Arabidopsis wild type (Columbia, Col-0) was used in this study. All the sterilized and cleaned seeds were put in 4°C for 2 days for vernalization. The 3–4 weeks old potato wild type (Desiree) aseptic seedlings were used in this study. For Arabidopsis and potato, all the materials were grown in growth chambers which maintained 22°C at a photoperiod of 16 h light (100-microm /m²/s fluorescence bulb light) followed by 8 h dark.

The Arabidopsis seeds were sown on half-strength Murashige and Skoog medium (0.5 × MS) (Murashige and Skoog, 1962) with different TOR inhibitors at varying concentrations for the time as described in the results section. The potato explants from 3 to 4 weeks old potato seedlings were subculture into 1 × MS with different TOR inhibitors at different concentrations for the time as indicated in the results section. Then all petri dishes were photographed next to a ruler. The root length and fresh weight were measured using ImageJ software and an electronic balance, respectively. The number of different root types was counted directly. Root hair was observed on the OLYMPUS MVX10 stereoscopic microscopes (Olympus, Japan).

GUS staining and GUS activity test were performed as described previously, respectively (Menand et al., 2002; Li et al., 2015). Western blotting method was following by Deng et al. (2016).

5 days old HS:: AXR3NT-GUS transgenic seedlings were incubated in 0.5 × MS liquid medium with TOR inhibitors for 48 h, DMSO was used as control. After that seedlings were heat shocked for 2 h at 37°C, and then supplied with NAA (10 μM) for 60 min at 22°C. Finally, the seedlings were stained 6 h for GUS staining.

Potato stem cuttings growing on the MS media for 48 h with different TOR inhibitors (rapamycin, KU, and their combination) and DMSO were collected and frozen in liquid nitrogen. Total RNA extraction used the RNAprep Pure Plant Kit (TianGen Biotech). RNA degradation and contamination were monitored on 1% agarose gels. RNA purity, concentration and integrity were measured by a NanoPhotometer spectrophotometer (IMPLEN, CA, USA), Qubit RNA Assay Kit in Qubit 2.0 Flurometer (Life Technologies, CA, USA) and the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA, USA), respectively. mRNA was purified from about 3 μg total RNA using poly-T oligo-attached magnetic beads. First strand cDNA and subsequent second strand cDNA synthesis were performed by using random hexamer primer. Sequencing adaptors were ligated to the fragments, which were enriched by PCR amplification. At last, PCR products were purified (AMPure XP system) and library quality was assessed on the Agilent Bioanalyzer 2100 system.

Construction of RNA-Seq libraries was performed on an Illumina Hiseq platform and 125/150 bp paired-end reads were generated. After removing the reads containing adapter, poly-N and low quality reads from raw data, the high quality clean reads were obtained. Paired-end clean reads were aligned to the reference genome using TopHat v2.0.12. The reads numbers mapped to each gene were quantified using HTSeq v0.6.1. FPKM (expected number of Fragments Per Kilobase of transcript sequence per Millions base pairs sequenced) of each gene was calculated based on the length of the gene and reads count mapped to this gene. Differential expression analysis between the treatment of TOR inhibitor and DMSO was performed using DESeq R package (1.18.0). Genes with an adjusted P < 0.05 found by DESeq were assigned as differentially expressed genes (DEGs). The DEGs were mapped to GO terms in the GO database (http://www.geneontology.org/) to calculate gene numbers in every term. KEGG (http://www.genome.jp/kegg/) is used to perform pathway analysis of DEGs. Statistical enrichment of DEGs in GO terms and KEGG pathway were implemented by the GOseq R package and KOBAS software, respectively. Significantly enriched GO terms and KEGG pathway (corrected p < 0.05) were identified based on a hypergeometric test.

Arabidopsis samples were prepared as describe elsewhere in the article. For potatoes, the different tissues (leaf, root, shoot, and stem) of the 4 weeks old potato were collected for measuring the expression level of StTOR, StFBKP12, StRAPTOR, and StLST8 by qRT-PCR. Some genes expression was tested by qRT-PCR as well in the potato samples, which were the same as that used for Illumina sequencing. Total RNA was isolated as described above. Approximately 1 μg of total RNA was used for reverse transcription with the PrimeScript RT Kit (Takara Biotech). The qRT-PCR was conducted using the TranStart Top Green qPCR SuperMix kit (Transgen) on a Bio-Rad CFX96 System. The amplification program consisted of the following cycles: 94°C pre-denaturation for 1 min, 40 cycles of 94°C 5 s, and 60°C 30 s. Plant actin was used as constitutive reference. The primers for qRT-PCR were designed using Primer premier 5 software and were listed in Table S1.

