Zebra leaf 15, a receptor-like protein kinase involved in moderate low temperature signaling pathway in rice

Background Zebra leaf mutants are an important resource for studying leaf colour in rice. In most such mutants, the zebra leaf phenotype results from defective chloroplast biogenesis. The molecular mechanism by which zebra leaves develop remains unclear, so additional zebra-leaf mutants need to be identified. Results We isolated a novel rice zebra-leaf mutant, named zebra leaf 15 (z15), which showed transversely striped leaves with yellow-green or white-green sectors, in which chloroplast structure was disturbed. Transmission electron microscopy revealed that the structure of various organelles was impaired in yellow/white sectors. Z15, a single-copy gene in the rice genome, encodes a receptor-like protein kinase. Subcellular localization analysis indicates that Z15 and z15 are localized on the plasma membrane. The expression of Z15 is induced by moderate low temperature (18 °C). The mutation of Z15 influenced the expression of two downstream genes, OsWRKY71 and OsMYB4, that were responsive to moderate low temperature. The results show that Z15 plays a crucial role in the early stages of the response to moderate low temperature in rice. Conclusions We identified a novel zebra-leaf mutant (z15) that impaired chloroplast structure in rice, LOC_Os05g12680, encoding a receptor-like protein kinase. Further study indiceted that Z15 plays a crucial role in the early stages of the response to moderate low temperature in rice.


Background
The "zebra leaf" trait of rice manifests as regular intermittent chlorosis of the leaves and sheaths. The entire leaf shows transverse green/yellow-striped segments, which are similar to zebra striping patterns. Under certain conditions, the zebra pattern may fade gradually and the leaves assume a normal appearance. Temperature is one important environmental factor that affects the zebra leaf trait in rice. The temperaturesensitive mutant tsc1 shows a white, light-yellow, and normal green leaf when grown at 23.1°C, 26.1°C, and 30.1°C, respectively (Dong et al. 2001). The temperature-conditioned mutants v3 and st1 produce bleached leaves at constant 20°C or 30°C, but leaves that are almost green develop under diurnal 30°C/20°C conditions (Yoo et al. 2009). These studies found variations in leaf colour at different temperatures. However, no-one has yet conducted a systematic investigation of the temperature-sensing mechanism of zebra leaf mutants.
Low temperature is a major environmental factor that limits plant growth, which is classifiable into chilling (0-15°C) and freezing (< 0°C) temperatures based on plant tolerance (Zhou et al. 2011). The most frequently applied low-temperature treatment is 4°C, to which many genes are reportedly responsive; examples are OPEN STOMATA 1 (OST1), BRASSINAZOLE RESISTANT 1 (BZR1), and COLD RESPONSIVE PROTEIN KINASE 1 (CRPK1) (Ding et al. 2015;Liu et al. 2017). Responses to moderate temperature (15-28°C) are rarely studied. Many plant species have evolved a series of mechanisms that minimize the negative effects of cold stress. These mechanisms are divided mainly into CBF-dependent and CBF-independent pathways. The CBF transcriptional regulatory cascade plays important roles in plant response to cold. CRPK1, a cold-activated plasma membrane protein, is involved in the CBFdependent pathway. CRPK1 phosphorylates 14-3-3 proteins, then the phosphorylated 14-3-3 proteins are transported from the cytosol to the nucleus, and ultimately through the CBF pathway through which the coldresponse signal is transmitted (Liu et al. 2017). CHILL-ING TOLERANCE DIVERGENCE 1 (COLD1) is an additional CBF-dependent pathway regulator that encodes a regulator of G-protein signalling. Interaction of COLD1 and its protein partner RGA1(rice G-protein α subnit 1) triggers a calcium influx that leads to the activation of crucial CCAAT motif-binding factor (CBF) transcription factors for chilling tolerance in rice (Ma et al. 2015). In addition to the CBF-dependent pathway, many genes involved in low-temperature regulation belong to CBFindependent pathways. The transcription factors MYB4 and MYB3R2 in rice can regulate the response to cold stress positively, but the mechanism is different from that of the CBF pathway (Dai et al. 2007;Vannini et al. 2004). BZR1 may act as a cross-regulator to regulate the cold response and other signalling pathways. In the non-CBF pathways, BZR1 binds directly to the promoter of the CBF-independent cold-regulated genes, such as WRKY6, PYL6, SOC1, SAG21, and JMT, which regulate the cold response . Information on CBFindependent pathways remains incomplete. Therefore, identification of additional genes involved in CBFindependent pathways is needed to provide a deeper understanding of the mechanism of the response to low temperature.
The receptor-like protein kinases (RLKs) gene family is one of the largest gene families in plants. It plays an important role in the regulation of plant growth and development, signalling networks, and disease resistance . Plant RLKs are involved in the response to cold stress and mediate signal transduction. A calcium-regulated RLK, CRLK1, modulates acclimation to cold by means of a MAP kinase cascade in plants and promotes the expression of cold-stress response genes, and ultimately plant adaptation to low temperature is regulated (Yang et al. 2010a;Yang et al. 2010b). Proline extension receptors (PERKs) are a class of plant RLKs with an extracellular domain rich in proline. Limited research on the PERK family has been conducted. PERK1 encodes a putative receptor protein kinase in Brassica napus L. and shows a similar structure to wall-associated kinases (WAKs). PERK1 plays a role in mediation of early events in the plant defence response to mechanical injury, perhaps by sensing changes in the cell wall via its extension-like extracellular domain, and a wound-response signalling cascade is triggered through its catalytic domain (Silva and Goring 2002). AtPERK4, which encodes a receptor-like kinase, plays a role in the early stages of abscisic acid (ABA) signalling to inhibit primary root cell elongation by perturbation of Ca 2+ homeostasis (Bai et al. 2009). This suggests that the PERK-based cascade may act by sensing changes in the signal and transduction of the signal to the cell wall during plant growth and development (Nakhamchik et al. 2004). Although previous research has shown that PERK family proteins perform many functions, no information is available on PERK functions in association with low temperature stress.
In the study reported herein, we isolated the zebra leaf 15 (z15) mutant. The Z15 gene encodes a receptor-like protein kinase of unknown function and belongs to the PERK family. Subcellular localization analysis indicates that Z15 and z15 are localized on the plasma membrane. The gene responds to a moderate low temperature of 18°C and involves in the moderate low temperature signaling pathway. Mutation of Z15 disrupts the response to cold stress and cell development in rice. The present results indicated that the PERK performs a novel function in leaf cell development, which provides insights into the underlying mechanism of the response in rice to stress induced by moderate low temperature.

