FRACTIONATION OF PROTEASES FROM COWTAIL RAY ( Trygon sephen ) VISCERA USING POLYACRYLIC ACIDS

Fractionation of proteases of cowtail ray (Tfygort, st;plwn) viscera using polyacrylic acids has been investigated. The viscera were ensiled using a mixture of propionic and formic acids (l:1, v/v) at a level of 3% (v/w) at 40"C for 5 days. Fractionation was carried out step wise at, 40C to produce highest possible activity of the enzymes at respective fractions. It was shown that an addition of 2Yo polyacrylic acids (PAA) solution to produce 0.275o/o PAA in the final solution followed by pH adjustment to pH 4.0 of fraction containing acidic protease, and to pH 8.0 ofthat cont,aining alkaline protease has been able to separate alkaline and acidic proceases. The process also purified the enzymes by 1.1and 4.7 fold for acidic and alkaline proteases, respectively.


INTRODUCTION
Manv methods have been studied and applied to separate proteins in solution from one another, and precipitation using polyelectrolytes is a method that is frequently used for such purpose.This technique is considered fast and safe especially to maintain the biological activity ofproteins such as enzlanes.
Oowtail ray (7.seplwn) viscera contain several different enzymes, some of which belong to proteases group.Ensilation has proved able to extract the enzymes (Poernomo, 1998), and separation of individual enzyrnes is necessary since it will assist the purification of the enzymes, especially if the enzymes have different charac.
teristics, such as those from fish viscera.This paper deals with the fractionation of alkaline and ncirlie proteases of cowtail ray viscera silage using PAA solution to facilitate subsequent purilication and characterisation of the enzymes.

MATERIALS AND METHODS
Oowtail ny (T.seph,e.n\viscerawere obtained from fish drying processors at Labuhan Maringgai, Lampung, Sumatera.Separation and washing of viscera were carried out at the process.ing sites.The viscera were road transported in ice (1:1, w/w) to the laboratory of the Slipi Research Station for Marine Fisheries, ,Iakarta which took about 8-10h.Upon arrival, theviscera were once more washed in fresh water.frozen and stored at -4500 until used.
Enzymes extraction was conducted by ensilation of the viscera.Ohopped (ca.lxl cm) viseera was added with a mixture of propionic and formic acids (l:1, v/v) at a level of ]Yo (v/w), and kept at 40"C for 5 days during which the mixture was stirred dailv.The silage was then eentrifugecl at 10,000x9 for l0 min.The sediment was separated from the silage, while the lipid on the surface was skimmed off.The supernatant, was frozen (-4500) in a chest freezer overnight and was stored at -25o to -ll0'0 until used.
Before used, the silage supernatant was thawed in running water and eentrifuged (10,000xg, l0 min) at 4'0 and then added with l/;l volume of ice.cold chloroform to remove the remaining fat.
Afterward.the mixture was left at 4"(l for l5 min, then the lower layer was separated using a separating funnel and finally recentrifuged to remove the remaining chloroform and fat.After centrifugation, the lower layer !\ras separated and referred to as crude enzyme extracts. The alkaline and acidic proteases were fractionated with PAA solution (IV[V il'5 million D alton/IvlD.Carbopol C9 34, BF'(;oodrich) th rou gh a series clf preliminary trials at 400.A volume of crude enzyme extract was mixed with 2% (w/v)   PAA solution to produce various concentrations (0.0125 to 0.i3%) of PAA in the final mixture.The mixtures were stirred for 30 min and then centrifugecl (10,000 x g. l0 min).The supernatant (S,) was separated and referred t'o as acidi'c Ttrote.oseext,rects.
The precipitates were redissolved in 0.4M Tris buffer pH 8.5 with an addition of lM OaCl, to produce 50 mM in the final mixture.The mixture was stirred for 10 min and recentrifuged.The subsequent supernatant (Sr) was separafecl ancl referred to as allmlirt,e protno^se ettracts.
The PAA concentration producing the highest.acidic protease activity and lowest alkaline prtltease activity in S, and lowest acidic protease ar:tivity and highest alkaline protease activity in S, was applied in the next trial.In the second trial, S, was adjusted to pH 4.0.pH B.5 and pH ;).0, while S" was adjusted to pH 8.5 and pH 9.0 using 2N tttll or 2N NaOH, respectively.Schematicnlly, the process is illustrated in Figure 1.
To a mixture of 0.5 mL assay buffer and 0.25 mL substrate equilibrated at 2500 was addecl 0.25 Silage supernatant € Supernatant (S,)   (acidic proteases) I'recipitates mL enzyme solution.After 60 min incubation at 250C, 5 mL 6% TCA was added to stop the reaction.Filtration was done after l5 min using Advantec filter paper no.131 (equivalent to   Whatman filter paper no.i3).The absorbance of the filtrate was read at 280nm and was corrected for the respective blanks.Blanks were obtained in a similar manner, but TCA was added before the enzyme.One unit of enzyme activity was defined as the increase of absorbance by 1.0 at 280 nm of TCA soluble material in 60 min at25'O.

