Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2004 Jul;135(3):1336–1345. doi: 10.1104/pp.104.043299

Expression in Multigene Families. Analysis of Chloroplast and Mitochondrial Proteases1

Galit Sinvany-Villalobo 1, Olga Davydov 1, Giora Ben-Ari 1, Adi Zaltsman 1, Alexander Raskind 1, Zach Adam 1,*
PMCID: PMC519052  PMID: 15266057

Abstract

The proteolytic machinery of chloroplasts and mitochondria in Arabidopsis consists primarily of three families of ATP-dependent proteases, Clp, Lon, and FtsH, and one family of ATP-independent proteases, DegP. However, the functional significance of the multiplicity of their genes is not clear. To test whether expression of specific isomers could be differently affected by growth conditions, we analyzed transcript abundance following short-term exposure to different environmental stimuli, using 70-mer oligonucleotide arrays. This analysis revealed variability in the response to high light and different temperatures within members of each family. Thirty out of the 41 tested genes were up-regulated in response to high light, including both chloroplast and mitochondrial isozymes, whereas only six and five genes responded to either high or low temperature, respectively. The extent of response was variable, ranging from 2- to 20-fold increase in the steady-state levels. Absolute transcript levels of the tested genes, compiled from one-channel arrays, were also variable. In general, transcripts encoding mitochondrial isozymes were accumulated to a lower level than chloroplastic ones. Within the FtsH family, transcript abundance of most genes correlated with the severity of mutant phenotypes in the relevant genes. This correlation was also evident at the protein level. Analysis of FtsH isozymes revealed that FtsH2 was the most abundant species, followed by FtsH5 and 8, with FtsH1 being accumulated to only 10% of FtsH2 level. These results suggest that, unlike previous expectations, the relative importance of different chloroplast protease isozymes, evidenced by mutant phenotypes at least in the FtsH family, is determined by their abundance, and not necessarily by different specific functions or specialized expression under certain conditions.

The proteolytic machinery of chloroplasts and mitochondria is essential for controlling the quality and turnover of these organelles' proteins and, thus, is important for their proper function. In Arabidopsis, the proteolytic machinery of chloroplasts consists primarily of three families of ATP-dependent proteases, Clp, Lon, and FtsH, and one family of ATP-independent proteases, DegP (for review, see Adam and Clarke, 2002; Sokolenko et al., 2002). Homologous enzymes are found also in mitochondria (Sarria et al., 1998; Adam et al., 2001; Halperin et al., 2001b; Kolodziejczak et al., 2002). All these families have well-characterized homologs in Escherichia coli (for review, see Clarke, 1999; Adam, 2000). Clp is a Ser protease that separates its two essential functions in two different polypeptides: a small subunit, ClpP, containing the proteolytic active site, and a larger regulatory ATPase subunit, either ClpA or ClpX (for review, see Gottesman, 1996). Lon protease is an ATP-dependent Ser protease in which the catalytic and ATPase domains reside in a single polypeptide (for review, see Gottesman, 1996). FtsH is the only essential ATP-dependent protease in E. coli. It is a membrane-bound metalloprotease, where the proteolytic and ATPase domains also reside in the same polypeptide. DegP is a peripheral membrane Ser protease located in the periplasmic side of the cytoplasmic membrane in E. coli (Skorko-Glonek et al., 1997).

Analysis of prokaryotic and eukaryotic genomes reveals that the number of genes encoding the aforementioned proteases has increased during evolution. For instance, the E. coli genome contains single genes encoding Lon, FtsH, ClpP, A and X, and three DegP-like encoding genes. The photosynthetic cyanobacterium Synechocystis has three ClpP genes, one copy each of ClpR (an inactive homolog of ClpP), ClpC, and ClpX, four FtsHs, and three DegPs (Sokolenko et al., 2002). Genomes of higher plants contain even more copies. The Arabidopsis genome contains 14 DegP, 4 Lon, 12 FtsH, and 20 Clp genes (Adam et al., 2001; Sokolenko et al., 2002; Peltier et al., 2004). With the exception of ClpP1, which is encoded in the plastid genome, all other genes are nuclear. Most of ClpP isomers, as well as their proteolytically inactive homologs ClpR, are found in chloroplasts, where two isomers of ClpC and one ClpD, the plant homologs of ClpA, are also located (Peltier et al., 2001; Zheng et al., 2002). One isozyme, ClpP2, is found in mitochondria, where it may associate with the regulatory subunit ClpX (Halperin et al., 2001b; Peltier et al., 2004). The four Arabidopsis isomers of Lon protease are predicted to reside in either chloroplasts or mitochondria (Sarria et al., 1998; Adam et al., 2001; Sokolenko et al., 2002), as are the different forms of the FtsH protease. FtsH1, 2, and 5 were found as integral proteins in the thylakoid membrane, with their ATP-binding domain and catalytic zinc-binding site facing the stroma (Lindahl et al., 1996; Chen et al., 2000; Sakamoto et al., 2002). FtsH6-9, 11, and 12 are also targeted to chloroplasts, whereas FtsH3, 4, and 10 are mitochondrial (Kolodziejczak et al., 2002; Sakamoto et al., 2003). DegP isomers are predicted to localize in several cell compartments. DegP1 and 2 are peripherally attached to the lumenal and stromal sides of the thylakoid membrane, respectively (Itzhaki et al., 1998; Haussuhl et al., 2001). Proteomic analysis revealed the presence of DegP5 and 8 in the thylakoid lumen (Peltier et al., 2002; Schubert et al., 2002), whereas the remaining isozymes are predicted to reside in mitochondria and various other cellular compartments (Adam et al., 2001; Sokolenko et al., 2002).

