9852
J. Phys. Chem. 1993, 97, 9852-9862
Monomeric C-Phycocyanin at Room Temperature and 77 K: Resolution of the Absorption and Fluorescence Spectra of the Individual Chromophores and the Energy-Transfer Rate Constants Martin P. Debreczeny and Kenneth Sauer. Chemical Biodynamics Division, Lawrence Berkeley Laboratory, and Department of Chemistry, University of California at Berkeley, Berkeley, California 94720
Jianhui Zhout and Donald A. Bryant Department of Molecular and Cell Biology and the Center for Biomolecular Structure and Function, Pennsylvania State University, University Park, Pennsylvania I6802 Received: April 23, 1993; In Final Form: June 29, I993@
At both room temperature (RT) and 77 K, the absorption and fluorescence spectra of the three individual chromophore types ((Y84, 884, and B l s 5 ) found in monomeric C-phycocyanin ( ( u ~ @isolated ~ ) , from the cyanobacterium Synechococcus sp. PCC 7002, were resolved along with the rates of energy transfer between the chromophores. The cpcB/CI55SmutantI whose PC is missing the fils5 chromophore, was useful in effecting this resolution. At RT, the single broad peak in the visible region of the absorption spectrum of (apcflpc) was resolved into its three-component spectra by comparing the steady-state absorption spectra of the isolated wild-type a subunit of PC (apc)(containing only the a 8 4 chromophore) with those of the monomeric PCs isolated from the mutant strain (aPC@*) and the wild-type strain (apcBpc).At 77 K, the visible region of the splits into two peaks. This partial resolution at 77 K of the chromophore absorption spectrum of (aPC@pc) spectra of (apcPpc)when compared with the 77 K absorption spectra of awl ,F, and ( P o * )provided a confirmation of our R T assignments of the chromophore absorption spectra. The individual fluorescence spectra of the chromophores and the rate constants for energy transfer between them in (apcPpc) were resolved by modeling the time-resolved fluorescence spectra of (containing the @1ss and 8 8 4 chromophores) and (apc@*). As with absorption, lowering the temperature to 77 K helped resolve the fluorescence spectra of the individual chromophores because the spectra were narrowed.
I. Introduction A crucial point in understanding the energy-transferprocesses occurring within a multichromophore system is the determination of the spectral properties contributed by each of the individual chromophores. If the chromophores are only weakly coupled, it may be sufficient to describe the absorption spectrum in terms of the sum of the spectra of the isolated parts. FBrster’s theory of energy transfer for the weakly coupled case predicts that the rate of energy transfer depends on the overlap of the donor chromophore’s emission spectrum with the acceptor chromophore’s absorption spectrum.’ The absorption spectra of chromophores that are more strongly coupled due to their proximity in space and energy may not behave additively. When the interaction energy between chromophores is larger than the individual chromophore absorption bandwidths, excitation is delocalized among chromophores, and a splitting of the energy levels of the individual chromophores (known as exciton splitting) is observed. It is still essential to know the properties of the chromophores separately in order to predict and understand the properties of the whole. Photosynthetic light-harvesting complexes almost certainly contain examples of chromophores involved in strong and weak coupling as well as intermediate cases. However, testing the applicability of energy-transfer theories to these protein-based chromophorecomplexes is complicated by the small amount of structural data available. Even when crystal structures are available, the determinationof the spectral properties of individual chromophores to model the propertiesof the whole is difficult for complexes containing many chromophores. As a result, much work in the area of photosynthetic energy transfer has relied on t Current address: Department of Plant Biology, University of Califomia, Berkeley, CA 94720. *Abstract published in Advance ACS Absrrucrs, September 1, 1993.
