The HAT2 gene, a member of the HD-Zip gene family, isolated as an auxin inducible gene by DNA microarray screening, affects auxin response in Arabidopsis


  • Shinichiro Sawa,

    Corresponding author
    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1, Hongo, Tokyo 113-0033, Japan,
      * For correspondence (fax +81 3 5841 4462; e-mail
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Maki Ohgishi,

    1. Laboratory for Genetic Regulatory Systems, Plant Science Center, RIKEN, Yokohama, 230-0045, Japan,
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Hideki Goda,

    1. Laboratory for Growth Regulation, Plant Science Center, RIKEN, Wako, Saitama, 351-0198, Japan, and
    Search for more papers by this author
  • Kanako Higuchi,

    1. Department of Biological Sciences, Graduate School of Science, Tokyo Metropolitan University, Hachioji, Tokyo, 192-0397, Japan
    Search for more papers by this author
  • Yukihisa Shimada,

    1. Laboratory for Growth Regulation, Plant Science Center, RIKEN, Wako, Saitama, 351-0198, Japan, and
    Search for more papers by this author
  • Shigeo Yoshida,

    1. Laboratory for Growth Regulation, Plant Science Center, RIKEN, Wako, Saitama, 351-0198, Japan, and
    Search for more papers by this author
  • Tomokazu Koshiba

    1. Department of Biological Sciences, Graduate School of Science, Tokyo Metropolitan University, Hachioji, Tokyo, 192-0397, Japan
    Search for more papers by this author

* For correspondence (fax +81 3 5841 4462; e-mail


The plant hormone, auxin, regulates many aspects of growth and development. Despite its importance, the molecular mechanisms underlying the action of auxin are largely unknown. To gain a more comprehensive understanding of the primary responses to auxin, we analyzed the expression of genes in Arabidopsis seedlings treated with indole-3-acetic acid (IAA) for 15 min. We identified a single gene that is downregulated early, and 29 genes that are upregulated early. Several types of typical transcription factors are identified as early upregulated genes, suggesting that auxin signals are mediated by a master set of diverse transcriptional regulators. Of the genes that responded to auxin, the expression of the homeobox gene, HAT2, was induced rapidly. Furthermore, we show that the expression of HAT2 is induced by auxin, but not by other phytohormones. To analyze the function of HAT2 in the plant's response to auxin, we generated 35S::HAT2 transgenic plants. These produced long hypocotyls, epinastic cotyledons, long petioles, and small leaves, which are characteristic of the phenotypes of the auxin-overproducing mutants, superroot1 (sur1) and superroot2 (sur2). On the other hand, 35S::HAT2 plants showed reduced lateral root elongation, and reduced auxin sensitivity compared to wild-type plants. Together with the results of RNA blotting and biochemical analyses, these findings suggest that HAT2 plays opposite roles in the shoot and root tissues in regulating auxin-mediated morphogenesis.


Auxin constitutes a class of plant hormones that have profound effects on plant development. Processes governed by auxin, in concert with other plant growth regulators, include the development of vascular tissues, the formation of lateral and adventitious roots, and the control of apical dominance and tropism. These processes are controlled by auxin-mediated changes in cell division, expansion, and differentiation. The molecular mechanisms underlying these processes are still largely unknown. Several studies have described the modulation of gene expression by auxin (Abel and Theologis, 1996; Sitbon and Rechenmann, 1997; Takahashi et al., 1995; Walker and Key, 1982). Immediate cellular responses to auxin treatment include alterations to gene expression; the levels of several mRNA species increase or decrease within minutes to hours after the application of auxin. Differential screening strategies applied to elongating tissues, tobacco protoplasts, and cell suspension cultures have isolated several families of early auxin-responsive genes and their cDNAs, including AUX/IAA, SAUR, and GH3 (Hagen et al., 1984; McClure and Guilfoyle, 1987; Theologis et al., 1985).

Genetic analyses have shown that some genes of the AUX/IAA family actually modulate auxin signal transduction. Partially dominant mutations of the Arabidopsis AUX/IAA genes, AXR3/IAA17, SHY2/IAA3, AXR2/IAA7, IAR2/IAA28, SLR/IAA14, and MSG2/IAA19, produce auxin-related phenotypes, including restrained root growth, reduced lateral root formation, and reduced gravitropism (Fukaki et al., 2002; Leyser et al., 1996; Nagpal et al., 2000; Rogg et al., 2001; Rouse et al., 1998; Tatematsu and Yamamoto, personal communication; Tian and Reed, 1999). To identify genes regulated by auxin in a comprehensive way, we used microarray analysis. We found one downregulated gene and 29 upregulated genes, including genes encoding transcription factors. Among them, the expression of HAT2 was rapidly upregulated by auxin treatment. HAT2 encodes a protein with a homeodomain and a leucine zipper domain. The HAT2 gene is classified as a member of the HD-Zip II subfamily. Another HD-Zip II subfamily gene, ATHB-2, is thought to mediate shade avoidance responses, acting as a negative regulator of gene expression (Ohgishi et al., 2001; Steindler et al., 1999). ATHB-2-induced hypocotyl elongation seems to depend on the auxin transport system. However, the possible functions of ATHB-2 and other HD-Zip II subfamily gene products involved in auxin signaling remain to be explored. To further characterize the function of HAT2 in the response to auxin, we analyzed the molecular structure of the HAT2 gene and the phenotypes of 35S::HAT2 transgenic plants. Our results suggest that HAT2 modulates auxin-mediated morphogenesis.


