Photochemistry and Photobiology: Simultaneous time resolution of the emission spectra of fluorescent proteins and zooxanthellar chlorophyll in reef-building corals[para][dagger]

FindArticles > Photochemistry and Photobiology > May 2003 > Article > Print friendly

Simultaneous time resolution of the emission spectra of fluorescent proteins and zooxanthellar chlorophyll in reef-building corals[para][dagger]

Gilmore, Adam M

ABSTRACT

Light is absorbed by photosynthetic algal symbionts (i.e. zooxanthellae) and by chromophoric fluorescent proteins (FP) in reef-building coral tissue. We used a streak-camera spectrograph equipped with a pulsed, blue laser diode (50 ps, 405 nm) to simultaneously resolve the fluorescence spectra and kinetics for both the FP and the zooxanthellae. Shallow water (

Abbreviations: CCD, charged coupled device; chl, chlorophyll; DCI, double-convolution integral; FP, fluorescent protein; FRET, Forster resonance energy transfer; LSGRG, large-scale general reduced gradient; PSII, photosystem II; [alpha]([tau]), pre-exponential amplitude factor as a function of the fluorescence lifetime.

INTRODUCTION

The sun-exposed tissues of reef-building corals may contain both fluorescent proteins (FP), which are related to the well-known green FP of Aequorea victoria, and photosynthetic algal symbionts known as zooxanthellae (1-7). Both the FP and the chlorophyll (chl)-containing algae may absorb visible light and consequently reemit the absorbed light as fluorescence at longer wavelengths. The FP and zooxanthellae density distribution varies considerably among coral genera, coral species and environmental circumstances (4,8-10). Recent concerns over rising global temperatures of coastal waters are associated with observations of environmentally stressed corals that lose their natural coloration because of changes in the content and composition of both zooxanthellae and FP (11-13). This loss of coloration is referred to as "coral bleaching" and is primarily associated with expulsion of the zooxanthellae from the coral tissue along with concomitant decomposition of the chromatophoric FP. Coral bleaching is believed to be closely associated with the deterioration of the physiological state of the zooxanthellae that is itself reflected in their diminished ability to use absorbed light for photosynthesis (8,14,15). The latter is of primary physiological concern because algal photosynthesis provides the primary source of energy for the host coral animal.

The photosynthetic efficiency of the zooxanthellae is routinely measured with techniques involving chl a fluorescence (8,14-17), which provides a nondestructive measure of the primary photochemical efficiency of photosystem II (PSII). A recent report by Salih et al. (8) showed a positive correlation between both FP content and color variation and PSII efficiency of the zooxanthellae under excess light stress. It was shown that photoinhibition of PSII in zooxanthellae due to high light was reduced in highly fluorescent color morphs with dense FP tissue populations compared with zooxanthellae from morphs of the same species lacking FP. These researchers proposed that the primary photoprotective mechanism of the FP for the zooxanthellae and vulnerable coral tissues involved light screening, energy dissipation via fluorescence and scattering of excess visible and UV light (8,18,19). The FP bodies in many cases are spatially arranged in the tissue and key organelles of the animal in such a way as to effectively filter or reflect excess light away from the symbiotic algae.

Studies have also proposed that in contrast to light screening and reflection, in deep waters the FP may act as accessory light-harvesting pigment bodies for the zooxanthellae (4,20,21). This hypothesis was recently discussed in the context of empirical observations from isolated protein and pigment systems that light reemitted by FP exhibits spectral overlap with, and thus exhibits possible energetic favorability for absorption by, the chl and carotenoid molecules in the algae's light-harvesting apparatus (4). The light-harvesting hypothesis further involves the possibility that FP may absorb light in the UV region, which is not absorbed readily by chl, and then reemit light at a longer wavelength (i.e. the emission is Stokes-shifted [22]) that is capable of being harvested by the endosymbiotic algae. In this manner the FP would function to expand the spectral range of photosynthetically active radiation from the normal range of 400-700 nm to shorter wavelengths.

In almost all cases FP comprise heterogeneous pools (8,10), normally of blue, blue-green and green emitters as well as of yellow and red emitters. It was considered that existence of FP mixtures with varying absorption and emission energy levels may function to transfer energy downhill within an FP array by a radiative energy transfer mechanism from bluer to greener or redder emitting-absorbing forms (8,19,23). This radiant energy transfer cascade was hypothesized to facilitate the conversion of damaging high-energy UV wavelengths (

In this study we simultaneously acquired and analyzed picosecond time-resolved fluorescence emission spectra of the FP and zooxanthellae in living, nonstressed coral tissue, including specimens collected from both Okinawa, Japan, and Sydney, Australia. The data were collected with a streak-camera spectrograph that simultaneously records both the time and wavelength dependence of the fluorescence emission, Excitation by a blue laser pulse simultaneously excited both FP and chl in the zooxanthellae photosystems. The data were analyzed using the double-convolution integral (DCI) method (24), which facilitates high-resolution time- and spectral-resolved FRET simulations because it substantially reduces the number of free model parameters compared with conventional decay-associated spectral analysis techniques. The results are discussed in light of the two main observations, namely, (1) the inefficient transfer of energy from the FP to the zooxanthellae chl; and (2) the evidence provided for FRET within the FP populations of all samples.

