Eur. J. Biochcm. 202, 485-491 (1991) Q FEBS 1991

Geotrichum candidurn produces several lipases with markedly different substrate specificities Christopher M. SIDEBOTTOM, Emmanuelle CHARTON, Paul P. J. DUNN, Gary MYCOCK, Christine DAVIES, Jane L. SUTTON, Alasdair R. MACRAE and Antoni R. SLABAS Unilever Research, Colworth Laboratory, Colworth House, Sharnbrook, England (Received April 24, 1991) - EJB 91 0535

We have purified and examined the substrate specificity of four lipases from two strains of the mould Geotrichum candidum, ATCC 34614 and CMICC 335426. We have designated the lipases I and I1 (ATCC 34614), and A and B (CMICC 335426). The enzymes are monomeric and have similar molecular masses and pl. Thus, lipases I and I1 have native molecular masses of 50.1 kDa and 55.5 kDa, and p l o f 4.61 and 4.47, respectively. Lipases A and B are very similar to lipases I and I1 with native molecular masses of 53.7 kDa and 48.9 kDa, and p l of 4.71 and 4.50, respectively. Treatment with endo-P-N-acetylglucosaminidasecaused a reduction in molecular mass of approximately 4.5 kDa for all four lipases, indicating that these enzymes are glycosylated. Western blotting shows that the lipases are related. However, lipase B from CMICC 335426 shows a remarkable specificity for unsaturated substrates with a double bond at position 9 (cis configuration), and this specificity is not exhibited by the other three lipases. No lipase of this unique specificity has previously been purified to homogeneity. Structural studies using these four lipases should allow insight into the molecular basis of this remarkable specificity.

Lipases (glycerol ester hydrolases) catalyse the hydrolysis of ester linkages in lipids with the release of the constituent alcohol and acid moieties. They act at the interface between a substrate phase, which has a very low solubility in water, and an aqueous phase in which the enzyme is dissolved [l]. These enzymes can differ considerably in their positional specificity for the hydrolysis of fatty acid from a triacylglycerol backbone [2, 31. Advantage is taken of this, industrially, in the production of cocoa butter equivalents from cheap starting materials [4]. In addition to positional specificity, lipases have been reported to have specificity for the fatty acids which are hydrolyzed. The most extreme example of this is an extracellular lipase isolated by Alford and Pierce from a strain of the mould Geotrichum candidum which has been reported to have a specificity for unsaturated substrates [5].This enzyme has not been purified and the original strain was thought to be no longer in existence [6]. However, in 1969 our laboratory received a strain of G. candidum from Dr Alford, and this has now been deposited at the Commonwealth Mycological Institute Collection as CMICC 335426. Tsujisaka et al. have purified a lipase from G. candidum ATCC 34614 to homogeneity [7], and this does not appear to show the same substrate specificity as the enzyme isolated by Alford and Pierce [8,9]. Several reports have appeared on the specificities of crude lipase preparations from G. candidum Correspondence to Dr P. P. J . Dunn, Unilever Research, Colworth LdbOratory, Sharnbrook, Bedford, MK44 1LQ, England Abbreviation. Endo H, endo-P-N-acetylglucosaminidase. Enzymes. Lipase (EC 3.1.1.3); Staphylococcus aureus V8 protease (EC 3.4.21.19); Streptomyces plicatus endo-8-N-acetylglucosaminidasc (EC 3.2.1.96).

[ 2 , 6, 10-241, but none show a very strong specificity for unsaturated substrates. This is not too surprising, when one considers that several extracellular lipases may be secreted by one organism, as shown for G. candidurn [15, 161, Candida antarctia [17], Rhizomucor miehei [I81 and Penicillium cyclopium [19], and these lipases could possess different substrate specificities. Specificity studies of crude extracts can therefore be complicated by the presence of more than one lipase. Sugihara et al. [16] have, however, reported the purification of two lipases from G. candidurn ATCC 34614, but the two enzymes appear to display the same substrate specificity towards a variety of triglycerides. To appreciate the relationship between substrate specificity and lipase structure, it is important to have appropriately detailed X-ray crystallographic data. There have been few reports of X-ray data from lipase crystals although the ‘major’ extracellular lipase from G. candidurn ATCC 34614 (lipase 1) has been crystallized and a 0.6-nm electron-density map is available [7, 201. It has been reported that 0.25 nm resolution has been obtained for this enzyme [21]. cDNA clones for this lipase and a ‘minor’ extracellular lipase have recently been isolated and their nucleotide sequences determined [22, 231. Lipases I and I1 are 84% identical at the amino acid level and these enzymes show some similarity (45%) with an extracellular lipase from Candidu rugosa [22- 341. There is no amino acid sequence similarity between this lipase and other lipases apart from a possible active site motif: Gly-Xaa-Ser-Xaa-Gly (22- 241. In view of the unique specificity for unsaturated substrates reported for a lipase secreted by G. candidum [5], we have endeavoured to purify extracellular lipases synthesized by ATCC 34614 and CMICC 335426 strains and to characterize

