() Photochemistry and Photobiology, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA1999, 69(5) 599 604 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Light dependent and Biochemical Propert[.]
zyxwvuts zyxwvu Photochemistry and Photobiology, 1999,69(5): 599-604 Light-dependent and Biochemical Properties of Two Different Bands of Bacteriorhodopsin Isolated on Phenyl-Sepharose CL-46 Francesco Lopezi, Simona Lobasso2,Matilde Colella2,Angela Agostiano113and Angela C~rcelli**,~ zyxwvutsr zy 'Dipartimento di Chimica and zDipartimentodi Fisiologia Generale ed Ambientale, Universita di Bari, Bari, Italy and 3Centro Studi Chimico Fisici sull'lnterazione Luce-Materia, CNR, Bari, Italy Received January 1999; accepted 19 February 1999 ABSTRACT We report a detailed description of the light-dependent and biochemical properties of two different bands of isolated and nearly delipidated bacteriorhodopsin obtained from chromatography on phenyl-Sepharose CL4B The two bands (BR I and BR 11) showed a number of markedly different spectroscopic and biochemical characteristics: different absorption maximums in the dark, different lighvdark adaptations, different M decay kinetics, different stabilities, different responses to titration with alkali in the dark and different circular dichroism (CD) spectra Organic phosphate contents of BR I and BR I1 were measured; we found that more than 90% of purple membrane organic phosphate was removed in the course of chromatography and that the phospholipidprotein molar ratio was always higher in BR I than in BR 11 In many functional aspects (high stability, response to light adaptation, spectral changes in the dark by alkali addition and bilobate CD spectrum) the first band appeared to be similar to the purple membrane We suggest that the functional differences between the two bands depend on the fact that the first band (BR I) contains mostly bacteriorhodopsin aggregates corresponding to purple membrane trimers, while the second band (BR 11) contains only bacteriorhodopsin monomers INTRODUCTION chromatography has been used to obtain delipidated BR used in recent crystallization studies (3,4) Another method of BR isolation has been developed using high performance size exclusion chromatography (5) A partial characterization of delipidated BR properties in various detergents has been also obtained by Milder et al (6) Here we report data illustrating the characteristics of bacteriorhodopsin isolated by means of phenyl-Sepharose C L B chromatography As previously reported for halorhodopsin (7,8), we show that BR also splits into two bands on phenyl-Sepharose and that these bands have a number of remarkably different light-dependent and biochemical properties The appearance of two BR bands on phenyl-Sepharose C L B was previously mentioned (8) Recently, other investigators who solubilize wild-type and mutant BR for the purpose of crystallizing it have observed that two bacteriorhodopsin peaks come off from the phenyl-Sepharose column (J K Lanyi, personal communication), but in the literature there are not other studies illustrating the different properties of the two bands or explaining what the splitting represents Although the behavior of BR in micellar systems has been described in the past (9,10), in this paper we present new data illustrating in detail the properties of solubilized BR in octylglucoside and, furthermore, we show that with this new isolation method it is possible to isolate and distinguish nearly delipidated BR fractions in the form of monomers and trimers zyxw zyxwvutsr MATERIALS AND METHODS Solubilization and isolation of lipid-depleted bacteriorhodopsin (BR)? was first achieved by Wildenauer and Khorana (1) Solubilization of the purple membrane (PM) with Triton X-100 followed by gel filtration on agarose in deoxycholate solution resulted in removal of 99% of endogenous lipids (2) Solubilization with octylglucoside followed by agarose zyxwvut Materials An engineered L33 Halobacterium salinarum strain, kindly provided by Richard Needleman, was used in this study (1 1) The growth medium, containing neutralized peptone (L34, Oxoid) and novobiocin (from Serva, at final pg/mL), was prepared as previously described ( 12) DNase and n-octyl-P-glucopyranoside (octyl glucoside or OG) were obtained from Sigma, sodium cholate from Serva and phenyl-Sepharose CL-4B