APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1981, p. 511-5170099-2240/81/02051 1-07$02.00/0

Vol. 41, No. 2

Microbiological, Biochemical, and Electron MicroscopicCharacterization of a Pectinatus Strain

AULI HAIKARA,'* LIISA PENTTILA, T.-M. ENARI, AND K. LOUNATMAA

Biotechnical Laboratory, Technical Research Centre of Finland, Espoo,1 and Department of ElectronMicroscopy, University of Helsinki, Helsinki, Finland

A spoilage organism isolated from turbid beer is described. The bacterium was

gram negative, catalase negative, strictly anaerobic, and rod shaped, havingflagella only on one side of the cell. The main metabolic product was propionicacid. In addition acetic, succinic, and lactic acids and acetoin were formed.Malonate inhibited the production of propionic acid by the strain studied and byboth Pectinatus and Propionibacterium strains. The guanine-plus-cytosine con-

tent of deoxyribonucleic acid was 36 mol%. Differences between this strain andPectinatus strains were 2 to 5 percentage points. Immunofluorescent staining andgel diffusion precipitin tests revealed that the antigenic structure differed fromthose of Pectinatus strains. The isolated organism can, despite some differences,be regarded as belonging to the genus Pectinatus.

Until relatively recently, it was thought thatonly lactic acid bacteria were capable of spoilinglager beer, because the low air content in bottledbeer makes the occurrence of acetic acid bacteriavery rare. Gram-negative bacteria had earlierbeen regarded only as contaminants of wort andof the early part of the fermentation process.Currently, increasing attention is being paid totheir appearance in the latter part ofthe process.Thus, among enterobacteria, some strains havebeen found which survive the main fermenta-tion. These can be transferred with the yeastfrom one fermentation to another and even mul-tiply under favorable conditions during second-ary fermentation (5). Moreover, Pectinatus cer-evisiophilus and Megasphaera spp., recentlydiscovered gram-negative bacteria, form a spe-cial group of contaminants. Pectinatus spp. wereisolated and found to be potential beer spoilageorganisms almost simultaneously in the U.S.A.(12), Germany (1, 20), and Scandinavia (1, 20).

In this paper, a gram-negative anaerobic bac-terial strain isolated from Finnish beer in thesummer of 1978 is described and compared withthe P. cerevisiophilus isolate named by Lee etal. (12) and with the strains isolated in Germany.

MATERIALS AND METHODSOrganisms. The bacterial strain VTT-E-79100,

isolated in the summer of 1978 from turbid beer (orig-inal gravity, ca. 10.5%) was studied. Pectinatus cerev-isiophilus ATCC 29359, DSM 20465, and DSM 20466,Propionibacterium thoenii DSM 20276, and Propi-onibacterium acidi-propionici DSM 20272 were usedas reference strains.

Isolation. Isolation of VTT-E-79100 by membranefiltration or pour plating and incubation under anaer-

obic conditions (GasPak anaerobic system; BBL Mi-crobiology Systems, Cockeysville, Md.) was not suc-cessful. A special anaerobic chamber containing 80%N2, 10% C02, and 10% H2 was necessary for the inoc-ulation steps. The dilution solutions used werestrongly reducing and saturated with N2. The incuba-tion was performed in an anaerobic jar at 28°C withMRS agar plates (Difco Laboratories, Detroit, Mich.),which were prereduced anaerobically for several days.

Media. VTT-E-79100 and the reference strainscould be grown in liquid media if oxygen was removedby boiling before use. In addition to MRS broth,peptone-yeast extract-glucose medium was used. Glu-cose was substituted with other sugars when testingthe utilization of carbon sources.A frozen cell suspension was used as the inoculum

in the respirometer studies. After cultivation, the cellswere washed with phosphate buffer and frozen inbuffered 15% glycerol.

Electron microscopy. The samples for thin sec-tioning and freeze-fracturing were prepared as de-scribed by Lounatmaa et al. (13, 14) and for negativestaining as described by Bamford et al. (2).

