- © 2016 American Society of Plant Biologists. All Rights Reserved.
Abstract
FTSH4 is one of the inner membrane-embedded ATP-dependent metalloproteases in mitochondria of Arabidopsis (Arabidopsis thaliana). In mutants impaired to express FTSH4, carbonylated proteins accumulated and leaf morphology was altered when grown under a short-day photoperiod, at 22°C, and a long-day photoperiod, at 30°C. To provide better insight into the function of FTSH4, we compared the mitochondrial proteomes and oxyproteomes of two ftsh4 mutants and wild-type plants grown under conditions inducing the phenotypic alterations. Numerous proteins from various submitochondrial compartments were observed to be carbonylated in the ftsh4 mutants, indicating a widespread oxidative stress. One of the reasons for the accumulation of carbonylated proteins in ftsh4 was the limited ATP-dependent proteolytic capacity of ftsh4 mitochondria, arising from insufficient ATP amount, probably as a result of an impaired oxidative phosphorylation (OXPHOS), especially complex V. In ftsh4, we further observed giant, spherical mitochondria coexisting among normal ones. Both effects, the increased number of abnormal mitochondria and the decreased stability/activity of the OXPHOS complexes, were probably caused by the lower amount of the mitochondrial membrane phospholipid cardiolipin. We postulate that the reduced cardiolipin content in ftsh4 mitochondria leads to perturbations within the OXPHOS complexes, generating more reactive oxygen species and less ATP, and to the deregulation of mitochondrial dynamics, causing in consequence the accumulation of oxidative damage.
Mitochondria play an important role in cellular metabolism and cell longevity, with the most notable function in the generation of ATP through oxidative phosphorylation (OXPHOS). During OXPHOS under unfavorable conditions, reactive oxygen species (ROS) are generated, which might lead to oxidative damages (Moller, 2001). To avoid the accumulation of nonfunctional proteins, particularly the formation of insoluble, harmful protein aggregates, mitochondria have evolved a hierarchically structured quality control (QC) system (Tatsuta and Langer, 2008; Baker et al., 2011; Fischer et al., 2012). The QC system links molecular, organellar, and cellular levels and includes (1) a protease/chaperone system (Voos, 2013); (2) mitochondrial fission and fusion processes (Twig et al., 2008; Osellame et al., 2012; Youle and van der Bliek, 2012; Elgass et al., 2013); and, in case the first two actions fail, (3) initiation of the intrinsic cell death program (Fischer et al., 2012; Gaspard and McMaster, 2015). It ensures the persistence of a healthy mitochondrial population and ultimately survival of the organism, especially during stress conditions (Baker et al., 2014).
A conserved intramitochondrial network of chaperones and proteases that maintain protein homeostasis (Tatsuta and Langer, 2007; Baker et al., 2011) constitutes the first level of the QC system. Key proteolytic enzymes are ATP-dependent proteases, which combine both proteolytic and chaperone-like activities (Voos, 2013). Typically, mitochondria contain three types of ATP-dependent proteases: Lon, Clp, and FtsH (Janska et al., 2010). The two first protease families are classified as Ser proteases, whereas FtsH proteases have a catalytic site characteristic for metalloproteases.
FtsH proteases, also termed AAA proteases (for ATPases associated with diverse cellular activities), form oligomeric complexes in the mitochondrial inner membrane with catalytic domains facing the intermembrane space (i-AAA) or the matrix side. Only one i-AAA protease is present in yeast (Yme1) and human (YMEL1), while plant mitochondria contain two i-AAA proteases: FTSH4 and FTSH11 (Urantowka et al., 2005). Interestingly, different from human YMEL1, Arabidopsis (Arabidopsis thaliana) FTSH4 protease does not complement the yeast homolog, suggesting its plant-specific function (Urantowka et al., 2005). Yeast i-AAA protease was reported to act as a QC enzyme and to degrade unassembled inner membrane proteins (Nakai et al., 1995; Pearce and Sherman, 1995; Augustin et al., 2005; Kambacheld et al., 2005) as well as misfolded small translocase of the inner membrane chaperones of the intermembrane space (Baker et al., 2012). Nonassembled respiratory chain proteins have been identified as proteolytic substrates for the human i-AAA protease (Stiburek et al., 2012). So far, there is no information about oxidatively damaged proteins being substrates of these i-AAA proteases. However, an increasing amount of evidence indicates their crucial role in mitochondrial QC (Baker et al., 2011). It seems that the i-AAA protease-mediated proteolysis controls mitochondrial dynamics in multiple ways. Yeast i-AAA was found to degrade the phospholipid transfer proteins UPS1 and UPS2, responsible for the distribution of cardiolipin (CL) and phosphatidylethanolamine (PE) in mitochondrial membranes (Potting et al., 2010). Both phospholipids have a critical impact on several mitochondrial functions, including fusion and fission (Joshi et al., 2012; Pan et al., 2014). Next, it was shown that Yme1 processes Atg32, a mitochondrial outer membrane receptor ensuring the specificity of mitophagy (Wang et al., 2013). Furthermore, in mammals and yeast, i-AAA proteases regulate mitochondrial morphology and dynamics by processing the conserved fusion mediator OPA1 (Mgm1 in yeast; Anand et al., 2013; Ruan et al., 2013). Finally, recent findings indicate that Yme1L regulates mitochondrial dynamics, mitochondrial cristae structure, and nucleoid organization by controlling the Mic60/Mitofilin homeostasis (Li et al., 2015). Other postulated biological roles for i-AAA proteases are dependent exclusively on the chaperone-like activity of these enzymes. Specifically, the yeast Yme1 protease has been proposed to act as a chaperone in the folding of numerous proteins in the intermembrane space (Fiumera et al., 2009; Schreiner et al., 2012) and in the assembly of complex V (Francis and Thorsness, 2011). Furthermore, Yme1 was shown to mediate protein import into the intermembrane space (Rainey et al., 2006).
In Arabidopsis, the loss of FTSH4 protease, one of the two i-AAA proteases in plant mitochondria, impairs development and leaf morphology at the late stage of rosette growth under short days but not under long days (Gibala et al., 2009; Kicia et al., 2010). These conditional, morphological, and developmental alternations correlated with elevated levels of ROS and carbonylated proteins and were accompanied by ultrastructural changes in mitochondria and chloroplasts (Gibala et al., 2009; Kicia et al., 2010). We also found that FTSH4 is significant for the assembly/stability of complex I and especially complex V (Kolodziejczak et al., 2007). More recently, Zhang et al. (2014a) confirmed our observations linking the lack of FTSH4 protease, oxidative stress, and alternations in plant development and architecture. These authors also revealed additional nonmitochondrial players such as cytoplasmic peroxidases and auxin homeostasis (Zhang et al., 2014a, 2014b). They proposed that FTSH4 mediates the peroxidase-dependent interplay between hydrogen peroxide and auxin homeostasis to regulate plant growth and development.
In this study, we compared the mitochondrial proteome and oxyproteome of the wild type and ftsh4 mutants grown under conditions inducing the phenotypic alterations in ftsh4 (short days at optimal temperature and long days under continuous moderate heat stress) and found that diverse proteins from various submitochondrial compartments were carbonylated in ftsh4 mitochondria, indicating a widespread oxidative stress. We postulate that this oxidative stress proceeds progressively and is associated mainly with the FTSH4 function to maintain a proper phospholipid content, especially CL, in the mitochondrial membrane.
RESULTS
ftsh4 Mutants Show Similar Morphological Changes under Short-Day Photoperiod at Normal Temperature and in Long Days under Continuous Moderate Heat Stress
Aging ftsh4-1 and ftsh4-2 Arabidopsis plants grown in short days (SD) at normal temperature (22°C) were shown previously to display severe morphological and developmental alterations compared with the wild type (Gibala et al., 2009; Kicia et al., 2010). In those studies, we further demonstrated that the postgermination growth of ftsh4 under a long-day photoperiod (LD) at 22°C was not affected. Here, however, we present data indicating that phenotypic abnormalities are noticeable even in LD when ftsh4 plants were grown at slightly higher temperature than optimal (30°C). Phenotypic features seen under both these regimens are a significantly reduced rosette size (Gibala et al., 2009; Fig. 1, A and C), a developmental delay in the appearance of true leaves (Gibala et al., 2009; Fig. 1B), and irregular serration of leaf blades (Gibala et al., 2009; Fig. 1D), the latter two more visible at the end of vegetative growth. These comparable defects suggest similar molecular alterations activated under two specific conditions in the ftsh4 mutant. We denote these conditions as inducing the phenotype.
