SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The Arabidopsis thaliana MERISTEM LAYER 1 (ATML1) gene is expressed in the epidermis of developing embryos and shoot meristems. To identify regulatory sequences necessary for epidermis-specific expression, three fusions of overlapping ATML1 genomic sequences to the GUS reporter gene were introduced into Arabidopsis plants. All fusion genes conferred epidermis-specific expression of both GUS mRNA and protein activity but varied in both the timing and relative levels of expression, suggesting partial redundancy of ATML1 regulatory elements. This study defines L1-specific regulatory sequences that are sufficient to direct foreign gene expression in a layer-specific manner.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Shoot apical meristems (SAMs) of higher plants are stratified into cell layers. Hanstein (1868) was the first to recognize that meristems have an epidermal layer, or L1, that covers distinct internal cell layers, or L2 and L3. The L1 is a single layer of cells, whereas L2 and L3 can each comprise more than one cell layer. Restrictions in the plane of cell division within the outer layers (L1 and L2) lead to the generation of two or three clonally distinct populations of cells in the SAM and its derivatives, as first shown in detail by Satina et al. (1940). In general, the L1 forms the epidermis and in some cases internal tissue at the margins of organs, the L2 gives rise to mesophyll and subepidermal layers of organs as well as to the germ line, and the L3 gives rise to the ground tissue and vasculature. However, meristem layer separation is not absolute, and there are many examples of cells in one layer invading into adjacent layers by periclinal divisions. The significance of SAM stratification is unclear, and it is not known whether it is causative of pattern or reflective of steric constraints. Furthermore, although the layers behave independently in their cell division patterns, they communicate during organ formation, as many studies have shown (reviewed in Szymkowiak & Sussex 1996).

Recently, several genes that are expressed in a meristem layer-specific manner have been described. For the L1, these include Arabidopsis thaliana MERISTEM LAYER 1 (ATML1) and LIPID TRANSFER PROTEIN 1 from Arabidopsis (Lu et al. 1996; Thoma et al. 1994). For L2/L3, these include KNOTTED1 from maize, A3 from tobacco, CENTRORADIALIS from snapdragon, and CLAVATA1 from Arabidopsis (Bradley et al. 1996; Clark et al. 1997; Jackson et al. 1994; Kelly & Meeks-Wagner 1995). The regulatory sequences responsible for layer-specific expression in meristems have not been identified for any of these genes.

The ATML1 locus encodes a homeodomain protein that is transcribed at high levels in the epidermis of developing embryos and the L1 of shoot and floral meristems (Lu et al. 1996). Expression begins in the apical cell of the one cell embryo, becomes restricted to the protoderm at the 16-cell stage of embryogenesis, and is expressed later in the L1 of shoot and floral, but not root, meristems (Lu et al. 1996). Here, we describe the region surrounding the ATML1 promoter and show that parts of this region are sufficient to drive epidermis-specific expression of a reporter gene in shoot, floral and root meristems.

Results and Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The alignment of ATML1 cDNA (Lu et al. 1996) and genomic sequences (Experimental procedures) suggests that the ATML1 transcription unit comprises 11 exons. The 5′ region of ATML1 is shown as a diagram in Fig. 1. The transcription start is tentatively defined as the 5′ end of the longest cDNA (Lu et al. 1996). The alignment shows that transcription starts at least 1.8 kb 5′ of the predicted first ATG, which lies in the third exon. The entire ATML1 genomic region has been sequenced (GenBank accession number AL035527).

image

Figure 1. Diagram of ATML1 and the map of fragments fused to GUS (➀, ➁, ➂).

Exons are indicated by open boxes, the putative transcription start site is labeled with an arrow, and the putative initiation ATG in the third exon is indicated. PAS76 and pAS85 are genomic subclones. Sites of primer 83 and primer 87 are indicated. B, BamHI; H, HindIII; X, XbaI.

