Digestive Endo-Proteases from the Midgut Glands of the Indian White Shrimp , Penaeus Indicus ( Decapoda : Penaeidae ) from Kenya

In order to provide information on the digestive capacity of marine crustacea of mariculture potential in Kenya with a view to aiding the development of suitable feeds to support the industry, a biochemical survey was made of enzymes of extracellular digestion in the Indian White shrimp, Penaeus indicus. Results showed midgut gland endo-proteases in wild adult shrimp from the Kenya coast to have optima between pH 7.2 and 8.5 (Trypsin pH 7.5-8.0, Chymotrypsin pH 7.2-7.8, Elastase pH 6.8-8.5) with maximum specific activities of 101-408, 37-516, 70-90 Units mg protein min for trypsin, chymotrypsin and elastase respectively. There was no pepsin. The North Sea Norway lobster, Nephrops norvegicus, was investigated to a lesser extent and found to have much lower trypsin activity than the shrimp and no chymotrypsin. In addition to the cited serine endo-proteases, significant activity in the shrimp was thought to originate from non-serine proteases. This situation may differ from other shrimps in which serine endo-protease activity, especially trypsin, is dominant. Diphenylcarbamyl chloride (DPCC) and 2-Nitro-4-Carboxyphenyl N,N-Diphenylcarbamate (NCDC) inhibited chymotrypsin but not trypsin, Soybean Trypsin Inhibitor (SBTI), BowmanBirk Chymotrypsin-Trypsin Inhibitor (BBSTCI), N-TosylL -Phenylalanine Chloromethyl Ketone (TPCK), 4-(2-Aminoethyl)-Benzenesulfonylfluoride Hydrochloride (AEBSF) and N-Tosyl-LLysine Chloromethyl Ketone/1-Chloro-3-Tosylamido-7-AminoL -2-Heptanone Hydrochloride (TLCK) inhibited both, while Phenyl Methanesulfonyl Fluoride/ Phenylmethyl Sulfonyl Fluoride/ α-Toluenesulfonyl Fluoride (PMSF) and Ovomucoid Trypsin Inhibitor (Ovomucoid/ OTI) precipitated shrimp homogenate. The effect of the former was inferred from the action of AEBSF which together with TLCK inhibited shrimp trypsin more than chymotrypsin. In contrast, TPCK inhibited shrimp chymotrypsin more than trypsin. These results indicate that relying on imported commercial feeds, usually developed for other species or strains of farmed shrimp in other parts of the world, may not only be too uneconomical but may not provide adequate nutrition to local animals if not efficiently digested. There is, therefore, greater need and urgency to establish detailed enzymic profiles and digestive capacities of locally cultured fin and shellfish. Such studies should parallel those prospecting for suitable feed ingredients while developing local capacity for feed technology.


INTRODUCTION
Declining shrimp fishery yields and increasing demand has stimulated the expansion of shrimp farming in the last 40 years (Weidner & Rosenberry, 1992).However, Africa has lagged behind in the exploitation of shrimp resources when over 35.2 million Africans rely on fisheries for a livelihood (Seki & Bonzon, 1993).Other than increasing local protein reserves, shrimp farming offers a way of exercising greater control over shrimp production.Greater income is possible over a long period with proper management to avoid environmental disasters experienced in some Latin American countries in the initial years (Currie, 1994).Because shrimp farming also enables a wider spread of the production base, employs more people directly and indirectly through associated industries, it can be a long term economic diversifier, especially since most African countries often rely on single-item sources for foreign exchange: cash crops, minerals and tourism or combinations.Unfortunately, the benefits of shrimp farming may not be realised as the indigenous technical expertise and the necessary research base required are not well developed.Few African countries (e.g.Madagascar, Mozambique, Guinea and Kenya) have attempted commercial shrimp farming with limited success.Species utilized include Penaeus indicus, P. semisulcatus, P. monodon, Metapenaeus monoceros, P. notialis and Parapenaeopsis atlantica (Garcia & Le Reste, 1981).However, there are no local, commercially available, specific artificial feeds for these species.The present study was undertaken to provide information on the digestive enzymes of P. indicus for the future development of appropriate feeds for its culture in Kenya where it constitutes 95% of the shrimp catch.Activities and pH-profiles of midgut gland enzymes important in digestion were investigated.Accounts of carbohydrase and general proteolytic activities were given in Omondi and Stark (1995, 1996, 2001).The present paper concentrates on the pH-response of endo-protease activities of wild shrimp from Kenya.The North Sea Norway lobster, Nephrops norvegicus, was used in some assays to indicate the likely situation in spiny lobsters in Kenya.