Relatively little is known about the specific functions of chloroplast proteases. Clp protease appears to be involved in degradation of unassembled proteins in the stroma (Halperin and Adam, 1996; Halperin et al., 2001a), in degradation of thylakoid membrane proteins in response to nutrient limitation (Majeran et al., 2000), and in shoot development (Kuroda and Maliga, 2003). FtsH is involved in the degradation of unassembled proteins in the thylakoid membrane (Ostersetzer and Adam, 1997), degradation of the oxidatively damaged D1 protein of PSII (Lindahl et al., 2000; Bailey et al., 2002; Sakamoto et al., 2002; Silva et al., 2003), and chloroplast development (Chen et al., 2000; Takechi et al., 2000; Sakamoto et al., 2002). The only reported activity of chloroplast DegP is the initial cleavage of the D1 protein of PSII during repair from photoinhibition (Haussuhl et al., 2001). Information about activities of mitochondrial proteases is even more scarce, with the involvement of Lon protease in the control of cytoplasmic male sterility (Sarria et al., 1998) being the only report to date.

Although it is evident that the number of genes encoding organelle proteases has increased during evolution, it is not clear what the functional significance of this multiplicity is. Two possibilities can be considered: (1) different isomers have different biochemical or physiological functions; and (2) homologous genes ensure sufficient expression under a wide range of developmental and environmental conditions. To start dissecting this issue, we sought to determine how individual members of the four gene families respond to different environmental conditions and to estimate the transcript level of each gene under optimal growth conditions. We demonstrate here that within each family, some individual genes respond to either exposure to high light or changing temperatures, and some do not. Variability between family members was observed for both the extent of the response and the absolute values of transcript abundance. For the FtsH family, we show that the relative in vivo importance of the different isozymes, as revealed by mutants' analysis, reflects differential expression, both at mRNA and protein levels, and not necessarily different functions.

RESULTS

Transcript Abundance in Response to High Light, Heat, and Cold

The relatively large number of genes constituting families encoding organelle proteases suggests that they might have specific roles under different environmental conditions. If so, they are expected to respond differently to changing conditions. To test this possibility, we exposed Arabidopsis plants that grew under optimal conditions to either high-light intensity (770 μE m−2 s−1 for 2.5 h), high temperature (42°C for 1 h), or low temperature (4°C for 18 h). RNA was isolated from treated and control plants, labeled, and hybridized to slides containing arrays of specific 70-mer oligonucleotides. The results of these experiments are summarized in Figure 1. Hybridization to control genes confirmed that the steady-state transcript level of the constitutively expressed gene Ubq10 was not changed with treatments. Expression of the stress-related gene Hsc70 increased more than 10-fold in response to high light, and approximately 4-fold in response to either low or high temperature. More than 2-fold increase in accumulation of Elip2 transcript, known to respond to changes in light or temperature, was also observed in response to the three treatments. These results confirmed the suitability of the experimental system for the study of transcript accumulation.

Figure 1.

Figure 1.

Open in a new tab

Effect of high light, heat, and cold on Clp (A), FtsH (B), DegP (C), and Lon (D) protease transcript level. Transcript levels were tested using arrays containing 70-mer specific oligonucleotides as described in “Materials and Methods.” Ratios between treatment and control levels are presented. Values for the high light experiment are in light gray, cold treatment are in dark gray, and heat in white. The lower, dashed horizontal line represents constitutive expression, and the line above it indicates the arbitrary cutoff for up-regulation. Values are means ± sd of 20 replicates.

In all gene families, exposure to high-light intensity appeared to be the most effective stimulus. Using 2-fold increase as a cutoff for up-regulation, 83% (35 out of 42) of the protease genes showed an increase in their steady-state transcript level (Fig. 1). The degree of up-regulation varied between the responsive genes; 29 showed 2- to 5-fold induction, and the remaining 6 were 5- to 20-fold up-regulated. The tested genes were much less responsive to temperature shifts. Only 5 and 6 genes showing increased mRNA accumulation in response to exposure to 42°C and 4°C, respectively. The extent of response was lower than observed for high light. None of the tested genes was down-regulated in response to any of the relatively short exposures to the different conditions.

Within individual protease families, genes encoding Clp subunits, both proteolytic and regulatory, were also responsive to high light, with 12 out of 15 showing 3- to 6-fold increase. Only three of these were up-regulated in response to low temperature, and none was affected by high temperature (Fig. 1A). (The plastid-encoded ClpP1 and the nuclear-encoded ClpS1, S2, B3, and T were not tested). Eleven out of 12 FtsH genes responded positively to high light, whereas only two and one were affected by high and low temperature, respectively. Two of the FtsH genes, FtsH4 and 8, showed remarkable 11- and 20-fold increases, respectively (Fig. 1B). More than half of DegP transcripts, 6 out of 11, were up-regulated by high light, although here the response was somewhat more moderate, with 2- to almost 4-fold increase (Fig. 1C). Four and three genes responded to low and high temperature shifts, respectively. Three of the four Lon genes showed 2- to 4-fold increase in transcript accumulation in response to high light, three were up-regulated by low temperature, and two by high temperature shift (Fig. 1D). It is interesting to note that although in most cases the response to high light was higher than that to changes in temperature, DegP4, 10, and 11 and Lon 4 responded more to temperature than to high light. Selected gene products were also subjected to reverse transcription (RT)-PCR analysis (Fig. 2). In accordance with their known characteristics, the level of the constitutively expressed ubiquitin transcript was not affected by exposure to high light, whereas the light-induced Elip2 did. Several protease genes that demonstrated up-regulation in response to high light in the array experiment behaved similarly also in the RT-PCR experiment (Fig. 2).

Figure 2.

Figure 2.

Open in a new tab

Effect of high light on accumulation of protease mRNA. Plants were treated with high light, and RNA was isolated and subjected to semiquantitative RT-PCR analysis, as detailed in “Materials and Methods.” C, Control; HL, high light.

Products of the studied genes were previously shown, or predicted, to be targeted to either chloroplasts or mitochondria (Adam et al., 2001; Sokolenko et al., 2002; Sakamoto et al., 2003). We therefore tested whether the effect of environmental conditions on expression could be related to the cellular location of the gene product. As shown in Table I, the different gene families exhibited different trends. Ten out of the 11 Clp genes, whose products are targeted to chloroplasts, are up-regulated by high light, whereas only two out the four mitochondria-targeted ones respond to this stimulus. Similar trend is observed in the DegP family. A lower proportion of the genes encoding mitochondrial isozymes responds to high light than the chloroplast-targeted ones. Within the FtsH family, almost all chloroplast-targeted gene products and all mitochondrial ones respond to high light, whereas in the Lon family only one out of three chloroplast-targeted gene products is up-regulated.

