Suzuki, Shimamura and Odake also used the phosphotungstic precipitation method and claimed to have prepared the crystalline antineuritic substance which they called oryzanin in the form of a crystalline picrate. Drummond and Funk repeated this work, but were unable to confirm the Japanese results. A group of British chemists (Edie, Evans, Moore, Simpson and Webster) obtained an active fraction from yeast and succeeded in separating this into a crystalline basic member belonging to the pyrimidine group which they called torulin.
None of these three preparations have stood the test of analysis however and their curative properties seem to lie in their greater or less contamination with the actual substance, whatever it is. Numerous modifications of the fundamental method for extracting the substance have been planned and executed. Funk for example has shown that if the phosphotungstic precipitate is treated with acetone it is possible to separate it into an acetone soluble and an acetone-insoluble fraction and that the curative fraction is in the latter. McCollum has reported that while ether, benzene and acetone cannot be used to extract the B vitamine from its source, benzene, (and to a slight extent acetone) will dissolve the vitamine if it is first deposited from an alcohol extract on dextrin. These observations have not yielded any further clew to the nature of the substance.
Recently Osborne and Wakeman have proposed a modification which yields a concentrate of high potency. Their method is to add fresh yeast to slightly acidified boiling water and continue the boiling for about five minutes. This process coagulates the proteins that are present and permits their removal by filtration. The protein-free filtrate appears to contain all of the vitamine originally present in the yeast but attempts to precipitate the vitamine fractionally from the evaporated filtrate by means of increasing concentration of added alcohol has been only partially successful. The method however yields a concentrated extract, and Harris has made use of this process to prepare tablets for medicinal purposes.
Seidell and Williams some time ago devised a procedure which seemed to give promise of good results. Their discovery was that when a filtrate from autolysed yeast is prepared, rich in the vitamine, and is shaken with a specially activated fuller's earth (the preparation produced by Lloyd and known as Lloyd's reagent has this power) in a proportion of 50 grams to the liter of extract the vitamine is absorbed by the earth and when the latter is filtered off it carries the vitamine with it. In their process they shake the mixture for about one-half hour and then remove the earth by filtration. Analysis of the yeast liquor after the extraction shows it to contain practically the same solids as originally present but to have lost practically all its vitamine. The latter is firmly attached to the earth and repeated washing with water fails to remove any appreciable amount of vitamine from it. Furthermore the vitamine-activated fuller's earth retains its active vitamine properties for at least a period of two years. Large amounts of the vitamine can be accumulated in this way and when fed to animals or infants the vitamine is liberated physiologically and produces the usual effects of a vitamine extract. When this discovery was made the discoverers thought that in the fuller's earth they had a means for arriving at the identification of the substance but attempts to recover the vitamine from the earth developed unexpected difficulties. Acids were found to split it off but they also split off aluminium compounds and left an impure mixture little better than the original extract for study. By using a dilute alkali they were able to obtain the substance without aluminium contaminations and by this method they actually obtained some microscopic fibrous needles which were curative. These needles however on recrystallization resulted in the production of a compound contaminated with adenin or rather in adenin contaminated with the curative substance and on standing for some time the adenin crystals gradually lost their curative power. These results led Williams to suggest an interesting hypothesis. By experiments conducted with the hydroxy- pyridines he believed that he had demonstrated a relation between tautomerism or changed space relations in these sort of substances and curative properties. He states his view as follows:
The vitamines contain one or more groups of atoms constituting nuclei in which the curative properties are resident. In a free state these nuclei possess the vitamine activity but under ordinary conditions are spontaneously transformed into isomers which do not possess an antineuritic power. The complementary substances or substituent groups with which these nuclei are more or less firmly combined in nature exert a stabilizing and perhaps otherwise favorable influence on the curative nucleus, but do not themselves possess the vitamine type of physiological potency. Accordingly it is believed that while partial cleavage of the vitamines may result only in a modification of their physiological properties, by certain means disruption may go so far as to effect a complete separation of nucleus and stabilizer, and if it does so will be followed by a loss of curative power due to isomerism. The basis for the assumption that an isomerization constitutes the final and physiologically most significant step in the inactivation of a vitamine is found in the studies of synthetic antineuritic products. This assumption is supported by evidence … of the existence of such isomerism in the crystalline antineuritic substances obtainable from brewer's yeast.
According to this view the active adenin obtained was not a contamination but an inactive isomer of the active substance. The hydroxy-betaines which Williams prepared in defense of his theory have been repeatedly tested but have in general failed to confirm his view which stands today as an interesting suggestion but without confirmatory evidence. Other attempts by these authors to fraction their alkaline extract of fuller's earth have been unsuccessful. It is of course well known that alkali acts upon the vitamine destructively. On this account the authors of this method operate as rapidly as possible and restore the alkali extract to a neutral or acid medium quickly. The aqueous extract obtained from the earth in this manner has been shown by Seidell to possess only about one-half of the vitamine originally present in the solid but the vitamine in it is shown to be fairly stable. Seidell has not yet determined how long it remains so. Attempts to recover the vitamine from such aqueous solutions have however totally failed to date. To quote Seidell from a recent publication:
By careful evaporation of the solution the products successively obtained show more or less activity by physiological tests but in no case does the resulting material possess the appearance or character which a pure product would be expected to show. Solvents such as benzene, ethylacetate and chloroform fail to effect a separation of active from inactive material. In all fractioning operations the vitamine tends to distribute itself between the fractious rather than to become concentrated in one or the other.
The difficulties encountered by Seidell in this fractioning study have led him to adopt Walsche's idea that vitamines are of the nature of enzymes and hence present all the difficulties of identification and isolation of those substances.
During 1920 Myers and Voegtlin attacked the problem. They have made a discovery that is useful as a separatory process. This that the "B" vitamine is not only soluble in water, but also olive oil and in oleic acid. By shaking an autolysed yeast extract with those solvents in the proportion of 1 cc. of solvent to which 4 cc. of extract the vitamine passes into the oil. When this activated oil is filtered and taken up with eight to ten volumes of ether it in possible to concentrate the ether extract in vacuo and extract from it with 0.1 per cent. HCl an active fraction. Aside from this observation however nothing further has been reported and the possibility of this method of concentration remains yet to be exploited. They did report other methods of fractioning which yielded crystals but failed to produce a pure active substance. Those results add nothing to what has been previously reported except a new method of fractioning and the elimination of the following substances as contributing nothing to vitamine activity (purines, histidine, proteins and albumoses). The crystals they obtained wore contaminated with histamine.
The World War has prevented full knowledge of the work of the German investigators but nothing has appeared that indicates any progress in this field with the exception of a paper by Aberhalden and Schaumann and some work by Hofmeister. The Aberhalden paper yields no new data of any moment and no active substances in pure condition are reported. The reports from Hofmeister are to the effect that he has isolated a very active solution belonging to the pyrimidine series. It yields a crystalline hydrochloride and double salt with gold chloride and has given it the formula C_5H_11NO_2.