Download figure to PowerPoint

To identify epidermis-specific transcriptional regulatory elements, we made three transformation vectors in which the GUS reporter gene (Jefferson et al. 1986) is transcriptionally fused to ATML1 promoter sequences (Fig. 1). The ML1::GUS 1 construct contains 200 bp of promoter sequence along with the first two exons and introns and part of the third exon, ML1::GUS 2 contains 3.5 kb of promoter sequence and part of the first exon, and ML1::GUS 3 contains 3 kb of promoter sequence along with the first two exons and introns and part of the third exon (Fig. 1). Except for the 5′ most 0.5 kb, ML1::GUS 3 combines the ATML1 sequences present in ML1::GUS 1 and ML1::GUS 2, which overlap for about 200 bp.

T2 siblings of four T1 lines of each construct in the Col-0 background were analyzed for accumulation of GUS mRNA in their inflorescence meristems and young floral buds using in situ hybridization. Table 1a gives the distribution of expression patterns found in individual ML1::GUS1, ML1::GUS2 and ML1::GUS3 lines. Figure 2 shows that lines from each construct express GUS mRNA specifically within the L1 layer, but that each line varies in the relative amount and timing of expression in the inflorescence and floral meristems. Four general classes of temporal expression patterns were observed. Class I lines showed expression in the inflorescence meristem and young stages of flower development (Fig. 2a). Class II lines showed expression in stage 1 and later floral primordia, and occasionally in the infloresence meristem at low levels (Fig. 2b). Class III lines showed expression after stage 1 of flower development (Fig. 2c), and class IV lines showed expression after stage 5 of flower development (Fig. 2d). These results suggest that ATML1 promoter activity increases from the meristem stage through to stage 5 of flower development.

Table 1.  . Distribution of staining patterns conferred by the three ML1::GUS constructs as assayed by (a) GUS RNA accumulation and (b) GUS activity. Class I, II, III and IV are the general staining patterns as described in the text
 ConstructnaClass IClass IIClass IIIClass IVNone
  • a

    Number of lines scored. In experiment (a), four T2 individuals per line were assayed. In experiment (b), 10 T2 individuals were stained and at least four of these 10 were sectioned.

(a) ML1::GUS1 425%75%
ML1::GUS2 450%25%25%
ML1::GUS3 425%50%25%
(b) ML1::GUS1 119%9%82%
ML1::GUS2 1436%7%43%14%
ML1::GUS3 1010%10%70%10%
image

Figure 2. GUS RNA accumulation in inflorescence tips of ML1::GUS lines and the four general staining classes.

(a) Class I, as in line ML1::GUS2.1; (b) class II, as in line ML1::GUS1.1; (c) class III, as in line ML1::GUS2.4; and (d) class IV, as in line ML1::GUS3.4. Numbers indicate stages according to Smyth et al. (1990). Scale bar: (a,b) 50 μm; (c) 100 μm; (d) 200 μm.

Download figure to PowerPoint

To compare the relative strengths of the three promoter fragments, T2 siblings of 10 lines of each construct were analyzed for GUS activity in their inflorescence meristems and young floral buds. All lines showed GUS activity in the epidermis, but varied in the level and onset of expression in the shoot apex. Figure 3 and Table 1b summarize the staining patterns found in the majority of siblings in each independent line. In general, ML1::GUS 2 lines had higher levels of GUS mRNA and enzyme activity than either ML1::GUS 1 or ML1::GUS 3 lines (Table 1b).

image

Figure 3. GUS activity accumulation in ML1::GUS lines and the four general staining classes.

(a) Class I, as in line ML1::GUS2.11; (b) class II, as in line ML1::GUS1.1; (c) class III, as in line ML1::GUS3.11; and (d) class IV, as in line ML1::GUS3.10. Numbers indicate stages according to Smyth et al. (1990). Scale bar: (a–c) 60 μm; (d) 120 μm.