Identification of early auxin-regulated genes

To identify genes that are transcribed immediately in response to auxin, 7-day-old seedlings were harvested after 15-min treatments with or without 1 µm IAA, and were analyzed using an Affymetrix oligonucleotide array of about 8000 Arabidopsis genes. At least 29 genes were upregulated and one gene was downregulated in response to auxin. The fold change (FC) values, which represent the ratios of the hybridization signals (the average difference values (AvDf)) of mock- and IAA-treated plants, were calculated using Microarray Suite software (Affymetrix). The downregulated and upregulated genes are listed with their corresponding FC values in Table 1. Up- or downregulated genes were defined as displaying a more than twofold difference in FC values either before or after signal amplification. Compilation of genes reproducibly regulated by IAA in three independent experiments yielded a final list of 29 upregulated genes and one downregulated gene (Table 1). These genes were classified into six functional classes based on the results of database analyses: (i) AUX/IAA, SAUR, and GH3 family genes; (ii) signal transduction-related genes; (iii) cell elongation-related genes; (iv) genes putatively involved in the production or degradation of phytohormones; (v) typical transcription factor genes; and (vi) genes of unknown function. Expression data for the 30 auxin-regulated genes that, together with the AUX/IAA family genes, were used to probe the DNA microarray are shown in Table 1.

Table 1.  Auxin early regulated genes
ClassGeneAccessionCoded protein featuresRatioNew/reported
Experiment 1Experiment 2Experiment 3
  1. Upregulated candidates are shown in red and downregulated candidates in blue.

  2. When genes have names on database, we adopted their names. Otherwise, we designated ADR1 (auxin downregulated1) and 15 AUR (auxin upregulated) genes.

1IAA1L15448AUX/IAA family2514.48.9Reported
IAA2AF027157AUX/IAA family43.74.8Reported
IAA3U18406AUX/IAA family3.43.61.6
IAA5U18407AUX/IAA family242432.6Reported
IAA6U18408AUX/IAA family3.53.34.5Reported
IAA7U18409AUX/IAA family−1.1−1.2−1.1
IAA8AC006340AUX/IAA family1.11.81.2
IAA9U18411AUX/IAA family1.11.31.4
IAA10U18412AUX/IAA family1.11.3−1.0
IAA11U18413AUX/IAA family2.91.92.3
IAA12U18414AUX/IAA family1.01.41.5
IAA13U18415AUX/IAA family1.41.32.0
IAA15U18417AUX/IAA family1.71.5−1.2
IAA16U49072AUX/IAA family−
IAA17AF040632AUX/IAA family1.51.6−1.0
IAA18U49074AUX/IAA family−1.2−1.31.1
IAA19U49075AUX/IAA family6.25.52.5New
IAA20U49076AUX/IAA family2.21.6−1.4
IAA26AF088281AUX/IAA family1.01.41.1
IAA28AF149816AUX/IAA family1.1−1.41.1
SAUR AC1S70188SAUR-AC18.36.86.5Reported
AUR1AC006201/T27K22.12SAUR-AC1 like10.64.82.5New
AUR2AL022373/T19K4.240SAUR-AC1 like10.28.02.8New
AUR3AL035601/F6G17.40GH3 like17.318.28.8New
AUR4AC002391/T20D16.20GH3 like13.717.55.2New
2ADR1AL035528/F18A5.290LRR (leucine-rich repeat) motif−2.8−3.3−2.4New
AUR5AC004165/T27E13.22Putative MAPKKK8.49.95.3New
AUR6AL030978/M4I22.90Putative Ca binding protein10.612.47.6New
AUR7AL031032/F17I5.110Putative protein phosphate 2C2.92.62.0New
3XTR6U43488Xyloglucan endotransglycosylase related4.46.64.9New
TCH4AF051338Xyloglucan endotransglycosylase related3.54.55.0Reported
AUR8AL021961/F28A23.90Extensin like2.93.42.1New
4AUR9AL035601/F6G17.20P450 like5.12.53.2New
AtACS-6M92354ACC synthase related5.25.011.0New
5HAT2U09335HD-Zip II subfamily15.515.62.5New
STZX95573Zn finger-like (C2-H2 type)
AUR10AF085279/T07M07.3Zn finger-like (C3-H type)
AUR11AC004683/T19C21.4Zn finger-like (WRKY type)
AUR12AC005896/F3G5.22Zn finger-like (C2-H2 type)
AtMYB50Z95773MYB like (R2R3)
AtMYB77Y14208MYB like (R2R3)
AUR13AC002388/T13E15.15AP2 like (EREBP homolog)
AtERF6AB013301AP2 like4.38.55.1New
6AUR14AC006921/F2H17.17No homology2.92.62.4New
AUR15AC007357/F3F19.2741 a.a. peptide5.75.52.1New

