MFR PAPER 1034 vents, u ually isopropyl alcohol, a nd several pilot plants and a few full -scale indu tria l pl ants have been constructed An enzymatic FPC process could (Ern t. 197 1). By olve nt e\.traction make profitable use of some of our FPC can be produced that is a bland. underutilized or latent resources . nearl y odorless. li ghtl y-colo red. water­ insoluble. but hi ghl y nutritive powder. FPC's pre pared by biological proce­ dures are usuall y mo re Aavorful and may have desira bl e functional proper­ Using Enzymes to Make ties. In ge neral the bi ologica l proce­ dures have not advanced beyond the Fish Protein Concentrates labo ratory o r small pilot plant stage. THE BIOLOGICAL METHOD The wo rk described in this paper MALCOLM B. HALE has been called the '" Bi ological Meth od " within the fish protein con­ centrate (FPC) program of the ation­ ABSTRACT a l Ma rine Fisheries Service. Biological proces ing st udies at the Coll ege Park Fish proTein CO ll cellTraT es having desirable JuncTional properTies call be pre­ Fishery Products Technology Labora­ pared by using selecTed enzy/ll es TO solubili::.e prOTein and release lipids. All en­ tory have been concentrated on the u e ::'Y IiUlTi c FPC process could /IIak e profiTable use oj sOlli e oj our IIl1denllili::.ed or of proteolytic enzymes to prepare 01- Im ellT resollrces. ubl e hydrolysates of fish protein. Th e relaTive aCTiviTies oj 23 cO lllll1 ercially ([I'ailab le proTeolYTi c enZY lll e prepar­ Fish proteins are complex molecules aTiollS aCTing 0 11 a fish proTein slibsTraT e were m easured. Pallcrea Tin , pepsin, lind consisting of chains of simpler mole­ papain had highes T aCTiviTy per UlliT COS T oJ ell <..Y lll e. cu les call ed amino acids lin ked togeth­ Soluble hydrolysaTes were prepared from red hake (Urophycis chuss) II sillg a er by peptide bonds. Proteolytic en­ l'arieTy oj en::.yme and digesTioll condiTions. ConcellTraTions oj Th e alllillo acids zy mes are also proteins and are struc­ Tr ypT ophan and hisTidin e ill Th e soluble producTs \I'ere criTical nliTriTioll ally and tured in such a way that they act as varied wiTh Th e pH oj hydrolysis. A soluble FPC ha ving a proTeill eJficiency raTio catalysts in the breakage of peptide equal TO ThaT oJ casein was prepared wiTh an alkalille baCTerial ell <..Y lll e aT pH 8.5. bonds by a process call ed hydrolysis. Pr odu c Ti on COS TS ha ve been eSTimaT ed Jor a soluble producT prepared Jro/ll Most of the shorter amino acid chains II'hole fish and a parTially soluble produ cT prepared frOIll presscake Through Th e broken off from the fish protein are lise oj ell zym es. water-soluble. Fish li pids (fat) are also released from the tissue during hydrolysis and can be physicall y sep­ INTRODUCTION FPC's may be produced by vari ous arated by cent ri fugation. processes to give products with differ­ Fish protein concentrates (FPC's) The enzyme systems employed in ent cost a nd properties. thus differe nt are made by processes which concen­ fish protein hydrolysis may be either appli cations. Most of th ese processes trate fi h protei n into a more stable the natural enzymes contained in the are described in a rece ntl y compiled form. As a form of animal protein, fis h (a uto lysis). purified commercial bibliography published by the Libra ry FPC's can supplement vegetable pro­ preparati o ns from variou biological of Congress (1970) . Most of the pro­ teins very effectively. Added in low sources (vegetable. animal. or micro­ ces ing method can be cl assifi ed as concentrations they can markedly im­ bial). or the enzymes supplied b, living chemical (solvent extractio n) o r bio­ prove the nutritive value of bread and cultures of microorgani ms. logical (enzymatic a nd microbi al) pro­ many other common foods by supply­ Development of a biological pro­ cedures. During recent years most ing certain essential amino acids defi­ cess was pursued because ( I) the efforts have involved the use of sol- cient in vegetable proteins. Produced product can have special properties from species of fish not normally used MaJcolm B. Hale is a member of (e.g.. olubility) w hi ch make it more as food , they are an inexpensive source the staff of the College Park suitable for certain appli cations than Fishery Products Technology is the o lvent extracted F PC. (2) it ma) of animal protein and could provide a Laboratory, National Marine profitable use for some of our presently Fisheries Service, NOAA, Col­ be possi ble to produce an acceptable underuti Iized or latent resources. lege Park, MD 20740. protein concentrate more cheapl,

