The Plant Cell, Vol. 2, 619-631, July 1990 O 1990 American Society of Plant Physiologists

Developmental and Environmental Regulation of a Bean Chalcone Synthase Promoter in Transgenic Tobacco

Jürg Schmid,".' Peter W. Doerner,' Steven D. Clouse,a92 Richard A. Dixon,b and Christopher J. Lamba,3 a Plant Biology Laboratory, Salk lnstitute for Biological Studies, 1 O01 O North Torrey Pines Road, La Jolla, California 92037

Plant Biology Division, Samuel Roberts Noble Foundation, P.O. Box 21 80, Ardmore, Oklahoma 73402

Regulatory properties of a 1.4-kilobase promoter fragment of the bean chalcone synthase CHS8 gene were examined by analysis of glucuronidase (GUS) activity in transgenic tobacco containing a CHS8-GUS gene fusion. The promoter was highly active in the root apical meristem and in petals, exclusively in those cells of the inner epidermis that accumulate anthocyanins. The gene fusion was only weakly expressed in other floral organs, mature leaves, and stems. The early stages of seedling development were characterized by an apparent wound induction of the promoter in the endosperm and strong expression in the immature root, which became localized to the apical meristem and perivascular tissue at the root-hypocotyl junction. The promoter became active during lateral root formation in both the new root and damaged tissue of the main root. The gene fusion was also expressed in greening cotyledons and primary leaves but not in the shoot apical meristem. Light modulated expression in the cotyledons and root-shoot junction but had no effect on other aspects of the developmental program. Wounding or funga1 elicitor treatment of mature leaves activated the promoter in a well-defined zone adjacent to the stress site. Stress induction occurred in mesophyll and vascular tissues as well as in the epidermis. We conclude that the CHS8 promoter contains cis-elements required to establish temporal and spatial control of flavonoid biosynthesis during development and in response to diverse environmental stimuli.

INTRODUCTION

Plants respond flexibly to environmental stimuli that act upon a plastic program of development. A key feature of many of these responses is the synthesis of flavonoid natural products, which have diverse functions in devel- opment and interactions with the environment. Thus, fla- vonoids are important for the pigmentation of flowers and seed coats and in protection against UV irradiation (Hahl- brock and Scheel, 1989). In legumes, pterocarpans and isoflavonoids derived from flavonoid precursors function as phytoalexins (Dixon et al., 1983), and simple flavonoids such as naringenin and luteolin are rhizosphere signals for the induction of nod genes in Rhizobium (Long, 1989). Moreover, recent data indicate that certain flavonoids in- hibit the binding of naphthylphthalmic acid to its cellular receptor and, hence, may have regulatory functions as modulators of polar auxin transport (Jacobs and Rubery, 1988).

Chalcone synthase (CHS) catalyzes the stepwise con-

' Current address: lnstitut für Pflanzenwissenschaften, ETH, Son- neggstrasse 5, 8092 Zürich, Switzerland. * Current address: Department of Biology, San Diego State Uni- versity, San Diego, CA 921 82. To whom correspondence should be addressed.

densation of three acetyl units from malonyl-COA with 4- hydroxycinnamoyl-COA to give naringenin chalcone, which is the first committed step in the branch pathway of phenylpropanoid metabolism specific for flavonoid biosyn- thesis. CHS mRNA and enzyme levels are highly regulated during development associated with the tissue- and cell type-specific accumulation of flavonoid pigments and in response to environmental stimuli for the synthesis of flavonoids involved in adaptation or protection (Hahlbrock and Scheel, 1989; Lamb et al., 1989). Thus, CHS is a key metabolic control point and provides an excellent system for analysis of the molecular mechanisms governing natural product biosynthesis. Run-on transcription assays in iso- lated nuclei and transient expression of CHS-reporter gene fusions in electroporated protoplasts have shown that UV irradiation of parsley cell cultures and elicitor treatment of bean cell cultures stimulate CHS transcription to initiate the synthesis of UV protectants and phytoalexins, respec- tively (Lawton and Lamb, 1987; Dron et al., 1988; Lip- phardt et al., 1988; Schulze-Lefert et al., 1989). Functional assays of the activities of mutated promoters in electro- porated protoplasts have identified the region of a bean CHS promoter required for elicitor induction (Dron et al., 1988) and defined cis-acting elements involved in UV in-

620 The Plant Cell

duction of parsley and antirrhinum CHS promoters (Lip-phardt et al., 1988; Schulze-Lefert et al., 1989).

In contrast, relatively little is known about the transcrip-tional properties of CHS promoters in intact plants. Run-on transcription assays indicate that wounding and fungalinfection stimulate CHS transcription in bean hypocotyls(Lawton and Lamb, 1987) and an antirrhinum CHS pro-moter-chloramphenicol acetyltransferase gene fusion islight regulated in transgenic tobacco (Kaulen et al., 1986).We are interested in how these responses to environmen-tal stimuli are integrated at the gene level within thedevelopmental program of natural product biosynthesis.The regulation of CHS genes from legumes is of particularinterest in this regard because of the additional biologicalfunctions of flavonoid derivatives in legumes as phytoalex-ins and nod gene inducers superimposed upon their ubiq-uitous functions as pigments and UV protectants. In thispaper we have examined the regulatory properties of thepromoter of the bean CHS8 gene by analysis of ,3-glucu-ronidase (GUS) activity in transgenic tobacco plants con-taining a CHS8-GUS gene fusion. CHS8 is one of sevenCHS genes in the bean genome and was selected for thepresent study because it is a highly expressed member ofthe gene family (Ryder et al., 1987). Our data demonstratethat the CHS8 promoter contains the c/s-acting elementsrequired to establish genetic control of tissue- and celltype-specific biosynthesis of flavonoids during develop-ment and also to respond to diverse environmental stimuli.

