- Email: [email protected]
Conformational study of the acetyl group and cyclohexenone in progesterone interacting with phosphatidyllipid by means of circular dichroism
J. steroid Eiochem. Vol. 31, No. 5, pp. 853-860, Printed in Great Britain.All rights reserved
1988
0022-4731/88
$3.00 + 0.00
Copyright0 1988PergamonPressplc
CONFORMATIONAL STUDY OF THE ACETYL GROUP AND CYCLOHEXENONE IN PROGESTERONE INTERACTING WITH PHOSPHATIDYLLIPID BY MEANS OF CIRCULAR DICHROISM SATORUWATANABE*,TAIICHI SAITO*and MITSUNORIIKEDAt *Departments of Pharmacology and tBiochemistry, Kawasaki Medical School, Kurashiki, Okayama 701&O],Japan (Received 28 July 1987; received for publication 22 March 1988)
Summary-The interaction between the A-ring and the 17-acetyl groups of progesterone (PROG) and various concentrations of distearoyl-, dipalmitoyl-, dioleoyl- and diarachidoyl-L-a-phosphatidylcholines, and dipalmitoyl+a-phosphatidyl-DL glycerol in methanol and chloroform solutions and its preferred conformational assignments in the presence of those lipids were examined qualitatively by circular dichroism on the basis of PROG spectra in the wavelength regions of 26&4OOnm. PROG did not interact with saturated distearoyl and dipalmitoyl phosphatidylcholines, but did with unsaturated dioleoyl and diarachidoyl phosphatidylcholines, and saturated dipalmitoylphosphatidylglycerol. The interacting moieties of PROG were an a$-unsaturated cyclohexenone of the A-ring for oleoyl and glycerol lipids, and the 17-acetyl group for unsaturated and glycerol lipids. The interaction with these lipids, the rotational conformations of the 17-acetyl group, and invertible conformations of the cyclohexenone of PROG were discussed on the basis of the elliptical strength of the Cotton effect and energy estimation of the preferred conformers. Oleoylphosphatidylcholine caused an increase in slightly energetically unstable conformers of the acetyl group and stable conformers of the a$-unsaturated cyclohexenone. Glycerol lipid, on the other hand, caused an increase in energetically unstable conformers of cyclohexenone, but it was similar to the effect of oleoyl lipid on the 17-acetyl group. Diarachidoylphosphatidylcholine, with eight double bonds, on the other hand, increased the number of energetically stable conformers of the 17-a&y1 group, but had no effect on the conformation of cyclohexenone. It became apparent that the double bond of hydrocarbon moiety as well as the head group of choline and glycerol in lipids were closely related to the conformational populations of both groups of the PROG molecule. The specific effect on the conformations of the acetyl and a,/?-unsaturated cyclohexenone of PROG of various lipids with different substitutions in their heads or hydrocarbon moieties might in part explain the nongenomic action of the steroid.
INTRODUCTION The receptor mediated pathway of steroid hormone action is not sufficient to fully account for the known effects of steroids. Some investigators [l-3], for example, have demonstrated that progesterone and other steroid molecules can promote the maturation of Xenopus fuevis oocytes, although these cells do not contain steroid receptors. Kaya and Saito [4] also recently observed that progesterone strongly inhibited hemolysis of human erythrocytes against osmotic shock within several minutes after exposure to the agent. These findings suggest that, in addition to its well known genomic action, progesterone operates through other mechanisms; particularly through direct action on the plasma membrane. It has been said that such a direct effect of steroids must be related to a change in the membrane function, which can be modified by a partition coefficient difference between lipid and water [5,6], cell excitability [7, 81, transport [9-121, the release effect on acid phos-
phatase from lysosomes and increases in cation leakage from liposomes [13], membrane expansion [ 141, and/or an enzymatic effect [15]. In addition to these factors of cell function, the direct interaction between steroids and lipid molecules which can cause alteration in cell function is significant, because the plasma membrane consists mainly of various phospholipids that stabilize its architecture. Consideration of the conformations of the moieties of the A-ring and 17-acetyl group during the interaction between progesterone and lipids should not only be useful in revealing the direct action mechanism, but also in developing new steroids that may be, of biological significance. Although circular dichroism appears to be more suitable for the monitoring the conformational alteration of optical active compounds because of its peculiar sensitivity to subtle conformational alterations, it has not been used for investigation of steroid-lipid interaction prior to our study except for some measurements of systems containing steroids and proteins [ 16, 171.
853
SATORUWATANABE et
854
al.
Table I. CD results of interaction between 100 CM PROG and various lipids at different concentrations
diarachidoyl-L- u phosphatidylcholine
0000x
All solutions contained 100 PM of progesterone.
006OQOO 0: Unaltered CD. 0: Altered CD. x : Insolubility.
EXPERIMENTAL
Reagents
All the chemicals used were analytical grade reagents employed without further purification. Distearoyl-, dipahnitoyl-, dioleoyl- and diarachidoyl-La-phosphatidylcholines and dipalmitoyl-L-a-phosphatidyl-DL-glycerol as well as progesterone and testosterone were purchased from Sigma, St Louis, U.S.A. Circular dichroism measurements
CD measurements were made at 25°C with a Union Giken Mark III spectropolarimeter calibrated with (+)-D-camphorsulfonic acid and equipped with a computerized data processing system. All spectra were recorded in a quartz cell with lO.Omm path length at a scanning range of 26tX-400 nm using a full scale deflection of 0.02 and a spectral band width of 2.0 nm. The spectra in each figure have been recorded at the same day on the same chart using a freshly prepared sample to minimize errors. Results were expressed as molar ellipticities, [e] (deg. cm2 dmol-‘), calculated with reference to the PROG concentrations. The PROG concentration was constantly 100 PM throughout the experiment.
both solvents, and to those of dipalmitoyl-L-a -phosphatidyl-DL-glycerol and diarachidoyl-L-a-phosphatidyl-choline in methanol regardless of the concentrations of lipids, with the exception of one case of insolubility. Consequently, CD spectra of PROG’s interaction with dioleoyl- and diarachidoyl-L-a -phosphatidylcholines, and dipalmitoyl-L-a -phosphatidylDL-glycerol in chloroform and dioleoyl-L-a-phosphatidylcholine in methanol is significant. The CD curve of PROG in methanol showed a negative Cotton effect at 320 and a positive one at 288 nm in agreement with a previously noted curve [ 181,as shown in Fig. 1. PROG has two optical
16
RESULTS The interaction between PROG and various concentrations of distearoyl, dipalmitoyl-, dioleoyl- and diarachidoyl-L-a-phosphatidylcholines, and dipalmitoyl-L-a-phosphatidyl-DL-glycerol in methanol and chloroform solutions and its preferred conformational assignments were examined qualitatively with a CD spectropolarimeter on the basis of PROG spectra in the wavelength regions of 260-400 nm at room temperature. Measurement results are summarized in part in Table 1. The CD spectrum of PROG was quite similar to curves in the presence or absence of distearoyl- and dipalmitoyl-L-a-phosphatidylcholines with saturated aliphatic moiety in
260
280
3al
320
340
360
uy)
Fig. 1. CD spectra for 100 pM of PROG and testonslkrone. The CD spectrum of testosterone was subtracted from that of PROG in methanol in the 26@4OOnm region. PROG: (---), testosterone: (--) and the difference: (----).
