Roles of a Cysteine Desulfhydrase LCD1 in Regulating Leaf Senescence in Tomato

Hydrogen sulfide (H2S), a novel gasotransmitter in both mammals and plants, plays important roles in plant development and stress responses. Leaf senescence represents the final stage of leaf development. The role of H2S-producing enzyme L-cysteine desulfhydrase in regulating tomato leaf senescence is still unknown. In the present study, the effect of an L-cysteine desulfhydrase LCD1 on leaf senescence in tomato was explored by physiological analysis. LCD1 mutation caused earlier leaf senescence, whereas LCD1 overexpression significantly delayed leaf senescence compared with the wild type in 10-week tomato seedlings. Moreover, LCD1 overexpression was found to delay dark-induced senescence in detached tomato leaves, and the lcd1 mutant showed accelerated senescence. An increasing trend of H2S production was observed in leaves during storage in darkness, while LCD1 deletion reduced H2S production and LCD1 overexpression produced more H2S compared with the wild-type control. Further investigations showed that LCD1 overexpression delayed dark-triggered chlorophyll degradation and reactive oxygen species (ROS) accumulation in detached tomato leaves, and the increase in the expression of chlorophyll degradation genes NYC1, PAO, PPH, SGR1, and senescence-associated genes (SAGs) during senescence was attenuated by LCD1 overexpression, whereas lcd1 mutants showed enhanced senescence-related parameters. Moreover, a correlation analysis indicated that chlorophyll content was negatively correlated with H2O2 and malondialdehyde (MDA) content, and also negatively correlated with the expression of chlorophyll degradation-related genes and SAGs. Therefore, these findings increase our understanding of the physiological functions of the H2S-generating enzyme LCD1 in regulating leaf senescence in tomato.


Introduction
Leaf senescence represents the final stage of leaf development, which is a genetically controlled process [1]. As leaves age, the decomposition of chloroplast is initiated, accompanied by the catabolism of macromolecules including nucleic acids, proteins, and lipids. The decomposed nutrients then transfer to other developing organs, such as young leaves and growing fruit [2]. Chloroplasts constitute approximately 70% of the total proteins in green leaves and chlorophyll degradation causes the first visible signs of leaf senescence [3]. Thus, the coordinated degradation of chlorophyll is crucial for the breakdown of chloroplasts. Terrestrial plants utilize two types of chlorophyll species (i.e., chlorophyll a and chlorophyll b) for photosynthesis [4]. Chlorophyll b has to be converted to chlorophyll a before it can be processed into the degradation pathway and NON-YELLOW COLORING 1 (NYC1) catalyzes the reduction of chlorophyll b to 7-hydroxymethyl chlorophyll a [5]. Chlorophyll a

Role of LCD1 in Regulating Tomato Leaf Senescence
To elucidate the possible involvement of LCD1 in regulating leaf senescence, two previously reported tomato lines, lcd1-7 and lcd1-9, with mutations near the PAM sequence were used as lcd1 mutants, and two lines (hereafter called LCD1-oe and LCD1-oe1) with an increased expression of LCD1 under the control of the CaMV 35S promoter were also used. The overexpression efficacy of five LCD1 overexpression lines was verified by RT-qPCR as shown in Figure S1 and the lines LCD1-oe and LCD1-oe1, which showed a higher LCD1 expression, were used in the present study. To confirm the role of LCD1 in catalyzing H 2 S production, the H 2 S producing rates were determined in leaves of lcd1 and LCD1-oe lines. The data in Figure 1B suggest that lcd1 leaves had a lower H 2 S producing rate compared with the wild type, while LCD1 overexpression induced a significantly higher level of the H 2 S producing rate. Besides, H 2 S production was also evaluated by lead acetate H 2 S detection strips, and the results showed that lcd1 leaves produced less H 2 S and LCD1 overexpression produced more H 2 S ( Figure 1C). After 10 weeks of growth, the LCD1 rate compared with the wild type, while LCD1 overexpression induced a significantly higher level of the H2S producing rate. Besides, H2S production was also evaluated by lead acetate H2S detection strips, and the results showed that lcd1 leaves produced less H2S and LCD1 overexpression produced more H2S ( Figure 1C). After 10 weeks of growth, the LCD1 mutation caused earlier leaf senescence compared with the wild type. In contrast, LCD1 overexpression significantly delayed leaf senescence ( Figure 1D). producing rate in the mature leaves from wild type, lcd1, and LCD1-oe lines of 10 weeks growth. (C) H2S production from the mature leaves of 10-week-old wild type, lcd1, and LCD1-oe lines detected by lead acetate H2S detection strips (Sigma-Aldrich). (D) The leaves of different tomato lines in (A) were detached and photographed. Data are means of three biological replicates ± standard deviation (SD). The symbols * and ** stand for p < 0.05 and p < 0.01 as determined by the Student's t-test, respectively.