Total RNA was extracted from yeast cells and Arabidopsis using Trizol (Invitrogen), following the manufacture's protocol. cDNA was synthesized by using the PrimeScript RT Kit (Takara). The full-length coding sequence of ScFKBP12 and AtTIR1 was amplified by TransStar Taq DNA Polymerase (Transgen) using the designed primers (Table S1). A NotI site at the 5′ end of forward primer and a Sbf I site at the 3′ end of reverse primer were introduced. The remains steps of plasmid construction were performed as described elsewhere (Ren et al., 2012). For the method of P35S::TIR1-GUS-K303 vector generation was described elsewhere (Li et al., 2015). The potato transgenic method was according to the book “Agrobacterium protocols” (Springer press, Second Edition; Millam, 2007). The floral dipping method was employed for generating transgenic Arabidopsis (Zhang et al., 2006).

The entire genome of a homozygous doubled-monoploid derived from a Solanum tuberosum group Phureja is sequenced and annotated (http://plants.ensembl.org/Solanum_tuberosum/Search/New?db=core; The Potato Genome Sequencing Consortium, 2011). The availability of the whole potato genome sequence allowed us to identify the putative components of TOR complex in potato. The full length of Arabidopsis TOR (AT1G50030.1) amino acid sequence was used to BLAST against the potato genome database. The unique putative TOR (StTOR) locus, locating on the Chromosome 1, was found in potato (Figure 1A). To confirm the sequence of StTOR, we amplified and cloned the full length CDS (coding DNA sequence) of StTOR from potato cultivar Desiree. The sequencing results showed that the CDS of StTOR is identical to the homologs of TOR in potato genome database, and the corresponding genome sequence of StTOR showed that it spans about 36 kb genomic region containing 55 exons and 54 introns and encodes a predicted ~280 KD protein with 2470 aa (XM_006346213.1; Figure 1A). All the five key domains of TOR were found in the StTOR protein with higher identities on main domains like the FRB and kinase domain (Figure 1B). Based on phylogenetic analysis, StTOR showed a closer evolutionary relationship with AtTOR than any other TOR proteins (Figure 1C). The catalytic base, catalytic loop stability and activation loop sites of TOR kinase domain are also conserved in potato, suggesting that the catalytic function of TOR in potato may be conservative, and TOR signaling may exist in potato (Figure 1D). Additionally, the other core components of TOR complex 1 (TORC1) such as RAPTOR and LST8 were also found in potato genome (Table 1). Furthermore, qPCR results showed that the expression of TOR complex was constitutive in potato seedlings (Figure S2D). However, RICTOR, a vital component of TORC2, was not found (Table 1), suggesting that the existence of a conserved TORC1 but not TORC2 in potato genome.

Classic “FKBP12-rapamycin-TOR” negative regulation system, an easy manipulated and universal accepted approach, has been successfully used for interpreting the function of TOR in yeast and mammals (Choi et al., 1996; Virgilio and Loewith, 2006; Benjamin et al., 2011). Choi et al. (1996) reported that rapamycin can interact with the FRB domain of TOR through close contacting with aromatic residues, such as Tyr¹⁹³⁴, Phe¹⁹³⁵, Trp²⁰⁰¹, Tyr²⁰⁰⁵, Phe²⁰⁰⁸, Ser¹⁹³¹, Leu¹⁹²⁷, Thr¹⁹⁹⁸, and Asp²⁰⁰² (sequence numbers followed by human TOR protein; Choi et al., 1996). We found these key AA residues were highly conserved between potato, Arabidopsis and human through aligning the amino acid sequence of FRB domain of TOR from different species (Figure S1A), indicating that the function of TOR in potato could be dissected by using the rapamycin-FKBP12-TOR negative regulation system.

Many detected plants, such as Arabidopsis, Vicia faba, Lotus (Lotus japonicus), Tobacco (Nicotiana henthamiana), Millet (Panicum miliaceum), and Rice (Oryza sativa) had no obvious growth defects after treating with rapamycin even at very high concentrations (up to 20 μM) in solid medium. In order to examine the functions of TOR signaling pathway in potato, rapamycin sensitivity assays were tested. Data showed that potato explants had no obvious growth defects after treating with rapamycin, even at very high concentrations (up to 20 μM) compared to DMSO (a dissolvent of rapamycin; Figures S1C,E,F).

