Introduction

The primordium of leaf blades is as a result of the peripheral section in shoot apical meristem (SAM). When the primordia of leaves form, it is necessary to build the proximodistal, mediolateral, and ab/adaxial axes (Hudson 2000). The ab/adaxial factor might be most crucial of all among these axes (Bowman et al. 2002).

Recently, a lot of researches have revealed a few genes that function in building ad-abaxial axes of leaf blades. In this homeodomain/Leu zipper (HD-ZIP) (class III) family, five members function in special adaxiality of lateral organs (Emery et al. 2003; Prigge et al. 2005; McConnell et al. 2001; McConnell and Barton 1998). With miR165 and miR166 targeting the mRNAs of this family genes; in abaxial cells, the down-regulation is determined (Williams et al. 2005; Kim et al. 2005; Bao et al. 2004; Tang et al. 2003). Under the Ler (Landsberg erecta) background, as2 mutant shows that the pattern of adaxial cells is transformed into the pattern of abaxial cells (Xu et al. 2003). However, its overexpressing lines (35S:AS2) exhibit obviously adaxial and radial leaf blades (Xu et al. 2003; Lin et al. 2003). In the AS2/LOB domains, the LOB domain is essential, and it cannot be shifted via that of other members in the family (Matsumura et al. 2009). bop1 bop2 double mutants exhibit that adaxial cell pattern is transformed into abaxial cell pattern (Ha et al. 2007). On the other hand, in KANADI (KAN) and YABBY (YAB) families, some members exhibit abaxial pattern (Eshed et al. 1999; Sawa et al. 1999; Siegfried et al. 1999; Kerstetter et al. 2001; Emery et al. 2003; Eshed et al. 2004; Candela et al. 2008).

Here, we report that ASL38/LBD41 gene is one member of the Arabidopsis LATERAL ORGAN BOUNDARY DOMAIN family, and studies its mode in expression level. Our data indicate that during Arabidopsis leaf growth and development, such expression mode is essential. We constructed sense 35S:ASL38/LBD41 to transform Arabidopsis, and attained 35S:ASL38/LBD41 plants that were both overexpression and cosuppression mutants. By studying these mutants, our findings indicate that ASL38 function is specification of adaxial cell fates. Furthermore, phenotypes observed in asl38-1 mutant suggested that ASL38/LBD41 might play a character of specialization of adaxial cell fate in lateral organs of Arabidopsis.

Materials and methods

Arabidopsis material, constructing vector and genetic transformation

Arabidopsis plants were sown in MS medium for 7 days, and then transferred to green room and grown under 21 ± 2 °C (Meng 2015; Meng et al. 2015).

The ASL38/LBD41 (At3g02550) codes a zone, which is a DNA fragment (214–1005 in the full-length Arabidopsis ASL38/LBD41). This DNA fragment is enlarged derived from 20-day-old plant cDNA of Arabidopsis wild-type (Col-0). And then the DNA was built in the sense direction and seated in downstream of the 35S promoter (Fig. 2a). By sequencing, we confirm the 35S:ASL38/LBD41 construct.

The 35S:ASL38/LBD41 plants were produced via vacuum infiltration of Arabidopsis Col-0 by usualizing GV3101 strain (Meng and Yao 2015). All 35S:ASL38/LBD41 transgenic lines (T2) were confirmed through PCR via usualizing an nptII primer (Meng et al. 2009a, b).

Molecular biological experiments

Total RNA was secluded via usualizing Trizol. To perform RT-PCR research, we synthesized cDNA by 1 μg of total RNA by usualizing an oligo dT 18 primers and MMuLV Reverse Transcriptase. To perform PCR experiments, in each cDNA, 1/4 volume was selected, and usualized them as a template.

