A Glucose-6-Phosphate/Phosphate Translocator Is Involved in Carbohydrate Metabolism and Starch Biosynthesis in Sweet Potato (Ipomoea Batatas (L.) Lam.)

Sweet potato (Ipomoea batatas (L.) Lam.) is a good source of carbohydrates, an excellent raw material for starch-based industries, and a strong candidate for biofuel production due to its high starch content. However, the molecular basis of starch biosynthesis and accumulation in sweet potato is still insuciently understood. Glucose-6-phosphate/phosphate translocators (GPTs) mediate the import of glucose-6-phosphate (Glc6P) into plastids for starch synthesis. Here, we report the isolation of a GPT-encoding gene, IbG6PPT1, from sweet potato and the identication of two additional IbG6PPT1 gene copies in the sweet potato genome. IbG6PPT1 encodes a chloroplast membrane–localized GPT belonging to the GPT1 group and highly expressed in storage root of sweet potato. Heterologous expression of IbG6PPT1 resulted in increased starch content in the leaves, root tips, and seeds and soluble sugar in seeds of Arabidopsis thaliana, but a reduction in soluble sugar in the leaves. These ndings suggested that IbG6PPT1 might play a critical role in carbohydrate accumulation in storage tissues and would be a good candidate gene for controlling critical starch properties in sweet potato.


Background
Sweet potato (Ipomoea batatas (L.) Lam.) is an important food crop that is cultivated in over 100 countries due to its stable yield, rich nutrient content, low input requirements, multiple uses, high yield potential, and adaptability under a range of environmental conditions (Burri, 2011;Hu et al., 2003;Mitra, 2012;Yang et al., 2020;Zhang et al., 2017). Sweet potato is grown mainly for its edible, starchy storage root, which is 50-80% starch by dry matter (Zhou et al., 2015). This high starch content renders sweet potato a good source of carbohydrates, an excellent raw material for starch-based industries, and a strong candidate as an inexpensive raw material for biofuel production (Koçar and  Starch is synthesized in plants through a complex pathway involving multiple enzymes and transporters (Lai et al., 2016;Schreiber et al., 2014;Zhang et al., 2017). Starch biosynthesis begins with the synthesis of sucrose, the major product of photosynthesis, in source tissues. During this process, sucrose can be converted to glucose-6-phosphate (Glc6P) and then imported into the plastid by glucose-6phosphate/phosphate translocators (GPTs), proteins belonging to the transporter subfamily of phosphate translocators (PTs).
Two functional GPT genes have been identi ed in Arabidopsis (Arabidopsis thaliana). GPT1 is localized on the inner envelope of the plastid membrane and transports Glc6P into plastids for use in starch biosynthesis, the oxidative pentose phosphate pathway (OPPP), and fatty acid biosynthesis (Kunz et al., 2010). GPT1 is also localized to peroxisomes, where it preferentially exchanges Glc6P for ribulose-5phosphate (Ru5P) (Baune et al., 2020).
GPTs play important roles in several physiological processes. Arabidopsis GPT1 is essential for the development of male and female gametophytes, embryos, and seeds Zhang et al., 2020). GPT1 is highly expressed at the late stages of pollen development, where it drives Glc6P from the cytosol and into plastids for fatty acid biosynthesis, and thus plays an important role in lipid body biogenesis during pollen maturation (Zheng et al., 2018). By contrast, Arabidopsis GPT2 is expressed when photosynthesis is increased by light, which allows increased net import of Glc6P from the cytosol to chloroplasts, thus facilitating starch synthesis during stochastic high-light conditions (Dyson et al., 2015;Weise et al., 2019). GPT2 responds rapidly by glucose and sucrose induced and plays an essential role in interpreting environmental signals Weise et al., 2019). GPT1 plays a major role in the regulation of starch synthesis in other plants. GPT1 in Narbonne vetch (Vicia narbonensis) is critical for starch synthesis and storage in developing seeds. In Vicia transgenic plants expressing antisense GPT1 via Agrobacterium-mediated transformation, amyloplasts developed later and were smaller in size, starch biosynthesis was reduced, and storage protein biosynthesis increased (Rolletschek et al., 2007). In rice, pollen grains from homozygous osgpt1 mutant plants fail to accumulate starch granules, resulting in pollen sterility (Qu et al., 2021). However, the role of GPTs in sweet potato has not been investigated.
In our previous work, we identi ed GPT genes in sweet potato using transcriptome analysis and showed that they have different expression patterns during storage root development and among sweet potato genotypes with different starch properties. Therefore, these GPTs are probably involved in starch accumulation and sucrose metabolism in sweet potato (Zhang et al., 2017). Here, we cloned two GPT genes and analyzed their protein localizations, sequence features, and functions. Our results provide important insights into the mechanisms underlying the starch properties of sweet potato.