MATERIALS AND METHODS

Sample collection, identification, storage and preparation. Samples of Plesiastrea versipora (blue and green color morphs) were collected from 5 to 9 m depth at Port Jackson (latitude 33[degrees]52'S), New South Wales, Australia, and were subsequently maintained in flow-through seawater aquaria in Sydney, New South Wales, Australia. Small samples (2-4 cm diameter) were shipped in less than 18 h from Sydney to either Canberra, Australian Capital Territories, Australia, or Nagoya, Japan, under room temperature and low light in tightly closed 50 mL plastic containers filled with seawater; this mode of transportation was previously found to be well tolerated by this species. Samples of P. versipora transported from Sydney to Canberra were analyzed within 24 h, during which they were stored in aerated natural seawater at room temperature in dim fluorescent room light.

Coral specimens were collected under a special permit from the Okinawa Prefectural Fisheries issued to R.V.W. in 2001. Small fragments (5-10 cm) of Acropora nasuta and A. digitifera were collected from 1.5 m depth at Cape Maeda (Okinawa, Japan). Each fragment was placed inside a small plastic bag underwater and then the bags were placed inside a 3 L plastic container with a tight-fitting lid. The corals were shipped to Nagoya within several hours after collection using the "dry method" described by Carlson (25). Specimens of A. nasuta and A. digitifera arrived at Nagoya, Japan, by airfreight on the same day from Okinawa, Japan. The specimens were immediately unpacked and placed in a skim-filtered, aerated, room temperature (minimum 20[degrees]C-maximum 25[degrees]C) aquarium filled with 18 L of artificial seawater. The aquarium was supplemented with commercially available purple photosynthetic bacteria cultures to reduce toxic effects of nitrogen (ammonia waste from the corals). The aquarium was illuminated with fluorescent lamps (

Contour mapping of excitation-emission fluorescence spectra. Contour maps of the instrument-corrected emission versus excitation spectra were obtained with an SLM 8100 Spectrofluorimeter (Spectronic Instruments, Rochester, NY) in photon-counting mode (26). Sample specimens were cut with bone-cutters (0.5-1 cm^sup 2^ front-surface area) and submersed in natural seawater at room temperature in a quartz cuvette that was positioned in a front-surface cell-holder in the light-tight sample compartment. Second-order emissions were blocked from the excitation monochromator by a filter (Schott LS-500) that prohibited all light >500 nm from illuminating the sample. Contour maps focusing on the FP emission were constructed by scanning the excitation monochromator (4 mm, 2 nm/mm slit) from 350 to 450 nm in 1.5 nm increments; the emission monochromator (2 mm slit) was scanned from 450 to 700 nm at a rate of 3 nm/s corresponding to an integration rate of 0.33 s/nm. Contour maps focusing on the zooxanthellae emission were constructed by scanning the excitation monochromator (16 mm slit) from 350 to 450 nm in 2 nm increments; the emission monochromator (4 mm slit resolution) was scanned from 650 to 750 nm at an integration rate of 0.33 s/nm. The contour maps were exported using the SLM software in ASCII format and imported into Microsoft Excel 97 wherein they were normalized to the peak emission value. Two integral profiles were constructed as functions of wavelength for each map corresponding to the excitation and emission axes and were normalized to their maximal intensities.

Fluorescence lifetime spectral image acquisition. Time-resolved fluorescence emission spectra were collected with a streak-camera spectrograph equipped with a blue laser diode emitting 50 ps pulses at 405 nm (Hamamatsu 4334, Hamamatsu Photonics Inc., Hamamatsu, Japan). The sample compartment consisted of a positionable x, y, z stage fitted to accept the coral samples in an otherwise light-tight compartment. The coral specimens were immersed in seawater at room temperature in a 1 by 1 cm quartz cuvette (or low-UV fluorescence plastic cuvette) in a front-surface configuration to receive the laser excitation. Diode emission was filtered before the sample by an interference filter (409 nm, 10 nm bandwidths, 90% transmittance). The total spectral region of the sample's fluorescence emission was controlled by adjusting the spectrograph's (Chromex 2501-S) central wavelength while maintaining the emission grating properties (grating = 50 grooves/mm, blaze = 600 nm and slit = 30 [mu].m). Separate images were collected to focus on either the FP emissions (central wavelength, between 500 and 545 nm) or both the FP emission and algal chl emission (central wavelength, 610 nm). In both cases sample emission was filtered with long-pass filters to exclude light below 420 nm, including the laser excitation. Sample data were collected in photon-counting mode with a 5 ns time window until a total of 25 000 or 125 000 excitation shots accumulated (at a rate of approximately 32.4 Hz) for the FP-only emission or FP plus zooxanthellar emission, respectively. The scattered excitation light profile of the laser pulse was collected (10000 shots) with the spectrograph center wavelength at 500 nm after placing both 1% and 2% neutral density filters between the sample and the spectrograph and removing the 420 nm long-pass filter. The 640 (wavelength axis) by 480 (kinetic axis) pixel charged coupled device (CCD) image was exported from the Hamamatsu software in ASCII format and imported into Excel 97. The pixels were binned and integrated at a rate of 5 pixels per time channel and 3 pixels per wavelength channel and subjected to a further three-channel adjacent averaging of the spectral profiles. The chosen spectral region of interest was restricted to 100 time channels ([lambda] = 1.11 nm/ch) and 160 wavelength channels (t = 31.25 ps/ch) for global analysis using the DCI method, as described below.

