Cloning of cDNA coding for dihydroflavonol-4-reductase (DFR) and characterization of dfr expression in the corollas of Gerbera hybrida var. Regina (Compositae)

Plant Molecular Biology 22: 183-193, 1993. © 1993 Kluwer Academic Publishers. Printed in Belgium. 183

Cloning of cDNA coding for dihydroflavonol-4-reductase (DFR) and characterization of dfr expression in the corollas of Gerbera hybrida var. Regina (Compositae)

Yrja Helariutta ~, Paula Elomaa 1 Mika Kotilainen 1, Pauli Sepp~inen 2 and Teemu H. Teeri ~ Institute of Biotechnology, University of Helsinki, Karvaamokuja 3, SF-00380 Helsinki, Finland," 2 Kemira 0 Y, Espoo Research Centre, P.O. Box 44, SF-022 71, Espoo, Finland

Received 2 November 1992; accepted in revised form 9 February 1993

Key words: Compositae, dfr, flavonoid genes, flower development

Abstract

We are approaching corolla differentiation in Compositae by studying the regulation of flavonoid pathway genes during inflorescence development in gerbera. We have cloned a dfr cDNA from a ray floret corolla cDNA library of Gerbera hybrida var. Regina by a PCR technique based on homologies found in genes isolated from other plant species. The functionality of the clone was tested in vivo by complementing the dihydrokaempferol accumulating petunia mutant line RL01. By Southern blot analysis, G. hybrida var. Regina was shown to harbour a small family of dfr genes, one member of which was deduced to be mainly responsible for the DFR activity in corolla. Dfr expression in corolla correlates with the anthocyanin accumulation pattern: it is basipetally induced, epidermally specific and restricted to the ligular part of corolla. By comparing the dfr expression in different floret types during inflorescence development, we could see that dfr expression reflects developmental schemes of the outermost ray and trans florets, contrasted with that of the disc florets.

Introduction

The inflorescence of the Compositae family is typically a large regulatory unit with developmen- tally, anatomically and functionally polymorphic flowers in which the organs are highly specialized [4, 21]. We are studying gene expression related to corolla differentiation in Compositae using gerbera (Gerbera hybrida) as a model. In gerbera there are three floret types (ray, trans and disc),

all with distinct corolla anatomy in some varieties (Fig. 1).

A number of flavonoid pathway genes have been isolated and their expression characterized during corolla development in several plant species (for reviews, see [11, 32]). Typically, their expression is transient in the corolla and devel- opmentally regulated, spatially following the pigmentation pattern. In Petunia hybrida and Anti- rrhinum majus the dfr gene, encoding an enzyme

The nucleotide sequence data reported will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Database under the accession number Z 17221.

184

e~

specific for the anthocyanin synthesis part in the pathway, has been shown to be a target for mul- tiple spatial regulation [3, 5, 16, 24].

By comparing different gerbera varieties, it is obvious that anthocyanin accumulation is specifically regulated at the levels of floral organ and floret type and also differentially controlled in the proximal and distal parts of the ligule. This indicates that the regulatory machinery leading to en- zymatic pigment production has components with spatial specificity. We are approaching corolla differentiation in gerbera by studying the regulation of the flavonoid pathway genes during inflorescence development.

In this article we describe the cloning of the dfr cDNA of gerbera (var. Regina) by PCR techniques based on homologies found in genes isolated so far. We tested the functionality and catalytic properties of the corresponding enzyme by introducing the gene into petunia line RL01 [25]. We also describe here the spatial and temporal pattern of dfr expression during the ray floret corolla development and the temporal expression pattern in the corollas of the different floret types during inflorescence development in the variety Regina.

Materials and methods

Plant material

Gerbera hybrida var. Regina was obtained from Terra Nigra BV. It was grown under standard

185

greenhouse conditions. Developmental stages of the inflorescence are shown in Table 1.

Isolation of plant DNA and RNA

Total DNA was isolated by the method of Del- laporta et al. [8]. Total RNA was isolated essen- tially as described [17]. Poly(A) + RNA was isolated by oligo(dT) cellulose affinity chroma- tography [27]. The concentrations of nucleic acids were determined by spectrophotometry.

