Global Renal Gene Expression Profiling Analysis in B2-Kinin Receptor Null Mice: Impact of Diabetes

Diabetic nephropathy (DN), the leading cause of end-stage renal failure, is clinically manifested by albuminuria and a progressive decline in glomerular filtration rate. The risk factors and mechanisms that contribute to the development and progression of DN are still incompletely defined. To address the involvement of bradykinin B2-receptors (B2R) in DN, we used a genome wide approach to study the effects of diabetes on differential renal gene expression profile in wild type and B2R knockout (B2R−/−) mice. Diabetes was induced with streptozotocin and plasma glucose levels and albumin excretion rate (AER) were measured at predetermined times throughout the 23 week study period. Longitudinal analysis of AER indicated that diabetic B2R−/−D null mice had a significantly decreased AER levels compared to wild type B2R+/+D mice (P = 0.0005). Results from the global microarray study comparing gene expression profiles among four groups of mice respectively: (B2R+/+C, B2R+/+D, B2R−/−C and B2R−/−D) highlighted the role of several altered pathological pathways in response to disruption of B2R and to the diabetic state that included: endothelial injury, oxidative stress, insulin and lipid metabolism and inflammatory process with a marked alteration in the pro-apoptotic genes. The findings of the present study provide a global genomics view of biomarkers that highlight the mechanisms and putative pathways involved in DN.


Introduction
Diabetic nephropathy (DN) is a major health epidemic and is the main cause of morbidity and mortality in diabetes. It is the single most common cause of end-stage renal failure [1,2]. A very characteristic and initial event of the development of DN is glomerulosclerosis, which is featured by increased thickness of the glomerular basement membrane, and widening of the mesangium with accumulation of extracellular matrix (ECM). Furthermore, the degree of mesangial expansion is strongly related to the clinical manifestations of diabetic nephropathy, such as albuminuria and decreased glomerular filtration rate [3,4]. Even though inherent susceptibility seems to influence the rate at which glomerular injury develops, hyperglycemia seems to be the primary driving force for cellular damage [5]. In this regard, intensive control of glycemia in type I diabetic patients was associated with a significant reduction in the development and progression of nephropathy [6].
Although, the underlying biochemical and cellular mechanisms that promote renal injury in diabetes are still undefined, accumulating evidence supports a relationship between the activity of the kallikrein-kinin system (KKS) and renal impairment. It has been shown that type I diabetic patients with hyperfiltration as well as diabetic rats with increased glomerular filtration rate (GFR) and renal plasma flow (RPF) are associated with increased active kallikrein excretion rate [7,8]. In addition, treatment of hyperfiltering diabetic rats with aprotinin, a kallikrein inhibitor, or with a B 2 -kinin receptor (B 2 R) antagonist, increases the renal vascular resistance and reduces GFR and RPF [9]. Furthermore, previous findings from our lab have shown that increased plasma prekallikrein activity is associated with increased albumin excretion rate; these data have been demonstrated in DCCT/EDICcohort of type 1 diabetic patients [10].
While most of the physiological actions of the KKS are attributed to the generation of BK and activation of B 2 R, the intracellular signaling pathways initiated upon activation of B 2 R leading to expression of prosclerotic factors that ultimately result in glomerular injury are just beginning to be defined. Activation of B 2 R by BK results in marked induction of connective tissue growth factor (CTGF), collagen I and transforming growth factor-b type II receptor (TGF-ßRII) in mesangial cells. Inhibition of B 2 R by Icatibant significantly reduced the increase in collagen I and CTGF mRNA levels in response to BK challenge [11]. Of interest, it has been shown that the glomerular expression of B 2 Rs are increased in diabetes and a targeted deletion B 2 R protects against the development of DN [12,13]. Furthermore, diabetic B 2 R 2/2 null mice display reduced albumin excretion rate (AER), as well as reduced glomerular and tubular injury compared to diabetic B 2 R +/+ mice [13].
In this study, we employed a global microarray analysis coupled with systems biology study to investigate the differential gene expression in wild type control (B 2 R +/+ ) and diabetic (B 2 R +/+ D) mice as well as in B 2 R knockout-control (B 2 R 2/2 ) mice and in B 2 R knockout-diabetic (B 2 R 2/2 D) mice in order to identify candidate genes that may be involved in the development of diabetic nephropathy. The objective of our study was to determine 1) whether deletion of B 2-receptors will result in alteration in specific gene expression profiles whose specific functions can shed light on the role(s) of B 2-receptors, and 2) whether diabetes will result in differences in the patterns of gene expression and pathways between B 2 R +/+ D and B 2 R 2/2 D mice that can be linked to the pathological manifestation observed after the induction of DN.

