COMPLEMENTARY CHROMATIC ADAPTATION IN A …

COMPLEMENTARY CHROMATIC ADAPTATION

IN A FILAMENTOUS BLUE-GREEN ALGA

ALLEN BENNETT and LAWRENCE BOGORAD

From The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02'138 . 1)r .Bennett's present address is the Biology Department, Brookhaven National Laboratory, Upton,Long Island, New York 11978 .

ABSTRACT

Fluorescent and red light environments generate greatly different patterns of pigmentationand morphology in Fremyella diplosiphon . Most strikingly, red-illuminated cultures containno measurable C-phycoerythrin and have a mean filament length about 10 times shorterthan fluorescent-illuminated cultures . C-phycoerythrin behaves as a photoinducible con-stituent of this alga. Spectrophotometric and immunochemical procedures were devisedso that C-phycoerythrin metabolism could be studied quantitatively with [14C]-phenylala-nine pulse-chased cultures . Transfer of red-illuminated cultures to fluorescent light initiatesC-phycoerythrin production by essentially de novo synthesis. C-phycoerythrin is not de-graded to any significant extent in cultures continuously illuminated with fluorescent light .Transfer of fluorescent-illuminated cultures to red light causes an abrupt cessation ofC-phycoerythrin synthesis. The C-phycoerythrin content of cultures adapting to red lightdecreases and subsequently becomes constant . Loss of C-phycoerythrin is not brought aboutby metabolic degradation, but rather by a decrease in mean filament length which iseffected by transcellular breakage . In this experimental system, light influences intra-cellular C-phycoerythrin levels by regulating the rate of synthesis of the chromoprotein .

INTRODUCTION

Environmental conditions exert a strong influenceupon the pigment composition of many algae(Halldal, 1970) . Complementary chromatic adap-tation is a spectacular response of some blue-greenand red algae to alterations in the energy distribu-tion in the visible light environment . As a conse-quence of this phenomenon, the pigment whichabsorbs the incident wavelengths of light moststrongly becomes predominant. Gaidukov (1902)first described complementary chromatic adapta-tion in blue-green algae . He observed that Oscilla-toria sancta assumed a red coloration after growthunder green light and a blue-green tint aftergrowth under orange light. Gaidukov (1902)ascribed these color variations to the synthesis of

different kinds of pigments . Kylin (1912) andBoresch (1919, 1921), further, correctly perceivedthat such color changes in blue-green algae areprimarily a specific consequence of alterations inthe relative proportions of the red and blue phyco-biliprotein constituents . The demonstration thatphycobiliproteins function as accessory pigments(Engelmann, 1883; Emerson and Lewis, 1942 ;Haxo and Blinks, 1950; Duysens, 1951 ; French andYoung, 1952) provided a rationale for complemen-tary chromatic adaptation : alterations in the levelsof these chromoproteins permitted maximal utili-zation of the available light energy for photosyn-thetic purposes .

The principal concern of the work reported in

THE JOURNAL OF CELL BIOLOGY • VOLUME 58, 1973 • pages 4 1 9-435

419

Dow

nloaded from http://rupress.org/jcb/article-pdf/58/2/419/1386619/419.pdf by guest on 30 D

ecember 2021

this communication was to define, in broad meta-bolic terms, how light influences C-phycoerythrin(PE)' metabolism during complementary chro-matic adaptation in Fremyella diplosiphon . The netintracellular levels of individual phycobiliproteinsmay decrease during complementary chromaticadaptation in growing blue-green algae . Suchlight-induced decreases could be brought abouteither by acceleration in the rate of phycobilipro-tein degradation or by deceleration in the rate ofphycobiliprotein synthesis and dilution withgrowth. It is impossible a priori to distinguish be-tween these alternatives . Experimentally, "syn-thesis-versus-degradation" problems have beenmost readily approached with inducible cellularconstituents, such as 0-galactosidase in Escherichiacoli (see Hogness et al ., 1955). The levels of in-ducible components are very low in the absence ofappropriate enviror4nental stimuli. Spectral re-sponse curves for complementary chromaticadaptation in Tolypothrix tenuis (Fujita and Hattori,1962) indicated that quanta with wavelengthsgreater than about 585 nm do not stimulate PEsynthesis . This suggested that it might be possibleto utilize restricted conditions of illumination togive PE the status of an inducible cell constituentduring complementary chromatic adaptation .

The filamentous blue-green alga F. diplosiphonwas selected for these experiments because it showsstriking complementary chromatic adaptation andbecause its trichomes, unlike those of T. tenuis andmany other filamentous blue-green algae, do notclump together into large masses or adhere to thesides of the culture flasks . This latter feature greatlyreduces the variability between replicate samplesobtained from liquid cultures .

Numerous differences in pigmentation andmorphology exist between F. diplosiphon adapted tofluorescent and to red light . The kinetics of thechanges in pigmentation and morphology whichoccur in Fremyella during complementary chro-matic adaptation were studied by transferring fluo-rescent light-adapted cultures to red light, and viceversa. The metabolic stability of PE was investi-gated in growing cultures adapted to fluorescentillumination and also during the adaptation of

I Abbreviations used in this paper : APC, allophycocyanin ;chi a, chlorophyll a ; IgG, y-globulin ; PC, C-phyco-cyanin ; PE, C-phycoerythrin ; PMSF, phenylmethyl-sulfonyl fluoride ; R av , at average radius of centrifu-gation ; SDS, sodium dodecyl sulfate; Sequestrene,sodium ferric diethylenetriamine pentaacetate .

420

THE JOURNAL OF CELL BIOLOGY • VOLUME 58, 1973

fluorescent-illuminated cultures to red light. Theseexperiments made use of a microscale purificationprocedure to obtain PE for specific activity meas-urements from pulse-chased cultures . It is shownthat fluorescent light may be utilized to "induce"PE production in red light-adapted cultures, andthat red light may be employed to turn off PEsynthesis in fluorescent light-adapted cells .

EXPERIMENTAL SECTION

General Culture Conditions

F. diplosiphon (B. and F .) Drouet (strain 481) wasobtained in axenic culture from the collection at In -diana University (Starr, 1964) . Dr . F. Drouet (per-sonal communication) now considers F. diplosiphon tobe a growth form of Calothrix parietina (Nag.) Thur .The organism was grown at 32 f 2 °C in cotton-plugged, shaker-agitated, 2800-ml Fernbach flasks inmodified Medium C (Kratz and Myers, 1955) . So-dium ferric diethylenetriamine pentaacetate (Seques-trene) (Geigy Chemical Corp ., Ardsley, N. Y.) wasadded to a final concentration of 40 mg/liter of me-dium as a substitute for Fe2(SO4)3-6H20, and theKNO3 concentration was increased to 2 g/liter . ThepH of the medium was adjusted to 7 .0 with 1 NNaOH before autoclaving . Fluorescent illumination(250 foot candles [ft-c]) was provided by a bank ofSylvania F 30 T12-W-RS lamps . Incandescent illu-mination (100 ft-c) was derived from a rack of 60-W,120-V Sylvania bulbs . Red illumination (26 ft-c) wasobtained by interposing a 1/8-inch thick EvrKleertransparent amber cast acrylic filter (no . 2422, Rohmand Haas Co., Philadelphia, Pa .) between GeneralElectric 60A21 /R red lamps and the culture flasks .

Light Measurements

Absorption spectra of whole cells and crude ex-tracts were determined with a Cary model 14 record-ing spectrophotometer equipped with a model 1462scattered transmission accessory (Cary Instruments,Monrovia, Calif.) . Light-source emission spectra weremeasured with an ISCO spectroradiometer (Instru-mentation Specialties Co ., Lincoln, Neb .) . Incidentlight intensities at the shaker surfaces were deter-mined either by manual integration of light-sourceemission spectra from 400 to 730 nm or with a Westonmodel 756 sunlight illumination meter (Weston Elec-trical Instrument Corp ., Newark, N .J .) .

