British Joumal of Dermatology (1976) 94, Supplement 12, 3.
Steroid structure and steroid activity J.ELKS Glaxo Research Ltd, Greenford, Middlesex, UB6 OHE
SUMMARY
The diverse biological effects associated with naturally occurring steroids are reviewed, with particular reference to the effect upon biological activity of changes in chemical structure. The motives for synthetic modification of the natural products are discussed and are illustrated by reference to the development of systemic and topical anti-inflammatory agents, with improved potency and reduced side-effects.
The name 'steroid' has tended to become identified with the relatively small group of anti-inflammatory compounds, based on cortisone, that has achieved prominence in the last twenty years or so. This, of course, is a misconception; the steroids comprise a very large group of natural products— widely distributed in both animals and plants—as well as their synthetic analogues.* They are remarkable as I shall try to show, for the diversity of their biological effects. In some instances, these effects are a reflection ofthe natural function ofthe compound; in others, the effect is pharmacological, rather than physiological.
The essentials ofthe steroid structure are shown in Formula i: it is an array of seventeen carbon atoms in four fused rings—three of them six-membered, the other five-membered—with another two carbon atoms protruding from angular positions. This arrangement of carbon atoms is common * Chevrcul, in 1816, applied the name 'cholesterine' (Greek XO^^HJ bile and crTspsos, solid) to the crystalline compound isolated from human gall-stones. The name 'steroid' is now used generically for compounds ofthe same chemical class.
4
J.Elks
to all but a few of the compounds that I shall be mentioning and, indeed, to the vast majority of compounds classified as steroids; the associated hydrogen atoms will vary from compound to compound, as substituents are introduced. The conventional representation of the steroid molecule and the method of numbering the carbon atoms are shown in Formula 2. It should be noted that the structure has no elements of symmetry and that each of the 19 positions is, therefore, distinguishable from each of the others; hence, with a relatively few substituent groups permuted among these 19 positions, one can derive a very large number of different chemical compounds. The possibilities for variation are even greater than this: the framework of carbon atoms is three-dimensional, with 'the front* quite different from 'the back'. For any given position in the rings, a monovalent substituent can be attached in the 'a'-configuration, projecting towards the rear, or in the '/''-configuration, projecting towards the front of the molecule; these alternatives will be chemically and biologically distinct from one another. Because the steroid skeleton is a rigid structure, even a small change in the position of a substituent will, very often, result in a large change of biological activity. This specificity is presumably a consequence of the necessity for the molecule to interact in a particular way with a highly structured protein receptor in order to exert its effect. I shall start my survey with cholesterol (3), which is Nature's own point of entry into steroids proper. It will be seen that cholesterol has the steroid skeleton of Formula r, together with a branched eight-carbon side-chain in the 17-position. In both animal and plant tissue, a very remarkable scries of enzymic reactions builds up the twenty-seven carbon atoms of this compound, with the two carbon atoms of acetic add as the fundamental building block. In turn, cholesterol is subjected to a variety of enzymic reactions to give most of the other naturally occurring steroids, and it is important for its central position in the biogenesis of this class of compound. Although it occurs widely in the lipid components of animal tissue and is a major component of such intrusions as gall-stones and atherosclerotic plaques, cholesterol itself would appear to function as a structural material rather than as a physiologically active compound.