486 their substrate specificity.In this paper we report on the purification and characterization of four lipases and their effect on a variety of substrates. One of the lipases secreted by CMICC 335426 has a unique specificity for unsaturated substrates and probably represents a major component of the lipase activity in the crude extracts investigated by Alford and his co-workers [5, 10, 1I].

eluant equivalent to 1 mg protein. These were bound to the column, equilibrated in 10 mM sodium acetate, pH 5.1, and eluted with a gradient to 0.1 M NaCl in 60 min, at a flow rate of 1.0 ml/min. Purification of extracellular lipasesfrom CMlCC 335416

The culture filtrate (3.4 1) was adjusted to pH 5.8 with NaOH and to 0.5% (by vol.) Triton X-100 and batch-bound to 300ml of the anion exchanger Q Sepharose Fast Flow MATERIALS AND METHODS (Pharmacia) equilibrated in 50 mM sodium phosphate, Microorganisms pH 5.8. After 5 min of mixing at room temperature, an aliquot G. candidum ATCC 34614 was obtained from the American was centrifuged briefly and the supernatant was assayed for Type Culture Collection, Rockville, Maryland, USA. The lipase activity to confirm binding. The matrix was washed Colworth strain was donated by Dr Alford of USDA labora- extensively on a scintered-glass funnel with equilibration tory, Beltsville, Maryland, USA and deposited at the Com- buffer until the eluant was clear. The Q Sepharose with bound monwealth Mycological Institute Culture Collection (Kew, lipase was packed into a column (2.4 cm x 66 cm) and protein was eluted with a 1-1gradient up to 0.5 M NaCl at 1.O ml/min Surrey, UK) as CMICC 335426. in 11-ml fractions. Active fractions from the Q Sepharose column were further purified by HPLC. Routinely, 5 ml of Growth and maintenance of organisms the pooled active fractions was diluted with 5 ml water to Spores of G. candidum were inoculated into a medium reduce the conductivity and filtered using a 2.2-pM Millipore containing the following in distilled water: 1.5 g KH2P04/1, filter. Chromatography was performed on a Gilson HPLC 1.0 g NH,CI/I, 1.2 g MgS04 . 7 H20/1, 20.0 g Difco Beta Lab system using a TSK DEAE 3SW column (7.5 mm x 75 mm). yeast extract/l, 25.0 g olive oil/l, 17 mg ZnS04/1, 17 mg The column was equilibrated in 50 mM sodium phosphate, MnS04/1 and 17 mg FeSO/l. All media constituents were ster- pH 5.8. Protein was eluted with a 60-ml gradient up to 0.2 M ilised together at 121"C for 30 min. Media were supplemented NaCl at a flow rate of 1 ml/min. Fractions of 1 ml were colwith 0.25 g MgS04/1 at 24 h post-inoculation. The fermen- lected and assayed for activity. tation temperature was 30°C, pH initially 6.4, air flow 0.16 vol. air . vol. medium-' . min-', oxygen supplement Determination of' native molecular mass 0.32 vol. air. vol. medium-' . min-l and agitation was Native molecular masses of the lipases was determined 400 rpm using a series of three Rushton pitched blades. The pH of the fermentation medium was maintained at 4.5 during using two in-line calibrated Superose 12 FPLC gel-filtration the period of lipase production with ammonia solution. The columns (Pharmacia). The columns were equilibrated in level of biomass generated during the vegetative growth of the 50 mM sodium phosphate, pH 7.5, and 0.15 M NaC1, and fungus necessitated that aeration and agitation be altered to initially coated with bovine serum albumin to prevent noninaintain the dissolved oxygen above 20%. The parameters specific binding. The following standards (Pharmacia) were were increased to a combined air and oxygen flow rate of used to calibrate the column: catalase (232 kDa), aldolase 0.83 vol. air . vol. medium-' . min-l and a maximum agi- (1 58 kDa), phosphorylase (97.4 kDa), ovalbumin (45 kDa), tation rate of 800 rpm. Foaming was controlled by the use of cc-chymotrypsinogen (25 kDa) and ribonuclease (13.7 kDa). Samples were loaded via a 200-p1 loop, eluted at 0.3 ml/min Silcolapse 5000 (ICI). From 18 h post-inoculation, every 2 h, 20 ml samples were and 20O-pl fractions were collected. Lipases were located by withdrawn from the fermentation to assay for lipase activity. absorbance at 280 nm and enzyme activity. After assaying an aliquot of the fermentation broth another sample was centrifuged briefly so that lipase secreted into the Polyacrylamide gel electrophoresis medium could be assayed. Proteins were analysed by SDSjPAGE according to the method of Laenimli [26] using 6.5% or 10% slab gels in a BioPurification of extracellular lipasesjiom ATCC 34614 Rad mini Protean I1 apparatus. The purification of two lipases from G. candidum ATCC 34614 has recently been reported by Sugihara et al. 1161 and Western blotring by Veeraragavan et al. [25]. We have developed a simpler Proteins were separated by SDSjPAGE and electromethod to purify these enzymes. Harvested culture medium (5 1) was clarified by filtration through Whatman no. 1 paper phoretically transferred to nitrocellulose (Schleicher and and NaCl was added to a final concentration of 2 M. The Schuell) as described by Yen and Webster [27]. After blocking filtrate was batch bound to 200 ml phenyl-Sepharose with 3% (mass/vol.) haemoglobin the nitrocellulose filters (Pharmacia) equilibrated in 50 mM Bistris, pH 6.0, and 2 M were incubated with rabbit anti-lipase antibody followed by NaCl with gentle agitation for 16 h at 4°C. The matrix was 2 pCi '251-labelled donkey anti-(rabbit IgG) IgF(ab), fragthen extensively washed on a scintered-glass funnel with ments (Amersham) and processed for autoradiography. 50 mM Bistris, pH 6.0, and 2 M NaC1, packed into a column (2.4 cm x 44 cm) and further washed until the eluant was clear. An ti-lipase an tibody The column was developed with 60% (by vol.) ethylene glycol Rabbits were immunised by intramuscular injection at one in water and a single peak of activity was recovered. Further purification was achieved on a FPLC mono Q ion-exchange site in the hind leg with 100 pg lipase in 100 pI 0.85% (mass/ column (Pharmacia) with aliquots of the phenyl-Sepharose vol.) NaCl plus 100 p1 Freund's complete adjuvant. Rabbits

