488
VERNON H. CHELDELIN, CHIH H. WANG, AND TSOO E. KING
are oxidizable by Acetobacter species at the position marked with
asterisk (*). This is reasoned from the fact that the CH 3 CHOH group
acts as if it were a CH 2 OH group. One of the interesting tests for the
Bertrand-Hudson rule is the oxidation of D-volemitol. The polyol contains two positions, namely C-2 and C-6, labile to oxidation. Although
the diketo derivative has not been found, two ketoses, D-mannoheptulose
and sedoheptulose, have been isolated in approximately equal yields
from the action of A. suboxydans on the polyol (330).
After critical examination* of the literature, McCorkindale and Edson
(333) have further generalized the Bertrand-Hudson rule. They have
advanced a unified theory that the enzyme responsible for reactions
prescribed by the rule is really due to a DPN linked, nonparticulate
soluble catalyst (330, 334) and have named it D-mannitol dehydrogenase.
However, this hypothesis has been challenged by the simultaneous
discoveries by Edson's group (335) and by this laboratory (310, 318,
322) on another pathway in the oxidation of sorbitol. The formation of
L-sorbose from D-sorbitol catalyzed by Acetobacter has been known
since the first experiment of Bertrand. In fact, this reaction is the only
commercial source for this keto sugar. In 1956, it was found that the
oxidation of sorbitol by A. suboxydans yields not only L-sorbose, but
also D-fructose (318, 322, 335, 336). In a cell-free system, the pathway
is determined by the kind of pyridine nucleotide present; the reaction
that forms sorbose is TPN dependent whereas the other requires DPN.
Here the "D-mannitol dehydrogenase" of McCorkindale and Edson as a
DPN-linked soluble enzyme cannot be applied.
This dual pathway of sorbitol oxidation is further characterized by
the partial separation of the two enzymes. The products are also identified by conventional means, such as melting points, paper chromatography, polarimetry, and the crystalline form of the osazones. If these
observations are correct, then the oxidation of sorbitol to fructose by A.
suboxydans is the first real exception to the well-known Bertrand rule.
A similar dual pathway has been found in Candida utilis in the oxidation of sorbitol (335) to fructose and sorbose. This organism also resembles A. suboxydans in the oxidation of mannitol to fructose, in the
oxidation of xylitol to xylulose, and in the failure to attack dulcitol.
Mention must be made that the oxidation of sorbitol in A. suboxydans
can be catalyzed by another enzyme (cf. Fig. 12). This enzyme (335,
337) differs from the two previously described in that it occurs in the
particulate fraction, shows a pH optimum at 6, and is pyridine nucleotide
independent. The product is presumably L-sorbose.
* See Hudson (332a) and Carr (332b) for the chemistry of, and early studies
on, metabolism of polyols.
VERNON H. CHELDELIN, CHIH H. WANG, AND TSOO E. KING
are oxidizable by Acetobacter species at the position marked with
asterisk (*). This is reasoned from the fact that the CH 3 CHOH group
acts as if it were a CH 2 OH group. One of the interesting tests for the
Bertrand-Hudson rule is the oxidation of D-volemitol. The polyol contains two positions, namely C-2 and C-6, labile to oxidation. Although
the diketo derivative has not been found, two ketoses, D-mannoheptulose
and sedoheptulose, have been isolated in approximately equal yields
from the action of A. suboxydans on the polyol (330).
After critical examination* of the literature, McCorkindale and Edson
(333) have further generalized the Bertrand-Hudson rule. They have
advanced a unified theory that the enzyme responsible for reactions
prescribed by the rule is really due to a DPN linked, nonparticulate
soluble catalyst (330, 334) and have named it D-mannitol dehydrogenase.
However, this hypothesis has been challenged by the simultaneous
discoveries by Edson's group (335) and by this laboratory (310, 318,
322) on another pathway in the oxidation of sorbitol. The formation of
L-sorbose from D-sorbitol catalyzed by Acetobacter has been known
since the first experiment of Bertrand. In fact, this reaction is the only
commercial source for this keto sugar. In 1956, it was found that the
oxidation of sorbitol by A. suboxydans yields not only L-sorbose, but
also D-fructose (318, 322, 335, 336). In a cell-free system, the pathway
is determined by the kind of pyridine nucleotide present; the reaction
that forms sorbose is TPN dependent whereas the other requires DPN.
Here the "D-mannitol dehydrogenase" of McCorkindale and Edson as a
DPN-linked soluble enzyme cannot be applied.
This dual pathway of sorbitol oxidation is further characterized by
the partial separation of the two enzymes. The products are also identified by conventional means, such as melting points, paper chromatography, polarimetry, and the crystalline form of the osazones. If these
observations are correct, then the oxidation of sorbitol to fructose by A.
suboxydans is the first real exception to the well-known Bertrand rule.
A similar dual pathway has been found in Candida utilis in the oxidation of sorbitol (335) to fructose and sorbose. This organism also resembles A. suboxydans in the oxidation of mannitol to fructose, in the
oxidation of xylitol to xylulose, and in the failure to attack dulcitol.
Mention must be made that the oxidation of sorbitol in A. suboxydans
can be catalyzed by another enzyme (cf. Fig. 12). This enzyme (335,
337) differs from the two previously described in that it occurs in the
particulate fraction, shows a pH optimum at 6, and is pyridine nucleotide
independent. The product is presumably L-sorbose.
* See Hudson (332a) and Carr (332b) for the chemistry of, and early studies
on, metabolism of polyols.