CHAPTER 1: Historical Context of Vitamin B

From ancient times, people were aware of the presence of a specific type of disease, beriberi, which affected people mainly in the East and South Asia. As early as 2600 bc, a Chinese account described that beriberi was caused by long-term rice eating but could be prevented by taking rice bran simultaneously. This finding was not recognized in modern medicine. However, in a similar context, the antiscorbutic effect of citrus fruits was already empirically known in the 17th century when James Lind of the Royal Navy systematically carried out experiments to demonstrate the beneficial effect of orange and lemon. Although he never thought that citrus juice is the only solution to scurvy, the surgeons of the Royal Navy were by experience convinced of the efficacy of citrus juice even if the reason was unknown. In the early 1880s, the Surgeon General of the Imperial Japanese Navy, Kanehiro Takagi, noticed that beriberi was common among crews and lower rank officers but not among officers who ate a Western-style diet. He considered that the low-protein diet was the cause of beriberi and performed an experiment in 1884 in which crews of a battleship were given bread and meat during a nine-month mission. Only 16 out of 333 developed beriberi and no one died, compared with a similar mission in the previous year, in which 169 out of 376 developed the disease and 25 died. The Japanese Navy adopted the Western-style diet (bread was later replaced by rice cooked with barley) and eliminated beriberi after 1885. This important lesson, however, was dismissed by the surgeons of the Imperial Japanese Army, who considered beriberi as an infectious disease and criticized Takagi for insufficiency of data and lack of theory that could explain the results. As a result, the army lost 28 000 soldiers due to beriberi in the Russo-Japanese War.

Independently of this, in 1887 a Dutch physician Christiaan Eijkman found an animal model for beriberi when he was in Dutch East Indies (Indonesia). As a former student of Robert Koch, he had been trying to isolate and infect chickens with ‘beriberi bacteria’. As he had expected, all the chickens developed polyneuritis gallinarum, the bird counterpart of human beriberi, but soon recovered spontaneously. He noticed that chickens were sick while they were fed polished rice, but recovered after the diet was inadvertently changed to unpolished rice. He conducted detailed experiments to exclude other possibilities, such as that polished rice promotes the growth of bacteria during storage, and concluded around 1895 that what made the difference was the presence or absence of the silver layer of rice. He hypothesized that rice contains some toxin and a substance in the silver layer—he called this the anti-beriberi factor—neutralizes its virulence (Eijkman 1897). In 1901, Gerrit Grijns, who was an assistant to Eijkman, observed that chickens fed on raw meat did not develop polyneuritis whereas those exclusively fed on meat that had been heated long enough at 120°C developed the disease. This clearly showed that neither polyneuritis nor beriberi was caused by some substances in rice. From this result, and by taking Takagi's observations into account, he interpreted Eijkman's results in another way: polyneuritis is a deficiency syndrome of a still unknown substance that is essential for life and destroyed by moist heat (Grijns 1901). His theory was later adopted by Eijkman in 1906.

In the 19th century, chemists knew that food contains carbohydrates, proteins and lipids. They imagined that it could be possible to make artificial food by properly mixing these nutrients. Ironically, it became a practical issue to test this idea during the Siege of Paris in 1870. A French chemist Jean Dumas made an artificial milk, but infants fed the milk did not survive. This observation attracted the attention of Gustav von Bunge who, believing that minerals are critical for preparing efficient artificial food, ordered his student Nicholas Lunin to study the effect of salt content on mice maintained on artificial food. The mice, however, could not survive for very long, irrespective of the salt content of the food. Lunin concluded that ‘a natural food such as milk must therefore contain besides these known principal ingredients small quantities of unknown substances essential to life’ (Lunin 1881). Unfortunately, his view was not supported by Bunge and the true significance of this report was not recognized in the scientific community, partly due to the title (‘On the importance of inorganic salts in the diet of the animal’) reflecting the thought of Bunge's school. In 1905, Cornelius Pekelharing, who had been a predecessor of Eijkman in Java, carried out similar experiments and reached the same conclusion. His report was published in Dutch and was not circulated widely. It was only in 1912 that the idea of ‘unknown substances essential to life’ was made widely known by the famous paper of Sir Frederick Hopkins (Hopkins 1912). Hopkins was studying the nutritional effect of tryptophan, which he had discovered in 1901, and noticed that animals fed with the tryptophan-deficient ‘synthetic’ diet could not live long, even if tryptophan was added to the diet. Perhaps without knowing the works of Lunin and Pekelharing (Hopkins 1929), he proposed the notion of ‘deficiency diseases’—beriberi, scurvy and rickets as distinct entities. He further forecast that there were many other nutritional errors dependent on unknown dietary factors.