Sample collection
The over 100 wild adult P. indicus used in this study were caught by local fishermen from the areas around Tsunza beach and Port Reitz on the Kenya coast.Fifty adult Nephrops were obtained from the Firth of the Forth estuary on the east coast of Scotland.The animals were frozen in liquid nitrogen and then stored in a deep freezer at -22.4°C until assayed.Some lobsters were kept for several months in recirculating sea water (35 ppt) aquaria held at 15°C and fed on a shrimp diet (Omondi & Stark, 1995) indicating that they could be easily cultured.

Preparation of Enzyme Extracts
Homogenate was prepared as in Omondi & Stark (1995) but samples from N. norvegicus were diluted 1 / 5 -1 / 10 w/v because of very low enzyme activities.After centrifugation, fat mats were removed by glass wool filtration.Enzyme extract was prepared by filtering homogenate supernatant (38,000 x g) through Whatman glass micro fibres (1µM, BDH).

Estimation of Enzyme Activities
Enzyme numbers are as given in Webb (1992).Protein, pH, and absorbance measurements as well as buffers and statistical analyses were as described in Omondi & Stark (1995).
To corroborate the results of the TAME assay, the chromogenic substrate BApNA (0.001 M) was used and the increase in absorbance over a 10minute period measured at 410 nm.The digest mixture (DM) consisted of substrate (2.0 ml), Tris buffer (1.3-1.4 ml) and extract (0.1-0.2 ml; 1 / 5 [lobster], 1 / 40 [shrimp] w/v).The final CaCl 2 concentration (0.02 M) was slightly higher than that used with TAME (0.015 M).The use of TEA buffer (0.2 M, CaCl2 0.02 M, pH 7.8) and a more concentrated substrate solution (0.004 M) was tried (405 nm, Geiger & Fritz, 1984) and abandoned because BApNA formed a milky-white suspension in water that stuck to the glass container.There was also considerable non-enzymatic substrate autolysis.As the BApNA Km for trypsin activity in the H/P of P. indicus is 0.0003 M (Honjo et al, 1990), there was excess substrate in the present study.Following the method of Erlanger et al (1961), BApNA was completely dissolved in a small quantity (1.0-2.0 ml) of DMSO (Dimethyl sulfoxide) before the addition of buffer or distilled water to make a clear solution which kept well for at least one week at 30°C.Self-hydrolysis set in earlier and to a greater extent in buffer (pH 7.7-8.1)than in water (pH 5.6-6.2).
N-CBZ-L -Ala-pNPE(N-Carbobenzoxycarbonyl Lalanine p-nitrophenyl ester) (1.0 ml, 0.009M, 0.003M DM) was suspended with extract (0.5 ml) and buffer (2.0 ml), incubated for Ω -1 h and p-NP released measured at 420 nm (Geneste & Bender, 1969).In fixed time assays, K 2 CO3 or KHCO 3 (1.0 ml, 0.1 M) was added to supernatant after centrifugation of the DM to enhance the colour of product but in initial velocity measurements, it was not possible to add the alkaline solution.

Enzyme Inhibition Studies
The following inhibitors were used in the present study : To distinguish between different activities, inhibitors were pre-incubated with extracts before assays.AEBSF, SBTI, BBSCTI and OTI were dissolved in water, TLCK in ethanol and TPCK in methanol and water.DPCC was dissolved in acetone or propanol while NCDC was used as a suspension in water.PMSF was dissolved in ethanol but did not completely dissolve in propanol.The PMSF-alkanol solution formed a viscous silvery-grey non-particulate precipitate with extract and a suspension in water was employed.The extent of inhibition was expressed as a percentage relative to the specific activity of control digest mixture (DM) containing no inhibitor.

