FOLIC ACID DERIVATIVES SYNTHESIZED DURING GROWTH OFDIPLOCOCCUS PNEUMONIAE

FRANCIS M. SIROTNAK, GLORIA J. DONATI, AND DORRIS J. HUTCHISONDivision of Experimental Chemotherapy, Sloan-Kettering Institute for Cancer Research, and Sloan-

Kettering Division, Graduate School of Medical Sciences, Cornell University Medical College,New York, New York

Received for publication 25 October 1962

ABSTRACT

SIROTNAK, F. M. (Sloan-Kettering Institutefor Cancer Research, New York, N.Y.), GLORIAJ. DONATI, AND DORRIS J. HUTCHISON. Folic acidderivatives synthesized during growth of Diplo-coccus pneumoniae. J. Bacteriol. 85:658-665.1963.-Under cultural conditions permittingsynthesis of folic acid in an amount greatly inexcess (20- to 30-fold) of that required formaximal growth of Diplococcus pneumoniae,85 to 90% of the growth factor accumulated aspolyglutamates. Approximately equal amountsof mono- and diglutamates made up the remain-ing 10 to 15% found in culture material. Most ofthe polyglutamates occurred intracellularly, in aproportion of triglutamates to higher glutamatesof about two to one. Only 10 to 15% of all folicacid derivatives (mono-, di-, and polygluta-mates) found had folinic acid (5-formylfolate-H4) activity for Pediococcus cerevisiae. Practi-cally all synthesis of the glutamyl-peptidemoieties of folate seems to occur at an enzymaticstep prior to folic acid, since no appreciablepeptide formation occurred under conditionsblocking folate synthesis. Sulfanilamide inhibi-tion of growth by a block in folate synthesis wasreversed by the addition of dihydrofolic acid,but not folic acid. The examination of twogenetically distinct amethopterin-resistant mu-tant strains has revealed no gross differences infolate accumulation during growth when com-pared with the wild strain.

Concurrent with studies of resistance to folicacid antagonists exhibited by various mutantstrains of Diplococcus pneumoniae, we have ex-amined the wild strain for folic acid derivativesbiosynthesized during growth. Of particularinterest was the fraction accumulated as poly-glutamyl conjugates.

The biosynthesis of polyglutamate forms offolic acid has been shown to occur in a number ofmicroorganisms. For example, derivatives havebeen found in yeast (Bird et al., 1946; Pfiffneret al., 1946; Doctor and Couch, 1953), in severallactic acid bacteria (Hendlin, Koditschek, andSoars, 1953), and in Neurospora crassa (Swend-seid and Nyc, 1958). A triglutamate form withactivity for Pediococcus cerevisiae was identifiedas a major constituent in cells of Bacillus subtilis(Hakala and Welch, 1955), in cells of Strepto-coccus faecalis (Zakrzewski and Nichol, 1955),and, along with a heptaglutamate derivative, inClostridium cylindrosporum (Wright, 1955), andin a number of clostridia by Rabinowitz andHimes (1960).

In the following report, evidence is presentedfor the accumulation in cultures of D. pneumoniaeof significant quantities of at least three poly-glutamyl forms of both folic acid and the reducedcoenzyme form. Additional data are also givenwhich suggest the manner in which these deriva-tives originate during growth.

MATERIALS AND METHODS

The organism used as a source of folic acid de-rivatives for most of this work was a roughvariant of D. pneumoniae strain R6 (obtainedthrough the courtesy of Rollin D. Hotchkiss, TheRockefeller Institute, New York). Two geneti-cally distinct amethopterin-resistant mutants,R6Amb and R6Amd, were also used. Each has asingle genetic marker, but at a different chromo-somal locus (unpublished data), determining 100-and 500-fold increases in resistance, respectively,and derived from mutant strains described pre-viously (Sirotnak, Lunt, and Hutchison, 1960a).