Table I.

Number of chloroplast and mitochondria protease genes responding to different environmental stimuli

Protease Family Cellular Location and Number of Proteins Number of Genes Responding to
HL Cold Heat
C 11 10 3 0
Clp M 4 2 0 0
C 9 8 1 2
FtsH M 3 3 0 0
C 4 4 0 0
DegP M 6 1 1 2
C 3 1 1 0
Lon M 1 1 0 1
Open in a new tab

Cellular locations of the different proteases were compiled from Adam et al. (2001), Sokolenko et al. (2002), and Sakamoto et al. (2003). Number of genes responding to environmental stimuli were taken from the data presented in Figure 1. C, Chloroplast; M, mitochondria; HL, high light.

Relative Transcript Abundance within Gene Families

Gene arrays that are based on cDNAs are usually used to compare the level of expression between two different conditions or backgrounds. They give relative values for each gene, but comparison between expression of different genes under a given condition or background cannot be derived from such experiments. The issue of transcript abundance is important in the analysis of gene families because differential accumulation of different transcripts within a family may have implications on relations between the different isozymes and their function. To assess the abundance of different transcripts within a gene family in relation to each other, we compiled data from two different Affymetrix 25 K-chip experiments that were available on the Internet (see “Materials and Methods”). We retrieved hybridization values only for the genes that were represented in our 70-mer oligonucleotide chips and only the data from control rosette leaves, regardless of the biological experiment that the treated plants were subjected to. Analysis of the data revealed that as expected, the transcript for Lhb1B2, encoding a highly abundant PSII antenna protein, is expressed at a relatively high level (Fig. 3). Ubq10, encoding a constitutively expressed component of the ubiquitin system, is also highly abundant. Much less abundant is Hsc70, which is known to accumulate upon exposure to stress conditions. Consistent with the well-documented lack of expression of Elip2 under optimal growth conditions and also with our RT-PCR verification experiment (Fig. 2), the transcript for this gene is undetectable in the Affymetrix data. Within the Clp family, ClpC1 transcript is relatively highly abundant, accumulating to a level of approximately 75% of Ubq10 (Fig. 3A). It should be noted that its highly homologous isomer, ClpC2, is not represented on the Affymetrix array. ClpP5 and ClpR1 and 4 are also relatively abundant, and somewhat less abundant are the other Clp transcripts. The least-expressed transcripts in this family are ClpP2 and ClpX2, both encoding subunits of the mitochondrial Clp protease.

Figure 3.

Figure 3.

Open in a new tab

Absolute transcript abundance of protease genes. Publicly available data from expression studies, using Affymetrix 25 K chips, were analyzed. Values from control rosette leaves in two different experiments, performed in different laboratories, for protease and control genes (Ubq10, Lhb1B2, and Hsc70-3), were compiled. Data from both experiments are presented. A, Clp subunits; B, FtsH; C, DegP and Lon.

The most abundant FtsH transcript is that of FtsH2 (Fig. 3B), demonstrating a level similar to ClpC1. FtsH1 and 5 are also relatively abundant, each accumulating to levels of about 50% to 60% of FtsH2 transcript. Other FtsH transcripts are much less abundant, with FtsH6 not accumulating at all under optimal growth conditions. DegP and Lon transcripts are much less abundant than those of Clp and FtsH. Within the DegP family, DegP1 is the most abundant, followed by DegP2, 5, and 8 (Fig. 3C). Other DegPs, as well as Lon transcripts, accumulate only to a very low level. It is also evident that within each family, transcripts encoding proteases that are targeted to chloroplasts are more abundant than those destined to mitochondria.

Since our arrays were composed of 70-mer oligonucleotides with similar melting temperatures, we reasoned that, unlike cDNA arrays, the signal associated with each spot should be proportional to the absolute abundance of the corresponding gene product. Thus, analysis of the data from the control channel should give an estimate of transcript abundance in control plants. To test the validity of this assumption, we compiled our raw data from the control channel in all experiments and averaged them to get a mean value for each gene. We then compared these values to the above Affymetrix data. Unfortunately, correlation between the two sets of data was relatively low (data not shown). It appears that although the 70-mer arrays are very useful for comparison of transcript abundance in different treatments, they cannot be reliable for assessing the relative transcript abundance within a family. This is probably due to variations in pin geometry that result in different amount of printed DNA (Schuchhardt et al., 2000).

Correlation between Transcript Abundance, ESTs, and Mutant Phenotype

Another indication for the relative transcript abundance within families might be the number of expressed sequence tags (ESTs) representing a given gene product in nonnormalized cDNA libraries. Thus, we compiled tag numbers or data from massively parallel signature sequencing, available in three different publicly opened sources (TAIR, MIPS, and MPSS, respectively) and compared them to transcript abundance derived from Affymetrix data. This comparison revealed a positive correlation between the number of tags and transcript abundance (Fig. 4). Interestingly, genes that did not give a statistically significant signal in the different Affymetrix experiments were not represented in the different tags data. These included FtsH6, Lon3 and 4, and DegP4 and 11 to 14. These results suggest that both Affymetrix data and any of the tags data are a reliable source for estimation of absolute transcript abundance.

Figure 4.

Figure 4.

Open in a new tab

Correlation between transcript abundance values from different sources. Number of ESTs, MPSS data, and values from Affymetrix data, corresponding to the different protease genes, were compiled and correlation between them was calculated. Correlation between Affymetrix data and ESTs found in TAIR and MIPS are shown in A and B, respectively, and MPSS data in C. Correlation coefficient values are presented in the table.