Results
The z15 mutant showed the zebra-leaf phenotype and cell development was impaired Under field conditions, the wild type 'Jinhui 10' rice plant developed green leaves throughout the growth period except in the later mature stage ( Fig. 1a-1, Additional file 1: Figure S1A). The z15 mutant plants produced leaves striped transversely with yellow-green or white-green sectors from the seedling stage to the tiller stage, but the leaf sheath showed normal pigmentation ( Fig. 1b-1, Additional file 1: Figure S1B). As the leaf developed, the defective phenotype was gradually restored and disappeared after the tiller stage. To determine the timing of appearance of the zebra leaf phenotype in the z15 mutant, young leaves at seedling stage that had not emerged from the sheath were observed. The zebra leaf phenotype developed before the leaf emerged from the sheath compared with the green leaves in the wild type under the temperature changes from 15°C to 28°C (Fig. 1, a-2 and b-2). To investigate whether Z15 affects leaf chloroplast development, white areas (M1), yellow-green areas (M2), and white areas (M3) of the leaves in the z15 mutant were observed by transmission electron microscopy (TEM). In the corresponding portions of the wild-type leaf (W1, W2, and W3), all developed complete cells, and the chloroplasts were well developed and well organized in the mesophyll cells. In contrast, in the z15 mutant the white areas (M1 and M3) contained mostly empty cells and the chloroplasts were impaired, whereas the yellow-green (M2) cells were normal. At the tiller stage, in the wild type, the cells developed normally and contained a complete suite of well-developed organelles (Additional file 1: Figure S2A). In the z15 mutant, most cells contained a small number of chloroplasts and the structure of some chloroplasts was impaired; however, some cells were empty and the organelles suffered severe damage (Additional file 1: Figure S2B). In addition, we observed that the timing of emergence of the yellow or white portions of the leaves in the z15 mutant was random by drawing line in the morning and night (Additional file 1: Figure S1B). At the mature stage, the phenotype of the z15 mutant was identical to that of the wild type with normal green-pigmented leaves (Additional file 1: Figure  S1A).
Agronomic traits at maturity were investigated. Compared with those of the wild type, plant height, grain number per panicle, percentage seed set, and 1000-grain weight were significantly reduced in the z15 mutant, in average by 5.26%, 9.99%, 16.96% and 2.76%, respectively (Additional file 1: Figure S3, A, D, E and F), whereas the number of productive panicles per plant and panicle length showed no significant difference from the wild type (Additional file 1: Figure S3, B and C). These results indicated that the mutation of Z15 disrupts cell development, thereby impairing chloroplast structure and causing a zebra leaf phenotype under field conditions.