PCR cloning

The first strand of cDNA was synthesized from 1/~g of poly(A) + RNA using the cDNA Synthe- sis System Plus (Amersham). One tenth of the first-strand cDNA preparation was used as a template for amplification of the dfr gene fragment as described by [27]. Taq polymerase (USB) was used, conditions for PCR being 94 °C/75 s, 55 °C/2min, 72 °C/2min, 25 cycles. Partially degenerate primers including the Barn HI sites CAAAGGATCCGAGAATGAAGT(A/G)AT- (A/C/T)AA(A/G)CC and AGAAGGATC- CAAAATACATCCATCC(A/C/G/T)GTCAT corresponding to the peptides ENEVIKP and MTGWMYF, respectively (Fig. 2), were used. The major PCR product (207 bp) was cloned into pSP73 plasmid (Promega) and it was sequenced as described below.

Fig. 1. Left: in gerbera there are three types of florets. Ray florets (left) and trans florets (middle) lack stamens; disc florets (right) are hermaphrodite. All corolla forms can be divided to tube (unpigmented) and zygomorphically bilabiate lips (epidermally anthocyanin and carotenoid pigmented). During inflorescence development, the two internal lips and one external lip (usually termed ligule) develop in a floret-specific way. In var. Regina, the ligule of the ray florets is large and the two internal lips rudimentary, in disc florets the two internal lips are together almost the same size as the ligule, and in trans florets the ligule is intermediate in size between those of ray and disc florets (terminology as in Drennan et aL [9]). Right: the outermost floret types (ray and trans) and the outermost ring of the disc florets develop simultaneously whereas the disc florets develop centripetally. The rapidity of late development of the disc florets is evident from the abrupt change in corolla anatomy of the rings with fully developed corolla compared to those with still unopened corolla.

Fig. 3. Left: the phenotype of a typical flower of an untransformed RL01. Right: the phenotype of a transformant with pHTT372.

Fig. 8. The basipetal dfr expression in the ray floret ligule by northern blot analysis. The ligule was divided to five parts of the same length along the axis in a given stage, b, the most proximal part; m, the middle part; d, the most distal part. 4, 5, 6 and 9: different developmental stages (Table 1).

186

Table 1. Developmental stages of the inflorescence and the appearance of anthocyanin pigmentation in different floret types. Transition from stage 1 to 9 and from stage 9 to 11 each requires about two weeks in summer.

Stage Length of ray Description floret corolla (rnm)

1 < 5 2 5-10 3 10-15 4 15-23

5 23-26 6 26-35 7 35-40

8 40-50 9 50-55

10 55-60 11 55-60

Ray floret corolla shorter than the bracts, the distal parts of the bracts touch the disc florets Ray floret corolla shorter than the bracts, the distal parts of the bracts are above the disc florets Ray floret corolla about as long as bracts Ray floret corolla longer than bracts, the tip of the ray and trans floret ligules faintly pigmented with anthocyanin, disc florets unpigmented The distal half of the ray and trans floret ligules pigmented The proximal part of the ray and trans floret ligules pigmented only above veins Also the proximal part of the ray and trans floret ligule is pigmented, the tip of the outermost disc floret corollas faintly anthocyanin pigmented, the inflorescence is half-opened The inflorescence is fully opened, anthers not visible in the disc florets Anthers visible in two outermost rings of disc florets, the corolla of the innermost disc florets still greenish The corolla of the innermost ray florets reddish, the anthers of these florets not visible The anthers of the innermost ray florets visible

Construction and screening of the cDNA library

The ray floret corolla (stage 5 and 6) poly(A) + R N A was used to construct a c D N A library in the plasmid p U E X 1 (Amersham). The screening of the library was performed according to standard techniques [27].

DNA sequencing

Prior to sequencing the c D N A was subcloned into plasmid pSP73 (Promega). Nucleotide sequencing was carried out by constructing a nested set of deletions using exonuclease III as described [ 14]. The deletion plasmids were sequenced by dideoxy termination [28] using double-stranded templates [ 13] and T7 D N A polymerase (Seque- nase, USB). The computer analyses were performed with the PC/Gene D N A and protein analysis package (Intelligenetics Inc.).