CD study of progesterone-phosphatidylhpid
280
300
320
340
absorption moieties of an a$-unsaturated ketone with the A-ring and of the 17-acetyl group. Judging from the absorption wavelength and the n - II* transition, it seemed very reasonable to assign 288 and 320nm to the saturated ketone absorption and unsaturated ketone absorption, respectively. In addition, when discussing changes of the rotational strength in association with conformational changes in respective ketones, it is
important to clarify whether negative and positive CD curves overlap or cancel each other out with opposite sign ellipticity. The degree of overlap can be best seen in the difference spectrum presented in Fig. 1. Here, the testosterone spectrum was subtracted from the PROG spectrum obtained at the same concentration in the same solvent, the only structural difference between PROG and testosterone being a substitution of the 17-acetyl group for the hydroxyl group, which does not contribute to elliptical absorption at all. The resulting CD peak at 288 nm can almost certainly be attributed to net absorption of the acetyl group, and can be consequently justified by the above-mentioned assignment of the PROG peaks. The negative elliptical strength of the subtracted CD curve at 288 nm was partially cancelled by the negative absorption of the A-ring but it was practically nil at 330 nm. This suggests that the unsaturated ketone on the A-ring absorption of PROG had no influence, in spite of a slightly positive Cotton effect due to the 5.0 nm shift. Thus the CD curve of PROG was composed of negative and positive extrema at 288 and 320 nm. These extrema were assigned to the independent absorptions of the 17-saturated acetyl and a$-unsaturated ketone groups, respectively. Accordingly, the rotational strengths of the 288 and 320 nm bands responsible for conformational change may be used to study the interaction between the PROG molecule and various lipid molecules and its preferred conformation assignments in their presence. The CD spectra of 100pM PROG in the presence of various concentrations of dioleoylphosphatidylcholine were composed of negative and positive bands with maxima at 330 in methanol and
260
300
320
855
340
3
360 nm Fig. 2. CD spectra of PROG in the presence of various concentrations of dioleoyl-L-a-phosphatidylcholine (left) and in the absence of 1.2 and 9.6 mM lipid concentrations in methanol in the 260-360 nm region subtracted from spectra in their presence (right). Left: no lipid: (-), 1.2 mM: (---) and 9.6mM: (----). right: 1.2 mM: (---) and 9.6 mM: (----). 260
280
interaction
288 nm in chloroform, as shown in Figs 2 and 3, respectively. When the concentration of the lipid was increased gradually in the order of 0.1, 0.3, 0.6, 1.2, 2.44.8 and 9.6 mM against a constant concentration of PROG, a decrease in both positive and negative elliptical strengths dependent on the concentration difference began from 4.8 in methanol and 1.2 mM in chloroform. Therefore, the difference in the effect of the lipid on the starting point of the conformational change of PROG may be dependent on a solvent polarity. Such a solvent polarity could influence the interactions of lipid-PROG and intramolecular PROG, since the polar solvent accessible to the oxygen of the polar functional group in the ketones probably prevents lipid attacks on this moiety at low
I 16 .
t
260
280
MO
320
YO
360
3aa nm
Fig. 3. CD spectra of PROG in the presence of various concentrations of dioleoyl+a-phosphatidylcholine and in the absence of 2.4, 4.8 and 9.6 mM lipid concentrations in chloroform in the 26WOOnm region subtracted from spectra in their presence. No lipid: (-), 2.4mM: (---), 4.8 mM: (---) and 9.6mM: (---). The difference: 2.4mM: (---). 4.8 mM: (----) and 9.6 mM: (----).
SATQRU WATANABE et al.
856
concentrations. Modifications in rotational strength due to absorption of the A-ring and the acetyl group of PROG by lipids were first observed by this author in the current work, and provide evidence of molecular interaction between PROG and lipids. The spectra seen on’ the left in Fig. 2 cannot be fully ascribed to such modification because of noise when PROG interacted with the lipid in methanol. The change produced can be better seen on the right in the same figure. Here, the spectrum in the presence of 100 PM of PROG was subtracted from the spectra in the presence of various concentrations of oleoyl lipid. The resulting curves showed a positive Cotton effect around 330 nm and a negative one at 288 nm in a concentration of 9.6 mM of the lipid. However, the subtracted spectrum at 1.2 mM exhibited a curve with no peaks. The curve of PROG in the presence of 1.2 mM of lipid therefore appears to actually be the same curve as that in the absence of the lipid, although a slight shift must be allowed for the line. In contrast to effect of the saturated lipid on the CD curve of PROG, the decrease in the rotational strength caused by the oleoyl lipid, which is substituted for the unsaturated hydrocarbon group in the hydrocarbon moiety of saturated distearoyl- and dipalmitoyl-phosphatidylcholines, evidently indicates that the double bond for rotational modifications in the conformation of the acetyl group and for inversional modifications in the a$-unsaturated cyclohexenone in PROG plays important role. Figure 4 shows the CD spectra of PROG in the presence of various concentrations of dipalmitoylphosphatidylglycerol measured at the 26&400 nm region in chlo16
roform. The rotational intensity of the CD curves at 288 nm decreased with increases in the lipid concentration, while that at 330nm on the negative band increased slightly. The concentration needed for conformational modification was lower for the acetyl group than for cyclohexenone. The structural difference between dipalmitoylphosphatidylglycerol and dipalmitoylphosphatidylcholine is the moiety of glycerol. This modification of the curve by the lipids thus seems to reveal a specific contribution of the glycerol group to the absorption due to conformational change in the acetyl group and the cyclohexenone ring in PROG. Choline moiety, on the other hand, had no influence. Since the CD curves of the acetyl group conformationally modified by glycerol lipid were very similar to those of oleoyl lipid, and since there are marked structural and positional differences between oleoyl and glycerol lipids, it is easy to conclude that there is a different mechanism at work. The marked structural and positional differences also may provide a reasonable explanation for the different effect to the oleoyl and glycerol lipids on the CD curves of the a&?-unsaturated cyclohexenone ring. The CD spectra attributable to the acetyl and cyclohexenone groups of PROG in the presence of various concentrations of diarachidoylphosphatidylcholine differed significantly from those coexisting with the oleoyl or glycerol lipids tested in this experiment, as shown in Fig. 5. That is to say, the rotational strength of the 288 nm band increased with increases in the arachidoyl lipid concentration, but there were no changes at 330 nm peak regardless of the increase 16,
12
8
-4
t 260
280
3m
320
340
360
380 “rll
Fig. 4. CD spectra of PROG in the presence of various concentrations of dipalmitoyl-r.-a-phosphatidyl-DL-glycerol and CD spectra in the absence of 2.4,4.8 and 9.6 mM lipid concentrations in chloroform in the 260-4OOnm region subtracted from spectra in their presence. The concentration conditions for each curve were-the same as in Fig. 3. The CD soectrum of PRGG in 9.6 mM of lipid was almost the sameas that in 4.8 mM, although it was &htly weaker and stronaer at oositive and neaative absorotions. resnectivelv.
260
280
300
320
340
360
380 nm
1
Fig. 5. CD spectra of PRGG in the presence of various concentrations of arachidoyl-L-a-phosphatidylcholine and CD soectra in the absence. of 0.6, 1.2 and 2.4mM lipid concentrations in chloroform in the 260-4OOnm region subtracted from spectra in their presence. No lipid: (-), 0.6 mM: (---), and 1.2 mM: (---) and 2.4 mM: (----). The difference: 0.6 mM: (--), 1.2mM: (----) and 2.4 mM: (----). The CD spectra of 4.8 and 9.6 mM were quite similar to that of 2.4 mM.
