The Characterization of Arabidopsis mterf6 Mutants Reveals a New Role for mTERF6 in Tolerance to Abiotic Stress

Exposure of plants to abiotic stresses, such as salinity, cold, heat, or drought, affects their growth and development, and can significantly reduce their productivity. Plants have developed adaptive strategies to deal with situations of abiotic stresses with guarantees of success, which have favoured the expansion and functional diversification of different gene families. The family of mitochondrial transcription termination factors (mTERFs), first identified in animals and more recently in plants, is likely a good example of this. In plants, mTERFs are located in chloroplasts and/or mitochondria, participate in the control of organellar gene expression (OGE), and, compared with animals, the mTERF family is expanded. Furthermore, the mutations in some of the hitherto characterised plant mTERFs result in altered responses to salt, high light, heat, or osmotic stress, which suggests a role for these genes in plant adaptation and tolerance to adverse environmental conditions. In this work, we investigated the effect of impaired mTERF6 function on the tolerance of Arabidopsis to salt, osmotic and moderate heat stresses, and on the response to the abscisic acid (ABA) hormone, required for plants to adapt to abiotic stresses. We found that the strong loss-of-function mterf6-2 and mterf6-5 mutants, mainly the former, were hypersensitive to NaCl, mannitol, and ABA during germination and seedling establishment. Additionally, mterf6-5 exhibited a higher sensitivity to moderate heat stress and a lower response to NaCl and ABA later in development. Our computational analysis revealed considerable changes in the mTERF6 transcript levels in plants exposed to different abiotic stresses. Together, our results pinpoint a function for Arabidopsis mTERF6 in the tolerance to adverse environmental conditions, and highlight the importance of plant mTERFs, and hence of OGE homeostasis, for proper acclimation to abiotic stress.


Introduction
The increased salt content in arable soils severely compromises plant growth and productivity. This is due to osmotic stress, which promotes water loss and hinders its uptake by plant roots, and to ionic stress (Na + and Cl − in most cases), which generates toxicity and hinders the recruitment of other ions [1]. The development of new varieties of more halotolerant crop plants requires unravelling the genetic and molecular mechanisms that underlie tolerance to salinity. It has been proposed that chloroplasts could act as sensors capable of sensing environmental stress, and, by retrograde signalling (from the chloroplast to the nucleus), could coordinate the expression of nuclear genes altered response of the mterf6 mutants to the stress conditions assayed, which is consistent with the substantial changes in mTERF6 expression we found in silico after exposing the wild-type to different abiotic stresses.

The mterf6-2 and mterf6-5 Mutants Are Hypersensitive to NaCl and Mannitol
In order to assess whether the mterf6-5 mutant that we previously identified [22] exhibited altered sensitivity to abiotic stresses, we first analysed its sensitivity, and that of the wild-type Col-0, to the ionic stress produced by NaCl and osmotic stress due to mannitol. We also included the mterf6-2 mutant, allelic of mterf6-5, in the analysis (see above). For this purpose, we first examined the ability of mterf6-2, mterf6-5 and Col-0 seeds to germinate and to form fully expanded green cotyledons (seedling establishment) in the first 2 weeks after seed stratification in the presence of 0, 150, or 200 mM of NaCl or 350 mM of mannitol. In the non-supplemented culture medium, mutants mterf6-2 and mterf6-5 respectively yielded, to some extent, lower and similar seed germination ratios than Col-0 (we considered germinated those seeds in which radicle emergence through the seed coat was observed) (Figure 1a). The supplementation of growth medium with NaCl (150 mM) or mannitol (350 mM) did not affect wild-type seed germination, but lowered the mterf6-2 and mterf6-5 germination rates, especially those of the former, and the effect was more pronounced from 1 to 5 DAS (days after stratification; Figure 1c,e).
