* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Myb genes enhance tobacco trichome production
Gene therapy wikipedia , lookup
Gene expression programming wikipedia , lookup
Genome evolution wikipedia , lookup
Primary transcript wikipedia , lookup
Point mutation wikipedia , lookup
Nutriepigenomics wikipedia , lookup
Minimal genome wikipedia , lookup
Genome (book) wikipedia , lookup
Genetically modified crops wikipedia , lookup
Epigenetics in stem-cell differentiation wikipedia , lookup
Genetic engineering wikipedia , lookup
Gene therapy of the human retina wikipedia , lookup
Polycomb Group Proteins and Cancer wikipedia , lookup
Genome editing wikipedia , lookup
Site-specific recombinase technology wikipedia , lookup
Gene expression profiling wikipedia , lookup
Helitron (biology) wikipedia , lookup
Epigenetics of human development wikipedia , lookup
Microevolution wikipedia , lookup
Therapeutic gene modulation wikipedia , lookup
Artificial gene synthesis wikipedia , lookup
Designer baby wikipedia , lookup
Vectors in gene therapy wikipedia , lookup
671 Development 126, 671-682 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 DEV3883 Heterologous myb genes distinct from GL1 enhance trichome production when overexpressed in Nicotiana tabacum Thomas Payne, John Clement, David Arnold and Alan Lloyd* Department of Botany, and Institute for Cellular and Molecular Biology, University of Texas at Austin, Austin, TX 78713, USA *Author for correspondence (e-mail: [email protected]) Accepted 20 November 1998; published on WWW 20 January 1999 SUMMARY Myb-class transcription factors implicated in cell shape regulation were overexpressed in Arabidopsis thaliana and Nicotiana tabacum in an attempt to assess the extent to which cellular differentiation programs might be shared between these distantly related plants. GLABROUS 1, a myb gene required for trichome development in Arabidopsis, did not alter the trichome phenotype of the tobacco plants in which it was overexpressed. MIXTA, which in Antirrhinum majus is reported to regulate certain aspects of floral papillae development, did not complement the glabrous 1 mutant of Arabidopsis. However, 35S:MIXTA transformants of N. tabacum displayed various developmental abnormalities, most strikingly production of supernumerary trichomes on cotyledons, leaves and stems. INTRODUCTION Trichomes are specialized structures which originate and project from the above-ground epidermal tissues of most plants. A vast array of morphologies occur in addition to the hair-like shape suggested by the term. Trichomes may be composed of one or more cells, may be unbranched or branched, and may or may not possess glands which accumulate or secrete alkaloids like nicotine, terpenoids such as menthol and camphor, or other compounds repellent or toxic to phytophagous insects (Johnson, 1975; Levin, 1973; Rodriguez et al., 1984). Different trichome types may occur on the same plant of a given species, with type and spacing determined by the plant organ or organ surface (ie, adaxial vs. abaxial, leaf vs. stem) from which the trichome extends, and the developmental phase of the plant at the time of organ initiation. The trichome is a useful cell type for studying regulation of cell differentiation in plants. It is morphologically distinct from surrounding epidermal cells, and its genesis and development from this population are readily observed due to its accessibility to visual inspection and its relatively large size. In the laboratory, trichomes are non-essential and thus can be genetically manipulated without a reduction in plant viability. A number of mutants lacking or producing fewer, aberrant, or mis-positioned trichomes have been isolated from Arabidopsis, In addition, floral papillae were converted to multicellular trichomes. CotMYBA, a myb gene which is expressed in Gossypium hirsutum ovules and has some homology to MIXTA, was also overexpressed in the two species. A similar but distinct syndrome of abnormalities, including the production of cotyledonary trichomes, was observed in 35S:CotMYBA tobacco transformants. However, CotMYBA did not alter trichome production in Arabidopsis. These results suggest that the trichomes of Arabidopsis and Nicotiana are merely analogous structures, and that the myb genes regulating their differentiation are specific and separate. Key words: CotMYBA, MIXTA, Trichomes, Nicotiana tabacum and the order of action in the differentiation pathway of the genes so identified has been deduced by epistatic analysis (Hülskamp et al., 1994). Wild-type Arabidopsis leaf trichomes are unicellular, typically have three branches, and lack glands. Progression through the stages of leaf trichome morphogenesis in this species is correlated with overall cell enlargement and endoreduplicative DNA increase. Three rounds of endomitosis occur initially, concommitant with extension growth from the epidermal surface. A fourth round of endomitosis occurs after primary branching of the developing trichome in a proximodistal orientation to the leaf primordium surface has begun. Subsequently the distal branch itself branches, and the nucleus migrates to the branching point from its previous position below it. Trichome cells are regularly spaced in the mature Arabidopsis leaf. Initiation and maturation of trichomes proceeds in a generally basipetal direction over the leaf primordium, but additional trichomes may be initiated between mature trichome cells as organ growth continues (Hülskamp et al., 1994). Mutations at either of two loci, TRANSPARENT TESTA GLABRA 1 (TTG1) or GLABROUS 1 (GL1), result in the virtual absence of leaf trichomes from A. thaliana. TTG1 is thought to have roles in trichome precursor selection, trichome initiation, and in a nearest-neighbor inhibition of initiation in surrounding protodermal cells. Mutations at this locus have 672 T. Payne and others pleiotropic effects: in addition to the absence of trichomes, homozygous mutants fail to produce anthocyanin pigments and seed coat mucilage (Koorneef, 1981), and, conversely, produce supernumerary root hairs on root epidermal cells that are normally hairless (Galway et al., 1994). Previously, it was found that overexpression of the maize R gene (a mychomologous transcription factor which regulates anthocyanin synthesis in that species) under control of the CaMV 35S promoter could suppress the various phenotypes associated with the ttg1-1 mutant. In wild-type Arabidopsis overexpression of R leads to production of extra trichomes on leaves and stems, and in some transformants to trichomes on organs which normally lack them, such as petals, stamens and pistils (Lloyd et al., 1992). TTG1 has recently been cloned and the gene is reported to encode a protein with WD-40 repeats and no myc homology (Walker et al., 1997). Presumably, TTG1 regulates or acts in concert with the activity of a myc-class transcription factor or factors. We have recently identified two endogenous Arabidopsis myc-class genes. When constitutively expressed in transgenic plants both suppress to varying extents the trichome and other defects of the ttg1-1 mutation, albeit less dramatically than R (Lloyd et al., unpublished). Both GL1 and TTG 1 are necessary for normal Arabidopsis trichome initiation. The GL1 gene was cloned by T-DNA tagging (Marks and Feldmann, 1989). It encodes a putative transcription factor with myb-type DNA binding and Cterminal acidic activation domains (Oppenheimer et al., 1991). By itself GL1 is not sufficient to initiate trichomes in Arabidopsis. Overexpression from a 35S:GL1 construct does not suppress the ttg1-1 mutation, but it does lead to the production of an occasional supernumerary trichome on the cotyledons of wild-type transformants. In fact, a reduction in leaf trichome number is observed, and this paradoxical result has been attributed to a squelching phenomenon (Larkin et al., 1994) akin to that described by Ptashne and Gann (1990). Little is known about the molecular basis of trichome differentiation in other species, although the inheritance of indumental characters correlated with pest resistance has been studied in some crop plants such as cotton (Stephens and Lee, 1961; Wannamaker, 1957), soybean (Broersma et al., 1972), and tomato (Gentile et al., 1969; Goffreda et al., 1990). In his 1954 monograph, Goodspeed recognized five basic morphological classes of trichomes variously distributed within the genus Nicotiana, and assessed their transmission to F1 interspecific hybrids and naturally occurring amphidiploids. Such experiments are difficult to interpret, but differences in the inheritance of trichome types, especially in regard to organ distribution and degree of structural elaboration, were documented. Such differences are likely to be indicative of separate regulatory mechanisms for the various trichome classes. Noda et al. (1994) have cloned the gene encoding another myb-class transcription factor, MIXTA, from