Solubility prediction and cytotoxicity of lipidic microemulsion: HSPiP modeling and synergism with ciclopirox olamine

Abstract

Cutaneous cancer is a constantly progressive health challenge as life-threatening in men and women around the world. This study addressed repurposed ciclopirox olamine (COA) through the utilization of different essential oils limonene (LIM), eugenol (EU), and olive oil (OLO) as natural lipids possessing innate anticancer potential. HSPiP software provided preliminary screening of oils and solvents based on the theoretical solubility of COA. Placebo microemulsions (ME1-ME3) were prepared after optimizing S_mix at varied ratios (surfactant to co-surfactant ratio) and evaluated for size, zeta potential, and polydispersity index (PDI). EU, LIM, and OLO (as microemulsions) were used to assess anticancer potential against A431, A375, and B16-F10 cell lines. Moreover, HSF (human fibroblast cells) was used to assess the safety of these excipients at the same concentrations. It was required to carry out concentration-dependent cytotoxic potential of EU and LIM against A431 and B16-F10 as compared to COA. Finally, an apoptosis study was conducted to understand the mechanistic perspective of EU and LIM against A431 and B16-F10. Results showed an excellent correlation (r² = 0.97) of the predicted theoretical solubility of COA in the explored excipients to the experimental solubility data. UV scanning of the COA-solubilized oils negated any chemical interactions. ME1-ME3 showed tween-80 (10.9–21.6%) dependent size reduction (510–339 nm) and PDI values (0.28–0.31). In the cytotoxicity study, EU, LIM, and OLO elicited concentration-dependent cellular killing. LIM was found to be relatively sensitive to A431 and B16-F10 compared to EU. Notably, LIM resulted in dramatically aggressive induced early and late apoptosis in both cell lines at the explored IC₅₀. Thus, the topical nanoformulations of COA containing anticancer essential oils can offer a promising approach for treating cutaneous melanoma.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-18141-4.

Introduction

Among all types of cancer worldwide, melanoma stands as the most lethal skin malignancy, characterized by its high aggressiveness and significant mortality rate. Skin cancer is one of the most common cancer among people. In 2022, 1.5 million new cases (330,000 new cases of melanoma) were diagnosed, whereas 60,000 people died from cutaneous cancer¹. Melanoma occurs commonly in men than women, and the highest incidence cases were observed in Australia, New Zealand, Western Europe, North America, and Northern Europe. Based on the population explosion in the world and incidence rate pattern of 2020, 510,000 new cases (a roughly 50% increment) of melanoma per year and 96,000 deaths (a roughly 68% increment) are expected by 2040². Regular exposure to sun light causes cutaneous cancer incidence and death. This can be correlated to the joint statement by the WHO (World Health Organization) and International Labour Organization as “Nearly 1 in 3 deaths occurs from non-melanoma skin cancer by working under the sun”³.

Chemotherapy is commonly used to treat squamous cell carcinomas with or without radiation or surgery. Cutaneous cancer rarely spreads to systemic cancer. 5-fluorouracil (5-FU), cisplatin, carboplatin, and paclitaxel are usually used as a standalone or in combination therapy (carboplatin with paclitaxel). High doses of toxic chemotherapy are associated with multiple side effects (alopecia, nausea, vomiting, loss of appetite, diarrhoea, and gastrointestinal distress) and long-term toxicities (bone marrow toxicity and skin toxicity)⁴. Moreover, surgery and chemotherapy approaches are required to manage early and late stages of melanoma, respectively. Detrimental post-surgery and chemotherapy result in inevitable consequences. Furthermore, conventional modes of therapy (surgery, radiation, and chemotherapy) have challenged the efficacy of chemotherapeutic agents due to the emerging resistance and adverse effects. Thus, scientists have gained interest in developing an alternative strategy with high effectiveness at low cost and minimum toxicities in the treatment of melanoma⁵. Natural therapy as a standalone or in combination with the standard chemotherapeutics can be a promising strategy to treat skin cancer and related complications. Several natural oils have been explored to have potential in vitro and in vivo anti-melanoma activities (models). VOSviewer analysed the co-occurrence of the keyword “melanoma (essential oil as treatment)” so far (2015–2025) (Fig. 1).

Fig. 1.

VOSviewer resulted in 60 publications with 3 clusters, 32 items, 386 links, and 1326 total links analysed on 17th January 2025 using PubMed central (occurrence of the keyword: melanoma “essential oil as treatment”).

Frankincense essential oil exhibited dose (300 and 600 mg/kg body weight) and time-dependent reduction in cell viability (B16-F10 and FM94) without toxicity to normal cells (normal human epithelial melanocytes) through downregulation of Bcl-2 and Bax cascades⁶. Natural terpenes are well-known for their potential ability as anticancer terpenoids due to tumour growth inhibition (citral-rich lemongrass oil), antioxidant, anti-proliferative, cell cycle trafficking (carvacrol), anti-mutagenic, inducing apoptosis in cancer cells, and modulating signalling pathways^7–9. These may be attributed to major constituents in essential oils as terpenes (limonene, carvacrol, linalool, taxol, thymoquinone, α-pinene and so on). Pharmacologically, these differ in terms of diverse mechanisms of action at the molecular level (interrupted signalling pathways, ion channels interaction, and cascade inhibitions)¹⁰. Pavithra et al. reported a detailed review on “essential oil to prevent and treat skin cancer”. Eugenol, limonene, carvacrol, perillyl alcohol, α-santalol, geraniol, α-pinene, terpinen-4-ol, camphene, β-elemene, β-caryophyllene, citral, thujone, zerumbone, and menthol are well-explored anti-melanoma monoterpenes. The authors provided an up-to-date and concise review on the efficacy of essential oils and its components as anti-melanoma and as anti-non-melanoma agents¹¹. Eugenol, orange oil, and limonene were reported as anticancer essential oils against A-431, A-375, HSF, and B16-F10. In the study, we addressed few essential oils as potential excipients to tailor microemulsion. These may execute a beneficial role in ameliorating cutaneous cancer due to their unique innate properties and biocompatibility. The selection of these essential oils was based on the established properties and therapeutic potentials. These were screened due to their potential bioactives with innate anticancer potential, possessing multi-model mechanisms of action, and safer alternatives compared to synthetic excipients. The studied essential oils are known to have flavonoids, terpenes, and phenolic compounds, responsible for inducing apoptosis and anti-proliferation of skin cell lines, reducing cancer progression by preventing ROS (reactive oxygen species) generation, and reducing release of inflammatory mediators^12,13. The essential oils may enhance dermal permeation of lipophilic drugs such as ciclopirox for a synergistic detrimental effect if laden in microemulsion. The oils with innate anticancer potential may offer reduced side effects, drug dose, dose-related toxicity, reduced drug resistance, and avoid unnecessary introduction of excipients in the patient’s body.