Molecular cloning and identification of Z15 and subcellular localization of Z15 protein
The Z15 gene has been mapped to a region of about 268 kb on chromosome 5 ). In our study, the location of Z15 was narrowed to a physical distance of 134 kb between the insertion/deletion markers nSSR516 and Z15-13. The interval included 23 annotated genes (http://www.gramene.org/). Sequencing analysis identified a signal-nucleotide substitution from G to A within the LOC_Os05g12680 gene, which caused an amino acid mutation of Gly-409 to Asp-409 ( Fig. 2a). To confirm whether the mutation of LOC_Os05g12680 resulted in the mutant phenotype, we performed a complementation experiment by transforming into the z15 mutant a 9085-bp wild-type DNA fragment that contained LOC_Os05g12680. In the transgenic plants, the phenotype was recovered completely (Fig. 2b). Observation by TEM showed that the mesophyll cells and organelles were similar to those of the wild type (Fig. 2b). In addition, the contents of chlorophylls and carotenoids in the transgenic plants were almost identical to those of the wild type (Fig. 2c), and the contents of chlorophylls and carotenoids in the z15 were significantly lower than that of wild type and the transgenic plants (t-test, p < 0.01). Taken together, these results confirmed that LOC_Os05g12680 corresponded to the Z15 gene.
To examine the subcellular localization of the Z15 and z15 proteins, vectors were constructed to produce fusion proteins with the green fluorescent protein (GFP), Z15-GFP and z15-GFP. In addition, we used ESL4-GFP (Xing et al. 2018) which is located on the plasma membrane, as a positive control. The vectors were introduced into rice protoplasts and the proteins transiently expressed. In cells that expressed GFP alone, green fluorescence was detected consistently throughout the cytosol, while the Z15-GFP and z15-GFP fusion proteins were localized in the cell periphery, similarly to the positive control, consistently with expected plasma membrane localization (Fig. 3a). These results indicate that Z15 is a plasma membrane protein and that the substitution Gly-409 to Asp-409 does not change the protein localization.

Z15 encodes a receptor-like kinase
Protein sequences for Z15 were downloaded from the National Centre of Biotechnology Information databases (https://www.ncbi.nlm.nih.gov/) and aligned for phylogenetic analysis. Z15 encodes a receptor-like protein kinase that belongs to the RLK family. The RLK family is a class of single transmembrane proteins located on the plasma membrane that sense and transmit a variety of signals to regulate plant growth and development (Shiu and Bleecker 2001). The phylogenetic tree indicated that Z15 is highly conserved in many organisms, ranging from unicellular organisms, to mosses and lichens, to higher plants (Fig. 4). In addition, the Z15 protein cluster showed the highest similarity to the PERKlike receptor protein kinase family. The conserved domain of the protein sequence contains a STKc_IRAK domain, which is a catalytic domain of Ser/Thr kinases (Additional file 1: Figure S4). We performed a predictive analysis of the kinase domain and observed that the proton acceptor is close to the mutation site (Additional file 1: Figure S5, A, B and C). The mutation of Z15 caused an amino acid mutation of Gly to Asp (Fig. 2a), i.e., substitution of a neutral amino acid with an acidic amino acid, which may be due to the changes in kinase activity. So we performed an in vitro enzyme activity assay to determine whether Z15 and z15 have kinase activities. We constructed pet32a vectors to express proteins containing the Z15 and z15 intracellular domains. The kinase activity of the recombinant proteins were determined by measuring ATP consumption in presence of a classical kinase substrate, the myelin basic protein (MBP) and using a luminescent assay kit. In presence of Z15, the relative light unit measured decreased over time (Fig.  3b), which indicates that ATP was continuously consumed during the enzymatic reaction and that the Z15 intracellular domain is catalytically active. In contrast, in presence of the z15 intracellular domain, the relative light unit was almost unchanged (Fig. 3c), which indicates that mutation of Z15 results in a loss of kinase activity.