Study Design
To address the contribution of B 2 R to the development of diabetic nephropathy, we studied B 2 R knockout mice (B 2 R 2/2 ) and their wild type littermates (B 2 R +/+ ). Male B 2 R 2/2 mice (strain # B6 129S-BdKrb2, Jackson Laboratories, Bar Harbor, ME) and B 2 R +/+ mice (strain # B6 129 SF2/J, Jackson Laboratories, Bar Harbor, ME) weighing 20-30 g were used in our studies. Mice were housed three per cage in a light and temperature controlled room and had free access to food and water. Diabetes was induced by daily intraperitoneal injection of streptozotocin (50mg/kg body weight) for 3-5 days. Diabetes was confirmed in STZ-treated mice by tail vein plasma glucose levels. We used a total of 12 mice for this study divided into 4 groups, 3 mice in each group. Group 1, wild type non-diabetic-controls (B 2 R +/+ C); group 2, wild typediabetic (B 2 R +/+ D); group 3, B 2 R knockout-control (B 2 R 2/2 C) and group 4, B 2 R knockout-diabetic (B 2 R 2/2 D). Glucose levels and body weights were measured at predetermined intervals to characterize the diabetic state and to ensure adequate metabolic control. Every week mice were placed in metabolic cages (Nalgene) for 24 h to acclimate, and then 24h urine collections were obtained from all mice to measure albumin excretion rate. The mice were sacrificed 6 months after the induction of diabetes. The studies were done in line with the Guide for the Care and Use of laboratory Animals published by the National Institutes of Health (NIH Publication No 85-23, revised 1996). The study was approved by the Institutional Animal Care and Use Committee at the Medical University of South Carolina.

RNA Extraction
Kidneys from control and diabetic mice (B 2 R +/+ C, B 2 R +/+ D, B 2 R 2/2 C and B 2 R 2/2 D mice, n = 3 per group) were removed under anesthesia and cortexes were cut off to extract RNA. For RNA extraction and purification, a method combined Trizol (Cat. No.15596-018, Invitrogen Life Technologies) and RNeasy Midi Kit (Cat. No.75144, QIAGEN) for total RNA isolation from animal tissue was used. Briefly, the cortexes were homogenized using an appropriate volume of Trizol (1ml of Trizol/100 mg tissue). Then chloroform (0.2 ml/1 ml Trizol used) was added to separate the aqueous phase from protein phase. Total RNA was dissolved in the aqueous phase. RNA purification followed the protocol of RNeasy kit handbook. The RNA concentration was determined in a spectrophotometer (ultraspec III, Pharmacia) by absorbance at 260 nm. The ratio of A260 to A280 was calculated to check the purification of RNA, and the rRNA ratio of 28S/18S using 2100 Bioanalyzer (Aglilent) was measured to check the quality of RNA.
Synthesis of Double-stranded cDNA from Total RNA Total RNA (10 mg) from each sample was used to synthesize ds-cDNA. In primer hybridization, 10 mg of RNA, T7-(dT) 24 primer (100 pmol/ul, HPLC purified) and DEPC-H 2 O were added to the tube and incubated at 70uC for 10 min. Next, 56 first strand cDNA buffer 4 ml, DTT (0.1 M) 2 ml dNTP (10 mM) were added to each tube, incubated at 42uC for 2 min. and followed by addition of SuperScrip II RT (200 U/ml) 2 ml and incubated at 42uC for 1 hour to synthesize the first strand of cDNA. The final volume for the first strand cDNA synthesis was 20 ml. In order to synthesis the second strand, the following reagents were added to the first strand synthesis tube: DEPC-treated water 91 ml, 56 second strand cDNA reaction buffer 30 ml, 10 mM dNTP mix 3 ml, 10 U/ml E.coli DNA ligase, 10 U/ml E.coli DNA polymerase I 4 ml and 2 U/ml E.coli RNase H. The final volume of the second strand reaction was 150 ml. The reaction tubes were incubated at 16uC for 2 hours in a cooling water bath. After the incubation, 2 ml of [10 U] T4 DNA polymerase was added to the reaction tube, incubated at 16uC for 5 min, followed by addition of 10 ml of 0.5 M EDTA to complete synthesis of the second cDNA strand.