Cholaalaiot
Cholacalcifarol
•ai.2S-0t*rona
Antharldrol
Steroid structure and activity
$
Vitamin D3 (cholecalciferol: 4) is a dose structural relative of cholesterol, with the unusual feature that one ofthe rings has been opened. It is now known that the vitamin is metaboHzed, first in the liver, then in the kidneys, to tx,25-dihydroxycholecalciferol (5), in which two new oxygen atoms have been introduced; it is believed that it is this compound that is responsible for controlling absorption and transport of calcium (Holick & De Luca, 1974). A very different sort of hormonal control is exerted by the ecdysones, a recently discovered group of compounds that control the moulting process in insects and crustaceans. Typical of the group is ecdysterone (6), in which the cholesterol skeleton is intact but has a number of oxygen atoms characteristically distributed around the molecule (Wyatt, 1972). Yet another type of hormonal action is shown by a related compound, antheridiol (j), which acts as a sex hormone in the aquatic fungus, Achlya bisexualis\ it is secreted by the female mycelium and induces formation of anthcridial hyphac in the male strains ofthe plant (Raper, t97i). As will appear later, the steroids that act as sex hormones in higher animals are of a quite different type. :0,H
ChsnodaoKychollc acid
Cholic acid
HO
ONHCH,CH,S0,H
CONHCH,CO,H
HO'
TaurochoUc acid (10)
GlycocholJc acid
A very important group of compounds, the bile acids, are derived from cholesterol by loss of three carbon atoms from the end ofthe chain, and introduction of oxygen atoms in various positions in the ring as well as at the end ofthe chain. The principal bile acids in man are cholic acid (%) and chenodeoxycholic acid (9). There are many more variations on this theme: the precise compounds and the ratios in which they occur vary from species to species, and there appears to be some evolutionary significance to this fHaslewood, 1967). The bile acids occur chiefly in conjugated form; for example, taurocholic acid (10) and glycocholic acid (11), in which cholic acid is combined with taurine and giycine respectively. The biological function of these conjugates depends upon their physical properties; they are detergent-like and they facilitate dispersion and absorption of fats and other lipids, including cholesterol. They do not seem to have any highly specific activity. The shortening of the side-chain is taken a step further in the cardiac glycosides, a group of
J.Elks
DigoKigenin
[12]
compounds of plant origin. Their function in the plant is obscure, but they are of great clinical importance for their tonic action on the failing heart. Again, they comprise a large series of closely related compounds, of which digoxin (13) is one ofthe few that are used in medical practice. To the usual steroid skeleton is now attached, at the 17-position, a branched chain of four carbon atoms, bridged by an oxygen atom to produce the lactone ring that is characteristic of this class; equally characteristic is the complex carbohydrate structure attached at the 3-position. This last feature is important for its effect upon the distribution characteristics of the drug but is not essential for its activity. Thus, digoxigenin (12), the steroid moiety of digoxin, itself has cardiotonic activity but it is not clinically useful because its action is very transient. The next step in the loss ofthe cholesterol side-chain takes us to the pregnanes, in which a chain of only two carbon atoms is attached to the 17-position of the steroid nucleus. This is a series in which rather small modifications of substituent groups result in big changes in biological activity. The simplest ofthe series is progesterone (14), carrying ketonic oxygen atoms at positions 3 and 20. This hormone is secreted by the ovary and is essential, in combination with oestrogen, for the maintenance of pregnancy. Introduction of another oxygen atom, at the 21-position, confers a quite different type of activity; ability to control electrolyte balance, and, with it, water balance fmineralocorticoid activity). Deoxycorticosterone (15) is an example of this type of structure; this compound is, indeed, formed in the adrenal cortex, but the circulating hormones are corticosterone (16) and aldosterone (17) which arc derived from deoxycorticosterone by ii-mono-and 11, i8-di-oxygenation, respectively. The introduction into corticosterone of yet another oxygen atom, this time in the 17-position, produces hydrocortisone (19). This compound retains a little of corticosterone's effect on mineral metabohsm but a number of other effects are now found; they include ability to increase breakdown of protein and deposition of glycogen in the liver (glucocorticoid activity), and inhibition of the inflammatory and the immune response. These and other properties of hydrocortisone and its analogues are considered in much more detail in Dr Snell's paper. Cortisone (18) is readily interconvertible with hydrocortisone and shares its biological properties in most situations. Finally, in the body's production of hormonally active steroids, it removes the side-chain altogether to produce the male and female sex hormones. Thus, testosterone (20) arises from progesterone (14) by loss ofthe two-carbon side-chain and introduction of an oxygen atom in its place. For the formation ofthe phenolic compounds, oestradiol (21) and oestrone (22), one ofthe original nineteen carbon atoms (see Formula i) is lost.