Consistent with the stunted growth of the mterf6-2 and mterf6-5 individuals [20,22], the seedling establishment of both mutants was delayed compared with Col-0 (e.g., at 6 DAS in the MS control medium, 10%, 37%, and 99% of the mterf6-2, mterf6-5 and Col-0 seeds yielded seedlings with fully expanded green cotyledons, respectively (Figure 1b)). However, at 10 DAS, 93% and 100% of seedling establishments were achieved for the mterf6-5 and the wild type, respectively, whereas mterf6-2 reached a maximum value of 77% at 11 DAS (Figure 1b). The mterf6-2 and mterf6-5 seeds yielded substantially lower seedling establishment rates than those of Col-0 in the presence of 150 mM of NaCl or 350 mM of mannitol (Figure 1d,f). Accordingly, the presence of the mterf6-2 and mterf6-5 seedlings with green expanded cotyledons could be scored only from 10 DAS in the presence of NaCl or mannitol, while the seedling establishment for Col-0 was observed from 4-5 DAS under the same conditions (Figure 1d,f). Notwithstanding, the mterf6-2 mutant was more sensitive than mterf6-5 to NaCl. In line with this, the maximum seedling establishment values for mterf6-2 and mterf6-5 in 150 mM NaCl were 14% and 62%, respectively, which were reached at 13 DAS, whereas Col-0 yielded~100% ( Figure 1d). However, a similar strong hypersensitive response to mannitol was found for both mutants throughout the study period ( Figure 1f).
We also investigated the response of Col-0 and mterf6-5 to a higher salt concentration by supplementing the culture medium with 200 mM of NaCl. We found that this condition significantly delayed mutant germination (e.g., at 5 and 10 DAS, 98% and 99% of the wild type and 12% and 80% of the mterf6-5 seeds germinated, respectively, in 200 mM of NaCl; Table S1), and completely abolished the Col-0 and mterf6-5 seedling establishments, as we were unable to identify any individual that displayed green expanded cotyledons.
Taken together, our results revealed enhanced sensitivity to salt and osmotic stress during germination, and mainly in the cotyledon greening stage, for the studied mterf6 mutants.
We evaluated the response of mterf6-5 to salinity by exposing plants to stress after germination and seedling establishment. To this end, 5 DAS wild-type and mutant seedlings were transferred from the non-supplemented medium to the media supplemented with NaCl (125 or 150 mM), and root length was determined 8 days after transfer (13 DAS; see Materials and Methods; Table S2). The mterf6-5 plants were significantly less sensitive than the wild-type ones to the inhibition of root growth caused by the presence of either 125 mM of NaCl or 150 mM of NaCl (Table 1; Table S2). We evaluated the response of mterf6-5 to salinity by exposing plants to stress after germination and seedling establishment. To this end, 5 DAS wild-type and mutant seedlings were transferred from the non-supplemented medium to the media supplemented with NaCl (125 or 150 mM), and  The values correspond to the root length inhibition percentages of the plants transferred 5 DAS to the media supplemented with either 125 or 150 mM of NaCl or 5 or 10 µM of ABA, which refers to those of plants of the same genotype, which were transferred to the non-supplemented media. Eight days after transfer (13 DAS), the main root length was determined per plant to evaluate their tolerance to these stress conditions (see Materials and Methods). Each value is the mean ± SD of the main root length of at least 20 plants per genotype and condition. The values significantly differed from the Col-0 at ** p < 0.01 according to a Student's t-test.
To study whether a low mTERF6 expression altered tolerance to moderate heat stress, the wild-type Col-0 and mterf6-5 mutant seedlings were exposed 13 days to a higher (28 • C) than normal culture temperature (20 • C). We also compared the response of mterf6-5 with that of mterf mutants mda1-1 and mterf9. The mterf6-5 mutant was hypersensitive to heat stress because paleness markedly increased and seedling growth was severely impaired, and even arrested, when grown at 28 • C ( Figure S1). In contrast, the growth of mda1-1 and mterf9 was enhanced at 28 • C, but to a lesser extent than in Col-0 ( Figure S1).

