Introduction

Diabetes mellitus is a chronic, complicated metabolic and endocrine disorder characterized by either absolute or relative lack of insulin due to destruction of the β cells of the islet of Langerhans, insufficient insulin production by the β cells of the islet of Langerhans, or resistance to the action of insulin [1]. The resultant effect of uncontrolled diabetes mellitus is hyperglycemia (fasting blood glucose > 126 mg/dl or > 7 mmol/L), leading to severe complications. It is the most typical endocrine–metabolic disorder characterized by chronic hyperglycemia giving rise to the risk of microvascular (retinopathy, nephropathy, and neuropathy) and macrovascular (ischemic heart disease, stroke, and peripheral vascular disease) damages, with associated reduced life expectancy and diminished quality of life [2]. According to the World Health Organization (WHO) statistics, the prevalence of diabetes has been rising more rapidly in middle- and low-income countries, and a majority of those living with diabetes live in low- and middle-income countries. The number of people living with diabetes has increased from 108 million in 1980 to 422 million in 2014, and the global prevalence of diabetes among adults over 18 years of age has risen from 4.7% in 1980 to 8.5% in 2014 [3]. In 2015, an estimated 1.6 million deaths were directly caused by diabetes, and another 2.2 million deaths were attributable to high blood glucose in 2012. Every six seconds, a person dies from diabetes, and 76% of diabetes-related deaths are in people under the age of 60 [4], and WHO projects that diabetes will be the seventh leading cause of death in 2030 [3]. A recent update by the International Diabetes Federation (IDF) atlas 2019, about 463 million adults (20–79 years) worldwide are living with diabetes, and this is projected to rise by about 51% to 700 million by 2045 [4]. According to the IDF, Africa, the region with the highest proportion of undiagnosed diabetes, albeit the lowest prevalence (3.9%), is projected to increase by 143.3% by 2045 [4].

Metformin hydrochloride is a biguanide and the most widely prescribed oral antihyperglycemic agent in managing type 2 diabetes mellitus [5]. However, despite its widespread clinical use, it is fraught with several problems such as poor absorption, low bioavailability, frequent dosage, high dose, short biological half-life, and gastrointestinal tract (GIT) side effects [6]. But by tactical engineering of lipid drug delivery systems, such as solidified reverse micellar solution (SRMS)-based solid lipid microparticles (SLMs) [5], nanostructured lipid carrier (NLCs), lipid drug conjugates (LDCs), and solid lipid nanoparticles (SLNs), these problems could be surmounted.

Attempts have been made to enhance the delivery of metformin; some of which includes the use of liposomes [7], PEGylated-mucin for oral delivery [8], alginate nanoparticles [9], hyaluronic acid nanoparticles [10], chitosan nanoparticles [11], chitosan/eudragit nanocomplex [12], niosomes [13], topical gel for derma delivery [14], solid lipid microparticles [5], solid lipid nanoparticle for transdermal delivery [15], reactive oxygen species (ROS)-responsive organosilica nanocarrier and metformin-pluronic-based polyurethanes for controlled drug delivery. The utilization of NLCs in drug delivery systems is because lipid matrices are safe, biocompatible, nontoxic, easily manufactured industrially, and versatile, thus bestowing its suitability for different routes of administration [16,17,18]. Nanostructured lipid carriers (NLCs) are second-generation lipid-based nanocarriers formed from a mixture of solid and liquid lipids and have unstructured matrix due to the different moieties of the constituents of NLCs [19]. The solid/liquid lipid blend ratio may vary from 70:30 to 99.9:0.1 [20,21,22]. The total lipid content of the NLC may vary from 5 to 40% [20,21,22,23], and the nanoparticles are stabilized in an aqueous medium using at least one surfactant in a concentration ranging from 0.5 to 5% w/w [5, 24, 25]. They were developed from solid lipid nanoparticles and have advantages over traditional lipid nanoparticles, such as improved drug incorporation and release properties [26]. In addition, they are more advantageous than any other nanoparticle preparation because of their oil-rich core, making for more significant drug entrapment [27]. Nanostructured lipid carriers came as a result of some of the failures/limitations of solid lipid nanoparticles, such as low payload, unpredictable gelation tendency, polymorphic transition, drug expulsion during storage [28], and high water content [17]. SLNs are prepared using solid lipids (i.e., lipids that are solid at room temperature as well as at body temperature) that are biocompatible and biodegradable [29]. The crystallinity of solid lipids affects the release properties of the SLNs derived [30], unlike in the matrix of NLCs, which is composed of a mixture of spatially different lipid molecules, which makes more imperfection in the matrix to accommodate more drug molecules than SLN [28, 31,32,33].

Davies and Abuchowsky first described PEGylation in the 1970s. It is the modification of a protein, peptide, or non-peptide molecule by linking one or more polyethylene glycol (PEG) chains. The polymer is usually nontoxic, non-immunogenic, non-antigenic, highly soluble in water, and FDA-approved [34]. It is a promising and extensively studied strategy for improving the performance of drugs and bioactives; it is mainly characterized by the use of very small quantities of substances for therapy [35]. The PEGylation of nanoparticles significantly changes a metabolic drug process in blood by protecting it from metabolism [36,37,38], conferring stealth properties on the nanoparticles [39], reducing the tendency of elimination by the reticuloendothelial system (RES), improving the bioavailability and half-life of drug used in the formulation [40]. PEG is hydrophilic; thus, its hydrophilic properties could prevent interaction with other nanoparticles via sterical hindrance, thereby improving nanoparticles’ stability [41]. PEGylated products have shown numerous advantages: a prolonged residence in the body, decreased degradation by metabolic enzymes and reduced or eliminated protein immunogenicity [34]. The novelty embodied in this study is the use of sunseed-oil-based PEGylated nanoengineered lipid carrier to enhance the antidiabetic activity of metformin via improved half-life, reduced dosage and dosing frequency as well as through enhanced permeability across the biological membrane. Although we had earlier utilized sunseed oil in the development of PEGylated formulations for enhanced delivery of antifungal agents[35], to the best of our knowledge, there is currently a dearth of information in the literature on the use of sunseed-oil-based PEGylated nanoengineered lipid carrier for oral delivery of metformin for enhanced treatment of diabetes mellitus. Therefore, the investigation of sunseed-oil-based PEGylated nanolipid carrier for enhanced metformin delivery to optimize the efficacy, intestinal absorption, bioavailability, biological half-life, and reduce drug dose, dosing frequency and drug adverse effects, informs the aim of this study. Thus, the objective of this study was to formulate, characterize and evaluate in vitroand in vivo pharmacodynamic properties of metformin-loaded sunseed-oil-based PEGylated nanostructured lipid carrier (PEG-NLC) for improved delivery of metformin hydrochloride.

