Introduction

Periodontitis is a chronic plaque-induced inflammatory disorder that affects the periodontium, teeth-surrounding structures such as gingiva, alveolar bone, periodontal ligament, and cementum. Although the principal etiologic factor of periodontitis is bacteria, other factors such as environmental, genetic, and behavioral, as well as host immune response, are involved in the progression of the disease [1]. The disease begins when the anaerobic gram-negative bacterial species adhere to the teeth surfaces and attack the gingival sulcus, a region between the teeth and the gums [2]. After that, bacterial populations multiplicate and co-aggregate in complex communities to form biofilms. As a result of further calcification of plaque, the depth of the gingival sulcus increases and gingival margin migrates along the tooth surface forming periodontal pockets (PDPs). PDPs are filled with extracellular transudate called gingival crevicular fluid (GCF) which is similar to extracellular fluid [2,3,4].

After that, host immune defense mechanisms like phagocytosis are initiated to attack the periodontal pathogens. Throughout phagocytosis, a nonmitochondrial O2 consumption increases up to 10 or 20 times more than that at normal situation with consequent generation of free radicals (FRs) and reactive oxygen species (ROS) [5]. Moreover, overproduction of ROS is induced by the periodontal pathogens themselves. Consequentially, oxidative stress (OS) arises due to insufficient cellular antioxidant defense to totally deactivate the generated ROS. Substantial direct and indirect tissue destruction is caused by ROS through several mechanisms, comprising DNA damage, enzyme oxidation, lipid peroxidation (LPO), and protein degradation [6, 7]. Continued release of ROS causes the release of inflammatory mediators followed by gradual degeneration of connective tissue, periodontal ligaments, and alveolar bone surrounding the teeth. Ultimately, untreated cases may end with tooth loosening and loss [2, 4, 5].

Mechanical removal of calculus and plaque by scaling and root planing (SRP) is considered as the key element for controlling and treatment of periodontitis. Nevertheless, the anatomical complexity of the teeth and instrumental interference retard accessibility to deeper points in PDPs. Hence, SRP alone might not produce complete elimination of pathogens nor control the condition. Such situations have fortified the use of an adjunct strategy such as locally delivered drugs parallel to the mechanical technique [3, 8]. Due to the postulated role of OS in the progression of periodontitis, usage of antioxidants may be an efficient aide for additional control of periodontitis [1, 7, 9].

Coenzyme Q10 (Q10) is an endogenous, vitamin-like, and oil-soluble antioxidant, and it is essential for every cell in the living organism. It can guard the mitochondrial membrane, proteins, and DNA from ROS damage [10]. While Q10 can be produced in the living organism, conditions may arise in which the production capacity is not enough to meet Q10 requirements. Deficiency of Q10 is reported to occur in a greater extent in metabolically active cells such as gingival cells. Gingival biopsies exhibited subnormal tissue level of Q10 in about 80% of patients with periodontal disease [11]. Outstanding safety and tolerability of Q10 was reported for even large oral doses of about 3000 mg per day. Moreover, no adverse effects were detected with daily doses extending from 600 to 1200 mg [12]. However, Q10 has extremely poor aqueous solubility and insignificant oral bioavailability due to its hydrophobicity [10].

Up to our knowledge, local delivery of Q10 for the management of periodontitis was employed in a few studies in the form of gel preparations [8, 13,14,15]. Reduction of particle size is definitely considered one of the best significant possible tactics to increase the bioavailability of some hydrophobic drugs such as Q10 [13]. Actually, nanotechnology progressively provides a suitable key for the treatment of many dental problems comprising periodontal disease. Nanosized polymeric systems provide several advantages as compared to other conventional fibers, gels, strips, films, etc. Nanosystems improve the stability of chemically unstable active ingredients, eliminate side effects, increase aqueous dispersibility, and sustain drug release over prolonged periods of time [3]. For controlling periodontitis, there have not been any reported attempts to study the extra advantageous effect of Q10 in the form of a nanosystem. Hence, such survey encouraged us to dedicate our study via designing a novel and simplistic strategy for fabricating nanomicelles. By luck, PDPs offer reservoirs for the introduction of drug delivery systems (DDS). GCF acts as a leaching medium that may enhance further release of a drug from DDS followed by its spreading throughout PDPs [2].

Polymeric nanomicelles (NMs) are nanocarriers of amphiphilic polymers arranged in the form hydrophobic core and hydrophilic shell. In aqueous media, NMs are self-assembling system at the critical micelle concentration (CMC) of the polymers [16]. The main advantages of NMs are the enhanced aqueous solubility of hydrophobic drugs, the inhibition of chemical decomposition of drugs due to external factors such as oxidation, and the enablement of targeted and site-specific drug delivery to a required site [17]. Hence, the much smaller of size of NMs renders them able to penetrate sites such as PDPs below the gingival line that may be inaccessible to other DDS. Moreover, nanosystems such as NMs are appropriate for combination into a hydrogel matrix to facilitate application [2, 18, 19]. Kolliphor® P 407 (K407) was selected as an amphiphilic polymer for preparation of NMs. K407 is a triblock copolymer of two polyethylene oxide moieties flanking one polypropylene oxide moiety (PEO101–PPO56–PEO101) that is commonly used to improve solubility, stability, and encapsulation efficiency of some drugs. Its hydrophilic–lipophilic balance (HLB) ranges from 18 to 23. K407 was classified as biocompatible and it is approved by the FDA [20].

Therefore, the aim of this study was to formulate and evaluate a nanomicellar carrier of Q10 and then transfer such innovative system from just a benchside formula to clinical practice. First, we intended to get initial information about its potential antioxidant effect as an adjunct to SRP for the management of chronic periodontitis. These early clinical results could be used as a guide to design and conduct larger-scale comparative clinical studies in our future plan.

Methodology

Materials

Q10 was obtained from MEPACO Pharmaceutical Company, Egypt. Kolliphor® P 407 (K407) was kindly supplied by BASF, Germany. Carbopol 934 (CP) and hydroxypropyl methyl cellulose K4M (HPMC) were obtained from Colorcon, UK. Methyl cellulose, high substitution (MC), was attained from BDH Chemicals Ltd., Poole, England. Triethanolamine (TEA) was obtained from Nice Chemicals Pvt. Ltd., Kerala, India. Tween 80 (T80) and potassium dihydrogen orthophosphate were obtained from ADWIC, Egypt. Total antioxidant capacity (T-AOC) kit and lipid peroxide or malondialdehyde (MDA) kit were purchased from BioDiagnostic, Egypt. PerioPaper™ strips were purchased from Oraflow Inc., Plainview, NY, USA, and deionized water was provided from Central Lab, Faculty of Pharmacy, Mansoura University.