Knock-Down of mTERF6 Alters the Response to ABA
The abscisic acid (ABA) hormone plays a fundamental role in seed germination and in the responses of plants to abiotic stresses [23]. The Arabidopsis mutants deficient in ABA signalling or biosynthesis also exhibited enhanced tolerance to salt stress [24][25][26]. Therefore, given the enhanced sensitivity of mterf6-2 and mterf6-5 to salt and osmotic stress, we investigated whether they also exhibited an altered response to ABA by growing the mterf6 mutant and Col-0 seedlings in the presence of ABA. As shown in Figure 1a,g, 3 µM of ABA substantially delayed mterf6-2, mterf6-5 and Col-0 germination, but from 3 to 5 DAS both mutant seeds exhibited higher levels of radicle emergence through the seed coat than those of the wild-type. However, when the mterf6-5 and Col-0 individuals were grown on 6 µM of ABA, seed germination was greater in mterf6-5 than in Col-0 only at 5 DAS, but both genotypes yielded very low germination values (6% and 3%, respectively; Table S3). In contrast, we found that mterf6-5 was hypersensitive to ABA from 6-13 DAS. Accordingly at 6, 7, 10, and 13 DAS, 44%, 62%, 99%, and 100% of the Col-0 seeds, and 22%, 36%, 60%, and 82% of the mterf6-5 seeds germinated, respectively (Table S3).
As regards seedling establishment, exposure to 3 µM ABA considerably reduced it in Col-0 (e.g., up to 48% of the wild-type seedlings under the control condition at 13 DAS), and completely abolished it in mterf6-2, while only 2% was found for mterf6-5 (Figure 1h). When grown on 6 µM ABA, 18% and 42% of the Col-0 seeds yielded seedlings with green expanded cotyledons at 10 and 13 DAS, respectively. As expected, no mterf6-5 seedlings showing green expanded cotyledons were found from 3 to 13 DAS (Table S3).
We allowed the Col-0 and mterf6-5 seedlings to grow on the ABA-supplemented medium. At 17 DAS, 6.4% and 1.4% of the mutant seeds (n = 150) yielded individuals that displayed two very tiny leaves in 3 and 6 µM of ABA, respectively. In contrast, 27.7% and 16.2% of the Col-0 seedlings (n = 150) displayed two small leaves in 3 and 6 µM of ABA, respectively.
Taken together, these results indicate that the mterf6-2 and mterf6-5 mutants are hypersensitive to ABA principally during seedling establishment.
As we did for NaCl (see Section 2.1), we also investigated the sensitivity of mterf6-5 to ABA after germination and seedling establishment. To this end, 5 DAS wild-type and mutant plants were transferred from the non-supplemented medium to the media supplemented with ABA (5 or 10 µM). Root length was determined 8 days after transfer (13 DAS; Table S2). As well as for NaCl, the root growth of the mterf6-5 individuals was significantly more tolerant than that of the Col-0 plants to 5 µM of ABA, whereas inhibition of root length only slightly decreased in 10 µM of ABA (Table 1).

