This Article Abstract Full Text (PDF) All Versions of this Article: 138/2/715    most recent pp.105.063289v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (20) Citing Articles via Google Scholar Google Scholar Articles by Baldwin, A. Articles by Jarvis, P. Search for Related Content PubMed PubMed Citation Articles by Baldwin, A. Articles by Jarvis, P. Agricola Articles by Baldwin, A. Articles by Jarvis, P.

CELL BIOLOGY AND SIGNAL TRANSDUCTION

A Molecular-Genetic Study of the Arabidopsis Toc75 Gene Family¹

Department of Biology, University of Leicester, Leicester LE1 7RH, United Kingdom (A.B., A.W., R.P., P.D., S.K.P., D.T., P.J.); and Department of Plant Sciences, University of California, Davis, California 95616 (K.I.)

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
Toc75 (translocon at the outer envelope membrane of chloroplasts, 75 kD) is the protein translocation channel at the outer envelope membrane of plastids and was first identified in pea (Pisum sativum) using biochemical approaches. The Arabidopsis (Arabidopsis thaliana) genome contains three Toc75-related sequences, termed atTOC75-I, atTOC75-III, and atTOC75-IV, which we studied using a range of molecular, genetic, and biochemical techniques. Expression of atTOC75-III is strongly regulated and at its highest level in young, rapidly expanding tissues. By contrast, atTOC75-IV is expressed uniformly throughout development and at a much lower level than atTOC75-III. The third sequence, atTOC75-I, is a pseudogene that is not expressed due to a gypsy/Ty3 transposon insertion in exon 1, and numerous nonsense, frame-shift, and splice-junction mutations. The expressed genes, atTOC75-III and atTOC75-IV, both encode integral envelope membrane proteins. Unlike atToc75-III, the smaller atToc75-IV protein is not processed upon targeting to the envelope, and its insertion does not require ATP at high concentrations. The atTOC75-III gene is essential for viability, since homozygous atToc75-III knockout mutants (termed toc75-III) could not be identified, and aborted seeds were observed at a frequency of approximately 25% in the siliques of self-pollinated toc75-III heterozygotes. Homozygous toc75-III embryos were found to abort at the two-cell stage. Homozygous atToc75-IV knockout plants (termed toc75-IV) displayed no obvious visible phenotypes. However, structural abnormalities were observed in the etioplasts of toc75-IV seedlings and atTOC75-IV overexpressing lines, and toc75-IV plants were less efficient at deetiolation than wild type. These results suggest some role for atToc75-IV during growth in the dark.

Components of the Toc complex include Toc34, Toc75, and Toc159, which were first identified in pea (Pisum sativum) using biochemical approaches (Hirsch et al., 1994; Perry and Keegstra, 1994; Schnell et al., 1994; Seedorf et al., 1995; Tranel et al., 1995). These three proteins make up the core Toc complex, which was recently characterized with respect to structure and component stoichiometry (Schleiff et al., 2003b). Toc34 and Toc159 are related GTPases that share considerable sequence identity within their GTP-binding domains (Hirsch et al., 1994; Kessler et al., 1994; Seedorf et al., 1995) and which mediate preprotein recognition at the chloroplast surface (Perry and Keegstra, 1994; Kouranov and Schnell, 1997; Sveshnikova et al., 2000b; Becker et al., 2004). Following recognition by Toc159 and Toc34, precursor proteins are transferred to Toc75, the protein that forms the outer envelope translocation channel (Schnell et al., 1994; Tranel et al., 1995; Hinnah et al., 2002). Simultaneous translocation through the Toc and Tic complexes, in an energy-consuming process, subsequently occurs (Keegstra and Cline, 1999; Chen et al., 2000; Hiltbrunner et al., 2001a; Jarvis and Soll, 2001; Jarvis and Robinson, 2004).

Structural and electrophysiological studies on heterologously expressed pea Toc75 (psToc75) reconstituted into liposomes demonstrated that the protein forms a -barrel channel with a pore size of approximately 14 to 26 Å (Hinnah et al., 2002). Thus, preproteins must have a completely or partially unfolded conformation to pass through the pore. Topological studies on psToc75 employing proteolytic digestion, amino acid sequencing, hydrophobicity, and computer modeling data suggested that Toc75 possesses either 16 -strands (Sveshnikova et al., 2000a) or 18 -strands (Schleiff et al., 2003a). Interestingly, it seems that the role of Toc75 is not confined to preprotein conductance. Electrophysiology measurements on reconstituted Toc75 revealed that the protein is able to specifically recognize a transit peptide without the aid of other Toc components (Hinnah et al., 2002).

Evidence that Toc75-related proteins are present in plant species other than pea was provided by Summer and Cline (1999), who used an anti-psToc75 antibody to identify a 75-kD protein in chromoplasts from red bell pepper. Since then, Toc75 homologs have been identified in many different plants, including both monocotyledonous and dicotyledonous species (Dávila-Aponte et al., 2003). Toc75 can also be found in different plastid types and throughout plant development (Tranel et al., 1995; Summer and Cline, 1999; Dávila-Aponte et al., 2003). In pea, Toc75 is expressed at a high level throughout the photosynthetic tissues and at a lower level in stems and roots (Tranel et al., 1995). Thus, Toc75 is thought to be a constitutive component of the protein import pathway in diverse plastid types and in all higher plant species.

