The plant Apolipoprotein D ortholog protects Arabidopsis against oxidative stress

Background Lipocalins are a large and diverse family of small, mostly extracellular proteins implicated in many important functions. This family has been studied in bacteria, invertebrate and vertebrate animals but little is known about these proteins in plants. We recently reported the identification and molecular characterization of the first true lipocalins from plants, including the Apolipoprotein D ortholog AtTIL identified in the plant model Arabidopsis thaliana. This study aimed to determine its physiological role in planta. Results Our results demonstrate that the AtTIL lipocalin is involved in modulating tolerance to oxidative stress. AtTIL knock-out plants are very sensitive to sudden drops in temperature and paraquat treatment, and dark-grown plants die shortly after transfer to light. These plants accumulate a high level of hydrogen peroxide and other ROS, which causes an oxidative stress that is associated with a reduction in hypocotyl growth and sensitivity to light. Complementation of the knock-out plants with the AtTIL cDNA restores the normal phenotype. On the other hand, overexpression enhances tolerance to stress caused by freezing, paraquat and light. Moreover, this overexpression delays flowering and maintains leaf greenness. Microarray analyses identified several differentially-regulated genes encoding components of oxidative stress and energy balance. Conclusion This study provides the first functional evidence that a plant lipocalin is involved in modulating tolerance to oxidative stress. These findings are in agreement with recently published data showing that overexpression of ApoD enhances tolerance to oxidative stress and increases life span in mice and Drosophila. Together, the three papers strongly support a similar function of lipocalins in these evolutionary-distant species.


Background
Lipocalins are small ligand-binding proteins found in bacteria and in invertebrate and vertebrate animals. Over 40 lipocalin members have been identified from all king-doms [1]. They show a simple tertiary structure which gives them the ability to bind small, generally hydrophobic, molecules. Animal lipocalins play important roles in the regulation of immunological and developmental processes, and they are involved in the responses of organisms to various stress factors and in signal transduction pathways. It was recently shown that the insect glial Lazarillo lipocalin (GLaz) possesses a protective role against oxidative stress conditions and that its absence increases lipid peroxydation, reduces life span and accelerates neurodegeneration in Drosophila [2]. On the other hand, its overexpression protects against the effects of starvation, hypoxia and hyperoxia, and extends the fly's life span [3].
Data mining of genomic databases and bioinformatic predictions allowed us to determine that plants also possess lipocalins, which were classified as temperature-induced lipocalins (TILs) and chloroplastic lipocalins (CHLs) [4]. The TIL genes are induced by high and low temperature (LT) and their level of expression is associated with the plant's capacity to develop freezing tolerance (FT). Furthermore, the accumulation of TIL transcripts is limited to photosynthetic tissues and oscillates during the diurnal cycle. TILs do not show any targeting peptide, but the proteins are found at the plasma membrane. Sequence, structure and phylogenetic analyses revealed that TIL proteins share homology with three evolutionary related lipocalins: the bacterial lipocalin (Blc), the mammalian apolipoprotein D (ApoD), and the GLaz proteins.
In spite of the accumulating knowledge regarding the molecular features of plant lipocalins, very little is known about their function at the cellular and biochemical levels. However, the TIL properties, their association with the plasma membrane in photosynthetic tissues, and their accumulation in response to temperature stress support the hypothesis that these proteins may act as scavengers of potentially harmful molecules known to be induced by temperature stress and excess light [5]. In this report, we characterized knock-out, complementation and overexpression lines to determine the cellular and biochemical functions of the AtTIL lipocalin.