Column Chromatography
Glass columns (90 cm x 2.5 cm, Wright Scientific Ltd, England), packed with Sephadex G-200 previously equilibrated and degassed in phosphate buffer (0.02 M, pH 6.9), were eluted with the same buffer at 14-20 ml h -1 and fractions (6.3 ml) collected in polyethylene tubes using a Biorad model 2110 fraction collector (Biorad Laboratories Ltd., Hemel Hempstead, Herts, England).For enzyme activity assays of fractions, the extract volume was increased and the incubation time also extended in some cases.

Pepsin-like Activity
After an incubation period of 60 minutes, the wild P. indicus midgut gland/hepatopancreas (H/P) extracts indicated the lack of pepsin activity and thus an absence of this enzyme from the digestive gland of the shrimp (see Fig. 1).

Trypsin-like activity
An analysis of relative enzyme activity on TAME and BapNA as illustrated in Fig. 2 shows that the optimum pH for P. indicus H/P trypsin was around 7.8-8.0,with significant hydrolysis throughout the pH range 7-8.The highest specific activity with TAME was 103.704±14.800U mg protein -1 min -1 .Whereas there was some apparent activity with TAME in strongly acid media (pH 1.5), there was none above pH 10. OTI severely reduced shrimp TAME trypsin activity but was not as effective against Sigma trypsin.The importance of calcium for Sigma and shrimp homogenate trypsins was made obvious by the significant decrease in rates in digest mixtures (DMs) lacking additional CaCl 2 (0.011 M) even when the substrate concentration was doubled.BApNA was a much more convenient and sensitive way of assaying trypsin activity and was stable for several weeks at 30°C.The pH optimum was 7.8-8.0.Supplemental calcium was not as crucial as with TAME and no significant differences were found with shrimp homogenates at pH 7.0.However, at pH >8.5, no activity was observed probably due to unfavourable pH and low calcium amounts.As with carbohydrase (Omondi & Stark, 1995) and total protease activities (Omondi & Stark, 2001), N. norvegicus had very low BApNA trypsin activities of 1.446-1.652± 0.056

Chymotrypsin-like activity
BTEE was problematic due to haziness and precipitation from calcium phosphate and shrimp extract.The pH-activity profile is shown in Fig. 3 where the optimum was pH 7.8-8.2.SApNA was a better substrate with the optimum pH between 7.2-8.0.(see Fig. 3).Use of Triton resulted in higher activities and the molarity of Tris buffer was significant as 0.1 M was better than 0.005 M. It is possible that observations made by some authors that serine protease activities in Tris buffer was less than in citrate phosphate (Schwert & Tanaka, 1955) may depend on Tris concentration.Additional calcium did not significantly affect chymotrypsin activity at all pH values.SApNA chymotrypsin activity was not found in the lobster.
Attempts to distinguish chymotrypsin and trypsin activities using specific substrates and inhibitors were partly successful.NCDC at 1624 µM (614.3µgml -1 ) did not significantly inhibit chymotrypsin activity determined with SApNA.This chemical is reportedly a competitive inhibitor/ substrate of mammalian chymotrypsin (Erlanger & Edel, 1964) but assays using it as a chromogenic substrate in the present study did not show measurable activity.However, its chloride, DPCC, was a very potent selective inhibitor of chymotrypsin activity as expected (Erlanger & Cohen, 1963, Erlanger et al, 1963) inhibiting up to 99 % of shrimp activity within 3 h at optimum acidity.Almost complete inhibition of mammalian Sigma chymotrypsin (Bovine pancreatic chymotrypsin Type IV, 52 BTEE U (µmoles) mg -1 min -1 ), occurred within 30 minutes.In contrast, the hydrolysis of BApNA by shrimp homogenate was hardly affected even after long-term pre-incubation with these inhibitors.
The inhibitor TLCK (464 µM; 171.4 µg ml -1 ) failed to inhibit chymotrypsin activity whereas TPCK inhibited 68 % of that activity.On the other hand , PMSF, did not inhibit either activity significantly.This inhibitor, when dissolved in  (Fahrney & Gold, 1963, Gold & Fahrney, 1964, Glass, 1989), produced a silvery grey/white viscous precipitate when added at 3278 µM to homogenate and was used suspended in water.In contrast, within 10-20 minutes, BBSTCI caused a 99-100 % inhibition of chymotrypsin and trypsin activities at concentrations as low as 40µg ml -1 but it did not inhibit Sigma trypsin.Finally, AEBSF inhibited 97 % of shrimp trypsin activity but only 27 % of the chymotrypsin, while SBTI inhibited both activities completely.