The sequencing of the Arabidopsis (Arabidopsis thaliana) genome has revealed that this species contains multiple, homologous isoforms of many of the Toc and Tic components originally identified in pea (Jackson-Constan and Keegstra, 2001). For example, the Arabidopsis genome encodes two Toc34 homologs, termed atToc34 and atToc33 (Jarvis et al., 1998; Gutensohn et al., 2000), and four Toc159 homologs, termed atToc159, atToc132, atToc120, and atToc90 (Bauer et al., 2000; Hiltbrunner et al., 2001a; Ivanova et al., 2004; Kubis et al., 2004). Reverse-genetic strategies have been used to examine the roles of these different isoforms, and it appears that they do exhibit some degree of functional specialization. In the case of the Toc34 family, the atToc33 isoform is preferentially involved in the import of photosynthetic precursors, while the atToc34 isoform is most likely involved in the import of nonphotosynthetic precursors (Gutensohn et al., 2000; Kubis et al., 2003; Constan et al., 2004). Interestingly, there are three different Toc75-related sequences in the Arabidopsis genome, termed atTOC75-I, atTOC75-III, and atTOC75-IV according to their chromosomal locations (Jackson-Constan and Keegstra, 2001). In this study, we used molecular, genetic, and biochemical techniques to investigate the functions of these three sequences. The much more divergent gene, atTOC75-V/AtOEP80, was not included in our analysis, since it does not appear to be a true member of the Toc75 family (Eckart et al., 2002; Inoue and Potter, 2004).

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Expression and Structure of the Different Arabidopsis Toc75 Gene Family Members

The sequencing of the Arabidopsis genome has enabled the identification, in silico, of putative homologs of proteins originally identified in other species. For example, it was previously reported that there are three psToc75-related sequences in the Arabidopsis genome: atTOC75-I, atTOC75-III, and atTOC75-IV (Jackson-Constan and Keegstra, 2001). The atTOC75-III gene is known to be expressed since there are currently 24 different atTOC75-III expressed sequence tags (ESTs) present in the databases. Conversely, there are no atTOC75-I or atTOC75-IV ESTs in the databases (although a partial cDNA sequence was recently released for atTOC75-IV; accession no. BX827624). To assess expression of the putative atTOC75-I gene (no. At1g35860), RNA isolated from whole, wild-type Arabidopsis seedlings was analyzed extensively by reverse transcription (RT)-PCR, using six different primer combinations (Fig. 1A). In each case, no evidence for atTOC75-I gene expression could be detected (data not shown).

View larger version (49K): [in this window] [in a new window]   Figure 1. Structural characteristics of the Arabidopsis Toc75 gene family. A, Schematic diagrams depicting the three Arabidopsis Toc75-related sequences. Protein-coding exons are represented by black boxes, and untranslated regions are represented by white boxes; introns are represented by thin lines between the boxes. Mutations in the atTOC75-I pseudogene are indicated as follows: F, frame-shift; S, nonsense; J, splice junction; I, missing intron. A 75-bp insertion and 48-bp direct repeat, both in exon 1, are represented by a black bar and two black arrows, respectively. Locations of forward (priF1, priF2, and priF3) and reverse (priR1 and priR2) primers used for RT-PCR analysis of atTOC75-I are shown. The gray diagram underneath atTOC75-I illustrates how the pseudogene is currently annotated as two separate genes (At1g35880 and At1g35860) in the GenBank database; vertical, dotted lines indicate four putative splice junctions that are identical in the two gene models. The locations of T-DNA and transposon insertions are indicated precisely, but the insertion sizes are not to scale. ATG, Translation initiation codon; Stop, translation termination codon; p(A), polyadenylation site; LB, T-DNA left border. B, Amino acid sequence alignment of pea Toc75 (ps) with the three Arabidopsis proteins (at-III, at-I, and at-IV). The atToc75-I sequence was derived by conceptual translation of a partially corrected version of the open reading frame illustrated in A. The locations of frame-shift (F) and nonsense (S) mutations in the atToc75-I sequence, a 25-residue insertion (IN), and a 16-residue direct repeat (DR) are indicated. Residues identical in at least three sequences are highlighted in black, whereas similar residues are highlighted in gray. The location of the first Asp residue of mature psToc75 is shown (Mat), as are the 16 predicted transmembrane domains of psToc75 (Tranel et al., 1995; Sveshnikova et al., 2000a). C, Phylogenetic analysis of Toc75-related proteins from Arabidopsis and other species. Full length pea, Arabidopsis, and cyanobacterial amino acid sequences were aligned, together with sequences from maize and rice and used to produce a phylogenetic tree. Numbers of mutations are given above the clades with bootstrap values below. Gene and accession numbers for the sequences used are as follows: psToc75 (L36858, S55344); atToc75-III (At3g46740, AAM83239; atToc75-I (BK005428); atToc75-IV (AY585655, AAT08975; atToc75-V/AtOEP80 (At5g19620, NP_568378); zmToc75 (AY106148); osToc75 (AK070010); SynToc75 (slr1227, NP_440832). Species of origin is indicated as follows: pea, ps; Arabidopsis, at; maize, zm; rice, os; Synechocystis PCC 6803, Syn. The cyanobacterial protein, SynToc75, was used as the outgroup.

Next, we investigated the activity and structure of the atTOC75-IV gene (no. At4g09080). Expression analysis by RT-PCR provided clear evidence that the gene is active (Fig. 2A), and the nucleotide sequence of the PCR product obtained revealed that atTOC75-IV transcripts undergo accurate intron splicing (data not shown). Since no publicly available EST or cDNA clones were available for atTOC75-IV when we started these studies, we initiated experiments to identify our own atTOC75-IV cDNA clone. Attempts to identify such a clone by library screening were unsuccessful, presumably due to the low level of expression of the gene, so we instead generated a full-length clone using a combination of RACE-PCR and RT-PCR. The structure of the transcript at its 5' and 3' ends was determined by RACE-PCR, and this information was used to design primers for amplification of a full-length cDNA by RT-PCR. The RT-PCR product was cloned and sequenced (information deposited in GenBank; accession nos. AY585655 and AAT08975), and its structure is illustrated in Figure 1A. This experimentally determined structure differs slightly from the predicted structure given as part of the genome sequence annotation (accession no. NM_116977; Inoue and Potter, 2004).

