Wilfred Niels Arnold

Lancet 1996; 347: 1811-13

---

Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City, Kansas, 66160-7421, USA (Prof W N Arnold PhD)

Rationale
Experiments
The illness of King George
Conclusion
References

Attending physicians to King George III (1738-1820) recorded unusual colours in his urine on at least four occasions.¹ Sir George Baker, on Oct 18, 1788, wrote in his diary that the urine was "bilious" (golden brown) and on Aug 26, 1819, Dr Matthew Baille and Dr John Willis reported on "bloody water". Those departures from the normal yellow of urochrome can be explained with a 20th century hypothesis1--namely, that the king had a form of porphyria, in which excessive amounts of 5-aminolevulinic acid and porphobilinogen are excreted during medical crises. These compounds are colourless but can cause pigmentation of the urine upon ageing.² The formation of brown, reddish-brown, or purple hue is due to non-enzymic reactions yielding mixtures of porphyrins and other pigments such as porphobilin.³ The final colour of acute porphyric urine depends on initial concentrations of the compounds undergoing polymerisation, light intensity, and storage time.

On the other hand, two reports by Sir Henry Halford about blue pigments were listed but not explained by Macalpine and Hunter.¹ On Jan 14, 1812, specimens containing 8 and 9 fluid ounces of "bluish" urine were registered on behalf of the King. In the bulletin of Jan 6, 1811, Sir Henry wrote, "The water is of a deeper colour--and leaves a pale blue ring upon the glass near the upper surface". Concurrent documentation showed that the patient was suffering from one of many intractable bouts of constipation and vomiting. Clearly, the blue compound was sparingly soluble and, because the precipitate developed near the surface, I reasoned that oxygen was involved. The hue pointed to indole chemistry.

The following working hypothesis may explain the early 19th century observations. Because the patient was constipated, putrefaction during stasis of the bowel led to indole formation from dietary tryptophan by the normal intestinal flora. Some indole was absorbed into the blood stream and converted in the liver first to indoxyl and then to the sulphate ester. The colourless, water-soluble indoxyl sulphate (metabolic indican in the older published research) was excreted in the urine. In the presence of a sulphatase of bacterial origin, either from a urinary tract infection or as an environmental contaminant of the chamber pot, indoxyl was slowly released by hydrolysis and oxidatively dimerised into indigo blue, which precipitated on the porcelain.

I have demonstrated a model for the latter steps in laboratory experiments. 1 mL samples of normal urine, with or without fortification by indoxyl sulphate
(1 mg/mL), were placed in porcelain crucibles (25 mm diameter at the top). Some pots were inoculated with
0·1 mL of a suspension (about 10¹⁰ colony forming units/mL) of Providencia stuartii cells, which had been flushed from the surface of an agar slant with fresh urine. Controls were samples of normal urine, and of urine plus indoxyl sulphate, that were not inoculated. The crucibles were placed on moist paper towels, in a covered jar, and the mixtures were incubated overnight at room temperature (23°C). The samples that contained both indoxyl sulphate and P stuartii developed a dark blue pigmentation (figure 1). The blue ring on the crucible itself was more obvious after the urine had been removed with a syringe. This pigment was insoluble in water but dissolved in chloroform and displayed an absorption maximum at 604 nm, consistent with indigo blue (indigotin).⁴

The formation of blue pigment depended on incubation time, and on the initial concentrations of indoxyl sulphate and of bacterial suspension. A bluish tinge was apparent after 1 to 2 days, even when only 25% of the indoxyl sulphate and 10% of the bacterial inoculum were used. The blue colour was more intense and developed faster if the urine was initially adjusted to pH 7·0 than if it were adjusted to pH 6·0 or lower; this accords with the chemistry of indigo blue formation, rather than that of indigo white or indigo red.^5,6

Figure 1: Simulation of Sir Henry Halford's observations of 1811

"The water is of a deeper colour" (left) . . . "and leaves a pale blue ring upon the glass" (right). Controls (left vessel of each panel) are included. The liquid from the vessel on the far right has been removed
to reveal the blue ring.

P stuartii was chosen as the test microorganism because it has been associated with urinary tract infections and aryl sulphatase activity (the species was named after C A Stuart, bacteriologist of Providence, Rhode Island, USA, and has no connection to the Royal House of Stuart). Of the eight isolates of this organism (supplied by William Bartholomew and Rebecca Horvat, University of Kansas, and Harry Mobley, University of Maryland), two contained enough aryl sulphatase to be useful in my experiments. Importantly, they also contained urease, which promoted alkalisation of the incubation mixtures and the yield of indigo blue. I also obtained a positive result by setting up the porcelain crucibles with 0·1 mL suspensions of P stuartii, drying them overnight at room temperature, and then adding the urine with indoxyl sulphate. P stuartii is not essential to the working hypothesis; any species with adequate aryl sulphatase and urease activities, and a tolerance of human urine, would perform similarly and generate a blue ring. The shape and material of the vessel, together with the kinetics of the chemical reactions, determine the disposition of the indigo and the extent of a "Halford blue ring".

