EFFECTS OF PULSE STIMULATION ON COLLAGEN INCREASE 417 Experiment 1: Electric stimulation with unipolar Gunpatsu pulse (Figure 1) (time course). A total of 50 rats were randomly divided into fi ve groups and housed individually. The control group (10 rats/group) did not receive any electric stimulation. The other four groups (10 rats/group) received the unipolar Gunpatsu pulse stimulation (see Figure 1) for 0.5, 2, 5, and 10 min every day for one week. Experiment 2: Electric stimulation with bipolar Gunpatsu pulse (combined with a VC solution). A total of 40 rats were randomly classifi ed into four groups and were housed individually. The control group (10 rats/group) was treated with a saline solution without any electric stimulation. The other three groups (10 rats/group) were treated with (a) the VC deriva- tive (aminopropyl ascorbyl phosphate, pH 7.4) solution without any Gunpatsu pulse stimulation, (b) bipolar Gunpatsu pulse stimulation (see Figure 1) for 10 min with saline, or (c) bipolar Gunpatsu pulse stimulation for 5 min with the VC solution. Gunpatsu pulse stimulation was performed at 4800 Hz and 5 V every day for one week. MEASUREMENT OF THE AMOUNT OF COLLAGEN IN SKIN Most methods for determining collagen levels are based on measurements of the quanti- ties of proline and/or hydroxyproline, the principal amino acid components of collagen. Recently, we developed a simple and highly sensitive method of measuring the amount of hydroxyproline in skin by using isocratic HPLC with NBD-F (9). REAGENTS NBD-F was purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan). The standards for HPLC were hydroxyproline (Wako Pure Chemical Industries Ltd.) and a standard amino acid mixture, Type AN-2 (Wako Pure Chemical Industries Ltd.). The Wako standard amino acid solution is an equimolar mixture of Asp, Cys, Glu, Ser, Gly, Figure 1. Typical pulse pattern of either unipolar or bipolar stimulation of Gunpatsu pulse. Unipolar stim- ulation shows the only positive site of 4800 Hz, and the bipolar stimulation shows the both positive and negative sites at each 1.5 sec. Electric stimulation (4800 Hz, 5 V, duty: 50% and 1.2 mA) was obtained by the new device (see photograph).

JOURNAL OF COSMETIC SCIENCE 418 His, Arg, Thr, Ala, Tyr, Val, Met, Ile, Leu, Phe, Lys, Pro, and hydroxyproline. All other chemicals were of HPLC grade. Hydrochloric acid (HCl), acetonitrile, potassium dihy- drogenphosphate, sodium tetra borate decahydrate, and boric acid were purchased from Wako Pure Chemical Industries Ltd. Ethylenediamine tetra acetic acid disodium salt dehydrates (EDTA-2Na) was purchased from Pharmacia Biotech AB (Uppsala, Sweden). HYDROLYSIS OF PROTEIN AND DERIVATIVE Tissue samples were hydrolyzed for 22 h at 108°C in evacuated fl ame-sealed Pyrex tubes with 400 μl of 6 N HCl, to which 1% phenol was added. The hydrolysates were dried in vacuo. For the HPLC analysis, the dry residue of each hydrolysate was re-suspended in 500 μl of 50 mM borate buffer containing 20 mM EDTA-2Na (pH 8.0). The resuspended samples were diluted to 1/2000 with a 50-mM borate buffer containing 20 mM EDTA- 2Na and 7.5 μM of L-homoserine (pH 8.0). Next, 30 μl of either the hydrolyzed sample or a mixed amino acid standard solution diluted to 1/100 with a 50-mM borate buffer containing 20 mM EDTA-2Na (pH 8.0) was poured into a 500-μl conical tube. Then 10 μl of NBD-F (10 mM in acetonitrile, freshly prepared) was added to this solution and the tube was capped and covered with aluminum foil. The vessel was heated to 60°C for 1 min, and after it had cooled in ice water, 40 μl of 0.05 M HCl was added to the reaction mixture. Finally, 10 μl of the solution was injected into the column. HPLC CONDITIONS Briefl y, an EP-300 (Eicom Corp., Kyoto, Japan) pump equipped with an L-7200 auto sampler (Hitachi, Tokyo, Japan) was employed. A guard column (200 mm × 3.9 mm) and a main column of SC-5ODS (200 mm × 4 mm, 10 μm Eicom Corp., Kyoto, Japan) were used. All solvents were fi ltered and degassed prior to use. The fl ow rate was 0.8 ml/min, and the column temperature was kept at 30°C using a column oven. A Hitachi L-7485 spectrofl uorometer equipped with a 12-μl fl ow cell was used with an excitation wave- length of 470 nm and emission at 540 nm. A phosphate buffer (pH 5.0) containing 8% acetonitrile was used as the mobile phase. STATISTICAL ANALYSES Results are presented as the mean ± SE. All results were analyzed with Dunnett’s multi- ple comparison test of the Pearson correlation after an analysis of variance. RESULTS During this experiment, there was no signifi cant difference between the groups in body weight or quantity of water ingested. No adverse effects were observed in any of the rats (data not shown). All results, particularly for the analysis of hydroxyproline content, ob- tained by HPLC are clearly documented in Figure 2. The retention times of hydroxypro- line and Pro were about 8 and 34 min, respectively. L-homoserine, as an internal standard, was eluted for about 12 min. Experiment 1: Changes in hydroxyproline levels in the skin and increase in body weight. In the control group (Figure 3 control: the electrode for Gunpatsu pulse was tied around the