RESULTS
The results of fractionation of alkaline and acidic proteases are presented in Figure 2 and Table 1.In the first trial (Figure 2), acidic proteases always remained in S, at levels of more than about 80% of the original and precipitated (contained in Sr) at levels lower than 5%.Alkaline proteases, however, remained in S, at levels higher than those IFR Journal Vol.V No.1, 1999   in the precipitate (in Sr) for PAA final concen- trations below 0.1%.Atfinal concentrations higher than 0.l%, the alkaline protease activities were higher in the precipitate (in Sr) than in S,.
A PAA final concentration of 0.2% resulted in the highest acidic protease activity in S, (about 98%).However, at this concentration, the supernatant still contained about ilO% alkaline protease activity, while in S, the alkaline protease activity was still low (about 50%) although the acidic protease activity was also low (about ll%).At a PAA final concentration of 0.275%, highest alkaline protease activity (about 65%) in S, was obtained.At this concentration, about 80% of acidic protease activity was retained in S, while the alkaline protease activity in S, was the iowest.
This treatment was used in the subsequent fractionation step.
The pH adjustment of S, and S, was bo inacti- vate the alkaline protease in S, and acidic protease  1 enzyme unit was defined as the increase Of tne absorbance at 280 nm of TCA soluble material in 60 min at 25oC using casein or hemoglobiri as substrate t<* Final concentration is 0.275 *** l.Jsing 2N HCI or NaOH in S..As shown in 'Iable l, the pH adjustmeiit"of S" did not significantly change the alkaline piotease activity.However, the acidic protease activity of this fraction was eliminated.The alkaline protease activity in Srwas zero after pH adjustment, but the acidic protease activity declined as the pH was reduced to ll.5 (76% remained) and :1.0 (l2.lo/o remained).At pH 4.0, the acidic protease activity increased to 99.5% of the origrnal.
Fractionation also partly purified the enzymes, and in a separate study it was shown that fractionation steps such as those in this study have been able to purify the enzymes by 1.1 and 4.7 fold for aci<lic and alkaline proteases, respectively (Iable 2).

