APPLIED AND ENVIRONMENTAL MICROBIOLOGY, OCt. 1976, p. 585-591Copyright © 1976 American Society for Microbiology

Vol. 32, No. 4Printed in U.S.A.

Quantitative Requirements for Exponential Growth ofAlcaligenes eutrophusR. REPASKE* AND A. C. REPASKE1

National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda,Maryland 20014

Received for publication 21 April 1976

Quantitative nutrient requirements for unrestricted autotrophic growth ofAlcaligenes eutrophus were determined. Minimum saturating concentrations ofMg2+, S042, P043, Fe3+, and Na2+ for an optical density increase of 2 were 10-M, 8 x 10-5 M, 5 x 10-4 to 6 x 10-4 M, 10-5 M, and 10-7 to 2 x 10-7 M,respectively. Trace metal requirements for cobalt, chromium, and copper werealso demonstrated, but minimum concentrations could not be determined be-cause other reagents contributed a high background of these metals. Undercertain conditions an apparent response to zinc was observed, although otherexperiments suggest the zinc salt contained another metal that was required forgrowth. Poly-,8-hydroxybutyrate biosynthesis was shown to be initiated by amagnesium or sulfate deficiency as well as by a nitrogen or phosphate defi-ciency.

When large quantities of cells are required,the simplest expedient is to increase normalculture volumes. However, this usually resultsin a relatively small increase in cell yield. Con-tinuous cultures are another alternative, butconsiderable time is required to establishproper conditions. In principle it should also bepossible to grow high-density cultures in largervolumes than reported by Gerhardt's group (4,15, 17). Alcaligenes eutrophus was eminentlysuited for these studies because it grows auto-trophically in a defined inorganic medium withno significant production of metabolic wasteproducts.Media currently being used for A. eutrophus

produce 6 x 109 to 8 x 109 cells per ml in batchcultures before one or more nutrients becomegrowth limiting (13). Arbitrarily increasing theinitial concentration of all culture nutrientsdoes not result in a proportional increase ingrowth; higher concentrations of some nutri-ents are toxic and others cause precipitate for-mation. Periodic supplements of specific nutri-ents are required. If additions of individual nu-trients could be made before their concentra-tions became gro-.th rate limiting, the expo-nential-growth phase might be extended tohigher optical densities. Recognition of an im-pending nutritional deficiency requires know-ing the relationship between a concentration ofnutrient and the extent of exponential growth

' Present address: Department of Bacteriology, Ameri-can Type Culture Collecton, Rockville, MD 20852.

that that concentration would support. This ar-ticle presents results of experiments designedto determine that relationship. During thesestudies three new trace metal requirementswere found. The application of these data to 23-liter batch cultures is presented in an accompa-nying article (13); exponential growth withminimum doubling time was extended to opti-cal densities in excess of 40 (8 x 1010 cells perml).

MATERIALS AND METHODSHydrogenomonas eutropha, now designated Alca-

ligenes eutrophus (3, 5), was grown autotrophicallyas 100-ml liquid cultures in indented 500-ml Erlen-meyer flasks and shaken at 31°C under a gas atmos-phere of H2, 02, and CO1 (7:2:1). Conditions forgrowth and methods for continuously supplying thegas atmosphere have been described (10, 11).

Media. Media were prepared with deionized wa-ter and reagent-grade chemicals. The medium usedpreviously was modified by deleting NaCl, changingthe concentration of several macronutrients, andreplacing the trace metal solution with three re-quired metals. The new medium was prepared with95 ml of 0.03 M potassium phosphate buffer, pH 6.5,to which the following nutrients were asepticallyadded to give the indicated final concentrations:NH4C1, 1.8 x 10-2 M; CaCl2, 7 x 10-5 M; NaHCO3,1.2 x 10-2 M; MgSO4, 5 x 10-4 M (or MgCl2 andK2SO4, each at 5 x 10-4 M); and FeCl3 in 0.001 MHCI or Fe(NH4)2(SO4)2, 6 x 10-5 M. To these ma-cronutrients were added 3 x 10-7 M NiCl2, 4 x 10-7M CuCl2, and 5 x 10-7 M CrCl3. The former tracemetal mixture (10) provided: CoC12, 8 x 10-9 M,MnCl2, 2 x 10-5 M; CuSO4, 8 x 10-8 M; Na2MoO4, 2

585

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

586 REPASKE AND REPASKE

X 10-7 M; and ZnSO4, 3 x 10-7 M. This mixtureminus ZnSO4 was used in several experimentswhere noted.

