Cold-shock-induced cytosolic calcium response changes in parallel in leaves and drupes, and no change in cytosolic calcium, but a reduction in ion leakage from cell membranes, occur in both organs when cold acclimation is reached
Cold-responsiveness was verified in leaves and drupes of open-air grown specimens of Canino, Leccino, Moraiolo, Frantoio, and Taggiasca genotypes at WAF10 and WAF19, i.e., at the beginning and end, respectively, of the oleogenesis, and related FAD activities, in the drupes (Matteucci et al. 2011), by monitoring [Ca2+]cyt-transients in their protoplasts under ΔT/Δt = 10 °C/60 s cold-shocks (D’Angeli et al. 2003, and “Materials and methods”).
Transient [Ca2+]cyt-rises were always present at WAF10, but remained present only in Frantoio, Taggiasca and Moraiolo at WAF19, showing that Canino and Leccino had acquired acclimation during this WAF-interval, whereas the other genotypes had remained cold-sensitive, as exemplified by Canino and Moraiolo in Fig. S1a–d. At WAF22, also leaves and drupes of Frantoio did not show any [Ca2+]cyt-rise after the same cold-shock (Fig. S1e), showing acclimation acquisition during the interval between WAF19 and WAF22, in accordance with previous results for the drupes (D’Angeli et al. 2013).
Based on these results, the following analyses were focussed on Canino and Moraiolo because representative of the cold responsiveness of the other genotypes, after having confirmed the acquisition (Canino), and the non-acquisition (Moraiolo) of cold acclimation at WAF19 by the application of the ion leakage test (see “Materials and methods”). In fact, the mean percentage of ion leakage was 68 % (±1) in the leaves, and 79 % (±0.1) in the drupes of Canino, and 86 % (±0.1) in the leaves, and 91 % (±0.2) in the drupes of Moraiolo. The statistical analysis of data showed that the leakage percentage from the leaves was significantly (P < 0.0001) lower than from the drupes in both genotypes, however it was significantly (P < 0.0001) reduced in Canino in comparison with Moraiolo independently of the organ.
At WAF10, cold-shocks on protoplasts isolated from plants exposed to A1- and A2-types of cold treatment resulted into the persistence of [Ca2+]cyt-rises, i.e., presence of cold-responsiveness, in both genotypes (Fig. S2a, b). An A3-type cold treatment to the plant was still not-sufficient to cause cold-acclimation in Moraiolo, but sufficient in Canino, because, under the cold-shock, only the drupe and leaf protoplasts of Moraiolo continued to show calcium rises [10.1 (±0.8) and 8.6 (±0.8) mean values, respectively] (Fig. 1a, b). The result was confirmed by the electrolyte leakage test, because the leaves and drupes of Canino showed a mean percentage of ion leakage highly (P < 0.0001) reduced in comparison with Moraiolo under the same A3-type cold treatment to the plant (Fig. S3). Also when the plants were exposed to the B-type cold treatment before protoplast cold-shocking, Moraiolo remained cold-sensitive, differently from Canino, because only the cold-shocked protoplasts of the former showed the [Ca2+]cyt-rises (Fig. 1c). The ion diffusion from the membranes of leaves and drupes coupled with this result, because the mean percentage of ion leakage in both organs was significantly (P < 0.0001) higher in Moraiolo than in Canino (Fig. S3). Collectively, results from cytosolic calcium response in the protoplasts, and ion leakage from cell membranes after the B-type cold treatment to WAF10-plants, demonstrate that the sudden exposure of Canino plants to 0 °C for 2 h, coming from an open-air temperature of 25 °C, followed by a long exposure to a higher temperature, which however was fifteen degrees lower than the open-air temperature, is sufficient to induce and maintain a sudden cold acclimation in the genotype. In accordance, perennial plants without winter dormancy, as olive tree, are known to acclimate rapidly and maintain cold hardiness even under periods of unseasonably higher temperatures (Guy 1990).
To verify whether Moraiolo might become cold-acclimated later than WAF19, the cytosolic-calcium response induced by the cold-shock in the leaf protoplasts (the drupes were fallen from the tree) was evaluated at WAF26 (full winter) in the open-air grown specimens. Calcium rises continued to be present, sustaining that Moraiolo had remained cold-sensitive (Fig. S2c). When, at WAF26, the plants were exposed to a C-type cold treatment before protoplast isolation, the inability of Moraiolo to acquire cold-acclimation was confirmed, because its protoplasts still showed presence of [Ca2+]cyt-rises under cold-shocking [4.1 (±0.5) mean value, Fig. 1d)].
