ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 236, No. 1, January, pp. 130-139, 1985

Characterization of Light-Harvesting Mutants of Rhodopseudomonas sphaeroides. I. Measurement of the Efficiency of Energy Transfer

from Light-Harvesting Complexes to the Reaction Center

STEVEN W. MEINHARDT, **l PATRICIA J. KILEY,? SAMUEL KAPLAN,? ANTONY R. CROFTS,* AND SHIGEAKI HARAYAMA$

*Department of Physiology and Biophysics, and tDepartment of Mtiobiology, University of Illinois at Urbana-Champaign, 407 S. Goodwin, Urbana, Illinois 61801; and $Department de Biochimie Medical,

Faculte de Medecine, Universite de Geneve, C.M. V Geneva, Switzerland

Received May 10, 1984, and in revised form August 3, 19&1

Light-harvesting mutants of Rhodopseudomonas sphaemides lacking either the B800- 850 complex or the B8’75 complex have been characterized by their absorption spectra in the visible and near-infrared region, and by their ability to transfer energy from the light-harvesting complexes to the reaction center. A new method of measuring the relative efficiency of energy transfer from the light-harvesting complexes to the reaction center is described. The B875- mutant had absorption maxima in the near- infrared at 800 and 849 nm with no evidence of an 875-nm shoulder. The efficiency of energy transfer from the light-harvesting complexes to the reaction center in the B875- mutant was 24% of the value measured for the wild-type strain and the B800- 850- mutant. Yet, despite the fact that the efficiency of energy transfer for the B800- 850- mutant and the wild-type strain were the same, there was a large difference in their photosynthetic unit size. These results are discussed in the context of a model in which light energy captured by the B800-850 complexes is transferred through the B875 complexes to the reaction center. CC’ 1985 i\eadcmic Press, Inc

In Rhodopseudowumas sphaeroides the light-harvesting complexes are the major pigment-protein components found in the photosynthetic membrane (1). These com- plexes absorb light energy and transfer excitation energy to the reaction centers where the primary photochemistry occurs (2-5). The light-harvesting complexes have been classified into two groups based on their near-infrared absorption maxima (5). The first type of light-harvesting com- plex, the B875 complex, occurs in the membrane in a fixed stoichiometry of ap- proximately 30 B875 BCh12 per reaction

1 To whom correspondence should be addressed. ’ Abbreviations used: BChl, bacteriochlorophyll;

Mops, I-morpholinepropane sulfonic acid; SDS- PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; DAD, 2,3,5,6-tetramethyl-pphenyl- enediamine.

0003-9861/85 $3.00 Capyright Q 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

130

center, whereas the amount of the second type, the B800-850 complex, varies in- versely as a function of the incident light intensity (6). Although the composition of both light-harvesting complexes has been described (5), the in wivo organization of the pigment-protein complexes is not well understood.

Energy transfer has been shown to oc- cur between different light-harvesting complexes (7-9) and between light-har- vesting complexes and reaction centers (9-11). From fluorescence measurements of chromatophore membranes, Monger and Parson (9) suggested that reaction centers are surrounded and interconnected by B875 complexes, and that the B800- 850 complexes are arranged peripherally around this network. This “lake” model for organization of chromatophore pho-

ENERGY TRANSFER IN R. sphaeroides LIGHT-HARVESTING MUTANTS 131

tosynthetic units emphasizes that reaction centers share a common pool of light- harvesting complexes, and suggests that the B875 complex is an essential inter- mediate for efficient energy transfer. In order to define further the role of the individual light-harvesting complexes in the photoconversion of light energy to chemical energy, we have studied two mutants of R. sphaertides, each missing one of the light-harvesting complexes. We have measured the relative efficiency of energy transfer from the light-harvesting complexes to the reaction centers, and have observed the effect of these muta- tions on cell growth.