Due to the highly conservation of TOR FRB domain, the resistance to rapamycin of potato might result from loss function of the StFKBP12, which bridge the interaction between rapamycin and TOR. The analysis of relative expression level of StFKBP12 showed it was constitutive expression in potato, and the highest expression level was occurred in leaf, followed by shoot, and stem (Figure S1D). The constitutive expression of StFKBP12 allows us to mine the changeable structure of StFKBP12 to interpret the resistance to rapamycin of potato. The alignment of FKBP12 AA sequences showed that the five important AA in forming the complex “rapamycin-FKBP12-TOR” in StFKBP12 were similar to Arabidopsis, which may affect its function (Figure S1B). Early studies showed that the FKBP12 protein from yeast and human could restore the rapamycin sensitivity in Arabidopsis (Mahfouz et al., 2006; Sormani et al., 2007; Ren et al., 2012). We therefore introduced yeast FKBP12 (ScFKBP12), driving by a constitutive (cauliflower mosaic virus 35S) promoter, into potato. More than 20 individual transgenic lines (BP12-OE) were identified by leafy PCR (Figure S2A). No obvious morphological phenotypes appeared in any of transgenic plants. The ScFKBP12 expression level was detected by qRT-PCR analysis (Figure S2B). Four plants with high ScFKBP12 expression levels were examined by western blot (Figure S2C). Three of them were selected for rapamycin sensitivity assay. BP12-OE lines showed significant growth retardation on fresh weight and adventitious root formation (adventitious root numbers) compared with WT plants in a rapamycin dose dependent manner. With the increasing rapamycin concentration, the severer growth inhibition was observed (Figure 2).

Our previous studies showed that the synergism inhibitory effects can be generated by combining the first and second generation inhibitors of TOR in Arabidopsis (Xiong et al., 2017). To better understand TOR signaling in potato, asTORis were employed to examine TOR signaling in potato. The dosage-dependent inhibitory effects of asTORis on the growth of cutting explants were observed and the IC50 of AZD and KU were about 2 and 10 μM in potato, respectively (Figure S3). Moreover, the combination of rapamycin and KU was examined in BP12-OE17 and WT and the fresh weight was measured on the 10th day. Through computing combination index (CI) values, data showed there were synergistic effects on inhibiting the potato growth by using of rapamycin + KU simultaneously (Figure S4). The results were similar to Arabidopsis BP12-2 (a well-established rapamycin sensitive transgenic material by introduced a ScFKBP12 gene into Arabidopsis in our previous study) treated with these drug combination (Ren et al., 2012; Xiong et al., 2017). And then the potato explants were subculture into MS medium with TOR inhibitors for 3 weeks to observed TOR functions in potato seedling growth. The results show that suppression of TOR could strongly inhibit the growth of potato seedlings (Figures 3A,D). Under rapamycin + KU treatment, the newly emerging leaves from BP12-OE17 potato seedling displayed a yellowed phenotype and the number of adventitious roots was also reduced (Figures 3A,B,F). Consistent with our previous studies in Arabidopsis, root hairs were also strongly suppressed under TOR inhibitors treatment in potato (Figures 3C,E; Ren et al., 2012; Deng et al., 2016).

To examine the functions of TOR in adventitious root formation, explants (stem cuttings) of potato seedlings were generated and grown on 1 × MS culture medium with different TOR inhibitors for 10 days. Data showed that adventitious root formation was retarded in BP12-OE cuttings compared with WT, when the cuttings grown on the medium with rapamycin (Figure 4A). Consistently, the synergistic inhibitory effect of rapamycin + KU was also observed in adventitious root formation of BP12-OE lines (Figure 4A). The adventitious roots hardly occurred in BP12-OE lines under rapamycin + KU treatment (Figures 4B,C). We further used the previously constructed PTOR:: GUS to observe the expression of TOR during adventitious root formation. The expression of TOR in the beginning of adventitious root formation has been induced, and then with the development of adventitious root primordium, the TOR expression increased gradually (Figure 4D). These results suggested TOR may play an important role in adventitious root formation.

Next, the RNA-seq assay was performed to dissect the possible pathways associated with adventitious root formation which were regulated by TOR signaling in potato. The potato stem cuttings were treated with Rapamycin, KU and their combination RAP + KU for further RNA-seq assay. To investigate the early molecular events in response to TOR suppression in potato, the RNA were extracted from the samples treated with rapamycin, KU and their combination after 48 h, respectively, and then sequenced with Illumina Genome Analyzer (II). An overview of the sequencing and assembly was outlined in Table S3.