Synthesis of Biotin-labeled cRNA
Before Synthesis of biotin-labeled cRNA, double-strand cDNA was cleaned according to the GeneChip Sample Cleanup Module. The following reagents were used in the final reaction volume (40 ml): 4 ml of 106HY reaction buffer, 4 ml of 106Biotin-labeled ribonucleotides, 4 ml of 106DTT, 4 ml 106RNase inhibitor mix, 2 ml 206T7 RNA polymerase and distilled water. All of the reagents were mixed and incubated at 37uC for 5 hours, with gentle mixing of the tube every 30 min. The biotin-labeled cRNA was cleaned according to the GeneChip Sample Cleanup Module before quantification.

cRNA Fragmentation and Microarray Procedure
To reach a final concentration of 1 mg/ml, 20 mg cRNA and 8 ml of 56fragmentation buffer were incubated at 94uC for 35 min. A total of 15 ml of each sample (1.0 mg/ml) was used for preparation of hybridization cocktail that was loaded onto the GeneChips (Mouse Expression Array 430 A, Affymetrix) and hybridized for 16 h at 45uC in the Affymetrix GeneChip hybridization oven 640. Following this, the chips were loaded into the Affymetrix GeneChip Fluidics Station 400 with double stain antibody amplification solution for washing and staining. Finally, the GeneChips were scanned using the Hewlett Packard GeneArray Scanner 2500.
Expression values were derived using RMA (for normalization and background subtraction) as executed by the software RMAexpress (University of California, Berkeley). Expressed genes were determined according to the following criterion: any gene for which a sample had an average detection p-value (MASS) .0.04 (standard threshold for MASS ''presence'' call); all other genes were excluded from further consideration. RMA expression values were converted from log-base 2 and imported into dchip. Dchip was used to perform comparisons for all desired group comparisons. Criteria for comparison were: Fold change of 1.8; 90% confidence bound of fold change was used; T-test with p-value ,0.05; false discovery rate was calculated as the median number genes discovered in 50 iterations of permutated samples.