Expression pattern of Z15
Quantitative real-time PCR (qPCR) analysis revealed that Z15 was expressed in a variety of tissues of the wild type, including the root, stem, young leaf, mature leaf, sheath, and young panicle (Fig. 5a).
To gather further information on Z15 expression, a binary vector containing the Z15 and β-glucuronidase (GUS) genes driven by the Z15 promoter (Z15::GUS) was constructed. GUS expression was detected in the root, stem, leaf, and panicle (Fig. 5b), whereas Z15 was expressed predominantly in vascular bundles in the root, panicle, and stamen (Fig. 5, b-1 , b-4, and b-5). Histochemical localization revealed that GUS expression was restricted exclusively to vascular bundles and epidermal cells in the stem and mesophyll cells in leaf (Fig. 5, b-6 and b-7).
For a more detailed analysis of the expression pattern of Z15, transverse sections of various tissues were examined for Z15 signals by in situ hybridization (Fig.  5c). Consistent with the GUS expression results, the Z15 signal was observed in the vascular bundle of the root ( Fig. 5c-1). In the stem, sheath, and young panicle, signals were detected in the vascular bundles and epidermal cells (Fig. 5c-2, c-6, c-4, c-9, c-5, c-10, and c-11). Interestingly, Z15 signals were concentrated mainly in mesophyll cells in young leaves (Fig. 5c-7 and c-8).
Z15 is involved in moderate low temperature signaling pathway As the phynotype of z15 was expressed under field conditions, we examined the expression of Z15 under different temperatures by qPCR. The expression of Z15 was not affected significantly under high temperature (34°C) and low temperature (4°C) from 30 min to 24 h, compared with the control under 28°C (Fig. 6a). However, when treated at 18°C, expression of Z15 was induced distinctly at 1 h (Fig. 6b), and thereafter the level of expression of Z15 fell gradually and stabilized. For a more detailed examination of the response of Z15 to temperature, seedlings were treated at 15, 18, 21, and 24°C, respectively. The highest level of expression of Z15 was observed at 18°C (Fig. 6c). We have also analyzed the expression of Z15 in z15 mutant under 18°C. We found that the expression of Z15 was less induced in the z15 mutant compared to the wild type (Fig. 6b). Taken together, these results indicated that the To obtain insights into the function of Z15, the transcriptomic profiles of seedlings of the z15 mutant and wild type treated at 18°C for 30 min were characterized. Comparisons of the differentially expressed genes (DEGs) of the different treatment groups, namely z15 (30 min) vs z15 (CK) and the wild type (30 min) vs the wild type (CK), are presented (Additional file 1: Figure  S6, B and C). The two comparisons shared 2038 DEGs (Additional file 1: Figure S6A), which indicated that these common DEGs continued to function during 30 min of cold-stress treatment. Among the common 2038 DEGs, the expression of certain genes involved in the cold-stress response changed significantly, which indicates that cold tolerance was affected strongly by the mutation of Z15 (Additional file 2: Tables S2 and S3). Previous studies indicated that OsWRKY71 and OsMYB4 were induced by cold stress specifically (Kim et al. 2016;Vannini et al. 2004). In our study, the expression of these two genes increased significantly after 6 h of 4°C treatment (Additional file 1: Figure. S6, D and E), and the genes were expressed at an early stage of 18°C stress ( Fig. 6d and e). These results indicated that OsWRKY71 and OsMYB4 are responsive not only to 4°C, but also to 18°C. In addition, qPCR verified that the levels of expression of OsWRKY71 and OsMYB4 were lower in the z15 mutant than in the wild type ( Fig. 6d and e), which indicated that Z15 influenced the expression of OsWRKY71 and OsMYB4. Taken together, the results suggested that Z15 is involved in moderate low temperature signaling pathway in rice.