The next step was, undoubtedly, to isolate these substances. Umetaro Suzuki was probably the first to extract the anti-beriberi factor in a concentrated form from rice bran. He presented his results, in December 1910, at the meeting of the Tokyo Chemical Society, and published them in the January issue of the Society's journal in 1911. Because it was written in Japanese, his work was not known outside Japan. In December 1911, Casimir Funk published a paper describing the extraction of the factor using a similar method to Suzuki (Funk 1911). In the following year, Suzuki published in German a paper combining his collective works (Suzuki et al. 1912). At this point, he was aware that the material he had previously extracted was largely nicotinic acid and that this had misled him to name the compound ‘aberic acid’. He could precipitate the effective component with picric acid and succeeded in concentrating the active substance. He then corrected the name to ‘oryzanin’ (from the Latin oryza meaning rice). Funk published a paper in the same year describing the concept of ‘deficiency disease’. Although it was not different from that proposed by Hopkins, the name vitamine (‘vital amine’) he used in his paper was soon adopted by researchers in this field. However, as Funk himself admitted these essential substances (a fat-soluble substance necessary for preventing xerophthalmia had been found by Elmer McCollum) did not need to be organic bases. Therefore, in 1920 Jack Drummond recommended dropping the final ‘e’ of ‘vitamine’. He also proposed discontinuing the use of the term ‘fat-soluble A’ and ‘water-soluble B’ suggested by McCollum as the unidentified substances necessary for growth, and instead calling these substances ‘vitamin A’, ‘vitamin B’, etc., until their true structures were determined.

The pure crystalline vitamin was obtained in Java in 1927 by Barend Jansen and William Donath, who were both students of Eijkman (Jansen and Donath 1927). Robert Williams developed an effective method to isolate vitamin B from rice bran, proposed its structure, and confirmed it by synthesizing the compound (Williams and Cline 1936). The compound, then already given the name vitamin B₁, had a thiazole ring and therefore was named ‘thiamine’ (Figure 1.1). (As with ‘vitamin’, ‘thiamine’ later lost its ‘e’ to ‘thiamin’, although the spelling ‘thiamine’ is still frequently used.)

Although there had been a belief that the ‘water-soluble B’ necessary for growth was identical with the antineuritic vitamin, people became aware of the possibility that vitamin B is not a single entity. In 1927, following the crystallization of the antineuritic vitamin, the British Committee on Accessory Food Factors distinguished the heat-labile and heat-stable components of vitamin B, and named the former B₁ and the latter B₂.

In 1933 Richard Kuhn, Paul György and Theodor Wagner-Jauregg isolated from egg white a yellow pigment having vitamin B₂ activity, which they crystallized and named ‘ovo-flavin’ (Kuhn et al. 1933). They noticed that the absorption spectra of ovo-flavin resembled those of the ‘yellow enzyme’ isolated by Warburg and Christian (see below). In the same year they also crystallized lacto-flavin, which had been reported by P. Ellinger and W. Koschara, and showed it was identical with ovo-flavin. Alkaline hydrolysis of lumiflavin (C₁₃H₁₂N₄O₂), the photolysis product of lacto-flavin (C₁₇H₂₀N₄O₆) in alkaline medium, afforded urea and a compound having the chemical formula C₁₂H₁₂N₂O₃, which yielded on thermal decomposition CO₂ and C₁₁H₁₂N₂O. These reactions were similar to those of alloxazine, which gave urea, CO₂ and 2-hydroxyquinoxaline-3-carboxylic acid (C₈H₆N₂O), suggesting that lumiflavin is a trimethyl derivative of alloxazine. Combining the observation by E. Holiday and K. Stern of the spectral similarities between lacto-flavin and alloxazine and its derivatives (Holiday and Stern 1934), Kuhn proposed the structure of lumiflavin to be 7,8,10-trimethylisoalloxazine (Kuhn and Rudy 1934). In 1935, Kuhn's group (Kuhn et al. 1935) and Karrer's group (Karrer et al. 1935) synthesized independently 7,8-dimethyl-10-d-1′-ribitylisoalloxazine, and confirmed its identity with lacto-flavin. Once the structure (Figure 1.2) was finally determined, the vitamin was called riboflavin thereafter.