Steroid structure and activity CH,OH
Progesterone
DeoxycorlicoBterone [15)
Corlicosterone (16)
CO ;-0H
Aldostarone (17)
Testosterone (20)
Cortisone
{,8)
Hydrocortlson* (19)
Oeatrsdiol (21)
C")
These are some of the more important ways in which the steroid skeleton is used, in Nature, as a template upon which are impressed more or less minor structural changes, often accompanied by major changes in biological activity. The list is far from being comprehensive; the steroidal alkaloid, malouetine, is a muscle-relaxant, while other naturally occurring steroids and their close relatives have anti-bacterial, anti-fungal, anti-amoebic and anti-tumour activity. With this rich variety of biological effects already available in naturally occurring steroids, it might be asked why the chemist bothers to look further. There are several reasons for his doing so, the more important of which are: 1. The steroid structure may be modified in the hope of finding yet further types of activity, unknown in the natural compounds; although this has led to some active compounds, none has found practical use, so far. 2. Inhibition of the action of a natural hormone may have therapeutic uses, and a competitive antagonist may be found among compounds structurally related to the hormone in question, presumably because they compete for active sites in the receptor. The outstanding example in the steroid scries is the use of spironolactone (24) as a diuretic. Spironolactone is a competitive antagonist to the water-retaining compound, aldosterone (23), possibly because it mimics its structure (Kagawa, 1964)3. A compound of known and useful activity may be modified in order to improve its pharmaco-
JMlks
SCOCH,
Aldosterone
Sptronolactone (24}
kinetics. Thus, some of the steroid hormones are not absorbed from the gastrointestinal tract but relatively minor modification has left the activity unaffected while allowing the oral use ofthe drug. Prolongation of action may be desirable and there are fairly well-understood methods for achieving that end. Again, modification of physical properties may be necessary in order to allow absorption and action at a selected site, so improving the efficiency of use and reducing the likelihood of sideeffects resulting from systemic administration. 4. The chemist's most usual motive in undertaking structural modification of a naturally occurring steroid is to improve on its activity, qualitatively or quantitatively, or both. Increase in potency may be helpful, particularly if it reduces the cost of treatment, but it is not as important as the separation of activities. Most ofthe hormones exert a complex of effects and, for their therapeutic use, some of these actions may be unnecessary or positively harmful. Because of our ignorance ofthe detailed mode of action of most of these compounds it is not always easy to know, other than by trial and error, which actions are separable. None the less, a great deal has been achieved, largely by such empirical methods. CH,OCOCH, CO
CO
HO'
Progailaron*
Alphaxolona
Alphadolona Acalaia
In such work, the aim is usually to emphasize the predominant action ofthe parent compound and to suppress the minor effects. There is at least one example ofthe successful application ofthe converse method. Progesterone (25) has a feeble hypnotic action in addition to its characteristic progestational effect. Modification ofthe structure has led to compounds in which the hypnotic activity has been greatly intensified and the hormonal activity eliminated; the intravenous anaesthetic, Althesin, is a composition containing two such compounds, alphaxolone (26) and alphadolone acetate (27) (Phillipps, 1975). The development ofthe anti-infiammatory steroid drugs in use today serves to illustrate some ofthe principles that I have outlined. The announcement, by Hench and his colleagues in 1949, of the
Steroid structure and activity CHfiH
CH,OH
CH.OH I CO ';-0H
CO -OH
Cortisone
4,5p-Di hydrocortisone
Hydrocortisona
-OehydrocorllcosteronA (32)
e-Dahytfrocortisono (30)
2r-D«oxvcortlson* (33)
effect of cortisone (28) in rheumatoid arthritis was followed by an explosive growth in research in steroid chemistry; this was largely directed towards the search for economic methods of synthesis of cortisone but some analogues of the hortnone were synthesized and tested for anti-rheumatic activity. For a number of years the only other compound to display such activity was hydrocortisone (29) and it was early established that the two compounds arc readily interconvertible in the body. Other changes, even changes that seemed quite trifling, had the effect of reducing the activity, often to the point of extinction. Examples include modification of cortisone (28); (a) by removal of two hydrogen atoms from positions 6 and 7 to give Compound 30, (b) by addition of two hydrogen atoms at positions 4 and 5 to give Compound 31, and (c) by removal of an oxygen atom from position 17 or 21 to give Compounds 32 and 33, respectively. Similar loss of activity resulted from modification of hydrocortisone (29); (a) by removal of an oxygen atom from position 11 or 17 to give Compounds 34 and 35 respectively, or (b) by simply changing the OH group at the 11 position from 'the front' to 'the back' ofthe molecule, to give ii-epihydrocortisone (36). The situation was transformed in 1953 when Fried & Sabo synthesized the 9a-bromo- (39) and 93!