Inocula and cells used for the preparation ofextracts were grown at pH 7.4 in a modification(Sirotnak et al., 1960b) of the partially syntheticmedium of Adams and Roe (1945) in which the

658

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FOLIC ACID SYNTHESIS BY D. PNEUMONIAE

chief source of nitrogen was supplied in the formof a vitamin-free enzymatic digest of casein(General Biochemicals, Inc., Chagrin Falls,Ohio). The inoculum size was adjusted so thatmaximal growth (OD 0.75 as measured at 600mA in a Beckman model B spectrophotometer)occurred after approximately 10 hr of incubationat 37 C. Cell yield (dry weight) was about 400mg/liter. Cells in the late exponential phase(around OD 0.6) or in the stationary phase ofgrowth were harvested by centrifugation andresuspended in one-tenth the original volume incold 0.1 M phosphate buffer (KH2PO4-Na2HPO4)at pH 6.5. Both the resuspended cells and thesupernatant culture broth were autoclaved im-mediately, for 15 min at 121 C, and adjusted totheir original volume with distilled water; thesuspensions were centrifuged to remove cellulardebris. The heat treatment releases growth fac-tors from the cells and mediates the conversionof labile one-carbon substituted derivatives, suchas N'0-formyl- and N5-10-methenyl-folate-H4, tothe stable N5-formylfolate-H4. Folate-H2 andfolate-H4, presumably, are oxidized to folic acid.When extracts were examined by the Bratton-Marshall (1939) procedure for the presence ofp-aminobenzoylglutamic acid (PABG), a cleavageproduct of folate-H2 and folate-H4, none wasdetected. In addition, when synthetically pre-pared folate-H2 was added to cell extracts pre-pared by sonic treatment and then heated, noBratton-Marshall positive material could bedetected. Such material, however, was detectedalmost immediately in controls prepared ineither distilled water or phosphate buffer; acommensurate loss of activity for S. faecalis orLactobacillus casei was also noted for such con-trols. It appears, therefore, that materials in thecell afford considerable protection to the reducedfolates and that under these conditions the resultfrom heating is one of oxidation to folic acid withlittle or no cleavage of the molecule to PABG anda pteridine derivative.The heat extracts and the similarly heated

supernatant medium, were stored at 5 C undertoluene. To convert polyglutamates to readilyutilizable growth factors available for micro-biological assay, a portion of each sample, alongwith a portion of the culture supernatant, wastreated for 17 hr at 37 C with a conjugase prepara-tion from chicken pancreas or hog kidney. AfterpH adjustment to 6.5, 0.5 mg of dried chicken

pancreas (Difco) was added per 10 ml of extract,and 0.2 ml of fresh hog-kidney suspension, madeaccording to the method of Bird et al. (1945), wasadded to 10 ml of extract after it was adjusted topH 4.5. Dried hog-kidney preparation (Difco)was found to be inactive under the same condi-tions.

Prior to bioautographic analysis, samples wereconcentrated tenfold by lyophilization and thentreated with an equal volume of absolute ethylalcohol. The resulting precipitate was removedby centrifugation, and the supernatant wasstored at -20 C until used.

Folic acid derivatives were measured micro-biologically. S. faecalis ATCC 8043 and L. caseiATCC 7479 were used for assays of total folicacid, with synthetic pteroylglutamic acid as astandard; P. cerevisiae ATCC 8081 was used fordetermination of the substituted folate-H4 de-rivatives, with synthetic N5-formylfolate-H4(folinic acid) as the standard. Media and condi-tions of assay were essentially those of Flynnet al. (1951) for S. faecalis and L. casei and thoseof Sauberlich and Baumann (1948) for P.cerevisiae. Growth response was measured tur-bidimetrically at 600 m,u with a Beckman modelB spectrophotometer.

For the preparation of bioautograms, usually0.01 ml of the concentrated samples or a suitabledilution was used. This was spotted on filterpaper strips (Eaton-Dikeman no. 613, 0.5 in. indiameter) for ascending chromatography with 1%K2HPO4 plus 0.2% ethylenediaminetetraaceticacid (EDTA) as the solvent, during a period ofapproximately 7 hr at room temperature. Thepaper strips were then air-dried and placed on thesurface of the appropriate assay medium, solidi-fied by adding 1.5% agar, and seeded with one ofthe three assay organisms shown above. After20 hr of incubation at 37 C, the paper strips wereremoved and areas of growth traced to be usedfor RF calculations. All folic acid compounds usedas standards in assay procedures and as controlsfor bioautograms (see Fig. 1) were generouslysupplied by Thomas H. Jukes, AgriculturalDivision, American Cyanamid Co. Folate-H2was prepared by the method of Futterman (1957)as modified by Blakley (1960).