a Tam4-tagged allele of Antirrhinum majus. MIXTA appears to be a positive regulator of a directional cell wall synthesis which results in the conical-papillate shape of epidermal cells on wild-type adaxial petal surfaces. This cell shape is thought to affect the optical properties of the petal epidermis such that incident light passes through the pigmented cell before it is reflected, and thus perceived intensity of pigmentation is enhanced. Epidermal cells of mixta mutant petals are flat by comparison, but retain other characters normally associated with the conical cell type (pentagonal symmetry, striations), indicating that the role of the MIXTA gene in this differentiation pathway is specific and late. Recently, Glover et al. (1998) have published a characterization of the MIXTA overexpression phenotype in Nicotiana tabacum. The data presented herein are in good agreement with their results which indicate that the transgene can also regulate trichome differentiation in tobacco. Suppression of the pleiotropic Arabidopsis ttg1-1 mutant by overexpression of the R gene from maize, a monocot, dramatically demonstrated the potential utility of heterologous gene expression as a tool for manipulating and studying the regulation of cellular differentiation processes in plants. Such an approach necessarily assumes that the regulatory mechanisms of similar pathways and developmental programs have been at least partly conserved over evolutionary time. At the outset of the present study we hypothesized that this conservation would be reflected by heterologous functioning of known epidermal cell shape regulators. Herein we describe an unsuccessful attempt to replace the trichome-initiating function of the Arabidopsis GL1 gene by overexpression of MIXTA, a myb-class cell shape regulator from another dicotyledonous plant, Antirrhinum, and CotMYBA, a closely related myb-class gene from cotton. We further describe the MIXTA overexpression phenotype in Nicotiana tabacum, which, in contrast to our results with Arabidopsis, includes the production of supernumerary trichomes. Overexpression of CotMYBA in Nicotiana also activates trichome initiation, additionally amplifying a distinct class of multicellular trichomes. Since overexpression of either GL1 or R has no effect on trichome initiation in Nicotiana tabacum, our data indicate that the trichomes of Arabidopsis and Nicotiana are not homologous structures and that their differentiation is controlled by distinct regulatory genes. Furthermore, the development of papillate cells and certain classes of trichomes in Nicotiana appears to be controlled by the same regulatory pathway while separate myb-elements can differentially regulate the cell-fate decision to produce separate, partially overlapping subsets of trichomes. MATERIALS AND METHODS Plant strains and growth conditions Arabidopsis thaliana. pMXCD, pCMAX41 and pLBJ4 constructs were used to transform the wild-type WS ecotype, the ttg1-1 mutant in Landsberg erecta outcrossed to WS one time, and the gl1-1 mutant of ecotype Columbia. The R gene was previously overexpressed in the wild-type WS, wild-type RLD, and in the Landsberg erecta ttg1-1 mutant outcrossed to either WS or RLD one time (Lloyd et al., 1992). Plants were potted in Sunshine Mix no. 1 (Premier Horticulture Inc.) and grown under continuous fluorescent illumination at 22°C. Nicotiana tabacum. All constructs were used to transform tobacco introduction TI1347 ‘Praecox’ pMXCD was also transformed into strain NC744. Both were supplied by the US Tobacco Germplasm Collection. Previously, R and C1 overexpression phenotypes were also evaluated in the ‘Xanthi’ cultivar. Potting mix consisted of a 2:1:1 mixture of Baccto, Sunshine no. 1, and vermiculite. Plants were grown in a warm greenhouse with supplemental fluorescent light to achieve 16-hour days. Myb genes enhance tobacco trichome production Vector construction pLBJ4. The oligonucleotide GL1-C (5′-GGGGGGGGCTGCAGATTAAACTAAAGGCAGTACTC-3′) was used to prime reverse transcription of the GL1 cDNA from Col-0 RNA. The cDNA was amplified by PCR using the primers GL1-C and GL1-A (5′-GGGGGGGGGAATTCATGAGAATAAGGAGAAGAGATG-3′), and then cloned into pBluescript II SK− (Stratagene) as an EcoRI-PstI fragment to create pSRV2. The ends of the insert from pSRV2 were sequenced and subsequently cloned into the binary plant vector pKYLX71 (Schardl et al., 1987) as a HindIII-XbaI fragment. The gl11 mutation has been complemented by transformation of this construct into Arabidopsis (Lloyd et al., unpublished). pMXCD. The MIXTA cDNA was cloned by ligating a HindIII-SacI restriction fragment from the pUC-18-derived plasmid pJAM980 (Glover et al., 1998) into pKYLX71. pMMKY-1. Oligonucleotides MXM5 (5′-GGAAGCTTGAATTCATGGTAAAAAAAGGGCCATGGACTG-3′) and MXM3 (5′GGGAGCTCGTCGACCTAACGCTTCTTTAGATGAGTGTTCCAG3′) were used to amplify the first 356 nucleotides of MIXTA from pJAM980. MXM3 incorporates a stop codon after the arginine residue which terminates the R3 myb repeat. The PCR product was cloned into pBluescript II KS+ (Stratagene) as a SacI-HindIII fragment to create pMXM1-1. The pMXM1-1 insert was completely sequenced from M13 forward and reverse primers, and then subcloned into pKYLX71 (Schardl et al., 1987) as a SacI-HindIII fragment. pCMAX41. The CotMYBA gene was amplified by PCR from genomic Gossypium hirsutum cv. Texas Marker-1 DNA using the primers COTMYBA-5 (5′-GCGAATTCCTCGAGCCCATGGGACGATCACGTTGT-3′) and COTMYBA-3 (5′-GCGGATCCTCTAGACCTATGAATCCAAGGGTCTAC-3′). After digestion with XbaI and XhoI, the PCR product was ligated into pBluescript II KS+ (Stratagene) to create pBSCMAX4. The ends of this insert were sequenced and then the insert was subcloned into pKYLX71 as an XhoI to XbaI restriction fragment. pCMANX2. This overexpression construct contains an intronless version of CotMYBA amplified by RT-PCR from a pCMAX41transformed TI1347 line using the COTMYBA-5 and -3 primers described above. The PCR product was cloned into pBluescript II KS+ (Stratagene) as an XbaI to XhoI fragment to create pCMP2. The insert was completely sequenced and compared to the GenBank cDNA accession to verify correct splicing. The cDNA insert from pCMP2 was then subcloned into pKYLX71 as an XhoI to XbaI fragment. TI1347 plants transformed with pCMANX2 displayed the same range of developmental abnormalities seen in pCMAX41 transformants (data not shown). The construction of pAL144, 35S:R, and pAL71, 35S:C1, is described elsewhere (Lloyd et al., 1992). Plant transformation Binary constructs were introduced into Agrobacterium tumefaciens strain GV3101 containing pMP90 (Koncz and Schell, 1986) by electroporation. Arabidopsis was transformed using the protocol of Valvekens et al. (1988), except in the case of the pCMANX2 construct, which was vacuum infiltrated essentially as described by Bechtold et al. (1993). Nicotiana tabacum transformation was accomplished essentially as described by Horsch et al. (1985), except that explants were derived from hypocotyls of one-week-old seedlings. With the exception of pMMKY, a minimum of fifteen to twenty-five independent transformants were produced for each construct. Where overexpression phenotypes were not observed, transcription of the transgenes was verified for subsets of transformants using either northern blot analysis or RT-PCR. Scanning electron microscopy Plant materials were fixed overnight in 2% glutaraldehyde and 0.1 M cacodylate, then taken through an alcohol dehydration series, once in 25, 50, 75, 85, 95% and twice in 100% ethanol, for at least 2 hours 673 per step. Specimens were critical point dried in a Tousimis Samdri790 and sputter coated with a gold-palladium alloy using a Ladd instrument. Specimens were visualized with a Phillips 515 scanning electron microscope and photographed with Polaroid film. Plant sectioning and light microscopy Tissue samples were fixed and stained as described by Mauseth et al. (1984). Sections were photographed through an Olympus BX60 microscope with an Olympus PM-C35DX camera. Image processing Images for figures were scanned with a Sharp JX325 High Resolution Color Scanner. Figures were constructed using Adobe Photoshop 4.0. Colors were corrected to match the living specimens and adjustments in brightness and contrast were made using this program. Images were printed with a FUJIX Pictrography 3000 printer. RESULTS Overexpression of MIXTA in Arabidopsis thaliana does not complement the gl1-1 mutation Given the structural similarities between GL1 and MIXTA and their roles in the control of epidermal cell differentiation, we hypothesized that MIXTA might be able to substitute for the function of the Arabidopsis trichome regulator. The MIXTA cDNA was placed under the transcriptional control of the CaMV 35S promoter in a binary vector and used to transform by means of Agrobacterium tumefaciens wild-type Arabidopsis and the gl1-1 and ttg1-1 mutants. No differences in trichome