Sir Henry Halford (1766-1844) was created a baronet by King George III. He had been summoned to attend Princess Amelia, who died in 1810, and was then retained by the King, who was 72 years old. Halford spent almost half the year of 1811 at Windsor Castle where he joined in sympathetic consultations with Dr Matthew Baille and Dr William Heberden and had an uneasy relationship with Dr John Willis and Dr Robert Willis, who were in favour of more restraining treatments for the King.⁷ Halford was not involved with any of the previous eight episodes of serious illness, the first of which occurred in the King's third decade.

Figure 2: The formation of indigo blue from dietary tryptophan

Enzymically catalysed steps occur in intestinal bacteria, liver, and urinary bacteria, in this sequence. The final chemical reaction is a
non-enzymic oxidative dimerisation.

Halford speculated that King George's illness was primarily due to ageing, and that the medical crises were associated with "climacterics" occurring at intervals with mystical multiples of seven.⁸ Subsequent studies, most notably by Ray in 1855 and Guttmacher in 1941, did not advance any real understanding of the King's illness because they embraced only selected aspects of the records and presented the case for manic depressive psychosis by excluding all but mental symptoms.¹

Macalpine and Hunter were the first to address the well-documented physical signs and symptoms as well as the psychotic episodes. Thus, the crises involving colicky abdominal pain accompanied by constipation and vomiting, paralysis, peripheral neuritis, tachycardia, sweating, brown or red urine, insomnia, delirium, the exacerbation of symptoms due to infections (caused by cupping), and the absence of any permanent neurological deficit after each crisis were noted and reasonably seen as acute intermittent porphyria.¹ I believe that the age of onset of first symptoms, the dynamics of all severe episodes, and the deleterious effects of some of the medications, all exemplified by modern examples of acute intermittent porphyria,² should also be added to that list.

Macalpine and Hunter subsequently modified their original working hypothesis to variegate porphyria, primarily to accommodate skin sensitivity, for which the evidence was meagre.^9,10 Objection to their retrospective categorical diagnosis and to specifics of variegate porphyria, and a preoccupation with prevalence rates of the disorder in the general population, became the thrusts of subsequent criticisms, which were often vehement and remarkably uninterested in the total medical picture of King George III (see for example Dean11 and references therein). Some sort of toxic psychosis is indicated within which a metabolic disorder such as acute intermittent porphyria, which will accommodate all of the neurological symptoms as well as the exacerbation factors,² remains a strong possibility. Indeed, Rimington (quoted in Dean,¹¹
p 168) opined that there was "so much in the records being suggestive of porphyria and difficult to explain on any other basis". But what about the blue urine?

The normal bacterial flora of the colon includes species with tryptophanase, which catalyses the formation of indole, pyruvic acid, and ammonia from dietary tryptophan.¹² Indole is rapidly absorbed from the colon.¹³ The sequential biochemical steps in intestinal bacteria, liver, and urinary bacteria, leading to indoxyl in the urine and thence by non-enzymic oxidative dimerisation to indigo blue are summarised in figure 2. Excretion of indoxyl sulphate (indican) up to 140 mg per day is not uncommon in healthy adults.¹⁴ The bag attached to a urinary catheter sometimes develops a purple colouration, which is due to excretion of indican, increased presentation time to adventitious bacteria with sulphatase, and an affinity of the plastics for indigoid pigments.¹⁵

Some nutritional abnormalities and disease states are attended with substantial increases in indican. For example, Rodnight and McIlwain¹⁶ reported that a young boy with pellegra and psychotic symptoms had a six-fold increase of indican in his urine before the condition was resolved with dietary supplements of nicotinamide. Patients with Hartnup disease, caused by a congenital abnormality in the absorption of some aminoacids including tryptophan, may produce 400 mg/day of indican.¹⁷ A case of sprue was marked by increased concentrations of indole derivatives in the urine.18 A familial disease in which hypercalcaemia and nephrocalcinosis are associated with decreased transport of intestinal tryptophan is accompanied by excessive indole production and abnormal indican concentrations in the urine.¹⁹

At least two reports on acute intermittent porphyria may be of immediate relevance. Ludwig and Epstein20 studied a family with an especially high degree of penetrance, and observed that some of the family members showed increased urinary indican even in remission. Price and colleagues²¹ observed an abnormal response to tryptophan loading in some of their porphyria patients. It is noteworthy that the sporadic reports of increased concentrations of various indole derivatives in the urines of mentally ill patients do not show a direct relation with mental disturbance. Rather, indicanuria is most often associated with bowel stasis and influenced by dietary protein composition and intestinal bacteria.²²