At least within the family of FtsH protease, absolute transcript abundance correlated fairly well with mutant visual phenotypes. It has been demonstrated previously that mutations in FtsH2 and FtsH5 lead to leaf variegation, with FtsH2 mutants being more variegated than FtsH5 (Chen et al., 2000; Takechi et al., 2000; Sakamoto et al., 2002). By contrast, T-DNA insertions in the highly homologous genes encoding FtsH1, 6, and 8 do not cause any visual phenotype (Sakamoto et al., 2003). Among FtsH genes, the transcript of FtsH2 is the most abundant, FtsH5 and 1 are approximately 50% less, FtsH8 is more than 80% less abundant, and the transcript for FtsH6 appears to accumulate to a very low level, if at all. Thus, the severity of the variegated phenotype in FtsH2 and 5 and the lack of phenotype in FtsH6 and 8 correlate well with their respective transcript level under optimal growth condition. The only exception to this rule is FtsH1, whose transcript level is similar to that of FtsH5, but its loss is not manifested visually under normal growth conditions.

Accumulation of FtsH Isozymes

Studying expression in gene families is usually easier at the transcript level than the protein one. However, it is obvious that transcript abundance only approximates protein levels. In light of the apparent correlation between absolute transcript abundance and mutant phenotype, it was interesting to evaluate the levels of individual FtsH isozymes. Attempts to generate specific antibodies to some of the chloroplast FtsH isozymes revealed that antibodies against FtsH2 cross-reacted with FtsH8, and those against FtsH5 cross-reacted with FtsH1 (Sakamoto et al., 2003). This cross-reactivity correlated well with the degree of identity between each pair of proteins—FtsH1 and 5 are duplicated genes, and so are FtsH2 and 8 (Sakamoto et al., 2003). Thus, although useful for other purposes, these antibodies could not be used for quantification of the different isozymes. To overcome this difficulty, a specific isoelectric focusing-SDS-PAGE system was developed. Thylakoid membranes were first treated to remove lipids and pigments, and proteins were precipitated and then separated on a long isoelectric focusing strip with a very narrow pI range (one pH unit), prior to conventional second-dimension separation by SDS-PAGE. As shown in Figure 5, several protein spots can be observed in the region corresponding to FtsH isozymes' molecular mass and pI, and most of these cross-react with the FtsH antibody. Spots corresponding to these were identified by electrospray ionization tandem mass spectrometry. Interestingly, only FtsH1, 2, 5, and 8 were identified, whereas the other putative thylakoid isozymes (FtsH6, 7, 9, 11, and 12) were not. Both FtsH2 and 8 resolved into three different spots each. This is not an uncommon observation. Multiple spots on two-dimensional gels corresponding to single proteins were observed also for Clp subunits (Peltier et al., 2004). These could result from either posttranslation modifications occurring in vivo or modifications that occurred during the experimental procedure.

Figure 5.

Figure 5.

Open in a new tab

Identification of FtsH isozymes accumulating in thylakoid membranes. Thylakoid membrane proteins were separated by two-dimensional PAGE as described in “Materials and Methods.” The proteins were either silver stained or immunodetected with an antibody against FtsH protease. Sections of the stained and blotted gels are shown in the top part of the figure. Examples of diagnostic peptides that enabled identification of individual protein spots are shown in the bottom part of the figure. The corresponding sequence in the closest homolog is also shown, with the different amino acid residues, which allow the unequivocal identification, shown in bold.

Silver staining is considered a nonquantitative staining method, and thus estimation of abundance of the different thylakoid FtsH isozymes could not be derived from the gel itself. However, it could be estimated from the immunoblot (Fig. 5). The antibody used was generated against a 16-amino acid synthetic peptide that corresponds to a highly conserved sequence in all FtsH proteins. Thus, the intensity of the spots on the blot should be proportional to the level of the different isozymes. Quantification of spots on the blot revealed that FtsH2 was the most abundant species, FtsH5 accumulated to a level approximately 60% of FtsH2, FtsH8 to approximately 50%, and FtsH1 to approximately 10%. Thus, the highest abundance of FtsH2 at the transcript level (see Fig. 3) is mirrored at the protein level as well, and its loss is manifested most severely in specific mutants. This correlation is valid for FtsH5 as well.

DISCUSSION

A striking feature of chloroplast and mitochondrial proteases is that they are encoded by multiple gene families. This raises the question whether the gene products have specific or redundant functions. Part of the answer was revealed by the observation that different gene products from the same family are targeted to either chloroplasts or mitochondria. Alternative targeting was predicted by specific algorithms (Adam et al., 2001; Sokolenko et al., 2002), and several predictions were confirmed experimentally. For instance, ClpP2 and ClpX were detected immunologically in mitochondria but not in chloroplasts (Halperin et al., 2001b). ClpP2 was found in mitochondria by proteomic analyses as well (Millar et al., 2001; Peltier et al., 2004). Other ClpPs and ClpRs were found only in chloroplasts (Sokolenko et al., 1998; Nakabayashi et al., 1999; Peltier et al., 2001, 2004; Zheng et al., 2002). The pea homolog of FtsH3 was detected in mitochondria but not in chloroplasts (Kolodziejczak et al., 2002). Transient expression assays of green fluorescent protein fusions revealed that Arabidopsis FtsH3, 4, and 10 were indeed targeted to mitochondria, whereas all other nine FtsHs were targeted to the chloroplast (Takechi et al., 2000; Sakamoto et al., 2002, 2003).

Another possibility to account for multiple protease gene products in each compartment might be specialized expression under different environmental conditions, i.e. different genes are responsible for expression of the protease under specific conditions. Our analysis (Fig. 1) reveals a complex situation. In the great majority of genes that responded to environmental stimuli, exposure to high light was more effective than temperature, suggesting that the expression of these genes could not be defined as temperature specific. However, DegP4, 10, and 11 and Lon4 were up-regulated more by temperature shifts than high light, suggesting that these might have a more specific role under temperature-stress conditions.