Cold-acclimation couples with increased epicarp/leaf cutinisation, and OeOSM expression/immunolocalization
In both Canino and Moraiolo, the epicarp outer cell wall was thickened, and with anticlinal pegs, already at WAF3 (Fig. 2a), and thickening continued during the following WAFs (Fig. 2b). Thickness became significantly higher in Canino than Moraiolo at WAF19 (Table 2), i.e., when Canino drupes had become cold-acclimated under open-air, but not those of Moraiolo (Fig. S1d). In Canino, in particular, cutinized pegs were highly extended (Fig. 2c), and the cell walls crossed by micro-channel-like striations (Fig. 2e, arrows). Only in Canino drupes, thickening continued (20 % mean increase at WAF22 in comparison with WAF19). In both genotypes, oil bodies (OBs) were present in the epicarp and adjacent mesocarp cells (Fig. 2c, d).
As for drupes, at WAF19, cutinisation of the outer cell walls of adaxial leaf epidermis was significantly higher in Canino than Moraiolo (Table 2). All together results showed that even if a thicked cuticle as an adaptation response to summer drought in numerous species of the Mediterranean region, including olive tree (Bacelar et al. 2004), there was a further increase in both the aerial organs related to cold acclimation acquisition. However, thickening in the outer cell walls of adaxial leaf epidermis did not continue at WAF22 in any genotype.
The OBs were present in the cytoplasm of both epidermal and adjacent palisade cells (Fig. 2f, g), and micro-channel-like striations appeared in the inner cutinized outer cell wall of Canino, in particular (Fig. 2f). To exclude that the increase in thickening of the outer cell wall of Canino leaf/drupe was due to peculiarities of the genotype unrelated with the acquired cold-acclimation, the thickness of the outer cell wall was also measured in Frantoio, in the absence (WAF19) and presence (WAF22) of acclimation (Fig. S1e). Results confirmed that the increased cutinisation was related to acclimation independently of the genotype, but also showed that the thickness reached at acclimation was genotype-dependent. In fact, in comparison with Canino and Moraiolo at WAF19 (Table 2), in Frantoio, at the same WAF, the outer cell walls of the epicarps and adaxial leaf epidermides showed a thickness comparable to that of the non-acclimated Moraiolo (i.e., 20.5 ± 0.9 and 9.2 ± 0.2 μm, in the two organs, respectively), whereas at WAF22 values significantly (P < 0.0001) higher (i.e., 27 ± 0.2 and 16 ± 0.6 μm, in the two organs, respectively), even if statistically lower (P < 0.0001) than those of the WAF19-acclimated Canino.
Moreover, in Canino and Moraiolo, at WAF19, OeOSM was present within the cells of the leaf adaxial epidermis and upper palisade (Fig. 2h–j), and in the epi-mesocarp (Fig. 2k–l). The immuno-localization signal was also present in the inner cutinized outer cell wall (Fig. 2h–l, arrows), including the micro-channel-like striations, and with higher intensity in Canino than Moraiolo (Fig. 2h, j, arrowheads).
During the B-type cold treatment inducing artificial acclimation in Canino plants (Fig. 1c), OeOSM transcripts sharply increased in its drupes already at day 7 (P < 0.0001 in comparison with days 0–2), and did not change further (Fig. 3a). Conversely, OeOSM transcripts widely fluctuated in Moraiolo drupes, with significant (P < 0.0001) increases at days 2–7 in comparison with day 0, but with a strong decrease, up to day 0 value, at day 14 (Fig. 3a).
To verify whether OeOSM transcripts changed in leaves that had experienced long periods at low temperatures under open-air, an experiment was carried out at WAF26 using the Ca2+-ionophore A23187, because it causes [Ca2+]cyt-transients similarly to cold-shocks in the absence of acclimation (D’Angeli et al. 2003). OeOSM levels were similar in Canino, and always significantly (P < 0.0001) higher than in Moraiolo (Fig. 3b). In the latter, OeOSM expression was calcium-modulated, with progressive rises (P < 0.05 increases at 48 h in comparison with 0 h, P < 0.01 at 72 h in comparison with 48 h, Fig. 3b).