Culture and Harvesting Procedures forMetabolism Experiments

Pulse-chased, fluorescent light-adapted cultures formetabolic experiments were initiated by adding 15

Dow


ecember 2021

µCi of L-[ 14C]phenylalanine and a small liquid inocu-lum containing 3-4 mg dry weight of healthy incan-descent light-grown cells to 500 ml of medium . Thecultures were allowed to proliferate under fluorescentillumination until the cells in I ml of medium con-tained about 5 .1 pg of chlorophyll a (chi a) . Freshmedium (500 ml) containing a 2,000-fold molar ex-cess of unlabeled L-phenylalanine was then added tothe cultures. Chasing under fluorescent illuminationproceeded until the chi a level again reached about5.1 pg/ml . The protocol for preparing red light-adapted, pulse-chased cultures was identical, exceptthat (a) inocula were obtained from red light-adaptedcultures, (b) the cultures were grown under red light,and (c) the chase with unlabeled L-phenylalanine wasstarted when a chi a level of 3 .3 pg/ml was attained .Both L-[ 14C]phenylalanine (uniformly labeled) (NewEngland Nuclear [Boston, Mass.], NEC-284, sp act472 mCi/mmol, stated radiochemical purity >99.6%) and unlabeled L-phenylalanine (NutritionalBiochemicals Corp., Cleveland, Ohio) were dissolvedin 10 ml of medium and added to the culture flasksvia a sterile Swinnex-13 filter unit containing a 0.22pm filter (Millipore Corp ., Bedford, Mass .) .

At each sampling point, duplicate aliquots wereremoved for estimation of phycobiliproteins (20 ml),for estimation of chi a and total counts incorporated(5 ml), and for estimation of growth (2 ml), and weretransferred to graduated 15-m1 conical glass centrifugetubes. During sampling, sterile technique was usedand red light-illuminated cultures were manipulatedonly under red light or in darkness. Samples could beharvested and washed quantitatively if the filamentswere pelleted with an IEC model UV centrifuge(Damon/IEC Div., Needham Hts., Mass., no . 240head, 2,000 rpm, 5 min, 23 °C) and the supernatantswere carefully removed by hand with a Pasteurpipette. Harvested cell pellets were washed rapidlyeither three times (phycobiliprotein, growth aliquots)or four times (chi a aliquots) with 10 ml of distilledwater .

Chl a and Phycobiliprotein Quantitation

Blue-green algae contain chi a only (Bogorad,1962) . Chi a was quantitatively removed from washedcell pellets by extraction with two 5 .0-m1 aliquots ofabsolute methanol in dim light at about 23 °C. Thechi a content of the combined extracts was deter-mined from the optical density of the solution at 665nm. The extinction coefficient of chi a at 665 nm inabsolute methanol is 74.5 ml/mg-cm (MacKinney,1941) . Absorption spectra of methanol extracts indi-cated that pheophytinization did not occur to anysignificant extent during chi a extraction .

To estimate phycobiliprotein contents, washed cellpellets were resuspended to a total volume of 6 .0 mlwith 0.01 M sodium phosphate, pH 7 .0, 0 .15 M NaCl

(buffered saline). The cell suspensions were trans-ferred to Beckman no. 302235 tubes (Beckman In-struments, Inc., Spinco Div ., Palo Alto, Calif.), quick-frozen in liquid nitrogen, capped with Parafilm, andstored at -20°C in darkness before disruption .Thawed cell suspensions were immersed in a stirredice bath and sonicated with the microtip of a HeatSystems model W185D sonifier (Heat Systems-Ultra-sonics, Inc ., Plainview, N.Y.) operated at power set-ting 6 . Each aliquot of cells was exposed to 4 min ofsonication, with alternate 1-min periods of sonicationand cooling. Light microscope examinations demon-strated that this procedure quantitatively disrupts thecells . Over the cell concentration ranges used in theseexperiments, the amount of each water-soluble phyco-biliprotein released into the buffered saline solutionwas directly proportional to the number of cells dis-rupted. Sonicates were transferred to Beckman poly-carbonate tubes and centrifuged at 4 °C and 81,000 gat average radius of centrifugation (R ay) for 45 min .This procedure pellets the membrane fragments ofthe cells, which contain chi a and most of the carot-enoids . Phycobiliproteins could not be detected inmembrane fragments which were resuspended in buf-fered saline. The supernatant (crude extract) containsthe phycobiliproteins and other soluble proteins . Phy-cobiliproteins are responsible for essentially all of thevisible light absorption above 500 nm in the crudeextracts. The upper 5 .5 ml of the supernatants werecarefully withdrawn and their absorption spectraweredetermined from 730 to 500 nm . Insertion of theoptical densities of the crude extracts at 562, 615, and652 mn into equations derived from extinction co-efficients of the purified phycobiliproteins permittedthe PE, C-phycocyanin (PC), and allophycocyanin(APC) concentrations in the crude extracts to beascertained .

The extinction coefficient of each of the three F.diplosiphon phycobiliproteins, purified by the proce-dure of Bennett and Bogorad (1971), was determinedin buffered saline at 562, 615, and 652 run . Thesewavelengths correspond to the absorption maxima ofPE, PC, and APC, respectively, in buffered saline .This solvent was selected because it provides optimalconditions for precipitin reactions with rabbit anti-sera (Kabat and Mayer, 1967) . The method of Lowryet al . (1951) was used to estimate phycobiliproteinconcentrations for extinction coefficient calculationsand also to estimate total soluble protein concentra-tions in crude algal extracts . Standard curves for theLowry protein measurements were obtained fromsolutions prepared with cytochrome c which had beenthoroughly desiccated over P205 in vacuo. Equationsfor calculating phycobiliprotein concentrations incrude extracts were derived by combining the extinc-tion coefficients with three simultaneous equationshaving the form ODa = (ePE, XCPEI) + (EPC, a'

BENNETT AND BooORAD Chromatic Adaptation in a Blue-Green Alga

421

Dow


ecember 2021

CPci) + (EApe, XCAPCI), in which A equalled 562,615, and 652 nm (see Table I). All of the extinctioncoefficients were constant over phycobiliprotein con-centration ranges yielding optical densities from 0 .05to 1 .0 at their absorption maxima, and the absorptionspectra of mixtures of purified phycobiliproteins werestrictly additive .

Dry Weight and Growth Determinations

To estimate dry weights, washed cell pellets werequantitatively transferred to preweighed glass tubesand dried to constant weight in a vacuum oven at100 °C. When determining pigment fractions of thetotal cellular dry weight, duplicate aliquots of washedcells were prepared . These samples were analyzedquantitatively with regard to their phycobiliproteinand chi a content, and their dry weights were deter-mined directly . Turbidity measurements were madeto estimate growth during metabolic experiments .Washed cell pellets were brought to volume with dis-tilled water and the optical density of the cell suspen-sions was determined at 730 nm in 1-cm cuvets witha Gilford spectrophotometer (Gilford InstrumentLaboratories, Inc ., Oberlin, Ohio) . Under these con-ditions, cellular dry weight in mg/ml ~-_ 0 .265(OD730), between 0 .075 and 0 .275 OD7 36 units .

Immunological Procedures

Antibodies to each of the purified phycobilipro-teins were prepared for use in metabolic experimentsand in Ouchterlony double-diffusion tests . For pri-mary immunization, female New Zealand White rab-bits were each injected intracutaneously in the backwith 5-10 mg of purified phycobiliprotein in a thickemulsion prepared from equal volumes of Freund'scomplete adjuvant (Difco Laboratories, Detroit,Mich.) and buffered saline . After 6 wk, a secondaryresponse was effected by the injection of 3-5 mg ofpurified phycobiliprotein in buffered saline into themarginal ear vein . Major bleedings were carried out6-9 days after the secondary response was stimulatedif the antibody titers were sufficiently high, judgingfrom the three-tube test of Chase (1967) . Quantitativeprecipitin curves indicated that specific antibodylevels of 4-5 mg/ml of serum were usually elicited bythis immunization protocol . Antisera from the firstbleeding after the secondary response were used inall experiments .