Morphology of ftsh4 and wild-type (WT) plants growing under LD at 30°C. A, Two-week-old wild-type and ftsh4 plants grown on agar plates. B, Delay time (in days) in leaf emergence of ftsh4-1 and ftsh4-2 compared with the wild type. Plants were staged as described by Boyes et al. (2001). C, Rosette diameter of plants grown in soil at the indicated time points after sowing. Mean values ± sd from three measurements are shown. Significant differences are indicated by asterisks (one-sample Student’s t test; *, P < 0.05). D, Five-week-old rosette leaves of plants grown in soil.
The Mitochondrial Proteome Is Altered in ftsh4 Mutants under Conditions Inducing the Phenotype
To gain insight into molecular alterations activated in ftsh4 mutants in conditions inducing the phenotype, we analyzed their mitochondrial proteomes using two-dimensional fluorescence difference gel electrophoresis (2D-DIGE). Mitochondrial proteins of 9-week-old ftsh4-1 and wild-type plants grown in soil under SD at 22°C as well as 3-week-old ftsh4-1, ftsh4-2, and wild-type seedlings grown hydroponically under LD at 30°C were differentially labeled using G-Dyes (DyeAgnostics) and subjected to isoelectric focusing (IEF)/SDS-PAGE. Thus, three experimental setups were analyzed (SD, 22°C for ftsh4-1; LD, 30°C for ftsh4-1; and LD, 30°C for ftsh4-2), and protein spots with fold changes in abundance of greater than ±1.2 (P ≤ 0.05) between ftsh4 and the wild type in at least one experimental setup were picked from the gel and identified by matrix-assisted laser-desorption ionization time of flight (MALDI-TOF) peptide mass fingerprinting (PMF). Overall, 63 unique and three redundant proteins (SERINE TRANSHYDROXYMETHYLTRANSFERASE1 [SHM1], protein spot numbers 28, 29, and 30; ATP SYNTHASE SUBUNIT 2 [ATP2], protein spot numbers 6, 7, 8, and 9; and HEAT SHOCK PROTEIN60 [HSP60-2], protein spot numbers 38 and 39) with significant Mowse scores were identified and classified into different functional categories (Table I; Supplemental Table S1; Supplemental Fig. S1). Most of these proteins were common to two tested setups, and six were common to all three variants assayed. Four of these common proteins are OXPHOS subunits, and two are associated with the tricarboxylic acid cycle. These proteins are indicated in a representative analytical 2D-DIGE gel shown in Supplemental Figure S2.
Collectively, several subunits of complexes I and V and the majority of enzymes of the tricarboxylic acid cycle were lower in abundance in mitochondria of FTSH4-deficient plants, while chaperones, antioxidant enzymes, and proteins involved in transport accumulated (Table I). A decrease in abundance of the components of OXPHOS and the tricarboxylic acid cycle and induction of the expression of proteins that have an antioxidant and stress-response function are characteristic for mitochondrial proteomes under sublethal doses of oxidative stress (Sweetlove et al., 2002). Thus, the proteomic data indicate that ftsh4 mutants suffer an endogenous oxidative stress. The comparative proteomic analysis also revealed that, in contrast to the several enzymes of amino acid metabolism that are decreased in ftsh4, isovaleryl-CoA dehydrogenase, which is involved in the catabolism of branched-chain amino acids and Lys, accumulated in both ftsh4 mutants in LD at 30°C (Table I). Degradation products of these amino acids can provide electrons to the respiratory chain via the ETF complex (Araújo et al., 2011). One can assume that, in ftsh4 mutants, this catabolic pathway compensates for the reduced electron supply from the tricarboxylic acid cycle at least in LD at 30°C.
Proteins Carbonylated to a Higher Extent in ftsh4 Than in the Wild Type
Extensive carbonylation of mitochondrial proteins was observed in ftsh4 mutants growing under LD at 30°C (Fig. 2B), confirming data on plants grown under SD at 22°C (Gibala et al., 2009). To prove a direct and/or indirect role of FTSH4 in preventing the accumulation of carbonylated proteins, we used a revertant line of ftsh4-1 (ftsh4-1-FTSH4) constitutively expressing full-length FTSH4 cDNA under the control of the cauliflower mosaic virus (CaMV) 35S promoter. As expected, the level of carbonylated proteins in ftsh4-1-FTSH4 was comparable to that in wild-type plants, and the revertant plants had lost the morphological aberrations characteristic for ftsh4-1 seedlings in LD at 30°C (Fig. 2).
A, Morphology of 2-week-old wild-type (WT), ftsh4-1, and ftsh4-1-FTSH4 (revertant) seedlings grown under LD at 30°C on agar plates. B, Carbonylated proteins in a total protein extract of the above genotypes. Immunodetection with anti-DNP antibodies and quantification of carbonylated proteins in total protein extract, separated by one-dimensional gel electrophoresis, are shown. Anti-DNP signals from entire lanes were quantified densitometrically. In each experiment, the values for the relative carbonylated protein amount were calculated as a percentage of the value determined for the wild-type plants (set to 100%). Mean values ± sd from at least three independent experiments are shown. Statistically significant differences in abundance between wild-type, ftsh4-1, and ftsh4-1-FtsH4 plants are indicated by asterisks (one-sample Student’s t test; *, P < 0.05).
To identify the proteins carbonylated in the absence of FTSH4, mitochondrial protein samples isolated from wild-type and ftsh4 plants grown under conditions inducing the phenotype were first separated according to their pI, derivatized in gel with 2,4-dinitrophenylhydrazine, and then resolved by SDS-PAGE. After blotting, the membrane was probed with polyclonal anti-DNP-adduct antibodies (Fig. 3). A protein spot was considered to be significantly increased in carbonylation level if the oxidation fold of its immunological signal in ftsh4 to that on the wild-type gel was 1.2 or greater. Protein spots found to be increasingly carbonylated in ftsh4 were matched to the corresponding spots on an IEF/SDS-PAGE gel stained with Coomassie Blue and identified by mass spectrometry. The carbonylated proteins accumulating in ftsh4-1 under SD at 22°C and in ftsh4-1 or ftsh4-2 under LD at 30°C are listed in Supplemental Table S2.
Comparison of mitochondrial carbonylated proteins from wild-type (WT; left) and ftsh4-1 (right) plants growing in LD at 30°C. Proteins separated by IEF/SDS two-dimensional gel electrophoresis were transferred on polyvinylidene difluoride (PVDF) membranes to subsequently detect carbonylated proteins using the OxyBlot technique. Arrowheads indicate protein spots accumulating in ftsh4 mitochondria, which are identified in all tested setups. Protein spots are listed in Table II and Supplemental Table S2.
In total, 26 mitochondrial proteins were more heavily carbonylated in ftsh4 than in wild-type plants. They represented several functional groups located in different mitochondrial compartments (Supplemental Table S2; Supplemental Fig. S3). Nine of those 26 proteins were identified in all analyzed conditions: two subunits of complex V (ATP1 and ATP2), two components of the tricarboxylic acid cycle (FUMARASE1 [FUM1] and SUCCINYL COENZYME A LIGASE [SuCoA]), three associated with photorespiration (lipoamide dehydrogenase [MTLPD1], aminomethyltransferase [GDC-T], and SHM1), a stress-related protein (MANGANESE SUPEROXIDE DISMUTASE1 [MSD1]), and a protease (MITOCHONDRIAL PROCESSING PEPTIDASE SUBUNIT β [MPPBETA]; Supplemental Table S2). Furthermore, a number of proteins like ACONITASE2 (ACO2), SHM1, GDC-T, ATP1, and ATP2 were found to be more extensively carbonylated in more than one spot on two-dimensional OxyBlots. The different spots of ACO2 and SHM1 showed differences in pI, while those of GDC-T, ATP1, and ATP2 showed differences in Mr; thus, the last three are typical examples of protein degradation products (Supplemental Table S2).
