Twelve-week-old male C57BL/6 mice from the Experimental Animal Center of the Fourth Military Medical University were housed five/cage under a temperature of 25 ± 1 °C, 50 ± 5 % humidity, with an alternating 12 hrs light–dark cycle and free access to food and water ad libitum. The type of housing facility was specific pathogen free (SPF), and the cage is 30 cm (width) × 40 cm (depth) × 20 cm (height). For STZ induced diabetes model, mice were injected intraperitoneally with streptozocin (150 mg/kg, Sigma-Aldrich) dissolved in a mix of citrate buffer (citric acid and sodium citrate, pH 4.8) or vehicle (citrate buffer) after overnight fast similar as described before [9]. Blood glucose was checked 5 days later via tail vein; mice with a blood glucose level above 13.5 mmol/L were considered diabetic. As a control, mice were injected with equal volume of citrate buffer. In total, 35 mice were include in this study, which were divided into 7 groups with 5 animals in each group: including control, NAC only, DM (diabetes without NAC treatment), and 4 different NAC treatment groups. The 4 NAC treatment groups, namely NAC1, NAC3, NAC5 and NAC7, define the start time point when NAC treatments start. For example, in the NAC1 groups, diabetic mice were treated with NAC (A9165, Sigma-Aldrich) from 1 week after STZ induction at the dose of 1.0 g/kg body weight per day in drinking water. In the NAC only group, control mice were further treated with NAC for five weeks. No obvious adverse events were seen in each experimental group. The detailed procedure described in Fig. 1.

Echocardiography was performed from week 12 after STZ injection. Transthoracic 2-dimensional (2D), M-mode and Doppler echocardiographic studies were performed with Mylab 50 (Esaote, Italy) using a high-resolution transducer (SL3116) with frequency of 22 MHz. Briefly, each mouse was anesthetized by injecting intraperitoneally with 10 % chloral hydrate at the dose of 350 mg/kg body weight before echocardiographic study [10], which had an onset of sedation within 5–10 minutes and was maintained for about 30–40 minutes. Heart rates were monitored and generally maintained around 450 beats per minute. The chest hairs were removed using Depilatory creams. The mouse was then placed on a warm pad to keep the body temperature around 36 ± 0.5°C. Warmed echo gel was placed on the shaved chest as a coupling medium while the mouse lay on the warm pad at a supine position. Images were acquired and analyzed by an operator blinded to mouse treatment.

Interventricular septal thickness and LV posterior wall thickness during diastole (IVSd, LVPWd), LV internal dimensions during diastole (LVIDd) and systole (LVIDs) were measured from M-mode images at the level of the papillary muscles at LV short-axis view (Additional file 1: Figure S1a). Representative images were digitally acquired and stored on the internal hard disk and USB Mass Storage Device for off-line analysis. LV ejection fraction (EF), LV fractional shortening (FS) were calculated according to the recommendation of the American Society of Echocardiography Committee [11].

Transmitral inflow Doppler was obtained from the apical 4-chamber view. The sample volume was placed just below the level of the mitral annulus and adjusted to the position at which the velocity was maximal. The angle correction was kept less than 20 degree. LV diastolic function was evaluated using the methods described previously [12]. In brief, the left ventricular isovolumic relaxation time (IVRT) and the acceleration and deceleration times of the early peak (E) wave (E_AT and E_DT, respectively) were derived respectively from the Doppler waveform (Additional file 1: Figure S1b).

After echocardiography, the heart was excised from the chest, trimmed of atria and large vessels and weighed. Half of the hearts (in the long axis view) were formalin-fixed for Hematoxylin Eosin (HE) and Trichrome staining, while the other half were mounted with OCT directly for ROS staining. For histological analysis, excised hearts were washed with saline solution, placed in 10 % formalin, and embedded in paraffin. Then, 5-μm thick sections were prepared and stained with Masson Trichrome staining for detecting the myocardial fibrosis [13]. To determine myocardial ROS generation, dihydroethidium (DHE) staining was included by probing for the ROS on the 5-μm frozen myocardial sections [14].

Fibroblasts were isolated from the hearts of normal P7 (postnatal day 7) male C57/Bl6 mice similar as previously described [15]. Briefly, 3 hearts were isolated and vessels and atria were removed before transferred to 1 mL of collagenase buffer. In the buffer, the ventricles were quickly minced into small pieces and digested for about 1 hour. Cell suspension were filtered with 100 μm filter and then centrifuged. The cell pellet was re-suspended and plated on a T75 tissue-culture flask (Corning Corp) in full medium supplemented with 10 % of fetal bovine serum (HyClone) and antibiotic-antimycotic solution. Non-adherent cells were removed after overnight culture, and adherent cells were cultivated as cardiac fibroblast. Only fibroblasts at passage 1 to 5 were used for the following experiments.

Cardiac fibroblast cells were cultured in the serum free medium containing either 5.5 mM (normal glucose, NG) or 25 mM glucose (HG) with 10 ng/ml TGFβ1 and without insulin for 24 hrs. In the HG group, cells were further added with control or NAC (5 mM). RNA was isolated with TriZOL (Invitrogen). Reverse transcription was performed with the Superscript III First Strand Synthesis kit (Invitrogen). SYBR Green Mix I (Takara) was used for amplification, and samples were run on an ABI7500 Instrument (AB, USA). Gapdh was used as internal control. 2− ΔΔCt method was used for analysis (n = 3). The primers are listed as follows: Gapdh forward, 5′-TGGCCTTCCGTGTTCCTACCC-3′, Gapdh reverse, 5′-AGCCCAAGATGCCCTTCAGTG-3′; Col1a1 forward, Col1a1 reverse, 5′-GGAATCCATCGGTCATGCTCT-3′; CTGF forward, 5′-CCACCCGAGTTACCAATGACA-3′, CTGF reverse, 5′-CTTGGCGATTTTAGGTGTCCG-3′.

Cardiac fibroblast cells were seeded in 96-well plates at a density of 1.5 × 10³ cells per well and treated as indicated. Cell numbers were analyzed by Cell Counting Kit-8 (Sigma-Aldrich) at 450-nm absorbance.

All data were expressed as mean ± SD. The mean data of six groups were compared with one-way ANOVA. The intra-and inter-observer variability were analyzed using 2-tailed Student’s t-test and linear regression analysis. A P-value < 0.05 was considered statistically significant.