Antibody purification was carried out at 4 °C. Glob-ulin fractions were obtained by 50% ammonium sul-fate fractionation (Goldman, 1968) . IgG fractionswere prepared by chromatography of globulin frac-tions on DEAE-cellulose following the procedure ofKoshland et al. (1962) . Globulin and IgG fractionwere dialyzed extensively against buffered saline,centrifuged at 10,000 g for 15 min, and stored in 5-ml

42 2 THE JOURNAL OF CELL BIOLOGY • VOLUME 58, 1973

TABLE I

Extinction Coefficients (ml/mg-cm) of Purified F .diplosiphon Phycobiliproteins in 0 .01 M SodiumPhosphate, pH 7 .0, 0.15 M NaCl and EquationsUtilized for Determining Phycobiliprotein Concen-trations (mg/ml) in Fremyella Crude Extracts Pre-pared in this Solvent

Extinction coefficientsPhycobili-protein

r = 562 nm

X = 615 nm a = 652 nm

PC 2.41 5.92 1 .23APC

0.849

2 .68

5 .65PE

9.62

Phycobiliprotein concentration equations

[PC] = OD615 - 0.474(OD652)5 .34

[APC] =ODe - 0.208(OD 6t;,)

5 .09

[PE] _OD ;,62 - 2.41(P C) - 0.849(APC)

9.62

aliquots at -20 °C. Globulin and IgG fractions wereused in Ouchterlony double-diffusion tests and pre-cipitin reactions, respectively .

Quantitative precipitin curves were determined ac-cording to Kabat and Mayer (1967, pp . 73-76) . Thetotal protein content of immunoprecipitates was esti-mated by a micro-Folin procedure (Kabat andMayer, 1967, pp. 557-558), using cytochrome c asstandard . Precipitin reactions were carried out inbuffered saline in conical 15-m1, graduated Kimblepolycarbonate centrifuge tubes (Kimble ProductsDiv ., Owens-Illinois, Inc ., Toledo, Ohio) . Precipitinmixtures were capped with Parafilm and left standingin the dark in an ice bath for 24-36 h before theprecipitates were harvested . Precipitates were washedtwice with two 6-ml volumes of ice-cold bufferedsaline . Quantitative precipitin reaction supernatantswere tested for residual soluble antigen (PE) with aBlak-Ray ultraviolet lamp (Ultra-Violet Products,Inc ., San Gabriel, Calif.) . Because of its intense orangefluorescence, PE may be unambiguously detected bythis procedure at concentrations greater than 0 .5µg/ml.

To prepare Ouchterlony double-diffusion plates,1 g Difco bacto agar and 20 mg Merthiolate weredissolved with stirring in 100 ml buffered saline,using a double boiler arrangement. Level, 0 .5% agar-coated Kodak slide projector plates (3% x 4 inches,Eastman Kodak Co ., Rochester, N .Y.) were eachoverlaid with 20 ml of the liquified agar solution .Antigen and antibody wells were cut into solidified

Dow


ecember 2021

agar layers with a no. 5 cork borer, and a distance of1 .8-1 .9 cm separated their centers . Plates were al-lowed to develop for 36 h at 23 ° C after the additionof 100 µl antibody and antigen aliquots, in bufferedsaline, to various wells. Results of double-diffusiontests were not qualitatively altered from those illus-trated when antigen concentrations were either in-creased or decreased 10-fold .

Preparation and Counting ofRadioactive Materials

Washed immunoprecipitates containing [14C]phen-ylalanine-labeled PE were carefully resuspended withseveral drops of 0.5 M NaOH and quantitativelytransferred to numbered Whatman filter paper disks(2 .4 cm diameter, no. 3 qualitative) . After drying,the disks were processed by the method of Mans andNovelli (1961) to selectively extract coprecipitatingnucleic acid and lipid components of the cell extracts .The 90 °C, 10% TCA extraction was limited to 15min. For total soluble protein radioactivity measure-ments, crude extracts were applied to filter paperdisks and processed identically .

The magnitude of radioactive "pools" remainingat the end of chase periods was estimated by extractingwashed cells twice with 3 .0 ml of ice-cold 5% TCAand determining the fraction of the total counts whichwere soluble . Total incorporated radioactivity meas-urements during metabolic experiments were deter-mined with cells which had been extracted withmethanol for chl a determinations. It was necessaryto utilize cells so extracted in order to eliminatecarotenoid quenching . Methanol-extracted cells andaliquots of 5% TCA extracts were transferred to glassvials, air-dried for 24 h, and processed for countingby the method of Tishler and Epstein (1968) . Thesesamples, as well as the nucleic acid and lipid extractedimmunoprecipitate and total soluble protein disks,were counted in a Picker 220 liquid scintillationcounter (Picker Corp., Cleveland, Ohio) after 10 mlof toluene-Omnifluor (New England Nuclear) solu-tion was added to the counting vials .

Filament Length Measurements

Small culture samples, obtained with a sterile in-oculating loop, were gently transferred to slides . Thefilaments were allowed to settle for several minutesand then photographed, without the use of a coverslip, at 15X on 35-mm film . Prints were made on8 X 10 inch paper . The lengths of at least 100 fila-ments from each sample were determined with aKeuffel and Esser no. 620305 map measure (Keuffel& Esser Co., Morristown, N.J .) . A stage micrometerscale was magnified to the same extent as the fila-ments in order to calibrate the map measure .

Immunoprecipitates Containing19C02-Labeled PE

Several methods were used to check anti-PE speci-ficity in precipitin reactions with crude extracts . Thespecific activity (ca . 500 cpm/µg) of PE from fluores-cent-illuminated, [14C]phenylalanine-labeled cultureswas not sufficiently high to permit anti-PE specificityto be evaluated by analysis of the distribution ofradioactivity in dissociated antigen-antibody com-plexes on sodium dodecyl sulfate (SDS) gels. PEhaving a high enough specific activity for this purpose(ca. 10,000 cpm/µg) was obtair ~yd lay allowing asmall inoculum of F. diplosiphoa to proliferate underfluorescent illumination in a closed system in thepresence of 5% 14CO2 . Ba14CO3 (New England Nu-clear, NEC-009, sp act 54 .9 mCi/mmol) was mixedwith sufficient nonradioactive BaCO3 to reduce thespecific activity to 125 µCi/mmol, and 14C02 wasgenerated by acidification of the mixture with 2 MH3P04 .

Solutions containing 8 M urea, 10% SDS, 0 .72 M2-mercaptoethanol, 0 .002 M phenylmethylsulfonylfluoride (PMSF) (see Schulze and Colowick, 1969),4.8% ethanol, and 0 .5 M Tris-HCI, pH 8.5, wereused to dissociate and reduce the antigen-antibodycomplexes . This solution (300 )Al) was added towashed immunoprecipitates containing 100 µg of la-beled PE and the mixture was placed in a boilingwater bath for 1 min to solubilize the precipitates.The solutions were then covered tightly with Parafilmand incubated at 37°C for 3 h to permit sulfhydrylreduction. The solutions were cooled to 23 °C andmixed with 250 µl of a solution containing 5% SDS,20% (wt/vol) iodoacetamide (Calbiochem, SanDiego, Calif.) and 0.5 M Tris-HCI, pH 8.5. The tubeswere then capped with Parafilm and left in darknessat room temperature for 1 h to permit alkylation ofthe dissociated, reduced complexes. Alkylation wasstopped by the addition of 25 µl of 14 .4 M 2-mercap-toethanol . The mixtures were then dialyzed againstseveral 50-m1 volumes of 2% SDS, 0.01 M sodiumphosphate, pH 7.0 . Dialysates were concentrated ap-proximately twofold by placing the bags on smallmounds of Sephadex G-200 .

