The Post-World War II Era Race to Synthesize Cortisone

          In mammals, the adrenal glands are responsible for releasing hormones in response to stress. The adrenal cortex, situated along the perimeter of the adrenal gland, is responsible for the biosynthetic production of mineralocorticoids, which control electrolyte and water levels, glucocorticoids, which possess historic anti-inflammatory properties, and androgen hormones.
          In the fall of 1941, just prior to the entry of the United States into World War II, adrenal cortical hormones were sought as potential therapeutics for the treatment of shock and battle fatigue. This was, in part, due to a bizarre rumor, erroneously circulated by allied intelligence, which suggested that Germany was buying the adrenal glands of cattle from Argentinian slaughterhouses and administering extracts to their Luftwaffe pilots. The prevailing thought at the time was that German scientists had unraveled the secret of the adrenal glands and were providing the Nazis with an adrenal product that enabled pilots to fly at unusually high altitudes without suffering ill effects from lack of oxygen. At the behest of the US government, industrial, university and foundation laboratories were joined in a massive effort to uncover and provide allied forces with this coveted ‘miracle drug’ substance of unknown molecular composition.
            By 1941, 26 cortical steroids had been isolated from the adrenal cortex by the independent laboratories of E. C. Kendall, T. Reichstein and Wintersteiner. Only a few milligrams of these compounds can be isolated from a ton of adrenals and, thus, synthesis is required to procure meaningful amounts. During the war, a 30-step partial synthesis of Kendall’s ‘Compound A’ was accomplished by Merck and Co. By 1944, nearly 100 grams of Kendall’s A, 11-dehydrocorticosterone, had been synthesized. The molecule was shown to be devoid of any significant biological activity. However, in the same year, L. H. Sarett (depicted below) of Merck prepared the 17-hydroxylated analog of Kendall’s A by a 39-step synthetic sequence starting from cholic acid (vida infra). Kendall referred to this substance as ‘Compound E,’ but to avoid confusion with ‘vitamin E,’ Kendall agreed to call it ‘cortisone.’ Cortisone failed in the treatment of adrenal insufficiency (Addison’s Disease).
          Since the late-1920s, the nobel laureate-to-be, P. S. Hench of the Mayo Foundation, had astutely observed that the condition of women suffering from rheumatoid arthritis improved during pregnancy. He hypothesized that the improvement was due to the release of some hormone. In 1948, acting solely on the basis of this rather vague hypothesis, Hench injected 100 mg of Sarett’s synthetic cortisone into a patient suffering from a serious case of rheumatoid arthritis. The dramatic results are now legendary. In a few days the previously bedridden patient went downtown on a shopping spree and the event was covered extensively by the popular press of the day. Demand for the new wonder drug for the treatment of inflammatory diseases became enormous.

Lewis Sarett (left) and Max Tishler (right) of Merck

          Merck’s process chemistry laboratories, then led by Max Tishler (depicted above), began to develop Sarett’s synthetic route to cortisone for industrial-scale manufacturing. By 1949, only one year after Hench’s historic discovery, one kilogram of synthetic cortisone was obtained through optimization and execution of Sarret’s ‘bile acid process,’ starting from animal-derived deoxycholic acid (7-DCA). By 1950, one ton of cortisone acetate had been prepared. As a direct result of this research and development effort, between the years of 1951 and 1960, the price of cortisone decreased from $200/gram to $1.50/gram.

The Merck bile acid process required 39 linear synthetic steps to convert cholic acid into cortisone. The route proceeded in three phases. The sole purpose of phase 1, highlighted in the reaction scheme above, was to transpose oxygen from carbon position 12 to 11 of the cyclopentenophenanthrene steroid C-ring. This was necessary because, at the time, no known, abundant, natural steroid with oxygen at C11 was available. A key synthetic feature of phase 1 of Merck’s bile acid process involved the formation of an oxo-bridge between the seemingly distant carbon positions 3 and 9, forged by intramolecular S_N2' displacement of the allylic bromide 1 by the a-oriented C3 hydroxyl group. Next, stereoselective bromination of the resultant intermediate 2 gave predominantly the 11b, 12a-dibromide 3. Subsequent regioselective displacement of the more reactive 11b-bromide and further oxidation with sodium dichromate then secured the ketone 4 and completed the critical task of transposition of oxygen from carbon position 12 (of cholic acid) to C11.
           The second phase of Merck’s manufacturing process transforms the cholate side chain to the requisite dihydroxyactone of cortisone. In the course of phase 2, the C17 side chain is first converted into a diene (5) and then oxidatively cleaved to furnish the pregnane dione 7. Seven subsequent conversions are then required for additional oxidative elaboration of the C17 substituent. And, finally, phase 3, which simply introduces unsaturation into the A-ring, is completed in four steps, marking the completion of Merck’s cortisone manufacturing process.

