Development of a fish oil–nanoemulsion gel as a drug-delivery system to prevent capsular contracture

Preparation and characterization of N3G

Omega-3 PUFAs, mainly EPA and DHA, present in fish oil have anti-inflammatory and antifibrotic properties. To exploit these beneficial effects, fish oil was encapsulated and delivered to prevent capsular contracture. However, fish oil, being lipophilic, is incompatible with aqueous living systems and requires modification. Therefore, we modified it into a nanoemulsion formulation via low-energy phase inversion composition (PIC), which results in the formation of an o/w emulsion45. To solubilize fish oil in water, surfactants were first added. We used two small-molecule non-ionic surfactants, TWEEN 80 and SPAN 80, which can form small nanoparticles and are non-toxic46. A mixture of TWEEN 80 and SPAN 80 was added to the oil phase during emulsification (Fig. 1a). TWEEN 80 is a hydrophilic surfactant with an HLB value of 4.3 and SPAN 80 is a hydrophobic surfactant with an HLB value of 15. A mixture of these two surfactants enabled the emulsification of fish oil into an o/w nanoemulsion with small oil droplets. A TWEEN 80 to SPAN 80 ratio of 4:1 generated a nanoemulsion with very small particles and a low PDI (Fig. S1).

Fig. 1
figure 1

Preparation of NE-ω3-gel (N3G). (a) N3G was prepared using a two-step method. The oil phase (fish oil and surfactants) was mixed with DI water. (b) The mixture was vigorously stirred to form a nanoemulsion. (c) In the second step, a mixture of poloxamers (PF127 and PF68) was added to the diluted nanoemulsion to produce N3G. The final formulation was thermoreversible; it was liquid at room temperature and a gel at 37 °C.

The addition of PEG 400 to the nanoemulsion minimized the oil droplet size and maximized stability (Fig. 1b). As shown in Fig. S2, the optimum weight ratio was 3.4%; higher or lower weight ratios resulted in larger oil particles. This amount was henceforth used to prepare NE-ω3. The ionic and non-ionic surfactants stabilized the fish oil and facilitated emulsification. An absolute zeta potential value of less than –30 mV indicates good stability47. Following the same protocol, we prepared a formulation with identical components and another lacking PEG 400. The zeta potentials of the formulations were determined using a Zetasizer Nano ZS90 (Malvern Instruments Ltd., Korea). For the latter sample, the absolute zeta potential value was lower (–1.75 ± 0.1) than the former sample (–34 ± 1.21) (Fig. S4). Therefore, PEG 400 is required to generate a stable nanoemulsion.

A portion of the emulsion was added to a poloxamer mixture (PF127:PF68). They are amphiphilic nonionic triblock copolymers containing hydrophilic polyethylene oxide (PEO) and hydrophobic polypropylene oxide (PPO) regions. The PEO and PPO blocks in the poloxamer molecule form a micellar structure in water with hydrophobic regions at the core and hydrophilic regions at the end, which makes it a liquid at room temperature. As the temperature increases, the sol-to-gel transition occurs, and the micelles rearrange into an orderly gel network (Fig. 1c). The two-step preparation method ensures that fish oil is trapped within the hydrophobic cavities of the poloxamers.

Next, the N3G formulation was optimized in terms of particle size, PDI, and zeta potential to maximize the ω3 content in N3G. To determine the maximum amount of fish oil that could create a successful nanoemulsion, 10%, 13.8%, 20%, and 25% fish oil were emulsified according to our protocol. The quantities of the other components—surfactants (TWEEN 80, SPAN 80), co-surfactant (PEG 400), and DI water—were the standards (wt%). An aliquot of nanoemulsion was diluted prior to DLS. As shown in Table 2, the particle sizes of 10% and 13.8% fish oil-containing nanoemulsions differed slightly (Fig. S5). Nanoemulsions with 20% and 25% had large particle sizes, possibly because of agglomeration of small particles to large aggregates. Although the particle size in the 10% fish oil nanoemulsion was smaller than that in nanoemulsion containing 13.8% fish oil, the latter had a lower PDI value, indicating a more homogeneous particle size. Thus, 13.8% fish oil was used in subsequent experiments.