The 1930s saw the dawn of coenzyme research. In 1932, Otto Warburg obtained a yellow enzyme from yeast and showed that the yellow dye reversibly underwent oxidation and reduction while conducting the oxidation of glucose 6-phosphate (Warburg and Christian 1932). Hugo Theorell, working in Warburg's laboratory, crystallized the enzyme and showed that the pigment could be reversibly removed from the enzyme protein and the enzyme which lost the pigment was inactive (Theorell 1935). This was the first discovery of a ‘coenzyme’. Theorell also determined the correct structure of the coenzyme, the phosphate ester of riboflavin, flavin mononucleotide (FMN) (Threorell 1937). The more abundant and complex form of the coenzyme was found by Warburg and Christian in D-amino acid oxidase, and was determined to be flavin adenine dinucleotide (FAD) (Warburg and Christian 1938).

Dating back to 1911, Neuberg had discovered that bacteria and plants have an enzyme that catalyses the decarboxylation of pyruvate to acetaldehyde and CO₂. He named the enzyme ‘carboxylase’, which in today's nomenclature refers to an enzyme undergoing carboxylation but at that time meant decarboxylase. In 1932, Ernst Anhagen observed that yeast ‘carboxylase’ lost activity when treated with alkali (Auhagen 1932). The activity was restored by the addition of a heated solution of yeast extract. He then speculated that the ‘carboxylase’ contains a non-proteinous low-molecular weight compound, named ‘cocarboxylase’. At the same time, R. Peters and colleagues observed that the brain extract of pigeons deficient of thiamin showed a decreased rate of lactate degradation, but that the rate was restored by the addition of ‘concentrated vitamin B₁’ (crystalline thiamin was expensive) to the extract (Meiklejohn et al. 1932). Based on these observations, Lohman and Schuster isolated cocarboxylase from yeast and proposed its structure to be thiamin diphosphate (Lohman and Schuster 1937).

The history of the pursuit for vitamin B₁ and B₂ paved the way for the research on other vitamins. Pellagra is a disease which was epidemic in southern Europe in the 19th century and in the southern USA in the early 20th century. In 1915, Joseph Goldberger reported that pellagra was not an infectious disease and could be ascribed to maize eating (Goldberger and Wheeler 1915). This view was consistent with the findings of Willcock and Hopkins in 1906 that mice fed with maize protein, which is low in tryptophan, as the sole protein source could not grow (Willcock and Hopkins 1906). Thus Goldberger suspected that tryptophan deficiency was the cause of pellagra. Although this is known to be correct today, the history of the pellagra-preventing factor took a curious path. In 1920, Carl Voegtlin and associates showed that pellagra could be cured by administering yeast extract, suggesting that vitamin B was the pellagra-preventing factor (Voegtlin et al. 1920). However, with the isolation of vitamin B₁ and B₂, it was soon acknowledged that neither of these vitamins had pellagra-preventing activity.

Independently of the pursuit of the pellagra-preventing factor, several studies on yeast fermentation provided for related fields of research. In 1906, Arthur Harden and William John Young found that fermentation by yeast juice required both a ‘heat-labile and nondialysable fraction’ and a ‘heat-stable and dialysable fraction’, the latter of which was termed ‘co-ferment’. In the early 1930s, co-ferment was resolved into several components. One was shown by Otto Meyerhof and colleagues as ATP (Meyerhof et al. 1931). Warburg and colleagues showed that nicotinic acid was present in co-ferment in the form of niconitic acid amide conjugated with two pentoses, three phosphoric acids and an adenine (Warburg et al. 1935). This is known today as NADP (Figure 1.3). Shortly thereafter, another nicotinic acid derivative with two phosphoric acids, now known as NAD (Figure 1.3), was discovered by Warburg and Christian and independently by a rival group led by Hans von Euler-Chelpin, a notable student of Warburg's father, Emil Warburg (Schlenk and von Euler 1936; Warburg and Christian 1936).