The higher proportion of genes responding to high light than to temperature may represent a higher need for proteases under the former conditions. This might have to do with the nature of damage incurred to proteins by the different conditions. Heat-denatured proteins are usually protected by molecular chaperones from aggregation (Wickner et al., 1999). Once normal temperature is restored, they will most likely resume their native conformation. By contrast, exposure to high light leads to oxidative damage that may result in irreversible conformational changes and protein inactivation. To deal with such damaged proteins, an excess of proteolytic enzymes may be needed, and this is probably why these enzymes are up-regulated in response to high light.

When it comes to gene families, relative abundance data in itself may not be sufficient for assessing the function and importance of individual gene products. Up-regulation of one low-abundance gene product may still result in its lower abundance compared with another gene product whose level is not increased, but its abundance is much higher to begin with. Thus, absolute abundance of individual genes needs to be considered. Analysis of one-channel arrays data might be a useful resource for this purpose (Ferl et al., 2003). Compiling Affymetrix data of protease genes revealed that data from control plants, grown and analyzed in different laboratories, showed very little variability between experiments, as suggested (Ferl et al., 2003). Transcript levels of the protease genes were lower than highly abundant transcripts. The most abundant ones, such as ClpC1, ClpP5, and FtsH2, accumulated to one-third to one-half of the level of highly expressed genes such as Lhb1B2 and Ubq10 (Fig. 3). Other gene products accumulated to much lower levels. Such analysis demonstrates the utility of assessing the relative expression level under a given condition within gene families and between different families or individual genes.

The absolute transcript level within a family might be useful in analysis of the results obtained in the stress experiment as well. For instance, the FtsH8 transcript shows a dramatic increase (approximately 20-fold) upon exposure to high light, but its absolute level under normal growth conditions is relatively low. FtsH2 transcript level under normal conditions is 3- to 6-fold higher than FtsH8, and it increases in response to high light by 5-fold. Thus, it is expected that FtsH2 and 8 will be similarly abundant under light-stress conditions.

Another source for estimation of absolute transcript abundance could be EST databases. Theoretically, transcripts that are more abundant in vivo should be represented by more ESTs in nonnormalized cDNA libraries. To assess this assumption, we tested the correlation between transcript abundance as revealed by Affymetrix data, the number of ESTs in two databases, and results from MPSS database, all available on publicly open resources. The high correlation between the Affymetrix data and the other resources suggests they are all suited for estimation of absolute transcript abundance of a given gene. Thus, EST databases might be a useful tool in most plant species where EST databases, but not full genome sequences, are available.

Extrapolation from mRNA to protein level is a common, but not always justified, practice. To evaluate how reliable this practice is in the case of organelle proteases, we studied the accumulation of thylakoid FtsH isozymes. Given the fact that the nine putative thylakoid FtsHs have very similar molecular masses and pIs, distinguishing between the different isozymes posed a technical challenge. However, the results presented in Figure 5 show that it is feasible. Out of the nine putative isozymes, only FtsH1, 2, 5, and 8 were identified. It is not known whether the other ones do not accumulate at all, or else their level is just too low for detection. Nevertheless, there is an apparent correlation between the absolute transcript abundance of the different FtsHs (Fig. 3) and the identification of their products at the protein level (Fig. 5)—three out of the four identified isozymes have relatively high transcript level. Interestingly, the relative level of the different isozymes within the FtsH family can be inferred also from analysis of specific mutants. Mutations in FtsH2, whose transcript and protein levels are the highest within the family, cause severe leaf variegation (Chen et al., 2000; Takechi et al., 2000). Mutations in FtsH5, whose transcript and protein levels are lower, lead only to slight variegation that disappears with development (Sakamoto et al., 2002; Sakamoto, 2003). Protein levels of FtsH1 and 8 are lower, and, in accordance, mutants of these are indistinguishable from wild-type plants. These observations suggest that the different thylakoid FtsH isozymes might have similar functions, and their respective importance in vivo reflects their differential expression.

The suggestion that the different FtsHs have similar functions is supported by the ability of overexpressed FtsH8 to complement the mutation in FtsH2 (Yu et al., 2004) and the interchangeability of FtsH1 and 5 (S. Rodermel, personal communication). Since FtsH2 and 5 form a mixed complex (Sakamoto et al., 2003), it is likely that FtsH1 and 8 also participate in this complex, and all four forms are interchangeable. However, this will have to be further confirmed by complementation experiments in double and triple mutants with the different FtsH genes.

MATERIALS AND METHODS

Plant Material

Wild-type Arabidopsis plants, ecotype Columbia, were grown under controlled conditions (20°C, 100 μE m−2 s−1 light intensity and 70% relative humidity). For testing environmental effects on expression, 6-week-old plants were transferred to either light intensity of 770 μE m−2 s−1 for 2.5 h, 40°C for 1 h, or 4°C for 18 h.

70-mer Oligonucleotide Arrays

Gene-specific 70-mer oligonucleotides with 5′-aminolinker were purchased from Operon (Qiagen, Chatswort, CA). Oligonucleotide sequences were chosen from the 3′ end of the genes, to be as free as possible from homology with other genes. The oligonucleotides were designed to have a melting temperature of 74°C ± 3°C and were checked for lack of secondary structures. The oligonucleotides were dissolved in 5× SSC to a final concentration of 70 μm, and spotted at five repetitions twice, on the two halves of SuperAmine-coated glass slides (Telechem, Sunnyvale, CA) using an arraying robot with 16-pin print heads (BioRobotics, Cambridge, UK). Arraying and the scanning described below were carried out at the Microarray Unit, Department of Biological Services, The Weizmann Institute.

RNA Extraction, cDNA Preparation and Labeling, Hybridization, and Scanning

Total RNA was isolated from 1 g of fresh leaves of control and treated plants, using the RNeasy kit (Qiagen, Chatswort, CA), according to the manufacturer's instructions. One hundred micrograms of total RNA was subjected to reverse transcription reaction using dNTPs mix containing 5-3 aminoallyl dUTP. cDNA was labeled indirectly with succinimidyl ester Cy3/Cy5 as described previously (Guterman et al., 2002). For each one of the two biological repetitions, two hybridizations with swapped dye labeling reactions were performed (see Guterman et al., 2002). Separate images for each fluorescent probe were acquired using ScanArray 4000 software (Packard BioScience) at a resolution of 10 mm per pixel, adjusting the photomultiplier and laser power to achieve an optimal distribution of signals without minimal saturation. Initial image analysis was performed using QuantArray version 3 software (Packard BioScience, Boston). Spots were quantified using the adaptive method, measuring the mean of pixels encompassing the spot and subtracting the local background areas. Average expression ratios, including sd, were calculated from the multiple measurements. The treatment to control ratio of two was arbitrarily used as a cutoff for up-regulation.

