,lollrald of

AND

B:BIOLOGY

E L S E V I E R Journal of Photochemistry and Photobiology B: Biology 34 (1996) 73-77

Chromophore-protein interaction controls the complexity of the phytochrome photocycle

Peter Schmidt, Ulf H. Westphal, Karl Worm, Silvia E. Braslavsky, Wolfgang G~irtner *, Kurt Schaffner

Max-Planck-lnstitut fiir Strahlenchemie, P. O. Box 101365, D-45413 Mtilheim an der Ruhr, Germany

Received 25 September 1995; accepted 13 November 1995

Abstract

A new protocol for the preparation of recombinant phytochromes results in significantly higher yields which, for the first time, have made kinetic studies possible. Flash photolysis with nanosecond laser excitation reveals that, in recombinant and native phytochromes, the decay kinetics of the primary photoproducts I7oo i and the kinetics of the formation of the Pfr form are similar. Phycocyanobilin-containing recombinant phytochrome, however, shows only a monoexponential decay of the I7o o intermediate with a time constant of approximately 90/xs, and a biexponential formation of the Ply form, albeit with time constants (approximately 13 and 100 ms) somewhat shorter than those from native phytochrome. Thus the seemingly small structural modification of the chromophore (substitution of the native vinyl for an ethyl group) has a profound influence on the availability of protein conformational rearrangement pathways. The result is therefore of general interest in chromoprotein dynamics.

Keywords: Phytochrome intermediates; Kinetic analysis; Phycocyanobilin; Phytochromobilin; Recombinant; Pichia pastoris

I. Introduction

The plant photoreceptor phytochrome is ubiquitous in all higher plants, in many ferns and mosses, and in green algae [ 1 ]. It consists of an apoprotein of approximately 125 kDa to which a chromophore (phytochromobilin) is covalently bound (formation of a thio ether bound to a cysteine residue, Fig. 1 ). Phytochrome converts long-wavelength light energy into intracellular signals inducing photomorphogenesis. This physiological function is mediated via a light-induced con- version of the parent state Pr ( r , red absorbing) into the final product Per (fr, far-red absorbing) [3]. The primary reaction of this conversion consists of photoisomerization of the chro- mophore, which is followed by a series of thermally driven conformational changes of protein and chromophore [4-8]. Starting from Pr (/~max = 667 nm), a set of first intermediates of this photoprocess, 17oo 1'2, is formed within several pico- seconds. The I7(x~ 1"2 intermediates are further converted in a complex manner into a second set of intermediates with a low absorption coefficient (Ibli; bl, bleached) on the microsecond time scale [6-9]. The thermally stable final product of the photoreaction, the Pfr state (Amax = 730 nm), is formed with

* Corresponding author.

1011-1344/96/$15.00 © 1996 Elsevier Science S.A. All rights reserved SSDI 1 01 1- 1 3 4 4 ( 9 5 ) 0 7 2 6 9 - 1

two distinct rate constants within several milliseconds (Fig. 2) [8]. So far, this kinetic behaviour has only been observed with phy-A-type phytochrome which is the most abundant species in etiolated plants. De-etiolation drastically reduces the concentration and thus prevents the extraction of material in sufficient amounts for spectroscopic analysis.

The kinetic analysis of two recombinant phytochromes has now revealed that a single modification in the substitution of the chromophore, namely vinyl at C- 18 as in the native phy- A-type phytochrome vs. ethyl in phycocyanobilin-containing phytochrome, dramatically influences the complexity of intermediate formation without affecting the principle of the photoconversion, i.e. the overall Pr ~ Pfr transformation.

2. Materials and methods

2.1. Protein and chromophore preparations

Full length oat phytochrome phy-A cDNA was cloned into the EcoRI site of the yeast vector pHIL-D2 (Invitrogen) which was introduced into the methylotrophic, inducible yeast Pichia pastoris. Transformed yeast cells ( 1 L cell cul- tures) were grown for about 1 day to A6o o = 1-2, and were

74 P. Schmidt et aL / Journal of Photochemistry and Photobiology B: Biology 34 (1996) 73-77

AI01 Ala-Pro- His-- Ser- Cys(321)- His-- Leu- Gin-- Tyr ......

