ARTICLE IN PRESS

Microbiological Research 164 (2009) 43—48

0944-5013/$ - sdoi:10.1016/j.

�Correspondfax: +81 3 5477

E-mail addr

www.elsevier.de/micres

SHORT COMMUNICATION

Isolation and characterization of actinomycetesstrains that produce phospholipase D having hightransphosphatidylation activity

Yozo Nakazawaa, Masataka Uchinoa, Yoshimasa Saganeb,Hiroaki Satoa, Katsumi Takanoa,�

aDepartment of Applied Biology and Chemistry, Faculty of Applied Bioscience, Tokyo University of Agriculture,1-1-1 Sakuragaoka, Setagaya-ku, Tokyo 156-8502, JapanbSars International Centre for Marine Molecular Biology, Bergen High Technology Centre, Thormøhlensgt. 55,N-5008 Bergen, Norway

Received 21 August 2006; received in revised form 30 October 2006; accepted 3 November 2006

KEYWORDSTransphosphatidyla-tion;Phospholipase D;Actinomycetes;Phospholipid modifi-cation

ee front matter & 2007micres.2006.11.003

ing author. Tel.: +81 3 52619.ess: takano@nodai.ac.j

SummaryThe present study was conducted to screen microorganisms that producephospholipase D (PLD), and we especially focused on the strains having hightransphosphatidylation activity. Eighty bacterial strains were isolated from soilsamples by a screening method utilizing a preliminary selection medium withphosphatidylcholine (PC) as the sole carbon source. The culture supernatants werethen assayed for PLD activity. The finding of dual PLD activities in cultures revealedthat the hydrolytic and transphosphatidylation activities were correlated. Conse-quently, six strains were selected as stably producing PLD enzyme(s) duringcontinuous subcultures. The culture supernatants of selected strains synthesizedphosphatidylglycerol, phosphatidylserine and phosphatidylethanolamine from PCwith high conversion rates. These isolated strains will be made available to carry outphospholipid modification through the efficient transphosphatidylation activity ofthe PLD that they produce.& 2007 Elsevier GmbH. All rights reserved.

Introduction

Lecithin, which can be defined as a mixture ofphospholipids (PLs) with other minor components,

Elsevier GmbH. All rights rese

477 2325;

p (K. Takano).

is an abundant by-product yielded from industrialoil production processes that utilize crude vegeta-ble oils or animal fats. PLs are polar lipids havingsurface-active properties; hence lecithin has beenwidely utilized in food products as an emulsifier,stabilizer and antioxidant (Doig and Diks, 2003).The dominant PL component in native lecithin isphosphatidylcholine (PC), and lecithin-derived PC

rved.

ARTICLE IN PRESS

Y. Nakazawa et al.44

has been included in cosmetics and pharmaceuti-cals due to its emulsification property and lipo-some-formation ability (Miura et al., 2004; Nii andIshii, 2004; Ishii and Nii, 2005). On the other hand,native lecithin also contains other PLs as minorconstituents, such as phosphatidylglycerol (PG),phosphatidylserine (PS), phosphatidylethanolamine(PE) and phosphatidylinositol (PI). According todata regarding modified lecithins, the minor PLcomponents can provide emulsification stability tothe lecithin over a wide range of pH and saltconditions (Kudo and Kuroda, 1990). Additionally,PS, which can also be found in animal brains, hasbeen proposed as a food supplement to decreaseand/or prevent senile dementia (Suzuki et al.,2001; Jorissen et al., 2002). However, the composi-tional variety of PLs in native lecithin is limited,and the yields of such PLs from various naturalresources produce only meagre amounts. Althoughchemically synthesized PLs are available, theyare probably too expensive to use in daily foodproducts.

Phospholipase D (PLD; EC.3.1.4.4) is an enzymethat catalyzes the hydrolysis of PLs, releasing aphosphatidic acid (PA) and an alcohol. While in thepresence of an additional alcohol, PLD preferen-tially catalyzes the transphosphatidylation reac-tion, in which a polar-head modified phosphatidylalcohol is synthesized from substrate PLs that act asphosphatidyl-moiety donors. The transphosphatidy-lation reaction carried out by the PLD enzymeshould be, therefore, a powerful tool for achievingPL modifications economically. The reactions by thePLD can occur with different degrees of selectivitybetween hydrolysis and transphosphatidylationdepending on the enzymatic sources. PLD is a fairlyubiquitous enzyme that is widely distributed inbacteria, fungi, plant and vertebrate species(Morris et al., 1997). Particularly, the PLD enzymesproduced by soil bacterial strains appear to displayhigh transphosphatidylation activity (Takami et al.,1994; D’Arrigo et al., 1996; Hirche and Ulbrich-Hofmann, 2000), and accordingly, such PLDs arealready used in industrial fields.