Concentrated dialysates were subjected to SDS gelelectrophoresis and the gels were fixed, stained, andscanned at 620 nm following the procedures of Ben-nett and Bogorad (1971), with several modifications :(a) electrode and gel buffer strengths were halved,(b) electrophoresis was carried out for 6 .5 h, (c) themethylene bisacrylamide concentration was doubled,(d) a 10 mg/ml ammonium persulfate solution wasused, and (e) gel tubes were immersed in a stirred23 °C water bath during polymerization . With thesemodifications, the resolution in the 25,000-15,000molecular weight region is greatly increased . Gels

BENNETT AND Boa0RAD Chromatic Adaptation in a Blue-Green Alga

423

Dow


ecember 2021

were calibrated with ,B-galactosidase (130,000), glu-tamate dehydrogenase (53,000), glyceraldehyde-3-phosphate dehydrogenase (36,000), carbonic anhy-drase (29,000), and ribonuclease A (13,700) . Sourcesand molecular weight references for these standardproteins are listed in Weber and Osborn (1969) .Stained scanned gels were cut into approximately1 .1-mm thick slices with a device formed from analternating series of razor blades and washers . Sliceswere transferred to glass counting vials, air-dried,solubilized, and counted following the procedure thatwas described for methanol-extracted cells . Back-ground was determined with identically processedslices of gels devoid of radioactivity . Channel ratiomeasurements indicated that the solubilized gel sliceswere counted with a uniform efficiency of about 86% .

RESULTS

Pigmentation and MorphologicalCharacteristics of F. diplosiphon Adapted toFluorescent and to Red Illumination

Fig. 1 illustrates the emission spectra of thefluorescent and red light sources used, togetherwith absorption spectra of intact cells and crudeextracts of F. diplosiphon grown under these condi-tions of illumination .

In vivo spectra were obtained with suspensionscontaining approximately 0 .68 mg cellular dryweight per ml . Chi a, PC, and PE are primarilyresponsible for the peaks occurring in vivo at678-680, 624-629, and 569 nm, respectively. AnAPC peak is not observed in vivo . However, APCabsorption is responsible for the prominentshoulder at 649-653 nm in crude extracts of F.diplosiphon grown under fluorescent light . Peaks inthe in vivo absorption spectrum are also found atabout 490 nm (carotenoids) and 430 nm (Soretband of chl a) (see Emerson and Lewis, 1942) . Thefluorescent light emission peak is very close to theabsorption maximum of PE in vivo . Absorptionspectra of intact F. diplosiphon adapted to red lightdo not contain any suggestion of a PE peak .

Table II enumerates the quantitative differencesin pigmentation between F. diplosiphon adapted tofluorescent and to red light . Phycobiliprotein con-tents were estimated from analyses of absorptionspectra of crude algal extracts with the informationcontained in Table I . The values listed in Table IIwere determined with cultures having a cellulardry weight of 0 .25-0.35 mg/ml when harvested .Per unit of cellular dry weight, cultures adapted tofluorescent illumination have at least 205 times

424


00 770

-2

-0450

500

350

600

650

Wavelength. (nm)

FIGURE 1 Emission spectra of fluorescent and red lightsources and absorption spectra of crude extracts andwhole cells of F. diplosiphon adapted to these conditionsof illumination. In vivo spectra were obtained withwashed cells which were resuspended in distilled waterto give a dry weight of 0.68 mg/ml. Crude extracts wereprepared in 0 .01 M sodium phosphate, pH 7 .0, 0 .15 MNaCl .

more PE, approximately 3 .1 times less PC, 1 .7times less APC, and about 1 .5 times more chl athan cultures adapted to red illumination . Theinformation in Table II also demonstrates thatphycobiliproteins constitute the bulk of the solubleprotein in F. diplosiphon under the culture condi-tions used in these experiments .

The absence of PE from red light-adapted cul-tures may be demonstrated with Ouchterlonydouble-diffusion tests . The results of such tests withpurified phycobiliproteins and crude extracts offluorescent- and red light-adapted cultures areillustrated in Fig . 2 . Both of the PE precipitin lineswere red and fluoresced orange, and the PE linesboth were blue and fluoresced red . No cross-reaction was observed between any of the phyco-biliproteins, which indicates that the protein com-ponents of PE, PC, and APC from F. diplosiphonare unrelated . The subunit molecular weights ofthese phycobiliproteins are also distinct (Bennettand Bogorad, 1971) . Lines of identity were ob-

Dow


ecember 2021

TABLE II

Intensity Differences between the Fluorescent and Red Light Sources and Differences inPigmentation and Morphology between F . diplosiphon Adapted to these

Conditions of Illumination

Mean filament length

FIGURE 2 Ouchterlony double-diffusion tests withcrude extracts and purified phycobiliproteins from F.diplosiphon . Key to antibody and antigen wells: aPE,anti-C-phycoerythrin ; aPC, anti-C-phycocyanin ; aAP,anti-allophycocyanin ; PE, purified C-phycoerythrin ;PC, purified C-phycocyanin ; AP, purified allophyco-cyanin ; FCE, crude extract of cells adapted to fluores-cent illumination ; RCE, crude extract of cells adaptedto red illumination.

460 pm

50µm

* Emission spectra of the fluorescent and red light sources are illustrated in Fig . 1 .$ Obtained by manual integration of emission spectra from 400 to 730 nm .§ Measured with a Weston model 756 sunlight illumination meter .1) Determined by serological methods .

tained between purified and impure phycobili-proteins, suggesting that the antigenic properties ofthese compounds are not altered in the course ofpurification . No precipitin lines ever formed be-tween wells containing crude extracts of red light-adapted cultures and anti-PE . Double-diffusiontests are more sensitive than spectral measurementswith crude extracts as a means of estimating thePE content of red light-adapted cultures . If theanti-PE preparation was adequately diluted,orange-fluorescing precipitin lines formed indouble-diffusion tests with 100-p1 PE aliquotshaving a concentration as low as 0.02 mg/ml .Orange-fluorescing lines failed to form in paralleldouble-diffusion tests in which 100µl of a concen-trated crude extract from a red light-adapted cul-ture (total protein concentration, 17 .9 mg/ml)was added to the antigen well . Serial dilutions ofthe concentrated crude extract also failed to giveprecipitin lines, demonstrating that antigen excessdid not inhibit the precipitin reaction . On the basisof these results, it was concluded that PE is less

BENNETT AND BOGOBAD Chromatic Adaptation in a Blue-Green Alga

425

Illumination*

Fluorescent Red

Incident light intensity 666 µW/cm 2 $, 282 µW/cm 2 $,250 ft-c§ 26 ft-c§

Percentage of total soluble proteinPE 21 .0 <0 .11~~PC 14.3 47 .3APC 9 .5 15 .9

Total 44 .8 63 .3

Percentage of total cellular dry weightPE 6 .17 <0 .0311PC 4.40 13 .79APC 3 .17 5 .37Ch1 . a 2 .42 1 .58

Total 16.16 20 .77

Dow


ecember 2021

than (0.02/17.9)100 = 0.11% of the soluble pro-tein in red light-adapted cultures .There are striking morphological differences

between F. diplosiphon adapted to fluorescent andto red illumination . Fluorescent-illuminated cul-tures have a mean filament length of about 460µm. The average filament contains approximately45-60 cells. Filaments of red-illuminated cultures,on the other hand, have a mean length of about50 µm and each of these filaments contains 10-15cells . Cell size and shape are altered by the differ-ent conditions of illumination (see Fig. 3) . Fluores-cent-illuminated cells are larger, more cylindri-cally shaped, and have less pronounced constric-tions at their cross-walls than red light-illuminatedcells . After numerous liquid transfers under redillumination, conspicuous cyanophycin granulesaccumulate adjacent to the cross-walls betweencells . The light-induced changes in pigmentationand morphology in F. diplosiphon which have beendescribed are fully reversible with growth .

