The xylem and phloem transcriptomes from secondary tissues of the Arabidopsis root-hypocotyl

Abstract

The growth of secondary xylem and phloem depends on the division of cells in the vascular cambium and results in an increase in the diameter of the root and stem. Very little is known about the genetic mechanisms that control cambial activity and the differentiation of secondary xylem and phloem cell types. To begin to identify new genes required for vascular cell differentiation and function, we performed genome-wide expression profiling of xylem and phloem-cambium isolated from the root-hypocotyl of Arabidopsis (Arabidopsis thaliana). Gene expression in the remaining nonvascular tissue was also profiled. From these transcript profiles, we assembled three sets of genes with expression significantly biased toward xylem, phloem-cambium, or nonvascular tissue. We also assembled three two-tissue sets of genes with expression significantly biased toward xylem/phloem-cambium, xylem/nonvascular, or phloem-cambium/nonvascular tissues. Localizations predicted by transcript profiles were supported by results from promoter-reporter and reverse transcription-polymerase chain reaction experiments with nine xylem- or phloem-cambium-biased genes. An analysis of the members of the phloem-cambium gene set suggested that some genes involved in regulating primary meristems are also regulators of the cambium. Secondary phloem was implicated in the synthesis of auxin, glucosinolates, cytokinin, and gibberellic acid. Transcript profiles also supported the importance of class III HD ZIP and KANADI transcription factors as regulators of radial patterning during secondary growth, and identified several members of the G2-like, NAC, AP2, MADS, and MYB transcription factor families that may play roles as regulators of xylem or phloem cell differentiation and activity.

Figure 1.

Three tissue samples can be isolated from the root-hypocotyl. A, Extensive secondary growth is evident in the root-hypocotyl of an 8-week-old Arabidopsis plant. Lignified vessels and fibers of secondary xylem are stained blue with TBO. Nonlignified primary cell walls of cells in secondary xylem, secondary phloem, and nonvascular tissues are stained pink with ruthenium red. B, Nonvascular tissue of the outer bark can be separated from secondary phloem, yielding the nonvascular sample for expression profiling. C, Secondary phloem can be separated from secondary xylem, yielding the phloem-cambium and xylem samples for expression profiling. Free-hand transverse sections were prepared from fresh tissue just prior to staining. Tissues were dissected after staining. nv, Nonvascular; pc, phloem-cambium; sp, secondary phloem; sx, secondary xylem; x, xylem. Bars = 50 μm.

Figure 2.

Ratio-intensity plots of between-tissue comparisons with statistically biased genes (≥3.29 sds, P ≤ 0.001) highlighted. The vertical axis plots the log₂ of the ratio of the normalized MSIs for the same gene between two tissues: X versus PC (A), X versus NV (B), and PC versus NV (C). The horizontal scale specifies the log₁₀ of the product of the two MSIs for the same gene between two tissues. Genes with between-tissue means below 50 (0.1 times the among-genes mean) were considered insignificant regardless of ratio due to the noise in that region. Correlations (R² values) are shown in the lower right corner for each between-tissue comparison.

Figure 3.

“Triangle plot” of the relative MSIs among three tissues. Each gene is represented by a point whose proximity to each of the tissue-labeled corners reflects the relative expression in that tissue (see “Materials and Methods” for the numerical formula). Only genes with MSI values ≥200 are shown. Genes expressed in a single tissue lie in an extreme corner, while those expressed equally in all three lie in the center. Genes with signals significantly (P ≤ 0.001) higher in one tissue relative to both other tissues are highlighted in red (so-called “one-tissue genes”). Genes with signals not significantly different for two tissues but higher in both of those tissues relative to the third are highlighted in green (so-called “two-tissue genes”). Genes with signals significantly biased in only one pair-wise comparison or those not significantly biased in any pair-wise comparison are shown as large and small black points, respectively.

Figure 4.

Promoters for CLV1, MYR1, ZFWD1, XND1/ANAC104, and At1g20160 direct expression of reporters in vascular tissues as predicted from genome-wide transcript profiles of isolated vascular tissues. A, GFP expression driven by the promoter for CLV1 localized to the cambial zone and secondary phloem. A, Inset, The secondary phloem of wild-type control plants exhibited no detectable green fluorescence. B and C, GUS activity driven by the promoter for the phloem-cambium-biased, G2-like transcription factor MYR1 was detected in the secondary phloem of the root (B) and inflorescence stem (C). D and E, GUS activity due to the MYR1 promoter was also detected throughout the vascular tissue of the leaf (D), where it localized predominantly to the abaxial (phloem) side of the vascular bundle, as shown in a transverse section through a midvein (E). F, GUS activity driven by the promoter for the xylem-biased gene ZFWD1 was limited to vascular tissues. G, Higher magnification of the area within the black box (F) revealed that ZFWD1p::GUS expression was associated with xylem cells (arrow). H, XND1/ANAC104 promoter-driven GUS activity visible beneath the bark of roots on 8-week-old plants (H, inset) was localized to xylem vessels, shown here following isolation of two adjacent vessels (one mature GUS-negative and one immature GUS-positive) from secondary xylem. I, XND1p::GUS expression in the shoot was limited to senescing leaves, where it was evident in xylem cells surrounding mature vessels (arrows) and immature vessels (arrowhead) in the midvein. J, XND1p::GUS expression in the primary tissues of seedling roots was limited to TEs, as shown for metaxylem cells (arrow) adjacent to mature protoxylem cells. K, GUS activity driven by the promoter for a xylem-biased subtilisin-like Ser protease At1g20160 was localized predominantly in xylem cells surrounding mature vessels (arrows) and additional cells on the adaxial (xylem) side of the midvein. L, Vascular tissue-localized expression for At1g20160p::GUS is shown in a representative leaf. A, The free-hand transverse section was prepared from a 6-week-old (or 5-week-old for inset) hypocotyl, and cell walls were counterstained with propidium iodide prior to detection of GFP by confocal microscopy. B, C, E, I, and K, Free-hand transverse sections of root-hypocotyl, stem, or petiole were prepared following histochemical staining for GUS activity. D, Whole-mount leaf from a 5-week-old plant. F and G, Whole-mount cotyledon from a 3-d-old seedling. H, Xylem vessels were isolated from a root similar to that shown in the inset following histochemical staining for GUS activity. J, Whole-mount root from a 3-d-old seedling. L, Whole-mount leaf from a 4-week-old plant. cz, Cambial zone; nv, nonvascular tissue; p, phloem; x, xylem. Bars = 100 μm (A, inset in A, and B); 50 μm (C); 25 μm (E, G, H, I, J, and K); 500 μm (inset in H). B, Broken lines indicate the limit of phloem or nonvascular tissues.

Figure 5.

RT-PCR results for selected tissue-biased genes are consistent with predictions from genome-wide transcript profiles of isolated vascular tissues. Ethidium bromide-stained gels show products of RT-PCR for tissue-biased genes selected from xylem (X; At1g02250 and At1g32770), phloem-cambium (PC; At2g03500 and At3g04030), and nonvascular (NV; At3g26450) gene sets. Numbers of PCR cycles used were 27 for At3g26450 and actin (ACT7) and 30 for At1g02250, At1g32770, At2g03500, and At3g04030. Different PCR cycle numbers and annealing temperatures were evaluated before these representative experiments were selected for presentation.