However, the proteins identified as carbonylated more strongly in ftsh4 than in wild-type plants using OxyBlot analysis showed diverse changes in abundance as determined by the 2D-DIGE assay, some being up-regulated and some being down-regulated (Table I; Supplemental Table S1). Therefore, to find proteins carbonylated to a higher extent in ftsh4 than in the wild type, oxidation indexes were determined for each spot of interest. To calculate the oxidation index, the carbonyl immunoreactivity of each spot was divided by the relative protein abundance estimated by the fluorescence difference gel electrophoresis (DIGE) analysis (Supplemental Table S3). Regardless of the conditions tested, ACO2 was found to be more heavily carbonylated in ftsh4 compared with wild-type plants (Table II; Supplemental Table S3). ACO2 represents a group of proteins with a high oxidation index (1 or greater) and a decreased protein amount estimated by DIGE (Table II; Supplemental Table S3). This set of proteins includes complex I (75-kD subunit) and complex V (ATP1, ATP2, and ATP synthase subunit Fad) subunits, tricarboxylic acid cycle components (ACO2, SuCoA subunit β, and MALATE DEHYDROGENASE2), enzymes of Cys (CYSTEINE SYNTHASE C1) and Glu (GLUTAMATE DEHYDROGENASE2 [GDH2]) metabolism, and MPPBETA. Analysis of the transcript levels of these proteins using quantitative PCR (Fig. 4A) indicates that, except for ATP2, other proteins did not show any decrease in mRNA expression. Thus, it is conceivable that these particular proteins are extensively oxidatively damaged and thus degraded in plants lacking FTSH4. In contrast, the remaining proteins with a high oxidation index were found to accumulate in ftsh4 plants according to 2D-DIGE. This group of proteins comprises an enzyme of the tricarboxylic acid cycle (FUM1), four photorespiration enzymes (GDC-T, SHM1, MTLPD1, and formate dehydrogenase), GDH1, and a key enzyme in the elimination of mitochondrial superoxide radicals, MSD1. Quantitative reverse transcription-PCR analyses indicated that the accumulation of these proteins is not accompanied by an increase in their transcript level (Fig. 4A). Most transcripts exhibited statistically insignificant changes in ftsh4-1 and ftsh4-2 compared with the wild type, with a log2 ratio below 0.5. Unexpectedly, the transcript level of GDC-T and SHM1, proteins that accumulate in our experimental conditions in the ftsh4 mutants, decreased substantially (Fig. 4A).
Spot quantification (percentage volume of each spot) was performed on two-dimensional OxyBlots from wild-type and ftsh4 mitochondria using ImageJ Fiji software and on 2D-DIGE gels using DeCyder 2D 6.5 software. The oxidation fold was normalized by the corresponding protein fold to assess the oxidation index.
Relative transcript levels for selected genes encoding mitochondrial proteins in ftsh4-1 and ftsh4-2 mutants growing in LD at 30°C compared with wild-type (WT) plants. A, Levels of transcripts for genes encoding proteins identified by DIGE analysis. B, Levels of transcripts for genes encoding proteins usually up-regulated by oxidative stress. The relative abundance of transcripts is expressed as log2 ratios. Mean values ± sd from at least three independent experiments are shown. The dotted lines indicate cutoff values ± 0.5 (log2) of the ratio corresponding to the threshold levels for significant up- and down-regulation of the transcripts in ftsh4. Full names of genes are given in Table I and Supplemental Table S1. TCA cycle, Tricarboxylic acid cycle.
Increased Expression Level of ATP-Dependent Proteases in ftsh4 Mitochondria
In yeast and mammals, it was shown that carbonylated proteins are removed by ATP-dependent proteases, presumably to avoid the formation of harmful aggregates (Anand et al., 2013). Therefore, we examined the expression level of known mitochondrial ATP-dependent proteases in the ftsh4 mutant compared with wild-type plants. No significant changes were observed in the expression of mitochondrial ATP-dependent proteases in ftsh4 plants grown under LD at 22°C, neither at the transcript nor at the protein level (Fig. 5A). However, under conditions inducing the phenotype, when carbonylated proteins accumulated in mitochondria of ftsh4, we observed a significant increase of FTSH10 protease at the transcript and protein levels and a higher abundance of FTSH3 at the protein level (Fig. 5). The accumulation of LON1 detected by 2D-DIGE agrees with its transcriptional up-regulation (Fig. 5A). The abundance of other plant mitochondrial proteases (FTSH11, LON4, and CLP) because of a lack of corresponding antibodies was estimated only at the transcript level, which did not change significantly in the ftsh4 mutant compared with wild-type plants (Fig. 5A). Taken together, the accumulation of carbonylated proteins under conditions inducing the phenotype in ftsh4 was correlated with elevated expression of at least three ATP-dependent proteases: LON1, FTSH3, and FTSH10.
A, Relative transcript levels for genes encoding mitochondrial ATP-dependent proteases in ftsh4-1 and ftsh4-2 mutants grown in LD at 22°C and LD at 30°C compared with wild-type (WT) plants. The relative abundance of transcripts is expressed as log2 ratios. Mean values ± sd from at least three independent experiments are shown. The dotted lines indicate cutoff values ± 0.5 (log2) of the ratio corresponding to the threshold levels for significant up- and down-regulation of the transcripts in ftsh4. B, Representative images of the immunodetection of selected mitochondrial proteases in the ftsh4-1 mutant compared with wild-type plants growing in LD at 22°C, LD at 30°C, or SD at 22°C. C, Densitometric quantification of immunoblots presented in B. The intensity of bands was estimated using ImageQuant software (Molecular Dynamics). Data for ftsh4-1 are expressed as percentages of the value for wild-type plants. Mean values ± sd from at least three experiments are shown. Significant differences in abundance between the wild type and the ftsh4-1 mutant are indicated by asterisks (one-sample Student’s t test; *, P < 0.05).
ftsh4 Contains a Decreased Abundance and Activity of Complexes I and V as Well as a Reduced Intramitochondrial Pool of ATP
Previously, using a combination of blue-native (BN)-PAGE and histochemical staining, we observed that loss of FTSH4 was associated with a decreased abundance and activity of complexes I and V in plants growing in SD at 22°C (Kolodziejczak et al., 2007). Using the same approach, we observed reduced amounts and activities of complexes I and V in ftsh4-1 and ftsh4-2 mutants growing in LD at 30°C as well (Fig. 6, B and C). Additionally, a slight (approximately 20%) but statistically significant decrease of activity of the analyzed complexes also was detected in the ftsh4 mutants in LD at 22°C (Fig. 6, A and C). Moreover, DIGE analysis confirmed these results: lower amounts of the NADH-ubiquinone oxidoreductase 75-kD subunit belonging to complex I, as well as ATP1 and ATP2 of complex V, were found in the ftsh4 mutants compared with the wild type in all experimental setups (Table I). A clear exception was γ-CARBONIC ANHYDRASE2, a subunit of complex I, which accumulated in the mutants. Thus, two experimental approaches, BN-PAGE and 2D-DIGE, indicate that FTSH4 is essential for the stability/assembly/activity of complexes I and V under conditions inducing the phenotype.
Amounts and activities of respiratory complexes (A–C) and the level of ATP (D) in ftsh4 and wild-type (WT) plants. Mitochondria were isolated from 3-week-old wild-type and mutant plants (ftsh4-1 and ftsh4-2) growing hydroponically in LD at 22°C and LD at 30°C. A and B, Coomassie Brilliant Blue (CBB) and in-gel activity staining of complex I (C I) and complex V (C V) after BN-PAGE. C, Quantification of the activities of complexes I and V. The intensity of bands was estimated by densitometric analysis using ImageQuant software (Molecular Dynamics). Relative complex activity from mutant mitochondria was calculated as a percentage of that in wild-type plants. Differences in activity between the wild type and mutants are in all cases statistically significant (one-sample Student’s t test; P < 0.05). Mean values ± sd from three experiments are shown. D, ATP contents in wild-type and ftsh4 mitochondria. Mitochondria were isolated from 3-week-old wild-type, ftsh4-1, and ftsh4-2 seedlings grown under LD at 30°C. The ATP concentration was determined as described in “Materials and Methods.” An unpaired Student’s t test was used to estimate the P values: *, P < 0.05. Error bars correspond to sd (n = 6).