Materials and Methods

Materials

Metformin hydrochloride pure sample was obtained as a gift from May and Baker PLC (Ikeja, Lagos State, Nigeria). Phospholipon® 90H (P90H) (Phospholipid GmbH, Köln, Germany), sorbitol (Caesar & Loretz, Hilden, Germany), polyethylene glycol 4000 (PEG 4000) (Ph. Eur. Carl Roth GmbH + Co. KG Karlsruhe, Germany), beeswax (Carl Roth, Karlsruhe, Germany), Polysorbate 80 (Tween® 80) (Acros Organics, Geel, Belgium), sunseed oil (double refined), vitamin A fortified (Kewalram Chanrai Group, Lagos, Nigeria), Alloxan (Merck KGaA, Darmstadt, Germany), distilled water (Lion water, University of Nigeria, Nsukka, Nigeria), and other solvents and reagents were used as procured from their manufacturers without further purification. Adult albino Wistar rats of both sexes were procured from the Faculty of Veterinary Medicine, University of Nigeria, Nsukka.

Preparation of PEGylated Lipid Matrices

PEGylated lipid matrices were prepared by the fusion method [42, 43] using beeswax (BW) and Phospholipon® 90H (P90H) (as solid lipids) in combination with super-refined sunseed oil (as liquid lipid) followed by PEGylation. The solid lipids and liquid lipid were used at a 7:3 ratio (i.e., 21.0 g of BW/P90H admixture and 9.0 g of super-refined sunseed oil). First, 21.0 g of beeswax and 9.0 g of P90H were weighed using an electronic balance (Mettler H8, Switzerland), poured into a crucible, and melted together at 70 °C on a thermoregulated oil bath (liquid paraffin) and stirred thoroughly to obtain an adequate mixing. Next, the homogenous mixture of the lipid matrix (LM1) was stirred further at room temperature and then allowed to cool and solidify. After 24 h, this lipid matrix was melted in the thermoregulated bath at a temperature of 80 °C, followed by the addition of 9.0 g (9.017 ml) of super-refined sunseed oil. The mixture was stirred continuously until a homogenous, transparent white melt was obtained. Next, the homogenous mixture of the lipid matrix (LM2) was stirred at room temperature until solidification. More so, after 24 h, various quantities (90, 80, and 60%w/w) of the prepared lipid matrix (LM2) were melted together with corresponding amounts of polyethylene glycol (PEG 4000) (10, 20, and 40%w/w) incorporated at 80 ºC over the oil bath to give PEGylated lipid matrices containing 1:9, 2:8, 4:6 ratios of PEG: lipid matrix, respectively, which were appropriately stirred and allowed to solidify.

Preparation of Drug-Loaded PEGylated Lipid Matrices

Representative drug-loaded PEGylated lipid matrices were prepared by fusion using the PEGylated lipid matrices and metformin. With target PEGylated lipid concentration of 5.0%w/w and target drug concentrations of 1.0%w/w of metformin in the PEGylated nanostructured lipid carriers to be developed, 2.5 g of each of the lipid matrices was melted in the thermoregulated oil bath at a temperature of 80 ºC followed by the addition of 0.5 g of metformin. Each mixture was stirred continuously until a homogenous, transparent white melt was obtained. Then, the drug-loaded lipid matrices were allowed to cool and solidify at room temperature.

Differential Scanning Calorimetry (DSC) Analysis of Plain and Drug-Loaded PEGylated Lipid Matrices

About 5 mg of each sample was weighed into an aluminum pan and hermetically sealed, and the thermal behavior was determined in the range of 20–350 °C at a heating rate of 5 °C/min. The temperature was held at 80 °C for 10 min and, after that, cooled at the rate of 5 to 10 °C/min. Baselines were determined using an empty pan, and all the thermograms were baseline-corrected.

Fourier Transform Infrared (FT-IR) Spectroscopic Analysis of Drug-Loaded PEGylated Lipid Matrices

FTIR spectroscopic analysis was conducted on metformin and representative metformin-loaded sunseed-oil-containing non-PEGylated and PEGylated beeswax-based P90H-modified lipid matrices (MT-loaded LM2 and MT-loaded PEG-LM2 or drug-loaded LM2 and drug-loaded PEG-LM2) using a Shimadzu FT-IR 8300 Spectrophotometer (Shimadzu, Tokyo, Japan). The spectrum was recorded in the wavelength region of 4000 to 400 cm−1 with a threshold of 1.303, sensitivity of 50, and resolution of 2 cm-1. A smart attenuated total reflection (SATR) accessory was used for data collection. The potassium bromate (KBr) plate used for the study was cleaned with a tri-solvent (acetone–toluene–methanol at 3:1:1 ratio) mixture for baseline scanning. A 0.1 g of each sample was mixed with 0.1 ml nujul diluent. The solution was introduced into the potassium bromate (KBr) plate and compressed into discs by applying a pressure of 5 tons for 5 min in a hydraulic press. The pellet was placed in the light path, and the spectrum was obtained. Spectra were collected in 60 s using Gram A1 spectroscopy software, and the chemometrics was performed using TQ Analyzer1.

Preparation of PEGylated Nanostructured Lipid Carriers

PEGylated nanostructured lipid carriers encapsulating metformin (D0, D10, D20, D40) were prepared using the drug, PEGylated lipid matrices, Polysorbate® 80 (Tween® 80) (mobile surfactant), sorbitol (cryoprotectant), and distilled water (vehicle) by the high shear hot homogenization method [44, 45]. In each case, the PEGylated lipid matrix was melted at 80 °C in the thermoregulated heater (IKA instrument), and metformin was introduced into the melted lipid and stirred thoroughly. Next, the aqueous phase containing sorbitol (4%w/w) and polysorbate® 80 at the same temperature was added to the molten lipid matrix with gentle stirring using a magnetic stirrer (SR 1 UM 52,188, Remi Equip., India), and the mixture was further dispersed using an Ultra-Turrax T25 (IKA-Werke, Staufen, Germany) homogenizer at 1000 rpm for 5 min. Finally, the obtained pre-emulsion was homogenized at 15,000 rpm for 30 min and allowed to cool and re-crystallize at room temperature. The formulation compositions of the PEGylated NLCs are shown in Tables 1 and 2.