RESULTS

Heat extracts were prepared in the mannerdescribed above from a number of cell samples of

659VOL. 85, 1963

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SIROTNAK, DONATI, AND HUTCHISON

TABLE 1. Accumulation of folic acid derivativesduring growth of Diplococcus pneumoniae as

measured by microbiological assay

Assayorganism

Streptococcusfaecalis

Pediococcuscerevisiae

Sam-ple

III

III

III

III

Folic acid derivativesa(m,ug/100 ml of culture)b

Cells

Noad di-tion

5962133

58

21

2,0101,8401,300

148226177

i+

520669

30

Supernatant

380310455

302816

860740966

605326

a Measured as equivalents of pteroylglutamicacid with S. faecalis and as equivalents of theactive stereoisomer of synthetic N5-formyl-folate-H4 with P. cerevisiae. Average of duplicatedeterminations done at several sample dilutions.

I Cell yield is approximately 40 mg (dry wt)/100ml.

c Treated with dried chicken-pancreas prepara-tion at pH 6.5.

d Treated with fresh hog-kidney preparation atpH 4.5.

D. pneumoniae harvested at late exponential orstationary phases of growth. The results ofmicrobiological assays of the conjugase-treatedand untreated samples for total folio acid withS. faecalis and for the one-carbon substitutedfolate-H4 derivatives with P. cerevisiae of threerepresentative samples are shown in Table 1.

Total folic acid accumulating in the cultureduring growth was uniformly high (approxi-mately 2,500 mAug/100 ml of culture containingan equivalent of 40 mg of dried cells). Thisrepresents approximately a 20- to 30-fold excessof the amount required for maximal growth ofthe organism, and was estimated by growing cellsin an amount of sulfanilamide (p-aminoben-zenesulfonamide, Merck and Co., Inc., Rahway,N.J.) just allowing maximal growth and com-paring assay values with others obtained oncultures grown in the absence of the drug. Otherassays, performed on cultures during variousstages of growth, indicate a proportionate synthe-sis of folic acid derivatives, with a continualincrease in total accumulation for several hoursafter cessation of growth.

Assay values with P. cerevisiae, expressed interms of equivalents of "active" synthetic folinicacid, indicate that at most only 10 to 15% of thetotal amount present existed as a one-carbonsubstituted folate-H4 (see Table 1).From values obtained before and after treat-

ment of samples with chicken-pancreas conjugase,it can be seen that no more than 15 to 20% ofboth folate and the one-carbon substituted folate-H4 derivatives existed in a form utilizable byS. faecalis or P. cerevisiae, with the majority ofthis found extracellularly in the culture medium.The proportion of utilizable forms to nonutiliz-able polyglutamyl conjugates remained es-sentially the same during growth. Treatmentwith the fresh hog-kidney extract was consider-ably less efficient than the chicken-pancreaspreparation in converting nonutilizable materialto forms readily used by the assay organisms. IDaddition, treating samples consecutively withboth conjugase preparations did not increase thevalue obtained with the chicken-pancreas prep-aration alone. Some heated samples were also

TABLE 2. Intracellular content of folic acid deriva-tives as measured by microbiological assay

Assayorganism

Streptococcusfaecalis

Lactobacillus casei

Pediococcuscerevisiae

Sample

IVVVI

IVVVI

IVVVI

Folic acid derivatives'(mAg/100 mg of dry weight)

Noaddition

5589202

738708

1,500

201046

+CPeb

8321,2301,790

9061,1801,760

193108303

Nonuti-lizablec

7771,1411,588

168472260

17398257

aMeasured as equivalents of pteroylglutamicacid with S. faecalis and L. casei and equivalentsof the active stereoisomer of synthetic N5-formyl-folate-H4 with P. cerevisiae. Average of duplicatedeterminations done at several sample dilutions.bAssay values obtained after treatment of

samples with chicken-pancreas preparation atpH 6.5.

c Calculated as that portion of the total notavailable for growth by the assay organism priorto treatment with chicken-pancreas conjugase.