Transformation of the RL01 petunia

The gerbera dfr c D N A (gdfrl) was cloned under the control of the 35S promoter in the plant vec-

tor pHTT202, itself a derivative of pGSJ250 (a kind gift from J. Bot terman) where the selectable marker nos-nptlI was transferred from pGV- neo1103 [12]. The resulting plasmid pHTT372, very similar to p H T T 3 7 0 in [ 10], was mobilized to Agrobacterium harbouring pGV2260 [7] where it is stabilized by crossing over into the disarmed Ti plasmid. The Agrobacterium-mediated transformation was done according to Horsch et al. [15 ], but without using tobacco nurse culture on the regeneration plates.

RNA and DNA blot analyses

5 #g total R N A or 10/~g digested D N A was loaded per lane. The electrophoresis and hybrid- izations were done as described [27].

In situ hybridization

In situ hybridization was performed according to Cox and Goldberg [6]. A plasmid having a 315 bp insert from the 5' end of the gene was used as a source for SP6/T7 system R N A probes.

187

MEEDSPATVCVTGAAGFIGSWLVMRLLERGYVVHATVRDPGDLKKVKHLLELPKAQTNL MPLHLRCSATVCVTGAAGFIGSWLVMRLLERGYNVHATVRDPENKKKVKHLLELPKADTNL

MSPTSLNTSSETAPPSSTTVCVTGAAGFIGSWLVMRLLERGYTVRATVRDPGNMKKVKHLIELPKADTNL MEGGAGASEKGTVLVTGASGFAGSWLVMKLLQAGYTVRATVRDPANVGKTKPLMDLPGATERL

MDGNKGPVVVTGASGFVGSWLVMKLLQAGYTVRATVRDPANVEKTKPLLELPGAKERL

KLWKADLTQEGSFDEAIQGCHGVFHLATPMDFESKDPENEIIKPTIEGVLSIIRSCVKAKTVKKLVFTSS TLLKADLTVEGSFDEAIQGCQGVFHVATPMDFESKDPENEVIKPTVRGMLSIIESCAKANTVKRLVFTSS TLWKADMTVEGSFDEAIQGCEGVFHLATSMEFDSVDPENEVIKPTIDGMLNIIKSCVQAKTVKKFIFTTS SIWKADLAEEGSFHDAIRGCTGVFHVATPMDFLSKDPENE¥IKPTVEGMISIMRACKEAGTVRRIVFTSS SIWKADLSEDGSFNEAIAGCTGVFHVATPMDFDSQDPENEVIKPTVEGMLSIMRACKEAGTVKIVFTSS

AGTVNGQEKQLHVYDESHWSDLDFIYSKKMTAWMYFVSKTLAEKAAWDATKGNNISFISIIPTLVVGPFI AGTLDVQEQQKLFYDQTSWSDLDFIYAKKMTGWMYFASKILAEKAAMEEAKKKNIDFISIIPPLVVGPFI GGTVNVEEHQKPVYDETDSSDMDFINSKKMTGWMYFVSKILAEKAGMEAAKENNIDFISIIPPLVVGPFI AGTVNQEERQRPVYDEESWTDVDFCRRVKMTGWMYFVSKTLAEKAALAYAAEHGLDLVTIIPTLVVGPFI AGSVNIEERPRPAYDQDNWSDIDYCRRVKMTGWMYFVSKALAEKAAMEYASENGLDFISIIPTLVVGPFL

TSTFPPSLVTALSLITGNEAHYSIIKQGQYVHLDDLCECHIYLYENPKAKGRYICSSHDATIHQLAKIIK TPTFPPSLITALSLITGNEAHYCIIKQGQYVHLDDLCEAHIFLYEHPKADGRFICSSHHAIIYDVAKMVR MPTFPPSLITALSPITGNEAHYSIIKQCQYVHLDDLCEGHIFLFEYPKAEGRYICSSHDATIYDIAKLIT SASMSPSLITALALITGNAPHYSILKQVQLIHLDDLCDAEIFLFENPAAAGRYVCSSHDVTIHGLAAMLR SAGMPPSLVTALALITGNEAHYSILKQVQLVHLDDLCDAMTFLFEHPEANGRYICSSHDATIHGLARMLQ

