INSTITU'I'E FOR MUSCLE RESI~ARCH, MARINE BIOLOGICAI~ L4BORATORY, WOODS HOLE, MASSACHUSETTS INTRODUCTION If a bullock lens is illuI~~in~ted by near EV, it emits a. brilliant blue light, which is due to its proteins. 1 This is unexpected because two of the three aromatic amino acids-phenylakmine and tyrosine-have no visible fluorescence, and tryptophane has but a very weak one. h similar blue light is emitted if' the lens is tooled in dry ice or liquid Nz, In neither case does the lens show an afterglow. If, however, the lens is illu~ninated by shorter UV, then ZL long-lasting afterglow can be observed. The monomolecular kinetics of the decay indicate that the Light, in this case, was emitted from a metastable triplet states2 In order to arrive at a better understand- ing of these phellorneila, we studied the optical properties of the aromatic amino acids and a number of proteins better defined than the proteins of tho lens. Our results can be summed up by saying that the optical behavior of the lens and other proteins, under our ~onditioIls, is do~li~~ated by their tryptoph~~~e and that trypto- phane, in a frozen watery solution, also shows two emissions---a short-lived one elicited by near fiV and n long-lived one elicited by a shorter wave length. This indicates that the tryptophsne within the protein is present in two different states which can be identified by t,heir different excitability and light emission. The analogous two states can be detected also in frozen watery solutions of this amino acid. RXPFiRIMENTAL Absorption spectra were measured with the Beckman DK-I automatic recording spe~~trophotoIneter. A variable-path quartz absorption cell was used in the absorp- tion-spectra measurements of the concentrated protein preparations. VOL. 44, 1958 BIOCHEMISTRY: STEELE AND KZENl'-GYCjRGYI 541 Room-temperature emission spectra of the prot!eins were determined with the Beckman DK-1 spectrophotometer, using the technique described by Gemmill. For low-temperature emission studies (77" K.) this technique was modified by positioning an unsilvered Dewar flask containing liquid nitrogen in the DK-1 lamp housing compartment. The samples under study were placed in quartz test tubes and precooled in an alcohol dry-ice bath and then inserted into the liquid nitrogen. Exciting light was supplied by a Henovi:l high-pressure xenon arc (Model 10-C-l). The grave-length region for exultation was isolated with a Leiss singl.+prism (fused- quarLz) monoc~~roma~tor. It should be recognized that, with the exceptions noted below, none of the emission spectra for prot,eins and the detailed study of trypto- r 60 I L I t I 300 350 400 450 500 XI n mu FIG. Z.--Phosphorescent. spectra of (a) ph~n~laianine, (b) tyrosine, (c) tryptophane. Aqueous 0.001 M solutions containing 0.5 per cent glucose. I' = 77' K. FIG. 1 .-Molar est,inction wave numbers for the aromatic amino acids: a, tryptophane; b, tyrosine; c, phenylalanine. phane were corrected for the spectral-response sensitivity of the photomultiplier detector in the DK-1 spectrophotometer or for the prism efficiency of the mono- chromator of the DK-1. This uncorrected feature applies as we11 to the trans- mitted exciting light curves. It is to be noted, however, that the photomultiplier was an RCA 1p28 multiplier tube with S-5 response* with peak sensitivity at ap- proximately 340 my. In practice, on the short-grave-Iength side of the 340-rnp maxi- mum, the recorded curves (including the transmitted exciting light curves) are shifted approximately 5 m,u toward longer wave lengths, while on the long-wave- length side of 340 mu the recorded curves are shifted approximately 5 mu toward shorter wave lengths. These shifts are not important for any of the conclusions re- ported in this paper. The fluorescent and phosphorescent spectra reported in Figure 2 were obtained using Farrand grating monochromntors to isolate the exciting Iight (xenon arc) and to analyze the emitted spectra. These curves were corrected for the spectral sensi- tivity of the RCA IF'28 photomultiplier tube. RESULTS Aromatic Am&o Acids,---Figure 1 gives the absorption spectra of the three aromatic amino acids which are responsible for most of the 2300-3100 A absorption of protein. There is no apparent absorptio~~ at a longer wave length than 32,000 cm.-' (3120 A). The longest-wave-length absorption of tryptophane appears to be the unresolved peak masked under the shoulder at 34,000 cm.-' (2,900 A). Our measurements of the fluorescent emission of these amino acids are in good agreement with those of Teale and Weber,5 the maxima being at 284, 302.5, and 348 mu for phenylslanine, tyrosine, and tryptophane, respectively. These emissions 600 500 400 300 280 A In my FIG. 3.