LCD1 Participates in Dark-Induced Senescence
Leaf senescence is an important phenomenon in the growth and development of plant leaves, and darkness is widely used as a tool to induce senescence in detached leaves. To study the role of LCD1 in dark-induced senescence, the mature leaves without visible senescence of 6-week-old wild type, lcd1 mutant, and LCD1-oe were stored in darkness for 8 days. As shown in Figure 2A, lcd1 showed the obvious syndrome of the leaf yellowing phenotype after 5 and 8 days in dark stress, whereas LCD1 overexpression still maintained the green phenotype. To study the kinetics of tomato leaf H2S production during Phenotype of 10-week-old wild-type (WT), lcd1-7, lcd1-9, LCD1-oe, and LCD1-oe1 plants. (B) H 2 S producing rate in the mature leaves from wild type, lcd1, and LCD1-oe lines of 10 weeks growth. (C) H 2 S production from the mature leaves of 10-week-old wild type, lcd1, and LCD1-oe lines detected by lead acetate H 2 S detection strips (Sigma-Aldrich). (D) The leaves of different tomato lines in (A) were detached and photographed. Data are means of three biological replicates ± standard deviation (SD). The symbols * and ** stand for p < 0.05 and p < 0.01 as determined by the Student's t-test, respectively.

LCD1 Participates in Dark-Induced Senescence
Leaf senescence is an important phenomenon in the growth and development of plant leaves, and darkness is widely used as a tool to induce senescence in detached leaves. To study the role of LCD1 in dark-induced senescence, the mature leaves without visible senescence of 6-week-old wild type, lcd1 mutant, and LCD1-oe were stored in darkness for 8 days. As shown in Figure 2A, lcd1 showed the obvious syndrome of the leaf yellowing phenotype after 5 and 8 days in dark stress, whereas LCD1 overexpression still maintained the green phenotype. To study the kinetics of tomato leaf H 2 S production during senescence, H 2 S production in the leaves at different developmental stages-young leaves (YL), mature leaves (ML), senescent leaves (SL), and late senescent leaves (LS)-was evaluated and the H 2 S detection strips showed browning with senescence, suggesting H 2 S production increased during leaf senescence ( Figure S2). Moreover, H 2 S production in leaves of wild-type (WT), lcd1, and LCD1-oe tomatoes were also determined during dark-induced senescence ( Figure 2B). Generally, an increasing trend of H 2 S production was observed in all samples during storage, while LCD1-oe leaves showed a higher H 2 S production compared with the wild-type control. In addition, the lcd1 mutant produced a significantly lower level of H 2 S compared with the wild type. Therefore, it can be concluded that LCD1 deletion caused a lower H 2 S release and the attenuated H 2 S release may cause an accelerated senescence in the lcd1 mutant. Overall, the present results indicate that LCD1 plays a negative role in leaf senescence in both developmental and dark-induced senescence. senescence, H2S production in the leaves at different developmental stages-young leaves (YL), mature leaves (ML), senescent leaves (SL), and late senescent leaves (LS)-was evaluated and the H2S detection strips showed browning with senescence, suggesting H2S production increased during leaf senescence ( Figure S2). Moreover, H2S production in leaves of wild-type (WT), lcd1, and LCD1-oe tomatoes were also determined during dark-induced senescence ( Figure 2B). Generally, an increasing trend of H2S production was observed in all samples during storage, while LCD1-oe leaves showed a higher H2S production compared with the wild-type control. In addition, the lcd1 mutant produced a significantly lower level of H2S compared with the wild type. Therefore, it can be concluded that LCD1 deletion caused a lower H2S release and the attenuated H2S release may cause an accelerated senescence in the lcd1 mutant. Overall, the present results indicate that LCD1 plays a negative role in leaf senescence in both developmental and dark-induced senescence.