48 7 were boosted at 4 weeks with 100 pg lipase in incomplete adjuvant subcutaneously in the same hind leg and boosted at 8 weeks intravenously. Rabbits were bled 5 -8 days later. Test bleeds were collected prior to each inoculation.

fatty acids on glycerol were determined as described by Gunstone [29] using pig pancreatic lipase to release fatty acids from C1 and C2; the 2-acylglycerol was separated from free fatty acid by TLC. Fatty acids were methylated and analysed as described above.

Endo-fl-N-acetyl~lycosaminidasedigestion Proteins were deglycosylated using Streptomyces plicatus endo-j-N-acetylglucosammidase (endo H) purchased from ICN Immunobiologicals. This enzyme releases N-linked mannose residues from glycoproteins [28]. Native and denatured samples of lipase were treated with endo H. Denatured samples were derived by treating native lipase with 0.2% (mass/vol.) SDS at 90°C for 2 min. Protein samples were incubated with endo H (0.08 U/mg protein) at 37°C for 4 h in 50 mM sodium acetate, pH 5.5. 1 mM phenylmethylsulphony1 fluoride was included to prevent proteolysis. The reaction was terminated by the addition of Laemmli sample buffer [26] and heating to 90' C for 2 min.

RESULTS AND DlSCUSSION Production of extracellular lipase by G. candidum

The largest increase in titre of lipase found within the supernatant appeared to coincide with sporulation in both ATCC 34614 and CMICC 335426 strains of G. candidurn. The harvesting time from the point of inoculation varied over 24 32 h, depending upon the amount of lipase secreted. Purijication of the lipases