660 J . BACTERIOL .

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FOLIC ACID SYNTHESIS BY D. PNEUMONIAE

treated at pH 4.5 and 6.5 with extracts preparedby sonic treatment to determine the presence ofconjugase enzymes in cells of D. pneumoniae.Under these conditions, no increase in growth forassay organisms which could be due to conjugaseactivity was observed.

For a more detailed estimation of polyglu-tamyl-conjugated derivatives present intra-cellularly, microbiological assays employing L.casei in addition to S. faecalis and P. cerevisiaewere performed on several other samples. Dataobtained with a representative group of threesamples are shown in Table 2. Total folic acidvalues in both cases (i.e., with S. faecalis and L.casei) were nearly identical, indicating that un-conjugated folic acid compounds stimulatory toonly one of the two assay organisms (viz., pteroicacid, N'0-pteroic acid, prefolic A) do not accumu-late to a significant extent in cells of D. pneu-moniae. Of even greater interest are the valuesobtained in each case prior to treatment with thechicken-pancreas conjugase preparation. WithL. casei, the assay value is nearly ten times higherthan that obtained with S. faecalis. The pro-portion calculated as nonutilizable conjugatedmaterial in this case is considerably less andindicates the presence of large quantities of tri-glutamate derivatives utilized only by L. casei.Assay values derived with P. cerevisiae, show-

ing the proportion of free to nonutilizable conju-gates, were surprisingly like those obtained withS. faecalis. Although both L. casei and P. cere-visiae, but not S. faecalis, were reported (Hakalaand Welch, 1957) to utilize the triglutamatederivative of one-carbon substituted folate-H4,it seems that the above data reflect a greatervariation than that reported for the efficiency ofutilization of this conjugate among the threeorganisms and that smaller amounts of higherconjugates also exist, since an increase in L. caseiactivity was also obtained after chicken-pancreasconjugase treatment.We have attempted to differentiate chromato-

graphically among the various folic acid conju-gates. The results obtained with samples treatedas described above are shown in Fig. 1, 2, and 3.Bioautograms obtained with heat extracts re-vealed several zones of growth with all threeorganisms. These gave RF values of 0.3, 0.55, 0.7,0.8, and 0.9, comparable to those obtained withmono-, di-, and triglutamate derivatives of bothfolic acid (not stimulatory for P. cerevisiae) and

0.9

0.8

0.7

0.6

Rf 0.5

0.4

03 PGA

0.2

0.1Pteroic

0 acid

Controls HE HE HE4'CPe *IKe

FIG. 1. Bioautogram with Streptococcus faecalis.Controls, which were synthetic derivatives of folicacid, were spotted in the amount of 2.5 mAg each.They included pteroic acid, PGA (pteroylglutamicacid), PG3A (pteroyltriglutamic acid), N5-formyl-PGA -H4 (N5-formyltetrahydropteroylglutamic acid),N5-formyl-PG3A-H4 (N5-formyltetrahydropteroyl-triglutamic acid). Synthetic preparations of PG2A(pteroyldiglutamic acid) and N5-formyl-PGA2A-H4(N5-formyltetrahydropteroyldiglutamic acid) werenot available. However, RF values for these two com-pounds are shown in the figure and are those re-ported by Hakala and Welch (1957). Extracts werespotted in the amount of 0.01 ml. HE = heat extractsof cells, CPe = chicken-pancreas conjugase, HKe =hog-kidney conjugase.

N5-formylfolate-H4. In addition, a small zoneoccurred at an RF less than 0.1 with S. faecalis,which is probably pteroic acid. Distinct zones atRF 0.7 and 0.9 with L. casei were obtained onlywhen dilution of the sample was carried out tothe extent that other zones of growth did notappear. This confirmed findings shown above andagain indicated the presence of large quantitiesof triglutamate derivatives of both folic acid(RF 0.7) and a compound having an RF (0.9)identical to the triglutamate of N5-formylfolate-H4. All of the control compounds shown, whenstimulatory, gave equivalent growth zones, with