Physiological examinations. The growth ofVTT-E-79100 in PYG after 1 week of incubation wasstudied over the pH range of 3.7 to 6.0 and temperaturerange of 10 to 45°C. Alcohol tolerance was tested inthree different types of beer containing 3.0 to 3.7, 3.7to 4.5, and 5.6 to 5.7% (wt/vol) ethanol.Biochemical tests. Peptone-yeast extract medium

(pH 6.5) containing 2% of the sugar to be tested wasused when studying the utilization of various carbonsources. After 1 week of incubation at 28°C, a turbidculture with pH below 5.0 was regarded as a positivereaction.

Respirometer tests were performed in a Gilson vol-umetric respirometer under nitrogen at 30°C, usingresting cells suspended in Sorensen phosphate buffer,pH 6.0. Buffered glucose solution (final glucose con-centration, 1%) was used as substrate for all strains

511

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

tested. Eleven other carbon sources were also studiedfor VTT-E-79100. CO2 was determined volumetrically,and acids formed were determined by gas chromatog-raphy. Because the assay time in the respirometer wasshort, the metabolic products formed by resting cellsin buffered 2% glucose solution during 1 day of incu-bation were studied.

Malonate inhibition tests were performed at pH 3.0and 4.0 in buffered 2% glucose solution containing 0.1M malonic acid. Metabolic products of VTT-E-79100and Propionibacterium species were analyzed by bothgas and liquid chromatography; those of the otherstrains were analyzed only by gas chromatography.

Volatile fatty acids (C2 to C5) were determined in aPerkin-Elmer Fll gas chromatograph equipped witha hydrogen flame ionization detector. A glass columnpacked with Carbopack C-Carbowax 20-M phosphoricacid (Supelco Inc., Bellefonte, Pa.) was used (7).

Acetoin, malonic, acetic, succinic, fumaric, and pro-pionic acids were determined in a Waters AssociatesM 6000 A liquid chromatograph with RI detector andMC 18 column (Waters Associates, P/N 27324). Dis-tilled deionized water adjusted with phosphoric acidto pH 2.3 was used as the solvent. The flow rate was0.7 ml/min, and the injection volume was 25 !d. Lacticacid and ethanol were determined enzymatically(Boehringer Mannheimn GmbH). In addition, the API20 anaerobe test was used for identification (API sys-tem S.A. Montalieu, Vercieu, France).Determination of DNA base composition. The

guanine-plus-cytosine content of deoxyribonucleicacid (DNA) was determined as described by Schild-kraut et al. (16).Immunological methods. Serum against VTT-E-

79100 was produced by injecting 0.3 to 0.6 ml of aformaldehyde-killed, washed bacterial suspension (ca.109 cells/ml) subcutaneously into rabbits. The bacte-rial suspension was emulsified in Freund incompleteadjuvant (1:1). Blood samples were taken during im-munization to control the level of antibodies. After 3.5months and seven injections, the rabbits were bledand the serum was freeze-dried. Immunofluorescentstaining was performed as described earlier (8, 9) byusing serum dilutions of 1:30 to 1:1,500. The gel diffu-sion precipitin test was performed according to themethods of Ouchterlony (15) and Lancefield (11).

RESULTSMorphology and electron microscopy.

The cells of strain VTT-E-79100 were gram neg-ative, nonsporeforming, and motile. The size ofthe cells was 0.4 to 0.5 by 2 to 3 um, but veryelongated cells were found in old cultures. Thecolonies on MRS agar were circular, entire,beige, and glistening, with a diameter of 2 mm.

Electron microscopy of a negatively stainedpreparation showed that the rod-shaped bacte-ria had many flagella, approximately six per cell(Fig. 1). The flagella of VTT-E-79100 appearedto be located on one side of the bacterial cellonly, thus resembling Pectinatus cerevisiophi-lus strain ATCC 29359 (12). In strain VTT-E-79100, however, there were fewer flagella.

In thin section, the cell envelope structurecould be seen (Fig. 2). The outer membrane(OM) of VTT-E-79100 undulated, and some ves-iculation of the OM was evident. The electron-translucent area just beneath the OM possiblycorresponds to a periplasmic space. The cyto-plasmic membrane (CM) and the thick, poorlyelectron dense layer, a peptidoglycan layer, wereclearly seen.

