Oct. 20, 1954
ROTATORY DISPERSIONS OF STEROIDS, AMINOACIDSAND PEPTIDES
[CONTRIBUTION FROM THE
DEPARTMENT O F BIOCHEMISTRY, COLLEGE
5037
O F PHYSICIANS AND SURGEONS, COLUMBIA
UNIVERSITY]
Rotatory Dispersions of Some Steroids, Amino Acids and Peptides Using a New Photoelectric Spectropolarimeter BY ERWINBRAND,’ELIZABETH WASHBURN, BERNARD F. ERLANGER~, ERICE L L E N B O G E N , 3 JANET DANIEL, FREDERICH LIPPMANN AND MARION SCHEU RECEIVED FEBRUARY 22, 1954
A new photFelectric spectropolarimeter is described. The precision and accuracy of the instrument were determined from 2400 to 7500 A. using a standard quartz plate. The standard deviation of each reading of the quartz plate was found to vary from fO.OO1 to f0.004” depending upon the wave length of the incident light. Dispersion data are also presented for several steroids, amino acids, and peptides and three constant equations are developed to fit the results.
Rotatory dispersion studies are valuable for purposes of identification and as criteria for purity,4 We have obtained rotations from 2500 to 7500 A. of several biochemically interesting substances using a new photoelectric spectropolarimeter which was developed by one of the authors (B) in cooperation with 0. C. Rudolph and Sons of Caldwell, N. J. The light from a Western Union K-100 zirconium arc lamp in quartz envelope, which is interchangeable with a General Electric Na-1 sodium lamp or a Hanovia 16A13 mercury quartz burner, is passed through a Beckman DU monochromator into a polarimeter. The monochromator has a special base and accessories to facilitate proper alignment of light source, monochromator and polarimeter. The polarimeter has quartz Rochon polarizer and analyzer prisms and an analyzer circle reading to O.O0lo arc. No Lippich prism is u‘sed. Light intensity is measured by a photoelectric attachment consisting of a photoelectric tube housing allowing rapid interchange of RCA 1P21 and 1P28 photomultiplier tubes and a Photovolt photometer with special scale.5 Optical rotations are measured by the method of symmetrical angles.6 The instrument is Rudolph Model 200 S-80. The precision and accuracy of this new polFrimeter were tested between 2300 and 7500 A. a t 25” using a standard quartz plate from the National Bureau of Standards (Table I). The calculated values were obtained from the formula of Lowry, aO/mm.= (9.5639/X2 - 0.0127493) - (2.3113/X2 0.000974) - 0.1905’ where X is expressed in microns. The thickness of the plate was 0.0377 mm. This was determined from the reading a t 5890 A. as Lowry reports the rotation of quartz a t this wave (1) Deceased July, 1953. (2) Department of Microbiology, College of Physiciansand Surgeons,
length to be 21.729°/mm.8 The temperature correction for quartz a t 5890 A. is given by the C ~ O D (0.000143 [f, - 20]).9 expression Q’D = azo^ The corrections a t other wave lengths were considered to be of the same relative size. A quartz plate of low rotation was chosen because most of the compounds measured a t convenient concentrations had rotations of this same order of magnitude. Twenty-six readings were taken a t each wave length and ten readings for each blank. The results were analyzed statistically to obtain the precision of each reading. For the 12 wave lengths between 2500 and 7000 A. the average difference between the calculated and observed values was only 0.003’ and the average standard deviation of each reading between 2400 and 7500 A. was A 0.002”.
+
TABLE I ROTATORY DISPERSIONS OF QUARTZASD SUCROSE Wave length,
aQ/.0377mm.
Quartz Std. Dev. Reading Blank
k.
Ca1cd.a
Obsd.
7500 7000 6500 6000 5890 5500 5460 5000 4500 4000 3500 3000 2800 2500 2400 2300
0.493 ,570 ,666 ,789 ,820 ,949
0.503 ,572 ,664 ,783 ,819 ,946
0.003 ,002
1.163 1.461 1.894 2.562 3.683 4.360 5.825 6.495 7.296
1.166 1.467 1.896 2.561 3.685 4.367 5.828 6.521 7.340
,002
,002 ,002
.003 ,002 ,003 ,001 ,002 ,002 ,002
,003 ,004 ,016
0.003 ,001 ,003 ,004 .003
,001 ,001 ,003 ,001 ,002 ,002
,004 ,002 ,003 ,005
Sucrose Total Dev. 0.006
,003
LTQ
Ca1cd.b
Obsd.