View larger version (19K): [in this window] [in a new window]   Figure 2. Expression studies on the Arabidopsis Toc75 genes. A and B, Analysis of atTOC75-IV expression in mutant and transgenic lines by RT-PCR. Total RNA samples isolated from the toc75-IV-1 mutant (A) and from two independent 35S-atTOC75-IV transgenic lines (B) were analyzed by RT-PCR, along with corresponding wild-type samples, using primers specific for the indicated genes (atTOC75-IV, atTOC33, and eIF4E1). In each case, the rosette leaves of a single mature plant were analyzed; the 35S-atTOC75-IV plants used were both homozygous for the transgene. For the experiment shown in A, PCR amplification was conducted over a total of 40 cycles. For the experiment shown in B, amplification was conducted using only 25 cycles. PCR products were resolved by agarose gel electrophoresis and stained with ethidium bromide. C, Expression profiles of atTOC75-III and atTOC75-IV during seedling development and in different tissues of Arabidopsis. Total RNA samples isolated from Arabidopsis tissues were analyzed by semiquantitative RT-PCR. RNA was isolated from wild-type seedlings grown in vitro for 5 or 10 d in the light (5 d L and 10 d L, respectively), or 5 d in the dark (5 d D), and from three different tissues of 28-d-old wild-type plants grown on soil (flower buds, rosette leaves, and roots). Amplifications were conducted under nonsaturating conditions using gene-specific atTOC75-III, atTOC75-IV, and eIF4E1 primers, and the products were quantified by hybridization with corresponding 32P-labeled cDNA probes. The data for atTOC75-III and atTOC75-IV were normalized for eIF4E1 and then expressed as a percentage of the maximum level observed for each gene. Values shown are means (±SD) of four independent measurements.

The structures of the three Arabidopsis Toc75-related sequences, shown schematically in Figure 1A, are remarkably similar. All three genes have essentially the same intron-exon boundaries, with only two exceptions: exon 1 is absent from atTOC75-IV, and the penultimate intron is absent from the atTOC75-I pseudogene. The three Arabidopsis homologs are also very similar to each other, and to psToc75, at the amino acid level (Fig. 1B). Percentage sequence identities shared between the Arabidopsis proteins and the pea protein are as follows: 73% (atToc75-III), 60% (atToc75-IV), and 55% (atToc75-I). The 5' truncation of the atTOC75-IV gene means that the atToc75-IV protein includes only eight of the 16 transmembrane domains (nos. 7 and 10–16; Fig. 1B) predicted to be present in psToc75 (Sveshnikova et al., 2000a).

Phylogenetic Analysis

To determine the evolutionary relationships between the different Arabidopsis Toc75 homologs, we conducted a phylogenetic analysis using amino acid sequences from several different species. In this analysis, we included all three Arabidopsis sequences, the psToc75 sequence, two Toc75 sequences from monocotyledonous species, as well as distantly related sequences from Synechocystis PCC 6803 (SynToc75) and Arabidopsis (atToc75-V/AtOEP80); SynToc75 is a cyanobacterial protein that shares only approximately 21% to 22% sequence identity with psToc75 (Bölter et al., 1998; Reumann et al., 1999), and atToc75-V/AtOEP80 shares limited sequence similarity with canonical Toc75 proteins and SynToc75 (Eckart et al., 2002; Inoue and Potter, 2004). Our analysis produced a single most parsimonious tree of length 1,872, containing three main clades (Fig. 1C). The outgroup containing SynToc75 and atToc75-V/AtOEP80 comprises 362 mutations that are present in both proteins but not in Toc75 proteins from the other two clades. With the exception of atToc75-V/AtOEP80, the Toc75 proteins from dicotyledonous species were grouped together in the same clade. Toc75 proteins from the two monocotyledonous species, maize and rice (Oryza sativa), formed a separate clade with 89 distinct mutations. The Arabidopsis Toc75 proteins formed a separate subclade, excluding psToc75, which was supported by a bootstrap value of 57%. Applying a cutoff value of 50%, all clades were supported by bootstrap analysis.

The phylogeny suggests that the Arabidopsis Toc75 gene family formed after the pea and Arabidopsis species diverged and implies that pea and other species do not necessarily contain genes closely related to either atTOC75-IV or atTOC75-I. After speciation, an Arabidopsis TOC75 progenitor duplicated to form atTOC75-III and another TOC75 gene. Duplication of the latter then created progenitors of atTOC75-I and atTOC75-IV. Inactivation of atTOC75-I (due to the insertion of the gypsy/Ty3 retrotransposon or the accumulation of other mutations), and the truncation of atTOC75-IV, may have occurred after the final duplication event. Alternatively, retrotransposon insertion may have occurred prior to the duplication event that gave rise to atTOC75-I and atTOC75-IV.

Expression Profiles of the Arabidopsis Toc75 Genes

Since the members of other Arabidopsis Toc component gene families have been shown to exhibit differential expression patterns and some degree of functional specialization (Bauer et al., 2000; Gutensohn et al., 2000; Kubis et al., 2003; Constan et al., 2004; Ivanova et al., 2004; Kubis et al., 2004), it seemed possible that similar specialization might exist within the Arabidopsis Toc75 family. In an initial attempt to gain insight into the roles of the two expressed Arabidopsis Toc75 genes, atTOC75-III and atTOC75-IV, we analyzed their expression patterns using semiquantitative RT-PCR. The use of RT-PCR was necessary since the atTOC75-IV transcript could not be detected by RNA gel-blot analysis (data not shown). Expression data for the Arabidopsis Toc75 genes were normalized using similar data for a uniformly expressed control gene (translation elongation factor, eIF4E1; Rodriguez et al., 1998) and then expressed as a percentage of the maximum value obtained in each case (Fig. 2C).

The data shown in Figure 2C do not provide any information on the relative levels of expression of the two Toc75 genes. However, it is clear that atTOC75-III is expressed at much higher levels than atTOC75-IV for several reasons. First of all, amplification of comparable amounts of the atTOC75-III and atTOC75-IV transcripts by RT-PCR required 20 and 25 cycles, respectively (Fig. 2; data not shown). Assuming that each cycle of amplification produces a 2-fold increase in product yield, this difference in cycle number equates to a 32-fold difference in template concentration. These data are broadly consistent with recently reported microarray data (Vojta et al., 2004); although atTOC75-IV expression was on the limit of detection in this study, the data indicated a difference in expression between the two genes of approximately 60-fold (Vojta et al., 2004). Finally, while there are 24 different ESTs currently available for the atTOC75-III gene, none is available for atTOC75-IV.