Preparation of Q10-loaded and Q10-free nanomicelles

The Q10-loaded nanomicelles (NMQ10) were prepared by nanoprecipitation according to the method reported by El-Far et al. after some modifications [21]. Initially, both Q10 (0.1 g) and K407 (0.2 g) with Q10:polymer ratios of 1:2 were dissolved in 30 ml of acetone to form organic phase. After that, the organic phase was added dropwise to 100 ml of distilled water under continuous magnetic stirring. The obtained dispersion was left overnight to permit complete evaporation of acetone with subsequent formation of a concentrated nanosuspension. NMQ10 was collected after lyophilization of the concentrated nanosuspension under vacuum at − 80 °C (freeze dryer, SIM FD8-8T, SIM International, USA) into a dry powder. Finally, the lyophilized NMQ10 were weighed and kept at 4 °C for further evaluation. Q10-free nanomicelles (NMP) were prepared using the same procedure except that the drug was not included.

Characterization of NMQ10

Determination of encapsulation efficiency, loading efficiency, and percent yield

The encapsulation efficiency (EE%) was determined using the ultrafiltration method as described by many investigators that encapsulated Q10 in different nanosystems [22,23,24,25]. The encapsulated amount of Q10 was calculated by the indirect method. Standard calibration curve of Q10 had been constructed following its quantification spectrophotometrically at 275 nm (Spectro UV-VIS double beam, Labomed Inc., USA). After formation of NMQ10, free and encapsulated Q10 were separated by ultrafiltration using centrifugal filter tubes (Millipore, Billerica, MA, molecular weight cutoff of 10 kDa) at 5000 rpm for 30 min. Free Q10 in the ultrafiltrate (free Q10) was measured spectrophotometrically at 275 nm. Also, NMP was used as parallel control under the same conditions to eliminate the impacts of the background. The experiments were achieved in triplicate and a mean value ± SD was obtained. Loading efficiency (LE%), EE%, and percent yield (Y%) of the NMQ10 were calculated according to the following equations [26]:

$$ \mathrm{EE}\%=\frac{\mathrm{Total}\ {\mathrm{Q}}_{10}-\mathrm{Free}\ {\mathrm{Q}}_{10}}{\mathrm{Total}\ {\mathrm{Q}}_{10}}\times 100 $$
$$ \mathrm{Y}\%=\frac{\mathrm{Wt}\ \mathrm{of}\ \mathrm{lyophilized}\ {\mathrm{NM}}_{\mathrm{Q}10}}{\mathrm{Wt}\ \mathrm{of}\ {\mathrm{Q}}_{10}+\mathrm{Wt}\ \mathrm{of}\ {\mathrm{K}}_{407}}\times 100 $$
$$ \mathrm{LE}\%=\frac{\mathrm{Total}\ {\mathrm{Q}}_{10}-\mathrm{Free}\ {\mathrm{Q}}_{10}}{\mathrm{Wt}\ \mathrm{of}\ \mathrm{lyophilized}\ {\mathrm{NMQ}}_{10}}\times 100 $$

Analysis of hydrodynamic size, polydispersity index, and zeta potential

Hydrodynamic size (Dh), polydispersity index (PDI) and zeta potential (ζ potential) of freshly prepared NMQ10 were determined by dynamic light scattering (DLS) using Malvern Zetasizer Nanoseries (Malvern Instruments Limited, UK) with a scattering angle of 90° at 25 °C. Before measurement, all samples were suitably diluted with DIW to get a homogeneous distribution in the measurement cell as well as an appropriate scattering intensity.

Transmission electron microscopy

The morphology of NMQ10 was identified using transmission electron microscopy (TEM) (JEOL JEM-2100, JEOL Ltd., Tokyo, Japan). From the original NMQ10 suspension, 1 ml was diluted properly with distilled water to obtain a diluted sample. Then, one drop of the diluted sample was transferred onto a 400-mesh carbon-coated copper grid. Using a Whatman filter paper, excess liquid was removed and the grid was allowed to dry in the air at room temperature. The TEM images were captured using Digital Micrograph and Soft Imaging Viewer software.

Fourier-transform infrared spectroscopy

Spectroscopic studies of Q10, K407, physical mixture, NMP, and NMQ10 were performed by using Thermo Fisher FT-IR spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Physical mixture of Q10 and K407 at a concentration of 1:2, respectively, was prepared simply by mixing using a spatula. The scanning scale ranged from 500 to 4000 cm−1 at room temperature. Then, the absorbance peaks were examined based on peak frequencies and domain.

Stability studies

The physical stability of NMQ10 dispersions was monitored under the refrigerated conditions for 12 months. The NMQ10 dispersions were freshly prepared, transferred to glass bottles, and maintained at 4 ± 2 °C without any agitation or direct light. The physical appearance, Dh, PDI, ζ potential, and percent Q10 leaked out of the NMQ10 (% LQ10) were analyzed at different time (t) intervals of 0, 1, 3, 6, and 12 months. For NMQ10, the entrapped amount of Q10 was determined as mentioned under the section “Determination of encapsulation efficiency, loading efficiency, and percent yield” using NMP dispersions that were concomitantly prepared and stored under the same conditions and time to serve as a control. Percent leaked of Q10 (% LQ10) was calculated as stated by the following equation [27]:

$$ \%{\mathrm{L}}_{\mathrm{Q}10}=\frac{{\mathrm{EE}}_0-{\mathrm{EE}}_t}{{\mathrm{EE}}_0}\times 100 $$

