Type 1 diabetes promotes disruption of advanced atherosclerotic lesions in LDL receptor-deficient mice

Johansson et al. 10.1073/pnas.0709958105.

Supporting Information

Files in this Data Supplement:

SI Table 1
SI Table 2
SI Figure 6
SI Figure 7
SI Table 3
SI Table 4
SI Table 5
SI Table 6
SI Figure 8
SI Materials and Methods
SI Figure 9
SI Table 7

Table 1. Diabetes does not significantly stimulate necrotic core formation, smooth muscle accumulation, or other features of advanced plaques in the BCA of LDLR^-/-;GP mice fed a low-fat diet

Unless otherwise indicated, results represent frequencies of lesion characteristics, expressed as the % of BCA cross-sections positive for a given feature. There were no significant differences between nondiabetic and diabetic mice (Mann-Whitney test). ND, not determined; SM, smooth muscle.

Table 2. Body weights, endogenous and exogenous (total) plasma insulin and SAA levels in low-fat fed and medium-fat fed nondiabetic and diabetic mice

SI Figure 6

Fig. 6. Diabetes does not increase the Sudan IV-positive aortic lesion area in LDLR^-/-;GP mice with preexisting lesions. LDLR^-/-;GP mice were fed a high-fat diet for 16 weeks (a; baseline), and then switched to a low-fat diet (a) or a medium-fat diet (b) for an additional 14 weeks. Diabetes (D) was induced by injecting LCMV at week 18, whereas nondiabetic (ND) littermates were injected with saline. At the same time, some mice were injected with HD-Ad-mVLDLR as a tool to aggressively lower plasma lipid levels, HD-Ad-0, or saline. At the end of the 30-week study, total aortic lesion area was measured on en face preparations stained with Sudan IV. Percentage of the aorta covered by lesions was quantified by National Institutes of Health software Image J. Statistical analysis was performed by using one-way ANOVA followed by Tukey's post hoc test. ^¶, P < 0.01; *, P < 0.05.

SI Figure 7

Fig. 7. HD-Ad-mVLDLR treatment results in long-term expression of the VLDL receptor in livers of nondiabetic and diabetic mice. Expression of VLDLR at the end of the 12-week study period was analyzed in hepatic membrane preparations from mice injected with HD-Ad-mVLDLR or controls, as described in ref. 4. (Left) VLDLR was detected in membrane samples (60 mg per lane) by using 2 mg/ml of an anti-VLDLR rabbit polyclonal antibody. (Right) The membrane was stained with Ponceau S (Sigma) as a loading control. Samples from representative mice are shown. Molecular mass markers (kDa) are indicated on the left.

Table 3. LCMV does not affect plasma parameters or lesion characteristics in nondiabetic LDLR^-/-;GP mice fed a medium-fat diet

Unless otherwise stated, results represent frequencies of lesion characteristics, expressed as the % of BCA cross-sections positive for a given feature. No significant differences were found (Student's t test or Mann-Whitney test).

Table 4. HD-Ad-0 does not affect plasma parameters or lesion characteristics in nondiabetic or diabetic LDLR-/-;GP mice fed a medium-fat diet

Unless otherwise stated, results represent frequencies of lesion characteristics, expressed as the % of BCA cross-sections positive for a given feature. The P values represent comparisons between HD-Ad-0-treated nondiabetic and diabetic mice (Student's t test). No significant differences were found (Mann-Whitney test) between HD-Ad-0-treated mice compared with saline-treated mice. ND, nondiabetic; D, diabetic

Table 5. Aggressive lipid lowering, achieved by hepatic overexpression of the VLDL receptor reduces lateral macrophage accumulation in medium-fat-fed diabetic mice, but does not significantly affect other characteristics of advanced lesions

Results represent frequencies of lesion characteristics, expressed as the % of BCA cross-sections positive for a given feature. *, P<0.05 compared with diabetic mice (Mann-Whitney test).