MATERIALS AND METHODS

The isolation, and physiological and biochemical characterization of the mutant strains will be detailed elsewhere (Kiley et al, manuscript in preparation). Both wild-type cells and mutants were routinely grown aerobically (25% Ox, 1% COx, 74% N,), in Sistrom’s minimal medium (12), and were shifted at a culture density of 1.5 X 10s cells/ml to anaerobic photoheterotrophic conditions through a sequential decrease in Oa concentration (10% 02, 2.5% CO*, 87.5% Na, 30 min; 5% O,, 2.5% CO,, 92.5% NB, 60 min; 2.5% Oa, 2.5% COs, 95% Na, 30 min). Cultures were illuminated 15 min after reaching the 0% Oa, 97.5% Nz, 2.5% CO, atmosphere. The incident light intensity was varied (10, 25, 50, or 110 W m-a), and was measured with a Yellow Springs-Kettering Model 6.5A radiometer through a Corning eolored- glass filter (CS no. ‘7-69, 620-1100 nm). Cell growth was measured with a Klett-Summerson photometer (no. 66 filter), where 1 Klett unit = 1 X 10’ cells/ml. Exponentially growing cells were routinely harvested at l-l.5 X lo9 cells/ml.

Chromatophores were prepared from wild-type and mutant cells by one passage of cells through a French pressure cell at 13,000-14,000 Psi. Large cell fragments and unbroken cells were removed by centrifugation at 17,OOOg for 20 min and 35,000~ for 45 min (4°C). Chromatophore membranes were then pelleted by centrifugation at 161,OOOg for 1 h. The membrane pellet was resuspended in 100 mM KCl, 50 mM Mops, pH 7.0, and pelleted a second time. The final pellet was resuspended in a minimal volume of the same buffer. Reaction center turnover was measured at 542 nm to avoid the background absor- bance due to the light-harvesting complexes. The reaction center concentration was measured at 542 nm, using an extinction coefficient of 10.3 mM-' cm-i (14). Chromatophores were poised at 405 + 5 mV to avoid fast rereduction of the reaction center by

cytochrome q. Valinomycin and nigericin were pres- ent to prevent the accumulation of any membrane potential or pH gradient, and to eliminate electro- chromic absorbance changes. BChl concentrations were estimated as described (15).

The actinic flash was provided by a xenon arc flash lamp with a custom-built flash tube (T. W. Wingent Ltd., Cambridge, U. K.), which gave a flash width of approximately 25 ps at half-maximal am- plitude. The flash was filtered by a Schott RG-695 red-glass filter. The intensity of the flash was varied by the use of combinations of Schott smoked-glass, neutral-density filters NG-11, NG-4, and two NG-3 filters which transmitted 69, 33, 12, and 2% of the light at 630 nm, respectively. The flash lamp emission spectrum was obtained by measuring the flash tran- sient with a photodiode detector as described (13). The light was passed through a Schott RG-695 filter, and then through a SPEX Industries Model 1672 Doublemate double monochromator with additive dispersion with 1.25-mm slits and gratings blazed at 500 nm with 1200 grooves/mm. The transients were measured every 2 nm from 700 to 900 nm. At each point, the total intensity was obtained by integration of the flash profile for the duration of the flash, and the integrated value was plotted to give the spectra shown below (Fig. 5A). Values between the measured points were estimated by linear interpolation.

RESULTS AND DISCUSSION

Absorption Spectra

In the process of screening for mutants in the photosynthetic apparatus, we have isolated two mutants in the light-harvest- ing pigment-protein complexes; RS104, a blue-green mutant missing the B800-850 light-harvesting complex and carotenoids, and RS103, a red-brown mutant missing the B8’75 light-harvesting complex. In Figs. IA and B are shown the absorption spectra of the two mutants and the parent wild-type strain, RS2. In the near-infrared region the mutant RS104 showed a single major absorption maximum at 87’0 nm, with a slight shoulder at 848 nm and a minor absorption at 805 nm. The RS103 mutant had an absorption spectrum showing peaks at 800 and 849 nm. There was no evidence of a shoulder at 875 nm. The wild-type strain, RS2, showed char- acteristic absorption maxima at 800 and 850 nm, with a large shoulder on the 850- nm peak which appeared to have a max- imum between 875 and 880 nm.