-chloro-derivative (38) of hydrocortisone and found them to have glucocorticoid activity in laboratory animals: the chloro-compound was considerably more active than cortisone. The corresponding
CH,OH CO
Reichsleins Substance S (34)
Corlicosterone (35)
II-Epi hydrocortisone (36)
J.Elks
10
9a-fluoro-compound (37) proved to be still more potent: 8 times as active as hydrocortisotie, itself, in an anti-inflammatory assay and 12 6 times as active in the liver-glycogcn assay for glucocorticoid activity. Unfortunately, the fluorine atom was still more effective in increasing mineralocorticoid activity, so that 95!-fluoro-hydrocortisone could not be used in the clinic for its anti-inflammatory action. This work was important, none the less, for its demonstration that anti-inflammatory activity was not special to cortisone and hydrocortisone. CH^H c-OH
(37}:
X = F ; Oa-lluorohydrocorii«ona
(36):
X = CI;ea-chlorohvdrocorllton«
(39):
X = Br: Soi-bromohydracortisona
HO
Piadniaon*
Prednlaoloni
A second key discovery, announced by Herzog et al. in 1955, was that introduction of a 1,2double bond into cortisone or hydrocortisone (i.e. removal of hydrogen atoms from positions i and 2) resulted in about a 4-fold increase in glucocorticoid and anti-inflammatory potency; in this instance, however, the mineralocorticoid effects of cortisone and hydrocortisone were reduced rather than increased. The new compounds, prednisone (40) and prednisolone (41), were the first synthetic analogues ofthe natural hormones to be used in the clinic for their anti-inflammatory properties and they are, of course, still widely employed. Another very important observation was that certain groups, particularly i6s-hydroxy, i6a-methyl and i6/i-methyl, had the effect of nulhfying the potentiation by 9x-fluorine of mineralocorticoid activity, without greatly affecting its potentiating effect on antiinflammatory activity. The combination, then, of these features led to three compounds, triamcinolone (42), dexamcthasone (43) and betamethasone (44), which remain among the most important of the compounds used as systemic anti-inflammatory agents. Other groups that enhance glucocorticoid and reduce mineralocorticoid effects include 6a-methyl and 63:-fluorine, which are found in methylprednisolone (45) and paramethasone acetate (46), respectively. These compounds (42-46) vary in clinical potency, from about equiactive with prednisolone to about 6 times this activity in the case of betamethasone and dexamethasone; they are all devoid of sodium-retaining properties. (The earher work on modiflcation of cortisone and hydrocortisone is very fully reviewed by Sarett, Patchett & Steelman, 1963.) However, there still remain side-effects that hamper the chronic use of all the systemic steroids currently available: they are discussed in Dr Snell's paper and I shall not detail them here. It remains to be seen whether some, at least, of these unwanted activities are separable from the anti-inflammatory action by new modiflcations ofthe molecule j this, certainly, remains a very important goal for further research. It has been known, since the early days of corticosteroid therapy, that many ofthe systemically active anti-inflammatory steroids are effective in the treatment of inflammatory conditions ofthe skin when they are applied topically. Some 15 years ago, it became apparent that certain structural modifications of these steroids resulted in compounds that were more effective than the parents. Since then, a number of new topical preparations have been marketed: Formulae 47-52 show the structures of the steroid components of a few of them. They vary somewhat in their potencies and in their
Steroid structure and activity
zz CH,OH
CH,OH
HO
-OH
Triamcinolona (42)
D
Steroid structure and steroid activity J.ELKS Glaxo Research Ltd, Greenford, Middlesex, UB6 OHE
SUMMARY
The diverse biological effects associated with naturally occurring steroids are reviewed, with particular reference to the effect upon biological activity of changes in chemical structure. The motives for synthetic modification of the natural products are discussed and are illustrated by reference to the development of systemic and topical anti-inflammatory agents, with improved potency and reduced side-effects.
