Plant Physiol. (1989) 90, 516-5210032-0889/89/90/051 6/06/$01 .00/0

Received for publication October 17, 1988and in revised form January 23, 1989

Coordinate Expression of Rubisco Activase and Rubiscoduring Barley Leaf Cell Development1

Raymond E. Zielinski*, Jeffrey M. Werneke, and Michael E. Jenkins2Department of Plant Biology, University of Illinois, Urbana, Illinois 61801

ABSTRACT

We have utilized the cellular differentiation gradient and pho-tomorphogenic responses of the first leaf of 7-day-old barley(Hordeum vulgare L.) to examine the accumulation of mRNA andprotein encoded by the ribulose-1,5-biphosphate carboxylaseholoenzyme (rubisco) activase gene (rca). Previous studies haverevealed a pattem of coordinate expression of rubisco subunitpolypeptides during development. We compared the expressionof nubisco polypeptides and mRNAs with those encoded by rca.The mRNAs encoding both rubisco activase and rubisco areexpressd exclusively in leaf tissue of 7-day-old barley seedlings;mRNAs and polypeptides of rca accumulate progressively fromthe leaf base in a pattem that is qualitatively similar to that ofrubisco subunit mRNAs and polypeptides. The parallel pattem ofrca protein and mRNA accumulation indicate that a primary con-trol of rca gene expression in this system lies at the level ofmRNA production. Light-induced expression of rca in etiolatedbarley follows a different pattem from that of the acropetal barleyleaf gradient, however. Etiolated, 7-day-old barley seedlings con-tain levels of rca mRNA near the limit of detection in Northemblot hybridization assays. White light induces a 50- to 100-foldaccumulation of rca mRNA, which is detectable within 30 minafter the onset of illumination. In contrast, steady state levels ofmRNAs encoding the small rubisco subunit are affected little bylight, and mRNAs encoding the large subunit accumulate about5-fold in response to illumination. While rca mRNA levels are lowin etiolated barley leaves, levels of the protein are approximately50 to 75% of those found in fully green leaves.

The C3 photosynthetic carbon reduction pathway, initiatedby ribulose-l,5-bisphosphate carboxylase/oxygenase (rub-isco),3 is highly regulated in vivo. Regulation of net carbonassimilation is complex and is accomplished via a number ofdiverse factors. These factors potentially range from, for ex-ample, water relations of the whole plant (8) to the relativeconcentrations of individual Calvin cycle metabolite mole-

' Supported by a grant from the U. S. Department of Energy (DEFG02 88ER13900) and by a National Institutes ofHealth BiomedicalResearch Support Grant to the School of Life Sciences, University ofIllinois (to R. E. Z.).

2 Present address: Department of Molecular and Cellular Biology,University ofArizona, Biosciences West Building, Tucson, AZ 85721.

3Abbreviations: rubisco, ribulose-1,5-bisphosphate carboxylaseholoenzyme; rbcL, ribulose-1,5-bisphosphate carboxylase large sub-unit gene; rbcS, ribulose-1,5-bisphosphate carboxylase small subunitgene; rca, rubisco activase gene; cab, light-harvesting Chl a/b-bindingprotein gene.

cules in the chloroplast (5). Recently, one potential source ofregulation over carbon assimilation was uncovered when itwas demonstrated that rubisco is activated to catalytic com-petency by a soluble chloroplast protein, rubisco activase (rcaprotein) (19, 22). ATP (26), Mg2+, and C02 (21) are requiredfor rubisco activation by rca protein in vitro. In the presenceof rca protein, millimolar levels of RuBP do not inhibitrubisco activation (21), as is the case when rubisco is activatedspontaneously. However, little is known about the nature ofthe physical interaction between rubisco and rca proteinduring activation. Using antibodies prepared against purifiedspinach rca protein as probes, it was shown that activaseconsists oftwo immunologically related polypeptides ofabout46 and 41 kD (29), and that they can be detected immuno-logically in extracts of all higher plant species examined (23).We recently utilized these monospecific antibodies to clonecDNAs encoding the spinach and Arabidopsis thaliana rcagene products (29). We initiated this study to use these toolsto ask whether accumulation of rca and rubisco subunitmRNAs and polypeptides is coordinated during the course ofplastid development.

