Amino Acid Biosynthesis in the Spirochete Leptospira: - Journal of

JOURNAL OF BACTERIOLOGY, Jan. 1974, p. 203-211Copyright O 1974 American Society for Microbiology

Vol. 117, No. 1Printed in U.S.A.

Amino Acid Biosynthesis in the Spirochete Leptospira:Evidence for a Novel Pathway of Isoleucine Biosynthesis

NYLES W. CHARON,' RUSSELL C. JOHNSON, AND DAVID PETERSON

Department of Microbiology, University of Minnesota, Minneapolis, Minnesota 55455

Received for publication 7 August 1973

Radioactive carbon dioxide was incubated with growing cells of Leptospirainterrogans serotypes semaranga and tarassovi, and the specific activities anddistribution of the label within the cellular amino acids were determined. Theorigins of the carbon skeletons of all the acid-stable amino acids except isoleucinewere found to be consistent with known biosynthetic pathways for these aminoacids. Experiments using radioactive carbon dioxide and other tracers indicatedthat most of the isoleucine was synthesized by a pathway not involvingthreonine. The origin of the carbon skeleton of isoleucine consisted of tworesidues of pyruvate (carbons 2 and 3) and acetate of acetyl-coenzyme A bythis pathway. Isotope competition studies indicated that the pathway was

regulated by isoleucine. The results are discussed in relation to two proposedpathways of isoleucine biosynthesis involving citramalate as an intermediate.

Leptospira is an aerobic spirochete with sim-ple nutritional requirements. Although L. inter-rogans is the only recognized species of Lepto-spira, the genus is divided into two groups orcomplexes (31, 35). Leptospires of the parasiticcomplex are associated with infection, parasit-ism, and disease, whereas leptospires of thebiflexa complex are isolated from soil and waterand are generally not pathogenic (31, 35). Withfew exceptions, long-chain fatty acids serve asthe obligate major carbon and energy sourcesfor both groups of leptospires (12, 14). Exoge-nous fatty acids are required by the leptospiresfor membrane biosynthesis (15, 30), and evi-dence suggests that fatty acids are also metabo-lized via beta-oxidation and the tricarboxylicacid cycle (3, 10, 12, 15).

In the course of studies related to carbondioxide metabolism in Leptospira, [14C]O2 inthe form of [14C]NaHO, was fed to growingleptospires, and the specific activity of each ofthe amino acids recovered in the protein frac-tion was determined. This type of informationhas proven helpful in determining the origin ofthe carbon skeleton of many of the amino acidsin other organisms (25). As reported in thiscommunication, our tracer studies indicate thatrepresentatives of both complexes of Leptospirasynthesize isoleucine by a novel pathway. Theorigins of the carbon skeletons of the otheracid-stable amino acids are consistent with the

l Present address: Department of Biological Sciences,Stanford University, Stanford, Calif. 94305.

pathways common to those found in otherprocaryotes.

MATERIALS AND METHODSOrganisms. L. interrogans serotypes semaranga

Veldrat Semarang 173 and tarassovi Mitis-Johnsonwere obtained from A. D. Alexander, Walter ReedArmy Institute for Research, Washington, D.C. Al-though originally isolated from a rat (35), all evidencepoints to serotype semaranga as being a member ofthe biflexa complex (8, 13, 14). Serotype tarassovi is ofhuman origin and is characterized as belonging to theparasitic complex (35).Growth medium. A semi-defined medium consist-

ing of 0.5 mM palmitate and oleate, 2% chloroform-methanol-extracted bovine serum albumin, vitamins,and salts was used for these studies. The followingstock solutions in solution (grams per 100 ml) wereprepared and, except for FeSO4, stored at -20 C:NH4Cl, 25; ZnSO4 7H20, 0.4; CaCl2 2H20,MgCl2 6H20, 1 each; FeSO4 7H20, 0.5 (made fresh);CuSO4 5H20, 0.3; thiamine, 0.5; and vitamin B12,0.02. Sterile stocks of 10 mM sodium oleate (HormelInstitute, Austin, Minn.) and 0.3 mM CaCl2, MgCl2were also prepared.A number of steps were involved in preparation of

the medium. A lOx basal medium was made bycombining the following compounds, adjusting thepH to 7.4, and autoclaving: Na2HPO4, 1.0 g; KH2PO4,0.3 g; NaCl, 1.0 g; NH4Cl, 1.0-ml stock; thiamine,1.0-ml stock; and distilled water, 100 ml. The al-bumin supplement was prepared by dissolving 40 g ofchloroform-methanol-extracted albumin as preparedby the procedure of Johnson et al. (15) in 100 ml ofdistilled water and heating it at 50 C for 1 to 3 h. Thefollowing stock solutions (in milliliters) were slowly

203

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

CHARON, JOHNSON, AND PETERSON

added to the cooled albumin solution while it wasbeing stirred with a magnetic stirring bar: CaCl2,MgCl2, 2.0; ZnSO4, 2.0; FeSO4, 20; CuSO4, 0.2; andvitamin B12, 2.0. The final volume of the albuminsupplement was made up to 200 ml with distilledwater, and the pH was adjusted to 7.40. Sterilizationwas accomplished by pressure filtration.The complete medium was prepared in 220-ml

volumes. A suspension of 160 ml of water containing0.1 mmol of sodium palmitate (Hormel Institute,Austin, Minn.) was autoclaved in a 200-ml borosili-cate glass bottle. When the sterile palmitate wascooled to 50 C, 10 ml of warm (50 C) oleate stocksolution and 20 ml of warm albumin supplement wereadded. After 1 h at 50 C, the solution was cooled to23 C, and 20 ml of 10x basal medium and 10 ml ofsterile stock CaCl2, MgCl2 were added. No ninhydrin-positive migrating material in methanol: pyridine: -water (20:5:1) was detectable in this medium. Aslittle as 1 Ag of free isoleucine or leucine per ml isdetectable by this assay.Maintenance and growth of cultures. Cultures

were maintained in glass tubes with stainless steelclosures on Leighton tube racks at 30 C. Growth wasmonitored by using these tubes and nephelo flasks ona model 9 Coleman photonephelometer. Readings onthe nephelometer were correlated to total cell count(T. Auran, M.S. thesis, Univ. of Minn. Minneapolis,1968) and to viable counts. Viable counts were ob-tained from plates incubated in an atmosphere of2.5% CO2 and 97.5% air at 30 C.

Chemicals. L-threonine-U-1'C, DL-aspartic acid-i-"C, DL-aspartic acid-4-"C, sodium pyruvate-l-"C,L-glutamic acid-U-14C, glycine-2-"4C, and toluene-U-"4C were purchased from New England Nuclear Corp.,Boston, Mass. Sodium pyruvate-3-14C and[14C]sodium bicarbonate were purchased -from Amer-sham Searle, Chicago, Ill. Palmitate-U-_4C was ob-tained from Applied Science Laboratories, State Col-lege, Pa. Nonradioactive compounds used for isotopecompetition studies included DL-citramalate (AldrichChemical Co., Inc., Milwaukee, Wis.), mesaconate (K& K Laboratories, Inc., Plainview, N.Y.), #l-DL-methylaspartate (Sigma, St. Louis, Mo.), and L-isoleucine (General Biochemicals, Chagrin Falls,Ohio).

Preparation of labeled leptospires. A generalprocedure was used to obtain "4C-labeled leptospires.Thirty milliliters of an exponentially growing culturewere shaken at 30 C in a water-bath shaker. Thedoubling time of serotype semaranga was 4 to 6 h, andthat of serotype tarassovi was 12 to 16 h under theseconditions. Sterile radioactive compounds were addedto the culture at a turbidity reading of 75 (2.5 x 108cells/ml). In experiments with ["4C]NaHO,, flaskswere closed with rubber stoppers, and the[14C]NaHO, was injected into the flask by using asyringe. To obtain leptospires labeled by palmitate-U-14C, leptospires were grown to a turbidity of 120 (4.0x 108 cells/ml), centrifuged, and resuspended inprewarmed medium containing palmitate-U-_4C andone-fifth the normal concentration of oleate andpalmitate. In isotope competition studies, the non-radioactive compounds were added during the expo-

nential phase of growth before the addition of theradioactive materials. Labeled leptospires were har-vested at a turbidity reading of 150 (5.0 x 108cells/ml) by centrifugation at 4 C, 17,500 x g for 30min. The pelleted leptospires were washed three timesby repeatedly suspending them in 15 ml of cold lxbasal medium and centrifuging at 4 C.

Fractionation. A procedure similar to that ofRoberts et al. was used to fractionate the radioactiveleptospires (25). The four fractions isolated includedthe cold trichloroacetic acid fraction (metabolic inter-mediates), ethanol-ether fraction (lipid), hot trichlo-roacetic acid supernatant fraction (nucleotides), andthe hot trichloroacetic acid residue fraction (proteinand cell wall). Samples for radioactive determinationwere taken throughout each fractionation analysisand were counted in the solution of Bray (6) in aPackard Tri-Carb scintillation spectrometer.

Analysis of "C-labeled amino acids. A proceduresimilar to that of Piez was used to determine thespecific activity (dpm/nmol) of the "4C-labeled aminoacids (23). Amino acids obtained by acid hydrolysis ofthe residue fractions were separated with a Beckmanmodel 120B amino acid analyzer by using a Beckman4-h, two-column procedure (4) slowed down to abuffer flow rate of 40 ml/h and a ninhydrin flow rate of20 ml/h. The effluent from the columns was passedthrough a 1-ml flow-through cuvette containing an-thracene crystals before reacting with ninhydrin. Thiscuvette was placed in a Beckman LS-100 scintillationcounter with both a strip chart recording and aprintout of each minute of counts collected to deter-mine the total radioactivity for each amino acid. Thecounting efficiency of this coupled system was 44% asdetermined with amino acids of known specific activ-ity. The purity of purchased "4C-labeled amino acidswas also checked by this procedure. In decarboxyla-tion experiments, a single-column resin system wasused to separate the amino acids (Durrum ChemicalCorp., Palo Alto, Calif.).Amino acid decarboxylation. "C-labeled amino

acids were quantitatively decarboxylated with the aidof the amino acid analyzer. The protein hydrolyzatewas analyzed as before, except 4-ml fractions werecollected from the effluent fluid after reaction withninhydrin. This fluid contains the ninhydrin aldehydereaction products of the amino acids (21). An eventmarker marked the chromatography chart with eachchange of tubes to help identify which amino acidreaction products were in each tube. To assay thisfluid for radioactivity, 1 ml of each fraction was addedto 5 ml of Triton X-100 scintillation solution (5). Oneliter of the scintillation solution contained 500 ml oftoluene, 500 ml of Triton X-100 (Packard InstrumentCo., Inc., Downers Grove, Ill.), 8.0 g of 2,5-diphenyloxazole, 0.1 g of 1, 4-bis-2-(5-phenyloxazolyl)-benzene. The counting efficiency was determined tobe 79% by internal "C-t9luene standards. By compar-ing the specific activity of each amino acid before andafter decarboxylation, the percentage of label residingin- the C-1 carbon was obtained. Control experimentswith amino acids labeled in specific positions con-firmed the validity of this method. For example, only2% of the radioactivity of aspartate-1-"4C was recover-

204 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

AMINO ACID BIOSYNTHESIS IN LEPTOSPIRA

able in the effluent fractions. In the case of glycine-2-"C, essentially all of the radioactivity was accountedfor in the effluent fractions.

RESULTS[14C]02-labeling pattern in leptospiral

amino acids. Leptospires of serotypes sema-ranga and tarassovi were fed [14C]02 for one

generation and fractionated according to theprocedure of Roberts et al. (25). The radioactiv-ity was found to be distributed among all of thefractions, including the hot trichloroacetic acidresidue fraction. This fraction was hydrolyzedwith 6 N HCl, and the specific activities of theamino acids were determined. The results of oneexperiment with serotype semaranga are pre-

sented in Table 1. Trace amounts of methio-nine, histidine, diaminopimelate, glucosamine,and muramic acid were also detected and were

radioactive. As discussed below in detail forserotype semaranga, the data indicate that allof the amino acids listed except isoleucine are

synthesized via pathways common to those ofother organisms (see 18 and 32 for reviews).Because serotype tarassovi had a similar['4C]02-labeling pattern with respect to relativespecific activities (Table 2), the discussion isalso relevant for this serotype.

TABLE 1. Specific activity and decarboxylation ofamino acids obtained from ['4C]02-labeled serotype

semarangaa

Sp act Sp actbefore after Fraction

decarbox- decarbox- of labelAmino acid ylation ylation residing

(dpm/nmol (dpm/nmol in C-1of amino of amino position

acid) acid)

Alanine.357 357 4 0.99Serine.349 349 15 0.96Glycine.357 357 26 0.93Valine ........... 343 2E 0.92Leucine .......... .0 17

Glutamate 291 291 8 0.97Proline .332 332 3 0.99Arginine .1137 1,137 786 0.31

Tyrosine ......... 949 670 0.29Phenylalanine ... 909 602 0.33

Aspartate ........ 649 15 0.90bLysine.645 645 306 0.53Threonine 623 623 295 0.53Isoleucine 16 16 17

aThe culture contained 30 nmol of added[14C]NaHO, per ml (specific activity, 57 nCi/nmol).

Both the C-1 and the C-4 of aspartate are decar-boxylated by ninhydrin.

TABLE 2. Specific activity of amino acids obtainedfrom ["C] O,-labeled serotype tarassovia

Sp actAmino acid (dpm/nmol of

amirno acid)

Alanine ........ 1,148Serine ........ 1,055Glycine ........ 1,207Valine ........ 1,225Leucine ........ 0

Glutamate ........ 807Proline........ 945Arginine ........ 3,693

Tyrosine ........ 3,014Phenylalanine .. ...... 3,161

Aspartate ........ 1,984Lysine ........ 2,034Threonine ..... 1,893Isoleucine..... 23

a The culture contained 87 nmol of added["4C]NaHO, per ml (specific activity, 57 nCi/nmol).

The amino acids are grouped according tofamilies similar to those presented by Roberts etal. for Escherichia coli. Members within afamily were found to share a common precursorin E. coli (25). One reflection of having acommon precursor is possessing a similar spe-cific activity. Alanine, serine, glycine, and va-line had approximately equal specific activities(Table 1). Decarboxylation of these amino acidswith ninhydrin demonstrated that greater than90% of the radioactivity resided in the C-1position of these amino acids (Table 1, columns3 and 4). Pyruvate and phosphoenolpyruvateare the major precursors to this group of aminoacids in other organisms (18, 25, 32). Theseintermediates exclusively labeled in the C-1position would yield this labeling pattern inalanine, serine, and glycine. The existence ofradioactivity in the C-1 position of valine butnot in any of the leucine carbons is in agreementwith pyruvate being exclusively labeled in theC-1 position, and also with the branch pathwaysof leucine and valine biosynthesis found in otherorganisms. According to these pathways, thecarbon derived from the C-1 position of pyru-vate in the common intermediate a-ketoisoval-erate is lost on the leucine branch but retainedon the valine branch (18, 32).The specific activities and distribution of the

label in glutamate, proline, and arginine sug-gest that this family of amino acids is made byknown pathways. Glutamate is a precursor toproline in other organisms (18, 25, 32). Both of

N-OL. 117, 1974 205

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

CHARON, JOHNSON, AND PETERSON

these amino acids had approximately the samespecific activity and were exclusively labeled inthe C-1 position (Table 1). Arginine had a highspecific activity, and approximately one-thirdof the label resided in the C-1 position. Theseresults are consistent with the carbamyl-phos-phate-ornithine pathways found in other orga-nisms (18, 32). An almost identical labelingpattern in this family of amino acids was foundwith ["C]02-labeled E. coli (25).

Tyrosine and phenylalanine had similar spe-cific activities, and approximately one-third ofthe label resided in the C-1 carbons (Table 1).Because phosphoenolpyruvate was deduced tobe labeled in the C-1 position by ["C]02, theradioactivity in the C-1 position of these aminoacids agrees with the common aromatic acidpathways. The C-1 of these amino acids isderived from the C-1 of phosphoenolpyruvate inother organisms (18, 32).

Aspartate, lysine, and threonine had approxi-mately equal specific activities. Greater than90% of the label resided in the carboxyl groupsof aspartate, and 53% resided in the C-1 positionof threonine and lysine. These results are inagreement with the aspartate semialdehydepathway of threonine biosynthesis and the dia-minopimelate pathway of lysine biosynthesis(18, 32). Aspartate, lysine, and threonine werefound to have equal specific activities in E. colifed [14C]02 (25).The data reported in Table 1 are not consist-

ent with the threonine dehydratase pathway ofisoleucine biosynthesis. Four of the carbons ofisoleucine are derived from threonine via thispathway (18, 25, 32). Thus, in a number oforganisms fed ["4C]02, the specific activity ofisoleucine was found to be approximately thesame as that of threonine (19, 25). The specificactivity of isoleucine was 3% of the specificactivity of threonine (Table 1). Additional ex-periments were carried out in attempting toaccount for these results.Aspartate-4-"4C- and threonine-U-"4C-la -

beling patterns. To further test whether threo-nine is a major precursor to isoleucine in sero-type semaranga, aspartate-4-"4C- and threo-nine-U-'4C-labeled leptospires were analyzed.Other organisms chiefly synthesizing isoleucinevia the threonine dehydratase pathway haveapproximately the same specific activity in cel-lular threonine and isoleucine when fed thesetracers (18, 25). The specific activity of iso-leucine was 4% of the specific activity of threo-nine in aspartate-4- "4C-labeled leptospires(Table 3). Only threonine and isoleucine were la-beled in the threonine-U-"4C experiment, andthe specific activity of isoleucine was 17% of thespecific activity of threonine. Because the spe-

TABLE 3. Specific activity of amino acids obtainedfrom aspartate-4-"4C-, threonine-U-I"C-, andpalmitate-U-14C-labeled serotype semarangaa

Radioactive precursor

Aspartate- Threonine- Palmitate-4-14C U-14C U-14C

Amino acid Sp act perSp act Sp act carbon

(dpm/nmol (dpm/nmol atomof amino of amino amino acidacid) acid) (dpm/C

atom)b

Alanine .......... 11 0 52Serine ........... 16 0 61Glycine .......... 11 0 57Valine ........... 14 0 67Leucine .......... 0 0 59

Glutamate ....... 20 0 61Proline .......... 18 0 48Arginine ......... 27 0 68

Tyrosine ......... 34 0 55Phenylalanine 32 0 68

Aspartate ........ 123 0 55Lysine ........... 123 0 52Threonine ....... 116 41 50Isoleucine ........ 5 7 66

aThe cultures contained 290 nmol of aspartate-4-"4C per ml (specific activity, 2.1 nCi/nmol), 6 nmol ofthreonine-U-"4C per ml (specific activity, 164 nCi/nmol), and 100 nmol of palmitate-U-'4C per ml(specific activity, 4.4 nCi/nmol).

b Mean specific activity, 58 dpm/C atom.

cific activity of isoleucine was considerably lessthan that of threonine in both experiments, thethreonine dehydratase pathway is not the chiefsource of cellular isoleucine. However, the find-ing of a trace amount of radioactivity in isoleu-cine in the threonine-U-"C experiment suggeststhat this pathway operates to a minor extent.Palmitate-U-"4C-labeling pattern. Two

possible explanations could account for the lowspecific activity found in isoleucine relative tothat of threonine in the above experiments. Theleptospires could synthesize most of the isoleu-cine via a pathway not involving threonine. Onthe other hand, isoleucine could conceivably bepreferentially assimilated from the albuminpresent in the medium. To test these possibili-ties, leptospires of serotype semaranga were fedpalmitate-U- 4C. If the leptospires were prefer-entially acquiring isoleucine from the medium,then isoleucine should have a significantly lowerspecific activity on a per carbon basis than theother amino acids. The specific activity percarbon atom of isoleucine approximated the

206 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

AMINO ACID BIOSYNTHESIS IN LEPTOSPIRA

mean specific activity per carbon atom of theother amino acids (Table 3). These resultssupport the first hypothesis and indicate thatmost of the isoleucine is synthesized by apathway not involving threonine.A number of possible pathways were tested in

determining the origin of the carbon skeleton ofisoleucine in serotypes semaranga and tarassovi(N. Charon, Ph.D. thesis, Univ. of Minn.,Minneapolis, 1972). This was done by feedingleptospires postulated radioactive precursors ofisoleucine and analyzing the relative specificactivities of the amino acids. Two of theseproposals were supported by subsequent experi-ments and will be discussed here. In the firstproposal (Fig. la), L(+)-citramalate, mesacon-ate, and beta-methylaspartate are possible in-termediates for isoleucine biosynthesis. Thispathway has been suggested for Acetobactersuboxydans (R. T. Belly, S. Greenfield, and G.W. Claus, Bacteriol. Proc., p. 141, 1970). Reac-tions similar to those found for the leucinepathway could also yield alpha-ketobutyrate forisoleucine biosynthesis (Fig. lb). Thus, a con-densation of acetyl-S-coenzyme A (CoA) withpyruvate could yield D(-)-citramalate, citra-conate, and eventually alpha-ketobutyrate.There is some evidence to suggest the synthesisof alpha-ketobutyrate by this series of reactionsin certain organisms (9, 11, 20, 29). The carbonskeleton of isoleucine consists of two C-2's andC-3's of pyruvate and acetate of acetyl-S-CoAby both pathways.

a

COO-

C=O

CH3

Pyruvate

+

ICH3C-0

S-CoA

Acetyl-S- CoA

b

COO-HO-C-CH3H-C-HCOO-

LO Citramablite

I ~~~~~~~~~CH3COO CH3 3

13 ~~H-C-HH3C-C-H C2 H-C-H active aldehyde -I

COO- COOI H-C-NH3COO-

COO-

H3C-C-OHH-C-H

COO-

D(-) Citramalate

H3C\ 0Coo-

C

C

H COO~

eCitraconate

3-Methyloxalacetate <- Ketobutyrate

COO-

.- HjC-C-HHO-C-H

COOD.rythro- t-Methyl-D-malate

L- Isoleucine

FIG. 1. Proposed pathways of isoleucine biosynthesis.

Pyruvate-3-'4C-, pyruvate-l-'4C-, and glut-amate-U-14C-labeling patterns. To test thepathways proposed in Fig. 1, the residue frompyruvate-3-'4C- and pyruvate-1-14C-labeledleptospires were analyzed. The ratio of thespecific activity of a given amino acid to that ofalanine was used as an index of the number ofpyruvate 3 or 1 carbons contributing to thecarbon skeleton of that amino acid. One C-3 ofpyruvate contributes to the structure of isoleu-cine in the threonine dehydratase pathway,whereas two contribute via the pathways pro-posed in Fig. 1. The results of the pyruvate-3-"C and pyruvate-1-_4C experiments are pre-sented in Table 4. In both experiments, thelisted amino acids accounted for greater than90% of the total radioactivity incorporated intoall of the acid-stable amino acids. As can beseen for pyruvate-3-" C-labeled leptospires, thespecific activity of lysine was approximatelyequal to that of alanine, and the specific activi-ties of leucine and valine were twice that ofalanine. These results are consistent withknown pathways of these amino acids (18, 32).The specific activity of isoleucine was approxi-mately twice that of alanine. These data sup-port the pathways proposed in Fig. 1. Almostidentical results with pyruvate-3-14C were foundwith serotype tarassovi.The results of the pyruvate-1-1 4-C analysis are

in agreement with the conclusions drawn fromthe ['4C]02 experiment. These results indicatedthat the C-1 of pyruvate does not contribute to

OOC\/CH3 coo-C

> H3C-¢-H/ \COO

_C-NHH C00 c-L

Mesaconate threo-(3-Methyl-L-aspartate

VOL. 117, 1974 207

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

CHARON, JOHNSON, AND PETERSON

the carbon skeleton of the isoleucine. As can beseen from Table 4 for pyruvate-1-"4C-labeledleptospires, the specific activities of valine andalanine were approximately equal, and thespecific activity of lysine was 40% that ofalanine. Leucine had a specific activity lessthan 1% that of alanine. These results reflectthe known pathways of these amino acids (18,32). The specific activity of isoleucine was

approximately 1% that of alanine, indicatingthat the C-1 of pyruvate does not appreciablycontribute to the carbon skeleton of isoleucine.Thus, apparently two C-3's but no C-l's ofpyruvate contribute to the carbon skeleton ofisoleucine.An isotope competition-type experiment was

done to test whether two of the carbons ofisoleucine are derived from acetyl-S-CoA as

proposed in Fig. 1. Serotype semaranga was

simultaneously fed glutamate-U-'4C and non-

radioactive fatty acids and analyzed as before.Studies done by Johnson and Rogers (16) haveshown that Leptospira incorporates ['4C]gluta-mate into all of the cellular fractions and many

amino acids. Because fatty acids are metabo-lized by beta-oxidation in these organisms (10,15), acetyl-S-CoA should have a relatively lowspecific activity in glutamate-U-_4C-labeledleptospires. Leucine is the only acid-stableamino acid made by conventional pathwayswhich directly incorporates acetate from acet-yl-S-CoA into its carbon structure (18, 25, 32).As a consequence, the specific activity per

carbon atom of leucine should be less than thatof the other amino acids. If acetate from acet-yl-S-CoA is incorporated into isoleucine as pro-

posed in Fig. 1, isoleucine should also have a

relatively low specific activity per carbon atom.

TABLE 4. Specific activity of amino acids obtainedfrom pyruvate-3-14 C- and pyruvate-l -"C-labeled

serotype semarangaa

Radioactive precursor

Pyruvate-3-"C Pyruvate-I-14C

Amino acid Sp act Sp act(dpm/ Sp act/ (dpm/ Sp act/nmol of alanine nmol of alanineamino sp act amino sp actacid) acid)

Alanine .... 1,016 1.0 1,961 1.0Valine ... .. 1,995 2.0 1,991 1.1Leucine ... 2,015 2.0 5 < 0.01Lysine ..... 1,091 1.0 875 0.4Isoleucine 2,141 2.1 27 0.01

a The cultures contained 57 nmol of pyruvate-3-'4Cper ml (specific activity, 14.4 nCi/nmol) and 160 nmolof pyruvate-1- "C per ml (specific activity, 5.3 nCi/nmol).

The results of the experiment are presented inTable 5. Glutamate, proline, and arginine hadthe highest specific activities, indicating thatthe external glutamate was preferentially as-

similated into this family of amino acids.Among the others, isoleucine and leucine had a

significantly lower specific activity on a percarbon basis (t test, P < 0.001). These dataindicate that acetate contributes to the carbonskeleton of isoleucine as proposed in Fig. 1.Thus, two molecules of pyruvate and one mole-cule of acetate contribute to the carbon skeletonof isoleucine in Leptospira.

Isotope competition experiments. Otherwhole-cell tracer studies were not successful indetermining the precise intermediates in the iso-leucine pathways in both serotypes. These ex-

periments included isotope competition studieswith pyruvate-3-"C and some of the nonradi-oactive intermediates proposed in Fig. 1. Exoge-nous DL-citramalate, mesaconate, and f3-DL-methylaspartate did not dilute the label inisoleucine (Table 6). The inability of these com-

pounds to specifically dilute the label in iso-leucine could be due to reasons other thanthese compounds not being on the pathway(e.g., transport; see 32 for discussion). Theseresults suggest that a genetic biochemical ap-proach will be necessary to determine the path-way. We note from Table 6 that exogenousisoleucine inhibited the radioactivity from py-

ruvated-3-14C from being incorporated into iso-

TABLE 5. Specific activity per carbon atom of aminoacids obtained from glutamate-U-'4C-labeled

serotype semarangaa

Sp act percarbon atom

Amino acid of amino acid(dpm/Catom)

Glutamate .............................. 434Proline ................................ 451Arginine ................................ 393

Alanine ................................ 100Serine .................................. 86Glycine ................................ 85Valine ................................ 75Tyrosine.82Phenylalanine ........................... 84Aspartate ............................... 86Lysine .................................. 84Threonine ............................... 91

Leucine ................................ 50Isoleucine ............................... 55

a The culture contained 17 nmol of glutamate-U-4C per ml (specific activity, 47 nCi/nmol).

208 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

AMINO ACID BIOSYNTHESIS IN LEPTOSPIRA

Alanine Isoleucine Isoleucinesp act sp act sp act/(dpm/ (dpm/ alaninenmol) nmol) Sp act

2,0152,7862,332

2,2772,293

4,5806,1435,443

5,168111

2.32.22.3

2.30.05

a The cultures contained 5 pmol of the competitorcompound per ml and the same pyruvate-3-'4C con-

centration as in Table 5.

leucine by 98%. Severe inhibition of this typehas been shown to be due to feedback inhibitionalong an amino acid biosynthetic pathway inother organisms (32, 33). These results imply a

regulation of the isoleucine pathway by isoleu-cine in Leptospira.

DISCUSSIONThe experiments reported in this communica-

tion were performed on L. interrogans serotypessemaranga and tarassovi. Because these sero-

types are respective members of the biflexa andparasitic complexes, conclusions drawn here arelikely to be relevant to other serotypes of bothcomplexes. As pointed out in the Results, the["4C]2-labeling patterns in the serotypes stud-ied were similar in many respects to that foundfor E. coli (25). Thus, the relative specificactivities and distribution of the label in aspar-tate, threonine, lysine, glutamate, proline, andarginine were almost identical in the two spe-cies. Because this labeling pattern in theseamino acids has been shown to be related to theoperation of a tricarboxylic acid cycle in anumber of organisms, including E. coli (25), theresults reported here suggest that presence ofthis cycle in Leptospira. Enzymological andfatty acid oxidation studies have previouslyindicated the existence of this cycle in Lepto-spira (3, 12).The phosphoenolpyruvate and aromatic fam-

ilies of amino acids were extensively labeled inboth serotypes of Leptospira by ["4C]O,, butonly traces of label were found in these aminoacids in E. coli fed [14C]O2 (25). This differencecan reasonably be accounted for by an expectedanaplerotic- and gluconeogenic-type metabo-lism of Leptospira growing on fatty acids com-

pared with a glucose catabolic-type metabolismof E. coli growing on glucose. The anaplerotic

reactions would result in a randomization of thelabel between the C-1 and the C-4 of oxalace-tate after an initial carboxylation into the C-4position, and the gluconeogenic reactions wouldgenerate phosphoenolpyruvate labeled in theC-1 position. This scheme is consistent withaspartate, threonine, and lysine having approxi-mately twice the specific activity of the phos-phoenolpyruvate and glutamate families ofamino acids, and also the distribution of thelabel within these amino acids.The origins of the carbon skeletons of all of

the acid-stable amino acids except isoleucineare in agreement with known pathways of theseamino acids. This was shown by using a varietyof tracers including ["C]O, aspartate-4-" C,pyruvate-3-14C, pyruvate-1-4C, and glutamate-U-14C. The combined results of the ["4C]02 and[14C]pyruvate experiments are consistent withthe in vivo operation of pyruvate kinase. Thus,serine, glycine, and the aromatic amino acidswere extensively labeled in [14C]02-labeled lep-tospires, but only traces of radioactivity werefound in these amino acids in cells labeled withpyruvate-1-14C and pyruvate-3-14C. It should bementioned that pyruvate kinase has been re-ported to be absent in crude extracts of Lepto-spira (3). Accordingly, an enzyme assay forpyruvate kinase was conducted in our labora-tory, and activity was detected in crude extractsof serotype semaranga (W. F. Touminen and N.Charon, unpublished results; N. Charon, Ph.D.thesis, Univ. of Minn., Minneapolis, 1972).The specific activity of isoleucine relative to

the specific activity of threonine was unexpect-edly low in [14C]02-labeled leptospires. Otherorganisms synthesizing isoleucine via the threo-nine dehydratase pathway have the same spe-cific activity in isoleucine as in threonine (19,25). Exceptions to the threonine dehydratasepathway have been reported for a number oforganisms. These include rumen bacteria (car-boxylation of 2-methylbutyrate [26]), E. coli fedbeta-methylaspartate (beta-methylaspartasepathway [1]), E. coli Crooks strain (glutamatemutase, beta-methylaspartase pathway [22]),and Acetabacter suboxydans (R. T. Belly, S.Greenfield, and G. W. Claus, Bacteriol. Proc.,p. 141, 1970; Fig. la). This pathway is a combi-nation of reactions drawn from the glutamatefermentation of Clostridium tetanomorphum(2) and the beta-methylaspartase pathwayfound in E. coli (1).A pathway similar to that for leucine biosyn-

thesis is proposed and diagramed in Fig. lb. Infact, some of the leucine enzymes in otherorganisms carry out these analogous reactionsin vitro (17, 20, 24, 27-29, 34) and in some

TABLE 6. Specific activity of alanine and isoleucineobtained from serotype semaranga grown in thepresence of possible isoleucine intermediates and

pyruvate-3-14C4

None ..............DL-Citramalate ....

Mesaconate .......s-DL-Methylaspar-

tate .............L-Isoleucine .......

VOL. 117, 1974 209

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

CHARON, JOHNSON, AND PETERSON

instances possibly in vivo (9, 11, 20, 27-29).Direct evidence has not been found for theseenzymes or this sequence of reactions playing arole in isoleucine biosynthesis in other orga-nisms. However, isopropylmalate synthase, thefirst enzyme on the leucine pathway, is inhib-ited by both leucine and isoleucine in Sal-monella (7), Pseudomonas (24), and Sac-charomyces (28). The basis for the isoleucineinhibition is not clear, but one possibility is thatthe leucine enzymes could have functioned forboth isoleucine and leucine biosynthesis in theevolution of these organisms.The radiotracer studies reported here demon-

strate that, although the threonine dehydratasepathway functions to a minor extent in Lepto-spira, another apparently regulated pathway isresponsible for most of the cellular isoleucine.This pathway was found to incorporate twoC-3's of pyruvate and acetate of acetyl-S-CoAinto the carbon skeleton of isoleucine. Becausethe radioactive pyruvate experiments indicatedthat pyruvate is metabolized as a unit or as aC-2 and C-3 fragment, the other two carbons ofisoleucine are expected to be derlved frompyruvate C-2. Although the precise intermedi-ates in the pathway are not yet known, theseresults are in agreement with the pathwaysoutlined in Fig. 1.

ACKNOWLEDGMENTSWe appreciate the encouragement and suggestions made

by T. Auran, R. Bernlohr, P. Chapman, S. Dagley, and P.Rogers.

This investigation was supported by Public Health Servicegrant A106589 from the National Institute of Allergy andInfectious Disease. This work was presented in part at theAnnual American Leptospirosis Research Conferences held inChicago, Dec. 2-3, 1970 and Dec. 1-2, 1971.

LITERATURE CITED

1. Abramsky, T., and D. Shemin. 1965. The formation ofisoleucine from ,-methylaspartic acid in Escherichiacoli W. J. Biol. Chem. 240:2971-2975.

2. Barker, H. A., V. Rooze, F. Suzuki, and A. A. lodice.1964. The glutamate mutase system. Assays and prop-erties. J. Biol. Chem. 239:3260-3266.

3. Baseman, J. B., and C. D. Cox. 1969. Intermediate energymetabolism of Leptospira. J. Bacteriol. 97:992-1000.

4. Benson, J. V., Jr., and J. A. Patterson. 1965. Acceleratedautomatic chromatographic analysis of amino acids ona spherical resin. Anal. Chem. 37:1108-1110.

5. Bransome, E. D., Jr. 1970. The current status of liquidscintillation counting. Grune & Stratton, Inc., NewYork.

6. Bray, G. A. 1960. A simple efficient liquid scintillator forcounting aqueous solutions in a liquid scintillationcounter. Anal. Biochem. 1:279-285.

7. Calvo, J. M., M. Freundlich, and H. E. Umbarger. 1969.Regulation of branched-chain amino acid biosynthesisin Salmonella typhimurium: isolation of regulatorymutants. J. Bacteriol. 97:1272-1282.

8. Cinco, M., and N. Petalin. 1970. Serogroups and sero-types in water-Leptospira strains. Trop. Georgr. Med.22:237-244.

9. Guymon, J. F., J. L. Ingraham, and E. A. Crowell. 1961.

The formation of n-propyl alcohol by Saccharomycescerevisiae. Arch. Biochem. Biophys. 95:163-168.

10. Henneberry, R. C., and C. D. Cox. 1970. #-Oxidation offatty acids by Leptospira. Can. J. Microbiol. 16:41-45.

11. Ingraham, J. L., J. F. Guymon, and E. A. Crowell. 1961.The pathway of formation of n-butyl and n-amylalcohols by a mutant strain of Saccharomyces cerevis-iae. Arch. Biochem. Biophys. 95:169-175.

12. Johnson, R. C., and N. D. Gary. 1962. Nutrition ofLeptospira pomona. II. Fatty acid requirements. J.Bacteriol. 85:976-982.

13. Johnson, R. C., and V. G. Harris. 1967. Differentiation ofpathogenic and saprophytic leptospires. I. Growth atlow temperature. J. Bacteriol. 94:27-31.

14. Johnson, R. C., V. G. Harris, and J. K. Walby. 1969.Characterization of leptospires according to their fattyacid requirements. J. Gen. Microbiol. 55:399-407.

15. Johnson, R. C., B. P. Livermore, J. K. Walby, and H. M.Jenkins. 1970. Lipids of parasitic and saprophyticleptospires. Infect. Immunity 2:286-291.

16. Johnson, R. C., and P. Rogers. 1964. Metabolism ofleptospirae. I. Utilization of amino acids and purineand pyrimidine bases. Arch. Biochem. Biophys.107:459-470.

17. Kohlhaw, T. R., and T. R. Leary. 1969. a-Isopropylma-late synthase from Salmonella typhimurium. Purifica-tion and properties. J. Biol. Chem. 244:2218-2225.

18. Meister, A. 2nd ed., 1965. Biochemistry of the amino acids,vol. 2. Academic Press Inc., New York.

19. Moses, V. 1957. The metabolic significance of the citricacid cycle in the growth of fungus Zygorrhynchusmoelleri. J. Gen. Microbiol. 16:534-549.

20. Nakano, H., K. Sasaki, Y. Kurokawa, and H. Katsuki.1971. Metabolism of ,-methylmalic acid by a soilbacterium. J. Biochem. 70:429-440.

21. Olson, A. C., L. M. White, and A. T. Noma. 1968.Scintillation counting of the ninhydrin-amino acid-C"reaction products from an automatic amino acidanalyzer. Anal. Biochem. 24:120-127.

22. Phillips, A. T., J. I. Nuss, J. Moosic, and C. Foshay. 1972.Alternate pathway of isoleucine biosynthesis in Esche-richia coli. J. Bacteriol. 109:714-719.

23. Piez, K. A. 1962. Continuous scintillation counting ofcarbon-14 and tritium in effluent of the automaticamino acid analyzer. Anal. Biochem. 4:444-458.

24. Rabin, R. I., I. Salamon, A. S. Bleiweis, J. Carlin, and S.J. Ajl. 1968. Metabolism of ethylmalic acids byPseudomonas aeroginosa. Biochemistry 7:377-388.

25. Roberts, R. B., P. H. Abelson, D. B. Cowie, E. T. Bolton,and R. J. Britten. 1955. Studies of biosynthesis inEscherichia coli. Carnegie Institution of Washingtonpublication no. 607, Washington, D.C.

26. Robinson, I. M., and M. J. Allison. 1969. Isoleucinebiosynthesis from 2-methylbutyric acid by anaerobicbacteria from the rumen. J. Bacteriol. 97:1220-1226.

27. Sai, T. 1968. (-) Citramalic acid formation by respirationdeficient yeast mutants. IV. Inhibition of (-) citra-malic acid formation bv L-leucine and the inducedformation of (-) citramalic acid from citraconic acid.Bull. Brew. Sci. 14:61-72.

28. Sai, T., K. Aida, and T. Uemura. 1969. Studies on(-)-citramalic acid formation by respiration-deficientyeast mutants. V. Purification and some properties ofcitramalate condensing enzymes. J. Gen. Appl. Micro-biol. 15:345-363.

29. Sasaki, K., H. Nakano, and H. Katsuki. 1971. Enzymaticisomerization of ,-methylmalate to (-) citramalate bya soil bacterium. J. Biochem. 70:441-449.

30. Stern, N. E., E. Shenberg, and A. Tietz. 1969. Studies onthe metabolism of fatty acids of leptospira: the biosyn-thesis of zV- and A" monounsaturated acids. Eur. J.Biochem. 8:101-108.

31. Turner, L. H. 1967. Leptospirosis I. Trans. Roy. Soc.

210 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.

VOL. 117, 1974 AMINO ACID BIOSYNTE

Trop. Med. Hyg. 61:842-855.

32. Umbarger, H. E., and B. D. Davis. 1962. Pathways ofamino acid biosynthesis, p. 167-251. In I. C. Gunsalusand R. Y. Stanier (ed.), The bacteria, vol. 3. AcademicPress Inc., New York.

33. Umbarger, H. E. 1961. Feedback control by end productinhibition. Cold Spring Harbor Symp. Quant. Biol.26:301-312.

iESIS IN LEPTOSPIRA 211

34. Webster, R. E., and S. R. Gross. 1965. The a-isopropylma-late synthetase of Neurospora. I. The kinetics and endproduct control of a-isopropylmalate synthetase func-tion. Biochemistry 4:2309-2318.

35. World Health Organization. 1967. Current problems inleptospirosis research: report of a WHO exper group, p.32. World Health Organ. Tech. Rep. Ser. no. 380(Geneva).

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

03

Febr

uary

202

2 by

121

.163

.35.

197.