The Colors of Firefly Bioluminescence: Enzyme Configuration and Species Specificity by H

Home , Firefly luciferase, Firefly luciferin, Photinus pyralis, Pyrophorus (beetle)

THE COLORS OF FIREFLY BIOLUMINESCENCE: ENZYME CONFIGURATION AND SPECIES SPECIFICITY BY H. H. SELIGER AND W. D. MCELROY MCCOLLUM-PRATT INSTITUTE, JOHNS HOPKINS UNIVERSITY Communicated May 25, 1964 We have previously reported on an unusual stereospecificity of firefly luciferase for a D(-) isomer of firefly luciferin.' While both the D(-) and the L(+) form will react with ATP to liberate pyrophosphate in the reaction E + LH2 + ATP =- E. LH2AMP + PP, (1) only D(-) LH2AMP will react further, in the presence of oxygen, to produce bioluminescence and an oxidized product. There is also a strong pH dependence of the color of the emitted light;2 in acidic buffer solutions, pH < 6.5, the intensity of the normal yellow-green emission, peaking at 562 ml,, decreases markedly and a low intensity red emission is observed, peaking at 616 miu. This is evidence that enzyme configuration is important in determining the resonance energy levels of the excited state responsible for light emission. Further Evidence for Configurational Changes.-Except for the partial denaturation of the enzyme in acidic buffer, the pH effect on the emission spectrum shift is completely reversible. We have been able to observe these same reversible red shifts in emission spectra by increasing the temperature of the reaction, by carrying out the reaction in 0.2 M urea and at normal pH values (7.6) in glycyl glycine buffer, by adding small concentrations of Zn++, Cd++, and Hg++ cations, as chlorides. The normalized emission spectra of the in vitro bioluminescence of purified Photinus pyralis luciferase for various Zn++ concentrations are shown in Figure 1. The

1.0

FG1.7Nraie.3 msinspcr ftei ir P. pyralisbouiecnea aiu .2

4763 4925 5087 5250 5412 5574 5736 5898 6061 6223 6385 6547 6709 6871 7033 7195 WAVELENGTH (ANGSTROMS) FIG. 1.-Normalized emission spectra of the in vitro P. pyralis bioluminescence at various concentrationsofZn++. (a)Norm~a1,noaddedZn++. (b) 1.33 X 1O4MZn++. (c) 3.95 XIO-4 M Zn+ (d) 2.3 X 10O3 M Zn+ 75 Downloaded by guest on September 24, 2021 76 BIOCHEMISTRY: SELIGER AND MCELROY PROC. N. A. S.

curves do not show relative changes in intensities in going from the yellow-green to the red emission; there is a marked decrease in the efficiency of the light reaction under conditions where red light is emitted. The maximum red shifts obtained at low pH, with Zn++, and with Cd++ are shown in Figure 2, compared with the normal yellow-green emission. The temperature, urea, and Hg++ effects are essen- tially the same. In the case of Hg++ the concentrations required for the red shift were 100 times smaller than for both Zn++ and Cd++. By the use of suitable blue-green transmitting filters and red transmitting filters we have been able to establish that there are two pH optima and two temperature optima for the bioluminescence emission, one each for the red and the yellow-green emissions. These are shown in Figure 3. As would be expected, the red emission has a lower pH and a higher temperature optimum. Evidence for Species Enzyme Specificity in Color of Light Emission.-It has been known for some time that different species of fireflies emit different colors of bioluminescence, ranging from green through bright yellow. These are valid observa- tions and are not visual artifacts due to selective cuticle absorption in the light organ. We have recently measured the in vivo emission spectra of 20 species of firefly, 16 Jamaican and 4 native American species.3 The large range of in vivo spectral variations is summarized in Table 1. From the symmetric shapes and TABLE 1 FIREFLY SPECIES AND WAVELENGTH OF MAXIMUM INTENSITY (PEAK WAVELENGTH)* Species Peak wavelength Photuris pennsylvanica 5524 A Pyrophorus plzgiophthalamus (dorsal organ) 5430 A Diphotus sp. 5550 A Photuris jamaicensis c, 9 5550 A Photinus pardalus 5600 A Photinus pyralis c, 9 5621 A Photinus commissus 5640 A Photinus marginellus 5646 A Photinus pallens 5650 A Photinus xanthophotus 9 5670 A Photinus leucopyge 5690 A Lecontea sp. 5700 A Photinus lobatus 5700 A Photinus evanescens 5700 A Photinus melanurus 5700 A Photinus nothus 5700 A Photinus (new species) 5700 A Photinus morbosus-ceratus 5710 A Photinus gracilobus 5720 A Photinus scintillans 9 5748 A Photinus scintillans c? 5751 A Pyrophorus pliophthalamus (ventral organ) 5820 A * Arranged in order of increasing wavelength. reasonably constant half-widths of the curves obtained, it is improbable that peak emissions intermediate between green and yellow are due to mixtures of a green- emitting molecular species and a red-emitting molecular species. The emission is most likely due to a single excited enzyme-substrate complex. We succeeded in collecting sufficient numbers of Photuris pennsylvanica (U.S.), Photuris jamaicensis (Jamaica), and Pyrophoru splagiophthalamus (Jamaica), to Downloaded by guest on September 24, 2021 VOL. 52, 1964 BIOCHEMISTRY: SELIGER AND MCELROY 77

