Treatment of insulin resistance in obesity-associated type 2 diabetes mellitus through adiponectin gene therapy

Abstract

Global rise in obesity-associated type 2 diabetes mellitus (T2DM) has led to a major healthcare crisis. Development of efficient treatments to treat the underlying chronic inflammation in obesity-associated T2DM, is an unmet medical need. To this end, we have developed a plasmid adiponectin (pADN) based nanomedicine for the treatment of insulin resistance in type 2 diabetes mellitus. Adiponectin is a potent anti-inflammatory/anti-diabetic adipokine, which is downregulated in obesity. In this study, nanomicelles comprising chitosan conjugated to oleic acid and adipose homing peptide (AHP) were developed to deliver pADN to adipocytes. Cationic chitosan-oleic-AHP micelles were 112 nm in size, encapsulated 93% of pADN and protected gene cargo from DNase I mediated enzymatic degradation. In vitro, the nanomicellar formulation significantly increased adiponectin production compared to free plasmid as well as standard transfecting agent FuGENE®HD. Single dose subcutaneous administration of pADN-chitosan-oleic-AHP to obese-diabetic rats, resulted in improved insulin sensitivity for up to 6 weeks, which matched the glucose disposal ability of healthy rats. Serum adiponectin level in pADN-chitosan-oleic-AHP treated rats was comparable to healthy rats for up to 3 weeks post treatment. Overall, the results indicate that pADN-chitosan-oleic-AHP based therapy is a promising treatment approach for obesity-associated T2DM.

Graphical Abstract

1. Introduction

Type 2 diabetes mellitus (T2DM) is a major global public health crisis. The highest prevalence of the disease is in Middle Eastern and North African countries, and the highest number of adults afflicted with the disease is in the Western Pacific region (Kharroubi and Darwish 2015). In the United States, diabetes is the 7^th leading cause of deaths, affecting about 30.3 million people (9.4% of the population) (Centers for Disease Control 2017). T2DM is characterized by hyperglycemia due to peripheral insulin resistance, impaired hepatic glucose homeostasis and deteriorating pancreatic β-cell function (Mahler and Adler 1999). It has been reported that about 90% T2DM patients are either obese or overweight (American society for metabolic and bariatric research 2013). Obesity-associated T2DM is characterized by hyperplastic and/or hypertrophic growth of adipose tissues (Gustafson, et al. 2015; Schuster 2010). The primary role of adipose tissue hypertrophy/hyperplasia is adaptive, i.e., to store excess nutrients as lipid depositions for proper energy homeostasis (Corvera and Gealekman 2014). However, continual stress due to excess calories, leads to adipose tissue dysfunction and development of T2DM (Hajer, et al. 2008). Mounting evidence has suggested that T2DM is intricately linked with systemic low-grade inflammation that is initiated and propagated by the adipose tissue (Bluher 2016; Burhans, et al. 2018; Kohlgruber and Lynch 2015; Richardson, et al. 2013; Stafeev, et al. 2017).

Adiponectin (ADN) secreted by adipocytes plays a key role in glucose and lipid metabolism in insulin sensitive tissues. It improves insulin sensitivity and free fatty acid oxidation, while decreasing hepatic glucose output and vascular inflammation (Chandran, et al. 2003). In obese individuals, serum ADN level is significantly reduced, which is strongly associated with development of insulin resistance (Aleidi, et al. 2015; Kadowaki, et al. 2006). Expectedly, ADN administration leads to improvement in insulin sensitivity and decrease in plasma glucose levels (Achari and Jain 2017; Ruan and Dong 2016; Zoico, et al. 2009). In a recent clinical meta-analysis study, higher ADN levels were associated with decreased T2DM risk in Chinese and other populations (Wang, et al. 2018). Additionally, ADN is known to suppress production of pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF-α), while inducing production of anti-inflammatory cytokines such as interleukin 10 (IL-10) (Tilg and Wolf 2005). Conversely, high levels of TNF-α significantly downregulates ADN production and leads to development of underlying chronic inflammation in obesity-associated T2DM (Lihn, et al. 2005).

We developed a gene based targeted nanoparticle formulation to stimulate ADN production in the adipose tissue and improve insulin sensitivity in T2DM. Several cationic polymers including chitosan, poly(etheyleneimine), poly(L-lysine), poly(amdioamine) and green tea catechin, amongst others have been investigated for gene delivery (Barua, et al. 2011; Shen, et al. 2018; Sun and Zhang 2010). We utilized hydrophobically modifed chitosan nanoparticles that are excellent gene carriers due to their superior transfection ability (Layek and Singh 2012; Layek and Singh 2013a; Layek and Singh 2013b; Layek, et al. 2015; Sharma and Singh 2017b). Similar lipid-polymer self-assemblies or hydrophobically modified cationic polymers have been used as non-viral gene delivery vectors with relatively weak gene interactions that enable efficient gene transfection both in vitro and in vivo (Liu, et al. 2010; Wang, et al. 2015). In this study, free amino groups on chitosan-oleic acid polymer were covalently conjugated to adipose homing peptide (AHP), for improving internalization into adipose tissues. AHP targets prohibitin-1, a membrane protein expressed specifically in endothelial cells of white adipose tissue vasculature (Kolonin, et al. 2004; Salameh, et al. 2016; Thovhogi, et al. 2015). AHP mediated targeted delivery of nanoparticles to adipocytes has been successfully conducted in various preclinical studies (Kolonin, et al. 2004; Won, et al. 2014; Xue, et al. 2016). The focus of this paper was to develop a nanoparticle formulation capable of improving ADN concentration in the body for a prolonged period, thereby treating insulin resistance in obesity-associated T2DM.