Although reagent-grade chemicals contained onlysmall percentages of impurities, the micronutrientcontribution from this source was significant whenrelatively high concentrations of macronutrientswere used. Therefore, in specified experiments,NH4Cl, CaCl2, MgCl2, K2SO4, and phosphate bufferwere treated with a chelating agent, 8-hydroxyquin-oline (8HQ), to reduce trace metal impurities. Con-centrated stock solutions of each of the above com-pounds were adjusted to pH 5.5 (phosphate was pre-pared from KH2PO4 and adjusted with KOH to pH5.5), and one-fourth volume of 1% 8HQ was added,mixed, and allowed to stand for 2 h. 8HQ metalchelates and unreacted chelator were repeatedly ex-tracted with chlorform until the chloroform wascolorless. The pH was then raised to 7.0 (KOH), andresidual chelator was removed by additional chloro-form extractions. Traces of chloroform remaining inthe aqueous phase were vaporized during autoclav-ing.

Inocula. Inocula were 18-h cultures that had beenincubated at 31°C. The same inoculum could be usedfor several weeks, since stored liquid cultures re-tained undiminished viability for at least 30 daysand no "aging" effects occurred during storage (12).For macronutrient assays, inocula were grown incomplete untreated media. For micronutrient tests,inocula were grown in (i) untreated media withoutadded trace metals or (ii) in 8HQ-treated mediawithout added trace metals.

Assay. The minimum saturating concentration ofeach nutrient was found by determining the mini-mum concentration of nutrient which supported thesame growth as a control culture containing excessnutrient. The assay end point occurred when controlcultures had reached an optical density (OD) of 2.0to 2.5 and were still growing exponentially. Thisdensity was reached after 18 h under standardizedassay conditions, namely, a washed inoculum resus-pended in 0.001 M P04, pH 7, inoculated cultureswith a starting OD of 0.020, and an incubation tem-perature of 31°C. Six concentrations of a nutrientwere tested simultaneously; four were limiting, onewas at or near the minimum saturating concentra-tion, and one, the control, was in excess. Assayresults were independent of inoculum size (*+three-fold) if cultures were removed when the control cul-ture reached the required OD. The background con-centrations of C02+, Cu2+, and Cr3+ were too high todetermine their minimum requirements. Effects ofthese micronutrients were based on total culturegrowth in 24 h as noted.The OD was determined with a spectrophotome-

ter at 660 nm in a standard 1-cm light path cuvetteagainst medium as a reference. Cell suspensionswere diluted as necessary to an OD of 0.020 to 0.400,the range of direct proportionality between OD andcell numbers. Optical density correlated with platecounts in autotrophic or heterotrophic media (2 x109 to 3 x 109 cells per ml per OD) and cell protein-nitrogen values. Enlarged nutrient-deficient cellscontaining poly-,B-hydroxybutyrate (PHB) had a

higher OD than a similar concentration of normalcells, but the spurious contribution to OD was nogreater than 10% (18) and did not affect the generalvalidity of OD as a measure of growth.PHB determination. PHB was determined by a

