2580
SIDNEY M. Fox, VICTORE. ORICONIAND LELAND L. SMITH
Vol. 82
aqueous ethanol and dried in the dark, m.p. 113.4-119' with the product sublimes without melting. The analytical previous sintering, XgiH278 mp (e 2270) with shoulders a t sample was sublimed a t 225' (lo-' mm.), vgf: 1604 (s, 225 ( e 8070) and 284 mp ( E 2060). COZ-), 220 ( E 8180), 279 ( E 2150) and 287 m l (e Anal. Calcd. for CIIHI~NOZ:C, 69.09; H, 6.85; N, 2010). Anal. Calcd. for C I ~ H ~ I N OC,~ :08.41; H, 8.04; N, 7.32. Found: C,69.33; H,6.56; N,7.19. 5.32. Found: C,68.49; H, 8.02; iY,5.24. 1,2,3,4-Tetrahydro-6-methoxy-2-naphthylamine(IVb).Methyl N-2-( 1,2 ,3,4-Tetrahydro-6-methoxynaphthyl)A solution of 9.0 g. of 6-methoxy-2-tetralone oxime in 200 ml. of methanol (saturated with ammonia a t 0-5") was hy- a-aminoisobutyrate (Ve).-In this preparation, the same procedure and apparatus was employed as in the synthesis of drogenated in the presence of 1.5 teaspoons of W2-Raney ( 1,2,3,4-tetrahydronaphthyl)-cu-arninoisobutyrate nickel and an initial pressure of 1470 p.s.i. for 4 hours a t methyl B above. 55". The oil obtained by removal of the catalyst and con- byhprocedure mixture of 3.5 g. of N-2-( 1,2,3,4-tetrahydro-6-methoxycentration of the filtrate was dissolved in dilute hydrochlo- naphthyl)-a-aminoisobutyric acid and 23.0 ml. of 0.589 N ric acid. The resulting solution was washed with ether, hydroxide was concentrated to a made strongly basic with sodium hydroxide and the regen- tetramethylammonium thick paste which wns transferred to a distilling flask. U-ater erated amine was extracted with ether. The dried ether removed, finallv a t 120" (0.2 mm.) for 4 hours. The extract was distilled giving 4.72 g. (57%) of 1,2,3,4-tetrahy- was resulting solid was melted using a micro burner. Gas evodro-6-methoxy-2-naphthylamine, b.p. 108-110° (0.2 mm.) lution was vigorous and on continued heating 2.29 g. (62Yc) which was characterized as the hydrochloride. of the crude product distilled, b.p. 160" (0.2-0.3 mm.). 1,2,3,4-Tetrahydro-6-merhoxy-2-naphthylamine hydroanalytical sample of Ve had b.p. 138-140' (0.06 mm.). chloride was prepared in 94yc yield by saturating an ethereal The n z 7 1.5233, ~ v::!' 172G (s, CO), A$:" 279.5 ( e 2180) and solution of the amine with hydrogen chloride, m.p. 234-236" 288 mp ( e 2060) with a plateau a t 216-226 n i l ( e 7700). with previous softening; 221 ( E 7960), 279 ( e 2180) Anal. Calcd. for C1SH23K08:C, 69.28; H, 8.36; Ii, and 287 mfi (e 2020). ' , 5.00. Anal. Calcd. for CllHl&lNO: C, 61.82; H, 7.55; N, 5.05. Found: C, 69.38; H , 8.42; h 6.55. Foimd: C,61.74; H, 7.54; N,6.41. Acknowledgment.-We wish to thank Dr. V. A. N-2-( 1,2,3,4-Tetrahydro-6-methoxynaphthyl)-cr-amino- Drill and his associates of the Division of Biological isobutyric Acid (Vd).-To a mixture of 4.72 g. of 1,2,3,4-tet- Research of G. D. Searle and Company for bioasrahydr0-6-1nethoxy-2-naphthylamine and 2.27 g. of acetone says of some of the compounds. Compounds Vbcyanohydrin, which had been allowed to stand overnight, was added 300 ml. of concentrated hydrochloric acid (satu- Ve showed essentially no lipodiatic, estrogenic, rated with hydrogen chloride a t 0-5'). The resulting mix- androgenic or anti-inflammatory activity. Howture was allowed to stand a t room temperature for 15 hours ever, the hydrochloride salt of Vb exhibited a and then was heated under reflux on the steam-bath for 3 positive anti-inflammatory activityz3 a t a level hours before being concentrated to dryness. The residue was dissolved in 50 ml. each of water and methanol, the solu- similar to that of Butazolidine. tion was clarified with h-orit and adjusted to pH 5.5-6 with (23) J . J. Selitto and L. 0. Randall, Federation Proc., Abstract No. potassium carbonate. The precipitate was collected on a 1323 (1964). filter, washed and dried, giving4.97 g. (70%) of N-2-(1,2,3,4-tetraliydro-6-metlioxynaphthyl)-cr-aminoisobutyric acid; CAMBRIDGE 39, MASS.
[CONTRIBUTION FROM
THE
CHEMICAL PROCESS IMPROVEMENT DEPARTMENT, LEDERLE LABORATORIES, AMERICAN CYANAMID Co.]
16a-Hydroxy Steroids. V.'