2. Materials and Methods

2.1. Materials

Human recombinant insulin (Cell Prime™ r-insulin) was purchased from EMD Millipore Corporation (Burlington, MA, USA). Lipopolysaccharides from Escherichia coli (O111:B4), 3-Isobutyl-1-methylxanthine (IBMX), dexamethasone, Deoxyribonuclease I (DNase I) from bovine pancreas were procured from Sigma Aldrich (St. Louis, MO, USA). Plasmid DNA encoding beta-galactosidase (gWiz-βGal) was acquired from Aldevron LLC (Fargo, ND, USA). Beta-galactosidase enzyme assay kit was procured from Promega (Madison, WI, USA). Chitosan (Mw 20 kDa, 90% deacetylated) was purchased from Glentham Life Sciences (Corsham, WT, UK). Oleic acid was obtained from Spectrum Chemical (New Brunswick, NJ, USA). pADN (RN205822, CW304812) was acquired from Origene Technologies Inc. (Rockville, MD, USA). The nucleotide sequence of pADN is provided in supplementary file. Hoechst dye 33342 was purchased from AnaSpec (Fremont, CA, USA). AHP (peptide sequence: CKGGRAKDC) was obtained from Zhejiang Ontores Biotechnologies Co., Ltd (Hangzhou, Zhejiang, China). 3T3L1 pre-adipocyte cells, dulbecco’s Modified Eagle’s media (DMEM) and fetal bovine serum (FBS) were bought from ATCC (Manassas, VA, USA). Micro bicinchoninic assay (BCA) kit was purchased from ThermoFisher Scientific (Waltham, MA, USA). Rat adiponectin and TNF-α enzyme linked immunosorbent assay (ELISA) kits were purchased from BosterBio (Pleasanton, CA, USA). 4 weeks old male Wistar rats and hypercholesteremic diet (45% Kcal from fat) were obtained from Envigo RMS, Inc. (Indianapolis, IN, USA). Contour NEXT ONE glucometer and strips were used for blood glucose measurements. All other chemicals used were of analytical grade.

2.2. Preparation and characterization of chitosan-oleic-AHP particles

2.2.1. Preparation and evaluation of particle size, polydispersity index and charge of nanoparticles

Synthesis of chitosan-oleic-AHP particles was carried out as illustrated in Figure 1. Oleic acid or AHP were grafted onto chitosan via carbodiimide-based conjugation of amines present in chitosan with carboxyl group of oleic acid or peptide, as discussed in prior publication (Sharma and Singh 2017b). Briefly, chitosan (500 mg) was dissolved in deionized water (10 mL, pH 5, using glacial acetic acid). Oleic acid or AHP (0.06 mol/mol of monomer unit of chitosan) was dissolved in 5 mL of ethanol followed by the addition of 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (5 mol/mol of oleic acid) and N-hydroxysuccinimide (5 mol/mol of oleic acid). This mixture was added drop-wise to chitosan solution and allowed to react for 6 h at 25 °C. Purification was performed by dialysis using 3.5 kDa molecular weight cut off dialysis membrane (Thermo Scientific, IL, U.S.A.) against deionized water for 48 h. Dialyzed product was freeze dried and washed 3–4 times using ethanol to remove unreacted oleic acid to obtain final chitosan-grafted-oleic acid or chitosan-grafted-AHP polymer. Chitosan-oleic-AHP conjugation was performed similarly by using chitosan-oleic acid as the reactant. Degree of substitution was determined using trinitrobenzene sulfonic acid reagent for conjugation of oleic acid and BCA kit for conjugation of AHP, using manufacturer’s protocols. For preparation of micelles, the grafted or non-grafted chitosan polymers were dissolved in 20 mM glacial acetic acid solution (1 mg/mL) in deionized water (pH 6.5) and added drop-wise to pDNA containing solution (N/P ratio of 20). Thereafter, the mixture was vortexed for 5 mins and the solution was extruded using an Avanti Mini Extruder (Avanti Polar Lipids, Inc., AL, U.S.A.) through a 0.2 μ Whatman Nucleopore polycarbonate membrane (GE Healthcare, PA, U.S.A.) to obtain homogenous micelles.

Figure 1.