modification of the method of Law and Slepecky (9).The variation between replicate determinations was±1% when PHB standards were used. When wholecell samples were used, the variation was ±+5%,apparently because of difficulty in digesting gram-negative cells. Highest yields and most consistentresults were obtained when the washed cell pelletwas frozen and thawed several times before Chloroxwas added. Chlorox digestion was continued for 1 hat 37°C, after which PHB granules were sedimentedby centrifugation. Traces of chlorine were removedwith a small amount of sodium thiosulfate (19)added to the PHB resuspended in water. Successivewashes in water, acetone, and absolute ethanolwere followed by gentle heating to evaporate resid-ual alcohol. Hot-chloroform extractions were notnecessary because there was no detectable interfer-ence by light-absorbing contaminants in the criticalspectral region. Five milliliters of concentratedH2SO4 was added to the dry pellet; the tubes werecapped with a glass marble and heated in a sandbath at 100°C for 20 min to oxidize PHB to crotonicacid. Samples were then diluted as necessary inconcentrated H2SO4, and the absorption spectrum ofcrotonic acid was obtained between 200 and 340 nmwith a Cary model 14 spectrophotometer. A molarextinction of 1.55 x 104 at 235 nm was used to deter-mine the concentration of crotonic acid and to calcu-late the original PHB concentration (9).

Nitrogen determinations. Ammonium nitrogenwas determined by the Nesslers reaction (6), andcell protein was determined by the biuret method (8,16) with bovine serum albumin as a reference stan-dard. Cell protein-nitrogen was calculated from biu-ret protein values on the assumption that cell pro-tein averaged 16% nitrogen.

RESULTS

The modified medium used in these studiescontained sufficient nutrient to permit expo-nential growth to continue to OD 3.5 and totalgrowth to reach OD 4.5. Subsequent growthwas limited by an ammonium-nitrogen defi-ciency; cultures provided with additionalNH4C1 reached a final OD of about 6 beforeanother nutrient limited growth (13).

If exponential growth could be extended toobtain higher culture densities, it would re-quire making individual supplements of eachnutrient before depletion reduced the concen-tration to a growth-limiting level.To anticipate a requirement, the amount of

exponential growth that can be expected from agiven concentration of nutrient must be known.That relationship could be determined by ob-serving culture growth with a series of nutrientconcentrations or by selecting an increment of

APPL. ENVIRON. MICROBIOL.

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

EXPONENTIAL GROWTH OF A. EUTROPHUS 587

growth and determining the minimum nonlim-iting concentration for that amount of culturegrowth. The latter approach was chosen. Agrowth increment of OD 2 to 2.5 (5 x 10-9 to 7 x109 cells per ml) was selected because at thisdensity cultures with excess nutrient were stillgrowing exponentially. The minimum concen-tration of nutrient that duplicated growth of acontrol culture containing excess nutrient wasdefined as the minimum saturating concentra-tion for unrestricted growth of 2 OD units ofcells. This amount of growth was sufficient todemonstrate nutrient-dependent growth withmost nutrients.

Macronutrients. A growth response curvefor each macronutrient except NH4Cl andCaCl2 was determined in complete medium con-taining various concentrations of the nutrientbeing tested. Since nitrogen limitation did notoccur before OD 3.5, it was not limiting duringthe assay; calcium had no demonstrable effecton growth, but its presence in most salts com-promised any assessment of a requirement.The effect of iron, sulfate, magnesium, and

phosphate concentrations on growth defined bythe assay is shown in Fig. 1. In phosphate-limited cultures, the buffer was 0.03 M Tris-chloride, pH 6.5, which had no deleterious ef-fect on growth. A direct relationship betweengrowth and nutrient concentration was shownby plotting the logarithm of culture densityagainst the logarithm of nutrient concentra-tion. The comparibility of the results were seenmore readily if the density of each culture in aseries was related to the maximum density ofits control culture; therefore, the logarithm ofthe percentage of maximum growth was used.Growth increased proportionally with highernutrient concentration until saturation oc-curred; thereafter additional nutrient had noeffect. The minimum saturating concentrationswere: Fe2+ or Fe3+, 10-5 M; SO43-, 8 x 10-5 M;Mg2+, 10-4 M; and PO43-, 5 x 10-4 to 6 x 10-4 M.At the lowest macronutrient concentrations

tested, cultures were deficient for the greaterpart of the incubation period, and secondaryeffects caused by the deficiencies were ob-served. One of these effects was culture foam-ing, which only occurred with an iron defi-ciency. Several hours after foaming began,iron-limited cultures produced a yellow-greenpigment visible in culture filtrates. The pig-ment increased in intensity with continued in-cubation. Cells from iron-deficient culturescharacteristically were about one-half the sizeof normal cells, and they stained evenly. Incontrast, cells from nitrogen-, phosphate-, mag-nesium-, or sulfate-deficient cultures were