About the same time, in 1937, Bert C.J.G. Knight found that nicotinic acid is a Staphylococcus growth factor. Inspired probably by these findings, Elvehjem and associates were able to cure black-tongue disease, a dog counterpart of pellagra, by administering nicotinic acid (Elvehjem et al. 1937). There remained, however, a mystery: diet that induced black tongue in dogs was richer in nicotinic acid or its amide than normal milk. In 1945, Elvehjem's group showed that tryptophan can fully substitute the vitamin action of nicotinic acid, indicating that nicotinic acid is synthesized from tryptophan (Krehl et al. 1945). Henderson and Ramasarma revealed that quinolinic acid is formed from 3-hydroxyanthranilic acid, a metabolite of tryptophan (Henderson and Ramasarma 1949). However, they were unable to explain how quinolic acid is converted to nicotinic acid. The solution to this problem was given in 1963 by Nishizuka and Hayaishi, who showed that quinolinic acid first reacts with 5-phosphoribosyl-1-pyrophosphate (PRPP) to form a nucleotide before being decarboxylated to form nicotinic acid mononucleotide (Andreoli et al. 1963). Thus contrary to the earlier belief, the direct biosynthetic product was not nicotinic acid but its phosphoribosylated form.

The vitamins discovered in the 1930s were soon found to have profound physiological effects on microorganisms. In an opposite manner, pantothenic acid was initially discovered in microorganisms and later in mammals. Since the beginning of the 20th century, it was known that yeasts require “bios”, which is present in malt, yeast extract, etc., for growth. Bios was later resolved into I, IIa and IIb. Bios I was identified as meso-inositol. Roger Williams found in 1933 the ubiquitous occurrence of bios IIa in the tissues of many organisms; he named it pantothenic acid and isolated it in 1938 (Williams et al. 1938). Owing to the feature of ‘universal biological occurrence’, pantothenic acid was recognized as a vitamin (vitamin B₅) found earlier than pyridoxine (B₆). In the following year, T. Jukes and Elvehjem's group independently announced that deficiency in pantothenic acid caused skin lesion in chickens (Jukes 1939; Woolley et al. 1939).

Analogously to the ‘artificial diet’ for humans, microbiologists attempted to prepare a ‘purified medium’ for microorganisms comprising all the necessary nutrients in chemically pure form. In 1939, Esmond Snell observed that, when riboflavin was added to the medium for Lactobacillus casei, in which the peptone was treated with alkali in advance to destroy the inherent riboflavin, no growth was observed. This indicated that some nutritional factors were removed by the alkali treatment. In 1939 he purified the factor 1000-fold from liver extract and showed it to be identical with pantothenic acid. Based on this, he developed a microbiological assay of riboflavin (Snell and Strong 1939).

In 1936, F. Kögl and B. Tönnis succeeded in crystallizing 4 mg of bios IIb from 1000 duck egg yolks and named it biotin (Kögl and Tönnis 1936). The chemical structure (Figure 1.4) was determined by V. du Vigneaud and associates in 1942, and synthesized by Folkers’ group in 1945. Paul György had been studying egg white injury and in 1939 partially purified ‘vitamin H’, which prevents the disease in rats, from bovine liver. The identity of vitamin H and biotin was shown in 1940 by a collaboration of György and du Vigneaud. In 1940, Snell developed a yeast assay system for biotin and using this Snell and Williams isolated the protein in egg white that tightly binds biotin and causes ‘egg white injury’ (Eakin et al. 1940). This was named ‘avidin’ because of its peculiar biotin-binding capacity.

Through these investigations the idea that microorganisms and animals share many vitamins in common became gradually accepted and the microbial bioassay system of vitamins became an essential tool in vitamin research. We will appreciate the significance of the method in the following sections.

In 1934, György found that there is a specific type of skin lesions in rat which was protected by an unidentified ‘vitamin B₆’. In 1938, Samuel Lepkovsky reported its crystallization. Slightly later, four other groups, Keresztesy and Stevens, György, Kuhn and Wendt, and Ichiba and Michi, independently announced the crystallization of vitamin B₆ from various sources (summarized in György 1964; György admits Lepkovsky, Keresztesy and he were in contact with each other just before publication). The structure of vitamin B₆ was solved independently by S.A. Harris and Folkers and Kuhn and associates in 1939 (György 1964). The compound, first named ‘adermin’ for its dermatitis-preventing activity, was soon renamed ‘pyridoxine’ derived from its chemical structure (Figure 1.5).