Schematic of preparation of chitosan-oleic-AHP micelles complexed with DNA

For determination of particle size, charge and polydispersity index (PDI), chitosan, chitosan-oleic, chitosan-AHP and chitosan-oleic-AHP particles were complexed with plasmid DNA encoding for β-galactosidase (pβgal), as a model pDNA, of comparable size as pADN. Hydrodynamic particle size distribution, PDI and charge of the nanoparticles were measured using Zetasizer Nano ZS 90 (Malvern Instruments, Malvern, UK) at room temperature.

2.2.2. DNA encapsulation efficiency

The association of pADN with chitosan, chitosan-oleic, chitosan-AHP and chitosan-oleic-AHP particles was evaluated by centrifuging polyplexes (N/P 20) at 30,000 g for 30 min at 4 °C, followed by quantification of the unbound pADN in the supernatant solution. Free pADN in the supernatant was stained with Hoechst dye 33342 (1 μg/mL in 1:1 v/v) and the fluorescence intensity was measured at 450 nm using a spectrofluorometer (Spectramax M5, Molecular Devices, San Jose, CA, USA), following excitation at 350 nm. Encapsulation efficiency was calculated as per the equation below.

$$\text{Encapsulation efficiency }\left(\%\right)=\left[\text{pADN }\left(\text{total}\right)-\text{pADN}\left(\text{in supernatant}\right)/\text{pADN }\left(\text{total}\right)\right]\times 100$$

2.2.3. DNase I protection assay

The protection afforded by various chitosan formulations to complexed pDNA against enzymatic degradation was evaluated using DNase I protection assay using pβgal as model plasmid DNA (Dos Santos Rodrigues, et al. 2019). For the study, 1 μg naked pDNA was complexed with chitosan, chitosan-oleic, chitosan-AHP or chitosan-oleic-AHP particles and incubated with 0.5 unit DNase I for 60 min at 37 °C. Naked pDNA with or without DNase I treatment was also simultaneously run as negative and positive controls, respectively. Reaction was halted by addition of 5 μL of EDTA (100 mM) and the pDNA was released from the complex upon 2 h incubation with 20 μL heparin (5 mg/mL) at 25 °C. Integrity of the released pDNA was thereafter examined using agarose gel (0.8% w/v) electrophoresis at 80 V for 90 min.

2.3. In vitro efficacy study

Pre-adipocytes, 3T3L1 cells were cultured in 24 well plates and allowed to differentiate into adipocytes, as per ATCC’s differentiation protocol. Briefly, cells were grown to 100% confluence, and incubated with DMEM containing 10% v/v FBS for 48 h at 37 °C and 5% CO₂ for two additional days, to obtain cells in growth arrest phase. Thereafter, the media was replaced with differentiation media containing DMEM with 10% FBS, 1 μM dexamethasone, 500 μM IBMX and 1 μg/mL insulin. After 48 h, the differentiation media was removed and adipocyte maintenance media containing DMEM with 10% FBS and 1 μg/mL insulin was added. The cells were cultured for 14 additional days and replenished with fresh maintenance media every 2 days. After a total of 18 days of culturing, the media was replaced with fresh DMEM only containing physiological diabetic concentration of glucose (8 mM) and mannose (50 μM). The cells were then treated with pH 7.4 phosphate buffered saline (PBS), pADN (1 μg/well) in PBS (free pADN), pADN in chitosan-oleic particles (1 μg pADN/well), empty chitosan-oleic-AHP particles, pADN in chitosan-oleic-AHP nanoparticles (1 μg pADN/well) or pADN in FuGENE®HD (1 μg pADN/well), and incubated at 37 °C and 5% CO₂. Following 4 h of incubation, the media was replaced with DMEM containing 10% FBS and cells were cultured for an additional 72 h. Thereafter, the supernatants were collected and cells were lyzed using 0.1% Triton X-100. ADN concentration in the supernatant were quantified using ELISA following manufacturer protocol and normalized to protein content determined using micro BCA assay.

2.4. In vivo efficacy study

All animal studies were conducted in conformity with the North Dakota State University animal care committee guidelines and to the Guide for the Care and Use of Animals of the Institute of Laboratory Animal Resources, National Research Council. Controlled temperature and 12 h light/dark cycle was provided to the rats. Rats were divided into high fat diet group (hypercholesteremic diet) and normal diet (age-matched healthy control) groups and given food and water ad libitum for 10 months. Weight gain in the rats was monitored every month. At the end of 10 months, insulin sensitivity was determined in rats that were fasted for 6 h, but given free access to water. For determination of insulin sensitivity, 1 U/kg insulin was injected intraperitoneally, and blood glucose was monitored every 0.5 h for a total of 2 h. On the subsequent day, blood was collected from the tail vein for serum adiponectin and TNF-α quantitation, using respective ELISA, as per manufacturer protocol. The group of rats that demonstrated significant decrease in insulin sensitivity compared to control (normal diet group), were subcutaneously injected with 1 mg/kg dose of either pADN in saline or pADN in chitosan-oleic-AHP nanoparticles. The group of rats that were fed with high fat diet but did not demonstrate loss in insulin sensitivity at the end of 10 months, were subcutaneously injected with empty chitosan-oleic-AHP nanoparticles. Healthy control group was subcutaneously injected with saline. Subsequent insulin sensitivity tests and blood collection through the tail vein were performed at 1, 3 and 6 weeks after injection and the rats were thereafter euthanized.