110-Esters of Triamcinolone
BY SIDNEY M. Fox, VICTORE. ORIGONIAND LELAND L. SMITH RECEIVED SEPTEMBER 30, 1959 Acetylation of the 1lp- and 17a-hydroxyl groups of triamcinolone and of 16a-hydroxyhydrocortisone is readily accomare plished with warm acetic anhydride and pyridine. Both a llp,l6cr,21-triacetate and a llp,16cu,17~~,21-tetraacetate formed from triamcinoloiie. Hydrocortisone and Sa-fluorohydrocortisone were acetylated a t the lla-hydroxyl also. Some chemical proof for the assigned structures is presented.
It is commonly held that the l l p - and the 17ahydroxyl groups of the active corticosteroids are not readily acetylated with acetic anhydride and pyridine, neither a t room tempexature nor a t slightly elevated temperatures. Occasional exceptions in other steroid series have been notedI2 but such treatment is generally regarded as a poor means of acetylation of these groups. We have found that acetic anhydride-pyridine smoothly acetylates the 110- and 17a-hydroxyl groups of triamcinolone3 (9a-fluoro-1 ID, 16a,17a,21-tetrahydroxy- 1,4-pregnadiene-3,2O-dione)(II a) yielding (1) Paper IV, L. L. Smith, J. J. Garbarini, J. J. Goodman, hl. Mars and H. Mendelsohn, THISJOURNAL, 83, 1437 (1960). (2) M. Steiger and T.Reichstein, Helu. Chim. Acta, ZO, 817 (1937); A. D. Kemp, A. Kappas, I. I. Salamon, F. Herling and T. F. Gallagher, J . Bioi. Chem., 210, 123 (1954). (3) S. Bernstein, R. H. Lenhard, W. S. Allen, hl. Heller, R. Littell, S. M. Stolar, L. I. Feldman and R. H. Blank, THISJOURNAL, 78, 5693 (1956); 81, 1689 (1959).
the llp,l6a,21-triacetate I11 and 11/3,16a,17a,21tetraacetate IV derivatives. Triamcinolone triacetate (111) and tetraacetate (IV) were encountered unexpectedly in preparations of triamcinolone 16q21-diacetate (IIb) obtained v i a microbiological dehydrogenation of 16a-hydroxy-9a-fluorohydrocortisone 16cr,21-diacetate (I). Nocardia corallina dehydrogenates I3 but also hydrolyzes partially the diesters involved, yielding a mixture of diacetates, monoacetates and free alcohols. Reacetylation of the fermentation extract residues without isolation of the purified steroids, using large excesses of acetic anhydride and pyridine and inadvertently warming on a steam-bath, gave the diacetate I I b which was contaminated with a major proportion of a new, more mobile component (paper chromatographic analyses) together with traces of a still more mobile component and unaltered substrate I.
May 20, 1960
1~P-ESTERS OF TRIAMCINOLONE
2581
Resolution of the mixture by partition chroma- conditions. Using excessive amounts of acetic tography yielded the more mobile components I11 anhydride and pyridine and heating to about 80" and IV in the first fraction, mixed with non-ster- gave products I11 and IV, as identified by paper oidal fermentation impurities. From this first chromatographic examination. Partition chromafraction was isolated in weight yield (based tography of the mixture gave IV as the major on the weight of I charged) a triacetate 111, con- product, with the triacetate 111 as a minor product. taminated with traces of IV. Unaltered substrate Analysis of IV indicated it to be a tetraacetate, 11~,16a,17a,21-tetraacefoxy-9a-fluoro-1,4I was recovered in the middle fractions (8% by ;.e., weight), and triamcinolone diacetate I I b in the pregnadiene-3,20-dione. Infrared and ultraviolet absorption spectra and papergram mobility support final fractions in weight yield. Purified I11 was recognized as a triacetate, this structural assignment. Hydrolysis of the C Z ~ H S ~ Oon ~ Fthe , basis of elemental analysis, tetraacetate I V with base also afiords the monoacepapergram mobility and infrared absorption spec- tate V, previously obtained from the triacetate 111. The facile acetylation of both the 110- and 17atra. Alkaline hydrolysis of I11 gave a monoacetate V, Cz3Hz9O7F, recognized as an 110-monoacetate hydroxyl groups of triamcinolone without evidences on the basis of mode of formationI4optical rotational of dehydration, isomerization or other artifact increments (see Table I), and a hypsochromic shift formation, while not anticipated, was accomplished of 2 mp in the ultraviolet spectra previously as- in good yield. Extension of the experiments to ll/?,l6a,l7a,21-tetrahydroxy-20-ketone, sociated with acetylation of the lla- or 110-hy- another droxyl group of A*-3-ketonesand A4.6-3-ketones.4as5 16a-hydroxyhydrocortisone, afforded a tetraaceThe present instance is the first reported case tate VI, assigned the stiucture ll6,16aI17a,21of the hypsochromic effect on the ultraviolet ab- tetraacetoxy-4-pregnene, 3,20-dione. Acetylation of both hydrocortisone 21-acetate Recently the sorption of 9a-fl~oro-A~~~-3-ketones. ultraviolet spectra of a series of 110-esters of 9a- (VIIa) and 9a-fluorohydrocortisone 