In a freeze-fractured preparation, both theconvex and concave fracture faces of the OMand CM were seen. The OM concave face wasdensely covered by 8- to 10-nm intramembra-nous particles (Fig. 3), similar to Salmonellatyphimurium (14), Escherichia coli (19) andSynechocystic strain CB3 (T. Vaara and K.Lounatmaa, FEMS Microbiol. Lett., in press).Similarly, the CM concave face was covered byonly a few particles (Fig. 3). The OM convexface was smooth, and CM convex face had anetlike arrangement of intramembranous parti-cles (Fig. 4). The ultrastructures of two otherPectinatus strains studied (ATCC 29359 andDSM 20465) in both thin sections and freeze-fractured cells (not shown) appeared to be sim-ilar.Physiological and biochemical charac-

terization. VTT-E-79100 grew well between pH4.5 and 6.0 and at temperatures between 15 and40°C, with optimal growth occurring at about30'C. It did not grow in beer with an alcoholcontent of 5.6 to 6.7% (wt/vol).

Strains VTT-E-79100, ATCC 29359, DSM20465, and DSM 20466 were catalase negative,did not liquefy gelatin, and did not produceindole. VTT-E-79100 and DSM 20465 wereurease negative; ATCC 29359 and DSM 20466urease positive. Arginine was not hydrolyzed,and the Voges-Proskauer test was positive forall strains tested.The utilization of different carbon sources by

Pectinatus spp. was determined by growth testin peptone-yeast extract medium (Table 1). Theformation rates of carbon dioxide, acetic acid,and propionic acid from fructose, glucose, glyc-erol, and mannose by VTT-E-79100 in the res-pirometer are given in Table 2. The amounts ofacids produced from buffered glucose by restingcells of the strains examined are presented inTable 3. All strains produced acetic, propionic,lactic, and succinic acids as well as acetoin.Ethanol was not produced.Addition of malonate decreased considerably

the production of propionic acid by all strains atpH 4.0 and inhibited production almost com-pletely at pH 3.0. The production of acetic acidincreased correspondingly (Table 4). Accumu-lation of fumaric acid was not observed. Theguanine-plus-cytosine content of DNA was 36,

512 HAIKARA ET AL.

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

CHARACTERIZATION OF A PECTINUS STRAIN

FIG. 1. Negatively stained dividing bacterium. Six flagella per cell located on one side of the cell ofPectinatus strain VTT-E-79100. Bar is 0.5 um.

FIG. 2. Thin section of Pectinatus strain VTT-E-79100. The CM, peptidoglycan layer (PG), and OM areclearly seen. The cell on the right side has been cut perpendicularly, and that on the left has been cutobliquely. Bar is 0.2 um.

39, and 34 mol% for VTT-E-79100, ATCC 29359,and DSM 29465, respectively.Immunological analysis. Serum prepared

against VTT-E-79100 was found to be very spe-cific. ATCC 29359 and DSM 20466 gave no

reactions with serum dilutions of 1:30 and 1:50,and DSM 20465 gave a weak fluorescence. Witha serum dilution of 1:1,000, only the cells ofVTT-E-79100 fluoresced brightly.The results of gel diffusion precipitin tests

VOL. 41, 1981 513

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

514 HAIKARA ET AL.

TABLE 1. Utilization of different carbon sources byPectinatus strains

FIG. 3. Outer concave fracture face (OM) withclosely packed 8- to 10-nm particles and the innerconcave fracture face (CM) covered by few particlesonly. Bar is 0.2 ,tm. Arrow represents direction of theplatinum shadowing.