15 hecau,c of thc addcd c ff cct o f na tive It, h c n /)- IllC, I hc rei at l v c effect IVl:nc.,\ o f pl: p' ln. fo r In , tance. wa., 10'" Wi th ACID ",holc fi,h hccau,c thl: nat lll: l:n / jlllc\ BASE arc not act lvc at p H 2 ",h lc h 1\ o ptl ­ IllUIll for pl:r'>ln Over a rerlod of time. Illany rroducl'> "'ere prl:ran;d In ~­ HYDROLYSIS VESSEL Itter hatchl:'" (I Igurl: I) U'>lng di ff erent en/) me, and rroce,,,ng condition., \erclgc ) leld, of dry ,oluhlc rroduct \\Ith each t\ pc or Cnl) mc arc li'll:d in T ahle I. Amino aCid anal),e, and BONES & SCALES Table 1 - Average Yi elds o f dry so lids from red hake so lubilIZ e d by proteolytiC enzymes

Y'e d dry l OIL SO u bles/ Ne Number fish °0 - Sid Enzyme(s 01 runs de'w'lallon UNDIGESTED SOLIDS AutolYSIS 100 + 0 4 PepSin 1 109 Papain 5 113., 0 8 BPN 4 117 +010 Bromelin FICin 3 123·03 Rhozyme P-ll 3 12 5 - 075 Pancreal n 7 138= 13 Alkaline pro eases 5 143- 10 SPRAY DRY PRODUCT, DRYER SOLUBLE FPC llmlteJ feeJlng 'tuJle Indicated that h)drol)\" at p H , ~I\lth an al~altne rrt1 teu..,e 01 8m illill IIIhlilil re,ulteJ fn Figure 1.-Basic process ou tline for enzymatic hydrolysis of fish . the he,t nutritlle lalue a\ Ilell a\ the hlghe ... t ) lelJ of product. pol) peptide~. and ~ome Iree amino through biological means than through HYDROLYSIS CONDITIONS acids. An alternate product. ea,tI) dl,­ solvent extraction. and (3) a biological AND AMINO ACIDS method with proper controls ",ould be per,iblc but onl) part lall) ,oluble. I, more suitable for u e on shipboard or prepared by elimina ting the centrifuga­ RaIl fl,h i~ h) Jrol) zed ~o that the for use in remote areas. tion step. Either product could be ~old actll It) of nat" e enz) me.., i cOlllblned A process has been developed for as a concentrated paste Illth elimina­ I\ith that of aJded enz)llle . In initial the preparation of a nutritious. totally tion of the dry i ng ~tep. ru ns the ti h 1\ a~ coo~eJ a.' a control Illea ure rrior to hydrol)~i~. Both water-soluble concentrate deri \ ed SELECTION OF ENZYMES from fish protein", hich may be u~ed ) ielJ and nutritive lalue \\ ere poor. to advantage in foods such as soups In early research the relati\e prote- n e'l.perlment shO\\ed that the degree a nd beverages. Although proc es~ing 01) tic activities of 23 commerciall) )f solubilization ach ie l ed \\ ith ral\ costs have been considered throu gho ut available enz) me preparations I\ere ha ke coulJ be obtained 1\ it h cooked the work. the major concern has been compared. A speciall) prepared fish hake onl) if si 'l. times as much com­ for improvement of the amino acid protein ub trate was used in sta ndard mercial e n z) me "'ere a dded . pattern and nutritive value of the sol­ hydrolysis tests of I-ho u r a nd 2.+- hour Solu ble h ydrolysates pre pa red fro m uble product. durations . Fici n was m ost active for a cooked ha ke ha d low nutritil e values The basic bi ologica l process is de­ I-hour period. O ver the lo nger h ydroly­ as the sole source of prote in in feeding picted in Figure 1. Enzymatic diges­ sis period the e nzymes pepsin. papain . studies w ith ra ts. The co ncentra ti o n o f tion of a whole fis h slurry with control and panc reatin ra nked hi ghest o n th e the essentia l a mino acid try ptopha n of p H and temperature is followed by basis of acti vity per unit cos!. These was quite lo w in the e pro ducts , Treat­ screening to remove bones and scales. results were de ta il ed in a previo us pu b­ ing ra w h ake unde r sli ghtl y alkaline Insoluble solids are separated by cen­ li cation (H ale. 1969). conditi o ns gave a good yie ld of pro ­ tr ifugation and a c larified hydrolysate When who le red ha ke ( Uropliycis duct conta inin g adequa te try ptopha n . is spray-dried (Figure 2) to yie ld a cllllss) was used the re la ti ve effecti ve­ The positi ve co rrela ti o n between the soluble p rodu ct consisting of peptides. ness o f the vari ous e nzym es changed level o f trypto phan a nd the yield of