OeFAD8 transcript levels changed with cold differently in the presence and absence of cold-acclimation, as C18:3-lipids
The content of the α-linolenic acid (C18:3) in the PL, FFA, and TAG fractions was determined in epi-mesocarp and leaves at WAF19, i.e., when differences in cold-acclimation had appeared in Canino and Moraiolo plants grown under open-air. The total levels of the unsaturated-FA were conspicuously higher in the cold-acclimated genotype than in the sensitive one, independently of the organ, being more than threefold higher in the drupes and more than sixfold in the leaves in the former genotype in comparison with the latter (Fig. S4). Moreover, consistent differences in the total α-linolenic acid content were observed between the two organs in Canino, with the leaves exhibiting a content about fourfold higher than the drupes, whereas only slight differences were observed in Moraiolo (i.e., a 1.5-fold higher content in the leaves compared with the drupes, Fig. S4). Moreover, the percentage distribution of C18:3 in the fractions of each organ changed significantly between the genotypes, with the compound mainly present in Canino in the TAGs of the drupe and PLs of the leaf, and in Moraiolo in the PLs of the drupe and FFAs of the leaf (Table 3).
The presence of the oxylipin 13S-hydroxy-9Z,11E,15Z-octadecatrienoic acid (13-HoTre), a C18:3 oxidation-product, was also evaluated, because it is a stress-induced compound and a possible cutin component (Montillet et al. 2005; Blée et al. 2012). In Canino and Moraiolo, it was present in the drupes at not significantly different micromolar levels, i.e., 0.37 (±0.03) and 0.52 (±0.18), whereas at highly significantly (P < 0.0001) different levels in the leaves [2.24 (±0.2) and 25.75 (±0.33), respectively].
OeFAD8 was investigated as promising candidate to induce the observed changes in the C18:3 compounds related to acclimation (see “Introduction”). OeFAD8 expression was low in the drupes at WAF10, without significant differences between genotypes, but with further drupe growth its levels became higher in Canino than in Moraiolo (P < 0.0001 differences at WAF12 and WAF16, P < 0.05 difference at WAF19, in comparison with WAF10, Fig. 4a).
When the plants were exposed to A2-type cold-treatment at WAF10, i.e. at the time, and conditions, not causing cold-acclimation (Fig. S2b), a significant (P < 0.01) reduction in OeFAD8 transcripts occurred in the drupes of both genotypes (Fig. 4b).
When artificial acclimation was induced at WAF10 in Canino and not Moraiolo by the B-type cold treatment (Fig. 1c), in the drupes of the former, after a small, but significant (P < 0.05), increase in OeFAD8 levels at day 2 in comparison with day 0, there was a P < 0.0001 rise at day 7, but no further significant variation (Fig. 4c). By contrast, transcript levels transiently changed in Moraiolo drupes, with a highly significant (P < 0.0001) increase at day 2, a strong decrease at day 7 (up to day 0 value), and a new increase (P < 0.01) at the treatment-end (Fig. 4c). When the naturally (open air)-cold-acclimated WAF19-plants of Canino, and the same-aged, but cold-sensitive, plants of Moraiolo, were exposed to an A3-cold treatment, OeFAD8 levels were very low in Moraiolo, and many-fold higher (P < 0.01 differences with Moraiolo), and stable, in Canino (Fig. 4d).
In both genotypes grown under open-air from WAF10 to WAF19, OeFAD8 levels in the leaves were higher than in the drupes, but, at WAF19, they were significantly (P < 0.01) increased in Canino, and decreased in Moraiolo, as in the drupes (Figs. 4a, 5a). Moreover, the expression at WAF19 of OeFAD3 and OeFAD7, other ω3-FAD genes, was detected only in traces, excluding any possibility of their cooperation with OeFAD8, and the activity of its protein, in producing the C18:3 levels necessary for natural acclimation. When, at WAF26, the plants were exposed to a C-type cold treatment, OeFAD8 mRNA remained unchanged in Canino leaves, whereas fluctuated in Moraiolo (P < 0.001 increases at 12, 37, 55, and 72 h in comparison with 0 h). However, the gene transcripts were highly lowered in comparison with WAF19 (Fig. 5a, b).