Microemulsions are thermodynamically stable oil (ensuring longer shelf-life and resistance to phase separation) in water type of biphasic drug delivery system with unique properties over vesicular systems, such as liposomes and niosomes¹⁴. The lipid-based vesicular systems are related to limited physical and chemical stability under storage conditions. The drug leakage from the vesicular lipid bilayer and limited permeation across the stratum corneum are the challenging issues for developing a stabilized nanoformulation¹⁵. On the other hand, microemulsions with nanoscale globules with high surface area are capable of delivering lipophilic drugs maximally (increased fluidity and high drug diffusion by disrupting stratum corneum) across stratum corneum and stable than the vesicular system¹⁶. Microemulsions are capable of solubilizing hydrophobic drugs due to an isotropic blend of oil, surfactant, and co-surfactant, forming nanodroplets. Moreover, the essential oils with innate anticancer potential may exert a synergistic impact for the management of cutaneous cancer. No report has been published to incorporate ciclopirox olamine in microemulsion employing biocompatible essential oil possessing innate anticancer potential so far. Thus, the studied systems offer high drug solubilisation, improved stability, ease in scale up, and higher skin delivery of ciclopirox olamine compared to niosomes and liposomes, making them an ideal carrier for cutaneous delivery.

Materials

Ciclopirox olamine (99%, COA) was procured from Sigma Aldrich, USA. Various food grade oils such as eugenol (≥ 98% pure, Sigma Aldrich, USA), clove oil (100% pure with density of 1.04 g/mL at 25 °C, Sigma Aldrich, USA), orange (≤ 100 pure, Merck, USA), lemon (≥ 89% FCC), olive (containing ≤ 85% oleic acid, ThermoFisher Scientific, USA), coriander (< 100%, Sigma Aldrich, USA), bergamot peel oil (< 100%, Sigma Aldrich, USA), and anise (98%, ThermoFisher Scientific, USA) were purchased from a local medical shop (Riyadh 11451, Kingdom of Saudi Arabia). Buffer reagents (for preparing buffer solution) were obtained from in-house chemical warehouse of College of Pharmacy, King Saud University, Riyadh. Distilled water was used from our laboratory (double distillation unit). For sterilized water, water for injection was used in the experiment wherever required (cytotoxicity study). A d-limonene oil was procured from Getchem, Co. Ltd., China. Analytical grade methanol and acetonitrile were purchased from Carolina Chemical Corp, North Carolina, the United State of America. Various cell lines [A375 (ATCC^® CRL-1619), B16-F10, HSF, and A431] were purchased from AlMokattam, Nawah Scientific, Cairo, Egypt.

Methods

HSPiP-based predicted Hansen solubility parameters (HSP) and theoretical solubility of COA

HSPiP is a predictive software for the theoretical solubility of a solute in a particular solvent or a polymer in a solvent or the miscibility of two or three immiscible solvents. It is based on the principle of total innate energy present in a material or solvent. Hansen solubility parameters are dispersion energy, polarity, and H-bonding energy, expressed as δ_d, δ_p, and δ_h, respectively. Mathematically, the total energy is the sum of the square of the individual Hansen parameter, as shown in Eq. (1).

1

In the Hansen sphere, Ra and Ro are the Hansen space parameter and the Hansen sphere radius, respectively¹⁷. The program categorized all materials (surfactants, solvents, and polymers) into two classes. These are bad or good solvents. Relative energy difference (RED) was used to identify the right solvent for a solute (like a drug) or polymer for a drug loading. Therefore, RED was defined as in Eq. (2).

2

Thus, a solvent or polymer with RED < 1 is considered as “good” and vice versa¹⁷. A solute is supposed to be suitably soluble in a solvent possessing HSP values close to the HSP values of the selected drug. For this, SMILE file was used to estimate HSP values for the solute and the solvents (oils and solvents). The software provides various tabs for the estimation of the solvent or solute properties using the same SMILE data. To simulate the predictive solubility data, it was imperative to conduct the experimental solubility values in the predictive excipients. Finally, a correlation was established using a solver program.

Experimental solubility of COA in the predicted excipients

The experimental solubility was carried out in the predicted excipients (solvents and oils) at 40 °C. In brief, an accurately weighed amount (150 mg) of COA was added to a glass vial previously containing excipient (5 mL). Methanol, ethanol, acetonitrile, eugenol, d-limonene, orange oil, and tween-80 were used for the study. Each excipient was individually used in a separate glass vial. The mixture was placed inside a water bath shaker previously maintained at a fixed temperature (40 ± 1 °C) and shaking speed (200 rpm)¹⁸. The assembly was allowed to operate till 72 h to get an equilibrium. Then, the mixture was removed and centrifuged to settle down the undissolved drug. The supernatant was used to estimate the dissolved drug in the investigated excipient using UV Vis spectrophotometer (UV-1601PC, Shimadzu, Japan) at 304 nm¹⁹. The study was replicated to get average and standard deviation values.