0.480 ,554
0.480 ,560
,798
,797
,939 1.136
,939 1.131
1.872 2.566 3.780
1.868 2.566 3.782
6.303
6.288
,005 ,006
,006 ,003
,003 ,006 .002 ,004
,004 ,006
.005 ,007 ,021
a Calculated values are corrected for 25”. Calculated values are corrected for 25” and a concentration of 1.2 g. per 100 ml.
339 (1940).
The accuracy of our determinations was further checked using 3 ml. of standard sucrose solutions a t a concentration of 1.2 g. in 100 ml. This concentration was chosen because i t was the lowest a t 2 X 0.002” would which the instrument error of never be more than 1% of the actual reading of the sucrose solution. The calculated values were obtained from the equation [a] = 21.648/X2 0.021310 where [a] = lOOa/cZ. The temperature correction is independent of wave length and the
(5) A more detailed description of this equipment and the methods of measurement will appear in separate papers in forthcoming issues of the J . Opt. SOC.Am. (6) Cf.ref. 4b,p. 969. (7) Cf.ref. 4a, p. 258.
(8) Cf. ref. 4a, p. 257. (9) F. J. Bates and Associates, “Polarimetry, Saccharimetry and the Sugars,” Circular of the National Bureau of Standards, C440, United States Government Printing Office, Washington, 1942, p. 92. (10) Cf. ref. 4a, p. 131.
Columbia University. (3) U. S. Public Health Service Postdoctorate Research Fellow,
1949-1951. (4) General references in optical activity and rotatory dispersion: (a) T. M. Lowry, ”Optical Rotatory Power,” Longmans, Green and Co., New York, 1935; (b) W.Heller in A. Weissberger’s “Physical Methods of Organic Chemistry,” Vol. 2, first edition, Interscience Publishers, New York, N. Y., 1946, pp. 869-987; (c) P. A. Levene and A. Rothen in H. Gilman’s “Organic Chemistry,” Vol. 2,John Wiley and Sons, Inc., New York, N. Y.,1938, pp. 1779-1850; (d) “Optical Rotatory Power,” A General Discussion of the Faraday Society, April, 1930; (e) J. Kauzmann, J. E. Walter and H. Eyring, Chem. Reus., 26,
503s
BRAND,~ ~ i l S H R U R YERI. ',
'.SGFK,
ELLENBOGEN,
DANIEL,LIPPMANN AND SCHEU
Vol. i t ;
11
T.4BLE
ROTATORS DISPERSIOSSOF SEVERAL STEROIDS ,--~~. ~
\\.n\c
Cho1e.t erui cIr t = 25 , c l
c;""
I-nfth.
.i. 7500 7 )11i1
Ohcd. b
L'.?lCd.
-90.7 -106 - 15id - 227
-89.4
- 104 - Id4 - 228 - 405
3Wj 3OilO
4000
3500 3400 3000 2800 2600
t
Calcd.
- 594
- 40jd - 592
-71 - 105 - 156 -280 -414
- 981 - 1266
- 9X4d - 1248
- 693 -901
10,A.
