PSEUDOMONAS AUREOFACIENS NOV. SPEC. AND ITS PIGMENTS

(1)

PSEUDOM1ONAS

AUREOFACIENS NOV. SPEC. AND ITS PIGMENTS

A. J. KLUYVER*

Laboratory for Mlicr-obiology, Technological Unimersity,Delft, Holland Received for publication March 12, 1956

It is well knowni that the genus Pseudomonas

comllpirises several

p)igment-producing

species,

an(l that in all cases in whiich the constitution

of these pigments lhais l)een estal)lishedl

thle-have leen found to l)e

plhenlzine

(lerivatives.

Wrede and Strack (1929) were thefirst to prove

this foI

pyocyanine,

the pigment produce(l by

Pseudomonas aeruginosa (P. pyocyanea).

Pyocv-anine is now usually consider(ed to be a

reso-nancehybridofthemesomeric formsofn-methyl

1-hydrioxyphenazine (Hillemann, 1938). K6gl

and Postowsky (1930) established that

chloio-raphine, the green

pigiment

of Pseudomonas

chlororaphis, is the semiquinoine of phenazine-a-carboxylic acidamide, while Clemo aind

Daglish

(1950) proved io(linin, the pigment of

Pseudo-nm,onas iodina, to l)e

1:6-dihydroxyphenazine-di-n-oxide.

In view of these

(latta

it seems worthwhile to

look out for further

pigment-producing

species within thegenus Pseuidomonas in order to check

if still other phenazine (lerivatives occur as

natural products.' The

opportunity presented

itself in 1936 when MrI. H.

Bouman, working

in this laboratory, incidentally came across a

bacterium which was easily identified as a

Pseudomnonas species, and whichattracted

atten-tion byanample formation ofyellow

crystals

in

its colonies. Mr. Bouman showed that a

crys-talline pigment could also be isolated from

various liquid medlia in which the bacterium

had grown.

Some years later his observations were

ex-tended, first by MrI. J. J.

Ghijsen,

and next by

Mr. C. J. P. Spruit. They obtained convincing

evridence

that the pigment as first isolated by

Bouman was

phenazine-a-carboxy

lic acid, and

therefore closely related to

chlororaphine.

At

that time warinterfered with the work, and the

resultsobtained remained unpublished.

*Deceased May1956.

l In this connection the pseudomonad strain C 50studiedbyChinn (1954), which formsgreenish

crystalsinagarmedia, also asks forattention.

Interest in the organism in question and its

products was revived a year ago when during a

visit at the Noorthern Utilization Research

Branch at Peorii, Ill., I)r. W. C.

Haynes

and

Dr. F. H.Stodola informed the authorregarding

their investigations on a new pigment-producing

pseudomonad for w-hich they also had

estab-lished production of phenazine-a-carboxylic

acid. After exchange of cultures it soon became

evident that the American and theDutchstrains

were either idenitical or at least very closely

rielated. The American woorkers and the present

authorthen (leci(led on simultaneous publication

oftheresults obtaine(l in each ofthetwo

labora-tories. In the meantime there also had been an

exchange ofexperimental data, which,of course,

led to a mutual influence in the later phase of

the investigation. The experimental work

per-formed in Delft during this second period was

carried out by the following students who

suc-cessively devoted themselves for a shorttime to

the problems involved: Tjoa Hin Soey, E. van

der Slooten, WY. Klop, Sie Hian Jang, and E. J.

Vles. The following expose should be considered

as a condensedl collective report on the

fragmen-tary observationsmadeoverthewholeperiod of

almost twentyyears; it isimpossible to

acknowl-edge here the separate individual contributions.

RESULTS

Isolation andculturalproperties of the organism.

For reasonswhich can be left out of consideration

here i\h.Boumanin1936

Imade

someexperiments

in which kerosene was treated with clay. A liter

ofheat-sterilized kerosene was poured into a 2-L Erlenmeyer flask and 100 g clay-from the river MIaas-was added. The flask was kept in an

incu-batorat25Cforthree weeks. The clay was

resus-pended in the kerosene from time to time by

shaking. At the end of the period the kerosene

was poured off, and a suspension of the clay in

sterile water was streaked on anagarplate

con-taining 1percentpeptone and 2 per cent glucose.

After twodays,incubation at 30 Cseveral colonies

406

at BIBLIOTHEEK TU DELFT on January 20, 2009

jb.asm.org

(2)

15.A'.AUUEOF lCIENS AND ITS P'IGMlCII,NTS4 had developed on the plate. Amongst these,

colo-nies with a marked brownish yellow color

pre-vailed. From one of these colonies a pure culture

was

easily

obtained by

replating

on the same

medium. The pure culture was maiintained on

plain peptone agar (1 per cent

peltone).