extract and partially purify both luciferase and luciferin from each species. The Jamaican elaterid beetle, Pyrophoru-s plagiophthalamus, has been described by Harvey4 as being unique among the fireflies in possessing light organs which emit two different colored lights. A symmetric pair of dorsal organs on the anterior

1.0

.8 FIG.2.7Normalized emission spectra of the in vitro P. pyralis bioluminescenceP.pyrsiis

Z .6 w Z -5

I- .4 -LJ

4763 4925 5087 5250 5412 5574 5736 5898 6061 6223 6385 6547 6709 6871 7033 7195 WAVELENGTH (ANGSTROMS) FIG. 2.-Normalized emission spectra of the in vitro P. pyralis bioluminescence showing the maximum red shifts obtained with acid pH, Zn++, and Cd++. (a) Normal, pH 7.6, no added Zn++ or Cd++. (b) Maximum effect of acid pH (pH 5). (c) Maximum effect of added Zn++ (2.3 X 10-3M). (d)MaximumeffectofaddedCd++(1.2 X 1O-3M).

0) A P. pyro/ls(in vitro) I 2 o- RED EMISSION(Corning Filter 2-64) *-GREEN EMISSION (Corning Filter 5-60) I Phototube -EMII. ,95588 I 6.8 7.0 7.2 7.4 7.6 7.8 8.0 8.2 8.4 8.6 8.8 9.0 Xj pH 10 ~8

> * . 4

2 TEMPERATURE FLASH HEIGHTS MEASURED AT pH 7 60

14 16 8 20 22 24 26 28 30 32 34 36 38 TEMPERATURELC FIG. 3.-Temperature and pH optima curves for the yellow-green and red in vitro bioluminescence emissions of P. pyralis. The short wavelength side of the yellow-green emission was isolated with a Corning 5-60 filter, and the long wavelength side of the red emission was isolated with two Corning 2-64 filters. Downloaded by guest on September 24, 2021 78 BIOCHEMISTRY: SELIGER AND MCELROY PROC. N A. S.

thorax light up with a constant bright green glow when the insect is resting or walk- ing, or disturbed in captivity. These are extinguished when the insect is in flight. There is in addition a single large ventral organ, completely shielded in a cleft between the thorax and the first abdominal segment. When the insect is in flight, the elytra are extended, the first abdominal segment is flexed, opening the cleft, and a constant bright yellow light is emitted. In a dark field a low-flying Pyro- phorus illuminates the ground below much as a downward-directed landing light of an airplane. When the insect alights or is batted down, the elytra close, closing the ventral cleft. The yellow light goes off, and the green "parking lights" go on. For this reason and because of the position and shape of the anterior thoracic organs, P. plagiophthalamus is also called the "automobile bug." On the basis of paper chromatography, absorbance and fluorescence spectra, and the pH dependence of the relative fluorescence yield, all of the isolated luciferins, including those isolated from the two different organs of P. plagiophthalamus, appear to be identical with Photinus pyralis luciferin, which we have been able to synthesize in the laboratory."' In the in vitro light reaction regardless of the source of luciferin, the spectral distribution of the light emitted by each species corresponded with the in vivo emission measured previously. The data for P. pyralis are shown in Figure 4. Similar studies have been made using P. plagiophthalamus dorsal and ventral organ luciferase. In each case the species enzyme determined the spectral distribution of the emitted light. The spectra in Figure 4 are displaced one above the other in order to show the relative distribu-