CD
study of progesterone-phosphatidyllipid
interaction
857
Table 2. Energy of the interactions between the acetyl group and the D-ring moiety of PROG estimated for the rotational conformations of the acetyl group of PROG, where 0 indicates the dihedral angle as viewed from C,, to C,
1O:o”
30’
60
90”
0.4 0.2 3.4
0.8
0.4
0.8 0.8
18-Me/20-Me 18-Me/C=0
2.1
0.7
0.1
total
3.7
interaction
C,,-C,,
torslon
C,,/20-Me
C,J20- Me
4.7
6.0
3.4
6.9
3.8
120’
0.8
0.8
150’
180”
210’
240’
270”
300”
330’
0.4
0.8
0.4 0.2
0.8
0.4 3.4
0.8 6.0
0.4 3.4 0.2
0.2 2.1
0.7 0.3
3.7
1.1
0.1 1.1
2.1
0.3 0.9
1.9
2.7
2.0
4.8
8.9
7.7
8 A4
Unit:
1
kcal/mol
0
\
17.20
13 K
in the lipid concentrations. The unsaturated arachidoyl group with four double bonds in the lipid affected the conformation of the acetyl group in a different way than the oleoyl group with one double bond. However, it did not affect the conformation of the cyclohexenone ring. Therefore, the number and/or position of the double bond of the hydrocarbon moiety of the lipids appears to govern the rotational and inversional conformations of PROG. DISCUSSION
Advances in both CD instrumentation and the interpretation of ellipticity have reached a stage at which circular dichroism has become very useful as a probe in the conformational analysis of molecules as well as in the study of interaction between proteins and steroids [16, 171. CD is also a convenient and efficient technique for preferred conformational analysis of the acetyl group [19-211 and a$-unsaturated cyclohexenone [22] of steroids. The asymmetric conformation around the chromophore of the acetyl group and cyclohexenone is changed by interaction, which theoretically can lead to differences in the Cotton effects of PROG, giving rise to negative and positive Cotton effects at 330 and 288 nm attributable to cyclohexenone and the acetyl group, respectively. The rotational strength of these Cotton effects shows an increase or decrease depending on conformational changes, when the concentration and/or structure of the lipid changes. Changes in the strength of the CD curve in the presence or absence of lipids make it possible to study the specific events taking place in the acetyl group and the @-unsaturated cyclohexenone of PROG. In conformational analysis of the interaction between PROG and lipids, it is necessary to consider three factors, inversions of the cyclohexenone ring, inversions of the B, C and D rings, and a rotation of the acetyl group. Throughout the following discussion of conformational analyses, however, the
/
16
conformation of alicyclic moiety, except for that of the A-ring of the compound is assumed to be un-
changeable because of its rigid structure. Accordingly, only the conformations of cyclohexenone and the acetyl group were studied by CD. With regard to conformational analysis of the acetyl group, we must consider the problem of the bond joining of a carbony1 group to a tetrahedral carbon in a molecule of the acetone type in the rotational barrier. As can be seen in Table 2, the energy of the acetyl group intramolecularly interacting with C,,-C& (torsion), C,,/20-Me, C,,/20-Me, 18-Me/20-Me, and 18-Me/ C=O around the C,,-C,, bond was estimated by using the energy function given for calculation of the energy of interactions for a -acetylcyclohexanois [19,23] and pregnan 20-one [34], where 0 corresponds to a dihedral angle [25] of the Newman projection [26]. Figure 6 shows each conformer of the acetyl group, each of which was successively rotated at 0 = 60” intervals. The plot of the total energy values versus the dihedral angle 0 from Table 2 is shown in Fig. 7. The signs of the Cotton effect shown were predicted according to the octant rule [27] by examination of the model [25] for each conformer. Conformers with dihedral angles of O-60” and 240-360” should exhibit a negative Cotton effect, while those angles 60-240” should display a positive one. Judging from the energy of interactions, the conformers from AC to Ae with dihedral angles of approx 120-124”, may be relatively preferred for the acetyl group of the compound (Fig. 6). Needless to say, other factors, such as temperature and solvents, also should affect the preference of rotational conformation of the acetyl group. The CD curves caused by the absorption of the acetyl group of PROG showed only a positive Cotton effect, suggesting that the predominant conformer was AC, regardless of solvent polarity or the kind of lipid, although the elliptical strength of the curves was evidently dependent on each lipid concentration for the unsaturated and glycerol lipids. The directionality of increases or decreases in the
858
SATORU WATANABE et al.
Ab:(KP
Ac:lPO’
x5-- P: 0
Atw80’
0
A.: 240’
At 300'
Fig. 6. Possible rotational conformations of the 17-acetyl group of PROG which were successively rotated at 0 = 60 intervals of the dihedral angle in Table 2.
band was distinctly regulated by only lipids with unsaturated hydrocarbon moiety or by the glycerol group. Decreases in the positive CD band with increases in lipid concentration were observed in the case of dioleoyl-L-a-phosphatidylcholine in nonpolar and polar solvents, and in dipalmitoyl-L-a-phosphatidylDL-glycerol in chloroform, while increases in the positive CD band were observed with increases in the concentration of diarachidoyl-L-a-phosphatidylcholine. The dioleoyl and glycerol groups in chloroform influenced ‘the increase in population of the energetically stable minor conformer Ae (240”) which predicted non-Cotton effects (Fig. 7), but not the energetically unstable conformers at 0 = O-60”. The interactions between the acetyl group and the two double bonds of the oleoylphosphatidylcholine, and two hydroxyl groups of glycerol moiety mainly dominate the conformation of the a&y1 group. However, the dioleoyl group in methanol contributes little to conformational change in the acetyl group because of the destructive effect on the polar-polar interaction of the group and methanol which decreases the amount of accessible lipid, as can be seen in the subtracted curves in Fig. 2. Therefore, the dioleoyl and glycerol moieties of the lipids were forced to shift from the stable conformer AC to conformers that may exist around 0 = 240” and this contributed to a decrease in the positive elliptical strength. In contrast to the CD curve in the presence of dioleoyl and glycerol lipids, CD curves of PROG in the presence of the diarachidoyl group continuously showed increases with increases in the lipid concentrations (Fig. 5). That is to say, the energetically stable conformer AC, with a positive Cotton effect, became predominant as a result of interaction between the acetyl and arachidoyl groups. The only difference between the arachidoyl and oleoyl groups, other than their chain lengths, is the number of double bonds in
Fig. 7. Energy curve of interactions calculated for the rotational conformations of the acetyl group of PROG as a function of 0 in Table 2. The predicted signs of the Cotton effect presented by the Newman projection are shown on the outer circle, and the inner and middle circles represent 5.0 and 10.0 Kcal/mol lines, respectively.
their hydrocarbon moiety. Therefore, the difference in effects between the arachidoyl and oleoyl groups on the CD curve of the acetyl group can be attributed to the double bond in the hydrocarbon moiety of the lipid. The arachidoyl group, with four double bonds versus the oleoyl group with one double bond, provides eight possible interactive positions in the lipid. Thus, a lipid of the arachidoyl group has 4 times more freedom in the interactive position than one of the oleoyl group. The number of double bonds which can interact with the acetyl group decides the directionality of the CD curve because the amount of freedom of interaction between the double bonds and the a&y1 group regulates conformational disorder. That is to say, multi-interictions at various positions of the eight double bonds in the arachidoyl lipid may force a shift in the conformational equilibrium from a mixture of the main preferred conformer AC with a positive sign and a small number of unstable 0 = O-30” conformers contributing a negative sign to the positive stable rotatary conformer AC at the sacrifice of minor conformers. On the other hand, three conformers and predictable signs of the Cotton effect of the a$-unsaturated ketone moiety of PROG are shown in Fig. 8. The conformer Ba predicts a negative sign while the conformers Bb and Bc predict a positive one according to the a$-unsaturated octant rule [22] by examination of the Dreiding model [25]. Therefore, the negative Cotton effect of PROG observed at the wavelength of 330 nm should be assignable to absorption of the conformer Ba. Neither the a,/?-unsaturated A-ring absorption or the acetyl group of PROG were affected by saturated phosphatidylcholine with stearoyl and dipalmitoyl groups (Table 1), or unsaturated phosphatidylcholine with the arachidoyl group (Fig. 5). This absorption, however, decreased with an increase in the concen-
CD study of progesterone-phosphatidyllipid interaction
Ba
Bb
859
SC
Fig. 8. Possible preferred conformations of the x,/J-unsaturated cyclohexenone ring of PROG and predicted signs of the Cotton effect as viewed from the oxygen of the carboxyl group to carbon C,.