Selection of S_mix ratio for microemulsion

In general, oils cannot be readily dispersed into aqueous media used for cell growth. Therefore, ME1-ME3 were prepared to ensure uniform exposure to the cell lines. Several pseudo ternary phase diagrams were constructed using varied ratios of tween-80 to PEG400. Oils (EU, LIM, and OLO) were selected based on the drug solubility and biocompatibility. For each oil, four different ratios of tween-80 to PEG400 (1:1, 2:1, 1:2, and 1:3) were evaluated. The aqueous phase containing a blend of tween-80 to PEG400 (S_mix) at a particular ratio (1:1, 2:1, 1:2, and 1:3), was added to the organic phase (oil without drug) under constant stirring (500 rpm) and temperature (40 °C) (hot magnetic stirrer)²⁰. In slow titration, oil-to-S_mix ratios ranging from 1:9 to 9:1 were employed, achieving phase diagrams with varied ME zones. A formulation with a maximum delineated zone of microemulsion and the optimal S_mix concentration was selected. The desired MEs (ME1, ME2, and ME3 at S_mix ratio of 1:2) were further probe-sonicated (QSONICA, Newtown, CT, USA), resulting in isotropic microemulsions. The microemulsions without sign of instability were selected for further investigation.

Preparation of placebo microemulsions

For the cell line study, it was imperative to develop oil-in-water microemulsions employing the selected oil and the S_mix at 1:2. Briefly, the selected oil was emulsified in an aqueous phase containing the S_mix. Three microemulsions were formulated as ME1, ME2, and ME3 using eugenol (EU), limonene (LIM), and olive oil (OLO), respectively. ME1 contains 11.2% of EU, 21.6% of tween-80, and PEG400 (10%). ME2 was comprised of LIM (11.2%), tween-80 (10.9%), and PEG400 (8.6%). Similarly, ME3 was formulated using 11.2% of OLO, 17.5% of tween-80, and 10.2% of PEG400. Emulsification was carried out using the probe sonication method using ON/OFF cycle of 5 sec.²¹. Several microemulsions were prepared with varying oils of different types (EU, LIM, and OLO) at constant experimental conditions (power and sonication time). The composition was varied depending upon the required dose of COA for topical delivery. Tween-80 was selected due to high HLB (hydrophilic-lipophilic balance) value (~ 14) for the oil-in-water type of microemulsion preparation. The addition of PEG400 as a co-surfactant offers several advantages in microemulsion preparation. The rationale behind adding PEG 400 was to minimize the total content of the surfactant needed for getting stable microemulsion. The co-surfactant lowers the interfacial tension more effectively than a surfactant alone for forming a stable microemulsion. The blend of both increases the interfacial film flexibility for stabilized droplets without phase separation even at low content of tween-80²². Higher content of tween-80 can result in cellular toxicity (or skin irritation)²³. Employing PEG400 with tween-80 assists maintaining stability while mitigating adverse effects related to excess concentration of tween-80²⁴. For developing microemulsions, PEG400 may contribute improved solubility capacity of hydrophobic drug like ciclopirox olamine, if laden within microglobules.

Characterizations of microemulsions

Each microemulsion was evaluated for globular size, zeta potential, and polydispersity index using a Malvern Zetasizer (Nano ZS Malvern Zetasizer, UK). The Zetasizer works on the principle of dynamic light scattering (DLS) wherein the mean diameter is recorded. Notably, the sample was diluted 100-fold with distilled water before analysis. It was required to avoid interference in analysis. For the assessment of zeta potential, a U-shaped capillary sample holder was used. The sample (undiluted) was taken into a syringe, and the capillary was filled with the sample without leaving any air space. The sample was wiped off from the sample holder surface or electrode area to avoid any electrical spark. The study was replicated to get a mean and standard deviation.

Two concentrations quick test: preliminary screening

In vitro anticancer activities against A375, A431, and B16-F10 cell lines

The A375 and A431 are two human skin melanoma cell lines. Similarly, B16-F10 is a mouse melanoma cell line. Thus, these three cell lines were exploited to investigate the anticancer potential of ME1, ME2, ME3, and COA. Therefore, these were transformed into the microemulsions using surfactant (tween-80) and co-surfactant (PEG400) without anticancer potential at the explored concentration. To investigate the cytotoxicity of the explored excipients, human skin melanoma (A375) cell lines were cultured, and the cells were maintained in the DMEM (Dulbecco’s Modified Eagle’s media) media supplemented with streptomycin (100 mg/mL), penicillin (100 units/mL), and heat-inactivated fetal bovine serum (10%). The culture suspension was prepared in the sterilized water for injection to avoid any contamination. The culture was grown in humidified (5% v/v) and CO₂ enriched conditions at 37 ± 1 °C^25,26.

Cytotoxicity study (SRB assay, sulforhodamine B assay) was carried out for ME1, ME2, and ME3 against A375. A fixed volume of the culture (5 × 10³ cells/100 µL) was transferred into 96 well-plates. The growth media (100 µL) was transferred into the wells followed by incubation for 24 h at 37 ± 1 °C. The growth media contained varied concentrations of the test sample (microemulsion). Then, the cells were fixed by replacing the growth media with 150 µL of TCA (trichloroacetic acid) solution (10%) and incubated for 1 h at 4 °C. After incubation, the TCA solution was removed, and the treated cells were washed thrice with sterilized distilled water. Now, an aliquot of 70 µL of SRB solution (0.4%) was added into it for incubation in the dark room (10 min). Then, plates were again washed with 1% acetic acid and dried overnight (air-dried). Finally, 150 µL of TRIS solution (10 mM) was added to precipitate protein-bound SRB stains. Each plate was individually analysed and the absorbance was measured at 540 nm using a microplate reader (Infinite F50 microplate reader, TECAN, M200, Switzerland). The same procedure and experimental conditions were applied for A431 and B16-F10 cell lines. The percent viability was expressed by considering viable cells without treatment as 100%. Thus, the relative % value was estimated and reported for each treatment.

