Tonoplast Sugar Transporters (SbTSTs) putatively control sucrose accumulation in sweet sorghum stems

ABSTRACT

Carbohydrates are differentially partitioned in sweet versus grain sorghums. While the latter preferentially accumulate starch in the grain, the former primarily store large amounts of sucrose in the stem. Previous work determined that neither sucrose metabolizing enzymes nor changes in Sucrose transporter (SUT) gene expression accounted for the carbohydrate partitioning differences. Recently, 2 additional classes of sucrose transport proteins, Tonoplast Sugar Transporters (TSTs) and SWEETs, were identified; thus, we examined whether their expression tracked sucrose accumulation in sweet sorghum stems. We determined 2 TSTs were differentially expressed in sweet vs. grain sorghum stems, likely underlying the massive difference in sucrose accumulation. A model illustrating potential roles for different classes of sugar transport proteins in sorghum sugar partitioning is discussed.

Abbreviations

TSTs

Tonoplast Sugar Transporters

SUTs

Sucrose Transporters

qRT-PCR

quantitative reverse-transcription polymerase chain reaction

Sucrose harvested from plants represents a multi-billion dollar (US) annual industry, with great interest in expanding production for food and biofuel uses.^1-5 Multiple crops have been independently bred to store high concentrations of sucrose in terminal storage organs, namely, the taproots of sugar beet (Beta vulgaris L.), and the stems of sweet sorghum (Sorghum bicolor L. Moench) and sugarcane (Saccharum officinarum L.).^5-9 However, the sucrose contents of these crops appear to be approaching maximal levels attainable from breeding efforts^10,11; therefore, new approaches are needed to increase sucrose accumulation in storage organs. Hence, characterizing the genes that function in sucrose transport and storage will reveal potential new targets for future manipulations to enhance crop yields.^3,4,12-16

Different sorghum genotypes have been selectively bred to store carbohydrates in contrasting storage organs: sweet sorghums accumulate large quantities of soluble sugars, mostly sucrose, in stem tissues, whereas grain sorghums primarily store carbohydrates as starch in the seeds.^1,2,6,17,18 The molecular basis for the difference in carbohydrate partitioning between these sorghum types is unknown. Previous research found that sucrose accumulation within sweet sorghum stems was not correlated with the activities of enzymes involved in sucrose metabolism, invoking sucrose transport proteins as potentially controlling sucrose content.¹⁹ Transport experiments using asymmetrically radiolabeled sucrose determined that sucrose movement into stems likely included an apoplasmic transport step.²⁰ Subsequent dye transport studies suggested the phloem tissues within sorghum stems are symplasmically isolated from surrounding tissues, supporting that sucrose phloem unloading occurs apoplasmically, and thus requires sucrose transport proteins.²¹ However, other studies support a possible symplasmic transport route from phloem sieve elements to storage parenchyma cells in mature stems.²² Quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analyses indicated that Sucrose Transporters (SUTs), which function as H⁺/sucrose symporters to transport sucrose across membranes, were not differentially expressed in the stem of a flowering sweet sorghum line, UNL71-2011, a sweet sorghum derived from cultivar Wray, in comparison to a similarly staged grain sorghum line, UNL3016, selected from cultivar Macia.²¹ These data suggest other types of sucrose transport proteins may underlie sucrose accumulation within sorghum stem tissues.

