CDP, the plasma from which cryoprecipitate has been removed, was obtained from Blood Components Production Section, National Blood Centre, Thai Red Cross Society (NBC-TRCS). For the preparation of CDP, a unit of citrated whole blood was spun using a heavy spin in the centrifuge, and the plasma was removed within 6 hours after collection. Fresh Frozen Plasma (FFP) was then prepared by snap freezing the plasma unit at -70 ^oC and stored at below -30 ^oC. FFP was slowly thawed overnight at 4°C and centrifuged to separate the cryo-supernatant from the insoluble cryoprecipitate. The insoluble cryoprecipitate was removed and the remaining plasma (CDP) was refrozen.

RVV-X, obtained from Snake Bite and Venom Research Unit, Faculty of Medicine, Chulalongkorn University, Thailand, was purified from crude Daboia russelli siamensis venom by a modification of the procedure of Kisiel et al. (1976) [18] using sequential column chromatography. The specific Factor Xa activity was 1.240 nkat/ng.

Thrombin was prepared by a modification procedure of Biggs R. (1972) [19]. The starting volume 100 ml of CDP was diluted to 1000 ml with distilled water, the pH was adjusted to 5.3 with 2% acetic acid and followed by centrifuge. The precipitate was dissolved in 25 ml of 0.85% sodium chloride and the pH was adjusted to 7.0 with 2% sodium carbonate. This was followed by the addition of 0.25 M calcium chloride with or without RVV-X, and incubated for full thrombin formation or without incubation. The coagulated fibrin was removed and acetone was added to the thrombin crude solution (volume 1:1) at room temperature to eliminate electrolytes and to denature some of the protein impurities [20]. The solution was centrifuged to separate the precipitate. The precipitate was finally extracted with 0.85% sodium chloride and followed by centrifuge. The supernatant was collected as thrombin solution and stored in a freezer at a temperature below -20 °C for its stability before the properties were tested.

Thrombin was prepared using the procedure described above in different conditions varying 4 related factors including: 1) RVV-X; 2) 0.25 M calcium chloride volume; 3) incubation time; and 4) 0.85% sodium chloride final extraction volume. The study compared different conditions step by step to determine the optimum preparing condition as the following methods: M1.0 vs. M2.0; M2.0 vs. M2.1; M2.0 vs. M2.2; and M2.0 vs. M2.3 (Table 1). The prepared thrombin was then further analyzed by different methods for the following properties: 1) fibrin clot formation; 2) total protein; and 3) Thrombin Clotting Time (TT). The prepared thrombin from two optimum conditions was analyzed and compared for: 1) Molecular Weight (MW) and protein patterns; and 2) thrombin concentration.

The prepared thrombin was preliminarily tested for its activity by observation of fibrin clot formation after mixing with cryoprecipitate (volume 1:1) obtained from Blood Components Production Section, NBC-TRCS.

Kjeldahl method (determination of nitrogen by sulfuric acid digestion) was used for total protein assay of prepared thrombin according to the method described by European Pharmacopoeia 8.0 (2014) [21].

The prepared thrombin was diluted to 1:50 with distilled water. One hundred microliters of diluted thrombin was added to 100 microliters of fresh plasma in the test tube at 37 °C. The tube was gently shaken and tilted back and forth. The time taken for the first appearance of a fibrin clot was recorded.

The Molecular Weight (MW) of the prepared thrombin was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using standard protein molecular weight markers.

The prepared thrombin was determined by the automated coagulation analyzer (Sysmex^® CA-560) compared to Japanese Pharmacopoeia Reference Standard Thrombin (690 units/ampoule, Control: THR0104) from the Society of Japanese Pharmacopoeia, for its activity in the International Units (IU/ml).

Prothrombin is the coagulation proenzyme present in the highest concentration blood (0.07-0.1 mg/ml) and was recognized very early as a prime contributor to the blood coagulation process [22]. By the common pathway, activated factor X (factor Xa) along with its cofactor (factor V), tissue phospholipids, platelet phospholipids and calcium forms the prothrombinase complex which converts prothrombin into thrombin [23]. In our study, factor X has been enzymatically activated by a coagulant protein (heterotrimeric metalloproteinase) [24] in Russell's viper venom (RVV-X) to help make the conversion of prothrombin more complete than adding calcium to the plasma alone.