To perform ASL38/LBD41 and ASL37/LBD40 PCR amplification, conditions below is used, that is, at 94 °C for 4 min (for denaturation), and then for 40 cycles (for cosupression of ASL38/LBD41) and 33 cycles (for overexpression of ASL38/LBD41) of 94 °C for 45 s, 60 °C for 45 s, 72 °C for 1 min, at last at 72 °C for 7 min. The PCR primers: ASL38/LBD41-F (5′-tcggaaagggtgtagtgag-3′) and ASL38/LBD41-R (5′-aggacgaaggtgattgggac-3′); ASL37/LBD40-F (5′-tacgaaaaggctgcagtgaa-3′) and ASL38/LBD41-R (5′-ggtaccaccacgtgatttcc-3′) (Shuai et al. 2002).

For assaying the expressing change of abaxial polarity genes in 35S:ASL38/LBD41 lines, RT-PCR was implemented, its conditions are below in details: at 94 °C for 4 min (for denaturation), for 35 (for FIL, YAB and PHB) cycles of 94 °C for 40 s, 58 °C for 40 s, and 72 °C for 1 min, and at last, a incubation at 72 °C for 7 min. Gene-specific primers used were described by Lin et al. (2003).

Morphological analyses

We performed morphological experiments, which refer to methods in Fuchs (1963) and Baum and Rost (1996). Also we completed experiments of scanning electron microscopy using Quanta 200 FEG SEM (Alvarez et al. 1992). FIL probe was produced via cDNA and synthesized by using DIG-labeled antisense RNA via usualizing T7 RNA polymerase. FIL probes were produced, which refer to methods in Eshed et al. (2001), and ASL38/LBD41 probes were produced via usualizing ASL38/LBD41 F (5′-tcggaaagggtgtagtgag-3′) and R (5′-aggacgaaggtgattgggac-3′).

Results

Site accumulation of ASL38/LBD41 transcripts during vegetative growth

In the wild-type Columbia ecotype, similar to previous report (Shuai et al. 2002), ASL38/LBD41 gene transcripts can be detected, at variable levels, in all tissues examined, for example, cauline leaves, roots, rosette leaves, inflorescence stems and blossom flowers (Fig. 1a), suggesting ASL38/LBD41 was expressed throughout Arabidopsis tissues. Furthermore, for better detecting ASL38/LBD41 transcripts in Arabidopsis, in situ hybridization of ASL38/LBD41 was constructed. In transverse sections of 10-day-old Arabidopsis (Col-0), relatively weak signals were detected in terms of ASL38/LBD41 transcripts in this zone of the expected primordium on leaf blades (P0) and early primordia (P1 and P2) can be seen. With primordia developed, stronger signals in adaxial region were detected. In general, with leaf blades developed, the relative signals became weaker (Fig. 1b). These data suggest that ASL38/LBD41 might regulate adaxial leaf blade of Arabidopsis.

Fig. 1
figure 1

Expression of some genes in wild-type and 35S: ASL38/LBD41 plants. a In rosette leaves, cauline leaves, roots, inflorescence stems and blossom flowers, ASL38/LBD41 can all be detected, at variable levels. RL rosette leaves, CL cauline leaves, RT roots, IS inflorescence stems, BF blossom flowers. b in situ hybridization with an ASL38/LBD41 probe in wild-type. a–f are serial sections of a vegetative meristem with 10 um interval. Bars 100 um. c ASL38/LBD41 mRNA levels are completely inhibited in sense 35S:ASL38 plants using primers of coding regions, in contrast to those of wild-type (Col-0). RNA was used as a control