In vitro cytotoxicity study against normal human skin fibroblast cell line (HSF)

ME1 (11.2 µg/mL), ME2 (0.05 µM), and ME3 (0.05 µM) contained oil as a major component to be investigated. The selected oil may cause toxicity to the normal cells. Therefore, it was imperative to negate its toxicity to the normal tissue by performing a cytotoxicity assay using HSF cells. The procedure and the experimental condition were adopted as per the aforementioned sections. COA suspension was used as the control sample for comparison. The study was replicated (n = 3, ± SD)²⁷.

Concentration dependent cytotoxicity study of ME2, ME3, and COA suspension against A375 (human) and B16-F10 (mouse)

Based on preliminary tests, A375 and B16-F10 were the selected cell lines as the most sensitive melanomas (human and mouse) to ME2 and ME3. Therefore, the concentration-dependent cytotoxicity study was carried out against the same cell lines under the same experimental conditions except for concentration. Different concentrations were prepared for ME2, ME3, and COA. The experiment provides calculations for IC₅₀ values.

Apoptosis study

A431 and B16-F10 cancer cell lines were cultured separately at 37 °C in a fully humidified incubator with 5% CO₂. The cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (Cultilab, Campinas, SP, Brazil), along with penicillin (100 units/mL) and streptomycin (100 µg/mL) to prevent microbial contamination. All experiments were conducted in replicates to ensure reproducibility.

The test samples were freshly prepared in a sterilized water-based microemulsion prior to treatment. Both cell lines were exposed to the respective samples for 24 h. Following incubation, the cells were washed with 500 µL of phosphate-buffered saline (PBS) per well, harvested (including any non-adherent or suspension cells), and stained for apoptosis analysis. The cells were incubated with 2 µM Annexin V-FITC (Sigma-Aldrich, St. Louis, USA) and 10 µM propidium iodide (PI; Life Technologies, Carlsbad, USA) in PBS for 30 min at 37 °C in the dark to prevent photobleaching. After staining, the samples were analysed using a FACScalibur flow cytometer (Becton Dickinson, San Jose, CA), and data were processed and interpreted using FlowJo software²⁸.

Statistical analysis

Statistical analysis was carried out using a GraphPad Prism 7.0 software (a trial version). The experiments were triplicated to get a mean and standard deviation (± SD). A value was considered to be significant at p < 0.05. A good correlation was established by comparing the actual and the predicted values using correlation coefficient r² ˃ 0.97. ANOVA (analysis of variance) was applied for comparison using repeated measures for a normally distributed data within the same group and corrected the data using Turkey test. Similarly, the mean of two different groups were compared using on way ANOVA with Dunnett test. For the non-normally distributed data, the Kruskal-Wallis test was used for correction.

Results and discussion

HSPiP-based predicted Hansen solubility parameters (HSP) and theoretical solubility of COA

The software (HSPiP) predicted various solvents, oils, and surfactants for the choice of excipients required for lipid-based nanoformulation development. Table 1 summarized HSP values of the explored excipients for comparative analysis against the drug. It is apparent that hydrophobic COA can be maximally solubilized in the excipient possessing HSP values near to it. Moreover, RED values less than unity, as an indication of good solvent, of eugenol and d-limonene are the lowest among them. Therefore, it can be recommended as a “good” oil for maximum solubility. This may be attributed to unique chemical structure and dispersion force as a prime driving factor. The δ_d values of eugenol and d-limonene are 18.5 MPa^1/2 and 17.7 MPa^1/2, respectively, which are closely related to the drug (18 MPa^1/2). Moreover, the program predicted various binary and ternary solvent mixtures with RED value < 1 as presented in Table 1. Thus, eugenol and d-limonene were expected to be miscible at a 1:1 ratio whereas the inclusion of water into it, resulted in a swift increase in RED value (from 0.46 to 0.73) due to greater δ_h (16 MPa^1/2) and δ_p (42 MPa^1/2) values of water. Chemically, this can be explained based on its structure and functional groups present in the drug, d-limonene, and eugenol as shown in figures A–E. Ciclopirox is associated with one hydrogen bond donor count (HBD) and 02 hydrogen bond acceptor counts (Fig. 2A and B). In contrast, its salt (olamine) was synthesized to improve its HBD (03) and HBA (04) counts for improved solubility through hydroxyl (-OH) and amine (-NH₂) groups as the prime interactive forces (Fig. 2B). HBA counts are predominantly interacting with eugenol as compared to limonene (due to lack of HBD and HBA counts) (Fig. 2C and D). The benzene ring of eugenol and the pyridine ring of COA are responsible for solubilization through the π-π interaction whereas –OH groups are involved in H-bonding-based solubility (Fig. 2E). Therefore, COA was anticipated to be relatively more soluble in eugenol than d-limonene. D-limonene is an optically active terpene with low molecular weight (136 g/mol) and boiling point (169 °C) as compared to eugenol (BP = 272.5 °C and molecular weight of 164 g/mol as predicted in HSPiP).

Table 1.

Various parameters of HSP using HSPiP software.

	δd (MPa1/2)	δp (MPa1/2)	δh (MPa1/2)	δt (MPa1/2)	MVol (cc3/mol)	ψRED	Solubility (mg/mL)
COA	18	2.1	12.1	21.7	185.9	-	-
Methanol	14.7	12.3	22.3	29.9	41.1	3.6	5.9
Water	15.5	16	42.3	47.6	18	1.9	0.021
Acetonitrile	15.3	18	6.1	24.3	53.4	0.6	4.4
Eugenol	18.5	5.6	9.5	21.5	156	0.8	25.9
d-limonene	17.7	1.9	3.2	17.1	161.5	0.7	24.9
Olive oil*	16.9	0.6	4.2	17.08	915–950	1.7	19.3
Tween-80	16.7	6.5	9.4	26.4	1237	-	12.9
	HSPiP suggested binary /ternary systems
Limonene + water: 7:3	16.7	6.1	15.7	23.71	-	0.64
Limonene + eugenol:1:1	18.1	4.7	8.7	20.63	-	0.46
Limonene + eugenol + water: 4:3:3	17.2	7.8	18.3	26.29	-	0.73
	SMILE Text
COA	CC1 = CC(= O)N(C(= C1)C2CCCCC2)O
Methanol	CO
Water	OH
Acetonitrile	CC#N
Limonene	CC1 = CC[C@@H](CC1)C(= C)C
Eugenol	COC1 = C(C = CC(= C1)CC = C)O
Olive oil	-

*[^{29, 30}]; ^ψRED = The relative energy difference (Ra/Ro) where Ro = 9.3.

