STUI>1ES ON THE FILAMENTOUS SHEATHED IRON BACTERIUMSPHAEROTILUS NATANS

J. L. STOKES'Department of Bacteriology, Indiana University, Bloomington, Indiana

Received for publication September 4, 1953

The extensive literature on the iron bacteriaand the present status of our knowledge of thisinteresting group of microorganisms recentlyhave been reviewed critically and exhaustivelyby Pringsheim (1949a). It is clear from this re-view that every aspect of the iron bacteria-ecology, isolation, culture, morphology, nutrition,biochemistry, and taxonomy-requires exten-sive investigation if there is to be a sound knowl-edge of these microorganisms. The present in-vestigations were begun as part of a systematicattempt to fill in some of the. existing gaps.The studies which are reported here are con-

cerned with one of the commonest of the ironbacteria, Sphaerotilu natans. This orgaimsm wasdescribed and named first by Kttsing (1833) whonoted its presence in large numbers in pollutedwaters. It was rediscovered later by Cohn (1875)under similar conditions, but because the fila-ments appeared to be dichotomously branched,although falsely so, Cohn thought that he hadfound a new organism and accordingly named itCladothrix dichotoma. Both names have appearedin the subsequent literature and were consideredby some investigators (Linde, 1913) to apply toone and the same organism and by others (Zikes,1915) to organisms that are different. But Prings-heim (1949b) has shown that S. natans exhibitsdichotomous false branching whenever it isgrown in media of low organic matter content.It seems best, therefore, to consider C. dichotomaas being the cladothrix form of S. natans. Inthis paper, therefore, we shall use the nameS. natans exclusively.Zopf (1882) claimed that S. natans is highly

pleomorphic and described the occurrence ofmicrococci, vibrios, spirilla, and spirochetesalong with the usual rod shaped cells. But thiswas based on observations of impure culturesand has not been supported by subsequent stud-ies with pure strains. Biisgen (1894) was the

1 Present address: Western Regional ResearchLaboratory, Albany, California.

first to obtain such pure cultures by streakingfilaments of S. natans, originally obtained fromflowing conUminated water, on meat extract-gelatin plates. Since then many other investiga-tors (H6flich, 1901; Linde, 1913; Zikes, 1915;Cataldi, 1939; Lackey and Wattie, 1940; Prings-heim, 1949b) also have isolated pure cultures byemploying the usual bacteriological techniques,and investigations made with these strains havesupplied a great deal of information on the cul-ture, morphology, nutrition, and general phys-iology of S. natans.

In the present investigation nine strains ofS. natans were isolated from a variety of aqueoushabitats. Their morphology, heterotrophic nutri-tion, and general physiology were studied. Theresults obtained confirm, modify, and extendthose of previous investigators. No significantdifferences were observed among the nine strainsalthough each one was isolated from a differentsource.

EXPERIMENTAL METHODS AND RESULTS

Islation of cultures. Enrichment cultures wereprepared in a medium which consisted simply ofone per cent alfalfa straw in tap water. Thestraw had been cut previously into pieces aboutone inch in length and extracted three or fourtimes by boiling with large volumes of water inorder to remove most of the soluble organic mat-ter. The presence of more than small amounts oforganic matter in the medium permits the growthof a large number of varieties of microorganisms,and this makes the subsequent isolation ofSphaerotilus more difficult (Cataldi, 1939). Theextracted hay medium was distributed in 50 mlquantities in 125 ml Erlenmeyer flasks, and eachflask was inoculated with about 10 ml of waterfrom streams, ponds, and ditches. The enrich-ment cultures were incubated at room tempera-ture and examined microscopically at intervalsfor the presence of filaments of Sphaerotilus.These appeared in considerable numbers, usually

278

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

STUDIES ON IRON BACTERIUM S. NATAN'S

after about a week, and were obtained with mostof the water samples. Tufts of filaments werefound attached to the pieces of straw and also tothe sides of the flasks at or near the surface of themedium; single filaments in small numbers weredistribute(d throughout the liquid portion of themedium.

Positive cultures were streaked on a solidmedium composed of glucose, 0.1; peptone, 0.1;MgSO4 7H20, 0.02; CaCl2, 0.005; FeCl3 6H20,0.001; agar, 1.25 per cent in tap water and in-cubated at 28 C. Within two or three days, col-onies of Sphaerotilus several millimeters in di-ameter could be seen and tentatively identifiedby theii characteristically flat, dull, white,cottony appearance. The edges of the coloniesare very irregular due to the presence of curledfilaments of growth which extend from the centerin all directions (figures 1-3). They resemblesomewhat young, unsporulated mold coloniesbut are not as dry or coarse. Identification wasconfirmed microscopically by finding chains ofrod shaped cells enclosed in sheaths. The latterfit closely about the cells and therefore are noteasily seen except when the cells have moved outof a portion of the filament and have le a sec-tion of empty sheath. Even then they are notvisible readily with the ordinary light microscopebut can be seen more easily if the slide prepara-tion is allowed to become almost completely dryor if a drop of one per cent aqueous crystal violetis added to the preparation. In the latter case thecells stain a deep purple and the sheaths lightlyso. The sheaths, however, can be seen easily inunstained preparations with the phase micro-scope under dark contrast.