I'
Cholesterol CHiOH
= 250, c 0.24
- 1238
Obsd. b
- 74 - 109
-69P - 908 - 1243
-lXd -280d
-425
Cortisone acetate CHCId t = 2 S 0 , c 0 2.; Calcd. Obsd. C
515 607 936 1469 2996 5152 6205
Cortisone acetate CH2OH t = 23 , c 0.2a Calcd. Obsd. C
527 603 936d 1469'' 3086 5152d 5997
463 546 847 1342 2869 5274 3523
466 546
845d 1339
2864d 5329 3523d
18GO 2480 2720 The chloroform and niethanol used were spectro grade Eastrnan 88-337 and S-467. Average of six determinations. Average of two determinations which did not differ by more than 1%. Indicates values used to calculate equations. 1830
a
+
coefficient is -0.000184 0.0000063 (t - 20).11 The concentration correction a t 5890 A. and 20' is GG.462 0.00870 (c) - 0.000235 (c*).~~The correction a t other wave lengths was considered to be of the same relative size. From 3000 through 7500 A.the largest difference between the observed and calculated readings was 0.005". The average difference was 0.002". Tbe comparatively large deviation of 0.015° a t 2500 A. is probably due to the effect of stray light which passes through the wider slits necessary in the ultraviolet. Drude expressed the relationship between optical rotation and wave length as a: = k / X 2 - Xi where k is known as the rotation constant, X the wave length of the incident light, and XO the dispersion constant which corresponds to the wave length of the nearest optically active absorption band. The rotation of a compound can be considered as the s u m of the contributions of the partial rotations each of which is caused by a different optically active absorbing center, a: = Bk/X2 - 1%. The dispersion is defined as normal if the rotatory power increases numerically with decreasing wave length and if cy, dcr/dX, and d2cy/dX2remain constant in sign throughout the range of wave lengths to which the medium is transparent. Anomalous dispersion in a region outside an absorption band is produced by partial rotations of opposite sign. \Tithin an optically active band the unequal indices of extinction for dextro and levo circularly polarized rays cause anomalous dispersion, an effect known as the Cotton effect. Rotatory dispersion data may be analyzed algebraically and graphically. The Drude equation in its reciprocal form is an equation for a straight line and it is convenient to plot 1 / a against X2. A straight line is usually obtained in the visible region of the spectrum; some deviation from linearity usually occurs in the ultraviolet regions where the rotation is influenced by more than one optically active absorbing center or where the Cotton effect may begin to be seen. The value of XO may be obtained from the x(or X2) intercept of the extrapolated curve. The shape of the curve determines the number oi terms in the Drude equation necessary
+
(11) Cf ref 9 p. 93 (12) Cf ref 9, p. 82
for correct expression of the data.l3Vl4 'I'woJ three and four constant equations of the type CY=
can be calculated. Usually no more than two terms are needed to express the results, one corresponding to the partial rotation of the band with the longest wave length and the other to the sum of the other partial rotations. The XO values have no physical significance unless the equation expresses the results close to the region of absorption. The foliowing compounds were studied from 2500 to 7500 A. : c h o l e ~ t e r o l cortisone ,~~ acetate,16 the L-isomers of alanine, glutamic acid, lysine, ornithine and t y r o ~ i n e ' ~and several di- and tripeptides. The results are expressed algebraically. All of the curves obtained, except that of tyrosine, show normal dispersion with a small deviation from linearity in the ultraviolet. The tyrosine curve is anomalous. IVith decreasing wave length the rotation goes through a minimum, becomes zero, changes sign and approaches a. Good agreement is obtained between observed values and those calculated from 3 constant Drude equations. The values of A0 calculated from the equations are compared with the wave lengths of maximum absorption reported in the literature. The results with steroids (Table 11) indicate that the presence of many asymmetric centers is not incompatible with normal dispersion. LZethanol as well as chloroform was used as a solvent since absorption by ghloroform does not perinit readings below 2800 A. The dispersion equations for cholesterol, calculated from results a t thrcc wave lengths, are
+
[M]CHCli =
71.6 - X* -~ -0.0335
26.8
-!-
7' 50.0 - x__ + 19x 29 2 - 0 0347
[hf]CH,OH =
.
(13) H. Hunter, J. Chcm. Soc., 125. 1198 (1924). (14) Cf.ref. 4a, pp. 419-422. (15) Kindly supplied by Dr. Louis Fieser of Harvard University. H e reports [ h l l n in CHCls = -161. ( 1 G j Kindly supplied b y Dr. M a x 'l'ishler o f lllerck a n d Co., K . i b way. N. J. 1171 Kindly supplied b y Dr. R', H . Stein of Rockefeller I n s t i t u t e , rf. l ' i t ~ sJ O I I R N . ~ L 64, 721 (ISlOi, desrrihed as commercial sample r e crystallized from hydrochloric acid a n d ammonium acetate.