Real-Time PCR
Total RNA (2 mg) was converted to cDNA using MLV Reverse Transcriptase (Promega, Madison, WI) according to the manufacturer's protocol at 37uC for 1 hr. To determine the validity of Figure 1. Plasma glucose levels (A) and body weights (B) in diabetic (B 2 R +/+ D and B 2 R 2/2 D) and control (B 2 R +/+ C and B 2 R 2/2 C) mice. (A) Plasma glucose levels were significantly increased two weeks after STZ injection in both diabetic groups (B 2 R +/+ D and B 2 R 2/2 D) compared to B 2 R +/+ C and B 2 R 2/2 C (P,0.001) and remained significantly elevated for the duration of the study. (B) Initial body weights were not significantly different between diabetic and control mice. However, B 2 R 2/2 D mice had significantly reduced bodyweight after 14 weeks and B 2 R +/+ D after 20 weeks compared with B 2 R +/+ C and B 2 R 2/2 C mice and this reduction in body weight was maintained for the duration of the study (P,0.001 vs. B 2 R +/ + C and B 2 R 2/2 C). doi:10.1371/journal.pone.0044714.g001 Figure 2. Albumin excretion rate (AER) in diabetic (B 2 R +/+ D and B 2 R 2/2 D) and control (B 2 R +/+ C and B 2 R 2/2 C) mice. AER was significantly higher in B 2 R +/+ D mice compared to B 2 R 2/2 D ({P,0.05) and to B 2 R +/+ C and or B 2 R 2/2 C (*P,0.001), as early as two weeks after induction of diabetes and remained elevated for the duration of the study period. doi:10.1371/journal.pone.0044714.g002 primers and appropriate Tm for Real Time PCR, the primers were first amplified in a PCR reaction to ensure that only one band is amplified. The following primers were designed so that all of the PCR products are within 75-150 bp (Integrated DNA Technologies Inc). b-actin: 59-actgccgctcctcttcctc-39; 59ccgctcgttgccaatagtga-39; Growth hormone receptor: 59ttctgggaagcctcgattcaccaa-39, 59:cagcttgtcgttggctttcccttt-39; Insulin growth factor binding protein-1(IGFBP1) 59: agatcgccgacctcaagaaatgga-39, 59-tgttgggctgcagctaatctctct-39; IGFBP4: 59-tcggaaatcgaagccatccaggaa-39, 59-tgaagctgttgttgggatgttcgc-39; Extracellular superoxide distmutase (EC-SOD) 59-tgcatgcaatctgcagggtacaac-39, 59-aagagaaccaagccggtgatctgt-39; Flavin containing monooxygenase 2 (FMO2) 59-caacgcactgtctttgacgctgtt-39, 59-atggaaa-tactggcttcggaacct-39; Glutathione-S-transferase a-2 (GSTa-2) 59atgacaaggactaccttgtgggca-39, 59-ggctggcatcaagctcttcaacat-39. For each target gene, a standard curve was established. This was achieved by performing a series of 3-fold dilutions of the gene of interest. Negative control was made using the same volume of Rnase-free water instead of sample. The master mix was prepared as follows: 26 SYBR Green Supermix (cat. No. 170-8880, BIO-RAD) 12.5 ml, forward and reverse primer 0.25 ml respectively and ddH2O 12 ml. For each well, 22 ml of master mix was loaded first, followed by 3 ml of sample, mixed well to get total reaction volume of 25 ml. For plate setup, SYBR-490 was chosen as fluorophore. The plate was covered with a sheet of optical sealing film. PCR conditions were 95uC for 3 min, followed by 40 cycles of 95uC for 10 sec, 58uC for 1 min for ß-actin and for all the other genes 60uC for 1 min, then 95uC for 1 min, 55uC for 1 min and 100 cycles of 55uC for 10 sec. All of the reactions were done in duplicate. The correlation coefficient is between 0.98-1, PCR efficiency is between 75-130%. The mRNA levels were expressed relative to ß-actin mRNA. Realtime PCR using iCycle TM iQ optical system software (version 3.0a) was used in our studies.

Urinary Albumin Excretion Rate
The urinary albumin excretion rate was measured with a murine microalbuminuria ELISA kit (Exocell Inc., PA) according to the manufacturer's suggestions.

Systems Biology Analysis
The microarray differential expression of the wild type B 2 R vs. knockout (B 2 R 2/2 ) in control and diabetic phenotypes was further analyzed using a systems biology approach to assess the altered pathway(s) relevant to differential B 2 R knockout (B 2 R 2/2 ) phenotype mice and its contribution to the development of Diabetes. PathwayStudio software (v 9.0; Ariadne Genomics, Rockville, MD, USA) was applied for the systems biology analysis. This software helps to interpret biological meaning from differential gene expression, build and analyze pathways, and identify altered cellular processes and molecular functions involved. PathwayStudio comes with a built-in resource named ResNet, which is a database of molecular interactions based on natural language processing of scientific abstracts in PubMed.
For gene ontology analysis including differential molecular function and biological processes involved, PANTHER software (Protein ANalysis THrough Evolutionary Relationships; http:// www.pantherdb.org/genes/batchIdSearch.jsp) was utilized to classify proteins into distinct categories of molecular functions and biological processes. Panther software uses published scientific experimental evidence and evolutionary relationships abstracted by curators with the goal of predicting function even in the absence of direct experimental evidence. Proteins are classified into families and subfamilies of shared function, which are then categorized using a highly controlled vocabulary (ontology terms) by biological process, molecular function and molecular pathway.

Power Analysis
Sample size calculation for our study was determined by using the formula by Hedeker D et al, for longitudinal data [14]. In this study we assumed 80% power, significance of 5%, repeated measure correlation of 0.5, 9 measurement time points, within subject variance of 4.2, and medium effect size of 0.3. This resulted in 2.3 mice per group, and accounting for possible attrition effect we inflated our sample size by 20% so the sample size in each group will be 2.76 mice.