production or other epidermal characters could be discerned. Transcription of the transgene was verified by northern blot analysis (data not shown). Since the MIXTA gene was isolated from Antirrhinum majus based on its effects on floral papillation, we also examined the petals of pMXCD transformants. Scanning electron micrographs of petal epidermis (data not shown) revealed no differences between 35S:MIXTA and untransformed plants of any of the three genotypes. No published transformation protocols for Antirrhinum exist, and the reciprocal experiment, overexpression of GL1 in that species, was not attempted. Mixta overexpression in Nicotiana tabacum results in supernumerary trichome production Like Antirrhinum majus and unlike Arabidopsis thaliana, Nicotiana tabacum produces flowers pigmented by anthocyanins. Concurrent with the Arabidopsis experiment, we transformed N. tabacum with pMXCD to evaluate the effects of MIXTA overexpression on floral pigmentation. Noda et al. (1994) have previously shown that in Antirrhinum the gene exerts its effects on floral phenotype by altering epidermal cell shape and thus perception of pigmentation rather than quantity of pigment. One floral phenotype of 35S:MIXTA transformants is shown in Fig. 1D. Several other transformants produced mottled pink flowers different from the one shown. The adaxial floral epidermis of the transformants was examined by scanning electron microscopy. The SEM shown in Fig. 1H is from the same transformant that produced the small flower in Fig. 1D. Remarkably, many of the papillae have elongated to produce what appear to be multicellular, uniseriate trichomes. These results are similar to those presented by Glover et al., (1998). For comparison, both a flower and floral epidermis 674 T. Payne and others from untransformed TI1347 are shown in Fig. 1A,E, (Fig. 1L). Hypocotyls and true leaves also produce respectively. The red flower in Fig. 1B is from a plant supernumerary trichomes. These same seedlings were fixed, overexpressing the maize R gene. Note that R overexpression sectioned, and stained for light microscopy as above. up-regulates anthocyanin production in flowers but does not Longitudinal sections through untransformed TI1347 and alter floral papillation (Fig. 1F). 35S:MIXTA seedling cotyledons are compared in Fig. 3A and Effects of MIXTA overexpression in tobacco were, however, B, respectively. Note that the 35S:MIXTA cotyledon is smaller visible prior to flowering. After transfer to soil, rooted shoots than the untransformed cotyledon; only the distal portion of the began to produce abnormal concave leaves with a spoon-like latter is shown. In general, cell expansion appears to be reduced appearance. This leaf curvature was reversed in later leaves, in 35S:MIXTA cotyledons, perhaps due to premature cell wall which were often convex. Leaf blades also showed a thickening. The papillate epidermis (both adaxial and abaxial) characteristic variegation, dark green tissue associated with is visible in Fig. 3B; associated debris is derived from veins against pale green ground tissue, the latter becoming trichomes projecting across the plane of the section. increasingly yellow as the leaves aged. The leaf epidermis in 35S:MIXTA root phenotypes have not been extensively the vein-associated dark green areas possessed a sheen which characterized, but a pronounced tendency toward reduced was absent from the rough-looking epidermis of the lighter elongation has been documented and increased cell wall green areas. This variegation and its corresponding uneven thickening is suggested by Fig. 4B and D, as compared witho pattern of reflectivity are seen in a close-up photograph of the A and C respectively. This is in contrast to the findings of adaxial leaf blade in Fig. 2B. To determine whether this Glover et al. (1998), who note that they were unable to variegation might be due to sub-epidermal cell shape changes, correlate a root phenotype with MIXTA overexpression. thick sections (25 µm) through adult leaves were sputter coated When GL1 was overexpressed in Nicotiana tabacum the and examined by SEM, data not shown. No gross differences phenotype of transgenic plants was indistinguishable from wild in sub-epidermal morphology or wall thickness judged to be type (compare Fig. 1C, G and K, with A, E, and I). sufficient to account for the variegation were seen (data not Transcription of the 35S:GL1 transgene was confirmed by RTshown). Thinner sections (10 µm) from the same specimens PCR (data not shown). This result is further evidence that GL1 were stained with safranin and counterstained with fast green, and MIXTA have distinct regulatory activities. In addition, R then examined by light microscopy. Adaxial epidermis with a was not observed to affect cotyledonary or leaf trichome pale, rough green appearance was primarily composed of development (Fig. 1,J). papillate cells and trichomes, whereas dark green areas were The myb gene CotMYBA affects trichome composed of mostly rounded cells and occasional trichomes differentiation when overexpressed in Nicotiana like those seen in the leaf epidermis of untransformed controls. tabacum This differential pattern of trichome occurrence and papillation A Blast search (Altschul et al., 1990) of the National Center for was subsequently confirmed in SEMs of adaxial leaf epidermis Biotechnology Information database using the complete (Fig. 2C-G). Leaves and petals are presumed to be Antirrhinum MIXTA cDNA sequence as a query, calls up a long homologous structures (Esau, 1977), so it is perhaps not surprising that MIXTA overexpression alters the epidermal characters of both, but our results are particularly interesting in that they suggest these alterations can affect perception of pigments with absorption spectra as diverse as anthocyanins and chlorophylls. Tobacco plants transformed with the 35S:MIXTA construct were selfed and the resulting T2 seedlings examined by SEM. The cotyledons of N. tabacum are normally devoid of trichomes except at the very base of the petiole (Fig. 1I), but in plants overexpressing MIXTA most of the cells of the cotyledonary epidermis have a papillate appearance, and Fig. 1. Floral and seedling phenotypes of Nicotiana tabacum overexpressing heterologous myc- and mybmany differentiate into class transcription factors. First row: flowers, all the same magnification, flower in A is approximately 2.25 glandular trichomes which are cm wide. Second row: SEMs of adaxial petal epidermis. Scale bar, 0.1 mm. Third row: SEMs of 2-weekeither unicellular or old seedlings showing adaxial epidermis of cotyledons (to the left and right) and first 2-3 true leaves. Scale bar, 1 mm. (A,E,I) Untransformed TI1347; (B,F,J) 35S:R. (C,G,K), 35S:GL1; (D,H,L), 35S:MIXTA multicelluar and uniseriate Myb genes enhance tobacco trichome production Fig. 2. Textural variegation of Nicotiana tabacum adaxial leaf epidermis. (A) Photograph of untransformed adaxial leaf epidermis. The midvein is vertical. (B) 35S:MIXTA leaf. (C-G) SEMs of adaxial leaf epidermis. (C) Untransformed. (D) 35S:MIXTA, from region adjacent to vein, corresponding to reflective areas in (B). (E) 35S:MIXTA, from interveinal, rough-looking regions as seen in B. Scale bar, 0.1 mm. (F) Detail of non-papillate cells seen in D. (G) Detail of papillate epidermis as in E. Area shown in F and G is 1 mm2. list of myb-homologous plant genes. One of the highest scoring but presumably non-orthologous sequences (Clement, Payne, and Lloyd, unpublished) is a GenBank accession, CotMYBA (L04497), isolated from Gossypium hirsutum 3-day pre-anthesis ovules. Cotton fibers are trichomes which are initiated and begin elongation from the epidermis of the cotton ovule at about the time of anthesis (Basra and Malik, 1984). Cotton fiber-specific promoters have previously been shown to direct transcription of reporter genes in trichomes of N. tabacum (Dang et al., 1996), suggesting that the two trichome types share some ontogenetic similarities. To explore the possibility that CotMYBA is a positive regulator of trichome differentiation, we overexpressed the genomic sequence (see Materials and Methods) in tobacco. SEMs of tobacco seedlings overexpressing the CotMYBA transgene are shown in Fig. 4. Whole 10-day-old T2 seedling shoots seen in A-D demonstrate the variation in overall severity of the phenotype observed in progeny of independent transformants. In all cases the cotyledons are distorted in comparison to untransformed plants (see Fig. 1I), with many surface irregularities. Cotyledonary margins are wavy and 675 bordered by elongated, turgid-looking cells. The exaggerated appearance of these border cells is emphasized in Fig. 4E,F. In untransformed seedlings they are similarly discernible as a double layer of transluscent cells at the cotyledonary margin, but lack the turgid appearance of the transformants. The angle of attachment of cotyledons to hypocotyl varies; some