JOURNAL OF BACTERIOLOGY, May, 1965Copyright © 1965 Amtierican Society for Microbiology

Vol. 89, No. 5Printed in U.S A.

Spectrophotometric Characterization of PigmentsProduced by Staphylococcus epidermidis Strains

Isolated from Bovine UddersOLAV SANDVIK' AND RICHARD W. BROWN

National Animal Disease Laboratory, U.S. Department of Agriculture, Agricultural Research ServiceAmes, lowa

Received for publication 2 December 1964

ABSTRACTSANDVIK, OLAV (National Animal Disease Laboratory, Ames, Iowa), AND RICHARD

W. BROWN. Spectrophotometric characterization of pigments produced by Staphylo-coccus epidermidis strains isolated from bovine udders. J. Bacteriol. 89:1201-1208.1965.-Spectrophotometric analysis of methanol-extracted pigments of 70 isolates,representing 48 strains of Staphylococcus epidermidis, exhibited absorption curves thatwere classified into seven types, designated I, II, III, IV, V, VI, and a S. aureus type.Two subtypes were included in types I and III. All nonpigmented cell extracts wereclassified as type I. One of the six numbered type absorption curves was obtained withthe majority of isolates. Three isolates gave an absorption curve that was similar tothe curve produced by extracts of five of seven S. aureus strains, thus the designationS. aureus type. The differences in pigment complexes indicated by the various ab-sorption curves of methanol extracts were substantiated by column-chromatographystudies. Generally, pigments of types II to VI, as produced by representative strains,were of a xanthophyllic nature, whereas, the S. aureus type pigment studied wascarotenelike. Analyses of representative strains showed that the type of spectral ab-sorption curves did not change whether the organisms were carried in vitro and testedthrough 3 months, or isolated repeatedly from infected udders for periods up to 8months. The method of determining the spectral absorption curves of whole-cell metha-nol extracts provides an additional tool for differentiating strains of S. epidermidisthat can be used in enzootiological studies of udder infections.

In a newly assembled herd of 20 heifers, Staphy-lococcus epidermidis was found to be the principalcause of udder infections. This organism has oftenbeen considered nonpathogenic, but in recentyears it has been realized that it should not beneglected as a cause of mastitis. Although it doesnot cause a severe mastitis, it can cause a leuko-cytosis in milk and lowered milk production insome infected quarters. To study the patho-genicity and spread of this organism in the herd,a differentiation of strains was considered neces-sary. One property that appeared to be charac-teristic and constant for individual strains was thepigmentation of colonies. Pigmentation of staphy-lococci has long been of taxonomic interest, al-though it is generally considered of little value inclassifying these organisms. Recent studies, how-ever, indicate it may have some value for identi-fication of S. aureus strains in epidemiological

I Present address: Department of Microbiologyand Immunology, Veterinary College of Norway,Oslo.

investigations (Sompolinsky, 1962; Willis andTurner, 1962). In addition, different carotenoidshave been demonstrated in the extracts of variouspigmented bacteria and yeasts by chromato-graphic and spectrophotometric methods (Sobinand Stahly, 1942; Peterson et al., 1954; Fox andLewin, 1963). Such methods have been used byEllinghausen and Pelczar (1955) to differentiateNeisseria species. In view of these results, a studyof the pigments of S. epidermidis by spectropho-tometric methods was conducted with the hopethat it might provide a means of differentiatingstrains.

MATERIALS AND METHODSCultures. We studied 74 isolates representing 50

strains of S. epidermidis. The organisms were iso-lated from milk samples taken weekly from an ex-perimental herd of 20 cows. A fresh sample fromeach quarter was streaked on a bovine blood-agarplate and incubated for 24 hr at 37 C and for 24 hrat room temperature. Except when indicated,transfer was made from single representative

