Overexpression of OsPIN9 Impairs Chilling Tolerance via Disturbing ROS Homeostasis in Rice

The auxin efflux transporter PIN-FORMED (PIN) family is one of the major protein families that facilitates polar auxin transport in plants. Here, we report that overexpression of OsPIN9 leads to altered plant architecture and chilling tolerance in rice. The expression profile analysis indicated that OsPIN9 was gradually suppressed by chilling stress. The shoot height and adventitious root number of OsPIN9-overexpressing (OE) plants were significantly reduced at the seedling stage. The roots of OE plants were more tolerant to N-1-naphthylphthalamic acid (NPA) treatment than WT plants, indicating the disturbance of auxin homeostasis in OE lines. The chilling tolerance assay showed that the survival rate of OE plants was markedly lower than that of wild-type (WT) plants. Consistently, more dead cells, increased electrolyte leakage, and increased malondialdehyde (MDA) content were observed in OE plants compared to those in WT plants under chilling conditions. Notably, OE plants accumulated more hydrogen peroxide (H2O2) and less superoxide anion radicals (O2−) than WT plants under chilling conditions. In contrast, catalase (CAT) and superoxide dismutase (SOD) activities in OE lines decreased significantly compared to those in WT plants at the early chilling stage, implying that the impaired chilling tolerance of transgenic plants is probably attributed to the sharp induction of H2O2 and the delayed induction of antioxidant enzyme activities at this stage. In addition, several OsRboh genes, which play a crucial role in ROS production under abiotic stress, showed an obvious increase after chilling stress in OE plants compared to that in WT plants, which probably at least in part contributes to the production of ROS under chilling stress in OE plants. Together, our results reveal that OsPIN9 plays a vital role in regulating plant architecture and, more importantly, is involved in regulating rice chilling tolerance by influencing auxin and ROS homeostasis.


Introduction
Due to extreme weather events, plants are progressively subjected to various abiotic stresses, such as chilling stress. Chilling stress restricts the geographical distribution and affects plant productivity [1,2]. Rice (Oryza sativa L.) originates from tropical and subtropical regions and is more sensitive to chilling stress than other cereal crops [3][4][5]. Previous data showed that rice yield decreased by 30-40% due to cold stress in temperate areas [6]. Therefore, improving rice chilling tolerance for maintaining food security is an urgent target [7]. In recent years, excellent progress has been made in understanding the physiological and molecular mechanisms underlying chilling stress [1,[8][9][10]. Chilling stress triggers Ca 2+ influx, which plays a key role in chilling tolerance [1,[11][12][13]. For example, COLD1 (CHILLING-TOLERANCE DIVERGENCE 1) and OsCNGC9 (CYCLIC NUCLEOTIDE-GATED CHANNEL 9) have been demonstrated to regulate rice chilling tolerance by mediating Ca 2+ influx under chilling stress [7,14]. Growing evidence shows trafficking of auxin efflux carriers substantially influences auxin homeostasis under cold conditions [29]. Further research demonstrated that GNOM, a SEC7 containing ARF-GEF, which is closely associated with endosomal trafficking of auxin efflux carriers, can positively regulate Arabidopsis cold tolerance [62]. Collectively, these results strongly suggest that auxin homeostasis and transport are implicated in regulating plant chilling tolerance. However, little is known about the underlying molecular mechanisms involved in the role of auxin homeostasis and transport in chilling stress.
Previous studies demonstrated that OsPIN9 functions in polar auxin transport and is involved in regulating adventitious root number and tiller number [55]. Recently, Manna et al. (2022) [63] reported that salt and drought treatment significantly induced OsPIN9 expression, indicating the potential role of OsPIN9 in abiotic stress adaptation. In line with this, our previous study showed that mutation of OsPIN9 by CRISPR/Cas9 technology confers enhanced chilling tolerance in rice [64]. However, whether and how OsPIN9 overexpression regulates rice chilling tolerance is unknown. In addition, since offtarget effects or newly formed mutation proteins derived from CRISPR/Cas9 technology have the potential to cover the true function of target genes, it is necessary to further verify gene function by upregulating its expression. Here, we functionally examined the role of OsPIN9 in modulating chilling tolerance by overexpressing technology. Our results showed that overexpression of OsPIN9 not only leads to the alteration of rice architecture but also impairs rice chilling tolerance. We provided strong evidence that OsPIN9 plays a crucial role in modulating rice chilling tolerance and shed light on the relationship between polar auxin transport and chilling tolerance in rice.

Expression Patterns of OsPIN Genes under Chilling Stress
Increasing evidence shows that plant chilling tolerance is tightly associated with polar auxin transport [29,59,61,62]. However, the mechanism by which OsPIN genes respond to chilling stress is still largely unknown. To this end, quantitative real-time polymerase chain reaction (qRT-PCR) was performed to evaluate the responses of OsPIN genes in rice roots upon chilling stress. Nine OsPIN genes (OsPIN1b, OsPIN1c, OsPIN1d, OsPIN2, OsPIN5a, OsPIN5b, OsPIN9, OsPIN10a, and OsPIN10b), which can be detected in rice roots [39,45,46], were selected to analyze the expression profiles under chilling stress. OsPIN1b displayed a different expression pattern under chilling treatment compared to OsPIN1c and OsPIN1d, albeit belonging to the same subfamily. In detail, OsPIN1b was induced to the highest expression after chilling for 6 h and then decreased sharply, while OsPIN1c and OsPIN1d were suppressed after chilling for 3 h and kept to a relatively constant level thereafter. OsPIN2 remained at a constant level under chilling treatment. Similar to OsPIN1b, OsPIN5a and OsPIN5c also showed increased expression first and then decreased. Three monocotspecific PIN genes, OsPIN9, OsPIN10a, and OsPIN10b, were greatly inhibited after chilling treatment and gradually decreased with treatment duration (Figure 1). Intriguingly, the paralogous OsPIN genes, OsPIN1c/1d, OsPIN5a/5c, and OsPIN10a/10b, showed a similar expression profile under chilling stress, implying that they might play a synergistic role in chilling tolerance.