RESULTS

Organ-Specific Expression of Bean CHS Genes

Poly(A)+ RNA isolated from leaves, stems, roots, flowers,and floral buds of greenhouse-grown bean plants wassubjected to RNA gel blot analysis. The coding sequenceof the CHS15 gene (Ryder et al., 1987) was used as nick-translated probe with hybridization and washing conditionsthat did not distinguish between different CHS transcripts.Figure 1 shows that the highest levels of CHS transcriptswere found in roots, several times less in stems, and evenless in leaves, mature flowers, and floral buds. The contri-bution of CHS8 transcripts to overall CHS expression indifferent organs was determined by S1 nuclease protectionanalysis. In previous studies induction of CHS8 transcriptswas observed in elicitor-treated cell suspension culturesand in wounded hypocotyls (Ryder et al., 1987). Figure 2shows that there was a very strong accumulation of CHS8transcripts in roots. Low levels of the CHS8 transcriptwere present in stems and floral buds, whereas transcriptlevels in fully expanded leaves and mature flowers weretoo low to be detected. Shorter protected fragments cor-responding to other CHS transcripts showed distinctlydifferent patterns of accumulation, indicating differential

1 2 3 4 5 6I I I I I I

kb

_ 9.5— 7.5

— 4.4

— 2.4

— 1.4

— 0.24

Figure 1. RNA Gel Blot Analysis of RNA Isolated from DifferentBean Organs.Poly(A)+ RNA (2.5 M9) from leaves (lane 1), stems (lane 2), roots(lane 3), flowers (lane 4), floral buds (lane 5), and total RNA (6 M9)from wound-induced bean hypocotyls as control (lane 6) weresubjected to RNA gel blot analysis using the nick-translatedCHS15 gene as hybridization probe. Molecular weights wereestimated from the migration distances of co-electrophoresedRNA size markers.

regulation of CHS8 transcripts compared with transcriptsof other members of the gene family. For example, themajor CHS transcript in floral buds corresponded to theCHS4 cDNA.

Transformation of Tobacco with CHS8 Constructs

The CHS8 promoter containing 1250 bp upstream of thetranscription start site and 162 bp of the untranslatedleader sequence extending to 1 bp 5' of the translationinitiation site (Figure 3A) was fused to the GUS codingsequence in the vector pB1101.1 (Figure 3C) (Jefferson etal., 1987). In addition, the complete CHS8 gene was ligatedinto the polylinker of the binary vector BIN19 (Figure 3B)(Sevan, 1984). Both constructs were transferred to to-

Chalcone Synthase Promoter Activity 621

B1 2 3 4 5 6 7 8 9 10 11 12 13 14

Figure 2. S1 Nuclease Protection Analysis.

Poly(A)+ RNA isolated from different bean tissues and in vitrotranscribed control RNA were subjected to S1 nuclease protectionanalysis to determine the contribution of CHS8 transcripts to theoverall accumulation of CHS transcripts. Control "sense" RNAwas generated with T7 RNA polymerase after linearization ofplasmids pCHS1 3' 25, pCHS4 3' 25, pCHS4a 3' 25, pCHS143' 25, and pCHS17 3' 25 (Figure 3E) with EcoRI. The probe forS1 nuclease protection analysis was made by excision of a 475-bp BspMI-Pvull fragment (Figure 3E) from pCHS14 3' 25 andfilling in the end with radiolabeled nucleotide triphosphates. Gen-omic clone CHS8 corresponds to cDNA clone CHS14 (Ryder etal., 1987).(A) Labeled probe (lane 1), 123-bp ladder (lane 2), and probehybridized without added RNA and subjected to S1 digestion(lane 3).(B) Probe hybridized with the following sense RNAs before S1nuclease digestion: CHS1 (lane 4), CHS4 (lane 5), CHS4a (lane6), CHS14 (lane 7), and CHS17 (lane 8). The lengths of protectedprobe are as expected from CHS cDNA sequence data (Ryder etal., 1987).(C) Probe hybridized with 2.5 Mg each of poly(A)* RNA fromleaves (lane 9), stems (lane 10), roots (lane 11), flowers (lane 12),floral buds (lane 13), and total RNA (6 Mg) from wound-inducedbean hypocotyls (lane 14). The 370-bp fragment protected by theprobe corresponding to the CHS8 transcript is marked by thearrow. (C) was exposed 50 times longer than (B).

bacco by leaf disc transformation (Rogers et al., 1986),and plants were regenerated under kanamycin selection.Of 10 kanamycin-resistant CHS8-GUS transformants, nineshowed GUS activity in root extracts. Three independenttransformants selected for further study (designated

CHS8-GUS-5, CHS8-GUS-7, and CHS8-GUS-8) segre-gated at a ratio close to 3:1 for GUS activity, indicatingthat these transformants contained one functional locus ofthe transgene per haploid genome (data not shown).

Transcription start sites were mapped by RNase protec-tion experiments (Sambrook et al., 1989) using total RNAfrom roots of sterile-grown transgenic plants (Figure 4).The size of protected fragments using RNA from trans-genic plants containing the CHS8-GUS gene fusion or theentire CHS8 gene were identical to the fragments pro-tected by RNA from wound-induced bean hypocotyls,indicating that the CHS8 promoter directed transcriptionin tobacco from the same start site as that involved inexpression of the endogenous CHS8 gene in bean.

Sru I Pva\ flspMI Erofll

M ' I t-L.

Figure 3. Restriction Maps of Genomic Clone CHSX8 and Con-structs Used for Tobacco Transformation and Molecular Analysis.

(A) Partial restriction map of genomic clone CHSX8. The 3.9-kbEcoRI-Hindlll fragment contains 1.4 kb of 5'-untranslated se-quences followed by a coding region of 1270 bp and a 3'-untranslated region of 1.2 kb. A single intron (boxed) of 103 bp islocated toward the 5' end of the coding sequence.(B) The 3.9-kb EcoRI-Hindlll fragment subcloned into BIN19.(C) The chimeric CHS8-GUS gene construct comprising the 5'-untranslated sequences from the EcoRI site up to a Ddel site 1bp upstream of the initiation codon ligated in front of the GUScoding region (GUS) followed by a nos terminator (NOS-T) in thevector pB1101.1.(D) The same promoter fragment ligated into Smal-restrictedpSP64.(E) CHS cDNA clone pCHS14 (Ryder et al., 1987) restricted atthe conserved Stul site and at the EcoRI site created during cDNAsynthesis and ligated into Smal-EcoRI-cut plBI25. Correspondingconstructs were made for pCHS1, pCHS4, pCHS4a, and pCHS17(Ryder et al., 1987). Scale bar applies to (A) through (D). (E) is 2times this scale. The arrows above the gene indicate the tran-scription start site.