The application of the Ca2+-ionophore to the leaves at WAF26 showed that OeFAD8 expression was calcium-modulated in Moraiolo, with strong rises in expression (P < 0.0001 increases at 48 and 72 h in comparison with 0 h), whereas it did not change significantly in Canino (Fig. 5c). The possibility that OeFAD7 and OeFAD3 might support OeFAD8 in sustaining leaf acclimation was investigated. However, no relationship of the two genes with cold acclimation was found, because OeFAD3 expression continued to be in traces, and OeFAD7 highly fluctuated also in Canino, and even when the plants experienced a C-type cold treatment (Fig. S5).
A LIP TF is present in olive-tree genome
The presence of a TF homologous to cold-induced b-Zip LIP genes of other species was investigated. Total cDNA, synthesized starting from RNA samples of Canino, was used as template for PCR amplifications of partial regions of OeLIP, as described in “Materials and methods”. Amplicons were sequenced and OeLIP nucleotide succession (Table 4) loaded in GenBank (KR360744). The amino acidic sequence of OeLIP was also aligned and compared with nineteen well-documented b-ZIP proteins (including LIP proteins) registered in GenBank. Multiple sequence alignment clearly indicated that OeLIP presented the typical features of b-ZIP proteins: the basic domain (underlined sequence, rich in basic amino acids and able to bind the target DNA regions), and the leucine zipper (evidenced by arrows, a series of repeats of leucines or other hydrophobic residues at every seventh position that are responsible for the dimerization of b-ZIP TFs) (Fig. S5). The similarity matrix produced by the comparisons of the b-ZIP protein successions (Table S1) revealed that OeLIP showed a percentage identity of its sequence with the other b-ZIP proteins ranging between 63.37 % (when related with the accession no. JN021399; 67 identical aa and 79 similar aa) and 30.39 % (in correlation with the accession no. KC951877; 34 identical aa and 52 similar aa). According to this matrix, a phylogenetic radial tree was also constructed; OeLIP branch was perfectly inserted in the genetic structure among the dicot b-ZIP proteins (Fig. 6).
OeLIP transcript levels change with cold in drupes and leaves, but differently in the presence and absence of acclimation, and with a trend similar to OeFAD8
OeLIP expression at WAF10 was similar in Moraiolo and Canino drupes, but further, fluctuated in the former reaching very low levels at WAF19. Conversely, in the latter genotype significant increases (P < 0.0001) occurred at WAF12/16, and were followed, at WAF19, by a slight decrease in level, which however, remained higher than WAF10 (P < 0.001) (Fig. 7a).
When WAF10-plants were exposed to an A2-type cold-treatment, a P < 0.01 reduction in OeLIP transcripts occurred in the drupes of both genotypes in comparison with the control plants (Fig. 7b). A B-type cold treatment applied to the plants starting from WAF10 caused a P < 0.001 increase in OeLIP transcripts in Canino drupes at day 2, and constant values after (Fig. 7c). Conversely, highly significant (P < 0.0001) OeLIP levels were reached in Moraiolo at days 2–7, but were followed by a strong (P < 0.0001) decrease (day 14) (Fig. 7c). When the plants were exposed to an A3-type cold treatment at WAF19, OeLIP levels were quite negligible in Moraiolo, and many-fold higher, and constant, in Canino (Fig. 7d), similarly to OeFAD8 (Fig. 4d).
During WAF10 to WAF19 interval, OeLIP levels in the leaves of both genotypes growing under open-air were more abundant than in the drupes, and in Canino, in particular (Fig. 8a). Moreover, Canino OeLIP abundance increased significantly (P < 0.01) at WAF19 in comparison with WAF10, whereas decreased (P < 0.01) in Moraiolo (Fig. 8a), as observed for OeFAD8 (Fig. 5a).
At WAF26, OeLIP transcripts in the leaves of both genotypes were many-fold lower than at WAF19 (Fig. 8a, b). However, when the plants were exposed to a C-type cold treatment, OeLIP mRNA remained constant in Canino, whereas fluctuated in Moraiolo (Fig. 8b). The ionophore application to the leaves showed that OeLIP expression was highly enhanced by calcium-influx in Moraiolo, with progressive rises, i.e., at 48 h (P < 0.001 difference with 0 h), and 72 h (P < 0.0001 difference with 0 h), whereas did not change significantly in Canino (Fig. 8c), with a response similar to that of OeFAD8 (Fig. 5c).