2.5. Histology

Rats on high fat diet were subcutaneously injected with saline or 1 mg/kg pADN in chitosan-oleic-AHP particles and euthanized after 24 h or 1 week following injection. Various organs and tissues including adipose tissue, liver, heart, lungs, kidney, spleen, and stomach were harvested for hematotoxylin and eosin (H&E) staining. Tissue sectioning and H&E staining were performed at the advanced imaging and microscopy core laboratory facility at North Dakota State University.

2.6. Data analysis

Data is represented as mean ± standard deviation (S.D.) for particle characterization studies and as mean ± standard error (S.E.) for in vitro and in vivo studies. Statistical analyses were conducted using two-tailed student’s t-tests and the difference was considered significant at p < 0.05. Graphs were plotted using Graphpad, prism 6.0 (GraphPad Software, La Jolla, CA).

3. Results

3.1. Characterization of chitosan nanoparticles

Chitosan formulations complexed with pDNA demonstrated size in the range of 112 – 194 nm (Table 1). Particularly, chitosan-oleic-AHP micelles exhibited the smallest size of 112 nm and good particle homogeneity with a PDI of 0.2. All formulations, with the exception of chitosan particles, exhibited near neutral charge ranging between 3 to 5.5 mV. A high positive charge of 21.5 mV was observed in chitosan formulation owing to the presence of free/unconjugated positively charged amino groups. Overall, chitosan-oleic-AHP nanomicelles demonstrated excellent characteristics of small size, low PDI and near neutral charge.

Table 1.

Characteristics of chitosan-based nanoparticles complexed with pDNA: particle size, PDI and charge. Data represented as mean ± S.D. (n = 4)

Formulations	Size (nm)	Polydispersity index	Zeta potential (mV)
Chitosan	193.73 ± 15.54	0.2 ± 0.04	21.46 ± 1.86
Chitosan-oleic	127.08 ± 10.24	0.24 ± 0.04	2.95 ± 0.93
Chitosan-AHP	164.65 ± 5.40	0.24 ± 0.07	5.53 ± 1.10
Chitosan-oleic-AHP	112.21 ± 10.75	0.2 ± 0.02	3.13 ± 0.85

3.2. pADN encapsulation efficiency

The encapsulation efficiency of pADN in chitosan-based nanoparticles is shown in Table 2. pADN associated well with chitosan-oleic-AHP particles, demonstrating ~93% encapsulation. Similar extent of pADN encapsulation ranging between 93 – 95% was observed with formulation controls including chitosan, chitosan-oleic and chitosan-AHP particles.

Table 2.

Encapsulation efficiency of pADN in chitosan based nanoparticles. Data represented as mean ± S.D. (n = 6)

Formulations	% Encapsulation efficiency
Chitosan	95.63 ± 2.25
Chitosan-oleic	93.18 ± 1.44
Chitosan-AHP	93.56 ± 0.79
Chitosan-oleic-AHP	92.62 ± 0.57

3.3. DNase I protection assay

DNase I protection assay was conducted to determine whether chitosan formulations are able to protect complexed plasmid from DNase I mediated enzymatic degradation. Naked pDNA incubated with DNase I completely degraded in 60 min, as noted by the absence of band in the gel electrophoresis pictograph (Figure 2). However, all chitosan formulations were able to protect complexed pDNA from DNase I mediated degradation, as evident from the presence of bright bands in the gel pictograph. Moreover, the distance travelled by pDNA in the gel were similar to naked pDNA not subjected to DNase I incubation, suggesting that pDNA in the chitosan formulations remained intact.

Figure 2. Protection against DNase I mediated degradation.

Ability of various formulations to protect complexed pDNA against DNase I based degradation. Lanes 1–6 represent naked pDNA, naked pDNA incubated with DNase I, pDNA complexed with chitosan particles incubated with DNase I, pDNA complexed with chitosan-oleic micelles incubated with DNase I, pDNA complexed with chitosan-AHP particles incubated with DNase I and pDNA complexed with chitosan-oleic-AHP micelles and incubated with DNase I, respectively.

3.4. In vitro efficacy study

Differentiated 3T3L1 adipocytes were treated with various formulations to determine their efficacy in upregulating pADN production. Basal ADN level was obtained in PBS treated cells (14,915 ± 544 pg/mg protein), which did not change significantly upon treatment with empty chitosan-oleic-AHP micelles (13,136 ± 637 pg/mg protein), (Figure 3). Treatment with free pADN, pADN-chitosan-oleic and pADN-FuGENE increased supernatant ADN concentrations compared to basal levels (18,900 ± 881, 18,367 ± 1158 and 19,533 ± 884 pg/mg protein, respectively). However, ADN concentration in the supernatant was most significantly enhanced in the pADN-chitosan-oleic-AHP treated cells (24,869 ± 1458 pg/mg protein) compared to all other formulations tested, including standard transfecting agent FuGENE®HD. This could be attributed to superior transfection capability of chitosan-oleic-AHP micelles compared to all other formulations tested. The results demonstrated the potential of pADN-chitosan-AHP micelles in significantly upregulating ADN levels in adipocyte culture in vitro, which was further validated in vivo using obese diabetic rats.