2 1-acetate chloro- and Sa-bromo-A' k4-3-ketoneswere reported, (VILIa) afforded the respective ll/l,21-diacetates which spectra show the typical hypsochromic shift VIIb and VIIIb. In each instance a hypsochromic shift in the ultraviolet spectra of ca. 2 mp, toof ta.1-2 mp.6 Rearrangement of the steroid during alkaline gether with increased papergram mobility and arguhydrolysis was ruled out by acetylation of the mono- ments of molecular rotation and biological inacetate V, which yielded the original triacetate activity, were used t o suppott the assigned llp111. These evidences, coupled with spectral acetate ester structure. The molecular rotational increments associated data, rotational data and papergram behavior, together with the complete loss of glucocorticoid with 1I@-acylation of 1l~-hydroxy-3-ketosteroids activity associated with triamcinolone 16a,21- are given in Table I. For several 110-acetates and diacetate, lead to the assignment of the structure 110-formates the A[MID ranges from $3" t o l l g , 16a. 21-triacetoxy- 9 a -fluor0 -17a-hydroxy - 1,4- +165", with the one exception of -73" for the pregnadiene-3.20-dione for the triacetate I11 and 118-acetate of 110,17a-dihydroxy-5p-prepane-3,2017a,21- dione. The recently reported 116-esters of 9aof the structure 1l~-acetoxy-9a-fluoro-l6a, trihydroxy - 1,4 - pregnadiene - 3,20 - dione for the chloroprednisolone 21-acetate6 (21-acetoxy-9achloro - ll/3,17a- dihydroxy - 1,4-pregnadiene-3,20monoacetate V. Since the initial observations of a relatively dione) conform to this range ( + 1 1 8 O to +146'), facile acetylation of the 116-hydroxyl group were but the 116-esters of the Sa-bromo analogs6 have made on impure materials isolated directly fiom significantly higher A[M]D in the range of +202" The 110-acetates of triamcinolone fermentation sources, further chemical confirmation to +255'. of the suggested structures was made using pure range within these groups, being +157" to $279' triamcinolone and better controlled acetylation (see Table I). Some further chemical work was done to estab(4) 116- and lla-acetates are resistant t o alkaline hydrolysis; c f . lish the structure of 111, particularly to exclude (a) E. P. Oliveto, C. Gerold, L. Weber, H. E. Jorgensen, R. Rausser the alternate possibility of a 16a,l7a,21-triacetate. and E. B. Hershberg, THISJOURNAL, 75, 5486 (1953); (b) E. P. This alternate possibility is ruled against by Oliveto, C. Gerold and E. B. Hershberg, Arch. Biochcm. Bio$hys., 43, 234 (1953); (c) J. Romo, G. Rosenkranz, C. Djerassi and F. Sondmolecular rotational arguments' by the formation of heimer, THISJOURNAL, 75, 1277 (1953). an alkali-stable monoacetates and by the ultra( 5 ) (a) J. Romo, A. Zaffaroni, J . Hendrichs, G. Rosenkranz, C. violet absorption spectra which show the typical Djerassi and F. Sondheimer, Chemislry 6' Indusfry, 783 (1952); (b) hypsochromic shift associated with an 110-acetate D. H. Peterson, S. H. Eppstein, P. D. Meister, B. J. Magerlein, H. C. Murray, H. M. Leigh, A. Weintaub and L. M. Reineke, THIS JOURgroup; however, the following reactions establish NAL, 75,412 (1953); (c) P. D. Meister, D. H.Peterson, H. C. Murray, the 11@,16a,21-triacetatestructure firmly. G. B. Spero, S. H. Eppstein, A. Weintaub, L. M. Reineke and H. M. The monoacetate V formed from the triacetate Leigh, ibid., 7 5 , 416 (1953); (d) D. H. Peterson, A. H. Nathan, P. D. I11 (or from the tetraacetate IV) forms an acetonide Meister, S. H. Eppstein, H. C. Murray, A. Weinberg, L. M. Reineke and H. M. Leigh, ibid., 7 5 , 419 (1953); (e) R . Antonucci, S. Bernderivative I X with acetone-hydrochloric acid,3-g stein, M. Heller, R. Lenhard, R. Littell and J. Williams, J . O r g . Chem., 18, 70 (1953); (f) E. P. Oliveto, C. Gerold and E. B. Hershberg, Arch. Biochcm. Biophys., 49, 244 (1954); (9) A. L. Nussbaum, G. Brabazon, E. P. Oliveto and E. B. Hershberg, J. Org. Chcm., 22, 977 (1957); (h) E. P. Oliveto, R. Rausser, C. Gerold, E. B. Hershberg, M. Eisler, R. Neri and P. L. Perlman, ;bid., 23, 121 (1958). (6) C. H. Robinson, L. Finckenor, M. Kirtley. D. Could and E. P. Oliveto, THIS JOURNAL, 81, 2195 (1959). See also S. G. Levine and M. Wall, ihid., 81, 2826 (1959), for further examples of synthesis of grr-bromo-110-acetoxy steroids.
(7) Molecular rotational contribution for acetylation of the 17ac f . R. B. Turner, hydroxyl group is of the order of -209 to -299'; ihid., T 5 , 3604 (1953). (8) Alkaline hydrolysis of 17a-acetates proceeds as easily as the hydrolysis of 21-acetates; c f . Huang-Minlon, E. Wilson, N. L. Wendler and M . Tishler, ibid., 74, 5394 (1952). (9) (a) J. Fried, A. Borman, W. B. Kessler. P. Grabowich and E . F. Sabo, ihid., 80, 2338 (1958); (b) S. Bernstein, Rrcenf Progress i n Hormone Research, 14, 1 (1958).