Characterization of the Non-PEGylated and PEGylated NLCs

Determination of Particle Sizes, Polydispersity Indices and Surface Charges

Mean diameter, Z. Ave (nm), and polydispersity indices (PDI) of the formulations were measured using a zeta sizer nano-ZS (Malvern Instrument, Worcestershire, UK) equipped with a 10-mw He–NE laser employing the wavelength of 633 nm and a backscattering angle of 173° at 25 °C. To obtain a suitable scattering intensity, the sample was diluted with double-distilled water before photon correlation spectroscopic (PCS) analysis.

Furthermore, the stability of the formulations was ascertained using zeta potential measurement. Briefly, zeta potentials or surface charges (ζ) of metformin-loaded PEGylated and metformin-loaded non-PEGylated NLC formulations were estimated using dynamic light scattering (DLS) (Malvern Instruments, Japan). Each sample was diluted with deionized water to avoid multiple scattering and to maintain the number of counts per second in the region of 600, and measured at angle of 90° and temperature of 25 °C. Replicate measurements were taken and averages taken.

Compatibility Study by Fourier Transform Infrared (FTIR) Spectroscopy

Fourier transform infrared (FT-IR) spectroscopic analysis was conducted using a Shimadzu FT-IR 8300 Spectrophotometer (Shimadzu, Tokyo, Japan), and the spectrum was recorded in the wavelength region of 4000 to 400 cm−1 with a threshold of 1.303, the sensitivity of 50, and resolution of 2 cm−1 range. A smart attenuated total reflection (SATR) accessory was used for data collection. The potassium bromate (KBr) plate used for the study was cleaned with a tri-solvent (acetone–toluene–methanol at 3:1:1 ratio) mixture for baseline scanning. A 0.1 ml volume of each formulation was mixed with 0.1 ml nujul diluent. The solution was introduced into the potassium bromate (KBr) plate and compressed into discs by applying a pressure of 5 tons for 5 min in a hydraulic press. Then, the pellet was placed in the light path, and the spectrum was obtained. Spectra were collected in 60 s using Gram A1 spectroscopy software, and the chemometrics was performed using TQ Analyzer1.

Determination of Encapsulation Efficiency

The encapsulation efficiency of each formulation was determined. A 5 ml volume of each formulation was placed in a centrifuge tube and centrifuged for 30 min at an optimized speed of 4000 rpm to obtain two phases (the aqueous and lipid phases). A 1 ml volume of the aqueous phase was measured with the aid of a syringe and then diluted 10, 000-fold using distilled water. The absorbance of the dilutions was taken using a UV spectrophotometer (Jenway 6405, UK) at a wavelength of 231.5 nm, and the EE % was calculated using the formula below:

$$EE\%= \frac{Actual\;drug \;content}{Theoritical \;drug \;content}\times 100\%$$
(1)

Loading Capacity

Loading capacity of the drug in lipid carriers depends on the type of lipid matrix, solution of drug in melted lipid, miscibility of drug melt and lipid melt, chemical and physical structure of solid lipid matrix, and the polymorphic state of the lipid material [42]. Therefore, loading capacity (LC) is expressed as the ratio between the entrapped drug by the lipid and the total quantity of the lipids used in the formulation. It was calculated using the formula below:

$$\begin{aligned}Loading \;capacity&=\frac{Total\; quantity \;of \;drug \;entrapped \;by \;the \;lipid }{Total \;quantity \;of \;lipid \;in \;the \;formulation}\\&\times 100\end{aligned}$$
(2)

In Vitro Drug Release Studies

The USP XXII rotating paddle apparatus (Erweka, GmbH Germany) was employed for this release study. The dissolution medium consisted of 500 mL of freshly prepared phosphate buffer maintained at 37 ± 1 ℃ utilizing a thermoregulated water bath. The polycarbonate dialysis membrane used as a release barrier was pre-treated by soaking it in the dissolution medium for 24 h before the commencement of each release experiment. In each case, 2 ml of the drug-loaded PEGylated NLC was placed in the dialysis membrane, securely tied with a thermoresistant thread, and then immersed in the dissolution medium under agitation provided by the paddle at 200 rpm. At predetermined time intervals (30, 60, 120, 180, 240, 300, 360, 420, 480, and 600 min), 5 mL portions of the dissolution medium were withdrawn and replaced with an equal volume of the medium to maintain a sink condition, filtered with a pore size of 0.22 mm (Millipore filter, Delhi, India) and analyzed using a spectrophotometer (Jenway, UK) at 231.5 nm.

In Vivo Antidiabetic Studies

Experimental Animals

Wistar strain albino rats of both sexes weighing between 150 and 200 g were bred in our institution. The animals were housed in standard environmental conditions, kept at a body temperature of 37 °C using warming lamps, and left for one week to acclimatize with the new laboratory environment while being fed a standard laboratory low chow diet. All the animals were fasted for 12 h but were allowed free access to water before the commencement of the experiments.

Induction of Diabetes

Adult Wistar rats (either sex) were used to evaluate the antidiabetic effects of the formulations. The rats were divided into groups of six animals, and each group of animals was housed in a separate cage. Diabetes was induced by a single intraperitoneal injection of a freshly prepared solution of alloxan (150 mg/ kg) in normal saline for all the groups. After one h of alloxan administration, the animals were fed freely, and 5% dextrose solution was also given orally in a feeding bottle for a day to overcome the early hypoglycemic phase. The animals were kept under observation, and after 48 h, blood was withdrawn from the tail vein of the animals, and the blood glucose level was measured with a glucometer (Accu-check, Roche, USA). The diabetic rats (glucose level above 200 mg/dl) were separated into ten groups of six animals (n = 6).