Full size image

Identification of 35S:ASL38/LBD41 Arabidopsis seedlings

For explaining the function of ASL38/LBD41, we performed the sense- expressing status of the ASL38/LBD41 in the 35S promoter, which is a strong promoter (Fig. 2b). The 35S:ASL38/LBD41 transgenic plants were made. Using PCR assay, seventy-two transgenic plants containing the nptII reporter gene were proved (Meng et al. 2009a, b). In these identified seedlings, 30 with narrow rosette leaves (Class I) and 22 with radially- symmetric needle-like leaves (Class II) were found. While residual 20 identified seedlings are proved via PCR, they were not obviously diffinct with Col-0 (data not shown). RT-PCR findings in these 52 (52/72) transgenic plants with aberrant phenotypes were performed. In 12 with narrow rosette leaves (Class I) and 9 with radially-symmetric needle-like leaves (Class II) (Fig. 3a–d, g, h; arrow), RT-PCR findings indicated that both transgenic ASL38/LBD41 and endogenous were entirely inhibited, for example, not any expression of endogenous ASL38/LBD41 mRNA compared with Col-0 was observed (Fig. 1c). These findings indicate that in the ASL38/LBD41 plants, the cosuppression is observed. These identified plants (9 + 12/30) of RT-PCR analysis were selected for further analysis. Besides these, 18 with narrow rosette leaves and 13 with radially-symmetric needle-like leaves (Fig. 3l, m; arrow) showed overexpression of ASL38/LBD41 by RT-PCR analysis (Fig. 2a). In this same construct, independent transgenic seedlings always fluctuate >100-fold in transgene expression level and cosupression frequently evokes gene silencing (Chen et al. 2007; Holtorf et al. 1995; Jones et al. 1985). Thus, it was not surprised that both cosupression and overexpression were triggered in 35S:ASL38/LBD41 seedlings.

Fig. 2
figure 2

Expression of some genes in 35S: ASL38/LBD41 plants. a ASL38 expression levels are enhanced in overexpressing 35S:ASL38 plants, in contrast to those of wild-type. b Scheme of the sense 35S:ASL38/LBD constructs. The vector used for the introduction of the 793-bp Arabidopsis 35:ASL38/LBD41cDNA sense orientation into Arabidopsis. The sense ASL38/LBD41cDNA insert is flanked by the cauliflower mosaic virus 35S promoter at the 5′ end and by the transcriptional terminator at the 3′ end. LB left border, RB right border, nptII neomycin phosphotransferase gene whose expression confers plant resistance to kanamycin. c In overexpression of ASL38/LBD41, expression of FIL and YAB3 genes is decreased, whereas expression of PHB genes is enhanced, in contrast to that of wild-type (Col-0). d ASL37 expression leaves of cosupressing 35S:ASL38/LBD41 plants were compared with those of wild-type. e 35S:ASL38/LBD41 cosupression enhances expression of FIL and YAB3 genes, whereas it decreases expression of PHB genes, in contrast to those of wild-type (Col-0). Levels of RNA of genes above are determined by RT-PCR analysis using 1 ug of total RNA from each sample. RNA was used as a control

Full size image
Fig. 3
figure 3

Phenotypes of 35S:ASL38/LBD41 plants and asl38-1 mutants. a, d A 14-day-old cosupressing 35S:ASL38/LBD41 plant with a narrow rosette leaf. Arrow indicates a narrow rosette leaf. b, c A cosupressing 35S:ASL38/LBD41 narrow rosette leaf grown for 14 and 30-day-old, respectively. Arrow indicates a narrow rosette leaf. e Vascular-pattern of 35S:ASL38/LBD41 leaf. f Arabidopsis thaliana ecotype Columbia wild-type. g Cosupressing 35S:ASL38/LBD41 plant with the radial symmetric needle-like leaf. h The radial symmetric needle-like leaf (box) of g shows a higher magnification. i Cauline leaves of cosupressing 35S:ASL38/LBD41 plants curled downward. j Vascular-pattern of wild-type leaf. k Wild-type grown in flowerpot. l Overexpressing 35S:ASL38 plant with the radial symmetric needle-like leaf. m Overexpressing 35S:ASL38 plant with a narrow rosette leaf. Arrow indicates a narrow rosette leaf. n, o asl38-1 mutants with narrow rosette leaves. Arrow indicates a narrow rosette leaf. Bars 1.0 mm

