*2.10. E*ff*ect of Chilling Stress on Epidermis of Rice Leaves during Greening*

To investigate whether low temperature affected the epidermal characteristics, we observed the epidermis cells after 48 h of greening. The upper and lower epidermis layers of rice are mainly composed of stomata apparatus and epidermis cells. In addition, there were some trichomes. The shapes and sizes of upper and lower epidermis cells showed no significant difference between the seedlings at 28 ◦C and low temperatures (Figure 11), while stomata under low temperature were smaller than those at 28 ◦C (Table 1). Meanwhile, chilling stress had no significant effect on the number of trichomes (Table 1).

**Figure 11.** The epidermis cells development of control and chill-stressed rice leaves. Six day old etiolated seedlings were treated with 18 ◦C or 12 ◦C chilling stress. Seedlings were harvested at 48 h of greening and their epidermal cells' characteristics were observed. Red arrows and blue arrows represent trichomes and stoma, respectively.


**Table 1.** Effect of chilling stress on stomatal characteristics and trichomes number of rice seedlings.

The values are expressed as mean ± SD (*n* = 50); different letters represent significant difference (*p* < *0.05*).

#### **3. Discussion**

Chlorophyll biosynthesis is affected by various biotic and abiotic factors. Previous studies have reported that water and salt stresses lead to the severe inhibition of chlorophyll biosynthesis during de-etiolation [1,30]. Temperature is one of the major environmental factors that can inhibit chlorophyll biosynthesis and chloroplast biogenesis, and thus affect photosynthesis [31–33]. Our previous study also showed that cold stress dramatically decreases the net photosynthetic rate, stomatal conductance, intercellular CO2 concentration, and water use efficiency in rice seedlings [34]. In this study, our results demonstrated that chlorophyll was significantly increased with the period of light exposure during greening at 28 ◦C (Figure 2C). However, chlorophyll biosynthesis was obviously inhibited in chill-stressed rice seedlings. Our study further found that ALA synthesis was significantly inhibited in chill-stressed seedlings (Figure 3A). That is to say, the early step of chlorophyll biosynthesis was inhibited by the low temperature, which ultimately led to a significant reduction of chlorophyll and heme contents. Similar changes have also been observed in water- and salt-stressed rice/wheat seedlings during early seedling development [1,30]. Meanwhile, GSA-AT activity was reduced by low temperature during de-etiolation (Figure 4A), suggesting that decreased ALA synthesis in chill-stressed rice seedlings might have been due to the downregulated GSA-AT activity.

In the whole chlorophyll synthesis process, Mg-proto IX and Chlide levels were reduced significantly in stressed seedlings (Figure 3F,I,J), and Mg-chelatase and POR activities were decreased synchronously in chill-stressed seedlings (Figure 4C,D). Declines in Mg-chelatase and POR activities were also observed in water-stressed rice [1]. In plants, Mg-chelatase is composed of three non-identical subunits that are encoded by *CHLI*, *CHLD*, and *CHLH* [35,36]. Previous studies have demonstrated that the mutations of *Chl1* and *Chl9* genes which encode CHLD and CHLI could reduce Mg-chelatase activity, and thus inhibit chlorophyll synthesis [37]. Additionally, ATP is required for the catalytic activity of Mg-chelatase, while ferrochelatase is inhibited by ATP. More Pchlide is allocated to the Mg branch when ATP levels are higher in the light; conversely, the Mg branch is inhibited in the dark [8]. In addition, we found that *CHLH* expression level was much higher at 18 ◦C than that at 28 ◦C, but its expression was the lowest at 12 ◦C. *CHLH* expression in etiolated seedlings was suppressed with lower histone acetylation levels by PIF3, but increased rapidly during greening with higher acetylation levels of histones [38]. Chilling stress might repress histone acetylation levels of *CHLH* at 12 ◦C. Numerous studies have indicated that post-translational regulation plays an essential role in chlorophyll biosynthesis [39–41]. The caseinolytic protease activity counteracts the binding of GluTR-binding protein to assure an appropriate content of GluTR and an adequate ALA synthesis at the post-translational level [42]. This might be why the increased transcriptional levels of *DVR* and *CHLH* under chilling stress were not directly proportional to their corresponding products (Figures 2 and 5). Previous studies have also shown that in water- and chill-stressed rice/cucumber seedlings, Mg-chelatase activity and its gene/protein expression were downregulated [1,43]. The light-dependent POR is a plastid (pro)thylakoid-membrane-associated protein, which binds to NADPH and Pchlide to form a ternary complex in etioplasts [44]. There are three POR isoenzymes in *Arabidopsis thaliana*. *PORA* transcripts accumulate in etiolated seedlings, but the expression of *PORA* is strongly downregulated when exposed to light. The *PORB* transcript can be detected throughout the growth and development