Generating OsPIN9-Overexpressing Transgenic Rice and Phenotypes of Transformants
The monocot-specific OsPIN9, which was dramatically suppressed under chilling treatment, was chosen for further study. To obtain insight into the function of OsPIN9 in chilling tolerance, a plasmid carrying the OsPIN9 ORF, driven by a strong constitutive Ubiquitin promoter (pUbi) (Figure 2A), was introduced into rice cultivar Nipponbare by Agrobacterium. The transgenic plants were confirmed by PCR using genomic DNA as a template with specific primers. qRT-PCR analysis showed that the OsPIN9 transcript significantly increased by 3.6-1237 times in the five tested transgenic line roots compared with that in WT plants ( Figure 2B). Among which, A4 and B14 represented the high expression levels, A1 represented the moderate expression level, and A2 and B15 represented the low

Generating OsPIN9-Overexpressing Transgenic Rice and Phenotypes of Transformants
The monocot-specific OsPIN9, which was dramatically suppressed under chilling treatment, was chosen for further study. To obtain insight into the function of OsPIN9 in chilling tolerance, a plasmid carrying the OsPIN9 ORF, driven by a strong constitutive Ubiquitin promoter (pUbi) (Figure 2A), was introduced into rice cultivar Nipponbare by Agrobacterium. The transgenic plants were confirmed by PCR using genomic DNA as a template with specific primers. qRT-PCR analysis showed that the OsPIN9 transcript significantly increased by 3.6-1237 times in the five tested transgenic line roots compared with that in WT plants ( Figure 2B). Among which, A4 and B14 represented the high expression levels, A1 represented the moderate expression level, and A2 and B15 represented the low expression levels. Three lines, A4, A1, and B15, which represented high, moderate, and low expression levels of OsPIN9, were employed to further analyze the expression of OsPIN9 in leaves, roots, and stem bases. The highest increase in expression was detected in leaves, followed by roots and stem bases ( Figure 2C). Two T2 homozygous transgenic lines, A4 and B14, with high expression levels of OsPIN9, designated as OE1 and OE2, respectively, were used for further investigation.

Generating OsPIN9-Overexpressing Transgenic Rice and Phenotypes of Transformants
The monocot-specific OsPIN9, which was dramatically suppressed under chilling treatment, was chosen for further study. To obtain insight into the function of OsPIN9 in chilling tolerance, a plasmid carrying the OsPIN9 ORF, driven by a strong constitutive Ubiquitin promoter (pUbi) (Figure 2A), was introduced into rice cultivar Nipponbare by Agrobacterium. The transgenic plants were confirmed by PCR using genomic DNA as a template with specific primers. qRT-PCR analysis showed that the OsPIN9 transcript significantly increased by 3.6-1237 times in the five tested transgenic line roots compared with that in WT plants ( Figure 2B). Among which, A4 and B14 represented the high expression levels, A1 represented the moderate expression level, and A2 and B15 represented the low expression levels. Three lines, A4, A1, and B15, which represented high, moderate, and low expression levels of OsPIN9, were employed to further analyze the expression of OsPIN9 in leaves, roots, and stem bases. The highest increase in expression was detected in leaves, followed by roots and stem bases ( Figure 2C). Two T2 homozygous transgenic lines, A4 and B14, with high expression levels of OsPIN9, designated as OE1 and OE2, respectively, were used for further investigation.  Overexpression of OsPIN9 significantly decreased shoot height, root length, and adventitious root number after germination for 7 d in comparison with those in WT plants ( Figure 3A). In detail, compared with the WT plants, the shoot height, root length, and adventitious root number of the OE plants were significantly decreased by 23-26%, 11-15%, and 17-18%, respectively. After germination for 14 d, shoot height and adventitious root number of OE lines declined by 21-25% and 13-14%, respectively, compared to those in WT plants. However, the root length was similar to WT plants at this stage ( Figure 3B), indicating that rice plants can dynamically adapt to the change in auxin transport to some extent.  Figure 3A). In detail, compared with the WT plants, the shoot height, root length, and adventitious root number of the OE plants were significantly decreased by 23-26%, 11-15%, and 17-18%, respectively. After germination for 14 d, shoot height and adventitious root number of OE lines declined by 21-25% and 13-14%, respectively, compared to those in WT plants. However, the root length was similar to WT plants at this stage ( Figure 3B), indicating that rice plants can dynamically adapt to the change in auxin transport to some extent. . The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; ***: p < 0.001.

Overexpression of OsPIN9 Disturbs Auxin Homeostasis in Rice
It was reported that overexpression of OsPIN9 accelerated auxin transport from shoots to roots and resulted in an increase and decrease of IAA in roots and shoots, respectively [55]. N-1-naphthylphthalamic acid (NPA) is an inhibitor of polar auxin transport and can interact with PIN carriers directly [65]. To further assess the role of OsPIN9 in auxin transport, germinated wild-type and transgenic seeds were cultured in Kimura B complete nutrient solution [39] supplemented with 0.5 μM NPA. After 7 d treatment, NPA obviously inhibited plant growth; the shoot height of OE lines was significantly decreased compared to WT plants both under normal conditions and NPA treatment. Contrastingly, the root length of OE lines was shorter than that of WT under normal conditions, while it increased significantly more than that of WT plants under NPA treatment. In detail, the root length of OE lines was significantly decreased by 10-12% when compared with WT under normal conditions, whereas it was about 39% and 49% longer compared with control plants after NPA treatment, indicating the roots of OE lines were more resistant to NPA than those of WT plants ( Figure 4). Overexpression of OsPIN9 significantly lowered the adventitious root number under control conditions, while NPA treatment severely decreased the adventitious root number and completely abolished the difference in adventitious root number between WT and OE lines. In addition, the ability to grow upright in some OE plant shoots was obviously inhibited and showed a partially impaired negative gravitropism compared with that in WT plants, indicating that the shoots of OE lines are probably more sensitive to NPA than those of WT plants ( Figure 4). Collectively, these results demonstrate that auxin homeostasis is disturbed in OE lines. Values are means ± standard deviation (SD) (n = 24). The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; ***: p < 0.001.