Full size image

Overexpression and cosuppression of ASL38/LBD41 show Ab-adaxial polarity defects

The vasculature of rosette leaves displayed a reticulate pattern (Fig. 3j) in wild-type plants, whereas midvein often disappeared and secondary vein drastically simplified in the cosupressing 35S:ASL38/LBD41 rosette leaves (Fig. 3e). Since the establishment of a vascular system requires the existence of different ab/adaxial axes (Lin et al. 2003), these alternations of the vasculature of the cosupressing 35S:ASL38/LBD41 leaves may result from the loss of polar gene expression. For further explaining the function of ASL38/LBD41, we analyzed the dissecting characteristics of these 35S:ASL38/LBD41 rosette leaves. In rosette leaves of wild-type, an obvious ab-adaxial polarity is visible. Tightly packed, extended palisade mesophyll cells locate at the adaxial side, while irregular or round spongy mesophyll cells, loosely packed, at the abaxial side (Fig. 4a). However, in overexpressing and cosuppressing 35S:ASL38/LBD41 radially-symmetric needle-like leaves (Class II), the polarity was disoriented; as the palisade mesophyll cells were well developed on abaxial side as well as adaxial side in overespressing 35S:ASL38/LBD41 leaf blades (Fig. 4n), and the spongy mesophyll cells were well developed on adaxiality except as on abaxiality in cosuppressing 35S:ASL38/LBD41 leaves (Fig. 4d).

Fig. 4
figure 4

Cross-section of 35S:ASL38/LBD41 plants and asl38-1 mutants. a Cross-section through the rosette leaf blade of a wild-type. b Cross-section through the rosette leaf petiole of a wild- type. c is the magnified views of the regions boxed in b. d, i Cross-section through the radial symmetric needle-like rosette leaves of cosupressing 35S:ASL38/LBD41 plants. h, j are the magnified views of the regions boxed in d and i, respectively. e Cross-section through the narrow rosette leaves of cosupressing 35S:ASL38 plants. f the magnified views of the big box regions in e. g Cross-section through the narrow rosette leaf blade of asl38-1 mutants. k Cross-section through the narrow rosette leaf petiole of asl38-1 mutants. l is the magnified views of the regions boxed in k. m, o Cross-section through the narrow and radial symmetric needle-like rosette leaves of overexpressing 35S:ASL38/LBD41 plant, respectively. n, p The magnified views of the box regions in m, o, respectively; Bars 50 um in a, b, d, e, g, i, k, m and o; Bars 20 um in c, f, h, j, l, n and p. ab abaxial, ad adaxial, p palisade mesophyll, s spongy mesophyll, ph phloem, x xylem

Full size image

In the elementary vascular tracts of Arabidopsis Col-0 leaf blades, ab/adaxial antagonism was obvious via the positioning of distinct tissues. Xylem located adaxially, while phloem positioned abaxially (Fig. 4b, c). In a transverse section of overexpressing and cosuppressing 35S:ASL38/LBD41 rosette leaves, regardless of narrow leaf blades (Class I) or radially-symmetric needle-like leaves (Class II), the primary tracts showed alterable xylem/phloem array, for example, xylems twined by phloems of cosuppressing 35S:ASL38/LBD41 leaves (Fig. 4d, h, e, f, i, j), and phloems situated amid xylems of overexpressing 35S:ASL38/LBD41 leaves (Fig. 4m–p).

Like the internal tissues, polar deficiency was also obvious in the epidermal cells of 35S:ASL38/LBD41 leaf blades. By scanning electron microscopy analysis, the wild-type abaxial epidermis was arranged via a wavelike surface containing no uniform measured cells, which contains long cells (Fig. 5a; arrow), whereas the control (Col-0) adaxial surface of leaf blades was arranged through a flat epidermis containing consistently measured cells (Fig. 5b). In comparison, the abaxial epidermises of overexpressing 35S:ASL38/LBD41 narrow rosette leaves (Class I) reveal a compound of ab/adaxial leaf blades (Fig. 5l). Their adaxial surface was not distinct from that of control (Col-0) leaves (data not shown). Moreover, the epidermis of the radially-symmetric needle-like leaves (class II) reveal adaxial blade traits, as the epidermis containing consistently measured cells can be observed (Fig. 5h). On the contrary, the adaxial epidermises of cosuppressing 35S:ASL38/LBD41 narrow rosette leaves (Class I) reveal an undulating surface resembling those of wild-type abaxial blade; and on this side, long cells were well observed (Fig. 5c, g; arrow). Their abaxial epidermises were not distinct from that of control (Col-0) leaves (data not shown). Moreover, the distal, middle and proximal epidermis in the radially- symmetric needle-like leaves (class II) reveal abaxial blade traits in cosuppressing 35S:ASL38/LBD41 narrow rosette leaves, as a lot of long cells can be observed (Fig. 5i–k; arrow).