Where EE0 is the amount of Q10 entrapped at the preparation time (0 time) and EEt is the amount of Q10 that was still entrapped after the specified time t of storage. Moreover, to express relatively the potential micelles’ photoprotective effect of Q10, ethanolic solution (EtQ10) was prepared using the same concentration of Q10 as NMQ10 [22, 28]. EtQ10, NMP, and NMQ10 dispersion were freshly prepared, sealed, and exposed to natural day light at room temperature for a period of 144 h. Samples were withdrawn at 0, 1, 2, 4, 6, 8, 12, 24, 48, 72, and 144 h after exposure to day light and assayed for Q10. In case of EtQ10, the concentration of Q10 was determined spectrophotometrically at 275 nm after appropriate dilution, while Q10 entrapped amount in NMQ10 was determined as mentioned under the section “Determination of encapsulation efficiency, loading efficiency, and percent yield” and % LQ10 of both NMQ10 and EtQ10 was calculated. All evaluations were performed in triplicate and the data then were analyzed as % retention of Q10 (% RQ10) as the following:

$$ \%{\mathrm{R}}_{\mathrm{Q}10}=100-\%{\mathrm{L}}_{\mathrm{Q}10} $$

Preparation of Q10 dispersion, NMQ10 dispersion, and in situ gelling systems

From the previously published data, it was found that various DDSs were prepared by incorporating a wide range of Q10 concentrations: 1% [29], 2.8% [22], 3% [30], 5% [31, 32], 6% [33], 10% [13], and 13% [34]. For all formulations in our study, Q10 was incorporated at concentrations of 5% w/w. Q10 or equivalent amounts of NMQ10 at a concentration of 5% w/w were simply dispersed in water under magnetic stirring for preparation of Q10 dispersion (FQ10) and NMQ10 dispersion (FNM), respectively. Table 1 shows the compositions of the prepared formulations.

Table 1 The compositions of the prepared formulations
Full size table

It was highly plausible to incorporate NMQ10 in a carrier for the purpose of local application. The association of nanocarriers and gels offers new therapeutic possibilities due to merging of the advantages of both nanotechnology and gel system. Gels support the nanocarriers, localize their administration to the target tissue, and may sustain their release [35]. Gels loaded with polymeric micelles have also been reported in the literature [36,37,38,39,40]. In situ gelling has been widely studied and optimized in the literature. Two techniques—thermoreversible (F1) and pH-responsive (F2)—were designated as carrier systems of NMQ10 (Table 1) [41, 42]. Thermosensitive systems are identified to undertake sol-to-gel transition as their temperatures are increased, and upon cooling, they return to the original sol phase. On the other hand, the pH-responsive in situ gels are characterized by their ability to gel at a characteristic physiological pH.

F1 was composed of CP (bioadhesive polymer) and HPMC (viscosity-modulating polymer) at concentrations of 0.2 and 0.4% w/w, respectively. It was prepared by allowing HPMC and CP to be soaked overnight using an adequate amount of water. Next, TEA at a concentration of 0.1% was added dropwise at a sufficient amount to neutralize CP. An amount of NMQ10 equivalent to 5% w/w of Q10 was dispersed in a part of water and incorporated to the obtained system [31, 32]. Finally, the total volume was completed and the system was stirred until homogeneity. On the other hand, F2 was composed of 1% w/w MC salted by 5% w/w NaCl. MC was dispersed in hot water at 70 °C and allowed to be soaked overnight with continuous stirring until a homogeneous polymeric mixture was obtained. After that, NaCl at a concentration of 5% was dissolved in a part of water and incorporated to the polymeric mixture. NMQ10 was incorporated and the procedure was completed as mentioned under F1.

In vitro release of Q10

Release of Q10 from FQ10, FNM, F1, and F2 was performed using modified vertical Franz diffusion cell [43]. Initially, each receptor half-cell was filled with 66 ml phosphate buffer pH 6.8 (PB6.8) containing 1% T80 as a receptor solution to maintain sink condition. Donor and receptor compartments were separated by a semipermeable cellulose membrane (Spectra/Por® Dialysis Membrane, molecular weight cutoff 12–14 kDa, Spectrum Laboratories, Inc., CA, USA) that was equilibrated overnight with PB6.8 prior to the experiment. Amounts of FQ10, FNM, F1, and F2 equivalent to 50 mg of Q10 were transferred to the donor compartments. Donor half-cells were covered with a wax foil (Parafilm® M, Bemis Company Inc., Oshkosh, WI, USA) to avoid evaporation of water. The entire diffusion cells were positioned in a shaking incubator (GFL, Gesellschaft fur Labortechnik GmbH, Burgwedel, Germany), kept at 37 ± 0.2 °C, and stirred at 100 rpm during the experiment. Samples of 1 ml each were withdrawn at predetermined time intervals and replaced with equal volumes of fresh receptor solution. The collected samples were filtered through 0.45 μm Millipore filter and the amount of Q10 in the receptor solution was analyzed spectrophotometrically at 275 nm (Spectro UV-VIS double beam, Labomed Inc., USA).

Kinetic modeling of release data

To determine the mechanism of Q10 release, different mathematical models including zero-order, first-order, Higuchi’s square root of time [44], and Korsmeyer–Peppas (K–P) equation [45] were used. The correlation coefficients (r2) were calculated from linear regression plots of ft versus t, log (100 − ft) versus t, and ft versus squared root of t, related to zero-order, first-order, and Higuchi’s model, respectively. In these plots, ft denotes the cumulative percentage of Q10 released at time t, and 100 − ft is the percent of Q10 nonreleased and remained after time t. The model representing r2 with the highest value was considered to illustrate Q10 release mechanism. For the K–P model, the equation was as follows:

$$ {f}_t/{f}^{\infty }={k}_{\mathrm{p}}{t}^n $$

Where ft/f∞ is the fraction of Q10 released after time t, n is a distinguishing exponent for the release mechanism, and kp is the K–P release rate constant. According to the K–P equation, the exponent n values ≤ 0.5 indicate Fickian diffusion mechanism, while n values that ranged from 0.5 to 1 are representative of an anomalous mechanism. Instead, zero-order release is reflected when n = 1. From our findings, it was inferred that F1 had successful and promising results that encouraged us to progress its investigation to clinical practice.

Clinical study

Patient selection

The Research Ethical Committee at Faculty of Dentistry, Mansoura University, Mansoura, Egypt, approved the protocols regarding human (NIH publication No. 85-23, revised 1985). Fifteen subjects were selected for this study from patients seeking periodontal treatment at the Department of Oral Medicine and Periodontology, Faculty of Dentistry, Mansoura University. Inclusion criteria include moderate periodontitis patients with at least 20 teeth in the mouth. PDP depth was 3–5 mm with radiographic evidence of bone loss and clinical attachment level 3–4 mm. Age of patients ranged from 35 to 55 years. On the other hand, the exclusion criteria encompass history of any systemic disease, periodontal treatment for at least 3 months before starting the study, pregnant patients, smokers, and patients with bleeding disorder. The selected patients were briefed about the treatment types (SRP and periodontal therapy), duration of treatment, and number of follow-up visits. After obtaining written consent from all the selected patients to participate in the trial, oral hygiene instructions were given.