Effect of LCD1 on Dark-Triggered Chlorophyll Degradation and Reactive Oxygen Species Accumulation in Detached Tomato Leaves
Chlorophyll degradation is the one of the most significant changes during leaf senescence; thus, chlorophyll contents were determined in wild-type, lcd1 mutant, and LCD1-oe leaves during dark-induced senescence. As shown in Figure 3A, the content of total chlorophyll in the wild type decreased gradually during storage in darkness for 8 days, whereas the content of chlorophyll in the lcd1 mutant showed an obvious decrease on days 5 and 8 under darkness, and the value on day 8 was about 32.6% of the initial value on day 0. In contrast, LCD1 overexpression maintained a relatively higher chlorophyll content compared with the wild type and the lcd1 mutant on days 5 and 8 under darkness. After 8 days in darkness, the chlorophyll content in LCD1 overexpression decreased to 84.6% of the initial value, suggesting the role of LCD1 in delaying dark-induced senescence. As shown in Figure 3B, there were minor changes in the chlorophyll a content between different groups during storage. Moreover, only a slight decrease in chlorophyll a was observed during dark-induced senescence, except for a significant decline found in lcd1 on day 8. Figure 3C shows the change pattern of chlorophyll b content in wild type, lcd1 mutant, and LCD1-oe during dark-induced senescence. With the increase of storage days, the chlorophyll b content decreased in each group. At day 0, chlorophyll b content in lcd1 leaves was about 53.4% of that in the LCD1-oe group, and decreased to 21.3% on day 8 compared with the value on day 0. Furthermore, the ratio of chlorophyll a/b was also evaluated in dark-stored detached leaves of wild-type, lcd1, and LCD1-oe tomatoes for 0, 2, 5, and 8 days. As shown in Figure 3D, the ratio of chlorophyll a/b in WT and lcd1 mutant leaves increased during storage, while LCD1 deletion caused the highest ratio compared with other groups. In contrast, the ratio of chlorophyll a/b in LCD1 overexpression almost remained unchanged. The above results indicate that the lcd1 mutation accelerated dark-induced leaf senescence and LCD1 overexpression delayed leaf yellowing and chlorophyll degradation.

Effect of LCD1 on Dark-Triggered Chlorophyll Degradation and Reactive Oxygen Species Accumulation in Detached Tomato Leaves
Chlorophyll degradation is the one of the most significant changes during leaf senescence; thus, chlorophyll contents were determined in wild-type, lcd1 mutant, and LCD1oe leaves during dark-induced senescence. As shown in Figure 3A, the content of total chlorophyll in the wild type decreased gradually during storage in darkness for 8 days, whereas the content of chlorophyll in the lcd1 mutant showed an obvious decrease on days 5 and 8 under darkness, and the value on day 8 was about 32.6% of the initial value on day 0. In contrast, LCD1 overexpression maintained a relatively higher chlorophyll content compared with the wild type and the lcd1 mutant on days 5 and 8 under darkness. After 8 days in darkness, the chlorophyll content in LCD1 overexpression decreased to 84.6% of the initial value, suggesting the role of LCD1 in delaying dark-induced senescence. As shown in Figure 3B, there were minor changes in the chlorophyll a content between different groups during storage. Moreover, only a slight decrease in chlorophyll a was observed during dark-induced senescence, except for a significant decline found in lcd1 on day 8. Figure 3C shows the change pattern of chlorophyll b content in wild type, lcd1 mutant, and LCD1-oe during dark-induced senescence. With the increase of storage days, the chlorophyll b content decreased in each group. At day 0, chlorophyll b content in lcd1 leaves was about 53.4% of that in the LCD1-oe group, and decreased to 21.3% on day 8 compared with the value on day 0. Furthermore, the ratio of chlorophyll a/b was also evaluated in dark-stored detached leaves of wild-type, lcd1, and LCD1-oe tomatoes for 0, 2, 5, and 8 days. As shown in Figure 3D, the ratio of chlorophyll a/b in WT and lcd1 mutant leaves increased during storage, while LCD1 deletion caused the highest ratio compared with other groups. In contrast, the ratio of chlorophyll a/b in LCD1 overexpression almost remained unchanged. The above results indicate that the lcd1 mutation accelerated dark-induced leaf senescence and LCD1 overexpression delayed leaf yellowing and chlorophyll degradation. Data are means of three biological replicates ± standard deviation (SD). Different letters above the columns stand for significant difference between two values (p < 0.05) at the same time point. Leaf senescence is usually associated with the excessive accumulation of ROS; therefore, the levels of H 2 O 2 and malondialdehyde (MDA) were monitored in wild-type, lcd1 mutant, and LCD1-oe leaves during dark-induced senescence. As shown in Figure 4A, there was no significant difference in H 2 O 2 content between the different groups on day 0. During the dark-induced senescence, the H 2 O 2 content in each group showed an increasing trend, of which the lcd1 group increased the fastest, followed by the wild-type and LCD1-oe group. However, H 2 O 2 content in the LCD1-oe group increased slowly compared with other groups. As shown in Figure 4B, the change of MDA content among the groups also showed a similar trend to H 2 O 2 . The content of MDA in lcd1 leaves during storage was the highest compared with other groups, and the lowest MDA content was observed in LCD1-oe leaves. Therefore, it can be concluded that the overexpression of LCD1 could reduce the accumulation of ROS and MDA in leaves under dark-triggered senescence. Leaf senescence is usually associated with the excessive accumulation of ROS; therefore, the levels of H2O2 and malondialdehyde (MDA) were monitored in wild-type, lcd1 mutant, and LCD1-oe leaves during dark-induced senescence. As shown in Figure 4A, there was no significant difference in H2O2 content between the different groups on day 0. During the dark-induced senescence, the H2O2 content in each group showed an increasing trend, of which the lcd1 group increased the fastest, followed by the wild-type and LCD1-oe group. However, H2O2 content in the LCD1-oe group increased slowly compared with other groups. As shown in Figure 4B, the change of MDA content among the groups also showed a similar trend to H2O2. The content of MDA in lcd1 leaves during storage was the highest compared with other groups, and the lowest MDA content was observed in LCD1-oe leaves. Therefore, it can be concluded that the overexpression of LCD1 could reduce the accumulation of ROS and MDA in leaves under dark-triggered senescence.

