TABLE 58. Rates of oxidation of alpha and beta sugars [35, 36, 37]

[In aqueous solutions at 0° C containing 0.05 mole of sugar, and approximately 0.08 mole of free bromine per liter and buffered with barium carbonate and carbon dioxide]

In many reactions the arrangement of the hydroxyls in relation to the plane of the ring is an especially important factor in influencing the rate and course of the reaction. Condensations in which two hydroxyls combine with another group take place when the hydroxyls lie on the same side of the ring. Thus the behavior of the sugars and their derivatives in their reactions with acetone, benzaldehyde, boric acid, and many other substances is largely conditioned by the cis or trans arrangement of the hydroxyl groups [50, 51, 28]. The facility with which boric acid combines with cis-hydroxyls on adjacent atoms. was mentioned with respect to the structure of alpha and beta glucose (p. 422). The property of increasing the electrical conductivity of boric acid appears to be characteristic of all sugars which have two hydroxyl groups on adjacent carbon atoms on the same side of the ring. Normally, condensation with acetone involves cis-hydroxyl groups on adjacent carbon atoms, making a cyclic link of five atoms [52], while condensation with benzaldehyde usually involves alternate cis-hydroxyls, making a cyclic link of six atoms [53].

In the presence of acid catalysts the free sugars react in the furanose or the pyranose form, depending upon which form is the more favorable for the condensation reaction. Thus glucose forms a 1,2-3,5-diacetone derivative, even though the free sugar exists in the pyranose modification. Correlations have been made also between the cis-trans configurations and the tendency to form anhydrosugars and sugar anhydrides. Thus a-d-glucose yields a glucosan by condensation between the hydroxyls of carbons 1 and 2; B-d-glucose, however, yields a glucosan by condensation between the hydroxyls of carbons 1 and 6 [54]. Apparently the configuration of carbon 1 determines the nature of the product formed. The importance of cis-trans configura

tional relationships in the formation of anhydrosugars and orthoacetates was reviewed recently by Isbell [55], who suggested that many carbohydrate reactions can be explained by Werner's concept [56] for the Walden inversion. Intramolecular condensation reactions involving Walden inversion appear to take place only on the side of the carbon opposite the departing group.

The influence of the cis-trans configuration of the sugars and their derivatives in relation to their biological behavior has been recognized for a long time. The sorbose bacterium, for example, oxidizes sugar alcoH H

hols to ketoses when they contain the grouping, C -C

-CH2OH,

OH OH with the CH(OH) groups having a cis configuration in the projectional formula [57]. The relationship between cis-trans configuration and rates of enzyme action has also been studied [58].

Even though many generalizations can be reached by considering only certain aspects of the molecule, each group influences to some extent the properties of the others, and therefore it is essential to consider also the configuration of the molecule as a whole. In the pyranoses, the carbon and ring oxygen atoms are tied up in a ring and cannot exhibit free rotation; consequently the pyranose ring forms a fundamental structure about which the hydroxyls and hydrogens are distributed according to the configurations of the constituent carbon atoms. The arrangement of the groups on either side of the ring makes up the thickness and general conformation of the molecule. The different configurations for the 5 asymmetric carbon atoms of the ring give rise to 32 isomeric pyranoses, which consist of 16 pairs of enantiomorphs. These fundamental configurational types are represented by the alpha and beta modifications of glucose, mannose, galactose, talose, gulose, idose, allose, and altrose and are illustrated by the formulas on p. 426. By substitution, all pyranose sugars are derived from these fundamental types, or from their enantiomorphs. The aldopentoses, methylpentoses, and heptoses differ from the aldohexoses in that the CH2OH group of the latter is replaced by hydrogen, the methyl group, or by the CHOH.CH2OH group, respectively. The ketoses differ from the aldoses in that the H on the first carbon of the aldose is replaced by a CH2OH group. If the sugars are considered in this manner, sorbose, xylose, B-galaheptose, maltose, lactose, and cellobiose are merely substituted glucoses, or levulose, perseulose, arabinose, a-mannoheptose, and B-guloheptose are merely substituted galactoses.

The tendency of a sugar or a sugar derivative to form five- or sixmembered lactol rings is dependent on the character of the substituent groups, and on the configuration of the carbon atoms involved in forming the furanose and pyranose rings. This subject will be considered more fully in the next chapter.

7. REFERENCES

[1] E. O. von Lippmann, Geschichte des Zuckers, 2d ed., p. 683 (Julius Springer, Berlin, 1929).

[2] Lowitz, Chem. Ann. von Crell 1, 218, 345 (1792).

[3] Chevreul, Ann. der Chym. 95, 319 (1815).

[4] Braconnot, Ann. Phys. Chim. 12, 172 (1819).

[5] F. J. Bates, Facts About Sugar 21, 250 (1926).

[6] J. B. Dumas, Traité de Chimie 6, 273 (1843) (Paris).

[7] A. Baeyer, Ber. deut. chem. Ges. 3, 63 (1870).

[8] B. Tollens and F. Mayer, Ber. deut. chem. Ges. 21, 1566 (1888) [9] H. Kiliani, Liebigs Ann. Chem. 205, 182 (1880).