Table 6. Blood glucose, cholesterol, body weights, and number of elicited peritoneal cells in LDLR^-/-;GP mice analyzed for S100A9 in thioglycollate-elicited peritoneal monocytes/macrophages

LDLR^-/-;GP mice were fed the low-fat diet for 4 weeks, after induction of diabetes. Results are shown as means ± SEM. Statistical analysis was performed by unpaired Student's t test.

SI Figure 8

Fig. 8. Mass spectrometric (MS) identification S100A9 secreted from macrophages from diabetic mice, but not from nondiabetic mice. Diabetes was induced by injecting LCMV into three mice; three littermate controls received saline injections. During the next 4 weeks, the mice were fed the low-fat diet. The mice were then injected i.p. with thioglycollate, and peritoneal cells were isolated, adhesion-purified for 1 h under serum-free conditions, and then incubated for an additional 6 h in the presence of serum-free RPMI. (a) S100A9 spectral counts identified in three MS experiments on conditioned media from diabetic and nondiabetic littermates. Granulin spectral counts are shown as control. Representative MS traces showing the absence of an S100A9 peak in a sample of conditioned medium from a nondiabetic mouse (b) and the presence of a S100A9 peak in conditioned medium from a diabetic mouse (c). cps, counts/second; m/z, mass-to-charge ratio

SI Materials and Methods

Generation of HD-Ad-mVLDLR and HD-Ad-0 Vectors. Helper-dependent adenoviral vectors with all viral coding sequences deleted were generated as described in ref. 1. HD-Ad vectors do not express any viral protein genes and have been shown to circumvent the problem of hepatitis in mice. An expression cassette devoid of the mVLDLR sequence, HD-Ad-0, was used as a negative control.

Measurements of Blood Glucose, Lipids, SAA, Insulin, and sICAM-1. Blood glucose, total cholesterol, plasma triglycerides, and lipoprotein profiles were analyzed as described in ref. 2. The method for measurements of plasma SAA levels has also been described (3). Plasma levels of soluble intercellular adhesion molecule-1 (sICAM-1) were measured as described in ref. 4 by immunoassays (Quantikine; R&D Systems). Plasma samples were diluted 1:50 before analysis according to the manufacturer's instructions.

Evaluation of Insulin Tolerance. Insulin tolerance tests were performed after 5-7 h fasting on a subgroup of mice fed the medium-fat diet at week 29 of the study. Mice received a s.c. injection 1 unit/kg body weight of human regular insulin (Eli Lilly). Saphenous vein blood glucose levels were quantified by using the Precision Q.I.D Complete Blood Glucose Monitoring System (MediSense) before insulin injection, and at 15, 30, 60 and 90 min after insulin injection.

Tissue Preparation and Histochemistry. The mice were perfused with 4% paraformaldehyde at physiological pressure (2). The entire BCA was embedded in paraffin and serially sectioned (5-mm sections). Every 20 mm, two sections were stained by using a modified Movat's pentachrome staining protocol (2, 4). Lesion morphology was scored in a masked fashion and verified with immunohistological and histological methods (4). All parameters were recorded as binary outcomes. The frequency of each parameter was determined for serial sections representing each 20 mm along the full length of each BCA (on average ~160 sections per mouse). Thus, a frequency of 20%, for example, indicates that an average of 20% of all BCA sections from an individual mouse were positive for a certain morphological feature. These frequencies were averaged across animals in each group. Medial elastin breaks were scored positive if there was degradation of any of the lamellae of the medial elastin underlying the lesion. The mean number of elastin breaks per cross-section was calculated for each mouse. We used a method similar to that described by Gough et al. (5) to evaluate fibrous cap breaks. An animal was scored positive for fibrous cap break when at least four sections over a continuous length of 80 mm showed a break in the cap that was also associated with intraplaque hemorrhage in the same area.

The aorta was used for en face evaluation of atherosclerotic lesion area in a masked fashion, as described in ref. 2.