VOL. 85, 1963 661

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SIROTNAK, DONATI, AND HUTCHISON

the exception of the triglutamate of N5-formyl-folate-H4 for P. cerevisiae. Since syntheticpreparations of this compound were not availableas controls, cell-free extracts of S. faecalis wereincubated with the triglutamate of folic acid toconvert it to the formylated, reduced derivative,which could be used to provide a known referencezone when bioautographed. A distinct growthzone at RF 0.9 (characteristic of the triglutamateof N5-formylfolate-H4) appeared on plates seededwith L. casei. With the same amount of material,no zone appeared with S. faecalis, and only afaint zone was detectable with P. cerevisiae.Microbiological assays indicated that the tri-glutamate was required in at least 50 times theconcentration of the other derivatives to producethe same level of growth for the latter organism.This is in agreement with the findings of Silver-man and Wright (1956), but not with those ofHakala and Welch (1957). Samples of the syn-thetic diglutamate derivative of folic acid were

0. (5N5formyl-

PG2 A-H4

N5 formyl ___0.7

PGA H4

0.6

0.5Rf

0.4

0.3

0.2

0.1

0 .Controls HE HE HE

+CPe +HKe

FIG. 2. Bioautogram with Pediococcus cerevisiae.Controls were spotted in the amount of 2.5 mrAg andare the same as in Fig. 1. Samples were spotted in0.01-ml amounts. HE = heat extract of cells, CPe =chicken-pancreas conjugase, HKe = hog-kidneyconjugase.

Rf

FIG. 3. Bioautogram with Lactobacillus casei.Controls were spotted in the amount of 2.5 mpg andare the same as in Fig. 1. Samples were spotted inthe amount of 0.01 ml (set A) or 0.0025 ml (set B).HE = heat extract of cells, CPe = chicken-pancreasconjugase, HKe = hog-kidney conjugase.

not available for use as a known compound.Consequently, we have assumed that RF valuesfor this compound and the diglutamate of N5-formylfolate-H4 are the same (0.55 and 0.8, re-spectively) as those reported for this solventsystem (Hakala aDd Welch, 1957). Some con-firmation is also provided by the fact that theproducts of hydrolysis by chicken-pancreasconjugase, reported as diglutamate derivatives ofboth folic acid and N5-formylfolate-H4 (Da-browska, Kazenko, and Laskowski, 1949), haveRF values of 0.55 and 0.8, respectively, in thissystem. Prior treatment of heat extracts withchicken-pancreas conjugase resulted in a markedincrease in zones with these same RF values.Similar treatment with hog-kidney conjugase, onthe other hand, brought about an increase inzones at RF 0.3 and 0.7, corresponding to folicacid and N5-formyfolate-H4. In addition, whendigestion with the hog-kidney preparation hadbeen complete (usually requiring at least twosuccessive treatments), all other zones of growthdisappeared. This is an indication of the absencein the cell extracts of large amounts of suchderivatives as NI0-pteroic acid and N10-formyl-folate, which have RF values equal to certainconjugated derivatives of folic acid.When bioautographs were made with extracts

prepared by sonic treatment or from acetone-

662 J. BACTERIOL.

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FOLIC ACID SYNTHESIS BY D. PNEUMONIAE

TABLE 3. Folic acid content of cells grown in thepresence of added folate-H2 under conditionspermitting and not permitting biosynthesis

of folic acid

Cultural conditions Folic acid derivatives'(m/Ag/10 mg of dry weight)Sample

Drug' Folate-H2b No. +CPed Nonuti-addition lizablee

Il - - 59 184 12512 + + 95 111 16

II, -_ 123 374 251112 - + 610 1,020 410113 + + 312 327 15

a p-Aminobenzenesulfonamide (100 ,ug/ml) wasused.

b The amount of folate-H2 was 10 ,ug/ml.c Measured as equivalents of pteroylglutamic

acid by microbiological assay with S. faecalis.Average of duplicate determinations done atseveral sample dilutions.

d Assay values obtained after treatment ofsamples with chicken-pancreas preparation atpH 6.5.

e Calculated as that portion of the total notavailable for growth by the assay organism priorto treatment with chicken-pancreas conjugase.

dried cells, no differences were noted (particularlyin the number of zones of growth) when comparedwith bioautographs made with heated material.The two genetically distinct amethopterin-

resistant strains described above were examinedto determine any differences in gross accumula-tion of folic acid. Microbiological assay andbioautographic data on culture material derivedfrom these strains revealed no significant dif-ferences in the amount or type of compoundsformed by these mutants and cultures of the wildstrain.