Pure cultures weere obtained by restreakingfresh plates with material from isolated colonies.It is important to dry the surface of the sterileagar medium. Otherwise contaminating bac-teria, attached to the gelatinous sheaths of theSphaerotilus filaments, are carried along with thelatter in the streaking process. Or if separation isobtained initially, the growths later become con-fluent because of the connecting surface waterfilm. This difficulty was eliminated by storing thesterile plates at 37 C overnight to remove excesssurface moisture. A total of nine pure culturesof Sphaerotilus was isolated and identified asstrains of S. natans by their morphological andcultural properties (Pringsheim, 1949b). Also,the cultures were sent to Dr. Pringsheim, and hehas confirmed ouIr identifications.

Morphology. Types of colonies. When isolatedfrom natural habitats, S. natans is recognizedmost easily by its characteristically filamentouscolonial form which is shown in figures 1-3. Thisform can be considered to be equivalent to therough or R type which is found so commonlyamong bacteria. But on cultivation in the labora-tory the R type dissociates and gives rise to asmooth or S type in which the colonies are strik-ingly different in that they are smooth, glistening,often hemispherical, and usually have a veryregular edge. S type colonies are shown in figure4. Intermediate forms in which the center of thecolony is smooth but the edges are irregular andfilamentous are common. The R type of colonyis composed predominantly of filaments whereasthe S type consists mainly of single cells. TheR type is best for studying the relationship be-tween cells and sheaths. The S type is most suit-able for studying the morphology and motilityof the individual cells. Also, because it can bemade into uniform cell suspensions, the S typeis very useful for manometric experiments.

Hoflich (1901) briefly mentions the occurrenceof filamentous and smooth colonies in cultures ofS. natans. And Pringsheim (1949b) has found Rto S dissociation in the related species, Sphaeroti-lus discophorus. The factors which control thedissociation phenomenon are not known. How-ever, we have found that a low content of or-ganic matter in the medium seems to favor theformation of the R type whereas the S form ap-pears more readily on rich media. The degree ofdryness of the agar surface does not seem to in-fluence dissociation.

It may be that the S type can also be isolateddirectly from enrichment cultures, but this isdifficult to determine because the cells of thistype look so much like other relatively shortmotile rods, i.e., pseudomonads, that it is impos-sible to identify them as cells of S. natans. Also,the frequent occurrence of the S type of growthin pure cultures makes it often difficult to decidewhether one has a dissociated or a contaminatedculture. Reisolations of the strains from singlecolonies were made whenever there was anysuspicion of contamination and also as a routinepreventative measure.

Structure of sheaths and cells. The most char-acteristic structures of Sphaerotilus are thefilaments. These consist of chains of rod shapedcells with rounded ends which are enclosed inclosely- fitting envelopes oIr sheaths. This is

1954] 279

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. L. STOKES [VOL. 67

-x!~~~~~~ ~ ~ ~ ~ ~ ~

... ..

X %..

~0

"0

0

C"0

0

~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~4

V biD~~~~~~~~ 0

~~~~~~~~,0

10eX,V.i-ni'Vi!l U - S H~~~~~~~~~~~~~~

:~~~~~~~~~~~~~~~~~4Ai1

~~~~~~~~~--~~~I~~~~~~~~~

< 1S

280

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

1954] STUDIES ON IRON, BACTERIUM S. NATANS 281

x

I..

..Q

....-~

~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . ~ . ~ -

Z'~~ ~ ~ ~ ~ ~ ~ - ~ ri

.~~~~~~0...:::::~`,:...

C12~~~~~ ~ ~

5Xo

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. L. STOKES

shown in figure 5. The sheaths are thin, trans-parent, and also coherent since they can bebent at very sharp angles without tearing. Theyare soft tubes and tend to collapse when theyare not distended by the cells. The sheaths lookvery much like pieces of cellophane. Theirchemical composition is not known. Accordingto Linde (1913) the sheaths are soluble in 50per cent H2SO4 but not in 5 per cent H2S04,60 per cent KOH, nor ammoniacal copper solu-tions. They may contain hemicellulose but giveno reaction for chitin (Linde, 1913; Zikes, 1915).There is a very close fit between cells and

sheaths. This is shown clearly in figure 6 whichis a photograph of living material taken with thephase microscope. Cell preparations that aredried and stained and also wet preparationsstained with nigrosin and allowed to dry fre-quently show large spaces between cells andsheath. This condition, in a nigrosin preparation,is shown in figure 7. But it is an artifact due tothe shrinkage of the cells on drying, and it wasused erroneously by Fischer (1895) as evidencethat there is sufficient space within the filamentsfor the cells to develop their flagella while stillenclosed in the sheaths although Fischer's mainconclusion appears to be correct (see below).

Occasionally two rows of cells are found withina common sheath as shown in figure 8. Thisphenomenon was first noted by Biisgen (1894)and later also by Linde (1913) who reportedthat as many as three rows of cells may be pres-ent in a single sheath.The dimensions of unstained cells taken from

three day old agar cultures ranged from 1.2 to1.8, in width and from 2.5 to 16 ,A in lengthalthough most cells were 3 to 8 IA long. Thesedimensions correspond fairly closely to thosereported by other investigators. Since the sheathis very thin and clings to the cells, the widthof the cells can be considered to be also thewidth of the sheaths and the filaments. Collapsed,empty sheaths, however, are broader by 50 percent or more.The cells contain prominent and numerous

refractile bodies of various sizes, and some arequite large. They are present both in young andold cultures and appear as black, globular bodieswhen viewed in the phase microscope underdark contrast (figure 9). They look like fat glo-bules and stain with Sudan Black B. Similarbodies have been noted in S. natans by mostinvestigators, and apparently they misled Zopf

(1882), Eidam (1879), and Butsgen (1894) intoconcluding that they are spores although noattempts were made to establish that they areresistant to heat.

