7390
Journal of the American Chemical Society
(26) Hydroxide ion participation was ignored on the basis of its low concentration under the reaction conditions. (27) Kice, J. L.; Rogers, T. E.; Warheit, A. C. J. Am. Chem. SOC. 1974, 96, 8020.
(28) Hibbert, F.; Awwal, A. J. Chem. SOC.,Perkin Trans. 2 1978, 939. (29) No isotope effect on the kl step is expected. (30) Bell, R. P.;Crooks, J. E. J. Chem. SOC.1962,3513.
/
101:24 J, November 21, 1979
(31) The kl values of Table 111 for reactions of secondary amines with 1, 4, and 6 and the k2 values of Table I for reactions of primary amines with 1-6 are assumed to be associated with the same nucleophilic attack step ( k l ) of Scheme I or II. (32) We believe that this loss of absorbance at 300 nm in more concentrated (0.02-0.08 M) mercaptoethanolsolution is due to Michael reaction of RSH with RSCH=CHCOCH3 to give (RS)~CHCHZCOCH~.
Acid- and Base-Catalyzed Dehydration of Prostaglandin E2 to Prostaglandin A2 and General-Base-Catalyzed Isomerization of Prostaglandin A2 to Prostaglandin B2 S. K. Perera and L. R. Fedor* Contribution from the Department of Medicinal Chemistry, School of Pharmacy, State Uniljersity of New York at Buffalo, Buffalo, New York 14260. Receiued June 29, I979
Abstract: Dehydration of prostaglandin EJ(PGE2) to prostaglandin A2 (PGA2) in aqueous solution is catalyzed by hydrogen , ion ( k d o ~ and ) , quinuclidine ( k d ~ )The . values of k d H and k d o are ~ very similar t a those for various ion ( k d ~ ) hydroxide known enolization reactions and k d ~ ( H z O ) / k d ~ ( D 2 = 0 ) 0.6, results that support a mechanism involving rate-determining enolization of PGE2 during dehydration. Isomerization of PGA2 to prostaglandin B2 (PCB2) is general base catalyzed by tertiary amines with k L ~ ( H 2 0 ) / k ' ~ ( D z = O )I . I . The rate constants k l A and k'OH for loss of PGAz arc equal to those for formation PGC2 of PGB7and it is suggested that C-12 proton abstraction is rate determining in the reaction sequence PGAz PG B2.
- -
The clinical usefulness of prostaglandin E2 (PGE2) in 0
bH PGE,
human r e p r o d u c t i ~ n ' -coupled ~ with its chemical instability in aqueous s ~ l u t i o n ~has - ' ~prompted a research activity directed toward preparation of PGE2 prodrugs and pharmaceutical formulation^'^-'^ that retain biological activity and possess greater stability. The chemical instability of E and A series prostaglandins was shown6-I3 to involve dehydration of the 9,l I-keto1 to PGAs in acidic and alkaline solution (eq l ) , isomerization of PGAs to PGCs in alkaline solution (eq l ) , isomerization of PGCs to PGBs in alkaline solution (eq l ) , 0
PGAs
0
PGCs
0
PGBs
epimerization of PGEl to 8-iso-PGEl in ethanolic potassium acetate, epimerization of prostaglandins to 15-epiprostaglandins in dilute acid, and allylic rearrangement of 15-epiPGA2 to the 13-hydroxy diastereomers of PGA2. Biochemical transformations of prostaglandins catalyzed by enzymes include dehydration of PGE2 to PGA2,1g-20isomerization of PGAs to P G C S , ~ I isomerization -~~ of PGCs to PGBS,~Oreduction of the 9-keto group to the ~ a r b i n o l , ' ~oxidation ,~~ of the C - I 5 carbinol to the k e t ~ n e ,and ~ ~reduction .~~ of the AI3 double Little is known\ of the chemistry of the dehydratase and isomerase enzymes except that they are inhibited by sulfhydryl reagents and they likely do not req&re cofactors. I 8-21,23 This study was undertaken to probe the nature of the acid0002-7863/79/1501-7390$01 .OO/O
-
-
base-catalyzed dehydration-isomerization sequence PGE2 PGA2 PGC2- PGB2 depicted in eq 1 with a view to contributing to a better understanding of the mechanisms of these transformations. Owing to the variety of prostaglandin chemistry that takes place simultaneously in aqueous solution a t any pH, the rate constants reported in this kinetics study do not strictly,represent the chemistry of eq 1: 8- and 15-epi-PGE2 and PGA2 should be formed during dehydration. However, these epimers are expected to have similar reactivities to PGE2 and PGA2 with respect to ring dehydration-isomeri'zation and we suggest that our conclusions of mechanism are little affected by these extraneous events.