622 The Plant Cell

1 2 3 4 5 6 7 M b P

— 492— 369

— 246

I— 123

Figure 4. RNase Protection Analysis.

Total RNA (50 MQ) isolated from the roots of axenically growntransgenic tobacco plants and from wounded bean hypocotylswas subjected to RNase protection analysis: probe not digestedwith RNase (lane 1); RNA from two independent transformantscarrying the entire CHS8 gene (Figure 3B): CHS8-4 (lane 2) andCHS8-13 (lane 3); RNA from tobacco transformed with vectorBIN19 (lane 4); RNA (6 ^g) from wounded bean hypocotyls (lane5); RNA from roots of tobacco transformant CHS8-GUS-8 carryingthe CHS8-GUS gene fusion construct shown in Figure 3C (lane6); tRNA control reaction (lane 7); end-labeled 123-bp ladder assize markers (lane M).

Organ-Specific Expression of the CHS8-GUS GeneFusion

The expression of the CHS8-GUS gene fusion in differentorgans of the transformants CHS8-GUS-5, CHS8-GUS-7,and CHS8-GUS-8 was monitored by fluorimetric assay ofextractable GUS activity (Table 1). In vegetative organsthere was a high level of GUS activity in roots, moderate

activity in stems, and low activity in extracts of matureleaves. In floral organs only very low levels of extractableGUS activity were observed with sepals, pistils, and sta-mens. High levels of GUS activity were found in petals,confined to the small apical regions that accumulate antho-cyanins. There was only very low GUS activity in theremaining nonpigmented regions, which comprise the ma-jor portion of the petal in tobacco.

Although the same overall pattern of expression ofCHS8-GUS was observed in each plant examined, therewas, as expected, some quantitative variation amongthese independent transformants. In particular, transfor-mant CHS8-GUS-7 exhibited less marked expression ofthe gene fusion in stems and in the pigmented regions ofpetals compared with transformants CHS8-GUS-5 andCHS8-GUS-8. However, the level of expression in roots ofCHS8-GUS-7 was comparable with that observed in theother transformants.

Spatial Pattern of CHS8-GUS Expression

The spatial pattern of CHS8 promoter activity underlyingthe strong expression of the CHS8-GUS gene fusion inroots and petals was examined by histochemical analysisof GUS activity in situ using the chromogenic substrate 5-bromo-4-chloro-3-indolyl /3-glucuronide. Figure 5A con-firms that the CHS8-GUS gene fusion was specificallyexpressed in the pigmented regions of petals. There wasan abrupt transition from tissue showing strong histochem-ical staining for GUS to tissue showing no appreciable GUSactivity. The limits of CHS8 promoter activity correspondedprecisely with the boundary between the pigmented andnonpigmented regions. To examine the spatial pattern ofCHS8 promoter activity at the cell level, petals were fixedafter completion of the GUS chromogenic reaction andthen embedded in paraffin for preparation of thin sections.Figure 5B shows that GUS staining was confined to cellsin the inner epidermis with no significant staining of theouter epidermis or internal vascular and parenchymal tis-sues. Moreover, at the junction between pigmented andnonpigmented tissues there was an abrupt transition zoneof only a few epidermal cells between those showing GUSstaining and those showing no GUS staining (Figure 5C).This junction corresponded exactly to the boundary be-tween pigmented epidermal cells, which show a centralconic projection, and the adjacent nonpigmented cells,which lack this morphological feature in tobacco.

Figure 5D shows that GUS activity could also be histo-chemically detected in pollen grains in situ. Only about halfthe pollen grains exhibited GUS activity, consistent withsegregation analysis of the gene fusion in progeny fromthe selfed primary transformant CHS8-GUS-8. The CHS8-GUS gene fusion did not appear to be expressed in thetapetal layer of the anther surrounding the pollen sac. NoGUS activity was detected in the pollen grains from trans-

Chalcone Synthase Promoter Activity 623

Table 1. Orqan Distribution of GUS Activity in Transformants ContaininQ the CHS8-GUS Gene Fusion

GUS Activity (pmol/min/mg of protein)

Orqan CHSSGUS-5 CHS8-GUS-7 CHS8-GUS-8

Leaves 22 3 48 Stems 31 5 70 150 Roots 6000 6900 7200 Sepals 2 4 9 Pistils 7 11 5 Stamens 7 11 17 Petals (pigmented tissue) 6000 480 8400 Petals (nonpigmented tissue) 5 14 15

genic plants transformed with the promoterless GUS plas- mid pBll 01.1 (data not shown).

Figures 5E to 5J show that the CHS8 promoter was active at the apical tip of the mature main root. The GUS activity was confined to a specific region of the root tip. Thus, no GUS activity was observed in the root cap, the root epidermis, or the root vascular system. Strong expres- sion of the chimeric construct was observed in the apical meristem and in the immediately adjacent zone of cell elongation. Expression of the gene fusion progressively weakened dista1 to the apical tip, becoming restricted to a group of cells cylindrically surrounding the developing vas- cular system. No significant GUS activity was detectable histochemically in more mature tissues behind the root tips. A similar spatial pattern of GUS activity was observed in the tips of mature lateral roots.

Developmental Regulation

The histochemical analysis of GUS activity in roots and petals indicated that the CHS8 promoter exhibited well- defined spatial patterns of activity in these organs. To trace the initial developmental program of CHS8 gene expression, we monitored promoter activity during germi- nation and the early stages of seedling development by histochemical analysis of GUS activity in the progeny of selfed transformant CHS8-GUS-8. Figure 6A shows that the CHS8 promoter was activated in the endosperm of germinating seeds, apparently as a wound response at the site of radicle penetration and emergence. This re- sponse was observed in all seed tested and did not segregate in the new seedlings, indicating that this repre- sented induction of the gene fusion in cells possessing the maternal genotype.