Quantification of Plaque Mac-2-positive Cells, a-Actin-Positive Smooth Muscle Cells, Erythrocytes, Ly-6C-positive Cells, and S100A9-Positive Cells. Lesion macrophages, smooth muscle cells, and the presence of erythrocytes were quantified immunohistochemically at the largest lesion site in the BCA. Mac-2 was used as a marker for macrophages, smooth muscle a-actin as a marker for a-actin-positive smooth muscle cells, and TER-119 as a marker for erythrocyte membranes, as described in ref. 4. S100A9 immunoreactivity was evaluated by using a goat anti-mouse S100A9 antibody (R&D Systems) at a final concentration of 0.1 mg/ml. A matched goat IgG (Zymed Laboratories) was used as negative control. The intensity of S100A9 immunoreactivity in Mac-2-positive cells was scored in a masked fashion by three investigators. Ly-6C immunoreactivity was visualized by a biotinylated rat monoclonal anti-Ly-6C antibody (BMA Biomedicals) at a final concentration of 0.2 mg/ml.

Analysis of Gene Expression, S100A9 Protein Expression, and MMP-9 Activity in Bone Marrow-Derived Monocytes/Macrophages. Bone marrow was harvested from femurs of adult C57BL/6 mice. The cells were plated in the presence of 10% FBS and 15% L-conditioned medium, as a source of M-CSF, and harvested at indicated time points. Semiquantitative reverse transcription PCR was used to evaluate mRNA levels of S100A9, Ly-6C, Mac-2 (galectin-3), CD11b, CD115, and b-actin (SI Table 7). MMP-9 activity was analyzed in conditioned media and cell lysates from bone marrow-derived mnoncytes/macrophages by substrate zymography, according to Gough et al. (5).

To evaluate the possible mechanism(s) whereby diabetes might alter monocytic cell expression of S100A9 and MMP-9, bone marrow was harvested from C57BL/6 mice and allowed to differentiate to monocytic cells in the presence of 15% L-conditioned medium, 1.5% FBS, and pathologically relevant concentrations of elevated glucose (25 mmol/liter), oleic acid (225 mmol/liter bound to fatty acid-free BSA at a 3:1 molar ratio), or palmitic acid at the same molar ratio. Oleic acid was selected because it is generally the most abundant fatty acid free in circulation, and in circulating triglycerides.

Analysis of Insulin Receptor b-Subunit Expression in Macrophages and Liver. Insulin receptor b-subunit expression was analyzed in liver and macrophages from both low-fat and high-fat fed nondiabetic and diabetic mice. Liver samples were obtained from the atherosclerosis study, whereas thioglycollate-elicited macrophages were harvested 4 weeks after induction of diabetes. Liver lysates were prepared by using a Dounce homogenizer, and macrophage lysates were prepared by sonication. Lysates were subjected to separation on 10% SDS polyacrylamide gels (40 mg per lane liver lysate, 50 mg per lane macrophage lysate). The proteins were transferred to Immobilon-P membranes (Millipore). Membranes were blocked with 5% milk in TBST for 1 h at room temperature and then incubated with an anti-insulin receptor b-subunit monoclonal antibody (Assay Design) at a concentration of 1 mg/ml for 2 h at room temperature. This was followed by an incubation with the secondary horseradish peroxidase-conjugate anti-mouse antibody (Amersham Pharmacia Biosciences) at a 1:5,000 dilution. Western blots were developed according to the manufacture's instructions (Millipore).