Experimental Section Apparatus. Gilford Model 2400, Beckman Model DBG, and Cary Model 1 18 spectrophotometers were used. Temperature was maintained with a Tamson T9 circulating water bath connected to thermospacers in the Gilford spectrophotometer. pH measurements were made with a Radiometer P H M 26 meter with GK2321B or GK2321C electrodes. Reagents. PGEz was a gift from Upjohn Co. All reagents were Fisher certified A C S grade except quinuclidine, quinuclidinol (Aldrich), triethylamine (Eastman), DzO, DCI (99.9% D) (Stohler Isotope Chemicals), K2HP04, and K3P04 (Sigma). Line distilled water was redistilled through a Corning AGla still before use. Kinetics. All solutions had a calculated ionic strength of 0.5 M (KCI) unless otherwise stated. The temperature of reactions was 30 f 0.1 OC. The pHs of reactant solutions were measured and found to be constant ( f 0 . 0 4 p H ) for all serial dilutions except HCI/KCI and KOH/KCI. Reactions were run under pseudo-first-order conditions and were initiated by addition of PGEz in absolute ethanol to aqueous solutions of reactants. The concentration of ethanol was ca. 1 % and that of PGE2 ca. 10-4-10-5M in the 3 3-mL cuvettes. pK, values were determined by the method of fractional neutralization. pD was determined from the pH meter reading by adding 0.4 to it.30 Hydroxide ion activity was determined from K,/aH where pK, = 13.83 at 30 "c3'and deuterioxide ion activity was determined from K D / a D where P K D = 14.65 at 30 0C.32 Reaction of PGE2 with hydroxide solution is characterized by an initial increase in optical density (OD) followed by a slower decrease at 224 nm. This decrease is accompanied by an increase in O D at 280
0 1979 American Chemical Society
Perera, Fedor
/
Dehydration of Prostaglandin E2
7391
Table I. Rate Data for Dehydration of PGE2 to PGA2 in Aqueous
Table 11. Rate Data for Isomerization of PGAz to PGBz in
Solution a
Aqueous Solutiono
k d z , M-' catalyst
min-'
fraction of base
c
1.48 f 0.07 1.60 f 0.03 2.23 f 0.03 16.0 f 0.7 0.004 31 f 0.000 02 0.006 97 f 0.000 54
Qh
Q Q OHHCld DCld
0.4 0.5 0.6
concn range of catalyst,
no. of
M
kobcd
0.002-0.02 0.001-0.02 0.002-0.02 0.01-0.1 0.05-1.0 0.5-1 .o
16 15 8 23 13 6
k'2, M-I catalyst6
min-'
Q' Q' Q' Q Q QO
1
e
1.13 f 0.26 0.92 f 0.06 0.73 f 0.07 1.43 f 0.08 1.09 f 0.07 0.005 48 f 0.000 04 0.0235 f 0.001 2 0.0324 f 0.0016 0.0327 f 0.0017 0.0266 f 0.000 44 1.59 f 0.05 1.69 f 0.19
concn range fraction of catalyst, no. of of base M kobcd 0.4 0.5 0.6 0.5 0.6 0.5 0.4 0.5 0.6 0.5
0.002-0.02 0.001 -0.02 0.002-0.02 0.02-0.2 0.02-0.2 0.04-0.2 0.05-0.5 0.05-0.5 0.05-0.5 0.1-0.5 0.0 1-0.1
20 16 9 9 9 6 IO IO IO 6 20 20
f' f = 30 f 0.1 OC, p = 0.5 (KCI), reactions monitored at 224 nm. All Q's are quinuclidine in 0.01 M phosphate buffer, pK,(Q) = 1 1.25. ' kd2 is a general expression for the second-order rate constant for dehydration of PGE2 catalyzed by Q ( k d A ) ,O H - ( k d o H ) ,and HCI, DCI ( k d H , k d o ) . p = 1.0 (KCI).