Table 2 Optimization of the amount of fish oil for nanoemulsion preparation.

To determine the amount of fish oil needed to generate a visible nanoemulsion without layer separation, different formulations with identical volumes of DI water were investigated. As shown in Table S1, a maximum of 20 g fish oil could be added before the oil layer separated. The quantities (wt%) of fish oil, surfactant, and co-surfactant were as in the standard formulation. This formulation contained approximately 3 mg fish oil per 1 g gel. DLS showed that the oil droplet size was larger (Zavg 5500 nm) than the standard formulation, and a macroemulsion was generated. Hence, this formulation was not considered further. The standard formulation (NE-ω3) contained 13.8% fish oil and had a very small particle size (mean 287 ± 8.599 nm) and a narrow particle size distribution (PDI 0.29 ± 0.047) according to DLS (Fig. 2a). The nanoemulsion had an average zeta potential of –34 ± 1.21 mV (Fig. 2b), indicating good physical stability and potential for drug delivery.

Fig. 2
figure 2

DLS analysis of the prepared nanoemulsion. (a) Particle size (Zavg) and PDI; (b) Zeta potential.

When surrounded by surfactants in a nanoemulsion, the activity of fish oil is preserved because of the absence of oxidation48. As shown in Fig. S5, there were no major changes in the Zavg, PDI, and zeta potential of NE-ω3, indicating preparation of a stable nanoemulsion.

After the preliminary stability study, we investigated the effects of temperature and pH on the stability of NE-ω3. Nanoemulsion samples were heated from room temperature to 37 °C. In addition, nanoemulsion samples were adjusted to acidic or basic, and the particle size, PDI, and zeta potential were evaluated. As shown in Table S2, the particle size increased by 9.78% and the PDI decreased by 8.31% after heating to 37 °C for 24 h. These changes may be attributable to Ostwald ripening, a common phenomenon in nanoemulsions49. In addition, the zeta potential of the acidic nanoemulsion was 99.661% lower than the neutral formulation, whereas that of the basic nanoemulsion was 47.3% higher. These results indicate that basic or acidic conditions can reduce the stability of NE-ω3.

NE-ω3 was converted into a thermoreversible gel. Thermoreversibility, which can be mediated by poloxamers, is the transition of a gel between liquid and semi-solid states in response to temperature changes and enables precise control of gel formation and drug release. Thermoreversibility is useful for implant coatings because the gel is liquid at room temperature but solidifies at a higher temperature (30–50 °C)35,36. PF127 was the primary gelling agent and formed a gel that encapsulated the fish oil micelles. However, this gel had a very low gelation temperature (< 20 °C). We added PF68 to the formulation to increase the gelation temperature. PF127 and PF68, at a variety of ratios, were added to 2 g diluted nanoemulsion and thermoreversibility was evaluated (Table 3). The addition of < 2 wt% PF68 resulted in thermoreversible N3G (Fig. S6). Solutions with PF127:PF68 ratios of 30:2 and 30:3 formed a gel at room temperature, but this gelation was disrupted by shaking. The optimum PF127: PF68 ratio for thermoreversible N3G was 30:1; this ratio was used in subsequent experiments. Gelation of N3G was optimal at 35 °C, indicating that it solidifies rapidly inside the body. The gelation time was ~ 5 min, enabling drug delivery to target locations.

Table 3 Determination of the optimum poloxamer ratio.

The effects of heat and pH on the stability of N3G were investigated via FTIR. N3G was stable at 50 °C for 24 h. However, the C=O vibrational band (1743 cm–1) disappeared at pH 5 and 9 (Fig. S7), indicating that fish oil decomposes under acidic and basic conditions.