seedlings are Y-shaped as in Fig. 4B, with cotyledons remaining at less than 45o relative to the long axis of the hypocotyl, while others assume a normal perpendicular orientation. Interestingly, transformant cotyledon emergence from the seed coat is delayed relative to untransformed seedlings, whereas radicle emergence is not. One of the most striking characteristics of tobacco seedlings overexpressing CotMYBA is “loss” of the cotyledonary petiole. The cotyledonary petioles of untransformed seedlings are narrower than the cotyledonary blade, and epidermal cells are elongate and occur in longitudinally aligned files (see Fig. 4G). In the most severe 35S:CotMYBA phenotypes (Fig. 4D,H) the cotyledonary petiole becomes a wider shoulder region in which tissue bulges out from the plane of the blade, epidermal cells have a rounded appearance, and cell files are difficult to recognize. Like the plants overexpressing Mixta, 35S:CotMYBA plants produce supernumerary epidermal trichomes on cotyledons and other organs. In many of the transformants a preponderance of short-stalked, multicellular “jingle bell” trichomes (corresponding in Goodspeed’s classification scheme to E1, also referred to as hydathodes) like those shown in Fig. 4I, occur. Supernumerary trichomes of this type were not observed in plants overexpressing MIXTA. Whereas 35S:MIXTA cotyledonary trichomes in general appear to be regularly spaced, trichome initiation in 35S:CotMYBA transformant cotyledons looks more random, at least partly as a consequence of the aforementioned epidermal irregularities. To date, only T2 plants with the mildest phenotypes (Fig. 4A) have survived transplantation to soil. This is not surprising given our observations of seedlings allowed to continue growth on nutrient medium without hormones for up to one month (Fig. 4J-L). The cotyledons of such seedlings may become extremely chlorotic, whereas cotyledons of untransformed seedlings of the same age maintained on the same medium remain green and healthy. Subsequent true leaves may or may not be green at emergence, but often pale as they age. Invariably they are distorted to some degree. In the most severely affected seedlings the epidermis coarsens and becomes increasingly disorganized as it ages, even to the point of being callus-like. Bulbous projections from the epidermis occur, and these are distinct from observed trichomes (compare Fig. 4I, with J and L). The “shoulder” regions of the cotyledons may become pronounced, flattened projections distinguishable from young first true leaves by their positions relative to the rest of the cotyledon (Fig. 4K). A section through a one-month-old 35S:CotMYBA cotyledon is seen in Fig. 3D. Extreme variation in cell size and shape is characteristic. Large cells lack chloroplasts. (The overall appearance of tissue prior to fixation was chlorotic.) Whereas some tissue is collapsed and apparently necrotic, there are also isolated pockets of adventitious mitotic activity (Fig. 3F compare with Fig 3C and E, untransformed). A distinctive root phenotype is also associated with overexpression of CotMYBA in tobacco. As seen in Fig. 5E, 35S:CotMYBA seedling roots produce lollipop-like root hairs 676 T. Payne and others with bulbous tips. Similar structures are occasionally seen on untransformed TI1347 seedling roots, but on 35S:CotMYBA seedling roots they are much more numerous, and their frequency is directly correlated with severity of the phenotype of the aerial organs. A tendency toward disorganization and uncoordinated cell expansion has also been observed as 35S:CotMYBA roots age (data not shown). Arabidopsis thaliana was also transformed with the pCMAX41 construct. Neither of the trichome mutants (ttg1-1, gl1-1) were suppressed. Initial experiments using the transformation protocol of Valvekens et al. (1988) resulted in a small number of regenerants which were stunted and produced aberrant flowers. Attempts to secure transgenic seed or correlate floral phenotypes with degree of papillation on petal epidermis were likewise unsuccessful. In a recent experiment, vacuum infiltration was used to transform wild-type Arabidopsis with the pCMANX2 construct, which would allow us to evaluate transgenic cotyledon phenotypes even in the absence of fertile transgenics. Cotyledonary abnormalities somewhat comparable to those seen in 35S:CotMYBA tobacco seedlings were observed in approximately one quarter of the V1 transformants, but no cotyledonary trichomes were seen. Of 56 kanamycinresistant V1 seedlings transplanted to soil, twenty-five survived to flowering. Survival on soil was inversely correlated with severity of the cotyledon phenotype. Among the survivors were six plants (derived from three different infiltration seed lots) which shared a syndrome of developmental abnormalities. These included small concave or convex rosette leaves (Fig. 6A) as well as floral defects like those seen in the earlier transformation experiment. In these plants first- and second-whorl floral organs fail to expand properly, or do so in an uncoordinated fashion. As a consequence, fourth- and (to a lesser extent) third-whorl organs are exposed prematurely (Fig. 6B). To date neither pollen dehiscence nor self-set siliques have been observed, although the plants appear to be fertile when pollinated with wild-type pollen. Further characterization of the 35S:CotMYBA Arabidopsis phenotype is planned for these out-crossed progeny of the surviving abnormal V1 seedlings once sufficient seed stocks have been accumulated. DISCUSSION We have shown that the independent overexpression of two myb-class transcription factors, MIXTA and CotMYBA, significantly perturb development of transgenic tobacco plants, engendering complex phenotypes which include the production of supernumerary trichomes. These same transgenes failed to alter the trichome phenotype of Arabidopsis. A known regulator of trichome initiation from Arabidopsis, GL1, likewise had no effect on the trichome complement of Nicotiana tabacum when overexpressed in that species. Previous experiments in which the myc-homologous R gene was overexpressed in the two species suggested that whereas the regulation of the anthocyanin pathway was conserved even between monocots and dicots and in both Arabidopsis and N. tabacum, the development of trichomes in these dicots was regulated differently. Present results corroborate this hypothesis by identifying two genes which can control trichome cell-fate in Nicotiana but not in Arabidopsis. Limited correlations between myb gene sequence homology and function are seen Fig. 7 shows an alignment of the myb DNA binding domains of the three plant myb-element transcription factors used in the Fig. 3. Nicotiana tabacum seedling sections. (A) Untransformed TI1347 cotyledon. (B) 35S:MIXTA cotyledon, same age as in A. (C) Longitudinal section of TI1347 seedling shoot apex showing parts of one cotyledon and two young leaves. (D) Longitudinal section of 35S:CotMYBA seedling showing abnormal cotyledon. (E) Longitudinal section through untransformed TI1347 cotyledon. (F) Detail of 35S:CotMYBA cotyledon seen in D, showing internal “islands” of small, darkly staining cells surrounded by larger, apparently quiescent, highly vacuolated cells. Scale bars (A-C,F), 20 µm; (D and E) 50 µm and 10 µm respectively. Myb genes enhance tobacco trichome production 677 Fig. 4. SEMs of untransformed and 35S:CotMYBA tobacco seedlings. (A-D) T1 35S:CotMYBA seedlings 10-14 days post-germination, demonstrating variation in severity of phenotype. (Compare with Fig.1I.) (E) Cotyledonary margin of untransformed TI1347 seedling. (F) Cotyledonary margin of 35S:CotMYBA seedling seen in A. (G) Cotyledonary petiole of untransformed TI1347 seedling. (H) Cotyledonary petiole region of 35S:CotMYBA seedling seen in C. (I) Detail of 35S:CotMYBA seedling first true leaf adaxial epidermis showing preponderance of trichomes of Goodspeed types A and E. (J) Detail of one-month-old 35S:CotMYBA seedling showing progressive epidermal distortion; a cotyledon is seen far right. (K) Detail of petiolar/“shoulder” region of another one-month-old 35S:CotMYBA seedling. (L) Closeup of rounded-out epidermal cell from cotyledon of seedling shown in K. Scale bars (A-D,G,H,J,K) 1 mm; (E,F,I,L) 0.1 mm. present overexpression experiments, as well as those of the maize C1 protein (Cone et al., 1986; Paz-Ares et al. 1987), AtMIXTA, a presumed MIXTA ortholog from Arabidopsis (Rabinowicz et al., 1996), and PhMYB1 from Petunia hybrida, which has now been shown to regulate papillae development in that species (Avila et al., 1993; van Houwelingen et al., 1998). Overexpression of C1 in concert with R results in the up-regulation of anthocyanin synthesis in Arabidopsis, but overexpression of C1 either alone or together with R does not alter trichome production in tobacco, nor does C1 enhance the exaggerated trichome phenotype of 35S:R Arabidopsis plants (Lloyd et al. 