In contrast, expression of GUS activity in developing seedlings segregated with the chimeric construct and rep- resented regulation of the promoter in the genomes of the progeny. In the germinating seedling a low leve1 of GUS activity was observed diffusely throughout the cotyledon and radicle (Figure 6B) until the emergence of the latter,

when strong expression of the gene fusion was Seen in the immature root (Figure 6B). Initially, expression was observed throughout the root from the tip region to the root-hypocotyl junction, but as the cotyledons began to green and expand, GUS staining in the root became re- stricted to the apical tip, in the characteristic tissue-specific pattern observed in mature roots (see above), and to the region of the root-hypocotyl junction (Figure 6C). GUS activity at this junction was predominantly localized to a group of cells cylindrically enveloping the vascular bundle and was absent from the epidermis (Figure 6D). GUS staining at the root-hypocotyl junction progressively dimin- ished and no activity was detectable at the developmental stage where the cotyledons were fully expanded (Figure 6E). In contrast, the root tip continued to show strong staining throughout the remaining course of seedling de- velopment and in the adult plant.

Figures 5K to 5N show that the CHS8 promoter was also activated in lateral root development. Localized GUS staining could be observed at an early stage (Figure 5K) as the new root was initiated from within the pericycle of the main root. During penetration through the cortex and epidermis of the main root, strong GUS staining was observed throughout the lateral root, and the surrounding damaged tissues of the main root also stained for GUS activity (Figures 5L to 5M). After emergence of the lateral root, GUS activity was confined to the tip of the new root in a pattern similar to that in the tip of the main root (Figure 5N).

The CHS8 promoter also became active in cotyledons during the expansion and greening of these organs (Figure 6F). When the cotyledons were fully expanded, GUS stain- ing was observed in blocks across the cotyledon not correlated with any morphological feature. The strength of staining increased over time as the cotyledons matured. Sectioned cotyledons showed CHS8 promoter activity throughout epidermal and parenchymal tissues within these patches (Figure 6G). Very strong GUS staining was also observed in the primary leaves as they emerged (Figure 6F). No GUS activity was histochemically detecta- ble within the apical meristem, which was in contrast to

624 The Plant Cell

SBcp

SD

5Gy T

5H 51 5J

'•"A

Figure 5. Histochemical Localization of GUS Activity in Floral Organs and Root Tissues of Transgenic Tobacco Plant CHS8-GUS-8.

Chalcone Synthase Promoter Activity 625

transgenic tobacco seedlings transformed with PAL2-GUS constructs (Liang et al., 1989), which show GUS activity within the tip of the apical meristem and also in the developing vascular system (Figure 6H).

Light lnteractions with the Developmental Program

Light had a marked effect on some aspects of this devel- opmental program but little effect on other aspects. Thus, in the cotyledons of 1 O-day-old etiolated seedlings, the CHS8 promoter was active only at the extreme tip and did not become active throughout the rest of the organ in the absence of a light signal (Figure 6K). If these etiolated seedlings were exposed to white light for 13 hr, the spatial pattern of GUS activity rapidly changed to give blocks of staining across the greening cotyledon, characteristic of the pattern observed in seedlings germinated and grown under a 16-hr light/8-hr dark cycle (Figure 6L). In devel- oping roots the progressive localization of CHS8 promoter activity at the apical tip did not appear to be light depend- ent, occurring in etiolated seedlings as in normal light- grown seedlings. However, the decay in promoter activity at the root-hypocotyl junction was delayed in etiolated seedlings (Figure SI). Exposure of etiolated seedlings to white light for 4 hr greatly accelerated the disappearance of GUS staining in this region, which was complete after 13-hr illumination (Figure 6J).

Stress Activation

The induction of the CHSSGUS gene fusion in seed tis- sues during germination and in tissues of the main root during emergence of a lateral root indicate that the CHS8 promoter was activated by tissue damage incurred during normal development. Figure 7 shows that the CHS8 pro- moter was also responsive to externally imposed stress.

Mature leaf tissue of transgenic plants containing the CHS8-GUS gene fusion contained very low levels of GUS activity as monitored by fluorimetric assay in leaf extracts (Table 1). Localized wounding of the tissue of a detached mature leaf caused induction of GUS activity in the rings of tissue immediately adjacent to the wound sites. While strong GUS staining was observed at all wound sites after 48 hr to 72 hr (Figure 7A), only weak staining at some sites was observed 16 hr after wounding. However, im- mediate application of funga1 elicitor to the wound site caused a marked induction of the CHS8 promoter after 16 hr to 24 hr compared with the response to wounding alone (Figure 7D). Strong GUS staining was observed in tissue immediately adjacent to the site of elicitor application (within 2 mm to 3 mm) and weaker staining in tissue up to 1 O mm to 15 mm distant. Activation of the CHS8 promoter at a distance was also observed in transgenic tobacco leaves infiltrated locally with an incompatible isolate of Pseudomonas syringae or treated with potassium oxalate (B. Stermer, J. Schmid, C.J. Lamb, and R.A. Dixon, un- published observations).

To examine tissue and cell type specificity of stress induction of the CHS8 promoter, leaf fragments containing a wound site were embedded and sectioned after the histochemical GUS reaction. These sections revealed in- ducible CHS8 promoter activity in a sharply delimited region surrounding the lesion (Figures 78 and 7C). Induc- tion of GUS activity was observed in leaf epidermal, me- sophyll, and vascular tissue within this zone.