Evaluation of Antidiabetic Activity

The different formulations of the metformin-loaded PEGylated NLC were administered orally to the animals according to their weight as follows:

  • Group A: received the test formulation (batch D0) equivalent to 100 mg/kg of metformin hydrochloride.

  • Group B: received the test formulation (batch D10) equivalent to 100 mg/kg of metformin hydrochloride.

  • Group C: received the test formulation (batch D20) equivalent to 100 mg/kg dose of metformin hydrochloride.

  • Group D: received the test formulation (batch D40) equivalent to 100 mg/kg dose of metformin hydrochloride.

  • Group E: received glucophage® equivalent to 100 mg/kg dose of metformin hydrochloride.

  • Group F: received pure metformin equivalent to 100 mg/kg dose of metformin hydrochloride.

  • Group G: received normal saline orally

All the samples for treatment were administered orally. In addition, blood samples of the animals were collected from the tail vein afterward at time intervals of 0, 1, 3, 6, 12, and 24 h and tested for blood glucose level using the glucometer.

The post-dose levels of the blood glucose were expressed as a percentage of the pre-dose level. The percent basal blood glucose concentration was plotted against time for the various groups.

$$\%glycemic \;change=\frac{Initial\; conc -Final \;conc}{Initial \;conc }\times 100$$
(3)

In Vivo Bioavailability Study

In this study, eighteen (18) rats were used. They were made diabetics as described in the preceding section on antidiabetic study, after which the rats were randomly divided into three groups of six rats each. Group one received the MT-loaded non-PEGylated NLC formulation (batch G0) equivalent to 100 mg/kg dose of metformin hydrochloride, Group two received the MT-loaded PEGylated NLC formulation (batch G40) equivalent to 100 mg/kg dose of metformin hydrochloride, and Group three received Glucophage® (reference sample) equivalent to 100 mg/kg dose of metformin hydrochloride (positive control). The formulations were administered orally, and thereafter blood samples were collected from the tail vein of the rats at time intervals of 0, 1, 3, 6, 12, and 24 h using heparinized hematocrit tubes, centrifuged at 5,000 rpm for 5 min to separate the plasma and then stored at -4 ºC till analyzed. Approximately 0.2 ml of the plasma sample was diluted with equal volume of acetonitrile and centrifuged at 2,000 rpm for 5 min after which 0.1 ml of the supernatant was diluted in distilled water and assayed spectrophotometrically (Unico 2102 PC UV/Vis Spectrophotometer, New York, USA) for drug content. For the preparation of calibration curve, plasma from the blood withdrawn at zero hour was similarly diluted and used as blank. The plasma concentration–time curve was then plotted and further evaluated to obtain the pharmacokinetic parameters such as the maximum plasma concentration (Cmax) and the corresponding time (Tmax) using Phoenix® WinNonlin (version 6.3; Pharsight, St Louis, MO, USA), based on the average blood drug concentration. Area under the curve (AUC) from 0 to 24 h was calculated using the program’s linear trapezoidal rule.

Safety Studies

Here, three groups of animals were selected from the antidiabetic studies in the previous section, i.e., one animal from each group (optimized formulation group (G40), diabetic but untreated rat group and non-diabetic/normal rabbit group) was utilized for the purpose of histopathological studies in line with earlier reports [6, 8]. Small pieces of liver tissues in each group were collected in 10% neutral buffered formalin for proper fixation for 24 h. These tissues were processed and embedded in paraffin wax. Sections of 5–6 um in thickness were cut and stained with hematoxylin and eosin (H & E). These sections were examined photomicroscopically for necrosis, steatosis and fatty changes of hepatic cell [6, 8].

Statistical Analysis

All experiments were performed in replicate for the validity of the statistical analysis. Results were expressed as mean ± SEM. Analysis of variance (ANOVA) using LSD post hoc multiple comparison test was performed on the data sets generated using SPSS 16.0. Differences were considered significant for p-values < 0.05.

Results and Discussion

Thermal Properties of Plain and Drug-Loaded Lipid Matrices

Lipids used in nanoformulations have different crystallization temperatures, and these polymorphic transitions affect the nanoformulations’ thermal behaviors. The energies required to melt substances vary depending on whether crystalline or amorphous. Amorphous substances lack a long-range order or are less ordered; thus, lesser energy is needed to melt the substance; however, crystalline substances have a perfect crystalline structure and require higher energy to overcome lattice forces. Therefore, higher melting enthalpy values suggest a higher-ordered lattice arrangement [45]. The slight presence or total absence of a melting peak in the DSC of a solid dispersion indicates that the drug is partly or wholly amorphous/molecularly dispersed [46, 47]. The thermogram of beeswax, Phospholipon 90H, structured lipid matrix (LM1), structured lipid matrix (LM2), PEG 4000, PEGylated lipid matrix (LM2), metformin, metformin-loaded non-PEGylated (drug-loaded LM2), PEGylated lipid matrices (drug-loaded PEG-LM2), metformin-loaded non-PEGylated (drug-loaded LM2), and PEGylated lipid matrices (drug-loaded PEG-LM2) in superposition are shown in Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10.

Fig. 1
figure 1

Differential scanning calorimetry (DSC) thermogram of beeswax (BW)

Full size image
Fig. 2
figure 2

Differential scanning calorimetry (DSC) thermogram of Phospholipon® 90H (P90H)

Full size image
Fig. 3
figure 3

Differential scanning calorimetry (DSC) thermogram of structured lipid matrix (BW: P90H) (7:3) (LM1)

Full size image
Fig. 4
figure 4

Differential scanning calorimetry (DSC) thermogram of structured phospholipid-modified beeswax-based lipid matrix structured with sunseed oil (LM2)

Full size image
Fig. 5
figure 5

Differential scanning calorimetry (DSC) thermogram of PEG 4000

Full size image
Fig. 6
figure 6

Differential scanning calorimetry (DSC) thermogram of PEGylated lipid matrix (PEG-LM2)

Full size image
Fig. 7
figure 7

Differential scanning calorimetry (DSC) thermogram of metformin

Full size image
Fig. 8
figure 8

Differential scanning calorimetry (DSC) thermogram of metformin-loaded lipid matrix (drug-loaded LM2)

Full size image
Fig. 9
figure 9

Differential scanning calorimetry (DSC) thermogram of metformin-loaded PEGylated lipid matrix (drug-loaded PEG-LM2)