Global analysis of the time-resolved emission spectra and FRET using the DCI method. The DCI method (27) constituted the foundation for the spectral kinetic models used to simulate the time-resolved spectra and FRET. The DCI method centers around two assumptions, namely, (1) at any point in time after an excitation event the fluorescence emission spectrum of any compound can be simulated assuming it is composed of a sum of a finite number of Gaussian spectral components; and (2) only the amplitude and not the spectral center or width of the component changes as a function of time. In the present analyses, the decay kinetics for each Gaussian spectral band were simulated using three to four Gaussian kinetic distribution modes, with either positive or negative amplitudes, to represent the integral distribution of the pre-exponential amplitude factor as a function of the fluorescence lifetime (([alpha]([tau]);). The integral of ([alpha]([tau]); was then used to calculate the fluorescence intensity as a function of time,

F(t)^sub i^= [integral operator][infinity]^sub 0^ [alpha]([tau])^sub i^-exp(-[t/[tau]])d[tau], (1)

for each spectral band^sub i^...n, given the respective assumption and physical constraints that both t and [tau] are > or =0 and that F(t)^sub i^ = 0 when t

I(t)^sub i^ = F(t)^sub i^ [x in circle] L(t), (2)

which represents the observed intensity of the Gaussian spectral band as a function of time. The fluorescence intensity as a function of both time and wavelength was obtained by convolving I(t)^sub i^ with a Gaussian spectral band function that predicts the intensity as a function of wavelength, I([lambda])^sub i^, to yield

I(t,[lambda])^sub i^ = I(t)^sub i^ [x in circle] I([lambda])^sub i^, (3)

the DCI. Practically, the double convolution is achieved for each Gaussian spectral band^sub i^...n by substituting the intensity calculated at time - t for the central amplitude factor in a normalized Gaussian spectral band equation (27) centered at a given wavelength with a bandwidth that is equal to the standard deviation of the wavelength distribution. Additional fitting parameters for each image simulation included (1) a constant offset applied to all time and wavelength coordinate channels; and (2) a time-axis shift factor to account for real time-axis differences between the laser intensity as a function of time, L(t), and the data f(i)^sub i^ because the profiles were collected as separate images.

To facilitate FRET-related kinetic estimations pertaining to energy transfers among the FP or zooxanthellae, spectral components corresponding to one or more Gaussian spectral bands were modeled by linking the kinetic distribution lifetime and width parameters for their kinetic modes while independently adjusting the kinetic amplitude and the band spectral center and width parameters. The method thus separates component spectra based on differences between their rise and decay kinetics. To specifically model spectral components corresponding to FRET donor and acceptor species, separate subsets of Gaussian spectral bands were globally linked such that donor species were assigned kinetic amplitudes corresponding to a primary decay process that was kinetically reciprocal in time and amplitude to the rise of the acceptor species. In this manner the kinetic distribution parameters corresponding to a coupled differential rate equation describing the respective decrease in the excited-state concentration of the donor and the increase and then decrease of the excited-stale concentration of the acceptor species were empirically and indirectly simulated. The empirical simulation was viewed as more appropriate than a direct "analytical" linking of the DCI parameters to a coupled equation. This is because the absorbance spectral shapes and absolute extinction coefficients of the donor and acceptor species, needed to estimate the Forster overlap integral, were unknown because of the nature of the sample's surface properties.

The spectral kinetic model was robustly fit (27,28) using the robust statistical norm and curve-fitting (L^sub 1^) method by minimizing the sum of absolute deviations, where the absolute deviation for a given time-wavelength channel coordinate was L^sub i^ = |D^sub i^ - M^sub i^|, where D^sub i^ is the datum and M^sub i^ is the model prediction of the datum. The global analysis program was written for use in Microsoft Windows NT4.0 (32 bit) and Excel 97, with Visual Basic for Applications97 and uses a large-scale general reduced gradient (LSGRG) minimization engine developed by Frontline Systems Inc. (Incline Village, NV). The LSGRG engine is capable of handling 4000 parameters and 4000 constraints and uses a sparse matrix representation. The LSGRG engine is capable of solving, in Excel 97 with a Pentium III computer with 500 Mb RAM and an 800 MHz processor, all 26 test problems prescribed by the National Institute of Standards and Technology website for nonlinear regression (http://www.itl.nist.gov/div898/strd/nls/ nls_info.shtml) with 10 digit precision for the sum of squares parameter. Standard errors on fluorescence lifetime mode center and width parameters longer than the instrument response function (~50 ps) were estimated from repeated simulations to be less than 10%, whereas lifetime modes and widths below this value were estimated to have ambiguity factors approaching 2.

RESULTS

Simultaneous time resolution of FP and zooxanthellae PSII fluorescence emission spectra