tration of dioleoyl phosphatidylcholine in both methanol (Fig. 2) and chloroform (Fig. 3), and increased in the presence of dipalmitoyl lipid, which was substituted a glycerol for the choline group in chloroform (Fig. 4). As for oleoyl lipid, the decrease in elliptical strength based on the structural difference in the lipids which were substituted for oleoyl in the palmitoyl or stearoyl groups was responsible for the double bond which interacted with the cyclohexenone ring. The relatively stable conformer Ba, with five carbons of C2, C, , C4, C, and Cl,, on the same plane, probably shifts to an equilibrium containing Bb and/or Bc as a result of the interaction of pi electrons localized on the double bonds in the cyclohexenone of the A-ring with those on one double bond contained in the oleoyl group. Electron interaction always takes place in the same position on oleoyl lipid because there is only one double bond in this group. The decrease in the rotational strength of PROG in the presence of oleoyl lipid might be due to the increase in the population of the unstable conformer Bb, but not to the conformer Bc. The Bc form may be slightly more unstable than the half chair conformer Bb which neglects the interaction of 19-Me and C=O as can be predicted in the case of Bc form with carbons of C,,C2, C4 and C5 on the same plane. In addition, the half chair conformer Bb is about 1.6 Kcal/mol[28] more energetically stable than the boat conformer Bc. In contrast with dioleoylphosphatidylcholine, however, the CD ellipticity of PROG at 330nm in the presence of glycerol lipid increased slightly with increases in its concentration. This increase might be explained by the stabilization force resulting from the intermolecular hydrogen bond between the carbonyl group of the A-ring and the hydroxyl group in the glycerol moiety of the lipid. This is because the energy resulting from the intermolecular hydrogen bond with the hydroxyl group of the lipid and the carbonyl group of the A-ring may prevent inversion of the ring in the conformer Ba. The results obtained from this investigation present a quite different picture of the effect of lipid on the conformations of both the rotational acetyl group and the inversional cyclohexenone of PROG inter-
acting with lipids. Specifically, the conformational modification power of lipids including their different positions and/or numbers of double bonds, is not only influenced by hydrocarbon moiety, but also by another moiety. These factors also characteristically regulate the distribution of the preferred conformations of PROG toward stable or unstable positions. The biomembrane is composed of many kinds of lipids with great variations with respect to structure, components and quantity. Some of the moieties of the lipids in biomembrane are forced into a particular conformation of the acetyl group and the cyclohexenone ring of PROG. The particularized conformations which may be predominant in a population as well as the basal lipids which make up the membrane architecture may affect specific interaction with other components, such as various enzymes with special conformation. The several membrane enzymes which interact with PROG might cause conformational change accompanying a modification in the membrane function. As to the significance of conformation for membrane enzyme action, it has been suggested that the membrane enzymes or receptors are unable to function properly unless they are surrounded by an appropriate phospholipid for the requirement of pertinent conformation [29-311. Therefore, the diversity of the biomembrane seems to be closely related to regulation of subtle conformational alterations in the various moieties of the steroid throughout interaction. This may cause functional modifications and may provide a reasonable explanation for the nongenomic effect of the steroid. Acknowledgements-We thank Misses N. Otsuki and J. Katayama for their help in the preparation of the manuscript and also Miss Otsuki for her excellent technical assistance. This study was supported in part by Research Project Grants Nos 61402 and 62-503 from Kawasaki Medical School. REFERENCES
1. Baulieu E. E., Godeau F., Schorderet M. and Schorderet-Slakine S.: Steroid-induced meiotic division
860
SATORUWATANABE er al.
in Xenopus laevis oocytes: surface and calcium. Nature 275 (1982) 593-598. 2. Smith L. D. and Ecker R. E.: Interaction of steroids with Rana pipiens oocytes in the induction of maturation. Dev. Biol. 25 (1971) 233-241. 3. Godeau J. F., Schorderet-Slatkine S., Hubert P. and Baulieu E. E.: Induction of maturation in Xenopus laeuis oocytes by a steroid linked to a polymer. Proc. nam. Acad. Sci. U.S.A. 75 (1978) 2353-2357.
4. Kaya H. and Saito T.: Effect of progesterone and its l’la-hydroxyl derivative on human erythrocyte membrane. Japan J. Pharmac. 39 (1985) 299-306. 5. Heap R. B., Symons A. M. and Walkins J. C.: Steroids and their interactions with phospholipids: solubility, distribution coefficient and effect on potassium permeability of liposomes. Biochim. biophys. Acta 218 (1970) 482-495.
6. Seeman P.: The actions of nervous system: drugs on cell membranes. In Cell Membranes, Biochemistry, Cell Biology and Pathology (Edited by G. Weissmann and R. Clairbone). HP Publishing, New York (1975) 239-247. I. McEwen B. S.: Gonadal steroids. Humoral modulators of nerve-cell function. Molec. Cell Endocr. 18 (1980) 151-164. 8. Stumpf W. E. and Sar M.: Steroid hormone target cells in the extrahypothalamic brain stem and cervical spinal cord: neuroendocrine significance. J. steroid Biochem. 11 (1979) 801-807. 9. Alivisatos S. G. A., Deliconstantions G. and Theodosiadis G. P.: Specificity of binding of cholesterol, steroid hormones and other compounds in synaptosomal plasma membranes, and their effect on ouabainsensitive ATPase. Biochim. biophys. Acfa 643 (1981) 65&658. 10. Wasserman W. J., Pinto L. H., O’Connor C. M. and Dennis Smith L.: Progesterone induces a rapid increase in [calcium ion] of Xenopus laeuis oocytes. Proc. natn. Acad. Sci. U.S.A. 77 (1980) 15341536. II. Moreau M., Vilain J. P. and Guerrier P.: Free calcium changes associated with hormone action in amphibian oocytes. D~LJ. Biol. 78 (1980) 201-214. 12. Livingston J. N. and Lockwood D. H.: Effect of glucocorticoids on the glucose transport system of isolated fat cells. J. biol. Chem. 250 (1975) 8353-8360. 13. Bangham A. D., Standish M. M. and Weissmann G.: The action of steroids and streptolysin S on the permeability of phospholipid structures to cations. J. molec. Biol. 13 (1965) 253-259. 14. Willmer E. N.: Steroids and cell surfaces. Biol. Reo. 36 (1961) 368-398. 15. Tomkins G. and Maxwell E. S.: Some aspects of steroid hormone action. A. Rev. Biochem. 12 (1963) 677-708. 16. Watanabe S.: Circular dichroism study of the interaction between conformationally altered human serum albumin and testosterone. J. steroid Biochem. 23 (1985) 177-183. 17. Ronald T. and Gene R. L.: Circular dichroism of
bovine serum albumin-testosterone system. J. steroid Biochem. 7 (1975) 1523-1525. 18 Beecham A. F. and Collins D. J.: The circular dichroism of a$-unsaturated ketones. 1. Circular dichroism in the singlet-singlet and singlet-triplet n +R l transitions of steroidal 4-en-3-ones. Aust. J. Chem. 33 (1980) 2187-2197. 19 Suga T. and Watanabe S.: Stereochemical studies of monoterpene compounds. XIII. The conformational analysis of the acetyl group of 4-methyl-lO-nor-8 oxomenthols by circular dichroism. Bull. them. Sot. Jap. 45 (1972) 24&245.
20 Snatzke G. and Schwinum E.: Circular dichroism. XVII. Low-temperature determinations on 12-substituted 20-oxopregnanes and -16-pregenes. Tetrahedron 22 (1966) 761-771.
21 Wellman K. M. and Djerassi C.: Optical rotatory dispersion studies. XCVIII. The conformation of the 1%acetyl group in steroids and its correlation with variable-temperature circular dichroism. J. Am. them. Sot. 87 (1965) 60-66. 22 Snatzke G.: Circulardichroismus-IX. Modifizierung der octantenregel fur a,/?-ungestlttigte ketone: transoide enone. Tetrahedron 21 (1965) 421-438. 23 Suga T., Shishibori T. and Matsuura T.: Stereochemical studies of monoterpene compounds. VI. Stereochemistry of 2-hydroxypinocamphone and its reduction products. Bull. them. Sot. Jap. 41 (1978) 1175-1179. 24 Allinger N. L., CarabblC P. P. and Perez G.: Conformational analysis. XLIX. The conformations of the acetyl side chains of some compounds related to pregnan-20-one. Tetrahedron 22 (1966) 1615-1623. 25 Examined by “Dreiding Stereomodels”, W. Buchi Manufacture of Glass Apparatus Flawil, Switzerland. 26. Eliel E. L., Allinger N. L., Angyal S. J. and Morrison G. A.: Conformational Analysis. Interscience Publishers, New York (1965) p. 52. 21. Moffit W., Woodward R. B., Moscowitz A., Klyne W. and Djerassi C.: Structure and optimal rotatory dispersion of saturated ketones. J. Am. them. Sot. 83 (1961) 40134018. 28. Hendrickson J. B.: Organic and biological chemistry molecular geometry. I. Machine computation of the common rings. J. Am. them. Sot. 83 (1961) 452334537. 29. Roleofsen B.: The (non)specifity in the lipidrequirement of calcium- and (sodium plus potassium)-transporting adenosine triphosphatases. Lr@ Sci. 29 (1981) 2235-2247. 30. Fleming J. W. and Ross E. M.: Reconstitution of beta-adrenergic receptors into phospholipid vesicles: restoration of 1251 iodohydroxybenzylpindolol binding to digitomin-solubilized receptors. J. Cyclic Nucl. Res. 6 (1980) 407419. 31 Loh H. H. and Law P. Y.: The role of membrane lipids in receptor mechanisms. A. Rev. Pharmac. Tox. 20 (1980) 201-234.