Fig. 5
figure 5

Epidermal Surface of 35S:ASL38/LBD41 Plants and asl38-1 mutants. a Abaxial epidermis of a wild-type (wt) rosette leaf showing an undulating surface composed of irregular cell size. Arrow indicates elongated cell. b Adaxial epidermal cells of a wild-type (wt) rosette leaf showing uniform cell size. e, f The adaxial epidermises of the asl38-1 narrow rosette leaf blades resembling a wild-type abaxial epidermis. Arrow indicates elongated cell. c, g The adaxial epidermises of the cosupressing 35S:ASL38/LBD41 narrow rosette leaf blades resembling a wild-type abaxial epidermis. Arrow indicates elongated cell. i–k The adaxial tip, middle and proximal epidermises of the cosupressing 35S:ASL38 radially-symmetric, needle-like leaf (arrowhead) resembling a wild-type abaxial epidermis. Arrow indicates elongated cell. d Abaxial epidermis of a wild-type (wt) rosette leaf with three to four branches. h The epidermises of the overexpressing 35S:ASL38/LBD41 radially-symmetric, needle-like leaf resembling a wild-type adaxial epidermis. l The abaxial epidermises of the overexpressing 35S:ASL38/LBD41 narrow leaf resembling a wild-type adaxial epidermis. Bars 50 um in a, b, f, g, h, j, k, and l; Bars 100.0 um in c–e and i

Full size image

To obtain molecular evidence for the ASL38/LBD41 function, by utilizing RNA in situ hybridization, we detected the expressing traits of the FIL polar gene. In Col-0 seedlings, transcripts of the FIL gene store up in younger leaves and their expressions decrease on the adaxial side with younger leaf growing (Sawa et al. 1999; Siegfried et al. 1999) (Fig. 6a; arrow). In cosuppressing 35S:ASL38/LBD41 plants with narrow (Class I) and needle-like rosette leaves (Class II), FIL expression was entirely detected in the ab/adaxial side of the developing leaf blades and the apex meristem (Fig. 6c, d; arrow), suggesting lateral organ of these 35S:ASL38/LBD41 plants was abaxialized patterning. However, in overexpressing 35S:ASL38/LBD41 plants with narrow (Class I) and needle-like rosette leaves (Class II), FIL was not detected in the ab-adaxial side of the developing leaf and the apex meristem (Fig. 5e, f; arrow), suggesting lateral organ of these 35S:ASL38/LBD41 plants was adaxialized patterning. In general, our data lead to the conclusion that the cosuppressing 35S:ASL38/LBD41 leaf blades closely link to the loss of adaxial polarity, but overexpressing 35S:ASL38/LBD41 leaf blades are closely relative to the loss of abaxial polarity.

Fig. 6
figure 6

RNA in situ hybridization analysis of FIL expression in 35S:ASL38/LBD41 Plants and asl38-1 mutants. In 12-day-old Col (a); asl38-1 mutants (b); cosupressing 35S:ASL38/LBD41 Class I seedlings (c); cosupressing 35S:ASL38 Class II seedlings (d); overexpressing 35S:ASL38 Class I seedlings (e); overexpressing 35S:ASL38 Class II seedlings (f). Arrows indicate expression of FIL in (a). Arrows indicate ectopic expression of FIL in b–d. Bars 100 um

Full size image

Aberrant phenotype of the overexpressing 35S:ASL38/LBD41 trichomes indicate that ASL38/LBD41 is multiple functions. There was a decrease in the divergent section of trichomes on the 35S:ASL38/LBD41 radial needle-like leaves (class II), to two (Fig. 5h) rather than the three to four branches of wild- type (Fig. 5d). Also, on 35S:ASL38/LBD41 leaves, both the size and the number of the trichome support cells were increased, and protruded from the leaf blade surface (Fig. 5h); in contrast to those of wild-type Col-0 (Fig. 5d).