Results
Two GPT-encoding genes were cloned from sweet potato To ensure that the full-length mRNA sequence of sweet potato GPT genes could be obtained, the RACE method was used for cloning. Two cDNA sequences encoding the sweet potato GPT genes were obtained, named IbG6PPT1 and IbG6PPT1-2. The obtained sequences contained 1763 and 1767 bp of mRNA, corresponding to 1191 and 1200 bp of ORFs and encoding 399-aa and 396-aa protein sequences, respectively. The two genes shared 96.627%, 98.083%, and 98.747% identity at the mRNA, CDS, and putative amino acid levels, respectively. The two proteins differed in only ve amino acids (Fig. 1), including a deletion of the L 37 P 38 A 39 sequence in the shorter GPT.
The sweet potato genome is annotated with three GPT1 gene members: IbG6PPT1 located on chromosome 3 (chr3), IbG6PPT1-2 located on chr2, and another IbG6PPT1-like gene also expected to be located on chr2. However, the sequence of this IbG6PPT1-like gene was not cloned from our cDNA library. Amino acid differences between IbG6PPT1 and IbG6PPT1-2 were not located at conserved domains or important transmembrane domains, indicating that these proteins are likely functional.
The protein sequence alignment and phylogenetic tree showed that IbG6PPT1 was more similar than IbG6PPT1-2 to GPT1 proteins from Ipomoea nil and Ipomoea triloba ( Fig. 1 and Fig. 2), indicating that IbG6PPT1 might match previous GPT1 ndings better in the Ipomoea genus. Thus, we focused on IbG6PPT1 for the remainder of this work.

IbG6PPT1 is likely a chloroplast-located GPT
We created a GFP-tagged version of IbG6PPT1 and transiently expressed it in N. benthamiana. The GFP signal surrounded the chloroplast marker uorescence, indicating that IbG6PPT1 localizes to the chloroplast membrane (Fig. 3). No signal peptide was predicted from the protein sequence, meaning that the protein wasn't secreted protein. TMPred and TMHMM predicted the presence of seven transmembrane helixes (Fig. 1), consistent with a role for IbG6PPT1 in Glc6P transport through the chloroplast membrane. Modeling of the three-dimensional (3D) structure of IbG6PPT1 predicted that two IbG6PPT1 proteins form a homodimer (Fig. 4). In addition, IbG6PPT1 contains a conserved sugar phosphate transporter domain (Jack et al., 2001). These results strongly suggest that IbG6PPT1 is a chloroplast membrane-localized, GPT family protein in sweet potato.

IbG6PPT1 is highly expressed in sweet potato storage root
We next attempted to determine whether and how strongly IbG6PPT1 is expressed in various sweet potato tissues by extracting RNA from the petiole, stem, leaf, and storage root and performing RT-qPCR to quantify gene expression. IbG6PPT1 was expressed in all tissues but showed its highest expression in storage roots, followed by the petiole, stem, and leaves (Fig. 5). Interestingly, the higher expression of IbG6PPT1 in roots than in leaves suggests that it may function in Glc6P transport in non-green tissues rather than in photosynthetic tissues.

Heterologous expression of IbG6PPT1 affects starch and sucrose content
To identify the function of IbG6PPT1, we transformed a p35S::IbG6PPT1-YFP construct into wild-type (Col-0) Arabidopsis. Analysis of IbG6PPT1-YFP expression by qPCR and western blotting showed that the fusion protein was heterologously expressed across four independent transgenic lines, designated OV-14, OV-30, OV-76, and OV-57, that were selected from the T 2 progeny, as well as in the wild type ( Fig. 6a and   6b). There were no differences in growth and development between the transgenic progeny and the wildtype control (Fig. 6c).
In contrast to their wild-type-like appearance, the soluble sugar content in the leaves of the transgenic IbG6PPT1-YFP lines was only 76.59-83.40% of control (Fig. 7a, Table S1). Meanwhile, the leaves of the 6 weeks transgenic T 2 plants had a measured starch content 1.65-to 2.75-fold higher than the control (Fig.   7b, Table S1), which was con rmed by iodine staining in 3 weeks seedlings (Fig. 7c). Surprisingly, the 1000 seed weights of the transgenic lines were 1.06-to 1.19-fold higher than in the control plants (Fig. 7d, Table S1). Further analyses showed that the soluble sugar content and starch content in the seeds of transgenic IbG6PPT1-YFP lines were 1.20-to 1.47-fold and 1.13-to 1.31-fold higher than in the control plants, respectively ( Fig. 7e and 7f, Table S1). In the root tips, iodine staining showed that the starch content of transgenic IbG6PPT1-YFP lines was higher than that in control plants ( Fig. 7g and 7h). Above all, heterologous expression of the IbG6PPT1 gene increased both starch and sucrose content in Arabidopsis.

