Genome-wide identification of the Tubby-Like Protein (TLPs) family in medicinal model plant Salvia miltiorrhiza

Data collection

Initially, based on the keyword Tubby like protein, already-known TLP protein sequences from A. thaliana, S. lycopersicum, and O. sativa were retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/) (Table S1). The S. miltiorrhiza genome (published) was downloaded from the China National Traditional Chinese Medicine Data Center (ftp://danshen.ndctcm.org:10402/) (Xu et al., 2016a).

Identification of TLPs in S. miltiorrhiza data collection

We used two methods to accurately identify TLPs in S. miltiorrhiza. First, we used the online software iTAK (http://itak.feilab.net/cgi-bin/itak/index.cgi) to directly identify 12 candidate SmTLPs (Zheng et al., 2016). Then, based on the hidden Markov model (HMM) of the Tubby domain (PF01167) obtained from the Pfam database (http://pfam.xfam.org/), we also identified 12 candidate SmTLPs using Hmmsearch (3.2.1) (Potter et al., 2018). Next, we used the Web CD-Search Tool (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) to identify and confirm the domains of candidate SmTLPs identified by the two methods. Finally, we identified 12 SmTLPs and named them SmTLP1–SmTLP12. The SmTLPs’ physicochemical characteristics were analyzed using the ProtParam tool (http://web.expasy.org/protparam/). Characteristics included the grand average of hydropathicity (GRAVY), instability index, theoretical isoelectric point (p I), and molecular weight (Mw). The WoLF PSORT (https://wolfpsort.hgc.jp/), Busca (http://busca.biocomp.unibo.it/) and Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/) were used to predict the subcellular localization of the SmTLPs in S. miltiorrhiza.

Analysis of gene structure and protein structure

TBtools was used to analyze the gene structure of SmTLPs (Chen et al., 2020). Conserved motifs in SmTLPs were identified using the MEME tool (http://meme-suite.org/tools/meme) and the following criteria: maximum number of motifs were set to 10; the minimum motif width was 6; and the maximum motif width was 200. Other parameters were set to default. Tubby domain protein alignment was performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) and verified by MEGA X (Kumar et al., 2018; Sievers & Higgins, 2018; Sievers et al., 2011). The SWISS-MODEL (https://www.swissmodel.expasy.org/) was used to construct the SmTLPs homologous protein model using the default parameters (Bienert et al., 2017; Guex, Peitsch & Schwede, 2009; Waterhouse et al., 2018).

Multiple sequence alignment and phylogenetic tree construction

All S. miltiorrhiza, A. thaliana, S. lycopersicum, and O. sativa TLPs were pooled into the Clustal Omega tool (https://www.ebi.ac.uk/Tools/msa/clustalo/) to perform multiple sequence alignments (Sievers & Higgins, 2018; Sievers et al., 2011). MEGA X (version 10.0.5) was used to perform a phylogenetic analysis of the aligned protein sequences using the Neighbor-Joining and the Maximum likelihood methods (Bootstrap = 1,000 replicates; other parameters: default) (Kumar et al., 2018). The phylogenetic tree we obtained was adjusted using Evolview (https://www.evolgenius.info/evolview).

Analysis of the promoter cis-regulating elements

A DNA sequence 2,000 bp upstream from 12 SmTLP family members was extracted from the genome database and defined as the promoter sequence. This submitted to PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) to identify the possible cis-regulating elements in the promoter sequence (Lescot, 2002). The visualization tool TBtools was used to map the distribution of cis-regulating elements in the promoters (Chen et al., 2020).