Periodontal evaluation

Each patient was thoroughly examined clinically to assess the gingival tissue health state in terms of color, size, texture, and contour. F1 was prepared by incorporating NMQ10 in a thermoresponsive carrier system of CP and HPMC as mentioned under the section “Preparation of Q10 dispersion, NMQ10 dispersion, and in situ gelling systems” and Table 1. A randomized split mouth design was followed where one side was assigned to receive F1 (study side), and the other side was considered as a control side [46]. Then, patients were subjected for periodontal evaluation of both sides (baseline of periodontal evaluation). Periodontal evaluation was performed using Williams periodontal probe (Nordent Inc. USA). Periodontal evaluation included assessment of gingival index (GI) [47], plaque index (PI) [48], probing pocket depth (PPD) in millimeters measured in six sites for each tooth, clinical attachment level (CAL), and bleeding on probing (BOP) [49]. Table 2 shows the scoring system for the periodontal evaluation. GI and PI were recorded at four sites per tooth: distofacial papilla, facial margin, mesiofacial papilla, and entire lingual gingival margin. PPD was measured from the gingival margin to the base of the pocket at six points: mesiobuccal, midbuccal, distobuccal, mesiolingual, midlingual, and distolingual around each tooth. CAL was measured from the cementoenamel junction to the base of the pocket. The score of each tooth was obtained by adding all the scores per tooth and dividing it by the number of sites. The index score for each patient was obtained by adding the index scores per tooth and dividing it by the number of teeth examined.

Table 2 Scoring system for the periodontal evaluation
Full size table

Biochemical assay

Collection of GCF samples

First, the study sites were isolated with cotton rolls and PDPs were gently dried with air. Then, GCF samples were collected from the pockets using PerioPaper™ strips over 30 s. Strips contaminated with saliva or blood were discarded [50]. Two samples were collected from the deepest pocket in the study side. Concurrently, two samples were obtained from the opposite PDPs in the control side. Samples were transferred to Eppendorf tubes each containing 500 μl of phosphate buffer pH 7.4 (PB7.4) and stored for laboratory investigations of MDA and T-AOC (baseline evaluation of MDA and T-AOC).

Determination of MDA and T-AOC

Both MDA as a biomarker of lipid peroxidation and T-AOC were assessed according to the instructions supplied by the assay kits. Their values could indirectly reflect the level of cellular damage [51]. MDA is the catabolite of lipid peroxide and it can react with thiobarbituric acid (TBA) to produce a red compound. This colored compound could be measured at its maximum absorption peak of 532 nm. The determination of T-AOC is based on the 2,2′-azino-bis-3-ethylbenzthiazoline-6-sulfonic acid (ABTS) method. In the presence of a proper oxidant, ABTS is oxidized to its green oxidized form (ABTS+). Therefore, the existence of antioxidants can inhibit the formation of ABTS+. The T-AOC of the sample can be measured and calculated by determining the absorbance of ABTS+ at 414 or 734 nm. After baseline evaluation of periodontal parameters, MDA and T-AOC, all patients were subjected to full mouth SRP.

Periodontal treatment of the study sides

After baseline evaluation that was followed by full mouth SRP, a 3-ml syringe with atraumatic needle was filled with F1. In the study side of each patient, the tip of the needle was inserted in the deepest point of the periodontal pockets until reaching its base. Then, 0.1 ml of F1 was injected gently to the pocket followed by slow withdrawal of the needle. The application was repeated each alternate day for a week. All the patients were directed not to rinse or drink for 1 h after application and also oral hygiene instructions were given to them. Periodontal parameters were re-evaluated and GCF samples were recollected from the study side and control side 6 weeks after finishing the treatment with F1. Also, MDA and T-AOC were redetermined as stated before. To calculate the supplementary benefit of F1 beyond SRP only, the percent reduction (% reduction) of each parameter was calculated using the following equation for both study and control sides.

$$ \%\mathrm{Reduction}=\frac{P_0-{P}_{\mathrm{re}}}{P_0}\times 100 $$

Where P0 is the value of a parameter at baseline and Pre is the value of the same parameter upon re-evaluation.

Statistical analysis

The parameters of periodontal evaluation, T-AOC, MDA, and % reduction, were represented as mean ± SD. The obtained results were statistically analyzed by Student’s t test at a significance level of P < 0.05 using GraphPad Prism software (version 6.00; GraphPad software, San Diego, USA).

Results and discussion

Characterization of NMQ10

Determination of encapsulation efficiency, loading efficiency, and percent yield

NMQ10 were prepared using the nanoprecipitation technique which is applicable for entrapment of hydrophobic drugs such as Q10. The encapsulation of this poorly water-soluble compound into nanosized particles may enhance its bioavailability. The nanoprecipitation technique resulted in the formation of NMQ10 as a colloidal dispersion with a strong yellow color. Hence, color uniformity during preparation could indicate homogeneous distribution and dispersion of Q10 in the aqueous medium [21].

Ultrafiltration was used to distinguish between Q10 that was entrapped inside NM and the other free one that was outside [22,23,24,25, 52]. Similarly, recent studies on other lipophilic drugs such as gemcitabine reported the measurement of EE% by the ultrafiltration method [53,54,55]. Indeed, we used centrifugal filter tubes with molecular weight cutoff of 10,000 Da at 5000 rpm for 30 min. The molecular weight of Q10 is 863.37 Da [28, 56] and the Dh of NMQ10 was found to about 150 nm as indicated later under the section “Analysis of Dh, PDI, and ζ potential.” According to the following equation of a spherical particle [57]:

$$ {R}_{\mathrm{min}}=0.066\ {M}^{1/3} $$

Rmin is the minimal radius of a sphere that could contain a given mass of M in Dalton. Only nonencapsulated Q10 could pass through the pores of the ultrafiltration membrane under centrifugation, whereas encapsulated Q10 could not. Hence, the free Q10 in the ultrafiltrate could be determined spectrophotometrically at 275 nm without further dilution.

