Effective treatment of diabetic kidney disease (DKD) remains a large unmet medical need. Within the disease’s complicated pathogenic mechanism, activation of the advanced glycation end products (AGEs)–receptor for AGE (RAGE) axis plays a pivotal role in the development and progression of DKD. To provide a new therapeutic strategy against DKD progression, we developed a vaccine against RAGE. Three rounds of immunization of mice with the RAGE vaccine successfully induced antigen-specific serum IgG antibody titers and elevated antibody titers were sustained for at least 38 weeks. In addition, RAGE vaccination significantly attenuated the increase in urinary albumin excretion in streptozotocin-induced diabetic mice (type 1 diabetes model) and leptin-receptor–deficient db/db mice (type 2 diabetes model). In microscopic analyses, RAGE vaccination suppressed glomerular hypertrophy and mesangial expansion in both diabetic models and significantly reduced glomerular basement membrane thickness in streptozotocin-induced diabetic mice. Results of an in vitro study indicated that the serum IgG antibody elicited by RAGE vaccination suppressed the expression of AGE-induced vascular cell adhesion molecule 1 and intracellular adhesion molecule 1 in endothelial cells. Thus, our newly developed RAGE vaccine attenuated the progression of DKD in mice and is a promising potential therapeutic strategy for patients with DKD.

According to the World Health Organization, the global prevalence of diabetes among adults has risen dramatically from 4.7% in 1980% to 8.5% in 2014, thus representing 422 million adults with diabetes worldwide (1). Over time, diabetes leads to serious microvascular complications and consequent damage to nerves, eyes, and kidneys. Diabetic kidney disease (DKD), defined as albuminuria or a decreased glomerular filtration rate, is present in >25% of people with diabetes (2,3). DKD is the leading cause of end-stage renal disease in approximately one half of incident cases (4) and increases cardiovascular events (5) and mortality among those with diabetes (6,7). Despite current approaches to the management of hyperglycemia and pharmacologic blockade of the renin–angiotensin system, the prevalence of DKD remains high (2). Therefore, to reduce burden of DKD worldwide, agents that directly target the pathogenic mechanism of DKD are needed.

Although the complicated mechanisms that drive the development and progression of DKD remain poorly understood, prolonged hyperglycemia is well known to lead to metabolic and hemodynamic derangements that involve the activation of the advanced glycation end products (AGEs)–receptor for AGEs (RAGE) axis (8), excessive production of NADPH oxidase–derived reactive oxygen species (9), and protein kinase C activation (10). Since the identification of RAGE (11), increasing evidence over the last several decades has indicated that the interaction between AGEs and RAGE plays a pivotal role in the development and progression of DKD. In fact, inhibitors of AGE formation (1214), agents that break cross-links within AGEs (15), and RAGE inhibitors, including monoclonal antibodies (16) and aptamers (17), all attenuate albuminuria in animal models of diabetes.

Given that DKD often progresses asymptomatically over decades and therefore must be treated with multiple prescribed medications over the long term (18), medication adherence and compliance remain important considerations in the treatment of DKD. Poor adherence to and compliance with antidiabetic medication regimens increases the risk of end-stage renal disease in patients with DKD (19). Therefore, the development of therapeutic strategies to improve adherence to and compliance with DKD treatment would be advantageous. In this regard, vaccination may be a promising approach because of its prolonged therapeutic effect and low frequency of administration.

In this study, we developed a therapeutic vaccine against DKD by targeting RAGE and demonstrated that three rounds of subcutaneous immunization with the RAGE vaccine successfully induced beneficial immune responses that attenuated the development of DKD in mice.

Animals

Male DBA/2JJcl and BKS.Cg-+Leprdb/+Leprdb/Jcl (db/db) mice were obtained from CLEA Japan, Inc. (Tokyo, Japan) and housed with ad libitum food and water under standard 12:12-h light/dark conditions. All experiments were performed in accordance with the Institutional Guidelines on Animal Experimentation at Keio University and were approved by the Keio University Institutional Animal Care and Use Committee (approval number 16071–1; Tokyo, Japan). To generate a type 1 diabetic model, 10-week-old DBA/2J mice were intraperitoneally injected with low-dose streptozotocin (STZ; 40 mg/kg) (Sigma-Aldrich Japan, Tokyo, Japan) in 0.1 mol/L sodium citrate buffer (pH 4.5) once daily for 5 consecutive days. We used db/db mice as a model for type 2 diabetes.

Preparation of RAGE Vaccine

A RAGE partial peptide (GenPept accession number NP_031451.2; amino acids 38–44) was synthesized by Eurofins Genomics (Tokyo, Japan) and coupled to the keyhole limpet hemocyanin (KLH) protein via the cross-linker m-maleimidobenzoyl-N-hydroxysuccinimide ester (Thermo Fisher Scientific, Waltham, MA). To form a water-in-oil emulsion of antigens for enhancing antigen-specific immune responses, we mixed 20 μg RAGE-KLH antigen (in 100 μL aqueous solution) with 100 μL Freund’s adjuvant (Sigma-Aldrich Japan). Complete Freund’s adjuvant was used for the first immunization; incomplete Freund’s adjuvant was used for subsequent immunizations.

Immunization Protocol

Male DBA/2J mice (age 4 weeks) were allocated into two groups and subcutaneously immunized with either the RAGE vaccine (20 μg RAGE-KLH antigen and Freund’s adjuvant; total volume, 200 μL/mouse) or vehicle (20 μg KLH and Freund’s adjuvant; total volume, 200 μL/mouse) on three occasions at 2-week intervals (i.e., at 4, 6, and 8 weeks of age). In addition, male db/db mice (age 10 weeks) were allocated into two groups and subcutaneously immunized at 10, 12, and 14 weeks of age with either the RAGE vaccine or vehicle in the same manner as the DBA/2J mice.