In young cultures the filaments appear to con-sist of long strands of protoplasm without anysubdivisions. As the cultures age, cross wallsappear and the filaments become chains of rodsconnected by thin bridges of protoplasm. Even-tually the rods separate completely, and gradu-ally large spaces appear between individual cellsand groups of cells as the rods begin to move outof the sheaths. Also, as the cultures age, thesheaths disintegrate and liberate cells. It is ourimpression that the latter process is the mainone for the liberation of the cells, at least underour experimental conditions. The situation maybe different in nature where one finds, in ochra-ceous deposits, enormous numbers of iron en-crusted empty sheaths. The iron coating mayprevent the sheath from disintegrating and thusforce the cells to migrate out. Only traces of ironwere present in our media. The disintegration ofthe sheaths into amorphous debris in our cultureswas so extensive and complete as to suggest thepossible participation of enzymes secreted by thecells.Form and arrangement of the flagella. The

individual cells liberated from the sheaths arefrequently actively motile. Fischer (1895) hasshown, by means of stained preparations, thatthese motile cells or swarmers possess a singlebundle of flagella which is composed usually of8 to 12 strands and that these arise from onelocus on the side of the cell, generally near therounded end. These observations have beenamply confirmed by subsequent investigators,but Hbflich and others have noted that theflagella may arise also from the ends althoughrarely at the apex. Also, the number of strandscomposing the fascicle appears to vary greatly.Pringsheim found only a single strand, but mostinvestigators report the presence of 5 to 8 flagella.This variation may be due, at least in part, tothe difficulty in obtaining a complete separationof the fascicle into individual strands since theflagella tend to stick together.Our strains of S. natans are also flagellated

lophotrichously as shown in figure 10. The flagellaare intertwined to such an extent that they ap-pear to be a single unit in most instances. Thetuft arises from one locus which may be on theside or end of the cell or occasionally on the

282 [VOL. 67

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

STUDIES ON IRON BACTERIUM S. NATANS

apex. But one cannot be certain of the point ofinsertion of the tuft because the flagella may

cling to the surface of the cell for some distancefrom the point of origin before they extend outfrom the cell. The fascicle is about 0.2 ,u wideand may be as much as 80 to 90 u long or fourto five times the length of the cell. It is wavy,

uniform in width, but tapers at the end. Thefascicle is composed of strands as can be seen on

the cell located at the bottom of figure 10; thetuft of this cell has separated into five strands.

Because some of the fascicles are wide enoughto be within the resolving power of the micro-scope, they are visible on living, motile cells inthe phase microscope under dark contrast. Thispermitted a study of the form and arrangementof the flagella without the danger of distortioninherent in making stained preparations. Itturned out, however, that in general the stainedpreparations gave a reasonably accurate pictureof the flagella. This is clear from a comparisonof the stained cells in figure 10 with the livingcells shown in figure 11. The flagella on livingcells also are twisted into a single, wavy struc-ture of uniform width but which tapers at theend. There is no sign of a separation of the tuftinto strands, and therefore the occurrence ofseparation in the stained preparations must beconsidered an artifact. The dimensions of theunstained and stained tufts are essentially thesame. The living material shows one aspect ofthe flagella which is not evident with the stainedcells, and that is the spiral shape of the fascicle.This can be seen in figure 12 but perhaps isobserved more clearly with living cells underdark field illumination as shown in figure 13.Also, the movement of the flagella as a singlehelicoidal unit can be seen very clearly with thedark field microscope. The flagella are alwaysbehind the moving cell and extend out from theapex.

Frequently we have seen motile cells, in thephase microscope, which have stopped moving,but their flagellar tufts continued to move. Thisapparently independent movement of the flagellaand their sharply defined and uniform morphol-ogy suggest strongly that the flagella of S. natansare truly organs of locomotion rather than inci-dental, trailing, strands of inert slime as claimedby Pijper for the flagella of many bacteria(cf Knaysi, 1951).The detailed structure of the flagellar tuft is

best seen under the high magnification of the

electron microscope. The drying process and thehigh vacuum imposed seem to favor the unravel-ling of the fascicle into its individual strands orflagella. The electron photomicrograph of thecells in figure 14 reveals that the fascicle may becomposed of more than 12 strands as reportedby Fischer (1895). It is difficult to make anexact count of the number of strands becausethe unravelling is not complete. But it is esti-mated that there are about 20 strands in thisparticular tuft. One cannot be certain, of course,that the visible individual strands are not com-posed of a number of even finer fibers. The cellin figure 15 has 12 strands of flagella and isparticularly interesting because the tuft arisesfrom the apex.

Since the cells can move out of the sheaths,it must be assumed that the flagella are formedwhile the cells are still encased. Indeed Fischer'sdrawings of stained preparations show flagellaon sheathed ctlls. We could not, however, seeflagella on sheathed living cells with either phaseor dark field microscopy although the flagella onliberated cells were visible. This is not surprisingbecause the tight fit between cells and sheathscould very easily obscure flagella pressed betweenthe two. But we did obtain just one stained pre-paration, shown in figure 16, which contained asegment of sheath with a single cell inside witha sharply defined flagellar tuft.

Nutrition. A survey was made of the carbon,nitrogen, and growth factor requirements of theSphaerotilus strains.