Values of EE%, LE%, and Y% are listed in Table 3. EE% was estimated to be 99.4% which indicated that almost all of Q10 was encapsulated. High EE% could be attributed to the high lipophilicity of Q10 as well as its very poor aqueous solubility [23]. Such sufficient drug entrapping is one of the prerequisites for subsequent delivery of the drug in a proper amount that is sufficient to exert its pharmacological effect [58]. The characteristics of K407 such as its amphiphilic nature might positively affect EE% [24, 58]. K407 was used at a concentration higher than its CMC that was reported to be 0.012 mg/ml [20]. The attained high EE% of Q10 was found to produce an excellent LE% of about 38%. In order to expect the scale-up of the preparation process, Y% of NMQ10 was determined by weighing the product gained after lyophilization. The preparation technique exhibited relatively high recovery percent (Y%) and was found to be 86.30 ± 3.033%.

Table 3 Physicochemical characteristics of the prepared NMs
Full size table

In our study, results of EE%, LE%, and Y% were analogous to the reported ones in other studies incorporating Q10 in several nanosystems. Arranz-Romera and coworkers reported that EE% of Q10 was 95.66% with high Y% of about 75% [59]. Optimized Q10-loaded nanostructured lipid carrier (NLC) for efficient management of wrinkle showed EE% of 93.20% [60]. The mucoadhesive nanoparticles for oral delivery of Q10 were prepared using chitosan and dextran sulfate sodium salt with entrapment efficiency reaching 98% [61]. The EE% and LE% of Q10-loaded micelles studied were found to be 99.39 and 13.77%, respectively [62]. TPGS–chitosome as effective oral delivery systems for improving the bioavailability of Q10 was investigated and EE% appeared to be high within the range of 76–84% in all three liposomal formulations [63]. Novel chitosan-coated liposomes containing Q10 exhibited EE% in the range from 95.4 to 97% [25]. Chen et al. formulated 17 Q10-loaded NLCs to find that EE% and LE% values ranged from 95.8 to 100% and from 2.25 to 2.68, respectively [22]. NLCs incorporating Q10 were also studied by Xia and Wang and showed that EE% of about 99.8% was obtained by all formulations [23]. In the case of Q10-loaded liposomes formulated by Zhang and Wang, Q10 was incorporated into the hydrophobic compartment of vesicles with 98% encapsulation efficiency [52]. The authors of the abovementioned studies demonstrated that the Q10 molecule comprises a quinone ring with a long isoprene side chain, which renders it lipophilic and favorable for incorporation within the core of the nanosystem to be kept at higher EE%.

Analysis of Dh, PDI, and ζ potential

The values of Dh, PDI, and ζ potential of NMQ10 were found to be 154.0 ± 6.3 nm, 0.108 ± 0.017, and − 31.67 ± 0.55 mV, respectively, as shown in Table 3. Regarding nanosystems, not only their Dh is prime but also the width of the particle size distribution expressed as PDI. These essential characters of nanosystems influence drug release rate, biodistribution, and colloidal dosage form upon storage. It can be inferred from our results that the small nanosize of NMQ10 might render them able to penetrate sites such as PDPs below the gingival line that may be inaccessible by DDS. Besides, PDI values showed narrow distribution, and thereby the prepared systems were considered to be homogeneous and could retain the colloidal stability without formation of aggregates.

Table 3 documents that both NMQ10 and NMP were negatively charged. In the structure of K407, both the PPO and PEO fragments were nonionic, so that the presence of Q10 might contribute to the surface negative charge of NMQ10. Such elucidation could be observed by increasing the absolute value of ζ potential after incorporation of Q10 from − 22.93 (NMP) to − 31.67 (NMQ10) mV. Moreover, ζ potential has a vital effect on the storage stability of a colloidal dispersion system. Aggregation of particles is likely to arise if ζ potential is too low to offer sufficient electric repulsion or to enhance steric barriers between particles. For high molecular weight stabilizers, ζ potential absolute values below 20 mV can afford sufficient stabilization. However, for sterically stabilized nanosuspensions, even lower values of ζ potential were perceived (down to − 3 mV). This further decrease of ζ potential could be attributed to the fact that the adsorbed layers of polymers/large molecules shift the plane of shear to longer distance from the particle surface. This in turn may reduce the measured value of ζ potential [64,65,66]. Hence, it could be indicated that the ζ potential of the prepared NMQ10 was acceptable.

Transmission electron microscopy

TEM micrographs of NMP and NMQ10 are illustrated in Fig. 1. It can be noted that both NMP and NMQ10 are spherical. It could be observed that the use of K407 (triblock copolymer) allows the formation of “flower-like” polymeric NMs especially in the case of NMQ10 (Fig. 1b). NMQ10 composed of small hydrophobic chains (PPO56) directed toward the core that encapsulate Q10. Each hydrophobic chain is flanked between two longer hydrophilic chains (PEO101) that form the shell. Hence, the outer shell was related to the hydrophilic PEO and the inner core could be attributed to hydrophobic PPO [67]. Such assembly gives NMs the chance to have several well-differentiated compartments [68]. Additionally, TEM micrographs confirmed the measured average Dh of NMQ10.

Fig. 1
figure 1

TEM micrographs of plain nanomicelles: NMP (a) and Q10-loaded nanomicelles: NMQ10 (b)

Full size image

By comparing NMP size obtained by means of DLS and TEM, a difference could be observed. Several studies documented a similar observation [69,70,71,72]. Although TEM is used to highlight the morphology, size, and shape of the particles, an extensive indication of particle size and size distribution could be obtained by the DLS. Because DLS measures a hydrodynamic z-average value of size, rather than a true physical size of individual particles by TEM, the DLS technique may possibly display a larger size. Hydrodynamic diameter includes the core plus any molecule attached or adsorbed on the surface of the micelle [72]. Moreover, the larger particle size of NMP could be attributed to some small aggregations in the aqueous dispersion [69]. In our work, we focused on DLS for measurement of z-average size and for stability study. For TEM, only one drop of the dispersion whole volume (in our study about 100 ml) was transferred onto a 400-mesh carbon-coated copper grid. TEM image might be not representative and so it was used to analyze morphology and shape rather than size.