132 MEINHARDT ET AL.

I,, I,,, 1, I, 1 I, I 750 800 850

(nm) ‘O”

\ k& RS104

, I I I I I I , I 400 440 480 520 5eo

(ml)

FIG. 1. Absorption spectra of RS104, RS103, and RS2, in the visible and near-infrared region. (A) Chromatophores from each strain were suspended in 50 mM Mops, 100 mM KCl, pH 7.0. The bar represents an absorbance of 0.156 for RS2, 0.625 for RS103, and 0.313 for RS104. (B) Conditions were the same as in (A) except that the bar represents a single absorption value.

We attempted to model the BChl ab- sorption spectrum of the wild-type strain by using a linear combination of the BChl absorption spectra obtained from the two mutants. A reasonable model of the wild- type spectrum was obtained only if the absorption spectrum of the blue-green strain, RS104, was shifted from 870 to 878 nm. The best fit was obtained when, in addition to the shift in the RS104 absorp- tion spectrum, the absorption spectrum of the RS103 mutant was shifted 1 nm to the blue (Fig. 2A). After normalization and addition of these spectra, a spectrum which showed a maximal deviation of +3% from the wild-type spectrum was obtained (Fig. 2B). From this composite spectrum and the spectra which were used to construct it, we were able to choose wavelengths which measured the absorp-

tion due to one complex while compensat- ing for the absorption due to the other complex. The amount of BChl in the B875 complex can be measured at 878 minus 820 nm, (E = 73 k 2.5 mM-’ cm-‘), whereas the concentration of BChl in the B800-850 complex can be measured at 849 minus 900 nm (t = 96 +- 4 mK’ cm-‘). These ex- tinction coefficients were obtained from the two mutant strains by measuring the BChl concentrations and then the absorp- tion spectrum at a known BChl concen- tration. The difference between the absor- bance at 850 and 901 nm was measured for RS103, and between 870 and 812 nm

JJ&

750 800 850 (nm) ‘O”

FIG. 2. Modeling of the wild-type spectrum from the spectra of the mutants. (A) (a) The composite spectrum obtained by the addition of spectra (b) and (c). (b) Spectra of RS103 from Fig. 1A which has been shifted 1 nm to the blue and multiplied by 0.2. (c) The spectra of RS104 shown in Fig. 1A which has been shifted by 8 nm to the red and multiplied by 0.38. (B) The wild-type spectrum from Fig. 1A is plotted with the composite spectrum from Fig. 2A which has been normalized at 800 nm and corrected for scattering at 725 nm. The difference between the two spectra is also plotted after being enlarged four times.

ENERGY TRANSFER IN R. sphaeroides LIGHT-HARVESTING MUTANTS 133

was measured for RS104, and was used to calculate the extinction coefficients shown. Since there was no change in the shape of the spectra, these differences should be equivalent to the values for the wild type after the shift in the spectra. These wave- lengths can be used for estimating the relative amounts of the two complexes only in wild-type strains since the absorp- tion maxima of the B875 complex appears to be shifted depending on the mutation present.

We have used these extinction coeffi- cients to estimate the total amount of BChl in the wild-type strain, RS2 (16). The estimated amounts of BChl were usu- ally lower than the measured amounts but were within an error of flO% (avg. of eight measurements). Extinction coef- ficients for the two light-harvesting com- plexes have been previously obtained for the B800-850 complex at 850 nm, and for the B875 complex at 875 nm (16). Since these values were obtained for a single wavelength they do not compensate for the other complex when both are present. Both of our extinction coefficients were approximately 25% lower than those val- ues reported by Clayton and Clayton (16), suggesting that there is a systematic dif- ference in the measurements. An obvious difference between the two measurements is that their values were obtained by using a combination of spectra obtained from detergent-isolated complexes and membranes while our measurements were obtained solely from untreated membrane preparations. The presence of detergents in the isolated complexes or a loss of BChl in the isolation process could have affected the absorption spectra and so the extinc- tion coefficients.