1992). The similarities calculated for the myb DNA binding domains of the putative orthologs MIXTA, AtMIXTA, and PhMYB1 are over 90% while the similarity for any other pair is from 60-68%. Our results suggest a role for the GIDPXXH motif (Avila et al., 1993) in cell shape regulation. Of the five myb genes implicated in plant cell shape control, four contain this motif. Most of the plant myb genes so far described possess two myb repeats. Based on amino acid sequence similarities, it is thought that these have been derived from the ancestral sequences which gave rise to the vertebrate myb protooncogene second (R2) and third (R3) myb repeats (Lipsick, 1996; Martin and Paz-Ares, 1997; Rosinski and Atchley, 1998). A comparative sequence analysis of 50 plant R2R3-type myb genes recently conducted in this lab (Clement, Payne and Lloyd, unpublished) revealed the existence of many previously unrecognized conserved motifs outside the myb DNA binding domain, some conserved between monocots and dicots. Since a majority of known plant myb genes have been cloned by DNA binding domain sequence homology (for example, Li and Parrish, 1995), gene function is often unknown, making the assignment of function to the motifs largely conjectural. Heterologous overexpression experiments such as those described herein, together with sequence analysis, may enable such assignments to be made with greater confidence, especially when the gene in question is derived from a species in which complementable or revertable mutants and transformation/regeneration technology is lacking. A short N-terminal motif occurs in 31 of 50 R2R3 plant myb genes examined. Immediately C-terminal to the myb DNA binding domain, 20 genes examined encode a GIDPXXH amino acid sequence, as described in Avila et al. (1993); of these, 18 also encode the small amino-terminal motif, including MIXTA and CotMYBA. These motifs are boxed in Fig. 7. None of the flavonoid biosynthesis regulating mybs possess the GIDPXXH sequence, whereas it is present in all mybs directly implicated in plant cell shape regulation except GL1. A role in plant cell shape regulation for another myb from Arabidopsis, AtMYB305, has been inferred from experiments 678 T. Payne and others Fig. 6. Leaf and floral phenotypes of Arabidopsis overexpressing CotMYBA. (A) Concave rosette leaf of V1 35S:CotMYBA A. thaliana seedling, WS ecotype. (B) Flower of same, showing uncoordinated expansion of floral organs. multicellular, which might imply that endoreduplication is unnecessary for Nicotiana trichome morphogenesis. If trichome genesis in the two species is as dissimilar as our experiments indicate, one would not expect to encounter homology between the implicated myb genes other than that which was minimally required for DNA binding domain function. Besides the common N-terminal motif, homology between CotMYBA and MIXTA outside the myb DNA binding domain is confined to an identically positioned GIDPXTH amino acid sequence, which suggests a role for the latter motif in cell shape regulation in tobacco. Fig. 5. Root phenotypes of tobacco seedlings overexpressing MIXTA and CotMYBA. (A) Longitudinal section of untransformed TI1347 seedling root. (B) Longitudinal section through 35S:MIXTA root. (CE) Light micrographs of living 10-day-old seedling primary root tips growing on MS agar. (C) Untransformed TI1347 seedling root. (D) 35S:MIXTA root. (E) 35S:CotMYBA root. Since both 35S:MIXTA and 35S:CotMYBA roots elongate less than untransformed, D and E show proportionally more of the primary seedling root than C. Scale bars (A,B) 10 µm; (C-E) 1 mm. in which Arabidopsis genes were expressed under an inducible promoter in Schizosaccharomyces pombe, and isolated based on their apparent ability to cause cell division anomalies in the yeast (Xia et al., 1996). However, Moyano et al. (1996) have reported that its most similar counterpart in Antirrhinum majus is a regulator of phenylalanine ammonia lyase and other phenylpropanoid biosynthetic genes, suggesting that the result in S. pombe is due to a coincidence in gene product structure not correlated with gene function in plants. AmMYB308, which also contains the GIDPXXH motif, may affect pallisade cell shape in Antirrhinum (Noda et al., 1994) and it is reported to inhibit lignin biosynthesis when overexpressed in tobacco (Tamagnone et al., 1998). As previously mentioned (see Introduction), the morphogenesis of unicellular Arabidopsis trichomes is associated with endoreduplicative DNA increase. Melaragno et al. (1993) have suggested that once a plant cell undergoes a round of endoreduplication it becomes incompetent to undergo further cell division. Endoreduplication may thus constitute cellular commitment to a terminal differentiation program. Although no attempt to quantify DNA content has been made in the present study, the tobacco trichomes affected by overexpression of MIXTA and CotMYBA are frequently The phenotypes of tobacco plants overexpressing MIXTA or CotMYBA are distinct The most parsimonious explanation for our results would be that in Nicotiana tabacum, a myb-class transcription factor with extensive homology to MIXTA or CotMYBA regulates trichome differentiation, and that overexpression of the homologous genes from other species can substitute for or augment its activity to initiate ectopic trichome production. But such a facile interpretation fails to take into account the complexity of the observed phenotypes and the superficiality of their resemblance. A hypothetical gene-for-gene relationship is a better fit for the MIXTA data than for CotMYBA, since the effects of its overexpression appear to be specific to the epidermis. As previously stated (see Results), overexpression of MIXTA in tobacco results in hyperpapillate leaf and petal epidermis, which in turn alters perception of plant pigments consistent with the reported function of the gene in Antirrhinum. However, in tobacco plants overexpressing MIXTA many of these papillate cells become what appear to be multicellular trichomes. One would imagine that natural selection by pollinators has optimized pigment perception in the Antirrhinum floral epidermis and thus the duration of the directional cell wall deposition signal presumably directed by MIXTA. In the 35S:MIXTA transgenic tobacco plants an unregulated exogenous signal has been applied and some epidermal cells have demonstrated a prolonged capacity to respond. The multicellularity achieved by some of these trichomes may reflect a general cellular mechanism set to maintain the nuclear DNA/cytoplasmic volume ratio of cells within viable parameters. In the case of CotMYBA, the effects of overexpression in Myb genes enhance tobacco trichome production 679 tobacco are global, implying that some fundamental cellular The results of overexpression studies are potentially process is disrupted or exaggerated, one result being the artifactual. Because the transgenes are constitutively production of supernumerary trichomes. These trichomes are transcribed from the strong CaMV 35S promoter, the results of several distinct morphological classes, again indicating that must be interpreted carefully. It is possible that the the process regulated by CotMYBA is less specific. In severely overexpression phenotypes of MIXTA and CotMYBA in affected 35S:CotMYBA transformants, non-trichome cells tobacco are the result of non-productive interactions with gene maintain the ability to elongate and expand long after normal promoters or endogenous regulatory proteins. The maize C1-I cotyledon development has ceased. Pattern maintenance as mutant allele of C1 is an endogenous dominant repressor of well as pattern formation is defective. It may be that in cotton anthocyanin structural gene transcription. The C1-I gene ovules CotMYBA positively regulates cell wall extensibility to product is truncated relative to C1, resulting in loss of the Callow for the considerable unidirectional fiber growth which terminal amphipathic alpha helix, which functions in the fulloccurs during development of the cotton boll. In Nicotiana length allele as an activator of transcription (Paz-Ares et al., tabacum, increased cell wall extensibility might trigger a 1990; Goff et al., 1992). C1-I retains the myb DNA binding certain subset of competent epidermal cells (perhaps those domain, and apparently represses anthocyanin synthesis by having conducively oriented cytoskeletal elements) to competing with the full-length C1 gene product for binding to differentiate into trichomes. structural gene promoters. C1-I may also titrate out In a preliminary characterization of tobacco plants transcriptional activators by formation of non-functional overexpressing both MIXTA and CotMYBA, F1 progeny heterodimers with C1 (Franken et al., 1994) or other regulatory seedlings closely resemble seedlings overexpressing MIXTA proteins. When the Arabidopsis CAPRICE (CPC) gene is alone (Payne and Lloyd, . . . . 