Chloroplast development has been a subject of intensiveresearch for a considerable time. Most of this work, however,has focused on development of the photosynthetic mem-branes and photoregulation of gene expression (for a recentreview, see ref. 17). Apart from the considerable work on thestructure and expression ofthe genes encoding rubisco subunitpolypeptides, little is known about the expression of genesencoding other stromal enzymes that function in the carbonreduction pathway.The leaves of monocotyledonous plants are particularly

well suited for studies of plastid and leaf development. Inthese plants, leaf cell division is restricted to a basal meristem.At increasing distances along the leaf axis from the basalmeristem, the cells and organelles are of increasing maturity.As a consequence, monocot leaves represent a developmentalcontinuum of cells whose spatial separation facilitates accessto gram quantities of developmentally similar material. Anumber of studies have exploited this system to examine thesynthesis and accumulation of several chloroplast polypep-tides and mRNAs including rubisco (reviewed in 13, 17). Inthis study, we utilized the naturally occurring developmentalgradient and the well characterized photomorphogenic re-sponses of the first leaves of barley (Hordeum vulgare L.) toask whether the accumulation of rca, rbcS, and rbcL mRNAs,and their polypeptide products is coordinated during leaf celldevelopment.

516

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.

RUBISCO ACTIVASE EXPRESSION IN BARLEY

MATERIALS AND METHODS

Plant Material and Growth Conditions

Barley (Hordeum vulgare L.) seedlings were grown for 7 din either a green house or a growth chamber on a soil/vermiculite mixture. For greenhouse plants, the photoperiodranged from 10 to 12 h d-', and for the growth chamberplants, it was 16 h d-'. In both cases, the temperature wasmaintained at 20 ± 2C. First leaves, ranging from 12 to 15cm in length, were harvested by cutting at the base, andremoving the coleoptiles and any second leaf material. Theleaves were dissected into six equal length segments (i.e. about2-2.5 cm each); the segments were processed immediately forsoluble protein extractions or were quick-frozen in liquid N2and stored at -80C until they were to be used for RNAisolation. Etiolated seedlings were raised for 7 d in a darkenedgrowth chamber at 20 ± 2C, and watered without illumina-tion. Under these growth conditions, there was no detectableaccumulation of Chl, nor was cab mRNA detectable. Green-ing was induced by exposure to a combination of fluorescentand incandescent lamps that provided 200 ME . m-2 * s-' at theuppermost regions of the seedling.

Preparation of Soluble Protein Extracts

Total soluble proteins were extracted from leaf tissue seg-ments into ice-cold buffer consisting of 20 mM Tris-HCl (pH8.0), 20 mM MgCl2, and 10 mM DTT or 10 mm Bis-Tris-propane/HCI (pH 7.0), 5 mM MgCl2, 1 mM Na2EDTA, 15mM DTT, 1 mm benzamidine, 1 mM PMSF, 10 Mm leupeptin,and 0.4 mM ATP. Five-hundred mg of tissue were cut intofine pieces with a razor blade, transferred to a glass homoge-nizer or mortar together with 2.5 mL of buffer, and groundfor about 1 min on ice. Extracted proteins were immediatelytransferred to a microcentrifuge tube, and insoluble materialswere removed by two successive 2-min centrifugations. Onehundred-ML aliquots were taken immediately and precipitatedwith 80% (v/v) acetone for total protein estimation (14) andSDS-PAGE analysis (12); untreated aliquots were used forrubisco isolation (10).

Antibody Preparation and Westem Immunoblotting

Polyclonal antibodies, recognizing both the large and smallsubunits of rubisco or rubisco activase, were prepared in miceaccording to Lacy and Voss (1 1). Rubisco holoenzyme waspurified from barley by sucrose density gradient sedimenta-tion (10); and rubisco activase was purified from spinachleaves by fast protein liquid chromatography (23). Approxi-mately 20 Mg of antigen were used for primary and boosterimmunizations. Antibody specificity was monitored by west-ern immunoblotting assays using alkaline phosphatase-con-jugated secondary antibody screening (2).