The expression of atTOC75-III was found to be strong in young, rapidly dividing photosynthetic tissues and significantly weaker in mature or slow growing tissues, such as 28-d-old rosette leaves, and roots (Fig. 2C). Unexpectedly, atTOC75-III expression was found to be particularly low in 5-d-old etiolated plants. This suggests that atToc75-IV, the only other close homolog of psToc75 that is expressed in Arabidopsis, may play a relatively more important role in etioplast biogenesis. The expression of atTOC75-IV is more uniform than that of atTOC75-III; while the expression of atTOC75-III varied 10-fold across the samples tested, atTOC75-IV expression varied just 2-fold (Fig. 2C). Like atTOC75-III, the pea Toc75 gene is also expressed most strongly in young, rapidly dividing tissues (Tranel et al., 1995). Additionally, immunoblot data indicated that psToc75 protein accumulates to a lesser degree in roots than in green tissues (Tranel et al., 1995). These various data on the level and pattern of expression of the Toc75 genes, in combination with the structural characteristics discussed earlier, strongly suggest that atTOC75-III encodes the main ortholog of psToc75 in Arabidopsis.

atToc75-IV Is an Integral Protein of the Plastid Envelope

Pea Toc75 is targeted to the chloroplast outer envelope membrane by a bipartite, amino-terminal transit peptide (Tranel et al., 1995; Tranel and Keegstra, 1996). Thus, psToc75 targeting involves two different proteolytic processing steps and a transitional molecule of intermediate size (Tranel et al., 1995; Tranel and Keegstra, 1996). Similarly, atToc75-III is also imported into chloroplasts in a two-step process involving an intermediate (Inoue and Keegstra, 2003; Fig. 3A). Since atTOC75-IV lacks the large first exon present in atTOC75-III (Fig. 1A), which encodes the bipartite transit peptide, we analyzed the atToc75-IV amino acid sequence using the transit peptide prediction program, TargetP (Emanuelsson et al., 2000). TargetP predicts that atToc75-IV does not have a transit peptide, and so it seems likely that atToc75-IV is not targeted in the same manner as atToc75-III and psToc75. It should be noted that it is not uncommon for outer envelope membrane proteins to be inserted into the membrane without the aid of a transit peptide (Schleiff and Klösgen, 2001). In fact, psToc75 and atToc75-III are unique in that they are the only outer envelope proteins known to utilize a transit peptide for insertion.

View larger version (52K): [in this window] [in a new window]   Figure 3. Import and localization studies on atToc75-IV using pea chloroplasts. A, High pH washes of imported proteins. In vitro translated, 35S-labeled precursor proteins (atToc75-IV, atToc75-III, psTic22, and Luciferase, Luc) were incubated with isolated pea chloroplasts for 30 min under import conditions. Intact chloroplasts were reisolated and then treated with either 10 mM HEPES, 10 mM MgCl2, pH 8.0 (bursting buffer), or 0.1 M Na2CO3 (high pH), before separation into pellet (P) and soluble (S) fractions by centrifugation. Samples were resolved by SDS-PAGE and visualized by fluorography. In each case, the first lane contains translation product equivalent to 10% of the amount added to each import reaction. The approximately 36-kD band below the main atToc75-IV translation product is presumably the result of translation initiation at Met-72, since this product was repeatedly observed in atToc75-IV translation mixtures. The precursor (p) and putative intermediate (i) and mature (m) forms of atToc75-III are indicated. B, Trypsin treatment of imported proteins. Import of four different precursors (atToc75-IV, atToc75-III, tp110-110N, and psTic22) was conducted as described in A. Reisolated chloroplasts (Imp) were then treated with three different concentrations of trypsin (1x, 2x, and 5x Trp indicates 0.1, 0.2, and 0.5 g trypsin/g chlorophyll, respectively, equivalent to 12.5, 25, and 62.5 ng trypsin/µL), or in the absence of trypsin (–), for 30 min on ice in the dark. Precursor (p), intermediate (i), and mature (m) forms of the various proteins, and a putative trypsin degradation of product of the psTic22 (*), are indicated. Bands corresponding to the imported proteins, in trypsin-treated and control samples, were quantified using ImageJ version 1.32j (http://rsb.info.nih.gov/ij/), and the data were used to produce the graph shown. Plotted values are means (±SD) derived from two independent experiments. The atToc75-III data were derived from the i and m forms, and the psTic22 data were derived from the p and m forms; it was previously demonstrated that cleavage of the psTic22 transit peptide is not essential for import of psTic22 to the intermembrane space (Kouranov et al., 1999). In each case, both bands exhibited essentially the same degree of trypsin resistance. C and D, Control trypsin treatment experiments. C, The in vitro translated atToc75-IV precursor (atToc75-IV IVT) utilized in A and B was treated in the absence (–) and presence (+) of 1x trypsin (Trp), as described in B, in the absence of isolated chloroplasts. D, Import of atToc75-IV was conducted as described in A and B. Samples were then treated with 2x trypsin, as described in B, in the presence or absence of the 1% (v/v) Triton X-100 (TX100), as indicated.