TEA TEA TEA T E A (D2O)f OH-d OH-
nm. The O D increase at 224 nm is due to formation of the conjugated ketone system in PGAz by dehydration of P G E z . ~The ) subsequent decrease in O D at 224 nm is due to loss of PGA2 by isomerization and the corresponding increase in O D at 280 nm is due to the formation of another conjugated ketone, PGB? (eq Kinetically these reactions are
a t = 30 f 0.1 "C; p = 0.5 (KCI); reactions monitored at 280 nm unless otherwise stated. Quinuclidine (Q), pK, = 1 I .25; quinuclidin01 ( Q o l ) , pK, 9.95; triethylamine (TEA), pK, = 10.84. Quinuclidine in 0.01 M phosphate buffer; reactions monitored at 224 nm. Reactions monitored a t 224 nm. kl2 is a general expression for the second-order rate constant for isomerization of PGA2 catalyzed by amines ( k ' A ) and hydroxide ion (k'OH).f pK, (D20) = 11.47.
ki
k2
A*B*C reactions,34 and kl and k2 ar'i pseudo-first-order rate constants. The concentration of B (PGA2) at any time is given by [B] = A o k l / ( k * - k ~ ) { ( e - ~-l I where A0 is the concentration of PGE2 at zero time. For targe values of f, k2 was determined from the slopes of plots of In (OD, - OD,) vs. t . The constant kl was determined from slopes of plots of In {e-k2r - [(OD, - OD,)/C]) vs. t. C i s a constant and is t h e y intercept of those plots used to determine k?. Plots were linear to 75% reaction or beyond. Reactions of PGE2 with hydroxide ion, triethylamine, quinuclidine, and quinuclidinol solutions were studied at 280 nm and were characterized by an initial lag period followed by a usual rectangular hypcrbolic O D increase vs. t due to the formation of PGB2. N o Michael addition of hydroxide ions to PGB2 was detected as evidenced by stable OD, values for these reactions under the conditions of the study. Pseudo-first-order rate constants were obtained from slopes of plots of In {(OD, - ODo)/(OD, - OD,)) vs. t . Plots were linear to 75% reaction or beyond when arbitrary t = 0 was taken beyond the initial lag period. Reactions of PGE2 with quinuclidine solutions, monitored at 224 nm, are experimentally complicated by the virtually instantaneous reaction of the product PGA2 with the amine as well as by the poor transparency of amine solutions at this wavelength. The reaction results in decreased absorption at 224 nm and this decrease is a function of amine concentration. W e ascribe this to addition of quinuclidine and a proton across the 10.1 I-olefin bond of PGA2.I5 However, dehydration of PGE2 and isomerization of PGA2 catalyzed by dilute quinuclidine solutions in 0.01 M phosphate buffer to maintain constant pH were studied at 224 nm. No catalysis by the phosphate buffer was detected, although no systematic study of phosphate buffer catalysis was made. Pseudo-first-order rate constants were calculated for an A B C sequence as described above. Optical density increase a t 224 nm vs. time was recorded to follow the dehydration of PGEz to PGA2 in aqueous HCI. Pseudo-first-order rate constants were obtained from the slopes of plots of In {(OD, ODo)/(OD, - OD,)1 vs. t and they were linear to 75% reaction or beyond. The OD, value decreased with time for reactions run in HCI solutions more concentrated than 3 M. The equilibrium constant K , = [PGA>]/[PGE*] = 161 was calculated from e224 595 M-' cm-' (five measurements) for PGE2, €224 I O 71 5 M-' cm-I for PGA2,35 PGA>), and O D at the concentration of added PGE2 (=PGE2 equilibrium .