We next tested whether this defect in the OXPHOS complexes of ftsh4 is associated with a reduced intramitochondrial pool of ATP. Therefore, using a bioluminescence assay, the ATP content was examined in mitochondria obtained from 3-week-old wild-type or ftsh4 plants grown under LD at 30°C. The amount of ATP in mitochondria of ftsh4 was approximately 30% to 40% lower than in the wild type (6.42 ± 2.01 for the wild type, 3.81 ± 1.48 for ftsh4-1, and 4.3 ± 0.72 for ftsh4-2, in pmol mg−1 protein; Fig. 6D). This significantly reduced intramitochondrial pool of ATP in plants lacking FTSH4 protease likely reflects a perturbation in ATP formation due to a defect in the stability/activity of complex V.
The Availability of ATP Limits the Degradation Rate of Carbonylated Proteins in Mitochondria of ftsh4
Given the apparent increase in the abundance of ATP-dependent proteases in mitochondria of ftsh4 grown in the phenotype-inducing conditions, it was surprising that carbonylated proteins were not efficiently removed. Thus, we hypothesized that, in the mutants, carbonylated mitochondrial proteins accumulated due to a lack of ATP, required for the proteolytic activity of the ATP-dependent proteases. To test this hypothesis, we performed in vitro time-course experiments monitoring the level of carbonylated proteins in mitochondria isolated from 2-week-old seedlings grown under LD at 30°C in the absence or presence of ATP (Fig. 7A). Densitometric analyses indicated that the total amount of carbonylated proteins in wild-type mitochondria decreased significantly (about 40%) when incubated without ATP for 16 h at 22°C, while in ftsh4-1 and ftsh4-2 mitochondria, the amount of carbonylated proteins did not change significantly under the same conditions (Fig. 7A). However, in the presence of ATP, the amount of carbonylated proteins was strongly diminished in ftsh4 mitochondria incubated for 16 h at 22°C, practically exceeding the level observed in the wild-type sample (Fig. 7A).
In vitro carbonylated protein degradation in wild-type (WT) and ftsh4 mitochondria. Immunodetection of carbonylated proteins separated by one-dimensional gel electrophoresis was estimated with anti-DNP antibodies. Anti-DNP signals from entire lanes were quantified densitometrically. Mean values ± sd from three experiments are shown. A, Mitochondria were isolated from 2-week-old wild-type, ftsh4-1, and ftsh4-2 seedlings grown under LD at 30°C and incubated at 22°C for 16 h in the absence or presence of 3.5 mm ATP. An unpaired Student’s t test was used to estimate the P values: *, P < 0.05. B, Mitochondria were isolated from 2-week-old wild-type, ftsh4-1, and ftsh4-2 seedlings grown under LD at 22°C and incubated at 22°C for 16 h in the presence of 5 mm succinate and 8 μm antimycin A, with or without of 3.5 mm ATP. For the protease inhibitor assay, inhibitors of Ser proteases (2 mm AEBSF) and metalloproteases (25 mm ortho-phenanthroline [O-Phe]) were added to the incubation medium. Mean values ± sd from three experiments are shown. An unpaired Student’s t test was used to estimate the P values: *, P < 0.05.
The same assay was performed on freshly isolated mitochondria from wild-type and ftsh4 seedlings grown in LD at 22°C, incubated with succinate and antimycin A, an inhibitor of complex III and a well-known inducer of oxidative stress (Fig. 7B). Treatment of ftsh4 mitochondria with antimycin A for 16 h in the absence of ATP resulted in a significant accumulation of carbonylated proteins, while supplementing the incubation medium with ATP induced a strong decrease of oxidatively modified proteins (approximately 60%; Fig. 7B). The addition of AEBSF, an inhibitor of Ser proteases, to the medium prevented the carbonylated protein content from decreasing in the wild type and ftsh4-1, indicating an important contribution of Ser proteases in eliminating oxidatively modified proteins in plant mitochondria (Fig. 7B). No such effect could be observed using ortho-phenanthroline, an inhibitor of metalloproteases (Fig. 7B).
ftsh4 Contains Altered Mitochondrial Morphology and Phospholipid Content
Several recent findings point to an essential role of yeast and mammalian i-AAA proteases in the mitochondrial QC at the organellar level (Stiburek et al., 2012; Li et al., 2015; Qi et al., 2016). To look into a putative role of FTSH4 in plant mitochondrial dynamics, we first examined the mitochondrial morphology of wild-type and ftsh4-1 plants grown under LD at 22°C (no visible phenotype) and LD at 30°C (visible phenotype) transformed with a construct expressing mitochondria-targeted GFP under the control of the CaMV 35S promoter. Analysis of confocal microscopy photographs revealed the existence of a heterogenous mitochondrial population in ftsh4-1 protoplasts compared with the wild type (Fig. 8A). This heterogenous population was characterized by the appearance of enlarged, spherical mitochondria, termed giant mitochondria, among healthy, oval forms (Fig. 8A). While the giant mitochondria were rare in ftsh4-1 protoplasts isolated from plants grown at 22°C, their number and area were highly increased when the plants were exposed to the moderately elevated temperature of 30°C (Fig. 8A). In addition, some of giant mitochondria displayed GFP voids, with reduced or absent fluorescence signal, the feature characteristic for oxidative stress (Fig. 8A; Logan, 2003). Even mitochondrial enlargement was reported to be caused by oxidative stress (Scott and Logan, 2008), but also by depletion of CL, a signature phospholipid of mitochondria (Pineau et al., 2013; Pan et al., 2014). To investigate if the aberrant mitochondrial morphology observed in ftsh4 plants is associated with perturbed CL abundance, we determined the lipid composition of mitochondrial membranes in the wild type and ftsh4 mutants grown under LD at 22°C and LD at 30°C using a comparative quantitative lipidomic approach (shotgun lipidomics; Fig. 8B). Figure 8B shows the quantification of CL and lipids involved in this phospholipid biosynthesis (phosphatidic acid, DAG, PI, and PG) as well as the most abundant lipids of the mitochondrial membrane (phosphatidylcholine and PE). The quantitative analysis of all identified lipid classes is presented in Supplemental Figure S4. Our results indicate that the level of CL was decreased slightly in mitochondria of ftsh4 plants grown under LD at 22°C. This effect was more pronounced under LD at 30°C compared with the wild type (Fig. 8B). Additionally, we detected an accumulation of PG in ftsh4 mitochondria under LD at 22°C, which was stronger under LD at 30°C. Furthermore, the levels of PI and DAG in ftsh4 were increased slightly (Fig. 8B). It should be emphasized that DAG, PI, and PG are involved in the CL biosynthesis pathway, and possibly their accumulation in ftsh4 mitochondria is related to the deficiency of CL; similar changes of the PG level were observed in an Arabidopsis mutant disrupted in the single-copy gene encoding CL synthase (Pan et al., 2014). While no significant difference was observed in the amount of PE in the wild type and ftsh4 grown under LD at 22°C, ftsh4 plants grown under LD at 30°C had lower levels of PE in their mitochondrial membranes (Fig. 8B). This effect might reflect a complex regulation of the phospholipid level, since both CL and PE are known to be involved in similar processes in mitochondrial membranes (Böttinger et al., 2012; Joshi et al., 2012). Taken together, these results demonstrate that FTSH4 protease affects the abundance of CL and, in turn, mitochondrial dynamics.