In the region in which the carotenoids absorb (Fig. lB), from 400 to 540 nm, the wild type had absorption maxima at 422- 424, 450, 476, and 508 nm. The absorption maxima of the three longer wavelength peaks of the mutant RS103 were shifted 2 to 3 nm to the red relative to the parent strain, and were at 452, 479, and 511 nm. The RS104 mutant showed no absorption bands for the carotenoids. The absorption maximum around 423 nm was probably

due to the cytochrome soret band. A shift in the carotenoid absorption maxima has been previously reported (17-19). Holmes et al. (18) reported that the magnitude of the shift was dependent on the amount of the B800-850 complex present relative to the B875 complex. RS103 represents what should be the limiting case for this shift since there was no detectable B875 com- plex present and the carotenoid absorption maxima were shifted the greatest amount reported to date. The carotenoids in the B800-850 complex have been proposed to be the only carotenoid species to undergo an electrochromic shift in response to the membrane potential (18,19). A comparison of the absorption spectra of RS103 in the carotenoid region to the spectrum obtained by the integration of the RS103 carotenoid band shift indicated that the carotenoids which exhibit the electrochromic shift had absorption maxima further to the red (data not shown). A similar 2-nm shift to the red in the carotenoid absorption max- ima has been reported in the reaction center mutant C71 (20). The shift in the carotenoid spectra may be related to the coupling of the light-harvesting complexes to the reaction centers.

Preliminary fluorescence experiments indicated that the B875- mutant had a maximal fluorescence intensity that was approximately ten times the intensity of the wild type. Uncorrected fluorescence emission spectra of all three strains showed the wild-type and the B800-850- strain to have a maximum near 870 nm. The B875- mutant, RS103, showed a single fluorescence maximum near 850 nm and no peak near 870 nm. These results indi- cate that most of the light energy ab- sorbed by the B800-850 complex in RS103 is emitted as fluorescence from the 850- nm BChl before it is transferred to the reaction center. The fluorescence emission from wild-type R. sphaeroides (strain Y) has been shown to consist of a constant component, with a peak centered at 865 nm, and a variable component centered at 890 nm (21). We were not able to resolve the different components and our peak position is shifted to the blue because the spectra have not been corrected for

134 MEINHARDT ET AL.

the instrument response. Similar light- harvesting mutants have been obtained in R. capsdata, and the fluorescence emis- sion spectra of the wild type (strain 37134) was shown to be composed of two peaks at 863 and 891 nm. A B800-850- strain, Ala+, exhibited a single fluorescence max- imum at 891 nm, while a B875- reaction center-mutant, Y5, showed a single max- imum at 865 nm (22). These results are qualitatively in agreement with ours.

The E#ect of Light Intensity on Growth Rates

The BChl spectral alterations identified in the mutants can be attributed to the absence of individual light-harvesting complexes. Such deficiencies should have a substantial effect on the photosynthe- tic capacity of the cell. In order to test this possibility, photoheterotrophic growth rates were measured over a range of in- cident light intensities optimal for the wild-type organism. Figure 3 shows that the mutant strains could grow as well as the wild type at high light intensities but, as the incident light intensity was lowered, the generation time of the mutant strains increased. Since the cells grew aerobically at rates comparable to the wild-type strain (3 h doubling time), it appeared that the slow growth rates of the mutants at low light intensities were probably due to aberrations in their light-gathering ability

I I 20 40 60 SO 100

Lighl Intens~iy(waits/mz)

FIG. 3. Generation times of R. sphaeroides wild type (A), B875- mutant (0), and B800-850- mutant (A) as a function of incident light intensity. Gener- ation times and light intensity were measured as described in text.

or possibly to some other undetermined defect in the photosynthetic process. We have measured the kinetics of electron transfer through the cytochrome b-cl complex in chromatophores derived from all of the strains, and have found that, at saturating light intensities, the electron transfer chain functions normally (data not shown). Therefore, the coupling of reaction center turnover to the cytochrome b-cl complex was unmodified, eliminating electron transport as a cause of the phe- notype for these mutants.

