Changes in cardiac heparan sulfate proteoglycan expression and streptozotocin-induced diastolic dysfunction in rats

This work was performed in compliance with the ARRIVE guidelines on animal research [18]. All experimental procedures were performed in accordance with the Guidelines for Ethical Care and Use of Experimental Animals and approved by the institution's ethics committee. Diabetes was induced in overnight-fasted rats by a single intravenous injection of 50 mg/kg of STZ in citrate buffer (Sigma Chemical Co., St. Louis, MO, USA). Diabetes induction by STZ injection is a well-characterized model of experimental Type 1 diabetes, where the selective destruction of beta-cells of pancreatic islets promotes a permanent hyperglycemic state and consequently diabetes complication [19]. After STZ injection, all animals were returned to their cages and kept in a ventilated shelf under a controlled 12-hour light/dark cycle and temperature (21-24°C).

Sixteen male Wistar rats obtained from the central animal facility of the University of São Paulo Medical School weighing 200-300 g were randomly divided into 4 groups of 4 animals each according to time of diabetes induction: control (sham injection with saline), 24 hours, 10 days, and 30 days after STZ injection.

All animals were sacrificed with a lethal dose of sodium pentobarbital (60 mg/kg i.p.) and had the heart and the gastrocnemius muscles immediately removed, washed in ice-cold 0.9% NaCl solution, weighed, frozen in liquid nitrogen, and stored at -80°C.

High-resolution echocardiography was performed in 4 diabetic rats after 10 days of induction and in 4 rats of the same age and sex from the control group just before sacrifice. Transthoracic echocardiography was performed after anesthesia with pentobarbital (40 mg/kg; i.p.) using a Sequoia 512 machine (Acuson, Mountain View, CA) equipped with a 13-MHz linear-array transducer. Images were stored digitally on magneto-optical discs [20].

Left ventricular end-systolic (LVESD) and end-diastolic dimension (LVEDD), interventricular septal thickness (IVST), and posterior wall thickness (PWT), both in diastole, were measured at the level of the papillary muscles on the short-axis view using 2-D guided M-mode imaging [20, 21]. All measurements were obtained according to American Society of Echocardiography recommendations [22]. Three representative cardiac cycles were analyzed and averaged for each measurement. Fractional shortening was calculated from the M-mode recordings, as previously described [23]. Left ventricle (LV) mass was calculated from M-mode recordings by using the uncorrected cube formula, assuming spherical LV geometry [20]. The LV mass index was determined as the ratio of LV mass in grams to body weight in grams.

The peak velocity of early (E) and late (A) diastolic filling and E/A ratio, deceleration time of the E wave (DT), and isovolumic relaxation time (IVRT) were obtained from the mitral inflow recordings, as previously described [20]. DT and IVRT were obtained in animals when the E and A waves were not fused. We also measured systolic (S'), early (E'), and late (A') diastolic peak velocities of Doppler tissue imaging with a sample volume placed at the septal side of the mitral annulus in the apical 4-chamber view.

Proteoglycans (PGs) were extracted from the cardiac or skeletal muscle pool obtained from all 4 animals of each experimental group according to the protocol described elsewhere [24]. After extraction, the supernatants were dialyzed and purified in DEAE Sephacel (Sigma) and protein concentration determined using Bradford's method [25]. Samples underwent digestion with a mixture of 0.06 unit of chondroitinase ABC and 1 unit of heparitinases (Sigma) at 37°C for 4 hours to remove glycosaminoglycans from the protein core [26].

Ten μg of PGs underwent 7.5% SDS-PAGE, followed by electrotransfer to an Immobilon-P membrane (Millipore, Belford, MA, USA). Membranes were blocked and incubated with the primary antibody, either antiglypican-1 C-18 or antisyndecan-4 N-19 (1:1000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), followed by the secondary antibody, goat anti-rat IgG-HRP (1:5000; Santa Cruz Biotechnology, Inc). The blots were developed using a chemiluminescent reagent (Western Blotting Chemiluminescence Luminol reagent; Santa Cruz Biotechnology) and exposed to Kodak MGX/Plus film (Eastman Kodak, Rochester, NY, USA). The PGs were quantified with the Eagle Eye analyzer system (Stratagene, La Jolla, CA, USA). Mean and SE were calculated using 4 independent experiments. The results were expressed as arbitrary optical density units (OD).