Two additional distinct classes of sucrose transport proteins have recently been described. SWEETs are a family of sugar transport proteins, with different family members preferentially transporting hexoses or sucrose.^23-28 Clade III SWEET proteins, which localized to the plasma membrane, have been proposed to function as uniporters that facilitate the transport of sucrose down a concentration gradient. A different class of sugar transport proteins, Tonoplast Sugar Transporters (TSTs, also known as Tonoplast Monosaccharide Transporters), is located on the tonoplast and function as H⁺/sucrose antiporters to transport sucrose into the vacuole.^29-32 Recently, a TST was shown to be responsible for sucrose accumulation within the sugar beet taproot.¹² Whether TST or SWEET genes have also been selected during the domestication of other major sucrose storage crops, such as sweet sorghum or sugarcane stem tissues, is not known. Since SbSUT genes were not differentially expressed between sweet and grain sorghum stem tissues,²¹ we decided to examine the expression of other predicted sucrose transport proteins, specifically, the clade III SbSWEET and the SbTST genes. For these studies, we compared gene expression between the sweet sorghum line UNL71-2011 at anthesis, when sugars are actively accumulating in the stem, with the equally staged grain sorghum line UNL3016, with low stem sugar content, to determine if any SbSWEET or SbTST genes are associated with sucrose accumulation in stem tissues. These lines are herein simply referred to as sweet and grain sorghums for clarity. As a point of reference, we found that the total solute levels, consisting primarily of sucrose, increased approximately 24-fold in sweet sorghum stems compared with grain sorghum stems during the ripening process from anthesis to physiological maturity.²¹

Bioinformatic analyses were used to identify SbTST and SbSWEET genes in the sorghum genome. Three SbTST genes and 20 SbSWEET genes were identified.²⁴ We next analyzed a sorghum gene expression database to determine which of these genes were expressed in leaf and stem tissues (Table 1).³³ SbSWEET13A was the most strongly expressed clade III gene within these tissues. Its expression was more than 10-fold higher than other clade III sweet genes; SbSWEET13B and SbSWEET13C had lower, but appreciable, expression relative to SbSWEET13A. The other clade III SbSWEET genes likely to transport sucrose were all very lowly or not detectably expressed (Table 1). Therefore, we selected the 3 SbSWEET13 genes and all SbTST genes for further expression analyses. Gene-specific qRT-PCR primer sets were validated for each gene (Table 2). The qRT-PCR experiments and statistical analyses were performed as previously described.²¹ In examining mature leaf and ripening stem tissues of both grain and sweet sorghum, we determined that SbTST1, SbTST2, and SbSWEET13A were reliably expressed, whereas SbTST3, SbSWEET13B, and SbSWEET13C were expressed at a much lower level (Figs 1–2). SbSWEET13B expression was at least 33-fold less than SbSWEET13A in all tissues examined. Similarly, SbSWEET13C was expressed at least 20-fold less than SbSWEET13A in leaves of both cultivars. SbSWEET13C was expressed approximately 6.7- and 29-fold lower than SbSWEET13A in grain and sweet sorghum stems, respectively; however, SbSWEET13C exhibited lower expression in sweet sorghum stems compared to grain sorghum stems, which is the opposite of what was hypothesized if the gene functioned to promote sucrose accumulation in sweet sorghum stems. Based on these results, we conclude that SbTST3, SbSWEET13B, and SbSWEET13C are minimally expressed in mature leaf and ripening stem tissues and therefore do not likely contribute substantially to sucrose accumulation within these tissues.

Table 1.

The expression level of the SbTST and SbSWEET genes by RNA-seq in the leaf and stem tissues. The numbers represent the average expression obtained from the FPKM (fragments per kilobase of transcript per million mapped reads) plots.