Statistical Analysis
Results are expressed as mean 6 standard error, unless stated otherwise. All data were analyzed using SAS (SAS Institute Inc., Version 8, Cary, NC). t-tests were used to analyze continuous outcomes versus each covariate separately. To compare means values across three or more groups, ANOVA was used. Generalized linear models and generalized estimating equations were used to compare albumin excretion rates, plasma glucose levels and body weights within mice and across groups over time. A longitudinal data analysis was conducted to assess the effect of group on the AER levels over time. A mixed model was fit and spatial data covariance structure was used to accommodate for the unequally-spaced measurement time points. In this context, a continuous-time model was employed using variance-covariance matrix with type = sp (pow) in SAS PROC MIXED. Bonferroni correction was used to adjust for inflated type I error when making multiple comparisons. Statistical significance was determined using a two-sided test and significance was assumed for P-values #0.05.

Characteristics of the Diabetic State
Plasma glucose levels were markedly elevated 2 weeks after STZ injection in both B 2 R +/+ D and B 2 R 2/2 D groups of mice compared to their non-diabetic controls, and remained elevated throughout the study period ( Figure 1A). On average plasma glucose levels increased by 205 mg/dl in B 2 R 2/2 D null mice and by 251 mg/dl in B 2 R +/+ D null mice compared to B 2 R +/+ C mice, P,0.001. No significant difference in plasma glucose levels was observed between B 2 R +/+ C mice and B 2 R 2/2 C mice, P = 0.276. No significant time effect on plasma glucose level was observed, P = 0.2647. Also no significant effect of group by time interaction on plasma glucose levels was detected, P = 0.28. Hence, the observed difference in plasma glucose levels across groups was primarily due to group effect.
Initial body weights were not significantly different between diabetic and non-diabetic mice. However, B 2 R 2/2 D mice had significantly reduced bodyweight after 14 weeks and B 2 R +/+ D after 20 weeks compared with B 2 R +/+ C and B 2 R 2/2 C mice and this reduction in body weight was maintained for the duration of the study ( Figure 1B). Body weight analyses revealed that there was no significant group effect on bodyweights over time, but there was a significant effect of time on bodyweights, P = 0.0011. In addition, there was interaction between time and group effect on changes in body weights P = 0.0011. Thus, the decrease in bodyweights in B 2 R 2/2 D null mice and B 2 R +/+ D mice compared to B 2 R +/+ C mice are a result of time effect.

Albumin Excretion Rate
The albumin excretion rate results are presented in Figure 2. Groups were defined as B 2 R +/+ C, B 2 R +/+ D, B 2 R 2/2 C and B 2 R 2/2 D. AER was modeled with a time and group main effect and a time by group effect. Since AER in each mouse was measured up to 10 times over 23 weeks, a longitudinal data analysis was conducted to assess the effect of group on the AER levels over time. A mixed model was fit and spatial data covariance structure was used to accommodate for the unequally-spaced measurement time points. Our results showed that there was a significant overall group effect with P,0.0001. In particular, when the wild type control group B 2 R +/+ C was considered as the reference group, we observed that B 2 R 2/2 D had a significant increase in the AER by 13.5 mg/24 h, P = 0.001. Overall, a significant increase by about 28.5 mg/24 h in AER was also observed for B 2 R +/+ D mice compared to B 2 R +/+ C mice P,.0001. No significant differences in AER was observed between B2R +/+ C and B2R 2/2 C, P = 0.1629.
Our result also showed that the B 2 R 2/2 D null mice had a significant decrease of 14.97 mg/24 h in the AER levels compared to wild type B 2 R +/+ D mice, P = 0.0005. Some minor time effect on the AER was also observed. In particular, we can estimate that overall the AER appeared to be decreasing with time at a slow rate of 0.547 mg/24 h, P,.0001. An interaction test was then performed which showed that there is no significant interaction between time and group (P-value = 0.24). Although there was some minor effect of time on AER, the observed changes in AER across groups was mainly due to group effect rather than an effect of time.

Hierarchical Clustering of Gene Expression
Differential gene expression profiles in the kidney were identified among four groups of mice: B 2 R +/+ C, B 2 R +/+ D, B 2 R 2/2 C and B 2 R 2/2 D. Each column represents one sample, and the color bars represent the median value of three array experiments for an individual mouse for that gene (Figure 3).