ROTATORY DISPERSIONS OF STEROIDS, AMINOACIDSAND PEPTIDES
Oct. 20, 1954
5039
TABLE I11 ROTATORY DISPERSIONS OF SEVERAL DIPEPTIDESIN 0.5 N HCl, [a]02f1.6-J.6 . ~.~ H.Ala-Bla.OH
7
\Wave length,
= C, 2
(LL) - .___ 61.0
- ,0085 '5
xa
A.
Calcd.
7500 7000 5890 5000 4000 3500 3000 2800 2700
-25.0 -36.9 -53.6 -91.7 -130 -196 -238
H*Glu-Glu*OH
A%
0bsd.a
+ (4 6.69
(LD)
+
50
Calcd.
37.3 16 [a]zm - .os45 x i c , 2.5 Calcd. Obsd. b
x2
10.0
A2
in 0.5 N HC1.
49.5 73.7 110 199 295 507 651
-
f
48.5 73.7'" 1101 198 294 498 651'
11.6 17.4'.' 26.4' 50.5 81.0 156' 227
30.1 35.1 51.8 76.5 136 198 324 414
(DL)
-26.6 f 11 ~ ~ 1 A.2 6- ,0328 7 Calcd. 0bsd.d
27.0 -12 - .0304 A 2 Obsd. C
[ a ]=
-30.6 -30.8 -35.7 -35.4 -52.9 -52.9'2' -78.4 -78.4 -140 -140' -207 -207 -343 -343' -443 -447 -512 -523 A 0 A. 920 1860 2280 1740 1810 Average of four determinations. Average of two determinations. One determination. One determination. ~ Specific rotatioil calculated on the basis of the free peptide. e Rotation previously reported (cf. ref. 22a) with [ a ]2 4 -37.3" -25.0 -36.8'3' -54.1' -92.4 - 131 194' - 227
9.50 11.5 17.4 26.4 50.5 80.0 156 230
H~Orn-Orn~OH~2HCl
+(LD)
-1.8 [a]= - .OS21 Xa c,3.6 Obsd. C Calcd.
30.9 35.0 51.8"' 76.5' 135 198 324' 413
Rotation previously reported (cf. ref. 22a) with [a]"D f74.1" in 0.5
c.2
N HCl.
0
Rotation previously
e12
f56.4" Rotation previously reported (cf. ref. 22b) with [cY]'~D
reported (cf.ref. 22b)with [a]?'D f18.2' in 0.5 N H C I . c.1-2
in 0.5
i Rotation
HC1.
e.1-2
previously reported (J. Polatnick, Doctoral Dissertation) with
[aIz3D
-52.3'
in 0.5 N HC1.
e,z
i
Indicates values used to calculate equations.
TABLE IV ROTATORY DISPERSIONS OF SEVERAL TRIPEPTIDES IN 0.5 N HCl,
Wave length,
(LDD) +2 c,2.5
A.
Calcd.
7500 7000 5890 5000 4000 3500 3200 3000 2800 3750 2700
53.9 62.8 93.0 137 245 361 480
Q
xz
-~+
54.8 62.1 92.5" 13ih 245h 361 480h
H~Ala-Orn-Ala~OH.HCl---
(3
t 5 ' O H 2 113.7 [ a ]= 13.3-_ ~ 2 [a] 3 = - .0~52 c2.4 x 2 --.040i x2 c2.1 Obsd." Calcd. 0osd.a Calcd.