Fourier-transform infrared spectroscopy

Compatibility and interaction of Q10 and K407 in NMQ10 can be tested by Fourier-transform infrared spectroscopy (FT-IR) analysis. It could be achieved by noticing changes that occurred in the positions of characteristic bands when compared to those of pure Q10 and K407. The FT-IR spectra of Q10, K407, physical mixture, NMP, and NMQ10 are demonstrated in Fig. 2. The spectrum of Q10 (Fig. 2a) disclosed all the distinctive bands corresponding to its functional groups. A small broad band at 3500 cm−1 (OH groups), a very high intense one at 3000 cm−1 (C=C), peaks from 1000 to 629 cm−1 (=C–H), 1611 cm−1 (the benzoquinone ring) and 1649 cm−1 (mono substituted isoprenoid) appeared. For K407 (Fig. 2b), its spectrum typically showed characteristic bands at 2877 cm−1 due to C–H stretching (aliphatic), 3446 cm−1 for O–H stretching, and 1107 cm−1 for C–O–C stretching. The spectrum of the physical mixture (Fig. 2c) shows the combined bands of both Q10 and K407 with reduced intensities of Q10 bands as a consequence of the dilution effect. In the same way, the FT-IR spectra of NMP and NMQ10 (Fig. 2d, e) coincide with that of the physical mixture. The FT-IR spectra exhibited a high degree of superimposition and the characteristic peaks of Q10 were clearly observed at their corresponding wave numbers. Hence, it was indicating high compatibility and nonappearance of any potential of chemical or ionic interaction between Q10 and K407.

Fig. 2
figure 2

FT-IR spectra of Q10, K407, physical mixture (Q10+K407), plain nanomicelles (NMP), and Q10-loaded nanomicelles (NMQ10)

Full size image

Stability studies

Stability is a major concern in the improvement of any formulation especially for unstable drugs such as Q10. The matter of size and ζ potential are more vital for nanosystems than other DDS because of the extensively higher surface area afforded by NMs [73]. The physical stability of NMQ10 was determined by observing changes that took place in the values of Dh, PDI, ζ potential, Q10 content, and appearance during the storage period. No degradation, obvious color change, phase separation, or other physical change was observed during the storage at refrigeration. Table 4 summarizes the values of Dh, PDI, ζ potential, and % LQ10 at the refrigerated conditions. The results elucidated that Dh retained the nanosize range with a unimodal homogeneous distribution (PDI did not exceed 0.209) throughout the storage period. ζ potential measurement was another method for the assessment of stability of NMQ10 nanosuspension. It was observed that the steric stabilization of NMQ10 was sufficient to exhibit satisfactory ζ potential during the storage period. Also, it can be concluded from Table 4 that there was no leakage of Q10 up to 3 months and % LQ10 did exceed 2.78% after storage for 1 year. This means that about 97.22% of the entrapped Q10 was still present in the core of NMQ10. Therefore, these findings were indicating proper physical stability of NMQ10 with homogeneous size, reasonable ζ potential, and their protection of Q10. Long swinging hydrophilic polyoxyethylene strands on the particle surface could offer an excellent steric hindrance, which precludes the particles from aggregating. Furthermore, the homogeneous NMQ10 might hinder the leakage of entrapped Q10 and the dissolution of smaller particles as well as the growth of larger particles, i.e., Ostwald ripening [66].

Table 4 Dh, PDI, and ζ potential of NMQ10 aqueous dispersions stored at refrigerated conditions
Full size table

Furthermore, Q10 is a light-sensitive compound, and upon exposure to light, it gradually decays and its yellow color turns darker [74]. The degradation profiles of Q10 in NMQ10 and EtQ10 exposed to day light for about 144 h are given in Fig. 3. It was observed that the photodegradation of Q10 in EtQ10 was very fast, whereas NMQ10 significantly contributed to the photoprotection of Q10. Moreover, 18 h after day light exposure left only about 39.2% of the initial concentration of Q10 in EtQ10. On the other hand, NMQ10 after 18 h still entrapped Q10, prohibited its leakage, and might protect it against photodegradation. After 144 h of exposure to day light, about 96% of the initial concentration of Q10 was still present in NMQ10 (Fig. 3). This indicated that the incorporation of Q10 into the hydrophobic inner core of NMs might remarkably improve its stability. This might be due to the stronger interaction between Q10 and the hydrophobic component (PPO) of K407, which helps to confine the drug molecules within the protective micellar compartment [75].

Fig. 3
figure 3

The degradation profiles of Q10 in encapsulated form (Q10-loaded nanomicelles; NMQ10) and in free form (ethanolic solution; EtQ10) upon exposure to day light

Full size image

It is reported that upon exposure to light, gel-forming polymers themselves such as carbomer and cellulose derivatives in their dispersion form undergo oxidation that is reflected in a decrease in their dispersion viscosity. Therefore, the stability of such dispersions to light usually can be improved by the addition of a water-soluble UV absorber [76, 77]. Practically, the prepared in situ gels were found to be transparent and colorless and could allow passage of light throughout the gel matrix with humble photoprotection. In other words, gel-forming polymers themselves could not protect their aqueous matrices versus light-induced oxidation which in turn could negatively interrupt the stability of a light-sensitive drug. So, we did not try to incorporate such a light-sensitive drug: Q10 as a “free” molecule in the gel formulation. As an alternative, we study the photoprotective effect of NMQ10 versus EtQ10.

The aforementioned results showed that the encapsulation of Q10 in NMs exhibited a relatively significant photoprotective effect when compared to free Q10 (EtQ10). This stability was enabling its efficacy of NMQ10 for prolonged protection period of Q10. Consistent with our observations, it was reported that the stability of Q10 is enhanced by nanoencapsulation [61, 78]. It is worthy to be highlighted that our aim of this relative determination of NMQ10 to EtQ10 was to attain an initial idea about the potential photoprotective effect of NMs. Hence, the stability of the prepared NMQ10 aqueous dispersion upon storage was promising and encouraged us to complete further studies investigating photoprotection in our future work.

In vitro release of Q10 and its kinetic analysis

Based on literature optimized work, two in situ gelling systems were selected to be used as NMQ10 carriers [41, 42]. First, carbopol 934 and HPMC at concentrations of 0.2 and 0.4% w/w were used for the preparation of F1. On the other hand, F2 was prepared using MC at concentration of 1% w/w containing 5% NaCl.

