Di-D-Fructose anhydrides

In 1952, Wolfrom and Hilton demonstrated that L-sorbose was also capable of forming dimeric dianhydrides,22 and they postulated sorbofuranosyl and pyra-nosyl cationic intermediates. In 1955, Boggs and Smith23 postulated a fructofu-ranosyl cationic intermediate in the formation of per-O-acetyl ot-D-Fru/-1,2 2,l -p-D-Fru/[di-D-fructose anhydride I (5)] from inulin triacetate. They indicated that three adjacent P-2,l -linked fructofuranosyl units would be required for formation of the dianhydride. 

In 1933, Schlubach and Knoop32 isolated a di-D-fructose dianhydride from Jerusalem artichoke and tentatively identified it as difructose anhydride I [a-D-Fru/-1,2 2,1 - 3-D-Fn / (5)]. Alliuminoside ( -D-fructofuranose- -D-fructofura-nose 2,6 6,2 -dianhydride) was isolated from tubers of Allium sewertzowi by Strepkov33 in 1958. Uchiyama34 has demonstrated the enzymic formation of a-D-Fru/-1,2 2,3 -(3-D-Fru/ [di-D-fructose anhydride III (6)] from inulin by a homogenate of the roots of Lycoris radiata Herbert. 

It has been mentioned that the mixture of trimethyl-D-fructoses from the acid hydrolysis of hexamethyl-di-D-fructose anhydride III, which are now known to be the 3,4,6- and 1,4,6-trimethyl-D-fructoses, shows a specific rotation near that of pure 3,4,6-trimethyl-D-fructose it is to be inferred accordingly that these two trimethyl-D-fructoses do not differ greatly in rotation. Montgomery76 has synthesized 1,4,6-tri-methyl-D-fructose and found its rotation in chloroform to be [a] d = + 29.7°, a value approximating that of 3,4,6-trimethyl-D-fructose (+ 27.7° in the same solvent). Pertinent also are the respective rotations of the acetone condensation products from these two trimethyl-D-fructoses. Montgomery found that 3,4,6-trimethyl-D-fructose under-... 

Schlubach, Knoop and Liu65 found indications that a more readily hydrolyzed di-D-fructose anhydride was an intermediate product in the hydrolysis of irisin. 

A compound represented by XVIII would yield 1,3,4-trimethyI-D-fructose, a crystalline substance with [aJD20 = — 51.4° (water), and 1,3,6-trimethyl-D-fructose (now unknown), upon hydrolysis of its hexa-methyl derivative. No evidence has been obtained to indicate the presence of this negatively rotating trimethyl-D-fructose and it is established that the mixture of trimethyl-D-fructoses from hexamethyl-di-D-fructose anhydride II rotates in the positive range of 25 to 30°. Unless the 1,3,6-trimethyl-D-fructose should prove to possess the unusually high rotation of about + 100°, which seems at least improbable, structure XVIII can be excluded. 

If di-D-fructose anhydride II has formula XIX, a mixture of 1,3,4-and 1,4,6-trimethyl-D-fructoses would be present in the hydrolytic product. Such a mixture would have a specific rotation of — 10° to — 20° (water) in contrast to the value of + 25 to + 30 found by McDonald and Jackson.76 The rotation of 1,4,6-trimethyl-D-fructose was measured by Montgomery76 in chloroform (+ 29.7°), but it has not been measured directly in water. However, the hydrolysis data of McDonald and Jackson76 for hexamethyl-di-D-fructose anhydride III show that 1,4,6-trimethyl-D-fructose has about the same rotation in water as in chloroform. The argument thus appears to exclude structure XIX. 

An anhydride whose formula is represented by XVII would yield a mixture of 3,4,6-trimethyl-D-fructose and 1,3,6-trimethyI-D-fructose when its hexamethyl derivative was hydrolyzed. Available data concerning di-D-fructose anhydride II are at least in accord with this structure, and if the present inferences are accepted, one may conclude that the rotation of 1,3,6-trimethyl-D-fructose is near those of 1,4,6- and 3,4,6-tri-methyl-D-fructose. 