Overexpression of OsPIN9 Disturbs Auxin Homeostasis in Rice
It was reported that overexpression of OsPIN9 accelerated auxin transport from shoots to roots and resulted in an increase and decrease of IAA in roots and shoots, respectively [55]. N-1-naphthylphthalamic acid (NPA) is an inhibitor of polar auxin transport and can interact with PIN carriers directly [65]. To further assess the role of OsPIN9 in auxin transport, germinated wild-type and transgenic seeds were cultured in Kimura B complete nutrient solution [39] supplemented with 0.5 µM NPA. After 7 d treatment, NPA obviously inhibited plant growth; the shoot height of OE lines was significantly decreased compared to WT plants both under normal conditions and NPA treatment. Contrastingly, the root length of OE lines was shorter than that of WT under normal conditions, while it increased significantly more than that of WT plants under NPA treatment. In detail, the root length of OE lines was significantly decreased by 10-12% when compared with WT under normal conditions, whereas it was about 39% and 49% longer compared with control plants after NPA treatment, indicating the roots of OE lines were more resistant to NPA than those of WT plants ( Figure 4). Overexpression of OsPIN9 significantly lowered the adventitious root number under control conditions, while NPA treatment severely decreased the adventitious root number and completely abolished the difference in adventitious root number between WT and OE lines. In addition, the ability to grow upright in some OE plant shoots was obviously inhibited and showed a partially impaired negative gravitropism compared with that in WT plants, indicating that the shoots of OE lines are probably more sensitive to NPA than those of WT plants ( Figure 4). Collectively, these results demonstrate that auxin homeostasis is disturbed in OE lines.

Overexpression of OsPIN9 Impairs Chilling Tolerance in Rice
Due to the fact that polar auxin transport is closely related to chilling stress [66], and PIN carriers perform a rate-limiting role in regulating auxin efflux [34], as well as the quick suppression of OsPIN9 under chilling stress (Figure 1), the role of OsPIN9 in chilling tolerance was evaluated in WT and OE plants. Fourteen day old rice seedlings were moved from 28 • C to 4 • C. After chilling for 36 h, most WT leaves remained normal, but almost all leaves of OE plants displayed rolling and wilting ( Figure 5A). The survival rate of OE plants was significantly lower than that of WT plants after 3 d of recovery at 28 • C ( Figure 5A). Consistently, after 57 h of chilling treatment followed by 3 d of recovery, the survival rate of WT was dramatically higher than that of OE plants ( Figure 5B). In detail, WT plants showed a 98% survival rate, while the survival rates of OE1 and OE2 were only about 17% and 26%, respectively. These results suggest that the expression of OsPIN9 is involved in negatively modulating rice chilling tolerance.

Overexpression of OsPIN9 Impairs Chilling Tolerance in Rice
Due to the fact that polar auxin transport is closely related to chilling stress [66], and PIN carriers perform a rate-limiting role in regulating auxin efflux [34], as well as the quick suppression of OsPIN9 under chilling stress (Figure 1), the role of OsPIN9 in chilling tolerance was evaluated in WT and OE plants. Fourteen day old rice seedlings were moved from 28 °C to 4 °C. After chilling for 36 h, most WT leaves remained normal, but almost all leaves of OE plants displayed rolling and wilting ( Figure 5A). The survival rate of OE plants was significantly lower than that of WT plants after 3 d of recovery at 28 °C ( Figure 5A). Consistently, after 57 h of chilling treatment followed by 3 d of recovery, the survival rate of WT was dramatically higher than that of OE plants ( Figure 5B). In detail, WT plants showed a 98% survival rate, while the survival rates of OE1 and OE2 were only about 17% and 26%, respectively. These results suggest that the expression of OsPIN9 is involved in negatively modulating rice chilling tolerance.  Values are means ± standard deviation (SD) (n = 24). The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