In addition, SDS-PAGE analysis of chromatophores prepared from strains RS103 and RS104 showed that, in each case where a spectral complex was absent, there was a corresponding loss of the polypeptides associated with that complex. This was determined by comparison of chromatophores derived from the mutants with wild-type-derived chromatophores, as well as with purified B800-850 and B875 spectral complexes (Kiley and Kaplan, manuscript in preparation). Also, both mutants reverted readily to an apparent wild-type phenotype, indicating that the only apparent difference in these strains is in their light-harvesting complexes (Kiley, unpublished results). Therefore, we examined the ability of these com- plexes to transfer energy to the reaction center.

Measurement of the Relative Eficiency of Energy Transfer

To measure the energy transfer from the light-harvesting complexes to the re- action centers, we have measured the amount of reaction center turnover after a single flash at varying flash intensities. In Fig. 4A are shown curves obtained in this manner for the three strains of bac- teria. The points have been normalized to the total amount of reaction center present in each sample. All three strains of bac- teria were normally suspended at a BChl concentration of 18 + 0.2 PM, although similar results were obtained using a lower BChl concentration of 5 PM. The wild-type strain showed 90% turnover of reaction center complexes at a light in-

ENERGY TRANSFER IN R. sphueroides LIGHT-HARVESTING MUTANTS 135

FIG. 4. Saturation properties of reaction centers in RS2, RS103, and RS104. (A) Chromatophores were suspended to a concentration of 18 f 0.2 pM BChl in 50 mM Mops, 100 mM KCl, pH 7.0, with 2 pM each of valinomycin, nigericin, and DAD, and 500 yM each of ferricyanide and ferrocyanide. Ten micromolar antimycin was present to prevent turnover of the & c1 complex. Points were measured 16 ms after the first flash from kinetic traces (average of four or eight lOO-ms sweep time, 200-ps instrument response time) obtained at 542 nm. The total amount of the maximal flash oxidizable reaction center change was measured by the changes from the base line after the fourth flash (16 ms between flashes) from the trace with the maximal flash lamp intensity. The symbols represent the wild type (A), the B875- mutant (0), and the B800-850- mutant (0). (B) The points were normalized by the change on the first flash at the maximal flash lamp intensity and plotted as the double reciprocal. Same symbols as in (A).

tensity considerably lower than that found for the B875- mutant, RS103, and the B800-850- mutant, RS104. The RS104 mu- tant reached 90% turnover on the first flash only at the maximal flash lamp light intensity, but with the B8’75- mutant, RS103, reaction center turnover as high as 90% on the first flash was never ob- tained at the maximal intensities available from our flash lamp. Limitations in our instrument would not allow us to work routinely at a concentration of RS103 that

would give 90% turnover of the reaction center population on the first flash. In each case the curves appeared to follow a simple saturation process. This is clearly shown in Fig. 4B, which is a double- reciprocal plot of the data from Fig. 4A. The data have been normalized to the maximal reaction center change on the first flash, instead of the total amount of reaction centers present. The reason for this normalization is explained below.

A simple saturation process can be de- scribed by

R = %a,(1 + A/&J, PI where R is the response being measured, R max is the maximal response, A is the amplitude of the stimulus applied, and Aso is the amplitude of stimulus necessary for 50% of maximal response. When this equation is rewritten in reciprocal form

1 40 1 1 -=-.- ~ R A En,, + &ax

PI

a plot of l/R versus l/A should yield a straight line.

In our system the analogous expression is given in

1 _ 40 1 1 R I &,a, + %a, ’

[31

where A is equal to the relative intensity of the flash lamp, I. The double-reciprocal plot of the normalized saturation curves, shown in Fig. 4B, all follow straight lines as predicted for a simple saturation pro- cess, and pass through the Y axis at 1 due to the normalization procedure used. At the point where the lines cross the Xaxis the intensity should be equal to -l/IsO.

