In the early 1700s in England 'nothing was so much feared or talk'd of as Rickets among Children'. We now know that this softening of the bones, is caused by a deficiency of vitamin D.
Rickets was common in large built up areas, where the climate was cold, dull and wet, pointing to a lack of sunlight as a possible culprit
- Identification of the structure of vitamin D led to its large-scale synthesis and fortification in a wide range of foods, especially milk products. Rickets became rare, until recently
An English medical student at Leyden University, Daniel Whistler, was the first to describe rickets in 1645. His brief pamphlet, perhaps based more on hearsay than direct observation, claimed it was a disease of English boys. Five years later Francis Glisson, a physician and regius professor of medicine at Cambridge, published his Treatise of the rickets and takes credit for the first full description of this disease.
However, though claimed as a new disease by Glisson, its symptoms were described by ancient Chinese, classical Greek and Roman physicians. Nonetheless rickets, once rare, became increasingly common, particularly in the Black Country and Glasgow in the 1700s. By 1850 medical opinion variously blamed heredity, early weaning, improper diets, dirty skin, impure air, and a northern climate. While there was a grain of truth in some of these, how they caused rickets remained a mystery.
The culprit and the cure?
Dr Schoepf Merei, a paediatrician working in Manchester in the 1850s, found rickets was common in large conurbations and near the coalfields, but rare in small settlements and agricultural areas. The medical geographer August Hirsch extended Merei's work and concluded that rickets was a disease of cold, wet climates, such as in Holland, the UK, Germany, and Northern Italy, but was absent in tropical and sub-tropical climates. Foul air, it appeared, was the culprit. Or was it?
In the late 1880s Theobald Palm, a Cumbrian GP, reflecting on the absence of rickets in Japan and in sunny climes from China to North Africa, suggested the problem was a lack of sunlight - after all, if it was so important for growing plants, could it not also affect growing children?2
Others were convinced that rickets was caused by something in food. In 1842 George Budd, professor of medicine at King's College, London, suggested that accessory food factors were essential components in food and their omission led to deficiency states. He gave scurvy (caused by a deficiency of vitamin C), xerophthalmia (caused by deficiency of vitamin A) and rickets as examples but, influenced by the chemist Justus von Liebig, ascribed rickets to insufficient calcium intake.3
In the early 19th century cod liver oil, which had been used towards the end of the 18th century to treat rheumatism, was being hailed as a potential treatment for rickets in Holland and later in Germany and France.4
In 1881 Nicolai Lunin, working in the Rhineland city of Basel, unsuccessfully attempted to raise mice on artificial diets of the basic components of milk (protein, carbohydrates, fats, water and salt), as did Utrecht professor of hygiene Cornelius Pekelharing in 1905. Both concluded that animals needed some additional factor, present in milk or other 'whole food', to grow properly. Others ignored their work, arguing that animals didn't thrive because of disinterest in such tasteless diets.
In 1921, however, Baltimore nutritionist Elmer McCollum noticed he could promote bone growth in rats by either exposing the animals to uv light, or supplementing their diets with cod liver oil. This, he reasoned, contained a previously unnoticed accessory food substance, which he named 'vitamin D', continuing the sequence of letters proposed the previous year for such 'accessory food substances'.5 Exactly how the uv light worked, however, wasn't clear. Even if the rats' cages (glass jars) were irradiated without the rats inside, this seemed to prevent them developing rickets. Was irradiated air enough to prevent rickets?
The mystery was solved three years later, in 1924, when McCollum realised that the jars were not entirely empty when irradiated, but contained food residues, bedding and faeces. The radiation transformed these materials in some way, which the rats ate.6
But what exactly was 'vitamin D'? At that time, to determine the structure of a mystery substance, chemists would isolate a reasonable amount of it, say 1-10 g, and then degrade it (chiefly by oxidation and ozonolysis) until the substance broke down into fragments of known structure (which could then be identified by qualitative tests, mixed melting point determinations and, rarely, optical rotation measurements). The overall jigsaw of the complete parent molecule could then be proposed by linking these fragments in a chemically plausible fashion. The 'icing on the cake' was to synthesise the proposed structure in the lab and see if it matched, absolutely, that supplied by Nature.
The problem with vitamin D was that it occurred in such tiny amounts. Milk, a much-revered source of the vitamin, contained only ca 0.3μg per litre, so isolating 1 g of the vitamin would require over three million litres of milk, even if such isolation were possible.
By 1924 there were two classes of fats and oils that were being used to treat rickets. The first included fish oils and fats obtained from milk, and cured the condition without the need for uv irradiation. The second group was ineffective until irradiated with uv light. The latter included rat brain tissue which is rich in cholesterol and, for a while chemists thought that cholesterol was the precursor to vitamin D, the 'pro-vitamin'.