Four separate extracellular lipases have been purified from two strains of C. candidum: two from ATCC 34614 and two from CMICC 335426. The two lipases from ATCC 34614 have Isoelectric,focusing been called I and I1 (by analogy with [36] and [25]),and those This was performed in a Pharmacia Phast gel system using from CMICC 335426, A and B. The new procedure which we have devised for the isolation of the ATCC 34614 lipases is calibrated pH 4.0- 6.5 gels according to the manufacturer's faster than the methods described by Tsujisaka et al. [7], instructions. Sugihara et al. [36] and Veeraragavan et al. [25] and, overall, results in very similar recoveries. Our method requires only Hydrolysis assays two steps in contrast to the 4 - 6 steps reported previously [7, Each sample of purified lipase was added to an emulsion of 16, 251. Table 1 shows the purification profile of a typical the substrate in 20 ml20% (massjvol.) gum arabic containing preparation of ATCC 34614 lipases. The specific activities of 0.5% (massivol.) CaClz at 40 "C. The emulsions were prepared the lipases were 2520 U/mg for lipase I and 1000 U/mg for by sonication for 3 min (speed 6) using a Lucas Dawe Ultra- lipase 11, with recoveries of 68% and 15% of the original sonic soniprobe. The release of free fatty acids was measured activity, respectively. lnitial experiments on the CMICC 335426 lipase indicated by automatic titration (Radiometer, Copenhagen) with 0.1 M NaOH to pH 8.0. 1 U lipase activity is defined as the amount that it had considerably different properties from that isolated of enzyme which would release, at a maximum rate, 1 pmol from the ATCC 34614 strain. Lipase activity from CMICC fatty acid/min from a 5% (mass/vol.) emulsion of olive oil 335426 bound only partially to phenyl-Sepharose yet bound irreversibly to a mono Q FPLC column. Culture filtrates of under the above conditions. The substrates chosen for specificity analysis were 2.6% the CMICC strain were somewhat cloudy and biological actriolein/tripalmitin in equimolar quantitites, 1% 1(3)-oleoyl- tivity was lost with the appearance of a skim of Fat in the 2,3(1)-dipalmitoylglycerol, 1% 2-oleoyl-I ,3-dipalmitoylgly- storage vessel. Irreversible binding to a mono Q column apcerol, 1YO2-oleoyl-I ,3-distearoylglycerol, 1% palm olein, and peared to be the result of a specific interaction with the back0.5% methyl esters of oleic, palmitic, vaccenic and elaidic bone of the mono Q column. Biological activity was, however, acids (run separately). Hydrolysis reactions of methyl esters retained if Triton X-100 (0.5% by vol.) was added to the were carried out for 3 min at 30°C using 5 U lipase. The rate culture filtrate which also served to clarify the filtrate. Lipase of hydrolysis of each methyl ester was calculated relative to activity did bind to Q Sepharose and this activity was eluted methyl oleate and was expressed as a percentage. Digestion in a single peak at approximately 100 mM NaC1. For the of triglycerides was terminated by the addition of 1 ml 1 M second step in the purification we attempted anion-exchange HC1/2-ml sample. The digestion products were extracted with chromatography using a DEAE 3SW column on the basis 40-60°C petroleum ether and separated by TLC on that its silica backbone would be considerably different from 20 cm x 20 cm glass plates coated with 1 mm silica gel G. the backbone of the mono Q column. Chromatography of Q Development was in 40 - 60°C petroleum ether/diethyl ether/ Sepharose lipase activity on the DEAE 3SW column resulted formic acid (70: 30: 1). Bands were visualised under ultraviolet in two peaks of biological activity, as shown in Fig. 1. Wc light after spraying with 0.01% (massivol.) phloxin in 50% have termed the early eluting peak (130 mM NaC1) lipase A, (by vol.) methanol. The fatty acids were extracted with diethyl and the second peak (150 mM NaCI) lipase B (Fig. 1 ) . Lipase ether and methylated using 14% boron trifluoride in methanol B is the major lipase secreted by G. candidurn CMICC 335426. Specific activities and a purification profile for lipases A as a catalyst. Gas/liquid chromatography was performed with a Pye Unicam PU4500 with a 1.83-mm, 4-mm internal diame- and B are presented in Table 1. Typical specific activities are ter, glass column containing 10% silica on 100/120 990 U/mg for lipase A (12% recovery) and 2455 U/mg for chromosorb. The temperature of the detector and injector was lipase B (39% recovery). The specific activities are similar to 230 'C. The column temperature programme was as follows: those for lipases I (2520 Ujmg) and I1 (1000 U/mg; Table 1). From analysis of lipases by SDSjPAGE and gel filtration, 180 C for 2 min, 180-230°C at 8"C/min and 230°C for 5 min. Peaks were integrated using Nelson Analytical software a number, including those from ATCC 34614 [7] and that from and quantification was achieved using the fatty acid content Rhizomucor mirhei [lX], have been reported to be glycosylated. of the sample derived by autotitration. The positions of the The majority of the carbohydrate associated with ATCC 34614