1988
0022-4731/88
$3.00 + 0.00
Copyright0 1988PergamonPressplc
CONFORMATIONAL STUDY OF THE ACETYL GROUP AND CYCLOHEXENONE IN PROGESTERONE INTERACTING WITH PHOSPHATIDYLLIPID BY MEANS OF CIRCULAR DICHROISM SATORUWATANABE*,TAIICHI SAITO*and MITSUNORIIKEDAt *Departments of Pharmacology and tBiochemistry, Kawasaki Medical School, Kurashiki, Okayama 701&O],Japan (Received 28 July 1987; received for publication 22 March 1988)
Summary-The interaction between the A-ring and the 17-acetyl groups of progesterone (PROG) and various concentrations of distearoyl-, dipalmitoyl-, dioleoyl- and diarachidoyl-L-a-phosphatidylcholines, and dipalmitoyl+a-phosphatidyl-DL glycerol in methanol and chloroform solutions and its preferred conformational assignments in the presence of those lipids were examined qualitatively by circular dichroism on the basis of PROG spectra in the wavelength regions of 26&4OOnm. PROG did not interact with saturated distearoyl and dipalmitoyl phosphatidylcholines, but did with unsaturated dioleoyl and diarachidoyl phosphatidylcholines, and saturated dipalmitoylphosphatidylglycerol. The interacting moieties of PROG were an a$-unsaturated cyclohexenone of the A-ring for oleoyl and glycerol lipids, and the 17-acetyl group for unsaturated and glycerol lipids. The interaction with these lipids, the rotational conformations of the 17-acetyl group, and invertible conformations of the cyclohexenone of PROG were discussed on the basis of the elliptical strength of the Cotton effect and energy estimation of the preferred conformers. Oleoylphosphatidylcholine caused an increase in slightly energetically unstable conformers of the acetyl group and stable conformers of the a$-unsaturated cyclohexenone. Glycerol lipid, on the other hand, caused an increase in energetically unstable conformers of cyclohexenone, but it was similar to the effect of oleoyl lipid on the 17-acetyl group. Diarachidoylphosphatidylcholine, with eight double bonds, on the other hand, increased the number of energetically stable conformers of the 17-a&y1 group, but had no effect on the conformation of cyclohexenone. It became apparent that the double bond of hydrocarbon moiety as well as the head group of choline and glycerol in lipids were closely related to the conformational populations of both groups of the PROG molecule. The specific effect on the conformations of the acetyl and a,/?-unsaturated cyclohexenone of PROG of various lipids with different substitutions in their heads or hydrocarbon moieties might in part explain the nongenomic action of the steroid.
INTRODUCTION The receptor mediated pathway of steroid hormone action is not sufficient to fully account for the known effects of steroids. Some investigators [l-3], for example, have demonstrated that progesterone and other steroid molecules can promote the maturation of Xenopus fuevis oocytes, although these cells do not contain steroid receptors. Kaya and Saito [4] also recently observed that progesterone strongly inhibited hemolysis of human erythrocytes against osmotic shock within several minutes after exposure to the agent. These findings suggest that, in addition to its well known genomic action, progesterone operates through other mechanisms; particularly through direct action on the plasma membrane. It has been said that such a direct effect of steroids must be related to a change in the membrane function, which can be modified by a partition coefficient difference between lipid and water [5,6], cell excitability [7, 81, transport [9-121, the release effect on acid phos-
phatase from lysosomes and increases in cation leakage from liposomes [13], membrane expansion [ 141, and/or an enzymatic effect [15]. In addition to these factors of cell function, the direct interaction between steroids and lipid molecules which can cause alteration in cell function is significant, because the plasma membrane consists mainly of various phospholipids that stabilize its architecture. Consideration of the conformations of the moieties of the A-ring and 17-acetyl group during the interaction between progesterone and lipids should not only be useful in revealing the direct action mechanism, but also in developing new steroids that may be, of biological significance. Although circular dichroism appears to be more suitable for the monitoring the conformational alteration of optical active compounds because of its peculiar sensitivity to subtle conformational alterations, it has not been used for investigation of steroid-lipid interaction prior to our study except for some measurements of systems containing steroids and proteins [ 16, 171.
853
SATORUWATANABE et
854
al.
Table I. CD results of interaction between 100 CM PROG and various lipids at different concentrations
diarachidoyl-L- u phosphatidylcholine
0000x
All solutions contained 100 PM of progesterone.
006OQOO 0: Unaltered CD. 0: Altered CD. x : Insolubility.
EXPERIMENTAL
Reagents
All the chemicals used were analytical grade reagents employed without further purification. Distearoyl-, dipahnitoyl-, dioleoyl- and diarachidoyl-La-phosphatidylcholines and dipalmitoyl-L-a-phosphatidyl-DL-glycerol as well as progesterone and testosterone were purchased from Sigma, St Louis, U.S.A. Circular dichroism measurements
CD measurements were made at 25°C with a Union Giken Mark III spectropolarimeter calibrated with (+)-D-camphorsulfonic acid and equipped with a computerized data processing system. All spectra were recorded in a quartz cell with lO.Omm path length at a scanning range of 26tX-400 nm using a full scale deflection of 0.02 and a spectral band width of 2.0 nm. The spectra in each figure have been recorded at the same day on the same chart using a freshly prepared sample to minimize errors. Results were expressed as molar ellipticities, [e] (deg. cm2 dmol-‘), calculated with reference to the PROG concentrations. The PROG concentration was constantly 100 PM throughout the experiment.
both solvents, and to those of dipalmitoyl-L-a -phosphatidyl-DL-glycerol and diarachidoyl-L-a-phosphatidyl-choline in methanol regardless of the concentrations of lipids, with the exception of one case of insolubility. Consequently, CD spectra of PROG’s interaction with dioleoyl- and diarachidoyl-L-a -phosphatidylcholines, and dipalmitoyl-L-a -phosphatidylDL-glycerol in chloroform and dioleoyl-L-a-phosphatidylcholine in methanol is significant. The CD curve of PROG in methanol showed a negative Cotton effect at 320 and a positive one at 288 nm in agreement with a previously noted curve [ 181,as shown in Fig. 1. PROG has two optical
16
RESULTS The interaction between PROG and various concentrations of distearoyl, dipalmitoyl-, dioleoyl- and diarachidoyl-L-a-phosphatidylcholines, and dipalmitoyl-L-a-phosphatidyl-DL-glycerol in methanol and chloroform solutions and its preferred conformational assignments were examined qualitatively with a CD spectropolarimeter on the basis of PROG spectra in the wavelength regions of 260-400 nm at room temperature. Measurement results are summarized in part in Table 1. The CD spectrum of PROG was quite similar to curves in the presence or absence of distearoyl- and dipalmitoyl-L-a-phosphatidylcholines with saturated aliphatic moiety in
260
280
3al
320
340
360
uy)
Fig. 1. CD spectra for 100 pM of PROG and testonslkrone. The CD spectrum of testosterone was subtracted from that of PROG in methanol in the 26@4OOnm region. PROG: (---), testosterone: (--) and the difference: (----).