While developing a microbiological assay system for pyridoxine, Snell noticed that the content of ‘pyridoxine’ in rat tissues as determined by growth of Streptococcus faecalis was several thousand times higher than those obtained by yeast growth, rat growth or colorimetric methods. Another important finding was that when media supplemented with pyridoxine was heat-sterilized, progressively smaller amounts of pyridoxine were required by S. faecalis as the autoclave period was increased. Snell proposed that a substance, which he called ‘pseudopyridoxine’, was formed from pyridoxine by autoclaving and that S. faecalis was much more sensitive to pseudopyridoxine than to pyridoxine. Pseudopyridoxine was also found to be formed from pyridoxine in vivo by human and rat bodies. In collaboration with Folkers, Snell showed that pseudopyridoxine was a mixture of the 4-aldehyde and 4′-amine derivatives of pyridoxine, which were named pyridoxal and pyridoxamine, respectively (Figure 1.5; Harris et al. 1944; Snell 1944). Using the differential effect of the three forms of vitamin B₆ on S. faecalis (pyridoxal and pyridoxamine are active), Lactobacillus casei (pyridoxal is active) and Saccharomyces carlsbergensis (all three forms are active), Snell developed a microbial quantification system to quantify the three forms separately.

The conversion of pyridoxal to pyridoxamine in heat-sterilized media was confirmed to be due to the reaction of pyridoxal with amino acids (Snell 1945). This was the first discovery of nonenzymatic transamination. The transamination reaction in animal tissues was first discovered in 1937 by Alexander Braunstein and associates as an amino group transfer between glutamate and alanine in pigeon muscle extract (Braunstein 1939). Irwin Gunsalus and associates showed that the tyrosine decarboxylase activity of S. faecalis was slightly stimulated by addition of pyridoxal but markedly stimulated by that of pyridoxal and ATP (Gunsalus et al. 1944). This led to the finding that pyridoxal phosphate (Figure 1.5) is the coenzyme. A detailed mechanistic study of pyridoxal was carried out in metal-ion catalysed model systems and a general mechanism of pyridoxal-catalysed diverse reactions, transamination, decarboxylation, aldol cleavage, β-replacement/elimination, γ-replacement/elimination, etc. was proposed (Metzler et al. 1954). These mechanisms were later found to operate in pyridoxal 5′-phosphate-dependent enzymes (Hayashi 1995; Eliot and Kirsch 2004).

In Bombay (now Mumbai) in 1930, Lucy Wills observed patients complicated with an unusual macrocytic anaemia (Wills 1931). The symptoms of the anaemia resembled those of pernicious anaemia, but unlike pernicious anaemia, it lacked neurological complications and quickly responded to yeast extract, which was then already known to be rich in ‘vitamin B complex’. Liver was also efficacious when administered orally, but not when injected as liver extract, which was a promising remedy for pernicious anaemia. This indicated that the curative factor was removed by preparation of liver extract. Wills could reproduce the anaemia in monkeys fed with the local food deficient in vitamin B complex. At the same time, similar observations were reported by Paul L. Day, who named the protective factor in brewer's yeast ‘vitamin M’, but could not continue the research owing to the restrictions that come with an assay system using monkeys. Albert Hogan developed an assay system using chickens, and with the help of a large research team, could obtain crystals of the vitamin, which was alternatively named ‘vitamin B_c’ (Pfiffner et al. 1943).