Discussion
IbG6PPT1 is present in several gene copies that may have different functions The sweet potato genome is allohexaploid (2n = 6x = 90), containing two B1 and four B2 component . Therefore, there may be up to six copies of each gene. In this study, we cloned two GPT1 genes that share a high level of identity in both the mRNA and protein sequences (Fig. 1). However, we found three potential IbG6PPT1 genes in the genome database, the two we cloned and another one on chr2 that might be a homolog or paralog of one of the cloned genes. During the evolution of sweet potato's polyploid genome, the IbG6PPT1 has similar functions to other GPT1 proteins GPT1 proteins transport Glc6P into plastids for starch and/or fatty acid biosynthesis, depending on the plant species (Zheng et al., 2018). A previous report demonstrated that starch biosynthesis is mediated by VnGPT1 in pea embryos (Rolletschek et al., 2007). Starch is the major carbon storage molecule of sweet potato, accounting for 50-80% of dry matter in the storage root, the organ that determines sweet potato's economic value as a crop (Zhang et al., 2017), whereas fatty acids are almost undetectable. In Arabidopsis, fatty acid biosynthesis in pollen is controlled by regulating AtGPT1 expression through the MKK4/MKK5-MPK3/MPK6 cascade and the downstream transcription factors WRKY2 and WRKY34 (Zheng et al., 2018). In rice, pollen grains from homozygous osgpt1 mutant plants fail to accumulate starch granules, resulting in pollen sterility (Qu et al., 2021). Like AtGPT1, which is expressed ubiquitously throughout Arabidopsis development (Niewiadomski et al., 2005), we found that IbG6PPT1 is expressed in both aboveground and underground organs in sweet potato (Fig. 5), suggesting potential functions in both autotrophic and heterotrophic tissues. The localization of IbG6PPT1 to the chloroplast membrane ( Fig. 3) implied that it may function in transporting Glc6P from the cytosol into plastids.
To better elucidate the function of the IbG6PPT1 gene in starch accumulation, we cloned IbG6PPT1, heterologously expressed it in Arabidopsis, and then measured starch accumulation in the resulting transgenic plants. IbG6PPT1 expression increased starch accumulation in Arabidopsis leaves, seeds, and root tips, suggesting that it promotes starch biosynthesis. Lipid bodies and protein are the major storage compounds in mature Arabidopsis seeds, each accounting for up to 40% of the dry weight (Andriotis et al., 2010), whereas starch are lower. The weight of 1000 seed we observed in IbG6PPT1-expressing plant is greater than control plant, indicating that IbG6PPT1 may also promote storage matter accumulation in Arabidopsis seeds.

IbG6PPT1 promotes carbohydrate accumulation in Arabidopsis storage tissues
Sucrose is a major end product of photosynthesis and the primary sugar transported within plants (Winter and Huber, 2000). In heterotrophic tissues, sucrose imported from photosynthetic tissues is converted to Glc6P, and some Glc6P can be transported into the plastid through GPTs for starch and/or fatty acid biosynthesis. Another portion of the Glc6P is metabolized in the cytosol to phosphoenolpyruvate (PEP), which is essential for the biosynthesis of lipids and other storage substances (Lee et al., 2017). In IbG6PPT1-expressing Arabidopsis, the starch content in the leaves increased signi cantly, while the soluble sugar content was reduced, compared to that in control plants (Fig. 7). Thus, heterologous expression of IbG6PPT1 promoted starch accumulation and sucrose metabolism, probably due to the high expression of GPT, which would be expected to increase the level of Glc6P imported into the chloroplast or amyloplast for starch synthesis. IbG6PPT1 expression in Arabidopsis would promote Glc6P transport into plastids for storage and thus contribute to the observed carbohydrate accumulation in transgenic seeds compared with controls. Therefore, it is also likely that IbG6PPT1 plays a critical role in metabolic distribution and carbohydrate accumulation in storage tissues. This also indirectly implies that IbG6PPT1 plays an important role in the accumulation of substances in the storage roots of sweet potatoes.
IbG6PPT1 enhances transport activity from sink to source In plants heterologously expressing IbG6PPT1, the starch contents in leaves, seeds, and root tips increased (Fig. 7), indicating that IbG6PPT1 might have high activity in Glc6P transport and play an important role in starch biosynthesis. Sweet potato is one of the top starch-rich root crops globally (Wang et al., 2019). To ensure starch storage in the storage root, any increases in photosynthetic activity depend on corresponding development of carbon sink capacity. Thus, Glc6P transporters such as IbG6PPT1 should exhibit a high capacity for photosynthate transport. Indeed, heterologous expression of IbG6PPT1 increased the starch content in the leaves, seeds, and root tips in Arabidopsis. However, expression did not affect the growth and development of transgenic plants, suggesting the potential of IbG6PPT1 to promote starch accumulation in other crops.