RNA Extraction and Northern Blot Analysis

High mol wt RNA was prepared from the first leaves of 7-d-old barley seedlings as described previously (24). RNA wasfractionated on agarose-formaldehyde gels (30), transferred tonitrocellulose or GeneScreen filters, and hybridized with 32p_

labeled probes. Hybridization was carried out in 50% (v/v)formamide, 5x SSPE (lx SSPE is 0.135 M NaCl, 10 mMNaH2PO4/Na2HPO4 [pH 7.4], 1 mM Na2EDTA), 1 to 5xDenhardt's solution (lx Denhardt's solution is 0.02% w/veach BSA, Ficoll, and polyvinyl pyrrolidone), 0.1% (w/v)SDS, and 32P-labeled probe at 10 to 20 ng/mL and 55C (forRNA probes) or 25 to 50 ng/mL and 42°C (for DNA probes).Final posthybridization washes were performed in 0.1X SSPE,0.1% (w/v) SDS at 65C (for RNA probes) or 50°C (for DNAprobes). The 32P-labeling ofDNA restriction fragment probeswas performed by either oligolabeling with DNA Polymerase1 (9) or in vitro transcription with T7 RNA polymerase (30).The recombinant probes used in this study were: (a) foractivase mRNA, spinach pRCA 1.6 (29) or barley rca cDNAs;(b) for small subunit mRNA, wheat pW9 (4) or barley rbcScDNAs (RE Zielinski, ME Jenkins, unpublished data); (c) forlarge subunit mRNA, an internal 0.8-kb EcoRI fragment ofmaize rbcL (16); and (d) for cab mRNA, pea pAB-96 (3).

Cloning of Barley rca cDNA Sequences

Polyadenylated RNA was prepared by poly(U)-agarose (PL-Pharmacia, type 6) chromatography of total high mol wtRNA isolated from 7-d-old barley seedlings (30). Double-stranded cDNA from this RNA was generated according toWerneke et al. (29), inserted into the EcoRI site of XgtlO orXgtl 1, and packaged in vitro. The cDNAs cloned in Xgtl 1were screened for expression of rca-lacZ fusion proteins usingpolyclonal antibodies to purified spinach rca protein, as de-scribed previously (29). The identities of positive clones wereconfirmed by hybridization, using a 32P-labeled cloned spin-ach rca cDNA, and by DNA sequencing. One such recombi-nant phage, XBrca-l9, containing a 1.4 kb insert, was isolatedand used as a hybridization probe in these studies. Detailsconcerning the isolation and characterization of these cloneswill be described elsewhere.

RESULTS

Tissue-Specificity of rca mRNA Expression

In our first series of experiments, we examined the tissuespecificity of rca expression. Figure 1 shows the results of twoNorthern blot hybridizations in which rca and rbcS steadystate mRNA levels were monitored in leaf, coleoptile, seed,and root total RNA fractions isolated from 4-d-old barleyseedlings. Both messengers are detected exclusively in the leafRNA fraction. The rca mRNA is estimated to be about 1.9kb (slightly larger than 18S rRNA), a value that is consistentwith the previously estimated sizes ofspinach and Arabidopsisrca messengers (29). Some spurious hybridization of the rcaprobe to rRNA was occasionally observed in the root RNAfraction, which is most likely to be a consequence of theprolonged hybridization time (48 h) used in those experi-ments. In favor of our interpretation, we note that no hybrid-izing RNA species in the root fraction corresponding to thesize of rca mRNA was ever observed, nor did we detect rcapolypeptide translation directed by coleoptile, seed, or rootpoly(A+) RNAs in vitro (RE Zielinski, unpublished results).

517

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 90,1989

b C.

LC S R L C S R LC S R

Figure 1. Tissue-specific expression of rubisco activase in barley.Total high mol wt RNA was isolated from 4-d-old barley leaf (L),coleoptile (C), seed (S), and root (R) tissues, and 2-,ug aliquots werefractionated by gel electrophoresis. Following transfer to nitrocellu-lose filters, the RNAs were probed with oligolabeled cDNA insertsencoding (panel a) barley rca, or (panel c) wheat rbcS. Panel b showsthe ethidium bromide staining pattern of the same RNA fractions toverify that comparable loadings were used.