Gelatin

was foundto be stronglyliquefiedl.

Mlicroseopic examination of young cultures

showed small motile rods which proved to be

gram negative. Usinig Gray's staiin, the presence

of polarflagella could be demionstrate(l. No

indi-cations of endospore formation were ever

en-countere(l. The bacterium

al)peare(l

tobe strictly

aerobic insofar as it did not grow in the usual

media in the absence of oxygen. No sugar

fer-mentation everoccuriedinsuclh

medlia.

In media

containing sucrose, ainamiple pro(luction oflevan took

place

However, a certaini amounit of ainaerobic

groNvth occurried in nutirient broth to which

nitrate was added. Undei thesecondlitions nitrite

accumulated. Becausea considerable part of the

nitrate inonepercentKN=O3nutrient broth

cul-ture remains untouched, inhibition of growth is

probably due to accumulated nitrite.

Some-times a very smallquantity of gas is

produced,

buttrue denitrifyingabilityisevidently lacking.

All these characteristics taken together leave no

doubtthat theorganismshould be classified as a speciesof thegenusPseudonmonas.

In view of the present confused state of the

taxonomy within the genus Pseudomonas a

fur-ther determination of the bacterium would have

been well-nigh impossible, had not its

charac-teristic pigment production set it quite apart

from the manv colorless species, as vellas from

those species which are characterized

by

the

production of a water soluble

yellow-green

fluorescentpigment.

After two to three

days

on peptone agar

smooth round colonies of about 2 mm diameter

are formed. They showa

yellowish-orange

color

which isespecially accentuated inthecenterand

the colorgraduallyturnsbrownish. The

pigment

alsodiffuses into the agar.

It was soon found that addition of

glucose

to

the medium

markedly

increases

pigment

pro-duction.After some time the colon-es on such a

medium

show,

on observation under the

micro-scope with low

magnification,

the presence of

brown crystalline

conglomerates

either in or in

the immediate vicinity of the colonies.

Figure

1

A

Ai

:~.A

Figulre 1. Colony of Pseudoinonas aureofaciens

on

peptone

agar + 1 per cenit gluicose with

crys-talline inclusions. MIagniification125 X.

givesal)ictuleofthe crystalsastheyappearina

streak culture on l)el)tone

agiar

+ 2

peC

(ent

glucose.

Not all colonies are characterized by such a

crvstalline deposit. It has, lhowvever, beeii

ob-served that in these cases the mere touching of

the colony w-ith a platinum needle often leads to

a cry-stallizatioll. In such a case we apparently

tare dealing with a (listurbed supersaturation.

Thenextobservationi wasthat onflooding the

l)late

Withl chlorofoim

the crvstals

quickly

dis-solved. On evaporating the yellow chloroform

solution, brownish-red crvstals were at once

obtaine(l. Thesecrvstals also l)roved to be

readily

solubleinbenzene,acetonean(l dilute alkali.

Conditions

fav,oring pigmient

production. Since

it seemed worthwhile to(leterimine the nature of

the pigmentformed, numerousexperiments were

carried out in or(ler to find coniditions favoring

pigment production.

In the first place it was established that good

production

of

pigment

could also take

place

in

liquidmedia. At first amedium containing I per cent

peptone,

0.5 per cent NaCl and 2 per cent

glucose was examined. In

stationary

cultures in

Erlenmeyerflasksit wasatonceevident thatthe

colorwas practicallyconfinied to theupper layer

of the medium,

probably

only

indicating

the

organism's need of oxygen for

proliferation.

In

the later phases of the

investigation

the

culti-vation was thereforeusuallycarriedout in

Erlen-meyerflasksplacedon arotaryshaker, although

1956] 407

jb.asm.org

(3)

Kl,UYVl[,Ot

soimietiiimes itwas

preferled

to cultivate inldeeper

layers ofmedium with continuous aeration.

Further experiments left nodoubt that a

tem-l)elatule of 30 C was

optimal

for both

growth

and pigment production. Growtth is still

satis-factory at 35 C, but,remnarkably, no pigment is

produced.

Next attempts were ma(le toincrease pigment

plroduction by varying

the

composition

of the

medium. There is no needl to sum up here the

numerous media which were used in these

com-parative

experiments.

It may suffice to mention

some ofthe results obtained.

In the first place it should be remarked that

thesubstitution ofglycerol forglucose proved to

befavorable, whilst the sameheldfor anaddition

of0.1 percentKNO3tothe medium.

