Characterization of differential expression and leader intron function of Arabidopsis atTOC159 homologous genes by transgenic plants

From an engulfed cyanobacterial ancestor, plastid is nowadays converted into three major groups in the plant cell, according to its function and location. Chloroplasts ubiquitously exist in the cells of green tissues, and are responsible for photosynthesis, the synthesis of amino acids and lipids, the synthesis of phytohormones, and the storage of starch and oil compounds. Chromoplasts appear in the cells of leaves, flowers and fruits, and are predominantly responsible for the synthesis of pigments. Amyloplasts majorly exist in the root cells and minor in the stem cells, functioning as the storage of starch and the gravity sensor. However, more than 90% of the plastid genes have since been transferred to the host nucleus during evolution (Martin et al., 2002). In consequence, these nuclear-encoded plastid-localized proteins must be correctly imported to the plastids during the processes of plastid biogenesis to guarantee proper plastid functions.

Except for a few outer membrane proteins, most of the nuclear-encoded plastid proteins are imported into plastids via a set of translocon components located at the outer/inner envelope membrane of chloroplasts (Toc and Tic; Inaba and Schnell, 2008). Toc34 and Toc159 are Toc core-complex components, and also identified as GTPases. They are in charge of recognizing preprotein, and thus are regarded as receptors for the preprotein (Jarvis, 2008). In Arabidopsis, Toc34 and Toc159 are encoded by multiple genes. AtTOC33 (At1g02880) and atTOC34 (At5g05000) are homologous to pea psTOC34 (Jarvis et al., 1998), whereas atTOC159 (At4g02510), atTOC132 (At2g16640), atTOC120 (At3g16620) and atTOC90 (At5g20300) are homologous to psTOC159 (Bauer et al., 2000). Genetic and biochemical analyses revealed that different receptor isoforms encoded by these gene families had distinct functions. AtToc33 and atToc159/atToc90 are probably the receptors specific for photosynthesis-related proteins (Jarvis et al., 1998;Bauer et al., 2000;Kubis et al., 2003;Hiltbrunner et al., 2004). However, atToc34 and atToc132/atToc120 are prone to recognize non-photosynthetic proteins and import them into plastids (Ivanova et al., 2004;Kubis et al., 2004).

Since atToc159 protein isoforms might recognize and allow to import diverse plastid proteins, atTOC159 homologous genes should be independently regulated during the process of vegetative and reproductive development. In fact, the regulation of atTOC159 gene members in several organs had been reported. Bauer et al. (2000) showed that the expression of the mRNA level of atTOC159 was five to ten times more than that of atTOC132 and atTOC120 in both 6-day-old etiolated and green seedling, indicating that atTOC159 is predominantly expressed in young seedling. Furthermore, Kubis et al. (2004) showed that atTOC159 was strongly expressed in 10-day-old leaves, atTOC90 was expressed at a uniformly moderate level in leaves/inflorescence/roots, whereas atTOC132 and atTOC120 were expressed at a uniformly low level but relatively prominent in inflorescence and roots. The expression level of atTOC90 in 28-day-old leaves, inflorescence and roots were approximately 3 times, 1.3 times and 5 times, respectively, higher than that of atTOC159. By contrast, the Gene Chronologer tool in Genevestigator revealed that atTOC159 had 4 to 5 times higher expression than atTOC90 in mature leaves and inflorescence (Zimmermann et al., 2004). Additional analyses are clearly needed to elucidate these conflicting results.

In this study, the differential expression of atTOC159 homologous genes regulated by their upstream sequence in the vegetative and reproductive tissues was analyzed. We first created stable transgenic plants expressing the GUS reporter gene driven by the upstream sequences of atTOC159 family members and determined the GUS activity of these transgenic plants. Our results suggest that atTOC159 and atTOC90 are preferentially expressed in 1-and 2-week-old leaves, whereas atTOC159 and atTOC132 have relative higher expression levels in 21-day-old roots. All atTOC159 homologous genes have increased expression level in flowers and siliques, especially a remarkably higher expression level of atTOC159 in later stage of fruit development. The intron effect on the transcription activity was also examined by analyzing GUS activity in the stable transgenic plants. Our results show that the expression yield of atTOC120 is dramatically up-regulated in root tip and flowers by its leader intron sequence (the intron nearest to the transcription start site), while that of atTOC90 is significantly up-regulated in cotyledons and anthers.

