COBALTAND BACTERIAL GROWTH, WITH SPECIAL REFERENCE TOPROTEUS VULGARIS'ARTHUR L. SCHADE2

Overly Biochemical Research Foundation, Inc., New York 1, New York

Received for publication September 20, 1949

For the past five years we have been engaged in an investigation of the effectof cobalt on the growth and metabolism of bacteria, especially of Proteus vul-garis. The emphasis on Proteus followed requests for aid in the control of growthof this penicillin- and sulfa-resistant organism in cases of peritonitis, cystitis,and eye infections.

Interest in cobalt as a growth inhibitor of microorganisms has been extremelylimited. Such studies as have been made deal primarily with the concentrationsof the metal necessary to kill the cells of bacteria, yeast, or paramecia after agiven period of exposure (Bokorny, 1905, 1913; Krauss and Collier, 1931; John-son, Carver, and Harryman, 1942). Apart from observations on the therapeuticuse of cobalt in the treatment of tuberculosis (Renon, 1915; Rondoni, 1920;Mascherpa, 1929), no attempt, so far as we are aware, has been made to studyin detail the nature of the action of cobalt on bacteria.

In this paper we shall consider only those aspects of the growth-inhibitoryeffect of cobalt on bacteria, particularly P. vulgaris, which are of a culturalrather than of a metabolic nature. The results of our metabolic studies willappear subsequently.

EXPERIMENTAL RESULTS

Bacterial spectrum. In the course of an investigation of the elements with whichconalbumin might combine to account for its growth-inhibitory action on aculture of Shigella dysenteriae (Schade and Caroline, 1944), we observed thatthe test organism failed to grow in nutrient broth to which 10 to 20 ppm cobalthad been added. This observation, as well as the lack of information regardingthe growth-inhibiting properties of cobalt, prompted us to make a preliminarysurvey of the response of a limited number of bacteria to cobalt additions totheir culture media.Table 1 lists the microorganisms tested, their source, and the concentration

of cobalt required to inhibit growth completely. Among the bacteriaare represent-ative gram-positive and gram-negative, as well as aerobic and anaerobic,species. The media in which the cultures are grown are likewise detailed, sincelater work has shown that the composition of the medium is important for theconsideration of the minimum amount of cobalt needed for effective growth in-hibition.

1 This investigation was supported in part by a research grant from the Division ofResearch Grants and Fellowships of the National Institute of Health, U. S. Public HealthService.

2 With the technical assistance of Leona Caroline and Marjorie Neyland.811

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

ARTHIUR L. SCHADE

That different strains of a given species may vary with respect to the concen-tration of cobalt required to effect complete inhibition of growth is readily

TABLE 1List of microorganisms tested for growth inhibition by cobalt

ORGANSM SOURCZ COBALT (PP)*

Proteus vulgaris (2 strains)Proteus vulgaris (3 strains)Escherichia coliEscherichia coli (2 strains)Salmonella paratyphi ASalmonella typhimuriumSalmonella cholerae-sui8Salmonella enteriditisSalmonella pullorumSalmonella newingtonShigella dysenteriae (3 strains)Shigella dysenteriae (3 strains)Shigella paradysenteriae (2 strains)Shigella paradysenteriae (1 strain)Shigella paradysenteriae (1 strain)Shigella sonneiShigella dysenteriae sp. (Newcastle)Alcaligenes faecalis (7 strains)Staphylococcus aureus (4 strains)Staphylococcus aureus (2 strains)Staphylococcus albusStaphylococcus albusStreptococcus faecalisStreptococcus viridansClostridium septicumClostridium chauveiClostridium pasteurianumSaccaromyces cerevisiaeLeptomitus lacteus (Roth) Agardh

CystitisSinusitisUlcerative colitisSinusitisStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureA.T.C.C.Stock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureStock cultureA.T.C. 7040Fleischman strain 139

20-3030-5010-2070-10020-3030-5010-2030-5010-2030-510-2020-301-10

10-2050-10050-10010-201-101

30-5010-2010-2075-10040-6040-60'10-20310-20'20-304100-125'5-10'

* Concentration of cobalt required to effect complete inhibition of growth in nutrientbroth (0.5 per cent meat extract and 1.0 per cent peptone in 0.5 per cent saline, pH 7.3)except when otherwise noted.