DKWPEYYIPTKFPGIDEELPIVSFSSKKLIDTGFEFKY-NLEDMFKGAIDTCREKGLLPYSTIKNHINGN EKWPEYYVPTEFKGIDKDLPVVSFSSKKLTDMGFQFKY-TLEDMYKGAIDTCRQKQLLPFSTRSAEDNGH ENWPEYHIPDEFEGIDKDIPVVSFSSKKMIGMGFIFKY-TLEDMVRGAIDTCREKGMLPYSTKNNKGDEK DRYPEYDVPQRFPGIQDDLQPVRFSSKKLQDLGFTFRYKTLEDMFDAAIRTCQEKGLIPLA .... TAAGG DRFPEYDIPQKFAGVDDNLQPIHFSSKKLLDHGFSFRY-TTEDMFDAAIHTCRDKGLIPLGDVPAPAAGG

HVNGVHHYIKNNDDDHEKGLLCCSKEGQ NREAIAISAQNYASGKENAPVANHTEMLSNVEV EPILNSLENNYNIQDKELFPISEEKHINGQENALLSNTQDKELLPTSEEKRVNGLESALLSKIQDKEVLP DGFASVRAPGETEATIGA KLGAL--AAGEGQA-IGAET

3 TS GVKHAKGQENALLPDIANDHTDGRI

Fig. 2. Alignment of the deduced amino acid sequences for the D FR proteins from G. hybrida (1), P. hybrida (2) , A. majus (3), Z. mays (4 ) and H. vulgate (5) . * - conserved amino acid residue in every species. The peptide sequences used for PCR cloning are marked as bold.

Results

Cloning and sequence analysis of dfr cDNA from G. hybrida

The PCR fragment from the endogenous dfr gene of gerbera was used as a probe in screening a cDNA library constructed in pUEX1 (Amer- sham) prepared from the developing ray floret corollas. Five positive clones were identified from a library consisting of 20 000 cfu. Three of these contained an insert of 1.3 kb and two of 1.1 kb. These were sequenced from the 3' and 5' ends. The clones appeared to be derived from the same

transcript as evidenced by the perfect match in their 3' untranslated region. The two 1.1 kb clones appeared to be shorter at their 5' ends. By comparing the different clones, three polyadeny- lation sites were identified. The longest, nearly full-length cDNA clone, gdfrl (1.3 kb) in pUEX1 was subcloned into pSP73 to give pCGD8.1 and subsequently sequenced. The translational start and stop codons were deduced from the cDNA sequence. The putative translational initiation codon is flanked by the plant consensus sequence AACAATGG [221.

The deduced amino acid sequence of the protein encoded by the dfr gene of the cloned cDNA

188

is 366 residues long, corresponding to a molecular mass of 41.1 kDa. Comparison of the derived amino acid sequence (Fig. 2) to those from P. hybrida [3], A. majus [3], Zea mays [29] and Hor- deum vulgare [20] revealed a high degree of ho- mology between the genes, except at the beginning and end of the sequence (70~ , 70~o, 58~o and 59~o identity, respectively).

Determination of the functional identity of the cloned dfr cDNA

The anthocyanidin species that accumulate during corolla development in different gerbera varieties are pelargonidin and cyanidin [2]. The accumulation of pelargonidin suggests that the cloned dfr cDNA from gerbera belongs to the class of dfr genes that code for enzymes using dihydrokaempferol as a substrate. To determine the functional identity of the cloned gene, the longest dfr cDNA (gdfrl) was subcloned into a plant expression vector, resulting in the construct pHTT372. By Agrobacterium-mediated transformation [ 15] the gene was introduced into the petunia line RL01 [25] which, due to a genetic block to B-ring hydroxylation, accumulates dihydrokaempferol. In 11 of the 15 transformants the pale colour of the petals was converted to brick red (Fig. 3) indicating that the cloned cDNA en- codes an active enzyme using dihydrokampferol efficiently as its substrate. In addition to the fully brick red phenotype we also observed one transformant with a sectored phenotype and three transformants with no alteration in their pigmentation. The presence of the transgene in the genome of the five brick red transformants was ver- ified by DNA blot hybridization (data not shown).