16 Figure 2.-Spray drying a partially soluble hydrolysate.

olu ble solids is shown in Figure -l. T he requ ired yield of 13 percent (of fis h weight) can only be obtained with added enz) me. not by a utolysis a lo ne. Hydrolysi with pa ncreatin at pH 8 gave good yield of product containing adequate levels of tryptophan but the nutntl\e \alue \\ as unsatisfactory. This corre ponded \\ ith a IO\~ le \ el of histidine. an amino acid which is essential for infants (or) o ung rats) a l­ th ough not for ad ult~ . Whe n an al l.. a­ lin e prote a~e de ri\ ed from th e bacteria B ..\II/1lili.1 \\' a~ acquired a nd tested at pH 8.5. both the hi tidine level a nd nutriti ve value of th e product were ~ig­ nificantl ) impro\ed. Pancreatin al~o ga\ e better re ult~ a t pH .5 th an at pH EVALUATION RUNS

Three enz) matlc proce se \\ere e \ aluated in replicate run using red hal..e. H ydrol),e~ with pancreatin. AI­ l calase . and b) autolysis were tested bj three S-Iller (1.3 gall on) batches plu a 20-gallon batch for each proce~s . Three processes were also tested in run with the fatty fish alewife (AlOIa p.\elldol/(/ren~III) uSing autolysIs and the all..aline protease Alcalase. A\erage ) leld and chemical composition~ of raw fish and of the hydrolysates pre­ pared from hal..e and alewi fe b) each of t\\ O proces~es are li~ted in Table 2. Protei n efficiency ratios (PER \) were determined by rat feeding trials with the hydrolysates of hal..e and a le­ wife a nd a re listed in Table 3. The products prepared from hal..e were in­ ferior to casein as a sole source of pro­ tein. but were tati tically equivalent to casein as a supplement to wheat flo ur. All o f the products prepared from alewife were sta ti stica ll y equi va­ lent to casein as a sole source of pro­ tein. T h us, we o bta ined a tota ll y ~o l ­ uble hydrolysate from alewife with

I Manufaclured by NOVO I ndustrles Reference 10 Irade names does nol Imply endorsement by Ihe Nallonal Marine Fisheries Service . NOAA