1 Five-tenth per cent sodium lactate, 0.1 per cent ammonium sulfate, 0.01 per cent mag-nesium sulfate heptahydrate, in 0.025 m phosphate buffer.

2 Nutrient broth plus blood serum.J Nutrient broth plus 0.1 per cent sodium thioglycolate and 0.1 per cent brain heart in-

fusion.4 One per cent malt extract plus 0.1 per cent yeast extract.6 Nutrient broth plus 0.05 per cent yeast extract and 1 per cent glucose.6 One-tenth per cent L-leucine, 0.05 per cent DL-alanine, 0.05 per cent glutamic acid,

0.0005 M MgSO4, and 0.005 M phosphate buffer, pH 6.6.

apparent from a study of the seven A.T.C. strains of Alcaligenes faecalis inves-tigated. Three strains (4741, 8749, and 8750) were inhibited by 1 ppm cobalt;

812 [VOL. 58

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

COBALT AND BACTERIAL GROWTH

one strain (213) by 2 ppm cobalt; two strains (8748 and 9220) by 4 ppm cobalt;and one strain (212) by 7 ppm cobalt. Further, we have been able to train anisolate of a strain of P. vulgaris whose growth is halted by 50 ppm cobalt to growwell at 150 ppm and to require at least 200 ppm cobalt for effective inhibition.Of the various bacteria listed in table 1, we chose to concentrate our attention

upon the species P. vulgaris as a test organism for an extensive study of theaction of cobalt on growth and metabolism. The strain of Proteus we employedhad been isolated from a cystitis case.3 When we found that this organism failedto grow in nutrient broth containing 30 ppm cobalt, it was of added interest tofind that its growth in urine was likewise inhibited by cobalt.

Figure 1. Growth rate of P. vulgaris in nutrient broth as affected by the addition of in-creasing concentrations of cobalt to cultures in early logarithmic phase of growth.

Growth-inhibiting concentrations of cobalt and composition of the growth medium.When increasing concentrations of cobalt were added to young cultures of P.vulgaris growing in nutrient broth, the subsequent growth rates of the cultureswere increasingly affected. Figure 1 summarizes typical results obtained withsuch cultures to which 0, 12.5, 25, 50, and 100 ppm cobalt had been added. Inaddition to the fact that greater concentrations of cobalt effect greater inhibitionsof the rate of growth, it is apparent that 50 ppm cobalt, though adequate finallyto inhibit growth completely, are not immediately effective as is a concentrationof 100 ppm.

a Kindly supplied by Dr. G. Schwartzman, Mt. Sinai Hospital, New York.

1949] 813

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

ARTHUR L. SCHADE

In a preliminary attempt to determine whether cobalt was exerting its effecton the growth of Proteus through interference with the utilization by the organ-ism of some nutrient in the medium essential to it as a growth factor, nico-tinamide (Peterson and Peterson, 1945) and magnesium sulfate (Knight, 1936)were added to the nutrient broth. Such additions had no effect on the concentra-tion of cobalt required to bring about cessation of growth. Additions of glucoseand of ammonium sulfate to the medium were also ineffective.That the concentration of cobalt needed to stop the growth of P. vulgaris in

nutrient broth was, however, related, in a large measure, to the combination ofcobalt with some constituent or constituents in the medium became evident whena determination was made of the cobalt needed for growth inhibition of thisbacterium cultured in a synthetic medium. Our strain was adapted to grow in thefollowing synthetic medium: 0.05 M glucose, 0.0075 M ammonium sulfate, 0.0004M magnesium sulfate heptahydrate, and 0.025 M potassium phosphate buffer atpH 7.3. A concentration of cobalt of about 0.25 ppm prevented growth in thismedium as judged by turbidity in comparison with about 100 times this concen-tration in nutrient broth and 1,000 to 2,000 times in brain heart medium (Difco).When brain heart infusion medium was dialyzed to remove low molecular

weight constituents and then added to nutrient broth, cobalt, at the same concen-tration as was required in the broth alone, inhibited growth of P. vulgaris. Ad-dition of the ash of the brain heart infusion medium was likewise without effecton the concentration of cobalt required to stop growth of the bacterium innutrient broth. These observations indicated that the amino acid component ofthe brain heart medium and, probably, of the nutrient broth itself might beresponsible through some cobalt-amino-acid complex formation for the need ofhigher concentrations of cobalt to effect inhibition of growth. To test this hy-pothesis we investigated the effect of the addition of casein hydrolyzate (Smacovitamin-free casein hydrolyzate) to nutrient broth at a concentration level of30 mg of hydrolyzate per ml of medium on the inhibition by 50 ppm cobalt ofthe growth of P. vulgaris at pH 7.2. Under these conditions, cobalt did not inter-fere with growth of the bacteria.With the effective concentration of casein as a guide, individual amino acids