1.8 Fe M

2 ~~~~~~~~~~~~~~~PO22 1.6-x ~~~~~~SO"

a1 1.4

12

10-M lO0sM 10-'M 1O-3MCONCENTRATION

FIG. 1. Minimum saturating concentration ofma-cronutrients for unrestricted growth ofA. eutrophusto OD 2.0 to 2.5. The assay was conducted with theindicated concentrations ofeach nutrient in an other-wise complete medium containing trace metals. Thebuffer in the phosphate experiment was 0.03 M Tris-chloride, pH 6.5. Incubation was under a 70% Hr20% 0210% CO2 atmosphere for 18 h at 31'C.

large, and large areas in the cytoplasm re-mained unstained. This appearance suggestedPHB inclusions, and it was confirmed by PHBanalysis. PHB synthesis resulting from nitro-gen or phosphate starvation was reported bySchlegel et al. (14). They found that continuedincubation of nitrogen-deficient cultures underH2, 02, and CO2 produced cultures containingas much as 65% dry weight PHB. Until recentlywe were not aware that any other laboratoryhad noted PHB synthesis as a consequence ofmagnesium or sulfate deficiency (7).PHB synthesized in cultures deficient in

phosphate, sulfate, or magnesium was com-pared with normal- and iron-deficient cultures(Table 1) in medium containing Ni2+ and tracemetal mixture. No attempt was made to selectoptimum conditions for the highest yield ofPHB. The high OD of the normal culture, no. 1(OD 4.840), is accounted for by a larger inocu-lum than was used in nutrient assays and alonger incubation period (20 h). The high ODindicated growth had exceeded the point of ni-trogen exhaustion, and the PHB content, al-though low, was elevated. No more than 1%PHB would be present if the OD were 3.5 orless. The selected concentrations of phosphate,sulfate, and magnesium produced cultures dif-fering in total growth, but the final PHB wasuniformly high, ranging between 22 and 34% ofthe total dry weight. PHB averaged about 50%of the total cell protein, or three to four timesthe cell protein nitrogen, and was highest insulfate-limited cultures. Normal and iron-defi-cient cultures had a 3.6-fold difference in totalgrowth; nevertheless, they had a similar lowcontent of PHB.

VOL. 32, 1976

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

TABLE 1. PHB synthesis and growth ofA. eutrophus in relation to nutrient deficienciesResidual PHB

Growtha.Limiting NH4-N Cells Cell pro-Culture no. (OD,,,) nutrient Concn (M) in me- (mg dry tein-N mg/mg of % dry

dium wt) (mg) mg protein- wt(mg) N w

Uninoculated None 26.21 4.840 None 4.5 174 20.3 5.9 0.29 32 1.640 P043- 2 x 10-4 18.6 59 5.9 19.6 3.32 333 0.730 So42- 2.5 x 10-5 24.1 26 2.3 9.0 3.91 344 2.580 Mg2+ 5 x 10-5 14.3 93 7.6 20.1 2.64 225 1.340 Fe3+ 3 x 10-6 18.1 48 7.1 2.9 0.41 6

a Data from 100-ml cultures incubated for 20 h.

Before the association of Mg2+ deficiency andPHB synthesis was recognized, Bongers (2)noted that NH4+- and Mg2+-starved culturescontinued to consume H2, 02, and CO2. Rein-troduction of Mg2+ inexplicably caused CO2 as-similation to cease for 30 min, after which therate of assimilation gradually increased, reach-ing its full capacity in the third hour. Thedemonstration of Schlegel et al. (14) of reassim-ilation of PHB after NH4+ was resupplied toNH4+-starved cultures and our observation ofthe same events in Mg2+-deficient cultures sug-gest that what Bongers observed was the pref-erential utilization of accumulated PHB at theexpense of CO2 assimilation.