The first study was conducted on the preparation of thrombin using Method M 1.0 and M 2.0 to compare the presence and absence of RVV-X in the preparation process. Thrombin was prepared from CDP (or cryo-supernatant), which is rich in prothrombin complex and has a low level of fibrinogen [25, 26], by the procedure described in 2.3. From starting volume 100 ml of CDP, we obtained about 25 ml of thrombin solution in the preparation batch. The solution was slightly turbid with some small particles due to no filtration step in the preparation process.

The preliminary test confirmed that there was thrombin in the prepared solutions, which can be observed from the reactions between thrombin and fibrinogen in cryoprecipitate to make the fibrin clot formation (Fig. 1).

Fig. (1). The appearance of fibrin clot after mixed the prepared thrombin with cryoprecipitate.

Total protein content analyzed by Kjeldahl method showed that the thrombin solution of Method M2.0 had higher protein content than of Method M1.0, as well as the shorter thrombin time of Method M2.0 than of Method M1.0. At this step, we concluded that the use of RVV-X in Method M2.0 resulted in a greater amount of thrombin than that of Method M1.0 without RVV-X (Table 2).

The reaction of coagulant protein on factor X is proteolytic, has an absolute requirement for Ca (II), and proceeds at physiological pH [27]. In this study, 0.25 M calcium chloride was reduced by half to 1.5 ml to see how it affects the preparation of thrombin.

At the half-dose of 0.25 M calcium chloride (Method M2.1), the total protein content of thrombin solution was lower than Method M2.0, and the thrombin time of Method M2.1 was twice longer than Method M2.0, indicating that the amount of thrombin prepared by Method M2.1 was much lower than of Method M2.0. In this step, it was suggested that the thrombin preparation process should have a sufficient amount of calcium in order to maximize the conversion of prothrombin (Table 3). This result corresponded to the previous study which indicated that prothrombin concentrates incubated with near-physiological levels of calcium appeared to correct the abnormal APTT to an increasing degree as the calcium concentration was increased [28].

For the next study, the different volume of 0.85% sodium chloride solution used to dissolve the final protein precipitate was compared between 25 ml (100%) in Method M2.0 and 10 ml (40%) in Method M2.2, to determine whether the sodium chloride extraction volume affected the properties of prepared thrombin or not.

The total protein content of Method M2.2 thrombin solution was higher than that of Method M2.0, and the thrombin time of Method M2.2 thrombin solution was nearly twice shorter than that of Method M2.0. These results indicated that thrombin concentration of Method M2.2 was highly concentrated, therefore resulted in higher activity than Method M2.0 (Table 4).

After the first three studies on RVV-X, calcium and sodium chloride final extraction volume, the most appropriate method for producing thrombin was Method M2.2 (using RVV-X 50 mcg, 0.25 M calcium chloride 3 ml, 0.85% sodium chloride volume 10 ml for final extraction, and incubation time 2 hours) as it gave the highest total protein and lowest thrombin time. The next step was to compare the incubation time factor between Method M2.2 (incubation time 2 hours) and M2.3 (no incubation time) to determine whether thrombin can be generated without incubation time. Theoretically, the RVV-X, a factor X activator, is capable of promptly converting prothrombin into thrombin [29]. Therefore, it was expected that Method M2.3 would result in thrombin amount not different from that of M2.2.

The results showed that the total protein content of thrombin solutions including the thrombin time of Method M2.2 and M2.3 was very similar. However, the preparation process of Method M2.3 was more convenient as it did not require incubation time up to 2 hours (Table 5).

The MW of the prepared thrombin from two optimum conditions (Method M2.2 and M2.3) and diluted CDP were determined by SDS-PAGE using standard protein molecular weight markers (Fig. 2). From SDS-PAGE results, CDP contained several proteins including 72,000 Da prothrombin (lane C). Upon activation with calcium and RVV-X, 36,000 Da protein band which is the size of thrombin protein, appeared (lane A and B) [30, 31]. Thrombin bands on the digest of Method M2.2 (lane A) and Method M2.3 (lane B) had a similar intensity even without incubation process in Method M2.3 due to the ability of RVV-X in activated prothrombin into thrombin promptly.

At this stage, thrombin concentration has been tested using a coagulation analyzer by comparison with the standard thrombin, and reported as a thrombin unit. It was found that the thrombin concentration of Method M2.2 and M2.3 was 269.4 and 295.2 IU/ml, respectively, which indicated that Method M2.3 was the most rapid preparation method to produce thrombin and gave the highest yield of thrombin as well (Table 6).

Fig. (2). SDS-PAGE of produced thrombin prepared from two optimum conditions. Lane M: molecular weight markers; Lane A: condition A (method M2.2); Lane B: condition B (method M2.3); Lane D: diluted CDP (10x).