214 H. G. WOOD, F. W. LEAVER VOL. 12 (1953) 



strate in each case. Thus only in the case of adonitol was there a production of ^^COg 

 which approached 50 mM as predicted from the scheme of Fig. i. However, this produc- 

 tion of CO 2 may not be intimately related to the formation of propionate as such but 

 rather be inherent in the fermentation of the 5 carbon substrate. The following example 

 is presented only as an illustration : 



3C5 — ^sCg +3C2;C2 — >2Ci 



2 C2 + 2 Ci — > 2 C3 



The overall reaction is 3 C5 > 5 C3 > products. 



The Cg +Ci reaction might be an alpha carboxylation such as the reverse of the so- 

 called phosphoroclastic reaction to yield pyruvate^'-^®'^^ or perhaps some modification 

 of the beta fixation with formaldehyde of which the condensation with glycine to form 

 serine serves as an example^"-^^. The fact that the propionic acid bacteria utilize for- 

 maldehyde and incorporate it in all positions of propionates^ and also produce f ormalde- 

 ]^y^g33,34 fj-om glucose and glycerol makes this latter hypothesis attractive. 



Inspection of the data from the glycerol fermentation in Table VI indicates clearly 

 the small CO2 turnover with this substrate. 6.95 mAf of glycerol were fermented and it 

 was converted almost entirely to propionate. According to the scheme of Fig. i, 6.95 mM 

 of CO2 would be produced and one-half would be of substrate origin, thus 3.5 mM of 

 ^^COg would be produced. There were only 2.27 mM of ^^COa present in the pool so that 

 direct dilution should have reduced the specific activity by more than half whereas it 

 was actually reduced from 78.4 to 65.2 cpm/per/xM or only 17%. It thus appears quite 

 certain than the major part of the propionate was not formed in this case by decarboxyla- 

 tion of succinate with formation of free COg. 



If the mechanism of propionate formation is by decarboxylation of succinate or a 

 C4-dicarboxylic acid it appears that a "Ci" other than COg must be formed. Evidence 

 pointing to this possiblity has already been presented in discussing the assumptions 

 made in the calculation of turnover and the idea of a "C^" not identical with COg was 

 suggested earlier by Leaver and Wood^^. Furthermore, there is good reason to reserve 

 judgment on the mechanism of propionate formation because there is evidence that 

 propionate can be formed by certain organisms without the occurrence of a C4 dicarb- 

 oxylic acid as an intermediate. Cardon and Barker^^ and Johns^^ have presented 

 results with Clostridium propionicum that indicate there is a direct reduction of lactate 

 to propionate, possibly via acrylate. Recently Leaver^' has provided additional evidence 

 supporting this possibility in that the fermentation of lactate-3-^'*C by C. propionicum 

 led to almost quantitative formation of propionate-3-^'*C. On the other hand with P. 

 arabinosum the ^'*C from lactate was almost completely randomized in the 2, 3 positions 

 of the propionate, as would be expected if a symmetrical C4-dicarboxylic acid were an 

 intermediate. However, the finding that propionate may undergo secondary reactions 

 in which the ^*C of propionate is randomized (Table V) makes it necessary to leave open 

 the possibility that the propionate is in part at least, first formed by direct reduction of 

 a C3 compound and only by secondary reactions is randomized. It likewise is possible 

 that formation of propionic acid by C. propionicum involves a C4-dicarboxylic acid but 

 it does not pass through a symmetrical form. 



These observations indicate that: (i) there is a biological mechanism for formation 

 of propionate by direct reduction, (2) in the propionic acid fermentation, the CO2 

 turnover is too low to indicate that all the propionate is formed by decarboxylation of 



References p. 221(222. 