[10] J. Meunier, Compt. rend. 111, 49 (1890).

[11] H. Kiliani, Ber. deut. chem. Ges. 19, 767 (1886).

[12] H. Kiliani, Ber. deut. chem. Ges. 19, 221 (1886).

[13] E. Fischer, Ber. deut. chem. Ges. 17, 579 (1884).

[14] L. Marchlewski and W. Urbanczyk, Biochem. Z. 262, 248 (1933).

[15] Dubrunfaut, Compt. rend. 23, 38 (1846).

[16] T. M. Lowry and G. F. Smith, Rapports sur les hydrates de carbone, p. 79, 10th Conference of the International Union of Chemistry (Liege, 1930).

[17] Colley, Compt. rend. 70, 403 (1870).

[18] B. Tollens, Ber. deut. chem. Ges. 16, 921 (1883).

[19] W. N. Haworth, The Constitution of Sugars (Edward Arnold & Co., London, 1929).

[20] H. S. Isbell and C. S. Hudson, BS J. Research 8, 327 (1932) RP418.

[21] H. S. Isbell, BS J. Research 8, 615 (1932) RP441.

[22] H. S. Isbell, J. Am. Chem. Soc. 55, 2166 (1933).

[23] H. S. Isbell and W. W. Pigman, J. Research NBS 20, 773 (1938) RP1104. [24] T. M. Lowry, Optical Rotatory Power (Longmans, Green and Co., London, 1935).

[25] M. A. Rosanoff, J. Am. Chem. Soc. 28, 114 (1906).

[26] E. Fischer, Ber. deut. chem. Ges. 27, 3211 (1894).

[27] E. Fischer, Uuntersuchungen über Kohlenydrate und Fermente (Berlin, 1909 and 1922).

[28] J. B. Böeseken, The Configuration of the Saccharides (translated by S. Coffey, page 58, A. W. Sijthoff, Leyden, (1923).

[29] E. F. Armstrong, J. Chem. Soc. 83, 1305 (1903).

[30] L. J. Simon, Compt. rend. 132, 487 (1901).

[31] E. L. Jackson and C. S. Hudson, J. Am. Chem. Soc. 59, 994 (1937).

[32] C. S. Hudson, BS Sci. Pap. 21, 241 (1926) S533.

[33] C. S. Hudson, J. Am. Chem. Soc. 31, 66 (1909).

[34] K. Freudenberg, Stereochemie, page 692 (Franz Deuticke, Leipzig and Wien, 1932).

[35] H. S. Isbell, J. Research NBS 18, 505 (1937) RP990.

[36] H. S. Isbell and W. W. Pigman, J. Research NBS 18, 141 (1937) RP969.

[37] H. S. Isbell, J. Chem. Education 12, 96 (1935).

[38] W. W. Pigman and H. S. Isbell, J. Research NBS 19, 189 (1937) RP1021. [39] J. H. Van't Hoff, The Arrangement of Atoms in Space (translated by Eiloart. Longmans, Green and Co., London, (1898).

[40] M. A. Rosanoff, J. Am. Chem. Soc. 28, 525 (1906); 29, 536 (1907).

[41] K. Freudenberg and W. Kuhn, Ber. deut. chem. Ges. 64, 703 (1931).

[42] C. S. Hudson, J. Am. Chem. Soc. 48, 1434 (1926).

[43] H. S. Isbell and H. L. Frush, J. Research NBS 24, 125 (1940) RP1274.

[44] C. S. Hudson, J. Am. Chem. Soc. 32, 338 (1910).

[45] E. Anderson, J. Am. Chem. Soc. 34, 51 (1912).

[46] P. A. Levene, J. Biol. Chem. 23, 145 (1915).

[47] P. A. Levene and G. M. Meyer, J. Biol. Chem. 31, 623 (1917).

[48] C. S. Hudson, J. Am. Chem. Soc. 39, 462 (1917).

[49] H. S. Isbell, J. Research NBS 14, 305 (1935) RP770.

[50] F. Micheel and H. Micheel, Ber. deut. chem. Ges. 63, 386 (1930).

[51] W. N. Haworth and E. L. Hirst, Ann. Rev. Biochem. 5, 82 (1936).

[52] R. G. Ault, W. N. Haworth, and E. L. Hirst, J. Chem. Soc. 1935, 1012.

[53] L. Zervas, Ber. deut. chem. Ges. 64, 2289 (1931).

[54] A. Pictet and P. Castan, Helv. Chim. Acta 3, 645 (1920). N. Cramer and E. N. Cox, Helv. Chim. Acta 5, 884 (1922). A. Pictet and J. Sarasin, Helv. Chim. Acta 1, 87 (1918). P. Karrer and A. P. Smirnoff, Helv. Chim. Acta 5, 124 (1922).

[55] H. S. Isbell, Ann. Rev. Biochem. 9, 65 (1940).

[56] A. Werner, Ber. deut. chem. Ges. 44, 873 (1911).

[57] M. G. Bertrand, Ann. chim. phys. 3, 181 (1904).