The PLDs can competitively react with newlysynthesized phosphatidyl alcohols as they hydrolyzethe de novo products. Thereby, the hydrolysisreaction by the enzyme may reduce the efficiencyof the transphosphatidylation reaction (Comfuriuset al., 1990). To avoid the hydrolysis by theenzyme, the reaction is often carried out in thebiphasic system consisting of an aqueous buffer andan organic solvent (e.g., diethyl ether or ethylacetate), in which the free water that would beemployed for the hydrolysis can be reduced. Butthe organic solvents for this system sometimes

include toxic compound, and thus this strategy maybe inappropriate for the food-processing purpose.Hence, the reaction system that allows effectivetransphosphatidylation even without the organicsolvent has been expected. In this aspect, anaqueous suspension system in which the lecithin,previously absorbed on fine powders, was used asphosphatidyl-moiety donor was developed recently(Iwasaki et al., 2003).

Besides the improvement of the reaction system,it seems to be very worthwhile to find the PLDhaving a high selectivity on the transphosphatidyla-tion activity even though in the system consists ofsimple aqueous suspension of the phosphatidyl-moiety donor, acceptor molecule and the PLDenzyme without using toxic organic solvent. How-ever, in most cases so far for the screening of thePLD-productive bacterial strains, the transpho-sphatidylation reaction of the PLD enzyme hasbeen assessed using biphasic system (Shimbo et al.,1989; Nakajima et al., 1994; Carrea et al., 1995;Hagishita et al., 2000). In this study, we focused onthe PLD enzyme that allows high transphosphatidy-lation capacity in the one-phase system in whichthe lecithin-derived PC is dispersed in an aqueousbuffer using small amounts of detergent, andconducted a screening method to isolate micro-organisms producing PLD enzyme that shows hightransphosphatidylation activity. That is, the micro-organisms that grow in lecithin-derived PC as thesole carbon source were isolated, and assayed theirculture supernatants for PLD activity, to selectmicroorganisms that produce the PLD enzyme(s).Then we examined the transphosphatidylationactivities of culture supernatants in the aqueousreaction system in which lecithin-derived PC isdispersed, in conjunction with various acceptorssuch as glycerol, L-serine, 2-aminoethanol hydro-chloride and myo-inositol. Additionally, we alsoexamined the hydrolysis of the PG, PS, PE and PI, aswell as PC, by the culture supernatants to assesswhether they would react to the de novo synthe-sized PL products by transphosphatidylation reac-tion.

Materials and methods

Screening and cultivation conditions

The bacterial strains used in this study wereisolated from 25 soil samples in Setagaya, Tokyo,Japan. For the screening of microorganisms, selec-tion plates consisting of 0.2% KNO3, 0.1% K2HPO4,0.05% NaCl, 0.05% MgSO4 � 7H2O and 1.5% agar were

ARTICLE IN PRESS

Figure 1. Transphosphatidylation conversion rates andPC-hydrolyzing activity in culture supernatants of PLDproducing microorganisms. The PC-hydrolyzing activitywas determined by measurement of released cholines.Symbols: K, PG-producing; &, PE-producing; and m, PS-producing. The value in vertical axis of each pointindicates the PL-producing transphosphatidylation con-version rate versus the PC-hydrolyzing activity of eachstrain’s culture supernatant; therefore, three points thatshow same PC-hydrolyzing activity represent each trans-phosphatidylation conversion rate of same strain.

Actinomycetes strains that produce PLD having high transphosphatidylation activity 45

used, and further, as the sole carbon source, 0.2%PC (PC-98N from egg yolk lecithin; kindly providedby Kewpie Fine Chemicals, Tokyo, Japan) wassupplied. The colonies that appeared on the plateswere separately inoculated and cultivated in 2.0mlof GYP medium composed of 1% glucose (Wako,Tokyo, Japan), 1% yeast extract (Difco, FranklinLakes, NJ), 1% polypeptone (Nihon Pharmaceutical,Tokyo, Japan), 0.2% K2HPO4 and 0.05% MgSO4 � 7H2O,at 28 1C with shaking at 200 rpm for 48h.

Measurement of PC-hydrolyzing activity andprotein concentration

PC-hydrolyzing activity of the culture super-natants was determined according to the methodsdescribed by Imamura and Horiuti (1978) using a0.1% (w/v) PC as a substrate and 4mM sodiumdeoxycholate (SDC) as a detergent. One unit (U)was defined, as the amount of enzyme thatliberated 1 mmol of choline per minute. Proteinconcentration was determined with a Bio-RadProtein Assay Kit (Bio-Rad, Hercules, USA) usingbovine serum albumin as a standard.