Figure 2 illustrates a comprehensive view of the time and wavelength dependence of both the FP and zooxanthellae fluorescence from a green color morph specimen of P. versipora harvested from Sydney Harbor, Australia. Figure 2a illustrates the decay kinetics of the clearly separate FP and chl emission components. The FP fluorescence bands (peak ~515 nm) emitted with stronger amplitudes (left side of main panel) and slower decay times compared with the algae's chl fluorescence bands (peak 683 nm, upper right of main panel). Figure 2b profiles the data and model fits on a log scale to illustrate the slower main FP emission component in direct comparison with the more rapid PSII component. Figure 2c illustrates the same spectral data as in Fig. 2a,b plotted as a function of time, where each time channel is normalized to the peak FP emission wavelength intensity. It is clear that all components of the chl exhibit kinetics that are independent of the FP emission. Figure 2d shows the main fluorescence lifetime decay distribution components with positive amplitudes used to simulate the chl (red) and FP emission (green) in Fig. 2b. The major chl distribution mode was significantly broader and centered at a much faster lifetime value compared with the narrow FP emission component. The model simulations all clearly indicated that there were no kinetic components resolved in the analysis to show that the excited-state population of the FP exhibited any significant correlation with that of the algae's chl, i.e. except for the original laser excitation pulse event. Interestingly, however, the raw data normalization procedure in Fig. 2c did clearly reveal that the initial kinetic phase of the FP rise, during and immediately after the laser excitation (0-300 ps), is associated with a small re-producible greenshift in the fluorescence. We interpret this shift to be consistent with rapid resonance energy transfer (FRET) from blue to greener FP proteins. The evidence for FRET is examined in more detail later.

Overall, Fig. 2 clearly indicates that there is little or no evidence in the form of reciprocal amplitudes or sustained excitation for either direct resonance or indirect radiative excitation of the chl bands by the FP emission, respectively. We assume that resonance energy transfer would be evident as a reciprocal decay of the FP emission and a rise and decay in emission from the zooxanthellae, whereas an indirect radiative transfer would exhibit kinetic behavior similar to a sustained excitation pulse with the same shape profile (as a function of time) as the long-lived FP decay component. As mentioned previously, we were able to simulate the zooxanthellae emission kinetics assuming that the only convolved excitation event, beginning from time t = 0, was from the laser. The caption of Fig. 2 provides evidence for the goodness of fit by the global Durbin-Watson d-statistic (an indicator of autocorrelation trends in the time series) and Runs test (an indicator of the randomness of sign and symmetric distribution of the errors around O) (28,29).

Table 1 compares the spectral band and fluorescence lifetime distribution parameters for the main decay functions of both the FP emission peaks and algal PSII emission peaks that were resolved simultaneously. The experimental conditions were as described in Fig. 2 and include data from images for the blue and green color morphs of P. versipora plus a specimen of A. digitifera. In several cases the symbionts' emission was too weak to easily resolve from the strong background FP emission. Positioning of the sample with the three-axis stage allowed us to focus the laser spot (3-5 mm^sup 2^ area), in the cases presented in Table 1, on areas enriched in chl emission, which were visually brown (30) as opposed to blue or green in color. The data indicate that the FP emission bands (after completion of FRET) are broader with respect to wavelength and significantly longer in lifetime center values (> 1990 ps) than PSII, which was under the laser illumination conditions less than 800 ps. Interestingly, Table 1 indicates that the main FP emission bands were all resolved with a very narrow lifetime distribution width (100 ps).

Excitation-emission contour maps for FP and zooxanthellae PSII

Figure 3 illustrates the relationship between the wavelength of exciting light and the fluorescence emission for both FP (right) and algal PSII (left) measured at steady state with a photon-counting fluorimeter. The data compare blue and green color morphs of P. versipora. The left panels correspond to contour maps of the chl emission as a function of excitation wavelength, whereas the right panels correspond to the FP emission for the same excitation spectral region. The excitation wavelength region (350-450 nm) was chosen for two reasons: (1) because it spans both the area of peak excitation for the blue FP (Fig. 3a,b) and an area where little or no excitation is observed for the green FP (Fig. 3c,d); and (2) because it spans an area of considerable spontaneous excitation corresponding to major chl and carotenoid absorption bands. Hence, we reasoned that if energy transfer were to occur between the FP and algal PSII, then there should be a correspondence between the excitation contours and integral profiles for the FP and zooxanthellae maps. The contour maps in the left panels (Fig. 3a,c) indicate that the excitation contours for PSII (683 nm region) show little, if any, difference aside from an overlap (~20% at 650 nm) of the broad blue FP emission tail with the entire chl emission spectrum. The overlap of FP and PSII emission was considerably less in the green morph with its narrower FP emission band. The normalized integral profiles (Fig. 3e,f) for both the excitation (solid lines) and emission (clashed lines) of the zooxanthellae and FP emission reinforce the observations in the contour plots. Most importantly, the chl emission and excitation profiles are virtually indistinguishable when comparing morphs, indicating that little, if any, chl excitation can be attributed to the FP.

Evidence for FRET among fluorescent proteins

Figure 4 outlines evidence for FRET among the FP obtained by analyzing the kinetic sequences of the normalized spectral profiles from the FP emission regions. The data in the left panels correspond to 10 spectral profiles resolved at time intervals of 31.25 ps for specimens of A. digitiferti (Fig. 4a,b), A. nasuta (Fig. 4c,d) and both a blue (Fig. 4e,f) and green (Fig. 4g,h) morph of P. versipora. The spectral profiles referred to as 0 ps (black lines and shaded areas) correspond to the initial time frame of the analysis that coincided closely with the rise of the emission when a spectral shape was first clearly resolved. The difference spectra in the right panels correspond to the difference between the final spectrum at 281.25 ps and all the previous spectra and hence represent the time-resolved difference between the initial donor species and the final acceptor species of the FP. It is clear that each specimen exhibited a similar pattern with a time-dependent shift from blue to greener fluorescence that appeared to be complete, when viewed in this normalized format, in less than 200 ps. The shift was strongest in the two acroporids (Fig. 4a-d) and weakest in the green P. versipora (Fig. 4g,h). The peak of the difference spectra coincided with the peak of the FP emission, being greenest in the green P. versipora morph. In general, the spectral change coincided with a 3-10 nm shift in the peak emission and a filling in of the greener wing of the spectral profile. Once complete, there was little, if any, evidence for further spectral changes in the FP region, suggesting excitation had transferred to a terminal acceptor pool.