ASL37 expression analysis in cosuppressing 35S:ASL38/LBD41 seedlings

While our data indicated that ASL38/LBD41 was probably involved in ab-adaxil polarity, it is reasonable that our findings may be produced via an artifact through the influence of the outcome of RNAs of ASL38/LBD41, influencing levels of undisclosed genes of the AS2/LOB family, which contains 42 members and unidentified members. In LBDs gene family, ASL37/LBD40 is the most similar to ASL38/LBD41. Since ASL38/LBD41 is highly homologous to ASL37 both in the AS2/LOB domains (58 % identical amino acids) and the C-terminal halves (38 % identical amino acids), it is possible that ASL38 and ASL37 would have the overlapping functions. Thus, RT-PCR analysis was performed for examining ASL37 expression levels in cosuppressing 35S:ASL38/LBD41 plants. Our data indicated that ASL37 expression levels in cosuppressing 35S:ASL38/LBD41 plants were consistent with those in control plants (Fig. 2d), suggesting that these aberrant phenotypes of cosuppressing 35S:ASL38/LBD41 plants were triggered by endogenous RNAs of suppressing ASL38/LBD41 other than those of suppressing ASL37 or ASL38 and ASL37. Previous reports (McHale and Koning 2004; Andersson et al. 2001; Oeller et al. 1991; Hamilton et al. 1990; Ecker and Davis 1986; Rothstein et al. 1987; Smith et al. 1988; Van der Krol et al. 1988) have proved that antisense mRNA and sense has been proved to be a helpful tool for suppressing the expression of particular genes; more importantly, the mRNA level of this requested protein in this gene family can be only altered, on the other hand, other genes in this family keep up no influence (Chen et al. 2007; Andersson et al. 2001; Ganeteg et al. 2001; Zhang et al. 1997).

T-DNA insertions of ASL38/LBD41 expression

In order to further assure that these aberrant phenotypes of 35S:ASL38/LBD41 plants do result from ASL38/LBD41 overexpression and cosuppression, we gained a asl38 mutant of T-DNA insertion, which is from the Salk collection (SALK-090708, T4; asl38-1). In 22 asl38-1 mutants, 5 with narrow rosette leaves were acquired; suggesting that ASL38/LBD41 might be participated in leaf enlargement (Fig. 3n, o). Remaining 17 asl38-1 mutants (17/22) displayed to be analogical leaf enlargement pattern to control (Col-0) plants at either younger or older stages (data not shown). Phenotypes analysis indicated that the polarity was disturbed in the subepidermis of the asl38-1 narrow rosette leaves, which indicates on the both abaxial and adaxial side, there are spongy mesophyll cells (Fig. 4g). Moreover, in arrangement of the primary vascular bundles, partially abaxialized rosette leaves appeared in asl38-1 plants, namely, leafstalks showed vasculature with half-moon-traits; where xylem was observed on the interior and phloem was observed on the external (Fig. 4k, l). This is interpreted as partially abaxialized vasculature (Ha et al. 2007). By scanning electron microscopy analysis, the adaxial epidermises of asl38-1 narrow rosette leaf blades showed an undulating surface resembling those of wild-type abaxial blade; and on this side, long cells were well observed (Fig. 5e, f; arrow). Their abaxial surface was not distinct from that of control (Col-0) leaves (data not shown). Furthermore, in asl38-1 with narrow rosette leaf blades, by utilizing RNA in situ hybridization, we detected the status of the FIL expression. In control (Col-0) seedlings, the FIL transcripts were seen in younger leaves, and when younger leaf grows their expressions were decreased on the adaxial surface (Siegfried et al. 1999; Sawa et al. 1999) (Fig. 6a; arrow), whereas in asl38-1 plants with narrow rosette leaves, FIL expression was entirely detected in the ab-adaxial side of the developing leaves and the apex meristems, which was never found in that of wild-type (Fig. 6b; arrow). At adult stages, asl38-1 plants displayed to be analogical to control (Col-0) seedlings (data not shown). These partially abaxilized phenotypes seen in asl38-1 mutant might be as a result of the redundant function in LBD/ASL family.