Fig. 2.

Chemical structures of ciclopirox olamine, limonene, and eugenol. (A) Ciclopirox olamine and hydrogen bonding donor/acceptor counts (π-π interaction due to pyridine ring), (B) 3D ball-stick structure of ciclopirox, (C) limonene without H-bonding acceptor/donor counts, (D) eugenol with HBD/HBA (π-π interaction due to benzene ring), and (E) Hansen solubility parameters of COA (ciclopirox olamine) based on chemical structure.

Figure 3 portrays various Hansen space parameters and Hansen spheres of different ratios of limonene, eugenol, and water. Moreover, eugenol and d-limonene are GRAS (generally regarded as safe excipients) excipients with volatile behaviour. It was imperative to understand the mixing mechanism and evaporation behaviour in the combination with or without water. Figures 3A–D demonstrates Hansen’s sphere of COA (green sphere at the centre) relative to the position of eugenol, limonene, and water, respectively, at studied volume ratio (ϕ₁ ≈ 0.1–0.9) as discussed earlier. HSPiP predicted eugenol and limonene immiscible at all ratios except ϕ₁ ≈ 0.26–0.39. The binary system of eugenol and d- limonene at ϕ₁ ≈ 0.5 was predicted as partially miscible (Fig. 3E). In this context, both excipients are selected as a binary system at a volume ratio of 0.3 to get a stable organic system for ferrying the solubilized form of COA. Moreover, the eugenol + water or limonene + water system is suitably predicted as immiscible due to an obvious reason (high free energy of change between water and the oil) (Fig. 3E). Notably, this was required for fabricating oil-in-water type of nanoemulsion/microemulsion to deliver COA. Due to the volatile nature of these oils, their relative rate of evaporation was predicted as a function of temperature and Ra (relative energy distance) as shown in Fig. 3F. At ϕ₁ ≈ 0.3, the total theoretical evaporation rate was quite low (< 20%) in the eugenol/limonene + water binary system. Conclusively, a combination of the explored oils at ϕ₁ ≈ 0.3 can be stable and suitable for lipid-based micro-and nanoemulsion formulation development. However, the theoretical and the experimental solubility values of the drug were simulated and correlated using HSPiP.

Fig. 3.

(A) Hansen sphere for two solvents (limonene + water at 7:3 ratio) system (green net sphere indicates COA), (B) Hansen sphere for limonene + water binary system at 1:1 ratio (RED = 0.43), (C) Hansen sphere for limonene + eugenol + water ternary system at 4:3:3 ratio (RED = 0.73), (D) HSPiP generated miscibility curve for limonene + eugenol system exhibiting partial miscibility at the peak of the curve where ϕ₁ ≈ 0.5, (E) HSPiP generated miscibility curve for limonene/eugenol + water system exhibiting immiscibility overall volume ratio (ϕ₁ ≈ 0.1–0.9), and (F) HSPiP generated % evaporation profile (overall evaporation behaviour of the system as indicated by blue line) as a function of Ra (relative energy distance) at the explored ratio (7:3 for S1 and S2).

Experimental solubility of COA in the predicted excipients

The theoretical solubility and the experimental solubility data were generated for the drug in the predicted excipients at 40 °C. HSPiP predicted the maximum and minimum solubility values in water and eugenol as 0.021 ± 0.004 and 25.9 ± 3.7 mg/mL, respectively, whereas the experimental values were found to be 0.03 ± 0.002 and 28.0 ± 4.1 mg/mL, respectively. These values were obtained at an equilibrium phase. The solubility of the drug in others is exhibited in Fig. 4A wherein exponential values were correlated. The experimental values and HSPiP-based predicted data were found to be in good correlation (Fig. 3B). The generated correlation value was r² ≈ 0.99 (regression value)²⁰. Fig. 4B provides a detailed depiction of statistical correlation data, illustrating the solubility order of COA in various excipients as follows: eugenol (28.2 ± 1.7 mg/mL) ˃ limonene (23.7 ± 1.3 mg/mL) ˃ olive oil (17.4 ± 1.7 mg/mL) ˃ tween-80 (15.6 ± 1.4 mg/mL) ˃ methanol (6.7 ± 0.8 mg/mL) ˃ acetonitrile (3.8 ± 0.2 mg/mL) ˃ water (0.03 ± 0.001 mg/mL). The highest solubility of COA in eugenol is likely due to the close matching of Hansen solubility parameters (HSPs), with δ_d values of 18.0 for COA and 18.5 for eugenol, and δ_h values of 12.1 for COA and 9.5 for eugenol. In the case of limonene, a low δ_h value created a greater disparity between the drug and the excipient, resulting in slightly lower solubility compared to eugenol (Hansen, 2007). Limonene, a monoterpene devoid of polar functional groups (-OH, -NH₂, -C = O), exhibits weak δ_p and δ_h values due to its polarizable electron cloud. Its feeble H-bond acceptability arises from electron density around the alkene’s double bond, influenced by conjugative delocalization. These HSP values stem from diverse intermolecular forces, including π-π interactions (“=” bonds of limonene responsible for possible π-π interaction with COA rendering it slightly more polar than others), dielectric effects, double bond delocalization, London dispersion (Van der Waals forces), and partial polarity induced by unsaturation^31,32. Consequently, HSPiP executes low δ_p and δ_h values.

Fig. 4.

Experimental solubility in various excipients: (A) HSPiP-based predicted theoretical and experimental solubility with high correlation established (r² = 0.986) and (B) Statistical analysis outcome of regression analysis using data analysis tab of microsoft excel.