(DLL)
+10.7 -.53 ,0440 xz 0bsd.u
x~ -
].[ = ---jOL9-- +11.5 cJ.7 ~2 - , 0 3 7 g X ' Calcd. 0bsd.a. b
20.8 24.2 35.8 52.9 94.6 160
20.1 24.2 35.4d 52.gh 94.6h 157
19.8 23.0 33.9 50.0 89.3 132
19.9 -38.4 23.3 -44.8 33.gEnh -66.6 50.2 -99.5 89.3* -180 133 -271
238 314
23Sh 311
228 305
22gh 301
-38.8 -44.1 -66.6'~~ -99.2 - 180h -270
-463 -615 -666
-463h -617 -668
[.I-
-
(LLL)
-
= -26.6 +I2 7 C,1.9 ,0264 A * Cacld. Ohsd a* h
-27.1 -31.6 -46.5 -68.4 120 - 174
-27.4 -31.4 -46.5g,h -68.3 - 120h - 173
- 278
- 278 -351h
-
-351
364 36 1 2000 2090 1950 1630 Calculated on the basis of the free peptide. One determination. Rotation prevously reported (cf. ref. 2%') with I ~ ~ +92.2" ~ D in 0.5 N HC1. Rotation previously reported (cf. ref. 22c) with [ a I z 54-34.5' ~ in 0.5 N HC1. E Rotation
xo, A. [ D
CY]^!,.^-^,^
H. Glu-Ala. OH
18x0
C.2
previously reported (cf. ref. 22c) with tion) with [ a ] 2 3 -65.4' ~ in 0.5
c,2
HCl.
[ a I z 3 D $32.8" c,2
in 0.5 -\' HC1.
0
in 0.5
N HC1.
3 78 135 x2m -~ 2 ;
The XO values are 2480 and 2720
A.,
Previously reported (J. Polatnick, Doctoral Disserta-
Previously reported (J. Polatnick, Doctoral Dissertation) with [ a J Z-46.7" 3~
Indicates values used to calculate equations.
The valu: of XO is equal to the square root of 0.0335 or 1830 A for the chloroform results, and to the square root of 0.0347 or 1860 A. for the methanol results. Heilbron, et aZ.,'* report that cholesterpl has no marked selective absorption down to 2000 A. The dispersion of cholesterol was reported in 1910 by TschugaefT a t four wave lengths between 4800 and 6500 A. The results a t 5890 A. were 15% lower than those found in this investigation.Ig The two equations calculated from the cortisone acetate results are [MICHCIZ f
c,2
f
respectively.
(18) I Heilhron, R. Morton and W. Sexton, J Chrm SOC..47 (1928). (19) L. Tschugaeff and W. Formin, Ann.. 575, 288 (1910).
c2
The reported absorption maximum is a t 2350 A.2o Five dipeptides and five tripeptides, including several sets of optical isomers, were analyzed. The specific rotations a t 25" and a concentration of 2 g. in 100 ml. are reported (Tables 111, IV).21.22 (20) L.H. Sarett, J. B i d . Chem., 162, 630 (1946). (21) The following abbreviations and symbols are used i r f . 11. Sachs and E. Brand, THISJOUKNAL, 7 6 , 4608 (1953)); Ala, NHCH(CHdCO, C:HsON; Glu. NHCH(CH,CHCOOH)-CO, CrH70sN; Om, NHCH(CIH~NHS)CO, CaHmONt; peptide linkage indicated by dash, -: configuration follows compound in parentheses, ( ). When the y-carboxyl group of glutamic acid is suhstituted, the following symbo1 is used for the residue: Glu. e.g., L-glutamyl-a-L-alanine-y-DLOH alanine: H*Glu-Ala*OH(LDL): a-1,-glutamyl-D-glutamic acid: H*Glu* LAla.OH Glu-OH (LD); L-glutamic acid: H*Glu*OH(L). (22) The peptides are reported in the following papers: (a) B. F. Erlanger and E. Brand, ibid.. 73, 3508 (1951): (b) H. Sachs and E. Brmd. i h i ~ l . , 7 6 , 4608 (1953); (c) H. Sachs and E. Brand, ibid., 76, 1811 (19.54); (d) J. Polatnick and E. Brand, to be published.