The efficiency of transfer from the light- harvesting complexes to the reaction cen- ter can be determined using

R = 4t - N- ~PC- 4i s [E(u) - A(~ [41

where R is the response (in our case the reaction center turnover); N is the number of light harvesting molecules (BChl per reaction center); & is the efficiency of transfer from the light-harvesting com- plexes to the reaction center, 4Pc is the efficiency of photochemistry; & is the ef- ficiency of energy transfer within the

complexes; and the terms in the integral represent the overlap in the spectrum of N 50-

the excitation light (E(V)) and in the ab- m ;;

sorption of the light-harvesting complexes <40-

U(u)). We have assumed that the mutations 30-

have resulted in phenotypes which have not altered the efficiency of reaction center 20-

photochemistry or the efficiency of energy transfer within the complexes. Therefore, 10-

$Pc and q$ are the same in all strains. The overlap integral is proportional to 0-

the integral of the product of the emission spectrum of the flash lamp and the ab- sorption spectrum of the light-harvesting complexes. In Fig. 5A is shown the ab- 20

B

720 760 600 840 880

sorption spectra of the three strains which were obtained at the same BChl concen- /

tration, and the emission spectrum of the flash lamp plotted in relative units. At the same BChl concentrations, the three strains had approximately the same in- tegrated absorbance in the region from 700 to 900 nm as shown in Fig. 5B. Nor- malizing the values by that of the wild- type strain, RS2, we obtained values of 0.96, 1.00, and 1.08 for the integral of the absorption for RS103, RS2, and RS104, respectively. Figure 5C shows the integra- tion of the product of the flash lamp

10-

emission spectrum and the absorption C

spectra of the three strains (the overlap 8-

integral). This product was obtained by digitization of the spectra followed by point-by-point multiplication, and finally the integration was shown. The overlap integral for the three strains, after being normalized by the value of the wild type, gave values of 0.93, 1.00, and 1.08 for RS103, RS2, and RS104, respectively.

To obtain relative efficiencies we can combine the terms &c, &, and the overlap integral into a single term, B, if we retain a normalizing factor, X, which is used to account for the small differences above. We have also included an intensity factor, 1, to include the effect of the neutral- density filters which would decrease the intensity of the flash lamp spectrum uni- formly because of their relatively flat ab- sorbance over the region from 720 to 900 nm. Combining all of these considerations, we obtained then integrated as in (B).

FIG. 5. Measurement of the overlap integral for RS103, RS104, and RS2. (A) The absorption spectra were obtained at 5 pM BChl for all three strains. The cells were grown at approximately 10 W m-‘. The emission spectrum of the xenon arc flash lamp (dotted line) is shown in relative units, and was obtained as described under Material and Methods. (B) The absorption spectra shown in (A) were inte- grated by a sequential summation procedure and plotted. (C) The absorption spectrum of each strain was digitized and multiplied, point by point, by the digitized emission spectrum of the flash lamp, and

136 MEINHARDT ET AL.

RSA04

ENERGY TRANSFER IN R. sphuertides LIGHT-HARVESTING MUTANTS 137

R = &IBNX. [51 By considering the case when the response is half-maximal and rearranging Eq. [5], we obtain

PI

where Iso is the flash lamp intensity nec- essary for half-maximal response. Since all experiments are normalized to the same maximal response, as in Fig. 4B, we can combine several of the terms into a single proportionality constant as shown in

Applying Eq. [7] to Eq. [6] for the special case of 50% maximal response, we obtain

tit = K.1 I&vX’ PI

Through the use of saturation curves as shown in Fig. 4B and from Eq. [3] we can obtain l/& values. From measurements of the total amount of BChl in the chro- matophores and the reaction center con- centration we can obtain the number of absorbing particles, N (the number of BChl per reaction center). By this proce- dure we are able to obtain the relative efficiency of transfer for each bacterial strain, and at the same time we can compensate for large changes in the pho- tosynthetic unit size.