of plants, while the expression of *PORC* is induced by light and is predominantly present in fully matured green tissues [8,45]. However, there are only two isoenzymes in rice, namely PORA and PORB [1]. In the present study, POR activity was significantly inhibited in stressed seedlings, especially at 12 ◦C (Figure 4D). *PORA* transcription decreased dramatically when etiolated rice seedlings were exposed to light at 28 ◦C, but chill-treated rice seedlings had a relatively high *PORA* mRNA level (Figure 5). Similarly, POR activity and *PORB* transcript abundance were downregulated in water- and chill-stressed rice/cucumber seedlings [1,43]. However, we found that a slightly lowered temperature (18 ◦C) could significantly increase the expression of *PORB*. As another important tetrapyrrole, heme content showed a 30% drop in seedlings at 12 ◦C compared to that at 28 ◦C after 24 h of greening (Figure 3A), which was far less than the declined proportion of chlorophylls, indicating that Chl biosynthesis is more sensitive to low temperatures than heme synthesis. Meanwhile, this result indicated that more tetrapyrrole metabolic intermediates were allocated to the heme branch than to the chlorophyll branch under chilling stress. Thus, the inhibition of chlorophyll biosynthesis under chilling stress might be attributable to the blocked synthesis of ALA and the inhibition of conversion from Pchlide into Chls.

It has been found that the development of thylakoid is inhibited when chlorophyll biosynthesis is reduced [37]. Chlorophyll content was decreased and thylakoid membrane was not stacked in *porB porC* double mutants [46]. These results indicated that the inhibition of chlorophyll biosynthesis affected the biogenesis of chloroplasts, which led to the reduction of granum lamellae and thylakoid membrane proteins. Thus, a stable supply and correct stoichiometry of chlorophyll are necessary for chloroplast biogenesis. In the present study, chloroplast biogenesis was significantly affected by chilling stress (Figure 7C,D). The grana lamellae were disorganized at 18 ◦C and no grana lamellar structure was formed at 12 ◦C (Figure 7D). The synthesis of thylakoid membrane protein is of importance to the assembly of the photosystem during greening. Western blotting results showed that thylakoid protein synthesis was obviously inhibited under chilling stress (Figure 6). PSII is considered a primary target of photodamage, and the D1 protein is the most vulnerable component in the PSII reaction center under stress conditions [47]. In this study, the content of D1 was obviously lower under low temperatures, and was almost undetectable at 12 ◦C. Moreover, low temperatures greatly inhibited the content of the peripheral antenna proteins of PSII, including Lhcb1, Lhcb2, Lhcb3, Lhcb4 (CP29), Lhcb5 (CP26), and Lhcb6 (CP24) and peripheral antenna proteins of PSI, including Lhca1, Lhca2, Lhca3, Lhca4, and PsaD (Figure 6), especially at 12 ◦C.

Chlorophyll fluorescence analysis has been proven to be a powerful method for obtaining the functional status of PSII [48]. *OsAsr1* rice seedlings have a high value of Fv/Fm, which is correlated with an enhanced cold tolerance [33], suggesting that Fv/Fm could be an indicator of cold tolerance. The decline of Fv/Fm in stressed seedlings may be due to the partial inactivation of PSII reaction centers [49]. In the present study, after 48 h of greening, chill-treated seedlings showed an obvious reduction of Fv/Fm (Figure 8A,B), and the lower Fv/Fm might have been because the low temperature suppressed the assembly and formation of PSII. Meanwhile, the ultrastructural changes of the plastids also indicated that chilling stress affected grana stacking and thylakoid integrity (Figure 6C,D), thereby resulting in a decrease in PSII activity in the stressed plants. Nonphotochemical quenching is a self-protection mechanism in plants. Previous studies have shown that effective heat dissipation in plants can reduce the occurrence of photoinhibition induced by stresses [49]. In this study, the increase of NPQ in stressed seedlings indicated that more excess light energy needed to be dissipated because of the low activity of PSII at low temperatures.