2582
SIDNEY M. Fox, VICTORE. ORIGONIAND LELAND L. SMITH
Vol. 82
TABLE I MOLECULAR ROTATIONAL DATAFOR 1~D-ESTERS Steroid
Hydrocortisone 116-acetate 1lp,21-diacetate 1I@-formate21-acetate 21-acetate 5P-Dihydrohydrocortisone llp,21-diacetate 116-formate 21-acetate 21-acetate 116,1io-Dihydroxyprogesterone 116-acetate 116,17o-Dihpdroxy-5p-pregnane-3,2O-dione 116-acetate 4~-Bromo-5~-dihgdiohydrocortisone 116,21-diacetate 2 1-acetate 1l~-Hydroxy-4-androstene-3,17-dione 116-acetate 116-Hydroxytestosterone 1lff,li&diacetate 17P-acetate
Triamcinolone llp-acetate 116,16o,21-triacetate 16~,21-diacetnte 116-acetate 16a,17a-acetonide 1l@,Zl-diacetate16o,lia-acetonide 1601,17o-acetonide
[a1
$163" +163.2 +16i. 1 +l76.1 +35i.5
+ 92.8
+ 99
4- 86.6 $112.3 4-136 4-141.2 4- 69.6
+ 43.6 + 96.5
+ 99
4-203 $ 179
i-117.8 +121 to 125 67.1 97.5 55.4 f 96.1 22 63 4-128 -t138 109 104 92
+ + + + +
+
+ +
+
Solvent
[-\.l]D
Methanol Chloroform Chloroform Chloroform Dioxane
+ + +
Chloroform Chloroform Acetone Chloroform Acetone Chloroform Acetone
+416 4-430 352 408 +472 573 243
Chloroform Chloroform .ketone Alcohol Dioxane . ..... . .
Chloroform Methanol Methanol Chloroform Methanol Chloroform Methanol Methanol Methanol Chloroform Methanol Chloroform
+591° 687 746 761 +637
+ + ++ + 170 + 504 +480 +612 $615 $-456 +-I19 to 4-432 $264 +426 +289 500 105 300 $610 716 +474 +453 +437
+ + + +
A [ M ] D ~ ~ @Reference - ~ ~ Y ~ ~ ~ ~ ~ ~ d
f 96" 136 124
+
+
0
a ,b
C
d
+ 64 + 78 +165 (-73)
+ 24
a .b c
d
b
f b
d h
+
4-
3
I
31
I
h.i.i
1
+ 162 +184 +200
k
+157
k
k
rn
k
$279
k
n
k
21-acetate 16a,l7a-acetonide 901-Fluorohydrocortisone k 11@,21-diacetate 134 Methanol 4-622 f 82 21-acetate 127 Acetone 640 901-Bromoprednisolone 1lp,21-diacetate $159 Dioxane 4-832 240 P P 1I@-formate21-acetate 166 Dioxane 794 202 P 116-trifluoroacetate 21-acetate 141 Dioxane +814 222 1ID-diethylacetate 21-acetate 130 Dioxane +847 +255 P P 21-acetate 123 Dioxane $ 592 901-Chloroprednisolone P 116,2l-diacetate 4- 163 Dioxane i81 146 P llb-formate, 21-acetate 162 Dioxane i53 +ll8 21-acetate +I45 Alcohol 635 'I See ref. 3h. See ref. 4a. See ref. 5f. N. L. Wendler, R. P. Graber, R. E. Jones and M. Tishler, THISJOURNAL, 74, 3630 (1952). e R. H. Levin, et al., ibid.,75, 546 (1953). f E. P. Oliveto, T. Clayton and E. B. Hershberg, ibid., 75, 486 (1953). See ref. 5g. S.Bernstein, R. H. Lenhard and J. H. m'illiams, J . Org. Chem., 18, 1166 (1953). M. E. JOURNAL, 75, 5927 (1953). i 0. Mancera, G. Rosenkranz and F. Sondheimer, J . Chem. Soc., Herr and F. W. Heyl, THIS 2189 (1953). This report. See ref. 16. See ref. 3. See ref. 8a. J. Fried and E. F. Sabo, THISJOURNAL, 79, 1130 (1957). * See ref. 6. J. Fried, K. Florey, E. F. Sabo, J. E. Herz, A. R. Restivo, A. Borman and F. M. Singer, THIS JOURXAL, 77, 4181 (1955).
+ +
+
+ + + +
+
+ + +
+
+ + +
+
Q
which reaction is characteristic of cis-l,2-diols or 1,a-diaxial diols. l o From these possibilities the 16~~.17a-cyclic acetonide structure is assigned t o IX, as acetylation affords a reducing diacetate acetonide derivative X, t o which is assigned the structure 11@,21-diacetoxy-9a-fluoro- 16CY,17a-isopropylidenedioxy-l,4-pregnadiene-3,20-dione. hcetylation of triamcinolone 16a,l7a-acetonide (XI) 3s9
(10) (a) S. J. Angyal and C. G. Blacdonald, J . Chem. SOL., 686 (1952); (b) P. A. Sneeden and R. B. Turner, THISJOURNAL, 76, 3510
(1953).
with acetic anhydride and pyridine using an excess of reagent and warming, afforded the same diacetate acetonide X, which establishes the assigned structures of X and IX, and thus of V and I11 as 1lp-acetates of triamcinolone. In each instance ultraviolet absorption spectra, infrared absorption spectra, papergram mobilities and optical rotational data support the assigned structures. The ll/?,Bl-diacetate acetonide X could be hydrolyzed with hydrochloric acid to the 1I/?-acetate
1 EST ESTERS
May 20, 1960
CHzOR 1
CHrOAc I
P O
C=O
OF
2583
TRIAMCINOLONE CHzOAc
I
CHI~OAC
C=O
C=O
Xocnrdia
cornilina ___f
IIa, R = H
I
0
111
CHzOH I
VI
CHzOAc
C-0
& .