1201

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

SANDVIK AND BROWN

colonies into two tubes of beef infusion broth.After incubation at 37 C for 24 hr, one culture wasstored at -65 C, and the other was used for identi-fication. The term "strain," as used here, refers tothe organism causing a chronic infection of aquarter of an udder. The term "isolate" is usedto designate the organisms of the same strain thatwere repeatedly isolated from the same infectedquarter.The cultures were classified as S. epidermidis

on the basis of negative coagulase tests andanaerobic fermentation of glucose, but not manni-tol, according to Bergey's Manual. A tube methodwith citrated rabbit plasma was used for thecoagulase test. Readings were made after 4 and24 hr. The anaerobic fermentations were deter-mined by stab cultures, in Phenol Red Agar Base(Difco) with 1% carbohydrate; the cultures wereincubated for 5 days at 37 C in a Brewer jar.Twelve of the isolates were nonpigmented, and theothers varied in pigmentation from pale yellow toorange.For the study of pigment production, frozen

cultures were thawed, and 0.1 ml was transferredto a tube of beef infusion broth and incubated for20 to 24 hr at 37 C. Cultures were then streaked onbovine blood-agar and incubated for 24 hr at 37 C,and again at room temperature. A representativecolony was transferred to a Tryptose Agar (Difco)slant which was incubated for 24 hr at 37 C andthen held at 4 C until tested.

Seven strains of S. aureus isolated from bovineudders and two strains isolated from man werestudied for comparative purposes. All strains werecoagulase-positive and fermented both glucoseand mannitol anaerobically.

Preparation of pigment extracts. A modificationof the method described by Ellinghausen andPelczar (1955) was used for the preparation of pig-ment extracts. Cultures were inoculated into 10 mlof Trypticase Soy Broth (BBL) from fresh Tryp-tose Agar slant cultures and incubated at 37 C for20 to 24 hr. The 10 ml of culture were then trans-ferred to 200 ml of Trypticase Soy Broth in a 1-liter flask and incubated for 24 hr on a rotaryshaker (New Brunswick Gyratory Shaker modelS-3) at 80 rev/min. In some instances, the cultureswere incubated for 48 hr to determine the effectof time on pigment production. After incubation,the contents of a flask were centrifuged for 10 minat 8,000 X g. The supernatant fluid was discarded,and the cells were washed twice with saline(0.85%; w/v) and once with 70% acetone.After the acetone wash, 0.5 g of wet, packed

cells was immediately weighed and placed into a125-ml screw-capped Erlenmeyer flask. A 25-mlamount of absolute methanol, heated to boiling ina water bath (80 C), was poured into the flask con-taining the cells. The flask was placed in the waterbath (80 C) for 5.5 min and then held for 2 hrat 4 C. After refrigeration, the suspension wasbrought to its original volume of 25 ml with abso-lute methanol and filtered through Whatman no. 1filter paper. Extracts with little or no visible pig-

ment were concentrated three times by evapora-tion with a stream of nitrogen at room tempera-ture.

Spectral absorption curves of the extracts wererecorded over a wavelength range of 320 to 620 m,uin a Beckman model DB spectrophotometer witha Photovolt Linear/Log Varicord model 43 re-corder. The data were then replotted on equal di-vision graph paper. Absorption curves obtainedby measuring optical densities in a Bausch &Lomb Spectronic-20 colorimeter (model 340) indi-cated that the latter instrument was not suffi-ciently sensitive to provide the detail necessaryfor differentiating the pigments studied.

Fractionation of pigment extracts. Extracts wereprepared as described above, except at least 4 gof wet cells were extracted with 25 ml of absolutemethanol per gram of cells. The extracts were con-centrated two times by evaporation with a streamof nitrogen, and 1 g of NaOH was added per 10 mlof extract (Suzue, 1959). After the NaOH dis-solved, the mixture was left at room temperatureovernight. Twice the volume of distilled waterwas added, and the mixture was extracted threetimes by shaking with petroleum ether (bp, 30 to60 C). The ether fraction, which also containedsome precipitated pigmented material, was washedthree times with small amounts of distilled waterand then evaporated to dryness with streamingnitrogen. The residue was resuspended in pe-troleum ether.The petroleum ether suspensions were chro-

matographed on columns of packed magnesiumoxide and celite, 1:1 by weight (Fox and Lewin,1963). Colored zones were separated and elutedindividually by adding solvent that contained in-creasing amounts of absolute methanol from 0 to100%. Each fraction was evaporated to drynesswith streaming nitrogen, and the residue was dis-solved in absolute methanol for determination ofa spectral absorption curve.The main fractions isolated were tested by par-

tition between hexane and 95% methanol, accord-ing to the method described by Petracek andZechmeister (1956).