P. pyralis

47k3 4925 5087 5250 54;2 5574 5736 5898 6061 6223 6385 6547 6709 6871 7033 7195 WAVELENGTH (ANGSTROMS) FIG. 4.-Emission spectra of P. pyralis. (a) In vivo. (b) In vitro. P. pyralis luciferase + P. pyralis luciferin. (c) In vitro. P. pyralis luciferase + synthetic luciferin. (d) In vitro. P. pyralis luciferase + P. plagiophthalamus dorsal light organ luciferin. (e) In vitro. P. pyralis luciferase + P. plagiophthalamus ventral light organ luciferin. Downloaded by guest on September 24, 2021 VOL. 52, 1964 BIOCHEMISTRY: SELIGER AND MCELROY 79

tions more easily. In the case of P. plagiophthalamus there are slight shifts in the peak positions of the in vivo emissions relative to the in vitro emissions. This is probably due to the fact that the color of the light emitted by the particular in vivo specimens measured was actually slightly different than the average color due to the in vitro enzyme extract obtained from the light organs of approximately 100 different insects. We found, on a subsequent collecting trip where a large number of P. plagiophthalamus was caught, that in this presumably single species we can distinguish, in different insects, three separate colors of in vivo emission from the dorsal organs, namely, green, yellow-green, and lemon-yellow, and three separate colors of in vivo emission from different ventral organs, yellow-green, yellow, and orange. There is apparently no correlation between ventral organ emission color and dorsal organ emission color. In more than 1000 insects examined, both dorsal organs emitted identical colors of light without exception.

1.0

FIG.>5.Effects of pH an metal ontvitrobioluminscenceobP. pogiophthaoamus (dorSol)

Z 6

z .5 w

I- .4 -j

sallihCrauieae+snhtcLcfr(a omlrato.p .-n de n+ .2

4763 4925 5087 5250 5412 5574 5736 5898 6061 6223 6385 6547 6709 6871 7033 7195 WAVELENGTH (ANGSTROMS) FiG. 5.-Effects of pH and metal cations on in vitro bioluminescence of P. plagiophthalamus dorsal light organ luciferase + synthetic luciferin. (a) Normal reaction. pH 7.6; no added Zn + (b) pH 6.0; no added Zn++. (c) pH 7.6; 5.5 X 10-4 M Zn++. Effects of pH and Metal Cations.-The maximum effects of pH and metal cations on the in vitro emissions of P. plagiophthalamus dorsal organ luciferase are shown in Figure 5. Qualitatively, these spectral shifts are in the same direction as the P. pyralis effects, although the magnitudes of the shifts are much smaller. In none of our experiments were we able to observe a red shift in the emission from P. plagiophthalamus ventral organ luciferase. However, as shown in Figure 6, basic solutions and, separately, the addition of metal cations produced a "blue shift," opposite to that observed from both the dorsal organ and from P. pyralis. Discussion.-In view of the chromatographic, absorbance, and fluorescence evidence and since, in all cases tested, the in vitro bioluminescence emission in any particular enzyme extract was the same regardless of the source of luciferin, we are probably correct in assuming that all natural firefly luciferins have the same structure. In all of the firefly species that have been tested, luciferase from any species will Downloaded by guest on September 24, 2021 80 BIOCHEMISTRY: SELIGER AND MCELROY PROC. N. A. S.

1.0

FIG.6.-Effectsof pH and meP. plagiophthala.us (ventral)