Full size image
Fig. 10
figure 10

DSC thermographs of metformin, metformin-loaded non-PEGylated and PEGylated lipid matrices (drug-loaded LM2) and (drug-loaded PEG-LM2) in superposition

Full size image

The DSC thermograms of beeswax, PEG 4000, metformin, Phospholipon® 90H, and lipid matrix (LM1) showed single sharp endothermic melting peaks at 73, 74.5, 265.5, and 122.1 ℃ with enthalpies of -55.3, -3.8, -28.4, and -38.9 mW/mg, respectively, indicating the highly crystalline nature of these materials. Lipid matrix (LM1) and PEG lipid matrix (LM2) showed endothermic melting peaks at 88.6 and 93.1 ℃ with enthalpies of -32.5 and -3.9 mW/mg, respectively, and the reduced enthalpy indicates that the lipid matrices are less crystalline than the individual components used in the formulation. Lipid matrix (LM2), on the other hand, showed two peaks, an endothermic and an exothermic peak at 115.8 and 217.4 ℃ with enthalpy of -7.5 and 19.5 mW/mg; this indicates that lipid matrix (LM2) generates an imperfect matrix, which arises from a distortion of the crystal arrangement of the individual lipids after melting and solidification, thus creating more spaces for drug encapsulation [48, 49]. The exothermic peak at 217.4 ℃ could be due to the molecules arranging themselves to a lower energy configuration or chemical cure.

The DSC thermograms of metformin-loaded lipid matrix and metformin-loaded PEGylated lipid matrix (drug-loaded PEG-LM2) showed two melting peaks each. Although the metformin-loaded lipid matrix showed an endothermic peak at 94.5 and 258.0 ℃ with an enthalpy of -0.9 and -5.7 mW/mg, the melting peak was 258.0 ℃ is a weak peak, showing that metformin is either partly amorphous or molecularly dispersed in the lipid matrix. The metformin-loaded PEGylated lipid matrix (drug-loaded PEG-LM2) also showed two endothermic peaks at 89.0 and 258.0 ℃ with enthalpy of -0.7 and -0.4 mW/mg, indicating molecular dispersion of metformin in the lipid matrix. Furthermore, the DSC results illustrate that the drug-loaded lipid matrices had lower melting points than the pure metformin and did not show any transition of the drug in the thermogram, indicating that the drug exists in a solubilized form [50].

Fourier Transform Infrared Spectroscopy of Drug and Drug-Loaded Lipid Matrices

Solid-state characterizations are usually conducted to describe the properties of nanoformulations. The infrared (IR) spectrum of any given compound is always unique and characteristic of that compound [35]. FTIR spectroscopic analysis is usually carried out to evaluate possible molecular interactions between the drug and the matrices in the solid state [51]. The FTIR spectroscopic spectrum of metformin, metformin-loaded non-PEGylated (drug-loaded LM2), and PEGylated lipid matrices (drug-loaded PEG-LM2) are shown in Fig. 11. The pure metformin sample shows principal characteristic absorption bands at 3456.55 cm−1 (N–H stretching), 2931.90 cm−1 ((CH3)2-N absorption), 1643.41 cm−1 (N–H deformation), 1396.51 cm−1 (N–H deformation), 1219.05 cm−1 (C-O vibration), 1103.32 cm−1 (C-N stretching), 1130.74 cm−1 (C-N stretching), 771.56 cm−1 (N–H wagging), 601.81 cm−1 (C-H out of plane bending), and 547.8 cm−1 (C-N–C deformation). The spectrum of the metformin-loaded LM2 shows principal absorption bands at 3819.18 cm−1 (O–H stretching), 3471.98 cm−1 (N–H stretching), 2916.47 cm−1 ((CH3)2-N absorption), 2854.74 cm−1 ((CH3)2-N absorption), 2314.66 cm−1 (-C = C- stretching), 1735.99 cm−1 (-C = O vibration), 1643.41 cm−1 (N–H deformation), 1458.23 cm−1 (symmetric N–H deformation), 1396.51 cm−1 (N–H deformation), 1165.04 cm−1 (C-N stretching), 1095.60 cm−1 (C-N stretching), 1049.31 cm−1 (C-N stretching), 717.54 cm−1 (N–H wagging), 601.81 cm−1 (C-H out-of-plane bending) and 547.80 cm−1(C-N–C deformation). The spectrum of the PEGylated metformin-loaded lipid matrix (drug-loaded PEG-LM2) shows principal absorption bands at 3433.41 cm−1 (N–H stretching), 2916.47 cm−1 ((CH3)2-N absorption), 2854.74 cm−1 ((CH3)2-N absorption), 2314.66 cm−1 (-C = C- stretching), 1975.17 cm−1 (-C = O vibration), 1735.99 cm−1 (conjugated C = C bond vibration), 1651.12 cm−1 (N–H deformation), 1458.23 cm−1 (symmetric N–H deformation), 1226.77 cm−1 (C-O vibration), 1157.33 cm−1 (C-N stretching), 964.44 cm−1 (N–H out-of-plane bending), 840.99 cm−1 (NH2 rocking), 717.54 cm−1 (C-H out-of-plane bending), and 586.38 cm−1 (C-N–C deformation). From the results, it can be inferred that there was no interference between the drug, lipids, and polymers used in the formulation; this was seen from the fact that there was no significant change in the absorption peaks of metformin functional groups. The other absorption peaks present in the spectra were attributed to excipients; hence, the polymers and lipids used in the formulation are compatible with the drug.

Fig. 11
figure 11

FT-IR spectra of metformin, metformin-loaded non-PEGylated and PEGylated lipid matrices (drug-loaded LM2) and (drug-loaded PEG-LM2) in superposition

Full size image

Mean Particle Size, Polydispersity Indices, and Surface Charges of the NLC Formulations

The combined particle size analysis and polydispersity index of the formulations are shown in Fig. 12. The formulation with the most miniature particle size was D20 (290.6 nm), and formulation D10 showed maximum particle size (880.60 nm). The PDI value of batch D0 and D10 was 0.477 and 0.625, showing the minimum and maximum PDI in the formulation, respectively. The particle size distribution was normal, and the polydispersity index confirmed this. These results showed a unimodal size distribution of the different particle sizes, indicating stable formulations [5]. Polydisperse systems tend to aggregate than monodisperse systems; thus, D40 is the most stable formulation (Table 3).