RESULTSSpectral absorption curves of methanol extracts.

Spectrophotometric analysis of methanol-ex-tracted pigments of 68 isolates of S. epidermidisexhibited absorption curves that were classifiedinto 6 types, designated I, II, III, IV, V, and VI.Two subtypes were included in types I and III.Representative absorption curves of the varioustypes are shown in Fig. 1, with their main char-acteristics summarized in Table 1. For compari-son, the curve of one strain of S. aureus is shown(Fig. 2).The curves of types I and II start with com-

paratively high optical densities at 320 m,u anddecrease with increasing wavelengths. The con-tinuity of curves lb and II are interrupted by a

1202 J. BACTERIOL.

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

028

026

024

I- 022

0.18-0.16

0.14

° 0.12-

0.10

0.08

006-0.04

0.02-

3

030-

028-

026-

024-

t 022-

Z 020-

0 l8-

< 016-

1: 0 14-o 012-

o 10-

008-

006-

004-

00-

0.30-

028-

0.26

0.24

Q022Z 020

0.18

< 0.16

1 0.14

° 0.12

0.10

0.0e

0.06

0.04

0.02

i ... ............

D0 400 500

WSWELENGTH mjs

400 'LDOWAELENGTH m,A

TYPE MI CURVE

400 530WAVELENGTH m#

600 700

TYPE X CURVE

4W G500WAVELENGTH m

400 50 60WAVELENGTH m,u

600 700

FIG. 1. Spectral absorption curves of Staphylococcus epidermidis pigments (methanol extracts) repre-senting the curve types I, II, III, IV, V, and VI.

1203

TYPE I CURVE

0.32

030

0.28-

026

024-

t 022

Z 0.20

a ole4 0 16

F: 0 14-

o 012

0 10

008-

006-

004-

002-

076

074

072-

070-

066-

0.66-

0.64-

0.62

0.60-0.58-

0.56-

0.54-

052-0.50-

0.460.46-

0.44-

0.42-040-

038

036034-

0.32-

0.30-

028-

026-

0.24-

022-020 -

0.18-

0.16-

0.14-

0.12-

0.10-

0.06-0.06-0.04-

0.02-

0.38-

036 -

0.34-0.32-

0.30-

0.28-

0.26-

0.24-

0.22-Z 0.20-O 0.16

0.16-0-014-o 0,12-

0 10-

008-

0.06-

0.04-

0.02-

400 500

WAVELENGTH m,600 700 0

LJ

Qi4

WAVELENGTH mj,

1-

I-

1-

i-

1-

-I3C

1-

1-

1-

i-

1-

I

4so300

....m300

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

SANDVIK AND BROWN

peak at 400 m, and additional shoulders orsmaller peaks at higher wavelengths in the typeII curve. The pigments of one strain gave an ab-sorption curve that was classified as a type II

TABLE 1. Characteristics of the different spectralabsorption curves obtained with methanolextracts of Staphylococcus epidermidis

Type of No. of Other Shoulderabsorption isolates Main peak distinct or

curve peaks plateau

myp m,u msIa 8lb 4 400II 9 400 426,465IIIa 16* 402,425 447 381IIIb 11* 425 402,447 381,472IV 11 448 425 405,474V 8 451 338,476 430VT 4 407 425.-470v X v1 -vI

S. aureus 3 450-470 330-Not 3

classified

* Includestypes.

I--

U),z

-J

()

a-

0

0.52-

0.50

0.48-

0.46

0.44

0.42-

0.40

038-

038 -

0.34

0.32-

0.30

028

0.26-

0.24

Q22

020-

0.18-

0.16

0.14

0.12-

0.10-

0.08

0.06

0.04-

0.02-

300

three isolates that

400 500

WAVELENGTH mu.