Overexpression of OsPIN9 Impairs Chilling Tolerance in Rice
Due to the fact that polar auxin transport is closely related to chilling stress [66], and PIN carriers perform a rate-limiting role in regulating auxin efflux [34], as well as the quick suppression of OsPIN9 under chilling stress (Figure 1), the role of OsPIN9 in chilling tolerance was evaluated in WT and OE plants. Fourteen day old rice seedlings were moved from 28 °C to 4 °C. After chilling for 36 h, most WT leaves remained normal, but almost all leaves of OE plants displayed rolling and wilting ( Figure 5A). The survival rate of OE plants was significantly lower than that of WT plants after 3 d of recovery at 28 °C ( Figure 5A). Consistently, after 57 h of chilling treatment followed by 3 d of recovery, the survival rate of WT was dramatically higher than that of OE plants ( Figure 5B). In detail, WT plants showed a 98% survival rate, while the survival rates of OE1 and OE2 were only about 17% and 26%, respectively. These results suggest that the expression of OsPIN9 is involved in negatively modulating rice chilling tolerance. We then detected several physiological indicators, including cell death, electrolyte leakage, and malondialdehyde (MDA) contents, to evaluate the influence of overexpression of OsPIN9 on rice chilling tolerance. Trypan blue staining, which is an indicator of cell death, was performed to monitor the cell damage. There was no apparent difference in trypan blue staining in leaves between WT and OE plants under normal growth conditions, while staining was more intensive in the leaves of OE plants compared to that of WT plants  Figure 6A), indicating that the cell death in OE plants is greater than that in WT plants. In line with this, another two indicators, electrolyte leakage and MDA contents, were both significantly increased in OE plants compared to those in WT plants after chilling for 36 h and 57 h ( Figure 6B,C). Taken together, these results strongly suggest that overexpression of OsPIN9 substantially impairs rice chilling tolerance.
We then detected several physiological indicators, including cell death, electroly leakage, and malondialdehyde (MDA) contents, to evaluate the influence of overexpre sion of OsPIN9 on rice chilling tolerance. Trypan blue staining, which is an indicator cell death, was performed to monitor the cell damage. There was no apparent differen in trypan blue staining in leaves between WT and OE plants under normal growth cond tions, while staining was more intensive in the leaves of OE plants compared to that WT plants after chilling for 36 h and 57 h ( Figure 6A), indicating that the cell death in O plants is greater than that in WT plants. In line with this, another two indicators, electr lyte leakage and MDA contents, were both significantly increased in OE plants compare to those in WT plants after chilling for 36 h and 57 h ( Figure 6B,C). Taken together, the results strongly suggest that overexpression of OsPIN9 substantially impairs rice chillin tolerance.

The Role of OsDREB1 Regulon and Ca 2+ Signaling-Related Genes in Chilling Stress
The expression of OsDREB1 genes could be dramatically induced by low temper tures [19,67], and more and more evidence has demonstrated that OsDREB1 genes play vital role in chilling stress [14,19,67]. We then detected the expression levels of OsDREB1 OsDREB1B, and OsDREB1C in WT and OE plants upon chilling stress. Under normal co ditions, these OsDREB1 genes were significantly more induced in OE plants than in W plants ( Figure 7A). However, after chilling for 8 h, the expression levels of these genes OE plants were similar to or even lower than those in WT plants ( Figure 7A). Similarl trehalose-6-phosphate phosphatase 1 (OsTPP1), which has been demonstrated to play vital role in rice cold tolerance [9,68], displayed a similar expression profile with OsDREB genes in WT and OE plants under both normal and chilling conditions ( Figure 7B). Add tionally, increasing evidence shows that Ca 2+ signaling is involved in chilling toleran regulation [9,69], and COLD1 and OsCNGC9 have been demonstrated to facilitate cyt plasmic Ca 2+ elevation and positively regulate rice cold tolerance [7,14]. To evaluate th role of these Ca 2+ signaling-related genes in OE lines under chilling stress, we further e amined the expression of these two genes under normal and chilling conditions. The r sults showed that both COLD1 and OsCNGC9 in OE lines were slightly induced compare to those in WT plants under normal conditions, while strikingly suppressed under chillin stress ( Figure 7C). Collectively, these data suggest that the defense response might be tri gered in advance under normal conditions, and the DREB1 regulon and Ca 2+ signaling a likely to be implicated in regulating the chilling tolerance of OE lines.

The Role of OsDREB1 Regulon and Ca 2+ Signaling-Related Genes in Chilling Stress
The expression of OsDREB1 genes could be dramatically induced by low temperatures [19,67], and more and more evidence has demonstrated that OsDREB1 genes play a vital role in chilling stress [14,19,67]. We then detected the expression levels of OsDREB1A, OsDREB1B, and OsDREB1C in WT and OE plants upon chilling stress. Under normal conditions, these OsDREB1 genes were significantly more induced in OE plants than in WT plants ( Figure 7A). However, after chilling for 8 h, the expression levels of these genes in OE plants were similar to or even lower than those in WT plants ( Figure 7A). Similarly, trehalose-6-phosphate phosphatase 1 (OsTPP1), which has been demonstrated to play a vital role in rice cold tolerance [9,68], displayed a similar expression profile with OsDREB1 genes in WT and OE plants under both normal and chilling conditions ( Figure 7B). Additionally, increasing evidence shows that Ca 2+ signaling is involved in chilling tolerance regulation [9,69], and COLD1 and OsCNGC9 have been demonstrated to facilitate cytoplasmic Ca 2+ elevation and positively regulate rice cold tolerance [7,14]. To evaluate the role of these Ca 2+ signaling-related genes in OE lines under chilling stress, we further examined the expression of these two genes under normal and chilling conditions. The results showed that both COLD1 and OsCNGC9 in OE lines were slightly induced compared to those in WT plants under normal conditions, while strikingly suppressed under chilling stress ( Figure 7C). Collectively, these data suggest that the defense response might be triggered in advance under normal conditions, and the DREB1 regulon and Ca 2+ signaling are likely to be implicated in regulating the chilling tolerance of OE lines.