Accuracy of Affymetrix DNA microarray data

To test the reliability of the screening method, expression of the AUX/IAA family genes was examined. Oligonucleotides for the IAA 1–3, IAA 5–20, IAA 24, IAA 26, and IAA 28 genes were used to probe the Arabidopsis genomic array. In our experiment, five of these, IAA1, IAA2, IAA5, IAA6, and IAA19, were identified as genes induced by auxin, 15 min after the application of IAA. IAA3 and IAA11 showed a slight induction (Table 1), whereas the expression of the other IAA family genes (IAA7, IAA8, IAA9, IAA10, IAA12, IAA13, IAA15, IAA16, IAA17, IAA18, IAA20, IAA26, and IAA28) did not show a significant response to IAA treatment after 15 min. Regulation of the expression of genes IAA1–IAA14 by IAA has been reported previously (Abel et al., 1995). The expression of IAA1, IAA2, IAA5, and IAA6 was strongly induced by 15-min applications of IAA, whereas IAA3 and IAA11 were induced only slightly over the same period. Our results are consistent with previously reported data. This indicates that the DNA microarray system used in this work functions correctly and is suitable for identifying auxin-regulated genes.

Multiple transcription genes as the early targets of auxin signaling

The genes of the AUX/IAA family are known to be early auxin-upregulated genes, and probably function as transcriptional regulators specific to plants. Nine putative typical transcription regulators (Table 1) were identified in the present work as early auxin-upregulated genes, encoding various domains such as HD-Zip (HAT2), zinc finger (STZ, AUR10, AUR11, and AUR12), Myb (AtMYB50 and AtMYB77), and AP2 (AsUR13 and AtERF6). Of these genes, HAT2 was most markedly induced, and we further analyzed HAT2 function.

RNA blot analysis of homeobox HAT2 and related genes

To determine the complete gene structure of HAT2, the full-length cDNA was isolated (Figure 1). RNA blot analysis was performed to determine the temporal pattern of HAT2 induction by IAA. Because the HAT2 cDNA sequence is similar to that of other HD-Zip II subfamily genes (Figure 1), especially in the homeodomain and leucine zipper domains, the N-terminal region of HAT2 was used as the probe. Steady-state mRNA levels were measured in intact wild-type seedlings for up to 60 min (Figure 2a). The HAT2 gene responded to exogenous auxin within the first 10 min of IAA treatment, and HAT2 expression continued to increase for 60 min (Figure 2a). The early response gene, IAA2, was monitored together with the HAT2 gene, and its expression also started to increase after 10 min and continued for at least 60 min. We used a long time course experiment to follow the accumulation of HAT2 mRNA between 0.5 and 8 h after auxin treatment (Figure 2b). Maximal transcripts were observed at 2 h after the application of auxin, and were maintained for 8 h. The IAA2 gene showed a similar expression profile to that of HAT2 (Figure 2b).

Figure 1.

The deduced amino acid sequence of HAT2 (GenBank accession number AB067629), and of other HD-Zip II subfamily proteins.

Identical and similar amino acids are shown in red. The three α-helix motifs of the homeodomain are indicated in blue boxes. The leucine zipper motif is shown in green boxes.

Figure 2.

RNA blot analysis of HAT2 and other auxin-related genes.

(a) Induction of HAT2 and IAA2 expression by treatment with auxin for 0, 2, 5, 10, 15, 30, and 60 min. C: control plants were mock-treated for 60 min.

(b) Expression of the HAT2 and IAA2 genes in wild-type plants treated with auxin for 0.5, 1, 2, 4, and 8 h. C2, C8: Control plants were mock-treated for 2 and 8 h, respectively.

(c) HAT2 expression in wild-type seedlings treated with IAA, GA, BR, ABA, or ACC, and in wild-type (WT) and tir1 mutant plants treated with IAA. C: Mock-treated control plants. RNA fractions, separated electrophoretically on agarose gel and stained with ethidium bromide, are shown below the autoradiograms.

RNA blot analysis using total RNA extracted from plants treated with IAA, gibberellic acid (GA), brassinosteroid (BR), abscisic acid (ABA), or 1-amino-cyclopropane-1-carboxylic acid (ACC), was performed to analyze the regulation of HAT2 expression by these phytohormones. Of the phytohormones tested, only IAA controlled the transcription of HAT2 (Figure 2c).

Because it has been suggested that TIR is responsible for auxin signaling (Gray and Estelle, 2000), we also examined the effects of IAA treatment on the levels of HAT2 expression in the tir1 mutant. The induction of HAT2 expression by treatment with IAA was reduced in the tir1 mutant (Figure 2c), suggesting that HAT2 expression was modified by the tir1 mutation. RNA blot analysis was performed to examine the relative abundance of HAT2 mRNA in different plant organs. HAT2 is expressed ubiquitously, with readily detectable levels of mRNA in 4-week-old plants, flowers, stems, rosette leaves, cauline leaves, and roots (Figure 3).

Figure 3.

Preferential expression of the HAT2 gene in plant organs.

Northern blot analysis of HAT2 transcripts in various organs. 4WS, whole seedling 4 weeks after germination; fl, flower; st, stem; rl, rosette leaf; cl, cauline leaf; ro, root. The same RNA fractions, separated electrophoretically on an agarose gel and stained with ethidium bromide, are shown below the autoradiograms.