Figure 3.- Experimental system u s ed f or fish protein hydrolysis in 5-liter batch . Table 2. - Yiel ds and compos ition of hydrolys ates and raw fish . duc ti on costs fo r th e t\\ 0 t) pes o f lem Red hake hydrolysates Raw hake Alewife hydrolysates Raw alewife products thro ugh use o f a recentl y Enzyme Pancreatin Autolysis Alcalase Alcalase de\ eloped computer progra m (Almen­ 0 Concentrat on 0.5 0 0 .5°/0 0 .3% Type of product Soluble Whole slurry Solu ble Wh ole slurryl as et al.. 197:2l. Assumi ng a pl a nt processin g :2 50 to ns per day o f fi sh Number of runs 4 8 samples 3 3 3 samples 0 Average yield' 143 0 17.9°0 13 . 0~lo 18 7°~ costi ng 1.5c/lb a nd o pe rating fo r :2 00 days/)ear. th e estitll ated producti o n AvQ "0 Pretf:ln 77 23 77.85 15.89 7601 65.24 1716 °0 MOisture 4.44 4.02 79.09 4.89 421 71 .07 costs are 17 c/lb fo r a totall ) solu ble °0 Ash 1602 673 273 1652 9.32 274 °0 Fat 3 016 1241 306 0.21 2507 835 F PC a nd l:2c/ lb for the parti all ) sol­ uble presscake product . F or a :20 per­ I Prepared from alewife presscake , Dry product as weight percent of raw fish cent proAt afte r ta:-.e o n tota l capita l I Total I plds by the SAK method (Smith Ambrose and Knobl. 1964) in \estment. th e estimated selling Table 3.-Nutritive quality of enzy matic hydrol ysates of red hake and alewife .' prices are J8.-Ic/lb of solu ble r-pc a nd :26 .7c/lb of presscake product. Material Enzyme Product Protein efficiency ratio hydrolyzed added solublilly Actual ~'c, of casein Raw ha e Pancreatin Total "2.89 ± 0.07 822 NOAA TECHNICAL REPORT Raw hake (AutolYSIS) Partial 287 ± 007 817 Raw alewife Alcalase Total 344 ± 006 97.2 Alew fe presscake Alcalase Partia l 3.40 ± 008 960 The research su mmarized in this paper has been described in more de­ Each feeding sample was a blend of the spray-dried products from Ihree 5-llter hyd rolYSIS batches - Averaq plus standard error of the mean tail in a technical report \\hich also includes a literatu re survey of biologi­ cal method~ for F PC producti o n (H ale. 197:2). 1 0 combined eith er \\ ith a ~peci a l prop­ ert). such as solubil it). o r \\ ith pro­ cess simplicit) a nd a relati\el) 10\\ LI TERATURE CITED production cost. Almenas. K . K .. L. C. Duri ll a. R. C. rn,t The most promising immediate ap­ J . \Y. G e mr}. M . B. H ale. and J. M . M ar­ • chello. 1972. Engineering economIc model plication for biologica l FPC's is as for fhh protein concentralton procc,'c~ . . 5 . Dep . Commer .. NOAA Tech . R~p . a partial replacement for milk in the NMFSCIRC 367.175 p. diets of \\eanling cahes. The combina­ Ern ... l, R. c.. J r . 1971. !'- PC: the MFS c ... - penmem c~ demon,tration plant procc" . tion o f good nutriti\ e \alue and dis­ Com mer. FIsh. Rev. 33(2):22-28 . persibilit) ma ke Ash protein h)drol) ­ H ale. M. B. 1969. Relative acti\ ities of com­ merClally-a\ ailable enz\ mes In the hydro l}­ sates \\ell su ited for this use. Fish auto­ sis of fish protein. Food Tech no!.- 23( I) : I)sates ha\e been u~ed as milk replac­ 107-110. 197 2. Mah.lng fi,h protein conccntrate, b} ers in Fra nce. The presscake hydrolysis cnz) matic h) drolysis . A status rcpon on research and ,>ome processc.., and products process abo sho\\\ promise of econ­ ,tudlcd b} I F . .5. Dep. Commer .. o mic feasibilit \. . OAA Tech . Rep. NMFS SSR F 657.32 p . Figure 4.-Tryptophan concentration in soluble Illtlh. P .. Jr" i\1. E. AmbHhe. ,: nd G . M . solids versus percent yield of solubles from en­ Process fkl\\shcets ha\e been de­ K nob!. Jr. 196-1 . Improved rap!d Illctlwd zymatic "ydrolysls of red hake \eloped fo r e nz) matic processes using for determining tota l lipid, in .hh mc,tl. Commer. FI\h. Rc\ . 26(7) : 1-5 ra\\ whole fish and fish presscake. .5. Llbrar} of Congress. 1970. Fish prote in th.: J.: tr.:J PER 1.:\.:1 \\hlch \\e haJ ~ I a terial balances obtained e \ pe ri­ concentrate: A comprehensive biblio­ graph}. U.S. Llbr. Congr., Sct. T echno!. h.:.:n un thl.: tl' r.:a.:h \\ ith hake menta ll ) \\ere used to estimate pro- Div" Spec. Biliogr. Sect .. 77 p. PRODUCT USES AND COSTS

I III II.:J Il'"J r':~':Mch ~tuJte, ha\ e he.:n Cht n.:J ,'ut \\ Ith the en/\ matlc 11\"1'\11\ .tt.:~ "I \\hol.: fi\h f-or mo~t MFR Paper 1034 . From Mari ne Fish eries Review, Vol. IlIlIJ "pplle.ttl,'n, a hlanJ pn'duct " 36, No. 2, February 1974. Copies of this paper, in limited dL"" .. bk and a gl'l,J 'l,lubl.: protein numbers, are available from 0 83, Technical Information Illi t"~ III e.trh,'n.tt.:J hL\erag.:, \\ ould DiVISIon , Environmental Science Information Center, Cl l ll'l11.tn" .t pr.:tlllutll pric.: I la'Llr " NOAA , Washington, DC 20235. prllh.lhh the m.tJl'r 1II1"'h.:J rrl'bl.:m hut I P<. Ir"tll \\ hllk Ii h c"uld h.: ,,1' fl.tl \ .due Il'l u'e, In \\ hteh glll'd nu­ tlttl\L' .Ill" 'UPP1L'tllL·nt.tI \alu.: .11'':

18