Figure 5 outlines a model FRET simulation of the FP streak-camera image obtained with A. digitifera and corresponds to the data presented in Fig. 4a,b. The image was simulated assuming two types of FP species, namely a donor species (blue) associated with the spectral contours in the early rise and rapid decay components and acceptor species (green) associated with the main decay spectral contour. The kinetic profiles plotted on a linear scale in Fig. 5a correspond to the complete wavelength-dependent integral of the measured image for each of the FP species, the total data, the model and the laser excitation pulse. The subplot displays the randomly distributed weighted residual errors to show that there were no remarkable systematic deviations to indicate a significant kinetic discrepancy between the data and model simulation. As explained in the caption of Fig. 5, the goodness of fit was further reflected by the acceptable global Durbin-Watson (-d-statistic and Runs test (28,29).

It is clear from Fig. 5a that there is a kinetic correspondence between the decay of the donor and initial rise of the acceptor. Figure 5b plots the corresponding distributions of m([tau]) for the spectral bands assigned to the donor and acceptor species; the lifetime distribution data in Fig. 5b represent the total integral for all bands from both species. The donor and acceptor were modeled with four and five Gaussian spectral bands, respectively; the sums but not the individual bands for both species are illustrated for clarity. The rapid decay of the donor was accounted for by a minor rapid (

Figure 5c illustrates the correspondence between the spectral kinetic contours of the model donor (blue) and acceptor (green) species and the actual image data (dotted line). The rapid decay of the donor that behaves as one component is evident when plotted on this linear (unnormalized) scale and complete to the ground state in less than a 1 ns time frame. The bands assigned to the acceptor(s) indicated some kinetic heterogeneity in the form of bluer components accepting from the donor and other greener species, apparently directly excited by the laser pulse. The differences in rise components for the acceptor's component bands were accounted for by allowing model variation for the relative amplitudes of the negative rise components and the main decay component. Figure 5d shows a normalized format for the image (dotted lines) and model (colored contour and lines) similar to that shown in Fig. 2c. It is evident from both the normalized spectral contour plot (Fig. 5d) and the natural spectral contour plot (Fig. 5c) that the obvious spectral shift is reasonably well simulated by the FRET model. However, because of the nature of the tendency for reciprocal amplitude components to cancel each other in the same spectral region, we cannot prove that the solutions obtained by the model are kinetically unique. Nonetheless, we conclude that the data and model simulation, especially when considered together, provide irrefutable evidence for FRET among the FP in vivo.

DISCUSSION

Inefficient transfer of light energy from the FP to the zooxanthellae

This study documents that in the shallow water corals analyzed, the relative quantum yield of energy transfer of the blue to green FP fluorescence to the light-harvesting apparatus of the zooxanthellae is negligible. This is in comparison with the direct excitation by absorption of light energy of wavelengths greater than 350 nm. The result is clear despite the fact that both the blue and green FP emission spectra clearly overlap with the excitation spectra of the zooxanthellae. The three main pieces of kinetic evidence obtained from the simultaneous acquisition analysis supporting the main conclusion include that (1) the symbionts' spectral components from PSII indicate distinctly more rapid decay kinetic modes and also different fluorescence lifetime distribution width and center properties; (2) the model simulation is consistent with only one convolved excitation source for both the FP and zooxanthellae components, namely, the laser pulse; the kinetic model thus indicates a low quantum yield for reabsorption of the FP emission by the zooxanthellae; and (3) there is no evidence of reciprocal kinetic amplitudes between the FP and zooxanthellae to indicate dipole-dipole interactions required for FRET between the two emitters.

The lack of FRET between algal PSII and the FP is most likely explained by prohibitive distance factors between the chromophores in the respective organelles (A. Salih, unpublished). Consistent with the kinetic data and modeling, the steady-state spectral data clearly show that the FP emission and excitation contour maps and integral profiles do not correlate with those of the algae's chl components. The data thus indicate that under natural illumination conditions the blue to green FP fluorescence would not make any substantial contributions to light harvesting in shallow water-dwelling corals. One important technical note, as an indication for the relative quantum yield of direct excitation, compared with indirect or FRET mechanisms, is that the excitation pulse profiles were acquired under conditions where the laser emission was attenuated with neutral density filters by a factor of 2 x 10^sup -4^. In these experiments, the peak laser signal intensity was still acquired in roughly 10-25% of the time required to resolve a strong FP emission signal. Moreover, it was also obvious that the yield of the zooxanthellae PSII emission was considerably lower than the FP emission.

We thus propose, consistent with the suggestions of Dove et al. (4) and Salih et al. (8), that FP absorption in the UV to blue-green regions of the solar emission spectrum most likely serves to intercept, scatter and block excess visible and potentially damaging UV excitation from reaching the symbionts and other vital organelles of the host coral. It was clear and potentially physiologically relevant that the highly fluorescent blue FP morph of P. versipora exhibited higher light absorption, especially in the UV, compared with the less fluorescent green morph. In this light, our recent unpublished results indicate that the overall FP density in the coral tissue is typically considerably higher in the blue as opposed to green color morphs of P. versipora.