Analysis of gene expression

Two-year full-flowering S. miltiorrhiza plants with good and consistent growth condition were selected and divided into three groups. The control group was not given any treatment. The treatment group 1 was sprayed with 100 μmol L-1 ABA and the treatment group 2 was cultured at 4 °C. Each treatment consisted of three separate biological replicates. After treatment for 6 h, samples were taken, frozen in liquid nitrogen and stored in a refrigerator at −80 °C. Total RNA was extracted from roots, stems, leaves, and flowers using RNA extraction kit (Version 1.5) (DNase I) (Biofit Biotechnologies Co., Ltd) and three biological replicates were performed. cDNA was obtained by reverse transcription using the PrimeScript™ RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara, Beijing, China). qRT-PCR analysis was performed using SYBR Premix Ex Taq™II (TaKaRa, Japan) and the following cycling parameters: 95 °C for 30 s, and 40 cycles of 95 °C for 5 s, and 60 °C for 34 s. qRT-PCR data was calculated and analyzed using the 2 − ΔΔCt method. DNA and CDS sequences were used to design gene-specific primers for amplification (Supplementary Materials Data 2) using Primer 5 software. The error bars represent qRT-PCR result variability from the three replicates. Using the Student’s t test, P ≤ 0.01 (*) and P ≤ 0.05 (**) were compared with WT.

Identification of SmTLPs

Based on TLP sequences known to be found in A. thaliana we used the online iTAK online tool to identify 12 SmTLPs in the S. miltiorrhiza genome. Like with previous studies on A. thaliana, S. lycopersicum, and O. sativa, we named the 12 S. miltiorrhiza TLPs (Table 1).

The number of amino acids in the 12 SmTLPs ranged from 308 to 540. Physicochemical property analysis results revealed that the relative molecular weight of the 12 TLPs ranged from 34.08 to 60.64 kDa. The theoretical isoelectric points (p I) of all 12 TLPs were greater than 7, with TLP9 having the smallest p I (8.61) and TLP11 having the largest pI (9.76). The instability indice values for all SmTLPs were greater than 40, indicating that they are unstable proteins. In the grand average of hydropathicity, all SmTLPs’ values were negative. This indicates that SmTLPs are hydrophilic proteins. TLP4 had the best water solubility with a GRAVY value of -0.588. Subcellular localization prediction showed that all SmTLPs were located in the nucleus, suggesting that they play an important role there. This is consistent with previous studies on A. thaliana and S. lycopersicum (Lai et al., 2004; Zhang et al., 2020). The basic physicochemical properties of TLPs are shown in Table 1.

Phylogenetic analysis of SmTLPs

To investigate the phylogenetic relationship between S. miltiorrhiza TLPs, we constructed a phylogenetic tree including S. miltiorrhiza, A. thaliana, S. lycopersicum, and O. sativa TLPs using the Neighbor-Joining and Maximum likelihood methods (Fig. 1). This analysis clearly reveals that the 12 SmTLPs, 11 AtTLPs, 11 SlTLPs, and 14 OsTLPs are clustered and divided into three large subfamilies: A, B, and C. The A subfamily is further divided into AI and AII groups. This grouping is consistent with previous reports on S. lycopersicum (Zhang et al., 2020). The AI subfamily contains the most TLPs, including five SmTLPs (SmTLP1, SmTLP2, SmTLP3, SmTLP4, and SmTLP5) and 14 other TLPs. There are 15 TLPs in the A II subfamily, including three SmTLPs (SmTLP6, SmTLP7, and SmTLP8). The B subfamily has three SmTLPs (SmTLP9, SmTLP10, and SmTLP11). The C subfamily is the smallest subfamily, with only four TLPs (OsTLP13, AtTLP8, SlTLP11, and SmTLP12), and none of these TLPs contain F-box domains. The TLPs within each subfamily have high homology and close evolutionary relationships (Wang, Xu & Kong, 2018). However, the evolutionary concerns of S. miltiorrhiza, A. thaliana, and S. lycopersicum are closer.

Gene structure, conserved motifs, and conserved domains in SmTLPs

To improve our understanding of gene function and evolution, we analyzed the TLP gene structure (Fig. 2). Although some SmTLPs are very close phylogenetically, their exon number, distribution, and gene length greatly differ, especially between TLP1 and TLP2, and between TLP7 and TLP8. This is consistent with previous studies on S. lycopersicum, and may imply that the genetic structure of TLPs is easily altered (Zhang et al., 2020).