in concentrations equivalent to those found in casein were next added to nu-trient broth to observe which amino acid, if any, might be responsible for the"protection" afforded to the bacteria by casein against the inhibitive action ofcobalt. The following amino acids and amines were tested against 50 ppm cobaltin nutrient broth at pH 7.0 to 7.5 inoculated with P. vulgaris and incubated for24 hours at 37 C: DL-alanine, L-arginine, L-aspartic acid, L-cysteine, L-cystine,L-glutamic acid, glutamine, glycine, L-histidine, histamine, L-leucine, L-lysine,DL-methionine, DL-phenylalanine, L-proline, DL-serine, L-tryptophan, L-tyrosine,and DL-valine. In every case the molar ratio of amino acid to cobalt was no lessthan 4 to 1. Of these 17 amino acids, only L-cysteine and L-histidine preventedor reversed the growth inhibition by cobalt. An indication that the pH of themedium was an important factor in the tests as run became evident when his-tamine and L-glutamic acid were added to media at pH 8.0 and above. Under

814 [VOL. 58

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

COBALT AND BACTERIAL GROWTH

such alkaline conditions these compounds were likewise effective against theinhibitory action of cobalt.

Since cobalt has been shown to form a stable complex with cysteine (Michaelisand Barron, 1929), its effectiveness as a "protective" agent might well have beenanticipated. The results obtained with L-histidine, on the other hand, were notexpected. Since the casein hydrolyzate employed in these studies did not containcysteine, it is probable that the "protective" action of the hydrolyzate againstcobalt was greatly, if not completely, due to its histidine content;

Reversal of cobalt inhibition by histidine. The studies alreadly published on thecharacteristics of the cobalt-histidine complex as well as some of the physiologicalinvestigations that have made use of the cobalt-combining properties of histidine

TABLE 2Titration of L-histidine HClfor ability to overcome inhibition of 5apg cobalt per ml in a

synthetic medium

cNc.MOLECULAR RATIO MIST!-

CONC. COBALT CONC. ISTIDCINE DINZ:COBALT

Ogagmi pgs/mi0 0 4+:4+5 0 _5 5 0.28 0:05 10 0.57 0:05 15 0.84 0:05 20 1.12 0:05 25 1.40 0:05 30 1.68 0:05 35 1.96 4+:4+5 40 2.25 4+:4+

Four ml of synthetic media (0.05 M glucose, 0.0075 M ammonium sulfate, 0.004 M mag-nesium sulfate, and 0.025 M potassium phosphate buffer, pH 7.3) containing 6.25 pg cobaltper ml, as cobalt sulfate, and varying amounts of L-histidine hydrochloride monohydratewere inoculated in duplicate with 0.1-ml suspensions of P. vulgaris grown in syntheticmedium for 48 hours at 37 C and brought to a final volume of 5 ml with distilled water.Duplicate controls were run in synthetic medium minus cobalt. Readings were taken after96 hours of culture at 37 C.

(Burk et al., 1946; Hearon et al., 1947) had their genesis in the currently reportedbacteriological work and have summarized some of its findings. It is useful,however, to detail an example of the type of experiment which showed the"protective" effect of histidine against the growth-inhibitory action of cobalton P. vulgaris and indicated quite clearly the combining ratio of histidine tocobalt as 2:1 under the particular conditions given. Table 2 illustrates such anexperiment. Although in this case a synthetic medium was chosen for the test,similar results were obtained in straight nutrient broth or in nutrient broth di-luted 1 to 5. Changing the sequence of addition of cobalt and histidine had noeffect on the results.

Evidence that histidine is capable of reversing an established growth inhibi-tion by cobalt is presented in figure 2. Bacteria from an 18-hour culture of P.