Mitochondrial morphology and class of lipids involved in CL biosynthesis, differing in ftsh4. A, Mitochondrial morphology was observed by scanning single protoplasts expressing GFP targeted to mitochondria of the wild type (WT) and ftsh4-1 using a Zeiss LSM 510 Meta confocal microscope. Giant mitochondria are highlighted with white arrows. The star indicates a spherical mitochondrion displaying reduced GFP fluorescence, pointing out an occurrence of oxidative stress. Bar = 5 µm. B, Lipids of the wild type, ftsh4-1, and ftsh4-2 grown under optimal (22°C) and moderately elevated (30°C) temperatures were measured by mass spectrometry. Lipid classes are shown in mol % of total lipids in the sample and are sums of individually quantified lipid species. An unpaired Student’s t test was used to calculate the P values: *, P < 0.05. Error bars correspond to sd (n = 3). DAG, Diacylglycerol; PA, phosphatidic acid; PC, phosphatidylcholine; PG, phosphatidylglycerol; PI, phosphatidylinositol. Only selected classes are shown; for the complete data set, see Supplemental Figure S4.
DISCUSSION
In this study, we show that loss of the FTSH4 mitochondrial protease under conditions inducing the phenotype leads to oxidative damage of many mitochondrial proteins different in function and submitochondrial localization. We believe that the main cause of this phenotype is altered content of CL in the mitochondrial membrane, associated with perturbation of at least two processes: functionality of the OXPHOS system and mitochondrial dynamics. Deregulation of these processes disturbs the mitochondrial QC on the molecular and organellar levels and leads to the accumulation of oxidatively damaged proteins (Fig. 9). In other words, we argue here that the accumulation of carbonylated proteins found in ftsh4 plants is due to their more intensive generation, an ineffective system for their removal by ATP-dependent proteases, and an altered system of mitochondrial fusion/fission/mitophagy.
Hypothetical scenario of the events leading to the accumulation of carbonylated proteins in mitochondria of Arabidopsis in the absence of FTSH4. The direction of short green arrows specifies change in the abundance of CL, complex I (CI), complex V (CV), ROS, ATP, and carbonylated proteins in the absence of FTSH4. Black lines ending with arrowheads indicate activating effects, while perpendicular lines indicate inhibiting effects. The decreased content of CL as well as the absence of the chaperone-like activity of FTSH4 lead to lower stability/activity of complexes I and V, which in turn results in the accumulation of ROS and the decrease of ATP, respectively. The lower concentration of ATP restricts the activity of mitochondrial ATP-dependent proteases, which are not able to degrade all carbonylated proteins accumulating as a result of the elevated ROS level. The lower content of CL further restricts fission, which in turn causes the appearance of giant mitochondria and also blocks mitophagy, which is important to eliminate mitochondria damaged by oxidative stress. IMM, Inner mitochondrial membrane; IMS, intermembrane space of mitochondria; OMM, mitochondrial outer membrane; PHB, prohibitin.
Diverse Mitochondrial Proteins Undergo Carbonylation in ftsh4 Grown Under Conditions Inducing the Phenotype
We previously reported that aging ftsh4 plants growing in SD at 22°C experienced oxidative stress (Gibala et al., 2009; Kicia et al., 2010). We have now extended those studies by identifying the proteins to be excessively carbonylated in ftsh4 and linking their oxidation status with their abundance estimated by 2D-DIGE (Tables I and II). These experiments were not limited to ftsh4 growing under SD at 22°C; similar analysis also was performed for ftsh4-1 and ftsh4-2 growing under LD at 30°C. Diverse proteins from the OXPHOS system, tricarboxylic acid cycle, and photorespiration, as well as manganese superoxide dismutase and Cys synthase, were found to be more excessively carbonylated in the mutant than in wild-type mitochondria (Table II). It should be emphasized that all the oxidatively damaged proteins identified in ftsh4 have been reported earlier as targets for carbonylation in plant mitochondria (Kristensen et al., 2004; Smakowska et al., 2014). Those proteins, which were carbonylated predominantly in the absence of FTSH4, also were identified by 2D-DIGE analyses and found to be differentially expressed compared with the wild-type mitochondrial proteome (Table I). A decrease in abundance was observed for two enzymes of the tricarboxylic acid cycle (ACO2 and SuCoA), two subunits of complex V (ATP1 and ATP2), and the 75-kD subunit of complex I, while fumarase and two enzymes of photorespiration (GDC-T and MTLPD1) showed increased abundance (Tables I and II). The lower abundance and enhanced carbonylation of complex I and V subunits in mitochondria from ftsh4 presumably were caused by a defect in their assembly/stability (Kolodziejczak et al., 2007). This defect may increase the pool of unassembled/misfolded proteins, which will undergo carbonylation and subsequent degradation. We believe that, in ftsh4, the breakdown of ATP1 and ATP2, ACO2, and SuCoA occurs to a certain degree after oxidation, particularly since the level of their transcripts did not decrease significantly (Fig. 4).
The question arises why the carbonylated enzymes FUM1, GDC-T, and MTLPD1, in contrast to carbonylated ATP2 or ACO2, are present in higher abundance in ftsh4 (Table II). In light of the report showing that chloroplast FTSH proteases degrade functionally assembled proteins that have undergone oxidation (Lindahl et al., 2000), it seems likely that FTSH4 fulfills a similar function in mitochondria and its loss causes the accumulation of its substrates, a specific set of highly carbonylated proteins. However, there is a topological incompatibility: the catalytic domain of FTSH4 protease faces the intermembrane space, while GDC-T, MTLPD1, and FUM1 are matrix proteins. At this stage, we do not know how to explain this observation.
ftsh4 Displays Altered Metabolic Pathways as a Response to Oxidative Stress
The ftsh4 mutant seems to provide protection against the oxidizing conditions in mitochondria at several levels. Induction of nonphosphorylating alternative pathways is indicated by increased transcription of the genes encoding AOX1A, NDB2, and NDB4 (Fig. 4B). These pathways are known to play an essential role in the response of plant mitochondria to oxidative stress mainly by preventing the overreduction of the mitochondrial electron transport chain and thus the production of ROS (Moller, 2001; Vanlerberghe, 2013). Furthermore, the obvious up-regulation of isovaleryl-CoA dehydrogenase at the transcript and protein levels (Fig. 4A; Table I) suggests another nonclassical entry route of electrons to the respiratory chain to be activated, provided by the catabolism of branched-chain and aromatic amino acids (Araújo et al., 2010). This makes sense in the light of the elevated amino acid pool generated by the degradation of carbonylated proteins in the ftsh4 mutant. The activation of this pathway by oxidative stress was reported earlier by Lehmann et al. (2009). Using 2D-DIGE, we further confirmed our earlier findings (Gibala et al., 2009) that mitochondrial chaperones accumulate in the ftsh4 mutant (Table I). The up-regulation of HSP70 and HSP60 chaperones on transcript and protein levels in ftsh4 is consistent with their protective roles against the aggregation of mitochondrial proteins under oxidative stress conditions (Table I; Fig. 4A; Bender et al., 2011).
In ftsh4, the Defective OXPHOS System Is Linked to Enhanced Generation of ROS and Carbonylated Proteins
The increase in the steady-state level of ROS observed in ftsh4 plants under conditions inducing the phenotype (Gibala et al., 2009; Kicia et al., 2010) leads to the generation of carbonylated proteins. The trigger of the oxidative stress observed in ftsh4 most likely is the dysfunctional assembly/stability/activity of complexes I and V under conditions inducing the phenotype (Kolodziejczak et al., 2007; Fig. 6, B and C). Under optimal conditions (LD at 22°C), these alterations were mild (Fig. 6A), but with time (late phase of vegetative growth in SD at 22°C) or under conditions with a higher probability of electron leakage like moderate heat stress (LD at 30°C), harmful overproduction of ROS occurs. Defects in the OXPHOS system caused by the absence of FTSH4 could be explained by at least two not mutually excluding mechanisms: (1) the lack of chaperone-like activity of FTSH4 protease required for the formation/stability of complexes I and V, and (2) the low level of membrane CL documented in this study (Fig. 8). It is well known that CL is a structural component of many protein complexes and supercomplexes of the inner mitochondrial membrane in yeast, human, and plant mitochondria (Pfeiffer et al., 2003; Petrosillo et al., 2007; Gonzalvez et al., 2013; Pineau et al., 2013). Reduced content of the respiratory complex I/complex III supercomplex and, to a lesser extent, complex I was found in an Arabidopsis knockout mutant depleted in the final enzyme of CL biosynthesis (Pineau et al., 2013).