Effect of LCD1 on the Expressions of Genes Related to Chlorophyll Degradation in Detached Tomato Leaves
Chlorophyll degradation marks the senescence stage of leaves. In order to explore the molecular mechanism of the differences in chlorophyll content of lcd1, LCD1-oe, and wild-type leaves during dark-induced senescence, the expression levels of key genes NYC1, PAO, PPH, and SGR1 in the chlorophyll degradation pathway were analyzed by RT-qPCR. The present data showed NYC1 was transcriptionally induced during dark-induced senescence in all groups ( Figure 5A). In accordance with the early senescence phenotype of the lcd1 mutant and late senescence in LCD1-oe leaves, the expression of NYC1 was significantly higher in the lcd1 mutant and was less expressed in LCD1-oe leaves during dark storage. Three other genes-PAO ( Figure 5B), PPH ( Figure 5C), and SGR1 ( Figure  5D)-were also analyzed at the transcriptional level in wild-type, lcd1 mutant, and LCD1oe leaves during dark-induced senescence and similar changes to that of the NYC1 expression were observed. The higher expression of PAO, PPH, and SGR1 in lcd1 and lower expression in LCD1-oe again supported the role of LCD1 in delaying leaf senescence. The results suggest that LCD1 may delay the chlorophyll degradation by down-regulating the transcription of key genes in the chlorophyll degradation pathway. Data are means of three biological replicates ± standard deviation (SD). Different letters above the columns stand for significant difference between two values (p < 0.05) at the same time point.

Effect of LCD1 on the Expressions of Genes Related to Chlorophyll Degradation in Detached Tomato Leaves
Chlorophyll degradation marks the senescence stage of leaves. In order to explore the molecular mechanism of the differences in chlorophyll content of lcd1, LCD1-oe, and wildtype leaves during dark-induced senescence, the expression levels of key genes NYC1, PAO, PPH, and SGR1 in the chlorophyll degradation pathway were analyzed by RT-qPCR. The present data showed NYC1 was transcriptionally induced during dark-induced senescence in all groups ( Figure 5A). In accordance with the early senescence phenotype of the lcd1 mutant and late senescence in LCD1-oe leaves, the expression of NYC1 was significantly higher in the lcd1 mutant and was less expressed in LCD1-oe leaves during dark storage. Three other genes-PAO ( Figure 5B), PPH ( Figure 5C), and SGR1 ( Figure 5D)-were also analyzed at the transcriptional level in wild-type, lcd1 mutant, and LCD1-oe leaves during dark-induced senescence and similar changes to that of the NYC1 expression were observed. The higher expression of PAO, PPH, and SGR1 in lcd1 and lower expression in LCD1-oe again supported the role of LCD1 in delaying leaf senescence. The results suggest that LCD1 may delay the chlorophyll degradation by down-regulating the transcription of key genes in the chlorophyll degradation pathway.

Effect of LCD1 on the Expressions of SAGs in Detached Tomato Leaves
To further analyze the senescence-alleviating role of LCD1, we conducted an RT-qPCR analysis to evaluate the expression patterns of senescence-associated genes (SAGs) in lcd1, LCD1-oe, and wild-type leaves during dark-induced senescence. As shown in Figure 6, SAG12, SAG15, and SAG113 were transcriptionally induced during dark-induced senescence. Compared with SAG15 and SAG113, SAG12 showed more fold changes during leaf senescence, which was 109.6 times in the wild type on day 8 compared with day 0 ( Figure 6A). In accordance with the early senescence phenotype of the lcd1 mutant and late senescence in LCD1-oe leaves, the expression of the three SAGs was significantly higher in the lcd1 mutant and was less expressed in LCD1-oe leaves during dark storage, especially on day 8.