Identification of S100A9 Secreted from Monocytes/Macrophages by Liquid-Chromatography-Tandem Mass-Spectrometry (MS). In a subgroup of mice, we assessed the effects of diabetes on proteins secreted and/or shed from peritoneal monocytes/macrophages. Diabetes was induced by injecting LCMV into three mice; three nondiabetic littermates were injected with saline. During the next 4 weeks, the mice were fed the low-fat diet. Blood glucose and cholesterol levels in the groups are shown in SI Table 6. At the end of the 4-week study, the mice were injected i.p. with thioglycollate, and peritoneal cells were isolated 5 days later. There was no difference in the number of cells obtained from nondiabetic and diabetic mice (SI Table 6). Monocytes/macrophages were adhesion-purified for 1 h under serum-free conditions, and then incubated for an additional 6 h in the presence of serum-free RPMI. The conditioned media were collected for proteomics analysis of secreted or shed proteins.

The proteins in conditioned media were precipitated overnight at 4°C by using 20% TCA with 0.2% sodium deoxycholate as coprecipitant. Precipitated proteins were pelleted by centrifugation and washed twice with ice-cold acetone. Precipitates were reconstituted in 0.2% RapiGest (Waters Corp.) in 50 mmol/liter Tris buffer, pH 8.0, and protein concentrations were determined by the Bradford assay. An aliquot of 10 mg of protein was reduced with DTT, alkylated with iodoacetamide, and digested at 37°C with two aliquots of trypsin (1:50; wt/wt each) for 1 h and overnight, respectively.

Mass spectrometric analyses were performed by using a capillary HPLC (MS4B; Michrom Bioresources) coupled to an electrospray-linear ion trap mass spectrometer (LTQ, Thermo Electron Corp.). A tryptic digest (2 mg of protein) was injected onto a trap column (Paradigm Platinum Peptide Nanotrap; Michrom Bioresources), desalted for 5 min with 5% acetonitrile, 0.1% formic acid (50 ml/min), and then eluted onto an analytical column (0.150 ´ 150 mm, Magic C18AQ 5 mm; Michrom Bioresources), and separated by reversed phase linear gradient 5-37% B over 180 min (solvent A, 5% acetonitrile, 0.1% formic acid; solvent B, 90% acetonitrile, 0.1% formic acid) at a flow rate of 1 ml/min. Tandem mass spectra were acquired by using date-dependent acquisition with one survey full-scan mass spectrum followed by eight MS/MS scans on the eight most abundant peaks from the survey scan. The spectra were then searched against mouse IPI database, version 3.16, April 4, 2006 (6), using the Sequest search engine (parameters: no enzyme restriction, fixed Cys alkylation, and variable Met oxidation modifications). Peptide matches were further validated by using PeptideProphet (7) with a peptide probability ³0.9 and at least one trypsin cleavage terminus (estimated sensitivity 0.83 and error 0.008). The validated peptides were then assembled into proteins by using ProteinProphet (8) at protein probability ³0.95 (estimated sensitivity 0.83 and error 0.006). Only proteins passing these criteria and with at least two unique peptides in one of the samples were further considered. Spectral counting (9, 10) was used to evaluate protein relative abundance differences between the sample groups. Statistical comparisons of MS spectral counts were done according to Bolshev (11).

Supporting Results and Discussion

Diabetic Mice Are Not Markedly Insulin-Resistant. Insulin tolerance tests showed no significant differences between nondiabetic mice and diabetic mice fed the medium-fat diet, or nondiabetic and diabetic mice treated with HD-Ad-0 or HD-Ad-VLDLR (data not shown).

Because reduced levels of insulin receptors on macrophages might result in altered lesion morphology or lesion size (12, 13), we also analyzed insulin receptor b-subunit expression levels in both macrophages and liver by Western blot analysis. There were no significant differences between nondiabetic and diabetic mice (data not shown). Together, these results indicate that the differences in lesion morphology in the present study are unlikely to be due to altered insulin resistance or insulin receptor expression in macrophages or liver.