Previous genetic and biochemical studies suggested that atToc159 and atToc90 were important for the accumulation of photosynthetic proteins during chloroplast development. The knockout Arabidopsis mutants of atTOC159 and atTOC90 were named ppi2 (plastid protein import 2, or attoc159) and ppi4 (or attoc90), respectively. The ppi2 mutant resulted in an albino seedling that was not able to grow beyond the cotyledon stage on soil (Bauer et al., 2000). The ppi2 plastids failed to accumulate massive amount of photosynthetic proteins, and therefore lacked thylakoids and starch granules. Nevertheless, the residual amounts of non-photosynthetic proteins in ppi2 plastid were not less than those of wild type plastid, underlining the importance of atToc159 for chloroplast development (Bauer et al., 2000). The ppi4 mutant had no obvious phenotype (Hiltbrunner et al. 2004). However, ppi2/ppi4 double mutant appeared a paler phenotype than ppi2 single mutant, revealing a functional overlap between atToc159 and atToc90 (Hiltbrunner et al. 2004). Nonetheless, overexpression of atTOC90 partially complemented the ppi2 defect, demonstrating atToc90 could only contribute to the import route for some atToc159 client proteins (Infanger et al., 2011). The severity of phenotypic defect could be linked to our transgene results. The steady expression level of atTOC159 in light-grown young leaves is much higher than that of atTOC90 (Figure 2), suggesting that atToc159 plays a more significant roles as the receptor of photosynthetic proteins in leaves and cotyledons.

By contrast, atToc132 and atToc120 seemed to be more specific for recognizing and importing non-photosynthetic proteins (Ivanova et al., 2004;Kubis et al., 2004). AtToc132 (and atToc120) was selectively binding to the transit peptide of preE1α, prepyruvate dehydrogenase E1a subunit representing a non-photosynthetic protein in plastid, rather than binding to the transit peptide of preSSU, presmall subunit of Rubisco representing the photosynthetic protein in chloroplast (Ivanova et al., 2004). The plastids in root of homozygous attoc132/attoc120 double mutant had irregular shape and contained large cytoplasmic inclusions (Kubis et al. 2004). Stanga et al. (2009) recently reported that overexpression of atTOC132 rescued the defect of gravitropism sensing and signaling in arg1 (altered response to gravity 1) and mar2 (modifier of ARG1, identified as atToc132 now) double mutant, revealing atToc132 to involve in gravity signal transduction within the plastid. These studies clearly suggest that atTOC132 and atTOC120 indeed play important roles in regulating root development. From above, atToc132 and atToc120 appear to preferentially affect the import of non-photosynthetic preproteins. However, expression of atTOC120 failed to complement attoc132/attoc120 double mutant, suggesting atToc120 was not sufficient to recognize most of non-photosynthetic proteins. This is probably due to the low expression yield of atTOC120 in Arabidopsis. Our data in this study indicated that the expression level of atTOC132 in leaves and roots was higher than that of atTOC120 (Figure 2), pointing out that atToc132 devotes more in the machinery of non-photosynthetic protein import.

Recent large-scale proteomic studies have raised the question whether it is suitable to categorize atToc159 and atToc132 as major receptors of photosynthetic and non-photosynthetic preproteins, respectively. Based on the comparative analysis of plastid proteome from two attoc159 knockout mutants to wild type plant, many photosynthetic proteins accumulated but several metabolic proteins decreased in attoc159 plastids (Bischof et al., 2011). This suggested that atToc159 was responsible for recognition of both photosynthetic and non-photosynthetic proproteins. Therefore, the authors defined the plastid preproteins as atToc159-dependent and atToc159-independent precursor proteins. Furthermore, the comparative analysis of chloroplast proteome between attoc132 mutant and wild type plant showed that the abundances of four photosynthetic-related and five non-photosynthetic proteins increased 1.5 to 2.5 folds, but five non-photosynthetic proteins decreased approximately 2 folds (Kubis et al., 2004). The accumulation of most chloroplastic proteins remained the similar yield. Theses research work might underestimate the function of atToc90 and atToc120. Except for the pale phenotype of ppi2 (attoc159 mutant), attoc132, attoc120 and ppi4 (attoc90) single mutant did not have any visible phenotype. However, ppi2/ppi4 double mutant appeared a paler phenotype than ppi2 single mutant (Hiltbrunner et al. 2004), and attoc132/attoc120 double mutant showed pale green (Kubis et al., 2004) or were even embryo lethal (Ivanova et al., 2004), revealing a functional overlap between atToc159 and atToc90, or between atToc132 and atToc120. It might be more informative to characterize substrate specificity of these atToc159 isoforms by the plastid proteome from double mutant to single mutant or wild type plants.