Thin-layer chromatography analysis of thePLs

To examine the transphosphatidylation reactions,one millilitre of reaction mixture composed of 0.1%(w/v) PC, 4mM SDC, 20mM sodium acetate buffer(pH 5.5), 10mM CaCl2, 20% (w/v) each acceptor(with the exception of myo-inositol, which wasadded at 15% (w/v) concentration) and 0.1ml ofculture supernatant was incubated at 37 1C. For theexamination of PL-hydrolytic reactions, one milli-litre of reaction mixture composed of 0.1% (w/v)each substrate PL, 20mM sodium acetate buffer(pH 5.5), 10mM CaCl2, and 0.1ml of culturesupernatant was incubated at 37 1C. The reactionwas terminated by adding one-fifth volume of 0.1 NNaOH to the reaction mixture. The PLs in thereaction mixture were extracted using the methodof Bligh and Dyer (1959). The chloroform phase wasseparated and evaporated, and then the resultantlipid residue was dissolved with 100 ml of chloro-form. After centrifugation, 10 ml of the clearchloroform phase was applied onto a thin-layerchromatography (TLC) plate (Silica gel 60; MerckKGaA, Darmstadt, Germany). The plate was devel-oped with chloroform/methanol/acetic acid(40:15:6, v/v). Spots on the plate were visualizedby iodine vapour. The intensities of the visualizedspots were measured with Gel-Pro Analyzer soft-ware (version 3.1; Media Cybernetics, Silver Spring,

MD, USA). The transphosphatidylation conversionrate (%) was defined as [PX]� 100/[PX]+[PC]+[PA],and the selectivity (%) was defined as [PX]� 100/[PX]+[PA], where the PX in the formula stands forthe transphosphatidylation product. The hydrolysisrate against each PL was determined by followingthe diminution of the spot intensities correspondingto the substrate PLs that were revealed by the TLCexperiments.

Results and discussion

Screening of microorganisms that producePLD enzyme(s) having hightransphosphatidylation activity

First, we isolated 80 strains of microorganismsfrom soil samples on the selection medium withlecithin-derived PC as the sole carbon source. Theculture supernatants of 30 of these strains exhib-ited significant PC-hydrolyzing activity (Fig. 1). Thespecific PC-hydrolyzing activities of these culturesupernatants ranged from 0.09 to 12.2 U/mgprotein (data not shown). In the presence of gly-cerol, L-serine and 2-amionoethanol hydrochloride,the culture supernatants of all the strains exhibitedPG-, PS-, and PE-producing transphosphatidylation

ARTICLE IN PRESS

Figure 2. Time course of cell growth (A), PLD activity (B)and protein content (C) in the culture medium from theisolated strains. Symbols: K, 9-4; &, 10-1; m, 10-2; J,10-3; p, 10-6; and ’, 21-4.

Y. Nakazawa et al.46

activities, with conversion rates ranging from 18.9%to 90.1% for PG-, 8.7% to 85.7% for PS-, and 14.6%to 89.2% for PE-producing reactions, respectively(Fig. 1). On the other hand, we could not find anystrains that exhibited PI-producing transphosphati-dylation activity in the presence of myo-inositol(data not shown). Hitherto, only few PI-produciblePLD that were produced by spinach and poppy hadbeen reported, despite the high biochemical andbiotechnological interest in finding such a PLD(Mandal et al., 1980; Oblozinsky et al., 2005).

Production of PLD enzyme(s) by the selectedstrains

To particularly characterize the hydrolyzing andtransphosphatidylation activity that was producedby the strains; we selected six strains, 9-4, 10-1,10-2, 10-3, 10-6 and 21-4, because these strainscould stably produce the PLD activity during thecontinuous subcultures. Based on the colony mor-phology on the selection plate, their capability ofspore formation, and the characteristic morpholo-gical behaviour of the culture broth, the selectedstrains seemed to be actinomycetes strains. Eachstrain was cultured in 100ml of GYP medium at28 1C with shaking at 200 rpm, and then thebacterial cell amounts were measured by theirdry weights, while the PLD activity and proteinconcentration in the supernatant from each culturewere concomitantly measured. As shown in Fig. 2A,these strains achieved a stationary phase at 10 hfrom the initiation of cultivation. With the excep-tion of strains 10-2 and 10-3, the maximum PLDactivities in the culture were obtained 48 h afterinitiation of the cultures, and then these activitiesdecreased (Fig. 2B); these behaviours were inagreement with the measured protein contents ofthe cultures (Fig. 2C). The PLD activities in the 10-2and 10-3 cultures increased until 72 h after theinitiation of the culture, unlike the behaviour ofthe other strains. These results implied thatproduction curves of the PLD enzyme(s) differ,depending upon the strains. To characterize thetransphosphatidylation and PL-hydrolyzing proper-ties of PLD enzyme(s) in the culture supernatants,48 h cultivations were used in the following experi-ments.