Fig. 12
figure 12

Combined particle size analysis result of the metformin-loaded sunseed-oil-based non-PEGylated and PEGylated nanostructured lipid carriers

Full size image

The results of the zeta potential measurement surface charges as shown in Table 4 indicate good stability of the formulations since the zeta potential values were > 25 mV. The positive surface charges on the MT-loaded PEGylated and MT-loaded non-PEGylated NLC formulations imply that they would aid in easily transporting the formulations across cell membranes. In addition, results indicate that they can bind with negative charges on the mucosa walls of the GIT and ultimately prolong drug release for enhanced absorption, which is in agreement with the previous study [5].

Table 1 Optimized formula for the preparation of the non-PEGylated and PEGylated NLCs
Full size table
Table 2 Formulation composition of lipid matrices for PEGylated NLCs preparation
Full size table
Table 3 Particle size and polydispersity indices of metformin-loaded sunseed-oil-based non-PEGylated and PEGylated nanostructured lipid carriers
Full size table
Table 4 Zeta potentials or surface charges of the formulations
Full size table

Fourier Transform Infrared Spectroscopy of Drug-Loaded NLC Formulations

Figure 13 shows FTIR spectra of metformin, metformin-loaded non-PEGylated and PEGylated sunseed-oil-based nanostructured lipid carriers (D0, D10, D20, and D40) in superposition. The FTIR spectrum of the pure metformin sample shows characteristic principal absorption bands at 3456.55 cm−1 (N–H stretching), 2931.90 cm−1 ((CH3)2-N absorption), 1643.41 cm−1 (N–H deformation), 1396.51 cm−1 (N–H deformation), 1219.05 cm−1 (C-O vibration), 1103.32 cm−1 (C-N stretching), 1130.74 cm−1 (C-N stretching), 1130.74 cm−1 (C-N stretching), 771.56 cm−1 (N–H wagging), 601.81 cm−1 (C-H out-of-plane bending), and 547.8 cm−1 (C-N–C deformation). The FTIR spectrum of D0 shows principal absorption peaks at 3363.97 cm−1 (N–H stretching), 2090.91 cm−1 (carboxylic acid C = O vibration), 1643.41 cm−1 (N–H deformation), 1103.32 cm−1 (C-N stretching), and 578.66 cm−1 (C-N–C deformation). The FTIR spectrum of D10 shows principal absorption peaks at 3363.97 cm−1 (N–H stretching), 2106.34 cm−1 (carboxylic acid C = O vibration), 1635.69 cm−1 (N–H deformation), 1226.77 cm−1 (C-N stretching), 1080.17 cm−1 (C-N stretching), and 547.80 cm−1 (C-N–C deformation). The FTIR spectrum of D20 shows principal absorption peaks at 3363.97 cm−1 ((CH3)2-N absorption), 2924.18 cm−1, 2854.74 cm−1 (N–H stretching), 2121.77 cm−1 (carboxylic acid C = O vibration), 1643.41 cm−1 (N–H deformation), 1512.24 cm−1 (asymmetric N–H deformation), 1087.89 cm−1 (C-N stretching), and 532.37 cm−1 (C-N–C deformation). The FTIR spectrum of D20 shows principal absorption peaks at 3873.19 cm−1 (N–H stretching), 3302.24 cm−1 (asymmetric N–H stretching), 2924.18 cm−1, 2507.54 cm−1 ((CH3)2-N absorption), 2360.95 cm−1 (-C = C- stretching), 2083.19 cm−1 (carboxylic acid C = O vibration), 1620.26 cm−1 (N–H deformation), 1535.39 cm−1 (asymmetric N–H deformation), 1411.94 cm−1 (N–H deformation), 1033.88 cm−1 (C-N stretching), 794.7 cm−1 (N–H wagging) and 493.79 cm−1 (C-N–C deformation). From the results, it can be inferred that there was no interference in the drug, lipids, and polymers used in the formulation [52]; this was seen from the fact that there was no significant change in the absorption peaks of metformin functional groups, and the reduced peak intensities were due to the presence of hydrogen bonding [53]. The other absorption peaks present in the spectra were attributed to excipients; hence the polymers and lipids used in the formulation could be compatible with the drug.

Fig. 13
figure 13

FT-IR spectra of Metformin, Metformin-loaded non-PEGylated, and PEGylated sunseed-oil-based nanostructured lipid carriers (D0, D10, D20, and D40) in superposition

Full size image

Scanning Electron Microscopy (SEM) of Drug-Loaded Non-PEGylated and PEGylated Nanostructured Lipid Carriers

Figure 14 shows the scanning electron micrograph of the NLC formulation for D40 (optimized formulation). The image showed the presence of well-defined nanodispersions with particles that were segregated, uniform in size, and semispherical in shape, with similar sizes spread over the formulation, which is the typical morphological aspect of nanoparticles.

Fig. 14
figure 14

Scanning electron micrograph (SEM) of metformin-loaded sunseed-oil-based PEGylated nanostructured lipid carrier (batch D40)

Full size image

Encapsulation Efficiency and Loading Capacity of PEGylated and Non-PEGylated NLCs

The drug loading efficiency could be affected by the drug’s molecular weight, the volumetric size of the carrier, chemical interactions between the drug and the carrier, solubility of the drug in the carrier, miscibility in the lipid matrix, and the lipid phase polymorphic state [14, 17, 54]. Metformin has a high tendency to escape from the lipid matrix due to its high solubility in water [55] and may result in lower drug loading. When added in high concentration, PEG lipids may induce the formation of mixed micelles (PEGylated carriers), which may increase drug encapsulation efficiency [35]. This may account for the high encapsulation efficiency of the NLCs. Previous studies have shown a reverse relationship between particle size and loading parameters (loading efficiency and entrapment efficiency). Formulations with smaller particle sizes had higher encapsulation efficiency and drug loading capacity, and this is consistent with previous reports.