Elastase-like activity
The two synthetic peptides tended to show less alkaline optima.A longer incubation with SAAApNA resulted in greater activity without apparent product inhibition (Fig. 4).SAAApNA had a very sharp peak at pH 8.0 while N-CBZ-L -Ala-pNE had a very broad top from pH 5.6-8.2 but with peaks at pH 6.9 and 7.9-8.2(Fig 5).Both the modified natural substrates had an optimum at pH 8.2-8.5 and there was no activity above pH 9.3.N-CBZ-L -Ala-p-NE was a rather unstable substrate, significantly undergoing rapid hydrolysis even in the absence of extract.The other synthetic substrate was used because, although its insolubility in aqueous media made it inconvenient, initial reaction velocities could be followed in cuvettes in the spectrophotometer.The natural substrates could only be used for fixed time assays to estimate the initial velocities and to complement studies using SAAApNA.

DISCUSSION
The lack of pepsin in shrimp is consistent with the absence of a true (acidic) stomach in which pepsin activity is found.Many other crustaceans examined have also failed to reveal the presence of pepsin activity.Protein digestion in these animals appears to be undertaken by serine endoproteases aided by exoproteases with the possible involvement of cysteine proteases.
In both UV assays for serine proteases in the present study, substrates were in excess.For trypsin, calcium was present in sufficient concentration (0.011 M) as recommended (Hummel, 1959) and similar quantities were present in the chymotrypsin digest mixtures (DMs) where Tris was used.Supplemental calcium did not significantly affect the BApNA-hydrolysing activity of shrimp homogenate, a finding that agrees with the observations of (Honjo et al, 1990).The homogenate may contain adequate calcium unlike purified commercial mammalian enzymes that are salt-free.Since TAME is not hydrolysed by chymotrypsin (Hummel, 1959), it can be concluded that the activity measured is due to trypsin.Under the conditions of reaction for chymotrypsin, BTEE is relatively resistant to trypsin hydrolysis (Hummel, 1959), so it can be concluded that the shrimp homogenate also contains chymotrypsin-like activity.Results from the chromogenic substrate assays confirm this.Some authors (e.g.Erlanger et al 1961) regard BApNA as the most sensitive substrate of trypsin.One of the non-trypsin enzymes able to hydrolyse it is papain, which does not occur in shrimp homogenate.Whereas there may be an appreciable amount of BApNA hydrolysis by non-trypsins, the hydrolysis of SApNA was almost exclusively by shrimp chymotrypsin.This is further evidence for the existence of shrimp chymotrypsin.Serine proteases are now known to be significant in penaeid digestion of proteins (Hernandez-Cortes et al, 1997).In the present study, results of experiments using specific inhibitors were confirmatory to a large extent.
DPCC inhibited almost all (99 %) of the chymotrypsin activity but did not alter the hydrolysis of BApNA by shrimp homogenate indicating that the two activities were different, that both substrates are fairly specific for each activity and that DPCC is an effective, specific inhibitor of chymotrypsin.The rate of DPCC inhibition of human trypsin is reported to be 80 times slower than that of chymotrypsin and reactivation of the latter with nucleophilic reagents such as indole and isonitrosoacetone is also twice as fast as that of trypsin but complete reactivation occurs for both after 18h (Erlanger & Cohen, 1963).In the present study, shrimp trypsin recovered after 20 h but chymotrypsin did not.
Furthermore, as expected, TLCK failed to completely inhibit chymotrypsin activity confirming that the activity involved in the hydrolysis of SApNA was non-trypsin.Inhibitions of upto 55 % of mammalian trypsin and 71 % of lobster H/P protease activity using this inhibitor have been observed (Glass, 1989).At 500 µg ml -1 , a 83 % inhibition of P. indicus H/P trypsin activity is observed (Honjo et al, 1990).
PMSF dissolved in ethanol or propanol precipitated the homogenate, hence was used suspended in water and not surprisingly, did not inhibit either activity significantly contrary to the findings of others where 50-100 % inhibition using concentrations as high as 3000-5000 µM were reported (e.g.Glass, 1989, Honjo et al, 1990).PMSF may consequently be a less effective inhibitor of shrimp activity.
Inhibition (97.1 %) of P. indicus H/P trypsin activity observed with SBTI (200 µg ml -1 ) by Honjo et al (1990) is consistent with findings from the present study which also confirm observations that no inhibition of shrimp trypsin activity occurs with TPCK (500 µg ml -1 ).Whereas 25 % propanol reportedly elevates P. indicus H/P trypsin by 4 times at 30°C and 177 times at 4°C (Honjo et al, 1990), in the present study, ethanol, methanol, acetone and propanol concentrations < 5% did not significantly increase enzyme activities.