Discussion
We characterized a novel rice zebra leaf mutant, z15, which showed transversely striped leaves with yellowgreen and white-green sectors in which chloroplast structure is impaired. In addition, observation by TEM showed that the structure of a variety of organelles was damaged in the yellow or white areas of leaves (Fig. 1, Additional file 1: Figure S1). Z15 was determined to encode a RLK protein. The RLK family members play important roles in mediating the cell response to a variety of environmental cues. The expression of Z15 is induced by moderate low temperature (18°C). The mutation of Z15 influenced not only its own expression at moderate low temperature, but also the expression of downstream genes that respond to moderate low temperature. The present results showed that Z15 is involved in moderate low temperature signaling pathway, which plays a crucial role in the early response of rice to moderate low temperature.
In previous studies, a number of genes associated with the zebra leaf phenotype in rice were cloned. The zebra necrosis (zn) mutant produces transversely striped leaves with green and yellow sectors, and necrotic lesions often develop in the yellow sectors during leaf elongation. ZN encodes a thylakoid-bound protein of unknown function. The necrotic lesions in yellow sectors of zn mutant leaves develop as a result of excessive accumulation of reactive oxygen species (Li et al. 2010). The zebra leaf 2 (z2) mutant shows alternating transverse green and yellow sectors on the leaves, and the phenotype is affected by temperature and light. ZEBRA2 encodes a carotenoid isomerase, the key enzyme that catalyzes the conversion of cis-lycopene to all-trans-lycopene. A more severe phenotype of the zebra2 mutant manifests under high light intensity, which indicates that carotenoids biosynthesis is impaired because of the gene mutation, and thus causes photooxidation and ABA defects, and leads to the zebra-leaf phenotype (Chai et al. 2011). Additional studies suggest that leaf variegation in the zebra2 mutant is caused by photoperiodic accumulation of tetra-cislycopene and singlet oxygen (Han et al. 2012). The rice zebra 3 (z3) mutant shows transverse dark-green or green sectors in mature leaves and lacks the typical yellow or white sectors. Z3 encodes a putative citrate transporter that belongs to the citrate-metal hydrogen symport family. The z3 phenotype is caused by unbalanced accumulation of citrate in transverse sectors of the leaf (Kim et al. 2018). The rice zebra leaf 16 (z16) Fig. 4 Phylogenetic relationships of the rice Z15 protein. The tree was constructed using the maximum-likelihood method based on the Jones-Taylor-Thornton matrix-based model. Bootstrap support values calculated from 500 replicates are given at the branch nodes. The Z15 protein is highlighted in red phenotype is light-dependent and is expressed as chlorotic abnormalities in transverse sectors of the young leaves of seedlings. Z16 encodes a β-hydroxyacyl-ACP dehydratase involved in de novo fatty acid synthesis. The z16 mutation affects early stages of chloroplast development in leaves . In our study, plants of the z15 mutant produced transversely striped leaves with yellow-green or white-green sectors from the seedling stage to the tiller stage ( Fig. 1b-1, Additional file 1: Figure S2 B-2), but the leaf sheath showed normal pigment development. Under field conditions, the phenotype of z15 is sensitive to cold stress. In contrast to the zn mutant, no necrotic lesions develop on the leaves of the z15 mutant, and unlike the z2 mutant, the phenotype of z15 is unaffected by light intensity. With regard to subcellular structure, unlike the zn and z2 mutants, a variety of organelles were damaged in addition to plastids and chloroplasts (Fig. 1, Additional file 1: Figure S1). Therefore, z15 is a novel zebra leaf mutant. Z15 encodes a receptor-like protein kinase, which is distinct from ZN and Z2, thus Z15 is a newly identified gene. We suggest that Z15 is a novel regulatory factor that affects leaf development indirectly by regulating signalling in response to moderate low temperature.