10 . . . .2 0 . . . . 30 . . . . 40 . . . .5 0 . . . .6 0 unpublished) indicating that the A mM I X TA M V R S P C C D K V G V-- K K GPW TV DE DQ K LL AY IE E HG HG S WRS LP L KAG LQ R CG K SC RL RW A A tM I X TA M G R S P C C D K L G L-- K K GPW TP EE DQ K LL AY IE E HG HG S WRS LP E KAG LH R CG K SC RL RW T MIXTA transgene is epistatic to P hM Y B 1 M G R S P C C D K V G L-- K K GPW TP EE DQ K LL AY IE E HG HG S WRA LP A KAG LQ R CG K SC RL RW T the CotMYBA transgene. MIXTA C ot M Y BA M G R S P C C E K AH T-- N K GAW TK EE DQ R LI NY IR V HG EG C WRS LP K AAG LL R CG K SC RL RW I Z mC 1 M G R R A C C A K E G V-- K R GAW TS KE DD A LA AY VK A HG EG K WRE VP Q KAG LR R CG K SC RL RW L may function earlier than A tG L 1 M R IR R R DE K EN QEY K K GLW TV EE DN I LM DY VL N HG TG Q WNR IV R KTG LK R CG K SC RL RW M CotMYBA in cell morphogenesis or the expansion-promoting . . . . 70 . . . .8 0 . . . . 90 . . . 1 00 . . . 11 0 . . . 12 0 activities regulated by the two A mM I X TA N Y LR P D IK RGP FSL QE EQT II QL HA L LG NR WS A IA SH L PKR TD N EIK NY W NT H LK KR LT R transgenes may in fact be A tM I X TA N Y LR P D IK RGK FNL QE EQT II QL HA L LG NR WS A IA TH L PKR TD N EIK NY W NT H LK KR LV K P hM Y B 1 N Y LR P D IK RGK FTL QE EQT II QL HA L LG NR WS A IA TH L PKR TD N EIK NY W NT H LK KR LV K antagonistic. C ot M Y BA N Y LR P D LK RGN FTE EE DEL II KL HS L LG NK WS L IA GR L PGR TD N EIK NY W NT H IK RK LI S In Antirrhinum majus floral Z mC 1 N Y LR P N IR RGN ISY DE EDL II RL HR L LG NR WS L IA GR L PGR TD N EIK NY W NS T LG RR AG A A tG L 1 N Y LS P N VN KGN FTE QE EDL II RL HK L LG NR WS L IA KR V PGR TD N QVK NY W NT H LS KK LV G papillae, MIXTA is thought to regulate directional deposition of . . . 1 30 . . . 14 0 . . . 1 50 . . . 1 60 . . . 17 0 . . . 18 0 cell wall material (Noda et al., A mM I X T A M G I D P V T H K P H T HN IL GH- GQ PK DV A NL N H IA Q WE S A R LQ A ER R LVR E S R LA Q NN NK IG T 1994). Random, excessive A tM I X T A M G I D P V T H K P K N ET PL SSL GL SK NA A IL S H TA Q WE S A R LE A EA R LAR E S K LL H LQ HY QT K P hM Y B 1 M G I D P V T H K P K N DA LL SHD GQ SK NA A NL S H MA Q WE S A R LE A EA R LVR Q S K LR S NS FQ NP L deposition of cell wall material C ot M Y B A R G I D P Q T H R P L N QT AN TNT VT AP TE L DF RN SP T SV S K S SSI KN P SLD FN Y NE F QF KS NT D coupled with an epidermal cell’s Z mC 1 G A GA G G SW VVV A PD TG SHA TP AA TS G AC ET GQ N SA AH R ADP DS A GTT TT S AA A VW AP KA V normal tendency to expand might A tG L 1 D Y SS A V KT TGE DDD SP PSL FI TA AT P SS C H HQ Q EN IY E NI A KS F NGV V S A SY E DK PK QE L result in the genesis of a trichome, since the integrity and shape of the . . . 1 90 . . . 20 0 . . . 2 10 . . . 2 20 . . . 23 0 . . . 24 0 A mM I X TA I Q RR L T WP LCL DNE QS NHY HS AL LN S TS AV GL N QD NS F TNY SA R PDN NN I YD D YE VN NI M wall facing outward would not be A tM I X TA T S SQ P H HH HGF THK SL LPN WT TK PH E DQ QQ LE S PT ST V SFS EM K ESI PA K IE F VG SS TG V reinforced or constricted by the P hM Y B 1 A S HE L F TS PTP SSP LH KPI VT PT KA P GS PR CL D VL KA W NGV WT K PMN DV L HA D GS TS AS A C ot M Y BA S L EE P N CT ASS GMT TD EEQ QE QL HK K QQ YG PS N GQ DI N LEL SI G IVS AD S SR V SN AN SA E walls of surrounding cells. The Z mC 1 R C TG G L FF FHR DTT PA HAG ET AT PM A GG GG GG G EA GS S DDC SS A ASV SL R VG S HD EP CR S supernumerary trichomes A tG L 1 A Q KD V L MA TTN DPS HY YGN NA -- -- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- 35S:MIXTA plants produce are uniseriate vectors essentially . . . 2 50 . . . 26 0 . . . 2 70 . . . 2 80 . . . 29 0 . . . 30 0 A mM I X TA G M IE F N NN SNY FAD SL RLP GF VE GI T DI SS SN I VL GA G VLP SN S DNV VG Y FE E NW SS VL N perpendicular to the epidermis. A tM I X TA T L MK E P EH DWI NST MH EFE TT QM GE G IE EG FT G LL LG G DSI DR S FSG DK N ET A GE SS GG D The intriguing possibility exists P hM Y B 1 T V SV N A LG LDL ESP TS TLS YF EN AQ H IS TG MI Q EN ST S LFE FV G NSS GS S EG G IM NE ES E that both MIXTA and CotMYBA C ot M Y BA S K PK V D NN NFQ FLE QA MVA KA VC LC W QL GF GT S EI CR N CQN SN S NGF YS Y CR P LD S- -- Z mC 1 G D GD G D WM DDV RAL AS FLE SD ED WL R CQ TA GQ L A- -- - --- -- - --- -- - -- - -- -- -- perturb cellular processes such that A tG L 1 - - -- - - -- --- --- -- --- -- -- -- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- the physical properties of cell walls or cytoskeletal elements are . . . 3 10 . . . 32 0 . . . 3 30 . . . 3 40 . . . 35 0 . . . 36 0 changed, and that these physical A mM I X TA N V AS S S SM DSP DVL VN ASS SY KM Y- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- A tM I X TA C N YY E D NK NYL DSI FN FVD PS PS DS P MF -- -- - -- -- - --- -- - --- -- - -- - -- -- -- manifestations in turn trigger the P hM Y B 1 E D WK G F GN SST GHL PE YKD GI NE NS M SL TS TL Q DL TM P MDT TW T AES LR S NA E DI SH GN N expression of trichome genes C ot M Y BA - - -- - - -- --- --- -- --- -- -- -- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- Z mC 1 - - -- - - -- --- --- -- --- -- -- -- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- downstream of endogenous A tG L 1 - - -- - - -- --- --- -- --- -- -- -- - -- -- -- - -- -- - --- -- - --- -- - -- - -- -- -- trichome initiation regulators, bypassing them. A possible role for Fig. 7. Alignment of five plant myb proteins. The homology within the DNA binding domain is physical constraints in determining indicated by shading. The amino- and carboxy-terminal conserved regions are boxed. AmMixta, organ differentiation has been MIXTA protein from Antirrhinum majus; AtMIXTA, MIXTA homolog from Arabidopsis thaliana; demonstrated in Helianthus PhMYB1, myb protiein from Petunia hybrida, CotMYBA, myb protein from Gossypium hirsutum; AtGL1, GL1 protein from Arabidopsis thaliana, ZmC1, functional C1 allele from Zea mays. (Hernandez and Green, 1993). 680 T. Payne and others overexpressed in wild-type A. thaliana, an overall reduction in trichomes and an increase in root hair production are seen. It has been proposed that CPC, which encodes a single R2homologous myb motif and no obvious transcriptional activation domain, competes with GL1 in the regulation of the GLABROUS 2 gene since the phenotypes of gl2 mutants and CPC overexpressing wild-type plants are similar (Wada et al., 1997). It is worthwhile to note that like GL1, C1 does not alter the development of tobacco plants in which it is overexpressed (Lloyd, unpublished). MIXTA or CotMYBA might bind to heterologous negative regulatory elements in gene promoters or titrate regulatory proteins such that trichome structural gene transcription would be derepressed. Such a competition hypothesis was tested by overexpressing MIXTA from which the putative activation domain had been deleted. A preliminary characterization of N. tabacum plants transformed with pMMKY, a construct encoding a MIXTA truncation from which everything C-terminal of the myb repeats is deleted, has revealed no developmental abnormalities, indicating that putative C-terminal transcriptional activation domains are necessary for production of the 35S:MIXTA phenotype (data not shown). This result suggests that the overexpression phenotype of the full-length MIXTA transgene is unlikely to be an artifact caused by titration of a MIXTA partner or nonproductive occupation of downstream gene promoters. It is unlikely that MIXTA or CotMYBA alter the normal course or timing of overall plant development Changes in epidermal characters such as trichome production have been correlated with phase change and the transition of the shoot from vegetative to inflorescence development programs. In maize, glossy-15 mutants result in the precocious acquisition of adult epidermal characteristics, including the elaboration of hairs (Moose and Sisco, 1994). Arabidopsis leaf trichome distribution changes gradually from exclusively adaxial (on early rosette leaves) to exclusively abaxial (on bracts) as the shoot develops (Telfer et al., 1997). Plants overexpressing R do not violate this trend, implying that in Arabidopsis competence to respond to trichome-specific differentiation regulators is governed by higher-order regulatory factors which are themselves modulated by such environmental stimuli as photoperiod. Increases in stem pubescence are correlated with inflorescence development in Antirrhinum majus (Bradley et al., 1996). None of the genes known to alter the timing of such developmental changes in trichome distribution (including genes involved in floral induction) are myb-class transcription factors. Time to flowering of plants overexpressing MIXTA is similar to that of untransformed plants. Slow growth of surviving 35S:COTMYBA T2 seedlings makes evaluation of flowering time problematical, but delays are likely due to pleiotropic effects of overexpression not associated with transition to flowering per se. The leafy cotyledon (lec) mutants of A. thaliana bear a superficial resemblance to