We have obtained saturation curves for chromatophores from all three bacterial strains having a wide range of BChl to reaction center ratios. This variation was obtained by growing cells at different light intensities. Figure 6 shows a plot of the results from these experiments. The wild-type strain had a relative efficiency of 18.4 f 0.6, and this value was constant for a large variation in the photosynthetic unit size. The B800-850- mutant, RS104, displayed a similar, although perhaps slightly higher, relative efficiency of 18.9 + 1.1. Although in Fig. 4A, RS104 appeared to have a lower efficiency than the wild- type strain, when the photosynthetic unit size (BChl per reaction center) is taken into account the two strains have essen-

5

i I I I I I I I I

60 120 100 240 BChl IRC

FIG. 6. Efficiency of reaction center turnover versus photosynthetic unit size. The points were obtained from experiments such as those described in Fig. 4, and were calculated according to the equations in the text. Solid symbols represent 5 pM BChl; other- wise, data was obtained at 18 pM BChl. The numbers were multiplied by 100 to give the relative efficiencies plotted here. BChl to reaction center ratios were obtained as described under Materials and Methods. Same symbols as in Fig. 4.

tially the same efficiency of transfer to the reaction center (Fig. 5). The B875- mutant, RS103, showed an average effi- ciency of 4.5 + 1.9, which is about 24% that of the wild type and RS104. There also appeared to be a tendency toward higher efficiencies in RS103 at lower BChl to reaction center ratios, but this could not be adequately determined due to the scatter in the data. Also note that the BChl to reaction center ratio was consis- tently higher in RS103 than in RS2, even when the latter were grown at high light intensities.

The results which we have obtained can be easily explained by the model proposed by Monger and Parson (9). In cells which possess the B8’75 complex (i.e., RS104 and RS2), the efficiency of energy transfer was the same since the energy was be- ing transferred either directly from or through the B875 complex to the reaction center. The difference in the growth rates observed between the wild-type strain and the B800-850- mutant can then be ex-

138 MEINHARDT ET AL.

plained by the difference in the size of the photosynthetic unit at lower light intensities. When this difference is ac- counted for, as in our calculations, the two strains showed the same efficiency of energy transfer. The B875- mutant, RS103, was much less efficient in transferring energy to the reaction center at all light intensities, even though it possessed a larger photosynthetic unit (Fig. 6). This result is consistent with the model in which the B800-850 complex must transfer its energy through the B875 complex for efficient energy transfer to occur. We could not determine from our experiments if there was any direct energy transfer from the B800-850 complexes to the reaction centers in the B875- mutant. However, much of the light energy absorbed by the B800-850 complexes appears to be emitted as fluorescence. Some of this emitted light could, in turn, be absorbed by the BChl molecules in the reaction centers and so provide a very inefficient mechanism of reaction center turnover. Alternatively, part of the efficiency we observe may be a result of reaction centers being excited directly by the light. In this case, no energy would be transferred from the B800-850 complex, which would explain why the efficiency did not increase with more BChl per reaction center (Fig. 5). Actually, the B800-850 complexes should screen increasingly more of the light from the reaction centers under these condi- tions, although we were unable to estab- lish this effect due to the scatter in our data (Fig. 5). It is probable that the efficiency that we have measured for the B8’75- mutant was due to a combination of a number of processes which provide only a minor contribution to energy transfer process under normal conditions in wild type cells.

CONCLUSIONS

We have characterized the energy transfer capability of mutants of R. sphaeroides containing only the B875 com- plex or the B800-850 complex. Energy transfer from the B875 complex, and from the B800-850 complex through the B875

complex, to the reaction center have the same relative efficiency. In the absence of the B875 complex, transfer of energy from the B800-850 complex is reduced to ap- proximately 24% of the normal efficiency, and perhaps in part represents photon absorption by the reaction center directly and only a minor amount of energy trans- fer from the B800-850 complex. These results support the model proposed by Monger and Parson (9) for the mechanism of energy transfer from the light-har- vesting complexes to reaction center com- plexes. In this model, energy absorbed by the B800-850 complex is transferred to the reaction center through the B875 com- plex. Therefore, these mutants which we have characterized should be very useful in studying the processes involved in the transfer of energy from the light-har- vesting complexes to the reaction centers as well as the organization of these com- plexes within the photosynthetic mem- brane.