Based upon this, we predicted that if the configuration and conformation of di-D-fructose anhydride I established by Lemieux and Naja-rajan (9) was correct, the compound should be nearly tasteless (the dihedral angle between -glycol groups is 150° and 75° in the a- and / -furanoside rings, respectively). This was true. 

The only dimethyl-D-fructose which has been characterized, 3,4-di-methyl-D-fructose, has been prepared by McDonald and Jackson141 from di-D-fructose anhydride I. Tritylation of this anhydride gives the 6,6 -ditrityl derivative which is methylated to 3,4,3, 4 -tetramethyl-6,6 -di-trityl-di-D-fructose anhydride I. Removal of the trityl groups followed by hydrolysis yields liquid 3,4-dimethyl-D-fructose, [ ]d —60.66° in water. It has also been obtained, with 4-methyl-D-fructose, from the hydrolysis of methylated di-D-fructose anhydride III. The structure of this dimethyl-D-fructose follows from its method of preparation from di-D-fructose anhydride I whose structure is known.10 McDonald and Jackson also prepared 3,4-dimethyl-D-fructose from inulin by the following method inulin — monotrityl inulin — monotrityl inulin diacetate — dimethyl monotrityl inulin — dimethyl isopropylidene-D-fructose — methyl dimethyl-D-fructoside —> 3,4-dimethyl-D-fructose. Its structure was confirmed by its oxidation without loss of methyl to the same lactol of the dimethyl dibasic acid obtained from 1,3,4-trimethyl-D-fructose (see page 78). The phenylosazone made from 3,4-dimethyl-D-fructose has m. p. 126° that from 3,4-dimethyl-D-glucose has not been recorded. 

The preparation of liquid 4-methyl-D-fructose, [a]D —87.5°, from di-D-fructose anhydride III has already been mentioned (see page 80). The phenylosazone, m. p. 156°, is identical with that obtained from 4-methyl-D-glucose.14a... 

McDonald and Jackson141 prepared liquid 6-methyl-D-fructose [a]D +6.4° in water, by the following reaction sequence di-D-fructose anhydride I — 6,6 -ditrityl-di-D-fructose anhydride I —> 6,6 -ditrityl-di-D-fructose anhydride I 3,4,3, 4 -tetraacetate — di-D-fructose anhydride I 3,4,3, 4 -tetraacetate — 6,6 -dimethyl-di-D-fructose anhydride I 3,4,3, -4 -tetraacetate — 6,6 -dimethyl-di-D-fructose anhydride I — 6-methyl-D-fructose. The possibility of acetyl migration during methylation makes it dangerous to rely on this synthesis as a definitive one for 6-methyl-D-fructose, especially as the proof of structure depends largely on the method of formation. The only crystalline derivative prepared by... 

Taniguchi, T. and Uchiyama, T., The crystal structure of di-D-fructose anhydride III, produced by inulin d-fructotransferase, Carbohydrate Res., 107, 255-262, 1982. 

The same conclusion can be drawn from the physical properties of the hexaacetate of di-D-fructose anhydride, which has thus far been obtained only in the form of plastic sheets resembling glass. Table I compares the melting points and specific rotations of other di-D-fructose anhydride hexaacetates reported in the literature. 

Proof of the location of the carbon atoms involved in the formation of the monomeric anhydro-n-fructose was obtained by hydrolyzing the sirupy hexamethyldi-n-fructose anhydride. Phenylosazones were obtained which showed no loss of methyl groups from the trimethyl-D-fructose, while the rotation of sirupy trimethyl-n-fructose placed it in the pyranose family. This meant that hexamethyldi-n-fructose anhydride on hydrolysis yielded 3,4,5-trimethyl-n-fructose, and that the parent substance whose formula is given on page 119 was similar to the di-D-fructose anhydride described by Schlubach and Behre. ... 

---