The copy number of dfr genes in G. hybrida

To determine the complexity of dfr gene family in the gerbera genome, genomic DNA from var. Re- gina was digested with Hind III, Barn HI, Eco RI and Sca I, blotted and hybridized. A 241 bp fragment of the 5' primed end (an Nco I [in the

Fig. 4. Southern blot analysis of gerbera var. Regina. Probed with the 241 bp fragment. Washing conditions: 2 × S SC, 0.1% SDS, 58 °C.

pUEX1 adapter]-Hind III fragment)was used as a probe. The result is shown in Fig. 4. In the variety Regina, one to five bands were observed (with 2× SSC, 0.1~/o SDS, 58 °C washing conditions) indicating the presence of a small gene family in Regina.

Spatial dfr expression in the ray floret corolla in var. Regina

To determine the organ-specificity of the dfr expression in var. Regina during flower development, RNA was isolated from organs covering several developmental stages and different floret types. In the corolla, expression is very high, whereas in the stylar/stigmatic part of the carpel the expression is low. There is no detectable expression in ovary, stamen or pappus (Fig. 5A).

Figure 1 shows the anatomy of the ray floret corolla, consisting of an anthocyaninless tube and an elongated ligule pigmented epidermalty with anthocyanin. Figure 5B presents an RNA gel blot analysis of dfr expression in the corolla. Follow- ing the anthocyanin accumulation pattern, expres-

189

sion is detectable only in the ligular part of corolla. In situ, dfr mRNA accumulation is restricted to the epidermal cell layers of the ligule (Fig. 6). These results indicate that the regulation of dfr expression involves components which are specific for the epidermis of the ligular part of the corolla.

Fig. 5. Spatial specificity of the dfr expression during flower development by the Northern blot analysis. Organs covering several developmental stages and different floret types were used. a, organ specificity; b, region specificity in the corolla.

Temporal dfr expression in different floret types

To temporally compare the dfr expression in different floret types, the steady-state level of dfr mRNA was analysed at stages 1-11 in the ray, trans and outermost disc florets and at stages 4-11 in the innermost disc florets. Figure 7 shows that the developmental peak of dfr expression is simultaneous (at stage 7) in the ray, trans and outermost disc florets whereas in the innermost disc florets, dfr mRNA accumulates later (at stage 10). Furthermore, anthocyanin accumulation in ray, trans and outermost disc florets occurs simultaneously, whereas in the innermost disc florets, anthocyanin pigments appear later. The dfr mRNA accumulation reflects the developmental scheme of the inflorescence, with the outer rings consisting of different floret types developing simultaneously but the innermost rings with only disc florets developing centripetally (Fig. 1). The temporal dfr expression pattern in ray and trans florets also differs from that in disc florets in the onset ofmRNA accumulation, which is very rapid in disc florets compared to the gradual accumulation in ray and trans florets. This indicates the involvement of a floret type-specific component in the induction of dfr expression.

Fig. 6. In situ hybridization analysis of dfr expression in the ligule. Cross-section of the ligule of stage 6 (Table 1) was used. a, anatomy of the toluidine blue stained section (bar 50/~m; b, hybridization with antisense probe; c, hybridization with sense probe.

Basipetal dfr express&n in the ray floret ligule

In the elongated ligule of a ray floret, the anthocyanin accumulation and unrolling patterns are basipetal, indicating temporally different up- regulation of gene expression in different parts of the ligule. In Fig. 8, the basipetal anthocyanin accumulation together with an RNA blot analysis of dfr expression is shown. Following the appear-

190

Fig. 7. Temporal dfr expression in different floret types by the Northern blot analysis. Ray and trans florets from different rings were used. The outermost disc floret sample was collected from three outermost rings of disc florets, the innermost disc floret sample represents about a hundred innermost disc florets. 1-11, different developmental stages (Table 1).

ance of anthocyanin pigmentation, dfr expression is also basipetally induced, being first detectable only in the distal part of the ligule. Eventually the whole ligule expresses the gene. The directional extension of the dfr expression towards the proximal part of the ligule demonstrates the involvement of a regulatory system which can sense different parts of the apparently homogenous epidermis of the ligule.

Discussion

The structure of the cloned dfr cDNA

We isolated the dfr cDNA from var. Regina by a PCR technique based on homologies between the previously cloned dfr genes [3, 29] and subse- quent cDNA library screening. The deduced amino acid sequence (366 residues, Fig. 2) of the

DFR protein corresponds to a molecular mass of 41.1 kDa. We studied the functional identity and catalytic properties of the cloned gene by introducing it into petunia line RL01 which, due to a genetic block, accumulates dihydrokaempferol [25]. The function of the dfr gene of gerbera could be visualized in petunia by the conversion of pale petal colour to brick red (Fig. 3).