--
+
Results Dehydration of PGEz. Dehydration of PGE2 to PGA2 in dilute potassium hydroxide solution obeys the rate law
d[PGA2]/dt[PGE2] = kobsd = kdoHaoH
(2) where aOH is the activity of hydroxide ion determined from the
0.01-0.1
pH of reactant solutions and pK,. The constant k d o(Table ~ 1) was evaluated by dividing kobsd by aOH for unbuffered runs a t pH > 11. The equilibrium dehydration of PGE2 to PGA2 in dilute acid solution obeys the rate law d[PGA2]/dt[PGE2] = kobsd = kdHaH
(3) where aH is the activity of the hydrogen ion. The constant kdH (Table I) was obtained by dividing kobsd by aH for runs in dilute HCI/KCl solutions. From K, = 161, the rate constant for dehydration is 4.28 X M-' min-l and that for hydration is 2.66 X M-l min-I. The deuterium solvent kinetic isotope effect (KIE), H20/D20, is 0.62 (Table I). Dehydration of PGE2 to PGA2 catalyzed by dilute quinuclidine (Q) solutions buffered with 0.01 M phosphate follows the rate law5' d[PGA2]/dt[PGE21 = kobsd = k d p f ~ [ Q t l+ kdOHaOH (4) The constant k d (Table ~ I) was obtained by dividing the slopes of plots of kobsd vs. [Qtl, where [Qtl = [QI + [QHI, by f ~the , fraction of base form of Q present in the Q/QH buffer solution. Isomerization of PGAz. In dilute potassium hydroxide solution, isomerization of PGA2 followed a t 224 and 280 nm obeys the rate law -d[PGA2]/dt[PGA2] = d[PGBz]/dt [PGAz] = kobsd = k'OHaoH + k0 (5) The rate constant k l (Table ~ ~ 11) was evaluated for each wavelength as the slope of a plot of kobsd vs. aoH and the two constants are identical within the limits of experimental error. The intercept ko has thevalue 0.001 f 0.0019 min-I. We believe that ko is real, although its value is uncertain, because plots of k&d vs. concentration of amine for isomerization reactions run in aqueous amine solutions (vide infra) consistently gave intercepts a t a variety of pHs that were larger than those computed from k'OHaOH by ca. 1.5 X min-1.5* Isomerization of PGA2 to PGB2 in very dilute quinuclidine (Q) solutions buffered with 0.01 M phosphate and monitored a t 224 nm obeys the rate law -d[PGA2] /dt [PGA2] = kobsd = k'i;f~[QtI + k'oHaoH
+ ko
(6)
7392
Journal of the Aniericati Chemical Society
where f B and Q1are previously identified symbols (vide supra), The rate constant k'n (Table 11) was obtained by dividing the slopes Of plots of kob5dvs. [Qt] a t constant buffer ratio byfB. The rate law of eq 6 also holds for isomerization of PGA2 to PGB2 catalyzed by triethylamine (TEA) and quinuclidinol (Qol) monitored a t 280 nm. For the TEA-catalyzed reaction the deuterium solvent KIE, HzO/D20, is 1 . 1 , Isomerization of PGA2 to PGB2 followed a t 280 nm and catalyzed by higher concentrations of Q than were used to monitor the same reaction at 224 nm obeyed the rate law d[PGB2]/dt[PGA2,] = kcorob\,j = k ' h f ~ [ Q t l / ( K f ~ [ Q t l I H +f 1 1) ( 7 )
+
where [PGA2,] = [PGAz] [PGA2-Q adduct] (Experimental Section), kcorob\,j= k&\d - (k'oHUoH -I-ko),fB and Qt are as previously defined (vide supra), and K = [PGA2-Q adduct]/ [PGAl] [ Q t l f ~ [ H + ]The . constants k l A (Table 11) and K were evaluated by plotting l/kcorob\dvs. l/[Qt], which gave /