N. tabacum plants overexpressing MIXTA or CotMYBA, in that they produce cotyledonary trichomes (Meinke et al., 1994; West et al., 1994). However, the precocious germination phenotype characteristic of lec1-1 seeds is not seen in tobacco plants overexpressing either MIXTA or CotMYBA, and supernumerary trichomes are not confined to embryonic epidermis. A number of Nicotiana species produce cotyledonary trichomes as a matter of course, and a change in the pattern of expression of a trichome regulatory gene seems a simpler evolutionary explanation than homeotic conversion or heterochrony. Although the cotyledons of plants overexpressing CotMYBA are distorted, their progressive loss of photosynthetic capacity makes them decidedly unlike normal true leaves. Since radicle emergence is simultaneous, slight delays in cotyledonary escape from the seed coat relative to wild-type are probably due to these distortions rather than to abnormalities in GA or ABA metabolism or response, which one would expect to have an effect on germination. Apparent contradictions between evolutionary distance and heterologous function of transgenes probably reflect parallel conservation of posttranscriptional regulatory mechanisms Paradoxically, overexpression of the Antirrhinum majus MIXTA gene was not observed to perturb development of floral papillae in Arabidopsis thaliana, even though the latter genome contains a putative orthologue which is 91.3% similar through the myb DNA binding domain. A likely explanation might be that the two genes are insufficiently diverged to overcome posttranscriptional checks operating in A. thaliana. Sequencespecific DNA binding by the vertebrate MYB oncoprotein has been shown to be regulated by a variety of post-transcriptional means, including phosphorylation and redox state (Guehmann et al., 1992; Lüscher et al., 1990; Myrset et al., 1993). Capacity for interactions with other regulatory proteins may also be vital to myb gene function. In maize, the C1 (a myb) and B (one of several R orthologues in this species, a myc) gene products interact to regulate anthocyanin synthesis (Goff et al., 1992). Myb- and myc-class transcription factors also co-regulate anthocyanin synthesis in Petunia hybrida (Quattrocchio, 1994); recently, another anthocyanin regulatory locus, An 11, has been cloned (de Vetten et al., 1997). It encodes a cytosolic protein with WD-40 repeats which acts upstream of the myb gene An 2. The TTG1 locus of Arabidopsis is reported to encode a WD-40 protein as well (Walker et al., 1997). Preliminary results of experiments conducted in this lab indicate that while overexpression of the maize R gene can strongly suppress the phenotype of the ttg1-1 mutant, two endogenous myc-class homologues do so more weakly. Overexpression of these same genes in a non-mutant background results in both excess trichome and anthocyanin production equivalent to wild-type plants overexpressing R, indicating that TTG1 influences the functioning of the endogenous myc gene products. The monocot gene may be sufficiently diverged to circumvent restraints imposed by TTG1 on its activity. MIXTA from A. majus may be too similar to its Arabidopsis orthologue to overcome equivalently similar regulatory mechanisms. It is interesting to note in this regard that overexpression of DELILA, the R orthologue from A. majus, results in less anthocyanin accumulation in tobacco than does overexpression of its monocot counterpart, and will not suppress the ttg1-1 mutant phenotype in A. thaliana (Mooney et al., 1995). The term “trichome” has been used as a catch-all designation for plant epidermal protuberances that in fact vary widely in morphology and function (Mauseth, 1988). Arabidopsis trichomes are exclusively unicellular, and probably function as Myb genes enhance tobacco trichome production physical and chemical barriers to insect feeding (Mauricio, 1998). Of the five distinct morphological classes of trichomes produced by N. tabacum (Goodspeed, 1954), two (types C and E) are glandular. Type D trichomes are defined by the presence of specialized stalk cells which may also accumulate secretory products. The fact that these do not occur together in every species of the genus Nicotiana implies that development of each type proceeds from the activity of specific and in some cases separate sets of regulatory genes, which would not preclude sharing of structural genes or trichome building mechanisms between types. MIXTA appears to exert its effects primarily on trichome types A and C. Overexpression of CotMYBA also promotes the development of Type E trichomes, which may indicate that it regulates or perturbs some growth process or fatedetermining mechanism fundamental to all three. At least 23 myb gene sequences have already been cloned from Arabidopsis (Wang et al., 1997), and according to one estimate up to a hundred may be present in the genome (Martin and Paz-Ares, 1997). It has been suggested that proliferation of the myb regulatory gene family was correlated with the evolution in plants of specialized physiological functions (Martin and PazAres, 1997). Indeed, the generalization can be made that those mybs to which functions have been assigned regulate late or downstream events in plant organ development, such as the epidermal characteristics of petals or leaves. Of course this bias in recovery might stem from the lethality of mutations in more fundamental developmental pathways. Given the large size of the myb gene family and observed regulatory association of these genes with downstream developmental events, it is perhaps not surprising that myb genes have been implicated in the differentiation of trichomes as different morphologically as those of Arabidopsis and N. tabacum. Our results imply that these myb genes (GL1 vs. MIXTA or CotMYBA) are as distinct as the structures their activities give rise to. This is in sharp contrast to the well-documented conservation of regulatory genes which exists in flower development (Coen and Meyerowitz, 1991; Mandel et al., 1992). From a hierarchical standpoint this disparity is readily explained by the non-essential character of the trichome relative to the fundamental importance of the flower in the plant life-cycle. And since trichomes are simpler organs, one would suppose that fewer structural genes need be subjected to coordinate regulation in order to produce them. We thank Dr Cathie Martin of the John Innes Institute for pJAM 980, Barbara Goettgens and John Mendenhall of the University of Texas Cell Research Institute for assistance with SEMs and image processing, and Dr Verne Sisson of the Oxford Tobacco Research Station for supplying Nicotiana tabacum germplasm. This work was partially supported by the Institute for Cellular and Molecular Biology, UT-Austin, and the Texas Higher Education Coordinating Board. REFERENCES Altschul, S. F., Gish, W., Miller, W., Myers, E. W. and Lipman, D. J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403-410. Avila, J., Nieto, C., Cana, L., Benito, M. J. and Paz-Ares, J. (1993). Petunia hybrida genes related to the maize regulatory C1 gene and to animal myb proto-oncogenes. The Plant J. 3, 553-562. Basra, A. S. and Malik, C. P. (1984). Development of the cotton fiber. Int. Rev. Cytol. 89, 65-113. Bechtold, N., Ellis, J. and Pelletier, G. (1993). In planta Agrobacterium- 681 mediated gene transfer by infiltration of adult Arabidopsis thaliana plants. C. R. Acad. Sci. Ser. III Sci. Vie 316, 1194-1199. Bradley, D., Vincent, C., Carpenter, R. and Coen, E. (1996). Pathways for inflorescence and floral induction in Antirrhinum. Development 122, 15351544. Broersma, D. B., Bernard, R. L. and Luckmann, W. H. (1972). Some effects of soybean pubescence on populations of the potato leafhopper. J. Econ. Entomol. 65, 78-82. Cone, K. C., Burr, F. A. and Burr, B. (1986). Molecular analysis of the maize anthocyanin regulatory locus C1. Proc. Natl. Acad. Sci. USA 83, 9631-9635. Coen, E. S. and Meyerowitz, E. M. (1991). The war of the whorls: genetic interactions controlling flower development. Nature 353, 31-37. Dang, P. M., Heinen, J. L. and Allen, R. (1996). Expression of a “cotton fiber specific” gene, Gh-1, in transgenic tobacco and cotton. Abstract 135, American Society of Plant Physiologists meeting. de Vetten, N., Quattrocchio, F., Mol, J. and Koes, R. (1997). The an11 locus controlling flower pigmentation in petunia encodes a novel WD-repeat protein conserved in yeast, plants, and animals. Genes Dev. 11, 1422-1434. Esau, K. (1977). Anatomy of Seed Plants, pp. 375-378. John Wiley and Sons, New York. Franken, P., Schrell, S., Peterson, P. A., Saedler, H. and Wienand, U. (1994). Molecular analysis of protein domain function encoded by the mybhomologous maize genes C1, Zm 1 and Zm 38. The Plant J. 6, 21-30. Galway, M., Masucci, J., Lloyd, A., Walbot, V., Davis, R. and Schiefelbein, J. (1994). The TTG gene is required to specify epidermal cell fate and cell patterning in the Arabidopsis root. Dev. Biol. 166, 740-754. Gentile, A. G., Webb, R. E. and Stoner, A. K. (1969). Lycopersicon and Solanum spp. resistant to the carmine and the two-spotted spider mite. J. Econ. Entomol. 