acetone-HC1
0
O
IX
VIIa. R
*OH
=
H
b . R = Ac
CH20H
TH~OAC
c -0
J$y-jyy /
/
0
X
0
XI
b, R-Ac
acetonide I X , leaving both the lip-estex group and the cyclic ketal group intact. Mineral acid is known to be ineffective in hydrolyzing 1Ga,17aketals of this seriesga Previous understanding of the acetylation of the 11B-hydroxyl group of the corticosteroid molecule has been that strong acid-condensing agents be used for proper acetylation,11 and that acetic anhydride-pyridine is not suitable for such acetylations. A recent report on the novel preparation of 9a-bromo- and chloro-1lp-acylates of prednisolone by addition of the elements of the acyl hypohalite to the appropriate 9(11)-dehydro steroid points out that direct acetylation of the 11P-hydroxyl in the presence of the 17a-hydroxyl is not possible, and that use of p-toluenesulfonic acid-acetic anhydride-acetic acid produces the 11@,17a,21-triacetates.6 The reaction of triamcinolone and Sa-fluorohydrocortisone with warm acetic anhydride and pyridine a t the 1I@-hydroxylgroup clearly permits one t o prepare the 11p-acetates of these 9afluoro steroids without complications fiom the 17a-hydroxyl group. The facile acetylation of the lip-hydroxyl might be attributed to the 901fluoro atom; however, similar acetylation experiments with hydrocortisone and 1Ga-hydroxyhydrocortisone afford 11P-acetylated products and thus require an alternate concept. Acetylation of the 17a-hydroxyl group of triamcinolone and of 1Ga-hydroxyhydrocortisone is influenced by the presence of the 1Ga-hydroxyl
group in the molecule, as the 17a-hydroxyl groups of the non-1Ga-hydroxylated steroids VI1 and VI11 are not acetylated under the same conditions.a The order of decreasing ease of acetylation of the four hydroxyl groups of triamcinolone becomes: 21 > 16a > ll@> 17a. With the recently reported high yields (90%) obtained for the microbiological deacetylation of hydrocortisone ll@,21-diacetate to yield hydrocortisone, l 2 the combination of 110-acetylation with warm acetic anhydride-pyridine and microbiological hydrolysis becomes more attractive as a protective measure in partial synthesis schemes for such 118-hydroxy steroids. Whereas previously described corticosteroids acetylated a t the ll@-hydroxylgroup are not bio, 1 2 some of the triamcinolone 118logically acetates possess weak corticoid activities. Triamcinolone 1I@-acetate lGa,l7a-acetonide (IX) possesses glucocorticoid activity; liver glycogen, 2 X hydrocortisone (1-4, 95% confidence limits) ; thymus involution, 3 X hydrocortisone (1-5) ; pellet test, 1 X hydrocortisone (0.2-5). Triamcinolone 118-acetate, while possessing no glucocorticoid activity in adrenalectomized rats, does exhibit a weak fluid diuretic activity in adrenalectomized and in normal rats. None of the other 1I@-acetate derivatives showed any corticoid activity. Acknowledgments.-The authors are grateful to Drs. I. Ringler and J. Cummings of these laboratories for biological data, to W. Fulmor for infrared
(11) (a) R. B. Turner, THISJ O U R N A L , 74, 4220 (1952); hIoEett and H. V. Anderson, i b i d . , 76, 747 (1954).
(12) W. Charney, L. Weber and E. P. Oliveto, Arch. Biochem. Biop h y s . , 79, 402 (1959).
(b) R. R .
2584
SIDNEY M. Fox, VICTOR E. ORICONIA N D LELANDL. SMITH
VOl. 82
absorption spectra, to L. Brancone for microanalytical data, to W. Muller for ultraviolet absorption spectra, and T. Foell for papergram examinations. Experimental l 3
(B) Controlled Acetylation Source .-Two grams of triamcinolone 16a, 21-diacetate was dissolved in a mixture of 50 ml. of pyridine and 20 ml. of acetic anhydride and the solution heated at 80' for 20 hours. After quenching with 40 ml. of methanol the solution was evaporated in vacuo, the residue dissolved in ethyl acetate, filtered through Florisil adsorbant to remove colored impurities, washed with bicar11P,16a,2 1-Triacetoxy-9~-fluoro-17a-hydroxy-l,4-pregnabonate and brine solutions, and dried over anhydrous magdiene-3,20-dione (111). (A)Fermentation Source.-Thirty grams of 16a,2l-diacetoxy-9a-fluoro-ll~,l7a-dihydroxy-4nesium sulfate. The solution was evaporated in vacuum to pregnene-3,aO-dione was dehydrogenated with Nocardia yield a crude product which analyzed as a mixture of I11 corallina under aerated submerged fermentation conditions (minor component) and IV (major component) by paperin the manner described by Bernstein, et al.,8 the steroids gram. Partition of the mixed acetates on Celite diatomaextracted with ethyl acetate, and the solvent removed in ceous earth using the toluene-petroleum ether ( b .p. 30-60')methanol-water, 12:s: 13:7, system yielded 0.21 g. of trivacuo. The semi-solid residue so obtained was a mixture of triamcinolone diacetate, mixed monoacetates and triam- amcinolone llP,l6a, 21-triacetate (111), identical with I11 cinolone alcohol, together with unaltered substrate, as evi- prepared under (A) or (C). The tetraacetate IV was redenced by paper chromatography. Reacetylation of the covered, and is described later in this section. (C) From Triamcinolone IlP-Acetate.