In view of the rather large amount of folic acidaccumulating in culture material as polyglutamylderivatives, it was considered of interest todetermine, to some extent, the manner in whichthese compounds are synthesized during growth.An attempt was made to ascertain whetherpolyglutamate formation occurred at a bio-synthetic step preceding, or simultaneously with,the formation of folic acid (i.e., by the addition ofpreformed homopeptides) or by peptide synthe-sis after the formation of folic acid. Some informa-tion could be obtained by blocking the synthesisof folic acid with sulfanilamide but allowinggrowth by the addition of a reversing compound

to the medium. After a number of unsuccessfulattempts with folic acid, it was determined thatfolate-H2 was the derivative actually necessaryto bring about reversal of inhibition. Cells de-pleted of folic acid, by culturing them at a sub-inhibitory level of sulfanilamide, were inoculatedinto a medium containing (per ml) 100 ,l.g of thesame drug, 10 ,ig of folate-H2, and 1 mg ofascorbic acid. Growth under these conditionsproceeded at the same rate as that which oc-curred in the absence of the drug, whereas nogrowth occurred in controls containing only thedrug. Cells were harvested at the end of the expo-nential phase of growth, and heat extracts wereprepared in the usual fashion. The results areshown in Table 3 and Fig. 4. Under growth condi-tions where the total intracellular amount of folicacid accumulated was approximately the same inall cases (values for folic acid obtained with con-jugase-treated extracts), practically no polygluta-mates were found in cells grown during a blockin folic acid biosynthesis. That the relativelylarge amount of folate-H2 added did not have anonspecific physiological effect on polyglutamyl

Rf

0.9N5 formyl-PG2 A-H4

0.8

07N5 formyl-PGA

0.6

PG2 A0.5

0.4

0.3

0.1

0

FIG. 4. Bioautogram with Streptococcus faecalis.Controls were spotted in the amount of 2.5 mjAg andare the same as in Fig. 1. Samples were spotted in0.01-ml amounts. A = heat extract of cells, B =

heat extract of cells grown in the presence of sul-fanilamide (100 ,Ag/ml) and folate-H2 (10 jAg/ml),CPe = chicken-pancreas conjugase.

PGA

Pteroicacid

Controls A B A B-LCPe -CPe

VOL. 85, 1963 663

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SIROTNAK, DONATI, AND HUTCHISON

synthesis was shown by data obtained in thesecond experiment (1I2, Table 3). The amount ofpolyglutamates was not only equal to, butgreater than, that accumulated in the controlcells. Possibly an inhibition by sulfanilamide ofreactions combining additional glutamic acidresidues with folate-H2 in addition to those lead-ing to the synthesis of this intermediate, althoughsomewhat untenable, might also explain theseresults. However, under conditions permittingnear maximal growth, but where folate synthesiswas inhibited more than 95% by sulfanilamide,the proportion of utilizable to nonutilizable con-jugated folate accumulating remained unchangedfrom that normally obtained. When bioauto-grams were made with these extracts (Fig. 4), avery large zone of growth was seen which cor-responded to folic acid (RF 0.3) in contradistinc-tion to that obtained with control extracts. Inaddition, after conjugase treatment of extractsfrom cells grown in the presence of drug, onlyminute zones of growth appeared at RF valuesof 0.55 and 0.8, corresponding to end products ofconjugase digestion.

DISCUSSION

As reported for a number of other micro-organisms, D. pneumoniae during growth alsosynthesizes the majority of its folic acid as poly-glutamyl conjugates. The large proportion ofconjugated growth factor found intracellularly(as much as 95% conjugated to the extent ofthree or more glutamic acid residues), as com-pared with that found in the growth medium,could indicate that the conjugation process is ameans for maintaining the necessary intracellularconcentration of these important materials.These findings acquire additional significancewhen one considers the view expressed previouslyby Rabinowitz and Himes (1960) that folic acidderivatives occur naturally as polyglutamateswith a capacity for participation in one-carbontransfer reactions at least equal to, if not greaterthan, analogues containing one glutamic acidresidue. The relatively small accumulation ofsubstituted, reduced derivatives as comparedwith folic acid precursors could result from thepresence in the cell of a less active formylatingmechanism, similar to that reported for S. faecalis(Albrecht, Johnson, and Hutchison, 1962).