--Emission spectrum of intact bullock lens, (a) excited in the 340-rnp region; (b) excited in the 280-rnb region. In all figlwes the wuve line marks transmitted exciting light,. represent transitioils from the low- est excited singlet state to the singlet ground state.6 Figure 2 gives the phosphorescent spectra for the three amino acids. We have presented data2 which led us to interpret. the low-temperature emission for tyrosine and trypto- phane as a transition from the low- est excited triplet state to the singlet ground state. Though the phenyl- alsnine emission decay was too rapid for us to make kinetic studies with the equipment we had avail- able, we have interpreted the long- wave-length shift with its close cor- respondence to the tyrosine emission maximum as evidence that this emis- sion isalso triplet-singlet in character. ~2~~~~~~~ Lens.--At room temperature, excitation in the general region of the 280-rnH protein absorption band gave rise to an emission with a maximum at 340 rnp which we consider to be fluorescence origin~~.ting from tryptophalle residues in the protein. Shifting the exciting light to longer wave lengths elicited an intense emission with the same blue color as the phosphorescence we had observed from the same prepay&- tion at low tempera,ture (77" K,). Curve b in Fig. 3 gives the emission spectra of the bullock lens excited within the protein 280-rnp absorption band and by near UV at 77' K. (liquid NS). The 330-rnp emission band had no delayed emission observable visua.lly on our not too fast in- struments. The 450 rnp band, in contrast, exhibited a considerable aftergIow. When we first observed the intense blue emission elicited by excitation in the 340-rnp region, we suspected that we might, he inducing a direct singlet-triplet transi- tion from the ground state or else that we were exciting the protein in a heretofore un- determined absorption band. It might be reasoned that a singlet-triplet transition which would give rise to the intense emission we observed should manifest a reasonably strong absorption and that, consequently, a spectral examination of a co~~centrated protein preparation might reveal this fact. Such a study did, indeed, reveal a long-wave-length maxi- mum in the absorption spectrum at 340 rnp. A long-wave-length shoulder was observed in this general region for all the proteins we have examined to date, but we have not felt justified in publishing the data, for we have not assured ourselves that scattering was not considerable in our experiments. Suffice it to point out that other workers, notably Goodwin and Morton, 7 have noted an "irrelevant absorp- tion" in proteins in the 340-400-m/1 region, where the aromatic amino acids do not, absorb. When, however, the lens was excited in this absorption region at 77" K., the intense blue emission had no prolonged decay time. We may conclude that, if the low-temperature, long-wave-length emission elicited in the 340-niti region was origi- nating from a metssta,ble triplet level, it was not the same level that 5 is populated under the sa.me condi- `i tions by exciting the protein in the iI0 280-rnp absorption band. z = Homogenized lens exhibit'ed the s same spectral behavior as did the i 5 intact material. The addition of E glucose to the lens homogenate did not influence the lifetime of the 77' 700 500 400 300 K. blue emission elicited by excita- A In my tion in the long-wave-length 340-rnp FIG. $.---Emission spectrum of bovine serum al- region. bumin, excited by light of different wave lengths. 10 Albumin, Trypsin, per cent aqueous solution, containing 1 per cent Mypsin.- glucose. T = 77" K. Spectral data for bovine serum alb- umin, recrystallized trypsin, and myosin we represented in Figures 44. The emissions elicited by irradiation in the main protein absorption band around 280 FIG, S.----Emission spectra of ;w~y;sJl~d trypsin (0.245 gm. ml.-l). 000 600500 400 300 260 A inmy FIG. 6.-Xmission spectra of myosin at different exciting wave lengths. 0.24 per cent aqueous solution. 2' = 77" I(, 544 BIOCHEMISTRY: STEELE AND SZENT-GYCiRGYI I'RC)C. N. A. s. rnl.r were long-lived showing a considerable afterglow, while the emissions elicited by near UV around 340 rnp were short-lived. No afterglow could be seen visually or detected by our relatively slow instrumentation. The results are summed up in Table 1. With the exception of the long-wave-length emission of myosin, elicited by ex- citing in the 340-rnp region, the emission spectra for the several proteins mirror the emission properties of tryptophane more than that of phenylalanine or tyrosine. Parenthetically, the close similarity between the low-temperature emission proper- ties of the proteins and tryptophane, elicited by exciting in the absorption bands, supports the concept expressed by Teale and Weberb and ourselves,s that energy absorbed by phenylalanine or tyrosine is transferred to tryptophane, from which molecule emission occurs. The minimum in the emission spectra nround 400 rnp, where phenylalanine and tyrosine phosphoresce (see Fig. Z), also bears out this idea. It might be considered that the high molar absorbancy of tryptophane rela- tive to that of tyrosine or phenylalanine would result in tryptophane being the primary emitter. An examination of Table 2, where the molar ratios of the aro- matic amino acid residues in the proteins are given, shows that, when allowance is made for the ratio of molar absorbancies (approximately 1: 5 : 35 for phenylalanine, t,yrosine, and tryptophane, respectively; see Fig. l), phenylalanine and tyrosine should be almost as efficient emitters as tryptophane. Further, when the quantum yields are considered, tyrosine has approximately the same value as tryptophane, rlamelyj 0.21 and 0.19, respectively (Teale and Weber5). The yield for phenylala- nine is only 0.045, which may account for the nonobservance of an emission in the 300-rnp region, where phenylalanine and tyrosine fluoresce. Conversely, it may be TABLE I* LONG-Wnvm- LENGTH AUYORPTION IN AQUEOUS IkUORESCENCE PROTEIN SOLUTION 298" K. 7T' K. lhllock lens 340 345 335 Hovinc swum albumin 350 340 335 Trypsin 340 342 337 Myosin 340? 