The estimated molecular weights of shrimp trypsin and chymotrypsin (14.5-29 kD) from the present study are within the range for decapod trypsins (see Vonk & Western, 1984), P. setiferus H/P trypsin (Gates & Travis, 1969) and purified mammalian pancreatic trypsin and α-chymotrypsin (Hess, 1971;Keil, 1971) but are less than one of the two H/P trypsin iso-enzymes found in P. indicus of 36 kDa using Mono Q FPLC and SDS-PAGE (Honjo et al, 1990).However, Sephadex G-100 gel filtration revealed a molecular weight of 18 kDa (see Honjo et al, 1990), similar to that of the present study.In contrast to the present findings, these authors also showed that their enzyme had a broad pH optimum (6.5-11.0)and was most active at 45°C.The stability of the enzyme in alkaline pH has been confirmed by the present study.However, contrary to the findings of Honjo et al (1990), in the present study, there was no shrimp trypsin activity that was almost completely inhibited by PMSF.Such discrepancies may be due to the presence of at least 7 trypsin-like enzymes in P. indicus (Honjo et al, 1990).Indications from the present study are that there may also be several variants of chymotrypsin molecular weights.
Decapod and mammalian trypsins have broadly similar molecular weights, substrate specificities and general responses to inhibitors, but differences include greater temperature and pH tolerance and resistance to autolysis by the crustacean enzymes (see Vonk & Western, 1984).The pH optimum in the present study for P. indicus H/P trypsin (7.8-8.0) was not very different from that (pH 8.0-8.5) in crabs (Galgani & Nagayama, 1987).The BApNA trypsin pH optimum in the present study indicates that shrimp H/P trypsin has a similar pH optimum to the mammalian enzyme (Hummel, 1959, Erlanger et al, 1961).
The presence of chymotrypsin in P. indicus indicates that this endoprotease is of wider occurrence in shrimps than previously thought (Gates & Travis, 1969, 1973).The pH optimum is similar to that of the chymotrypsins of P. monodon, P. japonicus, P. penicillatus, Metapenaeus monoceros and Macrobrachium rosenbergii (Tsai et al, 1986).
The last serine endo-protease determined in the present study had the most alkaline pH optimum of all assayed enzymes, extending to pH 8.5.This is slightly lower than the pH optimum (8.8) of mammalian pancreatic elastase EC 3.4.21.36 but is still within the optimal pH range of leukocyte elastase EC 3.4.21.37 (Geiger, 1984b) and all the major shrimp H/P proteases.It must therefore contribute significantly to protein digestion as it is an enzyme known to hydrolyse a wide range of proteins (Clark et al, 1985b).In agreement with Glass (1989) and (Clark et al, 1985a), the pH optima varied with the different substrates.In P. kerathurus, elastase could not be demonstrated with unmodified Elastin but collagenase was present (Galgani et al, 1984).The shrimp elastase partly behaves like that of the Dover sole with respect to some substrates (Clark et al, 1985b).The pH optima with the two artificial substrates were almost identical between the shrimp and fish (pH 8.1-8.2) but with natural substrates, including undyed Elastin, fish pH optima were more alkaline (8.3-9.8)than shrimp ones <pH 8.5.Fish elastase was inhibited by PMSF (35-92 %), TPCK (16-23 %) and TLCK (12-41 %) and had a molecular weight of 19.5 kDa on Sephadex G-200.Unlike in fish where there exists a wide discrepancy in pH optima between the synthetic and natural substrates (Clark et al, 1985b) possibly indicating different enzymes, the present study found a lesser difference and may be measuring the activity of one enzyme.The minor discrepancies may be due to the difficulties associated with working with these insoluble substrates (Clark et al, 1985a(Clark et al, , 1985b)).More studies of crustacean, especially shrimp, lobster and crab elastase activity are necessary.
In the present study, the residual activity observed in shrimp in the presence of selective inhibition of the three serine proteases indicates the presence of other types of proteases, probably cysteine ones.It would appear that the latter may be more important in the lobster (low total and trypsin activities and no chymotrypsin) while the former dominates shrimp dietary protein digestion.This differential importance of serine and cysteine proteases between crustacean groups has been observed by several authors (e.g.Ceccaldi, 1998;Garcia-Carreno et al., 1997;Hernandez-Cortes et al., 1997;Le Moullac et al., 1997).