0022-3654/93/2097-9852S04.00/0
such techniques as (i) fitting a single broad absorption band to multiple Gaussians in an attempt to extract component spectra and (ii) using lattice models of chromophores in random orientations to describe antenna arrays. It is possible that such models can provide information about the overall patterns of energy migration within the whole antenna, but certainly to understand energy transfer at the molecular level more detailed structural and spectral information must be used. In this respect, the phycobilisome (PBS), a photosynthetic lightharvesting complex found in cyanobacteria, is well suited for studies of energy transfer. From electron microscopy, the PBS can be seen to consist of two regions: a core of cylindrical elements from which radiate several rods. The chromophore composition and structure of the rods, and to a lesser extent the core, are well ~haracterized.2.~The chromoprotein, C-phycocyanin (PC), a major component of the PBS rods, has been extensively studied for its light-absorbingand energy-transferringabilities. Crystal structures of PC from several organisms are a~ailable.~.~ Nonetheless, the spectral propertiesof the three different chromophore types found in this protein arestill uncertain. The three covalently bound open-chain tetrapyrrole chromophores, labeled a84, ,884, and 8155 according to the protein subunit ((YE or BE) and residues to which they are attached, are chemically identical. As the high-resolution crystal structures of this protein show,4 however, each chromophore is held in a unique protein binding pocket, which makes each chromophore spectrally unique. The visible regions of the absorption and fluorescence spectra of monomeric (aw,8w)and hexameric (aEOE)6PC at room temperature (RT) consist of a single broad feature, indicating that the spectra of the three chromophore types are strongly overlapping. Further complicatingthe spectral resolution of the chromophoresare the changes in absorption and fluorescence as a functionof the protein aggregationstate. The protein assembles @ 1993 American Chemical Society
Spectra of Monomeric C-Phycocyanin
The Journal of Physical Chemistry, Vol. 97, No. 38, 1993 9853
into the phycobilisome in the hexameric aggregation state in association with “linker” protein^.^.^ PC has been isolated in the monomeric, trimeric, and hexameric forms, and it has generally been observed that the fluorescence and absorptionmaxima shift to the red with increasing aggregation state. Additionally, the CY= and j3PC subunits of the protein, holding one and two chromophores, respectively, have been denatured, separated, renatured, and studied spectroscopically.”16 The isolated and renatured aPCsubunit has been used to study the spectroscopic properties of the a 8 4 ~hromophore.~J2J7The fact that the CYPCand I6pc absorption spectra add to give nearly the same spectrum as the monomer CY^/^^) encourages such analysis.9.12 The two chromophores on are not so easily resolved. Mimuro et al. have modeled the fluorescence polarization spectrum of BPc to try to resolve the 884 and 8155 chromophore absorption spectra.12 Siebzehnriibl er 41. used a mercurial compound that binds specifically to the j 3 1 cysteine ~ residue near the 884 chromophore to modify its properties and confirm Mimuro’s a~signment.’~ Polarized absorption measurements of partially oriented PC18 and of single crystals of PCl9 have also led to the assignment of the j3155 chromophore as the short-wavelength absorber and the 884 chromophoreas the longwavelength absorber based on the X-ray crystal structure? These methods were adequate to establish the energetic ordering of the j3155 and 884 chromophores, but the detailed absorption spectral shapes for the two chromophores are less certain. Attempts at resolving the fluorescence emission spectra of the chromophoreshave been made by Sauer et alamfor the purpose of calculating Fijrster rate constants among the chromophores and by Holzwarth et a1.21 for the purpose of modeling the timeresolved fluorescence of PC. In both cases, it was assumed that the two chromophores have the same fluorescence line shape and Stokes shift as the CY84 chromophore. Sauer et al.20used the relative absorption strengths (estimated by a deconvolution procedure which assumed the 0 1 5 5 and 894 line shapes to be the same as that of the a 8 4 chromophore) to approximate the relative fluorescence quantum yields of the chromophores. In this paper, we focus on resolving the absorption and fluorescence spectra of the three chromophoretypes in ( ~ ~ ~ j and the rate constantsfor energy transfer between them. Included among the techniques we use to achieve this resolution are sitedirected mutagenesis, time-resolved fluorescence spectroscopy, and low-temperature (77 K) absorption and fluorescence spectroscopy. We performed spectroscopic studies on (aPCj3PC) and onPC isolated froma mutant (cpcB/Cl55S)in which thecysteine at the j31ss position has been substituted with a serine.22 This was accomplished by site-directed alterations in the cpcB gene. The mutant strain PR6235 (cpcB/ClHS)was constructed by deleting the chromosomal copies of thecpcBand cpcA genes by interposon mutagenesis with aph2 genes of P n P a n d transcomplementation with the biphasic shuttle vector pAQE19 which carries the wildtype cpcA gene and the mutant cpcB gene. PC monomers isolated from this mutant strain PR6235 are referred to in the rest of the paper as (aPCj3’). The result of the mutation is that the &S chromophore cannot bind covalently and does not appear to be associated noncovalently with the j3 subunit. The absence of the j3155 chromophore in (aPCj3*) has allowed us to use absorption difference methods to resolve the individual chromophore spectra. Due to the rapid and efficient energy transfer among the chromophores of ( C Y K ~ ~steady-state X), emission experimentswere not sufficient to resolve the fluorescence spectra of the three chromophore types. Time-resolved emission spectra were recorded with up to 10-ps time resolution (after deconvolution) to observe emission of the chromophores prior to equilibration by energy transfer. Because the spectra of the phycobiliproteins are narrowed by lowering the temperature, we performed some additional experimentsat 77 K. The absorption spectrum of BPc splits into two bands at 77 K. The time-resolved emission of j3PC
at 77 K also shows two distinct peaks. These measurements help to confirm the spectral assignments based on room temperature studies.