If we assume that the difference among assaydeterminations performed with each organism onextracts prior to conjugase treatment is due to

the inability of S. faecalis and P. cerevisiae toutilize triglutamates efficiently, then the bulk(about 60%) of the growth factors, both folicacid and substituted folate-H4, synthesized byD. pneumoniae are triglutamate derivatives. Inaddition to the 10% synthesized as mono- or di-glutamate derivatives (both utilized equally aswell by assay organisms), the remaining 30%probably occurs as a higher glutamate derivative,viz., hexa- or heptaglutamates described inyeast (Pfiffner et al., 1946) and in C. cylindro-sporum (Wright, 1955), unless the greater activityobtained with the conjugase-treated extracts isdue to a more efficient utilization of the hydrolysisproduct (diglutamate) by L. casei. Indications tothe contrary were obtained in reports (Hakalaand Welch, 1955, 1957) which show that tri-glutamates are utilized equally as well as lesserconjugates by L. casei. We have also confirmedthis by obtaining no increase in L. casei activityof solutions of the triglutamate of folic acid andthe enzymatically synthesized triglutamate formof N5-formyfolate-H4, after chicken-pancreasconjugase digestion. However, in view of theincrease in activity shown with conjugase-treatedextracts, we have attempted to distinguish bybioautography between triglutamates and higherglutamates using L. casei. Extracts, which werediluted to the extent that only the triglutamatederivatives would appear on bioautographicplates, were spotted on paper strips and run inthe aqueous phosphate buffer described aboveand in an ammonia-ethanol-butanol-water solvent(Doctor and Couch, 1953). The strips wereplaced on agar medium seeded with L. casei,some of which contained purified chicken-pancreas conjugase prepared by the method ofLaskowski, Mims, and Day (1945). Although theresults were not definitely conclusive, areas ofgrowth other than those normally seen at RF 0.7and 0.9 (triglutamyl derivatives) were definitelyvisible on the plates containing the chicken-pancreas conjugase.The reversal of sulfanilamide inhibition by

folate-H2 and not by folic acid is a most interest-ing observation. Assuming no permeabilityfactor, this finding could be interpreted to meanthat the normal pathway of synthesis, similar tothat reported for Escherichia coli by Brown et al.(1961), does not proceed via folic acid, but rathervia a reduced form of this compound.The apparent origin in D. pneumonie of folic

acid derivatives with additional glutamic acid

J. BACTERIOLF.664

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FOLIC ACID SYNTHESIS BY D. PNEUMONIAE

residues by the utilization of preformed homo-peptides, does not entirely rule out an increase inpeptide length at a later stage. The presence ofsmall amounts of polyglutamates in cells grownduring a block in folic acid synthesis could beindicative of the existence of such a mechanismin this organism. Furthermore, a similar mech-anism must exist in microorganisms, such as S.faecalis and L. casei (Hendlin et al., 1953), whichare unable to synthesize folic acid but formconsiderable amounts of folic acid polyglutamatesfrom added folic acid.

ACKNOWLEDGMENTS

The authors are indebted to Alberta M. Al-brecht for providing the cell-free preparations ofS. faecalis and for assistance in the preparation ofthis manuscript.

This work was supported in part by contractno. SA-43-PH-2445 from the Cancer Chemo-therapy National Service Center, NationalCancer Institute, National Institutes of Health,Bethesda, Md., and by grant T-107C from theAmerican Cancer Society.

LITERATURE CITED

ADAMS, M. H., AND A. S. ROE. 1945. A partiallydefined medium for cultivation of pneumo-

coccus. J. Bacteriol. 49:401-409.ALBRECHT, A. M., A. H. JOHNSON, AND D. J.

HUTCHISON. 1962. Studies on mechanisms ofresistance to antifolics in Streptococcusfaecalis. Bacteriol. Proc., p. 105.

BIRD, 0. D., B. BRESSLER, R. A. BROWN, C. J.CAMPBELL, AND A. D. EMMETT. 1945. Themicrobiological assay of vitamin Bc conju-gate. J. Biol. Chem. 159:631-636.