Download figure to PowerPoint

Low levels of GUS staining were often observed in subepidermal layers of lines with the highest levels of expression. Staining in L2 and L3 increased when the concentration of ferri- and ferrocyanide salts in the GUS assay buffer was below 10 mm. To determine whether the subepidermal staining reflected low levels of GUS RNA expression, we conducted parallel experiments in which siblings from individual lines were either stained for GUS activity in varying concentrations of ferri- and ferrocyanide, or assayed for GUS RNA expression (Fig. 4). These experiments showed that GUS RNA expression was never detected in L2 and L3, while GUS staining was observed in subepidermal layers in a ferri- and ferrocyanide-dependent manner (Fig. 4). Potassium ferri- and ferrocyanide concentrations above 10 mm caused a decrease in GUS activity (data not shown). We conclude that the subepidermal staining is either due to leakage of the GUS enzyme or of the X-gluc reaction product.

image

Figure 4. Effects of potassium ferri- and ferrocyanide concentrations on apparent GUS activity.

Sixteen day T2 siblings of ML1::GUS2.3 were stained for either (a) GUS RNA accumulation; (b) GUS activity in 2 mm potassium ferri- and ferrocanide; or (c) GUS activity in 10 mm potassium ferri- and ferrocanide. Arrows in (c) show L2/L3 leakage. Scale bar: 100 μm.

Download figure to PowerPoint

To explore whether the 3.5 kb fragment could direct epidermis-specific expression during other stages of the Arabidopsis life cycle, 10 independent ML1::GUS2 lines were stained for GUS RNA or protein activity in embryos and 14-day-old plants undergoing the transition to reproductive development. In general, lines with strong L1-specific expression in the inflorescence shoot apex also showed high levels of L1-specific expression in the shoot meristems undergoing the transition from vegetative to reproductive development (Fig. 5f). Class II lines, which showed expression in floral, but not shoot meristems of reproductive apices, also lacked L1 expression in the transition meristem (Fig. 5e). Epidermis-specific GUS RNA expression in embryos was only observed with class I lines (Fig. 5a–d). Unexpectedly, GUS activity was also detected in the tips of the primary and lateral roots in most ML1::GUS 2 lines examined. This activity was confined to the epidermis of the root cap and meristem, and to the L1 of initiating lateral roots (Fig. 5g,h).

image

Figure 5. GUS RNA and activity during embryogenesis and vegetative growth.

(a–d) GUS RNA in developing ML1GUS2.3 embryos. Expression is not detected in the hypophosis. (e–h) GUS activity. (e) 14-day-old class II ML1::GUS 2.7; (f) 14-day-old class I ML1::GUS 2.1; (g) primary root of ML1::GUS 2.4; (h) lateral root of ML1::GUS 2.1. Scale bar: (a–d,g,h) 50 μm; (e,f) 100 μm.

Download figure to PowerPoint

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Strong epidermis-specific transcriptional regulatory elements are present in the promoter-proximal region of AtML1. The overlap of the three reporter constructs tested here includes a 200 basepair region that is located just upstream of the presumed transcription start and that probably contains either an L1-specific enhancer or an L2/L3-specific silencer. Redundant elements might lie in the –3.5 kb region and in the first or second intron. If such redundant elements are present, they do not appear to act additively.

The ATML1 promoter sequences described here should be useful for a variety of transgenic experiments. For example, several genes that control meristem and flower development have been suggested to act non-autonomously (Clark et al. 1997; Fletcher et al. 1999; Mayer et al. 1998) and their layer autonomy can be tested using the ATML1 promoter. The ATML1::GUS reporters on their own should also provide a convenient marker for L1 identity in embryonic mutants that have defects in layer organization (Mayer et al. 1991). Workers interested in studying protein trafficking in shoot and root meristems should also find ATML1 promoter sequences useful for studying protein movement between cells in L1 and L2. In the future, it may be feasible to engineer traits such as insect resistance into the epidermis of crop plants using regulatory regions of ATML1 orthologues from other species.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

ATML1 genomic clones

A 550 basepair fragment, generated by reverse transcription coupled to polymerase chain reaction (RT–PCR) and corresponding to nucleotides 926–1476 of the ATML1 cDNA (Lu et al. 1996), was kindly provided by Koji Goto (Kyoto University, Kyoto, Japan). Using this fragment, three overlapping clones were isolated from an Arabidopsis genomic library (λAS1, λAS2, and λAS6).