Gene Regulation in Response to Disruption of B 2 R
Upon deletion of B 2 R, There were a total of 14 altered genes (4 upregulated and 9 downregulated shown in Table 1); these include genes that code for ATPase activity, hemoglobin and enzymes involved in protein metabolism. Among the altered genes, Monoglyceride lipase (MGLL; EC 3.1.1.23) and lysine (K)specific demethylase 2B (KDM2B) were found to be downregulated due to B 2 R deletion. KDM2B gene encodes a member of the F-box protein family lysine (K)-specific demethylase 2B which function in phosphorylation-dependent ubiquitination while MGLL gene functions together with hormone-sensitive lipase to

Gene Regulation in Response to Diabetes
Upon Diabetes induction, a total of 9 genes were found to be upregulated and 16 genes downregulated compared to B 2 R +/+ C wild type mice. An enriched pathways analysis identified genes associated with potassium transport, cell cycles and lipid metab-olism as shown in Table 2). The biological processes depicting genes that are altered in response to diabetes are shown in Figures 5A and B.
Of great interest, in B 2 R 2/2 null mice, a total of 181 genes were regulated by diabetes including 91 upregulated genes and 90 downregulated genes, respectively ( Table 3). A thorough systems biology analysis of specific enriched pathways, several genes were found to be associated with: endothelial cellular injury, insulin & lipid metabolism, oxidative stress, cardiac and kidney toxicity as illustrated in the biological processes ( Figures 6A & B).    In B 2 R 2/2 D vs. B 2 R +/+ D mice, a total of 43 genes were upregulated and 66 genes were downregulated ( Table 4). Among these altered genes: IGFBP, GST, EC-SOD and GHR genes. In a detailed assessment of these genes, gene expressions of IGFBP-1(3.65 fold) and GST (Yc2, 2.05 fold; omega1, 2.43 fold) were elevated in the B 2 KR 2/2 D mice compared to the B 2 KR +/+ D mice. On the other hand, gene expressions of Insulin-like growth factor-binding protein-4 (IGFBP-4) (22.18 fold), EC-SOD (21.95 fold), FMO2 (21.94 Fold) and GHR (22.7 fold) were suppressed in the B 2 KR 2/2 D mice compared to the B 2 KR +/+ D mice, P,0.05. The biological processes depicting genes that are altered in response to diabetes +/2 B 2 R are shown in Figure 7A & B.

Validation of Specific Gene Expressions by Quantitative Real-time PCR Superoxide Dismutase 3, Extracellular (EC-SOD)
EC-SOD gene encodes a member of the superoxide dismutase (SOD) protein family which are antioxidant enzymes that catalyze the dismutation of two superoxide radicals into hydrogen peroxide and oxygen protecting from oxidative stress. EC-SOD expression tended to be suppressed by diabetes in the wild type mice. Interestingly, in the B 2 R 2/2 D mice, EC-SOD expression was increased up to 37% compared to that in the B 2 R +/+ D mice (*P,0.05 vs. B 2 R +/+ D, Figure 8A). Glutathione S-transferase, Alpha 2(Yc2) (GST-Yc2) GST-Yc2 catalyze the conjugation of reduced glutathiones and a variety of electrophiles, including many known carcinogens and mutagens. Our data indicated that the expression of GST was significantly higher in B 2 R 2/2 D mice compared to B 2 R +/+ D mice (*P,0.05 vs. B 2 R +/+ D, Figure 8B).