It has been reported that RLKs are involved in plant response to low-temperature stress and transmit lowtemperature signals (Liu et al. 2017). The PERK family (proline-rich, extension-like receptor kinases) belongs to the RLKs family. Z15 encodes a receptor-like protein kinase of unknown function, which shows homology with the PERK family (Fig. 4). Information on PERK receptor protein kinases is limited, but previous studies have shown that PERK-based signal cascades may act by sensing signal changes and transduction of signals to the cell wall during plant growth and development (Nakhamchik et al. 2004). In Brassica napus, BnPERK1 , c Expression patterns of Z15 indicated by in situ hybridization (C-1: root, C-2: stem, C-3: leaf, C-4: sheath, C-5: panicle, C-6 is an enlarged view of C-2, C-7 and C-8 are enlarged view of C-3, C-9 is an enlarged view of C-4, C-10 and C-11 are enlarged view of C-5, C-6 to C-11: bar = 100 μm) may be involved in the plant's early defence responses to mechanical stress and fungal attack (Silva and Goring 2002). AtPERK4, which encodes a receptor-like kinase, plays a role in the early stages of ABA signalling and inhibits root cell elongation by disrupting Ca 2+ homeostasis (Bai et al. 2009). In the study reported herein, we provide evidence that Z15 is involved in moderate low temperature signaling pathway. First, we showed that the zebra leaf phenotype of the z15 mutant manifested only under field conditions, and the time of emergence of the yellow or white sectors of the leaf was random (Additional file 1: Figure S1). Second, we observed that Z15 was less induced in z15 mutant than wild-type by moderate low temperature (18°C) (Fig. 6b). Third, Z15 encodes a receptor-like kinase, which localizes at the plasma membrane. The mutation in z15 protein doesn't affect the plasma membrane localization but abolishes the in vitro kinase activity (Fig. 3aand c). In addition, Z15 was expressed predominantly in vascular bundles in the root, stem, sheath, and panicle, and in mesophyll cells in the leaf (Fig. 5). Thus, a novel function of PERK family proteins was revealed that Z15 is involved in moderate low temperature signaling pathway. Finally, we identified two potential downstream cold-defence genes, namely OsWRKY71 and OsMYB4. In rice, overexpression of OsWRKY71 in rice results in increased resistance to disease (Liu et al. 2007). OsWRKY71 is also induced by cold stress (Kim et al. 2016). Similarly, our study showed that the OsWRKY71 expression level was decreased in the z15 mutant after treatment with moderate low temperature, compared with the wild type (Fig. 6d). Previous research has demonstrated the ability of OsMYB4 to positively regulate the transcription of genes involved in the cold response (Vannini et al. 2004). In Arabidopsis, AtMYB14 is down regulated under cold treatment and its involvement in freezing tolerance is Fig. 6 Influence of temperature on expression of Z15. a Expression of Z15 at 4°C and 34°C, respectively. b Expression of Z15 in wild type and z15 mutant at 18°C. c Expression of Z15 at 15, 18, 21, and 24°C, respectively. d The expression of OsWRKY71 at 18°C. e The expression of OsMYB4 at 18°C. Error bars represent the standard deviation (SD) of three biological repeats affected by the expression of CBF genes (Chen et al. 2013). Interestingly, the expression of OsMYB4 is consistent with OsWRKY71 in response to cold stress (Fig.  6e). Thus, the above-mentioned results indicated that Z15 may regulate the expression of the potential downstream cold-response genes OsWRKY71 and OsMYB4. Therefore, expression of the Z15 receptor kinase is indicated to be an early response to exposure to moderate low temperature.
The mechanism by which Z15 is involved in moderate low temperature signaling pathway was investigated. The present observations suggest that Z15 might play an important role in early stages of the response to exposure to moderate low temperature.