The Expression of the mTERF6 Gene Changes in Response to Abiotic Stresses
Given the altered sensitivity of the mterf6 mutants to NaCl, mannitol and ABA, we decided to perform an in silico analysis of the expression of the mTERF6 gene in response to different abiotic stress conditions. Hence we studied the stress-induced changes in the transcript levels of mTERF6 with the Arabidopsis AtGenExpress Visualization Tool ( [27]; available online: http://weigelworld.org/ resources.html) in the roots and aerial parts of the Col-0 seedlings under NaCl, osmotic and drought stresses. The expression values were plotted over time (0, 0.5, 1, 3, 6, 12, and 24 h after treatment started) to obtain a graphical representation of the response of mTERF6 to these conditions ( Figure S2). Compared with the untreated plants, mTERF6 expression was down-regulated in the green parts of seedlings after 3 h of NaCl (150 mM), mannitol (300 mM) and drought treatments, and mostly in the presence of NaCl and mannitol from 6 to 24 h. This repression peaked 24 h after treatment when the mTERF6 transcript levels lowered to 16% and 42% of the control plants in response to mannitol and NaCl, respectively ( Figure S2a). As regards roots, mTERF6 expression was down-regulated by salt stress from 1 to 24 h after treatment started. The difference to the control plants was maximum at 6 h (38.6% of the control plants), whereas mannitol slightly increased the mTERF6 transcript levels at 3 h (28% more than in the control plants), but lowered them from 6 to 24 h, especially at 6 h (63% of the control plants) (Figure S2b). Drought reduced mTERF6 expression at 1 and 6 h after exposure (77% and 74% of the control plants), but no appreciable differences were found for the remaining time points. As regards the effect of ABA, mTERF6 expression was down-regulated to 53.4% of the control plants by 10 µM ABA after 3 h, but no noticeable differences were found after 0.5 and 1 h. We also investigated the transcript levels of mTERF6 using the online data from the At-TAX Arabidopsis whole genome tilling array [28]. Consistently with the AtGenExpress results, we found that the 12-h exposure of the 10-day-old Col-0 seedlings to 200 mM of NaCl, 300 mM of mannitol or 100 µM of ABA markedly reduced mTERF6 transcript abundance to 30.4%, 45.5% and 45.2% of those of the untreated seedlings, respectively. Slighter differences between the treated and untreated plants were detected after 1 h of exposure under the same conditions. We experimentally tested by qRT-PCR whether mTERF6 expression may change in response to NaCl. To this end, RNA was extracted from Col-0 seedlings collected 10 DAS and grown in GM medium supplemented with 100 mM NaCl or in non-supplemented medium. The RNA was retro-transcribed and the cDNAs analyzed by qPCR. Though this condition was different from those used by the Arabidopsis AtGenExpress consortium (see above; [27]), we previously found that it delayed Col-0 growth [4]. We included as a positive control the RD29A gene which is induced by salinity [29]. In response to this moderate salt stress, RD29A was significantly upregulated (1.7 ± 0.3; p = 10 −3 ) whereas mTERF6 was slightly downregulated (0.8 ± 0.4; p = 0.2).

Discussion
In this work, we analysed the response of the mterf6-2 and mterf6-5 mutants to different abiotic stresses during germination, seedling establishment and for mterf6-5 later in development. We found that the mterf6 mutants displayed altered sensitivity to salt, osmotic stress, ABA, and moderate heat stress. Unlike the results obtained with other mterf -deficient mutants, such as mda1 and mterf9, which are more insensitive than the wild type to such stresses [4,5], mterf6-2 and mterf6-5 were hypersensitive to the inhibition exerted on germination and seedling establishment by high concentrations of NaCl, mannitol, or ABA. mterf6-2 was always more sensitive than mterf6-5 to the different abiotic stress conditions studied, which is consistent with its more severe morphological phenotype [22]. The susceptibility of mterf6 mutants to NaCl was similar to that of mterf10 and mterf11 [7], but unlike these mutants, which were as sensitive as the wild type was to ABA, mterf6-2 and mterf6-5 were also hypersensitive to this hormone, mainly during seedling establishment.
In line with this, the knock-down of mTERF6 also reduced seedling tolerance to moderate heat stress and led to impaired growth and development, whereas mda1-1 and mterf9 (this work), and mterf10 and mterf11 [7], did not show a significantly different response from that of the wild type under this condition. mTERF6 seemed to play a different role further in vegetative development because the deficient mTERF6 function significantly reduced the sensitivity of roots to the presence of NaCl or ABA in the growth medium. A different susceptibility to salt and ABA during germination and vegetative growth has been previously reported for mutants mda1 and mterf9 [4,5].