15 1 ~

CO2H CO2H

.N

CO2H CO2H

Fig. 1. Top: structure of the chromophore of phytochrome, phytochromobilin (PqbB, covalently attached to cys321 of oat phytochrome phy-A). Photo- isomerization takes place at the 15,16-double bond. The non-protonated pyrrolenine form of the chromophore is shown (for a discussion of this point, see Ref. [2] ). Bottom: structure of phycocyanobilin (PCB). Both chromophores are shown in the all-Z, syn-, anti-, syn conformation. After binding to the protein, the sole structural difference between the two bilins resides in the substituent at position 18 (vinyl vs. ethyl).

then placed into a minimal medium containing methanol (0.5%) as the sole carbon source for induction. This protocol differs from that of Li and Lagarias [ 11 ] with regard to the yeast employed and to induction. After further growth for 1 day, cells were harvested and lysed via three passages through a French press (1000-700 kPa). The lysis suspension was cleared by ultracentrifugation and the supernatant was con- centrated to 5 ml. Each of the chromophores was added to identical protein preparations at a final concentration of 5-7 /zM yielding approximately 1 nmol of recombinant phyto- chrome, calculated from Em,x = 125 000 M -1 cm -~ [12]. The chromophores phycocyanobilin (PCB) and phytochro- mobilin (PqbB) were isolated according to literature proto- cols [13,14].

2.2. Difference spectra

For the formation of the Pr forms of the reconstituted phy- tochromes, the preparations were irradiated with light from a slide projector bulb which was filtered through an RG9 cut- off filter (thickness, 4 mm) at a distance of approximately 25 cm under saturating conditions (3 min). The Pfr forms were generated on irradiation with light of 665 nm (using an interference filter with a half bandwidth of 7 nm) in the same irradiation regime. Spectra were recorded after each irradia-

h . v P r >

I oo 12oo

m m _ _ ~

- - -> P f r

< < h-v

Fig. 2. Simplified Pr ~ Pfr photocycle of phytochrome (for details on Pr --* Pf~, see Scurlock et al. [9], andon Pfr--* Pr, see Eilfeld and Riidiger [ 10l ). The superscript " i " in the intermediate Ib~ i indicates the presence of at least two intermediates (in this paper, we have not determined the nature of the kinetics of these steps, i.e., whether they are parallel or sequential).

tion using a Shimadzu spectrophotometer (UV-2102 PC) and the difference spectrum was calculated.

2.3. Laser-induced flash photolysis

The flash photolysis system consisted of a DCM dye laser (Aexc = 650 nm) pumped by the frequency-doubled pulse of an Nd:YAG laser (Aexc = 532 nm). The duration of the laser pulse was approximately 11 ns and the excitation energy was approximately 1 mJ cm - 2 (7% conversion for a single flash). For microsecond detection, the monitoring light beam (from a pulsed 150 W Xe arc) was passed through a high intensity monochromator before entering the sample cuvette, and through a second monochromator after the sample. For mil- lisecond detection, light from a tungsten halogen bulb ( 100 W) was used. A double beam detection system served to compensate for irregular lamp intensities. Detection of the time-resolved absorbance changes was accomplished by a photomultiplier (Hamamatsu, R3896) the signal of which was fed into a Tektronix TDS 520A oscilloscope. Further data processing was performed with a DEC-alpha worksta- tion. Further details of the highly sensitive detection system will be published elsewhere [ 15].

The recombinant phytochromes were studied by nanosec- ond flash photolysis (Aexc = 650 nm) at 10 °C using a set-up which provides kinetic traces over a large range of wavelengths after selective red light excitation of Pr. Tran- sient absorbance changes in the microsecond time range (i.e. the 17oo 1'2 decay) were detected at 680 nm (PCB sample) and 695 nm (P~B) . The formation of the Pfr state was probed in the millisecond domain at 720 nm. In each case, 80 single traces were collected and averaged. The time interval between single excitations was at least 30 s, allowing sufficient irra- diation of the sample with far-red light (A >/720 nm) in order to re-convert Pfr into Pr.

3. Results and discussion

In order to investigate the chromophore-protein interac- tions underlying the photochemical process of phytochrome, full length apoprotein of oat phy-A was produced from the

P. Schmidt et al. / Journal of Photochemistry and Photobiology B: Biology 34 (1996) 73-77 7 5

1 2

40

20 t~

O

o e - ra 0 £ O

e~

-20

-40

-60-~

500 600 700 800

Wavelength / nm

=ig. 3. Difference absorption spectra (P,- Pf,) of recombinant phytochromes reconstituted from yeast-expressed apoprotein and PCB (trace 1 ) or P~B (trace 2). The difference spectrum from trace 2 has been scaled by a factor of 1.1 to allow for better comparison of the absorbance maxima. The reconstituted samples were irradiated with light from a 200 W tungsten lamp. filtered through an interference filter (A =665 + 12 nm) or a cut-off filter (A>~730 nm) for 2 min. :;pectra recorded after each irradiation were used for difference spectra (P,- Pf,) calculation.

leterogeneous expression of the encoding cDNA in the nducible yeast P. pastoris. Reconstitution was achieved by ncubation of two aliquots of the apoprotein preparation with

w o different chromophores, i.e., PCB and the native chro- nophore, pqbB. The structural difference between the two :hromophores is confined to the C-18 substituents (ring D, !:ig. 1). This difference may seem negligible at first sight. =Iowever, the substituents differ in both space filling and ,~,lectronic properties, the latter being clearly evident from the difference absorption spectrum (see below). Furthermore, :he experimental results show that the influence of the two Cubstituents on the mechanism of the Pr--> Pfr transformation s dramatically different.