Figure 3. In vitro efficacy of pADN carrying formulations in stimulating ADN production in differentiated adipocytes.

Transfection efficiency of various formulations was determined in differentiated adipocytes. A significantly higher adiponectin concentration was obtained in supernatants of differentiated 3T3L1 adipocytes treated with pADN-chitosan-oleic-AHP formulation compared to all other formulations at 72 h (p < 0.01). Data represented as mean ± S.E. (n = 5 for PBS and n = 10 for free pADN, empty chitosan-oleic-AHP, pADN-chitosan-oleic, pADN-chitosan-oleic-AHP and pADN-FuGENE groups).

3.5. In vivo efficacy study

Rats on hypercholesteremic diet rapidly gained weight and showed significantly higher body weight compared to rats on normal diet from 3 months onwards (Figure 4A). At the end of 10 months, rats fed with high fat diet weighed an average of 937 ± 179.5 g, while those on normal diet weighed about 541 ± 15.3 g. Serum ADN levels were evaluated in all rats prior to treatment initiation. Significant differences in serum ADN levels were obtained between the high fat and normal diet fed groups (Figure 4B). Specifically, about 3.4 fold lower ADN was observed in the serum of rats on high fat diet compared to those on normal diet. Alongside, serum TNF-α levels were significantly higher (2.3 fold) in obese diabetic rats compared to age-matched healthy controls, which validates presence of low-grade systemic inflammation in the obese-diabetic rodent model (Figure 4C). Insulin sensitivity at the end of 10 months was also evaluated in all rats and a significant loss in insulin sensitivity was observed in high fat diet fed group compared to their healthy counterparts (Figure 5). Specifically, only about 16% drop in blood glucose level from initial level was noted in 2 h in high fat diet fed group in week 0 (prior to treatment commencement) vs a 52% drop in age-matched healthy rats. Upon treatment with pADN in chitosan-oleic-AHP micelles, insulin sensitivity was restored to normal within a week of single dose administration. More specifically, a 56% drop in blood glucose from initial levels was observed in 2 h in the pADN-chitosan-oleic-AHP treated group in 1 week, which was comparable to healthy rats. This improvement in insulin sensitivity was sustained till the end of the study at 6 weeks, where a 52% drop from initial levels was observed at 2 h in weeks 3 and 6. On the contrary, rats administered with free pADN did not show any significant improvement in insulin sensitivity compared to before treatment during the entire course of study. Blood glucose drop of about 21, 17 and 26.5% from initial levels were observed in 2 h in weeks 1, 3 and 6, respectively. Rats that did not develop insulin resistance despite being on high fat diet were administered empty chitosan-oleic-AHP formulation. This group did not show any change in insulin sensitivity in the 6 weeks after formulation administration (Figure 6). Blood glucose drop of about 52–56% from initial levels was noted in weeks 1–6 at 2 h, which was similar to 57% drop observed in this group in week 0.

Figure 4. Physiological changes in rats fed with high fat diet for 10 months.

A) Weight gain trajectory of rats on high fat or control diet. Age matched Wistar rats were placed on high fat or normal (control) diet for 10 months. A significant difference (p < 0.001) in weight gain was observed between both groups from 3 months onwards. Data represented as mean ± S.E. (n = 6 and 20 for normal diet and high fat diet fed groups, respectively). Serum levels of B) Adiponectin and C) TNF-α were determined before initiation of treatment in rats on normal or high fat diet for 10 months. A significant difference (p < 0.05) was observed between normal and high fat diet groups. Data represented as mean ± S.E. (n = 5 and 10 for normal diet and high fat diet fed groups, respectively).

Figure 5. Insulin sensitivity before and after treatment with various formulations.

Treatment with pADN in chitosan-oleic-AHP micelles led to similar trend in blood glucose drop as observed in healthy rats for up to 6 weeks. Rats treated with pADN in saline did not show improvement in insulin sensitivity. Data represented as mean ± S.E. (n = 5). Compared to week 0 high fat diet, a significant difference in 2 h blood glucose levels in the pADN micelle treated group at weeks 1 and 3 was observed (p < 0.05). Week 0 represents before treatment.

Figure 6. Insulin sensitivity before and after treatment with empty chitosan-oleic-AHP formulation.

Rats treated with empty chitosan-oleic-AHP micelles demonstrated similar insulin sensitivity compared to before treatment. These rats were placed on high fat diet for 10 months but did not develop insulin resistance during the period. Data represented as mean ± S.E. (n = 4). Week 0 represents before treatment. No significant difference in blood glucose levels at all time points in weeks 1, 3 and 6 was observed compared to week 0.