Furthermore it became cleai that the type of

l)eptone

used

played

an essential

role

in

pigment

piroduction.

Although

withall

brands

of

peptone

a good growth of the bacteiium resulted, there

wvas practically no pigment production with

peptone

(Witte),

nor

with

pl)etone

(Difco).

On

the contrary, the "peptone bacteriologique" and

thc "peptone pancreatique" of the "Usines

chiim-iques

de laboratoiresfrancais" were equally satisfactory, and tryptone (Difco) also gave

excellent results. On thc other hand, casamino

acids (Difco) were unsatisfactory, whilst a

vita-min-free peptone preparation again gave a good

Iresult. This proves that vitamins are needed

neither for growth nor for pigment synthesis.

Further it canbeconcluded that certain peptone

prep)arations

contain one or more constituents

which are lacking in other brands. Attempts

were

made to supplement the medium containing

l)eptone

(Witte)

by adding compounds

ofknown

constitution, such as vaIrious amino acids and

intermediates of

carbohydrate

metabolism, but

thisremained unsuccessful. Howevei, an addition

of 0.1 per cent yeast

autolysate

to the peptone

(Witte) restored pigment production. This was

not primarily due to trace elements present in

the autolysate, for the

addlition

of ash of the

autolysate wasonly slightly effective.

The general conclusion from the results

ob-tained inthese exploratory tests-extending over

some200media-wasthat thefollowing medium:

"peptone bacteriologique," 1 per cent; sodium

clhloiide,

0.5per cent; potassiumnitrate, 0.1 per

cent; glycerol, 2 per cent; pH adjusted to 7.2,

gave the most favorable and consistent pigment

production.

In later

experiments

this medium

hasalways been use(l.

Itshould beremiiarked that in thesepreliminary

tests pigment

plroduction

was always judged

visually. Such a (rude estimation seemed

suffi-ciently reliable foI the purpose, although it

should be ackinow-ledge(d that sometimes

diffi-culties arose, becaltuse theIe were marked

differ-ences in the colo( of the various culture fluids.

This color could vlary finoin an almost pure yellow

to orange, to red(lislh brown and even to dark

brown. The (lifferences weere such that they

strongly suggeste(d the production of more than

onepigment in certainmedia.

On the other hand, these differences might

have been causedIby variations infinal pHoIr in

oxidation-reduction potentials of the media. It

was clear that only the isolation of the

pig-ment-or pigments---ould give an answer to

these questions.

Isolation and

puirification

ofthe pigments. Even

in the earlier experiments it had become clear

that the separation of the pigment from a

well-colored medium was quite easy. The initial

observation that the crystals present in a plate

werereadilysoluble in chloroformledtoattempts

to extract the coloring substance from liquid media also with this solvent. It was, of course,

necessarytogive due attentionto the pHofthe

medium, and it soon appeared that removal of

the pigment from the water phase was attained

only if the culture medium were first acidified.

This is in marke(d contrast to the behavior of

pyocyanine, which is extracted from the culture

medium of P. aerutginosa aslong asthe medium

isdistinctlyalkaline.

Inthe presentcasetheaddition of some

hydro-chloric acid to the medium changes the usually

orange-red

color into yellow, whilst very soon a

yellow crystalline precipitate is formed. These

crystals dissolve in chloroform, giving a bright yellow solution. The pigment is again removed

from the chloroform by washing with dilute

alkali; in this operation the orange-red color is

restored. After re-acidification a second

extrac-tion with chloroformcanbecarriedthrough.The

chloroform solution thus obtained isfree from

lipids, since these have remained in the first

chloroform batch. The final chloroform solution

was driedovernight with anhydrous sodium

sul-fateand concentratedbydistillingoff the

chloro-408 [voi,. 72

jb.asm.org

(4)

1P. .-llAIEOF,,I(CIEA-S AND ITS PIGME'NTS

foirm in vacuo. This

a1lwa-s

led to a

cry-stalline

pigmentproduct.

The product was

usually

recrystallize(l a few

times fromethanol (95 per cent) till the crystals

showedaconstantmeltingpoint.

It soon became evident that the products

obtained by this simple procedure from various

cultures showed marke(d differences. The rather

pronounced difference in color of the culture

media was also manifest in the final products;

moreover,thisdifferencewasobviouslycorrelated

with differences in melting

point

an(l in cr-ystal shape.