Snell was at that time working on nutrition of Lactobaccillus casei. He found a factor necessary for the growth of L. casei in spinach, which was easily available in Austin, Texas. He and colleagues processed 4 tons of spinach and isolated the factor (Mitchell et al. 1941) and named the compound ‘folic acid’ (from the Latin folium meaning leaf). The ability of folic acid to prevent chicken anaemia was confirmed in 1942 by Elvehjem's group. On the other hand, E. L. Robert Stokstad crystallized the growth factor for L. casei from liver and showed its identity with vitamin B_c but not with Snell's ‘folic acid’ (Stokstad 1943). Later, he and colleagues discovered a bacterium that produced large amounts of a compound with vitamin B_c activity. Hydrolysis of the compound yielded a pteridine derivative, p-aminobenzoic acid, and glutamic acid. Reconstitution studies showed that the nature of vitamin B_c was pteroylglutamic acid ((Figure 1.6) Waller et al. 1948). Elucidation of the structure enabled explanation for the action of sulfa drugs; they act as analogues for p-aminobenzoic acid. ‘Folic acid’ present in plants was later found to be pteroylpolyglutamic acid, explaining the difference in the effect of Stokstad's and Snell's preparations on growth of L. casei. However, the name ‘folic acid’ became used as a term that encompasses both pteroylglutamic acid and pteroylpolyglutamic acid.

Pernicious anaemia was first described by James Combe in 1824. Thomas Addison described in detail the disease in 1849, although he posited some connection with the adrenal gland. No significant progress had been made until in 1897, when F. Martius and O. von Lubarsch reported an association of pernicious anaemia with achlorhydria. Thereafter, more attention was paid to the role of gastric function in the pathogenesis of the anaemia.

In 1925, Whipple found by chance that severe anaemia in dogs induced by exsanguination could be cured by administering liver. This observation was immediately applied by Minot and Murphy to the treatment of pernicious anaemia and a remarkable response was observed for patients eating large amounts of liver (Minot et al. 1928). Subsequently, William Castle found that ground beef partially digested in the gastric juice of healthy men was effective as a remedy for the anaemia. However, when beef or gastric juice was given alone, or when given successively, no remission was observed (Castle et al. 1930). He then postulated that haematopoiesis requires simultaneous administration of an ‘extrinsic factor’ present in meat and liver, and an ‘intrinsic factor’ secreted by the stomach. He also showed that the intrinsic factor was neither HCl nor pepsin, and was heat-labile.

A tough quest for the two factors began. Owing to the lack of appropriate animal models, isolation of the vitamin was difficult. However, after Mary Shorb showed that the vitamin, now called vitamin B₁₂, was also a growth factor for Lactobacillus lactis Dorner, a microbial bioassay system was developed. In 1948, Folkers’ group succeeded in crystallizing vitamin B₁₂ (Rickes et al. 1948), and showed it to be a cyano complex of Co³⁺. In today's view, it is considered that cyanocobalamin formed with internal cyanide source was preferentially obtained since it is the most stable vitamin B₁₂. The complete structure of vitamin B₁₂ (Figure 1.6) was finally solved in 1955 by Dorothy Hodgkin using X-ray crystallography (summarized in Hodgkin et al. 1956). Microbial bioassay system was, however, inappropriate for detecting the intrinsic factor activity. It was only in 1972 when the intrinsic factor was purified using the newly developed purification method, affinity chromatography, with vitamin B₁₂-Sepharose (Allen and Majerus 1972).

The function of vitamin B₁₂ was first recognized in 1958 when Horace Barker found that it acts as a coenzyme for interconversion between L-glutamate and 3-methylaspartate (Barker et al. 1958). The structure of the coenzyme, adenosylcobalamin, was solved again by Hodgkin in 1961.

The finding in 1946 by T. Spies and colleagues that thymine can substitute the functions of folic acid and vitamin B₁₂ led to the understanding that both folic acid and vitamin B₁₂ are involved in methyl transfer reactions (Vilter et al. 1950). Donaldson and Keresztesy showed that folic acid can exist in various forms with one-carbon group attached. The coenzyme methylcobalamin was discovered in 1964 (Lindstrand 1964).

The history of B vitamin research, like that of other vitamins, began with the investigations of the aetiology of various diseases that were related to dietetic problems. However, in the 19th century, the glorious works of Robert Koch and others might have induced people to think that most causes of diseases could be ascribed to pathogens such as bacteria and toxins. It was considered that nutritional problems were largely solved by the discovery of three major nutrients and minerals, and the remaining problem was to find an ideal ratio of these nutrients. Even after the discovery of an essential factor to prevent beriberi, time was required to make Eijkman start to believe that this was not a detoxificant but a factor essential for life. This gives us a lesson: as we all know, there can be no theory without experimental evidence—but we often forget this especially when our minds are caught by some great ideas or concepts.