Sources of carbon and energy. The basal mediumcontained 0.05 per cent casamino acids (caseinhydrolyzate) as the source of nitrogen; the follow-ing minerals, in per cent: MgSO4 7H20, 0.02;MnSO4 H20, 0.01; CaCl2, 0.005; FeCl3 - 6H20,0.001; M/100 K2HPO4-KH2PO4 buffer, pH 7.1;and distilled water. The medium was preparedwithout the phosphate buffer, adjusted to pH 7,distributed in 9.5 ml quantities into 50 ml Erlen-meyer flasks and autoclaved. To each flask wasadded 0.1 ml of sterile M/1 phosphate bufferand 0.5 ml of an autoclaved 5 per cent neutralsolution of the carbon source to be tested. Pyru-vate and alcohols, however, were sterilized byfiltration. The concentration of the carbon sourcein the medium, therefore, was 0.25 per centalthough in some instances lower and higherconcentrations were used.Phosphate was used in the form of buffer be-

cause it had been noted that cultures tended to

2831954]

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

284

.:.'

J. L. STOKES IVOL. 67

,_ .......

::x t*. .Ql ...... .* .i,. O

.:

:::: .:

.: :::. ^ ::.

wi: ::

e;X.'!ni.

*e.

:e. S.:

iN:

..-

- s ..e -i:i

!--

:

::.:.

.: g

.e

* oni o..... ..s 0

0

cdX X;

~-0*^

0

f. ci44

o o -C) a

sv ; oC)

ci Ci C)

Cs'

C) C)Q

ce

o oomC) C)

IoQ .o>,

C)0,

4 .+

C) C .

C c- C)40*oo$ . ._ w

_ C)ZC)

Q c3c

._ =

rn" °

X U-

Q M 0 C

O- _"

0,0 C)W. .

* o )oS"0 0-~

C)* ccv.z.!X.

..f:

Aw... ..... ... ... ..

Aw'

..

.. ...

.... ..

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

1954] STUDIES ON IRON BACTERIUM S. NATANS 285

..............

x x

00 0I

§~~~~~~~~~~- ,_. C)1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:.

C)C)