DISCUSSION

Transgenic plants containing the CHS8-GUS gene fusion exhibited specific and characteristic patterns of GUS activ- ity during plant development and in response to environ- mental stimuli. Although we cannot completely rule out

Figure 5. (continued). (A) lndigo dye deposits coincide with the pigmented part of the tobacco petal. (B) Petal cross-sections of 14 r m thickness. (C) Petal cross-section showing the sharp boundary of GUS activity in the inner epidermis. (D) GUS activity in pollen grains stained in situ within the anthers. Embedded anthers were sectioned to 10-rm thickness. (E) Longitudinal section (10 fim thickness) of a mature primary root tip. The arrows point to approximate locations of cross-sections (16 rm) in (F) to (J). (F) Root cap. (G) Section at root cap-apical meristem interface. (H) Apical meristem. (I) Section within zone of elongation with developing vascular tissue. (J) Maturation zone with morphologically distinct vascular tissue. (K) Lateral root initiation. (L) Growth of the primordium through the cortex of the main root. (M) Penetration of primary root epidermis. (N) Lateral root elongation. cp, conic projections; size bars = 1 O0 Gm.

626 The Plant Cell

am

6H

Figure 6. Histochemical Analysis of CHS8-GUS Expression during Seedling Development.

Chalcone Synthase Promoter Activity 627

7D

w

7B

7C

we

Figure 7. Stress Response in Leaves of Transgenic Plants after Wounding or Fungal Elicitor Treatment.(A) GUS activity in leaf 72 hr after wounding.(B) 12-Mm cross-section adjacent to wound site of leaf in (A).(C) 12-Mim cross-section at distal border of wound response.(D) Elicitor induction after application of 5 pL of fungal elicitor to wound site and incubation for 16 hr compared with induction by woundingalone.w, wounded; we, wounding plus elicitor; size bars =

some bias in the observed patterns of GUS activity at thecell level due to differential uptake and availability of thechromogenic substrate, the patterns we observed are verydifferent from those observed in transgenic plants contain-ing cauliflower mosaic virus 35S-GUS gene fusions or anumber of other plant promoter-GUS gene fusions includ-ing PAL2-GUS (Benfey and Chua, 1989; Bevan et al.,1989; Liang et al., 1989). Hence, the patterns of GUSactivity in the CHS8-GUS transgenic plants are establishedby the specific properties of the CHS8 promoter.

The bean promoter is appropriately regulated in trans-genic tobacco. Thus, the gene fusion was transcribed fromthe same start site as that utilized by endogenous beanCHS genes in wounded bean hypocotyls and by the entireCHS8 gene in transgenic tobacco. Moreover, the patternof CHS8 promoter activity in transgenic tobacco, as in-ferred from the induction of GUS activity, closely resembled

the activity of the endogenous promoter in bean, as mon-itored by the accumulation of CHS8 transcripts. Thus, thehigh levels of GUS activity in tobacco roots and the induc-tion of the gene fusion by wounding and fungal elicitorclosely paralleled the pattern of accumulation of CHS8transcripts in bean.

Likewise, anthocyanin pigments accumulated at a muchearlier stage of floral development in bean compared withtobacco, and, hence, the high levels of GUS activity in thepigmented regions of petals of transgenic tobacco couldbe correlated with the presence of CHS8 transcripts inbean floral buds even though the transcripts were nolonger detectable in mature petals. The very high levels ofGUS activity in the pigmented regions of mature petals ofthe transgenic tobacco plants, compared with the modestlevel of the CHS8 transcript in bean floral buds, may reflectprolonged translation of the GUS mRNA throughout petal

Figure 6. (continued).(A) Induction in the endosperm projecting from the testa.(B) Stages in seedling growth, shown from bottom to top. The bottom two seedlings were excised from seed; the radicle of the topseedling had just penetrated the seed coat.(C) Further course of CHS8-GUS expression during seedling development.(D) Cross-section (40 ^m) of stem-root junction.(E) CHS8-GUS expression in an 11-day-old seedling with fully developed cotyledons.(F) Apical tissues of the developing seedling with fully matured cotyledons and expanding primary leaves.(G) Sections (12 /um) of apical tissues.(H) Similar sections of tobacco transformed with a bean PAL2-GUS gene fusion.(I) Etiolated seedlings at different developmental stages (arrows highlight stem-root junction).(J) Equivalent seedlings after 13-hr illumination.(K) Cotyledons of etiolated seedlings.(L) Equivalent cotyledons after 13-hr illumination with white light.am, apical meristem; c, cotyledons; rt, root; st, stem; v, vascular tissue; size bars = 100 /um.

628 lhe Plant Cell

development after expression of the gene fusion at an early stage in flowering. Moreover, RNA was extracted from the whole floral bud. GUS activity was low in floral organs other than petals, and within petals, high GUS activity was restricted to the pigmented tissue such that the specific activity of GUS in extracts of nonpigmented peta1 tissue was at least 30-fold lower than in extracts exclusively from the pigmented tissue. Expression of the CHS8-GUS gene fusion exclusively in the cells of the inner epidermis that accumulate anthocyanins provides further evidence that the CHS8 promoter is appropriately regu- lated in tobacco flowers.

Overall, the present data indicate that the transcriptional activity of the CHS8 promoter is a major factor in deter- mining the pattern of accumulation of the corresponding transcript in intact plants during development and in re- sponse to light, wounding, or fungal elicitor. The close correlation between expression of the CHS8-GUS gene fusion and accumulation of flavonoid natural products is consistent with the hypothesis that selective transcription of the CHS8 gene is a major control site in flavonoid synthesis both during development and in response to environmental stimuli. Examples include, in addition to expression in petals associated with anthocyanin synthe- sis, strong expression in roots correlated with the synthe- sis of nod gene inducers and other root flavonoids, as well as stress induction associated with the synthesis of isofla- vonoid phytoalexins (Wiermann, 1981; Dixon et al., 1983). Interestingly, nodules form in the cortical tissues adjacent to the vascular cylinder rather than in more peripheral tissues (Long, 1989). Hence, the tissue-specific expression of CHS8 in roots may be important in establishing the site of nodule development by localized synthesis of nod gene inducers.