To further understand the process of TLP evolution, we analyzed the TLP motifs using MEME. A total of 10 motifs were identified (Fig. 3 and Fig. S1).

Conservative motif analysis results showed that all SmTLP members contained motif 1 and motif 4 (Fig. 3A). SmTLP members on the same branch of the phylogenetic tree had small differences in motif type, number, and position. SmTLP members on different branches of the phylogenetic tree had larger differences in motif type, number, and position. In particular, TLP12 had the fewest motifs, which may indicate that it has other unique functions. Closely related members on the phylogenetic tree often contain the same motif, indicating that TLPs in the same group have similar functions.

Proteins’ key conserved domains often confer function. To explore SmTLP function and structure we used the SMART database. The results show that, except for SmTLP9 and SmTLP12, SmTLPs have both a Tub domain (PF01167) and an F-box domain (PF00646) (Fig. 4 and Fig. S2). SmTLP structure was further explored by Tub domain 3D modeling. This analysis revealed that SmTLP11 Tub structure is incomplete, which may impact its function.

With additional 3D modeling and analysis of the TLPs’ Tub domain, we found that 11 SmTLPs have a complete Tub domain consisting of a closed β barrel with 12 anti-parallel strands and a central hydrophobic α helix (Fig. 5). We also found that TLP11 has an incomplete Tub domain, which may account for its special function, but this requires further experimental verification.

Analysis of SmTLP promoter sequences

The PlantCARE database was used to analyze 2,000 bp of DNA sequence upstream of the TLP genes’ transcription start site to determine the potential functions of TLPs. The main promoter was mapped and is displayed in Fig. 6. A variety of cis-elements that respond to environmental and hormonal signals have been identified in SmTLPs (Table 2). This implies that SmTLPs have a somewhat complicated mechanism of expression regulation. TLPs contain a variety of hormone response elements, including ABA, auxin, gibberellin, and MeJA response elements. Previous studies have shown that TLPs can also respond to light, temperature, hypoxia, and salt stress (Zhang et al., 2020). SmTLP10 had the most low-temperature response elements, which may indicate that it plays an important role in plant cold resistance (Fig. 7). TLP5 had the most meristem expression elements, indicating that it is important for plant meristem growth. Additional experiments are required to confirm these inferences.

The expression profiling of SmTLPs in different tissues of S. miltiorrhiza

Tissue-specific gene expression can be used to predict gene function. Therefore, we used qRT-PCR to detect SmTLP expression in various S. miltiorrhiza tissues (Fig. 8). The results showed that SmTLPs have distinct expression patterns. Four SmTLPs (SmTLP1, SmTLP2, SmTLP4, and SmTLP5) have unique expression patterns in the roots, which may imply that they function in stress resistance. This speculation requires further experimental verification. Four SmTLPs (SmTLP3, SmTLP9, SmTLP10, and SmTLP11) have unique expression patterns in the stems. The remaining four SmTLPs (SmTLP6, SmTLP7, SmTLP8, and SmTLP12) have unique expression patterns mainly in the flowers. Only SmTLP12 is expressed in leaves.

The expression of SmTLPs under ABA and cold treatment

We performed stress treatment on S. miltiorrhiza and used qRT-PCR to detect SmTLP expression to explore the response of SmTLPs to stress (Fig. 9). Under cold stress, the stem SmTLP3 expression decreased significantly, while SmTLP10 stem expression increased significantly. This suggests that these genes may play important roles in stem growth. After the flowers were subjected to cold stress, SmTLP6 expression decreased significantly and SmTLP8 expression increased significantly. This may indicate that they are involved in important physiological processes in flowers. Under cold stress, the SmTLP8 and SmTLP10 expression in roots significantly increased, indicating that they are involved in the cold-stress response in S. miltiorrhiza. Under ABA stress, SmTLP12 expression significantly increased, indicating that SmTLP12 may participate in the ABA pathway in S. miltiorrhiza.