Different 35S::HAT2 phenotypes in shoot and root tissues

To study the function of HAT2 in relation to its auxin inducible and spatially controlled expression, we generated transgenic plants that overexpress HAT2 under the control of the 35S cauliflower mosaic virus (CaMV) promoter. When the levels of HAT2 mRNA were determined in four of 10 independent transgenic plants, we found that the severity of the phenotype correlated with the level of HAT2 expression (data not shown). Because transgenic line #7 showed the highest degree of enhanced expression (Figure 4a), the phenotype of transgenic line #7 was analyzed further.

Figure 4.

Phenotypes of 35S::HAT2 and auxin-related mutants.

(a) HAT2 expression in four independent 35S::HAT2 transgenic plants of lines #6, #7, #9, and #10 (left), and endogenous HAT2 expression in a 35S::HAT2 transgenic plant of line #7 determined by semiquantitative RT-PCR (right). C, wild type; TUA4, internal control.

(b) A wild-type seedling 4 days after germination (DAG).

(c) A 35S::HAT2 seedling at 4 DAG with epinastic cotyledons and long hypocotyl.

(d) A sur1 seedling at 4 DAG.

(e) A sur2 seedling at 4 DAG.

(f) A wild-type seedling at 10 DAG.

(g) A 35S::HAT2 seedling at 10 DAG with epinastic cotyledons, long hypocotyl, and long petiole.

(h) SEM view of wild-type hypocotyl.

(i) Elongated hypocotyl cells of a 35S::HAT2 transgenic plant.

(j) A wild-type plant at 21 DAG.

(k) A 35S::HAT2 plant at 21 DAG.

(l) Lateral root elongation in a wild-type plant (left) and in a 35S::HAT2 plant (right) at 12 DAG.

Scale bar: 1 mm (b–g); 1 µm (h, i); 1 cm (j, k).

Transgenic 35S::HAT2 plants produced long hypocotyls, epinastic cotyledons, long petioles, and small leaf lobes compared with those of wild-type plants (Figure 4b,c,f,g,j,k; Table 2). Scanning electron microscopy showed that the long hypocotyls of 35S::HAT2 transgenic plants were caused by cell elongation rather than by cell division (Figure 4h,i). These phenotypes are similar to those of auxin-overproducing mutants, such as superroot1 (sur1), superroot2 (sur2) (Figure 4d,e), and yucca (Zhao et al., 2001), and the transgenic plants that overexpress the iaaM auxin-synthesizing gene from Agrobacterium (Romano et al., 1995). The sur1 and sur2 mutants produce many lateral roots that branch not only from the main root, but also from the hypocotyl. The formation of lateral roots is initiated by periclinal and anticlinal divisions in mature pericycle cells of the primary roots and is induced by treatment with auxin (Laskowski et al., 1995). The enhanced lateral root formation of sur1 and sur2 may result from elevated levels of IAA (Boerjan et al., 1995; Delarue et al., 1998). Plants were grown on agar plates set vertically to examine the effects of altered HAT2 gene expression on lateral root formation. Lateral root formation was reduced in 35S::HAT2 plants (Figure 4l; Table 2), and this reduction was not due to a reduction in primordia formation but to limited lateral root elongation. Levels of HAT2 transcripts in the root were also strongly upregulated by the application of IAA (data not shown), which suggests that HAT2 affects lateral root elongation. The mean length of the main roots of wild-type plants at this stage was 7.3 ± 0.3 cm, and that of 35S::HAT2 plants was 2.4 ± 0.3 cm, indicating that the HAT2 gene also regulates the elongation of the main root.

Table 2.  Phenotypes of 35S::HAT2 plantss
Data setsWild type (Col)35S::HAT2
  • *

    Seedlings were grown on the soil in a continuous light for 12 days after germination.

  • Twenty samples were measured.

  • Seedlings were grown on agar plates in a continuous light for 12 days after germination.

Hypocotyl length*, (mm)2.5 ± 0.35.1 ± 1.2
Elongated lateral root number cm−1,2.4 ± 0.40.2 ± 0.3
Leaf number8.5 ± 1.06.1 ± 0.6
IAA content (ng g−1 FW)8.1 ± 0.76.0 ± 0.1

Flowering was initiated in 35S::HAT2 plants at an earlier stage than in wild-type plants. We counted the numbers of rosette leaves after the emergence of the floral buds at the top of the inflorescence axis, as the number of rosette leaves is widely accepted to be the parameter that determines flowering time (Sawa et al., 1999). As shown in Table 2, flowering was initiated earlier in the 35S::HAT2 plants than in the wild-type plants. The adult 35S::HAT2 plants were smaller than the wild-type plants (Figure 4j,k). However, morphogenesis in 35S::HAT2 plants grown in the dark was normal. Gravitropism of shoots and roots were also normal (data not shown).

Elevated HAT2 levels affect the response to auxin

Because 35S::HAT2 plants displayed auxin-related phenotypes, we examined the effects of auxin on the elongation of the main root. When plants were grown on a medium containing 1-naphthaleneacetic acid (NAA), the growth of the main root was strongly inhibited in wild-type plants, whereas the degree of inhibition was less severe in 35S::HAT2 plants (Figure 5). This result indicates that the underground tissues of 35S::HAT2 plants are less sensitive to exogenously applied auxin than are those of wild-type plants. Under our experimental conditions, no significant effects of exogenously applied auxin were observed in above-ground tissues.

Figure 5.

Effects of exogenous auxin on 35S::HAT2 plants.