1949] 815

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

ARTHUR L. SCHADE

vulgaris were inoculated into nutrient broth and dispensed in appropriateamounts into Warburg respirometer vessels. Using the rate of respiration as anindex of the course of growth, we added cobalt (final concentration = 60 ppm) toall of the cultures other than the histidine-containing controls when the bacteriawere leaving their lag phase and entering logarithmic growth. At 1-hour intervals,an amount of histidine sufficient to counteract the concentration of cobalt em-ployed was added to the inhibited cultures. One set of cultures was maintainedas a cobalt-inhibited control.

Figure 2 shows that, in each of the four sets of cultures to which histidinehad been added, growth of P. vulgaris subsequently ensued. The cultures thatwere inhibited for 1 hour by cobalt required approximately 80 minutes following

0 40 80 120 160 200 240 280 320 360 400 440 480 520 560 600

Figure 2. Reversal by histidine of the bacteriostatic effect of 1 X 10-' m cobalt on P.vulgaris in nutrient broth. The histidine was added one (hi), two (h2), three (hs), and four(h4) hours after inhibition of growth by the cobalt had been effected.

histidine addition to re-emerge from the lag phase. Those that were inhibitedfor 4 hours required approximately 140 minutes to achieve the same stage ofdevelopment.

Reactions of cobalt and histidine. The addition of cobalt to a solution of his-tidine under aerobic conditions resulted in the formation of a progressivelydeepening yellow-brown color that could be dissipated by acidification and re-formed through subsequent neutralization. When the colored mixture waskept overnight at room temperature, its yellow-brown hue had changed toreddish. Under anaerobic conditions the yellow-brown or reddish color failed todevelop.in the mixture of cobalt and histidine, which remained very light pinkuntil oxygen was admitted to the reaction flask. Similar color changes occurred

[VOL. 58816

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

COBALT AND BACTERIAL GROWTH

upon the addition of cobalt to nutrient broth in the absence of added histidine.It was observed, further, that when cobalt was added to either nutrient brothor a solution of histidine, a fall in the pH of the mixture took place.The chemical studies that have been made of the cobalt-histidine complex

formation (Hearon, 1948; Hearon, Burk, and Schade, 1949) have elucidated thesignificance of these observations. In summary, two molecules of histidine com-bine immediately and reversibly with one atom of cobalt to yield one molecule ofcobaltodihistidine and two hydrogen ions. The formation of the cobaltodihis-tidine complex is responsible for the pink color of the cobalt-histidine mixtureunder anaerobic conditions and for its acid pH shift. In the presence of oxygen,two molecules of cobaltodihistidine combine rapidly and reversibly with onemolecule of oxygen to form one molecule of oxy-bis(cobaltodihistidine), whichcomplex imparts the yellow-brown color to the solution. Slowly, with or withoutthe intervention of additional oxygen, the oxy-bis(cobaltodihistidine) changes toan irreversibly oxygenated form with no evident oxidation of cobaltous to cobal-tic. This irreversible form is reddish pink in color.

Cobalt inhibition under aerobic and anaerobic conditions. Although it is clearfrom table 1 that aerobic and anaerobic bacterial species are sensitive to cobaltconcentrations of the same order of magnitude, it is not evident that a givenspecies grown under both aerobic and anaerobic conditions would respond tocobalt in a comparable manner. To investigate this point, studies were made ofthe sensitivity to cobalt of P. vulgaris and Staphylococcus aureus grown underaerobic and anaerobic conditions in nutrient broth. With both bacterial species,irrespective of the presence or absence of oxygen, the rates of growth are initiallyaffected by the same cobalt concentrations. To achieve complete inhibition,however, roughly two to three times as much cobalt is required under anaerobicas under aerobic conditions. Although it is possible that different metabolicsystems are affected by cobalt depending upon the presence or absence of oxygen,we believe that cobalt is exerting its inhibitive effect on a common metabolicsystem in the cells of these two bacterial species independently of oxygen.

Effect of pH ofmedium and temperature of incubation on cobalt inhibition. Todetermine the effect of pH of the medium on the growth-inhibitory action ofcobalt on P. vulgaris, we set up cultures of this organism in nutrient broth diluted1:5 with distilled water, and adjusted with HCl to pH's 7.3, 6.1, and 5.6. Toeach medium enough cobalt was added to give a final concentration of 10 ppm,which concentration had previously been shown to be just adequate to effectcomplete inhibition over a pH range of 7.0 to 7.5. Inocula were then added tothe pH-adjusted, cobalt-containing media and incubated at 37 C over a periodof 96 hours. The results of two experiments in which the tests were run in du-plicate failed to show any effect of the pH levels employed on the inhibitoryaction of cobalt.The effect of the temperature of incubation of P. vulgaris cultures on cobalt

inhibition of growth was investigated by setting up series of test tubes con-taining synthetic medium and concentrations of cobalt ranging from 0.2 to 5ppm. After inoculation, duplicate series of tubes were put at 37 C and 20 C to