Sir Henry Halford's bulletin of Jan 6, 1811, euphemised King George's constipation: "2·5 hours sleep after midnight . . . up to the w. closet twice with little or no effect". The doctor prescribed powdered Aloes socotrina (a cathartic) and extract of Taraxacum officinale (a laxative and diuretic). Thus, the King probably had a condition that would lead to overproduction of urinary indican. I have suggested a role for a bacterial sulphatase and shown a positive result with conditions simulating either urinary tract infection or poor hygiene. The amount of indoxyl sulphate in the test was within physiological limits, although evaporation would actually increase concentrations. I also suggest that Halford's observations of blue urine can be accommodated by a background of acute intermittent porphyria, because severe and recurrent constipation is a hallmark of that disease.

The incentive for this study came from the play by Alan Bennett (The madness of George III. London: Faber & Faber, 1992). About a quarter way, one of the King's pages remarks upon His Majesty's urine being blue and a second attendant argues purple. A similar exchange occurs in the penultimate scene. The subsequent film script fixed upon blue alone in the final scene, and raised the eyebrows of students of porphyria.

1 Macalpine I, Hunter R. The "insanity" of King George III: a classic case of porphyria. BMJ 1966; 1: 65-71.

2 Kappas A, Sassa S, Galbraith RA, Nordmann Y. The porphyrias. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease. 6th ed. New York: McGraw-Hill, 1989: 1305-65.

3 Westfall RC. Isolation of porphobilinogen from the urine of a patient with acute porphyria. Nature 1952; 170: 614-16.

4 Brode WR, Pearson EG, Wyman GM. The relation between the absorption spectra and chemical constitution of dyes. XXVII Cis-trans isomerization in indigo dyes. J Am Chem Soc 1954; 76: 1034-36.

5 Gehauf B, Goldenson J. Detection and estimation of nerve gases by fluorescence reaction. Anal Chem 1957; 29: 276-78.

6 Guilbaut GB, Kramer DW. Resorufin butyrate and indoxyl acetate as fluorogenic substrates for cholinesterases. Anal Chem 1965; 37: 120-23.

7 Munk W. The life of Sir Henry Halford. London & New York: Longmans, Green & Co, 1895.

8 Halford H. Essays and orations, read and delivered at the Royal College of Physicians. 2nd ed. London: John Murray, 1833.

9 Macalpine I, Hunter R. Porphyria: a royal malady. London: British Medical Association, 1968.

10 Macalpine I, Hunter R. George III and the mad business. London: Allen Lane, Penguin Press, 1969.

11 Dean G. The porphyrias: a story of inheritance and environment. 2nd ed. London: Pitman Medical, 1971.

12 Wood WA, Gunsalus IC, Umbreit WW. Function of pyridoxal phosphate: resolution and purification of the tryptophanase enzyme of Escherichia coli. J Biol Chem 1947; 170: 313-21.

13 Fordtran JS, Scroggie WB, Polter DE. Colonic absorption of tryptophan metabolites in man. J Lab Clin Med 1964; 64:
125-32.

14 Marko AM, Reynolds FB. A colorimetric method for measuring indican. Can J Biochem Physiol 1960; 38: 253-62.

15 Dealler SF, Hawkey PM, Millar MR. Enzymatic degradation of urinary indoxyl sulfate by Providencia stuartii and Klebsiella pneumoniae causes the purple urine bag syndrome. J Clin Microbiol 1988; 26: 2125-56.

16 Rodnight R, McIlwain H. Indicanuria and psychosis of a pellagrin.
J Ment Sci 1955; 101: 884-89.

17 Jepson JB. Hartnup disease. In: Stanbury JB, Wyngaarden JB, Fredrickson DS, eds. The metabolic basis of inherited disease. 3rd ed. New York: McGraw-Hill, 1972: 1486-503.

18 Rimington C. Indigoid pigments derived from a pathological urine. Biochem J 1946; 40: 669-74.

19 Drummond KN, Michael AF, Ulstrom RA, Good RA. The blue diaper syndrome: familial hypercalcemia with nephrocalcinosis and indicanuria. Am J Med 1964; 37: 928-48.

20 Ludwig GD, Epstein IS. A genetic study of two families having the acute intermittent type of porphyria. Ann Intern Med 1961; 55: 81-93.

21 Price JM, Brown RR, Peters HA. Tryptophan metabolism in porphyria, schizophrenia, and a variety of neurologic psychiatric diseases. Neurology 1959; 9: 456-68.

22 Sprince H. Indole metabolism in mental illness. Clin Chem 1961; 7: 203-30.