ho No

_ cdc''

~~~~~~~~~~~~~~~~~~~~~~~~_-1 -2ex, -t 4- .

_ ~~~~~~~~~~~t... -4' 4-u u

C)

00..AP, bO "r "

r1 r

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. L. STOKES

become somewhat acid, pH 5, during growthunless buffered. The buffer was added separatelyto the sterile medium because the medium wouldnot support growth when autoclaved togetherwith the phosphate.The MnSO4 is not essential for growth. It was

introduced originally into the medium whenattempts were being made to isolate strains ofS. discophorus from enrichment cultures. Themanganous ion is taken up by this organism,becomes oxidized, and therefore colors the organ-isms dark brown. This is of considerable help inrecognizing S. discophorus in a mixed population.Only one strain of S. discophorus was isolatedduring the course of the present investigations,and therefore detailed studies of this species weredeferred until more strains are available. Prings-heim (1949b) found that as little as 0.005 percent MnSO4 may inhibit the growth of S. natansalthough we failed to see any inhibition of ourstrains with as much as 0.01 per cent MnSO4.The latter was eliminated eventually from themedium towards the end of the experimentalwork because it gives rise to a slight undesirableprecipitate when the medium is autoclaved.Organic nitrogen was used rather than inor-

ganic salts, which would have been preferablein experiments of this type, because it had beennoted previously that little or no growth occurredwith either ammonium or nitrate nitrogen in amineral salts-glucose medium. This led us tobelieve, initially, that S. natans cannot use in-organic nitrogen for growth. But this conclusionhad to be modified later when carbon sourcesother than glucose were used (see below). Inany event, the small amount of casamino acidsin the medium supported very little growth inthe absence of additional, available carbon com-pounds. Therefore one could determine readilywhether a particular organic compound wasutilized by observing whether it supported moregrowth of Sphaerotilus than could be obtainedwithout it.The stock cultures were kept on a medium

consisting of 0.2 per cent peptone, 0.2 per centglucose, 0.02 per cent MgSO4-7H20, 0.005 percent CaC12, 0.001 per cent FeCl3-6H20, 1.25 percent agar and tap water, and adjusted to pH 7.The phosphate buffer was not required and wasnot used because it gave rise to a marked precipi-tate in the agar medium. The stock cultures weretransferred about once a month although theyremained viable for at least two months when

stored at room temperature. Growth is best iftransfers are made on freshly slanted agar tubeswhich contain some surface moisture.The flasks of complete medium and controls

without added carbon were each inoculated withone drop of a water suspension of S. natans,prepared usually from two day old agar slantcultures. The inoculated media were incubatedin a stationary state at 28 C, and the amount ofgrowth formed was estimated by direct visualobservation at intervals usually for a week. Mostof the growth occurs as a coherent, rather slimy,surface mat or pellicle. Some of the growth clingsto the sides and bottom of the flask in the formof a delicate film or as granules. The main bodyof the medium, however, remains clear althoughsmooth type cultures, especially when shakenduring growth, may produce a fairly uniformturbidity throughout the medium. Unless largeinocula are used, growth is slight during the firstday but is maximal usually within three days.And then the turbidity of the cultures begins todecrease gradually as the cells and sheaths dis-integrate.

In the manner described above, it was foundthat S. natans can use a large variety of organiccompounds as sources of carbon and energy.These include glucose, galactose, sucrose, mal-tose, mannitol, sorbitol, succinate, fumarate,butyrate, butanol, glycerol, sodium lactate,sodium pyruvate, sodium acetate, and ethanol;ethanol and butanol were tested in a concentra-tion of 0.1 per cent. Acetate and butyrate ap-peared to be somewhat toxic since they failed tosupport growth when used in a concentration of0.25 per cent but induced good developmentwhen the concentration was lowered to 0.1 percent.

Little or no growth was obtained with 0.25per cent lactose, xylose, arabinose, or 0.1 percent sodium benzoate, sodium propionate, pro-panol, or methanol. There were no significantdifferences among the nine strains.Our data, except for lactose, are in accord with

the results of Linde (1913) who found that glu-cose, galactose, lactose, sucrose, maltose, man-nitol, and glycerol are utilized by S. natans.And from Cataldi's investigations (1939) thelist of available carbon compounds can be ex-tended to include also citrate and asparagine.Linde reported, however, that polysaccharides-dextrin, gum arabic, starch, inulin, and cellulose-will not support growth of S. natans.

286 [VOL. 67

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

STUDIES ON IRON BACTERIUM S. NATANS

A few experiments were made with differentconcentrations of glucose to determine the effectof sugar concentration upon growth. More luxuri-ant growth was obtained by increasing the glu-cose concentration from 0.1 per cent to 0.2 percent. Growth was delayed, however, by one percent glucose and completely inhibited by twoper cent glucose. To eliminate the possibilitythat inhibition at the higher sugar concentrationsmight be due to the formation of toxic substancesin the medium during autoclaving, this type ofexperiment was repeated with separately auto-claved glucose. Under these conditions, luxuriantgrowth was obtained even with two per centsugar although the total amount of growth didnot seem to be greater than that obtainable with0.2 per cent glucose. The organisms can thereforetolerate reasonably high concentrations of or-ganic matter.

Sources of nitrogen. The general methodologyused in the survey of the carbon requirementsof S. natans was employed also for the determina-tion of the availability of a variety of nitrogencompounds, mainly amino acids, as sources ofnitrogen for growth. The basal medium was modi-fied by the elimination of the casamino acidsand the introduction of 0.1 per cent glucose.The nitrogen compounds were used in a concen-tration of 0.1 per cent. The control flasks in-cluded basal medium without added nitrogenand basal medium with casamino acids.The following compounds supported good

growth of S. natans: L-glutamic acid, L-asparticacid, L-asparagine, DL-alanine, L-leucine, L-Cys-tine, DL-tryptophan, L-arginine, and L-proline.Growth with the individual amino acids wasusually not as rapid or extensive as with themixture of amino acids in the form of casaminoacids. Leucine, proline, and tyrosine were thebest among the amino acids in supporting growth.Qualitatively, the results were uniform with allstrains although some strains grew better with aparticular amino acid than did others.

