Chemicals for DNA extraction were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shenyang, China). Cytosine and 5-methylcytosine for HPLC analysis were supplied by Sigma; HPLC-grade methanol was supplied by Fisher Scientific. Deionised H₂O (18 MX·cm^–1) was used for all experiments.

Rice (Oryza sativa L) were used for nucleic acid extractions.Nucleic acids were extracted from 2 to 3-week-old leaf tissue according to the CTAB based on Miao [6]. Deoxyribonucleic acid extractions purified were resuspended in TE buffer (0.2 m-MTris-HCl, 1 mM EDTA, pH 7.5), Nucleic acid concentration and purity was estimated by measuring the absorbance at 260 nm (assuming 1 OD equates to 50 µg·ml^–1) and samples were stored at −40˚C.

Nucleic acid samples (3 µl) of Oryza sativa L treated with RNases were mixed with 1 µl of loading buffer (0.25% w/v bromophenol blue). The agarose gel (1.2% w/v) of 100 ml, adding ethidiumbromide 4 µl（0.5 µg·ml^–1) before solidification, were loaded onto a 65 mm × 65 mm. Then it was prepared in TBE buffer (0.089 MTris, 0.089 M boric acid, 0.01 M EDTA, pH 8.0) and performed at 90 V, 15 and 30 min .Then observed and photographed electrophoresis band by using a Digital gel image analysis system of Jeda (Jiangsu, China).

Acid hydrolysis of DNA refered to [13]. DNA solution 100 μl (containing 30 μg DNA) was added 50 μl perchloric acid (70%), and then was hydrolysed for 60 minutes in boiled water. Acid hydrolyzates PH was adjusted to 3 - 5 and centrifuged with 12,000 rpm, 10 min. Then take supernate fluid filtered to perform HPLC program.

The C and 5-mC standard stock solutions refer to [14] weighing C 0.4444 mg and 5-mC 0.55145 g, and dissolving 0.1% perchloric acid, then concentrations of C and 5-mC mother liquor were 4 × 10² μmol/L and 20 μmol/L.

HPLC analysis was performed using Agilent LC 1200 with Quat pump (Germany), a manual injection valve fitted with a 20 µl loop, a Tcc UV absorbance detector (Germany) set to 285 nm. A column for Diamonsil C18 (250 × 4.6 mm, 5 μm) was used in this detection. The HPLC system was flushed with deionised H₂O (0.22 µm filtered) at 0.5 ml·min^–1 for 15 min to remove all traces of phosphate at the end of each day, followed by 15 min at the same flow-rate with 90% v/v methanol (0.22 µm filtered) to clean the column.

The level of 5-MedC present in the DNA samples was expressed as a percentage of the level of dC which was calculated using the following equation:

% 5-mC = [5-mC/dC + 5-MedC] × 100.

There are a small number of impurities in liquid supernatant of DNA regardless of sucking liquid supernatant carefully. The result could be observed on Centrifugal tube bottom. Consequently, another centrifugation is essential after extracting liquid supernatant of DNA (data not shown).

RNA in genome DNA should be cleansed as far as possible because the determination of total level on DNA methylation is effected by RNA of DNA. The most common method to remove RNA from DNA is to treat extractions with RNase A [15]. An agarosegel showed RNase A had effectively removed residual RNA from rice DNA. (Figures 1 and 2) The results of ultraviolet spectrophotometry showed that the number of DNA optical density (OD) treated with RNase A approachhed between 1.80 and 2.0 (1.80 < OD < 2.0) much more, but the number without RNase A is almost over 2.0 (OD > 2.0).

C18 column is currently used for HPLC analysis of DNA

methylation. The study showed that relative retention times and separation degree of cytosine and 5-methylcytosine had changed obviously (Figures 3 and 4). In this detection, C18 column improved the separation degree much better.

The study showed the pH of the mobile phase affected cytosine and 5-methylcytosine peak separation and shape (Figures 5 and 6). A lowering in pH from 5.8 to 4.5 reduced the retention time of cytosine and 5-methylcytosine (Figures 5 and 6). Gehrke et al. [16] found similar results when studying a narrower pH range of 4.0 - 5.0. In addition, the strongest effects of pH for improvement of peak separation occurred at pH values below 4.0 [8]. In conclusion, the PH of the mobile phase effected greatly in the retention time and peak shape of C and 5-mC.

The calibration lines were constructed by the dilution of the stock standard solutions for C and 5-mC, which were used to determine level of 5-MedC and dC in the hydrolysed DNA samples. The results show that C and 5-mC had the better linear relation between the peak area and concentration of the standard solutions (data not shown). In addition, C and 5-mC were separated in Actual sample chromatogram with the condition for 50 mmol·L⁻¹ KH₂PO₄, PH 5.8, flow rate 0.5 ml·min⁻¹ (Figure 7).