CD study of progesterone-phosphatidylhpid
280
300
320
340
absorption moieties of an a$-unsaturated ketone with the A-ring and of the 17-acetyl group. Judging from the absorption wavelength and the n - II* transition, it seemed very reasonable to assign 288 and 320nm to the saturated ketone absorption and unsaturated ketone absorption, respectively. In addition, when discussing changes of the rotational strength in association with conformational changes in respective ketones, it is
important to clarify whether negative and positive CD curves overlap or cancel each other out with opposite sign ellipticity. The degree of overlap can be best seen in the difference spectrum presented in Fig. 1. Here, the testosterone spectrum was subtracted from the PROG spectrum obtained at the same concentration in the same solvent, the only structural difference between PROG and testosterone being a substitution of the 17-acetyl group for the hydroxyl group, which does not contribute to elliptical absorption at all. The resulting CD peak at 288 nm can almost certainly be attributed to net absorption of the acetyl group, and can be consequently justified by the above-mentioned assignment of the PROG peaks. The negative elliptical strength of the subtracted CD curve at 288 nm was partially cancelled by the negative absorption of the A-ring but it was practically nil at 330 nm. This suggests that the unsaturated ketone on the A-ring absorption of PROG had no influence, in spite of a slightly positive Cotton effect due to the 5.0 nm shift. Thus the CD curve of PROG was composed of negative and positive extrema at 288 and 320 nm. These extrema were assigned to the independent absorptions of the 17-saturated acetyl and a$-unsaturated ketone groups, respectively. Accordingly, the rotational strengths of the 288 and 320 nm bands responsible for conformational change may be used to study the interaction between the PROG molecule and various lipid molecules and its preferred conformation assignments in their presence. The CD spectra of 100pM PROG in the presence of various concentrations of dioleoylphosphatidylcholine were composed of negative and positive bands with maxima at 330 in methanol and
260
300
320
855
340
3
360 nm Fig. 2. CD spectra of PROG in the presence of various concentrations of dioleoyl-L-a-phosphatidylcholine (left) and in the absence of 1.2 and 9.6 mM lipid concentrations in methanol in the 260-360 nm region subtracted from spectra in their presence (right). Left: no lipid: (-), 1.2 mM: (---) and 9.6mM: (----). right: 1.2 mM: (---) and 9.6 mM: (----). 260
280
interaction
288 nm in chloroform, as shown in Figs 2 and 3, respectively. When the concentration of the lipid was increased gradually in the order of 0.1, 0.3, 0.6, 1.2, 2.44.8 and 9.6 mM against a constant concentration of PROG, a decrease in both positive and negative elliptical strengths dependent on the concentration difference began from 4.8 in methanol and 1.2 mM in chloroform. Therefore, the difference in the effect of the lipid on the starting point of the conformational change of PROG may be dependent on a solvent polarity. Such a solvent polarity could influence the interactions of lipid-PROG and intramolecular PROG, since the polar solvent accessible to the oxygen of the polar functional group in the ketones probably prevents lipid attacks on this moiety at low
I 16 .
t
260
280
MO
320
YO
360
3aa nm
Fig. 3. CD spectra of PROG in the presence of various concentrations of dioleoyl+a-phosphatidylcholine and in the absence of 2.4, 4.8 and 9.6 mM lipid concentrations in chloroform in the 26WOOnm region subtracted from spectra in their presence. No lipid: (-), 2.4mM: (---), 4.8 mM: (---) and 9.6mM: (---). The difference: 2.4mM: (---). 4.8 mM: (----) and 9.6 mM: (----).
SATQRU WATANABE et al.
856
concentrations. Modifications in rotational strength due to absorption of the A-ring and the acetyl group of PROG by lipids were first observed by this author in the current work, and provide evidence of molecular interaction between PROG and lipids. The spectra seen on’ the left in Fig. 2 cannot be fully ascribed to such modification because of noise when PROG interacted with the lipid in methanol. The change produced can be better seen on the right in the same figure. Here, the spectrum in the presence of 100 PM of PROG was subtracted from the spectra in the presence of various concentrations of oleoyl lipid. The resulting curves showed a positive Cotton effect around 330 nm and a negative one at 288 nm in a concentration of 9.6 mM of the lipid. However, the subtracted spectrum at 1.2 mM exhibited a curve with no peaks. The curve of PROG in the presence of 1.2 mM of lipid therefore appears to actually be the same curve as that in the absence of the lipid, although a slight shift must be allowed for the line. In contrast to effect of the saturated lipid on the CD curve of PROG, the decrease in the rotational strength caused by the oleoyl lipid, which is substituted for the unsaturated hydrocarbon group in the hydrocarbon moiety of saturated distearoyl- and dipalmitoyl-phosphatidylcholines, evidently indicates that the double bond for rotational modifications in the conformation of the acetyl group and for inversional modifications in the a$-unsaturated cyclohexenone in PROG plays important role. Figure 4 shows the CD spectra of PROG in the presence of various concentrations of dipalmitoylphosphatidylglycerol measured at the 26&400 nm region in chlo16
roform. The rotational intensity of the CD curves at 288 nm decreased with increases in the lipid concentration, while that at 330nm on the negative band increased slightly. The concentration needed for conformational modification was lower for the acetyl group than for cyclohexenone. The structural difference between dipalmitoylphosphatidylglycerol and dipalmitoylphosphatidylcholine is the moiety of glycerol. This modification of the curve by the lipids thus seems to reveal a specific contribution of the glycerol group to the absorption due to conformational change in the acetyl group and the cyclohexenone ring in PROG. Choline moiety, on the other hand, had no influence. Since the CD curves of the acetyl group conformationally modified by glycerol lipid were very similar to those of oleoyl lipid, and since there are marked structural and positional differences between oleoyl and glycerol lipids, it is easy to conclude that there is a different mechanism at work. The marked structural and positional differences also may provide a reasonable explanation for the different effect to the oleoyl and glycerol lipids on the CD curves of the a&?-unsaturated cyclohexenone ring. The CD spectra attributable to the acetyl and cyclohexenone groups of PROG in the presence of various concentrations of diarachidoylphosphatidylcholine differed significantly from those coexisting with the oleoyl or glycerol lipids tested in this experiment, as shown in Fig. 5. That is to say, the rotational strength of the 288 nm band increased with increases in the arachidoyl lipid concentration, but there were no changes at 330 nm peak regardless of the increase 16,
12
8
-4
t 260
280
3m
320
340
360
380 “rll
Fig. 4. CD spectra of PROG in the presence of various concentrations of dipalmitoyl-r.-a-phosphatidyl-DL-glycerol and CD spectra in the absence of 2.4,4.8 and 9.6 mM lipid concentrations in chloroform in the 260-4OOnm region subtracted from spectra in their presence. The concentration conditions for each curve were-the same as in Fig. 3. The CD soectrum of PRGG in 9.6 mM of lipid was almost the sameas that in 4.8 mM, although it was &htly weaker and stronaer at oositive and neaative absorotions. resnectivelv.
260
280
300
320
340
360
380 nm
1
Fig. 5. CD spectra of PRGG in the presence of various concentrations of arachidoyl-L-a-phosphatidylcholine and CD soectra in the absence. of 0.6, 1.2 and 2.4mM lipid concentrations in chloroform in the 260-4OOnm region subtracted from spectra in their presence. No lipid: (-), 0.6 mM: (---), and 1.2 mM: (---) and 2.4 mM: (----). The difference: 0.6 mM: (--), 1.2mM: (----) and 2.4 mM: (----). The CD spectra of 4.8 and 9.6 mM were quite similar to that of 2.4 mM.