Total RNA was isolated from the heart apical region and skeletal muscle by using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) as described in the manufacturer's protocol. RNA integrity was determined by agarose gel electrophoresis [27].

cDNA was synthesized using total RNA from cardiac and skeletal muscles by the Superscript Pre-amplification System (Invitrogen) and amplified by PCR using the manufacturer's protocol. PCR primers used in this study are listed in Table 1 and were designed using previously published sequences [28‐30]. The samples were amplified for 35 cycles (denaturation at 94°C for 45 seconds; annealing at 58°C for 1.5 minutes, extension at 72°C for 1.5 minutes). All amplifications were carried out using a thermal cycler (MJ Research Inc. Watertown, MA, USA). β-actin was used as a control.

The PCR products were purified with a Concert Rapid Gel Extraction System (Invitrogen) and cloned with a pGEM-Easy cloning vector (Promega, Madison, WI, USA.). The clones were identified by blue/white selection, processed for plasmid purification with the Wizard Plus kit (Promega), and sequenced (ABI PRISM- 377 DNA Sequencer- Perkin Elmer, Foster City, CA, USA) by using DNA prepared according to the protocol of the sequencing kit DNA Big Dye Terminator (Applied Biosystem- Perkin Elmer, Norwalk, CT, USA). The identity of the products was verified by using the basic local alignment search tool (BLAST) on the GenBank database.

Total tissue RNA (10 μg) from 4 animals of each group (control, 24-hour, 10-day, and 30-day) was fractionated in 1% agarose-formaldehyde gel [31]. The RNA was transferred to a positively charged membrane (Duralon-UV Stratagene) and prehybridized in a Stratagene hybridization solution and hybridized with digoxigenin (Dig)-labeled probes, syndecan-4 and glypican, prepared according to instructions of the Dig DNA labeling kit (Roche Molecular Biochemicals, Mannheim, Germany). The blots were revealed with a Dig luminescent detection kit (Roche) and exposed to Kodak MXG/Plus film. The hybridization signals were quantified with the Eagle Eye system, and normalized to the amount of 28S rRNA in the samples. Results are presented as mean ± SE of the relative density (optical density of proteoglycan/optical density of 28S rRNA * 100) of samples from 4 rats in each group. The experiment was run in duplicate.

For histological analysis, 5-μm sections were obtained from formalin-fixed paraffin-embedded samples of cardiac and skeletal muscle from control rats and the 30-day group and routinely stained with hematoxylin-eosin.

Three-micrometer sections from these specimens were deparaffinized, rehydrated, and underwent antigen retrieval in 0.5% pepsin pH 1.8 for 30 minutes at 37°C. The specimens were then incubated in 3% aqueous hydrogen peroxide for 30 minutes to quench endogenous peroxidase activity. Incubation with 1% BSA and 5% fetal calf serum in Tris-HCl, pH 7.4 for 60 minutes at room temperature was performed to suppress nonspecific binding of subsequent reagents. The sections were then incubated with the primary antiglypican-1 C-18 and antisyndecan-4 N-19 (Santa Cruz Biotechnology Inc.) diluted 1:200. The Catalyzed Signal Amplification system (CSA system, HRP, Dako, Carointeria, CA, USA) was used and consisted of 15-minute sequential incubations with a biotinylated link, streptavidin complex, amplification reagent, and streptavidin peroxidase. Each of these 15-minute incubation steps was preceded by a 5-minute rinse with 1% Tween 20 Tris-HCl, pH 7.4. Staining was completed by 3-minute incubation with 3' -3 -diaminobenzidine tetrachloride (DAB; Sigma, St Louis, MO, USA), which resulted in a brown colored precipitate at the antigen sites. The specimens were then lightly counterstained with Mayer's hematoxylin, dehydrated, and xylene-based mounted under glass cover slips. Negative controls were treated as above, but a solution of 1% BSA in Tris-HCl, pH 7.4 replaced the primary antibody. Epithelium and blood vessels were considered internal positive controls.

All STZ-treated animals developed severe hyperglycemia (> 27.0 mmol/L), including the 24-hour group, compared with the control (nondiabetic) group (< 7.5 mmol/L). Body weight was significantly lower in 10-day and 30-day groups compared with control animals according to ANOVA followed by the Tukey Multiple Comparison Test: 217.4 ± 3.9 g and 222.4 ± 9.53 g vs 265.8 ± 10.9 g, P < 0.05.

Left atrium dimension, LVESD and LVEDD, IVST, and PWT, LV fractional shortening, and LV mass index were not statistically different in the control and 10-day groups (Table 2). Although LV fractional shortening was normal, S' peak velocity was decreased in the diabetic group (Table 2). This finding shows that Doppler tissue imaging is a useful tool for the assessment of systolic function in this model of diabetic myocardial disease and is altered earlier than the standard indices of global systolic function.