N3G was characterized by FTIR spectroscopy to determine its chemical composition (Fig. 3a). The FTIR spectrum of N3G was compared to its precursor materials (fish oil and poloxamers). Fish oil showed major peaks at 2922 cm–1 and 1743 cm–1 due to the presence of C-H and C=O bonds. The major peaks for poloxamers were at 1341 cm–1 and 1102 cm–1, corresponding tο Ο–Η and C–O bonds, respectively. N3G showed peaks at 2922 cm–1, 1743 cm–1, 1341 cm–1, and 1102 cm–1, as did fish oil and poloxamers. These overlapping peaks confirm the presence of all of the components in N3G and that the components are physically blended, intact, and have unaltered chemical structures.

Fig. 3
figure 3

Characterization of N3G. (a) FTIR spectra of N3G and its precursor materials; (b) DSC spectra of N3G and its precursor materials.

The thermal behavior of N3G was investigated via DSC. Figure 3b shows the DSC thermograms of N3G, fish oil, solid PF127, and solid PF68. The oil showed an endothermic peak at 56.10 °C due to the presence of saturated fatty acids such as myristic acid in fish oil50. PF68 and PF127 showed single sharp endothermic peaks at 55.68 °C and 57.43 °C, respectively, which are their melting points51,52. A single endothermic peak of N3G appeared at 55.44 °C during heating, suggesting that N3G has a semi-crystalline structure that melts at around 55.44 °C.

In vitro drug release study of N3G

In vitro drug release—i.e., the time required for the complete release of the encapsulated fish oil from the gel matrix—was investigated. A small silicone mini-implant was used to mimic the real-life situation. The surface of the mini-implant was coated with N3G and it was placed inside a perforated tube. The assembly was fully submerged in a beaker containing 50 mL DI water. The temperature was maintained at 37 °C using a water bath (Fig. 4a). Both ends of the tube were secured with weighted dialysis closures to ensure full immersion and inhibit leakage or floating.

Fig. 4
figure 4

In vitro drug release by N3G. (a) Experimental setup to analyze drug release by N3G; (b) drug-release profile of fish oil. The amount of fish oil released over time is shown as a percentage.

Capsular contracture typically develops during the first few months after implantation; therefore, sustained drug release to the breast capsule is important at an early stage2,53. Compared to previous studies on poloxamer as a drug carrier, N3G exhibited sustained release54,55, and UV–vis absorption analysis showed that > 96% of the fish oil was released gradually over 10 h (Fig. 4b). The consistently slow release of fish oil from the DDS indicates that N3G can mediate controlled release, which is crucial in therapeutic contexts. Release could be slower in vivo (e.g., tissues and extracellular matrix), which differs markedly from the experimental conditions, underscoring the gel’s potential for clinical applications. N3G persistently adhered to the mini-implant surface, demonstrating the potential of N3G for biomedical applications, such as implant coatings, based on its sustained drug release.

Animal study with implants and N3G

To evaluate the practical utility of N3G, we performed an animal study using N3G-coated mini-implants (Fig. 5). Thirty rats were randomly assigned to three groups (n = 10 per group): G1 (negative control) received only the implants installed; G2 (positive control) received fish oil via gavage together with the installed implants; G3 (experimental group) received implants coated with N3G. Initially, the rats were implanted with a 2 cc custom-made, mini-implants of 2 cm diameter made of silicone gel (HansBiomed Co. Ltd., Seoul, Korea). Only rats in G3 received implants coated with N3G. All of the rats showed stable weight gain and food intake, suggesting the biocompatibility and non-toxicity of N3G. The study was terminated 90 days after the implantation surgery. The implants and surrounding capsules were excised as a single unit for use in subsequent experiments.

Fig. 5
figure 5

Study design. (a) G1, untreated negative control; (b) G2, positive control (treated daily with Menhaden fish oil [80 µL/250 g] via gavage); (c) G3, experimental group (N3G-coated implants [equivalent to a 0.6 g dose]); (d) custom-made, smooth, round silicone gel mini-implant (2 cc, 2 cm in diameter); e) mini-implant coated with N3G by immersion; f) implants and surrounding fibrous tissue (capsule).