FIG. 2. Spectral absorption curvlcoccus aureus pigments (methanol e:

0.24-t- 0.22-U) -

Z 020-wL -

D 0.18:-J< 0.16-

0. -

o 0.120.10-

0.06-0.06-0.04-0.02-

300 400 500WAVELENGTH m,

600 700

FIG. 3. Type IIIa spectral absorption curves ofStaphylococcus epidermidis pigments (methanol ex-tracts) after 24 and 48 hr of incubation of the culture.

-338 curve; however, it differed by showing three addi-tional small peaks at 332, 347, and 367 m,u.The other type curves showed absorption

maxima at typical locations between 400 and 475produce both mw. The type IIIa curves showed variation in the

height of the peaks at 402 and 425 m,u. Some-times the two peaks reached the same height, but,in most instances, only one of the peaks pre-dominated (Fig. 1 and 3). The two type III curveswere differentiated by the fact that the 402-m,peak was higher than the 447-mu peak in theIIIa curves, whereas the reverse was true in theIIlb curves. The type IV curves, although similarin form to the type III curves, always showed themain absorption peak at 448 m,u. The type VIcurve resembled the type II curve, but differedby having the main peak at 407 instead of 400 mIA,followed by an irregular area or plateau of in-creased absorption between 425 and 470 m,u. Thetypes II and VI curves were obtained in allinstances from extracts that were concentratedthree times.

Three strains of S. epidermidis gave absorptioncurves similar to the S. aureus curve; however,the maximal optical densities were lower thanthe value shown (Fig. 2). The spectral absorptioncurves of three isolates representing two strainswere not classified, because poor yields of cellsfrom the shaken cultures provided insufficientpigment for satisfactory tests.

Experiments were performed to determinewhether the absorption curves represented con-stant phenotypic properties or varied after longerincubation periods, repeated testing, or repeated

600 700 isolation of the same strain. The effect of the in-cubation period is shown by comparing the type

e of Staphylo- IIIa absorption curves of a culture incubated forrtract). 24 and 48 hr (Fig. 3). Except for the increase in

i I. tj

1204 J. BACTERIOL.

I

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

PIGMENTS PRODUCED BY S. EPIDERMIDIS

height of the two main peaks at 402 and 425 mj.t

and the decrease in the minor peak at 447 m,4, theadditional 24 hr of incubation did not alter thecharacter of the curve. Similar results were ob-tained with another strain of the same subtype,except that the minor peak at 447 mA was lesspronounced and appeared as a shoulder at 24 hr.To determine the effect of repeated testing, a

strain exhibiting the type IIIa absorption curvewas subcultured, and extracts were preparedseven times during a 5-week period. Between eachtest, the strain was stored at room temperature onTryptose Agar slants. In six successive tests, thetype IIIa curve was obtained, but on the seventhtest, a IlIb curve developed. In experiments inwhich four other strains exhibited a type IIIcurve and were stored on agar slants for 12 to 94days, two changed from one subtype to the otherwhen retested. In similar experiments with strainsproducing pigments of types II, IV, and V, no

changes were demonstrated with 2 to 4 tests.Strains of S. aureus stored and subcultured undersimilar conditions showed some variations indetails of the absorption curves, although, withfive strains, the main peak always occurred be-tween 450 and 470 m,. Altogether, S. aureusseemed more variable and sensitive to changes inincubation time and aeration than S. epi&rmidis.To determine whether the same type absorp-

tion curve was obtained with a strain isolated re-

peatedly from the same source of infection, two

zILJ0

-J

P:0

0.34-

0.32-

0.30-

0.28-

0.26-

0.24-

0.22-

0.20-

0.18-

0.16-

0.14-

0.12-

0.10-

0.08-

0.06-

0.04-

0.02-

300

A-- B

C

400 500

WAVELENGTH m#FIG. 4. Type IV spectral absorpti

tained with pigments (methanol e

Staphylococcus epidermidis strain is(infected quarter four times at weeklthe order A to D.