Little or no growth was obtained with urea,acetamide, glycine, DL-serine, L-lysine, DL-methio-nine, DL-isoleucine, DL-threonine, DL-valine, L-his-tidine, and DL-phenylalanine.The only other detailed investigation of the

nitrogen requirements of S. natans with chem-ically defined media is that of Lackey and Wattie(1940). They reported good growth with DL-ala-nine and L-asparagine; fair growth with urea; andpoor growth with L-tyrosine, L-cystine, L-leucine,

and L-glutamic acid. There are obviously somedifferences between our results and those ofLackey and Wattie, and these undoubtedly aredue to differences in the strains used.Many attempts to obtain growth of our strains

with inorganic nitrogen in the form of ammoniumor nitrate nitrogen were almost uniformly unsuc-cessful; occasionally a small amount of growth oreven good growth was obtained with some of thestrains. This situation was somewhat surprisingsince good growth of S. natans was obtained byLinde (1913) with both ammonium and nitratenitrogen in a mineral salts-sucrose medium andby Cataldi (1939) with ammonium nitrogen in amineral salts-glucose medium and in a mineralsalts-citrate medium. Also Lackey and Wattie(1940) obtained fair growth with ammoniumand nitrate salts in a mineral salts-glucosemedium. Further investigation disclosed, how-ever, the interesting fact that the degree ofavailability of inorganic nitrogen for our strainswas dependent upon the type of organic carbonand energy source supplied in the medium. Ex-cellent growth was obtained with ammonium ornitrate nitrogen when sucrose, glycerol, or suc-cinate was the carbon source whereas very poorgrowth occurred with glucose. A biochemicalexplanation for this phenomenon is not apparent.

It can be concluded therefore that S. natanscan utilize a wide variety of nitrogen sourcesranging from inorganic nitrogen salts to individualamino acids, mixtures of amino acids, and com-plex nitrogenous substances such as peptone andmeat extract.

Growth factor requirements. As described above,our strains grew abundantly in chemically de-fined media which consisted of mineral salts,sugar, and inorganic nitrogen. A number of thestrains also were subcultured successfully, seriallythrough several transfers in a mineral salts,glucose, vitamin-free, casein hydrolyzate medium.Similarly, Linde (1913), Cataldi (1939), andLackey and Wattie (1940) have been able togrow S. natans in chemically defined media with-out the addition of growth factors. This organism,therefore, does not require an exogenous supplyof growth factors and presumably has the abilityto synthesize whatever growth factors it needs.

Pringsheim (1949b) reported that soil extractstimulates growth in dilute beef extract andother media. At the very beginning of our in-vestigations we also found that soil extract wasnecessary for good growth of S. natans in media

2871954]

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. L. STOKES

containing 0.1 per cent peptone and 0.2 per centglucose. It was also found that the soil extractcould be replaced completely by a mixture ofsmall amounts of MgSO4, CaCl2, and FeCl3; andthese three salts were incorporated, therefore,into all media subsequently used. These resultssuggest that the stimulatory effect of the soilextract is due to the minerals contained in it.The complex nitrogenous materials are reliedupon to supply essential minerals as impurities,but the quantities apparently become limitingfor growth when dilute beef extract, peptone, andsimilar media are used, and additional amountsof minerals must be supplied then in the form ofsoil extract or as pure salts for good growth.

General physiology. Effect of temperature. Themedium contained 0.1 per cent each of glucoseand peptone and also the usual mineral salts andphosphate buffer. All of the strains of S. natansgrew in the range of 15 C to 40 C; no growth oc-curred at 5 C and 46 C. On the basis of both therate and extent of development, the optimumtemperature is about 30 C. Growth is slow at15 C and especially at 40 C.These results are in good agreement with

those of other investigators. Hoflich (1901)obtained growth of S. natans between 15 C and35 C with an optimum at 25 C to 30 C. Accord-ing to Linde (1913), no growth takes place below10 C nor above 40 C, and the optimum lies be-tween 30 C and 35 C. Zikes (1915) obtained prac-tically no growth at 5 C and 39 C, and the opti-mum was between 25 C and 29 C. Cataldi (1939)reported a minimum growth temperature of15 C and an optimum of 37 C which is somewhathigher than that found by other investigatorsand also in the present studies.

Effect of pH. The peptone-glucose mediumdescribed above was used. The desired pH valueswere obtained by the addition of appropriate,sterile, phosphate buffer mixtures; and the finalconcentration of buffer was M/100. This was notvery much buffer, but it was sufficient to preventany appreciable changes in pH levels duringgrowth. No growth appeared at pH 5.5. Slowand not always full growth occurred at pH 5 8.Rapid and full development was obtained inthe range of pH 6.4 to pH 8.1. The results ofsome experiments suggested that growth can beinitiated even at pH 9 or pH 10.

Cataldi (1939) obtained growth of S. natansin the range of pH 5 to pH 9.8 and reported thatthe amount of growth increased with an increase

in pH value. Lackey and Wattie (1940) obtainedgrowth in the range of pH 5.5 to pH 8.0. It isevident that there is general agreement thatS. natans cannot tolerate a very acid environmentand grows best in a neutral or slightly alkalinemedium.

Requirement for oxygen. It is clear from theliterature that S. natans is an obligate aerobesince it will not grow without oxygen. And thisis supported fully by our investigations. But itwas not easy to establish, beyond doubt, theaerobic nature of the organism. We obtainedgrowth frequently, although not always, on agarplates incubated presumably under strictlyanaerobic conditions in desiccators with pyro-gallic acid and K2CO3 under an N2 atmosphere,in oat jars and in Brewer jars. Good growth wasalways obtained in stab cultures and throughoutthe depth of the agar medium. This occurredeven when the stab cultures were layered withparaffin or when the cotton plugs were saturatedwith pyrogallic acid and alkali and then sealedoff from the atmosphere with paraffined rubberstoppers. In contrast, growth never occurred inliquid media, containing as much as 0.5 per centglucose, in completely filled glass stopperedbottles or in stab or slant cultures supplied withalkali, pyrogallic acid, and rubber stoppers andrefrigerated overnight prior to incubation inorder to allow time for the pyrogallol to absorball of the oxygen present before the organismscould effectively compete for it.The difficulty of preventing growth of S. na-