Substrate

ArabinoseCellobioseErythritolFructoseGalactoseGlucoseGlycerolGlycogenLactoseMaltoseMannitolMannoseRaffinoseRhamnoseRiboseSorbitolSorboseStarchSucroseTrehaloseXyloseAdonitolAmygdalinDextrinDulcitolInositolInulinMelezitoseMelibioseSalicin

VTT-E-79100

+a

+

++

+

ATCC29359

+

++

DSM20465

+++++++

+++

+++

DSM20466

+

+++++

+

++

+++

+

FIG. 4. Outer convex fracture face (OM), which issmooth and quite particleless, and inner convex frac-ture face (CM), which has numerous netlike arrangedparticles. Bar is 0.2 ,um. Arrow represents directionof the platinum shadowing.

were in good agreement with those obtainedfrom fluorescent-antibody staining. A strongprecipitin band was found near the well ofVTT-E-79100 antigens (Fig. 5). This band probablyrepresented the interaction of a relatively spe-cific antibody with the respective antigen ofVTT-E-79100. The common antibodies presentin the antiserum gave precipitin bands with eachof the strains examined. This reaction was notvisible when immunofluorescent staining withdiluted antiserum was used. On the other hand,VTT-E-79100 and DSM 20465 showed two ad-ditional antigens which gave precipitin bandsmissing from ATCC 29359 and DSM 29466.These antigens are probably common to bothstrains. The reaction of these antigens and re-spective antibodies gave a moderate fluores-cence after staining if the serum dilution wasless than 1:150.

DISCUSSIONThe morphological properties of VTT-E-

79100 are quite similar to those of the strains

a +, Turbid culture with pH below 5.0.

TABLE 2. Rate offormation of metabolic productsby VTT-E-79100 in 1% sugar solution

Rate of formation (mol/h per mg, dry wt)Carbonsource Carbon diox- Acetic acid Propionic

ide acid

Fructose 16 x 10-7 16 x 10-7 46 x 10-7Glucose 18 x 10-7 12 x 10-7 51 x 10-7Glycerol 3 x l-8 2 x 10-7 5 x 10-7Mannose 12 x 10-7 10 X 10-7 36 x 10-7

isolated by Lee et al. (12) and Back et al. (1).The ultrastructures of those strains also appearto be identical. Thus, no differences among thestrains could be demonstrated, and all of thestrains studied seem to be closely related.

Physiological as well as most biochemicalcharacteristics of VTT-E-79100 proved to besimilar to those of Pectinatus strains ATCC29359, DSM 20465, and DSM 20466. Differencesin hydrolysis of urea were found among thestrains examined. The urease tests were negativefor VTT-E-79100 and DSM 20465 and positivefor ATCC 29359 and DSM 20466. Back et al. (1)reported that the urease tests of strains DSM20465 and DSM 20466 were negative.

APPL. ENVIRON. MICROBIOL.

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

CHARACTERIZATION OF A PECTINUS STRAIN 515

TABLE 3. Metabolic products formed by Pectinatus strains'

Concn (mmol dm-')Pectinatus strain

Acetic acid Propionic acid Lactic acid Succinic acid Acetoin

VTT-E-79100 12 116 1.0 0.4 17ATCC 29359 12 57 0.7 0.4 15DSM 20465 17 78 0.8 1.3 15DSM 20466 24 74 0.8 0.9 12

a Experiments were carried out with resting cells in a 2% glucose solution in phosphate buffer, pH 6.0.Incubation time was 1 day.

TABLE 4. Effect ofmalonate on the production ofacetic and propionic acids atpH 3.0

Concn (mmol dm-3)Bacterial strain Acid Malonic

Control acid added

Propionibacte- Acetic 1 28rium thoenii Propionic 13 3DSM 20276

P. acidi- Acetic 1 30propionici Propionic 11 2DSM 20272

VTT-E-79100 Acetic 5 33Propionic 25 3

ATCC 29359 Acetic 9 45Propionic 58 4

DSM 20465 Acetic 14 42Propionic 75 6

DSM 20466 Acetic 7 32Propionic 27 4

Slight differences were found in the utilizationofcarbon sources. VTT-E-79100 and DSM 20465differed only in fermentation of sorbose (Table1). DSM 20466 differed in utilization of lactose,maltose, sorbose, and xylose; ATCC 29359 dif-fered in utilization of maltose and mannitol.Differing results obtained by Lee et al. (12) andBack et al. (1) may be due to the differences inmethods used for testing the fermentation ofsugars.