To further confirm the proper targeting of imported atToc75-IV, and to provide some clues about its suborganellar localization, we conducted postimport protease treatment experiments (Fig. 3, B–D). Four different precursors were imported into isolated pea chloroplasts, and then the chloroplasts were treated with three different concentrations of trypsin, or in the absence of trypsin; trypsin is a protease that is able to penetrate the outer envelope membrane and so gain access to proteins exposed in the intermembrane space, but not the inner envelope membrane (Cline et al., 1984; Jackson et al., 1998; Kouranov et al., 1999). Bands corresponding to imported proteins were quantified, and the data from two independent experiments are shown in Figure 3B. As expected, atToc75-III was almost completely degraded by trypsin (Tranel and Keegstra, 1996; Inoue and Potter, 2004), whereas tp110-N110 (a truncated derivative of psTic110 that was previously shown to insert into the inner envelope membrane; Jackson et al., 1998) was completely trypsin resistant (Fig. 3B). In contrast with atToc75-III, atToc75-IV was substantially resistant to trypsin treatment, to about the same extent as the intermembrane-space-exposed protein, psTic22 (Kouranov et al., 1999; Fig. 3B). Since the atToc75-IV translation product was not inherently trypsin resistant (Fig. 3C), and since the protease tolerance of the imported protein was dependent upon membrane integrity (Fig. 3D), these data again suggest that atToc75-IV was properly targeted to the chloroplasts in these assays. The fact that atToc75-IV was more trypsin-tolerant than atToc75-III may be interpreted in two different ways: either (1) atToc75-IV is present in the outer envelope membrane, like atToc75-III, but takes on a topology (or forms an assembly with other components) that makes it less accessible to trypsin; or (2) atToc75-IV is located inside of the outer envelope membrane, like psTic22 and psTic110. Given that atToc75-IV is an integral membrane protein (Fig. 3A) and yet is not trypsin-tolerant to the same extent as tp110-110N (Fig. 3B), we favor the former possibility.

To determine if the targeting of atToc75-IV occurs in similar fashion in Arabidopsis, we conducted similar import experiments with isolated Arabidopsis chloroplasts. Like atToc75-IV imported into pea chloroplasts (Fig. 3A), atToc75-IV imported into Arabidopsis chloroplasts (as well as endogenous atToc75-III) was exclusively associated with the membranes and completely resistant to extraction under high salt and high pH conditions (data not shown). In an additional experiment, postimport treatment of the chloroplasts with thermolysin (a protease that cannot penetrate the outer envelope membrane; Cline et al., 1984) was conducted. While atToc75-IV was not proteolytically processed during the import assay (Fig. 4A, compare lanes 1 and 2), the imported protein was nevertheless substantially resistant to thermolysin treatment at the end of the assay (Fig. 4A, compare lanes 5 and 6), suggesting that the protein had become integrated into the chloroplast envelope.

View larger version (44K): [in this window] [in a new window]   Figure 4. Import and localization studies on atToc75-IV using Arabidopsis chloroplasts. A, Postimport thermolysin treatment experiment. In vitro translated, 35S-labeled atToc75-IV was imported into chloroplasts isolated from 10-d-old Arabidopsis plants grown in vitro. The first lane contains translation product equivalent to 10% of the amount added to each import reaction, and the second lane contains a standard, 20-min import reaction (Imp). The remaining four lanes show an associated thermolysin treatment experiment: 35S-labeled atToc75-IV was incubated for 20 min under import conditions in the presence/absence of chloroplasts (Chl), as indicated, before further treatment in the presence/absence of 60 µg/mL thermolysin (Th), as indicated. In each case, chloroplasts were recovered and the constituent proteins were resolved by SDS-PAGE and visualized by fluorography. B, Suborganellar fractionation experiment. In vitro translated,35S-labeled atToc75-IV protein was imported into isolated chloroplasts as described in A. Recovered chloroplasts (Chl) were then fractionated by centrifugation to generate thylakoid (Thy), total membrane (Env/Thy), and envelope (Env) fractions. Following resolution by SDS-PAGE, atToc75-IV was visualized by fluorography, and atToc75-III, atTic110, and LHCP were detected by immunoblotting. Each lane contains protein equivalent to one import reaction (i.e. 10 million chloroplasts). C, Immunoblot experiment. Chloroplasts were isolated from 20-d-old wild-type, toc75-IV-1, and transgenic 35S-atTOC75-IV, line 13 (35S-75-IV, 13) plants and then used to prepare envelope membranes as in B. Whole chloroplast (Chl) and envelope membrane (Env) samples (equivalent to 4 µg chlorophyll for the atToc75-IV blots and 2 µg chlorophyll for the control blots) were then analyzed by immunoblotting. Identical blots were probed with antisera derived from two independent rabbits inoculated with an atToc75-IV-specific peptide, and the corresponding preimmune sera (the different rabbits gave similar results and so data are shown for one rabbit only). The antisera specifically detected a protein of 44 kD (the predicted size of atToc75-IV and the apparent molecular mass of in vitro translated atToc75-IV) in Chl and Env samples of the 35S-atTOC75-IV line, as indicated by the arrowhead and label at right. Also indicated are bands corresponding to Rubisco large subunit (RbcL), two prominent nonspecific proteins in the Chl samples (indicated with asterisks), and three prominent nonspecific bands in the Env samples (indicated with bullet points); all of these nonspecific bands were also detected by the preimmune sera. Positions of molecular mass standards (kD) are shown at left. Antibodies that recognize atToc75-III, atTic40, and two thylakoid proteins, the 33-kD and 23-kD subunits of the oxygen evolving complex (OE33 and OE23, respectively), were used as controls.

To confirm that atToc75-IV is also localized to chloroplasts in vivo, we probed whole chloroplast protein samples with antibodies raised against an atToc75-IV-specific peptide. Two independent antisera recognized several different protein bands in wild-type and toc75-IV mutant chloroplasts, but no differences were observed between the genotypes in either case (Fig. 4C, lanes 1 and 3; data not shown). These data, together with the various lines of evidence indicating low atTOC75-IV transcript abundance in wild type (see earlier), suggested that atToc75-IV is a low abundance protein. To facilitate its detection, we also analyzed chloroplast samples from two transgenic lines overexpressing the atTOC75-IV cDNA (Fig. 2B). Chloroplasts from both overexpressor lines contained a protein of approximately 44 kD that was specifically detected by both atToc75-IV antisera, but not by the corresponding preimmune sera (Fig. 4C, lane 5; data not shown). The apparent molecular mass of the detected protein (44 kD) matched exactly the apparent molecular mass of the atToc75-IV in vitro translation product (data not shown) and the predicted molecular mass of the atToc75-IV protein (Fig. 1, A and B). The antibodies also detected an approximately 44-kD protein in envelope membranes isolated from the overexpressor line chloroplasts (Fig. 4C, lane 6; data not shown), which implies that the endogenous protein is localized in similar fashion to in vitro translated atToc75-IV (Fig. 4B). These data therefore support our conclusions drawn from the in vitro import experiments, i.e. that atToc75-IV is a chloroplast envelope membrane protein and that it is targeted without the aid of a cleavable transit peptide.