The property of the cloned gene to complement the genetic block in RL01 places the corresponding DFR enzyme into the group of enzymes ca- pable of using dihydrokaempferol as substrate. A comparison of the primary structures of the proteins deduced from the cloned dfr genes is pre- sented in Fig. 2. As expected from the taxonomy, the gerbera DFR is more similar to the proteins of dicot origin (P. hybrida [3] and A. majus [3]) than those of monocot origin (Z. mays [29], H. vulgare [20]).

The complexity of dfr genes in G. hybrida

From the number of bands hybridizing to the 241 bp probe which represents a conserved area in the dfr genes of different species, we can con- clude that in var. Regina there is a dfr gene family of up to five members. The presence of a gene family slightly complicates the expression analysis. We do not know how different the genes are and where they are expressed. All five clones isolated from the ray floret corolla cDNA library seemed to be derived from the same transcript. This indicates that at least in the corolla, one gene is more active than the others. This is parallel to petunia, where one of the three members of the family is mainly responsible for DFR activity during flower development [3]. In all the other species studied so far, the gene appears to be present as a single copy [20, 23, 29]. We have recently isolated a few genomic clones for the gerbera dfr genes (unpublished results), in order to charac- terize the promoter region of the dfr gene corresponding to the cloned cDNA.

Dfr expression in G. hybrida

From the analysis of dfr expression in different floral organs (Fig. 5A), we could see that the dfr expression is correlated with anthocyanin accumulation and is specifically up-regulated during carpel and corolla differentiation. There are also varieties in which all of the floral organs accumulate anthocyanin, suggesting that in certain gen- otypes dfr expression would be detected in other floral organs as well. Analysis of these varieties is in progress.

We further demonstrated that dfr expression, followed by the anthocyanin accumulation in the ray floret corolla, is basipetally inducible, epidermally specific and detectable only in the ligular part of corolla, but not in tube (Figs 5B, 6, 8). There is variation in the pigmentation pattern and spatial expression of flavonoid pathway enzyme genes in the corollas of different species. In A. majus, the structural genes of the anthocyanin pathway are expressed only in pigmented cells in

191

the epidermal cell layer of the lobe and of the tube [24, 16]. Anthocyanin synthesis occurs first in a ring at the base of the tube and in the lobes, after which the tube becomes fully pigmented [1 ]. In P. hybrida, chalcone synthase (chs), chalcone flavanone isomerase (chi) and dfr genes are expressed preceding pigmentation both in the tube and in the lobe [3, 18, 30]. The expression ofchs and chi genes has been shown (by promoter analysis with GUS gene fusions) to be expressed both in the epidermis and mesophyll (the latter unpigmented) [19, 31]. The differential regulation of flavonoid pathway genes during corolla development in different species reflects the diversification of corolla anatomy and development in the course of the evolution of flowering plants.

The basipetal induction of dfr expression together with anatomical observations of the basipetal development of the ligule (Fig. 8) suggest that there is a general basipetal gene up-regulation pattern during ligule development. We have also recently isolated and characterized chalcone synthase (chs) clones and observed their expression to be similarly basipetally inducible but beginning somewhat earlier than dfr expression (unpublished results). Weiss and Halevy [33] have shown that gibberellic acid is involved in the expression of flavonoid genes in petunia. This may indicate the involvement of a regulator gradient (e.g. gibberellic acid) extending in a basipetal manner over ligule and inducing different genes at different concentrations on this axis.

We further analysed dfr expression in different floret types during corolla development in the inflorescence. We found that dfr mRNA accumulation peaks simultaneously during inflorescence development in all floret types in the outermost rings of florets, whereas there is a clear difference in the timing of dfr expression between the outermost and the innermost disc florets (Fig. 8). We could also see differences in the expression kinet- ics of dfr mRNA accumulation between different floret types as demonstrated by the very fast induction in outermost disc florets compared to a more gradual induction in ray and trans florets (Fig. 7).