Inhibition of primary root growth in response to exogenous NAA. White bar, wild type; shaded bar, 35S::HAT2. Seedlings of wild-type and 35S::HAT2 plants were grown for 12 days on various concentrations of NAA on agar medium under continuous light. Twenty specimens were assessed for each experiment.

Molecular phenotypes of 35S::HAT2 transgenic plants

Because many examples of autoregulation have been found in animal homeobox genes and in the Arabidopsis ATHB-2 gene (Ohgishi et al., 2001; Serfling, 1989), we examined the levels of endogenous HAT2 transcripts in 35S::HAT2 transgenic plants. We used a semiquantitative reverse transcription-PCR (RT-PCR) experiment because there is sequence similarity between the HD-Zip II subfamily genes and because probes specific to the non-coding exon regions of the endogenous gene (endogenous gene-specific probes) did not give a clear signal in the Northern analysis. The PCR band corresponding to the 3′ untranslated region (UTR) of the endogenous transcript was only half as intense in the HAT2-overexpressing plants as that in the wild-type plants (Figure 4a and 6), indicating that the HAT2 gene negatively regulates its own expression. We further examined the transcript levels of other HD-Zip II subfamily genes. The PCR bands corresponding to HAT1, HAT3, HAT9, HAT22, ATHB-2, and ATHB-4 transcripts were less intense in the 35S::HAT2 plants than in the wild-type plants (Figure 6).

Figure 6.

Quantification of HD-Zip II subfamily gene transcripts in transgenic plants overexpressing the HAT2 gene.

Quantitative RT-PCR experiment using two pairs of PCR primers in the same reaction mixture: one to amplify the endogenous HAT2 gene (HAT2 3′ UTR), the HAT2 coding region (HAT2 coding), or specific regions of HAT1, HAT3, HAT9, HAT22, ATHB-2, or ATHB-4, and the other to amplify the internal control gene, TUA4. Only the TUA4 control for the HAT2 3′ UTR amplification is shown.

To determine whether HAT2 and other HD-Zip II subfamily genes are target genes of the HAT2 protein, we constructed a HAT2-derived transcription factor (H2-V-G; illustrated in Figure 7(a), consisting of the putative DNA-binding domain of the HAT2 HD-Zip domain, the transactivating domain of the herpes viral protein VP16 (Triezenberg et al., 1988), and the hormone-binding domain of the rat glucocorticoid receptor (GR) (Picard et al., 1988). This system is known to increase the transcript level of the target gene with the application of dexamethasone (DEX) and cycloheximide (CHX) (Ohgishi et al., 2001; Steindler et al., 1999). As shown in Figure 7(b), DEX increased the transcript levels of all the HD-Zip II genes examined (HAT2, HAT1, HAT3, HAT9, HAT22, ATHB-2, and ATHB-4), in both the absence and the presence of CHX, although the basal and induced levels varied with each gene. This suggests that the HAT2 is able to bind to the promoter and regulate the expression of all the HD-Zip II genes tested. CHX treatment also increased both the transcript levels and the magnitude of the upregulation induced by DEX in all HD-Zip II genes examined. On the other hand, DEX treatment or the overexpression of HAT2 did not alter the transcript levels of genes belonging to other subfamilies or the AUX/IAA family, including ATHB-5 (subfamily I), ATHB-14 (subfamily III), ATHB-10 (subfamily IV), IAA2, IAA5, IAA6, IAA7/AXR2, IAA8, IAA14/SLR, IAA19/MSG2, or IAA28/IAR2 (data not shown).

Figure 7.

Analysis of the in vivo target gene of HAT2.

(a) Structure of the H2-V-G gene.

(b) Quantification of HD-Zip subfamily gene transcripts. Lane C indicates RNA templates prepared from non-treated plants, and +CHX, +DEX, and +CHX +DEX indicate RNA templates prepared from plants treated with CHX, DEX, and CHX and DEX, respectively. Only the TUA4 control for the HAT2 3′ UTR amplification is shown.

IAA quantification in 35S::HAT2 plants

Because 35S::HAT2 plants exhibit some auxin-related phenotypes, we measured their IAA content by mass spectrometry (GC-SIM-MS). Transgenic 35S::HAT2 plants were germinated on hormone-free medium, and 10-day-old seedlings were harvested. As shown in Table 2, the mean value for three replicate measurements of free IAA in the 35S::HAT2 plants was 74% that of the wild-type plants. This result indicates that the phenotypes of 35S::HAT2 plants that include long hypocotyls, epinastic cotyledons, long petioles, and small leaf lobes were not caused by increased levels of IAA.


Novel auxin-regulated genes

The expression of hundreds of plant genes is thought to be regulated by auxin via the auxin signal transduction pathway. Recently, auxin-regulated genes have been comprehensively examined using Affymetrix DNA microarrays (Tian et al., 2002). These researchers treated Arabidopsis seedlings 6 days after germination with 20 µm IAA for 2 h, and identified 100 auxin-regulated genes. However, it is difficult to identify the early auxin-regulated genes in this way. Here, we have identified 30 genes for which an application of 1 µm IAA alters the expression levels within 15 min (Table 1). Of these genes, the transcript levels of HAT2 were most markedly induced in classes 2–6.