Conclusion
The changes in starch and sugar content of transgenic Arabidopsis plants showed that IbG6PPT1 might play an important role in starch and sucrose metabolism. It is probably an important gene controlling the starch properties in sweet potato. These ndings will help to elucidate the genetic basis and regulatory mechanisms underlying starch properties in sweet potato.

Plant material and growth conditions
The sweet potato varieties Xushu22 (XS22) was cultivated at temperatures of between 22 and 28°C in the experimental base of the Sweet Potato and Potato Research Institute, Southwest University, Chongqing, China. Leaf, stem, petiole, and root were sampled at 95 days after transplanting (DAP) and quickly frozen in liquid nitrogen then stored at -80°C until use for RNA extraction. All Arabidopsis thaliana and Nicotiana benthamiana plants were grown in a 22°C and 28°C climate chamber (16h light/8 dark) in Longping experimental building, Southwest University, Chongqing, China.
Expression pattern assay RNA (1 μg) was extracted from the leaf, stem, petiole, and root of 95 DAP in XS22 and was reverse transcribed in a 20 μL volume by the PrimeScript RT Master Mix (TaKaRa) according to the manufacturer's instructions. The expression pattern of GPT genes was detected using primers and RT-qPCR methods as previously described (Zhang et al., 2017). Fold changes of the GPT transcripts were calculated according to the 2 -△△Ct method with three samples.

Subcellular localization
The full coding sequence (CDS) of IbG6PPT1 was cloned into pCAMBIA1300, and a GFP tag was fused to the C terminus of the gene. This construct was transformed into Agrobacterium tumefaciens strain GV3101 and transiently expressed in Nicotiana benthamiana using syringe agroin ltration (Huy et al., 2016). GFP uorescence was observed using a Zeiss LSM780 confocal laser scanning microscope (Li et al., 2019). Signals were detected using excitation/emission wavelengths for GFP (488 nm/495-535 nm) and the chloroplast marker (633 nm/660-720 nm).

Heterologous expression of IbG6PPT1 in Arabidopsis
The full CDS of IbG6PPT1 was recombined from the Gateway entry vector pENTR-D-TOPO (see the cloning and sequence analysis method above) into the destination vector pEarleyGate101 (Earley et al., 2006), yielding the construct p35S::IbG6PPT1-YFP, which has an N-terminal YFP tag. The construct p35S::IbG6PPT1-YFP was transformed into Arabidopsis using the Agrobacterium tumefaciens-mediated oral dip method (Desfeux et al., 2000).
Positive transgenic lines were identi ed by PCR detection of YFP using the primers YFP-Fwd (5'-TGGTCGAGCTGGACGGCGACGTAAAC-3') and YFP-Rev (5'-TTCTCGTTGGGGTCTTTGCTCAGGGC-3') and by detection of the bar gene in the construct using the primers FBar (5'-TGGGCAGCCCGATGACAGCGACCAC-3') and RBar (5'-ACCGAGCCGCAGGAACCGCAGGAGT-3'). IbG6PPT1 expression in the transgenic Arabidopsis plants was detected using the RT-qPCR method described in the expression pattern assay section. YFP expression was detected by western blotting using an anti-GFP antibody (Welsch et al., 2019). Thousand seed weight (g) was determined for 1000 seeds from each sample with three replicates.

Starch and sugar measurement
The starch and soluble sugar contents of leaves and seeds in transgenic and control Arabidopsis plants were determined using a previously published method (Kunz et al., 2010). Brie y, leaves and roots of 3 weeks seedlings were stained with an iodine solution (2% KI+1% I 2 ) and examined under a light microscope (Nikon, Japan), and images were captured using NIS-Elements BR 4.30.00 software as previously described (Kunz et al., 2010

Con icts of Interest
The authors declare no con ict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its additional les. About proteins database could download from NCBI by their accession number.  Figure 1 Alignment of IbG6PPT1, IbG6PPT1-2, and Ipomoea genus GPT1 proteins. ItG6PPT1, Ipomoea triloba GPT1 (XP_031105621.1); InG6PPT1, Ipomoea nil G6PPT1 (XP_019193616.1). The amino acids underlined in red form transmembrane helixes based on prediction using TMHMM; black and grey highlighting indicate amino acid differences between the species.   Predicted three-dimensional structure models of IbG6PPT1. Two IbG6PPT1 proteins (shown in yellow and blue) form a dimer.

Figure 5
Expression of IbG6PPT1 in the petiole, storage root, stem, and leaf of the sweet potato variety Xushu 22, as determined by qRT-PCR. Each value is the mean±SD of at least three independent measurements.

Figure 6
Heterologous expression of IbG6PPT1 in Arabidopsis. (a) qRT-PCR detection of IbG6PPT1 expression; each value is the mean±SD of at least three independent measurements. (b) Western blot detection of