Transphosphatidylation and PL-hydrolyzingproperties of culture supernatants ofselected strains

With regard to the PG-, PS- and PE-producingtransphosphatidylation activities of the selected

strains’ culture supernatants, the conversion ratesand selectivity were examined during the 5 hreaction time. As shown in Table 1, the culture

ARTICLE IN PRESS

Table 1. Hydrolysis rate of various PLs, and conversion rates and selectivity of the various PL-producing-transphosphatidylation reactions of the culture supernatants from selected strains

Strain� Hydrolysis rate (%)y Transphosphatidylation conversion rate and selectivity (%)z

PC PG PS PE PI PG PS PE PI

9-4 30.5 2.5 0.6 35.6 0.2 86.2 (98.3) 75.2 (90.9) 74.6 (98.1) ND10-1 82.0 0.1 2.6 78.5 0.4 89.1 (91.6) 92.8 (92.8) 90.0 (99.8) ND10-2 62.1 2.8 1.1 71.7 0.9 94.4 (97.5) 84.0 (88.8) 86.1 (93.9) ND10-3 59.1 2.2 0.8 69.1 0.8 97.1 (97.7) 82.5 (88.9) 88.1 (98.3) ND10-6 82.6 0.4 1.4 63.9 0.4 89.4 (92.6) 84.7 (91.5) 85.0 (99.8) ND21-4 66.5 0.3 0.03 71.4 0.1 92.3 (95.0) 88.5 (93.3) 89.1 (99.9) ND

ND; not detected.�Each strain was cultivated at 28 1C with shaking at 200 rpm for 48 h in GYP medium.yConversion rate producing the hydrolysis product PA from each PL after 3 h reaction.zThese data refer to a 5 h reaction. The selectivity is shown in parentheses.

Actinomycetes strains that produce PLD having high transphosphatidylation activity 47

supernatants of all selected strains, excepting forthe strain 10-1, exhibited a relatively high PG-producing transphosphatidylation activity, followedby significant rates of PE- and PS-producing activity.The final transphosphatidylation conversion rates inthese reactions were 86.2–97.1% for the PG-,75.2–92.8% for the PS- and 74.6–90.0% for the PE-producing reactions, respectively. Furthermore,culture supernatants of all strains exhibited a highselectivity, with 88.8–99.9% of the products result-ing from the transphosphatidylation, rather thanthe hydrolytic reaction. Based on the previouslypublished data, transphosphatidylation conversionrates by the Streptomyces PLDs in the biphasicsystem are 74–99% for the PG-, 98–100% for PS- and92–96% for the PE-producing reactions, respec-tively (Juneja et al., 1989; Shimbo et al., 1989;Nakajima et al., 1994; Carrea et al., 1995).Therefore, PLD enzyme(s) produced by our strainsappeared to convert the PC into PG, PS and PE withhigh conversion rates even in the one-phasesystem, though the rates were slightly lower thanthose in biphasic system by the previously pub-lished PLDs.

To assess whether the PLD enzyme(s) producedby the strains would react to de novo synthesizedproducts by transphosphatidylation, the PL-hydro-lyzing properties of the culture supernatants werealso characterized (Table 1). The substrates PC andPE were significantly hydrolyzed by the culturesupernatants of the selected strains. The hydrolysisrates of strains 9-4, 10-2, 10-3 and 21-4 against thePE exceeded those against PC. On the other hand,the PG, PS and PI were scarcely hydrolyzed by theculture supernatants even after 3 h incubation.Thus, the PLD enzyme(s) themselves and/or otherco-existing PL-reactive enzymes in the culturesshould not react with de novo synthesized PG andPS. The selected strains described here may be

adaptable to procedures to produce polar-headmodifications, especially to produce PG and PS. It ishoped that the PLD enzyme(s) produced by theseselective strains will be useful for carrying out PLmodifications from the abundant lecithin-derivedPC. Therefore, further examination of the trans-phosphatidylation reaction carried out by theenzymes produced by these strains and identifica-tion of the PLD enzymes from these strains will beneeded, as future perspective. Additionally, toensure the safety on food industrial applications,we will need to examine other detergents todisperse the PL into the aqueous solution.