Metabolic Experiments

GENERAL INTRODUCTION : Investigations ofPE metabolism in F. diplosiphon were carried outwith the objective of answering three questions :

426

(a) Is PE a stable molecule in fluorescent light-adapted cultures? (b) When fluorescent light-adapted cultures are transferred to red light, is thepreexisting PE actively degraded, or is it simplydiluted with growth? and (c) Upon transfer of redlight-adapted cultures to fluorescent illumination,is PE synthesized de novo or from existing poly-peptides? To answer these questions, pulse-chasedfluorescent- or red-illuminated cultures wereprepared. During the post-chase period, fluores-cent-illuminated cultures were either left underfluorescent light or transferred to red light, andred-illuminated cultures were transferred to fluores-cent light . In each experimental situation, aliquotswere removed as the cultures proliferated . Thesesamples were quantitatively disrupted and theirPE contents were estimated spectrophotometri-cally. The addition of excess anti-PE to crude ex-tracts permitted the chromoprotein to be isolatedon a microscale for specific activity determina-tions . Measured specific activities were comparedwith those predicted simply on the basis of dilution(continuous fluorescent illumination), cessation ofsynthesis (transfer to red light), and de novo synthe-sis (transfer to fluorescent light) . Total incor-porated radioactivity levels and morphological

FIGURE 3 Morphological differences between filaments of F . diplosiphon adapted to fluorescent (upper)and red (lower) illumination . Bar = 10µm X 2,600 .

THE JOURNAL OF CELL BIOLOGY -

Dow


ecember 2021

features of the cultures were also taken into con-sideration. It was then possible to determine if

degradative metabolic forces play a significant rolein the regulation of PE levels and if other cellularproteins are utilized to any appreciable extentin the synthesis of PE when its formation is "in-duced" by fluorescent illumination .

The cultures were labeled with [ 14C]phenylala-nine because : (a) it was hoped that an amino acidwould be more or less specifically incorporated intoprotein constituents, and (b) Kiyohara et al . (1960)observed that reasonable quantities of phenylala-nine were taken up by T. tenuis during hetero-trophic growth . These workers found that argininesupported even better dark growth in T. tenuisthan phenylalanine, but arginine would probablybe a poor choice for metabolic labeling . Simon(1971) has demonstrated that blue-green algalcyanophycin granules, considered to be inherentlyunstable protein storage bodies, are copolymers of

arginine and aspartic acid .At the beginning of the sampling periods in all

metabolic experiments, less than 1 0/0 of the radio-activity in the cells was soluble in ice-cold 5%TCA. The chases were not 100% effective in anyof these experiments, perhaps because of incor-porated radiolytic or other contaminants in the[14C]phenylalanine preparation which were notdiluted by the addition of a large excess of un-labeled phenylalanine. It was established thatconcentration of the cultures by medium evapora-tion did not occur to any significant extent (lessthan I %) in the course of these investigations.

USE OF ANTIBODY TO PURIFY PE FROMCRUDE EXTRACTS Anti-PE (IgG fraction) wasemployed for the quantitative and selective re-moval of PE from crude extracts for specific ac-tivity determinations . Specific activities were cal-culated by dividing the corrected counts perminute in the precipitates by the number of micro-grams of PE in the precipitates . Several experi-ments were carried out in order to demonstratethat the precipitin reactions were both quantita-tive and selective .

Fig . 4 a illustrates the results of quantitative

precipitin tests with purified PE and anti-PE .

In these experiments, 1-ml volumes of buffered

saline, containing progressively increasing quanti-

ties of purified PE, were mixed with 100-µl volumes

of purified antibody. Control tubes received 100 µlof an IgG preparation, equal in protein concentra-

tion to the anti-PE solution, from nonimmunized

100-

Supernatant Test for PhycoerythrinO

IgG Control

i•1•120

40

60

80

100

Purified Phycoerythrin Added,µg

M

O

0.4vmCt.cQml

V

IgG Control

1

20

40

60

8o

µg Phycoerythrin Precipitated from14C-Labeled Crude Extract

FIGURE 4 (a) Quantitative precipitin curve for purifiedPE and anti-PE. The supernatant test consisted of ex-amining the solution with ultraviolet light and checkingfor the orange fluorescence of PE after the immunopre-cipitates had been harvested by centrifugation . (b)Quantitative precipitation of PE from a [14 Cjphenyl-alanine-labeled, fluorescent-illuminated F . diplosiphoncrude extract.

rabbits. Under the conditions of these tests, up toabout 80 µg of purified PE could be quantitativelyprecipitated out of solution .

In order to determine the antigen concentrationsrequired to yield antibody excess when precipitat-ing PE from crude extracts, precipitin reactionswere carried out with dilution series of crude ex-tracts from [ 14C]phenylalanine-labeled, fluorescentilluminated cultures. PE concentrations were

BENNETT AND B000RAD Chromatic Adaptation in a Blue-Green Alga

427

Dow


ecember 2021

determined from absorption spectra of the crudeextracts, using the information in Table I . Thewashed precipitates were extracted for lipids andnucleic acids and then counted, following proce-dures outlined in the Experimental Section. Theresults, illustrated in Fig . 4 b, indicated that PEcould be quantitatively precipitated from 1 ml ofcrude extract containing up to 65 µg of PE . Theabsence of PE in precipitin reaction supernatantswas confirmed spectrophotometrically . ParallelIgG control series indicated that nonspecificprecipitation of crude extract components duringincubation periods was minimal . Before precipitinreactions in metabolic experiments, all crude ex-tracts were diluted to give a PE concentration of35-40 µg/ml to insure antibody excess .

Direct proportionality in the labeled crude ex-tract precipitin reactions did not necessarily implymonospecific precipitation of PE, since this resultwould have been obtained if contaminating ma-terials precipitated in constant proportion with PE .To be assured of anti-PE specificity, it was there-fore necessary to determine the fraction of the totalradioactivity in the precipitates which was asso-ciated with PE .

Fig. 5 illustrates the results of SDS gel electro-phoresis of dissociated antigen-antibody complexeswhich were derived from crude extracts of fluores-cent-illuminated cultures grown in the presence ofMC02 . Four major protein peaks were observed .

428

FIGURE 5 SDS gel electrophoresis of a reduced alkyl-ated immunoprecipitate prepared by adding anti-PE inexcess to a 14C02-labeled, fluorescent-illuminated F .diplosiphon crude extract. Key to abbreviations : H-Ab,heavy y-globulin subunit ; GPDH, glyceraldehyde-3-phosphate dehydrogenase marker ; L-Ab, lighty-globulinsubunit ; H-PE, heavy C-phycoerythrin subunit ; L-PE,light C-phycoerythrin subunit .

THE JOURNAL OF CELL BIOLOGY . VOLUME 58, 1973

Molecular weight calibration of the gels demon-strated that these peaks corresponded to the heavyand light IgG subunits and the heavy and light PEsubunits . Approximately 80% of the total radio-activity in the gels was associated with the heavyand light PE subunits. The major radioactivecontaminant was associated with a band having anapparent molecular weight of about 29,000 . Beforestaining, this band demonstrated an orangefluorescence, characteristic of the heavy PE sub-unit, under ultraviolet illumination . This "con-taminant" may therefore be a fraction of the heavyPE subunits which, for some unknown reason, dis-played anomalous mobility during electrophoresis .

Cultures for metabolic experiments were labeledwith [14C]phenylalanine, not with 14C02. Theimmunoprecipitation specificity estimate obtainedby SDS gel electrophoresis of 14COrlabeled anti-gen-antibody complexes is probably low, because14002 undoubtedly labels nucleic acids and lipidsmore heavily than [14C]phenylalanine, and, fortechnical reasons, the washed precipitates couldnot be extracted for these constituents before elec-trophoresis .

Estimates of anti-PE specificity with [ 14 C] phenyl-alanine-labeled crude extracts were obtained byexperiments with artificial mixtures . Crude ex-tracts were prepared from red-illuminated culturesgrown in the presence of [14 C]phenylalanine . Asdiscussed previously, these extracts do not containdetectable levels of PE. Unlabeled, purified PEwas added to these crude extracts and then re-covered by the addition of anti-PE in excess . If theimmunoprecipitation were totally specific, noradioactivity should have been present in theprecipitates . The results of artificial mixture ex-periments are summarized in Table III . Approxi-mately 86-90% of the algal protein in precipitatesformed from crude extracts having low PE/PCratios was PE . Subtraction of counts obtained withIgG control tubes corrected for only about 20% ofthe nonspecific precipitation . The uncorrectedcontamination was probably due to coprecipitatingproteins which could not be removed duringprecipitate washing.