It was interesting to observe the impact of PEG400 in the aqueous system. The value of δp and δ_h were rapidly dropping. However, δ_d value was found to be slightly increasing up to ϕ₁ ≈ 0.2. Therefore, it was better to recommend PEG400 content < 20% in each microemulsion (Fig. 5).

Fig. 5.

Relationship of HSPs, MVol, and mass ratio of PEG400 in the binary system (PEG400 + water).

The stability of the drug was evaluated through UV scanning in various oils, as shown in supplementary Fig. S1. All oils displayed the same maximum absorption wavelength (λ_max) of 302 nm and negated any chemical interaction with the oils.

Preparation of placebo microemulsions

EU and LIM were selected based on the maximum drug solubility. Both were not readily dispersible into water at room temperature. Therefore, it was required to transform it into a microemulsion form for a cytotoxicity study in the explored cell lines. S_mix ratio was optimized based on the observed maximum delineated area of microemulsions and the safety of the surfactant (supplementary Figs. S2–S4). Addition of PEG400 resulted in a significant zone in the pseudo ternary phase diagrams. However, excess content of PEG400 led to high viscosity in microemulsion. Therefore, S_mix at ratio of 1:2 was selected as the most optimized content on ME1-ME3. Notably, ME1, ME2, and ME3 were formulated by varying the content of tween-80 and PEG400. Tween 80 and PEG400 do not exhibit any anticancer potential in the studied cell lines at the explored concentration³³. Purposely, an attempt has been made to investigate the cytotoxic potential of the emulsified oil against the cell lines. For this, the studied oils must be in emulsified form with the desired globular size for maximum accessibility around the treated cell lines. Table 2 provides a detailed description of the composition, globular size, polydispersity index (PDI as size distribution), and zeta potential. Tween-80 efficiently served as a surfactant (HLB = 14.5) to render a stable microemulsion with a size range of 339–510 nm (Table 2). ME1-ME3 were stable and isotropic. PEG 400, functioning as a co-surfactant, effectively reduces the surfactant concentration required within the formulations. Generally, a synergistic blend of tween-80 and PEG400 is frequently employed to enhance the solubility of poorly soluble drugs in emulsions. In these microemulsions (ME1-ME3), tween-80 plays a pivotal role in facilitating dispersion of the organic phase and mitigating phase separation (Oswald ripening) upon emulsification of the organic phase with the aqueous phase³⁴.

Table 2.

Various compositions and evaluation parameters of selected microemulsions (ME1-ME3).

Code	Oil (%w/v)	Tween-80 (%w/v)	PEG400 (%w/v)	Water (%w/v)	Size (nm)	PDI	ZP (mV)
ME1	EU (11.2)	21.6	10.0	57.2	339	0.28	-11.5
ME2	LIM (11.2)	10.9	8.6	69.3	510	0.49	-20.7
ME3	OLO (11.2)	17.5	10.2	61.1	465	0.31	-17.1

OLO: Olive oil, LIM: D-limonene, EU: Eugenol, PEG400: Polyethylene glycol 400, PDI: Polydispersity index, ZP: Zeta potential. Oil had no fixed molecular weight.

Characterizations of microemulsions

Each formulation was evaluated for globular size, PDI, and zeta potential as shown in Table 2. The globular size values were obtained in order as ME2 > ME3 > ME1. The variation in size may be attributed to the content of tween-80. The lowest size of ME1 can be correlated to its highest content of tween-80 (21.6%) whereas ME2 exhibited the highest value of size due to the low content of tween-80 (10.9%). A similar trend was observed for PEG400 content in each formulation. Thus, the values of size and PDI were the result of the combined impact of tween-80 and PEG400. In the case of zeta potential, the value was maximum for ME2, suggesting it was the most stable as compared to others. The negative nature of zeta potential can be related to fatty oil and PEG400. It reflects the intensity of repulsion among the adjacent microglobules³⁵. Thus, stability was attained through electrostatic repulsion among the dispersed microglobules.

Two concentrations (low and high) quick test: preliminary screening

In vitro anticancer activities against A375, A431, and B16-F10 cell lines

Limonene is well explored as an anticancer monoterpene in rats and mice^36–38. It alters the signalling pathways within cancer cells of various tissues (colon, stomach, pancreatic, skin, and liver). Therefore, it was imperative to investigate the anticancer potential of eugenol- and limonene-based microemulsions at the explored concentration. A detailed report is published for the limonene as the anticancer monoterpenes in literature. The authors reported that its anticancer activity was associated with the inhibition of tumour initiation, growth (increased the expression of p⁵³), and angiogenesis (by decreasing the activity of Ras/Raf/MEK/ERK and PI3K/Akt pathways). Moreover, it induces apoptosis (by activating the caspase pathway) in the cancer cells³⁹. However, limited data are available for its potential ability to have a detrimental impact on A431, A375, and B16-F10 cell lines⁴⁰. Petretto et al. reported lemon essential oil (obtained from leaves) containing limonene about 260 mg/mL, reduced the viability of A375 cells to 27% whereas Hela cells were reduced to 33% at 25 µM⁴¹. He reported the IC₅₀ value of limonene as 450 µM/L (0.0613 mg/mL) in a population of murine B16-F10 during Trial I phase⁴². Similarly, eugenol was effective against B16-F10 cells by reducing tumour size (almost 40% reduction in size) and preventing the tumour metastasis. Moreover, eugenol is anti-proliferative in the human malignant melanoma cell line by arresting the cancer cells in the S phase of the cell cycle⁴³. Jaganathan et al. reported various doses (80, 100, and 125 mg/Kg) of eugenol as effective doses to inhibit induced solid carcinoma in animal model. Notably, 125 mg/Kg was toxic to animal whereas 100 mg/Kg dose was effective in treating cancer⁴⁴.