A concomitant significant improvement in serum ADN was observed in pADN-chitosan-oleic-AHP treated group within 1 week of single dose administration, which was akin to those in normal diet group in week 1 (Figure 7). After 3 weeks of treatment, serum ADN level remained significantly elevated compared to before treatment and no statistically significant difference was observed between the formulation-treated and healthy groups. However, by 6 weeks, serum ADN decreased significantly compared to healthy controls. On the other hand, rats injected with free plasmid, had significantly lower serum ADN levels compared to healthy controls throughout the study period. The group injected with empty micelles did not demonstrate any significant modulation in ADN levels in 6 weeks after treatment compared to before treatment levels.

Figure 7. Serum adiponectin levels before and after treatment with various formulations.

A single dose administration of pADN in chitosan-oleic-AHP micelles increased adiponectin levels in rats on high fat diet that were analogous to healthy rats for up to 3 weeks after treatment. * represents significant difference (p < 0.05) in ADN concentration between rats on normal diet and other formulations in the specific week. # represents significant difference between pADN in saline and pADN in chitosan micelle group in that particular week. Data represented as mean ± S.E. (n = 5 for normal diet, pADN in saline and pADN in chitosan micelles; n = 4 for empty micelle group). Week 0 represents before treatment.

3.6. Histology

Histological examination of various organs harvested 24 h or 1 week after treatment with saline or chitosan-oleic-AHP (formulation) did not show any morphological changes between the groups (Figure 8). No structural damage was noted in various organ/tissue sections upon administration of the formulation when compared to saline.

Figure 8. Biocompatibility of pADN chitosan-oleic nanomicelles.

Photomicrographs of H&E staining of various organs harvested from high fat diet fed rats treated with saline or formulation (pADN in chitosan-oleic-AHP) after 24 h or 1 week of treatment. Pictures taken at 20 X magnification. Scale bar is 25 μm.

4. Discussion

Obesity-associated T2DM is considered an inflammatory disorder due to overwhelming evidence supporting a positive correlation between pro-inflammatory adipocytokine levels and insulin resistance (Banerjee and Singh 2019; Freitas Lima, et al. 2015; Kang, et al. 2016; King 2008). Several preclinical and clinical studies have shown an increase in TNF-α, IL-1β, IL-6 and monocyte chemoattractant protein 1 (MCP-1) levels in obese diabetic patients, along with a concomitant decrease in anti-inflammatory adipocytokines such as adiponectin and IL-10. Expectedly, neutralization of pro-inflammatory adipocytokines and supplementation with anti-inflammatory adipocytokines can result in significant alleviation of insulin resistance (Agrawal and Kant 2014; Banerjee and Singh 2019; Rabe, et al. 2008; Wieser, et al. 2013). Amongst various adipocytokines, ADN is the most abundant adipocytokine in healthy individuals. ADN possesses potent anti-inflammatory, anti-diabetic and anti-atherogenic functions (Aleidi, et al. 2015; Tilg and Wolf 2005). Specifically, ADN plays a vital role in lipid and glucose metabolism by stimulating glucose uptake in skeletal and cardiac muscles, inhibiting hepatic glucose production and augmenting fatty acid oxidation in insulin sensitive organs (Chandran, et al. 2003; Karbowska and Kochan 2006). In addition, this adipokine is known to suppress synthesis of pro-inflammatory cytokines such as TNF-α and IL-6 while upregulating production of anti-inflammatory molecules including IL-10 and IL-1 receptor antagonist (Tilg and Wolf 2005). Diminished production of ADN in obesity leads to hypoadiponectinemia, which is strongly correlated with insulin resistance (Aleidi, et al. 2015; Kadowaki, et al. 2006). Few studies have reported that exogenous administration of ADN as recombinant protein decreased hyperglycemia, adiposity and weight gain in obese-diabetic mice (Berg, et al. 2001); (Masaki, et al. 2003). However, a study by Novo Nordisk failed to achieve any blood glucose-lowering efficacy upon administration of recombinant ADN in diabetic mice and rats (Tullin, et al. 2012). They speculated that ADN might need processing by adipocytes to demonstrate bioactivity. In light of this observation, gene therapy targeted to adipose tissue is a promising approach to stimulate proper processing and production of ADN from adipocytes, to obtain potent anti-inflammatory/anti-diabetic efficacy in obesity-associated T2DM.