Table 1. Purijication profiles for G. candidum ( A ) ATCC 34614 lipases I and II and ( B ) CMICC 335426 lipases A and B Enzymes were purified and assayed as described in Materials and Methods

G . Candidum strain

Step

ATCC 34614

culture filtrate phenyl-Sepharose mono Q lipase I mono Q lipase I1

CMICC 335426

culture filtrate Q Sepharose DEAE 3SW lipase A DEAE 3SW lipase B

Volume

Protein

Total activity

Specific activity

Recovery

ml 5000 77

mg

kU

U/mg

YO

389 250 139

920 700 630 140

1800 2520 I000

100 76 68 15

290 105 132

840 523 105 324

1802 990 2455

100 62 12 39

3400 264

0.5 -

0.4 -

0.3m 0

2 0.20.1-

._

J

0-

-’ I

1

I

1

a

Fig

PH and proteinswere detected at 280 nm

lipase was reported to be mannose [7], therefore we have analysed the subunit molecular masses of lipases isolated from ATCC 34614 and from CMICC 335426 before and after deglycosylation using endo H (Fig. 2). All of the enzymes show an electrophoretic shift to lower molecular mass following endo H treatment in the presence of SDS. An identical electrophoretic shift is observed when deglycosylation is achieved in the absence of SDS, indicating that all the lipases are glycosylated and their glycosylation sites are externally available. The molecular masses of glycosylated and deglycosylated lipases are presented in Table 2. The enzymes have similar molecular masses and treatment with endo H causes a reduction in molecular mass of approximately 4.5 kDa. Thus, lipases I and I1 of ATCC 34614 were 62 kDa as glycoproteins, but 57.5 kDa following treatment with endo H (Table 2). Lipases A and B of CMICC 335426 were 62 kDa and 58.3 kDa, respectively, before endo H treatment and 57.5 kDa and 54.0 kDa, respectively, after. The molecular mass values of the four lipases determined by SDSiPAGE shown in Table 2 are similar to two extracellular lipases of G. candidurn recently described by Jacobsen et al. [15, 301, Sugihara et al. 1161 and Veeraragavan et al. [25]. Tsujisaka et al. [7] reported that the extracellular lipase of ATCC 34614 had a molecular mass of 53-55 kDa, determined by gel filtration. We have determined the native molecular masses of

II

A

B

mass kDa

92 69

43

-

Fig. 2. SDSjPAGE of glycosylated and deglycosylated G. candidum lipuses. 10% gels containing 0.1% SDS were used to analyse lipases I and I1 purified from ATCC 34614 and lipases A and B purified from CMICC 335426. Prior to electrophoresis, the lipases were either treated (+) with endo H or left untreated (-). Subunit molecular mass was determined using phosphorylase b (92 kDa), bovine serum albumin (69 kDa) and ovalbumin (43 kDa)

four lipases of G. candidum by gel filtration, and these values are given in Table 2 adjacent to the values determined by SDS/ PAGE. The values are in the range 49 - 55.5 kDa; they are

489 similar to those reported previously [7, 15, 16, 25, 301 and show that G. candidum extracellular lipases are monomeric. Isoelectric focusing

We have determined the p l of the lipases before and after treatment with endo H (Table 2). All four lipases are clearly distinguishable: lipases A and B of CMICC 335426 strain have p l of 4.71 and 4.50, respectively, and these values are not altered by treatment with endo H. Lipases I and I1 of ATCC 34614 have p l of 4.61 and 4.47 before endo H treatment. In the absence of carbohydrate, lipase I still has a p l of 4.61. In contrast, the p l of lipase I1 shifts from 4.47 before endo H treatment of 4.61 after endo H treatment. Treatment of lipase I1 with endo H clearly exposes charged residues normally concealed by carbohydrate or removes charged carbohydrate. This does not appear to be the case with lipase I of ATCC 34614 or lipases A and B from CMICC 335426 (Table 2). Very similar pf have been reported by other workers for extracellular lipases from ATCC 34614 [7, 15, 16, 251. These workers did not report the effects of deglycosylation on the P I .