CD
study of progesterone-phosphatidyllipid
interaction
857
Table 2. Energy of the interactions between the acetyl group and the D-ring moiety of PROG estimated for the rotational conformations of the acetyl group of PROG, where 0 indicates the dihedral angle as viewed from C,, to C,
1O:o”
30’
60
90”
0.4 0.2 3.4
0.8
0.4
0.8 0.8
18-Me/20-Me 18-Me/C=0
2.1
0.7
0.1
total
3.7
interaction
C,,-C,,
torslon
C,,/20-Me
C,J20- Me
4.7
6.0
3.4
6.9
3.8
120’
0.8
0.8
150’
180”
210’
240’
270”
300”
330’
0.4
0.8
0.4 0.2
0.8
0.4 3.4
0.8 6.0
0.4 3.4 0.2
0.2 2.1
0.7 0.3
3.7
1.1
0.1 1.1
2.1
0.3 0.9
1.9
2.7
2.0
4.8
8.9
7.7
8 A4
Unit:
1
kcal/mol
0
\
17.20
13 K
in the lipid concentrations. The unsaturated arachidoyl group with four double bonds in the lipid affected the conformation of the acetyl group in a different way than the oleoyl group with one double bond. However, it did not affect the conformation of the cyclohexenone ring. Therefore, the number and/or position of the double bond of the hydrocarbon moiety of the lipids appears to govern the rotational and inversional conformations of PROG. DISCUSSION
Advances in both CD instrumentation and the interpretation of ellipticity have reached a stage at which circular dichroism has become very useful as a probe in the conformational analysis of molecules as well as in the study of interaction between proteins and steroids [16, 171. CD is also a convenient and efficient technique for preferred conformational analysis of the acetyl group [19-211 and a$-unsaturated cyclohexenone [22] of steroids. The asymmetric conformation around the chromophore of the acetyl group and cyclohexenone is changed by interaction, which theoretically can lead to differences in the Cotton effects of PROG, giving rise to negative and positive Cotton effects at 330 and 288 nm attributable to cyclohexenone and the acetyl group, respectively. The rotational strength of these Cotton effects shows an increase or decrease depending on conformational changes, when the concentration and/or structure of the lipid changes. Changes in the strength of the CD curve in the presence or absence of lipids make it possible to study the specific events taking place in the acetyl group and the @-unsaturated cyclohexenone of PROG. In conformational analysis of the interaction between PROG and lipids, it is necessary to consider three factors, inversions of the cyclohexenone ring, inversions of the B, C and D rings, and a rotation of the acetyl group. Throughout the following discussion of conformational analyses, however, the
/
16
conformation of alicyclic moiety, except for that of the A-ring of the compound is assumed to be un-
changeable because of its rigid structure. Accordingly, only the conformations of cyclohexenone and the acetyl group were studied by CD. With regard to conformational analysis of the acetyl group, we must consider the problem of the bond joining of a carbony1 group to a tetrahedral carbon in a molecule of the acetone type in the rotational barrier. As can be seen in Table 2, the energy of the acetyl group intramolecularly interacting with C,,-C& (torsion), C,,/20-Me, C,,/20-Me, 18-Me/20-Me, and 18-Me/ C=O around the C,,-C,, bond was estimated by using the energy function given for calculation of the energy of interactions for a -acetylcyclohexanois [19,23] and pregnan 20-one [34], where 0 corresponds to a dihedral angle [25] of the Newman projection [26]. Figure 6 shows each conformer of the acetyl group, each of which was successively rotated at 0 = 60” intervals. The plot of the total energy values versus the dihedral angle 0 from Table 2 is shown in Fig. 7. The signs of the Cotton effect shown were predicted according to the octant rule [27] by examination of the model [25] for each conformer. Conformers with dihedral angles of O-60” and 240-360” should exhibit a negative Cotton effect, while those angles 60-240” should display a positive one. Judging from the energy of interactions, the conformers from AC to Ae with dihedral angles of approx 120-124”, may be relatively preferred for the acetyl group of the compound (Fig. 6). Needless to say, other factors, such as temperature and solvents, also should affect the preference of rotational conformation of the acetyl group. The CD curves caused by the absorption of the acetyl group of PROG showed only a positive Cotton effect, suggesting that the predominant conformer was AC, regardless of solvent polarity or the kind of lipid, although the elliptical strength of the curves was evidently dependent on each lipid concentration for the unsaturated and glycerol lipids. The directionality of increases or decreases in the
858
SATORU WATANABE et al.
Ab:(KP
Ac:lPO’
x5-- P: 0
Atw80’
0
A.: 240’
At 300'
Fig. 6. Possible rotational conformations of the 17-acetyl group of PROG which were successively rotated at 0 = 60 intervals of the dihedral angle in Table 2.
band was distinctly regulated by only lipids with unsaturated hydrocarbon moiety or by the glycerol group. Decreases in the positive CD band with increases in lipid concentration were observed in the case of dioleoyl-L-a-phosphatidylcholine in nonpolar and polar solvents, and in dipalmitoyl-L-a-phosphatidylDL-glycerol in chloroform, while increases in the positive CD band were observed with increases in the concentration of diarachidoyl-L-a-phosphatidylcholine. The dioleoyl and glycerol groups in chloroform influenced ‘the increase in population of the energetically stable minor conformer Ae (240”) which predicted non-Cotton effects (Fig. 7), but not the energetically unstable conformers at 0 = O-60”. The interactions between the acetyl group and the two double bonds of the oleoylphosphatidylcholine, and two hydroxyl groups of glycerol moiety mainly dominate the conformation of the a&y1 group. However, the dioleoyl group in methanol contributes little to conformational change in the acetyl group because of the destructive effect on the polar-polar interaction of the group and methanol which decreases the amount of accessible lipid, as can be seen in the subtracted curves in Fig. 2. Therefore, the dioleoyl and glycerol moieties of the lipids were forced to shift from the stable conformer AC to conformers that may exist around 0 = 240” and this contributed to a decrease in the positive elliptical strength. In contrast to the CD curve in the presence of dioleoyl and glycerol lipids, CD curves of PROG in the presence of the diarachidoyl group continuously showed increases with increases in the lipid concentrations (Fig. 5). That is to say, the energetically stable conformer AC, with a positive Cotton effect, became predominant as a result of interaction between the acetyl and arachidoyl groups. The only difference between the arachidoyl and oleoyl groups, other than their chain lengths, is the number of double bonds in
Fig. 7. Energy curve of interactions calculated for the rotational conformations of the acetyl group of PROG as a function of 0 in Table 2. The predicted signs of the Cotton effect presented by the Newman projection are shown on the outer circle, and the inner and middle circles represent 5.0 and 10.0 Kcal/mol lines, respectively.
their hydrocarbon moiety. Therefore, the difference in effects between the arachidoyl and oleoyl groups on the CD curve of the acetyl group can be attributed to the double bond in the hydrocarbon moiety of the lipid. The arachidoyl group, with four double bonds versus the oleoyl group with one double bond, provides eight possible interactive positions in the lipid. Thus, a lipid of the arachidoyl group has 4 times more freedom in the interactive position than one of the oleoyl group. The number of double bonds which can interact with the acetyl group decides the directionality of the CD curve because the amount of freedom of interaction between the double bonds and the a&y1 group regulates conformational disorder. That is to say, multi-interictions at various positions of the eight double bonds in the arachidoyl lipid may force a shift in the conformational equilibrium from a mixture of the main preferred conformer AC with a positive sign and a small number of unstable 0 = O-30” conformers contributing a negative sign to the positive stable rotatary conformer AC at the sacrifice of minor conformers. On the other hand, three conformers and predictable signs of the Cotton effect of the a$-unsaturated ketone moiety of PROG are shown in Fig. 8. The conformer Ba predicts a negative sign while the conformers Bb and Bc predict a positive one according to the a$-unsaturated octant rule [22] by examination of the Dreiding model [25]. Therefore, the negative Cotton effect of PROG observed at the wavelength of 330 nm should be assignable to absorption of the conformer Ba. Neither the a,/?-unsaturated A-ring absorption or the acetyl group of PROG were affected by saturated phosphatidylcholine with stearoyl and dipalmitoyl groups (Table 1), or unsaturated phosphatidylcholine with the arachidoyl group (Fig. 5). This absorption, however, decreased with an increase in the concen-
CD study of progesterone-phosphatidyllipid interaction
Ba
Bb
859
SC
Fig. 8. Possible preferred conformations of the x,/J-unsaturated cyclohexenone ring of PROG and predicted signs of the Cotton effect as viewed from the oxygen of the carboxyl group to carbon C,.