The results presented here also suggest that extended oxidative stress in ftsh4 creates other additional sources of ROS, which further accelerate this stress. Oxidatively damaged iron-sulfur proteins like aconitase or the 75-kD subunit of complex I could accelerate oxidative stress by the Fenton reaction (Møller et al., 2011). The finding of several mitochondrial enzymes of photorespiration being highly carbonylated (GDC-T, SHM1, and MTLPD1) combined with the decreased abundance of other photorespiration enzymes (GLDP1 and GLDP2) suggests that this process is at least slowed down in ftsh4 mitochondria (Tables I and II). Since photorespiration is considered to be important to prevent ROS accumulation (Voss et al., 2013), the postulated decreased activity of the GDH/SHMT complex could be an additional cause for oxidative stress in ftsh4. Furthermore, among the proteins carbonylated preferentially in ftsh4 is the manganese superoxide dismutase, a key enzyme for eliminating mitochondrial superoxide radicals (Wang et al., 2010). This enzyme is probably inactive in ftsh4 due to oxidative modification (Qin et al., 2009), and it was found to be highly carbonylated in all the conditions examined (Table II).
In ftsh4, the Defective OXPHOS System Is Linked to Limited ATP-Dependent Degradation of Carbonylated Proteins
The protease responsible for degrading oxidatively damaged matrix proteins in yeast (Bayot et al., 2010; Bender et al., 2011) and mammalian mitochondria (Bota and Davies, 2002) is the ATP-dependent Lon protease. However, recent studies in Arabidopsis indicate the mitochondrial LON1 protease to be nonessential for the turnover of oxidized proteins (Solheim et al., 2012). Nevertheless, in the absence of FTSH4, we observed an overexpression of mitochondrial ATP-dependent proteases, potentially able to degrade oxidized proteins, at the transcriptional (LON1 and FTSH10) and protein (FTSH3) levels (Fig. 5). Curiously, despite this increased expression of ATP-dependent proteases, we documented that the degradation of carbonylated proteins in mitochondria isolated from ftsh4 was less efficient than in the wild type in the absence of ATP but similar to the wild type after the addition of ATP (Fig. 7). Using appropriate inhibitors, we could support a previous study (Sweetlove et al., 2002) showing Ser proteases to be mainly responsible for the degradation of oxidatively modified proteins in plant mitochondria (Fig. 7). Thus, the accumulation of carbonylated proteins in ftsh4 seems to be caused by limited ATP-dependent proteolytic capacity in ftsh4 mitochondria. The lower ATP content found in ftsh4 mitochondria compared with the wild type (Fig. 6D) probably results from an impaired efficiency of the OXPHOS system, particularly the substantially decreased activity of the ATP synthase.
A Decreased Level of CL induces Giant Mitochondria and the Accumulation of Carbonylated Proteins in ftsh4
One of the most exciting conclusions rationalizing the accumulation of carbonylated proteins in ftsh4 is based on the observation of giant mitochondria and on CL deficiency in the mitochondrial membrane in the absence of FTSH4 (Fig. 8). In Arabidopsis, CL deficiency has been linked to restricted fission, which in turn causes the appearance of giant mitochondria (Pan et al., 2014). Ongoing cycles of fusion and fission of mitochondrial membranes are essential for an efficient defense against mitochondrial damage (Tatsuta and Langer, 2008). Furthermore, in animals, it was documented that giant mitochondria block mitophagy, the elimination of damaged mitochondria (Zhang et al., 2014c). Thus, we postulate that, in mitochondria of ftsh4, carbonylated proteins are not removed efficiently, not only because of the limited capacity of the ATP-dependent proteolytic system but also because of a restricted fission/mitophagy (Fig. 9).
It is well documented that the proteolytic activity of FTSH4 homologs in yeast and mammals controls the accumulation of key mitochondrial phospholipids by turnover of the phospholipid regulators (Potting et al., 2010, 2013). In yeast, the proteins Ups1 and Ups2 in the intermembrane space act as central regulators of mitochondrial phospholipid homeostasis. The turnover of Ups2 is mediated by the i-AAA protease Yme1, whereas Ups1 is degraded by Yme1 and the metallopeptidase Atp23. Our results point to FTSH4 being part of a regulatory pathway that influences the abundance of CL in plant mitochondria. It seems that this pathway is similar at least to some extent to that discovered in yeast, given that, in the Arabidopsis genome, there is a homolog of the Ups1 protein (At5g13070), while a homolog of Ups2 has not yet been found.
CONCLUSION
Overall, the results presented here indicate that the plant protease FTSH4 suppresses oxidative damage in mitochondrial proteins indirectly by controlling the abundance of CL, a key phospholipid within the mitochondrial membrane. The availability of CL has been shown previously to stabilize respiratory complexes and mitochondrial dynamics. Our findings indicate that both of these processes are impaired in mitochondria lacking FTSH4 and that, in consequence, plants suffer oxidative stress. Although we have been unable to identify carbonylated proteins as direct substrates for FTSH4, we believe that this protease could functionally resemble chloroplastic FTSH in this respect (Lindahl et al., 2000) but in different developmental stages and/or environmental conditions than used in our work.
MATERIALS AND METHODS
Plant Material and Growth Conditions
Arabidopsis (Arabidopsis thaliana) wild-type and T-DNA insertion lines were of the Columbia-0 ecotype. The transgenic lines ftsh4-1 (SALK_035107/TAIR) and ftsh4-2 (GABI_103H09/TAIR) were obtained from the Salk Institute and the Max Planck Institute for Breeding Research, respectively. The lines were characterized previously by Gibala et al. (2009). Plants were grown in growth chambers in a 16-h-light/8-h-dark (LD) photoperiod at 22°C and 30°C for 2 or 3 weeks (agar plates or hydroponic culture) as well as in soil in an 8-h-light/16-h-dark (SD) photoperiod at 22°C for 9 weeks and for 4 weeks in LD at 22°C or 30°C, with a light intensity of 150 μmol m−2 s−1.
Isolation of Mitochondria
Isolation of mitochondria from 9-week-old rosettes was performed as described by Urantowka et al. (2005) and from the seedlings as described by Day et al. (1985). Purified mitochondrial fractions were resuspended in a final wash buffer at a concentration of 10 to 20 mg mitochondrial protein mL−1. The protein concentration was determined using the DC Protein Assays (Bio-Rad).
Protein Leaf Extracts
Total protein extract was obtained from 3-week-old hydroponic cultures as described by Martínez-García et al. (1999). Protein concentration was determined using the DC Protein Assay.
Immunoblot Analysis
Equal amounts of proteins from wild-type and ftsh4 plants (25 μg per line) were separated by SDS-PAGE according to Laemmli (1970). After electrophoresis, proteins were transferred to PVDF membranes and immunostained with appropriate antibodies. The antibodies used were purchased from Agrisera: anti-AtFTSH10 (AS07 251) and anti-AtFTSH3 (AS07 204; Piechota et al., 2010). Immunodetection was performed using the WesternBright Quantum Western Blotting Detection Kit (Advansta). The membranes were documented using a chemiluminescence imager (G-BOX ChemiXR5; Syngene), and the optical density of the bands was quantified using ImageQuant software (Molecular Dynamics).
BN-PAGE and Catalytic Staining
BN-PAGE was performed as described by Kolodziejczak et al. (2007). The catalytic staining of mitochondrial complex I and supercomplex I + III2 was carried out according to Zerbetto et al. (1997), while the staining of complex V was according to van Lis et al. (2003).