ROS plays double roles under cold stress. On one hand, ROS as a signal can trigger stress-responsive gene expression and the MKK6-MPK3 signaling pathway [26]. On the other hand, excessive ROS in plants can directly induce membrane lipid peroxidation, cell integrity damage, and cell death [50]. PSI and PSII in chloroplasts are the major sources of ROS in plants. Numerous studies have indicated that ROS production increases significantly in plants and the balance of ROS is disturbed under stress conditions [49,51]. Cold signals can be sensed by rice cells through changes in membrane rigidity and osmotic pressure [52]. The membrane rigidity increases under cold stress, resulting in a high electrolyte leakage [53,54]. In accordance with ROS accumulation, the level of lipid peroxidation (MDA) and damage to the cellular membranes were also higher under chilling stress (Figure 9E,F). The increased EL indicated the enhancement of membrane rigidification, which is required for the cold-activated SAMK signaling cascade via cytoskeleton, Ca2<sup>+</sup> fluxes, and CDPKs [55]. Cold stress initially promotes Ca2<sup>+</sup> influx into the cytoplasm, which might be controlled by Ca2<sup>+</sup> channels that are activated by membrane rigidification [56]. Furthermore, the calcium signaling cascade interprets and amplifies the rice-sensing cold signal and subsequently activates the DREB-CRT/DRE pathway, which is important for the cold response [57]. At the same time, our observations showed that leaf cell death was significantly increased under the cold-stress condition during greening (Figure 10). Proline and soluble sugars served as osmoprotectants against oxidative damage, which are also considered indicators to assess the potential cold tolerance of plants [34]. Our previous study indicated that proline and soluble sugar accumulation is enhanced in cold-stressed rice seedlings [34].

Over the past decades, researchers have made extensive efforts to improve cold tolerance in crops, especially in rice. The increasing global food demand, together with rapid population growth and frequent occurrence of chilling forces scientists to speed up and push forward the improvement of rice cold tolerance. ALA synthesis is the rate-limiting step in the whole tetrapyrrole metabolic net, and it is obviously inhibited in chill-treated rice seedlings (Figure 3A). Based on this finding, we propose that application of exogenous ALA or overexpression of *HEMA* or *GSA* gene in plants may overcome inadequate chlorophyll biosynthesis and maintain the structural and functional integrity of chloroplasts, and thus improve cold tolerance. Because tetrapyrrole intermediates are easily activated by light, leading to photooxidative damage, the concentration of exogenous ALA should be applied accurately. Several investigations have recognized that ALA pretreatment enhances plants' tolerance to chilling by increasing the activities of antioxidant enzymes to eliminate excessive ROS and improving chlorophyll fluorescence and photosynthesis [58–60]. Thus, our results suggested the protective role of exogenous ALA and contribute to further illustrate its mechanism. Nevertheless, exogenous ALA increased chlorophyll accumulation in etiolated oilseed rape, but failed to enhance its cold tolerance [61]. Application of exogenous ALA is an exciting field to explore, and might be beneficial for increasing chlorophyll content and improving cold resistance of rice seedlings. Thus, more refined investigations of the effect of exogenous ALA in etiolated seedlings under chilling stress are still required.

Given that transgenic technologies have been developed intensively and almost all chlorophyll biosynthetic genes have been identified in many crops, genetically engineered crops provide a new opportunity to solve the threats from environmental stresses. Thus, genetic modification of chlorophyll biosynthetic genes might be a promising approach for improving plant cold tolerance. A previous study suggested that *CHLG*-over-expressing plants have increased ALA synthetic capacity and increased chelatase activity, indicating that overexpression of *CHLG* could stimulate chlorophyll biosynthesis [62]. Overexpression of *HEMA*, *DVR*, and *CHLG* in plants may be helpful to improve the cold tolerance of rice seedlings during greening by increasing chlorophyll biosynthesis.

### **4. Materials and Methods**

#### *4.1. Plant Material and Growth Conditions*

Rice (*Oryza sativa* L.) cultivar DM You 6188 was used as experimental material, and was purchased from Ya'an seed store. Seeds were sterilized with 3% (*m*/*v*) sodium hypochlorite for 10 min, washed five times with distilled water and soaked in distilled water for 36 h, then placed on moist filter papers with 1/4 strength Hoagland nutrient solution and grown in the dark at 28 ◦C before chilling treatment. After 6 days, the etiolated seedlings were transferred to vermiculite with 1/4 strength Hoagland nutrient solution in the dark. The seedlings were then exposed to light (120 μmol m−<sup>2</sup> s<sup>−</sup>1) and transferred to 28 ◦C, 18 ◦C, and 12 ◦C, respectively. The first leaf was used to measure physiological

and biochemical parameters at 0 h, 0.5 h, 12 h, and 48 h after exposure to light, and all experiments were repeated at least three times.