Micronutrients. A nickel requirement for A.eutrophus was first demonstrated by Barthaand Ordal (1) with 8HQ-extracted medium sup-plemented with Hoagland trace metal solution.Bongers (2) reported that nickel increasedsteady state growth rates in continuous cul-tures. The dependence of autotrophic growth onnickel, and the difficulty in extracting it to dem-onstrate the requirement, indicates that appre-ciable nickel contamination occurs in ordinarymedium. Nickel could be depleted by growingcultures successively in the same medium (cu-mulative OD of about 10) supplemented as nec-essary with NH4+, Mg2+, S042-, and Fe2+ andcorrected for pH changes. These results inci-dentally indicated the above reagents were notmajor contributors of adventitious nickel.Our results with 8HQ-extracted medium

components differed quantitatively from thoseof Bartha and Ordal (1); extracted mediumwithout added trace metals still permittedgrowth to OD 1 in 24 h. When 3 x 10-7 M Ni2+and trace metals were included, the ODreached 2, or only about one-half the cell den-sity obtained in untreated medium (see Tables4 and 5). 8HQ extraction apparently had re-moved some other necessary but unknowntrace metal(s), and this medium was not con-sidered satisfactory for testing the response toNi2+.

The nickel requirement could be satisfacto-rily demonstrated in an unextracted mediumcontaining only macronutrients if the inoculumhad also been grown in medium without addedtrace metals. With this inoculum, culturesgrew to an OD of 1.5 to 2 when nickel and othertrace metals were not provided and to an OD of3.5 in 24 h when Ni2+ was added. Full growth toOD 4.5 in 24 h occurred when copper and chro-mium (discussed below) were added with Ni2+.The effect of nickel on growth (Fig. 2) wasdetermined by the usual 18-h assay in this me-dium. Nickel saturation occurred at 10-7 to 2 x10-7 M added nickel, essentially the same con-centration found by Bartha and Ordal (1) in amedium showing complete growth dependenceon nickel.

Nickelous chloride contained 0.1% cobalt asan impurity and therefore provided 5 x 10-9 MCO2+ to the medium when 3 x 10-7 M NiCl2 wasused. Some uncertainty existed about the speci-ficity of the Ni2+ response, since cobalt at thisconcentration could conceivably satisfy a co-balt requirement. Nickelous chloride and an-other nickel compound, nickelous ammoniumchloride containing one-fifth less cobalt on amolar basis, were compared at limiting Ni2+concentrations (5 x 10-8 M) in a macronutrient

° 2.0

E 1.8

5 16

81O-7M

Ni - CONCl0- M

FIG. 2. Minimum saturating concentration ofadded nickel for unrestricted growth ofA. eutrophusto OD 2.0 to 2.5. The macronutrient medium con-tained 4 x 10-7 M Cr3+ and 5 x 1O-7 M Cu2+ asmicronutrients and the indicated concentrations ofNiC12. Refer to the legend to Fig. 1 for growth condi-tions.

588 REPASKE AND REPASKE APPL. ENVIRON. MICROBIOL.

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

EXPONENTIAL GROWTH OF A. EUTROPHUS 589

medium supplemented with copper and chro-mium (Table 2). Growth with both nickel saltswas the same. When 8 x 10-9 M CoCl2 (contrib-uting 3 x 10-1" M Ni2+) was included, growthwas stimulated 180% and was equivalent togrowth with 10-7 M Ni2+ without added cobalt.No significant stimulation was obtained withCo2+ alone. These results indicated that bothnickel and cobalt were required and suggestedthe saturating nickel concentration determinedin Fig. 2 may be excessive because it repre-sented the amount of NiCl2 needed to supplyboth Ni2+ and C02+. Further study of the Ni2+and Co2+ effects were not made because baseline growth without added Ni2+ was too high.Chromium and copper stimulation of A. eu-