In 1927, however, Ian Heilbron, professor of organic chemistry at Liverpool University, challenged this view. Using uv spectroscopy, which was just beginning to be used in diagnostic work, he compared the spectrum of allegedly pure 'rat brain cholesterol' with that from other sources and found the rat brain material had additional absorbance peaks, indicating two or three conjugated double bonds, as well as the single double bond of cholesterol itself. These three peaks (at 269, 280 and 293 nm) became valuable markers of success in the later separation of the pro-vitamin from its tightly bound cholesterol partner.
The year before Heilbron's breakthrough, in 1926, Adolf Windaus (at the University of Göttingen) had used high-vacuum distillation and selective adsorption on charcoal to isolate vitamin D substantially free from cholesterol, but in such tiny amounts that it ruled out resolving the structure by the degradative process. He presumed that the pro-vitamin and cholesterol must have near-identical structures, preventing efficient separation, so decided on a different approach. Acknowledging Heilbron's work, Windaus went on to examine every sterol he could find to see if:
- its uv spectrum contained the three Heilbron peaks;
- it had anti-rickets properties after uv irradiation.
Windaus examined 30 substances and found one, ergosterol, which matched both criteria. This was the pro-vitamin D, which uv irradiation converted to the vitamin itself. If the structure of ergosterol could be established, it would be an important step towards determining the structure of vitamin D. Fortunately, Nature provides far more ergosterol that it does vitamin D (dried brewers' yeast, for instance, contains up to 1.4 per cent ergosterol). This key decision to study ergosterol as a precursor to the structural elucidation of vitamin D allowed him, and his friendly competitor, Ian Heilbron, to use existing techniques to characterise ergosterol and thus, eventually, to gain insight into the structure of vitamin D itself.
The ergosterol link
Ergosterol had been isolated in 1889 by the Parisian apothecary Charles Tanret. His analysis (84.75 per cent C; 11.17 per cent H, with the remainder presumed to be oxygen) pointed to an empirical formula of C28H44O. In 1920, Ida Smedley-MacLean had titrated ergosterol against an iodine solution and showed that it contained three double bonds per molecule, which would agree with Heilbron's work. But identifying where these bonds lay in the molecule was a major challenge.
The first breakthrough came in 1932, by F. Reindel and H. Kipphan. They reacted ergosterol with ozone under mild conditions to give a triozonide, which broke down on hydrogenation and steam distillation to give a variety of aldehydes, including methyliso-propylacetaldhyde (1). The aldehydic carbon must have come from C = C bond scission, implying an alkene group (2) somewhere in the molecule.
The following year C. K. Chuang proposed a 'core' structure for ergosterol. For simplicity, he chose to work on ergostane, (the saturated parent hydrocarbon of ergosterol). Oxidation with chromic acid produced a novel organic acid. He found that he could produce an identical substance from allocholanic acid, a substance of known structure, using the Barbier-Wieland degradation to reduce the carbon chain of a carboxylic acid by one CH2 group. By knowing the structure of allocholanic acid and the effect of a Barbier-Wieland degradation, Chuang proposed the 'core' of ergosterol's carbon framework, as shown in Scheme 1.
In 1934 E. Fernholtz repeated Chuang's work using the acetate of ergosterol, confirming the framework shown in Scheme 1 and confirming the position of the OH group. Adding Reindel and Kipphan's 1932 work on the nature of the side chain gave (3) as the partial structure of ergosterol.
In 1932 the Windaus group had suggested one of the remaining two double bonds probably occupied the same position as cholesterol's sole alkenic linkage in ring B because, whether one starts with ergosterol or cholesterol, both precursors respond similarly to the series of reactions shown in Scheme 2. The final structure for ergosterol was proposed by Windaus in 1934,7 after he realised that ergosterol could form a Diels-Alder [4+2] adduct with maleic anhydride. The '4' in this reaction implies two conjugated double bonds within one ring (ring B), cis -butadienoid in nature, thus identifying the location of the remaining double bond, see structure (4).
Irradiating ergosterol produced vitamin D (later christened vitamin D2 because 'D' or 'D1' was a mixture of materials). This was isomeric with ergosterol, and contained at least three double bonds and a single -OH group. But there was one important difference: while ergosterol yielded 'Diels' hydrocarbon' (5) on dehydrogenation with selenium dioxide, vitamin D2 did not. This suggested that the vitamin, unlike its precursor, ergosterol, did not have four hydrocarbon rings connected one to another, as in 1,2-cyclopentenophenanthrene. Furthermore, a molecule of the vitamin absorbed almost four molecules of H2 on hydrogenation, which would not be possible if it had a four-ring, three double-bond structure like ergosterol.