Envelope Insertion of atToc75-IV Does Not Require ATP at High Concentrations

Most outer envelope membrane proteins are synthesized at their mature size without a cleavable transit peptide (Keegstra and Cline, 1999; Schleiff and Klösgen, 2001). Unlike stromal proteins, these outer membrane proteins do not seem to require high ATP concentrations or thermolysin-sensitive envelope components for targeting to the chloroplast, indicating that they do not utilize the complete Toc/Tic complex for import (Keegstra and Cline, 1999; Schleiff and Klösgen, 2001; Tu et al., 2004). In an attempt to shed light on the atToc75-IV targeting mechanism, we conducted import assays under ATP-limiting conditions (Fig. 5). In vitro translated, ATP-depleted atToc75-IV and Rubisco small subunit precursor (preSSU) were each incubated with ATP-depleted chloroplasts under import conditions (Olsen et al., 1989; Aronsson and Jarvis, 2002). The control protein, preSSU, was not imported and processed by the chloroplasts unless 5 mM ATP was included in the import buffer, confirming that ATP had been successfully depleted from both the chloroplasts and the in vitro translated precursors. By contrast, atToc75-IV localized to the membrane fraction whether or not ATP was included in the import buffer (Fig. 5), demonstrating that the membrane insertion of atToc75-IV does not require ATP at the high concentrations necessary for translocation into the stroma. While these data demonstrate that the targeting of atToc75-IV differs significantly from that of preSSU, they do not reveal a difference between the atToc75-IV and atToc75-III/psToc75 targeting mechanisms, since psToc75 targeting was also shown to proceed at low ATP concentrations (approximately 50 µM; Tranel et al., 1995).

View larger version (18K): [in this window] [in a new window]   Figure 5. Energetics of atToc75-IV membrane insertion. Chloroplasts isolated from 10-d-old Arabidopsis plants grown in vitro were incubated in the dark in the presence of 6 µM nigericin for 10 min to deplete endogenous ATP. In vitro translated, 35S-labeled atToc75-IV and preSSU proteins were passed through Sephadex G-25 to remove small molecules and then incubated with the chloroplasts under import conditions for 20 min in the absence (Imp –ATP) or presence (Imp +ATP) of 5 mM ATP. Chloroplasts recovered from separate import assays were treated for 1 h with 0.1 M Na2CO3 (high pH) or 10 mM HEPES, 10 mM MgCl2, pH 8.0 (bursting buffer), before separation into pellet (P) and soluble (S) fractions by centrifugation. Following resolution by SDS-PAGE, atToc75-IV and preSSU/SSU were visualized by fluorography. Bands corresponding to preSSU (p) and mature SSU (m) are indicated.

To investigate the roles of the atToc75-III and atToc75-IV proteins in vivo, we identified T-DNA or transposon knockout lines for each gene. Two independent alleles of each knockout mutation were identified (Fig. 1A). In each case, the independent alleles behaved identically, indicating that the mutant phenotypes observed were due to the T-DNA or transposon insertions, rather than secondary, unlinked mutations. Unless stated otherwise, the toc75-III-1 and toc75-IV-1 alleles were used to generate the data presented below.

All four knockout lines contained single-site insertions, based on the segregation of the insertion-associated marker genes (Table I). Families derived from heterozygous toc75-IV plants segregated three resistant plants for every one sensitive plant when plated on selective medium (Table I), indicating standard Mendelian inheritance. By contrast, families derived from heterozygous toc75-III-1 plants contained only two resistant plants for every one sensitive plant, and families derived from heterozygous toc75-III-2 plants segregated approximately two heterozygous plants for every one wild-type plant when tested by PCR (Table I). Furthermore, all toc75-III mutant plants tested by PCR (60 plants for toc75-III-1 and 35 plants for toc75-III-2) were found to be hemizygous for the corresponding T-DNA insertion. These data indicate that homozygous atToc75-III knockout mutations, unlike homozygous atToc75-IV mutations, are lethal during an early stage of development.

View larger version (36K): [in this window] [in a new window]   Figure 6. Phenotypic analysis of the toc75-III and toc75-IV mutants. A, Plants grown under long-day conditions in vitro for 6- or 12-d postgermination are shown. The presented genotypes are, from left to right: wild type (WT), toc75-III heterozygote (+/toc75-III), toc75-III toc75-IV double mutant (+/toc75-III; toc75-IV/toc75-IV), toc75-IV homozygote, toc75-IV ppi1 double homozygote, and ppi1 homozygote. B, The visible phenotypes of 12-d-old plants of the indicated genotypes were quantified by making chlorophyll measurements. Sibling analysis was conducted on families segregating for the toc75-III mutation, which were either wild-type (bars 1 and 2) or homozygous mutant (bars 3 and 4) at the atTOC75-IV locus. For bars 1 to 4, measurements were of samples containing two primary leaves of the same plant; the remainder of each plant was used for genotyping by PCR. For bars 5 to 7 (toc75-IV, toc75-IV ppi1, and ppi1), measurements were of samples containing eight primary leaves from different plants. Values shown are means (±SD) derived from multiple independent measurements: wild type (WT; n = 11), +/toc75-III (n = 19), +/toc75-III; toc75-IV/toc75-IV (n = 45), toc75-IV (n = 15), toc75-IV (n = 20), toc75-IV ppi1 (n = 6), ppi1 (n = 6). The data are expressed as a percentage of the wild-type chlorophyll concentration.

The atToc75-III Knockout Mutation Is Embryo Lethal

Since homozygous toc75-III seedlings could not be identified in the progeny of heterozygous toc75-III plants (Table I), we determined the stage of development at which the growth of toc75-III homozygotes terminates. In siliques of heterozygous toc75-III plants, aborted seeds were present at a frequency of approximately 25% (Table II; Fig. 7, K and L). This indicates that homozygous toc75-III mutations are embryo lethal, which is consistent with the notion that atToc75-III is the main Arabidopsis ortholog of psToc75 and with the fact that Toc75 plays a key, channel-forming role in the protein import mechanism.