trophus growth was first detected indirectly inlarger fermentor cultures (13). Supplements ofMgSO4 maintained the growth rate more effec-tively than equivalent concentrations of MgCl2and K2SO4. Analysis of the two magnesiumsalts (Bureau of Standards) revealed MgSO4contained between 0.001 and 0.01% of bothchromium and calcium as major impurities andless than 0.001% copper. No chromium wasdetected in the MgCl2, but copper and calciumwere present in amounts of 0.01 to 0.1%. Theeffect of nickel, chromium, and copper singlyand together in the macronutrient medium isshown in Table 3. The inoculum was grownwithout trace metals. Nickel stimulated growthtwofold, and Cr3+ and Cu2+ individually in-creased growth (with nickel) by an additional30%. Higher concentrations of Cr3+ or Cu2+ pro-duced no further stimulation. Neither Cr3+ nor

TABLE 2. Cobalt stimulation ofgrowth with limitingnickel

Addition to medium" Growth

None ............. 2.040Co2+ (8 x 10-9 M) ..... ....... 2.190Ni2+ (5 x 10-8 M) ...... ....... 2.490plus C02+............ 3.660

Ni2+ (10-7 M) .... ........ 3.750a Medium contained macronutrients and 4 x 10-7

M Cu2+ and 5 x 10-7 M Cr3+.

TABLE 3. Ni2+, Cr3+, and Cu2+ stimulation ofA.eutrophus growth

Addition to mediuma % StimulationbNone.+ Ni2+ (3 x 10-7 M) ........ ........... 104+ Cr3+ (5 x 10-7 M) ........ ........... 130+ CU2+ (4 x 10-7 M) ....... ........... 136+ Cr3+ and Cu+.. ...................... 174

Cu2+ had any effect without Ni2+. Additivestimulation by Cr3+ and Cu2+ produced culturesof the same densities as ordinary inocula inmedium containing the trace metal mixture.No detectable zinc stimulation (2) was foundwhen zinc was used instead of Cr3+ or Cu2+ or inconjunction with them.A reexamination of the 8HQ-extracted me-

dium seemed warranted once Cr3+ and Cu2+requirements had been demonstrated. Singlereplacements of extracted components for unex-tracted ones showed unextracted phosphatecontributed substances responsible for superiorgrowth in the unextracted medium (Table 4).Cr3+, Cu2+, and Zn2+ were tested in extractedmedium supplemented with C02+ and nickel todetermine if these were the critical metals re-moved by extraction (Table 5). The twofoldstimulation by Ni2+ discussed previously isshown in line 2. Chromium stimulated growthby an additional 20%, but Cu2+ stimulationfound previously in nonextracted medium (Ta-ble 3) was absent (not shown). Unexpectedly,growth was now influenced by Zn2+. Stimula-tion by Zn2+ was about the same as with unex-tracted phosphate (Table 4). Neither Cr3+ norCu2+ enhanced Zn2+ stimulation. Zinc stimula-tion of growth was observed only in 8HQ-ex-tracted medium and could not be demonstratedin unextracted medium, even after attemptingto deplete Zn2+ by growing cultures to an OD of35 (13) with no phosphate supplements.

Other trace metals contained in the originaltrace metal solution either were not requiredfor growth or adequate amounts were suppliedas impurities.

DISCUSSIONThe relationship between nutrient concen-

tration and the amount of exponential growthTABLE 4. Effect of nickel and unextracted phosphate

on growth medium extracted with 8HQAddition to mediuma A ODm,/24 h

None ............ ............ 1.200NiC12 (10-7 M) ...................... 2.520NiCl2 plus unextracted PO4 - .......... 4.060

a 8HQ-extracted macronutrients plus Fe3+.

TABLE 5. Cr3+ and Zn2+ effects in growth ofA.eutrophus in 8HQ-extracted medium

Additions to mediuma A OD,,/24 h

None ......... ............. 1.330Ni++ (3 x 10-7 M) .................... 2.420+ Cr3+ (10-7 M) .................... 2.890+ Zn2+ (10-7 M) .................... 3.440+ Zn2+ and Cr3+ ..................... 3.340

a 8HQ-extracted macronutrients plus Fe3+.aMedium contained macronutrients only.bIncubation was for 18 h.