After trimming and filtering, the number of clean bases in treatment of DMSO, rapamycin, KU and rapamycin + KU was 8.10, 7.01, 7.91, and 8.88 Gb, respectively (Figure 5A). We mapped the clean reads to the potato reference genome (The Potato Genome Sequencing Consortium, 2011). The proportion of total clean reads in the four transcriptome libraries mapped to the reference diploid potato genome ranged from 65.11 to 66.54% (Figure 5B). The abundance of all the genes (46,217) was normalized calculated by the expected number of fragments per kilobase of transcript sequence per millions base pairs sequenced (FPKMs) method using unique mapped reads (Table S4). Genes with FPKMs over 60 were considered to be expressed at a very high level, and genes with FPKMs in the interval 0–1 were considered to be present at very low levels or not to be expressed. The distributions of the expression levels of all the genes were similar for all four treatments. We found that about 46% of genes (46,217) were lowly expressed (FPKM ≥ 1), and more than 2,248 genes were highly expressed (FPKM > 60). To evaluate the validity of transcriptome sequencing analysis, six candidate genes were selected and detected by qRT-PCR. The expressions of six genes according to the qRT-PCR results agreed well with the data from Illumina sequencing analysis (Figure S8A).

The rapamycin + KU treatment group got the most DEGs (2,899), while the KU treatment group got the fewest (776) (Figure 5C). After getting rid of the genes with different expression tendency among rapamycin vs. DMSO, KU vs. DMSO, and rapamycin + KU vs. DMSO, the total of unique 3,281 DEGs were found (1,830 were up-regulated and 1,451 were down-regulated; Table S2). To observe the gene expression patterns, the hierarchical clustering of all the DEGs were performed based on the log₁₀(FPKM + 1) of the four treatments (Figure 5D). GO and KEGG pathway analysis showed the conserved TOR functions in potato, such as participating cell wall restruction, photosynthesis, plant hormone signaling pathway etc., suggesting TOR is a core regulator of growth and development in potato (Tables S5, S6).

Further analysis revealed that a large number of genes associated with root development were affected in three treatment groups, particularly in the RAP + KU treatment group, including phytohormone, protein synthesis and degradation, cell division and cycle, cell wall restruction, and peroxidase (Tables S7–S10). The changes of auxin related genes are particularly evident in phytohormones related genes (Table S7), and the previous studies have showed there were some interconnections between auxin and TOR (Dinkova et al., 2000; Beltran-Pena et al., 2002; Schepetilnikov et al., 2013; Dong et al., 2015; Deng et al., 2016). We further examined potential crosstalk between TOR and auxin in the regulating of adventitious root formation.

The transcriptome data showed that the most DEGs were observed in auxin signaling pathway among all hormone signaling pathways. Table S7 showed that 39 genes associated with auxin signaling transduction were differentially expressed in RNA-seq data, implying that TOR played vital roles in auxin signaling pathway in potato (Table S7). DR5::GUS is a well-established and most commonly used auxin marker. In adventitious root formation, it has been found that auxin accumulation occurred in early stage (Liu et al., 2014; Rovere et al., 2016). In previous studies, we have generated DR5/BP12, a rapamycin sensitive Arabidopsis plant (Deng et al., 2016). To understand the relationship among auxin, TOR and adventitious root formation, DR5/BP12 seedling cuttings were treated with TOR inhibitors. Results showed auxin accumulation was significantly suppressed when TOR was inhibited in Arabidopsis. Especially when DR5/BP12 seedling cuttings were subjected to the medium with rapamycin + KU, adventitious root formation was completely inhibited and no GUS signal was detected (Figure S5A). In adventitious root formation, auxin accumulation mainly depends on the biosynthesis and polar transport of auxin which regulated by some important genes such as YUCCA6, PIN1, ABCB19, and LAX3 (Sukumar et al., 2013; Della Rovere et al., 2015; Chen et al., 2016; Rovere et al., 2016). Therefore, the expression level of these genes under TOR inhibitor treatment were examined. The data showed that the expression of these genes were suppressed by TOR inhibitors, especially the combination of rapamycin, and KU (Figures S5B,C).

