7, γT + Tocomin had a higher EC50 (99.05), lower maximum efficacy (Emax = 69.73), and lower observed effect (E© = 0.70), suggesting that in this line, the combination does not improve and may even limit the effect. Taken together, these results indicate that the response does not depend exclusively on the compounds, but on the molecular system available in each cell (e.g., differences in apoptosis and signaling pathways). MCF7 is known to be deficient in caspase 3 due to a functional deletion of the CASP-3 gene in exon 3, which prevents the translation of the protein by itself [23,25] resulting in the execution of alternative pro-apoptotic signals such as caspase 7. Other studies by Germain et al. [26] support that MCF7 line with caspase 3 deficiency activates caspase 7 concomitantly with the cleavage of poly (ADP-ribose) polymerase (PARP) after external apoptotic stimulation, confirming that caspase 7 can execute cleavage in PARP [26,27]. However, other mitochondrial pathways are also involved; for example, Mooney et al. [24] show that PARP cleavage is observed in MCF7 and argue that caspase 9 activation is modulated by the apoptotic mitochondrial pathway, which could explain this cleavage even in the absence of caspase 3, with redundancy with the activation of other caspases such as caspase 6 [23,24].

The differences between HeLa and MCF7 may differ in the efficiency of uptake and intracellular trafficking of tocols, since the expression of lipid-associated receptors/transporters (e.g., SR-BI, CD36, and NPC1L1) and membrane composition are not identical between lines [28,29,30]. These differences may determine what fraction of γT or tocotrienols is effectively incorporated into the membrane, redistributed to target organelles, and reaches sufficient intracellular concentrations to activate cytostatic or cytotoxic pathways, directly affecting EC50 and Emax values.

The synergy of Tocomin with αT or γT could be related to the complementarity of targets, where Tocomin provides tocotrienols and a mixture of isoforms that modulate oxidative and inflammatory stress, while αT and γT cover components of membrane damage such as lipoperoxidation or nitrative stress. The antagonism observed between αT + γT was stronger in MCF7, which could be related to competition for membrane incorporation and serum carrier proteins; αT and γT compete for the same hydrophobic environment in the cell membrane and the same endocytosis pathway, which reduces the effective fraction of each, and both have antioxidant activity, which dampens redox signals; if cell death or cell cycle inhibition depends on oxidative stress and its signaling, this combined dampening may slow down the response and show antagonism between the two. Tocotrienols have reported actions in inhibiting hormonal pathways and estrogen receptors that act as mitogens, mainly in MCF7. If αT + γT does not modulate these pathways and competes in the membrane to enter the cell, the antagonism is greater.

The interaction patterns observed at the tested fixed ratio may be influenced by isoform-specific differences in cellular availability, including uptake, membrane partitioning, and intracellular delivery. Accordingly, Figure 4 presents a hypothesis-driven working model, informed by prior literature, illustrating plausible routes by which tocopherols and tocotrienols could contribute to synergistic or antagonistic outcomes under a constant mixture proportion. Importantly, the involvement of specific mechanisms (membrane organization and pro-survival signaling such as NF-κB-related signaling and the mevalonate/HMG-CoA reductase axis) was not directly assessed in the present study and should therefore be interpreted as hypothesis-generating. In addition, antioxidant context may influence oxidative stress-dependent cellular responses and could contribute to antagonistic patterns in some settings; thus, timing, dose, and cellular context are key variables to evaluate in future mechanistic and multi-ratio combination studies.

This work is intended as an in vitro, hypothesis-generating evaluation of antiproliferative activity and interaction patterns among vitamin E isoforms/mixtures. First, combination experiments were performed using a single fixed 1:1 (w/w) ratio (µg/mL). Because interaction outcomes can be ratio-dependent, the synergy/antagonism patterns reported here should be interpreted as specific to this mixture proportion and exposure conditions and require confirmation using multiple ratios and/or full dose–matrix (response-surface) designs. Second, the concentration range required to span no-effect and maximal-effect conditions in our MTT-based viability assays (25–800 µg/mL) is relatively high for in vitro studies; therefore, physiological extrapolation is limited and non-specific effects (e.g., solubility/partitioning effects typical of lipophilic compounds) cannot be fully excluded. Future studies should incorporate expanded lower concentration ranges and additional models, including tissue-matched normal cells, to strengthen translational relevance. Third, Tocomin is a mixture containing tocopherol and tocotrienol species; thus, attributing effects to individual constituents is constrained by mixture composition and analytical overlap under the chromatographic conditions used. To improve transparency, we provide chromatographic elution profiles (standards and Tocomin overlays) as Supplementary Materials; nonetheless, additional quantitative/orthogonal analytical characterization would further strengthen component-level interpretation. Finally, this study did not include mechanistic validation (WB/RT-qPCR or functional pathway assays) or in vivo testing. Consequently, mechanistic explanations and the proposed working model remain speculative, and translational relevance requires confirmation using complementary mechanistic endpoints and in vivo systems.

5. Conclusions

The different isoforms of vitamin E (αT, γT) and a commercial extract rich in tocopherols/tocotrienols (Tocomin) have differential effects on the viability of tumor cells (HeLa and MCF-7). Using both Loewe additivity and Chou–Talalay analyses, we observed that certain Tocomin-containing combinations display synergistic antiproliferative activity, whereas αT + γT showed antagonism across the evaluated conditions. Although the enhanced activity of Tocomin-based combinations is consistent with a complementary, hypothesis-driven framework—where tocotrienols may contribute membrane-related and signaling effects (NF-κB, mevalonate pathway) and tocopherols may provide antioxidant buffering—these mechanistic interpretations remain speculative because this study was limited to in vitro viability assays and did not include mechanistic validation (e.g., gene/protein expression) or in vivo testing. Future work will include targeted WB/RT-qPCR panels and complementary functional assays (apoptosis, cell-cycle analysis, oxidative stress readouts, and/or uptake measurements) to validate the proposed mechanisms.

Therefore, the translational relevance of these interaction patterns must be confirmed in tissue-matched normal models and in vivo systems, and future studies should directly test pathway modulation and the potential for tocotrienol-rich mixtures to either sensitize cells through survival-pathway inhibition or counteract ROS-dependent cell death that could potentiate the anticancer effects.