VOL. 32, 1976

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

590 REPASKE AND REPASKE

that could be supported was determined to pro-vide a basis for scheduling additions of nutri-ents to cultures before the concentration of anynutrient became growth rate limiting. The pur-pose was to extend the period of exponentialgrowth to increase cell yields in batch cultures.The minimum concentration of nutrient for un-restricted growth to a given population densitywas determined by an assay which comparedgrowth at various nutrient concentrations withgrowth in a control culture containing an ex-cess of the nutrient being tested. At the endpoint (OD 2 to 2.5) the control was still growingexponentially. The resulting growth responsecurves (Fig. 1 and 2) show a proportional re-sponse of growth to nutrient concentration. Therelatively small deviation in the fit of experi-mental points results from the fact that growthin individual cultures in each series was thesame until each culture reached nutrient limi-tation; growth then stopped without passingthrough extended transitions of gradually de-creasing rates.

Since nutrient uptake ultimately is concen-tration dependent, some residual nutrient mustremain when growth stops. It is assumed thatgrowth stopped in all flasks when the sameresidual nutrient concentration was reached. IfthAt residual concentration were high, thequantity of nutrient consumed would be calcu-lated from the change in concentration. If, how-ever, the residual concentration were very lowwhen growth stopped, the total nutrient addedcould be considered necessary for the observedcell growth.The two possibilities may be distinguished by

comparing ratios of total growth to nutrientconcentration with each nutrient. Constant ra-tios would indicate growth at each limiting nu-trient concentration was proportional to theavailable nutrient. It would also suggest littleresidual nutrient remained when growthstopped. If appreciable nutrient had remained,the ratios would increase with increased nutri-ent concentration. This would follow becauseproportionally more growth would be obtainedat higher nutrient concentrations where theresidual nutrient concentration would be asmaller percentage of total nutrient supplied.The ratios for different limiting concentrationsof Mg2+ and of SO42- varied by less than 20%,whereas for PO43- and for Fe2+ they weregreater (50%) and ratios decreased progres-sively with increasing nutrient concentration.This was contrary to expectations if residualconcentrations of nutrient remained high, andit suggested the two latter nutrients causedsome inhibition of growth at higher concentra-

tions. The general conclusion, therefore, wasthat residual nutrient was low and could bediscounted when estimating nutrient con-sumed. The minimum saturating concentra-tions of nutrients shown in Fig. 1 and 2 wereregarded as the quantity required to produce 2OD units of cells with no restriction on thegrowth rate.

Detection of required micronutrients was dif-ficult because metal impurities in reagents of-ten exceeded the biological requirements. Forexample, nickel salts provided sufficient cobaltto meet the growth requirements of the orga-nism, and cobalt stimulation could be demon-strated only when minimum nickel was used(Table 2). Stimulation by such a low concentra-tion of Co2+ indicated Co2+ itself was the spe-cific active metal. The situation with Cr3+ andCu2+ was not as clear. Although both metalsapparently were required (Table 3), the factthat either one could independently stimulategrowth suggested growth was promoted by ametal contaminant common to both salts. In-deed, zinc could be that contaminant basedupon the Zn2+ stimulation observed in ex-tracted medium (Table 5). However, if thatwere true, Zn2+ should have replaced or aug-mented Cr3+ or Cu2+ in unextracted medium(Table 3). It did not. The identity of the specificrequired metals from this group or the activemetals they contribute remain to be resolved.The simplified minimal medium containing

the macronutrients discussed and micronu-trients of nickel (with cobalt), chromium, andcopper salts supported growth to OD 4.5, anddoubling times were 1.7 to 2 h, which waswithin the range of minimum doubling timesreported by Bongers (2). Nutritional adequacyof the medium is further supported by the factthat 23-liter batch cultures could be grown toOD 60 or 1.2 x 1011 cells per ml with thesenutrients alone (13). Exponential growth inthose cultures was maintained to OD 40 byscheduling nutrient supplements based uponthe requirements presented in this report.