62, 834-836. Glover, B. J., Perez-Rodriguez, M. and Martin, C. (1998). Development of several epidermal cell types can be specified by the same MYB-related plant transcription factor. Development 125, 3497-3508. Goff, S. A., Cone, K. C. and Chandler, V. L. (1992). Functional analysis of the transcriptional activator encoded by the maize B gene: evidence for a direct functional interaction between two classes of regulatory proteins. Genes Dev. 6, 864-875. Goffreda, J. C., Szymkowiak, E. J., Sussex, I. M. and Mutschler, M. A. (1990). Chimeric tomato plants show that aphid resistance and triacylglucose production are epidermal autonomous characters. The Plant Cell 2, 643-649. Goodspeed, T. H. (1954). The genus Nicotiana. In Chronica Botanica (ed. F. Verdoorn), pp. 102-131. Chronica Botanica Co., Waltham, Massachusettes. Guehmann, S., Vorbrueggen, G., Kalkbrenner, F. and Moelling, K. (1992). Reduction of a conserved cys is essential for myb DNA-binding. Nucl. Acids Res. 20, 2279-2286. Hernandez, L. F. and Green, P. B. (1993). Transductions for the expression of structural pattern: analysis in sunflower. The Plant Cell 5, 1725-1738. Horsch, R. B., Fry, J. E., Hoffman, N. L., Eichholz, D., Rogers, S. G. and Fraley, R. T. (1985). A simple and general method of transferring genes into plants. Science 227, 1229-1231. Hülskamp, M., Misera, S. and Jurgens, G. (1994). Genetic dissection of trichome development in Arabidopsis. Cell 76, 555-566. Johnson, H. B. (1975). Plant pubescence: an ecological perspective. Bot. Rev. 41, 233-258. Koncz, C. and Schell, J. (1986). The promoter of TL-DNA gene 5 controls the tissue-specific expression of chimaeric genes carried by a novel type of Agrobacterium vector. Mol. Gen. Genet. 204, 383-396. Koorneef, M. (1981). The complex syndrome of ttg mutants. Arabidopsis Information Service 18, 45-51. Larkin, J. C., Oppenheimer, D. G., Lloyd, A. M., Paparozzi, E. T. and Marks, M. D. (1994). Roles of the Glabrous1 and Transparent Testa Glabra genes in Arabidopsis trichome development. The Plant Cell 6, 1065-1076. Levin, D. A. (1973). The role of trichomes in plant defense. Quart. Rev. Biol. 48, 3-15. Lipsick, J. S. (1996). One billion years of myb. Oncogene 13, 223-235. Lloyd, A. M., Walbot, V. and Davis, R. W. (1992). Anthocyanin production in dicots activated by maize anthocyanin-specific regulators R and C1. Science 258, 1773-1775. Li, S. F. and Parrish, R. W. (1995). Isolation of two novel myb-like genes from Arabidopsis and studies on the DNA-binding properties of their products. The Plant J. 8, 963-972. Lüscher, B., Christenson, E., Litchfield, D. W., Krebs, E. G. and Eisenman, R. N. (1990). Myb DNA binding inhibited by phosphorylation at a site deleted during oncogenic activation. Nature 344, 517-522. 682 T. Payne and others Mandel, M. A., Bowman, J. L., Kempin, S. A., Ma, H., Meyerowitz, E. M. and Yanofsky, M. F. (1992). Manipulation of flower structure in transgenic tobacco. Cell 71, 133-143. Marks, M. D. and Feldmann, K. A. (1989). Trichome development in Arabidopsis thaliana. I. T-DNA tagging of the Glabrous1 gene. The Plant Cell 1, 1043-1050. Martin, C. and Paz-Ares, J. (1997). Myb transcription factors in plants. Trends Genet. 13, 67-73. Mauricio, R. (1998). Costs of resistance to natural enemies in field population of the annual plant Arabidopsis thaliana. The American Naturalist 151, 2028. Mauseth, J. D. (1988). Plant Anatomy, pp. 189-193 The Benjamin/Cummings Publishing Co., Inc., Menlo Park, California. Mauseth, J. D., Montenegro, G. and Walckowiak, A. M. (1984). Studies of the holoparasite Tristerex aphyllus (Loranthaceae) infecting Trichocereus chilensis (Cactaceae). Can. J. Bot. 62, 847-857. Meinke, D. W., Franzmann, L. H., Nickle, T. C. and Yeung, E. C. (1994). Leafy Cotyledon mutants of Arabidopsis. The Plant Cell 6, 1049-1064. Melaragno, J.E., Mehrotra, B. and Coleman, A. W. (1993). Relationship between endopolyploidy and cell size in epidermal tissue of Arabidopsis. The Plant Cell 6, 1661-1668. Mooney, M., Desnos, T., Harrison, K., Jones, J., Carpenter, R. and Coen, R. (1995). Altered regulation of tomato and tobacco pigmentation genes caused by the delila gene of Antirrhinum. The Plant J. 7, 333-339. Moose, S. P. and Sisco, P. H. (1994). Glossy15 controls the epidermal juvenile-to-adult phase transition in maize. The Plant Cell 6, 1343-1355. Moyano, E., Martinez-Garcia, J. F. and Martin, C. (1996). Apparent redundancy in myb gene function provides gearing fro the control of flavonoid biosynthesis in Antirrhinum flowers. The Plant Cell 8, 15191532. Myrset, A. H., Bostad, A., Jamin, N., Lirsac, P., Toma, F. and Gabrielsen, O. S. (1993). DNA and redox state induced conformational changes in the DNA-binding domain of the Myb oncoprotein. EMBO J. 12, 46254633. Noda, K., Glover, B. J., Linstead, P. and Martin, C. (1994). Flower colour intensity depends on specialized cell shape controlled by a Myb-related transcription factor. Nature 369, 661-664. Oppenheimer, D. G., Herman, P. L., Sivakumaran, S., Esch, J. and Marks, M. D. (1991). A myb gene required for leaf trichome differentiation in Arabidopsis is expressed in stipules. Cell 67, 483-493. Paz-Ares, J., Ghosal, D, Wienand, U., Peterson, P. A. and Saedler, H. (1987). The regulatory c1 locus of Zea mays encodes a protein with homology to myb proto-oncogene products and with structural similarities to transcriptional activators. EMBO J. 6, 3553-3558. Paz-Ares, J., Ghosal, D. and Saedler, H. (1990). Molecular analysis of the C1-I allele from Zea mays: a dominant mutant of the regulatory C1 locus. EMBO J. 9, 315-321. Ptashne, M. and Gann, A. A. (1990). Activators and targets. Nature 346, 329331. Quattrocchio, F. (1994). Regulatory genes controlling flower pigmentation in Petunia hybrida. Ph.D. thesis, Vrije Universiteit, Amsterdam, The Netherlands. Rabinowicz, P. D., Roberts, C. and Grotewold, E. (1996). A gene encoding a putative Mixta-homologous myb-domain protein (Accession No. X99809) from Arabidopsis thaliana. Pl. Physiol. 112, 863. Rodriguez, E., Healey, P. L. and Mehta, I. (1984). Biology and Chemistry of Plant Trichomes. Plenum Press, New York. Rosinski, J. A. and Atchley, W. R. (1998). Molelcular evolution of the myb family of trasncription factors: evidence of polyphyletic origin. J. Mol. Evol. 46, 74-83. Schardl, C. L., Byrd, A. D., Benzion, G., Altschuler, M. A., Hildebrand, D. F. and Hunt, A. G. (1987). Design and construction of a versatile system for the expression of foreign genes in plants. Gene 61, 1-11. Stephens, S. G. and Lee, H. S. (1961). Further studies on the feeding and oviposition preferences of the boll weevil (Anthonomus grandis). J. Econ. Entomol. 54, 1085-1090. Tamagnone, L., Merida, A., Parr, A., Mackay, S., Culianez-Macia, F.A., Roberts, K. and Martin, C. (1998). The AmMYB308 and AmMYB330 transcription factors from Antirrhinum regulate phenylpropanoid and lignin biosynthesis in transgenic tobacco. Plant Cell 10, 135-154. Telfer, A., Bollman, K. M. and Poethig, R. S. (1997). Phase change and the regulation of trichome distribution in Arabidopsis thaliana. Development 124, 645-654. Valvekens, D., Van Montagu, M. and Van Lijsebettens, M. (1988). Agrobacterium tumefaciens-mediated transformation of Arabidopsis thaliana root explants by using kanamycin selection. Proc. Natl. Acad. Sci. USA 85, 5536-5540. Van Houwelingen, A., Souer, E., Spelt, K., Kloos, D., Mol, J. and Koes, R. (1998). Analysis of flower pigmentation mutants generated by random transposon mutagenesis in Petunia hybrida. The Plant J. 13, 39-50. Wada, T., Tachibana, T., Shimura, Y and Okada, K. (1997). Epidermal cell differentiation in Arabidopsis determined by a Myb homolog, CPC. Science 277, 1113-1116. Walker, A. R., Davison, P. A., James, C. M., Esch, J., Marks, M. D. and Gray, J. C. (1997). The TTG1 gene does not encode a myc transcription factor. Abstract 3-79, 8th International Conference on Arabidopsis Research. Wang, Z., Kenigsbuch, D., Sun, L, Harel, E., Ong, M.S. and Tobin, E. M. (1997). A Myb-related transcription factor is involved in the phytochrome regulation of an Arabidopsis Lhcb gene. The Plant Cell 9, :491-507. Wannamaker, W. K. (1957). The effect of plant hairiness of cotton strains on boll weevil attack. J. Econ. Entomol. 50, 418-423. West, M. A. L., Yee, K. M., Danao, J., Zimmerman, J. L., Fischer, R. L., Goldberg, R. B. and Harada, J. J. (1994). LEAFY COTYLEDON 1 is an essential regulator of late embryogenesis and cotyledon identity in Arabidopsis. The Plant Cell 6, 1731-1745. Xia, G. Ramachandran, S., Hong, Y, Chan, Y., Simanis, V. and Chua, N. (1996). Identification of plant cytoskeletal, cell cycle-related and polarityrelated proteins using Schizosaccharomyces pombe. The Plant Journal 10, 761-769.