The encapsulation efficiency (EE%) and loading capacity (LC) of the NLCs are shown in Fig. 15. The encapsulation efficiencies were in the range of 98.78–99.83% for the NLCs, with the formulation D0 showing the highest EE%, while D20 had the least. However, all batches of the NLCs had good EE% (98.78–99.83%). The loading capacity of all the batches of NLCs showed a similar pattern. The loading capacity (LC) of the NLC batches was in the range of 16.50–16.64 g of metformin per 100 g of lipid. Again, the NLC batch D0 had the highest loading capacity, while D20 had the least.

Fig. 15
figure 15

Bar chart representation of encapsulation efficiency and drug loading capacity

Full size image

In Vitro Drug Release From the Metformin-Loaded PEGylated and Non-PEGylated NLCs

The in vitro drug release study is a critical test to assess nanoparticle-based drug delivery systems’ safety, efficacy, and quality [56]. The in vitro release kinetics of nanoparticles provides vital information regarding the ability to modify the drug release; when performed correctly, they can be correlated to in vivo behaviors through predictive mathematical models [57, 58]. Many mathematical models have been proposed to predict release-kinetic models, such as the zero-order, first-order, Higuchi, Korsmeyer–Peppas, Weibull, Gompertz, Peppas–Sahlin, Bakers–Lonsdale, Hopfenberg, Hixson–Crowell [59]. Although there are numerous mathematical models, these models could not cover all possible release mechanisms [60, 61]. To identify a particular release mechanism, experimental data of statistical significance are compared to a solution of the theoretical model.

Predicting the release of the active ingredient from a formulation is a vital aspect of drug delivery. The release pattern is usually projected as a function of time using both sophisticated and straightforward mathematical models [62]. Kinetics is an essential factor in drug release studies because it provides information on drug concentration in plasma. Applying the kinetic model is critical in clarifying the release mechanism, which helps design the drug release control [4]. Five different mathematical equations were used to predict the release kinetics: zero-order equation, first-order equation, Hixson–Crowell, Higuchi, and Korsmeyer–Peppas models to describe the kinetics of metformin release from the PEG-NLC formulations. The most appropriate drug release model was selected based on the goodness-of-fit test. Although the formulations showed a strong correlation for the first order, Higuchi, Hixson–Crowell, zero-order, and Korsmeyer–Peppas had a good correlation coefficient (r2 ≥ 0.6899). The results showed that all the prepared formulations fit well with a strong correlation (r2 ≥ 0.9954), following the first-order kinetics. Therefore, the formulations predominantly followed the diffusion mechanism of drug release. The equation of first-order kinetics for batch D40 (representative sample) is:

$$y=-0.1043x+4.6026 \quad r=0.9954$$

Table 5 shows the values of the calculated squares of the correlation coefficient of the formulations for each model. The Korsmeyer–Peppas model showed the weakest correlation. According to this model, the determined release rate constants were between 0.7147 and 0.8336, and the coefficients ‘n’ of the equation did not exceed 1.0392 with the corresponding r2 values 0.6899–0.7337.

Table 5 In vitro kinetics of release of metformin from non-PEGylated and PEGylated NLC in phosphate buffer
Full size table

Figure 16 shows the release profiles of metformin from all the batches of PEG-NLC formulation in phosphate-buffered solution (PBS, pH 7.4). The in vitro release profiles of metformin in PBS indicate significantly (p < 0.05) greater release of metformin from optimized formulation (batch D40) when compared to the drug release from the marked formulation (Glucophage®). The different sets of the formulations showed a high release rate in the trend D40 = D0 > D10 > D20 > Glucophage®. The release rate of the drug was higher in the formulations than the release in Glucophage. Metformin was released in a sustained manner throughout the period of study. A similar drug release pattern was reported earlier [9]. This is expected as solid lipid nanoparticles slow down the mobility of drugs out of the particles [63]. A study by Xu et al. also showed an initial rapid release of metformin from lipid-based nanoformulations; this was predicted to be due to adsorption of metformin on the surface of the nanoparticles [64]. Also, as the concentration of PEG 4000 polymer used increases, the rate of drug release increases. Thus, PEGylation enhanced the dissolution and release rate of the drug from the formulations.

Fig. 16
figure 16

In vitro drug dissolution profile of metformin, metformin-loaded non-PEGylated and PEGylated sunseed-oil-based nanostructured lipid carriers (D0, D10, D20, and D40)

Full size image

In Vivo Antidiabetic Activity

The results of the antidiabetic studies of the formulations, in comparison with the pure drug and commercially available product (glucophage®), are shown in Fig. 17 and Table 6. The effect of the PEGylated nanostructured lipid carrier (PEG-NLC) on the blood glucose level of non-diabetic rats was assessed and compared to both the pure drug sample and commercially available product (glucophage®). The in vivo pharmacologic antidiabetic activity of the formulations (D0, D10, D20, and D40), pure drug, and the commercially available product was evaluated using percentage reduction in blood glucose level (Table 6). This was the same evaluation method used by Mishra et al. where a percentage reduction in blood pressure was determined for Carvedilol-embedded nanostructured lipid carrier formulations [65]. The curves obtained by plotting percentage blood glucose reductions versus time are presented below. The groups treated with batch D0 and D10 formulations showed a higher and more sustained decrease in glucose level than those treated with the batch D20; there was a significant difference (p < 0.05) in the percentage reduction in blood glucose level between the groups that received formulations D0 and D10 with those that received D20. This could be due to the explanation that Shi et al. postulated in their study on the effects of polyethylene glycol on the surface of nanoparticles for targeted drug delivery. They said that encapsulation efficiency is an essential factor in the overall quality of the nanoparticles and that the PEG in nanoparticles often affect the encapsulation efficiency of the loaded molecules [66]. As the concentration of the PEG in the lipid fraction increases, the membrane's permeability also increases; this leads to leakage of the encapsulated drug, which negatively impacts the delivery of the drug [67,68,69]. The encapsulation efficiency of drugs that are soluble in water decreases as the PEG content of the lipid matrix increases. This negative relationship is also true for the encapsulation efficiency of water-soluble drugs and the molecular weight of the PEG in the lipid matrix, the probable reason being that the PEG will saturate the volume of the nanoparticles and the bigger the molecular weight of the PEG, the more the volume of the nanoparticles that are taken [70, 71]. Also, these batches produced a similar decrease in blood glucose level than the animal group treated with the market brand (glucophage®). This corroborated the results of Gadad et al. which showed that the rosuvastatin embedded nanostructured lipid carriers improved its bioavailability by two-fold compared to the marketed product [72].