AtTIL delays flowering and maintains leaf greenness
For functional analyses, four T-DNA lines carrying insertions in the AtTIL gene were analyzed to identify knockout (KO) lines (Fig. 1a) [see Additional file 1] [6]. The SALK_136775 line carries a single insertion in the first exon of the AtTIL gene and shows no detectable AtTIL expression (Fig. 1b). Complementation of this KO line restored AtTIL protein accumulation to a level about twofold higher than the wild-type (WT) plants (Fig. 1b,  Comp). Overexpression of the AtTIL cDNA resulted in a 4fold accumulation of the protein compared to the WT (Fig. 1b, OEX). Under normal growth conditions, the downregulation of AtTIL expression (KO) has no visible effect on plant growth and development compared to the WT (Fig. 1c). In contrast, the Comp and OEX lines show a delay in flowering and a stay-green phenotype clearly visible at 45 and 52 days, respectively ( Fig. 1c and 1d, Table  1). These data suggest that an increased AtTIL level extends the vegetative phase and maintains leaf greenness. To determine if AtTIL protein accumulation is modulated during development, a western blot analysis was performed on samples prepared from tissues collected throughout the WT plant's life cycle. The results show that apart from the dry seeds stage, AtTIL is expressed at approximately the same level at all stages of development (Fig. 2a). Since there is no association between the protein accumulation and normal flowering, and because there is already a high level of protein before and during bolt elongation, we hypothesize that an excess of AtTIL may disrupt certain molecular events that ultimately delay the transition from the vegetative to reproductive phase.

AtTIL enhances tolerance to freezing and oxidative stress
There are no noticeable morphological differences between the different Arabidopsis lines when they are compared before (non-acclimated; NA) or after cold acclimation (CA) (Fig. 3a). After subjecting NA plants to a freezing test at -6°C, a survival rate of 75% was obtained for the wild-type (WT) and control (vector) plants (Fig.  3b). In contrast, only 50% of the KO plants survived while the survival rate of the Comp and OEX plants was more than 90%. Freezing treatment caused typical damage and necrosis to the leaves of surviving WT and KO plants. On the other hand, leaves of OEX plants showed no damage after freezing and resumed vigourous growth when transferred to normal growth conditions at 20°C (Fig. 3a). After 7 days of cold acclimation (CA7), the level of AtTIL protein increases in all lines except in KO plants (Fig. 3c). This indicates that AtTIL accumulation is associated with an increased freezing tolerance that provides protection against damages caused by a sudden drop in temperature. On the other hand, the difference in sensitivity to freezing between NA WT and KO plants is not observed when plants are cold-acclimated prior to exposure to freezing (CA7 -10°C, Fig. 3a). Additional freezing tests performed down to -13°C did not reveal any differences between WT and KO plants [see Additional file 2], suggesting that cold acclimation is sufficient to provide maximal FT, and that AtTIL is likely not essential when plants undergo the full acclimation process. Protoplasts isolated from non acclimated lipocalin transgenic plants showed a higher survival rate compared to WT after freezing to -10°C. This increase in survival rate was associated with a decrease in the incidence of dehydration-induced injury [7].
The freeze-induced lesions may be caused by harmful reactive oxygen species (ROS) generated when plants are exposed to freezing. To test the effect of ROS, plants were treated with the oxidant paraquat, a powerful herbicide that generates ROS [8][9][10]. After 7 days of treatment, KO plants showed more necrotic lesions than WT plants, whereas OEX plants were more resistant ( Fig. 4a and 4b).

Modulation of AtTIL protein level affects development
The effect of light stress on the different lines was monitored under different fluence rates (Fig. 5). Dark-grown KO plants show a reduction in hypocotyl elongation that is reversed in the Comp line, while OEX plants show longer hypocotyls (Fig. 5a, upper panel). There is no difference in AtTIL accumulation between plants grown under light or dark conditions (Fig. 2b). When darkgrown WT, Comp, OEX and control plants are transfered to normal photoperiod and light conditions, greening occurs and typical development resumes. In contrast, dark-grown KO plants transferred to light do not accumulate chlorophyll and die shortly after (Fig. 5a Fig. 5D. It has been shown that hydrogen peroxide accumulation causes growth inhibition in maize and pine hypocotyls [11]. We speculate that the higher basal level of peroxide and other ROS in the KO plants is likely responsible for the higher sensitivity to oxidative stress, which leads to a reduction in hypocotyl growth and increased sensitivity to light. It was reported that the inability to eliminate excess peroxide increases sensitivity towards environmental stresses and induces the expression of several stress-related genes [12].