All strains examined produced acetoin andpropionic, acetic, lactic, and succinic acids. Backet al. (1) could not, by chromatographic or en-

zymatic methods, detect any formation of lacticacid by strains DSM 20465 and DSM 20466. Theamounts of acetoin and of acetic, lactic, andsuccinic acids produced were of the same orderof magnitude with all four strains (Table 3).However, these results differ considerably fromthose reported by Lee et al. (12). According tothat group more acetic acid than propionic acidis produced, whereas in our experiments pro-pionic acid formation was always greater.

In propionic acid fermentation by propioni-bacteria, succinic dehydrogenase catalyzes thereduction of fumarate to succinate (10). Malo-nate is a specific competitive inhibitor of thisenzyme and can thus inhibit the production of

propionic acid. It was found that addition ofmalonate prevented the production of propionicacid by Propionibacterium thoenii, P. acidi-propionici, and strains VTT-E-79100, ATCC29359, DSM 20465, and DSM 20466. This indi-cates that the Pectinatus strains tested use thesame pathway for the production of propionicacid as do propionibacteria. The production ofacetic acid increases when malonate inhibits theformation of succinate, and pyruvate is metab-olized via an alternative pathway to acetic acid.The difference of 2 to 5 percentage points in

the DNA base composition of VTT-E-79100(36%), ATCC 29359 (39%), and DSM 20465(34%) indicates that these strains are probablyvery closely related but not identical. The resultobtained for ATCC 29359 is the same as thatreported by Lee et al. (12) (39.8%), but theguanine-plus-cytosine content of DSM 20465 isdifferent from the value 37.4% reported by Backet al. (1).The existence of an antibody reacting only

with antigens of VTT-E-79100 in immunodiffu-sion experiments suggests that this organismbelongs to a different taxonomic group than thereference strains. For example, the lactobacillicommonly encountered in breweries show differ-ent antigenic structures when they belong todifferent species or are variants of the samespecies (6). Further experiments with absorbedserum and sera against reference strains, how-ever, are needed to confirm the immunologicaldifferences between Pectinatus species. StrainVTT-E-79100 is closely related to the species P.cerevisiophilus described by Lee et al. (12) andcan be regarded as belonging to the genus Pec-tinatus.

Further work is also needed to facilitate thedetection ofPectinatus contaminants in breweryquality control. As quoted above, conventionalmembrane filtration is not practical. Since Pec-tinatus probably is the only bacterium in beerwhich produces propionic acid, its identificationcan be based on detection of propionic acid ingas chromatography (17). Preliminary studiesusing artificially contaminated beer showed thatpropionic acid can be detected independent ofthe initial level of Pectinatus cells in beer only1 day before the beer becomes visually turbid.

VOL. 41, 1981

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

FIG. 5. Gel diffusion precipitin test of antiserum against VTT-E- 79100 uwith extracts of Pectinatus strains.I, control; II, VTT-E- 79100; III, VTT-E- 79100; IV, ATCC 29359; V, DSM 20465; VI, DSM 20466.

Despite some early information in the litera-ture (3, 4) on a bacterium which fits the descrip-tion of Pectinatus spp., the simultaneous occur-rence of this species in different countries mustbe regarded as a new phenomenon. This devel-opment is probably due to changes in the brew-ing process and to new cleaning and disinfectionmethods. These changes have caused alterationsin the relative survival of different microflora.New microbes, previously not found in brewer-ies, are now able to grow under the changedconditions. For example, in the case of Pectina-tus, the low level of air now present in bottledbeer guarantees good growth conditions for thisbacterium (18). Equally plausible is that thedevelopment of anaerobic laboratory techniquesnow has made the isolation and identification ofthis organism possible, thus revealing a contam-inant which has occurred sporadically over theyears. More information is needed about theorigin, detection, and taxonomic position ofthese new beer spoilage microorganisms.

Pectinatus strain VTT-E-79100 has been de-posited in the American Type Culture Collectionunder the number 33332.

ACKNOWLEDGMENTSWe thank Hans Soderlund for the DNA base composition

analysis, Kari Juntunen for the preparation of antisera, andTuire Koro for skilled technical assistance.

T.-M.E. is a research professor of the Academy of Finland.

LITERATURE CITED

1. Back, W., N. Weiss, and H. Seidel. 1979. Isolierung undsystematische Zuordnung bierschadlicher gramnega-tiver Bakterien. II: Gramnegative anaerobe Stabchen/Anhang: aus Bier isolierte gramnegative fakultativStabchen. Brauwissenshaft 32:233-238.