Fig. 17
figure 17

Graphical representation of the in vivo antidiabetic effect (blood glucose-lowering effect) of the formulations (D0, D10, D20, and D40) vis-à-vis pure metformin, glucophage®, and distilled water

Full size image
Table 6 Blood glucose level of the treatment groups
Full size table
Table 7 Pharmacokinetic parameters of the NLC formulations after oral administration to rats (mean ± SD, n = 6)
Full size table

Furthermore, in the animal groups treated with the batches D0, D10, D20, and D40, the blood glucose reduction effectively commenced within an hour of oral administration with maximum blood glucose reduction (Tmax) achieved in 24 h. The blood glucose level produced by D0, D10, and D40 formulations reduced from 100% to 44.36%, 43.51%, and 44.41%, respectively, indicating that batch D40 was the best formulation, which was optimized for further in vivo bioavailability study. However, the blood glucose levels did not return to the initial level at zero hours within 24 h of the study.

In comparison with the pure drug and commercial sample, the formulations generally maintained the blood glucose level of the rats within the normoglycemic level for 12–24 h. Thus, it depicts that metformin could be effectively and efficiently delivered as sunseed-oil-based PEGylated NLCs. This also corresponded with the results of Gadad et al. which showed that the Rosuvastatin-embedded nanostructured lipid carrier was able to provide a sustained release of the drug over an extended period [68]. It can also be extrapolated that the blood-glucose-lowering effect was dependent on the polymers used and their concentrations; however, as the concentration of the PEG increases, this observed effect tends to decrease. The improved performance observed in the animal groups treated with the formulations corresponds with the improved dissolution observed.

In Vivo Oral Bioavailability

The pharmacokinetics results are shown in Fig. 18 and Table 7 as plasma concentration–time profiles of metformin and metformin-loaded non-PEGylated and PEGylated NLC and pharmacokinetic data, respectively. The mean AUC-24 values for optimized metformin-loaded non-PEGylated and PEGylated NLC indicate approximately threefold and sixfold increase in systemic bioavailability of metformin from metformin-loaded non-PEGylated and PEGylated NLC, respectively. Similarly, the mean Cmax values for non-PEGylated and PEGylated NLC formulations were significantly (p < 0.005) greater than that for the marketed sample.

Fig. 18
figure 18

Changes of metformin concentration in blood over 24-h study period, of rats orally administered with the optimized metformin-loaded PEGylated NLC (G40) and metformin-loaded non-PEGylated NLC (G0) in comparison with marketed formulation (Glucophage®) at equivalent dose

Full size image

Obviously, the developed formulations recorded greater increased at broader peaks of mean plasma concentrations than the marketed metformin formulation (Glucophage®, reference sample) after an oral administration, implying that plasma concentration of metformin obtained with the reference sample was not sustained compared with the developed formulations. Thus, metformin-loaded non-PEGylated and PEGylated NLC formulations maintained a steady slow decrease or gradual clearance of drug throughout the study compared with the rapid exponential decrease in the marketed sample (Glucophage®). The decrease in metformin concentration in the blood by the formulation could be attributed to an increase in the circulating half-life of metformin when administered as metformin-loaded PEGylated NLC. Thus, the PEGylated nature of batch G40 further delayed the clearance of metformin from the blood circulation. By implication, there was enhancement in the circulation longevity of metformin in the developed NLC formulations, but the effect was highest with metformin-loaded PEGylated NLC. This implies that lower doses of the drug in nanoscale (PEGylated NLC) would be employed at reduced dosing frequency to achieve the desired antidiabetic effect of metformin with reduced unwanted side effects, which is consistent with earlier reports [6, 8].

Safety of the Metformin-Loaded PEGylated NLC Formulations

The effects of the optimized formulation on the liver of the rats are shown in Fig. 19. The observation from the histological studies conducted on the liver of the rabbits from various groups showed significant hepatitis (arrow) on the liver of the rats in diabetic but untreated group (b). Photomicrograph of liver section of rats from experimental diabetic group that received metformin-loaded PEGylated NLC (optimized batch D40) (a) showed reduced periportal mononuclear infiltration of cells-periportal hepatitis (arrow), while the non-diabetic and untreated group (c) showed normal portal area and hepatocytes (H and E, mag. × 400). Overall, the results of histological study showed that short-term administration of developed metformin-loaded PEGylated NLC formulation did not cause any significant changes in the body regarding the relative liver morphology of the treated rats. By implication, short-term oral administration of sunseed-oil-based metformin-loaded PEGylated NLC will have no negative effects on somatic growth.

Fig. 19
figure 19

Photomicrographs of liver section of rats from experimental diabetic group that received metformin-loaded PEGylated NLC (optimized batch D40) (a), liver section of the rats in diabetic but untreated group (b) and liver of the rats in the non-diabetic and untreated group (c)

Full size image

Conclusions

This study evaluates the potential of metformin-loaded beeswax-based PEGylated nanostructured lipid carrier (PEG-NLC) for enhanced metformin delivery to treat diabetes mellitus. Metformin-loaded beeswax-based PEGylated nanostructured lipid carriers (PEG-NLCs) were successfully prepared by high shear hot homogenization method using structured lipids (beeswax and Phospholipon 90H), liquid lipids (sunseed oil), and PEG-4000 and evaluated for improved treatment of diabetes mellitus. Solid-state characterization performed on the lipid matrices used in preparing the formulations confirmed the suitability of the lipid matrices and their compatibility with metformin. There was sustained release of metformin from the PEGylated NLC formulation, which gave rise to enhanced antidiabetic activity and contributed to the circulation longevity of metformin in the blood, with formulation containing the highest amount of PEG-4000 (batch D40) (optimized formulation) recording sixfold increase in AUC than marketed sample (Glucophage®). The study suggests that nanostructured lipid carriers of metformin are suitable carriers for the oral delivery of metformin and other hydrophilic drugs. Using sunseed-oil-based PEGylated NLC as a delivery system for metformin achieves a steady-state drug concentration in the blood that is therapeutically effective, safe, and remains for an extended period. Sunseed-oil-based PEG-NLCs present a promising approach for improving metformin delivery and oral bioavailability.