1949] 817

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

ARTHUR L. SCHADE

incubate. After making due allowance for the greater length of time required forthe bacterial cells at 20 C to grow to the same extent as those incubated at 37 C,we found that the same concentration of cobalt was required to effect completeinhibition of growth regardless of the temperature of incubation.

Effect of cell number on cobalt inhibition. In the investigation of the severalfactors that affect the minimum concentration of cobalt required to bring aboutcomplete inhibition of the growth of P. vulgaris, the inocula employed werestandardized roughly with respect to the number of bacteria per ml of the testgrowth medium. Qualitative evidence had, indeed, suggested that the values ofthe limiting concentrations of cobalt varied with the initial number of cellspresent in the test medium. To obtain quantitative data on this point, we set upexperiments in which the initial number of cells varied from 2 X 104 to 2 X 107perml and from 2 X 102 to 2 X 104 per ml of synthetic medium at cobalt levels of1.2 ppm and 0.12 ppm, respectively. The results showed that the effectiveness ofany given cobalt concentration is a function of the number of cells initiallypresent in the inoculum.

Effect of cobalt on size and stainability of cells. We have observed microscopicallythe effect of cobalt upon the size and stainability of resting cells of P. vulgarisfollowing their inoculation into nutrient broth. When cells, grown for 24 hourson a nutrient agar slant at 37 C, were inoculated into nutrient broth and ex-amined over a 2-hour period of incubation at 37 C, they showed an increase insize approximately twice their initial volume in the first hour, usually withoutevident cell division. During the second hour, division of the cells occurred alongwith maintenance of their enlarged size. Under otherwise comparable conditions,cells inoculated into nutrient broth containing a concentration of cobalt sufficientto inhibit growth failed to show any increase in size or evidence of cell divisionover the 2-hour period of incubation. Cells from both the control and cobalt-treated series, on the other hand, when stained with crystal violet, showedsimilar increases in stainability by the end of the 2-hour incubation period.These results are in conformity with the finding (Levy, Skutch, and Schade,1949) that the ribonucleic acid concentration of the control and cobalt-in-hibited cells increases to approximately the same extent over the same period oftime.

Effect of cobalt on viability of cells. Since, as we have shown, histidine is capableof reversing the inhibition of growth of a culture of P. vulgaris by cobalt, thequestion arises to what extent, if any, does cobalt result in the death of theindividual members of the population of the bacterial inoculum. Further, isthe cell viability dependent upon the stage of the growth cycle from which theinoculum is prepared?

Studies were made of the effect of cobalt on the number of viable cells re-maining in nutrient broth to which a concentration of cobalt had been addedsufficient to bring about complete inhibition of growth of the culture. Colonyplate counts were made in quadruplicate at two dilutions of the culture at zerotime, 30, 120, and 240 minutes following cobalt addition to the broth maintainedat 37 C throughout the experiment. In one case, cobalt was added to broth simul-

818 [VOL. 58

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

COBALT AND BACTERIAL GROWTH

taneously with the addition of the inoculum of resting cells prepared from an 18-hour nutrient agar slant culture. In the second case, cobalt was added following30 minutes' incubation of the broth culture at 37 C, at which time the cells werestill in the lag phase. In the third case, cobalt was added following 120 minutes'incubation of the broth culture at 37 C, by which time the cells were in theirlogarithmic growth phase. The results of these experiments are summarized infigure 3. It is clear that the cells of P. vulgaris varied in the degree of their sensi-tivity to a growth-inhibiting concentration of cobalt depending upon the stage oftheir life cycle at the time of cobalt addition to the growth medium. The via-bility of cells in the resting stage was relatively little affected by cobalt after a4-hour exposure to the inhibitor compared to that of cells in their logarithmicgrowth phase. Cells in their lag phase were of intermediate sensitivity. Thecorrelation of viability sensitivity to cobalt with the stage of the growth cycle

n CELLS FROM RESTING PHASE120 - LAG PHASE

U | "LOGARITHMIC PHASE

80-

>60-

402

20-

O' 30' 120' 240'TIME OF EXPOSURE TO COBALT

Figure S. The relation of the viability of cells of P. vulgaris in their several growth phasesto the length of time of exposure to cobalt in nutrient broth.