LITERATURE CITED1. Bartha, R., and E. J. Ordal. 1965. Nickel-dependent

chemolithotrophic growth of two Hydrogenomonasstrains. J. Bacteriol. 89:1015-1019.

2. Bongers, L. 1970. Some aspects of continuous culture ofhydrogen bacteria, p. 241-255. In C. J. Corum (ed.),Developments in industrial microbiology, vol. 2. Gar-emond-Predemark Press, Baltimore.

3. Davis, D. H., M. Doudoroff, R. Y. Stanier, and M.Mandel. 1969. Proposal to reject the genus Hydrogen-omonas: taxonomic implications. Int. J. Syst. Bacte-riol. 19:375-390.

4. Gerhardt, P., and D. M. Gallup. 1963. Dialysis flask forconcentrated culture of microorganisms. J. Bacteriol.86:919-929.

APPL. ENVIRON. MICROBIOL.

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from

EXPONENTIAL GROWTH OF A. EUTROPHUS 591

5. Halaus, D., R. Y. Stanier, M. Doudoroff, and M. Man-del. 1970. Taxonomic studies on some gram negativepolarly flagellated "hydrogen bacteria" and relatedspecies. Arch. Mikrobiol. 70:1-13.

6. Johnson, M. J. 1941. Isolation and properties of a pure

yeast polypeptidase. J. Biol. Chem. 137:575-586.7. Kesler, T. G., Y. V. Voitovich, N. A. Anistratova, I. N.

Trubachev, and N. S. Eroskin. 1973. The growth ofhydrogen bacteria under conditions of biosynthesisblocked by biogenic elements. Mikrobiologiya 41:456-460.

8. Koch, A. L., and S. L. Putnam. 1971. Sensitive biuretmethod for determination of protein in an impuresystem. Anal. Biochem. 44:239-245.

9. Law, J. H., and R. A. Slepecky. 1961. Assay of poly-(3-hydroxybutyric acid. J. Bacteriol. 82:33-36.

10. Repaske, R. 1962. Nutritional requirements for Hydro-genomonas eutropha. J. Bacteriol. 83:418-422.

11. Repaske, R. 1966. Characteristics of hydrogen bacteria.Biotech. Bioeng. 8:217-235.

12. Repaske, R., C. A. Ambrose, A. C. Repaske, and M. L.DeLacy. 1971. Bicarbonate requirement for elimina-tion of the lag period ofHydrogenomonas eutropha. J.Bacteriol. 107:712-717.

13. Repaske, R., and R. Mayer. 1976. Dense autotrophiccultures ofAlcaligenes eutrophus. Appl. Environ. Mi-crobiol. 32:592-597.

14. Schlegel, H. G., G. Gottschalk, and R. V. Bartha. 1961.Formation and utilization of poly-,8-hydroxybutyrateby knallgas bacteria (Hydrogenomonas). Nature(London) 191:463-465.

15. Schultz, J. S., and P. Gerhardt. 1969. Dialysis culturesof microorganisms: design, theory and results. Bacte-riol. Rev. 33:1-47.

16. Strickland, L. H. 1951. The determination of smallquantities of bacteria by means ofthe biuret reaction.J. Gen. Microbiol. 5:698-703.

17. Tyrrell, E. A., R. E. MacDonald, and P. Gerhardt.1958. Biphasic system for growing bacteria in concen-trated culture. J. Bacteriol. 75:1-4.

18. Wilde, E. 1962. Untersuchungen uber Wachstum andSpeicherstoff synthese von Hydrogenomonas. Arch.Mikrobiol. 43:109-137.

19. Williamson, D. H., and J. F. Wilkinson. 1958. Theisolation and estimation of the poly-p-hydroxybutyr-ate inclusions of Bacillus species. J. Gen. Microbiol.19:198-209.

VOL. 32, 1976

on February 20, 2021 by guest

http://aem.asm

.org/D

ownloaded from