DISCUSSION
The specilic activity of alkaline protease in the supernatant ofcowtail ray viscera silage, as shown in Table 1. was lower than that of salmon viscera silage as reported by Reece (1988).This was probably due to the differences in species as well as the procedure of ensilation.Moreover, cowtail ray is an elasmobranch fish which does not have a distinct llyloric caeca as a source of alkaline proteases.The period of silage storage in the present study was 5 days at 4000 and this might have affected the stability of alkaline protease in acidic conditions.Reece (1988) further reported that alkaline proteases of cod and mackerelviscera were not stable under acidic conditions during 64 ensilation and only acidic proteases were recovered.
ln the present study, PAA was used since it has been used by Sternberg & Hershberger (197'l), Caygill et al, (198;l) and Reece (1988) to successfully fractionate alkaline and acidic proteases.ln addition' in the preliminary study, the use of another polyelectrolyte, ['EI, to precillitate acidic proteases prov ed unsatisfaetory' In the precipitation of protein by t'EI' salt must be added to aid the process (Bell ef ol" 191J3; Jendrisak, 1987) and salt was ftrund to affect enzyme activity (Ihongthai et al., 1990: ( lildberg'   199:l; Gildb erg et al. , 1 99 1 ; Xu e.t aL ' 1996) '   The use of low molecultrr weight (90.000IDI'AA was also tested in the preliminary study.However' the results were not satisfactory since [he contamination of acidic proteases was high in the precipitate.Although low molecular weight isro,ooo-zro,000D) PAA has been successful in precipitating some enzymes (Sternberg &  i-I."rhb.tger, 19?il; Sternberg' 19?6; Shieh &   (]latz, 1991; Boucher e.t al., 1992), it seems that low molecular weight PAA was not suitable fbr fractionation of alkaline and acidic proteases from cowtail ray viscera.ln the fractionatirln of transglucosidase and amyloglucosidase (SternLerg, 1975) PAA of at least lMD molecular weight was recommended.According to Sternberg (19?6) the formation of an insoluble protein' polyacrylate complex depends among others on the mollcular weight of PAA and the ratio of Using I'AA of 3-bMD molecular rveighl.the precipitation pattern of tlLe alkaline prot,ease, which was indicated by the enzyme activities, was asymmetrir:ally bell shapecl (Figure 2).A similar result was also shovr'n by Caygiil et o.l. (198i1) {trr plant proteases (taparn and ficin) ancl Boucher et   ol.(1992) in precipitating Streptontyces N174 chitosanase.It r'vas suggested that the decrease in recovery at higher polvelectrolyte dosage rvas due to the excess polyelectrolyte which was incorpo- rated into the already formecl protein-polv- electrolyte r:omplex, resulting in a soluble complex with a significant net charge, and thus hindering the aggregirtion of the complexes ((lavgill et o,1,.,   198,1; (llark & (ilal,z.1987, 1988; Ohen & Rerg,   199;l).
In their sl utlics.(lavgill et ol.(l9l3i3) emlllovetl PA{/protnin ratios of 0.02'0.5.rvhile Bou<:her el ol.(1992) trsctl ralios of up to 10. 'l'he optimum ratio {br plant.proleases was between 0.14'0.29 and 0.2-0.4Iirr pa;lain and ficin, respectivelv.while for chil,ositnase it was around 4. In the presenl.study.wit,h {'inal concentration of ['AA in the mixture ranging from 0.02|r to0.:lo/,' or a ['AV protein ratio of 0.006 to 0.075 the optimum ratio to achieve the highest alkaline protease activity in the preeipitate rvas 0.069.These ratios were much lower than those ol'the above studies which could be due to the different natures of the proteases as well as the PAA.Boucher etaI. (1992)  used PAA with molecular rveight of 2110,000I), while the one used in the ;lresent study was i:l- 5MD.The molecular weight of PAA userl bv Oavgill et rr.l. (198'l) lvas not reportetl.
A similar ratio of l'AA"/protein (0.066) was userl by Reece (198ti) to separate alkaline protease from acidic protease of salmon viscera silage.No investigati<ln of the optimum ratio was reported in Reece's study.However, it was rellorted that the nlkaline protease activity in the precipitate was 82.7%o of the original, while that of aciclic proteases in the supernatant was 100%.The data for the acidic protease in Reece's sbudy were reported to be modified to take into account of the co-purification of the alkaline protease.however, no explanation of the modifications were mentioned.No data on the activity of alkaline protease in the supernatant or acidic protease in the precipitates were reported.PAA precipitation in Reece's study increased the puritv of the acidic and alkaline proteases by 1.6 and 6.8 fold' respectivelv, while in the present study' activities were 1.1 and 4.7 fold, respeetively.

CONCLUSI ON AND RECOMMENDATION
This study showecl that,llolvacrvlie acirls were able t,o {racl,ionate ;rroteascs {rom t:tlwl.ailray viscera and at the same time increasetl tho purity of the enzymes.It was clemonstratotl that the optimum fractionation condition wAs an trrldition of 2% (w/v) PAA to procluce 0.275% in the final mixture, followetl by pH adiusbment of supernatant containing acidic l)rotease ftr ;tH 4.0.antl of that containing alkaline protease t,o pH fl.ir.This treatment increasecl the qlecific activitv rrf itlkaline and acidic proteases by 4.7 and l.l fbld.respect' ively.Praetical protocol on the fractionation is necessary to be developetl so that the pnnedure can be applied in higher scale of fractionation.
Fu rth er p urifi cation followin g fractionation neecls to be investigated so that the enzymes t:an be p rop erly characterizetl.
concentration(o/o, wlv) in final mixture

Figure 2 .
Figure 2. Distribution of acidic and alkaline proteases activity after PAA addition A= Acidicproteases; B= Alkalineproteasest A= Supernatant(S,); O=Precipitate(S") molecular rveight ['AA was then tested in the present study.