In this study, eugenol, limonene, and olive oil were individually formulated as microemulsions (ME1-ME3) to assess anticancer potential against A375, A431, and B16-F10. The drug solution was used as a positive control. The concentrations (at a low concentration of 11.2%, equivalent to 112 mg/mL for each oil and a high of 56% equivalent to 560 mg/mL) of ME1-ME3 were serially diluted to get the final concentrations of 0.112 mg/mL and 0.560 mg/mL to get a sensitivity at low and high concentrations against these cell lines. The result is illustrated in Fig. 6A–D. Figures 6A–C exhibited % cell viability after treatment of 48 h against A375, A431, and B16-F10 cell lines, respectively. It is obvious that eugenol-based ME1 resulted in the highest level of cellular killing of cancer cell lines at the explored concentration as compared to LIM and olive oil. Moreover, the detrimental effect is concentration dependent in all cases and the pattern is in good agreement with the published report⁴⁵. Thus, the explored cell lines are highly sensitive to eugenol, which may be attributed to its innate cytotoxic (inhibiting mitochondrial functionality) and anti-proliferative behaviour as reported in various literatures⁴⁶ ,⁴⁷ The potential anticancer ability (against B16-F10) of eugenol may be due to growth inhibition through downregulation of E2F1 transcription and related gene expression as shown in Fig. 6C⁴³. Fig. 5D elicits concentration dependent detrimental effects of COA against the explored cell lines (A431, A375, and B16-F10). The drug was highly effective against all cell lines at low concentrations (125 µM and 500 µM). This finding is quite interesting, informative, and has not been explored so far in the cell lines.

Fig. 6.

In vitro cell line study of ME1, ME2, ME3, and COA suspension against various cell lines: (A) % Cell viability against human melanoma (A375), (B) % Cell viability against human epidermoid skin carcinoma (A431), (C) % Cell viability against mouse melanoma (B16-F10), and (D) % Cell viability against mouse melanoma (B16-F10), A375, and A431 cell lines after ciclopirox olamine treatment at varied concentrations (125 and 500 µM). Statistical significance is denoted as follows: ^#p < 0.05, ^##p < 0.01 as compared to the lower concentration).

The obtained result is novel for developing a state-of-the-art therapeutic strategy to treat a cancer patient developing a fungal infection due to compromised immunity as well. Dual functionalities of COA can be a promising discovery in the treatment of cancer and fungal infection together.

In vitro safety study against normal human skin fibroblast cell line (HSF)

To negate the toxic potential of eugenol, limonene, and olive oils (as ME1, ME2, and ME3), it was necessary to investigate their impact on the normal cells. Therefore, the samples were treated against HSF²⁴. Figure 7 exhibits cellular viability of the treated HSF cells at low and high concentrations. Eugenol showed high cytotoxicity to the normal cells at high concentration (0.560 mg/mL) whereas limonene was found to be safe at explored concentration range. Therefore, low concentrations (0.112 mg/mL) of eugenol, limonene and olive oils are safe and biocompatible, as shown in Fig. 7. To develop a safe and effective formulation (micro or nanoemulsion), the recommended concentrations of eugenol and d- limonene must be low (0.112 mg/mL) for effective COA delivery as a synergistic strategy. Moreover, their ratio may be selected for maximum stability and miscibility in the aqueous phase as revealed in the HSPiP predicted report. Thus, a combined approach of both oils may produce a detrimental effect against the cancer cell lines at a safe concentration to avoid unnecessary introduction of the inert excipients into the patient body. The approach may reduce the cost burden and the related toxicity during chemotherapy. Thus, the lower concentration was recommended for further apoptosis study.

Fig. 7.

In vitro cell line study to investigate % viability against normal human skin fibroblast cell lines at the explored concentrations (CX and C5X) of ME1, ME2, and ME3. Statistically significance is expressed at ^#p < 0.05. Data are expressed as mean ± SD.

Concentration dependent cytotoxicity study of ME1, ME2, ME3, and COA suspension against A431 (human) and B16-F10 (mouse)

A range of diluted concentrations was prepared for the treatment as shown in Fig. 8A–C. The experiment provided IC₅₀ values of LIM, EU, and COA in the studied cell lines. The concentration of COA was prepared in molar concentration (µM) due to high sensitivity and potency against the studied cell lines. The calculated IC₅₀ values of EU against A431 and B16-F10 were 46.8 and 28.7 µg/mL, respectively (Fig. 8A,B) whereas these values for LIM were found to be 71.3 and 60.2 µg/mL, respectively (Fig. 8A,B). These findings were compared against the pure drug solution as shown in Fig. 8C. The estimated IC₅₀ values of COA against A431 and B16-F10 were obtained as 31.4 and 6.2 µg/mL, respectively (Fig. 8C). Notably, the result showed that the explored essential oils and the drug were profoundly effective in rendering detrimental impact against the studied cell lines (in vitro assessment). EU exhibited relatively high sensitivity against A431 and B16-F10 cell lines as compared to LIM. In the literature, the IC₅₀ value of eugenol against A431 and B16-F10 was reported as 100 µg/mL^46,47. However, various published reports revealed variable values of IC_50, which may be correlated to the difference in the experimental conditions, such as exposure time, the cell density, and the composition of the growth media. These variables change the IC₅₀ value. However, the obtained result is in good agreement with the published IC₅₀. Limonene is well known to have anticancer potential in several in vitro studies. In an another study, the IC₅₀ value of LIM against B16-F10 was reported as148 µg/mL, which was quite higher than the explored value (60.2 µg/mL)²⁵. Therefore, the inconsistencies among the published reports may be attributed to various unreported factors (especially experimental design, cell line population density, incubation time, exposed cells, method validation, and instrumental error). No report has been published for the IC₅₀ of d-limonene against A431 so far. Thus, limonene exhibited equivalent sensitivity against the explored cell lines (B16-F10 and A431). In the case of the drug, it is interesting that the drug is highly active against A431 and B16-F10. The estimated IC₅₀ values are significantly low as compared to the excipients (LIM and EU). Thus, the drug may produce a detrimental effect if laden with nanocarrier containing LIM or EU as excipients possessing innate anticancer potential.