OsPIN9-Overexpressing Plants Accumulated More H 2 O 2 Rather Than O − 2 under Chilling Stress
In addition to the well-known CBF/DREB regulon, reactive oxygen species (ROS) also play a crucial role in chilling stress [21]. To further investigate the underlying mechanism of OsPIN9 under chilling stress, ROS content assays were performed by diaminobenzidine tetrahydrochloride (DAB 4HCl) staining for H 2

OsPIN9-Overexpressing Plants Accumulated More H2O2 Rather Than under Chilling Stress
In addition to the well-known CBF/DREB regulon, reactive oxygen species (ROS) also play a crucial role in chilling stress [21]. To further investigate the underlying mechanism of OsPIN9 under chilling stress, ROS content assays were performed by diaminobenzidine tetrahydrochloride (DAB 4HCl) staining for H2O2 and NBT staining for O under normal and chilling conditions. The result showed that OE plants accumulated a similar level of H2O2 and O compared to WT plants before chilling treatment ( Figure 8A). After chilling for 36 h, the leaves of OE plants accumulated more H2O2 and less O compared to WT plants ( Figure 8B). Consistently, more H2O2 and less O were still detected in OE leaves compared to WT plants after chilling for 57 h ( Figure 8C). These results indicate that OE lines accumulate more H2O2 than O when compared with WT plants, which probably causes cell damage and impairs rice chilling tolerance.  Values are means ± standard deviation (SD) (n = 3). The data were analyzed by ANO Tukey's tests at a p < 0.05 significance level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

OsPIN9-Overexpressing Plants Accumulated More H2O2 Rather Than under Chilling Stress
In addition to the well-known CBF/DREB regulon, reactive oxygen species (R play a crucial role in chilling stress [21]. To further investigate the underlying me of OsPIN9 under chilling stress, ROS content assays were performed by diaminobe tetrahydrochloride (DAB 4HCl) staining for H2O2 and NBT staining for O under and chilling conditions. The result showed that OE plants accumulated a similar H2O2 and O compared to WT plants before chilling treatment ( Figure 8A). After for 36 h, the leaves of OE plants accumulated more H2O2 and less O compare plants ( Figure 8B). Consistently, more H2O2 and less O were still detected in O compared to WT plants after chilling for 57 h ( Figure 8C). These results indicate lines accumulate more H2O2 than O when compared with WT plants, which p causes cell damage and impairs rice chilling tolerance.

The Delayed Induction of Antioxidant Enzymes Probably Leads to the Accumulation of H 2 O 2 in OE Lines
ROS are vital signaling molecules that play a crucial role in biotic and abiotic stress responses, while their content should be carefully controlled in plant cells via an array of enzymatic and non-enzymatic antioxidants [23]. Given that ROS homeostasis is disturbed in OE lines, we then analyzed the activities of three antioxidant enzymes, including superoxide dismutase (SOD), catalase (CAT), and peroxidase (POD), which play a key role in ROS scavenging [10,23], before chilling treatment and after chilling for 36 h and 72 h. There were no significant differences in SOD and CAT activities in WT and OE plants, while POD activity was significantly increased in OE lines compared to WT plants under normal conditions ( Figure 9A). After chilling for 36 h, the activities of CAT and SOD decreased significantly in OE plants compared to those in WT plants. To our surprise, the POD activity was still strikingly higher in OE lines than in WT plants ( Figure 9B). After chilling for 57 h, the activities of the three enzymes in OE lines all increased greatly compared to those in WT plants ( Figure 9C). These results imply that it is likely that the decreased activities of SOD and CAT rather than increased POD activity led to the accumulation of H 2 O 2 in OE lines at the early chilling stage, and the delayed induction of SOD and CAT is ineffective to quench excess ROS at the late chilling stage and causes cell damage in OE lines. However, considering that SOD usually functions in the conversion of O − 2 to H 2 O 2 , the source of the accumulated H 2 O 2 during chilling stress needs further investigation.
in ROS scavenging [10,23], before chilling treatment and after chilling for 36 h an There were no significant differences in SOD and CAT activities in WT and OE while POD activity was significantly increased in OE lines compared to WT plant normal conditions ( Figure 9A). After chilling for 36 h, the activities of CAT and S creased significantly in OE plants compared to those in WT plants. To our surp POD activity was still strikingly higher in OE lines than in WT plants ( Figure 9B chilling for 57 h, the activities of the three enzymes in OE lines all increased great pared to those in WT plants ( Figure 9C). These results imply that it is likely that creased activities of SOD and CAT rather than increased POD activity led to the a lation of H2O2 in OE lines at the early chilling stage, and the delayed induction and CAT is ineffective to quench excess ROS at the late chilling stage and causes ce age in OE lines. However, considering that SOD usually functions in the conversio to H2O2, the source of the accumulated H2O2 during chilling stress needs further gation. Figure 9. Analysis of antioxidant enzyme activities in wild-type (WT) and OsPIN9-overex (OE) plants before and after chilling treatments. Enzyme activities were detected before stress (A) and after chilling for 36 h (B) and 57 h (C) in WT and OE lines. Values are means ± deviation (SD) (n = 6). The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 sig level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

OsRobh Genes Show Differential Expression in OsPIN9-Overexpressing Lines
Previous reports showed that respiratory burst oxidase homologs (Rbo known as NADPH oxidase, play critical roles in various cellular activities and re to abiotic stresses by regulating the production of ROS [70][71][72]. For the higher pro of H2O2 detected in OE lines, we assayed the expression of OsRboh genes before a Figure 9. Analysis of antioxidant enzyme activities in wild-type (WT) and OsPIN9-overexpressing (OE) plants before and after chilling treatments. Enzyme activities were detected before chilling stress (A) and after chilling for 36 h (B) and 57 h (C) in WT and OE lines. Values are means ± standard deviation (SD) (n = 6). The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