Reporter constructs (see Fig. 1)

Co-ordinates are relative to the start of the first exon (Lu et al. 1996).

ML1::GUS.1. A 3.8 kb XbaI fragment from λAS2 extending from –200 region to the beginning of exon 6 was subcloned into pBluescriptSK + (Stratagene) to create pAS85. A fragment extending from –200 to just upstream of the predicted first ATG was amplified from pAS85 by PCR with primer 83 (AAAAAG- CTTAGTCTCGAAATCCTTC) and a T7 primer, cut with HindIII and cloned into pBluescriptSK+ and pBI101.1 (Jefferson et al. 1986) to create pAS95 and pAS92 (ML1::GUS.1), respectively.

ML1::GUS.2. A 4.4 kb BamHI fragment from λAS6 extending from –4 kb to just downstream of exon 1 was subcloned into pBluescriptSK+ to create pAS76. A fragment extending from –3.5 kb upstream of the beginning of exon 1 was amplified from pAS76 by PCR using primer 87 (TTTAAGCTTAACCGG- TGGATTCAGGG) and a T7 primer, cut with HindIII and cloned into pBI101.1 to create pAS103 (ML1::GUS.2).

ML1::GUS.3. A fragment extending from –3 kb to just upstream of the predicted first ATG was cloned into the XbaI site of pBluescriptSK+ using a three-way cloning involving the 2.8 kb XbaI fragment of pAS76 and the 2.0 kb XbaI fragment of pAS95, creating pAS106. The 4.8 kb HindIII fragment of pAS106 was inserted into pBI101.1 to create pAS110 (ML1::GUS.3).

Constructs were introduced into the Columbia (Col-0) ecotype using Agrobacterium-mediated vacuum transformation (Bechtold et al. 1993). Transformants were selected on MS plates containing kanamycin.

GUS assays

Tissue was pre-fixed in ice-cold 90% acetone for 20 min on ice, rinsed with cold water for 5 min, vacuum infiltrated for 5 min on ice with staining solution (50 mm sodium phosphate buffer pH 7.0, 0.2% Triton-X-100, 10 mm potassium ferrocyanide, 10 mm potassium ferricyanide, 1 mm X-gluc) and incubated at 37°C for 12 h. Samples were changed through 30 min steps of 20% ethanol, 30% ethanol, 50% ethanol, FAA (50% ethanol, 5% acetic acid, 3.7% formaldehyde), dehydrated through an ethanol series into Histoclear (National Diagnostics), and embedded in Paraplast Plus. Eight μm sections were viewed after deparaffinization under Nomarski optics.

For comparisons of inflorescence staining among the three ML1::GUS constructs, for each construct the emerging inflorescence shoots of at least 10 T2 siblings from at least 10 lines were stained for GUS activity. Four to eight stained individuals from each line that showed GUS activity were embedded and sectioned (109 total) and the staining pattern of the majority of individuals was recorded. A similar strategy was used to compare the staining of 14-day-old seedlings among 12 independent ML1::GUS.2 lines (62 individuals sectioned).