HAT2 functions differently in auxin signaling in shoot and root tissues

We have shown that HAT2 expression is rapidly and strongly induced by auxin, but not by other phytohormones. Expression of the HAT2 gene was maintained for at least 8 h (Figure 2b), suggesting that HAT2 activity is required continuously in the auxin-mediated response processes for several hours.

Transgenic 35S::HAT2 plants display phenotypes with long hypocotyls, epinastic cotyledons, long petioles, and small leaf lobes, which are typical of the auxin response in above-ground tissues, indicating that the HAT2 gene is probably involved in auxin-mediated morphogenesis in these tissues. We have also shown that the IAA content of 35S::HAT2 plants is lower than that of wild-type plants. This indicates that the phenotypes typical for the auxin response were caused by modulation of the signaling mechanism, and not by the accumulation of IAA.

As for the phenotypes of underground tissues, 35S::HAT2 plants showed a reduced sensitivity to auxin compared to wild-type plants. By using the transgenic plants containing the β-glucronidase (GUS) gene under the control of a region of the HAT2 promoter (a 1661-bp genomic region upstream from the putative initiation codon), we observed that the HAT2 was preferentially expressed in the early stages of lateral root formation, and the GUS expression around the pericycle was maintained throughout the stages of lateral root elongation, although 1.6-kb promoter sequence probably does not contain sequences for auxin induction (Sawa et al., unpublished result). Conversely, it has been suggested that the formation of lateral roots is dependent on auxin transport. Treatment of seedlings with the polar auxin transport inhibitor, N-1-naphthylphthalamic acid, abolishes lateral root formation (Reed and Munday, 1996). Furthermore, overexpression of the ATHB-2 gene, which is homologous to HAT2, is thought to modulate auxin distribution by regulating auxin flux, resulting in the reduced formation of lateral roots (Steindler et al., 1999), a phenotype similar to that displayed by 35S::HAT2 plants. On the other hand, NAA treatment partially rescues lateral root elongation in 35S::HAT2 plants (Sawa et al., unpublished results). This may indicate that HAT2 negatively regulates lateral root elongation by modulating the distribution of auxin from the main root to the lateral root primordia.

T-DNA insertional mutations of the HAT2 gene did not produce any remarkable morphological phenotypes. Furthermore, these mutants responded to gravity, exogenous auxin, and auxin transport inhibitor in similar ways to wild-type plants (Sawa et al., unpublished results), indicating the presence of redundant functions among the HD-Zip II subfamily genes. Genetic analyses using overexpressing and T-DNA insertional mutants of some HD-Zip II subfamily genes, or probable dominant-negative derivatives of HAT2 (e.g. overexpression of the DNA-binding domain) should provide additional evidence for the role of HAT2 in the auxin-response pathway.

Molecular mechanisms of the HAT2 gene

The translational inhibitor CHX induces the expression of auxin-related genes, including some of the AUX/IAA family genes (Abel et al., 1995; Koshiba et al., 1995). Some of the AUX/IAA proteins might be substrates of the AXR1–TIR1 ubiquitin-dependent protein degradation pathway (Gray and Estelle, 2000; Gary et al., 2002). It has been suggested that the AXR1–TIR1 pathway controls the stability of negative regulators of the auxin-response pathway via ubiquitin-mediated protein degradation. Expression of the HAT2 gene is induced by CHX treatment. Furthermore, we have demonstrated that endogenous HAT2 transcript levels are reduced in 35S::HAT2 plants compared with wild-type plants. This indicates that HAT2 expression is negatively regulated by its own protein. In transgenic plants expressing H2-V-G, transcription of all the HD-Zip II subfamily genes examined was upregulated by treatment with DEX, even in the presence of CHX. This indicates that the putative DNA-binding domain of HAT2 directly recognizes the HD-Zip II genes in vivo. Because we assume that HAT2 acts as a transcriptional repressor, this result is consistent with the finding that levels of endogenous HAT2 mRNA and other HD-Zip II transcripts are reduced by the overexpression of the HAT2 gene. Furthermore, the HAT2 induction by auxin treatment was diminished by the tir1 mutation, suggesting that TIR1 is epistatic to the HAT2 in the auxin signaling. HAT2 function may be modulatd by the AXR–TIR1 pathway.

A characteristic of HAT2 function is that it positively regulates auxin signaling in shoot tissues, whereas HAT2 negatively affects auxin signaling in underground tissues. A HD-Zip II subfamily protein, ATHB-2, is thought to act as a dimeric complex that negatively regulates its own expression (Sessa et al., 1993; Steindler et al., 1999). HAT2 may also require dimerization for auxin signaling, and HAT2 may be able to bind to HAT2 and/or other proteins, such as the HD-Zip II subfamily gene products. We presume that different complexes can recognize different cis-acting elements resulting in different responses to auxin signaling. Under this hypothesis, different HD-Zip II subfamily proteins bind to HAT2 protein, to positively and negatively affect auxin signaling in the above-ground and below-ground tissues, respectively. The different roles of HAT2 in aerial and underground tissues are reminiscent of the opposing roles of GL1 in trichome and root hair development (Payne et al., 2000; Wada et al., 1997).

In this study, a number of typical transcription factor genes were identified as auxin inducible. Detailed analysis of the HAT2 gene further suggests an involvement of HAT2 protein in auxin-induced morphogenesis. These findings, together with further analysis of the other genes listed in Table 1, will clarify the molecular mechanisms underlying the action of auxin.