Because of the low-integrated intensity (

One final point of interest comparing the fluorescence lifetime distributions of the algal PSII and the FP was the clear differences in the resolved widths of the fluorescence lifetime distributions. As explained in earlier studies (32,34,35), the width of a fluorescence lifetime distribution of a protein-bound fluorophore, like the FP and algal chl, may be interpreted to indicate heterogeneity in the molecular environment of the fluorophore. These ideas are in accordance to the theories of protein conformational substates and dynamics proposed by Frauenfelder (36,37) and related theories regarding the kinetic behavior of protein-bound fluorophores promulgated by Alcala et al. (38,39). Because of the narrowness of the FP fluorescence lifetime distributions in this study, we interpret the main FP emission to emanate from a homogeneous population of fluorophores, probably representing proteins with fluorophores in very similar molecular environments. As explained before, PSII (35) lifetime distributions at room temperature are generally broad (>100 ps in many cases), and the width may be attributed to a variety of heterogeneity factors including the PSII chl-protein environment and dynamic PSII photochemical activities of the sample, inter alia.

Evidence for Forster resonance energy transfer among fluorescent proteins in vivo

It is clear that FP emission observed in all specimens of this study is influenced by minor, rapid spectral shift components that we attribute to FRET. The spectral shifts at room temperature always indicate downhill energy transfer as evidenced by reciprocal amplitudes decreasing in the blue and increasing in the greener wings of the spectrum. We speculate that in each color morph the initial donor species may relate primarily to minor pools of pigments that are likely to be in close spatial association with each other in oligomeric form (40). Perhaps the oligomers are present either in the granular FP bodies or in intracellular FP bodies that are not enclosed in FP granules (8). The donors may relate to the same or similar protein structures as the acceptors with the altered spectral properties perhaps attributable to altered chemical properties (pH) or environments (electrostatic factors) of the amino acid residues associated with the FP fluorophore (3,6), inter alia. The close structural similarity of the donor-acceptor pairs is reasoned because, as previously noted, the main FP emission component is a predictor for the initial peaks resolved in the early phases of the FP excitation.

Moreover, it is established from recent time-resolved fluorescence studies in vitro that oligomers comprising both immature green and mature red forms of the coral FP known as Discosoma Reel (Ds-Red) exhibit FRET (41,42). The reciprocal kinetic responses observed in vitro generally correspond with those obtained here in vivo with the blue and green FP. Notably, in comparison with one of the clearest case studies (41), our main decay component was faster for the blue-green acceptor

CONCLUSIONS

This study demonstrates the use of in vivo simultaneous time and wavelength fluorescence acquisition and DCI analysis (27) for answering important physiological questions, namely, (1) that the blue and green FP do not function in light harvesting for reef-building corals that dwell in shallow water; and (2) that FRET is a primary process in the excited-state behavior of FP in reef-building corals. This study provides a foundation of data from nonstressed corals from key geographical regions of interest with respect to coral bleaching (13,44,45) based on which future studies of stress conditions may be initiated and compared. Future work should be aimed at exploring the mechanism of the proposed photoprotective function of the FP and the physiological state of both the FP and zooxanthellae in response to coral-bleaching episodes. The fluorescence lifetimes and spectra should provide a valid quantitative indicator, especially under conditions where the tissue concentrations and chemical properties of either zooxanthellae or FP are likely to change simultaneously. This is because determination of the fluorescence intensities and ratios alone is complicated by concomitant changes in the concentration and excited-state lifetimes of fluorophores in the tissue. Finally, the methods used for identification of the new FP and FRET partners described in this study may be of interest to researchers who wish to isolate, clone and mutate FP for molecular technology, as is becoming increasingly common for the green and red FP from other cnidarians.

Acknowledgements-A.M.G. and S.I. thank the Japanese Society for the Promotion of Science for an invited visiting fellowship to A.M.G. Special thanks are offered to Dr. Mino Hiroyuki and Kana Sugiura in Nagoya, Japan, and Drs. Yasuko Sakihama and Shunichi Takahashi and Takashi Nakamura, M.Sc., in Okinawa, Japan, for experimental assistance with coral collection, shipment and storage. R.V.W. and H.Y. were supported by a Grant-in Aid for Scientific Research (B) (12480166) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan. A.M.G. thanks Dr. Marilyn Ball and the Ecosystem Dynamics group of the ANU RSBS for stimulating discussions.

[para]Posted on the website on 6 March 2003.

[dagger]The results of this study were presented in their entirety as a poster at the ComBio2002 conference held in Sydney, Australia, 29 September-3 October 2002.

REFERENCES

1. Kawaguti, S. (1944) On the physiology of reef corals. VI. Study on the pigments. Palao Trop. Biol. Stn. Stud. 2, 617-674.

2. Mazel, C. H. (1995) Spectral measurements of fluorescence emission in Caribbean cnidarians. Mar. Ecol. Prog. Ser. 120, 185-191.

3. Matz, M. V., A. F. Fradkov, Y. A. Labas, A. P. Savitsky, A. G. Zaraisky, M. L. Markelov and S. A. Lukyanov (1999) Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17, 969-973.

4. Dove, S. G., O. Hoegh-Guldberg and S. Ranganathan (2001) Major colour patterns of reef-building corals are due to a family of GFP-like proteins. Coral Reefs 19, 197-204.