View larger version (135K): [in this window] [in a new window]   Figure 7. Embryo lethality of the toc75-III mutation. A to J, Morphology of normal (A–E) and mutant (F–J) embryos from the siliques of toc75-III heterozygous plants. The normal developmental stage in each corresponding pair of images is as follows: A and F, four-cell embryo stage; B and G, eight-cell embryo stage; C and H, 16-cell embryo stage; D and I, 32-cell embryo stage. Embryo cell stages refer to the number of cells in the embryo proper. A normal, heart-stage embryo (E) and an aborted seed (J) are also shown. Bars = 10 µm (A–D and F–I) and 50 µm (E and J). K, The appearance of part of a typical silique from a mature toc75-III heterozygous plant. Arrows indicate aborted seeds. Bar = 0.5 mm. L, Higher magnification image of the two central aborted seeds shown in K. Bar = 0.1 mm.

Since homozygous toc75-IV seedlings were found to have no obvious visible phenotypes (Fig. 6), we sought to identify a mutant phenotype by conducting more detailed studies. First, we compared the ability of isolated wild-type and toc75-IV mutant chloroplasts to import two different precursor proteins (preSSU and the 33-kD oxygen evolving complex precursor, preOE33), using published procedures (Aronsson and Jarvis, 2002; Kubis et al., 2003). However, no significant differences were observed (data not shown). Next, we used electron microscopy to analyze the ultrastructure of chloroplasts in the primary leaves of 10-d-old, light-grown plants. Electron micrographs of approximately 20 different plastids each from wild type and toc75-IV-1 were studied and quantified with respect to cross-sectional area and overall shape of the plastid, the number of starch grains, plastoglobuli and granal stacks per plastid, and the number of thylakoid membranes per granum. These data were analyzed statistically, but no significant differences between the wild-type and mutant chloroplasts could be detected (data not shown).

The expression data presented earlier (Fig. 2 and related discussion) suggest that atTOC75-IV may play a relatively more important role in etioplasts. We therefore analyzed the ultrastructure of etioplasts from 3-d-old, dark-grown wild-type and toc75-IV seedlings. While the prolamellar bodies and prothylakoids of toc75-IV etioplasts looked normal, we observed that some etioplasts contained enclosed bodies of cytosol, each apparently surrounded by normal envelope membrane (Fig. 8). We examined a relatively large number of etioplasts from wild-type, toc75-IV-1, and toc75-IV-2 seedlings, scored the number of etioplasts with cytosolic inclusions in each case, and then statistically analyzed the data (Table IV). Although approximately 20% of the wild-type etioplasts contained cytosolic inclusions, both toc75-IV alleles contained significantly more cytosolic inclusions than wild-type etioplasts (Table IV). The data were analyzed using the chi-squared test, and, because there are only two categories of data (one degree of freedom), we applied the Yates' correction, which makes the analysis more stringent. The null hypothesis that there was no significant difference between the wild-type and mutant data sets tested was rejected (critical ²-value = 3.84, P-value = 0.05), and so we conclude that mutant toc75-IV etioplasts contain significantly more cytosolic inclusions than wild-type etioplasts.

View larger version (107K): [in this window] [in a new window]   Figure 8. Ultrastructure of etioplasts in the toc75-IV mutants. Etioplasts from wild-type (A) and toc75-IV-2 mutant (B) plants grown in vitro in the dark for 3 d; the toc75-IV plastid contains a large inclusion (i) of cytosolic material. In C, an enlargement of a typical cytosolic inclusion within a toc75-IV-1 mutant etioplast is shown. Bars = 0.5 µm.

The toc75-IV Mutants Exhibit Inefficient Deetiolation

Expression (Fig. 2) and electron microscopy (Table IV; Fig. 8) data pointed toward a role for atToc75-IV in the etioplasts of dark-grown plants. To corroborate these findings, we carefully examined etiolated toc75-IV mutant plants for evidence of growth and developmental defects. First of all, we simply measured hypocotyl growth in dark-grown plants. However, repeated measurements of plants grown in the dark for either 5, 6, or 9 d (in total, approximately 800 plants of each genotype were measured) revealed no hypocotyl length differences between the wild type, toc75-IV-1, and toc75-IV-2 (data not shown).

Since the most important function of etioplasts is the initiation of chloroplast development, following exposure to light, we next examined the ability of the toc75-IV mutants to undergo deetiolation. Wild-type and mutant plants were grown in the dark for 6 d and then transferred to continuous white light for a further period of 2 d. At the end of the 2-d light period, the plants were scored for visible evidence of deetiolation. Plants with green, expanded cotyledons were scored as having undergone deetiolation, whereas those with yellow, unexpanded cotyledons were scored as having failed to undergo deetiolation (Fig. 9A). This experiment was repeated 11 times, using multiple fresh batches of seed for each genotype (all of which were of the same age and produced by plants grown under identical conditions), including the same seed batches that were used to make the root and hypocotyl length measurements described earlier. Although the experiment was subject to substantial variability, presumably due to the stochastic nature of the deetiolation response and its acute sensitivity to environmental factors, the results revealed a significant difference between both toc75-IV mutants and the wild type (Fig. 9, B and C). In several individual experiments, a striking difference was observed between the mutants and wild type (e.g. Fig. 9B). In other experiments, this difference was less pronounced. Overall, however, the combined data of all 11 experiments (in total, >2,500 plants of each genotype were scored) revealed an approximately 50% reduction in deetiolation efficiency in the mutants by comparison with wild type (Fig. 9C). The combined toc75-IV-1 and toc75-IV-2 data sets were each statistically significantly different from the wild-type data set, as revealed by a Student's t test (P-values = 0.000016 and 0.0056, respectively); by contrast, the same test indicated that the two mutant data sets were not significantly different from each other (P-value = 0.073). These results demonstrate a physiological role for the atToc75-IV protein, most likely in etioplast biogenesis and/or the etioplast-chloroplast transition. They are also supportive of the ultrastructural data presented in Table IV and Figure 8, as well as of the data indicating that atToc75-IV is a plastidic protein shown in Figures 3 and 4.