The name 'steroid' has tended to become identified with the relatively small group of anti-inflammatory compounds, based on cortisone, that has achieved prominence in the last twenty years or so. This, of course, is a misconception; the steroids comprise a very large group of natural products— widely distributed in both animals and plants—as well as their synthetic analogues.* They are remarkable as I shall try to show, for the diversity of their biological effects. In some instances, these effects are a reflection ofthe natural function ofthe compound; in others, the effect is pharmacological, rather than physiological.
The essentials ofthe steroid structure are shown in Formula i: it is an array of seventeen carbon atoms in four fused rings—three of them six-membered, the other five-membered—with another two carbon atoms protruding from angular positions. This arrangement of carbon atoms is common * Chevrcul, in 1816, applied the name 'cholesterine' (Greek XO^^HJ bile and crTspsos, solid) to the crystalline compound isolated from human gall-stones. The name 'steroid' is now used generically for compounds ofthe same chemical class.
4
J.Elks
to all but a few of the compounds that I shall be mentioning and, indeed, to the vast majority of compounds classified as steroids; the associated hydrogen atoms will vary from compound to compound, as substituents are introduced. The conventional representation of the steroid molecule and the method of numbering the carbon atoms are shown in Formula 2. It should be noted that the structure has no elements of symmetry and that each of the 19 positions is, therefore, distinguishable from each of the others; hence, with a relatively few substituent groups permuted among these 19 positions, one can derive a very large number of different chemical compounds. The possibilities for variation are even greater than this: the framework of carbon atoms is three-dimensional, with 'the front* quite different from 'the back'. For any given position in the rings, a monovalent substituent can be attached in the 'a'-configuration, projecting towards the rear, or in the '/''-configuration, projecting towards the front of the molecule; these alternatives will be chemically and biologically distinct from one another. Because the steroid skeleton is a rigid structure, even a small change in the position of a substituent will, very often, result in a large change of biological activity. This specificity is presumably a consequence of the necessity for the molecule to interact in a particular way with a highly structured protein receptor in order to exert its effect. I shall start my survey with cholesterol (3), which is Nature's own point of entry into steroids proper. It will be seen that cholesterol has the steroid skeleton of Formula r, together with a branched eight-carbon side-chain in the 17-position. In both animal and plant tissue, a very remarkable scries of enzymic reactions builds up the twenty-seven carbon atoms of this compound, with the two carbon atoms of acetic add as the fundamental building block. In turn, cholesterol is subjected to a variety of enzymic reactions to give most of the other naturally occurring steroids, and it is important for its central position in the biogenesis of this class of compound. Although it occurs widely in the lipid components of animal tissue and is a major component of such intrusions as gall-stones and atherosclerotic plaques, cholesterol itself would appear to function as a structural material rather than as a physiologically active compound.