5. Fuchs, E. (2001) Separating the fluorescence and reflectance components of coral spectra. Appl. Optics 40, 3614-3621.

6. Labas, Y. A., N. G. Gurskaya, Y. G. Yanushevich, A. F. Fradkov, K. A. Lukyanov, S. A. Lukyanov and M. V. Matz (2002) Diversity and evolution of the green fluorescent protein family. Proc. Natl. Acad. Sci. USA 99, 4256-4261.

7. Lukyanov, K. A., A. F. Fradkov, N. G. Gurskaya and M. V. Matz (2000) Natural animal coloration can be determined by a nonfluorescent GFP homolog. J. Biol. Chem. 275, 25879-25882.

8. Salih, A., A. W. D. Larkum, G. Cox, M. Kuhl and O. Hoegh-Guldberg (2000) Fluorescent pigments in corals are photoprotective. Nature 408, 850-853.

9. Myers, M. R., J. T. Hardy, C. H. Mazel and P. Dustan (1999) Optical spectra and pigmentation of Caribbean reef corals and macroalgae. Coral Reefs 18, 179-186.

10. Papina, M., Y. Sakihama, C. Bena, R. Van Woesik and H. Yamasaki (2002) Separation of highly fluorescent proteins by SDS-PAGE in Acroporidae corals. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 131, 767-774.

11. Hoegh-Guldberg, O. (1999) Climate change: coral bleaching and the future of the world's coral reefs. Mar. Freshw. Res. 50, 839-866.

12. Fitt, W. K., B. E. Brown, M. E. Warner and R. P. Dunne (2001) Coral bleaching: interpretation of thermal tolerance limits and thermal thresholds in tropical corals. Coral Reefs 20, 51-65.

13. Knowlton, N. (2001) The future of coral reefs. Proc. Natl. Acad. Sci. USA 98, 5419-5425.

14. Warner, M. E., W. K. Fitt and G. W. Schmidt (1999) Damage to photosystem II in symbiotic dinoflagellates: a determinant of coral bleaching. Proc. Natl. Acad. Sci. USA 96, 8007-8012.

15. Jones, R. J. and O. Hoegh-Guldberg (2001) Diurnal changes in the photochemical efficiency of the symbiotic dinoflagellates (Dinophyceae) of corals: photoprotection, photoinactivation and the relationship to coral bleaching. Plant Cell Environ. 24, 89-99.

16. Jones, R. J., O. Hoegh-Guldberg, A. W. D. Larkum and U. Schreiber (1998) Temperature-induced bleaching of corals begins with impairment of the CO2 fixation mechanism in zooxanthellae. Plant Cell Environ. 21, 1219-1230.

17. Ralph, P. J., R. Gademann, A. W. D. Larkum and U. Schreiber (1999) In situ underwater measurements of photosynthetic activity of coral zooxanthellae and other reef-dwelling dinoflagellate endosymbionts. Mar. ecol. Prog. Ser. 180, 139-147.

18. Salih, A., O. Hoegh-Guldberg and G. Cox (1998) Photoprotection of symbiotic dinoflagellates by fluorescent pigments in reef corals. In Proceedings of the Australian Coral Reef Society (Edited by J. G. Greenwood and N. J. Hall), pp. 217-230. School of Marine Science, University of Queensland, Brisbane.

19. Salih, A. (2000) Mechanisms of light and temperature-induced bleaching of corals and the sunscreening role of their fluorescent pigments. Ph.D. dissertation, University of Sydney, Sydney.

20. Kawaguti, S. (1969) Effect of the green fluorescent pigment on the productivity of the reef corals. Micronesica 5, 313.

21. Schlicter, D., H. W. Fricke and W. Weber (1986) Light-harvesting by wavelength Iransformation in a symbiotic coral of the Red Sea twilight zone. Mar. Biol. 91, 403-407.

22. Lackowicz, J. (1999) Principles of Fluorescence Spectroscopy, 2nd ed. Kluwer Academic Publishers, Dordrecht, The Netherlands.

23. Cox, G. and A. Salih (2002) Confocal microscopy of GFP-like pigments in corals. In Focus on Multidimensional Microscopy (Edited by P.-C. Cheng and H. Kim). World Scientific Publishing Company, Singapore. (In press)

24. Gilmore, A. M. and H. Y. Yamamoto (2001) Time-resolution of the antheraxanthin- and [Delta]pH-dependent chlorophyll a fluorescence components associated with photosystem II energy dissipation in Mantoniella squamata. Photochem. Photobiol. 74, 291-302.

25. Carlson, B. A. (1999) Organism responses to rapid change: what aquaria tell us about nature. Am. Zool. 39, 44-55.

26. Gilmore, A. M. and M. C. Ball (2000) Protection and storage of chlorophyll in overwintering evergreens. Proc. Natl. Acad. Sci. USA 97, 11098-11101.

27. Gilmore, A. M., S. Itoh and Govindjee (2000) Global spectral-kinetic analysis of room temperature chlorophyll a fluorescence from light-harvesting mutants of barley. Phil. Trans. R. Soc. Lond. B 355, 1371-1384.

28. Draper, N. R. and H. Smith (1998) Applied Regression Analysis, 3rd ed. Wiley and Sons Inc., New York.

29. Straume, M. and M. L. Johnson (1992) Analysis of residuals: criteria for determining goodness-of-fit. Methods Enzymol. 210, 87-105.