is interestingly paralleled by the respiration sensitivity to cobalt of P. vulgaris(Schade and Levy, 1949). The respiration of resting cells added to nutrient brothcontaining cobalt was comparable to that of resting cells added to nutrient brothminus cobalt. The respiration of cells growing logarithmically in nutrient broth,on the other hand, was greatly reduced on the addition of cobalt. Figure 3 alsoshows that, when cobalt was added to a logarithmically growing culture, not allof the cells at least were immediately inhibited from dividing since the number ofviable cells observed one-half hour after cobalt addition to the cultures wasgreater than at zero time.

It was possible that the decrease in the number of viable cells in nutrientmedia containing cobalt might be a reflection of a normal rate of death of cellsunable, for one reason or another, to grow. To check this possibility, restingcells from an 18-hour culture grown on nutrient agar at 37 C were inoculated into

1949] 819

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

ARTHUR L. SCHADE

phosphate buffer and into nutrient broth plus cobalt. Colony plate counts ofviable cells from each medium were made over a period of 50 hours' incubationat 37 C. The results are given in table 3 (a). Comparison of the percentagedrops in viable cell count shows that cobalt significantly effected a reductionin viability over and above that observed in a medium deficient in nutrients.

Further evidence that cobalt affects viability in a manner other than just topermit continued maintenance of the cell while inhibiting its growth was ob-tained by suspending resting cells of P. vulgaris in phosphate buffer with andwithout added cobalt. Table 3 (b) summarizes the data. It is again apparent

TABLE 3Cobalt and the viability of P. vulgaris

(a)

INCUTION (HouRs)

0 6 26 so

Per cent viability in(a) Broth + 60 ppm cobalt............................. 100 15 0.4 0.05(b) 0.05 M phosphate buffer............................. 100 95 28 22

(b)

INClulON (HOtURIS)

0 5 26

Per cent viability in(a) 0.001 M phosphate buffer............................. 100 86.4 19.1(b) 0.001 M phosphate buffer + 10 ppm cobalt.100 43.2 0.001

Suspensions of a culture of P. vulgaris, grown on nutrient agar slant for 18 hours at 37C, were inoculated into the given media (pH 7.3) to give initial cell concentrations of 2.3X 106 and 2.2 X 106 for experiments summarized in (a) and (b), respectively. The inocu-lated media were incubated at 37 C and tested for viable cell number at the stated intervals,by use of the colony plate count method.

that cobalt does, in time, have a significant effect on the viability of these bac-teria.

SUMMARY

The growth of representative species of bacteria, both aerobic and anaerobicas well as gram-positive and gram-negative, is completely inhibitable by con-centrations of cobalt ranging from 1 to 100 ppm. The actual inhibiting cobaltconcentration depends upon the sensitivity of the individual strain of bac-terium, the number of cells per ml used as inoculum, and the constituents ofthe growth medium. As an example of the importance of the last-named factor,the concentration of cobalt necessary to inhibit the growth of Proteus vulgaris inmeat extract peptone broth is 100 times that required in a synthetic medium ofglucose and ammonium sulfate.

820 [VOL. 58

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

COBALT AND BACTERIAL GROWTH

Of 17 amino acids tested under physiological conditions of pH and tempera-ture, only histidine and cysteine are capable of overcoming the growth inhibitionof P. vulgaris by cobalt. For complete prevention or reversal of the growth in-hibition, the molar ratio of histidine to cobalt must be at least 2 to 1.The pH of the medium and the temperature of incubation have no effect

on the concentration of cobalt required to inhibit the growth of P. vulgaris.Irrespective of the presence or absence of oxygen, the rates of growth of thisspecies and of Staphylococcus aureus are initially affected by the same cobaltconcentrations. For complete inhibition, approximately two to three times asmuch cobalt is required under anaerobic as under aerobic conditions.