Effect of LCD1 on the expressions of SAGs in Detached Tomato Leaves
To further analyze the senescence-alleviating role of LCD1, we conducted an RT-qPCR analysis to evaluate the expression patterns of senescence-associated genes (SAGs) in lcd1, LCD1-oe, and wild-type leaves during dark-induced senescence. As shown in Figure 6, SAG12, SAG15, and SAG113 were transcriptionally induced during dark-induced senescence. Compared with SAG15 and SAG113, SAG12 showed more fold changes during leaf senescence, which was 109.6 times in the wild type on day 8 compared with day 0 ( Figure 6A). In accordance with the early senescence phenotype of the lcd1 mutant and late senescence in LCD1-oe leaves, the expression of the three SAGs was significantly higher in the lcd1 mutant and was less expressed in LCD1-oe leaves during dark storage, especially on day 8.

Correlation Analysis of Different Leaf Physiological Indexes and Senescence-Related Gene Expression
The correlation among total chlorophyll, chlorophyll a, chlorophyll b, chlorophyll a/b, H 2 O 2 , and MDA content and the gene expression of NYC1, PAO, PPH, SGR1, SAG12, SAG15, and SAG113 was analyzed to investigate the potential relations among the indexes. As shown in Figure 7, chlorophyll content was negatively correlated with H 2 O 2 , MDA content, and with the expression of chlorophyll degradation-related genes NYC1, PAO, PPH, and SGR1 and senescence-related genes SAG12, SAG15, and SAG113. Moreover, total chlorophyll and chlorophyll b showed a higher negative correlation to ROS content and senescence-related gene expression in comparison to chlorophyll a. The contents of H 2 O 2 and MDA were positively correlated with the expression levels of NYC1, PAO, PPH, SGR1, SAG12, SAG15, and SAG113. Among them, the expressions of PPH and NYC1 were highly positively correlated (r = 0.966), and the total chlorophyll content was highly negatively correlated with the H 2 O 2 content (r = −0.882). Moreover, chlorophyll a/b also showed a significant positive correlation with H 2 O 2 and MDA content. Through these analyses, the positive correlation between H 2 O 2 /MDA content and senescence-related gene expressions indicates that they may act synergistically to accelerate the senescence process of leaves. , lcd1, and LCD1-oe tomatoes stored in darkness for 0, 2, 5, and 8 days. Data are means of three biological replicates ± standard deviation (SD). Different letters above the columns stand for significant difference between two values (p < 0.05) at the same time point.

Correlation Analysis of Different Leaf Physiological Indexes and Senescence-Related Gene Expression
The correlation among total chlorophyll, chlorophyll a, chlorophyll b, chlorophyll a/b, H2O2, and MDA content and the gene expression of NYC1, PAO, PPH, SGR1, SAG12, SAG15, and SAG113 was analyzed to investigate the potential relations among the indexes. As shown in Figure 7, chlorophyll content was negatively correlated with H2O2, MDA content, and with the expression of chlorophyll degradation-related genes NYC1, PAO, PPH, and SGR1 and senescence-related genes SAG12, SAG15, and SAG113. Moreover, total chlorophyll and chlorophyll b showed a higher negative correlation to ROS content and senescence-related gene expression in comparison to chlorophyll a. The contents of H2O2 and MDA were positively correlated with the expression levels of NYC1, PAO, PPH, SGR1, SAG12, SAG15, and SAG113. Among them, the expressions of PPH and NYC1 were highly positively correlated (r = 0.966), and the total chlorophyll content was highly negatively correlated with the H2O2 content (r = −0.882). Moreover, chlorophyll a/b also showed a significant positive correlation with H2O2 and MDA content. Through these analyses, the positive correlation between H2O2/MDA content and senescence-related gene expressions indicates that they may act synergistically to accelerate the senescence process of leaves.

Principal Component Analysis of Different Leaf Physiological Indexes and Senescence-Related Gene Expression
The principal component analysis (PCA) was performed based on the data of chlorophyll, H 2 O 2 , and MDA content and the expressions of chlorophyll degradation-related genes and SAGs. As shown in Figure 8, PC1 and 2 contributed to 81.1% and 10.4% of the variability of the data, respectively. It can be seen that lcd1 2 d and lcd1 5 d clustered together, and lcd1 8 d was distributed separately from other groups. The variety showing the highest positive load value in the direction of PC1 was LCD1-oe 0 d and the variety that showed the lowest load value in the direction of PC2 was lcd1 5 d. Therefore, it could be concluded that a decrease in the endogenous H 2 S content in the lcd1 mutant caused significant changes during dark-induced senescence compared with other groups.