Semiquantitative RT-PCR

cDNA was prepared from 500 ng of total RNA using 1 μg oligo(dT)12-18 primers, 2.5 mm dNTPs, 0.1 m dithiothreitol, 5× RT buffer, and 200 units of SuperScript II reverse transcriptase (Invitrogen, Life Technologies, Carlsbad, UK) in a total volume of 20 μL. The reaction mixture was incubated at 65°C for 5 min, 42°C for 50 min, and then inactivated at 70°C for 15 min. The following primer pairs were then used for amplification by PCR: Ubq10, TTCACTTGGTCCTGCGTCTT and CAAGGCCCCAAAACACAAAC; Elip2, CAGTGTTCGCTGCTCCTTCC and TCGATGCCAACGTCAACAAC; ClpC1, TAACCCGAGCTATGGAGCAA and GCCACTTCCACCATTTAGCA; FtsH1, TGGACAAGTTGCTGTTGGTG and CTCGCACCTCAGCATCTACAAT; FtsH2, AGAAACTATTGGCGGTGACG and TGATGCTGGAGTTGTCGTTG; FtsH5, ATGTCATCGCAGAAGGATTACT and TGATCTCTTTCGCTCTCACG; FtsH8, CCATGGTCGCTAATGGATTC and GGTGTTGGTGTTGATGTGGA; DegP8, TTCGTAATGGAGCCCTTGTC and CGGCTTTGTTCTTCACAGGT. In addition to the specific primers, the reaction mixtures contained 2 μL of the cDNA and 1 unit of Taq polymerase (JMR Holding, Sevenoaks Kent, UK), in a total volume of 25 μL. Annealing temperature was 60°C, with the exception of Elip2, for which the temperature was 55°C. All reactions were subjected to 12 cycles, followed by separation of the products on 1.5% agarose gel, and the DNA was then transferred to Hybond N+ membrane. To visualize the products, probes were labeled by random priming with DIG High Prime and Detection Starter Kit II (Roche, Indianapolis), according to the manufacturer's protocol. Membranes were prehybridized with DIG Easy Hyb solution for 30 min at 42°C, followed by hybridization in the same solution containing denatured DIG-labeled probe at 42°C overnight. The membranes were washed to a final stringency of 0.5× SSC and 0.1% SDS at 68°C. The membranes were subjected to immunological detection with anti-digoxigenin-AP conjugate and disodium 3-(4-methoxyspiro{1,2-dioxetane- 3,2′ –(5′-chloro) tricyclo [3.3.1.1. 3.7] decan}- 4xyl) phenyl phosphate, followed by exposure to x-ray film (Fuji, Tokyo). Hybridization signals were analyzed using the ImageScaner (Amersham Biosciences, Piscataway, NJ).

Publicly Available Data Analysis

Data from Arabidopsis 25 K Affymetrix chips, available at http://nasc.nott.ac.uk, were searched in order to estimate the absolute transcript abundance of protease genes. Data from control channels of two different experiments done on rosette leaves were compiled and normalized to the Ubq10 value in each experiment. To assess expression based on numbers of Arabidopsis ESTs corresponding to protease genes, the TAIR, MIPS, and MPSS databases, available at http://www.arabidopsis.org/info/expression/index.jsp, http://mips.gsf.de, and http://mpss.ucdavis.edu/monjava.html, respectively, were searched, and the relevant data were compiled. Statistical analysis was performed using the JUMP program.

Two-Dimensional PAGE Analysis

Chloroplast thylakoid membranes were isolated according to Aronsson and Jarvis (2002), with minor modifications. Arabidopsis leaves were homogenized in grinding buffer (0.3 m sorbitol, 20 mm HEPES-KOH, pH 8.0, 5 mm MgCl2, 5 mm EDTA, 5 mm EGTA, 10 mm NaHCO3). Thylakoid membranes were precipitated by centrifugation at 1,000g for 10 min from filtered homogenate and further purified on 40% Percoll. Membranes were then washed twice with ice-cold 0.1 m Na2CO3 and once with 10 mm Tricine-KOH, pH 8.0, 10 mm NaCl, and 10 mm MgCl2. Washed membranes were precipitated again and resuspended in storage buffer (0.3 m sorbitol, 10 mm HEPES-KOH, pH 8.0, 10 mm MgCl2, 10 mm NaCl) to 1 to 2 mg chlorophyll/mL, and either used immediately or stored at −70°C. Proteins were extracted from membranes and separated from pigments by resuspension in 20% TCA and 1% β-mercaptoethanol in acetone, as described previously (Gorg et al., 1988), and solubilized in two-dimensional sample buffer (8 m urea, 2 m thiourea, 2% w/v CHAPS, 1% w/v NP-40, 100 mm dithiothreitol). Isoelectric focusing was performed using 18-cm IPG strips (Amersham Biosciences) covering the pI range of 5 to 6. Focusing was performed as follows: 6 h at 30 V, 6 h at 60 V, 1 h at 500 V, 1 h at 1,000 V, 2 h at 5,000 V, and then 8,000 V until 60,000 Vh was reached. Typically, focusing was completed in 24 h. In the second dimension, proteins were separated by 12% PAGE according to Fling and Gregerson (1986), and silver staining of proteins was performed as described previously (Yan et al., 2000). For immunodetection proteins were transferred to a nitrocellulose membrane by the semidry method using Trans-Blot DS (Bio-Rad Laboratories, Hercules, CA) according to manufacturer's instructions, reacted with a FtsH antibody (Lindahl et al., 1996), and visualized by the SuperSignal ECL kit (Pierce, Rockford, IL). Silver-stained protein spots, corresponding to the immunoreactive spots on the western blot, were excised and identified by electrospray ionization tandem mass spectrometry (Yates, 1998) at the Smoler Protein Research Center, Technion, Haifa, Israel.