Fig. 8.

Concentration-dependent cytotoxicity study: (A) Cell viability after treatment with EU (ME1) showing concentration-dependent killing of A431 cell lines and IC₅₀ value and % cell viability after treatment with LIM (ME2) showing concentration-dependent killing of A431 cell lines and IC₅₀ value, (B) Cell viability after treatment with EU (ME1) showing concentration-dependent killing of B16-F10 cell lines and IC₅₀ value, and cell viability after treatment with LIM (ME2) showing concentration-dependent killing of B16-F10 cell lines and IC₅₀ value, and (C) Cell viability and IC₅₀ values after treatment with COA showing concentration-dependent killing of A431 and B16-F10 cell lines using 96-well microtiter plates.

To elucidate the mechanisms of action of LIM, EU, and COA, a literature-based exploration of molecular signalling pathways and receptor-mediated interactions would provide a more comprehensive understanding. Limonene is a natural monoterpene found in citrus oils, known for its anticancer properties. It exerts cytotoxic effects against A431 (human epidermoid carcinoma) and B16-F10 (murine melanoma) cells by inducing apoptosis and inhibiting proliferation. Limonene disrupts cell cycle progression, particularly causing G1 phase arrest, and modulates key molecular pathways such as downregulating PI3K/Akt and Ras-MAPK signalling^48,49. It also generates oxidative stress within tumour cells, leading to mitochondrial dysfunction and activation of intrinsic apoptotic pathways. Moreover, eugenol, a phenolic compound derived from clove oil, also demonstrates significant anticancer effects against A431 and B16-F10 cells. It promotes apoptosis by increasing reactive oxygen species (ROS) production, causing mitochondrial membrane potential loss, and activating caspase cascades⁵⁰. Additionally, eugenol inhibits NF-κB signalling, reducing the expression of anti-apoptotic proteins, and interferes with tumour cell migration and invasion. Briefly, both limonene and eugenol may act through multiple pathways involving oxidative stress, mitochondrial dysfunction, apoptosis induction, and inhibition of survival and proliferation signals, making them promising candidates for cancer therapy against A431 and B16-F10 cells.

COA might exerted anticancer effects against B16-F10 and A431 cells primarily through iron chelation, oxidative stress induction, disruption of critical survival pathways, and promotion of apoptosis, making it a promising agent for cancer therapy repurposing^51,52.

Apoptosis study

In order to explore mechanistic insights into the treated cells (A431 and B16-F10), an in vitro cytotoxicity study was conducted at the explored concentration of excipients (LIM and EU) and COA. The cells were treated at the explored IC₅₀ value for 24 h. The result is depicted in Fig. 9A–D (against A431) and Fig. 10A–D (against B16-F10). It is revealed in the study that LIM and EU were effective against cancerous cell lines (A431 and B16-F10) at the treated concentration. Figure 9A illustrates the untreated control group wherein cells are normally alive with slight natural death (< 5%). The LIM-treated group showed a significant level of early (12.42%) and late apoptosis (10.68%) based on cancerous cell mortality (Fig. 9B). A similar observation was obtained for the EU-treated cells (Fig. 9C). The EU-based detrimental effect was comparable to the drug at the studied concentration. LIM elicits relatively more effectiveness against the treated cells compared to EU. Notably, LIM and EU are food-grade essential oils. Therefore, the cells were treated and compared to pure COA. The finding showed that the obtained IC₅₀ value for limonene against B16-F10 (60.2 µg/mL) was approximately 2-fold lower than the published value (148 µg/mL) (Fig. 11)²⁸. This may be attributed to variation in the population size of cells, the exposure time, and ethanolic solution used in the published report²⁸.

Fig. 9.

Apoptosis study against A431: (A) Control group of A431 cells cultured and grown, (B) LIM treated cells, (C) EU treated cells, and (D) COA treated cells.

Fig. 10.

Apoptosis study against B16-F10 cells: (A) Control group of B16-F10 cultured cells and grown, (B) LIM treated cells, (C) EU treated cells, and (D) COA treated cells.

Fig. 11.

The effect of pre-treatment of LIM, EU, and COA at the explored concentration against A431 and B16-F10 cell lines. Impact of the treatment on different stages of apoptosis and necrosis. *p < 0.05 and **p < 0.001 were considered as significant as compared to the control group in the study. (A) Treatment against A431 cell lines and (B) treatment against B16-F10 cell lines.

Conclusions

This study explored the anticancer potential of essential oils such as limonene (LIM), eugenol (EU), and olive oil against A431 and B16-F10 cancer cell lines. Among them, EU and LIM showed promising in vitro activity. Literature based information on their anticancer effects at the tested concentrations is limited. Additionally, COA demonstrated strong cytotoxicity against both cell lines. The viability assays indicated a concentration-dependent cytotoxic effect. EU exhibited slightly lower IC₅₀ values than LIM, suggesting higher potency, though LIM was nontoxic to normal HSF cells at both low and high doses. Thus, LIM may serve as a suitable excipient for COA-based formulations, enhancing therapeutic and synergistic effects in skin cancer treatment. Apoptosis studies revealed that LIM induced greater early and late apoptosis, as well as necrosis, compared to EU. Notably, B16-F10 cells were more sensitive to LIM than EU. While COA’s cytotoxicity aligns with existing reports, direct comparisons between COA and the excipients are not valid due to differences in treatment concentrations. Conclusively, LIM- based microemulsion offers a cost-effective and synergistic platform for COA loaded microemulsion, supporting its repurposing for topical cutaneous cancer therapy.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

O.A.A: Methodology, validation, analysis, and data curation; A.H: Conceptualization, methodology, drafting, editing, and review; M.A.A: Supervision, funding acquisition, resources, software, and visualization.

Funding

The authors acknowledge and extend their appreciation to the Ongoing Research Funding program (ORF-2025-524), King Saud University, Riyadh, Saudi Arabia.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.