For gene delivery, an ideal nanoparticle should (i) form a stable complex with the DNA, (ii) protect the gene cargo from endonuclease based degradation, (iii) promote efficient uptake of the polyplex in the target cells, (iv) assist endosomal escape upon internalization, (v) release gene load in the cytosol, and (vi) show minimal toxicity (Cao, et al. 2019). Chitosan exhibits most of these desirable characteristics and is therefore extensively used for gene delivery. Positive charge on chitosan’s primary amine backbone can form stable complexes with negatively charged DNA molecules through electrostatic interaction. In addition, chitosan demonstrates excellent biocompatibility (safety, biodegradability and low immunogenicity), as well as ease of chemical modification (Mansouri, et al. 2004). Moreover, chitosan greatly enhances the stability of complexed gene against endonuclease degradation and prevents lysosomal sequestration by proton sponge effect, thereby improving gene delivery (Layek and Singh 2013a; Mao, et al. 2001). However, a major shortcoming of chitosan-based gene delivery is its poor transfection efficiency due to formation of tight complexes with genes, resulting in poor intracellular release of complexed genes. In previous studies, we have significantly improved cellular uptake and transfection of chitosan nanoparticles by grafting chitosan polymer with hydrophobic molecules such as fatty acids or hydrophobic amino acids (Layek and Singh 2012; Layek and Singh 2013a; Layek and Singh 2013b; Layek, et al. 2014; Sharma and Singh 2017b). In many cases, hydrophobically modified chitosan demonstrated higher transfection efficacy compared to standard transfecting agent FuGENE®HD (Layek and Singh 2012; Layek and Singh 2013b; Layek, et al. 2014). Hydrophobic modifications prevent the formation of very tight chitosan-gene complexes and improve intracellular dissociation of loaded gene from the particles, leading to enhanced transfection. In addition, ω−3 fatty acids such as oleic acid can aid nanoparticle internalization, increase ADN production, promote fatty acid utilization in the adipose tissue and decrease expression of inflammatory adipokines such as MCP-1 (Rombaldova, et al. 2017; Spencer, et al. 2013). The grafting of fatty acids on chitosan was previously confirmed using proton nuclear magnetic resonance and fourier transform infrared spectroscopy (Sharma and Singh 2017a; Sharma and Singh 2017b). The critical micellar concentration of chitosan-oleic acid micelles was also previously evaluated using pyrene as a hydrophobic fluorescence probe and determined to be 65 μg/mL (Sharma and Singh 2017a; Sharma and Singh 2017b). In addition, prior atomic force microscopy studies have revealed that these nanomicelles are spherical and their sizes are in agreement with those obtained using dynamic light scattering (Sharma and Singh 2017a; Sharma and Singh 2017b). To achieve successful ADN based diabetes therapy, it is imperative to deliver pADN specifically to the adipocytes. Drug delivery to adipose tissue has been successfully carried out using white adipose tissue homing peptide (KGGRAKD; AHP) that binds to prohibitin (Kolonin, et al. 2004; Salameh, et al. 2016; Won, et al. 2014; Xue, et al. 2016). Prohibitin is abundantly expressed in adipose tissue vasculature and in the cell membrane of 3T3-L1 adipocytes (Ande, et al. 2009; Won, et al. 2014). AHP has been used in adipocyte-targeted delivery of short hairpin RNA against retinol binding protein-4 and rosiglitazone (Won, et al. 2014; Xue, et al. 2016). Factoring in the benefits of hydrophobically modified chitosan and adipocyte targeting via AHP, we attempted to utilize chitosan-oleic-AHP nanoparticles for adiponectin gene therapy.

Chitosan-oleic-AHP formulation demonstrated several favorable characteristics for efficient gene delivery such as small size, particle homogeneity, slight positive charge, ability to protect complexed DNA from enzymatic degradation and high entrapment efficiency. Ideally, particle size should be above 100 nm to prevent rapid lymphatic uptake of the particles after subcutaneous administration, which was the route used to administer these particles in vivo (Oussoren, et al. 1997; Xie, et al. 2009). Chitosan-oleic-AHP micelles were similar in size to chitosan-oleic micelles but significantly smaller than chitosan or chitosan-AHP particles. This may be attributed to formation of well-structured micellar assemblies encapsulating DNA compared to relatively unstructured chitosan nanoparticles associated with DNA. Meanwhile, it is beneficial for a gene vector to possess positive charge to form stable complexes with DNA and allow interaction of the polyplexes with negatively charged cell membranes, to aid cellular internalization. Upon receptor-mediated endocytosis, the positive charge on the carrier also helps in endosomal escape of the accompanying gene due to buffering effect of unprotonated amines in chitosan (Layek and Singh 2013a; Richard, et al. 2013). However, abundance of positive charges on the polymer backbone ensues a very tight association of gene with the carrier, which can impose difficulty in release of the gene material, thereby hindering gene transfection (Schaffer, et al. 2000). Grafting of fatty acids on chitosan leads to reduction in free amine groups, decrease in positive charge, weaker electrostatic interaction with DNA, and easy release of complexed DNA for transfection. In this study we observed that despite reduction in positive charges through incorporation of hydrophobic oleic acid in the chitosan polymer, the entrapment efficiency of pADN in chitosan-oleic-AHP micelles was relatively high (~93%) and comparable to unmodified chitosan (95%). These results suggest that hydrophobic modification in the chitosan did not affect pADN loading. Stability analysis of pDNA in the nanomicelles demonstrated that the nanomicelles could efficiently protect the loaded genes in the normal physiology i.e., in the presence of endonucleases. In vitro efficacy study performed on differentiated adipocytes asserted the superior transfection capability of chitosan-oleic-AHP micelles. Compared to basal level, about 2-fold higher ADN was obtained with chitosan-oleic-AHP treatment. ADN concentration upon treatment with chitosan-oleic-AHP was also significantly higher than that obtained using standard transfecting agent FuGENE®HD.

Under inflammatory conditions such as in obesity, a lower basal ADN level is expected, and treatment with pADN-chitosan-oleic-AHP can potentially bring about pronounced enhancement in ADN concentrations. We tested this postulation through studies in obese diabetic rats that present a model for hypoadiponectinemia, low-grade systemic inflammation and insulin resistance. Indeed, a large difference in insulin sensitivity, serum ADN as well as TNF-α levels were observed between rats on high fat and normal diets. However, a single subcutaneous injection of 1 mg/kg pADN in chitosan-oleic-AHP micelles completely offset insulin resistance and hypoadiponectinemia within a week of treatment. High serum ADN levels persisted for 3 weeks after treatment and thereafter declined. This gradual drop in serum ADN concentration was expected since chitosan and several other non-viral vector based transfections do not permanently incorporate the DNA into the host genome (Ramamoorth and Narvekar 2015). This transient transfection is desirable, as it allows the therapy to be modulated as needed. In this study, despite decrease in systemic ADN concentration in 6 weeks, the ADN levels were still significantly higher in the chitosan-oleic-AHP treated group compared to free pADN treated group, and insulin sensitivity was comparable to healthy controls till 6 weeks. This prolonged efficacy of ADN gene therapy using chitosan-oleic-AHP micelles demonstrates tremendous promise of this nanomedicine in improving efficacy of T2DM treatment. The current pharmacotherapy for T2DM does not adequately address treatment of underlying inflammation and requires frequent medication intake for diabetes management. Moreover, since the therapies are not targeted to insulin sensitive organs, a higher dose administration is typically needed, which raises concerns of toxicity due to frequent and off-target accumulation of drugs. For example, thiazolidinediones suppress inflammation but the therapy is plagued with several very severe adverse effects including hepatotoxicity, cancer and heart failure, resulting in many deaths (Rizos, et al. 2009; Scheen 2001).

This study presents a highly effective treatment approach for insulin resistance in obesity-associated T2DM. With our approach of supplementing insulin-sensitizing adipokine through targeted gene delivery to adipocytes, we achieved a significant modulation in insulin resistance in obese-diabetic rats for several weeks. Using hydrophobically modified chitosan nanomicelles and adipose homing sequence, we were able to overcome traditional limitations of gene delivery. We strongly believe that this novel therapy that treats the underlying propagator of insulin resistance, will have potential to improve treatment outcomes and quality of lives of millions of patients suffering from obesity-associated T2DM worldwide.

5. Conclusion

The study demonstrates the efficacy of adiponectin gene therapy using adipocyte targeted chitosan nanomicelles for the treatment of insulin resistance in obesity-associated T2DM. ADN possesses potent anti-inflammatory and anti-diabetic activity and is therefore very useful in sequestering chronic systemic inflammation, which spearheads insulin resistance in T2DM. Oleic acid conjugated chitosan nanomicelles were surface modified with adipose-targeted sequence to ensure that the gene is processed in adipocytes, and efficient transfection is achieved both in vitro and in vivo. A single dose administration of the formulation resulted in reversal of insulin resistance in obese-diabetic rats for up to 6 weeks and brought serum adiponectin levels akin to healthy rats for up to 3 weeks after treatment. The study indicates that pADN-chitosan-oleic-AHP nanomedicine can be very useful for mitigation of insulin resistance for prolonged period. Further studies involving multiple doses administered at different time points might be of relevance in optimizing the treatment plan. In addition, rigorous safety studies are paramount to successful development and potential clinical translation of the nanomedicine.

Highlights.

• Adipose tissue inflammation drives insulin resistance in type 2 diabetes mellitus

• Adiponectin gene therapy can offset inflammation, improve insulin sensitivity

• Adipocyte targeted chitosan-oleic acid micelles were loaded with adiponectin gene

• The nanomicelles elevated adiponectin levels for 3 weeks in obese-diabetic rats

• Insulin sensitivity improved for 6 weeks, matched healthy rats

Abbreviations:

ADN

adiponectin

AHP

adipose homing peptide

BCA

bicinchoninic assay

DMEM

Dulbecco’s modified eagle’s media

DNase I

deoxyribonuclease I

ELISA

enzyme linked immunosorbent assay

FBS

fetal bovine serum

H&E

hematotoxylin and eosin

IL

interleukin

IBMX

3-Isobutyl-1-methylxanthine

MCP-1

monocyte chemoattractant protein 1

pADN

plasmid adiponectin

PBS

phosphate buffered saline

pβgal

plasmid β-galactosidase

PDI

polydispersity index

pDNA

plasmid DNA

SD

standard deviation

SE

standard error

T2DM

type 2 diabetes mellitus

TNF-α

tumor necrosis factor α