-A solution of 4.75 residue was accomplished using 1000 ml. of acetic anhydride and 900 ml. of pyridine a t room temperature for 16 hours. g. of V in 25 ml. of pyridine and 3 ml. of acetic anhydride was kept a t room temperature for 6 hr. and then quenched The reaction mixture was diluted with 1500 ml. of methanol and 1000 ml. of toluene and the solvents then removed under with 10 ml. of methanol. After evaporation, etc., the resivacuum over a 4.5-hour period. During this period the tem- due was crystallized from ethyl acetate-petroleum ether, perature of the reaction mixture may have reached as high yielding 3.1 g. of triacetate 111, m.p. 221-223' dec., identias 80' inadvertently. The solids were partitioned on a 6- cal with I11 prepared by A or B above, as evidenced by ininch diameter column packed with Celite diatomaceous earth frared spectra and paper chromatography. 11@,16a, 1 7 ~l-Tetraacetoxy-9a-fluoro-1,4-pregnadiene~ ~ 2 using the solvent system ethylene glycol-petroleum ether3,20-dione (IV).-The reaction described under B above methylenechloride, 1:4:5. Aforerunof81. (0-1.0hold-back yielded 0.21 g. of the triacetate 111 on partition chromavolumes, HBV) contained the triacetate I11 and traces of IV and other tetrazolium blue-reducing impurities, together tography and 0.97 g. of the tetraacetate IV; after recrystalD with non-steroidal lipid fermentation impurities. Further lization from methanol, m.p. 217-219' dec., [ C Y ] ~ * +99.8", 235 mp ( e 15,760); : : :A 3.40, 5.72, 5.99, 6.13, 6.20, development of the column afforded fractions containing A, 7.30, 8.15, 9.11 r etc.; papergram mobility, system V14 triamcinolone 16a,al-diacetate and unaltered substrate. Rr 0.71. Isolation of steroids from the fractions I1 and I11 required Anal. Calcd. for CzoHasOloF: C, 61.91; H, 6.27; F , concentration in vacuum and crystallization from acetone3.38; acetyl, 29.88. Found: C, 62.01; H , 6.62; F , 3.16; petroleum ether. acetyl, 29.02. FracWt. isoPapergram 11P-Acetoxy-9a-fluoro-16a,17a,21-trihydroxy-l,4-pregtion HBV lated, g. Identity examination nadiene-3,2O-dione (V) .-Ten grams of triamcinolone was I 0-1.0 8.5 Triacetate I11 111 f traces of IV acetylated by the means described for preparation of the I1 1.0-1.5 2 . 4 5 Substrate I I only triacetate and tetraacetate (method B above). The product isolated, 14.1 g., which was mainly the tetraacetate but with 7.28 Diacetate I I b IIb only I11 1.5-3.0 some triacetate, was dissolved in 150 ml. of methanol, Isolation of the steroids from fraction I followed concen- purged of air with nitrogen, and 42.0 ml. of a 10% aqueous tration to a mobile oil and dilution with petroleum ether. solution of potassium carbonate added dropwise over 20 The precipitated solids were filtered, washed with petroleum minutes. After a further 15 minutes of stirring 4.2 ml. of ether, and dried, yielding 8.5 g. of tan colored solids which glacial acetic acid was added, then 300 ml. of 12.5% sodium analyzed on papergrams as a mixture of I11 (major) and IV chloride solution. The crystals were filtered, 8.7 g., and (trace) (Rrvalues in system V14 0.60 and 0.90-0.96; system recrystallized from %propanol; m.p. 228-230°, [ a l Z z D B1 of Bush's 0.41-0.50 and 0.90-0.97; system B3 of Bush15 +156.6', A,, 236 mp ( e 15,250);: : A: 2.93, 3.40, 5.73, 5.81 0.47-0.56 and 0.84-0.90). (shoulder), 5.99, 6.15, 6.20, 8.05, 8.12, 9.50, 11.26 r , etc.; The impure triacetate of fraction I was partitioned on papergram mobility in system 11, R f0.75 ( v s . triamcinolone Celite diatomaceous earth (2.0 g. of crude I11 on 200 g. of Ri 0.42). Celite diatomaceous earth) with the ethylene glycol-methylAnal. Calcd. for CZ~HZSO,F: C, 63.27; H, 6.70; F, ene chloride-petroleum ether, 1:5: 10 system, and the frac- 4.35; acetyl, 9.63. Found: C, 63.10; H , 6.90; F, 4.10; tion containing I11 (fractions 4-10, 50-ml. fractions col- acetyl, 10.7. lected) was concentrated in vacuo, the residue dissolved in Hydrolysis of pure triamcinolone 1lp,l6a,21-triacetate methanol and treated with charcoal and evaporated, yielding (111) with sodium methoxide in methanol gave the same prod650 mg. of 111, m.p. 218-220°, homogeneous on papergrams, uct, triamcinolone 1lp-acetate, although papergram analy[a]22D +55". After two recrystallizations from acetoneses of the products formed indicated some complete hydrolypetroleum ether the pure triacetate had m.p. 190.0-191.5', resolidifying by 195' with remelting 221.0-223.0' dec. sis to triamcinolone, and some partial hydrolysis to uniden~ (chloroform), A,, 236 mp ( e 14,900); tified diacetates, together with some isomerization to form (Kofler), [ a I z 24-55.4 triamcinolone isomer.lB The product isolated from the ( E : $ m )at , 15 min., 261 mp (260), 308 mp (114); a t 2 reaction was the 116-acetate V. hr., 260 m r (266), 308 mp (117); a t 20 hr., 260 m r (275), p , 16a,1701,2l-Tetraacetoxy4-pregnene-3,20-dione(VI). 