Discussion

LOB is considered as playing an important character in building boundary of plant lateral organ development. When the meristem and lateral organs initiate, LOB can establish boundary or communicate signaling (Shuai et al. 2002). LOB is known as a nuclear protein, which participates in regulation of transcriptional level (Husbands et al. 2007). Although we have well understood their some functions, such as AS2 (Xu et al. 2003; Lin et al. 2003), in this family, many genes are not fully studied.

Aberrant phenotypes of some transgenic plants are triggered by ASL38/LBD41 cosupression

Independent transformed plants containing the identical construct always fluctuate >100-fold in transgene expression level and cosupression frequently evokes gene silencing (Chen et al. 2007; Holtorf et al. 1995; Peach and Velten 1991; Jones et al. 1985). In our work, wild-type exhibits evident ASL38/LBD41 transcripts by RT-PCR analysis, whereas 21 (9 + 12) 35S:ASL38/LBD41 transgenic plants with similar traits yield no signal in the code regions corresponding to endogenous ASL38/LBD41 mRNA, strongly suggesting that these 35S:ASL38/LBD41 plants are co-suppressed (Chen et al. 2007; Dougherty and Parks 1995; van der Krol et al. 1990; Napoli et al. 1990). Interestingly, the 21 35S:ASL38/LBD41 plants show similar traits that xylems are surrounded by phloems in xylem/phloem arrangements (traits of abaxially lateral organ), suggesting cosuppression, but not overexpression, of sense 35S:ASL38/LBD41. Furthermore, had sense 35S:ASL38/LBD41 in transgenic plants overexpressed, xylem/phloem arrangements would have been different or even reversed (Lin et al. 2003; McConnell and Barton 1998; McConnell et al. 2001). Sense or antisense suppression is a helping instrument for building Arabidopsis plants which knocks out specific protein, more significantly, specific in the sense transgenic plants in which the mRNA expression of the requested protein is influenced, whereas other genes in this family are not influenced (Andersson et al. 2001; Ganeteg et al. 2001; Zhang et al. 1997). Sense cosupression can be induced by divergent mechanism that is, at present, still poorly understood. Chen et al. (2007) thought that RNA silencing was triggered by sense cosupression evoked by siRNAs. The mechanism of ASL38/LBD41 cosupression remains to be elucidated.

ASL38/LBD41 specifies adaxial cell fate

A lot of studies imply there is a link between cosuppressing 35S:ASL38/LBD41 and fate of abaxial cells or overexpressing 35S:ASL38/LBD41 and fate of adaxial cells. Cauline leaves of cosuppressing 35S:ASL38/LBD41 plants curled downward, toward the abaxial side, suggesting that these 35S:ASL38/LBD41 plants may be involved in abaxialized defect. More significantly, blade expansion of rosette leaves of 35S:ASL38/LBD41 overexpression and cosuppression were radically limited. The similar narrow leaf blades have been reported in 35S:AS2 overexpression, curling upward, hence are believed to be adaxialized organs (Lin et al. 2003). Mutations that related to ab-adaxial axes induce decreased blade elognation (Bowman et al. 2002), which agrees with the hypothesis that juxtaposition of ab/adaxial and abaxial regions is needed to leaf growth (Waites and Hudson 1995). In extreme cases, 35S:ASL38/LBD41 rosette leaves exhibit radially-symmetric, needle-like appearance. Similar to these extreme leaves of these 35S:ASL38/LBD41 plants, in the Antirrhinum, mutants of PHAN lead to the radial and symmetric needle-like leaf blades, which are thought to be the results of abaxialization (Waites and Hudson 1995), and loss-of-function mutant of phb-phv-rev triple exhibits a single radial cotyledon and hypocotyl, which are believed to be abaxialized organs (Emery et al. 2003). ae3-1 leaves have a radially-symmetric, needle-like look, which is considered resulting from the defective adaxial identity (Huang et al. 2006). These radial cotyledon, hypocotyl, and rosette leaves all show very similar phenotypes, namely, the radially-symmetric, needle-like traits are always closely related to ab-adaxial polarity, suggesting that the needle-like leaves of 35S:ASL38/LBD41 plants may be the adaxial-abaxial defective organs. The development with needle-like leaves cannot further produce laminae (Xu et al. 2003). Together, added losses with asymmetric growth are kept company via decreased blade expansion (Eshed et al. 2004). This phenomenon accords the classic theory that the build of ab/adaxial axis is necessary to laminae development (Sussex 1954, 1955).