We extracted the mTERF6 transcript levels from AtGenExpress [27] by selecting "AtGE Abiostress" as a data source. Consistent with altered tolerance to abiotic stresses, we found that mTERF6 expression was markedly down-regulated in response to salt, osmotic stress (mannitol) and drought, especially after prolonged exposure (12-24 h) to 150 mM NaCl and 300 mM mannitol. Interestingly, ABA treatment also repressed mTERF6 expression. We experimentally tested by qRT-PCR mTERF6 expression in 10 DAS plants grown in mild salt stress conditions (100 mM NaCl), and found that it was slightly but not significantly downregulated, which is likely due to the different stress conditions used to study mTERF6 expression. Together, our results suggest that the altered tolerance of mterf6-2 and mterf6-5 to the tested abiotic stresses could be attributed to its different sensitivity to ABA compared with the wild type, because this hormone plays a fundamental role in plants' response and adaptation to abiotic stress conditions. The involvement of mTERF6, MDA1 (affected in the mTERF5 gene), and mTERF9 (the mda1 and mterf9 mutants are less sensitive to ABA than the wild type) [4,5], and possibly of mTERF10 (since a modest overexpression of this gene leads to enhanced germination and growth in the presence of ABA) [7] in abiotic stress tolerance could take place through ABA signalling. Accordingly, several pieces of experimental evidence indicate a role for ABA in plastid-to-nucleus signalling (reviewed in [3]). Therefore, the impaired plastid gene expression may be due to a defective mTERF function perturbing the retrograde communication (from plastids to the nucleus) mediated by ABA under salt or other abiotic stress conditions. As a result, this would alter nuclear gene expression, and hence, tolerance to these environmental conditions. Similarly, Leister and Kleine [21] found that levels of the nuclear transcripts, which encode the chloroplast proteins involved in organellar gene expression (OGE), were affected in the weak mterf6-1 mutant. Notwithstanding, while some mTERF proteins (e.g., mTERF5, mTERF9 and mTERF10) would negatively modulate Arabidopsis salt tolerance as their down-regulation diminishes sensitivity to ABA and abiotic stresses, mTERF6 would play the opposite role by promoting such tolerance, at least during germination and seedling establishment. Consequently, it could be hypothesised that the outcome of the activity of different mTERF proteins, which act during germination and early vegetative development, might contribute to responses to abiotic stress in these developmental stages. The mTERF6 function in abiotic responses might be conserved in other plant species because the expression of the maize mTERF12 gene, the orthologue of Arabidopsis mTERF6, is substantially altered after NaCl or ABA treatments [30]. Interestingly, the transcript levels of other maize mTERF genes also change after exposing maize plants to light/dark treatments, salt, ABA or 1-Naphthaleneacetic acid exposure [30]. The altered levels of the mTERF6 transcripts after abiotic stress treatments found in silico might be interpreted as being necessary for plants to adapt to adverse environmental conditions. Nevertheless given currently available molecular information, we cannot rule out the notion that changes in the expression of mTERF6 and other mTERF genes under different abiotic stress conditions might result from the perturbation of certain biological processes. Chloroplast homeostasis is likely to be one of these processes altered in mterf -deficient mutants, because all the mTERFs involved in the response to salt stresses described to date are targeted to chloroplasts; they also belong to the "chloroplast cluster" (mTERF5, mTERF6 and mTERF9) or to the "chloroplast associated-cluster" (mTERF10 and mTERF11) of proteins by functioning in organelle gene expression, embryogenesis, gene expression, and/or protein catabolism in plants [13]. The altered OGE, and hence chloroplast homeostasis, would account for the delayed growth and greening of the cotyledons of the mterf6 individuals in relation to Col-0, even in the absence of abiotic stress. However, differences with Col-0 considerably increased when the mterf6 mutants were exposed to salt, mannitol, or ABA, which indicates that mterf6-2 and mterf6-5 sensitivity to abiotic stresses cannot be attributed solely to its defective growth.