Expression of rca during Barley Leaf Development

Expression of rbcS and rbcL polypeptides is highly coordi-nated in a variety of plant species, including monocot leaves,with little or no detectable free subunit accumulation, undernormal physiological conditions (7, 18). Since rca protein isrequired for optimal in vivo rubisco activity (25), we askedwhether rca expression is also coordinated with expression ofrbcS and rbcL at the steady state level ofmRNA and protein.We carried out a series of experiments designed to examineaccumulation of rca mRNA by hybridization of cloned spin-ach or barley rca cDNAs to agarose gel-fractionated RNAsisolated from six segments spanning the entire first leaf of 7-d-old barley seedlings (1 = base; 6 = tip; described further in"Materials and Methods"). Similarly to rbcS mRNA (6, 27),relative rca mRNA steady state levels increase acropetallyalong the barley leaf axis, reaching a maximum in segmentnumber five, and decline about 60% in the oldest cells at thetip of the leaf (Fig. 2a). Transcripts encoding rbcL accumulatewith a qualitatively identical pattern, although the decline inrbcL mRNA levels in the oldest region of the leaves is not as

dramatic as in the case of rca (Fig. 2b). In contrast, cabmRNA, which encodes the major light-harvesting thylakoidmembrane polypeptide, accumulates much earlier in devel-opment and its steady state level declines markedly in the leafregions expressing the highest levels of rca and rbcL mRNAs.The densitometer quantitations shown in Figure 2 are theresults of a single experiment in which autoradiographic ex-

posures of systematically varying length were scanned. Thescans shown in the figure represent ones taken from exposuresthat were determined to be within the linear response rangeof the film (when autoradiographic density versus time was

plotted). Similar trends were observed for two other inde-pendent RNA preparations examined on different occasions

I - ir,.)(j.-

-N

ro9..d k t'

a-.

3b.. .:"a

Figure 2. Northern blot hybridization assay of the relative levels ofrca, rbcL, and cab mRNAs along the barley leaf developmentalgradient. Poly(A+) RNA (1 M~g) or total RNA (10 Mg) was isolated fromsix segments spanning the first leaf (1 = basal, 6 = tip), fractionatedby electrophoresis in formaldehyde-agarose gels, transferred to nitro-cellulose and hybridized with (a) a spinach rca probe, or (b) a maizerbcL, or pea cab cDNA probe. Relative hybridization intensities wereestimated by laser scanning densitometry of autoradiographs of eachblot experiment in an exposure range that was empirically determinedto be linear for density versus time.

but, in these cases, comparable quantitations were notperformed.

It should be noted that the two rca polypeptides are encodedby distinct mRNA species that are produced by alternativemRNA splicing (JM Werneke, WL Ogren, manuscript inpreparation). These mRNAs differ by less than 25 nucleotides,and thus cannot be distinguished by size in Northern blotassays such as the one shown in Figure 2a.

It was previously shown (27) that rubisco protein accumu-lates in barley leaf cells in increasing amounts from the baseto the tip of the leaf. This pattern of accumulation is similarto that observed for rbcS mRNA, except that the oldest cellsnear the leaf tip retain high levels of rubisco, while rbcSmRNA levels decline. We compared the relative accumula-tion of rca protein, detected immunologically by Westernblotting, with the relative accumulation of rubisco assayed bysucrose density gradient purification (10) and SDS-PAGEfractionation. Accumulation of rca polypeptides qualitativelyparallels that of rubisco protein accumulation along the barleyleaf axis (Fig. 3), except at the leaf tip. In this region, rcapolypeptide levels decline slightly, although this decline wasnot as great as the one observed for rca mRNA (Fig. 2a).Similar results were observed when rubisco levels were assayedimmunologically in total soluble protein extracts (data notshown).

It should be noted that rca protein consists of two immu-nologically related polypeptides (29) of about 46 and 41 kDin every plant species examined (23). We initially observed

ZIELINSKI ET AL.518

W,

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.

RUBISCO ACTIVASE EXPRESSION IN BARLEY

bActivase

aTotal Protein Rubisco

123456 12345692.5-_ _66.2 -45 ^_

21.5-_14.4-

Figure 3. Accumulation of rca and rubisco holoenzyme along thebarley leaf developmental gradient. Equal amounts of total solubleproteins or equal proportions of sucrose gradient-purified rubiscofrom different regions of 7-d-old barley leaves were fractionated bySDS-PAGE. Total protein and purified rubisco were stained withCoomassie blue (panel a); rca was detected immunologically amongtotal soluble proteins using anti-spinach rca mouse ascites fluid andalkaline phosphatase-conjugated goat anti-mouse IgG (panel b).

reduced amounts of the 46 kD rca polypeptide comparedwith the 41 kD polypeptide (Fig. 3b). In later experiments, inwhich several protease inhibitors were included in the solubleprotein extraction buffer, the relative amounts of the 46- and41-kD rca polypeptides were more comparable. We attributethese differences to a greater susceptibility of the 46 kDpolypeptide to proteolytic degradation. In support of thisinterpretation, we note that the 46 kD rca polypeptide isparticularly unstable during purification of the protein fromspinach (JM Chatfield, personal communication) or barley(RE Zielinski, unpublished results). When soluble proteinswere isolated from the six acropetal leaf gradient segmentsusing extraction conditions that partially stabilize the 46-kDrca polypeptide, Western blot measurement of the relativelevels of the rca polypeptides revealed qualitatively similarresults compared with those shown in Figure 3b (i.e. rca

protein accumulated progressively from the base to the tip ofthe leaf, data not shown).