2D-DIGE
For 2D-DIGE analysis, mitochondrial proteins from the wild type and mutants (ftsh4-1 and ftsh4-2) were precipitated with cold acetone for 2 h at −20°C and pelleted at 18,000g and 4°C for 30 min. Mitochondrial protein pellets were resuspended in lysis buffer (8 m urea, 4% [w/v] CHAPS, 50 mm dithiothreitol [DTT], and 40 mm Tris-HCl, pH 7.5) at a concentration of 10 mg mitochondrial protein mL−1 and vortexed for 30 min at 4°C. The nonsoluble material was removed by centrifugation at 18,000g for 20 min. Each 25-μg biological replicate sample was labeled with 0.1 nmol of G-Dyes (G-200 or G-300; DyeAgnostics) on ice for 30 min in the dark according to the manufacturer’s instructions. The reaction was stopped with 1 nmol of Lys. A G-100-stained sample (25 μg of mitochondrial proteins), composed of equal amounts of wild-type and ftsh4-1 or ftsh4-2 mitochondrial protein extracts, constituted an internal standard used for the normalization of 2D-DIGE gels. Aliquots of 25 μg of each differentially labeled mitochondrial sample (G-100-, G-200-, and G-300-labeled samples) were pooled, mixed with rehydration solution (7 m urea, 2 m thiourea, 2% CHAPS, 20 mm DTT, and 0.6% IPG buffer 3-11 NL [GE Healthcare]), and loaded on a 24-cm, pH 3 to 11 NL Immobiline DryStrip (GE Healthcare). Rehydration of the strips and first-dimension IEF were conducted in an Ettan IPGphor isoelectric focusing system (GE Healthcare) at 50 V for 12 h (rehydration), 500 V for 3 h (step), 2,000 V for 2 h (step), 8,000 V for 1 h (gradient), and 8,000 V for 10 h (step) at a maximum setting of 50 μA per strip. After IEF, the strips were equilibrated for 15 min in the equilibration buffer (6 m urea, 30% glycerol, 50 mm Tris-HCl [pH 8.8], 2% [w/v] SDS, and trace of Bromophenol Blue) supplemented with 65 mm DTT and then for a further 15 min in the same buffer with 135 mm iodoacetamide. The strips were laid on top of 12.5% (w/v) polyacrylamide gels (26 × 20 cm) and sealed with 0.5% agarose in 25 mm Tris-HCl, 192 mm Gly, and 0.2% SDS. Second-dimension electrophoresis was performed overnight in an Ettan Dalt II electrophoresis unit (GE Healthcare) at 20°C and 1 W per gel. After electrophoresis, the gels were scanned with a Typhoon 9400 scanner (GE Healthcare) at the excitation wavelengths corresponding to each of the G-Dyes. Images of the gels from three independent biological repetitions were analyzed using DeCyder software version 6.5 (GE Healthcare) for normalization and statistical analysis. Protein spots that showed a significant difference in abundance (fold difference of ±1.2, P < 0.05) between ftsh4 and the wild type were picked from the gel and identified by MALDI-TOF PMF. In this approach, three experimental setups were analyzed (SD at 22°C for ftsh4-1, LD at 30°C for ftsh4-1, and LD at 30°C for ftsh4-2) in three independent biological replicates.
Detection of Carbonylated Proteins on One-Dimensional Gels
Carbonylated proteins were detected and analyzed following derivatization of protein carbonyl groups with 2,4-dinitrophenylhydrazine using the OxyBlot kit (Millipore). Immunodetection was performed with a primary antibody directed against dinitrophenylhydrazone using 25 μg of total protein extract or 30 μg of mitochondrial proteins from ftsh4 and wild-type plants per lane. The carbonylated proteins were visualized with the WesternBright Quantum Western Blotting Detection Kit.
Two-Dimensional OxyBlots
Mitochondrial protein extracts (150–200 μg) of wild-type, ftsh4-1, and ftsh4-2 lines were prepared as described above. After IEF, the strips were frozen at −80°C for 2 h. Derivatization of protein carbonyl groups was done by incubation of the strips in a buffer containing 2 m HCl and 10 mm 2,4-dinitrophenylhydrazine for 20 min at room temperature. Derivatized proteins were neutralized by repeated incubation of the strips in a solution of 2 m Tris base and 30% (v/v) glycerol for 20 min at room temperature. Afterward, the strips were equilibrated and proteins were resolved on second-dimension gels as described above. After electrophoresis, proteins were electrotransferred onto a PVDF membrane (Bio-Rad) and subjected to immunodetection of carbonyl groups using the OxyBlot kit (Millipore). Immunodetection was performed with a primary antibody directed against dinitrophenylhydrazone and the Amersham ECL Prime Western Blotting Detection System (GE Healthcare).
Quantitative Analysis of Two-Dimensional OxyBlots
Protein spots on two-dimensional OxyBlots were quantified using ImageJ Fiji software (Fiji). For different experimental setups and types of sample (wild type, ftsh4-1, and ftsh4- 2), each spot was quantified and normalized to the total background intensity of each OxyBlot. Then, the ftsh4/wild type oxidation ratio was calculated for each spot. sd was calculated and statistical significance was assessed using Student’s t test for three independent experiments (n = 3) for each experimental setup (ftsh4-1/wild type, LD at 30°C; ftsh4-2/wild type, LD at 30°C; and ftsh4-1/wild type, SD at 22°C). Oxidation indexes were calculated as the ratio between oxidation fold and protein abundance fold obtained by 2D-DIGE for each spot of interest.
Mass Spectrometry and Protein Identification
For protein identification, 150-μg protein samples consisting of equal amounts of wild type and ftsh4 mitochondrial protein extracts were run on preparative SDS gels. The gels were stained with colloidal Coomassie Blue, and protein spots showing significant differences in abundance between samples were manually or automatically excised from the gels with an Ettan Spot Picker (GE Healthcare). Gel pieces were washed briefly with deionized water and incubated for 1 h in a solution of 35% acetonitrile (ACN) and 20 mm NH4HCO3. Two further washes (each 10 min) were performed with 100% ACN to dehydrate the gels. In-gel protein digestion was performed with 5 ng μL−1 trypsin (Promega V5111) in 20 mm NH4HCO3/10% ACN overnight at 37°C. The peptides were extracted with 1% trifluoroacetic acid for 30 min at room temperature and occasional shaking. A volume of 1 μL of peptide mixture was spotted on an ABI-PerSpective Voyager DE STR MALDI plate (Applied Biosystems) covered with 1 μL of α-cyano-4-hydroxycinnamic acid as a matrix in a solution of 50% ACN and 0.1% trifluoroacetic acid. PMFs of the samples were acquired using a Voyager-DE STR MALDI-TOF (Applied Biosystems) mass spectrometer. A peptide calibration standard mixture in the mass range 800 to 4,000 D (Sequazyme; Applied Biosystems Siex) was used for calibration of the mass spectrometer. The PMFs were searched against TAIR. The fixed and variable modifications were Cys carbamidomethylation and Met oxidation, respectively, with a maximum number of missed cleavages of 2. Peptide mass precision tolerance error was set to 100 ppm. Proteins were identified with Voyager 5 Software (Applied Biosystems) using the Mascot 2.3 search engine (Matrix Science; www.matrixscience.com).
RNA Isolation and cDNA Synthesis
Total RNA was isolated from 3-week-old hydroponically grown seedlings and rosettes of plants grown in soil for approximately 9 weeks using the GeneMATRIX Universal RNA Purification Kit (EURx). The reverse transcription reaction was performed using up to 2 μg of total RNA and a reverse transcription kit (Applied Biosystems). The resulting cDNA was used as a template for quantitative real-time PCR.
Real-Time PCR Analysis
Real-time PCR analyses were performed in a LightCycler 2.0 instrument (Roche Applied Science). Real-Time 2x PCR Master Mix SYBR version B (A&A Biotechnology) was used. Reactions were carried out in a total volume of 15 μL with a final concentration of 0.5 μm primers. Material from wild-type plants served as the calibrator, and the PP2AA3 gene (At1g13320) was used as a reference. The amplification efficiency values of the analyzed amplicons were calculated based on standard curves generated for serial 2-fold dilutions of the cDNA samples. The amplification protocol comprised denaturation at 95°C for 1 min; amplification, 45 cycles at 95°C for 10 s, 55°C to 65°C (the annealing temperature was specific for the primers used) for 10 s, and 72°C for 20 s with single data acquisition; and cooling at 40°C for 30 s. The specificity of the amplification products was verified by melting curve analysis. The primers used are listed in Supplemental Table S4.