In metabolic experiments with pulse-chased, redlight-illuminated cultures transferred to fluorescentillumination, PE specific activities were estimatedwhen the PE/PC ratio became about 20% of thatin cultures fully adapted to fluorescent illumina-tion . The average soluble protein specific activitieswere determined, and counts in the precipitates

Dow


ecember 2021

Components of precipitin reaction mixtures

TABLE IIIEstimation of Anti-PE Specificity with Artificial Mixtures

Fraction non-PE algalPE/PC ratio

protein in ppt.

Fraction after IgG control

%

%

%

10

14

1120

10

8 .4

Crude extracts were prepared from [ 14 C]phenylalanine-labeled, red-illuminated F. diplosiphon . Such ex-tracts do not contain detectable levels of PE . Measured amounts of unlabeled, purified PE were added tothe crude extracts and then quantitatively recovered from the mixtures by the addition of anti-PE inexcess . From measurements of the radioactivity in the precipitates and the specific activity of the solubleprotein in the labeled crude extracts, the extent of nonspecific coprecipitation may be ascertained .* Contained 200 µg of PC ; average specific activity of protein in crude extract was 270 cpm/µg . Totalradioactivity was 112,640 cpm/ml .$ Averages of duplicates .§ Added to give 10% of the PE/PC mass ratio in fully adapted fluorescent-illuminated cultures .11 Added to give 20% of the PE/PC mass ratio in fully adapted fluorescent-illuminated cultures .

were corrected on the basis of 10 0/, nonspecificprecipitation . In metabolic experiments withpulse-chased, fluorescent-illuminated cultures,counts in the precipitates were corrected simply bysubtracting the counts present in IgG controltubes. Under these conditions, about 5 0/0 of thealgal protein in the precipitates was probably notPE. IgG controls did not correct for counts asso-ciated with these contaminants.CONTINUOUS FLUORESCENT ILLUMINA-

TION : The results of an experiment designed todetermine the stability of PE in cells fully adaptedto and growing under fluorescent illumination areillustrated in Fig . 6. Samples were removed froma pulse-chased, fluorescent-illuminated culturethat was left under fluorescent light . The radio-activity content of the cells per unit volume of cul-ture increased by 2% during thesampling period . The culture was healthy, judgingfrom the absence of vacuoles in the cells, and con-tinued to grow, as is demonstrated by the steadyincrease in dry weight per unit volume of culture(A-A-A). The PE/dry weight ratio increasedby about 6-8% over the sampling period. As the

culture proliferated, the PE content (O-0-0)increased and consequently the PE specific activity(not shown) decreased. The total radioactivityassociated with PE in the cells per unit volume ofculture (0-F-0) was obtained at each sam-pling point by multiplying the PE concentration(micrograms per milliliter culture) by the PEspecific activity (counts per minute per micro-gram). The least-squares line through these pro-ducts had essentially no slope . Scatter among theseproducts was due primarily to variability in esti-mates of PE concentration, not in PE specific ac-tivity .

The total radioactivity of the PE per unit volumeof culture was essentially constant ; this shows thatPE is a stable molecule under these conditions .New PE was synthesized, but was unlabeled,owing to the effectiveness of the chase . On theother hand, the PE labeled during the pulseperiod was not degraded . The PE concentrationtimes specific activity curve that would have beenobtained if PE had been degraded, without re-cycling, at 15% of the rate of the new PE increasehas been calculated and is plotted (+-+-+) .

BENNETT AND BocoRAD Chromatic Adaptation in a Blue-Green Alga 429

Red light crudeextract* Purified PE

Saline IgG Anti-PE cpm in precipitate :

M1

1 .0 100 X 100 X - 3281 .0 100 X 100 X 7061 .0 100 µl (30 .7 µg)§

- 100 X 1,3581 .0 100 µl (61 .4 µg)1 100 X 1,848

Dow


ecember 2021

0

PE SP ACI r PE Cox.

U.15% Degrodot,aa .

430

∎ Total Count ∎

III0

25

50

75

100

125TIME, HOURS

FIGURE 6 Estimation of the stability of PE in fluores-cent-illuminated F. diplosiphon cultures .

Although no measurable PE degradation could beobserved in this experiment, destruction at a rateof up to about 5% of the net PE accumulationwould probably have been undetectable .RED-TO-FLUORESCENT LIGHT TRANSFERS

The results of a red-to-fluorescent illumina-tion transfer experiment are shown in Fig. 7 .The pulse-chased, red light-illuminated culturewas left under red light for 20 h and then trans-ferred to fluorescent illumination . For undeter-mined reasons, it was particularly difficult to ob-tain effective chases with red-illuminated cultures .The total radioactivity in the cells increased byabout 10% during the overall sampling period inthis experiment . Transfer to fluorescent illumina-tion initiated PE synthesis (0-0-0). The rateof accumulation of chl a (X- X- X) was ac-celerated, while the level of PC (0-0-0) in-creased slightly and then became stationary . (Inseveral other experiments, the PC level continuedto increase at a slow rate over the entire samplingperiod .) These changes in rate of accumulation arein the direction of complementary chromaticadaptation. On a dry weight basis (see Table II),fluorescent light-adapted cultures have more PEand chl a and less PC than red light-adapted cul-tures . No net decreases in the phycobiliprotein orchl a levels of the cultures were ever observed inred-to-fluorescent light transfer experiments . Thelast sample taken in this experiment had a PE/PC

THE JOURNAL OF CELL BIOLOGY - VOLUME 58, 1973

Red Light Fluorescent Light

∎

∎

∎ PC

Total Co,Ms En

3a0

30

20

10

100

Eut

a80

-60

40

0.3

0.2

160

•

E•

0

ofU

-40

.4

1

FIGURE 7 Complementary chromatic adaptation aftertransfer of red light-adapted cultures to fluorescent il-lumination .

•

-120

ratio that was about 20% of that in fully-adaptedfluorescent-illuminated cultures . In this sample,the PE specific activity was 5.2 cpm/µg, whereasthat of the total soluble protein was 76 cpm/µg .

These data indicate that no more than about8 % of the PE synthesized in response to fluorescentlight could have been derived from preexistingproteins. However, in view of the 10 0/0 increase inradioactivity of the cells during the samplingperiod, a large fraction of the counts in the PE canprobably be attributed to the ineffectiveness of thechase .FLUORESCENT-TO-RED LIGHT TRANSFERS :

A fluorescent-to-red light transfer experimentis illustrated in Fig . 8 . While under fluorescentillumination, the dry weight, phycobiliprotein,and chl a levels of the culture all increased,but the total incorporated radioactivity remainedessentially constant. A decrease in the PE specificactivity (0-~-0) occurred because of thesynthesis of unlabeled PE . After transfer to redillumination, the PE specific activity becameconstant, indicative of a cessation of PE synthesis .The PC content of the culture (0-0-0) in-creased dramatically and, in this experiment, nodecrease in the level of PC was observed at anytime. However, net decreases in the levels of PE(0-0-0), chl a (X-X-X), dry weight

E ENE

5 600

3 Z

aE 500 E

0 UU24 150

3400

d E0

a-0.5 y

oI

E

E

300 WS

160 - 100.0.4 U -200,a

-0.3-150 -100 0

∎80 -50 0 .2 -1 2 -0

aO

01 -50

Dow


ecember 2021

-300-

FIGURE 8 Complementary chromatic adaptation aftertransfer of fluorescent light-adapted cultures to red il-lumination .

and total radioactivity (t-/--/)occurred in parallel shortly after transfer tored light. After these decreases, the PE, chl a andtotal radioactivity levels remained constant, butthe dry weight of the culture increased . Thechanges in pigmentation which took place aftertransfer to red light were exactly opposite to thoseobserved in red-to-fluorescent light transfer ex-periments (see Fig . 7) . In this experiment, the PElevel decreased by 16 .2 0/c and the total radioac-tivity decreased by 15.8%. Equivalent losses in PEand total incorporated radioactivity should beobserved as a consequence of lysis if the chase hasbeen effective, if PE synthesis has stopped, and ifthese two components are evenly distributedthroughout the cell population .MEAN FILAMENT LENGTH REDUCTION

DURING ADAPTATION TO RED LIGHT :There are morphological aspects to complemen-tary chromatic adaptation . The mean filamentlength of fluorescent-illuminated cultures is muchgreater than that of red-illuminated cultures (seeTable II) . Fig . 9 demonstrates that the long fila-ments of fluorescent-illuminated cultures frag-ment very soon after transfer to red light . Further-more, the decrease in mean filament length(+---1---+) occurs simultaneously with losses inPE and total incorporated radioactivity .