Regulation of rca mRNA Accumulation by Light

In addition to the acropetal cellular differentiation gradient,barley leaves are also a useful experimental system for exam-ining the effects of light on gene expression. We examinedthe steady state levels of rca, rbcS, and rbcL mRNAs in 7-d-old etiolated barley seedlings versus seedlings raised in diurnalcycles consisting of 16 h of light and 8 h of darkness. Totalhigh mol wt RNA was extracted from the distal two-thirds ofthe leaves (corresponding to segments three through six de-fined in the previous section) ofboth groups of seedlings. TheRNA fractions were assayed for relative steady state mRNAcontent by northern blot hybridization. Unlike the case oftheleafcellular differentiation gradient, where rca and rbc subunitmRNA expression appears to be tightly coordinated, pro-longed etiolated growth appears to elicit three different re-sponses in mRNA accumulation (Fig. 4a). Rca mRNA levelsare barely, if at all, detectable in etiolated barley seedlings.RNA fractions isolated from light-grown seedlings, however,contain at least 50- to 100-fold higher steady state levels of

G Ea

rca

G E G E

rbcS rbcL

b

o .5 1 3 6 12 B

Time of Illumination (hr)

Figure 4. a, Northern blot hybridization assay of the relative steadystate levels of rca, rbcS, and rbcL mRNAs in the primary leaf ofgreen (G) and etiolated (E) barley. Ten zg of total RNA were fraction-ated in formaldehyde-agarose gels, transferred to nitrocellulose, andprobed with 32P-labeled RNA or DNA probes. Two independent RNApreparations tested with the rca probe are shown. b, Kinetics ofaccumulation of rca mRNA upon greening of etiolated barley. Slotblot hybridization was performed on 0.5-gg (top panel) or 5-jsg(bottom panel) aliquots of total RNA isolated from seedlings exposedto white light for the indicated periods of time.

rca mRNA, as estimated by densitometry of Northern blotautoradiographs. In contrast to rca mRNA, and in agreementwith previous work (28), rbcS mRNA levels differ little inetiolated versus light-grown barley. Steady state levels of rbcLmRNA, on the other hand, accumulate during etiolatedgrowth to levels that are intermediate between the responsesseen for rca and rbcS messengers (about a 5-fold higher levelof accumulation in the light of this experiment). Thus, coor-dination of rubisco and rca mRNA expression is dependenton environmental conditions in barley.White light induces accumulation of rca mRNA in etiolated

barley leaves. The kinetics of this induction are shown inFigure 4b. Detectable amounts of rca mRNA consistentlyaccumulate in this system within 30 min after the onset ofillumination. Maximum levels of the messenger are attainedafter 6 to 12 h. In contrast, others have shown that cabmRNAs require at least 2 h of continuous white light toaccumulate to detectable levels (15).Although light induces accumulation of rca mRNA in 7-d-

old etiolated barley, it has a much less dramatic effect on thesteady state level of rca polypeptides detectable by Westernimmunoblotting (Fig. Sa). In this experiment, both the 46-and 4 l-kD rca polypeptides accumulated in the dark to levelsabout 50 to 75% of those observed in seedlings grown undernormal diurnal cycles. Rubisco subunit levels responded in asimilar manner under these experimental conditions (Fig. Sb).It should be noted that in this experiment, several proteaseinhibitors were included in the protein extraction buffer (see"Materials and Methods"), but apparent degradation of the46-kD rca polypeptide is still observed. This proteolytic deg-radation produces polypeptides with apparent mol wt ofabout44, 43, and more significantly, 41 kD. Thus, the differencesin the relative ratios of the 46- and 41-kD rca polypeptidesextracted from etiolated and green tissues seen in Figure 5amay be artifacts, and should be interpreted with caution.Clarification of this point awaits the establishment of extrac-