30. Kawaguti, S. (1966) Electron microscopy on the fluorescent green of reef corals with a note on mucous cells. Biol. J. Okayama Univ. 12, 69-80.

31. Van Kooten, O. and J. F. H. Snel (1990) The use of chlorophyll fluorescence nomenclature in plant stress physiology. Photosynth. Res. 25, 147-150.

32. Gilmore, A. M., T. L. Hazlett, P. G. Debrunner and Govindjee (1996) Photosystem II chlorophyll a fluorescence lifetimes are independent of the antenna size differences between barley wild-type and chlorina mutants: comparison of xanthophyll cycle-dependent and photochemical quenching. Photosynth. Res. 48, 171-187.

33. Hecks, B., C. Wilhelm and H.-W. Trissl (1996) Functional organization of the photosynthetic apparatus of the primitive alga Mantoniella squamata. Biochim. Biophys. Acta 21, 21-30.

34. Govindjee, M. Van de Ven, C. Preston, M. Seibert and E. Gratton (1990) Chlorophyll a fluorescence lifetime distributions in open and closed photosystem II reaction center preparations. Biochim. Biophys. Acta 1015, 173-179.

35. Gilmore, A. M. and Govindjee (1999) How higher plants respond to excess light: understanding energy dissipation strategies for photosystem II. In Concepts in Photobiology: Photosynthesis and Photomorphogenesis (Edited by G. S. Singhal, G. Renger, S. K. Sopory, K.-D. Irrgang and Govindjee), pp. 513-548. Narosa Publishing House, New Delhi.

36. Frauenfelder, H. A., F. Parak and R. D. Young (1988) Conformational substrates in proteins. Annu. Rev. Biophys. Biophys. Chem. 17, 451-479.

37. Frauenfelder, H. A. and D. T. Leeson (1998) The energy landscape in non-biological and biological molecules. Nat. Struct. Biol. 5, 757-759.

38. Alcala, J. R., E. Gratton and F. G. Prendergrast (1987) Resolvability of fluorescence lifetime distributions using phase fluorometry. Biophys. J. 51, 587-596.

39. Alcala, J. R., E. Gratton and F. G. Prendergrast (1987) Fluorescence lifetime distributions in proteins. Biophys. J. 51, 597-604.

40. Lounis, B., J. Deich, F. I. Rosell, S. G. Boxer and W. E. Moerner (2001) Photophysics of DsRed, a red fluorescent protein, from the ensemble to the single-molecule level. J. Phys. Chem. B 105, 5048-5054.

41. Schuttrigkeit, T. A., U. Zachariae, T. von Feilitzsch, J. Wiehler, J. von Hummel, B. Steipe and M. E. Michel-Beyerle (2001) Picosecond time-resolved FRET in the fluorescent protein from Discosoma Red (wt-DsRed). Chemphyschem 5, 325-328.

42. Cotlet, M., J. Hofkens, S. Habuchi, G. Dirix, M. van Guyse, J. Michiels, J. Vanderleyden and F. C. de Schryver (2001) Identification of different emitting species in the red fluorescent protein DsRed by means of ensemble and single molecule spectroscopy. Proc. Natl. Acad. Sci. USA 98, 14398-14403.

43. Yarbrough, D., R. M. Wachter, K. Kallio, M. V. Matz and S. J. Remington (2001) Refined crystal structure of DsRed, a red fluorescent protein from coral, at 2.0-A resolution. Proc. Natl. Acad. Sci. USA 98, 462-467.

44. Loya, Y., K. Sakia, K. Yamazato, H. Sambali and R. Van Woesik (2001) Coral bleaching: the winners and losers. Ecol. Lett. 4, 122-131.

45. Van Woesik, R. (2001) Coral bleaching: transcending spatial and temporal scales. Trends Ecol. Evol. 16, 119-121.

Adam M. Gilmore*1, Anthony W. D. Larkum2, Anya Salih3, Shigeru Itoh4, Yutaka Shibata4, Chiaki Bena5, Hideo Yamasaki6, Marina Papina7 and Robert Van Woesik8

1Ecosystem Dynamics Group, Research School of Biological Sciences, Australian National University, Canberra, Australian Capital Territory, Australia;

2Sydney University Biological Informatics and Technology Centre, School of Biological Sciences, University of Sydney, Sydney, New South Wales, Australia;

3Electron Microscope Unit and the Australian Key Centre for Microscopy and Microanalysis, University of Sydney, Sydney, New South Wales, Australia;

4Photobioenergetics Group, Department of Physics, Nagoya University, Nagoya, Japan;

5Department of Marine Environmental Science, University of the Ryukyus, Okinawa, Japan;

6Center of Molecular Biosciences, University of the Ryukyus, Okinawa, Japan;

7Laboratory of Biotechnology, Institute of Biology and Soil Sciences, Far East Branch of Russian Academy of Sciences, Vladivostok, Russia and

8Department of Biological Sciences, Florida Institute of Technology, Melbourne, FL

Received 23 September 2002; accepted 18 February 2003

*To whom correspondence should be addressed at: Ecosystem Dynamics, Research School of Biological Sciences, Australian National University, Institute of Advanced Studies, Canberra, Australian Capital Territory 0200, Australia. Fax: 612-6125-5095: e-mail: gilmore@rsbs.anu.edu.au