Liquid-Chromatography-Tandem Mass-Spectrometric Identification of S100A9. Of the proteins identified by this approach, S100A9 (IPI00222556) showed the greatest increase in samples from diabetic mice versus nondiabetic mice (SI Fig. 8). Another protein abundantly secreted by macrophages, granulin (IPI00124640), showed no difference between the groups. The S100A9 abundance change was further corroborated by extracted ion chromatogram of a selected peptide derived from S100A9 (peptide 1-40;SITTIIDTFHQYSR) for which a strong chromatographic peak was detected in all of the media samples conditioned by monocytes/macrophages from diabetic mice, whereas no peak was detected in any of those from nondiabetic mice (SI Fig. 8), indicating significant enrichment of S100A9 from diabetic animals.

1. Oka K, Pastore L, Kim IH, Merched A, Nomura S, Lee HJ, Merched-Sauvage M, Arden-Riley C, Lee B, Finegold M, et al. (2001) Circulation 103:1274-1281.

2. Renard CB, Kramer F, Johansson F, Lamharzi N, Tannock LR, von Herrath MG, Chait A, Bornfeldt KE (2004) J Clin Invest 114:659-668.

3. Lewis KE, Kirk EA, McDonald TO, Wang S, Wight TN, O'Brien KD, Chait A (2004) Circulation 110:540-545.

4. MacDougall ED, Kramer F, Polinsky P, Barnhart S, Askari B, Johansson F, Varon R, Rosenfeld ME, Oka K, Chan L, et al. (2006) Am J Pathol 168:2064-2073.

5. Gough PJ, Gomez IG, Wille PT, Raines EW (2006) J Clin Invest 116:59-69.

6. Kersey PJ, Duarte J, Williams A, Karavidopoulou Y, Birney E, Apweiler R (2004) Proteomics 4:1985-1988.

7. Keller A, Nesvizhskii AI, Kolker E, Aebersold R (2002) Anal Chem 74:5383-5392.

8. Nesvizhskii AI, Keller A, Kolker E, Aebersold R (2003) Anal Chem 75:4646-4658.

9. Liu H, Sadygov RG, Yates JR III (2004) Anal Chem 76:4193-4201.

10. Old WM, Meyer-Arendt K, Aveline-Wolf L, Pierce KG, Mendoza A, Sevinsky JR, Resing KA, Ahn NG (2005) Mol Cell Proteomics 4:1487-1502.

11. Bolshev LN (1962) Theory Probability Appl 7:113-114.

12. Baumgartl J, Baudler S, Scherner M, Babaev V, Makowski L, Suttles J, McDuffie M, Tobe K, Kadowaki T, Fazio S, et al. (2006) Cell Metab 3:247-256.

13. Han S, Liang CP, DeVries-Seimon T, Ranalletta M, Welch CL, Collins-Fletcher K, Accili D, Tabas I, Tall AR (2006) Cell Metab 3:257-266.

SI Figure 9

Fig. 9. Significant correlation between plasma triglyceride levels and a marker of inflammation in diabetic mice. By using the Quantikine ELISAs, plasma levels of sICAM-1 were measured in nondiabetic (filled squares) and diabetic (open circles) mice fed the high-fat diet for 16 weeks and then the low-fat or medium-fat diet for an additional 14 weeks. Statistical analysis was performed by using Pearson correlation tests. The P value indicates significant correlation between plasma triglycerides and sICAM-1 levels in diabetic mice.

Table 7. Primer sequences and product sizes

CD115, CSF 1 receptor; CD11b, integrin aM; Galectin-3, Mac-2.

• Early Edition
• Archives
• Online Submission

• Feature Articles

  (Expanded research articles of exceptional breadth)

• Commentaries

  (Expert commentary on leading research)

• Letters

  (Online-only letters to the editor)

• Inaugural Articles

  (Articles by recently elected NAS members)

• PNAS Profiles

  (Biographical profiles of Academy members)

• Sustainability Science

  (Interactions of human and environmental systems)

• Collected Papers

  (Editorials, Perspectives, Reviews, Colloquia, etc.)

• Special Features

  (Solicited articles on exceptional research topics)

• Sackler Colloquia

  (Scientific reports held under NAS auspices)