Fig. 10 a is a photomicrograph of F . diplosiphon

t I

∎ Toro/ Cants

BENNETT AND BOGORAD Chromatic Adaptation in a Blue-Green Alga

Worescent RedLight

Light

1t

t

0

40

80

120

160

200TIME, HOURS

FIGURE 9 Complementary chromatic adaptation aftertransfer of fluorescent light-adapted cultures to red il-lumination . This experiment was similar to the one forwhich data are given in Fig . 8 except that the courseof alteration in mean filament length in response to redlight was also determined.

filaments that were removed from a fluorescent-to-red light transfer immediately before the majorperiod of mean filament length reduction. Largenumbers of dark, biconcave cells, known as ne-cridia or separation disks, are scattered throughoutthe filaments . These structures disappear from thecultures after filament fragmentation and have notbeen observed in cultures fully-adapted to redlight. Fig . 10 b is a high power photomicrograph ofan infrequently observed necridium from a fluores-cent-illuminated culture .

Lamont (1969) has summarized evidence indi-cating that transcellular filament breakage almostinvariably occurs across necridia, resulting in lysisof the cell at the locus of breakage . Transcellularfilament breaks may also occur across vegetativecells (Lamont, 1968 ; F. Drouet, personal commu-nication) . If filament fragmentation during fluores-cent-to-red light transfer experiments wereachieved exclusively by transcellular breakage, asubstantial fraction of the cell population would belysed . Cells in filaments of fluorescent-illuminatedF. diplosiphon are about 8-10 µm long. Transcellu-lar breakage of filaments 450-600 µm long intofilaments having a length of about 50 µm wouldresult in the lysis of one out of every six or seven

431

EPZ

g

E

20

Ua E

PE E

2-1

a200

Z70 15

3Oa x

o

WE

E

0-0 .4 40 50 10

UV

100 -0.3 30 40

-0 .2 20 30 5

-0.1 10 20

10

600

Eh

E5 -500

P3

P3 E

W0. E

0

120 300 -400

80.200y0.4

a

of

U

40

b

EZ

J

E4-0.3 30

40 100 -0.2 20

O∎

f

1 -100-0.1 10

Dow


ecember 2021

FIGURE 10 (a) Photomicrograph of a sample from a fluorescent-illuminated F. diplosiphon cultureadapting to red light, obtained just before mean filament length reduction . Many dark, biconcavecells, (arrowheads), known as necridia, are scattered throughout the filaments . These cells serve as locifor transcellular filament breakage. X900 . (b) High power photomicrograph of a necridium in a fila-ment of F. diplosiphon adapted to fluorescent light. X 1,500 .

cells (16 .5 070 to 14 .3 01/0) . Losses in PE and totalradioactivity occur at about these levels in fluores-cent-to-red light transfer experiments and aretemporally preceded by morphological indicatorsof incipient lysis (necridia) . During the period ofmean filament length reduction, phycobiliproteinsmay also be observed in the culture medium . A lossin the morphological integrity of a fraction of thecell mass, rather than active metabolic degrada-tion, is responsible for the observed diminution

432


in the PE content of fluorescent-illuminated cul-tures adapting to red light.

DISCUSSION

PE is synthesized but not measurably degraded inF. diplosiphon adapted to and growing underfluorescent illumination . This fact is demonstratedby the constant total radioactivity associated withthe PE per unit volume of proliferating pulse-chased cultures. The PE specific activity remains

Dow


ecember 2021

constant after pulse-chased, fluorescent-illumi-nated F. diplosiphon cultures are transferred to redlight because synthesis of unlabeled PE is shut offand no additional dilution of radioactive PEoccurs. The PE content of F. diplosiphon culturesdecreases, however, during complementary chro-matic adaptation accompanying the shift fromfluorescent to red illumination. The decrease inPE level which is observed after transfer to redlight is due to cell lysis, not to measurable activemetabolic degradation . These results lead to theconclusion that the PE is a stable molecule in F.diplosiphon and that light influences the intracellu-lar levels of this chromoprotein solely by increasingor decreasing its rate of synthesis . The methodsemployed in these experiments must be regardedas providing minimum estimates for PE degrada-tion rates, since the extent of metabolic recyclingof degraded labeled constituents could not bemeasured. Destructive forces may play an impor-tant role in PE metabolism under different envi-ronmental circumstances and in stationary ornecrotic cultures, but PE is not degraded to anysignificant extent in growing cultures under theconditions of these experiments .

Transfer of pulse-chased, red-illuminated, PE-less cultures to fluorescent light induced PE pro-duction. When the PE/PC ratio reached 20 0]0 ofthat observed in cells fully adapted to fluorescentlight, the specific activity of PE was 8 % of thatobserved in the average soluble protein . Themajority of the PE synthesized was, therefore, notderived from other preexisting soluble proteins .

The mechanism by which light influences therate of PE synthesis is not known . Light does notappear to be absolutely required for PE synthesisin all blue-green algae . Nostoc muscorum A, whichdoes not adapt chromatically, continues to pro-duce PE when grown heterotrophically in com-plete darkness (Lazaroff, 1966) . Scheibe (1972)has recently provided limited evidence for theexistence of a green-red photoreversible pigment inT. tenuis. Referring to action spectrum data ob-tained by Fujita and Hattori (1962) and byLazaroff (1966), Scheibe has proposed a hybridmodel in which light modifies the properties of thephotoreversible pigment, which in turn "controls"phycobiliprotein synthesis and morphogenesis .According to this model, green light promotes PEsynthesis and growth as short filaments, and redlight promotes PE synthesis and long filaments .This relationship does not hold for F. diplosiphon,

since green light stimulates the production of longfilaments and red light generates cultures com-posed of short filaments .

Once PE synthesis is induced by light, furtherlevels of control come into play . Since measurablepools of phycoerythrobilin and PE apoproteinhave not been detected in the present work, thesynthesis of these moieties is probably tightly co-ordinated. It is also probable that a mechanismexists for the stoichiometric synthesis of the heavyand light PE subunits . It is not known at present ifthe protein components of the two subunits arerelated .

When attempting to determine action spectrafor phycobiliprotein synthesis in T. tenuis, Hattoriand Fujita (1959) did not observe any effect oflight upon the chi a content of the organism .Light has a strong influence upon both the chia and the phycobiliprotein content ofF. diplosiphon .Fluorescent illumination provides large quantitiesof light that the phycobiliproteins absorb well,but relatively little light in the wavelength regionsof maximal chi a absorption. Red light, on theother hand, has large quantities of energy thatmay be absorbed by the long wavelength bandof chi a, but comparatively little that is ab-sorbed by phycobiliproteins. The ratio of totalphycobiliprotein to chi a in F. diplosiphon culturesadapted to red illumination is 2 .15 times greaterthan that in cultures adapted to fluorescentillumination . Thus, changes in chi a content arealso a component of complementary chromaticadaptation in this system . Jones and Myers (1965)previously observed light quality to have a stronginfluence upon the chi a content of Anacystisnidulans .