In situ hybridization was performed according to Ferrándiz et al. (submitted). Digoxigenin-labeled GUS antisense RNA probes were generated according the manufacturer's specifications (Boehringer Mannheim) using pLS27 as a template (Blázquez et al. 1997).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Koji Goto for the ATML1 cDNA fragment; Cristina Ferrándiz for help with in situ hybridization; Miguel Blázquez for help with GUS staining; and Ella Mendoza for technical assistance. A.S. is a Department of Energy Fellow of the Life Sciences Research Foundation. D.W. is supported by the National Science Foundation (IBN 9723818). M.Y. is supported by the National Science Foundation (IBN9728402) and the National Institutes of Health (GM55328).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and Discussion
  5. Conclusion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Bechtold, N., Ellis, J. & Pelletier, G. 1993 In planta Agrobacterium mediated gene transfer by infiltration of adult Arabidopsis thaliana plants. C. R. Acad Sci., 316, 11941199.
  • Blázquez, M.A., Soowal, L., Lee, I. & Weigel, D. 1997 LEAFY expression and flower initiation in Arabidopsis. Development, 124, 38353844.
  • Bradley, D., Carpenter, R., Copsey, L., Vincent, C., Rothstein, S. & Coen, E. 1996 Control of inflorescence architecture in Antirrhinum. Nature, 379, 791797.
  • Clark, S.E., Williams, R.W. & Meyerowitz, E.M. 1997 The CLAVATA1 gene encodes a putative receptor kinase that controls shoot and floral meristem size in Arabidopsis. Cell, 89, 575585.
  • Fletcher, J.C., Brand, U., Running, M.P., Simon, R. & Meyerowitz, E.M. 1999 Signaling of cell fate decisions by CLAVATA3 in Arabidopsis shoot meristems. Science, 283, 191114.
  • Hanstein, J. 1868 Die Scheitelzellgrouppe im Vegetationspunkt der Phanerogamen. Festschrift der Niederrheinischen Gesellschaft f˛r Natur- u. Heilkunde. 109–134.
  • Jackson, D., Veit, B. & Hake, S. 1994 Expression of maize KNOTTED1 related homeobox genes in the shoot apical meristem predicts patterns of morphogenesis in the vegetative shoot. Development, 120, 405413.
  • Jefferson, R.A., Burgess, S.M. & Hirsch, D. 1986 B-Glucuronidase from E. coli as a gene fusion marker. Proc Natl Acad Sci. USA., 83, 84478451.
  • Kelly, A.J. & Meeks-Wagner, D.R. 1995 Characterization of a gene transcribed in the L2 and L3 layers of the tobacco shoot apical meristem. Plant J., 8, 147153.
  • Lu, P., Porat, R., Nadeau, J.A. & O'Neill, S.D. 1996 Identification of a meristem L1 layer-specific gene in Arabidopsis that is expressed during embryonic pattern formation and defines a new class of homeobox genes. Plant Cell, 8, 21552168.
  • Mayer, K.F., Schoof, H., Haecker, A., Lenhard, M., Jurgens, G. & Laux, T. 1998 Role of WUSCHEL in regulating stem cell fate in the Arabidopsis shoot meristem. Cell, 95, 805815.
  • Mayer, U., Torres Ruiz, R.A., Berleth, T., Miséra, S. & Jürgens, G. 1991 Mutations affecting body organization in the Arabidopsis embryo. Nature, 353, 402407.
  • Satina, S., Blakeslee, A.F. & Avery, A.G. 1940 Demonstration of the three germ layers in the shoot apex of Datura by means of induced polyploidy in periclinal chimeras. Am. J. Bot., 44, 311317.
  • Smyth, D.R., Bowman, J.L. & Meyerowitz, E.M. 1990 Early flower development in Arabidopsis. Plant Cell, 2, 755767.
  • Szymkowiak, E.J. & Sussex, I.M. 1996 What chimeras can tell us about development. Ann. Rev. Plant Mol Biol, 47, 351376.
  • Thoma, S., Hecht, U., Kippers, A., Botella, J., Devries, S. & Somerville, C. 1994 Tissue-specific expression of a gene encoding a cell wall-localized lipid transfer protein from Arabidopsis. Plant Physiol., 105, 3545.
Footnotes
  1. GenBank accession number AL035527.