alternative promoters are used for genes within maize ... · the plant cell, vol. 2, 323-333, april...

The Plant Cell, Vol. 2, 323-333, April 1990 O 1990 American Society of Plant Physiologists

Alternative Promoters Are Used for Genes within Maize Chloroplast Polycistronic Transcription Units

Jean Haley and Lawrence Bogorad’ Department of Cellular and Developmental Biology, Harvard University, Cambridge, Massachusetts 021 38

Many chloroplast genes are co-transcribed in polycistronic transcription units that give rise to numerous overlapping RNAs, but the significance of this pattern of transcript accumulation is not understood. An analysis of the transcripts of the adjacent and divergent maize psbE-psbF-psbL-ORF40 and ORF31- petE-ORF42 gene clusters indicates that transcription initiation at alternative promoters contributes to the generation of overlapping RNAs for both clusters. Furthermore, developmentally varying transcript ratios for the ORF31-petE-ORF42 gene cluster are determined at least in part by selective promoter usage. During light-induced plastid maturation, increased levels of primarily monocistronic petE transcripts accumulate from a promoter upstream of the interna1 petE gene. Dark-predominant and non-light-responsive bi- and tricistronic transcripts result from transcription initiation upstream of ORF31, the proximal gene of the cluster. In addition to the transcriptional overlap within gene clusters, divergent transcription units for the two gene clusters overlap and reciproca1 antisense RNAs accumulate. The organization of the transcription units in this region raises the possibility of promoter interdependence or other functional interaction between transcription units.

INTRODUCTION

Many chloroplast polycistronic transcription units are char- acterized by the accumulation of numerous heterogeneous and overlapping RNAs. Gene sequences can be carried on both mono- and polycistronic transcripts in various patterns of transcript redundancy that are not understood. The overlapping RNAs of a polycistronic transcription unit may accumulate to different steady-state levels, and ratios of transcripts may vary during light-induced plastid development (Rodermel and Bogorad, 1985; Hudson et al., 1987; Rock et al., 1987; Gamble et al., 1988; Westhoff and Herrmann, 1988; Woodbury et al., 1988; for review, see Gruissem, 1989).

The current model used to explain the generation of complex transcript accumulation patterns for co-transcribed chloroplast genes is largely generalized from analyses of the transcription of thepsbB-psbH-petB-petD gene cluster. The model assumes that heterogeneous transcripts accumulate as a result of extensive post-transcriptional processing of a large primary polycistronic transcript. A primary precursor transcript for the psbB gene cluster has been identified and mapped by in vitro transcription (Westhoff and Herrmann, 1988) and in vitro capping (Koh- chi et al., 1988); multiple mRNA processing events, including splicing, endonucleolytic cleavage, and 5’ and 3‘ end- trimming, have been deduced from transcript accumulation patterns for the cluster in maize, tobacco, and spinach

’ To whom correspondence should be addressed.

(Rock et al., 1987; Tanaka et al., 1987; Westhoff and Herrmann, 1988). Transcripts of other complex transcription units have been mapped (e.g., Hudson et al., 1987; Gamble et al., 1988), but the origins of the transcripts have not been analyzed. A corollary to the transcript-processing model derived from analyses of the psbB gene cluster is the proposal (Gruissem et al., 1988) that developmentally regulated transcript processing or stability, rather than initiation, accounts for developmental changes in the transcript ratios of complex transcription units.

We have analyzed the transcripts of two adjacent and divergent maize gene clusters. The psbE-psbF-psbL- ORF40 cluster encodes three polypeptide subunits of the photosystem II photosynthetic electron transport complex of the thylakoid membrane (Westhoff et al., 1985; Widger et al., 1985; lkeuchi et al., 1989; Webber et al., 1989; for review of photosystem II, see Rochaix and Erickson, 1988) and one unidentified polypeptide. The divergent ORF31- petE-ORF42 gene cluster encodes a polypeptide subunit of the photosynthetic electron transport cytochrome b6-f complex (Haley and Bogorad, 1989; for review of the cytochrome bs-f complex, see Hauska et al., 1983) and two unidentified polypeptides. We find that the 5’ hetero- geneity of some of the overlapping transcripts of both gene clusters results from transcription initiation at more than one promoter; other 5’ termini probably derive from the processing of precursor transcripts. During light-induced chloroplast maturation, increased levels of severa1 mono-

324 The Plant Cell

cistronic perE transcripts accumulate from a promoter interna1 to the transcription unit expressing the entire ORF31 -perE-ORF42 gene cluster. Thus, developmentally varying transcript ratios within this polycistronic transcription unit are determined in part by selective promoter usage. The overlap of transcription units for the two gene clusters in this region raises the possibility of functional interaction between them, including promoter interactions that may influence transcription.

R E SU LTS

Transcripts of the Divergent psbE-psbF-psbL-ORF40 and ORF31-petE-ORF42 Gene Clusters

Figure 1 shows the gene organization and the position in the maize chloroplast genome of the divergent psbE-psbF- psbL-ORF40 and ORF31 -petE-ORF42 gene clusters. A series of DNA probes spanning the 4.0-kb region containing the divergent, adjacent gene clusters was used for RNA gel blot analyses of transcript accumulation in maize seedling leaves. Figure 2 shows RNA gel blots of transcripts complementary to DNA probes A to H, which accumulate in leaves of dark- and light-grown seedlings (“D” and “L” lanes, respectively, top panel). The diagram in Figure 2 (bottom panel) indicates the approximate size, location, and relative abundance of the detected transcripts as well as their relative accumulation in dark- and light-grown seedling leaves. The transcripts are assigned to groups I to VI1 based on the positions of their 5’ termini as established by RNA gel blot and S1 nuclease mapping analyses shown in Figures 2 and 3, respectively. More than 25 transcripts of varying abundance and varying ratios of accumulation in the leaves of dark- and light- grown plants hybridize to this region. The eight transcripts in groups I1 and 111 contain sequences of the psbE-psbF- psbL-ORF40 gene cluster, whereas the 17 transcripts in groups I , IV, V, and VI contain sequences of the ORF31- petE-ORF42 cluster. Two transcripts in group VI1 map downstream of ORF42 and are not analyzed here. The tRNAs are transcribed divergently from the ORF31 gene cluster.

The 1 .O-kb intergenic sequence between the proximal genes of the divergent clusters, psbE and ORF31, is transcribed from both strands and is carried on severa1 dark-predominant, low-abundance transcripts that trav- erse the entire region in both directions and proceed through the flanking coding regions of both clusters (Figure 2, probes B to D, transcript groups I and 11). Thus, the initial 1 .O-kb leader sequences of the overlapping group I and I/ transcripts are reciproca1 antisense RNAs that accumulate primarily in the etioplasts of dark-grown leaves. After dark-grown seedlings have been illuminated for about 72 hr, these transcripts decrease in abundance to levels

y 120 20% \ ”=j-m Zea mays uKb* 80

70 6o 919’

Bam HI

D

Figure 1. Location on the Maize Chloroplast Chromosome and Organization of the Divergent psbE-psbF-psbL-ORF4O and ORF31 -petE-ORF42 Gene Clusters.

Filled boxes represent coding regions; those above the line are transcribed to the right, and those below the line are transcribed to the left, as depicted by arrows. The region analyzed in this paper cornprises the maize chloroplast BarnHl fragrnent 15’ (2760 bp) and two neighboring BamHl fragments of 835 bp and 320 bp not previously mapped (nucleotides 43080 to 46995 of the maize chloroplast map) (Larrinua et al., 1983). The rnaize coding sequences are organized similarly to the corresponding chloroplast genes of liverwort (nucleotides 62794 to 65155) (Ohyama et al., 1986), tobacco (nucleotides 66170 to 69418) (Shinozaki et al., 1986), and rice (nucleotides 61565 to 64756) (Hiratsuka et al., 1989).

fourfold to fivefold lower than those found in dark-grown leaves and similar to the levels in chloroplasts of light- grown leaves (Figure 2, probes B to D, compare “D” and “L” lanes; greening series not shown).

A number of more abundant and overlapping (same- sense) transcripts hybridize to the coding regions of the two clusters. The six tetracistronic transcripts of group 111 (Figure 2, diagram) contain the entire psbE gene cluster and have 5’ ends that map near the proximal psbE gene at a position approximately 1 kb downstream of the 5’ termini of the group /I psbE cluster transcripts. The major transcript of group 111, the 1 . l-kb transcript, accumulates to about the same leve1 in leaves of both dark- and light- grown plants (probes A and B) and exhibits only a minor

Alternative Promoters for Plastid Genes 325

A B C D E F Q H1 D L " D L' 'D L"D L"D L' ' o L' ' D L' ' D L '

«.o-|2.8-2.3-z.r1.7- '1.4-,1.1-l|0.8-

**

Probes:B C D E F G H

~u——ii———ii——ii————i r~rOU_ UJ

0= UO Q.

/ :21^

4.0 .2.6 •1.7 •

O— • 0.22

Figure 2. RNA Gel Blot Analyses and Schematic Diagram ofTranscripts.

Total leaf RNA (5 ng) isolated from dark-grown (D) or light-grown(L) maize seedling leaves was separated by gel electrophoresisand analyzed with a series of double-stranded DNA probes (A toH) containing different regions of the two gene clusters (shown attop and center). The schematic diagram (bottom) shows tran-scripts organized into groups / to VII based on the positions oftheir 5' termini and their direction of transcription [determined bysingle-stranded M13 probes (data not shown) and designated byarrows]. Sizes of transcripts (in kilobases) are shown at left (forpst>E-ps£>F-pst>L-ORF40 gene cluster transcripts transcribed tothe left) or at right (for ORF31-pefE-ORF42 cluster transcriptstranscribed to the right). The approximate relative abundance oftranscripts is indicated by the thickness of arrow lines and thesize of circles representing 5' transcript termini, both of whichincrease with increased relative abundance. The 5' circle alsoindicates whether a transcript accumulates predominantly in dark-grown leaves (•), predominantly in light-grown leaves (O), or tosimilar levels in both dark- and light-grown leaves (O).

transient increase during chloroplast maturation triggeredby the illumination of etiolated leaves (Sheen and Bogorad,1988). The other group /// transcripts are much less abun-dant and either have an accumulation pattern similar tothat of the 1.1-kb transcript or are present primarily inetiolated leaves.

Each of the three genes in the ORF31-pefE-ORF42cluster is carried on multiple heterogeneous transcripts,most of which are polycistronic. Levels of these transcriptseither decrease, increase, or do not change appreciably(other than a slight transient increase) upon illumination ofdark-grown seedlings (greening series not shown; abun-dance in leaves of dark- and light-grown plants shown in"D" and "L" lanes, probes B to H). In addition to the fourlow-abundance, dark-predominant group / transcripts thatcontain the coding regions of the gene cluster, there arethree groups of ORF31 cluster transcripts: (1) group IVtranscripts have 5' ends that map at several sites near theproximal ORF31 gene and are either more abundant indark-grown leaves than in light-grown leaves or accumu-late to similar levels in both types of leaves, (2) group Vtranscripts have 5' ends that map to a region in front ofthe internal pefE gene and share a pattern of an approxi-mately fourfold to fivefold relative increase in abundanceupon the illumination of dark-grown plants for 48 hr andlonger (greening series not shown; relative levels in dark-and light-grown leaves shown in Figure 2, probes E to H),(3) group VI transcripts have 5' ends that map upstreamof the distal ORF42 gene and accumulate similarly in dark-and light-grown leaves (probe H). Whereas transcripts ofthe psbE gene cluster accumulate to much higher levelsin mesophyll than in bundle sheath cells of maize (Sheenand Bogorad, 1988), members of all four groups of ORF31cluster transcripts accumulate to approximately the samelevel in both leaf cell types (data not shown).

Transcripts of Both Gene Clusters Are Initiated atMore than One Promoter

Figures 3, 4, and 5 show the S1 nuclease protectionmapping of 5' transcript termini and "Northern-Cross"hybridization analyses of in vitro capped plastid RNA.These approaches were used to determine the origins ofthe overlapping transcripts of the two gene clusters. Takentogether, data from RNA gel blot analyses, S1 nucleaseprotection mapping of transcript 5' termini, and the analy-sis of capped RNAs indicate that transcripts of both geneclusters are initiated at more than one promoter.

A Group III psbE Gene Cluster Transcript Arises byTranscription Initiation

The major 1.1-kb group /// transcript of the psbE genecluster has been observed in spinach (Westhoff et al.,

326 The Plant Cell

B

5'probe •—

mRNA —---

-• le)—• (ci

Figure 3. S1 Nuclease Protection Fine-Mapping Analysis of the 5' Termini of Transcripts.Total leaf RNA (50 ng) isolated from either dark-grown (DARK) or light-grown (LIGHT) maize seedling leaves was hybridized with excess5' end-labeled DMA probe as shown. Protected fragments were mapped against sequencing ladders; 5'-terminal bases of mappedtranscripts are aligned with the complementary ladder sequences.(A) 1.1 -kb psbE gene cluster transcript.(B) Light-induced group V transcripts.(C) 1.35-kb group IV transcript (?).(D) Dark-predominant group / transcripts.(E) 0.53-kb and 1.35-kb group IV transcripts.(F) 0.26-kb group VI transcript.

1985), Oenothera (Carrillo et al., 1986), and wheat (Webberet al., 1989). Figures 3A and 4A show that the 5' terminusof the maize 1.1-kb transcript maps to a position 138nucleotides (nt) upstream of the psbE initiation codon nearsequences resembling chloroplast promoter elements. Asimilar 5' terminus has been established for the wheat 1.1-kb transcript (Webber et al., 1989). The proximity of pro-moter-like sequences suggests that the mapped 5' termi-nus represents the transcription initiation site of the 1.1-kb transcript.

Northern-Cross hybridization analysis (Graham et al.,1986) of in vitro capped chloroplast and etioplast RNAwas used to identify and map primary transcripts. Thehybridization of in vitro capped (32P-GTP-labeled), electro-phoretically separated transcripts to unlabeled DNAprobes permitted the identification of cappable transcriptsby the simultaneous determination of their sizes, map

positions, and approximate locations of their capped 5'termini. Figure 5A shows that the major 1.1-kbpsbE genecluster transcript can be capped in vitro and is, therefore,a primary transcript initiated at the site of its mapped 5'terminus: an abundant, capped 1.1-kb chloroplast tran-script hybridizes uniquely to chloroplast DNA probes 1 and2, which contain thepsbE gene cluster (Figure 5A, -RNasepanel; probes shown in 5D); the capped 1.1-kb transcriptco-migrates with the 1.1-kb psbE cluster transcript de-tected by probe 1 in a standard RNA gel blot assay (atleft). In the second part of the experiment (shown in Figure5A, +RNase panel), the 5'-labeled capped terminus of the1.1-kb transcript is seen to be protected from RNasedigestion by hybridization with DNA probe 2, which ex-tends upstream of probe 1 and contains the region of theS1-mapped 5' terminus of the 1.1-kb group /// transcript,but not by hybridization with DNA probe 1 lacking this 5'


group 111

7 1 . 1 kb . -35 . -10 +1 40 A 5 I . . . AAAACTGGATTGCTGTGCCATAGGAAGGATAGCTATACTAATTCGGTATACT~-TACAC~~TGGTA~TTGA~TCTCACAAGG~TGAAATA

8 0 120 m b E TCAGTAATTTTCTATTTACTGCTGCATCCCATCTTTTTACGGAATCAATTCCTTTTTTGAAT~TTTTGGGAGCTCAGC ATG TCT GGA ... 3'

B a m H I Met Ser Gly

group I V

B 5 ' . . . GTATTATCT "-35" V I - 10 VI ORF31 , , I /

CCCTTC ATG CTT ACT...TGA MTGMTTGAATAGAAGAATCTTTCPTTTTGGATTCTTGGTATTCTAGACTCTTTTCCACACTMTTACCAATTCTTT Met Leu Thr

group v 1.35 kb 1?1

"-35" 11 - 10 *I 1 1°122kb @E TCTTGGTCATTGAGATTCGTGGGTAGACTATTATTTAT~AGAGATAGATCGTACCTCTTTTTTTATCCCCTCG~CAAATCGM ATG ATT GAA...)'

"light-induced" Met Ile Glu

group I I I 1.1 kb 7

(OPPOSITE S T W D ) 3'...AAACATGGmCCACATAAAACTCATATGGCTTMTCATATCGATAGG~GGATACCGTGTCGTTAGGTC...5'

- 3 3 -10 +l 40 t group I "dark"

+1 -10 -35

c 5 ' . . . TTCATCCTTGTGAGATTGTCM~TGTAC~GGTGTAT~TGAGTATACCGMTTAGTATAGCTATCCTTCCTATGGCACAGCMTCCAGTTTTGCTT

8 0 120 GGTCCCGAAACAGAATTCCT~CT~~G~CCTTGTCTATAGG-~ACATGTTATTCAAGGCATCAATAGMCCCCACAATTTTTTGGGTCCTA

160 200 ORF31 CTTATTTTCATTGTCTTCGGAATAGTAGMTM~AATTT~MTAGCGGCCAAGATCTTGGGAAAATCTA...875 nt... ATG CTT ACT ... 3'

BglII Met Leu Thr

grovp VI

n o.26 kb ORF 42 D 5 ' . . . AAATACAAAGGATCTTGGGCMGAGTATCTGATCATATATGTATTCCMTACGGAAGGAGGATTTTCA ATG CGG GAT . . . 3 ' Met Arg Asp

Figure 4. Nucleotide Sequences (RNA-Like Strand) Containing S1 Nuclease Protection Mapped 5' Transcript Termini of the Two Gene Clusters.

Arrows above the sequence indicate the mapped termini shown in Figure 3. Sequences resembling "-35" and "-10" chloroplast promoter elements (Hanley-Bowdoin and Chua, 1987) are underlined. The heptanucleotide sequence ATGA/TATT located near processed 5' termini is doubly underlined. Nucleotides are numbered in (A) and (C) beginning with the first base of the group /// and group / transcripts, respectively. (A) Sequence upstream of the psbE gene containing the 5' terminus of the major 1 .I -kb group /// transcript; the BamHl site used for S1 mapping (Figure 3A) is shown. (E) Sequences in the region of the adjacent ORF31 and petE genes containing 5' termini of group /V and the group V light-induced transcripts. (C) Sequence containing the 5' terminus of dark-predominant group / transcripts; the Bglll site used for SI mapping (Figure 3D) is shown. At top is shown the sequence of the opposite strand containing the 5' terminus of the divergent 1 .I-kb group /// transcript for the psbE gene cluster [(A)]. (D) Sequence upstream of the ORF42 gene containing the 5' terminus of the 0.26-kb group VI transcript.

region. The other, less abundant, psbE gene cluster transcripts of groups 11 and 111 are not detected as cappable chloroplast transcripts in vitro and may arise by the post- transcriptional processing of either a primary group 11 transcript (unidentified here) or the cappable group 111 1 . l - kb primary transcript. We conclude that transcripts of the

psbE gene cluster are initiated at two different promoters because: (1) in vitro capping of the major 1 . l-kb group 111 transcript indicates the existence of a functional promoter (PI I I ) upstream of its 5' terminus (shown in Figure 6), and (2) the approximate position of 5' termini for group 11 transcripts mapped by RNA gel blot analysis implies the

328 The Plant Cell

RNA Cappedgel blot Chloroplast RNA

D L -RNase + RNase

RNAgel blot Chloroplast RNA

RNase"

CappedEtioplast RNA

kb3)Z kb

m 1.35*0.95*

055*.

0.22*1 2

DNA —Probes

1 2

kb

DNA —Probes

f« t t2 5 6 2 5 6

DNA ———Probes

2———————— i

r~-*

56

34

oa

Figure 5. Northern-Cross Hybridization Analysis of in Vitro Capped Chloroplast and Etioplast RNA.

Capped, 32P-GTP-labeled RNA (75//g) was fractionated by gel electrophoresis and blotted without fixing onto a GeneScreen membrane;uncapped (unlabeled) total leaf RNA from dark-grown (D) or light-grown (L) seedlings was separated simultaneously in a companion laneand analyzed with labeled DNA probes as for standard RNA gel blots. (RNA gel blots are shown aligned at the left of each hybridizationanalysis above.) Unlabeled DNA probes 1 to 6, containing regions of the two gene clusters, were separated on a second gel, blotted, andfixed to Zeta-Probe membranes. Contact hybridization of the two membranes, oriented at right angles to each other, permitted thediffusion of labeled, capped RNAs and their hybridization to the fixed DNA probes. After autoradiography of the "cross-hybridized" blot,the membrane was treated with RNase A to digest the capped 5' termini of transcripts that were not protected because the hybridizingDNA probes lacked sequences complementary with the termini; cap signals remaining after RNase treatment are considered indicative ofhybridization-protected 5' caps.(A) Autoradiogram of capped Chloroplast RNA "cross-hybridized" to DNA probes 1 and 2 before (—RNase) and after (+RNase) RNase Atreatment of the membrane; RNA gel blot aligned at left is assayed with labeled DNA probe 1.(B) Autoradiogram of capped Chloroplast RNA "cross-hybridized" to DNA probes 3 and 4 before (-RNase) and after (+RNase) RNase Atreatment of the membrane; RNA gel blot aligned at left is assayed with labeled DNA probe 4.(C) Autoradiogram of capped etioplast RNA "cross-hybridized" to DNA probes 2, 5, and 6 before (-RNase) and after (+RNase) RNase Atreatment of the membrane; RNA gel blot aligned at left is assayed with labeled DNA probe containing the psbE-ORF31 intergenic region(corresponding to probes B to D, Figure 2).(D) Map of gene organization, indicating subregions contained in DNA probes 1 to 6 used in the "cross-hybridization" analyses shown in(A) to (C). The S1 -mapped origins of group /, group ///, and group V transcripts are shown (see Figures 3 and 4). Probes 1 and 2 containthepsbE gene cluster; they differ in length by a region containing the mapped 5' termini of group / and group /// transcripts (A = 360 bp).


P I

I- P P t-

c

* implied by existence of transcripts

Figure 6. Schematic Diagram of Locations of Promoter Regions Deduced from Transcript Mapping and Capping Analyses.

PI, PV, and PIII represent promoters shown to exist by the capping of members of transcript groups I , V , and / I / , respectively (Figure 5). PII and the tRNA promoter are implied by the existence of group I/ transcripts and the tRNAs, respectively (Figure 2).

existence of an upstream promoter (PII) (not mapped here) that initiates transcription of the group / I transcripts through the sequence of the downstream PIII promoter.

Group V and Group I ORF31 Gene Cluster Transcripts Arise by Transcription lnitiation

The S1 nuclease protection analysis shown in Figure 3B maps a heterogeneous 5’ terminus with a light-induced protection pattern approximately 40 nt upstream of the petE gene (Figure 48). This position corresponds to the location of the 5’ termini of group V light-induced transcripts as established by RNA gel blot analyses (Figure 2, probe E, and additional probes not shown). Anotherweakly light-induced 5’ terminus maps approximately 45 nt upstream of the strongly light-induced terminus (Figures 3C, 4B); this terminus may correspond to the origin of a second 1.35-kb group I V transcript that maps in this general region according to RNA gel blot analysis (Figure 2, increase in signal at 1.35 kb for probes F to H over signal for probe E). Promoter-like sequences, including several overlapping

“-10” as well as redundant “-35” elements, are found upstream of both termini. Figure 58 shows that the abundant 0.22-kb light-induced group V transcript is cappable in vitro and is, therefore, a primary transcript that arises at a promoter in front of thepetE gene. A capped chloroplast transcript that co-migrates with the 0.22-kb light-induced transcript (shown in the RNA gel blot at left) hybridizes to DNA probes 3 and 4, which contain the petE gene (Figure 5B, -RNase panel); its 5’4abeled capped terminus is protected from RNase digestion by hybridization to DNA probe 4, which contains the region of the S1-mapped light- induced terminus (Fig. 5B, +RNase panel), but not by hybridization to DNA probe 3, which lacks this region. Figure 58 shows three other less abundant, larger, in vitro cappable chloroplast transcripts that also map to this region and that correspond in size to transcripts mapped at this location by RNA gel blot analysis. These include two other group V transcripts, the 0.95-kb transcript and probably the low-abundance 0.55-kb transcript (Figure 2, probe F), and a 1.35-kb transcript that may arise just upstream of the strongly light-induced transcripts or may represent a co-migrating light-induced transcript. None of the four capped transcripts hybridizes to DNA probes specific for ORF31, and only the 1.35-kb capped transcript hybridizes to ORF42 coding sequences (data not shown). We conclude that there is at least one functional promoter (PV) in this region (shown in Figure 6) from which the 0.22- kb transcript and probably several other light-induced group V transcripts encoding primarily the petE gene are transcribed. The PV promoter is interna1 to the transcription unit(s) that promote the synthesis of group I and group I V transcripts through the PV promoter region. The hetero- geneity of the mapped light-induced terminus representing the origin of PV transcription initiation (Figure 38) may result from multiple initiation sites, although 5’ end mRNA trimming or experimental artifact cannot be ruled out.

Figure 3D shows that the dark-predominant, low-abundance group I transcripts encoding the ORF31 gene cluster (Figure 2, probes B to D) share a 5’ terminus that maps by S1 nuclease protection to a position 3 nt from the 5’ terminus of the divergent 1.1 -kb primary group 111 transcript containing the psbE gene cluster (Figure 4C). Redundant promoter-like sequences lie immediately upstream of the mapped terminus. A Northern-Cross hybridization analysis of in vitro capped etioplast RNA (Figure 5C, -RNase panel) detects two cappable transcripts corresponding in Size to the group I dark-predominant 2.6-kb and 1.55-kb transcripts (shown in the RNA gel blot at left in Figure 5C) that are not detected in capped chloroplast RNA (compare the

Figure 5. (continued). Probes 3 and 4 contain a portion of the ORF31 gene cluster including the petE gene and downstream sequences; they differ in length by a region containing the mapped 5’ terminus of the light-induced group V transcripts that extends 120 nt upstream of the petE gene (A = 155 bp). Probes 5 and 6 contain the entire ORF31 gene cluster; they differ in length by a region containing the mapped 5’ terminus of group I transcripts (A = 230 bp).

330 The Plant Cell

hybridization of capped etioplast RNA to DNA probe 2 in Figure 5C with that of capped chloroplast RNA to the same probe 2 in Figure 5A). The 5‘4abeled cap of the 2.6- kb etioplast transcript is largely protected from RNase digestion by hybridization with DNA probes 2 and 5 , which contain the region of the mapped dark-predominant terminus of the group / transcripts, but is unprotected by hybridization with DNA probe 6 lacking this sequence (Figure 5C, +RNase panel). The cap signal of the 1.55-kb transcript hybridized to probe 6 is not abolished by RNase treatment, but it is somewhat diminished by comparison to the 1.55-kb signal protected by hybridization with probes 2 and 5 ; secondary RNA structure may protect the cap from RNase digestion or there may be more than one co-migrating primary 1.55-kb transcript. We conclude that at least one dark-predominant group I transcript (2.6 kb) is initiated at a promoter (PI) (Figure 6). Figure 4C shows that the PI promoter may be located in the region of DNA transcribed from the divergent PlII promoter to yield the initial bases of the 1 . l-kb psbE cluster primary transcript. Thus, adjacent and divergent PI and PIII promoters for the two gene clusters may each initiate transcription within a region complementary to the promoter sequence of the other.

5’ Processing of Transcripts Encoding the ORF31- petE-ORF42 Gene Cluster

Severa1 major non-light-responsive or dark-predominant ORF31 cluster transcripts are not capped in vitro, and their 5’ termini, therefore, may well be determined by post- or co-transcriptional processing. Transcripts in groups IV and VI, whose 5’ termini map upstream of the ORF31 and ORF42 genes, respectively, fall into this category. Figures 3E and 3F show the major 5’ termini mapped in these regions of the gene cluster by S1 nuclease protection. RNA gel blot analyses with probes extending upstream from within the ORF31 gene (data not shown) allow the termini mapped in Figure 3E to be assigned to the 0.53- kb and one of the 1.35-kb group IV transcripts (Figure 48). The terminus mapped in Figure 3F corresponds to the 0.26-kb group VI transcript encoding ORF42 (Figure 4D). Upstream promoter-like sequences are not found near the 5’ termini of these transcripts. Although RNA gel blot analyses clearly indicate that a 1.55-kb dark-predominant transcript must also originate in the region just upstream of the ORF31 gene (Figure 2, increase in signal at 1.55 kb for probes E to H over that for probes B to D), no dark- predominant terminus was identified by S1 nuclease or primer extension mapping. RNA secondary structure may interfere with the mapping of this terminus.

The heptanucleotide sequence ATGA/TATT is found in the DNA sequence near the mapped locations of the 5‘ termini of major noncappable transcripts as well as near other minor S1 -mapped termini (double-underlined se-

quences in Figures 48 and 4D). The heptanucleotide is not found elsewhere in the 4.0-kb DNA region, and its correlation with the position of 5’-processed transcript termini suggests that it may have a role in their processing. This sequence has some resemblance to the hexanucleotide YGGAA/TY associated with the 5‘ termini of psbB gene cluster transcripts thought to be generated by endonucleolytic cleavage (Westhoff and Herrmann, 1988).

DISCUSSION

The results presented here indicate that the multiple heterogeneous transcripts that encode chloroplast gene clusters can arise by transcription initiation from more than one promoter. This is a new finding for chloroplast gene transcription and it not only has interesting implications for the expression of the genes analyzed here but also permits a refinement of the operative concept of the chloroplast polycistronic transcription unit. Each of the gene clusters we describe is not transcribed solely as one large transcription unit defined by a single proximal promoter but rather as overlapping units defined by tandem promoters. The use of more than one promoter for the transcription of chloroplast genes may be unique to the expression of the two gene clusters analyzed here but it seems likely to occur more generally. Thus, it may be instructive to analyze the transcript families within other polycistronic transcription units. This would include, for example, such transcripts as the internal light-induced transcripts of the psbD-psbC cluster (Gamble et al., 1988) and the families of transcripts arising upstream of atpl and atpH in the rps2-atpl-atpH- atpF-atpA cluster (Rodermel and Bogorad, 1985; Hudson et al., 1987).

For the ORF31-pefE-ORF42 gene cluster, where transcript ratios vary during the maturation of chloroplasts induced by the illumination of dark-grown seedling leaves, there is a general correlation between promoter usage and the developmental pattern of accumulation of the transcripts. That is, transcripts that accumulate predominantly in the etioplasts of dark-grown leaves arise from proximal promoters, whereas at least one major transcript that becomes more abundant during light-induced chloroplast maturation originates at a dista1 promoter. The major light- induced transcript encodes the internal petE gene, and- although we have no direct evidence on this point-its accumulation may serve to uncouple the expression of petE from that of ORF31 and ORF42 during and following plastid maturation. It has been suggested that the light- induced accumulation of barley chloroplast psbD-psbC transcripts during plastid maturation is required to maintain translation of the psbD and psbC gene products in mature chloroplasts (Gamble et al., 1988). A similar case may obtain for the expression of the 4-kD cytochrome-bs-f polypeptide encoded by petE.


It is less clear how dual promoter usage may influence the expression of genes in the psbE-psbF-psbL-ORF40 cluster. However, an intriguing characteristic of the transcripts that accumulate from the upstream PII promoter is their complementarity, over their initial 1 kb, to transcripts initiated at the divergent PI promoter of the ORF31 cluster. It is not known whether double-stranded RNAs of the complementary transcript sequences form in vivo; if so, they may provide a functional link between the two overlapping transcription units that is in some way involved in their regulation.

The light-induced increase in accumulation of primary transcripts from the interna1 promoter of the ORF31 gene cluster may result from developmental regulation of either transcript initiation or transcript stability. Currently, it is not known how chloroplast promoters are regulated or to what extent the specific regulation of initiation at a promoter determines the levels and patterns of transcript accumulation. It has been proposed that the differential developmental accumulation of chloroplast transcripts can be ac- counted for entirely by changes in transcript stabilities due to the interaction of developmentally regulated proteins with stem-loop structures at the 3’ transcript termini (Gruissem et al., 1988). The shortest and most abundant of the light-induced family of petE transcripts, the 0.22-kb transcript, may contain a predicted stem-loop structure at its 3‘ end (A G = -20.9 kcal) that could function in such a regulatory scenario for differential stability. However, other light-induced transcripts do not contain stable predicted stem-loop structures near their 3’ termini. If differential stability accounts for the light-induced increase in accumulation of group V transcripts, the mechanism for conferring this stability might be more likely to involve the 5’ sequences (or secondary RNA structures) that are common to all the transcripts. There are, in fact, data from other systems that indicate that 5’ transcript sequences may influence transcript stability (for review, see Brawer- man, 1989).

The accumulation of the petE gene-encoded 4-kD polypeptide during greening (Haley and Bogorad, 1989) paral- lels rather closely the accumulation of the light-induced family of primary petE transcripts. There is a relatively low level of the petE polypeptide in etioplasts, despite the association of many petE-encoding transcripts with poly- somes (data not shown). The petE gene is distal to the ORF31 gene on all the transcripts that accumulate in etioplasts except the relatively scarce group V monocistronic transcripts. Assuming that the 4-kD polypeptide is not turned over rapidly in etioplasts, the low level of its accumulation may be due to inefficient translation from transcripts on which it is not the proximal gene.

The extensive overlapping of transcription units in the region of the two chloroplast gene clusters analyzed here raises severa1 possibilities of promoter-promoter interactions that may regulate transcription initiation. In both prokaryotic and eukaryotic systems, interactions between

tandem or adjacent divergent promoters have been shown to regulate transcription initiation and affect transcript abundance (Adhya and Gottesman, 1982; Proudfoot, 1986; Biswas and Getz, 1988). These interactions include the steric hindrance of RNA polymerase binding at a promoter by its binding at a closely situated promoter and the inhibition of a promoter by transcription through it. In the region analyzed here, the adjacent promoters for group I and group 111 transcripts (PI and PIII, Figure 6) appear to have mutually exclusive RNA polymerase binding sites that might preclude the simultaneous use of both promoters on the same individual template. Furthermore, transcription through the distal promoters of both gene clusters (PIII and PV) might be expected to interfere with initiation at these promoters. Simultaneous convergent transcription of group I and group I/ transcripts and of the tRNAs and ORF31 cluster transcripts might also be expected to be incompatible. It is possible, however, that if promoter-promoter interactions do occur, they may oper- ate differently in plastids containing many copies of a genome than in cells having unicopy genomes. For example, interference between plastid promoters may play a role in determining how many individual chloroplast DNA templates in the total multicopy population are used for transcription at each promoter. It is not known whether all chloroplast chromosomes are functionally equivalent with respect to the transcription of any one gene or cluster at any one time.

METHODS

DNA Sequence

The DNA sequence of both strands of the maize chloroplast DNA fragment BamHl 15’ (Larrinua et al., 1983; derived from plasmid pZmc503) and two neighboring BamHl fragments of 320 bp and 835 bp (Figure 1) was determined by chemical cleavage (Maxam and Gilbert, 1980) and dideoxy chain termination (Sanger et al., 1977). The full sequence has been deposited in the GenBank@ EMBL Data Bank (accession no. J04502); partia1 sequences are shown in Figure 4. The t m P , tmW, and petE sequences have been published elsewhere (Lukens and Bogorad, 1988; Haley and Bogorad, 1989).

RNA Preparation and Gel Blot Analysis

Total leaf RNA was prepared from leaves of maize [Zea mays (FR9 cms x FR37) lllinois Foundation Seed] seedlings grown for 7 days in darkness or in the greenhouse. The apical 5 cm of leaves were harvested into liquid nitrogen and extracted with 4 M guanidinium thiocyanate (Maniatis et al., 1982). RNA was separated on 1.2% agarose Mops-formaldehyde gels and transferred to Zeta-Probe membranes in 50 mM NaOH. RNA gel blots were hybridized with DNA probes labeled by random hexamer priming (Pharmacia LKB Biotechnology Inc.) in 250 mM sodium phosphate (pH 7.2), 7% SDS, and 1 mM EDTA at 65°C for 16 hr to 30 hr.

332 The Plant Cell

S1 Nuclease Protection Assays

S1 nuclease protection of the 5' transcript termini was performed using kinased double-stranded (ds) DNA probes. Total leaf RNA was hybridized with excess denatured dsDNA probe in 80% formamide hybridization buffer (Favoloro et al., 1980). The RNA- DNA hybrids were treated with S1 nuclease, and protected DNA fragments were sized on denaturing acrylamide gels next to chemical cleavage sequencing ladders of the 5' end-labeled pro- tecting probes.

In Vitro Capping of FINA and Norihern-Cross Hybridization Analysis

Etioplast or chloroplast RNA (75 pg) was capped in vitro in a 40- pL reaction mixture containing 50 mM Tris-HCI (pH 7.9), 1.25 mM MgC12, 6 mM KCI, 2.5 mM DTT, 80 units of RNasin, 350 pCi of ~u -~~P-GTP (3000 Ci/mmol, Du Pont-New England Nuclear), and 1 O units of guanylyltransferase (Bethesda Research Laboratories) for 60 min at 37OC. A Northern-Cross hybridization method (adapted from that of Graham et al., 1986) was used to analyze the capped RNA. Capped, labeled RNA was separated electrophoretically across a 1.2% Mops-formaldehyde gel and transferred overnight in 10 x SSC to a GeneScreen membrane without subsequent immobilization. Unlabeled DNA fragments representing various subregions of the two gene clusters were separated electrophoretically across the width of an agarose gel and transferred (and fixed) in 0.4 M NaOH to a Zeta-Probe membrane. Contact-hybridization of the two membranes oriented at right angles to each other was performed for 16 hr to 20 hr at 42OC in 50% formamide, 5 x SSC, 50 mM NaP04 (pH 7.0), 0.2% SDS, and 250 pg/mL salmon sperm DNA. Following autoradiography of the "cross-hybridized membrane, the membrane was treated with RNase [250 pg of RNase A in 16.7 mL of incubation buffer containing 10 mM Tris-HCI (pH 7.5), 5 mM EDTA, and 300 mM NaCI] for 30 min at 37OC to digest labeled caps that were not protected by hybridization to the DNA probe sequences.

ACKNOWLEDGMENTS

We thank Drs. Alan D. Blowers, Alice Cheung, Steven R. Roder- mel, and Robert Troxler for helpful discussions. This work was supported in part by a research grant from the National lnstitute of General Medical Sciences.

Received August 7, 1989; revised February 4, 1990.

NOTE ADDED IN PROOF

Yao et al. [Nucl. Acids Res. (1989). 17,9583-95911 have recently reported finding two transcription initation sites in the psbD-psbC gene cluster of the tobacco chloroplast chromosome, and Wood- bury et al. [Curr. Genet. (1989). 16, 433-4451 have found that

this is also the case for the same cluster in the pea chloroplast genome. Data in the latter paper also indicate that there may be more than one transcription initiation site for the rps2-atpl-atpH- atpF-atpA cluster in pea.

REFERENCES

Adhya, S., and Gottesman, M. (1982). Promoter occlusion: Tran- scription through a promoter may inhibit its activity. Cell 29, 939-944.

Biswas, T.K., and Getz, G.S. (1 988). Promoter-promoter interactions influencing transcription of the yeast mitochondrial gene, Oli 1, coding for ATPase subunit 9:Cis and trans effects. J. Biol. Chem. 263,4844-4851.

Brawerman, G. (1989). mRNA decay: Finding the right targets. Cell 57, 9-10,

Carrillo, N., Seyer, P., Tyagi, A., and Herrmann, R.G. (1986). Cytochrome b-559 genes from Oenothera hookeri and Nicotiana tabacum show a remarkably high degree of conservation as compared to spinach. The enigma of cytochrome b-559: Highly conserved genes and proteins but no known function. Curr. Genet. 10, 619-624.

Favaloro, J., Treisman, R., and Kamen, R. (1980). Transcription maps of polyoma virus-specific RNA: Analysis by two-dimen- Sional nuclease S1 gel mapping. Methods Enzymol. 65,

Gamble, P.E., Sexton, T.B., and Mullet, J.E. (1988). Light-de- pendent changes in psbD and psbC transcripts of barley chloroplasts: Accumulation of two transcripts maintains psbD and psbC translation capability in mature chloroplasts. EMBO J. 7,

Graham, D.E., Xu, Y.-H., Ishii, S., and Merlino, G.T. (1986). Northern Cross hybridization for rapid identification of exon- containing restriction fragments. Gene 48, 241 -249.

Gruissem, W. (1 989). Chloroplast gene expression: How plants turn their plastids on. Cell 56, 161 -1 70.

Gruissem, W., Barkan, A., Deng, X.-W., and Stern, D. (1988). Transcriptional and post-transcriptional control of plastid mRNA levels in higher plants. Trends Genet. 4, 258-263.

Haley, J., and Bogorad, L. (1989). A 4-kDa maize chloroplast polypeptide associated with the cytochrome b6-f complex: Sub- unit 5, encoded by the chloroplast petE gene. Proc. Natl. Acad. Sci. USA 96,1534-1538.

Hanley-Bowdoin, L., and Chua, N.-H. (1 987). Chloroplast promoters. Trends Biochem. Sci. 12, 67-70.

Hauska, G., Huri, E., Gabellini, N., and Lockau, W. (1983). Comparative aspects of quinol-cytochrome c/plastocyanin oxi- doreductases. Biochim. Biophys. Acta 726, 97-1 33.

Hiratsuka, J., Shimada, H., Whittier, R., Ishibashi, T., Saka- moto, M., Mori, M., Kondo, C., Honji, Y., Sun, C.-R., Meng, B.-Y., Li, Y.-Q., Kanno, A., Nishizawa, Y., Hirai, A., Shinozaki, K., and Sugiura, M. (1989). The complete sequence of the rice (Oryza sativa) chloroplast genome: lntermolecular recombina- tion between distinct tRNA genes accounts for a major plastid DNA inversion during the evolution of the cereals. MOI. Gen. Genet. 217,185-1 94.

Hudson, G.S., Mason, J.G., Holton, T.A., Koller, B., Cox, G.B.,

71 8-749.

1289-1 297.


Whitfeld, P.R., and Bottomley, W. (1987). A gene cluster in the spinach and pea chloroplast genomes encoding one CF, and three CFo subunits of the H+-ATP synthase complex and the ribosomal protein S2. J. MOI. Biol. 196, 283-298.

Ikeuchi, M., Takio, K., and Inoue, Y. (1989). N-terminal sequencing of photosystem II low-molecular-mass proteins: 5 and 4.1 kDa components of the 02-evolving core complex from higher plants. FEBS Lett. 242, 263-269.

Kohchi, T., Yoshida, T., Komano, T., and Ohyama, K. (1988). Divergent mRNA transcription in the chloroplast psbB operon.

Larrinua, L., Muskavitch, K.M.T., Gubbins, E., and Bogorad, L. (1983). A detailed restriction endonuclease site map of the Zea mays plastid genome. Plant MOI. Biol. 2, 129-140.

Lukens, J.H., and Bogorad, L. (1 988). Nucleotide sequence containing the maize chloroplast proline (UGG) and tryptophan (CCA) tRNA genes. Nucl. Acids Res. 16,5192.

Maniatis, T., Fritsch, E.F., and Sambrook, J. (1 982). Molecular Cloning: A Laboratory Manual. (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory).

Maxam, A.M., and Gilbert, W. (1 980). Sequencing end-labeled DNA with base-specific chemical cleavages. Methods Enzymol.

Ohyama, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano, T., Sano, S., Umesono, K., Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi, H., and Ozeki, H. (1986). Chloroplast gene sequence deduced from the complete sequence of liverwort Marchantia polymorpha chloroplast DNA. Nature 322,

Proudfoot, N.J. (1 986). Transcriptional interference and termination between duplicated m-globin gene constructs suggests a nove1 mechanism for gene regulation. Nature 322, 562-565.

Rochaix, J.-D., and Erickson, J. (1 988). Function and assembly of photosystem II: Genetic and molecular analysis. Trends Biochem. Sci. 13, 56-59.

Rock, C.D., Barkan, A., and Taylor, W.C. (1987). The maize plastid psbB-psbF-petB-petD gene cluster: Spliced and un- spliced petB and petD RNAs encode alternative products. Curr.

EM60 J. 7,885-891.

65,499-560.

572-574.

Genet. 12, 69-77. Rodermel, S.R., and Bogorad, L. (1985). Maize plastid photo-

genes: Mapping and photoregulation of transcript levels during light-induced development. J. Cell Biol. 100, 463-476.

Sanger, F., Nicklen, S., and Coulson, A.R. (1977). DNA sequencing with chain-termination inhibitors. Proc. Natl. Acad. Sci. USA

Sheen, J.-Y., and Bogorad, L. (1988). Differential expression in bundle sheath and mesophyll cells of maize of genes for photosystem I1 components encoded by the plastid genome. Plant Physiol. 86, 1020-1 026.

Shinozaki, K., Deno, H., Sugita, M., Kuramitsu, S., and Sugiura, M. (1986). lntron in the gene for the ribosomal protein S16 of tobacco chloroplast and its conserved boundary sequences. MOI. Gen. Genet. 202, 1-5.

Tanaka, M., Obokata, J., Chunwongse, J., Shinozaki, K., and Sugiura, M. (1987). Rapid splicing and stepwise processing of a transcript from the psbB operon in tobacco chloroplasts: Determination of the intron sites in petB and petD. MOI. Gen. Genet. 209,427-431.

Webber, A.N., Packman, L., Chapman, D.J., Barber, J., and Gray, J.C. (1 989). A fifth chloroplast-encoded polypeptide is present in the photosystem I1 reaction centre complex. FEBS Lett. 242, 259-262.

Westhoff, P., and Herrmann, R.G. (1988). Complex RNA maturation in chloroplasts: The psbB operon from spinach. Eur. J. Biochem. 171,551-564.

Westhoff, P., Alt, J., Widger, W.R., Cramer, W.A., and Herr- mann, R.G. (1985). Localization of the gene for apocytochrome b-559 on the plastid chromosome of spinach. Plant MOI. Biol.

Widger, W.R., Cramer, W.A., Hermodson, M., and Herrmann, R.G. (1985). Evidence for a hetero-oligomeric structure of the chloroplast cytochrome b-559. FEBS Lett. 191, 186-1 90.

Woodbury, N.W., Roberts, L.L., Palmer, J.D., and Thompson, W.F. (1 988). A transcription map of the pea chloroplast genome. Curr. Gen. 14, 75-89.

74,5463-5467.

4, 103-1 1 O.

DOI 10.1105/tpc.2.4.323 1990;2;323-333Plant Cell

J Haley and L Bogoradunits.

Alternative promoters are used for genes within maize chloroplast polycistronic transcription

This information is current as of March 12, 2020

Permissions 298X

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

http://www.plantcell.org/cgi/alerts/ctmain

http://www.plantcell.org/cgi/alerts/ctmain

http://www.aspb.org/publications/subscriptions.cfm

alternative promoters are used for genes within maize ... · the plant cell, vol. 2, 323-333, april...

Documents