OsRobh Genes Show Differential Expression in OsPIN9-Overexpressing Lines
Previous reports showed that respiratory burst oxidase homologs (Rboh), also known as NADPH oxidase, play critical roles in various cellular activities and responses to abiotic stresses by regulating the production of ROS [70][71][72]. For the higher production of H 2 O 2 detected in OE lines, we assayed the expression of OsRboh genes before and after chilling stress treatment to evaluate the source of H 2 O 2 in OE lines. As shown in Figure 10, under normal conditions, most OsRboh genes in transgenic lines kept similar expression levels compared to those in WT plants, except for OsRboh1, OsRboh2, and OsRboh9, which showed a slight increase in OE lines. After chilling for 24 h, more OsRboh genes were induced, including OsRboh2, OsRboh4, OsRboh6, and OsRboh8, and showed a relative higher expression level in OE lines, which may contribute to the production of H 2 O 2 in transgenic lines under chilling conditions. chilling stress treatment to evaluate the source of H2O2 in OE lines. As shown in Figure  10, under normal conditions, most OsRboh genes in transgenic lines kept similar expression levels compared to those in WT plants, except for OsRboh1, OsRboh2, and OsRboh9, which showed a slight increase in OE lines. After chilling for 24 h, more OsRboh genes were induced, including OsRboh2, OsRboh4, OsRboh6, and OsRboh8, and showed a relative higher expression level in OE lines, which may contribute to the production of H2O2 in transgenic lines under chilling conditions. Figure 10. OsRboh expression analysis in wild-type (WT) and OsPIN9-overexpressing (OE) lines before and after chilling treatment. The expression level of OsRboh genes in WT was set to one. The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

Discussion
Although excellent progress has been made in the molecular mechanisms of plant cold adaptation [8][9][10], auxin, the first phytohormone discovered and one of the most important phytohormones in plant growth and developmental regulation, its role in modulating chilling adaptation is still elusive. In this present study, the role of a monocot-specific OsPIN9 gene, which mainly functions in regulating auxin transport basipetally, in chilling stress was investigated by overexpression technology.
Auxin plays a vital role in regulating almost all cellular processes, and in many cases, an appropriate auxin level and distribution within plant tissues are direct determinators affecting plant architecture [73]. PIN genes have been demonstrated to play a central role in auxin efflux and, consequently, influence auxin content and distribution in various tissues. Expression levels of PIN genes are usually closely associated with plant architecture [74,75]. The most representative is AtPIN1, and disruption of AtPIN1 seriously influences inflorescence development and leads to the formation of needle-like inflorescence in Arabidopsis [76]. The rice genome possesses 12 OsPIN genes [45,46], and although the possibility of redundant functions of OsPIN genes exists [49], mutation of most OsPIN genes substantially influences rice architecture [47,49,[53][54][55]77]. The previous report demonstrated that the OsPIN9 carrier functions in auxin transport basipetally, overexpression of OsPIN9 substantially increases auxin content in rice roots [55], so we speculate that the auxin content in OE plant roots should be increased compared to that in WT plants. In line with this, the OE plant roots displayed a more resistant phenotype under NPA treatment (Figure 4), implying an increased auxin content in the OE roots. It was reported that overexpression of OsPIN9 increased the adventitious root number [55]. Unexpectedly, we Figure 10. OsRboh expression analysis in wild-type (WT) and OsPIN9-overexpressing (OE) lines before and after chilling treatment. The expression level of OsRboh genes in WT was set to one. The data were analyzed by ANOVA and Tukey's tests at a p < 0.05 significance level. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