Experimental procedures

Plant material and growth conditions

Arabidopsis wild type (Columbia ecotype) and the transgenic 35S::HAT2 plant seeds were sown on the surface of vermiculite in small pots or on the surface of agar plates with half-strength MS medium, and incubated at 4°C for 3 days. Plants were grown in a laboratory room under continuous illumination of 50–100 µE m−2 s−1 at 22°C. For the DNA microarray analysis, 8 mg of wild-type seeds were sown in 40 ml of half-strength MS liquid medium including 1.5% sucrose (pH 5.8), and grown for 7 days. Samples were harvested after treatment with 0.1% DMSO with or without 1 µm IAA.

DNA microarray analysis

DNA microarray analysis was performed as described (Goda et al., 2002). Total RNA was isolated from seedlings by the acid–guanidinium–phenol–chloroform method (Sambrook et al., 1989) Polyadenylated RNA was purified with the Oligotex-dT30 kit (Roche) and converted into double-stranded cDNA with the use of a Super Script Choice cDNA Synthesis kit (Gibco BRL) and with an oligo(dT)24 primer containing a T7 polymerase promoter site at its 3′ end (Amersham Pharmacia Biotech). Biotin-labeled cRNA was generated from the double-stranded cDNA using the BioArray HighYield RNA Transcript Labeling Kit (ENZO), and was then purified with the use of the RNeasy RNA purification kit (Qiagen). Each cRNA sample (20 µg) was fragmented by incubation for 35 min at 94°C in fragmentation buffer (40 mm Tris-acetate (pH 8.1), 100 mm potassium acetate, 30 mm magnesium acetate). The hybridization mixture comprised of 15 µg of fragmented cRNA in 300 µl of a solution containing 100 mm MES, 1 m[Na+], 20 mm EDTA, 0.01% Tween 20, herring sperm DNA (0.1 mg ml−1), acetylated bovine serum albumin (0.5 mg ml−1), and control cRNA (Eukaryotic Hybridization Control Kit, Affymetrix). Portions of 200 µl of each mixture were subjected to hybridization with the Arabidopsis Genome Array (Affymetrix) for 16 h at 45°C with rotation at 60 r.p.m. Each array was then washed consecutively with non-stringent wash buffer (6X SSPE (Sambrook et al., 1989), 0.01% Tween 20, 0.005% Antifoam) and stringent wash buffer (100 mm MES, 0.1 m[Na+], 0.01% Tween 20). Hybridization complexes were then detected by consecutive exposure to phycoerythrin–streptavidin (Molecular Probes), biotinylated antibodies to streptavidin (Vector), and phycoerythrin–streptavidin, after which each array was washed again with non-stringent wash buffer. All washing and staining procedures were performed with a Fluidics Station 400 (Affymetrix). The array was scanned by a confocal microscope scanner (HP Genome Array Scanner, Affymetrix) at a wavelength of 570 nm.

DNA microarray technology is immature and still has technical disadvantages to be improved, such as a narrow signal dynamic range and a misdetection of cross-hybridization or noise. To solve these problems, we took the following methods. We chose GeneChip system (Affymetrix) as DNA microarray system. Protocols for data analysis of the GeneChip system as well as the issues of sensitivity and quantitation have been described (Lockhart et al., 1996). Briefly, each gene is represented on the array as a set of 16 oligonucleotide probes that match the sequence of the gene exactly. Specificity of hybridization is verified by inclusion on the array of the same set of probes each with a single nucleotide mismatch in the center of its sequence. The difference between the hybridization signal obtained with the matching set of probes and that obtained with the mismatched probes is proportional to the abundance of the corresponding transcript and is calculated as the AvDf value. Analysis of absolute and differential gene expression was performed with GeneChip software, Microarray Suite version 4.0 (Affymetrix). With standard GeneChip protocol, AvDf values of highly expressed genes appear to be affected by signal saturation when using antibody amplification. To achieve higher signal dynamic range, we scanned each chip before and after signal amplification using antistreptavidin antibody. Each chip was normalized relative to the total sums of AvDf values, and then each gene was compared between control and BL-treated samples. Genes that were upregulated or downregulated as reflected by a more than twofold difference in their AvDf values and were assigned to an ‘increase’ or ‘decrease’ in the Difference Call of the comparison analysis by Microarray Suite were identified. Furthermore, genes with an ‘absent’ value in the Absolute Call of baseline data and with a ‘decrease’ value in the Difference Call were excluded from the list. Conversely, genes with an ‘absent’ value in the Absolute Call of experimental data and with an ‘increase’ value in the Difference Call were also excluded from the list. To ensure the reproducibility of results, we performed three independent experiments with different plant samples and genes that showed the same responses in all three experiments were classified as genes regulated by IAA (Table 1).

Northern analysis

Total RNA isolation and RNA gel blot hybridization was performed as described (Oyama et al., 1997). The DNA fragments used as probes were chosen so that they hybridized specifically to each transcript. Their regions were determined with the position of the first nucleotide of each coding sequence is arbitrary set as HAT2: 1–469, IAA2: 268–525.