Even though atTOC159 gene members tend to differentially express in various vegetative tissues, they all participate in flower and fruit development (Figure 3). Our data suggested that these genes had high expression in flower and early fruit development, and significantly higher expression level in later fruit development (Figure 3). These results were supported by previous genetic studies. For example, the single mutant of each gene is able to produce seeds with normal viability. However, homozygous attoc159/attoc132 and attoc132/attoc120 double mutants cause embryo lethality (Ivanova et al., 2004;and Kubis et al., 2004). Therefore, we conclude that all atToc159 isoforms are not only contributed to the import of photosynthetic/non-photosynthetic related proteins during vegetative growth, but also modulate reproductive growth.

The enhancing effects of gene expression by introns were first reported by (Mascarenhas et al. 1990) and the function was defined as IME. Even though the mechanism of IME remains largely unknown, the effect of IME has been broadly observed in eukaryotes, including animals and plants (Jonsson et al., 1990;Palmiter et al., 1991;Rethmeier et al., 1998;Morello et al., 2002;Jeong et al., 2006;Chen and Sun, 2010). Our stable transformation assay suggested the intron sequence of atTOC132 did not alter its expression yield in cotyledons, root tips, rosette leaves, and flowers (132PUI and 132P in Figure 4). Nevertheless, the intron sequence of atTOC120 and atTOC90 indeed increased the expression yield obviously. The endogenous intron sequence up-regulated atTOC120 expression level specifically in root tips and flowers, but not in cotyledons (120PUI and 120P in Figure 4). In contrast with atTOC120, the intron sequence induced atTOC90 expression level specifically in cotyledons, leaves and anthers, but not in root tips and rest of flower tissues (90PUI and 90P in Figure 4). These findings imply that the intron effect of atTOC120 is more effective in roots and whole flowers, and atTOC90 is in cotyledons, leaves and anthers. We thus hypothesized that these introns might contain tissue-specific enhancers. Additional experiments are clearly required to prove this hypothesis.

The level of GUS activity in 1-or 2-week-old leaves and 3-week-old roots of 120PUI plant was similar to 1391Z control plant (Figure 2). This seemed to imply that 120PUI did not express GUS gene. However, 120PUI plant showed a moderate GUS expression level in the 5-day-old root tips and 2-week-old rosette leaves (Figure 4). These conflict results might be due to the binary vector pCAMBIA1391Z used in this transgenic study. Even though pCAMBIA1391Z contains a promoter-less GUS gene, a nearby cauliflower mosaic virus (CaMV) 35S promoter which drives the expression of hygomycin phosphotransferase II (HptII) gene, could also drive the expression of GUS gene. In consequence, the transgenic plants containing pCAMBIA1391Z show a low to moderate level of GUS expression in various tissues (http://www.cambia.org). Based on this rationale, 120PUI plant indeed expresses the GUS gene in leaves and root tips, and the expression level is similar to 1391Z control plant. Besides, the GUS activity of leaves in 120P plant was lower than 1391Z plant (Figure 4). It is possible that an actively expressed gene can suppress the expression of a close-by gene, especially a weakly expressed gene. For example, unlike most of polyubiquitin (UBQ) genes in Arabidopsis which had an abundantly expressed yield, the expression level of UBQ4 was barely detected in all tissues (Sun and Callis, 1997). Carter et al. (1998) indicated thereafter that UBQ4 expression was suppressed by a strongly expressed nearby gene, germin-like protein 3b (GLP3b), which was approximately 500 bp upstream of UBQ4 transcription start site. Altogether, we conclude that pCAMBIA1391Z vector may not be appropriate to determine the absolute expression level of weakly expressed genes, but still feasible to compare the relative expression level of these genes in transgenic plants.

Due to their recognition of different substrate proteins during protein import into plastids, the closely related atTOC159 homologous genes are independently or coordinatedly regulated in response to developmental signals to ensure the proper import of plastid proteins. Based on the results from stable transgenic plants expressing GUS reporter, we conclude that the upstream regulatory sequences of atTOC159 homologous genes govern the differential expression of these genes among vegetative and reproductive tissues. Furthermore, the endogenous leader intron sequences of atTOC120 and atTOC90 up-regulate gene expression in a tissue-specific manner.