ZL.6 a z C .5

H- .4 -LJ

4763 4925 5087 5250 5412 5574 5736 5898 6061 6223 6385 6547 6709 6671 7033 7195 WAVELENGTH (ANGSTROMS) FIG. 6.-Effects of pH and metal cations on in vitro bioluminescence of P. plagiophthalamu-s ventral light organ luciferase + synthetic luciferin. (a) Normal reaction. pH 7.6; no added Zn++or Cd++. (b) pH 9, noadded Zn++or Cd++. (c) pH 7.6; 5.5 X 10-4 M Zn++. (d) pH 7.6; 8 X 10- M Cd+ produce bioluminescence with luciferin from any other species. The species differences in the in vivo spectral emissions and the carrying over of these species differences to the in vitro reactions of enzyme extracts would indicate that the color of the emitted light is due only to the species enzyme. This idea is further strength- ened by the fact that the different enzyme extracts examined behave differently to pH and to the metal cations tested. The presence of two types of emission peaks in the P. pyralis bioluminescence could mean either that two different configurations are possible in the excited enzyme-substrate complex, or that there are two isoenzymes present. On the as- sumption that the pH and temperature optima give some measure of the lability of the enzymes involved, we have tried extensive heat and acid denaturation of P. pyralis enzyme. By the use of selective color filters we can assay the native enzyme at pH 7.6 and at room temperature to determine the ratio of red-light intensity to green-light intensity. If, after a high-temperature incubation or after acid incubation, more of the "green-emitting enzyme" has been denatured than the "red- emitting enzyme," this ratio, again measured at room temperature and at pH 7.6, should be different from that observed for the 100 per cent native enzyme before heat or acid denaturation. The red-to-green intensity ratio remains constant up to 95 per cent enzyme denaturation. Obviously much more information on the nature of the active sites on these enzymes, on the nature of the oxidized product, and on the effects of other luciferin derivatives must be obtained before we can present an explanation for these striking specificities. This work has been supported in part by the Atomic Energy Commission, the National Institutes of Health, and the National Science Foundation. We would like to thank Miss E. Kayser and Mr. W. Biggley for their invaluable technical assistance. Downloaded by guest on September 24, 2021 VOL. 52, 1964 BIOCHEMISTRY: SUSSMAN AND OSBORN 81

1 Seliger, H. H., W. D. McElroy, E. H. White, and G. F. Field, these PROCEEDINGS, 47, 1129 (1961). 2Seliger, H. H., and W. D. McElroy, Radiation Res., Suppl. 2, 528 (1960). 3 Seliger, H. H., J. B. Buck, W. G. Fastie, and W. D. McElroy, J. Gen. Physiol., in press. 4Harvey, E. N., Bioluminescence (New York: Academic Press, 1952). 6 White, E. H., F. McCapra, and G. F. Field, J. Am. Chem. Soc., 85, 337 (1963).

UDP-GALACTOSE POLYSACCHARIDE TRANSFERASE IN THE CELLULAR SLIME MOLD, DICTYOSTELIUM DISCOIDEUM: APPEARANCE AND DISAPPEARANCE OF ACTIVITY DURING CELL DIFFERENTIATION* BY M. SUSSMAN AND M. J. OSBORNt DEPARTMENT OF BIOLOGY, BRANDEIS UNIVERSITY, AND DEPARTMENT OF MICROBIOLOGY, NEW YORK UNIVERSITY SCHOOL OF MEDICINE Communicated by B. L. Horecker, May 8, 1964 Dictyostelium discoideum amoebae, upon entering the stationary growth phase, collect together into organized, multicellular aggregates which are ultimately converted into fruiting bodies that contain at least two differentiated components, spores and stalk cells. A schematic diagram of the morphogenetic sequence ac- companies the data shown in Figure 4. The syntheses of three polysaccharides are keyed to this sequence and their developmental kinetics are governed by the over-all rate of morphogenesis." 2 None is synthesized by mutant strains unable to accomplish the later stages of fruiting body construction. One of the three is an acid mucopolysaccharide, composed of galactose, galactosamine, and galacturonic acid; together these sugars constitute at least 92 per cent of the dry weight of the purest preparations. This polysaccharide is serologically reactive against anti-D. discoideum spore sera and anti-pneumococcal (type VII) serum, and reaction with either antiserum is blocked by galactose and lactose (50% as active at equimolar concentrations), but by no other carbohydrate tested. This component, as measured by quantitative com- plement fixation or by bound nondialyzable galactose or galactosamine, first makes its appearance during the transformation of cell aggregates into migrating pseudoplasmodia. It is synthesized rapidly during actual fruiting body construction and ultimately reaches a peak of 1-2% of the dry weight. These data led us to an examination of the enzymes that might be responsible for the synthesis of the mucopolysaccharide in order to determine if they, too, are controlled by the over-all morphogenetic program. Prior work3-5 has indicated that galactose-containing polysaccharides are synthesized by the following transfer reaction: UDP-Galactose + Acceptor Galactosyl-Acceptor + UDP. It has been possible to demonstrate the presence of a UDP-galactose polysaccharide transferase, catalyzing the above reaction, in cell-free extracts of D. discoideum. The enzyme does indeed appear to be under morphogenetic control. Some of its properties and its developmental kinetics are described below. Downloaded by guest on September 24, 2021