References

Bligh EG, Dyer WJ. A rapid method of total lipidextraction and purification. Can J Biochem Physiol1959;37:911–7.

Carrea G, D’Arrigo P, Piergianni V, Roncaglio S, Secundo F,Servi S. Purification and properties of two phospholi-pase D from Streptomyces sp. Biochim Biophys Acta1995;1255:273–9.

Comfurius P, Bevers EM, Zwaal RFA. Enzymatic synthesisof phosphatidylserine on small scale by use of a one-phase system. J Lipid Res 1990;31:1719–21.

D’Arrigo P, de Ferra L, Piergianni V, Selva A, Servi S, StriniA. Preparative transformation of natural phospholipidscatalyzed by phospholipase D from Streptomyces.J Chem Soc Perkin Trans I 1996;21:2651–6.

Doig SD, Diks RMM. Toolbox for exchanging constituentfatty acids in lecithins. Eur J Lipid Technol 2003;105:359–67.

Hagishita T, Nishikawa M, Hatanaka T. Isolation ofphospholipase D producing microorganisms with hightransphosphatidylation activity. Biotechnol Lett 2000;22:1587–90.

Hirche F, Ulbrich-Hofmann R. The interdependence ofsolvent, acceptor alcohol and enzyme source in

ARTICLE IN PRESS

Y. Nakazawa et al.48

transphosphatidylation by phospholipase D. BiocatalBiotransfor 2000;18:343–53.

Imamura S, Horiuti Y. Enzymatic determination ofphospholipase D activity with choline oxidase.J Biochem 1978;83:677–80.

Ishii F, Nii T. Properties of various phospholipid mixturesas emulsifiers or dispersing agents in nanoparticle drugcarrier preparations. Colloids Surf B Biointerfaces2005;41:257–62.

Iwasaki Y, Mizumoto Y, Okada T, Yamamoto T, Tsutsumi K,Yamamoto T. An aqueous suspension system forphospholipase D-mediated synthesis of PS withouttoxic organic solvent. J Am Oil Chem Soc 2003;80:653–7.

Jorissen BL, Brouns F, Van Boxtel MPJ, Riedel WJ. Safetyof soy-derived phosphatidylserine in elderly people.Nutr Neurosci 2002;5:337–43.

Juneja LR, Kazuoka T, Goto N, Yamane T, Shimizu S.Conversion of phosphatidylserine by various phospho-lipase D in the presence of L- or D-serine. BiochimBiophys Acta 1989;1003:277–83.

Kudo S, Kuroda A. Enzymatic production of transpho-sphatidylated lecithin and its use. Bio Industry1990;7:494–500 [in Japanese].

Mandal SB, Sen PC, Chakrabarti P. In vitro synthesis ofphosphatidylinositol and phosphatidylcholine by phos-pholipase D. Phytochemistry 1980;19:1661–3.

Miura S, Tanaka M, Suzuki A, Sato K. Application ofphospholipids extracted from bovine milk to the

reconstitution of cream using butter oil. J Am OilChem Soc 2004;81:97–100.

Morris AJ, Frohman MA, Engebrecht J. Measurement ofphospholipase D activity. Anal Biochem 1997;252:1–9.

Nakajima J, Nakashima T, Shima Y, Fukuda H, Yamane T. Afacile transphosphatidylation reaction using a culturesupernatant of actinomycetes directly as a phospho-lipase D catalyst with a chelating agent. BiotechnolBioeng 1994;44:1193–8.

Nii T, Ishii F. Properties of various phosphatidylcholines asemulsifiers or dispersing agents in microparticlepreparations for drug carriers. Colloids Surf B Bioin-terfaces 2004;39:57–63.

Oblozinsky M, Ulbrich-Hofmann R, Bezakova L. Headgroup specificity of phospholipase D isoenzymes frompoppy seedlings (Papaver somniferum L.). BiotechnolLett 2005;27:181–5.

Shimbo K, Yano H, Miyamoto Y. Two Streptomyces strainsthat produce phospholipase D with high transpho-sphatidylation activity. Agric Biol Chem 1989;53:3083–5.

Suzuki S, Yamatoya H, Sakai M, Kataoka A, Furushiro M,Kudo S. Oral administration of soybean lecithintransphosphatidylated phosphatidylserine improvesmemory impairment in aged rats. J Nutr 2001;131:2951–6.

Takami M, Hidaka N, Suzuki Y. Phospholipase D-catalyzedsynthesis of phosphatidyl aromatic compounds. BiosciBiotech Biochem 1994;58:2140–4.