Acknowledgments

We thank Dr. D. Chamovitz for Elip primers, E. Kapri for helpful discussion throughout the course of this study, and Dr. N. Ori for critical reading of the manuscript.

1

This work was supported in part by grants from the Israel Science Foundation and the U.S.-Israel Binational Agricultural Research and Development Fund (to Z.A.).

Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.104.043299.

References

  1. Adam Z (2000) Chloroplast proteases: Possible regulators of gene expression? Biochimie 82: 647–654 [DOI] [PubMed] [Google Scholar]
  2. Adam Z, Adamska I, Nakabayashi K, Ostersetzer O, Haussuhl K, Manuell A, Vallon O, Rodermel SR, Shinozaki K, Clarke AK (2001) Chloroplast and mitochondrial proteases in Arabidopsis. A proposed nomenclature. Plant Physiol 125: 1912–1918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Adam Z, Clarke AK (2002) Cutting edge of chloroplast proteolysis. Trends Plant Sci 7: 451–456 [DOI] [PubMed] [Google Scholar]
  4. Aronsson H, Jarvis P (2002) A simple method for isolating import-competent Arabidopsis chloroplasts. FEBS Lett 529: 215–220 [DOI] [PubMed] [Google Scholar]
  5. Bailey S, Thompson E, Nixon PJ, Horton P, Mullineaux CW, Robinson C, Mann NH (2002) A critical role for the Var2 FtsH homologue of Arabidopsis thaliana in the photosystem II repair cycle in vivo. J Biol Chem 277: 2006–2011 [DOI] [PubMed] [Google Scholar]
  6. Chen M, Choi Y, Voytas DF, Rodermel S (2000) Mutations in the Arabidopsis VAR2 locus cause leaf variegation due to the loss of a chloroplast FtsH protease. Plant J 22: 303–313 [DOI] [PubMed] [Google Scholar]
  7. Clarke AK (1999) ATP-dependent Clp proteases in photosynthetic organisms: A cut above the rest! Ann Bot 83: 593–599 [Google Scholar]
  8. Ferl GZ, Timmerman JM, Witte ON (2003) Extending the utility of gene profiling data by bridging microarray platforms. Proc Natl Acad Sci USA 100: 10585–10587 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fling SP, Gregerson DS (1986) Peptide and protein molecular weight determination by electrophoresis using a high-molarity tris buffer system without urea. Anal Biochem 155: 83–88 [DOI] [PubMed] [Google Scholar]
  10. Gorg A, Postel W, Domscheit A, Gunther S (1988) Two-dimensional electrophoresis with immobilized pH gradients of leaf proteins from barley (Hordeum vulgare): method, reproducibility and genetic aspects. Electrophoresis 9: 681–692 [DOI] [PubMed] [Google Scholar]
  11. Gottesman S (1996) Proteases and their targets in Escherichia coli. Annu Rev Genet 30: 465–506 [DOI] [PubMed] [Google Scholar]
  12. Guterman I, Shalit M, Menda N, Piestun D, Dafny-Yelin M, Shalev G, Davydov O, Ovadis M, Emanuel M, Wang J, et al (2002) Rose scent: genomic approach to discover novel floral fragrance-related genes. Plant Cell 14: 2325–2338 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Halperin T, Adam Z (1996) Degradation of mistargeted OEE33 in the chloroplast stroma. Plant Mol Biol 30: 925–933 [DOI] [PubMed] [Google Scholar]
  14. Halperin T, Ostersetzer O, Adam Z (2001. a) ATP-dependent association between subunits of Clp protease in pea chloroplasts. Planta 213: 614–619 [DOI] [PubMed] [Google Scholar]
  15. Halperin T, Zheng B, Itzhaki H, Clarke AK, Adam Z (2001. b) Plant mitochondria contain proteolytic and regulatory subunits of the ATP-dependent Clp protease. Plant Mol Biol 45: 461–468 [DOI] [PubMed] [Google Scholar]
  16. Haussuhl K, Andersson B, Adamska I (2001) A chloroplast DegP2 protease performs the primary cleavage of the photodamaged D1 protein in plant photosystem II. EMBO J 20: 713–722 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Itzhaki H, Naveh L, Lindahl M, Cook M, Adam Z (1998) Identification and characterization of DegP, a serine protease associated with the luminal side of the thylakoid membrane. J Biol Chem 273: 7094–7098 [DOI] [PubMed] [Google Scholar]
  18. Kolodziejczak M, Kolaczkowska A, Szczesny B, Urantowka A, Knorpp C, Kieleczawa J, Janska H (2002) A higher plant mitochondrial homologue of the yeast m-AAA protease. Molecular cloning, localization, and putative function. J Biol Chem 277: 43792–43798 [DOI] [PubMed] [Google Scholar]
  19. Kuroda H, Maliga P (2003) The plastid clpP1 protease gene is essential for plant development. Nature 425: 86–89 [DOI] [PubMed] [Google Scholar]
  20. Lindahl M, Spetea C, Hundal T, Oppenheim AB, Adam Z, Andersson B (2000) The thylakoid FtsH protease plays a role in the light-induced turnover of the photosystem II D1 protein. Plant Cell 12: 419–431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lindahl M, Tabak S, Cseke L, Pichersky E, Andersson B, Adam Z (1996) Identification, characterization, and molecular cloning of a homologue of the bacterial FtsH protease in chloroplasts of higher plants. J Biol Chem 271: 29329–29334 [DOI] [PubMed] [Google Scholar]
  22. Majeran W, Wollman F-A, Vallon O (2000) Evidence for a role of ClpP in the degradation of the chloroplast cytochrome b6f complex. Plant Cell 12: 137–149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Millar AH, Sweetlove LJ, Giege P, Leaver CJ (2001) Analysis of the Arabidopsis mitochondrial proteome. Plant Physiol 127: 1711–1727 [PMC free article] [PubMed] [Google Scholar]