These results demonstrate the contrasting de-

192

velopmental schemes of the outermost ray and trans floret populations versus disc florets. The outermost floret types develop simultaneously and gradually, whereas the disc florets develop centripetally ring by ring (spirally, in fact), the final stages of development of a disc floret occur- ring very fast (Fig. 1). The existence of a floret type specific component in the regulation of the anthocyanin accumulation is further suggested by many varieties where the coloration of any of the three floret types can be different than that of the two others. The molecular analysis of varieties of this kind is in progress.

During inflorescence development, dfr expression is under a detailed developmental regulation apparently having specific components at the cel- lular level in the corolla and at the levels of rio- ret type and floral organ. Future work will focus on dfr expression in varieties with mixed anthocyanin accumulation patterns, and isolation and characterization of the promoter of the dfr genes active in the corollas.

Acknowledgements

We thank Prof. P. Meyer for the petunia RL01 and Dr J. Botterman for the plasmid pGSJ250. We also thank Ms Eija Holma, Ms Marja Huovila and Ms Pirjo Rahkola for excellent tech- nical assistance; Jaana Korhonen M.Sc. and Tapio Heino M.Sc. for their help in in situ hybridization and Dr Alan Schulman for critically reading the manuscript. This work was partially funded by the Academy of Finland.

References

1. Almeida J, Carpenter R, Robbins TP, Martin C, Coen ES: Genetic interactions underlying flower color patterns in Antirrhinum rnajus. Genes Devel. 3: 1758- 1767 (1989).

2. Asen S: High pressure liquid chromatographic analysis of flavonoid chemical markers in petals from Gerbera flowers as an adjunct for cultivar and germplasm identifica- tion. Phytochemistry 23:2523-2526 (1984).

3. Beld M, Martin C, Huits H, Stuitje AR, Gerats AGM:

Flavonoid synthesis in Petunia hybrida: partial characterization of dihydroflavonol-4-reductase genes. Plant Mol Biol 13:491-502 (1989).

4. Burtt BL: Aspects of diversification in the capitulum. In: Heywood VH, Harborne JB, Turner BL (eds) The Biol- ogy and Chemistry of the Compositae, pp. 41-59. Aca- demic Press, London (1977).

5. Coen ES, AlmeidaJ, Robbins TP, Hudson A, Carpen- ter R: Molecular analysis of genes determining spatial patterns in Antirrhinum majus. In Verma DPS, Goldberg RB (eds) Temporal and Spatial Regulation of Plant Genes, pp. 63-82. Springer-Verlag, Wien (1988).

6. Cox KH, Goldberg RB: Analysis of plant gene expression. In: Shaw CH (ed) Plant Molecular Biology: A Prac- tical Approach, pp. 1-34. IRL Press, Oxford (1988).

7. Deblaere R, Bytebier B, De Greve H, Deboeck F, Schell J, Van Montagu M, Leemans J: Efficient octopine Ti plasmid-derived vectors for Agrobactenum-mediated gene transfer to plants. Nucl Acids Res 13:4777-4788 (1985).

8. Dellaporta SL, Wood J, Hicks JB: A plant DNA mini- preparation: Version II. Plant Mol Biol Rep 1:19-21 (1983).

9. Drennan D, Harding J, Byrne TG: Heritability of inflorescence and floret traits in Gerbera. Euphytica 35: 319- 330 (1986).

10. Elomaa P, Honkanen J, Puska R, Seppanen P, Helariut- ta Y, Mehto M, Kotilainen M, Nevalainen L, Teeri TH: Agrobacterium-mediated transfer of antisense chalcone synthase cDNA to Gerbera hybrida inhibits flower pigmentation. Bio/technology (in press).

11. Forkmann G: Flavonoids as flower pigments: The formation of the natural spectrum and its extension by genetic engineering. Plant Breed 106:1-26 (1991).

12. Hain R, Stabel P, Czernilofsky AP, Steinbiss HH, Her- rera-Estrella L, Schell J: Uptake, integration, expression and genetic transmission of a selectable chimeric gene by plant protoplasts. Mol Gen Genet 199:161-168 (1985).

13. Hattori M, Yoshiyuki S: Dideoxy sequencing method using denatured plasmid templates. Anal Biochem 152: 232-238 (1986).

14. Henikoff S: Unidirectional digestion with exonuclease III creates targeted breakpoints for DNA sequencing. Gene 28:351-359 (1984).