Cholaalaiot
Cholacalcifarol
•ai.2S-0t*rona
Antharldrol
Steroid structure and activity
$
Vitamin D3 (cholecalciferol: 4) is a dose structural relative of cholesterol, with the unusual feature that one ofthe rings has been opened. It is now known that the vitamin is metaboHzed, first in the liver, then in the kidneys, to tx,25-dihydroxycholecalciferol (5), in which two new oxygen atoms have been introduced; it is believed that it is this compound that is responsible for controlling absorption and transport of calcium (Holick & De Luca, 1974). A very different sort of hormonal control is exerted by the ecdysones, a recently discovered group of compounds that control the moulting process in insects and crustaceans. Typical of the group is ecdysterone (6), in which the cholesterol skeleton is intact but has a number of oxygen atoms characteristically distributed around the molecule (Wyatt, 1972). Yet another type of hormonal action is shown by a related compound, antheridiol (j), which acts as a sex hormone in the aquatic fungus, Achlya bisexualis\ it is secreted by the female mycelium and induces formation of anthcridial hyphac in the male strains ofthe plant (Raper, t97i). As will appear later, the steroids that act as sex hormones in higher animals are of a quite different type. :0,H
ChsnodaoKychollc acid
Cholic acid
HO
ONHCH,CH,S0,H
CONHCH,CO,H
HO'
TaurochoUc acid (10)
GlycocholJc acid
A very important group of compounds, the bile acids, are derived from cholesterol by loss of three carbon atoms from the end ofthe chain, and introduction of oxygen atoms in various positions in the ring as well as at the end ofthe chain. The principal bile acids in man are cholic acid (%) and chenodeoxycholic acid (9). There are many more variations on this theme: the precise compounds and the ratios in which they occur vary from species to species, and there appears to be some evolutionary significance to this fHaslewood, 1967). The bile acids occur chiefly in conjugated form; for example, taurocholic acid (10) and glycocholic acid (11), in which cholic acid is combined with taurine and giycine respectively. The biological function of these conjugates depends upon their physical properties; they are detergent-like and they facilitate dispersion and absorption of fats and other lipids, including cholesterol. They do not seem to have any highly specific activity. The shortening of the side-chain is taken a step further in the cardiac glycosides, a group of
J.Elks
DigoKigenin
[12]
compounds of plant origin. Their function in the plant is obscure, but they are of great clinical importance for their tonic action on the failing heart. Again, they comprise a large series of closely related compounds, of which digoxin (13) is one ofthe few that are used in medical practice. To the usual steroid skeleton is now attached, at the 17-position, a branched chain of four carbon atoms, bridged by an oxygen atom to produce the lactone ring that is characteristic of this class; equally characteristic is the complex carbohydrate structure attached at the 3-position. This last feature is important for its effect upon the distribution characteristics of the drug but is not essential for its activity. Thus, digoxigenin (12), the steroid moiety of digoxin, itself has cardiotonic activity but it is not clinically useful because its action is very transient. The next step in the loss ofthe cholesterol side-chain takes us to the pregnanes, in which a chain of only two carbon atoms is attached to the 17-position of the steroid nucleus. This is a series in which rather small modifications of substituent groups result in big changes in biological activity. The simplest ofthe series is progesterone (14), carrying ketonic oxygen atoms at positions 3 and 20. This hormone is secreted by the ovary and is essential, in combination with oestrogen, for the maintenance of pregnancy. Introduction of another oxygen atom, at the 21-position, confers a quite different type of activity; ability to control electrolyte balance, and, with it, water balance fmineralocorticoid activity). Deoxycorticosterone (15) is an example of this type of structure; this compound is, indeed, formed in the adrenal cortex, but the circulating hormones are corticosterone (16) and aldosterone (17) which arc derived from deoxycorticosterone by ii-mono-and 11, i8-di-oxygenation, respectively. The introduction into corticosterone of yet another oxygen atom, this time in the 17-position, produces hydrocortisone (19). This compound retains a little of corticosterone's effect on mineral metabohsm but a number of other effects are now found; they include ability to increase breakdown of protein and deposition of glycogen in the liver (glucocorticoid activity), and inhibition of the inflammatory and the immune response. These and other properties of hydrocortisone and its analogues are considered in much more detail in Dr Snell's paper. Cortisone (18) is readily interconvertible with hydrocortisone and shares its biological properties in most situations. Finally, in the body's production of hormonally active steroids, it removes the side-chain altogether to produce the male and female sex hormones. Thus, testosterone (20) arises from progesterone (14) by loss ofthe two-carbon side-chain and introduction of an oxygen atom in its place. For the formation ofthe phenolic compounds, oestradiol (21) and oestrone (22), one ofthe original nineteen carbon atoms (see Formula i) is lost.