2. Bamford, D. H., E. T. Palva, and K. Lounatmaa. 1976.Ultrastructure and life cycle of the lipid-containingbacteriophage 0 6. J. Gen. Virol. 32:249-259.

3. Chevalier, M. 1946. La bacteriologie et les bieres legeres.Petit J. Brass. 54:186-191.

4. Dachs, E. 1974. Neue biologische Gefahren in derBrauerei. Domensianer 14:19-21.

5. Dachs, E. 1978. Innovation in Bereich der biologischenBetriebskontrolle. Brauwelt 119:1519-1523.

6. Dolezil, L., and D. H. Kirsop. 1975. An immunologicalstudy of some lactobacilli which cause beer spoilage. J.Inst. Brew. London 81:281-286.

7. Du Preez, J. C., and P. M. Lategan. 1978. Gas chro-matographic analysis of C2-Cr, fatty acids in aqueousmedia using Carbopack B-Carbowax 20 M-Phosphoricacid. J. Chromatogr. 150:259-262.

8. Haikara, A., and T.-M. Enari. 1975. The detection ofwild yeast contaminants by the immunofluorescence

516 HAIKARA ET AL.

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from

CHARACTERIZATION OF A PECTINUS STRAIN 517

technique, p. 363-375. In Proceedings of the 15th Con-gress Eur. Brew. Conv., Nice.

9. Haikara, A., and V. Makinen. 1972. Nachweis von inBrauereien auftretenden wilden Hefen mittels des im-munologischen Fluoreszenzverfahrens. Brauwissen-schaft 25:266-271.

10. Hettinga, D. H., and G. W. Reinbold. 1972. The pro-

pionic-acid bacteria. A review. II. Metabolism. J. MilkFood Technol. 35:358-372.

11. Lancefield, R. C. 1933. A serological differentation ofhuman and other groups of hemolytic streptococci. J.Exp. Med. 57:571-595.

12. Lee, S. Y., M. S. Mabee, and N. 0. Jangaard. 1978.Pectinatus, a new genus of the family Bacteroidaceae.Int. J. Syst. Bacteriol. 28:582-594.

13. Lounatmaa, K., P. H. Makela, and M. Sarvas. 1976.Effect of polymyxin on the ultrastructure of the outermembrane of wild-type and polymyxin-resistant strainsof Salmonella. J. Bacteriol. 127:1400-1407.

14. Lounatmaa, K., and N. Nanninga. 1976. Effect of poly-myxin on the outer membrane of Salmonella typhi-murium: freeze fracture studies. J. Bacteriol. 128:665-667.

15. Ouchterlony, 0. 1958. Diffusion in gel method for im-munological analysis. Prog. Allergy 5:1-78.

16. Schildkraut, C. L., J. Marmur, and P. Doty. 1962.Determination of the base composition of deoxyribo-nucleic acid from its buoyant density in CsCl. J. Mol.Biol. 4:430-443.

17. Schisler, D. O., M. S. Mabee, and C. W. Hahn. 1979.Rapid identification of important beer microorganismsusing gas chromatography. J. Am. Soc. Brew. Chem.37:69-76.

18. Seidel, H., W. Back, and N. Weiss. 1979. Isolierung undsystematische Zuordnung bierschadlicher gramnega-

tiver Bakterien. III: Welche Gefahr stellen die in denbeiden vorausgegangenen Mitteilungen vorgestelltengramnegativen Kokken und Stabchen fur das Bier dar.Brauwissenschaft 32:262-270.

19. van Gool, A. P., and N. Nanninga. 1971. Fracture facesin the cell envelope of Escherichia coli. J. Bacteriol.108:474-481.

20. Weiss, N., H. Seidel, and W. Back. 1979. Isolierung undsystematische Zuordnung bierschadlicher gramnega-

tiver Bakterien. I: Gramnegative strikt anaerobe Kok-ken. Brauwissenschaft 32:189-194.

VOL. 41, 1981

on March 26, 2021 by guest

http://aem.asm

.org/D

ownloaded from