A surprising feature of the CHS8 promoter is the strong expression in the apical meristem of the root and at sites of lateral root initiation. However, the CHS8 promoter is not active in all meristems because no GUS staining was observed at the shoot apex, whereas the bean PAL2 promoter is active in this meristem (Figure 6H and Liang et al., 1989). Because certain flavonoids may modulate polar auxin transport, CHS8 expression in root meristems could be involved in determining the distribution of auxin and, hence, modulation of root morphogenesis. Analysis of the effects of exogenous auxin and auxin-insensitive mutants suggests that auxin plays a critical role in lateral root initiation (Wightman et al., 1980). At these sites, the CHS8-GUS gene fusion appears to be activated not only in cells of the new root but also in surrounding tissues of the main root. This is unlikely to reflect spreading of the GUS stain because GUS staining exclusively in the cells of the lateral root tip as it penetrates the tissues of the main root was observed in transgenic tobacco plants containing the hydroxyproline-rich glycoprotein HRGPnt3 gene pro- moter-GUS gene fusion (Keller and Lamb, 1989). lnduction of CHS8 in the damaged tissue of the main root may

represent a defense mechanism against microbial infection at the site of emergence of the lateral root.

Tobacco phytoalexins are furanocoumarin and sesqui- tespenoid compounds rather than flavonoids, and endog- enous CHS genes are not stress induced (Dixon et al., 1983). However, the bean CHS8 promoter is activated by wounding or elicitor in transgenic tobacco even though CHS has no function in stress metabolism in tobacco. Therefore, the CHS8 promoter responds to a conserved signal system for stress induction of defense genes that was established before the evolution of chemically distinct phytoalexins in legumes and solanaceous plants. Deposi- tion of lignin is a ubiquitous defense response, and induc- tion of genes encoding phenylpropanoid biosynthetic en- zymes involved in the synthesis of lignin monomers, e.g., PAL, 4-c0umarate:CoA ligase, and cinnamyl alcohol de- hydrogenase, by wounding, fungal elicitor, or infection has been observed in a wide range of monocot and dicot species (Dixon and Harrison, 1990). The evolution of fla- vonoid derivatives as phytoalexins in legumes may, thus, have been accompanied by the incorporation into the CHS8 promoter of cis-elements conserved in the PAL promoters of many species. It will be of interest to deter- mine whether genes encoding enzymes of the phenylpro- panoid branch pathway for coumarin synthesis from plants where furanocoumarins function as phytoalexins, e.g., par- sley (Dixon et al., 1983), are stress induced when trans- ferred into legumes.

The distinct zone of induction of the CHS8 promoter by wounding or fungal elicitor is reminiscent of the spatial pattern of accumulation of defense gene transcripts in response to infection of parsley leaves with Phytophthora megasperma (Schmelzer et al., 1989). Such a pattern, which our data indicate is established at the leve1 of gene transcription, cannot readily be accounted for by a single diffusible signal because this would be expected to give a gradation between strongly induced cells and noninduced cells. Hence, induction of the CHS8 promoter may involve two or more signals with different transmission properties or a signal that is transmitted by local seria1 progression, cell to cell. It is striking that within the response zone, the CHS8 promoter is stress activated in all tissues throughout the cross-section of the leaf. This is in marked contrast to the tissue- and cell type-specific expression observed in many facets of the developmental program of regulation. Hence, the promoter contains cis-acting elements respon- sive to externally applied signals that override or circum- vent regulatory elements that establish the strict tissue and cell type specificity exhibited in other contexts. Acti- vation in all tissues adjacent to a local wound may be crucial for effective protection against microbial attack after mechanical damage.

The expression of the CHS8-GUS gene fusion resem- bles the pattern of activation of the bean PAL2 promoter in transgenic tobacco with respect to wound induction and strong expression in roots and the pigmented regions of

Chalcone Synthase Promoter Activity 629

petals (Bevan et al., 1989; Liang et al., 1989). Moreover, the PAL2-GUS and CHS8-GUS gene fusions exhibit the same complex program of expression in the early stages of seedling development with GUS staining initially delo- calized throughout the root and then becoming restricted to the root tip and the root-hypocotyl junction (S. Ohl and C.J. Lamb, unpublished observations). However, in stems the PAL2 and CHS8 promoters exhibit markedly different patterns of expression. Thus, the CHS8 promoter is only weakly expressed in stems compared with roots, whereas the PAL2 promoter is active in stems at high levels com- parable with those in roots. This reflects strong PAL2 expression in differentiating xylem associated with lignin synthesis in which CHS is not involved. Moreover, whereas both promoters are active in root apical meristems, only the PAL2 promoter is active in the shoot apical meristem.

Some common sequence motifs in CHS and PAL pro- moters have recently been discerned that function in the induction of transcription in suspension cultured cells by UV irradiation and elicitors (Cramer et al., 1989; Lois et al., 1989; Lawton et al., 1990). Experiments to determine the functions of these and other elements in the regulation of CHS and PAL genes in transgenic plants are in progress. Moreover, these studies should reveal the functional ar- chitecture of these promoters by which exquisite tissue- and cell type-specific developmental regulation is com- bined with flexible responses to diverse environmental stimuli. The complex regulation of the CHS8 promoter demonstrated in the present study provides an experimen- tal basis for examination of the molecular mechanisms underlying key aspects of the unique developmental plas- ticity of higher plants.

METHODS

CHSB-GUS Gene Fusion and Probe Construction

The complete CHS8 gene on a 3.9-kb EcoRI-Hindlll fragment of the bean genomic clone CHS-A8 was ligated into the polylinker of the BIN19 binary vector (Bevan, 1984) as shown in Figure 3. A 1.4-kb EcoRI-Ddel fragment of this clone containing 1250 bp 5’ of the transcription start site and 162 bp of the untranslated leader sequence to 1 bp upstream of the translation initiation codon was ligated into the Smal site of the promoterless GUS expression vector pBI101.1 (Jefferson et al., 1987) after filling in the ends (Figure 3C). The same fragment was also ligated into the Smal site of plasmid pSP64 (Figure 3D) (Promega Biotec). Both con- structs were verified by restriction site analysis and sequencing (Sambrook et al., 1989) over the new junction with GUS-specific (Clonetech) or SP6-specific primers, respectively. Reagents for S1 nuclease protection analysis were made by restricting CHS cDNA clones pCHS1, pCHS4, pCHS4a, pCHS14, and pCHS17 (Ryder et al., 1987) at the conserved Stul site and at the EcoRl site created during cDNA synthesis. These fragments were ligated into Smal-EcoRI-cut plB125 (IBI), to give the plasmids pCHS1 3‘

25, pCHS4 3‘ 25, pCHS4a 3‘ 25, pCHS14 3’ 25, and pCHS17 3’ 25, respectively.