and also recognises lipases A and B in Western blots, demonstrating that they have common antigenic determinants (Fig. 3a). This antibody also cross-reacts with deglycosylated lipases, showing that the glycoprotein is not the major antigenic determinant (Fig. 3a). Anti-(lipase I) antibody appears to cross-react more strongly with lipase A from CMICC 335426 than with lipase B, suggesting homology between these two lipases. Fig. 3b shows that anti-(lipase B) antibody is cross-reactive in Western blots with glycosylated and deglycosylated lipases A, B, I and 11, all to the same extent. The Western blots shown in Fig. 3 were performed using denatured protein, indicating that these G. candidum lipases have amino acid sequence similarity. This has already been shown by the I

'+

A

B

'

-I-[+

A+B

-

Immunological cross-reactivity

In order to investigate the possibility of common antigenic determinants between the enzymes, an antibody was raised against isolated lipase B and also one against ATCC 34614 lipase I. Anti-(lipase I) antibody recognises lipases I and I1 Table 2. Stimmury qf molecular masses (native and denatured) andpI qf gljc'osyluted and deglycosyluted ( - CHO) lipases isolated ,from ATCC 34614 ( I und I I ) und CMICC 335426 ( A urid Bj Native molccular masses were determined by gel filtration using two calibrated, in-line Superose 12 FPLC columns. Denatured molecular masses were determined by SDSjPAGE with appropriate standards. pZ were determined by isoelectric focussing Lipase

Molecular mass

PI ~

native

denatured

kDa I I (-CHO)

50.1

11

55.5 -

I1 (-CHO)

A A (- CHO)

B B (-CHO)

-

53.7 48.9 -

62.0 57.5 62.0 57.5

4.61 4.61 4.47 4.61

62.0 57.5 58.3 54.0

4.71 4.71 4.50 4.50

b Fig. 3. Western blotting of glycosylated and deElycosyluted lipases. Purified lipases were run out on SDS/polyacrylamide gels (10%) then electrophoretically transferred to nitrocellulose. The blots were thcn incubated with (a) anti-(lipase I) antibody or (b) anti-(lipase B) antibody followed by '251-labelleddonkey anti-(rabbit IgC)IgF(ab) fragments and processed for autoradiography. I, A, B and A + B refer to purified lipases blotted and probed. (+) and (-) refer to lipases treated or not treated with endo H, respectively

Table 3. Distribution of fatty acid groups in palm olein Fatty acyl groups were quantified as described in Materials and Methods. tr, trace Distribution of fatty acid

Total Position 2 Positions 1 and 3

Relative amount of fatty acid 12:o

14:O

16:O

16: 1

18:O

18:l

18:2

18 : 3/20: 0

0.4 0.5 0.3

1 .o 0.4 1.3

40.2 8.1 56.2

0.2 tr 0.4

4.3 0.5 6.2

42.6 69.4 29.2

10.4 20.7 5.3

0.9 0.3 0.8

490 Table 4. Hydrdjsjs ofpalm olein by G . candidum lipuses Hydrolysis was measured as described in Materials and Methods and free Patty acid released were identified by GLC after separation by TLC. The time course of liberation of various free fatty acids from palm olein was studied using 30 (olive oil) U of each lipase preparation Lipase

Rate of release of free fatty acid

16:O

18:O

18:l

18:2

0.00 0.01 0.01 0.01

0.23 0.27 0.28 0.67

0.05 0.02 0.05 0.10

bmol/min

I

0.18 0.02 0.28 0.00

A/ B

A B

amino acid sequences deduced from cDNA clones of lipases I and 11 from ATCC 34614 [22,23]. Sprcifkity studies with G. candidum lipases

The biological activity of the ATCC 34614 type I lipase was investigated using an equimolar emulsion of triolein and tripalmitin. After 8 min, the free fatty acid composition of the mixture was 20% 16 :0 and 79% 18 : 1. From this observation it appeared as if the enzyme had a preference for the unsaturated substrate over palmitate, as reported for the crude G. candidum lipase [5, 8, 91, but not for that from the ATCC 34614 strain 171. However, the possibility exists that since tripalmitin is crystalline at 40"C, it is not available to the enzyme. This problem may be overcome using a substrate which is liquid at