3.05, 3.38, 5.71, 5.76 -A 1 1solution 308 mp (156) 375 m r (135); :;:1 of 0.5 g. of 16a-hydroxyhydrocortisone in 5.0 (shoulder), 5.98, 6.11, 6.17, 7.26, 8.07, 9.58, 11.16 f i , etc.; of pyridine and 2.5 ml. of acetic anhydride was heated at Rfsystem VI4 0.57; propylene glycol-toluene system, 0.83 ml. 80-85' for 26 hr., quenched with 10 ml. of methanol and cm./hr.; positive to tetrazolium blue. worked up as usual. The product, 0.28 g., crystallized from Anal. Calcd. for CnH3809F: C, 62.30; H, 6.39; F , acetone-petroleum ether, was recrystallized from aqueous 3.65. Found: C,62.12; H,6.84; F,3.51. 239 mp ( E acetone, m.p. 210-21l0, [ a ] 2 2 ~+57.4', A, 15,900): A":, 3.40, 5.70, 5.75, 5.97, 6.16, 7.28,8.10,9.42, (13) All melting points were taken in capillary except where noted 9.73, 10.62 ri-etc. (Kofler), which melting points were taken on a calibrated Kofler Anal. Calcd. for C2BH38010: C, 63.72; H , 7.01; acetyl, block under a microscope. Paper chromatographic examinations 31.5. Found: C, 63.48; H , 7.19; acetyl, 27.64. were conducted using systems already described." Optical rotations were made on 0 . 5 1 % solutions in methanol unless noted otherwise. 11p,2l-Diacetoxy-9~-fluoro-l7a-hydroxy4-pregnene-3,Infrared spectra were obtained on potassium bromide disks using the 20-dione (VIIIb) .-One gram of Sa-fluorohydrocortisone 21Perkin-Elmer model 21 double beam instrument. Ultraviolet abacetate was dissolved in 40 ml. of pyridine and 10 ml. of sorption spectra were obtained on absolute ethanol solutions using the acetic anhydride, warmed a t 80' for 20 hr. and worked up as Cary model 1 1 s recording spectrophotometer. usual. The mixture of 21-acetate and 11p,214iacetate was (14) L. L. Smith, T. Foell. R. De Maio and hI. Halwer, J. An%. Pharm. Assoc., 48, 528 (1959). (15) I. E. Bush, Biochem. J . . SO, 370 (1951).
(16) I,. L. Smith and M. Halwer, J . A m . Pharm. Assoc., 48, 348 (105'J).
May 20, 1960
D-ERYTHROSE WITH METHANOLIC HYDROGEN CHLORIDE
2585
Anal. Calcd. for C ~ H s ~ O ~C, F : 64.85; H, 6.80; F, partitioned on Celite diatomaceous earth using solvent system V14. A major fraction of 224 mg. of ll@,al-diacetate 3.66; acetyl, 16.21. Found: C, 64.43; H, 7.33; F, 3.97; and a fraction, 421 mg. of diacetate-monoacetate mixture acetyl, 15.58. were obtained. Recrystallization of the pure diacetate fracAcetylation of triamcinolone 11@-acetate 16a,l7a-acetotion from acetone-petroleum ether gave crystals, m.p. 215nide with acetic anhydride and pyridine afforded the same 2.90,3.40, 216, [ a l Z z D +134", Xmsx 236 mp ( e 15,800); llp,21-diacetate 16a,17a-acetonide, as evidenced by in5.73, 5.77(shoulder), 6.01, 6.15(shoulder), 7.29, 8.10, 9.07 frared spectra, papergram mobility and melting point crip , etc.; papergram mobility in system VI4 Rr 0.71. teria. 11~-Acetoxy-9~-fluoro-21-hydroxy-l6~t, 17a-isopropyliAnal. Calcd. for CBHIIOTF: C, 64.64; H , 7.16; F, 4.09; acetyl, 18.53. Found: C, 64.10; H , 7.39; F, 4.04; denedioxy-l,4-pregnadiene-3,20-dione(IX). (A). From Triamcinolone 11p-Acetate.-A solution of 1.5 g. of triamacetyl, 19.23. cinolone llp-acetate in 500 ml. of acetone containing 2.25 1lp,2 l-Diacetoxy-l7a-hydroxy-4-pregnene-3,20-dione ml. of concentrated hydrochloric acid was allowed to stand (VIIb).-One gram of hydrocortisone 21-acetate was dis- a t room temperature for 24 hr., a t which time the solution solved in 10 ml. of pyridine and 5.0 ml. of acetic anhydride, was neutralized with sodium bicarbonate solution, diluted warmed a t 80-85" for 24 hr., and worked up in the usual with water, and concentrated in vacuum to a volume of 200 manner. The residue obtained was crystallized from ethyl ml. The crystals formed were iiltered, washed with water, acetate-petroleum ether, yielding 0.96 g. of the 118,21-diace- and dried. The crvstals. 1.4 e.. were recrvstallized from tate, m.p. 185.0-187.5'' (Kofler), [a]2zD+168", 238 50% aqueous methinol, 'm.p.-215-217", [ o L I z 2 ~ +126.1°, mp ( e lF~,880).