Complementary chromatic adaptation has beenregarded heretofore as a rather specific alterationin the phycobiliprotein complement in responseto changes in the light environment's quality .However, quite marked morphological andprobably many other less obvious enzymaticchanges also occur in F. diplosiphon . These mani-fold changes may be direct or indirect conse-quences of light absorption by a "master switch"pigment or by several discrete photoreceptors .

The findings discussed here underscore diffi-culties that systematists face in their efforts tocreate a meaningful blue-green algal taxonomy .Alterations in either the availability of exogenouscombined nitrogen (Bennett, 1972) or in theenergy distribution in the visible light environ-

BENNETT AND BOGORAD Chromatic Adaptation in a Blue-Green Alga 433

Dow


ecember 2021

merit can convert F. diplosiphon into strikingly dif-

ferent morphological variants . If standard taxo-nomic criteria were used, these forms would beclassified as totally different organisms . Further-more, the experiments concerned with the adapta-tion of fluorescent-illuminated cultures to red lightillustrate how incorrect physiological conclusionsmay be reached; cell lysis could easily have beeninterpreted to mean PE degradation in this par-ticular situation if PE levels and specific activitieshad been the only parameters measured .

This work was supported by Grant GB 31114 fromthe National Science Foundation . A. Bennett was anNSF Predoctoral Fellow .

Received for publication 15 January 1973, and in revisedform 5 April 1973 .

BIBLIOGRAPHY

BENNETT, A. 1972. Metabolic and structural investi-gations of Fremyella diplosiphon phycobiliproteins.Ph.D. Thesis . Harvard University, Cambridge,Mass .

BENNETT, A., and L . BOGORAD. 1971 . Properties ofsubunits and aggregates of blue-green algal bili-proteins. Biochemistry . 10 :3625 .

BOGORAD, L . 1962 . Chlorophylls . In Physiology andBiochemistry of Algae. R. A. Lewin, editor. Aca-demic Press, Inc ., New York. 385 .

BORESCH, K . 1919. Ueber die Einwirkung farbigenLichtes auf die Farbung von Cyanophyceen . Ber .Dtsch. Bot. Ges . 37 :25 .

BoRESCH, K. 1921. Die komplementare chromatischeAdaptation . Arch . Protistenkd. 44 :1 .

CHASE, M. W. 1967 . Collection of antiserum . Meth .Immunol. Immunochem . 1 :237.

DUYSENS, L. N. M. 1951 . Transfer of light energywithin the pigment systems present in photosyn-thesizing cells. Nature (Lond.) . 168 :548 .

EMERSON, R ., and C. M. LEWIS . 1942. The photo-synthetic efficiency of phycocyanin in Chroococcusand the problem of carotenoid participation inphotosynthesis . J. Gen . Physiol . 25 :579 .

ENGELMANN, T. W. 1883. Farbe and Assimilation .Bot. Z. 41 :1 .

FRENCH, C . S., and V . K. YOUNG . 1952 . The fluores-cence spectra of red algae and the transfer of energyfrom phycoerythrin to phycocyanin and chloro-phyll . J. Gen . Physiol. 35 :873 .

FUJITA, Y., and A. HATTORI. 1962 . Photochemicalinterconversion between precursors of phycobilinchromoproteids in Tolypothrix tenuis. Plant CellPhysiol. 3 :209 .

GAIDUxov, N. 1902. Ueber den Einfluss farbigen

434

THE JOURNAL OF CELL BIOLOGY . VOLUME 58, 1973

Lichts auf die Farbung lebender Oscillarien . Abh .Preuss Akad. Wiss . (Berlin) . no . 5 .

GOLDMAN, M. 1968. Fluorescent Antibody Methods .Academic Press, Inc ., New York.

HALLDAL, P. 1970. The photosynthetic apparatus ofmicroalgae and its adaptation to environmentalfactors . In Photobiology of Microorganisms. P.Halldal, editor . John Wiley & Sons, Inc ., NewYork . 17.

HATTORI, A., and Y. FUJITA. 1959. Formation ofphycobilin pigments in a blue-green alga, Tolypo-thrix tennis, as induced by illumination with coloredlights. J. Biochem. (Tokyo) . 46 :521 .

HAXO, F . T., and L . R . BLINKS . 1950. Photosyntheticaction spectra of marine algae . J. Gen . Physiol. 33 :389 .

HOGNESS, D. S., M. COHN, and J. MONOD. 1955 .Studies on the induced synthesis of /3-galactosidasein Escherichia coli : the kinetics and mechanism ofsulfur incorporation . Biochim . Biophys. Acta. 16 :99 .

JONES, L . W., and MYERS, J . 1965 . Pigment variationsin Anacystis nidulans induced by light of selectedwavelengths . J. Phytol . 1 :7 .

KARAT, E. A. and M. M . MAYER. 1967 . Experimen-tal Immunochemistry . Charles C. Thomas, Pub-lisher . Springfield, Ill . 2nd edition .

KIYOHARA, T ., Y. FUJITA, A. HATTORI, and A .WATANABE. 1960 . Heterotrophic culture of a blue-green alga, Tolypothrix tennis . J. Gen. Appl. Micro-biol . 6 :176 .

KOSHLAND, M . E ., F. M. ENGLBERGER, and S . M .GADDONE. 1962 . The effect of iodination on thebinding of haptene to arsonic antibody . J. Immu-nol. 89 :517 .

KRATZ, W. A., and J. MYERS. 1955 . Nutrition andgrowth of several blue-green algae . Am. J. Bot . 42 :282 .

KYLIN, H. 1912. Ueber die Farbe der Florideen andCyanophyceen . Sven . Bot . Tidskr . 6 :531 .

LAMONT, H. C. 1968 . Trichome breakage in Oscilla-toriacean blue-green algae. Ph.D. Thesis. HarvardUniversity, Cambridge, Mass.

LAMONT, H. C. 1969 . Sacrificial cell death and tri-chome breakage in an Oscillatoriacean blue-greenalga : the role of murein. Arch . Mikrobiol . 69 :237 .

LAZAROFF, N. 1966 . Photoinduction and photorever-sal of the Nostocacean developmental cycle. J.Phycol. 2 :7 .

LOWRY, O. H., N. J . Rosebrough, A. L. Farr, andR. J. Randall . 1951 . Protein measurement with theFolin phenol reagent . J. Biol. Chem . 193 :265 .

MACKINNEY, G. 1941 . Absorption of light by chloro-phyll solutions . J. Biol. Chem . 140 :315 .

MANS, R. J ., and G . D. NOVELLI . 1961 . Measurementof the incorporation of radioactive amino acidsinto protein by a filter-paper disk method. Arch .Biochem . Biophys. 94:48.

SCHEIBE, J. 1972 . Photoreversible pigment : occur-

Dow


ecember 2021

rence in a blue-green alga . Science (Wash ., D.C.) .176:1037 .

SCHULZE, 1 . T ., and S. P. CoLOwscK . 1969. The mod-ification of yeast hexokinases by proteases and itsrelationship to the dissociation of hexokinase intosubunits . J . Biol . Chem . 244:2306.

SIMON, R. D. 1971 . Cyanophycin granules from theblue-green alga Anabaena cylindrica : a reserve mate-rial consisting of copolymers of aspartic acid andarginine . Proc . Natl . Acad. Sci. U.S.A . 68:265.

STARR, R. C. 1964 . The culture collection of algae atIndiana University . Am. J. Bot. 51 :1913 .

TISHLER, P. V ., and C. J . EPSTEIN . 1968 . A convenientmethod of preparing polyacrylamide gels for liquidscintillation spectrometry . Anal. Biochem. 22 :89.

WEBER, K. and M . OSBORN . 1969. The reliability ofmolecular weight determinations by dodecyl sul-fate-polyacrylamide gel electrophoresis . J. Biol .Chem . 244:4406.

BENNETT AND BOGORAD Chromatic Adaptation in a Blue-Green Alga 435

Dow


ecember 2021

Documents

COMPLEMENTARY CHROMATIC ADAPTATION IN A …