Transcriptome analysis of AtTIL knock-out plants
Expression profiles were determined for WT and KO plants grown under NA and CA conditions (7 days at 4°C). No differences in gene expression were observed between the CA WT and CA KO, except for the absence of AtTIL transcript in the KO line. However, 66 genes were differentially regulated by more than 2-fold in the NA KO compared to the NA WT (51 up and 15 down-regulated; Table 2). Among these 66 genes, 49 are regulated in a similar manner (up or down) by LT exposure in WT plants while 5 are regulated in an opposite manner. These genes encode known or putative transcription factors or signal transduction proteins, heat shock proteins, enzymes involved in carbohydrate metabolism or oxidative stress responses, and senescence and circadian clock-associated proteins. The remaining 12 genes are regulated only in the KO line (* in Table 2; KO-specific genes) and encode proteins that are associated with stress responses. This analysis indicates that without exposure to LT, the absence of AtTIL mimicks part of the LT response. The accumulation of hydrogen peroxide and other ROS in the KO plants ( Fig. 5) may act as a signal that triggers the changes observed in gene expression. A similar finding was reported for catalase-deficient plants under non-stressed conditions [13]. It is known that hydrogen peroxide accumulation results in differential gene expression [14].
Several genes related to the disease resistance pathway are overexpressed in the AtTIL KO plant. The WRKY54 gene is known for its association with the defense response and involves an increase in salicylic acid (SA), which accumulates during oxidative stress and mediates the induction of defense response genes [15,16]. Similarly, the SigA binding protein is a homologue of MSK1 which was shown to cause an accumulation of SA [17]. The At3g55450 protein is most homologous to the BIK1 kinase which is induced by ROS generators such as paraquat and proposed to play a role in regulating optimal SA levels [18]. Induction of the BIK1 kinase-like gene in AtTIL KO plants without exposure to any stress suggests that the level of ROS is higher than normal in these plants. The greater susceptibility to paraquat is also indicative that the AtTIL KO plants are less capable of fighting additional oxidative stress. Furthermore, the up-regulation of riboflavin biosynthesis protein in KO plants suggests that riboflavin becomes limiting when AtTIL is absent. Since riboflavin deficiency is associated with oxidative stress and is an important component of FAD-requiring enzymes such as glutathione reductase, it is possible that KO plants may try to fight oxidative stress by increasing their level of riboflavin.
The microarray analysis suggests that in the KO plants, starch synthesis is reduced while starch catabolism is accelerated to provide additional soluble sugars needed to fight oxidative stress. In addition, the Krebs cycle is inhibited under stress conditions, preventing the efficient utilization of the glycolysis products and thus affecting the respiratory machinery [19]. To overcome this situation, plants rely on the γ-aminobutyrate (GABA) shunt. Succinic semialdehyde dehydrogenase, the last enzyme of the GABA shunt, is upregulated in AtTIL KO plants. The GABA shunt is a bypass of the Krebs cycle which provides succinate and NADH to the respiratory machinery [20]. In plants, the activity of this pathway is enhanced in response to biotic and abiotic stresses [21,22] suggesting a potential role in reducing the impact of oxidative stress in mitochondria [19].
Animal lipocalins have been suggested to have a role in restoring membrane integrity caused by oxidative stress, likely by their binding to fatty acids [23]. This study provides evidence that the plasma membrane-associated AtTIL lipocalin plays a role in protecting plants against the oxidative stress induced by freezing, paraquat treatment and light. We hypothesize that during oxidative stress, lipocalins may bind and scavenge peroxidated lipids, and thus help restore membrane integrity. Our findings are in agreement with recently published data showing that overexpression of ApoD enhances tolerance to oxidative stress and increases life span in mice and Drosophila [24,25]. The available evidence therefore point towards a universal cellular function of lipocalins among species.