The examination of the internal tissues reveals that adaxial side of the rosette leaves of cosupressing 35S:ASL38/LBD41 plants are spongy mesophyll, which is different from that extended cells of the palisade mesophyll locate at the adaxial surface in wild-type; whereas abaxial side of overexpressing 35S:ASL38/LBD41 rosette leaves are palisade mesophyll. In a vascular-pattern of Arabidopsis leaf blades, ab-adaxial axis is organized via arranging distinct tissues. Xylem is positioned adaxially, whereas phloem abaxially. However, in a transverse section of cosuppressing 35S:ASL38/LBD41 rosette leaf blades, most primary bundles display distinct xylem/phloem arrangements, i.e., they typically show phloem-surrounding-xylem patterns; whereas overexpressing 35S:ASL38/LBD41 rosette leaves reveal xylem-surrounding-phloem patterns. In previous reports, a link has been found, for example, in kan1-kan2 or bop1-bop2 double mutant, phb-phv-rev triple mutant and phan mutant leaf blade growth (Ha et al. 2007; Huang et al. 2006; Eshed et al. 2004; Emery et al. 2003; Waites and Hudson 1995), vascular-pattern presents a pattern of xylem surrounded by phloem, and the vascular-shape deficiencies are always closely relative to the missing of adaxial axis. For another, in adaxial phb-1d/+ leaf blades, amphivasal positioning is observed, in which the tissue of phloem is enclosed via xylem (McConnell and Barton 1998); and in leaf blades of adaxial 35S:AS2 lines, phloem is enclosed via xylem tissue (Lin et al. 2003). The above observations demonstrate that there is a close linkage between the vascular patterning and ab/adaxial axis, and that the cosuppressing 35S:ASL38/LBD41 leaf blades closely link to the loss of adaxial polarity and overexpressing 35S:ASL38/LBD41 leaf blades closely link to the loss of abaxial polarity. By scanning electron microscopy analysis, the abaxial epidermises of overexpressing 35S:ASL38/LBD41 reveal adaxial blade traits of wild-type, as a even face made of uniformly sized cells can be observed. On the contrary, the adaxial epidermises of cosuppressing 35S:ASL38/LBD41 narrow rosette leaves show an undulating surface resembling those of wild-type abaxial blade; and on this side, long cells were well observed.

Phenotypic analyses of overexpression and cosupression of ASL38/LBD41 reveal that the ASL38/LBD41 is necessary to specify the adaxial position in leaf blades. To better understand if overexpression and cosuppression of ASL38/LBD41 is closely related to ab-adaxial later lateral organs of shoots, the molecular evidence of ASL38/LBD41 function was attained by examining the expressing modes of the FIL via utilizing RNA in situ hybridization. Our data showed that FIL was entirely detected in the abaxial and adaxial surfaces of the developing leaf blades and apex meristems; whereas FIL was not detected in the abaxial and adaxial surfaces of the developing leaves and the apex meristems in overexpressing 35S:ASL38/LBD41 plants, which was never found in that of wild-type.

Author contribution statement

Z. B Wang designed the research. J. P Song and Y. B Wang performed the research. Z. B Wang wrote the article.