Histological analyses of capsular tissues

Excised capsular tissue was subjected to H&E staining to measure capsular thickness. Capsular thickness was decreased around the implants (G1, 109.13 ± 13.58 μm; G2, 91.52 ± 11.78 μm; G3, 86.00 ± 9.90 μm) (Figs. 6 and 7a). ANOVA indicated a significant reduction in G2 compared to G1; this finding was substantiated through a Fisher’s PLSD test, which showed significantly reduced thickness in both G2 and G3 compared to G1, with no marked difference between G2 and G3. MT staining was performed to identify fibrosis and collagen around the mini-implant. It revealed a progressive decrease in fibrosis and collagen levels in all groups, which was most notable in G3 (Fig. 7b). IHC indicated a considerable decline in myofibroblasts, which are vital for the development of fibrosis, with G1 and G3 presenting the highest and lowest counts, respectively (Fig. 7c). Taken together, these results indicate that N3G reduces capsular thickness, fibrosis, and collagen deposition.

Fig. 6
figure 6

Average capsular thickness was 109.13 μm in G1, 91.52 μm in G2, and 86.00 μm in G3. Average capsular thickness differed significantly different between G1 and G2 according to post hoc analysis (*p < 0.05).

Fig. 7
figure 7

(a) H&E-stained sections of capsules from in G1 (left), G2 (center), and G3 (right); scale bar, 30 µm. (b) MT-stained sections showing fibrosis and collagen in G1 (left), G2 (center), and G3 (right); scale bar, 60 µm. (c) IHC sections stained with an anti-α-smooth muscle actin (α-SMA) antibody showing the myofibroblast distribution in G1 (left), G2 (center), and G3 (right); scale bar, 60 µm.

The lack of a significant difference in capsular thickness between G2 and G3 may be attributable to the oral administration of fish oil in G2 for 90 days. Although capsular thickness did not significantly differ among the groups, MT staining of fibrosis and collagen and IHC of myofibroblasts suggested that G3 showed greater reductions of fibrosis and myofibroblast counts around the implant site than the other two groups. Based on this and the potential challenges of long-term compliance with oral supplementation, the N3G formulation is the most feasible for preventing capsular contracture.

Gene expression analysis

The formation of thick fibrous scar tissue around a breast implant is a symptom of capsular contracture. Fibroblasts trigger fibrosis, which transform into myofibroblasts. Myofibroblasts form mesh-like scar tissue with blood vessels to seal the wound. Fibroblasts are responsive to inflammatory cytokines, such as COL1A2, TGF-β2, IFN-γ, IL-4, IL-6, and IL-1056,57. In qRT-PCR, extracted RNA is converted into complementary DNA (cDNA) by reverse transcription (RT). The cDNA is amplified and quantified by real-time PCR. During the RT step, a set of primers complementary to the mRNA of the gene of interest was used to specifically amplify the cDNA in the next step58.

The expression levels of inflammatory cytokines in capsule tissues are shown in Fig. 8. The expression of IFN-γ was significantly higher in G3 than G2 and G1 (7.582 ± 3.318, 2.719 ± 1.730, and 1.000 ± 0.000, respectively). IFN-γ plays a complex role in the development of capsular fibrosis around implants. It modulates the immune response, activates macrophages, and inhibits fibroblast activity, which are crucial for collagen production and extracellular matrix formation. It also regulates other inflammatory cytokines, thereby affecting fibrosis progression. It reduces myofibroblast viability and increases apoptosis, leading to decreased collagen formation and fibrosis59,60,61. The highest IFN-γ level in G3 is consistent with the antifibrotic property of ω3 and suggests that N3G has potential for preventing capsular contracture. IL-4 expression was significantly lower in G3 (0.444 ± 0.178) and G2 (0.689 ± 0.528) than G1 (1.000 ± 0.000), indicating lower IL-4 levels in G3 and G2 compared to G1. This reduction indicated a decrease in the inflammatory response, consistent with the regulatory role of IL-4 in M2 macrophage polarization and inflammation. This reduction might indicate less need for the anti-inflammatory and tissue remodeling functions typically mediated by IL-4. Given that M2 macrophages play a role in healing and anti-inflammatory processes, lower IL-4 expression may reflect an altered inflammatory response or reduced inflammation in the context of capsular contracture2,62.