Plant Transformation and Growth

The vectors containing the CHSB gene or the CHS8-GUS gene fusion were transferred from Escherichia coli HBl O1 into Agro- bacterium fumefaciens LBA4404 by a triparental mating with E. coli containing the mobilization plasmid pRK2013 (Ditta et al., 1 980). Tobacco leaf disc transformation and regeneration of trans- genic plants were performed as described in Rogers et al. (1986). Primary transformants were grown at 25OC under a 16-hr light (1 15 pE)/bhr dark cycle.

RNA Analysis

Total cellular RNA was isolated from bean tissues, wound-induced hypocotyls, and tobacco roots as described in Bell et al. (1986) and further purified over a CsCl gradient (Sambrook et al., 1989). Poly(A)’ RNA was isolated by oligo(dT)-cellulose affinity chroma- tography (Sambrook et al., 1989) and used for RNA gel blot and S1 nuclease protection analysis. For RNA gel blot analysis, RNA samples and an RNA ladder (Bethesda Research Laboratories) were electrophoresed through a 1.2% formaldehyde agarose gel, transferred to nitrocellulose, and processed further as described in Sambrook et al. (1 989). Prehybridization and hybridization were performed under standard conditions (Sambrook et al., 1989); the final wash of filters was at 6OoC in 0.3 X SSC, 0.5% SDS (1 X SSC = 150 mM NaCI, 15 mM Na citrate, pH 6.5). Autoradiography was performed with Kodak RP film and intensifier screen at -70°C.

S1 nuclease protection analysis was performed as described by Sambrook et al. (1989). “Sense” RNA was generated with T7 RNA polymerase after linearization of plasmids pCHSl 3‘ 25, pCHS4 3’ 25, pCHS4a 3’ 25, pCHS14 3‘ 25, and pCHS17 3‘ 25 with EcoRI. The probe for S1 nuclease protection analysis was made by excising a BspMCPvull fragment (Figure 3E) from pCHS14 3‘ 25 and filling in the end with radiolabeled nucleotide triphosphates. RNase protection was performed as in Sambrook et al. (1989). The template DNA for the “antisense” RNA probe (see Figure 3D) was linearized with EcoRl before in vitro transcrip- tion with SP6 RNA polymerase.

CHSB-GUS Expression

Developmental and environmental regulation of the CHSB pro- moter were studied in seedlings germinated from surface-steri- lized seeds of selfed primary transformant CHS8-GUS-8. After imbibition in H,O, seeds were kept at 4OC overnight, and further germination was carried out in basal MS medium (Linsmaier anc! Skoog, 1965) without glucose (Sigma) under the same light con- ditions as primary transformants or in complete darkness. Leaves of primary transformants were wounded either by gently crushing leaves with flat forceps or by pricking with an Eppendorf tip to give a 2-mm to 3-mm hole and treating the wound site immediately with 5 pL of sterile distilled H20 or 5 pL of funga1 elicitor (100 Wg of glucose equivalent/ml). The elicitor was the high molecular weight fraction heat released from washed mycelial walls of P. megasperma f. sp. glycinea (Ayers et al., 1976).

630 The Plant Cell

GUS Assays

GUS activity was assayed in tissue extracts by fluorimetric deter- mination of the production of 4-methylumbelliferone from the corresponding glucuronide (Jefferson, 1987). Protein was deter- mined by the method of Bradford (1976), and GUS activity was expressed as picomoles of product per minute per milligram of protein. Histochemical localization of GUS activity in situ was performed by incubation of plant tissues at 37OC in 50 mM sodium phosphate (pH 7.0) containing the chromogenic substrate 5- bromo-4-chloro-3-indolyl p-o-glucuronide (1 mM) after fixation in 0.3% formaldehyde as described by Jefferson (1 987). Tissues for embedding were subsequently dehydrated through solutions of increasing ethanol concentration. After thorough dehydration in 100% ethanol, samples were cleared in four changes of xylene, two changes of paraffin-saturated xylene, and four changes of molten paraffin. Embedded samples were then sectioned to 10- Frn to 1 6 - ~ m thickness. Ribbons were floated onto glass slides, dried, deparaffinized, and mounted with acrylic resin.

ACKNOWLEDGMENTS

Jürg Schmid and Peter Doerner contributed equally to this work and are joint first authors. We thank Dr. Jack Paxton for the fungal elicitor preparation and Xiaowu Liang for the RNA from wounded bean hypocotyls. This research was supported by grants to C.J.L. from the Samuel Roberts Noble Foundation and the U.S. Department of Agriculture, Competitive Research Grants Office (87-CRCR-1-2350 and 89-CRCR-3-7263). J.S. was sup- ported by the Swiss National Science Foundation, P.W.D. by the Deutscher Akademischer Austauschdienst. S.D.C. was a National Science Foundation Plant Biology Fellow (DCB-84-12386).

Received March 12, 1990; revised May 7, 1990.

REFERENCES

Ayers, AR., Ebel, J., Valent, B., and Albersheim, P. (1976). Host-pathogen interactions. X. Fractionation and biological ac- tivity of an elicitor isolated from the mycelial wall of Phyfophthora megasperma var. sojae. Plant Physiol. 57, 760-765.

Bell, J.N., Ryder, T.B., Wingate, V.P.M., Bailey, J.A., and Lamb, C.J. (1 986). Differential accumulation of plant defense gene transcripts in a compatible and an incompatible plant-pathogen interaction. MOI. Cell. Biol. 6, 1615-1623.