Amongstthe isolatedI products

there

were

two

which apparently prevailed;

pigmeint

I, wlhich

was obtained in fine

greenish-yellow

needles

with a melting point varying between 239 and

241C (uncorr.)and whichdissolved withayellow

color in dilute

alkali,

and pigment

IT,

which

showedorange-redtobright red irregular

crystals

whichonheatingsinteredat207 C

andl

melted in the neighborhood of 225 C.

However,

as

already

mentioned, in other cases preparations were

obtained with

differeint

shades anid melting

poiInts.

The inevitable conclusion is that at least two,

and possiblyevenmore, pigments canbeformed bythe

organism

inquestion.

This view could be confirmed with the ai(l of

column

chromatography

using A1203

(alumin-ium oxydatum, E.

Merck)

as an adsorbent.

The A1203 was

suspended

in chloroform and by

sucking

off the

liquid

a solid column was

pre-pared.

A solution of a crude preparcation in

chloroform was then

cautiously poured

oIn

the

topofthe

column,

and afterashorttimeseveral

differently

colored bands

appeared.

On washing

withasmallquantityofpurechloroform atleast

six clearly

separatedI

bands could be observed. No details will begiven here; the

followiing

may

suffice. The top

layer

was of a somewhat

dirty

brown color, but then a

broad,

beautiful red

zone occurred. The three lower

layers

all had a yellow color,

although

there was a

difference

in

tinge. From the different zones the

pigments

were extracted with dilute potassium

hydroxide

(5

per

cent),

and after acidification again taken

upinchloroform. It will onlybe mentioned that

fromtheredzone,pigmentII

(daIk

red crystals,

mp 224-225 C), and from the combined

yellow

layers, pigment I

(greenish-y-ellow

needles,

mp

238-240C) could be isolated.

P'reliminary

intestigations

on the nature of the

pigmients. From the very beginning the phenazine

chariacter of the

pigments

wasdeeme(1 probable.

The correctness of this idea wvas soon proved by

carrying out a zinc (lust (listillation of both

pig-meint I and pigment II in a suitably bent

hard-glass tube. In both cases a cirvstalline sublimate

wasobtained in the cooled bent airm of the tube.

The sublimate proved to have a melting point of

171 C, whilst a mixturie of this

p)rodluct

witlh a

synthetic phenaziine

preparalItion

(mp 170.5 C)

did inot show any

(lepiressioin

in melting poiint.

This result seems to offer sufficieint evidence for

theconclusion that in bothpigments we are

deal-ing withl phenaziine derivatives.

Identification

of pigimient I. The next step was determinatioin of the

elementary

composition of

pigment I. C anid( H were (teternmined by

com-bustion, N according to the micro-Dumas

methodl.

The result(averalge of two aInalyses with quite

satisfactory agreement) was: C, 69.4; H, 3.8;

N, 12.4;0, 14.4 (bydifference).

Accepting

two N atomiis to be present in the

molecule, this riesult indicates an empirical

formula: C13H8.4N202.

For this formula the calculated composition

is:C, 69.6;H,3.6;N, 12.5;0, 14.3.

The deviation between the experimental and

thetheoretical figuresis so slight thata formula

C13H8N202 for pigmient I becomes extremely probable. This formula immediately suggests the

conf guration of a phenazine carboxylic acid,

especially if we realize that chloioraphine, the

pigment of P.chlororaphis,hasbeenprovedtobe

the amide of phenazine-a-caiboxylic acid. K6gl

and Postowsky (1930) have

synthesized

two phenazinecarboxylic aci(ls, aan(l 3, with melting

points of 239 and 292 C respectively. This at

oncestrongly suggeste(l the identity of pigment

I and

phenazine-a-carboxylic

acid, and this

became practically certain when it was found

thatamixture ofpigment I and of thesynthetic

preparation kindly

provi(le(l

to us by Professor

Kogldid notshowany

melting point

depression.

Itcanbeadded that in other respects also both

products behave identically. On methylation

both products yielded a

monomethyl

ester with

identical melting point, 119 C, and a nitrogen

contentof 11.5percent

(theoret.:

11.8 percent).

A determination of the molecular

weight

of

pig-ment I by potentiometric titration of an

alco-195631 ₄₀₉

jb.asm.org

(5)

KLUYVER

holic solution yielded the value 211 (theoret.

forphenazinecarboxylic acid: 224).

Takingall together, the identity ofpigment I and

phenazine-a-carboxylic

acid may be con-sidered to be proved.

Attempts at identification of pigment II. Since

the attempts atidentificationofpigmentII have

not yetledtoafinalresult,only a very brief

sur-vey ofprovisional results will begiven. It should

be acknowledged that subsequent preparations

yieldedproductswithslightly varying properties,

even afterrepeated

crystallizations.