[58] B. Helferich, S. Winkler, R. Gootz, O. Peters, and E. Günther, Z. physiol. Chem. 208, 91 (1932).

XXIX. MUTAROTATION AND SUGARS IN SOLUTION

1. CHARACTERISTICS OF THE EQUILIBRIUM STATE Dubrunfaut [1] discovered that when glucose is dissolved in water, the optical rotatory power of the solution decreases on standing until finally it reaches a constant value. Subsequently, Pasteur [2], Erdmann [3], and others [4] found that the optical rotations of freshly prepared solutions of the reducing sugars in general change on standing, a phenomenon which came to be known as mutarotation.

In 1846, when Dubrunfaut discovered mutarotation, he advanced the hypothesis that the change in optical rotation is caused by a change in molecular structure. After many years this hypothesis was confirmed by Tanret's preparation [5] of two forms of glucose and lactose, one having a higher rotation than the stable solution and the other having a lower rotation. Fischer's preparation of two methyl glucosides and Armstrong's discovery [6] that on enzymic hydrolysis the more dextrorotatory glucoside yields a sugar solution with a rotatory power greater than the equilibrium value and that the less dextrorotatory glucoside yields a sugar solution with a rotatory power less than the equilibrium value furnished a clue as to the character of the change involved in the mutarotation reaction. Subsequent work has shown that the mutarotation reactions consist in the reversible interconversion of the various modifications of the sugars and that many mutarotation reactions are caused almost entirely by the interconversion of the alpha and beta pyranose modifications. Since the several modifications in the sugar solution have different physical properties, the mutarotation reaction may be followed by the changes in solubility, volume, refractivity, and energy, in addition to the optical rotations.

As mentioned on page 414, the equilibrium involves the structures found in the open-chain sugar, the alpha and beta pyranose modifications, and the alpha and beta furanose modifications. The relative proportions of these constituents vary greatly from sugar to sugar and with the experimental conditions. The separation of numerous open-chain derivatives is evidence for the presence of the open-chain modification, but the lack of strong characteristic aldehydo reactions indicates that the open-chain modification is not present in large quantity. In weakly alkaline solution some sugars show a faint absorption band in the region characteristic of the carbonyl group [7]. This absorption band is further evidence that the open-chain modification is present only in small quantity. The ease with which the alpha and beta pyranose modifications can be crystallized shows that they are the predominating constituents of the sugar solution. For alpha and beta furanoses there are no characteristic qualitative tests such as the absorption band of the open-chain modification. The preparation of the methyl furanosides and other furanose derivatives and the crystallization of a calcium chloride compound of a manno-furanose [8, 9] imply that the furanoses are present. This inference is further substantiated by the complex mutarotation reactions, and by the similarity of the mutarotation of levulose to the mutarotation of the furanose form of levulose liberated from sucrose by invertase [10].

The position of the equilibrium between the various modifications in sugar solutions appears to depend upon the configuration of the

sugar and upon the substituent groups. Aldoses which have the glucose, mannose, and gulose structures establish equilibrium states consisting almost exclusively of the alpha and beta pyranose modifications [11]. Aldoses which have the galactose, talose, and idose structures establish equilibrium states containing small but substantial quantities of the furanose modifications and larger quantities of the alpha and beta pyranose modifications [12, 13, 14, 15]. Sugars which have the fructose structure establish equilibrium states consisting for the most part of a single pyranose modification with a substantial proportion of a furanose modification [10], and sugars which have the allose, altrose, sorbose, and tagatose structures establish equilibrium states consisting largely of a single pyranose modification [16, 17, 18, 19]. Replacing the hydrogen of the aldehydo group by a CHOH group to give a ketose results in a large alteration in the equilibrium proportions of the various ring and openchain modifications of the resulting sugar. Isbell and Pigman [10] have shown that an equilibrium solution of levulose does not contain an appreciable quantity of the beta pyranose modification, even though the equilibrium solution of the structurally related aldose (d-arabinose) contains 73 percent of the beta pyranose modification. Furthermore, the proportions for the modifications of sorbose and of tagatose also differ widely from the proportions found for the configurationally related sugars, glucose and mannose. The aldoses and ketoses differ on the carbon atom involved in forming the lactol ring; that is, the structural difference is on the carbon involved in the ring-forming reaction. When viewed in this light, the differences in the equilibrium states for the aldoses and ketoses are understandable and comparable to the differences in the equilibrium states for the sugars and sugar acids.

The aldonic acids can be considered as derived from the aldoses by replacement of the hydrogen on carbon 1 by a hydroxyl. In aqueous solution, they establish equilibrium with the delta and gamma

lactones.

On account of differences in the chemical properties of the acids and lactones, the composition of the equilibrium solutions for the acids can be determined more readily than the composition of the sugar solutions. In marked contrast to the sugars, the equilibrium mixtures for the aldonic acids contain larger quantities of the five-membered ring modifications (gamma lactones) and smaller quantities of the sixmembered ring modifications (delta lactones). The marked difference in the equilibrium states of the aldoses, ketoses, and aldonic acids