Evaluation of Antibody Titers

Serum IgG antibody titers (n = 4/group) were determined using ELISAs. Each well of a 96-microwell plate (Immulon 1B; Thermo Fisher Scientific) was coated with 1.0 μg RAGE partial peptide (amino acids 38–44) conjugated with BSA in PBS and incubated at 4°C overnight. The plate was then incubated for 1 h at 25°C with blocking buffer (PBS containing the nonionic detergent Tween 20 and 1% BSA) before adding serum samples. Sera were diluted with blocking buffer (by serially diluting serum samples with an equal volume of blocking buffer for each dilution), added to each well, and incubated for 2 h at 25°C. After the plates had been washed with PBS containing the nonionic detergent Tween 20 and 1% BSA, goat anti-mouse IgG (dilution factor 1:5,000) (SouthernBiotech, Birmingham, AL) was added to each well, and the plate was incubated for 1.5 h at 25°C. Reactions were visualized using the TMB Microwell peroxidase substrate system (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD), and the optical density at a wavelength of 450 nm (OD450) was determined. The end point titer was expressed as the reciprocal log2 of the last dilution that gave an OD450 that was 0.1 unit greater than that of the negative control (20). In addition, antigen-specific serum IgM, IgA, and IgE antibody titers were evaluated in the same manner, by using goat anti-mouse IgM, goat anti-mouse IgA, and rat anti-mouse IgE (dilution factor 1:5,000) (SouthernBiotech).

Western Blotting Analyses

RAGE partial peptide (amino acids 38–44) conjugated with BSA, recombinant full-length RAGE protein (Sino Biological, Beijing, China), and KLH protein was separated by means of SDS-PAGE (4–12% gradient gel) and transferred onto a polyvinylidene fluoride membrane (Merck Millipore, Darmstadt, Germany) for Western blotting. These proteins were detected in the Western blots using serum from either KLH-treated or vaccinated mice (1:200) followed by peroxidase-conjugated anti-mouse IgG secondary antibodies (1:20,000; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). The protein bands were visualized using SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher Scientific) and detected with an ImageQuant LAS 4000 Mini chemiluminescence imager (GE Healthcare Bio-Sciences KK, Tokyo, Japan).

Biochemical Studies

Blood glucose was measured at 2 and 12 weeks after the injection of STZ in DBA/2J mice (i.e., at 12 and 22 weeks of age) and at 10 weeks of age in db/db mice by means of the glucose oxidase method and an automated glucose meter (GlutestAce R; Sanwa Kagaku Kenkyusho Co., Ltd., Nagoya, Japan). Urine samples (24 h) were collected from metabolic cages at 8 weeks after the injection of STZ in DBA/2J mice (n = 10/group) and at 10 weeks of age (just before the initiation of vaccination) and at 14 weeks after the final immunization in db/db mice (n = 6/group). Urinary albumin was measured by using a direct competitive ELISA according to the manufacturer’s protocol (Albuwell M; Exocell, Inc., Philadelphia, PA). Urinary creatinine, serum creatinine, serum urea nitrogen, AST, ALT, and serum glycoalbumin were measured enzymatically.

Light Microscopic Analyses

The right kidney was removed from DBA/2J mice at 12 weeks after the injection of STZ (n = 7–9/group) and from db/db mice at 14 weeks after the final immunization (n = 5/group), fixed in 4% paraformaldehyde, embedded in paraffin blocks, and sectioned. Paraffin sections were stained with periodic acid Schiff (PAS). To determine the glomerular volume in each mouse, the cross-sectional area was measured in at least 30 consecutive unselected glomeruli. The glomerular volume [V(G)] was obtained as: V(G) = β/κ × [Ā(G)]3/2, where Ā(G) is the cross-sectional glomerular area, β = 1.38 as pertains to spheres, and κ (a distribution coefficient) is 1.10 (21). In the same glomeruli, the mesangial matrix area was calculated as the area of positive PAS staining, expressed as a percentage of the total cross-sectional glomerular area.

In addition, paraffin sections were stained with Wilms tumor 1 (WT-1) antibody (Santa Cruz Biotechnology, Inc., Dallas, TX) as a podocyte marker followed by incubation with diaminobenzidine. In each mouse, the WT-1–positive podocytes in at least 30 consecutive unselected glomeruli were counted.

Electron Microscopic Analyses

At 12 weeks after the injection of STZ, the left kidney of DBA/2J mice (n = 4/group) was fixed with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.4) at 4°C overnight and postfixed with 2% osmium tetroxide in 0.1 mol/L phosphate buffer at 4°C for 2 h. The kidney sample was dehydrated in graded ethanol solutions, infiltrated with propylene oxide, embedded in 100% resin, and then incubated at 60°C for 48 h to polymerize the resin. The polymerized resin was ultra-thin sectioned at 70 nm by using an ultramicrotome with a diamond knife, and sections were mounted on copper grids. Each section was stained with 2% uranyl acetate at room temperature for 15 min and then secondary stained with lead stain solution (Sigma-Aldrich Japan) at room temperature for 3 min. The grids were observed under a transmission electron microscope (JEM-1400Plus; JEOL, Ltd., Tokyo, Japan) at an acceleration voltage of 100 kV. Digital images (3,296 × 2,472 pixels) were taken by using a charge-coupled device camera (EM-14830 RUBY2; JEOL, Ltd.).

The thickness of the glomerular basement membrane (GBM) was determined using the orthogonal intercept method (22). From each mouse, we captured 32–40 digital images at ×16,800 magnification (1 image per capillary loop, 8–10 capillary loops per glomerulus, and 4 glomeruli per animal). A grid mask with 16 intercepts was applied to each image using Photoshop CS4 (Adobe Systems, San Jose, CA). GBM thickness—from the basal endothelial to the basal podocyte plasma membrane—was measured at each intercept.