40'C, such as palm olein. Table 3 shows a positional fatty acid analysis ofthe palm olein which we used in the subsequent experiments, The time of liberation of various fatty acids from palm olein using four lipase preparations is shown in Table4. From this it can be seen that the ATCC type I lipase and CMICC 335426 lipase A have a similar specificity: 16:O is liberated at a similar rate to lX:l, while 1 X : O is not liberated at an appreciable rate. CMICC 335426 lipase B has a much greater activity against 18: 1 in comparison to 16:0 and, therefore, is specific for unsaturated substrates (Table 4). Like the other two lipases, it does not utilize 18:'O at any appreciable rate. Deglycosylation of lipase B does not alter its preference for 18 : 1 over 16:O (data not shown). The crude lipase, containing both the A and the B forms, showed a moderate svecificitv for unsaturatred substrates (Table 4). The activities of pure lipases A and B against other synthetic triglycerides [2-oleoyl1,3-diPalmitoYlglYcerol;1(3)-oleoYl-2,3(1)-diPalmitOYlglYCer01; 2-oleoyl-I ,3-distearoylglycerol] were evaluated and are shown in Table 5. These results confirm that livase B has a strong preference for unsaturated substrates. In stark contrast, lipase A hydrolysed at a similar rate saturated and unsaturated fatty acids from triglycerides; it even showed a preference for 16 : 0 from 2-oleoyl-l,3-dipalmitoylglycerol. There was no significant difference in the rate of release of 18 : 1 from 2-oleoyl-I ,3-dipalmitoylglycerol and l-oleoyl2,3-dipalmitoylglycerol by lipase B, which means that the enzyme has no noticeable positional specificity. The hydrolysis of 18 :0 from 2-oleoyl-1,3-distearoylglycerol by lipase B was particularly slow: 50 times slower than by lipase A (Table 5). We finally tested the four lipases on methyl esters of oleic acid and two of its isomers, elaidic and vaccenic acids, and on

Table 5. Specificity studies with lipases A and B purified from G. candidum CMICC 335426 The activities of pure lipases A and €3 against three synthetic substrates were studied. Hydrolysis was measured by titration, and released free fatty acids were identified by GLC after separation by TLC. Hydrolysis was catalysed by 30 (olive oil) U of each lipase Synthetic substrate

Rate of release of free fatty acids lipase B

lipase A ._

16:O

18:O

18:l

16:O

18:O

18:l

-

0.090 0.320

0.005 0.002

-

-

0.100

0.160

-

0.0004

0.500 0.400 0.370

pmol/min

1,2-Dipa~mityl-2-o~eylg~ycero~ 0.320 0.400 2,3-Dipalmityl-l(3)-oleyl-glycerol 1,3-Disteary~-2-o~ey~g~ycero~ -

~

Table 6.Relative rates of hydrolysis qf,fatty acid methyl esters by G. candidum lipases Hydrolysis of methyl cstcrs by crude or pure lipases was measured as described in Materials and Methods and expressed as a percentage, relative to the hydrolysis of methyl oleate. Unless otherwise stated, hydrolysis was catalysed by 5 (olive oil) U of each lipase Fatty acid ester

Methyl oleate Methyl elaidatc Methyl vacccnate Methyl palmitate

Relative hydrolysis with lipase I

crude CMICC

pure A

pure B

100 15.2 21.0 42.9

100 15.9 8.5 26.8

100 23.9 13.1 64.2

100

This very low rate is approximate; 200 (olive oil) U lipase B were used

0.7 0.01 0.001 a

49 1 methyl palmitate (Table 6). The results show that ATCC 34614 type I lipase, the crude CMICC 426 (contains lipases A and B), and lipase A hydrolysed all of these substrates at a similar rate. However, lipase B was found to be active on methyl oleate only, showing that this enzyme is highly specific for unsaturated methyl esters with a double bond at position 9 in the cis configuration. The results presented show that we have purified a lipase from a strain of G. candidurri which is very specific for unsaturated substrates with a cis double bond at position 9. Remarkably, this lipase appears to be related to a number of other extracellular lipases from G. candidurn which do not show the same degree of specificity toward unsaturated substrates. We have also looked at the chain-length specificity of lipase B as, well as its preference for the number and position of double bonds (Charton, C., and Marcrae, A. R., unpublished results). Elucidation of the structure of lipases A and B from CMICC 335426 by amino acid and nucleotide sequencing and, ultimately, X-ray crystallography, will give us an insight into the unique substrate specificity exhibited by lipase B. Due to its unusual specificity, lipase B may have important industrial applicatioiis in the isolation of pure fatty acids, the synthesis of fatty acids and the production of novel triglycerides by transesterification reactions. We would like to thank Helen Bambridge and Andrew Austin for help in the purification and the assays and John Irwin for the quantification of fatty acid groups.