Steroid structure and activity CH,OH
Progesterone
DeoxycorlicoBterone [15)
Corlicosterone (16)
CO ;-0H
Aldostarone (17)
Testosterone (20)
Cortisone
{,8)
Hydrocortlson* (19)
Oeatrsdiol (21)
C")
These are some of the more important ways in which the steroid skeleton is used, in Nature, as a template upon which are impressed more or less minor structural changes, often accompanied by major changes in biological activity. The list is far from being comprehensive; the steroidal alkaloid, malouetine, is a muscle-relaxant, while other naturally occurring steroids and their close relatives have anti-bacterial, anti-fungal, anti-amoebic and anti-tumour activity. With this rich variety of biological effects already available in naturally occurring steroids, it might be asked why the chemist bothers to look further. There are several reasons for his doing so, the more important of which are: 1. The steroid structure may be modified in the hope of finding yet further types of activity, unknown in the natural compounds; although this has led to some active compounds, none has found practical use, so far. 2. Inhibition of the action of a natural hormone may have therapeutic uses, and a competitive antagonist may be found among compounds structurally related to the hormone in question, presumably because they compete for active sites in the receptor. The outstanding example in the steroid scries is the use of spironolactone (24) as a diuretic. Spironolactone is a competitive antagonist to the water-retaining compound, aldosterone (23), possibly because it mimics its structure (Kagawa, 1964)3. A compound of known and useful activity may be modified in order to improve its pharmaco-
JMlks
SCOCH,
Aldosterone
Sptronolactone (24}
kinetics. Thus, some of the steroid hormones are not absorbed from the gastrointestinal tract but relatively minor modification has left the activity unaffected while allowing the oral use ofthe drug. Prolongation of action may be desirable and there are fairly well-understood methods for achieving that end. Again, modification of physical properties may be necessary in order to allow absorption and action at a selected site, so improving the efficiency of use and reducing the likelihood of sideeffects resulting from systemic administration. 4. The chemist's most usual motive in undertaking structural modification of a naturally occurring steroid is to improve on its activity, qualitatively or quantitatively, or both. Increase in potency may be helpful, particularly if it reduces the cost of treatment, but it is not as important as the separation of activities. Most ofthe hormones exert a complex of effects and, for their therapeutic use, some of these actions may be unnecessary or positively harmful. Because of our ignorance ofthe detailed mode of action of most of these compounds it is not always easy to know, other than by trial and error, which actions are separable. None the less, a great deal has been achieved, largely by such empirical methods. CH,OCOCH, CO
CO
HO'
Progailaron*
Alphaxolona
Alphadolona Acalaia
In such work, the aim is usually to emphasize the predominant action ofthe parent compound and to suppress the minor effects. There is at least one example ofthe successful application ofthe converse method. Progesterone (25) has a feeble hypnotic action in addition to its characteristic progestational effect. Modification ofthe structure has led to compounds in which the hypnotic activity has been greatly intensified and the hormonal activity eliminated; the intravenous anaesthetic, Althesin, is a composition containing two such compounds, alphaxolone (26) and alphadolone acetate (27) (Phillipps, 1975). The development ofthe anti-infiammatory steroid drugs in use today serves to illustrate some ofthe principles that I have outlined. The announcement, by Hench and his colleagues in 1949, of the
Steroid structure and activity CHfiH
CH,OH
CH.OH I CO ';-0H
CO -OH
Cortisone
4,5p-Di hydrocortisone
Hydrocortisona
-OehydrocorllcosteronA (32)
e-Dahytfrocortisono (30)
2r-D«oxvcortlson* (33)