Antibody Function Analyses

The ability of IgG antibodies from vehicle-treated or vaccinated mice to inhibit RAGE signaling in vitro was assessed using human umbilical vein endothelial cells (HUVECs) (PromoCell, Heidelberg, Germany). HUVECs were plated in low-serum (2% v/v) media including growth factors (ScienCell Research Laboratories, Carlsbad, CA) and cultured at 37°C in a 5% CO2 atmosphere. Serum polyclonal IgG antibodies from both vehicle-treated and vaccinated mice were obtained by binding to Protein G Sepharose (GE Healthcare Bio-Sciences KK), followed by elution with glycine-HCl (pH 2.5) and neutralization with 1 mol/l Tris-HCl (pH 9.0). Purified IgGs were added to HUVECs at 10 μg/mL, after which the cells were incubated for 2 h at 37°C and then stimulated with AGE-BSA (50 μg/mL) (Abcam, Cambridge, U.K.) or BSA (50 μg/mL) (Sigma-Aldrich Japan) for 4 h at 37°C in normoglycemic conditions partially according to the previous article (23). Total RNAs were isolated from these cells by using TRIzol reagent (Thermo Fisher Scientific); 1 μg total RNA was reverse transcribed using PrimeScript RT Master Mix (Takara Bio, Otsu, Japan). The resulting synthesized cDNA underwent quantitative PCR amplification by using SYBR Green Master Mix (Thermo Fisher Scientific); the oligonucleotide primers for this amplification are listed in Supplementary Table 1. We obtained relative mRNA expression data for vascular cell adhesion molecule 1 (VCAM-1) and intracellular adhesion molecule 1 (ICAM-1) by using the results of real-time quantitative PCR amplification of these target transcripts normalized against those for β-actin mRNA.

Statistical Analysis

Data are expressed as mean ± SEM. Statistical comparisons between two groups were made by using ANOVA followed by an unpaired two-tailed Student t test; P values <0.05 were considered statistically significant.

Data and Resource Availability

The data sets generated or analyzed during the current study are available from the corresponding author on request.

Selection of Antigen Sequence for the RAGE Vaccine and its Antibody Response

For use as an epitope, we selected the RAGE partial peptide comprising amino acids 38–44, because it overlaps the AGE-binding site and surface-exposed region (2426) and in light of its amino acid properties, such as hydrophilicity and charge (27); these features increase the likelihood of neutralizing antibody production. To increase its antigenicity, we coupled the RAGE partial peptide to KLH protein for use as the RAGE vaccine.

To confirm the immunogenicity of our RAGE vaccine, we subcutaneously immunized 4-week-old DBA/2J mice with three doses of vaccine, each containing 20 μg antigen. This vaccination protocol rapidly induced serum IgG antibodies against RAGE (reciprocal log2 titer 15.25 ± 0.83 at 8 weeks after the last immunization; n = 4). In addition, the anti-RAGE IgG titer increased after each vaccine dose, reached its peak at 8 weeks after the last immunization (16 weeks of age), and then gradually decreased over several months (Fig. 1A). In contrast to serum IgG levels, antigen-specific serum IgM was induced at a lower level than IgG, whereas IgA and IgE antibodies were not detected in immunized mice (Fig. 1B).

Figure 1

Antigen-specific serum antibody titers. Male DBA/2J mice were immunized with the RAGE vaccine at 4, 6, and 8 weeks of age. Sera were obtained from immunized mice, and antigen-specific antibody titers (n = 4) were determined by using an ELISA. A: Antigen-specific serum IgG antibody titers in immunized mice were examined until 46 weeks of age (i.e., 38 weeks after the final vaccination). The antibody titer increased rapidly after the second immunization (i.e., at 6 weeks), reaching its peak at 4 to 8 weeks after the final vaccination. Although it gradually decreased over time, an immune response was retained until 38 weeks after vaccination. B: At 4 weeks after the final immunization, antigen-specific serum IgG antibodies predominated. Antigen-specific IgM antibodies were induced also, but other Ig isotypes (i.e., IgA and IgE) were scarce. C: In the Western blotting analysis, sera from mice immunized with RAGE vaccine bound to epitopes within the vaccine antigen and full-length RAGE protein. In contrast, mice immunized with the KLH vehicle did not develop antibodies that specifically bound to either full-length RAGE protein or the vaccine antigen. All data are expressed as mean ± SEM.

Figure 1

Antigen-specific serum antibody titers. Male DBA/2J mice were immunized with the RAGE vaccine at 4, 6, and 8 weeks of age. Sera were obtained from immunized mice, and antigen-specific antibody titers (n = 4) were determined by using an ELISA. A: Antigen-specific serum IgG antibody titers in immunized mice were examined until 46 weeks of age (i.e., 38 weeks after the final vaccination). The antibody titer increased rapidly after the second immunization (i.e., at 6 weeks), reaching its peak at 4 to 8 weeks after the final vaccination. Although it gradually decreased over time, an immune response was retained until 38 weeks after vaccination. B: At 4 weeks after the final immunization, antigen-specific serum IgG antibodies predominated. Antigen-specific IgM antibodies were induced also, but other Ig isotypes (i.e., IgA and IgE) were scarce. C: In the Western blotting analysis, sera from mice immunized with RAGE vaccine bound to epitopes within the vaccine antigen and full-length RAGE protein. In contrast, mice immunized with the KLH vehicle did not develop antibodies that specifically bound to either full-length RAGE protein or the vaccine antigen. All data are expressed as mean ± SEM.

In the Western blotting analysis, sera from mice immunized with RAGE vaccine bound not only to epitopes within the vaccine antigen, but also to full-length RAGE protein (Fig. 1C). In contrast, mice immunized with the KLH vehicle only did not develop antibodies that specifically bound to either full-length RAGE protein or the vaccine antigen (Fig. 1C).

In terms of safety, we measured the levels of AST, ALT, urea nitrogen, and creatinine in serum from healthy nondiabetic mice immunized with RAGE-KLH vaccine. These levels were within normal limits, thus indicating no significant impairment of liver or kidney function in these immunized mice (Fig. 2).

Figure 2

Biochemical profile of vaccinated nondiabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle without the injection of STZ. Sera (n = 5/group) were obtained at 12 weeks after the final vaccination, and AST (A), ALT (B), blood urea nitrogen (BUN; C), and creatinine (Cr; D) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test.

Figure 2

Biochemical profile of vaccinated nondiabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle without the injection of STZ. Sera (n = 5/group) were obtained at 12 weeks after the final vaccination, and AST (A), ALT (B), blood urea nitrogen (BUN; C), and creatinine (Cr; D) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test.