Discussion
Although excellent progress has been made in the molecular mechanisms of plant cold adaptation [8][9][10], auxin, the first phytohormone discovered and one of the most important phytohormones in plant growth and developmental regulation, its role in modulating chilling adaptation is still elusive. In this present study, the role of a monocot-specific OsPIN9 gene, which mainly functions in regulating auxin transport basipetally, in chilling stress was investigated by overexpression technology.
Auxin plays a vital role in regulating almost all cellular processes, and in many cases, an appropriate auxin level and distribution within plant tissues are direct determinators affecting plant architecture [73]. PIN genes have been demonstrated to play a central role in auxin efflux and, consequently, influence auxin content and distribution in various tissues. Expression levels of PIN genes are usually closely associated with plant architecture [74,75]. The most representative is AtPIN1, and disruption of AtPIN1 seriously influences inflorescence development and leads to the formation of needle-like inflorescence in Arabidopsis [76]. The rice genome possesses 12 OsPIN genes [45,46], and although the possibility of redundant functions of OsPIN genes exists [49], mutation of most OsPIN genes substantially influences rice architecture [47,49,[53][54][55]77]. The previous report demonstrated that the OsPIN9 carrier functions in auxin transport basipetally, overexpression of OsPIN9 substantially increases auxin content in rice roots [55], so we speculate that the auxin content in OE plant roots should be increased compared to that in WT plants. In line with this, the OE plant roots displayed a more resistant phenotype under NPA treatment (Figure 4), implying an increased auxin content in the OE roots. It was reported that overexpression of OsPIN9 increased the adventitious root number [55]. Unexpectedly, we observed a lower number of adventitious roots in these two OE lines, which is different from the previous report [55]. Considering that adventitious root number is tightly associated with auxin content [55], and auxin regulates plant growth and development usually in a concentration-dependent manner, relative high auxin content accelerates plant growth, while excess auxin content strongly inhibits plant growth. Therefore, we speculate that the dramatic expression of OsPIN9 probably leads to excess accumulation of auxin in OE roots, which suppresses adventitious root number (Figure 3). In agreement with this, the relative lower expression of OsPIN9 (A2 line) significantly increased adventitious root number when compared to WT plants ( Figure S1).
Rice is sensitive to chilling stress [7,[78][79][80], and severe chilling disasters hamper rice growth and strikingly impair rice productivity [10]. Previous reports have suggested that chilling stress is implicated in auxin transport in plants [29,30,61,62]. However, there is no direct genetic evidence regarding the role of auxin transport in the regulation of chilling stress. In this study, we provided direct evidence that OsPIN9, which functions in auxin transport basipetally and could be quickly suppressed by chilling stress (Figure 1), is involved in rice chilling tolerance. The chilling treatment assay showed that OE lines are more sensitive to low temperatures than WT plants ( Figure 5); further experiments confirmed that OE lines possess more dead cells ( Figure 6A), higher membrane permeability ( Figure 6B), and increased MDA content ( Figure 6C) compared to WT plants after chilling treatment. Increased MDA content usually acts as a marker of lipid peroxidation [81], especially under various abiotic stresses [82,83], including chilling stress. Apart from rice, MDA levels were also employed to evaluate chilling tolerance in other plants, such as bermudagrass (Cynodon dactylon) [84] and Pyrus betulaefolia [85]. Collectively, these results strongly demonstrate that OsPIN9 is involved in regulating rice chilling tolerance, which is probably mediated by influencing auxin transport. Over the last decade, significant progress has been made in the dissection of the underlying molecular mechanisms of plant chilling adaptation. At least the CBF/DREB-dependent cold-signaling pathway, the Ca 2+ signaling pathway, the ROS-dominated adaptation mechanism, and the phytohormone regulation mechanism [1,9,86] have been demonstrated to regulate plant cold adaptation. Several lines of evidence have strongly suggested that overexpression of OsPIN9 notably impairs rice chilling tolerance ( Figures 5 and 6), while how OsPIN9 influences plant chilling tolerance is completely unknown. To this end, we first examined whether the impaired chilling tolerance of OE lines is associated with the CBF/DREB-dependent cold signaling pathway and the Ca 2+ signaling pathway by employing qRT-PCR technology. The expression of OsDREB1A, OsDREB1B, and OsDREB1C was detected before and after chilling stress due to the sharp induction of these genes under chilling conditions [19,67]. Surprisingly, the OsDREB1 genes, OsTPP1, COLD1, and OsCNGC9 were all induced before treatment and kept at similar or even lower levels in OE lines compared to those in WT plants after chilling treatment (Figure 7). Considering that OsDREB1 genes are mainly induced by low temperatures [19], and higher expression of OsTPP1, COLD1, and OsCNGC9 also facilitates plant cold adaptation [7,14,68], we speculate that a stress response probably occurred in OE lines before chilling treatment. In contrast, all genes show a similar or lower expression level in OE lines compared to WT plants under chilling conditions, suggesting that the DREB regulon and Ca 2+ signaling might be implicated in the impaired chilling tolerance of OE lines. Further investigations are needed to address this speculation.
Next, we assayed ROS levels to clarify the role of ROS homeostasis in WT and OE lines under chilling stress. Chilling stress triggers ROS accumulation [21,87]. Low-level H 2 O 2 can act as a signal to trigger the stress response at the early stress stage, whereas at the later stress stage, high-level H 2 O 2 can damage plant cells [10,87]. Surprisingly, we only observed the accumulation of H 2 O 2 in OE leaves at both early and late chilling stages, while O − 2 was not accumulated during the chilling stages (Figure 8), indicating that ROS homeostasis is disturbed in OE plant cells and that H 2 O 2 rather than O − 2 causes cell damage in OE lines. It has been demonstrated that ROS homeostasis plays a vital role in plant abiotic stress adaptations [88,89], and ROS content must be carefully controlled by many ROS scavenging and detoxification systems during stress conditions [10]. Antioxidant enzyme activity measurement showed that SOD and CAT activities were significantly decreased and increased in OE lines compared to those in WT plants at the early and late chilling stages, respectively, while POD activity was always strikingly higher than that in WT plants ( Figure 9B,C), even before chilling treatment ( Figure 9A), indicating that the prompt trigger of antioxidant enzymes, mainly SOD and CAT in OE lines, plays a vital role in quenching accumulated ROS at the early chilling stage. Additionally, although antioxidant enzymes showed significantly higher activities in OE lines than in WT plants at the late chilling stage ( Figure 9C), which could not detoxify excess H 2 O 2 quickly ( Figure 8C), indicating that the delayed induction of the antioxidant enzymes is not enough to scavenge ROS at the late chilling stage, and the constantly high levels of H 2 O 2 then damage cells.
To further evaluate the source of H 2 O 2 in OE lines under chilling conditions, we assayed the expression of OsRboh genes before and after chilling stress. Since only a slight increase of OsRboh1, OsRboh2, and OsRboh9 was detected in OE lines before chilling stress, it is reasonable that less ROS accumulation was observed under normal conditions. In contrast, more OsRboh genes were induced after chilling treatment, indicating that OsRboh is at least in part involved in ROS production after chilling stress. We also noticed that these induced OsRboh genes were still kept at low expression levels (up to about twofold) even after chilling treatment ( Figure 10). Although evidence has shown that Rboh plays a vital role in ROS production under various abiotic stress conditions [70,71], we cannot rule out the possibility that other ROS-production sources probably exist that contribute to the ROS burst after chilling stress in OE lines.

Plant Materials, Growth Conditions, and Chilling Treatment
Rice japonica variety Nipponbare was employed for the physiological experiments and rice transformation. Hydroponic experiments were performed according to our previous report [90]. Briefly, rice seeds were sterilized first and then cultured in darkness for 3-4 days at 30 • C. The germinated seeds were transferred to Kimura B complete nutrient solution in plant growth chambers with a 12 h of light (30 • C)/12 h of darkness (25 • C) photoperiod and 60-70% relative humidity.
To assay the responses of OsPIN genes to chilling treatment, 14 day old seedlings were transferred to low temperatures (4 • C), and root samples were collected at the indicated time points. The samples were then frozen quickly using liquid nitrogen and preserved at −80 • C for gene expression analysis.
To evaluate rice chilling tolerance, 14 day old WT and OE seedlings were transferred to 4 • C for 36 h and 57 h, followed by a 3 d recovery, and then the survival rate and physiological indicators were analyzed.