084-

082-

0 80-

0 78-

0 76

0 74

0 72-

070-

068

0.66-

0.64-

0.62

0.60

0.58-

0.56-

0 54-

0.52

0.50

0.48-

0.46-

z 0.44

0.42-

- 0.38

o 0.36

0.34-

0.32

0.30

0.28-

0.26-

0.24-

022 -

0.20-

0. 18 -

0.lb-

0.14-

0.12 -

0.10-

0.08-

0.06-

0.04-

0.02-

300 400 500

WAVELENGTH mA

-A--- 8

_~C---

600 700

-D FIG. 5. Type V spectral absorption curves obtainedwith pigments (methanol extracts) of a Staphylo-coccus epidermidis strain isolated from an infectedquarter four times at weekly intervals in the orderA to D.

,,TT7W7T1,,, cows, each infected in one quarter with S. epi-600 700 dermidis, were selected for study. One of the in-

fecting strains gave a type IV curve and theion curves ob-

other a type V curve. Milk samples from infectedxtracts) of a quarters were taken four times at weekly inter-olated from an vals and inoculated on bovine blood-agar plates.'y intervals in After incubation, a representative colony was

picked and inoculated into Trypticase Soy Broth.

VOL. 89, 1965 1205

.......

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

1206 SANDVIK AND BROWN

Extracts were then prepared, and the absorptioncurves were determined as described previously.The four absorption curves for the isolates fromthe two different quarters are shown (Fig. 4 and5). There was no change in the type of curvedemonstrated on repeated isolation of each strain,although quantitative differences between iso-lates were evident. Two analyses were also madeof 13 strains isolated from other chronically in-fected quarters at intervals of 7 days to 8 months.The absorption curves represented by the strainsweretypes I (1), III (7), IV (1), V (1), and VI (3).No changes in the type curves were demonstrated,although three strains changed from one subtypeIII to the other.

Fractionation of pigment complexes. The pig-ment complexes extracted from representativestrains were separated by column chromatog-raphy. At the beginning of the elution, the meth-anol content of the eluent was increased veryslowly to permit separation of pigment compo-nents that showed only slight differences inmobility. The results of these studies are sum-marized in Table 2. An example of the absorptioncurves of the main pigment fractions separated

J. BACTERIOL.

from the pigment complex responsible for a typeIV curve is presented (Fig. 6). At least two to fivecomponents were demonstrated from each of therepresentative pigment extracts. Usually, one ortwo of the components were quantitatively pre-dominant. The types IlIa and IIIb had whatappeared to be the same main pigment in fraction3, and also shared another component, fraction 1,with type IV. A high degree of similarity alsoexisted between fraction 2 of pigment type V andfraction 3 of the pigment produced by the repre-sentative S. aureus strain.Each pigment fraction was further charac-

terized as being xanthophyllic or carotenelike innature by partition between 95% methanol andhexane (Table 2). Some xanthophyllic fractionswere so strongly adsorbed to the column that theycould not be eluted with absolute methanol; how-ever, these fractions were of minor quantitativeimportance. The majority of fractions were green-ish-yellow in methanol, although fraction 1 oftypes IlIa, IlIb, and IV and fraction 2 of the S.aureus extract had a more reddish color.When the pigments were not completely ex-

tracted from the alkaline methanol-water solu-

TABLE 2. Characteristics of the main pigment fractions separated by chromatography of methanolextracts of Staphylococcus epidermidis strains giving representative type absorption curves

Partitiont

HypophasicHypophasic

Mainly epiphasicMainly hypo-

phasicHypophasicHypophasicHypophasicMainly epiphasicMainly hypo-

phasicHypophasicMainly epiphasicMainly epiphasicHypophasicHypophasicMainly epiphasicHypophasicHypophasic