An appropriate design to conduct in vitro release was necessary to create sink conditions for highly lipophilic Q10. The design of the release experiment was based on the literature reported data as discussed below. Q10 is poorly soluble in aqueous medium because of its unionizable nature. Thus, a release medium should be introduced to provide adequate aqueous solubility and sink conditions. The medium should have the ability to discriminate the release patterns of different formulations. By surveying literature, phosphate buffer was proposed as physiologically more relevant media for in vitro/in vivo correlation. Different pH values—5.8 [25, 60], 6.5 [79], 6.8 [30, 34], 7.4 [60], and 7.5 [29]—were used based on the physiological pH of the action site. Phosphate buffer at pH 6.8 was used as a release media to simulate pH of GCF that fills PDPs [80, 81].

Adequate release medium providing sink conditions was not obtained with only aqueous solutions within physiological pH [82]. For this reason, solubilization using surfactants in the release medium provides an attractive approach to obtain the desired medium [83, 84]. Commonly acceptable ionic and nonionic surfactants include sodium lauryl sulfate and Tweens. The nonionic surfactant Tween 80, being considered more biologically relevant, was our first choice as surfactant for in vitro release medium [82]. Among a number of concentrations, Liu and coworkers reported that the highest percentage of released Q10 was obtained in 1% Tween 80 [62]. A similar concentration of Tween 80 for release medium of Q10 was used by Ankola et al. [82]. Hence, we ultimately decided to use phosphate buffer pH 6.8 containing 1% Tween 80 as a receptor release medium to maintain sink conditions. Additionally, the volume of fresh release medium (total about 12 ml for each cell throughout the experiment) that was added after each withdrawal could guarantee the sink condition.

In vitro release profiles of Q10 from different formulations and free Q10 suspension in release medium are illustrated in Fig. 4. It could be observed that Q10 showed biphasic release with burst release within the first 24 h which was preceded by a gradual pattern up to 4 days. Such biphasic release has been generally familiar for pluronic micelles [20, 68, 85,86,87]. The initial burst Q10 release could be attributed to the drug part which may be found on the interface of the hydrophilic corona and the micelle hydrophobic center or found inside the movable micellar corona. This part of Q10 might be released via hydration of the interfacial Q10 molecules followed by passive diffusion. On the other hand, the other part of Q10 was entrapped within the inner core compartment. Hence, this part remained hidden in the micellar core, and its release was lagged to give the second phase of release pattern after the end of the burst release phase. Fortunately, the therapeutic concentration of Q10 could be delivered by such initial burst release, and at the same time, keeping the Q10 concentration at the therapeutic levels may be achieved by the subsequent slow release [87]. The release profile of coenzyme Q10 from FQ10 (aqueous dispersion) was biphasic. This initial burst release phase was followed by a second slower release phase. These two phases were likely related to the rapid release and diffusion of molecular coenzyme Q10 that was solubilized at the beginning of the release process. It is well identified that diffusion can be driven by the potential energy of a concentration gradient between a formulation and release medium [88, 89]. Then, initial release was followed by slower dissolution-limited release of coenzyme Q10 that still remained in particulate form in FQ10.

Fig. 4
figure 4

In vitro release profiles of Q10 from FQ10, FNM, F1, and F2 in phosphate buffer pH 6.8 (1% Tween 80) using modified Franz diffusion cell and dialysis membrane with molecular weight cutoff 12–14 kDa

Full size image

Regarding the release profiles and amounts of Q10, we found them comparable to the reported release data of Q10 in the literature [22, 62, 82, 83]. In the in vitro release study of Ankola et al., the amount of Q10 remaining in the nanoparticles was estimated to be about 84% at the end of 21 days. The amount remaining demonstrated that Q10 did not undergo any major degradation due to the protective effect of the nanosystem. For such a photosensitive drug, factors other than solubility are to be considered such as stability [82]. Moreover, the dialysis membrane might be an additional barrier despite the use of a high molecular weight cutoff (12,000–14,000 Da), hence slowing down the release rates [84, 90]. The cumulative Q10 released amounts (μg) were determined to be 1163.06, 1666.47, 1471.5, and 1250.25 μg for FQ10, FNM, F1, and F2. Among others, FNM and F1 exhibited comparatively higher cumulative released amounts of Q10 over 6 days. Based on our findings, F1 was selected to serve as an optimized formulation for further clinical studies.

Table 5 illustrates the results of kinetic analysis of Q10 release data. The results exposed that in vitro release of Q10 from all formulations was best expressed by the Higuchi model that suggested Q10 release was controlled by diffusion. It is evident that the hydrophilicity of a polymer can direct the quickness of water uptake during drug release. For NMQ10, the water uptake into micelles happened at the beginning due to the hydrophilic nature of PEO strands in K407 allowing the diffusion of Q10. However, the hydrophobic core of NMQ10 formed of chains of PPO delayed the diffusion of water into the core and hindered the dissolution of NMQ10 on further dilution with release medium. Accordingly, the general diffusion of Q10 from NMQ10 was gradual and donated a sustained release profile [87]. In the case of NMQ10, the release and diffusion of Q10 from NMs could be considered as the rate-limiting step. On the other hand, dissolution was often the rate-limiting step of release of a pure unprocessed lipophilic drug. It was reported that encapsulation of lipophilic drug molecules into NMs eliminates the significance of the dissolution step as the drug molecules are well dispersed within the core of NMs [91]. From our conclusions, it was inferred that F1 had successful and promising results that fortified us to progress its investigation to clinical practice.

Table 5 Kinetic analysis of the percentage Q10 released from FQ10, FNM, F1, and F2
Full size table

Clinical study

In vitro studies on Q10 were considered as a sort of investigational preclinical study [91]. Based on the preclinical results, the optimized formulation F1 was selected for further evaluation. Our pilot-scale first clinical trial aimed to get initial clinical information after topical application [92,93,94,95,96]. In the literature, similar procedures of niridazole biodegradable inserts for local long-term treatment of periodontitis [93], dexamethasone sodium phosphate-loaded chitosan-based delivery systems for buccal application [94], green tea strips as a local drug therapy on periodontitis patients with diabetes mellitus [96], liposomal-benzocaine gel formulation [95], and in vivo topical anesthesia in volunteers via administration of liposome preparation [92] were followed.