In the reconstitution experiments, Pq~B and PCB were 2raided to apoprotein samples prepared from the same batch :)f cells. The reconstitution took place within a few minutes 2tt ambient temperature and difference spectra of the expected ~hape were obtained (Fig. 3), albeit with about tenfold ~,reater amplitudes than described for the first preparation by Li and Lagarias [ I 1 ]. This calculation is based on similar ,~olumes of transformed yeast cell cultures. An estimation ~eveals that a certain portion of the apoprotein remains in insoluble, non-functional form, but the remaining soluble fraction reconstitutes almost quantitatively. In accordance

with the literature [ 11 ], the peaks of the difference spectra corresponding to the Pr and Pfr forms of the PCB-containing sample are blue shifted by approximately 15 nm relative to the absorbance of native oat phytochrome A (Ar,~x of Pr, 651 nm; Amax of Per, 715 nm; values derived from the difference spectra). The Pq~B-containing sample exhibits values similar to those for native oat phytochrome (An-~x = 663 and 728 nm). It should be noted that the considerably higher yield enabled us to carry out, for the first time, a kinetic study by flash photolysis of the Pr+ hu ~ I700 i ~ Ix i --> per conversion of recombinant phytochromes.

The traces for the I7oo i decay of the recombinant phyto- chromes (Fig. 4) show a significantly different kinetic behav- iour for the PCB- and PqbB-carrying chromoproteins. For the PCB protein, the experimental traces can be fitted satisfac- torily by a single exponential with a lifetime of 90/~s (see residuals), whereas the PqbB-containing protein exhibits biexponential kinetics similar to that observed for native phytochrome requiring lifetimes of 6 and 95/zs for a good fit. The rise kinetics of Pfr reveal biexponential behaviour for both recombinant chromoproteins. Similar to the Pfr forma- tion of wild-type phytochrome (time constants of approxi- mately 40 and 400 ms [8] ) , the recombinant PqbB phytochrome exhibits time constants of about 40 and 600 ms.

76 P. Schmidt et al. / Journal of Photochemistry and Photobiology B: Biology 34 (1996) 73-77

"7.

8 c

,<

'7 o

x

8

o . o , ,<

1.2

0.8

0.4

0.0

0.05 I 0.00 -0.05

-100

3.0

2.0

1.0

0.0

0.2 I 0.0

-0.2 -100

I l I

a I I I I i i i i

I ' I ' I ' I •

0 100 200 300

f f I I

I , I I I i l =

I , I , I , I ,

0 100 200 300

400

1 400

between the chromophore and the protein, which in the case of PCB are sufficiently different to result in a loss of one of the two reaction channels. The modified chromophore struc- ture is also evident in the accelerated formation of Per, again due to the less intimate interactions.

4. Conclusions

For the first time, a kinetic event in the photocycle of native phytochrome can be ascribed to interaction between the bind- ing domain and a particular chromophore substituent. It is noteworthy that this substituent is attached to ring D which is primarily reoriented by the cis ---> trans isomerization of the 15,16-double bond, and that the options for subsequent pro- tein conformational adaptation steps are affected by the seem- ingly small structural modification. This finding is in remarkable contrast to the wide experience with other chro- moproteins, in particular in the field of rhodopsin and bac- teriorhodopsin [ 16,17 ], where in most cases relatively small changes in the prosthetic group have little influence on the reaction kinetics provided that the molecular site modified does not directly participate in the primary process.

Time [~s] Fig. 4. Absorbance decay xlneucs at iv ~.. ui • ,,~u-~.aa,),m~ ~ tup) and P~B- carrying (bottom) recombinant phytochrome in the microsecond domain. Samples were prepared in 100 mM phosphate buffer (pH 8.0) containing 2 mM dithiothreitol. The data were fitted with a monoexponential (top) and a biexponential (bottom) function. The residuals between the data points and the fitted curves are shown at the bottom of each trace.

In the PCB protein, Pfr formation is more rapid and occurs with constants of 13 and 100 ms. Repeated investigations of recombinant samples during a time period of several days revealed a change in the kinetics such that (probably due to aggregation with contaminating proteins or other compo- nents) the processes in the millisecond time range became slower with time. All the values given here were proven to be identical for measurements performed with independently prepared, fresh samples. This result supports our formerly proposed reaction scheme based on two microsecond and two millisecond time constants, An even further complicated reaction scheme presented by Zhang et al. [6] , in which the two parallel pathways converge to a common intermediate before the appearance of Pf, demands further investigation.