AtTIL lines
Arabidopsis ecotype Columbia (Col-0) was the genetic background of all the lines used. T-DNA knock-out (KO) lines for the AtTIL gene were obtained from the Salk Institute Genome Analysis Laboratory http://signal.salk.edu/. Seeds were sown on agar plates containing 50 μg/ml kanamycin, stratified for 2 days at 4°C and grown at 22°C. Kanamycin-resistant plants were propagated as individual lines on potting medium consisting of two parts Arabidopsis growing medium PM-15-13 (LEHLE seeds), one part vermiculite and one part black earth. Based on the segregation analysis of plants grown on antibiotic-containing plates, lines putatively homozygous for the T-DNA insertion were subjected to PCR analysis using T-DNA-specific (LBb1) and AtTIL-specific (RP) primers, and the resulting PCR fragments were sequenced to determine the precise insertion sites.
To generate the complementation (Comp) and overexpressing (OEX) lines, the AtTIL ORF was first cloned into pRTL2, a vector that contains a double cauliflower mosaic virus (CaMV) 35S promoter and a 35S terminator [26]. The resulting PRO 35S :AtTIL:TERM 35S cassette was then cloned into the binary plant expression vector pPZP121 [27] and electroporated into Agrobacterium GV3101. KO and WT Arabidopsis plants were transformed with this vector using the floral dipping method [28] to generate the Accumulation of the AtTIL protein immunodetected with the anti-AtTIL antibody  Values shown are the mean number ± SD of rosette leaves at the time of flowering. At least 12 plants of each line were used, and the experiments were repeated at least twice. Statistical analysis was performed by one-way ANOVA. The asterisks (**) and (*) indicate differences that are significant at the P < 0.001 and P < 0.01 levels respectively. Acronyms are as described in Fig. 1.
Comp and OEX lines, respectively. Two independent OEX lines were selected for complete characterization. Since similar results were obtained, only one line is presented.
To generate a negative control, WT plants were trans-formed with the pPZP121 vector that does not contain the AtTIL cassette. Transformed plants of all lines were selected by growth on gentamycin-containing medium. Homozygous plants of the fourth generation after transformation were used for the experiments.

Plant growth conditions and treatments
Surface-sterilized seeds were sown on sterile half strength Murashige and Skoog medium (Sigma-Aldrich). Seeds were stratified for 2 days at 4°C then germinated and grown at 22°C/18°C (day/night) with a 16 hr photoperiod and a photon flux density of 90 μmol m -2 sec -1 . For soil-based analyses, seeds were sown directly on potting medium, stratified for 3 days at 4°C and grown for 11 AtTIL enhances tolerance to freezing stress  At1g68440 259856_at -2,1 -2,5 -4,6 -1,1 * 12 genes regulated (induced or repressed) only in the KO line (not regulated by exposure of the WT to LT) # 22 genes for which the regulation in KO is at least 50% of the regulation seen in WT exposed at 4°C † key regulators of the circadian clock 4°C. To determine temperature variability in the freezer, temperatures were monitored by 6 independent thermocoupled T probes distributed in the freezer and connected to an Agilent 3497-0A data acquisition/switch unit. Freezing regimes that showed more than 0.5°C discrepancies between the different probes were rejected. To minimize light stress effects after the freezing treatment, plants were thawed at 4°C for 24 hr in the dark and away from direct light in the growth chamber (20°C) for an additional 24 hr before returning to normal light conditions. Pictures were taken 3 weeks after the freezing treatment. Eighteen plants were frozen per line per assay, and the experiment was repeated 3 times with independent biological replicates.

Hypocotyl analyses
Seeds were sown on sterile half strength Murashige and Skoog medium (Sigma-Aldrich) and stratified in the dark at 4°C for 6 days. They were then exposed to white light (90 μmol m -2 s -1 ) at 20°C for 60 min and returned to darkness for 23 hr before being exposed to varying fluence rates at 20°C for 7 days. Hypocotyl length was measured using the Image J 1.36b software (mean ± SD; n = 40). H 2 O 2 and other ROS were detected in situ by 3,3'-diaminobenzidine (DAB) staining as described [29].