View larger version (26K): [in this window] [in a new window]   Figure 9. Deetiolation in the toc75-IV mutants. Wild-type and mutant Arabidopsis plants were grown in vitro in the dark, from germination, for a period of 6 d and then transferred to continuous white light of standard intensity for a further period of 2 d. At the end of the 2-d light period, plants were scored for evidence of deetiolation. This experiment was repeated 11 times. A, Typical plants from an individual experiment that revealed a striking difference in deetiolation efficiency between the toc75-IV mutants and wild type (data from this experiment are shown in B). The insets at right show the cotyledons of representative plants from the main image, at 2-fold higher magnification, and serve to illustrate the presence/absence of a deetiolation response. B, An individual experiment that revealed a striking difference in deetiolation efficiency between the toc75-IV mutants and wild type (representative plants from this experiment are shown in A). The data were derived from four separate petri plates per genotype, each one containing approximately 80 plants. Deetiolation frequencies were calculated on a per plate basis and then used to produce the mean values shown (±SD; n = 4). C, The combined data of all 11 experiments, each of which comprised three or four separate petri plates per genotype. Deetiolation frequencies were calculated on a per experiment basis and then used to produce the mean values shown (±SD; n = 11).

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

It is interesting to note that Schleiff et al. (2003b) observed a minimal number of smaller Toc particles during their studies of the pea Toc complex, since it is possible that these particles incorporate an atToc75-IV-like protein. However, our phylogenetic analysis revealed that the Arabidopsis Toc75 gene family most likely formed subsequent to the speciation of Arabidopsis and pea (Fig. 1C). Thus, pea and other species do not necessarily contain genes related to atTOC75-IV or atTOC75-I. The Arabidopsis genome has undergone extensive duplication and reshuffling, and it has therefore been suggested that Arabidopsis is a degenerate tetraploid (Blanc et al., 2000). While it was not possible to map the three Arabidopsis Toc75-related sequences to the major duplications reported by Blanc et al. (2000), it is nevertheless possible that these or similar processes gave rise to the Arabidopsis Toc75 gene family. Interestingly, atTOC75-I appears to have lost a single intron subsequent to its divergence from the other genes. The precise loss of one or more introns has previously been observed in plants and other organisms (Häger et al., 1996; Drouin and Moniz de Sa, 1997). One possible mechanism involves the RT of processed cellular mRNA to produce a cDNA copy of the gene in question, or of a closely related gene, which can then partially replace the genomic copy through homologous recombination or gene conversion (Derr et al., 1991; Drouin and Moniz de Sa, 1997).

Based on the level and pattern of atTOC75-III expression, and on the high sequence identity shared between atToc75-III and psToc75, we conclude that atToc75-III is the main Arabidopsis ortholog of psToc75. The unique importance of atToc75-III was also demonstrated by the embryo lethality of the homozygous toc75-III mutations (Tables I–III; Fig. 7). As has been reported previously, plastids are essential for embryo development, presumably because they synthesize many important products that are utilized by the rest of the cell, including carbohydrates, fatty acids, amino acids, and terpenoids (Uwer et al., 1998; Apuya et al., 2001). The central importance of plastids, and of the proper targeting of plastid proteins, was recently demonstrated by studies on Arabidopsis mutants lacking other components of the protein import apparatus (Hörmann et al., 2004; Kovacheva et al., 2005). Homozygous knockout mutants for the well-characterized Tic complex component, Tic110, resulted in embryo arrest at the globular stage (i.e. at a much later stage than toc75-III; Kovacheva et al., 2005), and similar mutations affecting a new, putative Tic complex component seemed to arrest development even later at the heart stage (Hörmann et al., 2004). It is interesting that, in a large-scale screen for T-DNA-induced mutations affecting seed development, >25% of the reliably identified loci were predicted to encode plastid-localized proteins (McElver et al., 2001), whereas, by contrast, only approximately 12% of all Arabidopsis nuclear genes are predicted to encode proteins of plastid-related function (Leister, 2003).

Various lines of evidence indicate that the atTOC75-III gene is expressed at much higher levels than atTOC75-IV (see "Results"). In Arabidopsis, the various homologous genes that encode other components of the Toc complex also display differing levels of expression. For example, atTOC159 is expressed at approximately 5- to 10-fold higher levels than its homologs, atTOC132 and atTOC120 (Bauer et al., 2000; Ivanova et al., 2004; Kubis et al., 2004), and maximal expression of atTOC33 is approximately 6-fold higher than that of atTOC34 (Jarvis et al., 1998; Kubis et al., 2003). In each case, the more highly expressed homolog (atTOC159 or atTOC33) has been implicated in the preferential import of photosynthetic precursor proteins (Bauer et al., 2000; Kubis et al., 2003, 2004; Ivanova et al., 2004). However, our results suggest that this type of functional specialization does not exist within the Toc75 family. Undifferentiated proplastids are inherited maternally by the plant zygote and do not develop into chloroplasts until the heart stage when the embryo begins to turn green (Apuya et al., 2001). Thus, the fact that homozygous toc75-III embryos do not develop beyond the two-cell stage (Fig. 7) strongly suggests that atToc75-III is essential for the import of nonphotosynthetic proteins in developing proplastids. The fact that the mutant embryos abort so early also implies that atToc75-III and atToc75-IV share only limited functional redundancy or that the atTOC75-IV gene is not expressed during early embryogenesis.

It is interesting that the toc75-III mutation causes embryo lethality, especially since other individual components of the Toc complex (Toc34- and Toc159-related proteins, which are involved in preprotein recognition) are not essential for embryo development (Jarvis et al., 1998; Bauer et al., 2000; Constan et al., 2004; Ivanova et al., 2004; Kubis et al., 2004). In yeast, while mutants lac