In previous studies, Transport Inhibitor Response 1 /Auxin Response F-box (TIR1/AFBs) were well-identified as auxin receptors through the biochemical and genetic approaches (Dharmasiri et al., 2005a,b; Kepinski and Leyser, 2005). In order to further understand the mechanism of TOR and auxin signal dependent adventitious root formation, we collected their single, double and quadruple mutants, tir1, tir1afb2, tir1afb3, and tir1afb1afb2afb3. The seeds of these mutants (tir1, tir1afb2, tir1afb3, and tir1afb1afb2afb3) and WT were grown on the 0.5 × MS in the dark to generate long hypocotyl for 4 days. Root-excised hypocotyls of mutants and WT were transferred to the mediums with AZD (1 μM), KU (5 μM) or DMSO under normal growth condition. Adventitious root formation was significantly inhibited when the mutants were growing on the medium with the asTORis (Figures 6A,B). Adventitious root formation was completely suppressed in quadruple tir1afb1afb2afb3 plants grown on medium supplemented with TOR inhibitors (Figures 6C,D). Furthermore, we collected two gain-of-function mutant axr2-1 and axr3-1 which laid in the downstream of auxin receptors in auxin signaling pathways. AXR2 and AXR3 belong to AUX/IAA protein family and act as a repressor of auxin signaling transduction. These two mutant showed more sensitivity to TOR inhibitors compared to WT (Figures S7A,B). We therefore used AZD and KU to treat HS:AXR3NT-GUS, a well-established marker for detecting the degradation of AUX/IAA, to observe the effects of TOR inhibitors on auxin signaling transduction. When using the TOR inhibitors to treat HS:AXR3-NT-GUS, we observed that both the AXR3-GUS signal and GUS activities were stronger than the control (Figures S6B–D). These data indicated that the degradation of AUX/IAA was repressed and the auxin signaling transduction was inhibited when the activity of TOR was suppressed. Therefore, we proposed that TOR inhibitors may regulate the auxin signaling transduction by inhibiting the expression of TIR1/AFBs genes. And then, BP12-2 a well-established rapamycin sensitive line was treated with different TOR inhibitors to detect the expression level of TIR1/AFBs genes (Ren et al., 2012). The data show that TIR1/AFBs genes were down regulated by TOR inhibitors in BP12-2 (Figure S6A). To understand the genetic links between TOR and AtTIR1 (Arabidopsis TIR1), the 35S::TIR1 construct was introduced into BP12-2 to generate TIR1 overexpression lines (called TIR1/BP12 OE). More than 30 TIR1/BP12 OE lines were identified with leafy PCR, semi-qPCR, and western blot (Figures S7C,D). Three individual lines were selected to treat with TOR inhibitors. The data showed that the growth of TIR1/BP12-2 OE lines were significant better than BP12-2 under TOR inhibitors treatment (Figures S7E–G). Overexpression of AtTIR1 in BP12-2 is able to rescue adventitious root formation of BP12-2 under TOR inhibitors treatment (Figures 7A–D). We also examined the expression of two auxin responsive genes, GH3.2 and SAUR16. And data showed that overexpression of AtTIR1 may increase the expression of these genes under TOR inhibitors treatment compared to BP12-2 (Figure 7E). These results indicate that overexpression of AtTIR1 can enhance the auxin signaling transduction in BP12-2 under TOR inhibitors treatment to promote adventitious root formation.

Several recent studies have found that the regulation of TIR1/AFBs stability played an important role in auxin signal transduction (Yu et al., 2015; Wang et al., 2016). In order to further observe whether TOR plays a role in regulating the stability of TIR1/AFBs, P35S::TIR1-GUS was introduced into BP12-2 to generate TIR1-GUS overexpression lines (called TIR1-GUS/BP12). Consistently, TIR1-GUS/BP12 also grew better than BP12-2 under rapamycin treatment (Figure 8A). However, the GUS signals of TIR1-GUS/BP12 were significantly decreased under TOR inhibitors treatment (Figure 8B). We also tested the expression of GUS and GUS activity under different treatment. The data showed that the expression of GUS was not affected, but the activity of GUS was significantly reduced (Figures 8C,D), which may suggest that TIR1-GUS was degraded. These observations indicated that the stability of TIR was also regulated by TOR signaling.

In order to investigate the relationship of TOR and auxin signaling pathway in potato adventitious root formation, we screened the homologous auxin receptors gene of potato by using AtTIR1 amino acid sequence in NCBI and potato genome database. Five candidates were found in potato genome and an evolutionary tree was constructed based on their amino acid sequences (Figure 9A). qPCR data showed that the expression level of these auxin receptor-like genes were also affected by TOR inhibitors (Figure 9B). To examine whether AtTIR1 can overcomes the inhibitory effect of asTORis on adventitious root formation in potato, AtTIR1 was introduced into potato to generate AtTIR1-OE lines. More than 10 individual lines was generated and identified with leafy PCR and semi-qPCR (Figures S8B,C). Three individual lines were treated with asTORis to observe their adventitious root formation ability. The results showed that AtTIR1-OE lines were partially overcomes the inhibitory effect of asTORis and able to promote adventitious root formation under asTORis treatment compared with WT (Figures 9C–E). Taken together, these results showed that TOR and auxin had a closely relationship in regulating adventitious root formation, and its mechanism may be conserved evolution in potato and Arabidopsis.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.