BKAND,L~ASHBUKN, EKLANGEK, ELLENBOGEN, UANIEL,LIPPMANN AND SCHEU
5040
Vol. 76
TABLE \.' Wave length,
a
7500 io00
6890 5000 4700 4500 4300 4000 3520 3500 3300
:moo
2800 2700 2600 2500
--
ROTATORY DISPERSIONS OF L-WAMINOACIDSIN 0.5 N HC1,"
-
H.Glu.OH 98 -332 X* - 0482 A * Calcd. Obsd
[ a ]=
Calcd
18.4 21.4 31.8 4T. 1
18.6 21.1 31.7' 47.2
8.00 9.40 11.3 22.1
85.3
85.3h
129.0
130.0
230 310
23 1
144
31Rh
391
393
Calcd
8.70 9.43 14.3' 21.9
15.9 17.6 2 5 . 8h 38.5
16.1 18.8 28.0 41.8
16.9 18.8 28. Ob 41.7
4.3 . 3
43. 3 h
i1.3
i 1 .3'
i6.i
76, ib
70.0
70.9
109
110
117
11;
220
146 220h
202 285
202 285b
214 300
213 300h
41'4
389
46 1
458
-8.20 -11.0 -13.0 -14.3 -14.0 -14,5 -1:3.1 -0.06 t-l.22 f23.2
A.
2200 2330 All results are the average of two determinations.
io,
Calcd.
14.8 17.3 25.8 38.6
h
864 ti50 2220 2210 Indicates values used to calculate equations.
'$he calculated values of XO lie between 900 and 2300 A. Saidel and Goldfarb report a peptide bond absorption band near lS50 A.23 The rotatory dispersions of the a-amino acids have been studied by Patterson and Brode between 4400 and 6600 A.24 M'e have extended this study to sixteen wave lengths between 2500 and 7500 K . (Table V). The first term of the calculated equations is positive for each of the five L-acids and the XO values are between 2200 and 2900 A. The absorption of amino acids, i.e., alanine, valine, leucine, glycine, cystine, in the ultraviolet starts a t approximately 2500 A. and continues to rise at 1800 A . 2 5 The tyrosine curve is an excellent example of anomalous dispersion outside an absorption band and is due to two rotatory contributions of opposite sign. The presence of an aromatic nucleus also causes anomalous dispersion in a series of alcohols, reported by Pickard and Kenyon.26 The dispersion of these aromatic alcohols was very susceptible to temperature changes. Tyrosine was run at 25 and 20". The results a t 2.5' are about l4yOhigher than those a t 20". The dispersion ratio is the same. The three-constant equation calculated
- 7 9; -11 O b -13 8 -14 3 - I -1 (i" -I* 1 -12 9 +l.lfi +3 76 f33 2"
2870
from the results at 25' does not express the data very well. X four-constant equation was attempted with three different sets of four values but each time an imaginary result was obtained. The Ceported absorption peak of tyrosine is a t 2740 A. 27 Experimental All dilutions were made in 3-ml. volumetric flasks. Onc decimeter glass polarimeter tubes with quartz end plates were used. The strain effect on the quartz was minimized by adjusting the end plates so that the empty tube never read more than 10.02O from the zero reading of the instrument at a n y wave length. The blank was read in the same tube without changing the strain 011 the end plates or the position of the tube in the trough. To eliminate the determination of a blank for each set of data several tubes with cemented quartz end plates were tried, but none was satisfactory. ,4 water jacketed trough was developed and used in conjunction with an Aminco constant temperature bath. All solvents were kept in the bath and the filled polarimeter tube was allowed t o stand in the trough for 20 minutes hefore readings were taken. About one hour is needed to ohtain data at 10 wave lengths.
Acknowledgment.-We wish to thank Dr. Alexandre Rothen of the Rockefeller Institute for Xedical Research for his interest and assistance in the preparation of this paper. This work was aided by a contract between the Office of Naval Research, Department of the Navy, and Columbia University (NR 119-035).
( 2 3 ) L. J. Saidel and R . Goldfarb, Sciewcc, 114, 156 i l R 5 l ) . 124) J . W. Patterson anrl LT'. R . Drode, 4 1 ' c h . Biochem., 2 , 217 (1443). Xmv YORK,SEW YORK -~ 8\23) H. Ley and B. Arends, %. p l i y s i k . Cheiiz., B1T. 1 7 7 (1932). I2ljl li. H. Pickard anrl 1.Kenyon, .i. (,'hem. S o c . , 105, 1113 ( I R ~ ~ J , ( 2 7 ) S. Kretchincr :ind R . 'l'aylor, T H I SJ < I U K Y . P I . , 72, 3L'III (10.X1 ~
~