Fig. 8
figure 8

qrt-PCR analysis of the expression in capsule tissues of collagen alpha-2(I) chain (COL1A2), transforming growth factor (TGF)-β2, interferon (IFN)-γ, interleukin (IL)-4, IL-6, and IL-10. IFN-γ expression was significantly higher in G3 than G2 and G1, and IL-4 expression was lower in G3 and G2 than in G1. TGF-β2 expression was significantly lower in G3 than in G1. Data are means ± standard error of the mean. Significantly different from G1 vs. G2 (ANOVA post hoc test: *p < 0.05), G1 vs. G3 (ANOVA post hoc test: x, p < 0.05) and G2 vs. G3 (ANOVA post hoc test: **p < 0.001).

TGF-β2 expression was significantly lower in G3 (0.569 ± 0.291) than in G1 (1.000 ± 0.000), with G2 showing a value of 0.874 ± 0.501. This significant reduction in TGF-β2 expression in G3 emphasizes the importance of TGF-β in fibrosis, particularly in the context of chronic inflammation. This reduction is particularly notable, given the involvement of TGF-β in collagen deposition and fibrosis during the foreign body response, which is often triggered during processes occurring in the acute phase of inflammation, such as platelet degranulation. This decrease in TGF-β expression aligns with previous reports that its suppression not only reduces scar formation but also myofibroblast contraction. Therefore, the lower TGF-β2 level in the treatment group indicate the effectiveness of the gel in mitigating fibrotic reactions around implants63,64,65,66. The expression levels of COL1A2, IL-6, and IL-10 were not significantly different among the three groups. COL1A2 expression was 0.811 ± 0.566 in G3, 0.605 ± 0.243 in G2, and 1.000 ± 0.000 in G1. IL-6 expression was 3.300 ± 2.055 in G3, 2.944 ± 3.096 in G2, and 1.000 ± 0.000 in G1. IL-10 expression was 0.918 ± 0.355 in G3, 0.758 ± 0.670 in G2, and 1.000 ± 0.000 in G1.

Mechanistic insight into ω3 PUFAs in inflammation and fibrosis

Ω3 PUFAs have anti-inflammatory and antifibrotic properties, which involve multiple biochemical pathways and cellular processes. They reduce inflammation and fibrosis by serving as precursors to specialized pro-resolving mediators such as resolvins, which inhibit pro-inflammatory eicosanoid production and cytokine expression, thus dampening the inflammatory response. In addition, ω-3 PUFAs compete with ω-6 fatty acids, displacing arachidonic acid and reducing the synthesis of inflammatory compounds67,68. In fibrosis, ω-3 PUFAs promote the degradation of Yes-associated protein and transcriptional co-activator with PDZ-binding motif in the liver and kidneys. This reduces hepatic stellate cell activation and enhances autophagy flux and AMP-activated protein kinase activation, thereby mitigating fibrosis30,69. They also upregulate nuclear factor erythroid 2-related factor 2, decreasing oxidative stress and apoptosis70. These mechanisms underline the therapeutic potential of omega-3 PUFAs in non-alcoholic steatohepatitis and rheumatoid arthritis, effects mediated by improvement of inflammation and tissue repair71,72.

Effectiveness of N3G for capsular contracture

The strength of this study is its comprehensive assessment of N3G for the prevention of capsular contracture. The detailed histological and gene expression analyses provide insight into the anti-inflammatory and antifibrotic effects of ω3 PUFAs. The use of an established rat model, with appropriate control groups, enhanced the reliability and validity of the findings. Moreover, the thermoreversible gel facilitated sustained drug release and localized treatment, highlighting its potential as a safer and more effective alternative to current therapies.

N3G has several advantages over existing treatments for capsular contracture, including significant reductions in capsular thickness, fibrosis, and myofibroblast count, comparable to pharmacological options such as LTRAs and steroids, but with fewer side effects. Furthermore, it could improve patient compliance and reduce the frequency of treatments. In addition, by preventing capsular contracture, N3G could reduce the need for costly revision surgeries, thereby minimizing the financial burden on patients and healthcare systems. These attributes underscore the innovative nature and clinical relevance of N3G.