ACKNOWLEDGMENTS

We thank Ms. Kally Webster and Mary Hadden for their technical assistance. We also thank Dr. Colin Wraight and Dr. Don DeVault for their helpful discussions. This work was supported by NIH Grants PHS R01 GM26305 and GM15590.

REFERENCES

1. KAPLAN, S., AND ARNTZEN, C. J. (1982) in Pho- tosynthesis: Energy Conversion by Plants and Bacteria (Govindjee, ed.), Vol. 1, pp. 65-151, Academic Press, New York.

2. AMESZ, J. (1978) in The Photosynthetic Bacteria (Clayton, R. K., and Sistrom, W. R., eds.), pp. 333-338, Plenum, New York.

3. ZANKEL, K. L. (1978) in The Photosynthetic Bac- teria (Clayton, R. K., and Sistrom, W. R., eds.), pp. 341-346, Plenum, New York.

4. PEARLSTEIN, R. M. (1983) in Photosynthesis: En- ergy Conversion by Plants and Bacteria (Gov- indjee, ed.), Vol. 1, pp. 293-330, Academic Press, New York.

5. COGDELL, R. J., AND THORNBER, J. P. (1980) FEBS Lett 122. 1-8.

6. AAGARD, J., AND SISTROM, W. R. (1972) Photo&em Photobiol 15, 209-225.

7. VAN GRONDELLE, R., HUNTER, C. N., BAKKER, J. G. C., AND KRAMER, H. J. M. (1983) Biochim Biophys. Acta 723, 30-36.

ENERGY TRANSFER IN R. sphaeroides LIGHT-HARVESTING MUTANTS 139

8. ZANKEL, K. L., AND CLAYTON, R. K. (1967) Pho- tochem Photobiol 9, 7-15.

9. MONGER, T. G, AND PARSON, W. W. (1977) B&him. Biophys. Acta 460, 393-407.

10. SAUER, K., AND AUSTIN, L. A. (1978) Biochemistry 17,2011-2019.

11. HEATHCOTE, P., AND CLAYTON, R. K. (1977) B&him. Biophys. Acta 459, 506-515.

12. LUEKING, D. R., FRALEY, R. T., AND KAPLAN, S. (1978) J. BioL Chem. 253, 451-457.

13. DEVAULT, D. (1978) in Methods in Enzymology (Fleischer, S, and Packer, L., eds.), Vol. 54, pp. 32-46, Academic Press, New York.

14. BOWYER, J. R., MEINHARDT, S. W., TIERNEY, G. V., AND CROFTS, A. R. (1981) B&him. Biophys. Acta 635, 167-186.

15. CLAYTON, R. K. (1963) in Bacterial Photosynthesis (Gest, H., San Pietro, A., and Vernon, L. P., eds.), pp. 397-412, Antioch Press, Yellow Springs, Colorado.

16. CLAYTON, R. K., AND CLAYTON, B. J. (1981) Proc. Nat1 Acad. Sci. USA 78, 5538-5587.

17. SCHMIDT, S., REICH, R., AND WITT, H. T. (1971) in Proceedings, Second Congress on Photosyn- thesis (Forti, G., Auron, M., and Melandri, B. A., eds.), pp. 1087-1095, Dr. W. Junk, The Hague.

18. MATSUURA, K., ISHIKAWA, T., AND NISHIMURA, M. (1980) Biochim. Biophys. Acta 590, 339- 344.

19. HOLMES, N. G., HUNTER, C. N., NIEDEKMAN, R. A., AND CROFTS, A. R. (1980) FEBS Lett. 115, 43-48.

20. JOLCHINE, G., AND REISS-HUSSON, F. (1983) Phu tobiochem. Photobiophys. 6, 75-84.

21. SEBBAN, P., AND MOYA, I. (1983) Biochim. Biqphys. Acta 722, 436-442.

22. FEICK, R., VAN GRONDELLE, R., RIJGERSBERG, C. P., AND DREWS, G. (1980) B&him. Biophys. Acta 593, 241-253.