We can explain the additional unsaturation in the vitamin if it has a three-ring structure. Heilbron investigated this in 1936. First, he treated vitamin D2 with ozone to form methanal. At 7.2 per cent, the yield was low but it indicated a = CH2 group somewhere in the molecule. Secondly, chromic acid oxidation yielded an oily aldehyde. This formed a crystalline semicarbazide that, on purification and analysis, pointed to a parent aldehyde with the formula C21H34O. Ultraviolet spectrophotometry of the semicarbazide derivative showed a peak at 2750 Å (= 2.750 × 10-7 m), almost exactly matching the derivative prepared from α-citral. From these results, Heilbron deduced a structural similarity: both his aldehyde and α-citral were α,β-unsaturated. Building on earlier studies on ergosterol, he proposed this structure for a major fragment of vitamin D2 (see Scheme 3).
Since the C = O in his aldehyde presumably came from a C = C precursor, Heilbron proposed positions for the second and third alkene groups (the first is presumed to be in the C9 H17'tail' identical to that in ergosterol). This left the fourth double bond to be located. Because this is = CH2 in nature (from the methanal fragment), it was a short step to connect this to ergosterol's CH3 group located between rings A and B. In 1936 Heilbron informed the world that he had discovered the correct structure for vitamin D2.8 But had he?
Chemists of the 1930s were generally unconcerned by stereochemistry. It was still convenient to think of the cyclohexane ring as planar and not worry about the precise orientation of substituents (axial or equatorial). Although they acknowledged the possibility of cis/trans isomerism across a double bond, Heilbron assumed the cis -butadienoid arrangement derived from ergosterol in his structure. It was left to Dorothy Crowfoot (Hodgkin) to examine derivatives of vitamin D2 by x-ray crystallography in 1948 and show the diene system derived from the B ring in ergosterol was transoid, rather than cis.9 So the structure of vitamin D2 is as shown in structure (7) or almost, because the bottom ring is slightly twisted out of coplanarity, to reduce steric strain.10
And the chappattis?
From the 1920s, treating and preventing rickets involved either regular cod liver oil - not the most pleasant of medicines - or the expensive irradiation of relevant foods with uv light to enhance their vitamin D content (and, unfortunately, alter their flavour). Understanding its structure allowed large-scale synthesis of the vitamin to fortify a wide range of foods, particularly milk products. Today this is generally through purification and treatment of cholesterol from animal products (such as lanolin from sheep wool) to produce the precursor 7-dehydrocholesterol. This is irradiated to give vitamin D3, which, like D2, is effective in preventing rickets.
Later work showed vitamin D3 is a prohormone rather than a true vitamin. It must be hydroxylated twice in the body, once in the liver and again in the kidneys, before it becomes biologically active.11 More recently, it seems that vitamin D may have other roles in the body, including effects on the immune system and control of cell division. This might lead to treatment of conditions such as psoriasis, in which skin cells divide too fast, and perhaps even in controlling cancers.12
Rickets, once common, became rare again - until, that is, Asian migrants to the UK began developing rickets and osteomalacia. A tendency to live in northern cities, spend little time in direct sunlight and, especially among women, avoid exposing their skin may increase reliance on dietary sources of vitamin D. A meat-free diet limits calcium intake, and chapattis and other phytate-containing foods bind calcium, preventing its absorption. Low levels of calcium increase vitamin D requirements, and osteomalacia (in adults) and rickets (in children) can result. It is, once again, the English disease.
Alan Dronsfield is professor of the history of science in the school of education, health and sciences, at the University of Derby, Kedleston Road, Derby DE22 1GB. Peter Ellis is professor of psychological medicine at the Wellington School of Medicine and Health Sciences, University of Otago, PO Box 7343, Wellington South, New Zealand.
American Society for Nutrition's Journal of Nutrition - article: Forgotten Mysteries in the Early History of Vitamin D
About Vitamin D
- W. Temple, Miscellanea II: an essay upon health and long life. Dublin: J. Swift, 1701.
- Quoted in A. Hardy, Int. J. Epidemiol., 2003, 32 (3), 337.
- R. E. Hughes, Med. Hist., 1973, 17 (2), 127.
- K. Rajakumar, Pediatrics, 2003, 112 (2), 132.
- L. Rosenfeld, Clin. Chem., 1997, 43, 680.
- K. J. Carpenter and L. Zhao, J. Nutrition, 1999, 923.
- A. Windaus, H. H. Inhoffen and S. V. Reichel, Annalen, 1934, 510, 248.
- I. M. Heilbron et al, J. Chem. Soc., 1936, 905.
- D. Crowfoot and J. D. Dunita, Nature (London), 1948, 162, 608.
- L. F. Fieser and M. Fieser, Steroids. London: Chapman & Hall, 1959.
- University of California, Riverside home page on Vitamin D.
- See, for example, M. F. Hollick, Vitamin D: molecular biology, physiology and clinical applications. New Jersey: Humana, 1999.