Themelting

point is not sharp, usually it is near 225 C, but

even atabout 200 C markeddiscoloration occurs.

.Meltingisaccompanied

by

gasevolution. Microanalysis of one of the preparations yielded the following results: C, 63.3; H, 3.3;

N, 11.4; 0, 21.9 (bydifference).

This means that the empirical formula

ap-proaches C13H803N2 for which compound the theoretical valuesare: C, 65.0; H, 3.3; N, 11.7;

0, 20.0. Assuming this formula to be correct,

the constitution could be either that of a

hydroxyphenazine carboxylic acid oraphenazine carboxylic acid n-oxide. A determination of the

molecular weight by Rast's method did not

yield reliable results, probably owing to decom-positioninthe meltedcamphor.Apotentiometric titrationdid notgivea

sharp

endpoint, themost

probable value of the molecular weight being

259(theoret. valueforeitherofthetwosuggested

configurations:

240)

Owing to the fact that the bacterial cultures have yielded only relatively small amounts of

pigment II, and that subsequent preparations showed variations in properties, a furtherstudy

ofthe chemicalnature ofthis pigment hadtobe postponed.

It is only certain that pigment II is also a

phenazine derivative; moreover, the presence of one carboxylic group is at least most probable, whilst the compound is oxidized as compared

withphenazine carboxylic acid.

DISCUSSION

The isolation of a previously undescribed

species of the genus Pseudomonas has been

reported.

It has been shown that this species produces

atleast two pigments of the phenazine series in

suitable culture media. One of these pigments has been identified as phenazine-a-carboxylic

acid,

whilethesecond onemaybesomeoxidized

derivative ofthis acid. In many cases chromato-graphic analysis showed the presence of still

other

pigments

of apparently related type. It

should, however, be taken into consideration

that both pigments are reversibleredoxsystems

and that chromatographic analysis may bring

abouta separation betweentheoxidizedand the

reduced component ofeach of the twosystems. Moreover, in view ofthe situation encountered in the caseofchlororaphine, it seems likely that the occurrence of semiquinones must also be

taken into account (K6gl and Postowsky, 1930;

K6gl and Tonnis, 1932; Elema, 1933). Only

furtherinvestigations canelucidate thesepoints.

It is proposed to name the new species

Pseudo-monas aureofaciens on account of the

golden-yellow color it produces on some agar media, a

property which led to its

discovery.

Dr. Haynes and Dr. Stodola and their col-leagues reporttheir observations in thefollowing

article.

SUMMARY

In 1936 iAIr. H. Bouman isolated a new

pseudomonad strain which attracted attention

by the production of a golden-yellow color on

certain solidmedia, and by the formation of

pig-ment crystals within its colonies. The organism

has now been described in detail and given the

namePseudomonas aureofaciens nov.spec.

It has been shown that P.aureofaciens produces

more than one pigment in varying proportions depending on the composition of the medium used and on external conditions. Two of these pigments have been isolated and both proved

to be phenazine derivatives. One pigment was

identified as phenazine ca-carboxylic acid. The

second pigment apparently is a closely related

compound, but its configuration still awaits

elucidation.

REFERENCES

CHINN, S. H. F. 1954 An antibiotic-producing

bacterium of the genus Pseudomonas. Can.

J.Microbiol.,1, 118-124.

CLEMO, G. R. AND DAGLISH, A. F. 1950 The

constitution of the pigment of

Chromobac-terium iodinum. J. Chem. Soc., 1960

1481-1485.

ELEMA, B. 1933 Oxidation-reduction potential

of chlororaphine. Rec. Trav. Chim., 52,

569-583.

410 [VOL.72

jb.asm.org

(6)

P. AUREOFACIENS AND ITS PIGMENTS

HILLEMANN, H. 1938 tber die Stellung der

Methylgruppe im Pyocyanin und fiber

Ver-suche zur Synthese von Isopyocyanin. Ber.

deut. Chem.Ges.,71B,46-52.

KOGL, F. AND POSTOWSKY, J. J. 1930 tber das

grine Stoffwechselprodukt des Bacillus chlo-roraphis. Ann. Chem., 480, 280-297.

KOGL,

F. AND TONNIS, B. 1932 Uber

Chloro-raphin und "Xanthoraphin",ein Beitragzur

Chemie der Chinhydrone. Ann. Chem., 497,

265-289.

WREDE,

F.AND

STRACK,

E. 1929

tUber

das

Pyo-cyanin, den blauen Farbstoff des

Bacilluts

pyocyaneus. Z. physiol.Chem., 181,58-76.

1956] ₄₁₁

jb.asm.org