Plasmid Construction
The original FTSH4 cDNA was cloned in pTZ57 R/T vector (Thermo Fisher Scientific), sequenced, and recloned into pENTRTM/D-TOPO (Thermo Fisher Scientific) using the primers listed in Supplemental Table S5. In the second step of cloning by the Gateway method, FTSH4 was introduced under the control of the CaMV 35S promoter to the destination vector pGWB514 (a kind gift from Dr. Tsuyoshi Nakagawa; Nakagawa et al., 2007) containing the hemagglutinin tag at the C terminus. The final construct was subjected to complementation analysis.
Agrobacterium tumefaciens-Mediated Transformation of Arabidopsis
The pGWB514 plasmid was introduced into A. tumefaciens strain LBA4404 by electroporation. The bacterial strain obtained was used for floral dip vacuum infiltration of the ftsh4-1 mutants as described by Desfeux et al. (2000). Transformants were checked for complementation of the developmental defects under appropriate conditions.
Lipidomic Analysis
Lipid extraction, mass spectrometric analysis, and data analysis were done by Lipotype. For shotgun lipidomics, lipids were extracted with chloroform and methanol from mitochondria isolated from wild-type, ftsh4-1, and ftsh4-2 2-week-old seedlings or leaf homogenates as described by Sampaio et al. (2011). Samples were spiked with known amounts of lipid class-specific internal standards prior to extraction, and lipid extracts were subjected to mass spectrometric analysis. Mass spectra were acquired on a hybrid quadrupole/Orbitrap mass spectrometer (Q-Exactive; ThermoFisher Scientific) equipped with an automated nano-flow electrospray ion source (Triversa Nanomate; Advion) in both positive and negative ion modes. Lipid identification using Lipotype Xplorer was performed as described previously (Herzog et al., 2011, 2012). Precursor ion intensity values from mass spectra were normalized to intensities of their respective internal standards to obtain pmol values. These values were converted to mol % (mole fraction) to show the stoichiometric relationship between lipids.
Plant Protoplast Isolation and Confocal Imaging
Protoplasts were isolated from leaves of 4-week-old wild-type and ftsh4-1 plants grown in soil under LD at 22°C and LD at 30°C and expressing mitochondria-targeted GFP under the control of the CaMV 35S promoter, using the method described by Yoo et al. (2007). In order to target GFP into mitochondria, the N-terminal targeting sequence from the gene encoding the mitochondrial F1F0 ATP synthase δ-subunit was used (Sakamoto and Hoshino, 2004). The images of protoplasts were acquired using a Zeiss LSM510 confocal laser scanning microscope equipped with a 40× water-immersion objective. The GFP signal was visualized with a 488-nm argon laser and a BP505-530 filter.
In Vitro Mitochondrial Protein Degradation Assay
To test the rate of degradation of carbonylated proteins in mitochondria of the wild type and ftsh4 mutants, freshly isolated mitochondria (30 µg) from 2-week-old seedlings grown in LD at 30°C on agar plates were resuspended in a washing buffer (0.3 m Suc and 10 mm TES, pH 7.5) and incubated for 16 h at 22°C with or without 3.5 mm ATP. In order to induce an oxidative stress, mitochondria obtained from seedlings grown in LD at 22°C were resuspended in a washing buffer containing 8 µm antimycin A and 5 mm succinate and incubated for 16 h at 22°C in the presence or absence of 3.5 mm ATP. For protease inhibitor assays, inhibitors of Ser proteases (2 mm AEBSF; Sigma) and metalloproteases (25 mm ortho-phenanthroline; Sigma) were added into the incubation medium, and the mitochondrial protein samples were incubated under the same conditions. After the incubation, the samples were centrifuged for 10 min at 21,000g and 4°C, and the mitochondrial pellets were further analyzed for protein carbonylation by immunodetection with anti-DNP antibodies using the OxyBlot kit (Millipore). The densitometry analyses of the carbonylated proteins were performed using ImageJ Fiji software (Fiji).
Measurement of Mitochondrial ATP
The ATP amounts in mitochondria of wild-type and ftsh4 plants were calculated from an ATP standard curve using the ATP Determination Kit (Thermo Fisher Scientific) following the manufacturer’s instructions. The measured mitochondrial ATP content was expressed as pmol mg−1 mitochondrial protein. Mitochondria were isolated in sterile conditions from 3-week-old seedlings grown in LD at 30°C. Statistical analysis of differences between the ATP amounts was performed by unpaired two-tailed Student’s t test.
Accession Numbers
Sequence data from this article can be found in the GenBank/EMBL data libraries under the accession numbers in Tables I and II and Supplemental Tables S1 to S4.
Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Proportion, number, and functional categories of mitochondrial proteins differing in abundance between ftsh4 and wild-type Arabidopsis plants growing under different conditions inducing the phenotype.
Supplemental Figure S2. Representative differential 2D IEF/SDS-PAGE gel of ftsh4-1 versus wild-type mitochondrial proteins from plants grown in LD at 30°C.
Supplemental Figure S3. Proportion, number, and functional categories of identified mitochondrial carbonylated proteins in the ftsh4 mutants.
Supplemental Figure S4. Total lipid contents and classes in mitochondria of the wild type, ftsh4-1, and ftsh4-2 grown under optimal (22°C) and moderately elevated (30°C) temperatures determined by mass spectrometry (shotgun lipidomics).
Supplemental Table S1. Identification of proteins differentially abundant in ftsh4-1 and ftsh4-2 in comparison with the wild type in 2D-DIGE.
Supplemental Table S2. Identification of carbonylated proteins accumulating in ftsh4-1 and ftsh4-2 in comparison with the wild type in two-dimensional OxyBlot analysis.
Supplemental Table S3. Protein abundance, oxidation folds, and oxidation indexes estimated for all tested experimental setups (LD at 30°C for ftsh4-1 and ftsh4-2 and SD at 22°C for ftsh4-1).
Supplemental Table S4. Primers used for quantitative reverse transcription-PCR.
Supplemental Table S5. Primers used for plasmid construction, cloning, and genotyping.
Acknowledgments
We thank Dr. Harald Aigner (Department of Chemistry, Umeå University) for help in designing and performing 2D-DIGE analyses, Dr. Michal Surma (Lipotype) for help, advice, and discussion concerning lipidomic analysis, and the Proteomics Facility of the Chemical Biological Centre of Umeå University for provided the facilities to perform the 2D-DIGE experiment and the MALDI-TOF mass spectrometry analyses.
Footnotes
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Hanna Janska (janska{at}ibmb.uni.wroc.pl).
H.J. and C.F. designed the research; E.S., R.S.-B., M.C., M.K., M.K.-O., and K.P. performed the experiments; H.J., E.S., M.C., R.S.-B., M.K., M.K.-O., and K.P. analyzed the data; H.J. wrote the article with input of all coauthors.
↵1 This work was supported by the National Science Centre, Poland (grant nos. 2011/03/N/NZ2/00221 and 2013/11/N/NZ3/00061), the Swedish Energy Agency (grant no. 2012–005889), Umeå University, the Foundation for Polish Science (START scholarship to M.K.-O.), and the Polish Minister of Science and Higher Education (scholarship to M.K.-O.).
↵2 These authors contributed equally to the article.
Glossary
- OXPHOS
- oxidative phosphorylation
- ROS
- reactive oxygen species
- QC
- quality control
- CL
- cardiolipin
- PE
- phosphatidylethanolamine
- SD
- short day
- LD
- long day
- 2D-DIGE
- two-dimensional fluorescence difference gel electrophoresis
- IEF
- isoelectric focusing
- PMF
- peptide mass fingerprinting
- MALDI-TOF
- matrix-assisted laser-desorption ionization time of flight
- CaMV
- cauliflower mosaic virus
- DIGE
- fluorescence difference gel electrophoresis
- BN
- blue-native
- DAG
- diacylglycerol
- PI
- phosphatidylinositol
- PG
- phosphatidylglycerol
- PVDF
- polyvinylidene difluoride
- DTT
- dithiothreitol
- ACN
- acetonitrile
- Received March 11, 2016.
- Accepted June 6, 2016.
- Published June 13, 2016.
REFERENCES
