While there is no doubt whatever of the ability of the animal to synthesize fat from carbohydrate, the mechanism of the process is far from clear. As expressed by Leathes, "the chemical changes involved are fascinating in their obscurity." Whatever the exact steps, the transformation of carbohydrate into fatty acid radicles must involve reduction of hydroxyl groups and condensations to form the long chains of the higher fatty acids. We have already seen that in what we believe to be the normal course of carbohydrate catabolism there occurs, either along with or quickly following the breaking of the glucose molecule into three-carbon compounds, a reduction of certain hydroxyl groups with transfer of the oxygen so that substances such as methyl glyoxal, pyruvic acid, and lactic acid are formed. From pyruvic acid or lactic acid acetaldehyde may be formed; two molecules of acetaldehyde may then undergo aldol condensation and the aldol be transformed (by simultaneous reduction and oxidation, or transfer of oxygen from the ft to the terminal carbon) into butyric acid. Such an hypothesis is consistent with reactions observed in vitro and with the well-known production of butyric acid in certain bacterial fermentations of sugar and of lactic acid. Leathes favors this hypothesis and comments upon it (in part) as follows: "The biochemical significance of the synthesis of butyric acid from lactic acid and from sugar by bacteria becomes greater, however, when it is remembered that in this fermentation normal caproic acid is simultaneously formed, and as Raper showed also, though in still smaller amount, normal octoic or caprylic acid. ... In butyric fermentation it seems that the reactions that lead to the synthesis of butyric acid may lead to the synthesis of acids I of longer chains but still unbranched and containing an even number of carbon atoms, in other words, that these acids may be produced by condensation of two, three, or four acetic aldehyde molecules. In higher organisms, plants or animals, this same condensation carried further would result as Nencki suggested in the formation of the series of acids with straight chains of even numbers of carbon atoms leading up to palmitic and stearic acid." Raper1 has shown experimentally that condensation of two molecules of aldol in alkaline solution yields a straight chain product which on oxidation and reduction by laboratory methods yields normal octoic (caprylic) acid.

Smedley has developed an alternative hypothesis regarding the mechanism of fatty acid synthesis from carbohydrate material.

According to Smedley,2 the most probable starting point is pyruvic acid.

As an intermediary step in the metabolism of carbohydrate, pyruvic acid is probably formed in large quantities in the body, though its reactivity may prevent it from accumulating in measurable amounts.

Pyruvic acid readily breaks down to acetaldehyde and carbon dioxide. It also condenses with aldehydes to form products which, under conditions similar to those existing in the body, undergo rearrangements (through simultaneous or successive oxidation and reduction) which result in the splitting out of carbon dioxide leaving an acid of two more carbon atoms than were contained in the original aldehyde; or an aldehyde of two more carbon atoms than the original aldehyde may be formed, and this in turn react with another molecule of pyruvic acid forming a fatty acid or aldehyde of two more carbon atoms.

Each of these hypotheses assumes as a starting point only substances which we have good reason to believe are regularly formed in carbohydrate metabolism, and both are consistent with the well-known fact that natural fats contain fatty acid radicles having all multiples of two carbon atoms from four to eighteen, but none containing uneven numbers of carbon atoms in the molecule.

1 J. Chem. Soc., Vol. 91, page 1831 (1907)- See also Leathes, The Fats, pages 106-109.

2 Journal of Physiology, Vol. 45, Proc. page 26; Biochemical Journal, Vol. 7, page 364.