Benfey, P.N., and Chua, N.-H. (1 989). Regulated genes in trans- genic plants. Science 244, 174-1 81.

Bevan, M. (1984). Binary Agrobacterium vectors for plant trans- formation. Nucl. Acids Res. 12, 8711-8721.

Bevan, M., Shufflebottom, D., Edwards, K., Jefferson, R., and Schuch, W. (1989). Tissue- and cell-specific activity of a phen- ylalanine ammonia-lyase promoter in transgenic plants. EMBO J. 8, 1899-1906.

Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254.

Cramer, C.L., Edwards, K., Dron, M., Liang, X., Dildine, S.L., Bolwell, G.P., Dixon, R.A., Lamb, C.J., and Schuch, W. (1989). Phenylalanine ammonia-lyase gene organization and structure. Plant MOI. Biol. 12, 367-383.

Ditta, G., Stanfield, S., Corbin, D., and Helinski, D.R. (1980). Broad range DNA cloning system for gram-negative bacteria: Construction of a gene bank of Rhizobium meliloti. Proc. Natl. Acad. Sci. USA 77,7347-7351.

Dixon, R.A., and Harrison, M.J. (1 990). Activation, structure and organization of genes involved in microbial defense in plants. Adv. Genet. 28, 165-234.

Dixon, R.A., Dey, P.M., and Lamb, C.J. (1983). Phytoalexins: Enzymology and molecular biology. Adv. Enzymol. Relat. Areas MOI. Biol. 55, 1-135.

Dron, M., Clouse, S.D., Dixon, R.A., Lawton, M.A., and Lamb, C.J. (1988). Glutathione and fungal elicitor regulation of a plant defense gene promoter in electroporated protoplasts. Proc. Natl. Acad. Sci. USA 85,6738-6742.

Hahlbrock, K., and Scheel, D. (1989). Physiology and molecular biology of phenylpropanoid metabolism. Annu. Rev. Plant Phys- iol. Plant MOI. Biol. 40, 347-364.

Jacobs, M., and Rubery, P.H. (1 988). Naturally occurring auxin transport regulators. Science 241, 346-349.

Jefferson, R.A. (1987). Assaying chimeric genes in plants: The GUS gene fusion system. Plant MOI. Biol. Rep. 5, 387-405.

Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUS fusions: p-Glucuronidase as a sensitive and versatile gene fusion rnarker in higher plants. EMBO J. 6, 3901-3907.

Kaulen, H., Schell, J., and Kreuzaler, F. (1 986). Light-induced expression of the chimeric chalcone synthase-NPTII gene in tobacco cells. EMBO J. 5, 1-8.

Keller, B., and Lamb, C.J. (1989). Specific expression of a nove1 cell wall hydroxyproline-rich glycoprotein gene in lateral root initiation. Genes Dev. 3, 1639-1646.

Lamb, C.J., Lawton, M.A., Dron, M., and Dixon, R.A. (1989). Signals and transduction mechanisms for activation of plant defenses against microbial attack. Cell 56, 21 5-224.

Lawton, M.A., and Lamb, C.J. (1 987). Transcriptional activation of plant defense genes by fungal elicitor, wounding, and infec- tion. MOI. Cell. Biol. 7, 335-341.

Lawton, M.A., Clouse, S.D., and Lamb, C.J. (1 990). Glutathione- elicited changes in chromatin structure within the promoter of the defense gene chalcone synthase. Plant Cell Rep. 8,

Liang, X., Dron, M., Schmid, J., Dixon, R.A., and Lamb, C.J. (1 989). Developmental and environmental regulation of a phen- ylalanine ammonia-lyase-p-glucuronidase gene fusion in trans- genic tobacco. Proc. Natl. Acad. Sci. USA 86, 9284-9288.

Linsmaier, E., and Skoog, F. (1965). Organic growth factor requirements of tobacco tissue cultures. Physiol. Plant. 18,

Lipphardt, S., Brettschneider, R., Kreuzaler, F., Schell, J., and Dangl, J.L. (1 988). UV-inducible transient expression in parsley protoplasts identifies regulatory cis-elements of a chimeric chal- cone synthase gene. EMBO J. 7,4027-4033.

561 -564.

100-1 27.

Chalcone Synthase Promoter Activity 631

Lois, R., Dietrich, A., Hahlbrock, K., and Schulz, W. (1989). A phenylalanine ammonia-lyase gene from parsley: Structure, reg- ulation and identification of elicitor and light-responsive cis- acting elements. EM60 J. 8, 1641 -1648.

Long, S.R. (1 989). Rhizobium-legume nodulation: Life together in - the underground. Cell56, 203-214.

Rogers, S.G., Horsch, R.B., and Fraley, R.T. (1 986). Gene trans- fer in plants: Production of transformed plants using Ti plasmid vectors. Methods Enzymol. 118, 627-640.

Ryder, T.B., Hedrick, S.A., Bell, J.N., Liang, X., Clouse, S.D., and Lamb, C.J. (1 987). Organization and differential activation of a gene family encoding the plant defense enzyme chalcone synthase in fhaseolus vulgaris. MOI. Gen. Genet. 210,

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd ed (Cold Spring Harbor, NY:

219-233.

Cold Spring Harbor Laboratory). Schmelzer, E., Krüger-Lebus, S., and Hahlbrock, K. (1 989).

Temporal and spatial patterns of gene expression around sites of attempted funga1 infection in parsley leaves. Plant Cell 1,

Schulze-Lefert, P., Becker-André, M., Schulz, W., Hahlbrock, K., and Dangl, J.L. (1989). Functional architecture of the light- responsive chalcone synthase promoter from parsley. Plant Cell

Wiermann, R. (1981). Secondary plant products and cell and tissue differentiation. In The Biochemistry of Plants. Vol. 7: Secondary Plant Products, E.E. Conn, ed (New York: Academic Press), pp. 86-1 17.

Wightman, F., Schneider, E.A., and Thimann, K. (1980). Hor- mona1 factors c,ontrolling the initiation and development of lat- eral roots. Physiol. Plant. 49, 304-314.

993-1 001.

1,707-714.