The observation that, for the PCB-containing phyto- chrome, one of the two kinetic components of the 17o 0 decay is missing may be due to the fact that this process is much faster as a result of the greater flexibility of the chromophore. In particular, the less intimate interaction with the ethyl group (instead of the vinyl group) may make the detachment of the chromophore from its environment more rapid, escaping detection in the microsecond range. Alternatively, the parallel pathways observed in native phytochrome and in the recom- binant derivative carrying the native chromophore can be understood as originating from very specific interactions

Acknowledgements

We are indebted to Professor P.H. Quail, Albany, CA, for providing oat phytochrome cDNA, and to Professor S. Beale, Brown University, Providence, Rhode Island, for a batch of Porphyridium cruentum, which served as a reference for opti- mizing the growth conditions of a culture obtained from the Algae Culture Collection, University of Grttingen, Germany. We thank T. Huestege and G. Koc for able assistance in the preparation and purification of the two chromophores.

References

[ 1 ] M. Furuya, Phytochromes: their molecular species, gene families, and functions, Annu. Rev. Plant Physiol. Plant. MoL Biol., 44 (1993) 617- 645.

[2] J. Matysik, P. Hildebrandt, W. Schlamann, S.E. Braslavsky and K. Schaffner, Fourier-transform resonance Raman spectroscopy of intermediates of the phytochrome photocycle, Biochemistry, 34 (1995) 10 497-10 507.

[3] P.H. Quail, Phytochrome: a light-activated molecular switch that regulates plant gene expression, Annu. Rev. Genet., 25 ( 1991 ) 389- 409.

[4] K. Schaffner, S.E. Braslavsky and A.R. Holzwarth, Photophysics and photochemistry of phytochrome, in D.H. Volman, G.S. Hammond and K. Gollnick (eds.), Advances in Photochemistry, Vol. ! 5, Wiley, New York, 1990, pp. 229-277.

[5] K. Schaffner, S.E. Braslavsky and A.R. Holzwarth, Protein environment, photophysics and photochemistry of prosthetic biliprotein chromophores, in H.-J. Schneider and H. D(irr (eds.), Frontiers in Supramolecular Organic Chemistry and Photochemistry, VCH Verlagsges., Weinheim, 1991, pp. 421-452.

P. Schmidt et al. / Journal of Photochemistry and Photobiology B: Biology 34 (1996) 73-77 77

[6] C.-F. Zhang, D.L. Farrens, S. BjOrling, P.-S. Song and D.S. Kliger, Time-resolved absorption studies of native etiolated oat phytochrome, J. Am. Chem. Soc., 114 (1992) 4569-4580.

[7] H. Kandori, K. Yoshihara and S. Tokutomi, Primary process of phytochrome--initial step of photomorphogenesis in green plants, J. Am. Chem. Soc., 114 (1992) 10 958-10 959.

[8] R.D. Scurlock, S.E. Braslavsky and K. Schaffner, A phytochrome study using two-laser/two color flash photolysis: 1700 is a mandatory intermediate in the Pr~Pfr phototransformation, Photochem. Photobiol., 57 (1993) 690-695.

[9] R.D. Scurlock, C.H. Evans, S.E. Braslavsky and K. Schaffner, A phytochrome phototransformation study using two-laser/two-color flash photolysis: analysis of the decay mechanism of 1700, Photochem. Photobiol., 58 (1993) 106-115.

[10] P. Eilfeid and W. Riidiger, Absorption spectra of phytochrome intermediates, Z. Naturforsch., 40c (1985) 109-114.

[ 11 ] L. Li and J.C. Lagarias, Phytochrome assembly, Z Biol. Chem., 267 (1992) 19 204-19 210.

[ 12] J.C. Lagarias, J.M. Kelly, K.L. Cyr and J.W.O. Smith, Comparative photochemical analysis of highly purified 124 kilodalton oat and rye phytochromes in vitro, Photochem. Photobiol., 46 (1987) 5-13.

[13] W. Kufer and H. Scheer, Studies on plant bile pigments, VII, preparation and characterization of phycobiliproteins with chromophores chemically modified by reduction, Hoppe-Seyler's Z. Physiol. Chem., 360 (1979) 935-956.

[ 14] J. Comejo, S.I. Beale, M.J. Terry and J.C. Lagarias, Phytochrome assembly. The structure and biological activity of 2(R),3(E)- phytochromobilin derived from phycobiliproteins, J. Biol. Chem., 267 (1992) 14 790-14 796.

[ 15] P. Schmidt et al., in preparation. [ 16] R.K. Crouch, Studies of rhodopsin and bacteriorhodopsin using

modified retinals, Photochem. Photobiol., 44 (1986) 803-807. [ 17] M. Ottolenghi and M. Sheves, Synthetic retinals as probes for the

binding site and photoreactions in rhodopsin, J. Membr. Biol., 112 (1989) 193-212.