effect of cortisone (28) in rheumatoid arthritis was followed by an explosive growth in research in steroid chemistry; this was largely directed towards the search for economic methods of synthesis of cortisone but some analogues of the hortnone were synthesized and tested for anti-rheumatic activity. For a number of years the only other compound to display such activity was hydrocortisone (29) and it was early established that the two compounds arc readily interconvertible in the body. Other changes, even changes that seemed quite trifling, had the effect of reducing the activity, often to the point of extinction. Examples include modification of cortisone (28); (a) by removal of two hydrogen atoms from positions 6 and 7 to give Compound 30, (b) by addition of two hydrogen atoms at positions 4 and 5 to give Compound 31, and (c) by removal of an oxygen atom from position 17 or 21 to give Compounds 32 and 33, respectively. Similar loss of activity resulted from modification of hydrocortisone (29); (a) by removal of an oxygen atom from position 11 or 17 to give Compounds 34 and 35 respectively, or (b) by simply changing the OH group at the 11 position from 'the front' to 'the back' ofthe molecule, to give ii-epihydrocortisone (36). The situation was transformed in 1953 when Fried & Sabo synthesized the 9a-bromo- (39) and 93!-chloro-derivative (38) of hydrocortisone and found them to have glucocorticoid activity in laboratory animals: the chloro-compound was considerably more active than cortisone. The corresponding
CH,OH CO
Reichsleins Substance S (34)
Corlicosterone (35)
II-Epi hydrocortisone (36)
J.Elks
10
9a-fluoro-compound (37) proved to be still more potent: 8 times as active as hydrocortisotie, itself, in an anti-inflammatory assay and 12 6 times as active in the liver-glycogcn assay for glucocorticoid activity. Unfortunately, the fluorine atom was still more effective in increasing mineralocorticoid activity, so that 95!-fluoro-hydrocortisone could not be used in the clinic for its anti-inflammatory action. This work was important, none the less, for its demonstration that anti-inflammatory activity was not special to cortisone and hydrocortisone. CH^H c-OH
(37}:
X = F ; Oa-lluorohydrocorii«ona
(36):
X = CI;ea-chlorohvdrocorllton«
(39):
X = Br: Soi-bromohydracortisona
HO
Piadniaon*
Prednlaoloni
A second key discovery, announced by Herzog et al. in 1955, was that introduction of a 1,2double bond into cortisone or hydrocortisone (i.e. removal of hydrogen atoms from positions i and 2) resulted in about a 4-fold increase in glucocorticoid and anti-inflammatory potency; in this instance, however, the mineralocorticoid effects of cortisone and hydrocortisone were reduced rather than increased. The new compounds, prednisone (40) and prednisolone (41), were the first synthetic analogues ofthe natural hormones to be used in the clinic for their anti-inflammatory properties and they are, of course, still widely employed. Another very important observation was that certain groups, particularly i6s-hydroxy, i6a-methyl and i6/i-methyl, had the effect of nulhfying the potentiation by 9x-fluorine of mineralocorticoid activity, without greatly affecting its potentiating effect on antiinflammatory activity. The combination, then, of these features led to three compounds, triamcinolone (42), dexamcthasone (43) and betamethasone (44), which remain among the most important of the compounds used as systemic anti-inflammatory agents. Other groups that enhance glucocorticoid and reduce mineralocorticoid effects include 6a-methyl and 63:-fluorine, which are found in methylprednisolone (45) and paramethasone acetate (46), respectively. These compounds (42-46) vary in clinical potency, from about equiactive with prednisolone to about 6 times this activity in the case of betamethasone and dexamethasone; they are all devoid of sodium-retaining properties. (The earher work on modiflcation of cortisone and hydrocortisone is very fully reviewed by Sarett, Patchett & Steelman, 1963.) However, there still remain side-effects that hamper the chronic use of all the systemic steroids currently available: they are discussed in Dr Snell's paper and I shall not detail them here. It remains to be seen whether some, at least, of these unwanted activities are separable from the anti-inflammatory action by new modiflcations ofthe molecule j this, certainly, remains a very important goal for further research. It has been known, since the early days of corticosteroid therapy, that many ofthe systemically active anti-inflammatory steroids are effective in the treatment of inflammatory conditions ofthe skin when they are applied topically. Some 15 years ago, it became apparent that certain structural modifications of these steroids resulted in compounds that were more effective than the parents. Since then, a number of new topical preparations have been marketed: Formulae 47-52 show the structures of the steroid components of a few of them. They vary somewhat in their potencies and in their
Steroid structure and activity
zz CH,OH
CH,OH
HO
-OH
Triamcinolona (42)
D