BIRD, 0. D., S. B. BINKLEY, E. S. BLOOM, A. D.EMMETT, AND J. J. PFIFFNER. 1946. On theenzymatic formation of vitamin Bc from itsconjugate. J. Biol. Chem. 157:413-414.

BLAKLEY, R. L. 1960. Crystalline dihydropteroyl-glutamic acid. Nature 188:231-232.

BRATTON, A. C., AND E. K. MARSHALL, JR. 1939.A new coupling component for sulfanilamidedetermination. J. Biol. Chem. 128:537-550.

BROWN, G. M., R. A. WEISMAN, AND D. A. MOL-NAR. 1961. The biosynthesis of folic acid. I.Substrate and cofactor requirements forenzymatic synthesis by cell-free extracts ofEscherichia coli. J. Biol. Chem. 236:2534-2543.

DABROWSKA, W.) A. KAZENKO, AND M. LASKOW-SKI. 1949. Concerning the specificity of chickenpancreas conjugate. Science 110:95.

DOCTOR, V. M., AND J. R. COUCH. 1953. Occurrenceand properties of a conjugated form of Leuco-nostoc citrovorum factor. J. Biol. Chem. 200:223-231.

FLYNN, L. M., V. B. WILLIAMS, B. L. O'DELL, ANDA. G. HOGAN. 1951. Medium for assay ofvitamins with lactic acid bacteria. Anal.Chem. 23:180-185.

FUTTERMAN, S. 1957. Enzymatic reduction of folicacid and dihydrofolic acid to tetrahydrofolicacid. J. Biol. Chem. 228:1031-1038.

HAKALA, M. T., AND A. D. WELCH. 1955. A conju-gated form of citrovorum factor synthesizedby Bacillus subtilis. Federation Proc. 14:222.

HAKALA, M. T., AND A. D. WELCH. 1957. A poly-glutamate form of citrovorum factor synthe-sized by Bacillus subtilis. J. Bacteriol. 73:35-41.

HENDLIN, D., L. K. KODITSCHEK, AND M. H.SOARS. 1953. Investigations of the biosynthesisof citrovorum factor by lactic acid bacteria.J. Bacteriol. 65:466-471.

LASKOWSKI, M., V. MIMS, AND P. L. DAY. 1945.Studies on the enzyme which produces theStreptococcus lactis R-stimulating factor frominactive precursor substance in yeast. J. Biol.Chem. 157:731-739.

PFIFFNER, J. J., D. G. CALKINS, E. S. BLOOM, ANDB. L. O'DELL. 1946. On the peptide nature ofvitamin B, conjugate from yeast. J. Am.Chem. Soc. 68:1392.

RABINOWITZ, J. C., AND R. H. HIMES. 1960. Folicacid coenzymes. Federation Proc. 19:963-970.

SAUBERLICH, H. E., AND C. A. BAUMANN. 1948. Afactor required for the growth of Leuconostoccitrovorum. J. Biol. Chem. 176:165-173.

SILVERMAN, M., AND B. E. WRIGHT. 1956. Micro-biological aspects of the diglutamyl deriva-tives of citrovorum factor and NI0-formyl-folic acid. J. Bacteriol. 72:373-377.

SIROTNAK, F. M., R. B. LUNT, AND D. J. HUTCHI-SON. 1960a. Genetic studies on amethopterinresistance in Diplococcus pneumoniae. J.Bacteriol. 80:648-653.

SIROTNAK, F. M., R. B. LUNT, AND D. J. HUTCHI-SON. 1960b. Transformation of resistance to8-azaguanine in Diplococcus pneumoniae.Nature 187:800-801.

SWENDSEID, M. E., AND J. F. NYC. 1958. Study offolic acid derivatives in Neurospora crassaand the recognition of a mutant resistant toaminopterin. J. Bacteriol. 75:654-659.

WRIGHT, B. E. 1955. Polyglutamyl pteridinecoenzymes. J. Am. Chem. Soc. 77:3930-3931.

ZAKRZEWSKI, S. F., AND C. A. NICHOL. 1955.Derivatives of pteroic and pteroylglutamicacids formed by Streptococcus faecalis. Federa-tion Proc. 14:311.

VOL. 85, 1963 665

on July 31, 2020 by guesthttp://jb.asm

.org/D

ownloaded from