from Pharmacia PM isolation Purple membrane was isolated as previously described (13) A concentrated suspension of cells in M NaCl was dialyzed overnight; the dialysate was washed four times with distilled water by centrifuging and the pellet (in distilled water) was layered over a step gradient (60% , and 15% sucrose) and centrifuged 18 h 100000 g The purple band was collected and sucrose was removed by repeated washing Finally, the collected PM was frozen (-20°C) BR isolation Membrane solubilization was performed by incubating 12 mg of PM at room temperature for 64 h with mL 0.1 M Na acetate, 1% Triton X-100 (pH ) An aliquot of the extract (about 55 nmol of BR) was diluted 10-fold with 0.4% sodium cholate buffer containing 0.1 M NaC1 buffered with 25 mM Tris/HCl (pH 7.2) and then loaded (0.2 mL/min) over a X cm column of *To whom correspondence should be addressed at: Dipartimento di Fisiologia Generale ed Ambientale, UniversitA degli Studi di Bari, Via Amendola 165/a 70126 Bari, Italy Fax: 39 80 544 3388; e-mail: a.corcelli@biologia.uniba.it tdbbreviations: BR, bacteriorhodopsin, BR I, first eluted band from phenyl-Sepharose chromatography; BR 11, second eluted band from phenyl-Sepharose chromatography; CD, circular dichroism; DA, dark-adapted; LA, light-adapted; OG, n-octyl-P-glucopyranoside; PGP-Me, phosphatidylglycerol phosphate methyl ester; PGS, phosphatidylglycerol sulfate; PM, purple membrane I999 American Society for Photobiology 003 1-8655/99 $5.00+0.00 zyx zyxwvut zyxwvutsrqpon 599 zy zyxwvut zyxwvutsrqpo zyxwvutsr zyxwvutsrq zyxwvutsrq zyxwvutsrqponmlk 600 Francesco Lopez et a/ n E 0.14 p p 0.4 v 0.12 C 0.3 Q) * 10 fractions 15 20 b: - 0.1 0.08 s: 0.06 0.04 zyxwvutsrqp u P ul- 0.02 350 Figure Chromatography of BR on phenyl-Sepharose CL-4B gel BR was solubilized in 1% Triton X-100 diluted in cholate buffer to obtain absorption on phenyl-Sepharose during loading and eluted in 0.5% octyl glucoside buffer (pH 7.2) 450 550 650 350 nm 550 450 650 nm zyxwvutsrqponm zyxwvutsrq AA phenyl-Sepharose CL-4B (2.5 mL bed volume) previously equilibrated with the same buffer I t was washed with 70 mL of cholate buffer (0.8 mL/min) and then eluted with octyl glucoside buffer containing M NaCI 25 mM Tris/HCI (pH 7.2) and 0.5% octyl glucoside (0.3 mL/min) Chromatography was performed at room temperature under dim light; collection of the colored fractions ( I -5 mL) started after elution of about 12 mL of octyl glucoside buffer Total recovery of BR from the column was 90-100%% The absorptions at 560 and 280 nm of each BR-containing fraction were measured spectrophotometrically (A28dA5601.3-1.6): BR fractions were stored in the dark at -20°C The same chromatography was repeated at pH using 0.1 M sodium acetate buffer instead of Tris/HCl in the washing and eluting solutions Pho.~phorusqicrintitatictn The content of organic phosphorus was measured using the method of Bartlett (14) Ab.rorptinn spectrosc-opy UV-visible absorption spectra were obtained with a Cary UV-visible spectrophotorneter Dark-adapted ( D A ) spectra were obtained on samples stored at room temperature in complete darkness for 12 h Light adaptation of PM and BR was accomplished using a 150 W illuminator A cuvette containing sample was irradiated for at a distance of 15 cm from the bulb The rate of dark adaptation of BR was measured by following absorption decrease at the wavelength in the visible region with the largest absorption difference between light-adapted (LA) and DA forms M decay The kinetics of M decay was followed using an Optical Multichannel ‘4nalyzer (OMA 111, model 1460 EG&G Princeton, N Y ) equipped with an EG&G Red Intensified Diode Array Detector model 1430-512-G Actinic flashes were provided by an EG&G xenon lamp (3.35 J discharge energy) and screened through two layers of Wratten gelatin filter giving a light pulse of ps duration at halfmaximal intensity The probe beam was screened with a blue (300500 nm) filter Deconvolution of multiexponential decays was performed by computer routines based on the Marquardt algorithm Samples for the kinetic studies were thermostatted (20°C) All samples were LA before measurements were taken Circulur dichroisnt Circular dichroism (CD) spectra were obtained with a Jasco spectropolarimeter; BR samples in octyl glucoside buffer having absorption visible in the range 0.5-0.8 were analyzed 