II. Materials and Methods G ~ w t b C ~ d t i ~ ~ a n d P C I ~ ~Thewild-typeand htion, mutant strains were grown as described by Gindt et ~ 1 . ~The 3 mutant strain PR6235 (cpcB/Cl55S) was grown in a medium that contained kanamycin (100 mg L-1) and ampicillin (2 mg L-1). PC was isolated using a preparation procedure slightly modified from that used for crystallography,” as described below. All buffer solutions contained 1 mM (Na) azide as a preservative. Cells, harvested by centrifugation at 3500 X g, were suspended in an equal volume of 5 mM (Na) phosphate at pH 7.0,25 OC. The cells were homogenized, and egg-white lysozyme and EDTA disodium salt were added to final concentrations of 1 mg mL-1 and 10 mM, respectively. This mixture was stirred for 1 h at 25 OC. Cells were then ruptured by passing them several times through a French pressure cell at 140 MPa. The broken cell suspension was centrifuged at 35 000 X gfor 30 min at 4 OC. The blue supernatant was ultracentrifuged at 300 000 X gfor 60 min at 4 OC. The supernatant was dialyzed overnight into 5 mM (Na) phosphate buffer, pH 7, at 4 OC and then loaded onto a DEAE-cellulose (Whatman DE52) anion-exchangecolumn that had been preequilibrated in the same buffer. The column was washed extensively with 5 mM (Na) phosphate, pH 7, and the proteins were eluted with a linear gradient of 5-250 mM (Na) phosphate, pH 7.0. Wild-type PC eluted from the column at around 100 mM (Na) phosphate. Allophycocyanin, the other main phycobiliprotein component isolated from this organism, eluted primarily between 150 and 200 mM (Na) phosphate. Separation of the a and /3 Subunits. High-performance liquid chromatography (HPLC) was used to separate the CY and 8 subunits of PC as described by Swanson and Glazer16with slight modifications. A preparative scale column (17 mm X 30 cm, Waters, Delta Pack C18-300 A) was used with a flow rate of 8 mL min-1. The elution profile was as follows: 0-2 min, 55% buffer A-45% buffer B; 2-37 min, linear gradient to 30% buffer A-70% buffer B. Buffer A contained water with 0.1% trifluo3roacetic ~ ) acid (TFA). Buffer B contained 2:l acetonitri1e:Z propanol with 0.1% TFA. For preparative scale runs, typically 10 mg of protein in 0.5 mL of 50 mM (Na) phosphate, pH 7, was mixed with an equal volume of 9 M urea, pH 3/HC1. For analytical runs (from which elution times and relative absorption areas were calculated), typically 0.5 mg of protein in 0.2 mL of 50 mM (Na) phosphate, pH 7, was mixed with 1.0 mL of 9 M urea, pH 3/HC1. This lower concentration of protein and higher concentration of urea were used in the analytical runs to avoid the partial precipitation of the j3 subunit that was observed in the preparative scale runs. Renaturation of the subunits following the HPLC separation was performed as described by Fairchild et al.25 Separation of the subunits of PC was also performed by cationexchange chromatography as described by Glazer and Fang.8 The samples used in the experiments described in this paper were separated using the cation-exchange procedure. The CY^ samples described throughout this paper were separated using the HPLC procedure. The absorption spectra of the samples isolated by these two different techniques were indistinguishable in the visible region. Determination of Protein Aggregation State. Gel filtration (column dimensions: 50 cm X 1.5 cm) through a size exclusion medium (Sephadex G-100, Pharmacia) was used to determine the aggregation state of (aPCBPC)and (aPC@*)and of the isolated CYPCand j3PC subunits. The filtration medium was pre-equilibrated in the same buffers used for the protein solutions: 5 mM (Na) phosphate, pH 7, in the case of the (YE and BPCand 1 M KSCN, 50mM (Na) phosphate, pH 7, in thecaseof and (aKj3*). The solution to be filtered was approximately 1.5 mL in volume
9854
The Journal of Physical Chemistry, Vol. 97, No. 38, 1993
and contained vitamin B- 12 (1 -35kDa, 0.1 mg) and the following proteins: equinemyoglobin (17 kDa, 0.5 mg), chickenovalbumin (44 kDa, 1 mg), bovine gamma globulin (158 kDa, 1 mg), and thyroglobulin (670 kDa, 1 mg) (Gel Filtration Standards, BioRad), in addition to 0.5 mg of the protein sample whose aggregation state was to be determined. The eluant fractions were characterized by absorption spectroscopy and SDS-PAGE. Steady-State Measurements. Absorption spectroscopy was performed on an Aviv 14DS UV-vis NIR spectrophotometer (Aviv, Inc., Lakewood, NJ). Steady-state fluorescence was measured on a Spex Fluorolog fluorimeter (Spex Industries, Edison, NJ). Samples used for fluorescencemeasurements were diluted to have a peak absorbance of less than 0.1 in a 1-cm cuvette. Samples were excited with an approximately 4-nm bandwidth of light, and the emission bandwidth was limited to