Principal Component Analysis of Different Leaf Physiological Indexes and Senescence-Related Gene Expression
The principal component analysis (PCA) was performed based on the data of chlorophyll, H2O2, and MDA content and the expressions of chlorophyll degradation-related genes and SAGs. As shown in Figure 8, PC1 and 2 contributed to 81.1% and 10.4% of the variability of the data, respectively. It can be seen that lcd1 2 d and lcd1 5 d clustered together, and lcd1 8 d was distributed separately from other groups. The variety showing the highest positive load value in the direction of PC1 was LCD1-oe 0 d and the variety that showed the lowest load value in the direction of PC2 was lcd1 5 d. Therefore, it could be concluded that a decrease in the endogenous H2S content in the lcd1 mutant caused significant changes during dark-induced senescence compared with other groups.
Recently it was reported that a des1 mutant displayed accelerated leaf senescence, while the leaves of OE-DES1 contained adequate chlorophyll levels, suggesting the role of DES1 in regulating leaf senescence [25]. In our previous work, mutation of an L-Cys desulfhydrase named LCD1 caused accelerated fruit ripening compared with the wild type [26]. In the present study, to elucidate the possible involvement of LCD1 in regulating leaf senescence, two previously reported tomato lines, lcd1-7 and lcd1-9, with mutations near the PAM sequence were used as lcd1 mutants, and two lines (hereafter called LCD1-oe and LCD1-oe1) with an increased expression of LCD1 under the control of the CaMV 35S promoter were also used. To confirm the role of LCD1 in catalyzing H 2 S production, the H 2 S producing rates were determined in leaves of the lcd1 and LCD1-oe lines; lcd1 leaves showed a lower H 2 S producing rate compared with the wild type, while LCD1 overexpression induced a significantly higher level of the H 2 S producing rate, suggesting the efficiency of LCD1 in producing H 2 S.
Leaf senescence is a highly programmed degeneration process during the final stage of leaf development. To study the role of H 2 S-producing enzyme LCD1 in regulating natural leaf senescence, we compared 10-week old tomatoes of lcd1 mutants, and the two lines of LCD1-oe with the enhanced expression of LCD1, and found that lcd1 developed more senescence syndrome while LCD1 overexpression maintained more functional leaves.
Prolonged darkness is often used to initiate rapid and synchronous senescence in detached leaves [13]. The roles of LCD1 on dark-induced senescence were evaluated in tomato leaves. The lcd1 mutant leaves showed an obvious syndrome of the leaf yellowing phenotype after 5 and 8 days in dark stress, whereas LCD1 overexpression still maintained the green phenotype ( Figure 2A). All this evidence suggests the role of H 2 S in alleviating the dark-induced senescence of detached leaves. In accordance with the phenotype of accelerated senescence in the lcd1 mutant and delayed senescence in LCD1-oe leaves, chlorophyll decreased significantly in lcd1, but the decrease was attenuated in LCD1-oe leaves. Interestingly, we found that chlorophyll b may contribute more to the decrease in total chlorophyll compared with chlorophyll a (Figure 3). By analyzing the ratio of chlorophyll a/b during leaf storage, it was found that the ratio in LCD1 deletion increased significantly, implying that more chlorophyll b was decomposed to chlorophyll a in the lcd1 mutant. Moreover, the ratio in LCD1 overexpression almost remained unchanged, suggesting the significant impact of H 2 S content on the ratio of chlorophyll a/b. Then, the expression levels of key genes NYC1, PAO, PPH, and SGR1 in the chlorophyll degradation pathway were analyzed by RT-qPCR. It was found that NYC1, PAO, PPH, and SGR1 transcript abundance increased during darkness in all groups, especially in lcd1 mutant leaves, whereas this response was significantly inhibited by LCD1 overexpression (Figure 5). Senescence-associated gene (SAG) 12, 15, and 113 are widely used as molecular markers for leaf senescence [28] and their transcriptions were also analyzed in detached tomato leaves. In agreement with the phenotype, significant increases in SAGs expression were observed in all groups, especially in lcd1 mutant leaves, and the increase was greatly attenuated by LCD1 overexpression (Figure 6). The results suggest that LCD1 may delay the chlorophyll degradation by down-regulating the transcription of key genes in the chlorophyll degradation pathway and SAGs.
Leaf senescence is often associated with the pronounced accumulation of ROS [13]. Recently, it was reported that transcription factor NAC075 delays leaf senescence by deterring ROS accumulation through directly activating the expression of the antioxidant enzyme gene catalase 2 (CAT) in Arabidopsis [14]. To unveil the relations between H 2 S and ROS metabolism in leaf senescence, H 2 O 2 and MDA contents were determined during leaf senescence in darkness. During the dark-induced senescence, the H 2 O 2 and MDA content in each group showed an increasing trend, but the overexpression of LCD1 could reduce the accumulation of ROS and MDA in leaves under dark-triggered senescence. Furthermore, the correlation analysis indicated that the ROS and MDA content showed a higher negative correlation to total chlorophyll and chlorophyll b in comparison to chlorophyll a. The contents of H 2 O 2 and MDA were positively correlated with the expression levels of NYC1, PAO, PPH, SGR1, SAG12, SAG15, and SAG113. The positive correlation between H 2 O 2 and MDA content and senescence-related genes indicates that they may act synergistically to accelerate the senescence process of leaves, whereas LCD1 overexpression delayed leaf senescence by inhibiting ROS accumulation and senescence-related gene expressions. Interestingly, increasing H 2 S production was observed in natural senescence leaves ( Figure S2). We hypothesize that endogenous H 2 S production was activated to counteract the effect of increasing ROS in senescence leaves. In our previous reports, exogenous H 2 S fumigation delayed the postharvest senescence of broccoli in a dose-dependent manner and H 2 S maintained higher levels of chlorophyll, carotenoids, anthocyanin, and ascorbate, suggesting the role of H 2 S in delaying the postharvest senescence of broccoli [21]. Moreover, H 2 S treatment effectively alleviates ethylene-induced banana peel yellowing and fruit softening [22]. The above results suggest that H 2 S is an effective signal in delaying the postharvest senescence of fruits and vegetables. In the present research, an increasing trend of H 2 S production was observed during leaf senescence, suggesting that H 2 S generation may be activated in response to leaf senescence as ROS metabolites (H 2 O 2 and MDA) accumulate during dark-induced senescence. Leaf senescence, once initiated, cannot be stopped. Though more H 2 S is produced during leaf senescence, leaves still undergo senescence during storage in darkness. Moreover, compared with the early senescence phenotype of the lcd1 mutant, LCD1 overexpression induced more H 2 S production and showed a delayed leaf senescence, clearly suggesting the role of H 2 S in delaying leaf senescence. In all, the data suggest that senescent leaves produced more H 2 S, but reduction in H 2 S production of the lcd1 mutant caused early senescence in both natural and dark-induced senescence. Moreover, the principal component analysis (PCA) in Figure 8 shows that lcd1 2 d and lcd1 5 d clustered together, and lcd1 8 d was distributed separately from other groups, suggesting that the decreased endogenous H 2 S content in lcd1 caused significant changes during dark-induced senescence compared with other groups.