REFERENCES 1. Borgstrom, B. & Brockman, H. L., eds. (1984) Lipases, Elscvier, Amsterdam. 2. Jensen, R . G . & Pitas, R. E. (1976) in Lipids (Paoletti, R., Porcellati, G. & Jacini, G. eds) vol. 1, pp. 141 -146, Raven Press, Ncw York. 3. Okumura, S., Iwai, M. & Tsujisaka, Y. (1976) Agric. B i d . Chem. 40,655 - 660. 4. Macrae, A. R. (1985) in Biocatalysts in organic synthesis (Tramper, J., van der Plas, H. C. & Linko, P., cds) Stud. Org. Chem. 22, 195-208. 5. Alford, J . A . &Pierce, D. A . (1961) J . Food, Sci. 26, 518-524. 6. Baillargeon, M. W., Bistline, R. R . & Sonnet, P. E. (1989) Appl. Microbiol. Biotechnol. 30, 92 - 96. 7. Tsujisaka, Y., Iwai, M. & Tominaga, Y. (1973) Agric. B i d . Chem. 37, 1457- 1464.

8. Aneja, R. & Hollis, W. H. (1983) J. Am. Chem. Soc. 105, 1861. 9. Aneja, R. (1987) J . Am. Oil Chern. Soc. 64. 645. 10. Jensen, R. G., Sampugna, J., Quinn, J . G., Carpenter, D. L., Marks, T. A. & Alford, J. A. (1965) J . Am. Oil Chem. Soc. 42, 1029 - 1033. 1 I. Marks, T. A , , Quinn, J. G., Sampugna, J. & Jensen, R. G. (1868) Lipids 3, 143 - 146. 12. Jensen, R. G. (1973) Lkid.9 9, 149-157. 13. Tdhoun, M. K . (1987) Fut Sci. Techtzol. 89, 318-320. 14. Franzke, C., Kroll, J. & Petzold, R. (1973) Nuhrutzg 17, I72 184. 15. Jacobscn, T., Olsen, J., Allerman, K., Poulsen, 0. M. & Hau, J. (1989) Enzyme Microb. Techno/. 11, 90-95. 16. Sugihara, A,, Shimada, Y. & Tominaga, Y . (1990) J . Binci7ern. (Tokyo) 107,426-430. 17. Heldt-Hanscn, H. P., Ishii, M., Patkar, S. A,, Hansen, T. T. & Eigtred, P. (1989) in Biocutalysis in agricultural biotechnology (Whitaker, J. R. &Sonnet, P. E., eds) pp. 158-172, American Chemical Society, Washington. 18. Huge-Jensen, B., Galluzzo, D. R. & Jensen, R. G. (1987) Lipids 22, 559-565. 19. Iwai, M., Okumura, S. & Tsujisaka, Y. (1975) Agr. Biol. Chem. 39, 1063-1070. 20. Hata, Y., Matsuura, A., Tanaka, N., Kakudo, M., Sugihara, A., Iwai,M.&Tsujisaka,Y. (1979)J. Biochem. (Tokyo) 86,1821 1827. 21. Hata, Y., Tanaka, N., Kakudo, M., Sugihara, A , , Iwai, M . & Tsujisaka, Y. (1981) Actu Crystallogr. A37, C38. 22. Shimada, Y., Sugihara, A., Tominagd, Y., Iizuma, T. & Tsunasawa, S. (1989) J . Biochem. Tokyo) 106, 383-388. 23. Shimada, Y., Sugihara, A,, Iizuma, T. & Tominaga, Y. (1990) J . Biochem. (Tokyo) 107, 703 -707. 24. Kawaguchi, Y., Honda, H., Taniguchi-Morimura, J. & Iwasaki. S. (1989) Nature 341, 164- 166. 25. Veeraragavan, K., Colpitts, T. & Gibbs, B. F. (1990) Bcoclzim. Biophys. Acta 1044, 26 - 33. 26. Laemmli, U. K. (1970) Nature 227,680-685. 27. Yen, T. S. B. & Webster, R. E. (1981) J . B i d . Chern. 256,1125911265. 28. Tarcntino, A . L. & Maley, F. J . (1974) J . B i d . Chem. 24Y, 81 I 817. 29. Gunstone, F. D. (1967) A n introduction to the chernistry atid biochemistry of.foods and their glycerides, 2nd edn, p. 165, Chapman and Hall. 30. Jacobsen, T., Jensen, B., Olsen, J. & Allermann, K. (1989) Appl. Microbiol. Biotechnol. 32, 256-261.

Geotrichum candidum produces several lipases with markedly different substrate specificities.

We have purified and examined the substrate specificity of four lipases from two strains of the mould Geotrichum candidum, ATCC 34614 and CMICC 335426...
724KB Sizes 0 Downloads 0 Views