tration of dioleoyl phosphatidylcholine in both methanol (Fig. 2) and chloroform (Fig. 3), and increased in the presence of dipalmitoyl lipid, which was substituted a glycerol for the choline group in chloroform (Fig. 4). As for oleoyl lipid, the decrease in elliptical strength based on the structural difference in the lipids which were substituted for oleoyl in the palmitoyl or stearoyl groups was responsible for the double bond which interacted with the cyclohexenone ring. The relatively stable conformer Ba, with five carbons of C2, C, , C4, C, and Cl,, on the same plane, probably shifts to an equilibrium containing Bb and/or Bc as a result of the interaction of pi electrons localized on the double bonds in the cyclohexenone of the A-ring with those on one double bond contained in the oleoyl group. Electron interaction always takes place in the same position on oleoyl lipid because there is only one double bond in this group. The decrease in the rotational strength of PROG in the presence of oleoyl lipid might be due to the increase in the population of the unstable conformer Bb, but not to the conformer Bc. The Bc form may be slightly more unstable than the half chair conformer Bb which neglects the interaction of 19-Me and C=O as can be predicted in the case of Bc form with carbons of C,,C2, C4 and C5 on the same plane. In addition, the half chair conformer Bb is about 1.6 Kcal/mol[28] more energetically stable than the boat conformer Bc. In contrast with dioleoylphosphatidylcholine, however, the CD ellipticity of PROG at 330nm in the presence of glycerol lipid increased slightly with increases in its concentration. This increase might be explained by the stabilization force resulting from the intermolecular hydrogen bond between the carbonyl group of the A-ring and the hydroxyl group in the glycerol moiety of the lipid. This is because the energy resulting from the intermolecular hydrogen bond with the hydroxyl group of the lipid and the carbonyl group of the A-ring may prevent inversion of the ring in the conformer Ba. The results obtained from this investigation present a quite different picture of the effect of lipid on the conformations of both the rotational acetyl group and the inversional cyclohexenone of PROG inter-
acting with lipids. Specifically, the conformational modification power of lipids including their different positions and/or numbers of double bonds, is not only influenced by hydrocarbon moiety, but also by another moiety. These factors also characteristically regulate the distribution of the preferred conformations of PROG toward stable or unstable positions. The biomembrane is composed of many kinds of lipids with great variations with respect to structure, components and quantity. Some of the moieties of the lipids in biomembrane are forced into a particular conformation of the acetyl group and the cyclohexenone ring of PROG. The particularized conformations which may be predominant in a population as well as the basal lipids which make up the membrane architecture may affect specific interaction with other components, such as various enzymes with special conformation. The several membrane enzymes which interact with PROG might cause conformational change accompanying a modification in the membrane function. As to the significance of conformation for membrane enzyme action, it has been suggested that the membrane enzymes or receptors are unable to function properly unless they are surrounded by an appropriate phospholipid for the requirement of pertinent conformation [29-311. Therefore, the diversity of the biomembrane seems to be closely related to regulation of subtle conformational alterations in the various moieties of the steroid throughout interaction. This may cause functional modifications and may provide a reasonable explanation for the nongenomic effect of the steroid. Acknowledgements-We thank Misses N. Otsuki and J. Katayama for their help in the preparation of the manuscript and also Miss Otsuki for her excellent technical assistance. This study was supported in part by Research Project Grants Nos 61402 and 62-503 from Kawasaki Medical School. REFERENCES
1. Baulieu E. E., Godeau F., Schorderet M. and Schorderet-Slakine S.: Steroid-induced meiotic division
860
SATORUWATANABE er al.
in Xenopus laevis oocytes: surface and calcium. Nature 275 (1982) 593-598. 2. Smith L. D. and Ecker R. E.: Interaction of steroids with Rana pipiens oocytes in the induction of maturation. Dev. Biol. 25 (1971) 233-241. 3. Godeau J. F., Schorderet-Slatkine S., Hubert P. and Baulieu E. E.: Induction of maturation in Xenopus laeuis oocytes by a steroid linked to a polymer. Proc. nam. Acad. Sci. U.S.A. 75 (1978) 2353-2357.
4. Kaya H. and Saito T.: Effect of progesterone and its l’la-hydroxyl derivative on human erythrocyte membrane. Japan J. Pharmac. 39 (1985) 299-306. 5. Heap R. B., Symons A. M. and Walkins J. C.: Steroids and their interactions with phospholipids: solubility, distribution coefficient and effect on potassium permeability of liposomes. Biochim. biophys. Acta 218 (1970) 482-495.
6. Seeman P.: The actions of nervous system: drugs on cell membranes. In Cell Membranes, Biochemistry, Cell Biology and Pathology (Edited by G. Weissmann and R. Clairbone). HP Publishing, New York (1975) 239-247. I. McEwen B. S.: Gonadal steroids. Humoral modulators of nerve-cell function. Molec. Cell Endocr. 18 (1980) 151-164. 8. Stumpf W. E. and Sar M.: Steroid hormone target cells in the extrahypothalamic brain stem and cervical spinal cord: neuroendocrine significance. J. steroid Biochem. 11 (1979) 801-807. 9. Alivisatos S. G. A., Deliconstantions G. and Theodosiadis G. P.: Specificity of binding of cholesterol, steroid hormones and other compounds in synaptosomal plasma membranes, and their effect on ouabainsensitive ATPase. Biochim. biophys. Acfa 643 (1981) 65&658. 10. Wasserman W. J., Pinto L. H., O’Connor C. M. and Dennis Smith L.: Progesterone induces a rapid increase in [calcium ion] of Xenopus laeuis oocytes. Proc. natn. Acad. Sci. U.S.A. 77 (1980) 15341536. II. Moreau M., Vilain J. P. and Guerrier P.: Free calcium changes associated with hormone action in amphibian oocytes. D~LJ. Biol. 78 (1980) 201-214. 12. Livingston J. N. and Lockwood D. H.: Effect of glucocorticoids on the glucose transport system of isolated fat cells. J. biol. Chem. 250 (1975) 8353-8360. 13. Bangham A. D., Standish M. M. and Weissmann G.: The action of steroids and streptolysin S on the permeability of phospholipid structures to cations. J. molec. Biol. 13 (1965) 253-259. 14. Willmer E. N.: Steroids and cell surfaces. Biol. Reo. 36 (1961) 368-398. 15. Tomkins G. and Maxwell E. S.: Some aspects of steroid hormone action. A. Rev. Biochem. 12 (1963) 677-708. 16. Watanabe S.: Circular dichroism study of the interaction between conformationally altered human serum albumin and testosterone. J. steroid Biochem. 23 (1985) 177-183. 17. Ronald T. and Gene R. L.: Circular dichroism of
bovine serum albumin-testosterone system. J. steroid Biochem. 7 (1975) 1523-1525. 18 Beecham A. F. and Collins D. J.: The circular dichroism of a$-unsaturated ketones. 1. Circular dichroism in the singlet-singlet and singlet-triplet n +R l transitions of steroidal 4-en-3-ones. Aust. J. Chem. 33 (1980) 2187-2197. 19 Suga T. and Watanabe S.: Stereochemical studies of monoterpene compounds. XIII. The conformational analysis of the acetyl group of 4-methyl-lO-nor-8 oxomenthols by circular dichroism. Bull. them. Sot. Jap. 45 (1972) 24&245.
20 Snatzke G. and Schwinum E.: Circular dichroism. XVII. Low-temperature determinations on 12-substituted 20-oxopregnanes and -16-pregenes. Tetrahedron 22 (1966) 761-771.
21 Wellman K. M. and Djerassi C.: Optical rotatory dispersion studies. XCVIII. The conformation of the 1%acetyl group in steroids and its correlation with variable-temperature circular dichroism. J. Am. them. Sot. 87 (1965) 60-66. 22 Snatzke G.: Circulardichroismus-IX. Modifizierung der octantenregel fur a,/?-ungestlttigte ketone: transoide enone. Tetrahedron 21 (1965) 421-438. 23 Suga T., Shishibori T. and Matsuura T.: Stereochemical studies of monoterpene compounds. VI. Stereochemistry of 2-hydroxypinocamphone and its reduction products. Bull. them. Sot. Jap. 41 (1978) 1175-1179. 24 Allinger N. L., CarabblC P. P. and Perez G.: Conformational analysis. XLIX. The conformations of the acetyl side chains of some compounds related to pregnan-20-one. Tetrahedron 22 (1966) 1615-1623. 25 Examined by “Dreiding Stereomodels”, W. Buchi Manufacture of Glass Apparatus Flawil, Switzerland. 26. Eliel E. L., Allinger N. L., Angyal S. J. and Morrison G. A.: Conformational Analysis. Interscience Publishers, New York (1965) p. 52. 21. Moffit W., Woodward R. B., Moscowitz A., Klyne W. and Djerassi C.: Structure and optimal rotatory dispersion of saturated ketones. J. Am. them. Sot. 83 (1961) 40134018. 28. Hendrickson J. B.: Organic and biological chemistry molecular geometry. I. Machine computation of the common rings. J. Am. them. Sot. 83 (1961) 452334537. 29. Roleofsen B.: The (non)specifity in the lipidrequirement of calcium- and (sodium plus potassium)-transporting adenosine triphosphatases. Lr@ Sci. 29 (1981) 2235-2247. 30. Fleming J. W. and Ross E. M.: Reconstitution of beta-adrenergic receptors into phospholipid vesicles: restoration of 1251 iodohydroxybenzylpindolol binding to digitomin-solubilized receptors. J. Cyclic Nucl. Res. 6 (1980) 407419. 31 Loh H. H. and Law P. Y.: The role of membrane lipids in receptor mechanisms. A. Rev. Pharmac. Tox. 20 (1980) 201-234.