The involvement of the mTERF family of genes in the acclimation and tolerance of plants to different abiotic stresses conditions [11,14] is further supported by recent findings in cotton (Gossypium barbadense). Accordingly, multiple stress responsive genes have been identified in G. barbadense using a normalised cDNA library, constructed after exposure to various abiotic (heat, cold, salt, drought, potassium, and phosphorous deficit) and biotic (Verticillium dahlia infection) stress conditions [31]. Remarkably, the mRNAs of 464 transcription factors (TF) have been enriched in this library, and mTERFs are one of the most abundant TF families to have been identified (3.7%) [31].

Germination and Growth Sensitivity Assays
For the germination assays, sowings were carried out as described in [4] on Petri dishes filled with GM agar medium (Murashige and Skoog (MS) medium containing 1% sucrose), supplemented with NaCl (150 and 200 mM), mannitol (350 mM) or ABA (3 and 6 µM). The seeds in which radicle emergence was observed were considered to be germinated, whereas seedling establishment was determined as those seedlings that exhibited green and fully expanded cotyledons. Seed germination and seedling establishment were scored from 1 to 13 DAS or from 1 to 24 DAS on Petri dishes, kept at 20 ± 1 • C with 72 µmol·m −2 ·s −1 of continuous light.
To determine the salt and ABA responses during vegetative growth after seedling establishment, seeds were sown on non-supplemented GM agar medium, and seedlings were transferred on 5 DAS to new Petri dishes supplemented with NaCl (125 or 150 mM) or ABA (5 or 10 µM), and vertically grown. Plant root length was determined after 8 days of NaCl or ABA treatment to evaluate their tolerance to these stress conditions by referring the values to those of the individuals transferred to the control (non-supplemented) media.

Quantitative RT-PCR (qRT-PCR)
Total RNA was extracted from 80 mg 10 DAS wild-type Col-0 plants grown in the presence or absence of 100 mM NaCl in the GM agar medium. The RNA was resuspended in 40 µL of RNase-free water and DNA removed using the TURBO DNAfree kit (Invitrogen, Waltham, MA, USA) following the manufacturer's instructions. The cDNA preparations and qPCR amplifications were carried out in an ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Waltham, MA, USA) as described in [4] using the oligonucleotides listed on Table S4. Each reaction mix of 20-µL contained 7.5 µL of the SYBR-Green/ROX qPCR Master Kit (Fermentas, Waltham, MA, USA), 0.4 µM of primers, and 1 µL of the cDNA solution. Relative quantification of gene expression data was performed by the 2 −∆∆Ct method as described in [4]. Each reaction was done in three replicates, and three different biological replicates were used. The expression levels were normalised to the CT values obtained for the housekeeping ACTIN2 gene [32], and a Mann-Whitney U-test was applied to the relative expression data obtained.

Computational Analyses
The expression responses of the mTERF6 gene under ABA, salt, osmotic, and drought stress were obtained from the AtGenExpress Visualization Tool (available online: http://jsp.weigelworld.org/ expviz/expviz.jsp) [27] by selecting the "AtGE Abiostress" as the data source and mean-normalised values. The mTERF6 expression in response to ABA was also visualised by extracting the tilling array data from TileViz (available online: http://jsp.weigelworld.org/tileviz/tileviz.jsp) [28] by selecting the "Abiotic Stress Dataset" and the mean-normalised values.

Conclusions
In summary, the results reported herein reveal a new function for the mTERF6 gene related to the emerging roles that have been recently proposed for the mTERF family in plants' response and adaptation to different environmental stress conditions. In the plant mterf mutants characterised to date which have exhibited altered sensitivity to abiotic stresses, the affected mTERF proteins are involved in OGE [11,13,14]. Hence, this pinpoints an important function for OGE and plastid homeostasis, likely by acting throughout retrograde signalling, in tolerance to adverse environmental conditions, as recently proposed [3]. Further molecular research on the effect of abiotic stresses on the mTERF6 function, and by extension on the remaining mTERFs, is required to shed more light on the contribution of this scarcely known family of genes for plants to cope with abiotic stresses.