Evaluation of the RAGE Vaccine in an STZ-Induced Type 1 Diabetic Mouse Model

To evaluate the renoprotective effect of the RAGE vaccine in a model of type 1 diabetes, DBA/2J mice were immunized at 4, 6, and 8 weeks of age with either RAGE vaccine (20 μg antigen) or KLH vehicle only; these mice then received low-dose STZ (to induce type 1 diabetes) at 2 weeks after the last vaccination. Blood glucose, glycoalbumin, and body weight did not differ significantly between vehicle-treated mice and vaccinated mice (Table 1). In contrast, subcutaneous immunization with the RAGE vaccine significantly attenuated albuminuria (333.6 ± 104.7 μg/mg creatinine in vaccinated mice vs. 850.6 ± 208.6 μg/mg creatinine in vehicle-treated mice; P = 0.04) (Fig. 3A) in the type 1 diabetic mouse model. In contrast, the RAGE vaccine had no significant effect on either serum creatinine or urea nitrogen levels (Fig. 3B and C).

Figure 3

Effect of RAGE vaccine on urinary albumin and serum urea nitrogen and creatinine levels in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. A: Urine samples (n = 10/group) were obtained at 8 weeks after the final injection of STZ, and the urine albumin-to-creatinine ratio was calculated. Sera (n = 7–9/group) were obtained at 12 weeks after the final injection of STZ, and serum urea nitrogen (B) and creatinine (C) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. *P < 0.05.

Figure 3

Effect of RAGE vaccine on urinary albumin and serum urea nitrogen and creatinine levels in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. A: Urine samples (n = 10/group) were obtained at 8 weeks after the final injection of STZ, and the urine albumin-to-creatinine ratio was calculated. Sera (n = 7–9/group) were obtained at 12 weeks after the final injection of STZ, and serum urea nitrogen (B) and creatinine (C) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. *P < 0.05.

Table 1

Blood glucose, body weight, and kidney weight of STZ-induced and leptin receptor–deficient (db/db) diabetic mice

Blood glucose (mg/dL)Glycoalbumin (%)Body weight (g)Kidney weight (g)
2 weeks after STZ12 weeks after STZ12 weeks after STZRightLeft
STZ-induced mice       
 Vehicle 358.7 ± 24.3 567.0 ± 9.9 33.6 ± 1.5 24.7 ± 0.8 0.31 ± 0.02 0.31 ± 0.02 
 RAGE vaccine 369.6 ± 21.6 538.7 ± 10.8 33.4 ± 2.1 23.6 ± 0.8 0.28 ± 0.01 0.28 ± 0.01 
db/db mice      
 Vehicle 383.2 ± 10.0  50.3 ± 1.8 0.30 ± 0.01 0.30 ± 0.01 
 RAGE vaccine 376.2 ± 10.0  52.7 ± 2.3 0.27 ± 0.02 0.24 ± 0.02* 
Blood glucose (mg/dL)Glycoalbumin (%)Body weight (g)Kidney weight (g)
2 weeks after STZ12 weeks after STZ12 weeks after STZRightLeft
STZ-induced mice       
 Vehicle 358.7 ± 24.3 567.0 ± 9.9 33.6 ± 1.5 24.7 ± 0.8 0.31 ± 0.02 0.31 ± 0.02 
 RAGE vaccine 369.6 ± 21.6 538.7 ± 10.8 33.4 ± 2.1 23.6 ± 0.8 0.28 ± 0.01 0.28 ± 0.01 
db/db mice      
 Vehicle 383.2 ± 10.0  50.3 ± 1.8 0.30 ± 0.01 0.30 ± 0.01 
 RAGE vaccine 376.2 ± 10.0  52.7 ± 2.3 0.27 ± 0.02 0.24 ± 0.02* 

Data are expressed as means ± SEM; differences between mean values of treatment groups were evaluated for significance using the Student t test.

*

P < 0.05.

To further confirm the protective effect of RAGE vaccine on DKD, kidneys were analyzed histologically at 12 weeks after the injection of STZ. Although immunization with RAGE vaccine had no effect on overall kidney weight (Table 1), the total glomerular volume of vaccine-treated mice (3.18 ± 0.17 μm3) was significantly less than that of vehicle-treated mice (5.28 ± 0.52 μm3; P < 0.01) (Fig. 4A). In addition to the decrease in glomerular hypertrophy, glomerular mesangial matrix expansion was attenuated in vaccine-treated mice (18.8 ± 0.8%) compared with vehicle-only controls (26.3 ± 1.6%; P < 0.01) (Fig. 4B). Vaccinated mice had significantly more WT-1–positive podocytes (8.77 ± 0.36 cells/glomerulus) than vehicle-treated mice (6.72 ± 0.27 cells/glomerulus; P < 0.01) (Fig. 4C). Furthermore, ultrastructural morphometric study indicated that GBM thickness was significantly suppressed in vaccine-treated mice (181.3 ± 1.8 nm) compared with vehicle-treated animals (215.3 ± 1.9 nm; P < 0.01) (Fig. 5).

Figure 4

Effect of RAGE vaccine on histologic changes in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. The right kidney was removed at 12 weeks after the last injection of STZ (n = 7–9/group). The light microscopic images show representative glomeruli stained with PAS; original magnification ×200. Glomerular volume (A) and mesangial area (B) were determined from the cross-sectional glomerular area of at least 30 consecutive unselected glomeruli. Paraffin sections were also stained with WT-1 antibody as a podocyte marker. C: WT-1–positive podocytes in at least 30 consecutive unselected glomeruli were counted. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Figure 4

Effect of RAGE vaccine on histologic changes in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. The right kidney was removed at 12 weeks after the last injection of STZ (n = 7–9/group). The light microscopic images show representative glomeruli stained with PAS; original magnification ×200. Glomerular volume (A) and mesangial area (B) were determined from the cross-sectional glomerular area of at least 30 consecutive unselected glomeruli. Paraffin sections were also stained with WT-1 antibody as a podocyte marker. C: WT-1–positive podocytes in at least 30 consecutive unselected glomeruli were counted. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Figure 5

Effect of RAGE vaccine on GBM thickness in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. The left kidney was removed at 12 weeks after the injection of STZ (n = 4/group). During electron microscopic analyses, 32–40 digital images (original magnification ×16,800) were captured per mouse, and a grid mask with 16 intercepts was applied to each image. GBM thickness (from the basal endothelial to the basal podocyte plasma membrane) was measured at each intercept. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Figure 5

Effect of RAGE vaccine on GBM thickness in STZ-induced diabetic mice. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. The left kidney was removed at 12 weeks after the injection of STZ (n = 4/group). During electron microscopic analyses, 32–40 digital images (original magnification ×16,800) were captured per mouse, and a grid mask with 16 intercepts was applied to each image. GBM thickness (from the basal endothelial to the basal podocyte plasma membrane) was measured at each intercept. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Evaluation of the RAGE Vaccine in a Mouse Model of Type 2 Diabetes

To evaluate the renal protective effect of our RAGE vaccine in a type 2 diabetic model, leptin receptor–deficient db/db mice were immunized at 10, 12, and 14 weeks of age with either RAGE vaccine (20 μg antigen) or the KLH vehicle only. Although body weight was similar between groups, kidney weight was slightly less in vaccinated mice (Table 1). At 10 weeks of age (that is, just prior to the first vaccine dose), db/db mice were already albuminuric (Fig. 6A). Whereas the urine albumin-to-creatinine ratio progressively increased in vehicle-treated db/db mice, this increase in urine albumin was suppressed in db/db mice immunized with the RAGE vaccine (737.7 ± 116.4 μg/mg creatinine in vehicle-treated mice vs. 367.0 ± 88.2 μg/mg creatinine in vaccinated mice; P = 0.03) (Fig. 6A). However, as seen in the type 1 diabetic model, neither serum creatinine nor urea nitrogen differed between vehicle-treated and vaccine-treated groups in the type 2 diabetic model (Fig. 6B and C).

Figure 6

Effect of RAGE vaccine on urinary albumin and serum urea nitrogen and creatinine levels in leptin receptor–deficient (db/db) mice. Male db/db mice were inoculated with the RAGE vaccine or KLH vehicle. A: Urine samples (n = 6/group) were obtained from mice just before immunization (prevaccination; age 10 weeks) and at 14 weeks after the final immunization (postvaccination; age 28 weeks), and the urine albumin-to-creatinine ratio was calculated. Sera (n = 5/group) were obtained from mice at 14 weeks after the final immunization, and serum urea nitrogen (B) and creatinine (C) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. *P < 0.05.

Figure 6

Effect of RAGE vaccine on urinary albumin and serum urea nitrogen and creatinine levels in leptin receptor–deficient (db/db) mice. Male db/db mice were inoculated with the RAGE vaccine or KLH vehicle. A: Urine samples (n = 6/group) were obtained from mice just before immunization (prevaccination; age 10 weeks) and at 14 weeks after the final immunization (postvaccination; age 28 weeks), and the urine albumin-to-creatinine ratio was calculated. Sera (n = 5/group) were obtained from mice at 14 weeks after the final immunization, and serum urea nitrogen (B) and creatinine (C) were measured. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. *P < 0.05.

Histologic analyses of db/db mice indicated that glomerular volume was significantly less in vaccine-treated animals (3.22 ± 0.23 μm3) compared with vehicle-treated animals (5.67 ± 0.18 μm3; P < 0.01) (Fig. 7A). In addition, the glomerular mesangial matrix area was significant smaller in vaccinated mice (23.6 ± 1.0%) than vehicle-treated mice (32.7 ± 2.3%; P < 0.01) (Fig. 7B).

Figure 7

Effect of RAGE vaccine on histologic changes in leptin receptor–deficient (db/db) mice. Male db/db mice were immunized with the RAGE vaccine or KLH vehicle. The right kidney was removed at 14 weeks after the final immunization (n = 5/group). Light microscopic images show representative glomeruli stained with PAS; original magnification ×200. Glomerular volume (A) and mesangial area (B) were determined from cross-sectional glomerular areas of at least 30 consecutive unselected glomeruli. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Figure 7

Effect of RAGE vaccine on histologic changes in leptin receptor–deficient (db/db) mice. Male db/db mice were immunized with the RAGE vaccine or KLH vehicle. The right kidney was removed at 14 weeks after the final immunization (n = 5/group). Light microscopic images show representative glomeruli stained with PAS; original magnification ×200. Glomerular volume (A) and mesangial area (B) were determined from cross-sectional glomerular areas of at least 30 consecutive unselected glomeruli. Data are expressed as mean ± SEM, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

In Vitro Examination of Antibody Effects on RAGE

To determine whether IgG antibodies elicited by immunization with RAGE vaccine directly inhibit RAGE signaling, we performed an in vitro assay. In endothelial cells, AGEs are known to induce VCAM-1 and ICAM-1 via RAGE (23,28). In this regard, AGE-BSA increased the mRNA expression of both VCAM-1 and ICAM-1 in HUVECs (Fig. 8). In addition, the elevated VCAM-1 and ICAM-1 mRNA levels were almost completely suppressed in HUVECs incubated with IgG antibodies obtained from the sera of vaccinated mice (Fig. 8). In contrast, treatment with IgGs obtained from vehicle-treated mice did not suppress VCAM-1 and ICAM-1 expression (Fig. 8).

Figure 8

Effect of IgG antibodies elicited by RAGE vaccination on AGE-induced VCAM-1 and ICAM-1 mRNA expression in HUVECs. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. IgG antibodies were purified from sera of these mice. After pretreatment with purified IgGs, HUVECs were stimulated with AGE-BSA (50 μg/mL) or BSA (50 μg/mL). Relative mRNA expression data were obtained for VCAM-1 (A) and ICAM-1 (B) by using results of real-time quantitative PCR amplification normalized against those for β-actin mRNA. Data are expressed as the mean ± SEM of three independent experiments, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

Figure 8

Effect of IgG antibodies elicited by RAGE vaccination on AGE-induced VCAM-1 and ICAM-1 mRNA expression in HUVECs. Male DBA/2J mice were inoculated with the RAGE vaccine or KLH vehicle. IgG antibodies were purified from sera of these mice. After pretreatment with purified IgGs, HUVECs were stimulated with AGE-BSA (50 μg/mL) or BSA (50 μg/mL). Relative mRNA expression data were obtained for VCAM-1 (A) and ICAM-1 (B) by using results of real-time quantitative PCR amplification normalized against those for β-actin mRNA. Data are expressed as the mean ± SEM of three independent experiments, with differences between mean values evaluated for significance using the Student t test. **P < 0.01.

In this study, we developed a novel therapeutic strategy against DKD. Specifically, we generated a vaccine based on a RAGE partial peptide, and we demonstrated that immunization with this RAGE vaccine attenuated albuminuria and early renal histologic changes in two mouse models of diabetes.

DKD remains an unmet medical need of high priority due to the lack of effective treatments that can prevent or reverse disease progression. Metabolic changes associated with diabetes alter renal hemodynamics and thus promote inflammation and fibrosis via incompletely understood mechanisms (29). Among the various pathogenic mechanisms of DKD, the AGE–RAGE interaction is well known to promote the pathogenesis and progression of DKD (30,31). In fact, disruption of the RAGE gene decreases albuminuria and attenuates microscopic lesions in a diabetic mouse model (30,32). The development of a therapeutic strategy against the AGEs–RAGE axis is a recent highlight in the field of DKD treatment (1217). The significance of our present study is its demonstration of a RAGE vaccine as a potential approach that effectively targets the AGEs–RAGE axis to prevent and treat DKD.

DKD is chronic and progressive, and long-term intensive treatment is needed to prevent the advancement of disease. However, to delay DKD progression in previous animal studies, anti-RAGE–neutralizing antibody had to be administered to mice intraperitoneally three times weekly (16), and anti-RAGE DNA aptamer needed to be infused continuously in rats (17). In terms of therapeutic adherence and compliance, vaccination is a promising approach for the treatment of DKD because of its prolonged therapeutic effect and low frequency of administration (33). In our present study, high titers of antibodies against RAGE were maintained for at least 38 weeks after three rounds of immunization (Fig. 1A), leading to the possibility of a long-term therapeutic effect against DKD. In fact, we noted beneficial effects of RAGE vaccination on histologic lesions of DKD at 14 weeks after the final immunization in mouse models of both type 1 and type 2 diabetes.

We selected a RAGE partial peptide (amino acids 38–44) as a vaccine antigen. RAGE is a cell-surface receptor for AGEs and consists of three extracellular Ig-like domains (the V, C1, and C2 domains), a transmembrane helix, and an intracellular C-terminal tail (34). Findings from an AGE–RAGE binding assay using nuclear magnetic resonance spectroscopy suggest that Cys-38 and Gly-40 are involved in binding the AGE Nε-carboxy-methyl-lysine and that Cys-38, Gly-40, Ala-41, and Lys-43 are components of the AGE–BSA interaction surface (24). Site-directed mutagenesis revealed that single-substitution mutants of Lys-43 and Lys-44 markedly decreased complex formation between RAGE and AGE–BSA (25). Therefore, the RAGE partial peptide that we used as a vaccine antigen (amino acids 38–44) likely is an important site for AGE binding.

In the current study, we evaluated the efficacy of our RAGE vaccine against DKD in two mouse models of diabetes: the low-dose STZ-induced model of type 1 diabetes and the genetically obese leptin receptor–deficient db/db mice, which are a model of type 2 diabetes. These mouse models develop albuminuria and histologic changes similar to those of early DKD in humans. However, due to the relatively short experimental period, both of these animal models usually lack the marked increases in serum creatinine (Fig. 3C and 6C) and advanced histologic lesions, such as nodular glomerulosclerosis, that are characteristic of advanced DKD in human patients (35). The short life span of animal models also makes it difficult to assess the therapeutic effects. Therefore, we focused on the preventive effect in the current study. Despite the innate difficulty of accurate DKD modeling in rodents because of their relatively short life span, low-dose STZ mice and db/db mice are the most widely accepted DKD mouse models (36).

Independent of blood glucose levels, immunization with the RAGE vaccine reduced albuminuria by 60.8% in STZ-induced diabetic mice and by 50.3% in db/db mice. The results of the current study were consistent with a previous study that demonstrated that deletion of the RAGE gene in OVE26 mice (a transgenic model of severe early-onset type 1 diabetes) decreased albuminuria by 56% compared with that in RAGE-bearing, wild-type OVE26 mice (32). In addition, the effects of RAGE vaccine on the attenuation of albuminuria were equal to or greater than those of renin-angiotensin system inhibitors (37,38), sodium–glucose cotransporter 2 inhibitors (39,40), and NE2-related factor 2 agonists (41,42).

In endothelial cells, the AGE–RAGE interaction causes enhanced formation of oxygen radicals and the release of proinflammatory cytokines and adhesion molecules (43). In the current study, AGEs increased VCAM-1 and ICAM-1 expression in cultured human endothelial cells; treatment with vaccination-induced IgG antibodies apparently interrupted the AGE–RAGE axis, consequently almost completely suppressing the increased expression of VCAM-1 and ICAM-1 (Fig. 8). Mice deficient in adhesion molecules are known to be resistant to diabetic nephropathy (44). In the in vivo condition, chronic hyperglycemia promotes the AGE formation. In contrast, in the in vitro condition, AGEs induce the expression of adhesion molecules in endothelial cells at least partially independent of glucose (11,23,28). Therefore, the renoprotective effect of RAGE vaccination, independent of its modulation of blood glucose levels, may be at least partially due to its downregulation of VCAM-1 and ICAM-1 expression.

In summary, albuminuria, glomerular hypertrophy, and mesangial expansion are characteristic of early DKD. Our RAGE vaccine markedly suppressed urinary albumin excretion and attenuated these DKD-associated histologic lesions in mouse models of both type 1 (STZ-induced) and type 2 (db/db mice) diabetes. Thus, our RAGE vaccine represents a new therapeutic strategy to slow or halt DKD progression.

This article contains supplementary material online at https://doi.org/10.2337/figshare.14791647.

Acknowledgments. ELISA experiments were performed with the aid of Collaborative Research Resources, Keio University School of Medicine.

Funding. This work was supported by JSPS KAKENHI grant JP18K16006, Takeda Science Foundation, Novartis Research Grants, and the Japan Foundation for Applied Enzymes (to T.A.).

Duality of Interest. No potential conflicts of interest relevant to this article are reported.

Author Contributions. T.A. designed the study, performed experiments, analyzed data, and wrote the manuscript. T.N., K.H., A.H., N.Y., R.N., and H.I. reviewed the manuscript. All authors provided final approval for submission. T.A. is the guarantor of this work and, as such, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
World Health Organization
.
Fact sheets/Diabetes
.
2.
Afkarian
M
,
Zelnick
LR
,
Hall
YN
, et al
.
Clinical manifestations of kidney disease among US adults with diabetes, 1988-2014
.
JAMA
2016
;
316
:
602
610
3.
Zelnick
LR
,
Weiss
NS
,
Kestenbaum
BR
, et al
.
Diabetes and CKD in the United States population, 2009-2014
.
Clin J Am Soc Nephrol
2017
;
12
:
1984
1990
4.
Tuttle
KR
,
Bakris
GL
,
Bilous
RW
, et al
.
Diabetic kidney disease: a report from an ADA Consensus Conference
.
Diabetes Care
2014
;
37
:
2864
2883
5.
Afkarian
M
,
Katz
R
,
Bansal
N
, et al
.
Diabetes, kidney disease, and cardiovascular outcomes in the Jackson Heart Study
.
Clin J Am Soc Nephrol
2016
;
11
:
1384
1391
6.
Fox
CS
,
Matsushita
K
,
Woodward
M
, et al.;
Chronic Kidney Disease Prognosis Consortium
.
Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without diabetes: a meta-analysis
.
Lancet
2012
;
380
:
1662
1673
7.
Afkarian
M
,
Sachs
MC
,
Kestenbaum
B
, et al
.
Kidney disease and increased mortality risk in type 2 diabetes
.
J Am Soc Nephrol
2013
;
24
:
302
308
8.
Tanji
N
,
Markowitz
GS
,
Fu
C
, et al
.
Expression of advanced glycation end products and their cellular receptor RAGE in diabetic nephropathy and nondiabetic renal disease
.
J Am Soc Nephrol
2000
;
11
:
1656
1666
9.
Susztak
K
,
Raff
AC
,
Schiffer
M
,
Böttinger
EP
.
Glucose-induced reactive oxygen species cause apoptosis of podocytes and podocyte depletion at the onset of diabetic nephropathy
.
Diabetes
2006
;
55
:
225
233
10.
Menne
J
,
Shushakova
N
,
Bartels
J
, et al
.
Dual inhibition of classical protein kinase C-α and protein kinase C-β isoforms protects against experimental murine diabetic nephropathy
.
Diabetes
2013
;
62
:
1167
1174
11.
Schmidt
AM
,
Vianna
M
,
Gerlach
M
, et al
.
Isolation and characterization of two binding proteins for advanced glycosylation end products from bovine lung which are present on the endothelial cell surface
.
J Biol Chem
1992
;
267
:
14987
14997
12.
Degenhardt
TP
,
Alderson
NL
,
Arrington
DD
, et al
.
Pyridoxamine inhibits early renal disease and dyslipidemia in the streptozotocin-diabetic rat
.
Kidney Int
2002
;
61
:
939
950
13.
Figarola
JL
,
Loera
S
,
Weng
Y
,
Shanmugam
N
,
Natarajan
R
,
Rahbar
S
.
LR-90 prevents dyslipidaemia and diabetic nephropathy in the Zucker diabetic fatty rat
.
Diabetologia
2008
;
51
:
882
891
14.
Hou
B
,
Qiang
G
,
Zhao
Y
, et al
.
Salvianolic acid A protects against diabetic nephropathy through ameliorating glomerular endothelial dysfunction via inhibiting AGE-RAGE signaling
.
Cell Physiol Biochem
2017
;
44
:
2378
2394
15.
Watson
AM
,
Gray
SP
,
Jiaze
L
, et al
.
Alagebrium reduces glomerular fibrogenesis and inflammation beyond preventing RAGE activation in diabetic apolipoprotein E knockout mice
.
Diabetes
2012
;
61
:
2105
2113
16.
Flyvbjerg
A
,
Denner
L
,
Schrijvers
BF
, et al
.
Long-term renal effects of a neutralizing RAGE antibody in obese type 2 diabetic mice
.
Diabetes
2004
;
53
:
166
172
17.
Matsui
T
,
Higashimoto
Y
,
Nishino
Y
,
Nakamura
N
,
Fukami
K
,
Yamagishi
SI
.
RAGE-aptamer blocks the development and progression of experimental diabetic nephropathy
.
Diabetes
2017
;
66
:
1683
1695
18.
Williams
AF
,
Manias
E
,
Walker
R
.
Adherence to multiple, prescribed medications in diabetic kidney disease: a qualitative study of consumers’ and health professionals’ perspectives
.
Int J Nurs Stud
2008
;
45
:
1742
1756
19.
Chang
PY
,
Chien
LN
,
Lin
YF
,
Chiou
HY
,
Chiu
WT
.
Nonadherence of oral antihyperglycemic medication will increase risk of end-stage renal disease
.
Medicine [Baltimore]
2015
;
94
:
e2051
20.
Azegami
T
,
Yuki
Y
,
Sawada
S
, et al
.
Nanogel-based nasal ghrelin vaccine prevents obesity
.
Mucosal Immunol
2017
;
10
:
1351
1360
21.
Hirose
K
,
Osterby
R
,
Nozawa
M
,
Gundersen
HJ
.
Development of glomerular lesions in experimental long-term diabetes in the rat
.
Kidney Int
1982
;
21
:
689
695
22.
Steenhard
BM
,
Isom
K
,
Stroganova
L
, et al
.
Deletion of von Hippel-Lindau in glomerular podocytes results in glomerular basement membrane thickening, ectopic subepithelial deposition of collagen alpha1alpha2alpha1(IV), expression of neuroglobin, and proteinuria
.
Am J Pathol
2010
;
177
:
84
96
23.
Anisuzzaman
DM
,
Hatta
T
,
Miyoshi
T
, et al
.
Longistatin in tick saliva blocks advanced glycation end-product receptor activation
.
J Clin Invest
2014
;
124
:
4429
4444
24.
Xie
J
,
Reverdatto
S
,
Frolov
A
,
Hoffmann
R
,
Burz
DS
,
Shekhtman
A
.
Structural basis for pattern recognition by the receptor for advanced glycation end products (RAGE)
.
J Biol Chem
2008
;
283
:
27255
27269
25.
Matsumoto
S
,
Yoshida
T
,
Murata
H
, et al
.
Solution structure of the variable-type domain of the receptor for advanced glycation end products: new insight into AGE-RAGE interaction
.
Biochemistry
2008
;
47
:
12299
12311
26.
Park
H
,
Adsit
FG
,
Boyington
JC
.
The 1.5 Å crystal structure of human receptor for advanced glycation endproducts (RAGE) ectodomains reveals unique features determining ligand binding [published correction appears in J Biol Chem. 2011;286:19178]
.
J Biol Chem
2010
;
285
:
40762
40770
27.
Hopp
TP
,
Woods
KR
.
A computer program for predicting protein antigenic determinants
.
Mol Immunol
1983
;
20
:
483
489
28.
Basta
G
,
Lazzerini
G
,
Massaro
M
, et al
.
Advanced glycation end products activate endothelium through signal-transduction receptor RAGE: a mechanism for amplification of inflammatory responses
.
Circulation
2002
;
105
:
816
822
29.
Alicic
RZ
,
Rooney
MT
,
Tuttle
KR
.
Diabetic kidney disease: challenges, progress, and possibilities
.
Clin J Am Soc Nephrol
2017
;
12
:
2032
2045
30.
Myint
KM
,
Yamamoto
Y
,
Doi
T
, et al
.
RAGE control of diabetic nephropathy in a mouse model: effects of RAGE gene disruption and administration of low-molecular weight heparin
.
Diabetes
2006
;
55
:
2510
2522
31.
Hagiwara
S
,
Sourris
K
,
Ziemann
M
, et al
.
RAGE deletion confers renoprotection by reducing responsiveness to transforming growth factor-β and increasing resistance to apoptosis
.
Diabetes
2018
;
67
:
960
973
32.
Reiniger
N
,
Lau
K
,
McCalla
D
, et al
.
Deletion of the receptor for advanced glycation end products reduces glomerulosclerosis and preserves renal function in the diabetic OVE26 mouse
.
Diabetes
2010
;
59
:
2043
2054
33.
Azegami
T
,
Itoh
H
.
Vaccine development against the renin-angiotensin system for the treatment of hypertension
.
Int J Hypertens
2019
;
2019
:
9218531
34.
Yatime
L
,
Andersen
GR
.
Structural insights into the oligomerization mode of the human receptor for advanced glycation end-products
.
FEBS J
2013
;
280
:
6556
6568
35.
Breyer
MD
,
Böttinger
E
,
Brosius
FC
 3rd
, et al.;
AMDCC
.
Mouse models of diabetic nephropathy
.
J Am Soc Nephrol
2005
;
16
:
27
45
36.
Azushima
K
,
Gurley
SB
,
Coffman
TM
.
Modelling diabetic nephropathy in mice
.
Nat Rev Nephrol
2018
;
14
:
48
56
37.
Katoh
M
,
Ohmachi
Y
,
Kurosawa
Y
,
Yoneda
H
,
Tanaka
N
,
Narita
H
.
Effects of imidapril and captopril on streptozotocin-induced diabetic nephropathy in mice
.
Eur J Pharmacol
2000
;
398
:
381
387
38.
Zhang
Z
,
Zhang
Y
,
Ning
G
,
Deb
DK
,
Kong
J
,
Li
YC
.
Combination therapy with AT1 blocker and vitamin D analog markedly ameliorates diabetic nephropathy: blockade of compensatory renin increase
.
Proc Natl Acad Sci USA
2008
;
105
:
15896
15901
39.
Lee
YH
,
Kim
SH
,
Kang
JM
, et al
.
Empagliflozin attenuates diabetic tubulopathy by improving mitochondrial fragmentation and autophagy
.
Am J Physiol Renal Physiol
2019
;
317
:
F767
F780
40.
Kamezaki
M
,
Kusaba
T
,
Komaki
K
, et al
.
Comprehensive renoprotective effects of ipragliflozin on early diabetic nephropathy in mice
.
Sci Rep
2018
;
8
:
4029
41.
Tan
SM
,
Sharma
A
,
Stefanovic
N
, et al
.
Derivative of bardoxolone methyl, dh404, in an inverse dose-dependent manner lessens diabetes-associated atherosclerosis and improves diabetic kidney disease
.
Diabetes
2014
;
63
:
3091
3103
42.
Huang
Z
,
Mou
Y
,
Xu
X
, et al
.
Novel derivative of bardoxolone methyl improves safety for the treatment of diabetic nephropathy
.
J Med Chem
2017
;
60
:
8847
8857
43.
Heidland
A
,
Sebekova
K
,
Schinzel
R
.
Advanced glycation end products and the progressive course of renal disease
.
Am J Kidney Dis
2001
;
38
(
Suppl. 1
):
S100
S106
44.
Okada
S
,
Shikata
K
,
Matsuda
M
, et al
.
Intercellular adhesion molecule-1-deficient mice are resistant against renal injury after induction of diabetes
.
Diabetes
2003
;
52
:
2586
2593
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/content/license.