EpiphasicMainly epiphasicHypophasic

Kind of pigment

XanthophyllicXanthophyllic

CarotenelikeXanthophyllic

XanthophyllicXanthophyllicXanthophyllicCarotenelikeXanthophyllic

XanthophyllicCarotenelikeCarotenelikeXanthophyllicXanthophyllicCarotenelikeXanthophyllicXanthophyllic

CarotenelikeCarotenelikeXanthophyllic

Remarks

Main fractionCould not beeluted

Main fraction

Main fraction

Main fraction

Main fractionCould not beeluted

Main fraction

Type ofabsorptioncurve*

II

IIIa

IlIb

IV

V

S. aureus

Frac-tionst

12

12

34512

31234123

123

Mainpeak

MIA400

423427

426399427423427

426423445421448427447

463450448

Otherdistinctpeaks

428

466446

402,447380,424

402466

402,443

402,447466

399,447477

338,477

425477

Shoulders orplateaus

388365,388,465

362,382360,446388,450

388385,465

362,382388

425-427379425

405,460424

330-338,482405,477,512

425-427

* Ia and lb type absorption curves were from nonpigmented strains.t Eluted in numerical order by increasing methanol gradient.t Partition between 95% methanol and hexane.

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

PIGMENTS PRODUCED BY S. EPIDERMIDIS

-0.22-

Z 0.20-

0.18-

< 0.16-

0.14 / UN-FRACTIONATED00-- FRACTION I

0 0.12- FRACTION 2

*........FRACTION 3

0.10 FRACTION 4

0.08-

0.046/0.02-0.0.300 400 500 600 700

WAVELENGTH m,s

FIG. 6. Spectral absorption curves of pigmentfractions chromatographically separated from thepigment complex of a Staphylococcus epidermidisstrain, curve type IV. Readings were made in abso-lute methanol.

tion by the petroleum ether, the methanol-watersolution was evaporated to dryness, after whichthe residue was taken up in petroleum ether andchromatographed as described above. Fractionsisolated in this way showed curves closely relatedto the main component of the original extract.

Extracts from strains producing the type VIcurve have not been studied with chromatog-raphy.

DISCUSSIONThe results of these experiments indicate that

strains of S. epidermidis, including nonpigmentedstrains, can be divided into at least six types byspectral analysis of methanol extracts of wholecells. In addition, some strains may exhibit spec-tral absorption curves similar to those of the S.aureus strains investigated. Differentiation of thestrains on the basis of the type of absorptioncurve was substantiated by chromatographicseparation of the pigment complexes produced byrepresentative strains. In only one instance, whena minor fraction was common to types Illa, IIIb,and IV, was a pigment fraction shared by differ-ent types. This may be expected in view of thesimilarity in shape of these three absorptioncurves in which some of the peaks and shouldersoccur at the same wavelengths. The fact that thesubtypes IIIa and b also shared the same mainpigment component is one reason for classifyingthem together. Their relationship is further indi-cated by the fact that some strains produced pig-ments of both subtypes on different tests. It isnot clear, however, if these changes are due togenotypic variations. The Ia and Ib curves areclassified together because (i) both were pro-

duced by extracts from nonpigmented strains,and (ii) the Ia curve was obtained, in some in-stances, before concentration of the extract, andthe lb curve was obtained after concentration.The type VI curve closely resembles the type

II curve, which may suggest that they be classi-fied together. The fact that no change from onecurve type to the other has been seen, and thatthe organisms which gave one or the other typecurve differed in their pigmentation on solidmedia and in various biochemical tests seems tojustify considering them different until furtherstudies show otherwise.The S. epidermidis strains produced pigments

that were primarily xanthophyllic in nature,whereas the pigments of the one S. aureus straintested was primarily carotenelike. It should beemphasized that, because of the carotene natureof the S. aureus pigment, the shape of the ab-sorption curve (Fig. 2) would be different if thepigment had been measured in petroleum etherrather than methanol. Although the pigment ex-tracts of three S. epidermidis strains gave absorp-tion curves similar to the S. aureus curve (Fig. 2),it was not determined by fractionation studieswhether the pigments of the former also wereprimarily carotenelike rather than xanthophyllicin nature.