Diastolic function indices allowed the differentiation between groups, because E peak velocity was decreased in the diabetic group, and IVRT was increased in the same group. No difference occurred between diabetic and normal groups for E/E' ratio (Table 2). Taken together, the diastolic function indices reflect an impaired relaxation with normal LV end-diastolic pressure.

We detected the presence of glypican-1 and syndecan-4 mRNA in normal cardiac and skeletal muscles by RT-PCR. The identity of the HSPGs cDNA was confirmed by sequence comparison with published sequences for these molecules [GeneBank accession n° L02896 and M81786].

A single mRNA band was detected for glypican in each tissue: 5.01 kb in heart tissue (Figure 1A) and 5.36 kb in skeletal muscle (not shown). Glypican mRNA expression in cardiac muscle revealed, in comparison to normal tissue, decreased values in the 10-day group (relative density: 30.40% ± 1.08% vs 44.90% ± 4.02%; P < 0.05) followed by a significant increase, almost 100%, in the 30-day group (83.80% ± 5.07% vs 44.90% ± 4.02%; P < 0.001). STZ treatment did not change mRNA levels in skeletal muscle (Figure 1B).

Protein analysis showed the presence of 2 bands of molecular mass (Mr) of 50 and 57 kDa in cardiac and skeletal muscles. The bands obtained in our study are probably related to the presence of reduced and nonreduced glypican core proteins. Similar data were reported for glypican extracted from tumor cell lines in an immunoblot analysis that had a band of 48 kDa at nonreducing conditions and 55 kDa at reducing, quite similar to the molecular masses obtained in our study [32, 33].

As for the mRNA expression, an increase was noted in the amount of glypican core protein in cardiac muscle accompanying the development of diabetes, although this elevation could be detected not only in the 30-day group but also in the 10-day group, OD: 2.219 ± 0.06 and 1.520 ± 0.031, respectively, vs OD: 0.742 ± 0.0422; P < 0.001 (Figure 2A and 2B). On the other hand, our results obtained from skeletal muscle showed a small increase in the 24 hour-group (OD: 0.618 ± 0.023 vs 0.517 ± 0.030; P < 0.05) followed by a decrease of this PG in the 10- and 30-day groups (OD: 0.333 ± 0.020 and 0.178 ± 0.011), respectively, compared with the control, OD: 0.517 ± 0.030; P < 0.05 (Figure 2A and 2B). The immunohistochemistry qualitative analysis confirmed the higher level of glypican in cardiac muscle of the 30-day group compared with the control group (Figure 2C). For skeletal muscle, as expected, the analysis of the sections showed the presence of a few spots of colored precipitate in the control and the absence of them in the 30-day group, likely because of the lack of sensitivity of this method for such a low amount of PG (Figure 2D).

RNA blot filters hybridized with a rat syndecan-4 specific DNA probe revealed the presence of 2 bands (6.57 and 3.34 kb, not shown) for adult cardiac and one band for skeletal muscle, 6.37 kb (Figure 3A). Quantification of syndecan-4 mRNA expression in skeletal muscle revealed an increase of almost 50% in the 30-day group (relative density: 125.00% ± 10.60% vs 84.80% ± 5.98%; P < 0.01) compared to controls. On the other hand, for cardiac tissue we did not detect any changes between the samples tested (Figure 3B).

The protein analysis revealed a band of Mr 59 kDa, with a similar pattern for cardiac and skeletal muscles (Figure 4A and 4B). This molecular mass may represent a homo- or heterooligomerization of the protein core, quite similar to that observed by Hamon et al. (Mr 56 kDa) [34]. Although the mean optical density value obtained for syndecan from heart muscle in the 30-day group almost doubled compared to that in control, it was not statistically significant, probably because of the high variability of the results in this group. On the other hand, for the 10-day group the increase was significant (OD: 2.971 ± 0.666 vs 1.435 ± 0.072; P < 0.01). The same pattern was detected for skeletal muscle with an increase in protein level in the 10- and 30-day groups, OD: 2.160 ± 0.433 and 2.845 ± 0.643, respectively, compared with the control group, OD: 0.011 ± 0.005, P < 0.001. It must be emphasized that, in skeletal samples, the amount of this PG was low not only in the control but also in the 24-hour group (Figure 4A and 4B). The qualitative analysis of immunohistochemistry (control and 30 days after induction) confirmed the increase observed in the Western blot for the 2 muscles analyzed (Figure 4C and 4D).

HSPG mRNAs changes induced by STZ treatment were not paralleled by alterations in ß-actin mRNA from heart and skeletal muscle (not shown).