0.02 0.01 -0.01 -0.02 350 450 550 nm 650 Figure Dark (solid line) and light (dotted line) adapted spectra of (a) BR I and (b) BR 11 c: Light-dark difference spectra of BR I (dotted line) and BR I1 (soliddotted line) are compared with the difference spectrum of PM in water (solid line) late buffer to remove PM lipids and then eluted by shifting to octyl glucoside buffer Figure illustrates the chromatographic profile of solubilized BR eluted on phenyl-Sepharose CL-4B in octyl glucoside buffer at pH 7.2 It can be seen that two different BR bands are obtained (BR I and BR 11) and that BR I represents about 30% of total loaded protein The ratio between the areas of the two bands does not change when half or double protein amounts are loaded on the column Furthermore, the profile was not significantly modified by running the chromatography at pH At lower pH the protein tends to remain bound to the gel longer and very broad bands are obtained In the following we illustrate the biochemical and spectroscopic characteristics of the two bands obtained at pH 7.2 No difference in the chromatographic profile was found between isolation procedures performed in room light and in the dark RESULTS Light/dark adaptation and M decay of the two bands BR splits in two bands in phenyl-Sepharose CL4B chromatography Figure 2a,b illustrates spectra in the dark and after light adaptation of BR I and BR I1 isolated at pH 7.2 at 22°C Remarkable differences in the dark spectra as well as in the light adaptation phenomena of BR I and BR I1 can be seen In the dark, the absorption maximum of BR I is at 555 nm, while the absorption maximum of BR I1 is blue-shifted at about 552 nm Furthermore, in the case of BR 11, besides the 552 nm peak another peak at 380 nm is present; the We isolated two different bacteriorhodopsin bands on phenyl-Sepharose gel by using essentially the same protocol developed years ago by Duschl et al ( 15) to isolate and purify halorhodopsin Purple membranes were solubilized in Triton and diluted with cholate buffer to obtain BR binding on phenyl-Sepharose BR was extensively washed on gel with cho- zy zyxwvutsrqpon zyxw zyxwvuts zyxwvutsrqp zyxwvuts zyx Photochemistry and Photobiology, 1999, 69(5) 601 Table Lightldark absorbance maximum of PM, Ttiton-solubilized PM and BR r (nm) XIa? (nm) 560 550 555 552 568 558 562 560 Ada* AA (nm) AA 8 I fll 0.8 h e 0.6 % 0.4 PM Solubilized PM BR I BR I1 “da = dark-adapted tla = light-adapted -6 t9 0.2 -9 50 100 150 200 time (min) Figure Kinetics of dark adaptation of ( ) BR and Both samples were at concentrations of about 3.5 k M height of this peak increases together with a decrease of absorption in the visible region when BR 11 is left at room temperature even for only a few hours The spontaneous conversion of BR into BR I1380can be reversed by lowering the pH to (not shown) BR I is instead quite stable and has only a negligible tendency to form the 380 nm species By keeping the sample at 0°C in the dark, either no formation of the 380 nm species or no increase of the 2801 560 absorption ratio was observed for BR I over a period of months from isolation, while in the same time the second band showed a 30% decrease in absorption at the peak in the visible region Interestingly, the extent of light adaptation of BR I is similar to that of PM, as after illumination a red shift of absorption maximum is observed, together with an increase in the absorption value (of about 10%); in contrast, light adaptation of BR I1 is considerably reduced The similarity in light adaptation between BR I and PM is also evident from an examination of difference spectra of BR I and BR I1 reported together with the difference spectrum of PM in water in Fig 2c On the other hand, the light-dark difference spectrum of BR I1 resembles that of alkalinized PM (not shown in Fig 2C) Table reports dark and light lambda max and the percentage changes in absorptions after illumination of PM, Triton solubilized PM and the two BR bands; the values are means obtained from several different spectra The return to the dark state is also different for BR I and BR 11 The process of dark adaptation for BR I can be well described by a biexponential equation, while dark adaptation of BR I1 is slower and difficult to fit, probably because it occurs together with a loss of absorption, i.e with some protein degradation due to instability of this BR fraction (see Fig 3) Figure illustrates the M decay of both BR I and BR 11; it can be seen that M decay of both BR forms is slower than that of PM Since BR I and BR I1 are expected to be delipidated as a consequence of the isolation procedure (see below), this result is in agreement with previous data obtained for delipidated PM (16,17); in both cases the traces are well fitted by biexponential equations whose life times are given in Fig Furthermore, it can be seen that M decay of BR I1 is faster than that of BR I Literature data describing the dependence of M decay kinetics on the flash intensity and the cooperative effect caused by a different excitation of monomers and trimers in BR can help to interpret this result (18), but our data are too preliminary to allow conclusions on the differences in the photocycle intermediates of BR I and BR 11 (0) BR 11 Different responses to alkali addition in the dark Figure (a,b) shows the absorption spectra of BR and BR I1 as a function of pH in the pH range 7.2-11; the pH was adjusted in the dark by addition of 1-5 pL aliquots of 0.1 N NaOH Figure (c-f) shows the difference spectra between the high pH spectra and the spectrum at pH 7.2 During the titration, the formation of three peaks at 460, 380 and 365 nm can be seen for BR I, while only the 380 and 365 nm peaks are seen for BR 11 In a narrow pH range, the spectral changes induced by alkalinization can be partially reversed for both BR I and BR I1 upon acidification (not shown) Furthermore, by plotting residual absorption in the visible region at increasing pH, it can be seen that BR I is more resistant to alkali addition than BR II; in fact, bleaching of BR I is complete at pH 11, while BR I1 is already completely bleached at pH 10 (see Fig 6) As expected, the accessibility of Schiff base to external OH- is greatly increased in both BR I and BR I1 compared with PM Furthermore, these results indicate that BR I also shows characteristics similar to those of PM in the titration with alkali, while the inability of BR I1 to give rise to the 460 nm species upon alkalinization offers another example of similarity of this band to the so-called “alkaline BR” (19,20) The functional similarity between alkaline BR and BR I1 prepared by delipidation of BR could be due to the fact that PM titration with alkali could induce a weakening of the lipid-protein interactions mimicking the delipidation effects in BR 11 zyxwvutsrq n E c 0.8 0.6 0.4 0.2 zyxwvutsrq 0 100 200 300 400 500 time (ms) Figure Comparison of the M decay kinetics of BR I (solid line) and BR I1 (dotted line) after flash excitation The samples were at the following concentrations: BR I, 4.3 pM BR 11, KM zyxwvutsrqp zyxwvutsrq zyxwvutsr zyxwvutsrqponm 602 Francesco Lopez et a/ Y 2 2 a, 0.1 0.0s 0.06 0.04 0.02 0.02 0.01 Table Organic phosphorus content of PM, Triton-solubilized PM and BR* zyxwvutsrqpo Phosphorus (nmol) PM 0.08 BR I BR I1 0.04 f0, o -0.04 -0.08 PhospholipidslBR molar ratio 6.0 16 5.7 0.0 18 6.4 8.5 18 1.1 7.5 29 0.5 zyxwvutsrqpo zyxwvutsrqponml Solubilized PM f -0.01 a -0.02 -0.03 -0.04 150.3 -C (12) 197.9 C (4) 31.9 (6 ) 21.4 (6) BR (nmol) *Phosphorus content is given as mean C SD of different determinations (number reported in parentheses) BR amount has been estimated by using A,,, ( e = 63 000) for PM, A560(E = 54 000) for Triton-solubilized PM and A,,,, ( e = 66000) for BR and BR 11 The phospholipidsBR molar ratio has been estimated considering that 30% of total phospholipids is represented by phosphatidylglycerosulfate (1 P/lipid) and the remaining 80% by phosphatidylglycerolinethylphosphate (3 P/lipid) zyxwvutsrqpon zyxwvutsrqp zyxwvutsrqp 400 500 nm 400 600 500 600 nm Figure a and b: Abwrption spectra of ( a ) BR I and t b ) BR I1 samples as a function of pH at room temperature The samples were DA at pH 7.2 and then adjusted to a given pH by the addition of NaOH I n the order o f decrease in absorption at the maximum: a: I pH 7.1: pH 7.58: pH 8.02; pH 8.39: , pH 8.87; 6, pH 9.27: 7, pH 10.29: pH 10.67: pH 10.81: 10 pH 11.04, b: 1, pH 7.1; pH 7.61 pH 8.07: 4, p H 8.41: pH 8.92; pH 9.31; 7, pH 10.29: , pH 10.68 c-f: Difference absorption spectra of (c and ci BR I and (ct and f ) UR I1 samples at pH, minus that at pH 7.2 c : I, pH, 7.58: pH, X.02; pH, 8.39; pH, 8.87 d: pH, 7.61: pH, 8.07: pH, 8.12: pH, 8.91 e: pH, 9.27; 6, pH, 10.39: pH, 10.67: pH, 10.87: pH, 11.04 f pH, 9.32; pH, 10.19: 7, pH, 10.68 Optical densities of both BR I and BR I1 at 550 nni in the dark ;it p H 7.2 were 0.09: pathlength of the cuvette I cm Residual phospholipids associated with the two bands Another important question to address is that of the lipid content of BR fractions eluted from the phenyl-Sepharose column Table reports the lipid phosphorus content together with the estimated phospholipid/BR molar ratio for the two bands The ratio values have been obtained taking into consideration that about 80% of PM phospholipids contain two phosphate groups per lipid molecule (as in the case of phosphatidylglycerol phosphate methyl ester [PGP-Me]) and that the remaining 20% contains only one phosphate group per lipid molecule (as in the case of phosphatidylglycerol sulfate [PGS]) At least one phospholipid molecule is still associated with BR I, while only one phospholipid per two BR molecules are available on average in BR 11 CD As BR I was found to have light adaptation similar to that of PM, we thought it reasonable to check for the presence of an exciton coupling effect in the CD spectrum of BR I Figure 7a,b shows the CD spectra of BR I and BR 11 We readjusted the isolation protocol in ordcr to obtain more concentrated BR I samples suitable for CD analyses in the visible region Both samples used in the experiment in Fig had absorbance at 560 nm of about 0.7; attempts to obtain more concentrated BR I fractions were unsuccessful Despite the noise, it can be seen that the CD spectrum of BR I is clearly bilobate, indicating interaction between retinal chromophores on adjacent BR molecules, as it likely occurs in PM trimers However, the possibility that the exciton coupling of BR I could be due to aggregates different from trimers (in particular, dimers) cannot be excluded (21) Only a positive band is instead present in the CD spectrum 1.5 10 11 PH Figure Titration curves of (*I BR I and ( ) BR 11 Relative absorption i n the visible region is plotted \'ersii.s increasing pH: data of t w o diffcrcnt titraiion experiments are reported zyxwvu -1.5 450 500 550 600 650 450 500 550 600 650 nm nm Figure Visible CD spectra for (a) BR and (b) BR 11 Optical densities at 560 nm of BR I and BR I1 were, respectively 0.685 and 0.616 zy zyxwvutsrqpon Photochemistry and Photobiology, 1999, 69(5) 603 Table Summary of the biochemical and spectroscopic properties of BR I and BR I1 zyxwvutsrqp zyxwvutsrqp Dark BR I not exclude the presence of residual glycolipids in the both BR I and BR 11 It has been reported that only 45% of glycolipids are removed by treatment of PM with Triton X-100 (25) Furthermore, recent data have indicated that both phospholipids and glycolipids are associated to delipidated BR fractions isolated with a different isolation method and used in crystallization studies (26) BR I and BR I1 could represent two different BWphospholipid complexes or aggregates already present in different proportions in the Triton solubilized membranes as a result of incomplete solubilization or as a consequence of an intrinsic heterogeneity of the PM This last possibility would be suggested by the fact that the ratio between the areas of BR I and BR I1 was found to be constant by changing the experimental conditions, e.g by lowering the pH of chromatography buffers, by increasing the Tritodprotein ratio in the solubilization step or by changing the conditions of dilution with cholate (not shown) However, it is also possible that although a basic homogeneity is present in the PM, the two different species of BR are produced in the course of the delipidation process due to the fact that it is extremely difficult to remove the last remaining phospholipid and that what is recovered is the residual last two components remaining in the conversion of BR I to BR I1 by phospholipid removal The nature of the residual phospholipid(s) bound to BR I and BR I1 is presently under investigation In conclusion, the BR isolation method described in this report offers the advantage of separating monomers and trimers of BR, resulting in final homogeneous BR preparations, which can be potentially useful in crystallographic studies Light h,, red shift AA > 555 nm BR I1 A, PhosphoCD lipids/ bilobate Stability BR Yes High 21 zy zyxwvutsrq zyxwvutsr blue-shifted red shift No Low