tans by the usual anaerobic procedures suggestsstrongly that the organism can grow quite wellwith very small amounts of oxygen. And suchtraces of oxygen, which are difficult to remove,were undoubtedly present in many of our sup-posedly anaerobic systems. Moreover, the cellsare not harmed by storage for relatively longperiods under strict anaerobiosis. Inoculatedplates which showed no visible growth after aweek of incubation in Brewer jars became cov-ered with abundant growth of S. natans whenremoved from the Brewer jars and incubatedaerobically.On the other hand, the marked preference of

S. natans for oxygen is indicated by its predom-inant surface growth in flasks and tubes ofliquid media.No growth was obtained in anaerobic bottle

cultures with media containing 0.2 per centKNO,. The latter therefore cannot replace oxy-

288 [voL. 67

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

STUDIES ON IRON BACTERIUM S. NATANS

gen in the metabolism of S. natans. This was alsothe conclusion of Linde (1913). Nitrates, how-ever, are reduced readily to nitrites by all of ourstrains.

Oxidative metabolism. Since S. natans is anaerobe, experiments were made to obtain informa-tion on its oxidative metabolism. The conven-tional Warburg apparatus and techniques wereemployed. The cells were grown in a mediumconsisting of 0.2 per cent each of casamino acidsand glucose plus the usual mineral salts andphosphate buffer. Most of the experiments weremade with the smooth form of S. natans, strain12, but some of the experiments were repeatedwith strains 4 and 11. Similar results were ob-tained with all three strains.

Cells were obtained from cultures grown forapproximately 16 hours at 28 C on a shaker.The cells were centrifuged, washed once withwater, and resuspended in sufficient M/50 phos-phate buffer, pH 7.1, to give a concentration ofone mg to two mg of cells, dry weight, per ml ofsuspension. The cell suspensions were aeratedfor 3 hours to 5 hours by bubbling a slow streamof air through them in order to reduce theirrather high rate of endogenous respiration. Thishigh rate may be due to the large amount offatty material stored in the cells. Aeration formore than 5 hours tended to destroy the oxidizingcapacity of the cells.Each Warburg vessel received two ml of cell

suspension in the main compartment, 0.1 ml of anM/50 neutral solution of substrate or two LAM ofsubstrate in the side arm, and 0.2 ml of 10 percent KOH in the center well for the absorptionof CO2. The gas phase was air, and the bathtemperature was about 30 C.The compounds investigated were mainly the

carbohydrates, amino acids, and the other sub-stances which were used in the experiments onthe carbon and nitrogen requirements. Data onthe rate and extent of oxidation of some of thesecompounds are given in table 1, and the kineticsof the oxidations of additional compounds areshown in figure 17. In general, the compoundswhich were utilized as sources of carbon or nitro-gen by S. natans also were oxidized readily bynonproliferating cell suspensions. It is noteworthythat large amounts of the substrates are assimi-lated oxidatively and these may be as high as 70per cent or even 80 per cent in the case of sugarsand sugar alcohols. The assimilation process isonly slightly inhibited by sodium azide and 2,4-

TABLE 1The rate and extent of oxidation of various

compounds by Sphaerotilus natans

SUBSTRATE

Glucose ...................Galactose ...................Sucrose .....................Mannitol ....................Sorbitol .....................Benzoate, Na................Succinate, Na...............Butanol .....................Lactate, Na.................Pyruvate, Na................Acetate, Na.................L-Glutamate, Na............L-Aspartate, Na.............L-Leucine ...................DL-Alanine ..................L-Proline ....................

QO * PER CENT02 OXIDATION

53 3037 3926 1926 3117 2952 4852 4115 37

118 3598 3855 5055 3349 7529 5049 5140 50

* /AL 02 consumed per mg (dry wt) of cells perhour; the value for the endogenous respiration,Qo227, has been substracted.

dinitrophenol-compounds which are usuallyeffective in preventing assimilation and syntheticprocesses in general. When used in suitable con-centrations, these poisons stimulated somewhatthe rates of oxidation by S. natans, but this effectwas mainly on the endogenous respiration.

In an experiment on the effect of pH on theoxidation of glucose, it was found that as thepH was increased from pH 6.6 to pH 8.1, therewas a concomitant increase in the apparent rateof glucose oxidation. But this increase could beaccounted for entirely as due to the stimulationof the endogenous respiration because when thehigher endogenous rates were subtracted, therates for glucose were identical at all pH levels.The oxygen consumption curves for galactose,

mannitol, sorbitol, and sucrose showed an initiallag which lasted about an hour. But this lag wasnot apparent when the cells used were grown withthe specific substrate rather than with glucose.These substrates, therefore, are oxidized byadaptively formed enzyme systems. Also benzoateappears to be oxidized by adaptive enzymes.There is a delay of about 40 minutes before thereis any appreciable increase in oxygen consump-tion with benzoate above the level of the endog-enous oxygen consumption. Although benzoateis oxidized readily, it could not be used as a source

2891954]

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. L. STOKES

Figure 17. Oxidation of two pM each of Na bu-tyrate, Na fumarate, and Na DL-malate by Sphae-rotilus natans.

of carbon and energy by S. natans. This lack ofcorrelation between growth and oxidation is notsurprising since growth is a much more complexprocess than mere oxidation. It is well known thatformate can be oxidized rapidly by many bac-teria although few organisms can use it as thesole source of carbon and energy.

Cell suspensions in bicarbonate buffer, underan N2-5 per cent CO2 atmosphere, did not formany acid from glucose. This indicates the ab-sence of a fermentation mechanism for carbo-hydrate decomposition and is in agreement withthe finding that the organism cannot grow with-out oxygen.

DISCUSSION

The distinctive feature of S. natans is the oc-.

currence of the cells in an organic tube or sheath.The chemical composition of the sheath is notknown, and because it is difficult to see on livingcells, little is known about how it is formed. Butthe sheath is normally a closed tube which closelyencases the chains of rod shaped cells, and itmust be permeable, therefore, to the nutrientsrequired by the cells since the nutrients mustpass through the sheath. In media containing