Flavin Containing Monooxygenase 2 (FMO2)
FMO2 family is NADPH-dependent enzymes that catalyze the oxidation of many drugs and xenobiotics. In the B 2 R +/+ D mice, FMO2 expression was decreased up to 34% compared to that in the controls. However, the expression FMO2 was significantly higher in B 2 R 2/2 D mice compared with B 2 R +/+ D mice (*P,0.05 vs. B 2 R +/ + D, Figure 8C).   IGFBP-1 gene is a member of the insulin-like growth factor binding protein (IGFBP) family and encoding proteins with an IGFBP domain and a thyroglobulin type-I domain. It binds both insulin-like growth factors (IGFs) I and II and circulates in the plasma prolonging the half-life of the IGFs. In our work, the deletion of B 2 R didn't change the expression of IGFBP-1. However, IGFBP-1 expression was decreased up to 33% by  diabetes in the wild type mice (P,0.05). Interestingly, in B 2 R 2/2 D mice, IGFBP-1 expression was upregulated significantly: up to 2.7fold increase compared to that in B 2 R +/+ D (*P,0.05 vs. B 2 R +/+ D, Figure 8D). We next performed a targeted analysis to identify the involvement of these selected validated genes in the most highlighted altered pathways (apoptosis, oxidative stress and inflammation). These genes were shown to be highly related to the aforementioned pathways as shown in Figure 9.
Systems Biology Analysis of Altered Genes in B 2 R 2/2 D and B 2 R +/+ D mice Pathway Studio 9.0 (2011, Ariadne Genomics, Rockville, MD) was also used to search for potential altered cellular processes, and related pathways for associations with gene alterations in our diabetic mice in the presence or absence of B 2 R. The network was generated using the ''direct interaction'' algorithm with the filters of ''Cellular process and Protein'' as Entity Type while the Relation Type parameter was set to ''Regulation Analysis'' to map altered pathways regulated by the identified (downregulation vs. upregulation) subsets of genes. Several processes believed to be central to the pathogenesis of DN included oxidative stress mechanisms (ROS generation & oxidative stress), cardiac injury mechanisms along with pronounced inflammatory process with a marked alteration in the pro-apoptotic genes as illustrated in Figure 10. Discussion A pivotal event initiated by DN is glomerular injury, characterized by mesangial deposition and podocyte loss. The degree of podocyte loss and mesangial expansion are strongly correlated with the clinical manifestations of DN, such as albuminuria and decreased GFR [3,4,15]. Microalbuminuria, an early marker of DN, signifies high risk for progressive renal failure and cardiovascular disease [16]. Microalbuminuria has also been associated with increased cardiovascular mortality in diabetic and non-diabetic populations and with generalized and glomerular endothelial dysfunction [17]. Identifying biomarkers and risk factors that contribute to the development of microalbuminuria may provide insights into the mechanisms of diabetic renal injury.
Few interventions have been shown to slow the progression of renal disease in diabetic patients. These include intensive glycemic control, blood pressure control and treatment with angiotensin converting enzyme inhibitors (ACEI) or angiotensin receptor blockers (ARBs) [6,18]. Despite these interventions and beneficial effects, diabetic patients progress with time to develop end stage renal disease. It is of significance to note here that a recent Interventional study aimed at blockade of the renin-angiotensin system (RAS) with ACE-inhibitors or ARBs, in patients with type 1 diabetes, did not slow nephropathy progression [19]. However, the Using Pathway Studio 9.0, altered genes relevant to diabetic induction with or without disruption of B 2 R. were analyzed. In B 2 R 2/2 D vs. B 2 R +/+ D mice, a total of 109 genes were found to be altered (43 upregulated and 66 downregulated). The network was generated using ''direct interaction'' algorithm to map cellular processes and interactions among altered genes. Of interest, global Pathway analysis revealed association of these genes to oxidative stress mechanisms (ROS generation & oxidative stress), cardiac injury mechanisms along with pronounced inflammatory process with a marked alteration in the pro-apoptotic genes. The upregulated genes are shown in green and downregulated genes are in red. doi:10.1371/journal.pone.0044714.g010 exact factors responsible for these maladaptive signals leading to renal failure are poorly defined.
Metabolic imbalances associated with high tissue glucose and abnormal lipid levels in the diabetic state influence many pathways that contribute to the pathogenesis of DN [20,21]. The modifiable factors engaged in these processes are yet to be identified but there is evidence for promotion of chronic low-grade inflammation, oxidative stress, endothelial dysfunction, stimulation of proliferative/apoptotic pathways, and deposition of extracellular matrix [22][23][24]. Importantly, inflammatory mediators and growth factors are increasingly recognized as key players in the pathogenesis of DN [25][26][27].