340 333 * Numbers stand for wave lengths of maxima in mu. EMISSION LONG-LIVID EIZXTI~ BY NEAR I'aosPIron- UV (E 340 mc) ESCENCE 298O K. iTo K. 450 440 440 460 470 470 450 450 450 440 ? 420 TABLE 2* Moms OF AMINO ACIDS IN lo6 GM. PROTEIN I'ROTEIN PHENYLALANINE TYROSINE T~u~rorrrnN~ Bullock lens 35 11 Bovine serum albumin ii: 29 3 Trypsin 23 Myosin ;"8 18 i Insulin 52 69 0 * compiled from date taken from R. J. Block and K. W. Weiss, Amino Acid Handbook (Springfield, III. : Charles C Thomas, 1956). argued that the low quantum yield of fluorescence is evidence for an increased transi- tion probability to the kiplet state, where it may be more readily degraded as heat due to the increased lifetime. We, however, are inclined to the idea that the excita- tion energy is transferred by resonance to tryptophane and then emitted.* Insulin and Tryptophane. -The close similarity in the emission properties of the proteins and tryptophane led us to a compa.rative study of the emission behavior of tryptophalle and of insulin, which contains no tryptophane, Since we have previously noted2 that the blue delayed emission of tryptophane (phosphorescence) was observable only after the addition of glucose to the system (in the absence of glucose what appeared to be the same blue emission was short- lived and decayed too rapidly for us to measure the decay kinetics), we examined the emission behavior with and without glucose. Though the kinetic behavior bore out our earlier observations, the emission spectra revealed interesting new details. These studies with watery solutions of tryptophane-no-glucose and tryptophane- glucose gave the following results: 1. Tryptophane-no-glucose excited in the absorption band at 77" K. displayed no fluorescent emission band, and the blue emission band with maximum at 435 rnp had no afterglow. 2. In the absence of glucose, 340-rnp excitation, at 77 O K., elicited 435-rnp emis- sion, just. as found for proteins; this emission also had no afterglow. 3. Tryptophane in the presence of glucose at 77O K. exhibited a fluorescent band with a maximum at 325 rnp, and the blue band emitted at 440 rng displayed a phos- phorescent afterglow. 4. Long-wave-length excitation of tryptophane in the presence of glucose failed to elicit the 435-m/l short-lived emission. In these experiments aqueous 10d4 M tryptophane solutions were used. The insulin was studied at 77" K., and the exciting light was varied from 260 to 450 ml*. When the protein was excited in the region of protein absorption (260-290 mp), there was evidence of emission in the 290-340-rnp region, which was suggestive of fluorescence from the phenylalanine and tyrosine residues, but the intensity was too low to make positive assertion. It is of interest to note, however, that we observed no 440-rnb emission band, no matter what exciting wave-length region was selected. That is, phosphorcscencc, elicited by exciting in the region of protein absorption, was absent, as was the short-lived long-wa~re-length emission elicited by exciting the protein in the 340-rnp region. Xnsulin contains no tryptophane. * This work was supported, in part, by grants from the -National Institutes of Health (RG 4643 and RG 5374). t Present address: Department of Biochemistry, School of Medicine, Tulane University, New Orfeans 18, Louisiana. 1 This work was supported, in part, by a grant from the Commonwealth Fund, the National Heart Institute (H-2042R), the Muscular Dystrophy Associations of America, Inc., the National Science Foundation, the American Heart Association, the Association for the Aid of Crippled Children, and the United Cerebral Palsy Foundation. 1 Szent-Gyorgyi, A., Biochim. et Biophys. Ada, 16, 167, 1955. 2 Steeie, It. II., and Scent-Gyijrgyi, A., these PROCEEDINGS, 43, 477-91, 1957. 3 Gemmill, Ch. L., Anal. Chem., 28, 1061-64, 1956. 4 Radio Corporation of America, Tube Handbook 6 Teale, F. W. J., and Weber, C., 3~och~. J., 65, 476-82, 1957. 6 For all spectra reported in this paper (with the exception of that reported in Fig. 2) quartz monochromators were used to isolate both exciting and emitted spectral regions, and nondescript transmission was no problem. 7 Goodwin, T. W., and Morton, R, A., Biochem. J., 40, 628-32, 1946. * Karreman, G., Steele, R. H., and Scent-Gyijrgyi, A., these PROCEEDINGS (in press).