iron, the sheaths readily take up the iron, andalthough they may continue to be colorless, thepresence of the iron can be demonstrated easilyby means of the Prussian blue reaction. In thistest the iron containing sheaths, on contact withdilute solutions of potassium ferrocyanide andHCl, turn blue. And according to Pringsheim(1949b), after long periods of incubation, suffi-cient iron may be laid down on the sheaths tocolor them brown and to give them the glassyappearance which is characteristic of and iden-tical with that of the Leptothrix ochracea sheathswhich are so commonly and abundantly foundin natural waters where iron is being deposited.

There is little doubt that S. natans is truly aniron bacterium in so far as it accumulates ironfrom solution in a characteristic and morpho-logically distinct form. But whether it, or indeedwhether any of the other classical filamentousand nonfilamentous iron bacteria can grow at theexpense of the energy liberated in the oxidationof ferrous to ferric iron, remains to be rigorouslyestablished. The long existing, confusing, andoften acrimonious controversies on the possibleautotrophy of the iron bacteria have been fullydiscussed by Molisch (1910), Cholodny (1926),Winogradsky (1949), Pringsheim (1949a), andothers. There is much strongly suggestive evi-dence in favor of autotrophy, but incontrovertibleproof is still lacking. There can be little doubt,however, that the unusual bacterium, Thiobacil-lus ferrooxidans, recently isolated by Templeand Colmer (1951) from the acid drainage watersof bituminous coal mines, can grow autotrophi-cally with ferrous iron. It develops well at pH 2or pH 3 in inorganic media with ferrous iron andC02, and growth occurs concomitantly with theoxidation of the iron and the fixation of CO2. Atthese low pH levels oxidation of the ferrous ironby the oxygen of the atmosphere cannot takeplace, and therefore the oxidation which occursin cultures must be due to the metabolic activi-ties of the bacteria. Because S. natans and othertypical iron bacteria develop only in an approxi-mately neutral environment in which directchemical oxidation of ferrous iron is rapid andextensive, it becomes difficult to determine whatrole, if any, the bacteria play in the oxidationprocess. This is a technical problem which doesnot appear to be insurmountable although it hasnot yet been solved in a fully satisfactory manner.

290 [VOL. 67

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

STUDIES ON IRON BACTERIUM S. NATANS

ACKNOWLEDGMENTS

I am glad to have this opportunity to expressmy thanks to Dr. Ruth Foster for her invaluableassistance in the experimental work, to Dr. E. G.Pringsheim for his help in the identification of thecultures, and to the National Science Foundationfor a grant which helped pay the costs of theseinvestigations.

SUMMARY

Nine pure cultures of the filamentous sheathediron bacterium, Sphaerotilus natans, were iso-lated from stream, pond, and ditch water. Thecultures gave rise to two types of colonies: arough, somewhat cottony type composed mainlyof filaments and smooth, glistening colonies com-posed principally of single cells. The filamentsconsist of chains of rod shaped cells in closelyfitting tubes or sheaths. The individual cells arefrom 2.5 to 16 Iu long and 1.2 to 1.8 ,t wide. Theyare motile, and each cell has a tuft of spirallycoiled flagella which is located usually subpolarlyor on the long side of the cell. All of these morpho-logical features are presented in a series of photo-graphs of stained and living cells taken withthe ordinary light, phase, dark field, and electronmicroscopes.

S. natans can grow with a wide variety oforganic compounds, including sugars, sugar alco-hols, and 4, 3, and 2-carbon compounds as sourcesof carbon and energy. Ammonium salts, nitrates,individual amino acids, and complexes such aspeptone and meat extract can supply the nitrogenneeds of the organism. An exogenous supply ofgrowth factors is not required.

All strains developed in the range of 15 C to40 C, and the optimum temperature is about 30 C.Growth occurred in the range of pH 5.8 to pH8.1, and the best development was in neutral orslightly alkaline media. S. natans cannot growwithout oxygen although growth will occur whenonly small amouints of oxygen are present; ni-trates cannot replace oxygen. Nonproliferatingcell suspensions oxidize a variety of carbo-hydrates, amino acids, and related compounds;and for some of these compounds the oxidizingenzymes are adaptively formed. As much as 70per cent to 80 per cent of carbohydrates may beoxidatively assimilated, and this process is veryresistant to inhibition by sodium azide and 2,4-dinitrophenol. Glucose is not fermented by cellsuspensions under anaerobic conditions.

REFERENCES

BtSGEN, M. 1894 Kulturversuche mit Clado-thrix dichotoma. Ber. deut. botan. Ges., 12,147-152.

CATALDI, M. S. 1939 Estudio fisi6logico ysistematico de algunas Chlamydobacteriales.Thesis, Buenos Aires.

CHOLODNY, N. 1926 Die Eisenbakterien, Beit-rage zu einer Monographie. Pflanzenfor-schung, 4, 69-133.

COHN, F. 1875 Untersuchungen uber BakterienII. Beitr. Biol. Pflanz., 1, 141-207.

EIDAM, E. 1879 tber die Entwickelung desSphaerotilus natans Ktz., sowie fiber dessenVerhaltnis zu Crenothrix und zu den Bac-terien. Botan. Zeitung, 37, 724-728.

FISCHER, A. 1895 Untersuchungen uber Bak-terien. Jahrb. wiss. Botan., 27, 1-163.

HOFLICH, K. 1901 Kultur und Entwickelungs-geschichte der Cladothrix dichotoma (Cohn).Oesterr. Monatschr. Tierheilk., 25, 4-23, 49-64.

KNAYSI, G. 1951 Elements of bacterial cytology.Chapter IX. Second Edition. ComstockPublishing Co., Ithaca, N. Y.

KUTZING, F. T. 1833 Sphaerotilus natans, eineneue Susswasseralge. Linnaea, 8, 385.

LACKEY, J. B., AND WATTIE, E. 1940 Thebiology of Sphaerotilus natans Kutzing in re-lation to bulking of activated sludge. PublicHealth Repts. (U. S.), 55, 975-987.

LINDE, P. 1913 Zur Kenntnis von Cladothrixdichotoma Cohn. Centr. Bakteriol. Para-sitenk., II Abt., 39, 369-394.

MOLISCH, H. 1910 Die Eisenbakterien. GustavFischer, Jena.

PRINGSHEIM, E. G. 1949a Iron bacteria. Biol.Revs., 24, 200-245.

PRINGSHEIM, E. G. 1949b The filamentousbacteria Sphaerotilus, Leptothrix, Cladothrix,and their relation to iron and manganese.Trans. Roy. Soc. (London), Series B, 233,453-482.

TEMPLE, K. L., AND COLMER, A. R. 1951 Theautotrophic oxidation of iron by a new bac-terium: Thiobacillus ferrooxidans. J. Bact.,62, 605-611.

WINOGRADSKY, S. 1949 Microbiologie du Sol.Masson et Cie, Paris.

ZIKES, H. 1915 Vergleichende Untersuchungenuber Sphaerotilus natans (Kutzing) und Clado-thrix dichotoma (Cohn) auf Grund von Rein-kulturen. Zentr. Bakteriol. Parasitenk., IIAbt., 43, 529-552.

ZOPF, W. 1882 Zur Morphologie de Spaltpflanzen(Spaltpilze) und Spaltalgen. Leipzig.

19541 291

on March 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from