Hormone and chemical treatment of Arabidopsis plants

Arabidopsis plants were germinated on agar plates and grown for 3 weeks. The plants were then transferred together with the agar onto a plastic mesh with 2–3 mm high spacers attached to the lower side. They were placed in a tall plastic container (PLANTCON®; Flow Laboratories Inc., Mclean, VA, USA) with just enough water to cover the bottom. The top of the plastic container was gradually removed over several days to adapt the plants to open-air conditions and enable them to grow vigorous roots in the water. For chemical treatment, water was replaced with a solution containing the following hormone and chemicals: IAA, 20 µm; GA4, 20 µm; BR, 0.2 µm; ABA, 20 µm; ACC, 200 µm; DEX, 30 µm; CHX, 30 µm each. The treatment was continued for 3 h, after which the plants were harvested.

SEM analysis

The surfaces of the hypocotyls of 35S::HAT2 and wild-type plants were observed using a variable pressure scanning electron microscope, model S-3500 N (Hitachi Ltd), equipped with a cooling stage.

Constructions and transgenic plants

The HAT2 coding sequence was obtained by PCR amplification with primers HAT2–5′ (5′-CCTCTAGAAGATGGGCAAAGAAGATCT-3′) and HAT2–3′ (5′-CCCTCGAGTCACGATCGTGGACGCA-3′) and fused behind the CaMV 35S promoter of the pBI121 binary vector (Clontech). All construct sequences were checked and introduced into the Arabidopsis thaliana by vacuum infiltration. T3 or T4 plants, homozygous with each transgene, were used in the experiments.

The construction of the H2-V-G gene was followed as described (Ohgishi et al., 2001), modifying that we used HAT2 HD-Zip domain instead of the ATHB-2 HD-Zip domain. Transgenic Arabidopsis plants of T3 or T4 plants, homozygous with each transgene, were used in the experiments.

RT-PCR analysis

RT-PCR analysis for the quantification of the endogenous HAT2 transcript in transgenic plants overexpressing a HAT2 transgene was performed according to the instructions for the First-Strand cDNA Synthesis Kit (Amersham Pharmacia Biotech, Piscataway, NJ, USA), using a set of primers specific to the internal control gene, TUA4 (5′-CTTCCTTGACTGCTTCTC-3′, 5′-TCATCGTCACCACCTTCA-3′), and a set of primers specific to the endogenous HAT2 gene (5′-TCAGTTCTACATGCACATGA-3′, 5′-ACATGTTAACAACTACATGC-3′) in the same reaction mixture. An aliquot (10 µl) of each PCR reaction mixture was subjected to electrophoresis through a 4% agarose–ethidium bromide gel. Ethidium bromide intensities in each band were quantitated using an alpha imager 1220 digital imaging system (Alpha Inotech Corp., San Leandro, CA, USA).

Quantification of IAA level

For gas chromatography, single-ion-monitoring mass spectrometry (GC-SIM-MS) analyses, fresh plant material (corresponding to a pool of 10 seedlings grown on agar plates for 10 days in a continuous light condition) was carefully weighed, frozen in liquid nitrogen, and stored at −80°C. For extraction of free IAA, the material was ground in liquid nitrogen using a mortar and pestle. After addition of 0.6 ng of [13C6] IAA (Cambridge Isotope Laboratory, MS, USA) as an internal standard, the material was extracted in 80% acetone and 0.1 mg ml−1 2,6-Di-tret.-butyl-4-methylphenol (BHT) for 60 min. After centrifugation, the supernatant was collected. The pellet was re-extracted again for 90 min, and the supernatant was brought to a water phase in a rotary evaporator. IAA was partially purified from the residual aqueous solution by partitioning, using ether, and then purified by HPLC connecting with a fluorometer (Hitachi FL Detector L-7485; 280 nm excitation, 355 nm emission). The HPLC was performed using a Nucleosil N(CH3)2 column (Senshu, Tokyo, Japan) with an isocratic system of 100% methanol and 0.03% acetic acid. Purified IAA fraction was dried under a nitrogen stream, and trimethylsilylated with N-Methyl-N-trimethylsilyltrifluoroacetamide (MSTFA) at 60°C for 15 min.

Splitless injections were made into a GC-SIM-MS system (QP5050A, Shimazu, Tokyo, Japan), equipped with a capillary column (DB-1, 0.25 mm i.d. ×30 m, film thickness 0.25 µm; J&W Scientific, CA, USA). A linear temperature gradient was applied from 80 to 280°C with an increase of 20°C min−1. The injection temperature of the GC was 250°C, the ion source temperature of the MS was 250°C, and a helium flow of 1.2 ml min−1 was applied. The ionization potential was 70 eV, and the scan time was 0.2 sec. The percentages of molecules of IAA labelled with 13C were calculated from the relative intensities of m/z 202–208, and 319–325 ions after subtraction of background.


We thank Arabidopsis Biological Resource Center at Ohio University and Dr Nam-Hai Chua for providing us some plant materials or GR-related plasmids. We also thank Kensuke Yamazaki and Hajime Sakai for technical assistance and valuable discussions. This work was supported by a grant to Kanagawa Academy of Science and Technology Research Grants, Saneyoshi Scholarship Foundation, Nissan Science Foundation, and Grant-in-Aid to S. Sawa (No. 12740442) from the Japanese Ministry of Education, Science and Culture.

Accession number; HAT2: GenBank accession No. AB067629.