*~ Infrared spectra and papergram mobilities Amax 234 mp ( e 15,490); Ai: 2.90, 3.40, 5.72, 5.80, 5.99, were consistent with the assigned 11@,2ldiacetatestruc- 6.10, 6.19, 6.91, 7.27, 8.13, 9.25, 9.55, 11.20, 11.67 p , etc. ture. papergram mobility in system VI4 Rf0.69. 1lp,2 l-Diacetoxy-9a-fluoro-l6a, 17a-isopropyldienedioxyAnal. Calcd. for CZ~HIIO~F: C, 65.53; H, 6.98; F, 1,4-pregnadiene-3,20-dione(X).-Acetylation of 0.6 g. of 3.98; 9.03. Found: C, 65.25; H, 7.13; F, 3.80; triamcinolone 1 6 a , l 7 a - a ~ e t o n i d ewith ~ - ~ 10 ml. of pyridine acetyl,acetyl, 12.00. and 2.5 mi. of acetic anhydride by heating a t 90" for 15 hr. (B) From Triamcinolone 16a,l7a-Acetonide llp,21-Diaceafforded 0.7 g. of crystalline 1lfi,2ldiacetate when worked tate.-A solution of 170 mg. of triamcinolone 16a,l7a-aceup in the usual manner. Recrystallization from acetonepetroleum ether gave crystals, m.p. 230-232", [ ( Y ] * ~ D tonide ll@,2l-diacetate in 16 ml. of methanol was diluted +138", Xmax 236 mp ( e 15,300); ::A: 3.40, 5.70, 5.75(shoul- with 6.1 ml. of water and 2.0 ml. of concentrated hydroder), 5.98, 6.10, 6.19, 7.26, 8.13, 8.57,.9.25, 9.53, 10.98, chloric acid, and the mixture was refluxed for 3 hours. After 11.20, 11.68 p , etc.; papergram mobiliy in system VI4 dilution with 15 ml. of water, the solution was concentrated in vacuum to about 15 ml. The crystals formed, 0.12 g., Rr 0.92. were recrystallized from aqueous methanol. Identity of the acetonide llp-acetates formed by methods A and B was (17) Hydrocortisone llD,Zl-diacetate has been c h a r a c t e r i ~ e d ~ ~ . ~ ~ established by melting point, infrared spectral and paper as follows: (a) m.p. 188-189, [ a ] 2 6 ~ +107.1° (chloroform), XmsI chromatographic evidences. 238 mr ( 6 17,200); (b) m. p. 191.0-191.8°, [ a ]+167.1" ~ (chloroform), PEARL RWER,N. Y . X 9 6 ~ ~ ~ 240 0 H mr (17,200).
[CONTRIBUTION FROM
THE
DEPARTMENT OF BIOCHEMISTRY, UNIVERSITY OF CALIFORNIA, BERKELEY, CALIF.]
Reaction of D-Erythrose and 2,4-O-Ethylidene-D-erythrosewith Methanolic Hydrogen Chloride BY CLINTON E. BALLOU RECEIVED OCTOBER 6, 1959 The reaction of 2,4-O-ethylidene-~-erythrose with methanol containing hydrogen chloride gives, as the main product, methyl 2,3-O-ethylidene-P-~-erythroside. The migration of the ethylidene group and the formation of two fused fivemembered rings apparently accounts for the stability of this product. Free D-erythrose reacts under the same conditions to give mostly methyl 6-D-erythroside, contaminated with a little of the a-anomer. Pure methyl j3-D-erythroside has a specific hydrolyzes completely in aqueous acid only if the acetaldehyde is allowed rotation of - 149". 2,4-O-Ethylidene-~-erythrose by the migration of the ethylidene to escape. There is a strong tendency for the formation of 2,3-O-ethylidene-~-erythrose group, but the same product is formed from free D-erythrose in aqueous acid containing 5% acetaldehyde. The reduction of the 2,3-O-ethylidene-~-erythrosegives 2,3-0-ethylidene-erythritol.
Several past investigations, particularly by Hockett and Maynard,* have been concerned with the reaction of the tetrose, erythrose, with methanol in the presence of an acid catalyst to give acetal derivatives. The only well defined product obtained has been a methyl D-erythroside; although the possible formation of acyclic acetals, as well as acetals of dimers of a dioxane type structure, has been the basis of conjecture. During an attempt to prepare methyl D-erythrowith side by treating 2,4-O-ethylidene-~-erythrose methanol containing 1% hydrogen chloride, we were surprised to find the rotation of the solution become strongly levorotatory. On the assumption (1) Presented in part at the Meeting of the American Chemical Society in Chicago, Ill., September, 1958. (2) R. C. Hockett and C. W. Maynard, Jr., THIS JOURNAL, 61, 2111 (1939); G. F. Felton and W. Freudenberg, ibid., 67, 1037 (1935).
that the ethylidene group was coming off by alcoholysis and the resulting free D-erythrose was being converted to methyl D-erythroside, the rotation indicated that the product must be primaIily a 0anomer. This would be a t variance with the observations of Hockett and Maynard that a solution of D-erythrose in methanolic hydrogen chloride became only slightly levorotatory. When the products from the above reaction of 2,4-0-ethylidene-D-erythrose were isolated, the major component was found to be methyl 2,3-O-ethylidene-~erythroside, indicating that the ethylidene group had migrated, but had remained, for the most part, attached to the sugar molecule. A lesser amount of a methyl D-erythroside also was obtained. The rotations of both of these products were strongly levorotatory and suggestive of p-anomeric forms.