  24. Nakabayashi K, Ito M, Kiosue T, Shinozaki K, Watanabe A (1999) Identification of clp genes expressed in senescing Arabidopsis leaves. Plant Cell Physiol 40: 504–514 [DOI] [PubMed] [Google Scholar]
  25. Ostersetzer O, Adam Z (1997) Light-stimulated degradation of an unassembled Rieske FeS protein by a thylakoid-bound protease: the possible role of the FtsH protease. Plant Cell 9: 957–996 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Peltier J-B, Ytterberg J, Liberles DA, Roepstorff P, van Wijk KJ (2001) Identification of a 350 kDa ClpP protease complex with 10 different Clp isoforms in chloroplasts of Arabidopsis thaliana. J Biol Chem 276: 16318–16327 [DOI] [PubMed] [Google Scholar]
  27. Peltier J-B, Emanuelsson O, Kalume DE, Ytterberg J, Friso G, Rudella A, Liberles DA, Soderberg L, Roepstorff P, von Heijne G, et al (2002) Central functions of the lumenal and peripheral thylakoid proteome of Arabidopsis determined by experimentation and genome-wide prediction. Plant Cell 14: 211–236 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Peltier JB, Ripoll DR, Friso G, Rudella A, Cai Y, Ytterberg J, Giacomelli L, Pillardy J, Van Wijk KJ (2004) Clp protease complexes from photosynthetic and non-photosynthetic plastids and mitochondria of plants, their predicted 3-D structures and functional implications. J Biol Chem 279: 4768–4781 [DOI] [PubMed] [Google Scholar]
  29. Sakamoto W (2003) Leaf-variegated mutations and their responsible genes in Arabidopsis thaliana. Genes Genet Syst 78: 1–9 [DOI] [PubMed] [Google Scholar]
  30. Sakamoto W, Tamura T, Hanba-Tomita Y, Sodmergen, Murata M (2002) The VAR1 locus of Arabidopsis encodes a chloroplastic FtsH and is responsible for leaf variegation in the mutant alleles. Genes Cells 7: 769–780 [DOI] [PubMed] [Google Scholar]
  31. Sakamoto W, Zaltsman A, Adam Z, Takahashi Y (2003) Coordinated regulation and complex formation of YELLOW VARIEGATED1 and YELLOW VARIEGATED2, chloroplastic FtsH metalloproteases involved in the repair cycle of photosystem II in Arabidopsis thylakoid membranes. Plant Cell 15: 2843–2855 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sarria R, Lyznik A, Vallejos CE, Mackenzie SA (1998) A cytoplasmic male sterility-associated mitochondrial peptide in common bean is post-translationally regulated. Plant Cell 10: 1217–1228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Schubert M, Petersson UA, Haas BJ, Funk C, Schroder WP, Kieselbach T (2002) Proteome map of the chloroplast lumen of Arabidopsis thaliana. J Biol Chem 277: 8354–8365 [DOI] [PubMed] [Google Scholar]
  34. Schuchhardt J, Beule D, Malik A, Wolski E, Eickhoff H, Lehrach H, Herzel H (2000) Normalization strategies for cDNA microarrays. Nucleic Acids Res 28: e47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Silva P, Thompson E, Bailey S, Kruse O, Mullineaux CW, Robinson C, Mann NH, Nixon PJ (2003) FtsH is involved in the early stages of repair of photosystem II in Synechocystis sp. PCC 6803. Plant Cell 15: 2152–2164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Skorko-Glonek J, Lipinska B, Krzewski K, Zolese G, Bertoli E, Tanfani F (1997) HtrA heat shock protease interacts with phospholipid membranes and undergoes conformational changes. J Biol Chem 272: 8974–8982 [DOI] [PubMed] [Google Scholar]
  37. Sokolenko A, Lerbs-Mache S, Altschmied L, Herrmann RG (1998) Clp protease complexes and their diversity in chloroplasts. Planta 207: 286–295 [DOI] [PubMed] [Google Scholar]
  38. Sokolenko A, Pojidaeva E, Zinchenko V, Panichkin V, Glaser VM, Herrmann RG, Shestakov SV (2002) The gene complement for proteolysis in the cyanobacterium Synechocystis sp. PCC 6803 and Arabidopsis thaliana chloroplasts. Curr Genet 41: 291–310 [DOI] [PubMed] [Google Scholar]
  39. Takechi K, Sodmergen, Murata M, Motoyoshi F, Sakamoto W (2000) The YELLOW VARIEGATED (VAR2) locus encodes a homologue of FtsH, an ATP-dependent protease in Arabidopsis. Plant Cell Physiol 41: 1334–1346 [DOI] [PubMed] [Google Scholar]
  40. Wickner S, Maurizi MR, Gottesman S (1999) Posttranslational quality control: folding, refolding, and degrading proteins. Science 286: 1888–1893 [DOI] [PubMed] [Google Scholar]
  41. Yan JX, Wait R, Berkelman T, Harry RA, Westbrook JA, Wheeler CH, Dunn MJ (2000) A modified silver staining protocol for visualization of proteins compatible with matrix-assisted laser desorption/ionization and electrospray ionization-mass spectrometry. Electrophoresis 21: 3666–3672 [DOI] [PubMed] [Google Scholar]
  42. Yates JR (1998) Mass spectrometry and the age of the proteome. J Mass Spectrom 33: 1–19 [DOI] [PubMed] [Google Scholar]
  43. Yu F, Park S, Rodermel SR (2004) The Arabidopsis FtsH metalloprotease gene family: interchangeability of subunits in chloroplast oligomeric complexes. Plant J 37: 864–876 [DOI] [PubMed] [Google Scholar]
  44. Zheng B, Halperin T, Hruskova-Heidingsfeldova O, Adam Z, Clarke AK (2002) Characterization of chloroplast Clp proteins in Arabidopsis: localization, tissue specificity and stress responses. Physiol Plant 114: 92–101 [DOI] [PubMed] [Google Scholar]

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES