[COKTRIBUTION FROM THE XORTHERN
REGIONAL RESEARCH LABORATORY']
Production of a-Ketoglutarate in Glucose Oxidation by Pseudomonas jluorescens2 BY HAROLD J. KOEPSELL, FRANK H. STODOLA AND EUGENE S. SIIARPE RECEIVED MARCH17, 1952 Fermentation of glucose in synthetic media by Psez~domonas~uorescens has led to yields of 0.50 to 0.55 mole of a-kctoglutaric acid per mole glucose consumed with indications that the maximum possible yield may he considerably higher. Gluconate and 2-ketogluconate accumulated as intermediates, suggesting that glucose was metabolized according to the hypothetical hexosemonophosphate shunt pathway. However, successive 1-carbon degradation up to the triose stage, as currently postulated for this pathway, is not consistent with the a-ketoglutarate yields obtained. A paper chromatogram procedure for following the course of fermentation is described.
Certain species of Pseudomonas bacteria have The predicted theoretical yields of pyruvate and been shown to .oxidize glucose to gluconate and 2- a-ketoglutarate conforming to intermediary proketogluconate successively in essentially quantita- duction of one mole triose and three moles of carbon tive yield.3 One species, Ps. JEziorescens NRRL dioxide are 1.0 and 0.5 mole, respectively, per mole B-G, was found to produce a-ketoglutarate (about 2-ketogluconate consumed. A previous paper5 0.2 mole per mole glucose oxidized) on continued has indicated that the sum of the yields of pyruvate oxidation.4 Pyruvate accumulation was detected and a-ketoglutarate obtained in 2-stage oxidative in the oxidation of gluconate by this organism, and gluconate fermentation by Ps. JEuorescens exceeded evidence suggesting that pyruvate is converted the predicted yields. This conclusion, however, to a-ketoglutarate was ~ b t a i n e d . ~ depended on disregarding that stage of the ferThe observed sequence of oxidative steps would mentation in which no pyruvate or a-ketoglutarate seem to indicate that glucose oxidation by this accumulated. The evidence would be more conorganism proceeds according to one version of the clusive if higher than predicted yields of either hypothetical "hexosemonophosphate shunt" (War- acid could be deinoiistrated in over-all fermentaburg-Liprnann-Dickens) pathway, which is cur- tion. Simple conditions for obtaining a-ketorently receiving considerable attention among glutarate yields a t or in excess of 0.5 mole (Table I) biochemists as being a pathway alternate to aerobic from glucose in fermentation of a synthetic medium glycolysis.6 This version comprises oxidation of (Table 11) are described here. In addition to their glucose-6-phosphate zjin (i-phosphogluconate and value from n theoretical standpoint, the conditions f~-phospho-2-ketogluconate to 5-phosphopentose. Ixovide a convenient and inexpensive source of This reaction sequence constitutes stcpwise 1- a-ketoglutaric acid. carbon degradation of the carbohydrate molecule, 'rAIjLE I and it is frequently postulated that the sequence is A N D PYRUVATE I N GJ,UCOSP; repeated until one mole of triose phosphate plus Y~srmsO P N-KETOCLUTARATE OXIDATION B Y Ps. $fEuorescens three moles of carbon dioxide have been p r ~ d u c e d . ~ Final The triose phosphate would then be dissiniilated PerInitial 2-ketoby known reactions. Ps. $uore.scens B-6, when mengluglucoPyruvate a-Ketoglutarate nate. Found Found tation cose, grown on glucose or gluconate in synthetic media, pM./ pM./ Molar phf./ Molar time, pM./ ml. yieldu ml. yield tcxpt. hours ml. ml. accumulates 2-ketogluconate rapidly and almost stoichiometrically, whereas further oxidation lead3 0.01 281 0 . 5 5 0 151 307 510 ing to pyruvate and a-ketoglutarate accumulation 0 .01 252 5 .40 1-52 386 513 is slow. An opportunity for studying broadly 5O(i 8 .01 260 ,62 ,7 310 131 the nature of reactions in 2-ketogluconate oxidation 9 .01 2,59 .54 310 ,ilO (i 15.2 is thus afforded, and iiiforni:ition applicable to the .OO 241 .48 0 0 1.57 !'(81 310 obscure reactions of the hesoscriionop1iosl)hlLte 132 288 j12 28 ,52 .11 227 .4.7 shunt pathway may be obtained. I46 263 510 102 ti2 . I 5 242 .59 (1) One of t h e laboratories of t h e Bureau of Agriculcurnl a n d Inriurtrial Chemistry. Agricultural Research Administration, .:I S . r k p a r t m r n t 01 Agriculture. Article ni,t copyrighted (2) Presented bcfore t h e 13ivision of Agricultural ; ~ n dPcml Chrtnistry a t thr 1 1 9 t h h'lecfinc of t h c tlrnrrican Clicinic~ilSociety, Cli O h i w April 8 12. 1951.
159 216 510 100 13 .03 223 .X A yicld of 0.5 mole a-ketoglutaric acid per mole glucose c.orrrsporitis to a yield of 40.6 g. per I00 g. of glucose. 'I'AR1.R S1.V I l j f ! l l C
I I, (iiir,, I.)) IT ifi)
u.
I
.
~ d l l ~, i I: ~ i r L
~
~
I . K n c ~ p s ~ lrh,,i l , , 186, 71:; , l!>r,il) V. 12. I'nttpr, Bdlr t i l F , ~ z L ~ w I I , 4 , 5 1 1
~
~
~
J . Bid. Chetn., 196, 133 (1052), a n d others.
~i?I./Inl.
~ (. ;lulY>w
( l ! J l $ i , :.
r
I n this version, as found in yeast, 6-pliosphogluconate is simultaneously dehydrogenated a n d Id J-phospliorii,ulosr, An i m p o r t a n t distinction i o n s is t h a t t h e pathway siil t h e 2-ketogluconate ester involves a primary a-oxidation, ivhrreas t h e pathway in y e a s t implies a poxidation.
.*,l If1
12 , ,')..2,',
c~+twlGL
i n 4 N.0 . K a p l a n , Aizu. Rcv. B c o r h r i i i . , 20, 513 (19.7>l). i f ) Another vrrsion of t h e I ~ c n o s e m o ~ ? o i ~ l ~ o i p shuiit l l a t c iintliwdy is r m e r g i n g from t h e recent work of D. I.. Iiorecker and P. Z.Smyrniotic.
IJ
?~~liIlllJM VI)R m - ~ F W O ~ ; I . LTARAI'I'; l
S
hlgSO4 IJe" CaC0'3"
-
)
f'ROnLlCI I O N
(;,
100 nll
!+ I1
l i i I) :i2 1I . (El
/I.
1 p.p.n1. 375 3.75 a Urea, sterilized separately, could he substituted for ammonium sulfate on an equimolar basis without affecting the results. * Supplied as ferrous ammonium sulfate. Sterilized separately. 0.08
Oct. 20, 1952
CY-KETOGLUTARATE IN GLUCOSE OXIDATION BY Pseudomonas jluorescens
The upper group of data in Table 1 are analyses of fermentations in which oxidation beyond the 2- . ketogluconate stage was complete, whiie the lower group represents analyses a t the time of peak aketoglutarate yield but before 2-ketogluconate utilization was finished. No residual glucose was detectable on paper chromatograms despite heavy spotting a t the time of analysis. It is seen that a-ketoglutarate yields were in the range of 0.47 to 0.59 mole per mole glucose consumed. These yields were verified by isolation of a-ketoglutaric acid from the termentation liquor in one experiment. The fermentations cited in Table I were conducted in erlenmeyer flasks stoppered with cotton plugs, and diffusion of gases (oxygen substrate and carbon dioxide product) precluded the obtaining of fermentation balances. Paper chromatography as described below was employed for following the course of fermentation and for noting the appearance of the various products. It was found that a considerable period of time, characterized by intense metabolic activity, expired before a-ketoglutarate accumulation began. Cell synthesis, as indicated by turbidity and by cell crop on differential centrifugation, commenced within 12 hours after inoculation and proceeded rapidly. By 40 hours there had been advanced conversion of glucose to gluconate and 2-ketogluconate. However, a-ketoglutarate and pyruvate were not detectable despite heavy spotting until between 40 and 64 hours. As in our experiments on gluconate oxidation,j these findings suggest that accumulation of a-ketoglutarate is the result of a partial breakdown of the normal mechanism for complete oxidation of glucose to carbon dioxide and water. The Chromatography procedure used also afforded an opportunity to test for the production of other organic acids or reducing sugars as major products of the fermentations. None was detected. The yields of a-ketoglutarate obtained, together with the fact that a-ketoglutarate is a transient product which is not produced in the first stages of glucose oxidation and that pyruvate is converted to a-ketoglutarate, suggest that a-ketoglutarate arises by splitting of a 5- or 6-carbon molecule into several fragments of two or three carbons each, followed by condensation of the fragments, rather than stepwise degradation of 2-ketogluconate by the contemporary version of the hexosemonophosphate shunt pathway. The conversion of 100 g. of glucose to about 41 g. of a-ketoglutaric acid, the inexpensiveness and simple composition of the media, and the ease with which a-ketoglutaric acid can be recovered will enhance the possibility of producing the substance on an industrial scale by fermentation methgds.
Experimental Fermentation Technique and Analysis.-The composition of the synthetic medium used is shown in Table 11. The yield of a-ketoglutarate obtained was affected markedly by the amount of nitrogen supplied as previously experienced i n gluconate oxidation.6 At low nitrogen levels (less than 25 ph1. per n i l calculated LLS 1 gram-Atom per mole) fermeiitation was incomplete, while at high levels (above 50 pM. per ml.) fermentation was very much more rapid but aketoglutarate yields were low or non-existent. A nitrogen level of 25 pM. per ml. was used for most expcriments. Sub-
5143
stitution of urea for ammonium sulfate had no detectable e f f ~ ~ e ~ ~ t ~flasks ~ ~containing ~ ~ ~ $2oo ~ ml, ; e ofr medium inoculated with Pseudomonas fluoresce% NRRL B-6 were incubated on a reciprocating shaker at 28'. Progress of the fermentations was followed by reducing-power analysis (detecting both glucose and 2-ketogluconate) according to Somogyi.s Pyruvate and a-ketoglutarate were estimated simultaneously by a modification of the method of Friedemann and H a ~ g e n . Analytical ~ values were corrected for evaporation during incubation. Fermentation was intense initially, but reducing power changes became progressively slower after about 60 hours. Ten t o fourteen days was usually required for complete reducing power disappearance. A silky precipitate of calcium a-ketoglutarate was found frequently after about the seventh day, and the medium became a thick slurry of crystals by the end of fermentation. Sometimes, however, the medium remained supersaturated and no precipitation occurred without seeding. Supersaturation or precipitation did not change the fermentation rate appreciably. Isolation of a-Ketoglutaric Acid.-A portion (260 nil. froiii two flasks in expt. 154) of whole culture which contained 10.4 g. of a-ketoglutaric acid by the analytical nicthod cited, was concentrated t o a small volume, acidified with hydrochloric acid, and extracted repeatedly with ethyl acetatc. Concentration of the extract gave a crystalline iiiass of crude a-ketoglutaric acid. Crystallization from ethyl acetatcpetroleum ether gave the following fractions:
1 8.684g. 2 0.777 3 0.289 9.750 [Calcd.
C
H
Xeut. equiv.
41.0 41.2 40.6 41.1
4.16 4.14 3.97 4.14
74 76 73 731
1T.p
,lo
OC
112-114 107-111 114-115
From the mother liquor could be obtained 792 lug. of @-ketoglutaric acid setnicarbazone (m.p. 197-198") (cquivalent t o 570 mg. of a-ketoglutaric acid). Calcd. for CsHeOsNI:C, 35.5; H, 4.47; iY,20.7. Found: C, 35.6; H, 4.39; S , 20.7. The total recovery of a-ketoglutaric acid, therefore, amounted to 10.32 g. or 99.2% of t h a t found by analysis by the colorimetric method. Detection of Fermentation Products by Paper Chromatography.-A number of procedures from the chemical literature were investigated for separating the fermentation products (gluconate, 2-ketogluconate, pyruvate and a-ketoglutarate) and other possible new products (e.g., succinate, fumarate, citrate) on paper. The following procedure gave well defined separations. Pyruvic acid, which is tnoderately volatile, could be detected, and the individual acids of concern could be identified by the colors produced on color development. A small amount (2 t o 3 ml.) of centrifuged fermentation liquor was freed of cations' by adding ion exchange resin until an oxalate test for calcium was negative. The liquor then was spotted on each of three sheets of filter paper. The papers were irrigated overnight by the descending technique with a solvent mixture consisting of 3 volumes of butanol, 2 volumes of pyridine and 1 . 5 volumes of water. This solvent formula was developed by Jeanes, Wise and Dimler" for the separation of mono- and disaccfiarides, and it served well in our separations of organic acids also. After drying, one of the three papers was sprayed with a 0.05% solutioii of brom phenol blue. Organic acids appeared as yellow spots on a blue background. (8) M. Sornogyi, J . B i d . Chem., 160, 61 (1945). (9) T. E . Friedemann a n d G. E. Haugen, ibid., 147, 415 (1943). T h e single-solvent procedure, with ethyl acetate a s solvent, was used. As applied t o these experiments, i t was found t h a t greater accuracy a n d more precise differentiation between pyruvate a n d a-ketoglutarate could be obtaiued b y reading t h e yellow color o l t h e hydrazones, after extraction iuto sodium carbonate solution, :rt 380 mp. h'aOH to be 0.4 N i n t h e solution was t h e n added, a n d t h e resulting reddish-brown colors were read a t 435 and 520 mp. All readings were against a/) propriate renyriit t>Iaiiks. ~ e s u ~ t s ' w e r calcubate