Vector Construction and Generation of the Transgenic Plants
The full-length OsPIN9 gene was amplified by PrimeSTAR HS DNA Polymerase (TaKaRa Biotechnology Co., Ltd., Dalian, China) from rice cDNA using the specific primers listed in Table S1. A plant expression vector, pCAMBIA1301-pUbi, which was kindly provided by Dr. Yao-Guang Liu (College of Life Sciences, South China Agricultural University, Guangzhou, China), was used for overexpressing OsPIN9. To insert into the expression vector pCAMBIA1301-pUbi, the restriction sites BamH I and Kpn I were added at the 5 end of forward and reverse primers, respectively. The recombinant construct was named pUbi:OsPIN9 and confirmed by restriction enzyme digestion and sequencing. The constructed plasmid was then transformed into Agrobacterium tumefaciens (EHA105) and employed for rice transformation according to the previous report [91]. Homozygous T 3 -generation transgenic plants were used for further studies.

Quantitative RT-PCR
qRT-PCR was performed according to a previous report [39]. In short, total RNA was extracted from the collected samples using RNAiso Plus (TAKARA Bio Inc., Dalian, China), and reverse transcription was conducted using HiScript III RT SuperMix for qPCR (Nanjing Vazyme Biotech Company, Ltd., Nanjing, China). The online website INTEGRATED DNA TECHNOLOGIES (https://sg.idtdna.com, accessed on 1 May 2019) was used to design the gene-specific primers. qRT-PCR was conducted using AceQ Universal SYBR qPCR Master Mix (Nanjing Vazyme Biotech Company, Ltd.) and the Lightcycle ® 96 system. At least three biological replicates and three technical repetitions were performed to assay the gene expression. The OsACTIN1 gene (Os03g0718100) was employed as an internal control. All primers used in this study are listed in Table S1, and gene names and ID numbers used for qRT-PCR in this study are listed in Table S2.

Exogenous NPA Treatment
NPA is usually employed as a classical inhibitor of polar auxin transport to elucidate the underlying mechanisms of plant growth and development associated with polar auxin transport [65]. For analyzing the responses of WT and OE lines to NPA treatment, germinated seeds were transferred into Kimura B complete nutrient solution containing 0.5 µM NPA and cultured for 7 d, and then the shoot height, root length, and adventitious root number were assessed. WT and OE lines cultured under normal conditions were used as controls in this experiment.

Physiological Analysis
Trypan blue staining was conducted as described previously [92] with minor modifications. Rice leaves were sampled before and after chilling treatment and stained with a lactophenol-trypan blue solution (10 mL lactic acid, 10 mL glycerol, 10 g phenol, and 10 mg trypan blue, mixed in 10 mL distilled water). The leaves were boiled in a water bath for 10 min in trypan blue solution and then kept at room temperature for 1 h. Decolorization was performed using chloral hydrate solution (25 g chloral hydrate dissolved in 10 mL distilled water). Electrolyte conductivity and MDA content were detected according to the previous report [93]. 3 3-diaminobenzidine (DAB) and nitro blue tetrazolium (NBT) staining were performed as previously reported [21]. Briefly, rice leaves were first cut into sections (about 2 cm in length) and then soaked in DAB and NBT solutions, respectively, overnight at 37 • C. Alcohol (95%) was used for decolorization. CAT, POD, and SOD activities were assayed according to the previous methods [94] with minor modifications. In short, 0.1 g of rice leaves were broken with liquid nitrogen and then homogenized in 1 mL of extraction buffer (50 mM phosphate, pH 7.8). After centrifugation, the supernatant was used for enzyme activity assays. CAT activity was measured in a reaction solution consisting of 50 mM phosphate buffer (pH 7.0), 0.2% H 2 O 2 , and enzyme extract. POD activity was measured in a reaction mixture (0.2% H 2 O 2 , 0.2% guaiacol, and enzyme extract) at 470 nm. SOD activity was assayed in a reaction mixture containing 50 mM phosphate, pH 7.8, 130 mM L-methionine, 750 µM NBT, 100 µM EDTA, 20 µM riboflavin, and enzyme extract at 560 nm. The Coomassie Brilliant Blue G-250 staining method was employed to measure protein content [95].

Data Analysis
Each experiment was biologically replicated at least three times. The one-way analysis of variance (ANOVA) method was used to statistically analyze experimental data by GraphPad PRISM 8 version 8.0.2 (GraphPad Software Inc., San Diego, CA, USA) at the significance levels of p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***), and all data are given as means ± SD.

Conclusions
In conclusion, in this report, we systematically monitored the responses of OsPIN genes to chilling treatment and generated transgenic rice plants with high expression levels of OsPIN9. The transgenic plants showed impaired chilling tolerance compared to WT plants, which is probably mainly caused by the sharp accumulation of H 2 O 2 and delayed induction of SOD and CAT. OsRboh genes, at least in part, contribute to the production of ROS in OE lines under chilling stress conditions. The DREB regulon and Ca 2+ signaling pathway are implicated in the chilling tolerance of OE lines, while the detailed functional mechanisms need further investigation. Several issues need further discussion, for example, how the relationship between auxin homeostasis and ROS homeostasis is derived and where the accumulated H 2 O 2 is derived from? What is the role of O − 2 and POD in OE lines under chilling stress? Further investigation focused on these detailed issues may lead to insight into the underlying molecular mechanisms of rice chilling tolerance and may provide potential targets for breeding chilling-resistant crops.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/plants12152809/s1, Figure S1: Adventitious root number analysis in the transgenic A2 line; Table S1: Primers used in this study; Table S2: Gene names and ID numbers used for qRT-PCR in this study.