519

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 90, 1989

b

1 2 3 4 1 2 3 4

rca

_46\ ~~~ ;;; ~rbc L

_.......~~~~~~~~~

b

Figure 5. Comparison of relative rca and rubisco levels in etiolated(lanes 1 and 3) and green (lanes 2 and 4) barley. Total soluble proteinswere extracted and fractionated by SDS-PAGE. a, Immunologicaldetection of rca; b, Coomassie blue staining of total protein. Extractsin this experiment were assayed on the basis of equal total protein(lanes 1 and 2), or equal fresh weight of starting tissue (lanes 3 and4). Unlabeled lanes contain mol wt standards.

tion conditions under which the structure of the 46-kD rcapolypeptide is quantitatively preserved.

DISCUSSION

In this study, we exploited both the positional gradient ofcellular differentiation and photomorphogenic response of 7-d-old barley leaves to examine the expression of the rubiscoactivase gene (rca), and to ask whether rca expression parallelsthat of rubisco, the enzyme it regulates in vivo. As expectedfor a protein that functions in the photosynthetic carbonreduction cycle, we found rca mRNA expression to be tissue-specific, with detectable steady state amounts ofthe messengerpresent only in leaves. Our results confirm and extend pre-vious reports (6, 7, 27), which showed that steady state levelsof rbcS mRNA and rubisco parallel one another during barleyleaf development. In the acropetal developmental gradient ofbarley leaves, rca, rbcS, and rbcL polypeptides and mRNAsdisplay qualitatively similar patterns of accumulation (Figs. 2and 3). The relative level of each polypeptide appears to be a

function of the steady state content of the correspondingmRNA along the barley leaf axis. An exception to this gen-eralization appears at the tip region of the leaf where mRNAlevels decline but polypeptide levels show little, ifany, change.We take these results to indicate that a primary control over

the expression of these proteins lies at the level of mRNAproduction and accumulation. In the oldest region of the leaf,rca and rubisco proteins are maintained at high levels, whilethe corresponding relative mRNA levels decline. This obser-vation is consistent with the idea that, in the oldest cells, therate ofone or more ofthe steps in production ofthese specificmRNAs is greatly decreased or the turnover rate of thesemessengers is greatly increased. Maintenance of high rca

protein levels in the face of dramatic reductions in mRNAcontent implies that rca polypeptides are relatively long livedcompared with their corresponding messengers in older leafcells. In contrast to the patterns of expression observed forrca and rubisco subunit messengers, mRNAs encoding cab(Fig. 2) and calmodulin (30) polypeptides accumulate in thebarley leafgradient in a completely different manner. In thesecases, maximum steady state mRNA levels are found in themidleaf and basal meristematic regions, respectively. These

observations support the idea that the coordination we ob-served between rca and rbc mRNA and protein accumulationin the barley leaf developmental gradient is not an experi-mental artifact.The 46- and 41-kD rca polypeptides are independently

capable of catalyzing rubisco activation in vitro (JM Werneke,J Shen, WL Ogren, manuscript in preparation), but the natureand extent of rca polypeptide interaction in vivo is not yetclear. The two rca polypeptides are produced by two distinctmRNAs of nearly identical size that are derived from one rcastructural gene by alternative mRNA splicing in spinach andArabidopsis (JM Werneke, JM Chatfield, WL Ogren, unpub-lished data), and probably in barley (RE Zielinski, and SJRundle, unpublished experiments); the two rca polypeptidesdiffer only in their carboxy-terminal amino acid regions.These observations originally prompted us to ask whetherdifferential expression of the two rca polypeptides serves as aregulatory mechanism to ensure optimal rubisco activation.In no instance, however, have we observed alterations in theamounts of the two rca polypeptides relative to one anotherthat could be attributed to developmental regulation. On theother hand, because ofthe proteolytic susceptibility of the 46-kD rca polypeptide, there may be subtle changes that wereundetectable under our assay conditions. Thus, a definitiveanswer to this question awaits development of extractionconditions that preserve the integrity of the 46-kD rca poly-peptide. A second, more subtle point in the design of ourexperiments, which may have obscured differential rca poly-peptide expression, is the cellular complexity of barley leaftissue. The leaf dissections performed in this work enabled usto examine gene expression in cell populations of similardevelopmental age. However, we note that each leaf segmentis composed of several different cell types. Thus, our resultsdo not preclude the possibility that the 46- and 41-kD rcapolypeptides are differentially expressed in a spatial manneralong the lateral dimension of barley leaves.