Conclusion
A novel zebra-leaf mutant was identified in rice, LOC_ Os05g12680, encoding a receptor-like protein kinase both of Z15 and z15 were located on plasma membrane, but the mutation of Z15 resulted in loss of kinase activity. The expression of Z15 was optimally induced by moderate low temperature, under which condition the transcriptional regulon of Z15 are mainly involved in response to low temperature (4°C). Among of them, the expression of two selected Z15-affecting genes was specifically induced only at the temperatures ranging from low to moderate low, confirming our results that Z15 is involved in moderate low temperature signaling pathway, which plays a crucial role in the early stage of freezing tolerance in rice growth. Results from this work suggest that Z15 is involved in moderate low temperature pathway, which plays a crucial role in the early stages of the response to moderate low temperature in rice.

Plant materials
The z15 mutant was derived from rice (Oryza sativa L. subsp. indica) 'Jinhui 10' treated with ethyl methanesulfonate. 'Jinhui 10' was used as the wild type in all experiments. All plants were grown in an experimental field at Chongqing, China, under natural conditions. We cultivated seeds with nutrient solution. 'Jinhui 10' seeds were cultured in a light incubator (RXZ-500) under 12:12 h light (400 μm/m 2 /s)/darkness cycle at temperature of 28°C (day and night). For temperature treatment, we set the temperature in advance, and put the same nutrient solution into the incubator so that the water temperature is the same as the incubator temperature.

Photosynthetic pigments analysis
Photosynthetic pigments were extracted from equal fresh weight (~0.1 g) of mature leaves from the wildtype, z15, and transgenic plants in 25 ml extraction buffer (ethanol:acetone, 1:1; v/v) for 24 h at room temperature in the dark. The concentrations of chlorophylls and carotenoids were determined with a UV-1800PC spectrophotometer (Mapada Co. Ltd., Shanghai, China) in accordance with the method of (Lichtenthaler 1987).

Agronomic trait analysis
Agronomic traits (comprising plant height, number of productive panicles per plant, panicle length, grain number per panicle, percentage seed set, and 1000-grain weight) for each of the z15 mutant and the wild type were analyzed at the mature stage with 15-20 replicates.

Transmission electron microscopy
Leaves of wild-type, z15, and transgenic plants were investigated by TEM using the method described by (Zhu et al. 2016), The middle portion of the leaves of the plants grown in the experimental field under natural conditions were collected and fixed in a primary fixative solution (3.5% glutaraldehyde) for 2 d at room temperature. Each sample was washed with 0.1 mol l − 1 phosphate buffer solution, then post fixed for 2 h with 1% osmium tetroxide. The tissues were then stained with uranyl acetate, dehydrated in ethanol, and embedded in Spurr's resin before ultrathin sectioning. Each sample was then restained and examined using a H-7500 transmission electron microscope (Hitachi High-Technologies).

Fine-mapping of Z15
The z15 mutant was crossed with 'Xinong1A' (bred by the Southwest University Rice Research Institute, Chongqing, China) to obtain the F 1 population, and the F 2 population was generated by self-fertilization of F 1 plants. The F 2 plants that manifested the mutant phenotype were selected and used to map the genomic location of Z15. Fine-mapping was performed using insertion/deletion markers that were developed from comparison of genomic sequences for 'Xinong1A' and 'Jinhui 10' in our laboratory. The sequences of primers used in fine-mapping are listed in Additional file 2: Table S1.

Method for verifying the complementation of the gene
Within the physical distance of 134 kb in our mapped location of Z15, we have sequenced the exon region of all the genes (23 in total), identified a single nucleotide substitution on the third exon of LOC_Os05g12680 in the z15 mutant compared to the wild-type (Fig. 2a). Therefore, we regarded LOC_Os05g12680 as a candidate gene. We then performed a complementation experiment by transforming the z15 mutant with a T-DNA construct containing a 9085-bp Z15 genomic DNA fragment (LOC_Os05g12680, including 2978-bp of the promoter), neighbored by an hygromycin resistance cassette and the GUS reporter gene under the control of the 35S promoter (Additional file 1: Figure S7). We obtained six transgenic plants. The method for determining full insertion of the T-DNA was firstly sequencing plant growth on hygromycin and PCR amplification of the hygromycin resistance gene. And secondly, GUS staining. We then observed the phenotype of T 0 and T 1 transgenic plants, and their phenotype was stable. Finally, we have used the primers F1 and R1 (Additional file 1: Figure S7) amplify and sequence the endogenous Z15 allele and verify that the transgenic plants were on the homozygous mutant background. We also used the primers F2 and R1 to verify that the WT copy was also present in the transgenic plants. By transmission electron microscopy and chlorophylls and carotenoids content determination, we determined that the phenotypes of these six positive transgenic plants were similar to those of the wild type (Fig. 2b). Taken together, this shows that the mutation in LOC_Os05g12680 is responsible for the observed phenotype in the z15 mutant.
Vector construction for genetic complementation and ProZ15::GUS To construct the Z15 complementation plasmid, a 9085bp genomic fragment containing the Z15 coding sequence, coupled with 3022-bp upstream and 1224-bp downstream sequences, was cloned into the binary vector pCAMBIA1301. To analyze the expression pattern of Z15, the ProZ15::GUS vector was constructed by amplifying the upstream 3026-bp promoter sequence of Z15 from 'Jinhui 10' genomic DNA, then cloning the sequence into the pCAMBIA1301 vector. The recombinant complementation plasmids were introduced into z15 using the Agrobacterium tumefaciens-mediated transformation method as described previously ). The primer pairs used for vector construction are listed in Additional file 2: Table S1.