Conclusions
In the present study, the role of a cysteine desulfhydrase LCD1 in regulating leaf senescence in tomato was explored. The LCD1 mutation caused an earlier leaf senescence, whereas LCD1 overexpression significantly delayed leaf senescence compared with the wild type in 10-week tomato seedlings. Furthermore, LCD1 was found to play a negative role in dark-induced senescence in detached tomato leaves. Further investigations showed that LCD1 delayed dark-triggered chlorophyll degradation and ROS accumulation in detached tomato leaves, and the increase in chlorophyll degradation and SAGs related gene expression was attenuated by LCD1 overexpression. Moreover, a correlation analysis indicated that chlorophyll content was negatively correlated with H 2 O 2 and MDA content, and also negatively correlated with the expression of chlorophyll degradation-related genes NYC1, PAO, PPH, and SGR1 and senescence-related genes SAG12, SAG15, and SAG113. Therefore, these findings increase our understanding of the physiological functions of the H 2 S-generating enzyme LCD1 in regulating leaf senescence in tomato.

Plant Material and Growth Conditions
Solanum lycopersicum cv. "Micro-Tom" was used as the control in this study. The mutants lcd1-7, which contained a T residue inserted near the PAM sequence, and lcd1-9, which had a deletion of G near the PAM as previously reported were used as lcd1 mutants [26]. The coding sequence of tomato cysteine desulfhydrylase LCD1 (LOC101258894) was ob-tained from NCBI (http://www.ncbi.nlm.nih.gov/, accessed on 11 September 2018) and the primers including the restriction enzyme sites (LCD1-F: CGCGGATCCTAATCCTAAAT GGAACCGGC; LCD1-R: CCGCTCGAGTTCTGAGTGAAGCATCTTAC, the underlines stand for BamHI and XhoI sites, respectively) were used to amplified the coding sequence of LCD1. Then, the coding sequence of LCD1 was inserted into the pBI121 vector at the sites of BamHI and XhoI and transformed into wild-type tomato by Agrobacterium tumefaciens, which contained the recombinant LCD1-pBI121. The efficiency of LCD1 overexpression was verified by RT-qPCR. The seeds of tomatoes were grown in a nutrient soil:vermiculite (3:1, v/v) in growth pots 10 cm in diameter in an environment-controlled growth room (23 ± 2 • C; 50-70% relative humidity, RH) under 16 h light/8 h dark and 250 mol/m 2 /s light.