Our published work has provided evidence for the involvement of the kallikrein-kinin system (KKS) in the initiation of DN [7,13]. In the current work, we performed longitudinal data analysis to assess the rate of change in AER levels over time among the 4 different groups. Our data indicated that targeted deletion of B 2 R in mice interferes with the progression of DN. Diabetic B 2 R 2/2 mice display reduced AER compared to diabetic B 2 R +/+ mice. Other investigators have also implicated a role for B 2 R in DN. Polymorphisms in the human B 2 R have been linked to increased albuminuria in diabetic patients and to the development of chronic renal failure [28,29]. In addition, blockade of B 2 R markedly reduced the proteinuria in STZ-diabetic mice and inhibition of B 2 R ameliorated the accelerated nephropathy in uninephrectomized db/db mice, lending support to the pathogenic role of B 2 R in DN [30,31].
Contrary to the aforesaid findings, Kakoki and Smithies have reported a protective role for B 2 R in DN. They have shown that the insulin Akita (Ins2 Akita ) mice crossed with null B 2 R (In2 Akita / B 2 R 2/2 ) or with double-null B 2 R and B 1 R (In2 Akita /B 2 R 2/2/ B 1 R 2/2 ) displayed increased albuminuria compared to Ins2 Akita mice alone [32,33].Other factors contributing to these apparent differences in the role of B 2 R in DN may be attributed to differences in the model of DN studied, genetic background of the animal models studied, severity and metabolic control of the diabetic state, specifics of the experimental design, the end points measured. It is noteworthy to point here that a confounding factor to be considered when using the Insulin Akita mouse is the propensity for these mice to develop mesangial deposits of IgG [34].
To investigate the underlying mechanisms and involved pathways linking the role of B 2 R genotype to the development/ progression of DN, we examined the contribution of B 2 R genotype on the global genomics level. We performed a global microarray study comparing gene expression profiles among four groups of mice respectively: (B 2 R +/+ C, B 2 R +/+ D, B 2 R 2/2 C and B 2 R 2/2 D). Findings from this work highlighted the role of several altered pathological pathways involved in the development of diabetes in the B 2 R 2/2 D vs. B 2 R 2/2 C mice which included: endothelial injury, oxidative stress, and insulin and lipid metabolism.
A detailed analysis of the top scoring biological processes data [Panther Analysis] reflected the central role of B 2 R to increased immune response/inflammation along with other cellular functions (transport, systems process and response to stimulus which can be linked to protective/compensatory mechanism. This is in accordance with a previous study by Bascands et al, in which a global microarray renal gene expression changes were examined in lipopolysacharide-treated wild-type and kinin B 1 receptorknockout mice to investigate underlying mechanisms of renal inflammation reflected the role of acute phase response and inflammatory process [35]. This is in contrast to the sole effect of diabetes induction in wild type mice which reflected more pronounced metabolic/cellular processes changes (metabolites precursor generation, cellular adhesion, and cellular communication) rather than inflammatory immune response mediated response. Of interest, is the upregulation of one of the genes, aquaporin 4, (AQP4, 2.24) due to diabetes. AQP4 functions as a water transport channel in the kidney and has been shown to be downregulated in mice lacking B 2 R [36].
These results validate existing published literature linking renal inflammation to early events of renal disease [37][38][39]. Furthermore, a global systems biology analysis among the diabetic mice with or without disruption of B 2 R (B 2 R 2/2 D vs. B 2 R +/+ D) illustrated the role of oxidative stress mechanisms (ROS generation & oxidative stress), along with inflammatory process with a marked alteration in the pro-apoptotic genes. Indeed, these results may reflect a pathologic exacerbative role of B 2 R in inducing cellular vascular injury mediated via apoptotic pathways in the presence of diabetes. These findings are in concert with other microarray studies involving B 1 and B 2 receptor knockout mice [40,41].
Taken together, the finding of this study investigates the contributing role of B 2 -receptors in either exacerbating or at least enhancing the occurrence of diabetic nephropathy. In conclusion, the present study investigates the impact B 2 R deletion on the development of DN. A critical analysis of the data hints that renal function is preserved in the B 2 R 2/2 D mice especially at the early stages of DN, compared to that of B 2 R +/+ D mice; these data were substantiated by the genomics/systems biology analysis. To the best of knowledge, this represents the first study that utilizes wide scale genomic/systems biology analysis in B 2 R 2/2 D mice. Finally, several of the identified genes (EC-sod, GST, IGFBP1 and FMO) were validated with RT-PCR to confirm gene alteration. Further studies including immunohistological analysis and assessment of protein levels and the activities of the antioxidants identified are certainly necessary to further evaluate the contributing role of the disruption of the B 2 -receptors.