CONCLUSIONS
In view of the recognition of the need for capacity building in aquaculture feed technology and nutrition in sub-Saharan Africa (FAO, 2001), the present study has demonstrated considerable proteolytic activity in P. indicus within the physiological pH (7-8) range and active at natural ambient water temperatures of the shrimp.Until locally formulated feeds are available, high-protein diets developed for other cultured shrimp species elsewhere may be tried as long as they are economically affordable and justifiable.
It may not be necessary to supplement shrimp enzyme levels by using live or freshly killed prey organisms.Organisms with much lower enzyme levels such as some bivalve molluscs and asteroidea may be tried as protein and mineral/ vitamin sources.
High and varied protease activity may preclude the use of supplemental amino acids in artificial diets (Lee et al, 1980) which reduces the costs of manufacturing these diets.However, such feed prices may still be unaffordable to prospective African shrimp farmers and efforts must be made to develop alternative aquafeeds using local ingredients.Examples might include aquatic floral resources like the currently problematic, but abundant, Lake Victoria water hyacinth (Eichhornia crassipes) and marine seaweeds of the East African coast.
The high enzymatic activity in P. indicus may be a source of commercial enzymes, which could be extracted from shrimp waste.By this further processing of shrimp after removal of conventional eatable parts, environmental pollution may be reduced.
Nephrops, and probably other lobsters, when cultured, may require considerable enzyme input from prey organisms with high enzyme activities such as other crustacea, echinoidea, polychaetes and cephalopods.Its diet may also need simpler food components such as dipeptides, amino acids and disaccharides.This means that artificial diets made from such components, and rich in protein precursors or protein, may be that much more expensive, limiting lobster culture unless final market prices can offset the high cost of production due principally to the high feed costs.

Fig. 1 .
Fig. 1.A time-course activity profile of mammalian (Sigma) pepsin with Ace-Phe-Dit as substrate at 30°C.Values are means ± Standard Errors of means

Fig. 2 .
Fig. 2. pH-activity profiles of digestive gland trypsin activity in wild P. indicus at 30°C assayed with TAME and BApNA as substrates.Values are means ± Standard Errors of the means

Fig. 3 .Fig. 4 .
Fig. 3. pH-activity profiles of digestive gland chymotrypsin activity in wild P. indicus at 30C assayed with BTEE and SApNA as substrates.Values are means ( Standard Errors of the means