Why are B vitamins necessary? If we consider the action of B vitamins as coenzymes, the answer is clear. There is no electrophilic functional group in the 20 amino acids of proteins. Coenzymes are therefore required for enzymes to carry out electrophilic and radical catalysis. Thiamin diphosphate is by itself a nucleophile, but it generates an electrophilic centre once it reacts with a carbonyl compound. The exceptions are CoA and tetrahydrofolate, which are more properly regarded as substrates rather than catalysts. However, since they are recycled, they are not usual substrates and share ‘reusability’ as a common property with other catalytic coenzymes.

Today, the catalytic mechanisms of coenzymes are understood in detail. However, only few, namely those of folic acid and vitamin B₁₂, can explain the symptoms of their deficiency. With regard to beriberi, our knowledge has not much improved since the time of Takagi and Eijkman. Certainly, B vitamins have functions other than those as coenzymes. The discovery of NAD-dependent ADP ribosylation was, in this sense, a breakthrough. Interestingly, the electrophilic nature of NAD is again exploited here. However, it may be appropriate to think that we should not be dazzled by the fantastic chemical properties of vitamins, for we learnt a lesson that too much adherence to theory caused the tragedy of unthinkable loss of life to beriberi in the early 20th century.

• This chapter focuses on the historical context of B vitamins.

• Beriberi was known to be related to diet in ancient China but was not recognized in early modern medicine.

• There was a debate over the cause of beriberi in Japan in the 19th century, causing large numbers of victims of the disease comparable to those of wars.

• The prevailing idea in the 19th century was that diseases were caused by extraneous pathogens such as bacteria and toxins.

• It was not until the early 20th century that the notion of essential micronutrients for life, vitamins, was established.

• Almost all the B vitamins were discovered in the 1920s to 1940s.

• A microbiological assay system contributed much to the isolation of vitamins.

• B vitamins were found to function as coenzymes in enzymatic reactions and metabolism.

• Only a part of the symptoms of B vitamin deficiency can be explained by coenzyme functions of vitamins.

• Exploring the functions of B vitamins other than as coenzymes that account for their deficiency symptoms is a goal for future work.

1. In the Russo-Japanese War, the death toll of the Imperial Japanese Army was 48 400 from battle and 37 200 from disease. Among the latter, 27 800 died of beriberi. The total number of beriberi patients in the Army reached 250 000. On the other hand, the Imperial Russian Army suffered from scurvy, which led to the fall of Port Arthur.

2. The efficacy of barley-blended rice against beriberi was known to the chief medical officers of many army divisions. However, their request to adopt barley-blended rice as army provision was neglected by Rintaro (Ogai) Mori, a famous novelist and the head of the Second Army Medical Corps, Tadanori Ishiguro, a former director already retired but still influential on the Medical Department of the Japanese Army, and Tanemichi Aoyama, Professor of Tokyo Imperial University.

3. The reason for the Japanese army not having adopted barley-blended rice is a complicated combination of several factors.

4. Some critics ascribe the reason to the aggressive and stubborn personality of Rintaro Mori, who did not admit that ‘beriberi bacillus theory’ was wrong until his death, although he seemed to be aware of it. They blame Mori as ‘the officer who killed more Japanese soldiers than any Russian officer’.

5. Others point out that it was the revenge of Ishiguro who was forced to resign after the Sino-Japanese War (1895) taking the responsibility for the beriberi pandemic among the soldiers, and the pride of Aoyama who considered himself as the authority of bacteriology in Japan.

6. There was also a logistic problem: production of barley was not enough to support soldiers whose number was several tens larger than that of sailors. Furthermore, barley is not as stable as polished rice on storage.

7. Most of the soldiers of the Japanese army were sons of poor peasants and eating polished rice was their desire even at the cost of their lives. It was probably the ‘mercy’ of army officers to let their soldiers eat polished rice as much as they wanted before leaving for the battlefield. Even in the Japanese navy, there was recurrence of beriberi in the period around World War II, because barley was discarded by cooks who disliked its flavour.

FAD

flavin adenine dinucleotide

FMN

flavin mononucleotide

NAD

nicotinamide adenine dinucleotide

NADP

nicotinamide adenine dinucleotide phosphate