For a clinical trial material (CTM) as stated by Stephon [97], current good manufacturing practices (CGMP) requirements are to be applied in an incremental approach as A, B, C, and D according to clinical development phase (I, II, III), including preclinical. The first level of CGMP (level A) includes preparation of intermediates and final forms of active pharmaceutical ingredient (API) on the small laboratory scale for preclinical explorative study (nongood laboratory practices non-GLP). The more advanced level of CGMP (level B) comprises the preparation of API intermediates and final API, bulk drug and finished product for initial pilot study, advanced preclinical studies, and preliminary stability batches (to support GLP nonclinical studies, advanced preclinical and phase I clinical studies) [97].

This work was designed to include preclinical formulation and evaluation of Q10 followed by initial pilot clinical study. Consequently, the CGMP requirements of levels A and B supporting GLP were required to be satisfied in our study. Actually, these measures had been followed during preparation. Also, preparation of F1 was to be considered as a mini-production operation of a topically applied and locally delivered API (Q10). In our lab, we applied all standard measures of aseptic manipulation, good documentation, good housekeeping, and good scientific practices. All formulations were prepared by the same operator.

In the case of our clinical study, we had estimated periodontal parameters and quantitatively assayed biochemical parameters such as MDA and T-AOC that had reflected Q10 antioxidant effect [1, 9, 25]. According to our design, the antioxidant effect of Q10 was aimed to be as an adjunct to SRP for the management of chronic periodontitis. Table 6 shows the parameters of the periodontal evaluation, T-AOC and MDA. After SRP, F1 was locally delivered into the periodontal pockets (the study side) and the other side was left without further treatment to serve as the control side. Hence, the effect of SRP only on the management of periodontitis could be detected at the control side, whereas the study side showed the extra effect of F1. SRP only demonstrated significant improvement (P < 0.0001) of all the periodontal parameters in excess to those before mechanical manipulation at baseline. The mean value of GI reduced from 2.07 to 1.06, the value of PI decreased from 2.13 to 0.97, and PPD also declined by about 26%. CAL and BOP showed significant decrease after 6 weeks to reach 1.83 and 0.96, respectively (Table 6). The degree of lipid peroxidation is measured by the levels of MDA that is stated to be an end product of lipid peroxidation. Laboratory examination of T-AOC and MDA revealed that their values diminished by 22.2% (P < 0.0001) and 23% (P < 0.05), respectively. This finding could be attributed to the early mentioned role of SRP as a key element for controlling and management of periodontitis [8]. However, it was found that some patients in the study showed marked recurrent increase of MDA even after SRP.

Table 6 Parameters of periodontal evaluation, T-AOC and MDA
Full size table

Local delivery of F1 offered improvement over baseline with significant reduction (P < 0.0001) of GI, PI, and PPD. Similarly, significant amending of CAL and BOP was found upon application of F1. GI values decreased from 2.04 to 0.88, PI decreased from 2.06 to 0.80, and PPD also shallowed by about 41%. Likewise, the values of BOP decreased from 2.72 to 0.81. As illustrated in Table 6, T-AOC and MDA levels of the study side were significantly reduced by about 51.59% (P < 0.0001) and 42% (P < 0.01), respectively. Table 6 also documents that upon re-evaluation, the study side had significant (P < 0.05) lower values of the determined parameters than the corresponding ones of the control side. Additionally, comparing the % reduction of each parameter at the study side and control side revealed that F1 significantly (P < 0.01) enriched the improvement and management of periodontitis over SRP alone (Table 6). Furthermore, we calculated the % reduction of the periodontal parameters of data in the literature [8, 13] after combination of SRP and locally delivered Q10 as a mixture of Q10 and vegetable glycerin base (1:9) (Perio-Q™ gel, PERIOQ INC, USA). It was found that our developed formulation F1 has much higher % reduction of the periodontal parameters than the published ones. Moreover, a study of Pitale et al. [14] revealed that application of Perio-Q™ gel associated with SRP was not significantly different than SRP alone. These published studies lacked the biochemical assay of GCF.

The results of the clinical study could be attributed to many factors that govern the in vivo behavior of API. To be effective in vivo, APIs must meet certain criteria: reach the site of action, be maintained there in a sufficient concentration, and be maintained long enough for the intended effect to occur. However, PDPs resulting from the progression of periodontal disease are filled with GCF. Fortunately, PDPs offer reservoirs for the introduction of DDS. The turnover rate of GCF was calculated to be 40 times per hour [98, 99]. Therefore, GCF acts as a leaching medium that may enhance further release of API from DDS followed by its spreading throughout PDPs [2]. Hence, our intrapocket DDS (F1) was established on the prospects of maintaining effective high levels of drug in the GCF for a prolonged period of time to produce the desirable clinical benefits [2, 99]. These features, together with the fact that the periodontal diseases are localized to the immediate environment of the pocket, are to be considered for the explanation of our good clinical findings. Moreover, it is highly plausible to attribute the significant improvement of periodontal parameters upon application of F1 to the promoting effect of Q10 in the form of NMs. It was reported that the application of Q10 encapsulated in nanoliposomes after tooth extraction decreased the inflammatory process through controlling myeloperoxidase activity and nitric oxide concentration and accelerate the healing processes [10].

It can be inferred from our results that our developed NMs might bring successful control of periodontitis and provide additional benefits as compared to other conventional gel. Up to the authors’ knowledge, incorporation of Q10 in the form of an innovative and simple nanocarrier for the management of periodontitis has not been carried out to date. Such findings, besides our clinical and biochemical outcomes, augmented the assumption that the innovative nanomicellar carrier of Q10 with K407 could be used an adjunct approach to SRP for the management of chronic periodontitis.

Conclusion

Periodontal diseases are worldwide health problems that negatively affect the lifestyle of many populations. The production of ROS by the host and the plaque-forming microorganism worsen the conditions. Mechanical removal of plaque with association of local application of an antioxidant could enhance the management of periodontitis. There was a pressing need to find a novel approach that can harness the potential use of nanotechnology in the clinical arena. NMs of Q10 were prepared, evaluated, and incorporated in an optimized gel formulation (F1) to serve our goal. In the light of our results, it was found that NMQ10 can be suggested as a promising nanosystem and its application in the form of F1 after SRP resulted in significant and advanced management of the periodontal parameters and antioxidant activity over SRP alone. These early clinical results could be used as a guide to design and conduct larger-scale comparative clinical studies in our future plan.