Determination of H 2 S Producing Rate and H 2 S Detection in Tomato Leaves
The H 2 S producing rate was measured as described previously [29]. Tomato leaves at 1 g were ground to a fine powder in liquid nitrogen and homogenized in 9 mL of 20 mM Tris-HCl, pH 8.0. After centrifugation at 12,000× g for 15 min, the protein content of the supernatant was sampled and the H 2 S producing rate was detected by monitoring the release of H 2 S from L-cysteine in the presence of dithiothreitol (DTT). The assay was performed in a total volume of 1 mL comprising 0.8 mM L-cysteine, 2.5 mM DTT, 100 mM Tris-HCl, pH 8.0, and 100 µL of protein solution. The reaction was incubated for 15 min at 37 • C, and terminated by adding 100 µL of 30 mM FeCl 3 dissolved in 1.2 N HCl and 100 µL of 20 mM N,N-dimethyl-phenylenediamine dihydrochloride dissolved in 7.2 N HCl. The formation of methylene blue was determined at 670 nm.
The end-point detection of H 2 S production from tomato leaves by lead acetate strips (cat. number WHA2602501A, Sigma, Darmstadt, Germany) were performed as previously described [30]. One gram of fresh tomato leaves was ground to a fine powder in liquid nitrogen and then homogenized in 10 mL of Phosphate Buffered Saline (pH 6.8) supplemented with 10 mM L-cysteine and 10 µM pyridoxal-5 -phosphate (PLP), and then the mixture was placed in petri dishes. The H 2 S detection strips were attached to the inner part of the upper lid of the petri dishes and incubated at 37 • C for 2-5 h until lead sulfide darkening of the strip occurred.

Dark Treatment of Tomato Leaves
For dark-induced leaf senescence experiments, detached mature leaves from 6-weekold wild-type, lcd1, and LCD1-oe transgenic plants were placed on filter papers which were moistened by 2 mL of sterile water in petri dishes with a 9 cm diameter and the adaxial side facing upwards. The petri dishes which contained 5-6 detached leaves were kept in darkness at 23 • C for 8 days. The leaves were sampled and rapidly frozen in liquid nitrogen and stored at −80 • C until analysis.

Determination of Chlorophyll Content
Tomato leaves at 2 ± 0.01 g were homogenized in liquid nitrogen and subsequently extracted ethanol and 80% (v/v) acetone solution in a ratio of 1:1 (v/v) according to the method in [31]. The absorbance of the supernatant was read at 663 and 645 nm. The experiments were repeated three biological times, and the chlorophyll levels were expressed as mg/g fresh weight (FW).

Determination of H 2 O 2 and Malondialdehyde (MDA) Content in Tomato Leaves
The contents of H 2 O 2 and malondialdehyde (MDA) were assayed as described by Ge et al. [22] and Hu et al. [27]. For the determination of H 2 O 2 content, 2.0 ± 0.01 g of tomato leaves were homogenized in 3 mL of precooled acetone, and centrifuged at 12,000× g for 30 min. The content of H 2 O 2 was measured by determining the absorbance value at 508 nm. For the determination of MDA content, 2.0 ± 0.01 g of tomato leaf powder was homogenized with 5% trichloroacetic acid, and the supernatant was obtained by centrifugation at 12,000× g for 30 min. The absorbance values were measured at 600 nm, 532 nm, and 450 nm, respectively. The experiments were repeated three times, and the contents of H 2 O 2 and MDA were expressed as µmol/g fresh weight (FW).

Gene Expression Analysis
Total RNA from 0.1 g of tomato leaves was extracted using an Extraction Kit (Tiangen, Beijing, China) and cDNA was synthesized by a reverse transcription kit (PrimeScript RT Master Mix; Takara, Kyoto, Japan). Then the cDNA products were used for gene expression analysis by quantitative polymerase chain reaction (qPCR) performed using a Bio-Rad IQ5 (Hercules, CA, USA). The specific primers used for qPCR were designed based on the coding sequence of the genes as shown in the SGN database (https://solgenomics.net/, accessed on 12 April 2021). Tubulin gene expression in control tomato plants was used for the normalization of the data. The experiments were repeated in three technical replicates.

Data Analysis
The statistical analysis of data was based on Student's t-tests. Significant differences were evaluated using multiple pair wise t-test comparisons at p < 0.05. The correlation among the contents of chlorophyll, H 2 O 2 , MDA, and the expression of chlorophyll degradation related genes and SAGs in tomato leaves and the principal component analysis (PCA) of the data above were analyzed by the tools on the OmicShare platform (https://www.omicshare.com/, accessed on 20 November 2021).