Seven-d-old, etiolated barley seedlings contain very lowlevels of rca mRNA, while maintaining relative levels of rcapolypeptides that are comparable to those found in seedlingleaves grown under a normal illumination regime (Figs. 4 and5). Under these growth conditions, accumulation of rcamRNA is strongly influenced by white light (Fig. 4b). Incontrast, rubisco subunit mRNAs are readily detectable afterprolonged etiolated growth. During etiolated growth, certainmRNAs, such as rca mRNA, may be programmed for morerapid turnover, or their transcription may be turned off inorder to conserve seed reserves. When a light signal sufficientto initiate photomorphogenesis is received by the etiolatedseedling, rca mRNA could once again be produced withoutdepleting the carbon and nitrogen reserves of the seed. Oneway to test this hypothesis would be to assay the rca mRNAlevels in etiolated leaves at different times following germi-nation. Our expectation is that rca messenger will be detect-able during the first few days after germination, after which itwill decline as the levels of seed reserves decline. Nuclear run-on transcription assays could then be used to infer whethercontrol is directed over transcription or a later stage in geneexpression.

Chloroplast biogenesis requires coordinate expression of

520 ZIELINSKI ET AL.

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.

RUBISCO ACTIVASE EXPRESSION IN BARLEY

both the plastid and nuclear genomes. The most extensivelystudied example of this coordination is the stromal enzyme,rubisco. Production of the two rubisco subunits is highlyregulated; little, if any, free subunit protein can be found invivo under normal physiological conditions (7, 18). The resultspresented in this study demonstrate that rca mRNA andpolypeptide accumulation is coordinated with that of rubiscosubunit mRNA and protein in the course of normal barleyleaf cell development. Our data suggest that a constant stoi-chiometry is maintained between rca, rbcS, and rbcL poly-peptides during this developmental process. Formal proof ofthis point, however, awaits quantitative measurement of thelevels of each polypeptide in question. Perhaps the moststriking feature of this apparent coordinate expression is thatit occurs even though rbcL is highly reiterated (20), rbcS ispresent as a small family of nuclear genes (e.g. 1), and rca isa single copy nuclear gene (29; RE Zielinski, unpublishedexperiments). In the future, manipulation of the level of rca

expression may provide insight into the mechanisms by whichcoordinate regulation of plant gene expression is achieved indeveloping leaves.

ACKNOWLEDGMENTS

We thank Tom Jacobs, Archie Portis, and Buddy Orozco forcritically reading the manuscript, and Gloria Coruzzi for kindlyproviding the wheat small subunit probe.

LITERATURE CITED

1. Bedbrook JR, Smith SM, Ellis RJ (1980) Molecular cloning andsequencing of cDNA encoding the precursor to the smallsubunit of chloroplast ribulose-1,5-bisphosphate carboxylase.Nature 287: 692-697

2. Blake MS, Johnston KH, Russel-Jones GL, Gotschlich EC(1984) A rapid, sensitive method for detection of alkalinephosphatase-conjugated anti-antibody on Western blots. AnalBiochem 136: 175-179

3. Broglie R, Bellemare G, Bartlett SG, Chua N-H, Cashmore AR(1981) Cloned DNA sequences complementary to mRNAsencoding precursors to the small subunit of ribulose-1,5-bis-phosphate carboxylase and a chlorophyll a/b binding polypep-tide. Proc Natl Acad Sci USA 78: 7304-7308

4. Broglie R, Coruzzi G, Keith B, Chua N-H (1983) Structuralanalysis of nuclear genes coding for the precursor to the smallsubunit of wheat ribulose-1,5-bisphosphate carboxylase. Bio-technology 1: 55-61

5. Brooks A, Portis AR (1988) Protein-bound ribulose bisphosphatecorrelates with deactivation of ribulose bisphosphate carbox-ylase in leaves. Plant Physiol 87: 244-249

6. Dean C, Leech RM (1982) Genome expression during normalleaf development. I. Cellular and chloroplast numbers andDNA, RNA, and protein levels in tissues of different ageswithin a seven-day-old wheat leaf. Plant Physiol 69: 904-910

7. Dean C, Leech RM (1982) The co-ordinated synthesis of thelarge and small subunits of ribulose bisphosphate carboxylaseduring early cellular development within a seven day wheatleaf. FEBS Lett 140: 113-116

8. Farquhar GD, Sharkey TD (1982) Stomatal conductance andphotosynthesis. Annu Rev Plant Physiol 33: 317-345

9. Feinberg AP, Vogelstein B (1983) A technique for radiolabelingDNA restriction fragments to high specific activity. AnalBiochem 132: 6-13

10. Goldthwaite JJ, Bogorad L (1971) A one step method for theisolation and determination of leaf ribulose-1,5-diphosphatecarboxylase. Anal Biochem 41: 57-66

11. Lacy MJ, Voss EW (1986) A modified method to induce poly-clonal ascites fluid in BALB/c mice using sp 2/0-Ag'4 cells. JImmunol Methods 87: 169-177

12. Laemmli UK (1970) Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227:680-685

13. Leech RM (1985) The synthesis ofcellular components in leaves.In NR Baker, WJ Davies, CK Ong, eds, Control of LeafGrowth, Cambridge University Press, Cambridge, UK, pp93-113

14. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Proteinmeasurement with the Folin phenol reagent. J Biol Chem 193:265-275

15. Mathis JN, Burkey KO (1987) Regulation of light-harvestingchlorophyll protein biosynthesis in greening seedlings. PlantPhysiol 85: 971-977

16. McIntosh L, Poulson C, Bogorad L (1980) Chloroplast genesequence for the large subunit of ribulose bisphosphate carbox-ylase of maize. Nature 288: 556-560

17. Mullet JE (1988) Chloroplast development and gene expression.Annu Rev Plant Physiol 39: 475-502

18. Nivison HT, Stocking CR (1983) Ribulose bisphosphate carbox-ylase synthesis in barley leaves. A developmental approach tothe question of coordinated subunit synthesis. Plant Physiol73: 906-911

19. Parry MAJ, Keys AJ, Foyer CH, Furbank RT, Walker DA(1988) Regulation of ribulose-1,5-bisphosphate carboxylase ac-tivity by the activase system in lysed spinach chloroplasts.Plant Physiol 87: 558-561

20. Possingham JV (1980) Plastid replication and development inthe life cycle of higher plants. Annu Rev Plant Physiol 31:113-129

21. Portis AR, Salvucci ME, Ogren WL (1987) Activation of ribu-losebisphosphate carboxylase/oxygenase at physiological CO2and ribulosebisphosphate concentrations by rubisco activase.Plant Physiol 82: 967-971

22. Salvucci ME, Portis AR, Ogren WL (1985) A soluble chloroplastprotein catalyzes ribulosebisphosphate carboxylase/oxygenaseactivation in vivo. Photosynth Res 7: 193-201

23. Salvucci ME, Werneke JM, Ogren WL, Portis AR (1987) Puri-fication and species distribution of rubisco activase. PlantPhysiol 84: 930-936

24. Schmidt GW, Bartlett SG, Grossman AR, Cashmore AR, ChuaN-H (1981) Biosynthetic pathways oftwo polypeptide subunitsof the light-harvesting chlorophyll a/b protein complex. J CellBiol 91: 468-478

25. Somerville CR, Portis AR Jr, Ogren WL (1982) A mutant ofArabidopsis thaliana which lacks activation of RuBP carbox-ylase in vivo. Plant Physiol 70: 381-387

26. Streusand VJ, Portis AR Jr (1987) Rubisco activase mediatesATP-dependent activation of ribulose bisphosphate carboxyl-ase. Plant Physiol 85: 152-154

27. Viro M, Kloppstech K (1980) Differential expression of the genesfor ribulose-1,5-bisphosphate carboxylase and light-harvestingchlorophyll a/b protein in the developing barley leaf. Planta150: 41-45

28. Viro M, Kloppstech K (1983) Gene expression in the developingbarley leaf under varying light conditions. Planta 157:202-208

29. Wemeke JM, Zielinski RE, Ogren WL (1988) Structure andexpression of spinach leaf cDNA encoding rubisco activase.Proc Natl Acad Sci USA 85: 787-791

30. Zielinski RE (1987) Calmodulin mRNA in barley (Hordeumvulgare L.): apparent regulation by cell proliferation and light.Plant Physiol 84: 937-943

521

www.plantphysiol.orgon October 6, 2020 - Published by Downloaded from Copyright © 1989 American Society of Plant Biologists. All rights reserved.