Protein isolation and immmunoblot analysis
The antibody raised against the wheat TaTIL lipocalin protein (4) does not cross react with the Arabidopsis AtTIL lipocalin, therefore it was necessary to raise a specific antibody against the latter. The AtTIL cDNA was cloned in the pTrc-HIS vector (Invitrogen) and electroporated into the E. coli strain DH5α. The recombinant HIS::AtTIL protein was produced by induction with 1 mM IPTG for 3 hr. The cells were collected, broken by lysozyme treatment and sonication, and HIS-tagged proteins were purified by immobilized metal affinity chromatography on Ni-NTA agarose (Novagen). The purified proteins were more than 90% pure and used to immunize a rabbit to obtain polyclonal antibodies.
Aerial parts of the Arabidopsis plants were cut and immediately frozen in liquid nitrogen. One hundred milligrams of leaf material was processed as one sample. Proteins were isolated using Tri Reagent according to the manufacturer's instructions (Sigma). Samples were separated on 12% SDS-PAGE gels and the rabbit anti-AtTIL antibody (1:10,000) was used for the immunoblot analysis. Detection was performed with a peroxidase-coupled anti-rabbit IgG secondary antibody (1:25,000) and the Western Lightning Chemiluminescence Reagent Plus (Perkin-Elmer).

PCR and Southern blot analyses
Genomic DNA was extracted from flower buds of the different Arabidopsis lines. PCR analysis was performed according to the SIGnAL protocol using the recommended primers designed with the SIGnAL iSect Primer Design sofware: LBb1 (5' GCGTGGACCGCTTGCT-GCAACT 3'), LP (5' CTGGATCCAGAGATGAAGTCG 3') and RP (5' AAGACGTGTATGGTACCGTCG 3'). For Southern analysis, the DNA (5 μg) was digested by PciI (New England Biolabs) and fractionated by electrophoresis on a 0.7% agarose gel. After electrophoresis, the gel was transferred to a positively charged nylon membrane and hybridized with 32P-labeled probes corresponding to the T-DNA of pROK2 (vector used to generate the SALK lines) or the T-DNA of the 35S:AtTIL vector. The following primer pairs were used for amplification of the fragments used as probes: 5'-CAGCAAAATCACCAGTAGCACCAT-TACCAT-3' and 5'-GCGATAGAAAACAAAATATAGCGCG-CAAAC-3' for the pROK2 T-DNA, and 5'-ACGAAACGTGGAGCAACGGGAAGAG-3' and 5'-TGCA-CATACAAATGGACGAACGGATAAACC -3' for the AtTIL T-DNA. All filters were washed at high stringency (0.1× SSC, 0.1% SDS), exposed to K screens (Kodak) and analyzed on a Molecular Imager FX (Bio-Rad).

Transcriptome analysis
Non-acclimated and cold-acclimated WT and KO plants were used for microarray analysis. For each sample, three independent biological replicates of 25 plants were harvested for RNA isolation. RNA quality assessment and hybridization to the Affymetrix GeneChip Arabidopsis ATH1 Genome Arrays were performed at the McGill University and Génome Québec Innovation Centre (Montreal QC Canada). Data accumulation and analysis were performed using the Robust Multi-array Average analysis (RMA) [30], and data was deposited in the ArrayExpress database [ArrayExpress:E-TABM-430]. An analysis of variance was performed using GraphPad InStat 3 to select genes that are significantly differentially expressed by at least two-fold under the conditions specified for each analysis (4°C vs 22°C and KO vs WT). Functional classification of the genes was done according to their annotation in the TAIR database http://www.arabidopsis.org/ tools/bulk/go/index.jsp, by BLAST search of the NCBI database http://www.ncbi.nlm.nih.gov/blast/Blast.cgi and by literature search performed using the gene name and annotation. RT-PCR analyses of AtTIL and randomly selected transcripts were performed to confirm the absence of AtTIL expression in the KO plants and to validate the microarray results [see Additional file 3].