15. Horsch RB, Fry J, Hoffmann N, NeidermeyerJ, Rog- ers SG, Fraley RT: Leaf disc transformation. In: Gelvin SB, SchilperoortRA, VermaDPS (eds) Plant Molecular Biology Manual, pp. A5/1-A5/9. Kluwer Ac- ademic Publishers, Dordrecht (1988).

16. Jackson D, Roberts K, Martin C: Temporal and spatial control of expression of anthocyanin biosynthetic genes in developing flowers ofAntirrhinum majus. Plant J 2: 425- 434 (1992).

17. Jones JDG, Duinsmuir P, Bedbrook J: High level expression of introduced chimeric genes in regenerated trans- formed plants. EMBO J 4:2411-2418 (1985).

18. Koes RE, Spelt CE, Mol JNM: The chalcone synthase multigene family of Petunia hybrida (V30): differential, light-regulated expression during flower development and UV light induction. Plant Mol Biol 12:213-225 (1989).

19. Koes RE, van Blokland R, Quattrocchio F, van Tunen AJ, Mol JNM: Chalcone synthase promoters in petunia are active in pigmented and unpigmented cell types. Plant Cell 2:379-392 (1990).

20. Kristiansen KN, Rohde W: Structure of the Hordeum vulgare gene encoding dihydroflavonol-4-reductase and molecular analysis of antl8 mutants blocked in flavonoid synthesis. Mol Gen Genet 230:49-59 (1991).

21. Leppik EE: The evolution of capitulum types of the Com- positae in the light of insect-flower interaction. In: Hey- wood VH, Harborne JB, Turner BL (eds) The Biology and Chemistry of the Compositae, pp. 61-89. Academic Press, London (1977).

22. Lutcke HA, Chow KC, Mickel FS, Moss KA, Kern HF, Scheele GA: Selection of AUG initiation codon differs in plants and animals. EMBO J 6:43-48 (1987).

23. Martin C, Carpenter R, Sommer H, Saedler H, Coen ES: Molecular analysis of instability in flower pigmentation of A. majus, following isolation of the pallida locus by trans- position tagging. EMBO J 4:1625-1630 (1985).

24. Martin C, Prescott A, Mackay S, Bartlett J, Vrijlandt E: Control of anthocyanin biosynthesis in flowers of Anti- rrhinum majus. Plant J I: 37-49 (1991).

25. Meyer P, Heidmann I, Forkmann G, Saedler H: A new petunia flower colour generated by transformation of a mutant with a maize gene. Nature 330:677-678 (1987).

26. O'Reilly C, Shepherd NS, PereiraA, Schwarz-Sommer

193

Z, Bertram I, Robertson DS, Peterson PA, Saedler H: Molecular cloning of the al locus of Z. mays using the transposable elements En and Mul . EMBO J 4:877-882 (1985).

27. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor (NY) (1989).

28. Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74:5463-5467 (1977).

29. Schwarz-Sommer Z, Shepherd N, Tacke E, Gierl A, Ro- hde W, Leclercq L, Mattes M, Berndtgen R, Peterson- PA, Saedler H: Influence of transposable elements on

the structure and function of the A1 gene of Zea mays. EMBO J 6:287-294 (1987).

30. van Tunen AJ, Hartman SA, Mur LA, Mol JNM: Regu- lation of chalcone flavanone isomerase (CHI) gene expression in Petunia hybrida: the use of alternative promoters in corolla anthers and pollen. Plant Mol Biol 12: 539-551 (1988).

31. van Tunen AJ, Mur LA, Brouns GS, RienstraJ-D, Koes RE, Mol JNM: Pollen- and anther-specific chi promoters from petunia: tandem promoter regulation of the chiA gene. Plant Cell 2:393-401 (1990).

32. van Tunen AJ, Mol JNM: Control of flavonoid biosynthesis and manipulation of flower colour. In: Grierson D (ed) Plant Biotechnology series, pp. 94-130. Blackie and Son, Limited, Glasgow (1991).

33. Weiss D, Halevy AH: Stamens and gibberellin in the regulation of corolla pigmentation and growth in Petunia hybrida. Planta 179:89-96 (1989).

Documents

Cloning of cDNA coding for dihydroflavonol-4-reductase (DFR) and characterization of dfr expression in the corollas of Gerbera hybrida var. Regina (Compositae)