It should be noted that Sarett of Merck (along with R. B. Woodward of Harvard) also completed a highly stereospecific total synthesis of optically active cortisone in 1951. The endgame of Sarett’s total synthesis is depicted in the scheme shown above. The protocol hinges on the formation of a key cyanohydrin intermediate, which is dehydrated upon exposure to phosphorous oxychloride in pyridine to produce 12. Dihydroxylation of the 17,20-olefin is then achieved by treatment of 12 with potassium permanganate and hydrolysis of the transiently formed oxidation product (13) gives cortisone acetate. Total synthesis, as is often the case, ultimately proved too costly for industrial-scale cortisone production.

Russell E. Marker

By the late 1940s, chemists at Syntex laboratories in Mexico City had initiated efforts to develop an alternative to Merck’s bile acid process. Syntex (from “Synthesis” and “Mexico”) was founded to develop and exploit a synthetic technology invented by Russell Marker, a steroid pioneer and former chemistry professor at Penn State University. Marker had discovered a simple method for degradation of the steroidal sapogenin side chain, which efficiently produces a synthetic precursor to the female sex hormone progesterone from abundant, plant-derived starting materials. In 1951, Carl Djerassi was recruited by Syntex to lead a program focusing on the synthesis of 11-oxygenated cortical steroids starting from a plant raw material derived from Mexican yams (cabeza de negro root), utilizing the Marker degradation (depicted in the scheme below).

          The development of an alternate synthesis of cortisone that did not start from a cattle-derived bile acid but, rather, an inexhaustible plant-derived starting material was the most crucial organic chemistry challenge of the 1940s and early 1950s. But since all of the plant sapogenins that were known at the time lacked an oxygenated substituent in the C-ring, the critical synthetic hurdle associated with this endeavor became introduction of oxygen into a naked steroidal ring C. Many synthetic solutions to this daunting problem were developed by Djerassi’s group at Syntex, but none of these saw application on an industrial scale. This was because an Upjohn group headed by O. H. Peterson reported in 1952 a culture of the fungus Rhizopus arrhizus from Kalamazoo air, capable of converting progesterone into 11a-hydroxyprogesterone. Culture development eventually raised the isolated yield of the process to ~90%, thereby solving the problem of oxygenating C11 of the steroid framework en route to cortical steroids. However, Upjohn’s microbiological process depended on the availability of tonnage quantities of cheap progesterone, which, in the early 1950s, were only available from Syntex using Marker’s diosgenin degradation. The intersection of the Marker degradation (shown above) with Upjohn’s microbial biotransformation of progesterone (see below, conversion of 19 to 20) provided the desired alternative to Merck’s celebrated bile acid production process, which was shut down in 1966.

          The Syntex route to cortisone, in brief, involved a novel, stereoselective catalytic hydrogenation of the 4,5-double bond of 20, to be followed by a highly challenging regio- and stereoselective reduction of the pregnane-trione 22, which produces the 3a-ol advanced intermediate 23 with moderate efficiency.

           The conversion of 23 into cortisone acetate (depicted in the scheme above) relied upon synthetic methodology that was previously described by Kritchevsky and co-workers at the Sloan-Kettering Institute. The route features selective epoxidation of the dienol acetate 24 to introduce the C17 dihydroxyacetone moiety followed by a bromination/elimination sequence to install the requisite A-ring unsaturation, culminating in synthetic cortisone.

Carl Djerassi at Syntex in 1951

           A few years later, Glaxo in England commercialized a cortisone process that effectively transpositioned the keto group of hecogenin, obtained from abundant and inexpensive East African sisal wastes, from C12 to C11. The process was based on an alternate synthesis of cortisone that was developed at Syntex and licensed to Glaxo.