Cells of P. vulgaris, incubated at 37 C for 2 hours in nutrient broth containingan inhibiting concentration of cobalt, fail to show any increase in cell size orevidence of cell division. By the end of the 2-hour incubation period, however,they do show an increase in stainability with crystal violet comparable to thatof the controls. If cobalt is added to nutrient media inoculated with cells intheir resting, lag, and logarithmic phase of growth, their viability after a periodof 4 hours is decreased to 75, 40, and 2 per cent of the initial value, respectively.The rate of reduction in the viability of resting cells effected by cobalt in anutrient medium or in phosphate buffer is greater than the rate of death ofcells in phosphate buffer alone.

REFERENCESBOKORNY, T. 1905 Nochmals tiber die Wirkung verdunnter Losungen auf lebende Zellen.

Arch. ges. Physiol. (Pgfigers), 110, 174-226.BOKORNY, T. 1913 Einwirkung von Fe, Mn, Zn, and CdSO4 auf die Vermehrung der Hefe.

Allgem. Brauer Hopfen Z., 53, 223.BuRK, D., HEARON, J., CAROINE, L., AND SCHADE, A. L. 1946 Reversible complexes of

cobalt, histidine, and oxygen gas. J. Biol. Chem., 165, 723-724.BURK, D., SCHADE, A. L., HESSELBACH, M. L., AND FISCHER, C. E. 1946 Cobalt inhibition of

tissue respiration, glycolysis, and growth. Federation Proc., 5, 126-127.HEARON, J. 1948 The configuration of cobaltodihistidine and oxy-bis(cobaltodihistidine).

J. Natl. Cancer Inst., 9, 1-11.HEARON, J., BURK, D., AND SCHADE, A. L. 1949 Physical chemical studies of reversible and

irreversible complexes of cobalt, histidine, and molecular oxygen. J. Natl. Cancer Inst.In press.

HEERON, J., SCHADE, A. L., LEVY, H., AND BURK, D. 1947 Cobalt inhibition of tumor respira-tion and protection by histidine. Abstract of paper delivered before the meeting of theAmerican Association for Cancer Research, Chicago, Ill., May 16-17, 1947.

JOHNSON, F. H., CARVER, C. M., AND HARRYMAN, W. K. 1942 Luminous bacterial auxano-grams in relation to heavy metals and narcotics, self-photographed in color. J. Bact.,44, 703-716.

KNIGHT, B. C. J. G. 1936 Bacterial nutrition. Material for comparative physiology of bac-teria. Med. Research Council (Brit.), Special Rep. Series, No. 210.

KRAUSs, F., AND COLLIER, W. A. 1931 tiber die biologischen Wirkungen von anorganischenStoffen. I. Mitteilung: Die Wirksamkeit verschiedener Schwermetallverbindungenauf Bakterien, Blutparasiten and den experimentellen Mausekrebs. Arch. exptl. Path.Pharmakol., 162, 452-462.

LEVY, H. B., SKuITCH, E., AND SCHADE, A. L. 1949 The effect of cobalt on the relationshipbetween nucleic acid concentration and growth rate in Proteus vulgaris. Manuscript inpreparation.

1949] 821

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

822 ARTHUR L. SCHADE [VOL. 58

MASCHIERPA, P. 1929 Kobalt und experimentelle Lungentuberkulose. Arch. exptl. Path.Pharmakol., 189-213.

MICHAELIS, L., AND BARRON, E. S. G. 1929 Comparative study of the complexes of cysteinewith the metals of the iron group. J. Biol. Chem., 83, 191-210.

PETERSON, W. H., AND PETERSON, M. S. 1945 Relation of bacteria to vitamins and othergrowth factors. Bact. Revs., 9, 49-109.

RENON, M. L. 1915 Le probleme de la chimiothdrapie de la tuberculose. Rev. gen. clin.therap., 29, 433-437.

RONDONI, P. 1920 Ricerche sperimentali sulla chemoterapia della tubercolosi con particulareriguardo ad alcuni composti del nichelio. (Lo Sperimentale T., 73, 1919, p. 93) Zentr.Bakt. Parasitenk., I, Ref., 70,427-428.

SCHADE, A. L., AND CAROLINE, L. 1944 Raw hen egg white and the role of iron in growthinhibition of Shigella dysenteriae, Staphylococcus aureus, Escherichia coli, and Sac-charomyces cerevisiae. Science, 100, 14-15.

SCHADE, A. L., AND LEVY, H. B. 1949 Cobalt and bacterial metabolism with special referenceto Proteus vulgaris. Manuscript in preparation.

on Septem

ber 9, 2019 by guesthttp://jb.asm

.org/D

ownloaded from