RNA extraction and quantitative real-time PCR analysis
Total rice RNA was extracted from the roots, stems, leaves, leaf sheaths, and young panicles using the RNAprep Pure Plant Kit (Promega Co., Ltd., Beijing, China). All reverse transcriptions were performed using the GoScript™ Reverse Transcription System (Promega). After cDNA synthesis, all samples were diluted 10 times and stored at − 20°C. Quantitative RT-PCR was performed using TB Green™ Premix Ex Taq™ II (Tli RNa-seH Plus) (TaKaRa, China) in a Bio-Rad Real-Time System (1000 Alfred Nobel Drive, Hercules, California 94,547, U.S.A). Three replicates were performed and the relative level of expression of each gene was calculated using the 2 −△△Ct method and expressed relative to OsActin expression. The primers used are listed in Additional file 2: Table S1.

Subcellular localization
The full-length coding sequence of Z15 and z15 without the stop codon were fused to the N-terminus of the GFP gene under the control of the enhanced Cauliflower mosaic virus 35S promoter in the SpeI and BamHI sites of the vector pAN580 to generate the pAN580-Z15-GFP and pAN580-z15-GFP construct. The vectors were introduced into rice protoplasts using a method described previously ). The GFP fluorescence was determined using an LSM710 confocal laser scanning microscope (Zeiss, Jena, Germany). The primer sequences are listed in Additional file 2: Table S1.
Determining the activities of Z15 and z15 kinase Pet32a vectors allowing expression of Z15 and z15 intracellular domains (from 181th amino acid to 675th amino acid) tagged with 6X His in Escherichia coli were built. The primer sequences used for cloning are listed in Additional file 2: Table S1.
The recombinant proteins expressed in the E.coli (DE3) were purified using Beaver Beads™ IDA-Nickel kit (cat70501-klo; Haitian Nano Technology Co, Suzhou, China).
Recombinant protein enzymatic activities were performed using a Kinase-Glo R Luminescent Kinase Assay kit (cat no.#M2295; Promega Biotechnology Co., Beijing, China). Enzymatic buffer and the substrate myelin basic protein (MBP) were added to start the reaction. Luminescence was measured as described in (Xing et al. 2018). The blank corresponds to the same reaction mixture (Enzymatic buffer and MBP), in absence of recombinant protein.

In situ hybridization
The 429-bp gene-specific Z15 probe was amplified and labelled using a DIG RNA Labeling Kit (SP6/SP7; Roche). Section pretreatment, hybridization, and immunological detection methods were performed using previously described methods . The sequences of the primers used are listed in Additional file 2: Table S1.

Transcriptome analysis
Leaves from 10-day-old plants of z15 and wild-type grown in an incubator at 28°C and treated at 18°C for 30 min were collected for RNA extraction and transcriptome sequencing with three biological replicates. The purity and concentration of RNA were measured using a NanoPhotometer spectrophotometer (Implen, Westlake Village, CA, USA) and the Qubit® RNA Assay Kit in a Qubit® 2.0 Fluorometer (Life Technologies, Carlsbad,