In spite of the complicated pigment composi-tion of the methanol extracts (Table 2), the ab-sorption curves for extracts of S. epidermidis werevery constant, even when the strains existed forlong periods in vitro or in vivo. Only minorchanges in the pattern of the absorption curveswere demonstrated when the cultures were sub-cultured repeatedly or incubated an additional 24hr. The effect of incubation was studied with onlytwo strains that exhibited type III absorptioncurves. It was felt that any effect would bedemonstrated more readily by these strains be-cause of their greater variability, as indicated bytheir tendency to produce both subtype curves.Since no basic change in the absorption curves de-veloped (Fig. 3), additional studies with strainsthat gave the other type of absorption curves werenot considered necessary to fulfill the aims of thisstudy. It is of particular interest to note thatstrains isolated repeatedly from chronically in-fected udders always produced pigments thatgave the same type curve. Thus, the differencesof the curves (Fig. 4 and 5) represent quantitativerather than qualitative changes in the pigments.This constancy in the absorption peaks andshoulders seems to indicate that certain environ-mental changes do not alter extensively the char-acter of the pigment complex extracted by boilingmethanol. Therefore, this character of pigmenta-tion seems to represent constant properties of the

VOL. 89, 1965 1207

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

SANDVIK AND BROWN

organisms. On the other hand, more varyingpatterns were obtained with S. aureus, which maybe due to the predominating content of carotene-like components in the pigment complex. Ourlimited studies with S. aureus agree with those ofSuzue (1959), who found the absorption curves

of the pigment components to vary considerablywith incubation time and amount of aeration.He suggested that the carotene components were

developed from a nonpigmented precursor viaseveral intermediate steps.The present investigations are based on the

spectrophotometric analyses of methanol ex-

tracts of whole cells and not on the completepigment complexes that may be obtained by theuse of other solvents or disintegrated cells. It isrecognized that different media and other incu-bation conditions also can influence the type ofpigments produced. Nevertheless, the methodproved satisfactory for demonstrating that differ-ent pigments are produced by S. epidermidisstrains, and that these differ from the pigments ofthe S. aureus strains studied. The method pro-

vides an additional tool for differentiating strainsof S. epidermidis and shows promise as a means

for studying the development and spread of udderinfections caused by this organism.

ACKNOWLEDGMENTThis investigation was supported by a Public

Health Service International Postdoctoral Fellow-ship (G-V-111).

LITERATURE CITED

ELLINGHAUSEN, H. C., JR., AND M. J. PELCZAR,JR. 1955. Spectrophotometric characterizationof Neisseria pigments. J. Bacteriol. 70:448-453.

Fox, D. L., AND R. A. LEWIN. 1963. A preliminarystudy of the carotenoids of some flexibacteria.Can. J. AMicrobiol. 9:753-768.

PETERSON, W. J., T. A. BELL, J. L. ETCHELLS,AND W. W. G. SMART, JR. 1954. A procedure fordemonstrating the presence of carotenoid pig-ments in yeast. J. Bacteriol. 67:708-713.

PETRACEK, F. J., AND L. ZECHMEISTER. 1956.Determination of partition coefficients of carot-enoids as a tool in pigment analysis. Anal.Chem. 28:1484-1485.

SOBIN, B., AND G. L. STAHLY. 1942. The isolationand absorption spectrum maxima of bacterialcarotenoid pigments. J. Bacteriol. 44:265-276.

SOMPOLINSKY, D. 1962. Chromogenesis by Staphy-lococcus aureus and typing according to pig-mentation. J. Lab. Clin. Med. 60:439-450.

SUZUE, G. 1959. Studies on carotenogenesis byMicrococcus pyogenes var. aureus. J. Biochem.(Tokyo) 46:1497-1504.

WILLIS, A. T., AND G. C. TURNER, 1962. Staphylo-coccal lipolysis and pigmentation. J. Pathol.Bacteriol. 84:337-347.

1208 J. BACTERIOL.

on May 11, 2021 by guest

http://jb.asm.org/

Dow

nloaded from