Gene name	Phytozome reference no.	Gene ID.	Leaf	Stem
SbTST1	Sobic.001G312900	Sb01G030430	105	149
SbTST2	Sobic.004G099300	Sb04G008150	110	130
SbTST3	Sobic.010G276100	Sb10g031000	Not detected	Not detected
SbSWEET11AIII	Sobic.007G191200	Sb07g026040	Not detected	10
SbSWEET11BIII	Sobic.002G259300	Sb02g029430	Not detected	Not detected
SbSWEET12III	Sobic.001G373600	Sb01g035490	Not detected	Not detected
SbSWEET13AIII	Sobic.008G094000	Sb08g013620	2250	200
SbSWEET13BIII	Sobic.008G094300	Sb08g013840	28	2
SbSWEET13CIII	Sobic.008G094400	Sb08g014040	120	20
SbSWEET14III	Sobic.005G123500	Sb05g018110	1	3
SbSWEET15III	Sobic.004G157100	Sb04g021000	Not detected	1
SbSWEET16IV	Sobic.001G377600	Sb01g035840	Not detected	Not detected
SbSWEET1AI	Sobic.003G377700	Sb03g041740	100	135
SbSWEET1BI	Sobic.009G143500	Sb09g020860	82	2
SbSWEET2AI	Sobic.003G182800	Sb03g024250	15	8
SbSWEET2BI	Sobic.003G269300	Sb03g032190	75	4
SbSWEET3AI	Sobic.009G080900	Sb09g006950	135	300
SbSWEET3BI	Sobic.003G015200	Sb03g001520	Not detected	1
SbSWEET4AII	Sobic.004G136600	Sb04g015420	85	35
SbSWEET4BII	Sobic.004G133500	Sb04g012910	20	30
SbSWEET4CII	Sobic.004G133600	Sb04g012920	Not detected	Not detected
SbSWEET5II	Sobic.009G252000	Sb09g030270	Not detected	Not detected
SbSWEET6II	Sobic.003G213000	Sb03g027260	30	10

The numbers are estimated from the FPKM plots obtained from the MOROKOSHI sorghum transcriptome database (http://sorghum.riken.jp/morokoshi).³³ The different upper-case roman letter superscripts indicate the clade to which each SbSWEET gene belongs. Not detected means that the gene had no detectable RNAseq counts in the corresponding tissue. The underlined genes are those selected for this study.

Table 2.

List of primers used.

Primer name	Primer sequence (5′ – 3′)	Product size (bp)
SbTST1-F	GATGGGCTGACCTGTTTG	175
SbTST1-R	GCAGAAGATGCGCTAAGG	175
SbTST2-F	TTGGAGGTTGGAGGAGAC	150
SbTST2-R	CTTGGAAGGTCGAGCAATC	150
SbTST3-F	CTGTTGCTTCGTCATGGG	146
SbTST3-R	TGACAGGAAGAGAGTAGGTG	146
SbSWEET13A-F	CGCTCACTACTGCTAAGTATTAT	96
SbSWEET13A-R	ACAGTAGTCTGGGATCGATTA	96
SbSWEET13B-F	CATGAGTCGAGTCCGAATG	116
SbSWEET13B-R	AGCTACGGTTGGATAAACG	116
SbSWEET13C-F	ACCCGTTTATCCAACCCTTAG	87
SbSWEET13C-R	TGAAATTCCTGCCTGGTTACA	87
Luciferase-F	CCAGGGATTTCAGTGGATGT	183
Luciferase-R	AATCTGACGCAGGCAGTTCT	183

Figure 1.

Expression levels of SbTST2 and SbTST3 relative to SbTST1 in grain and sweet sorghum mature leaves and stems. A, B show grain sorghum (black bars), and C, D show sweet sorghum (white bars); A, C are mature leaf tissues, and B, D are flowering stems. Values are means ± standard error of N = 5 plants, and an asterisk indicates significantly different means between the 2 genes at p ≤ 0.05. Relative gene expression is shown compared to exogenously added Luciferase RNA as a normalization control.²¹

Figure 2.

Expression levels of SbSWEET13B and SbSWEET13C relative to SbSWEET13A in grain and sweet sorghum mature leaves and stems. A, B show grain sorghum (black bars), and C, D show sweet sorghum (white bars); A, C are mature leaf tissues, and B, D are flowering stems. Values are means ± standard error of N = 5 plants, and an asterisk indicates significantly different means between the 2 genes at p ≤ 0.05. Relative gene expression is shown compared to exogenously added Luciferase RNA as a normalization control.²¹

Based on the previous results, we examined SbTST1, SbTST2, and SbSWEET13A to determine if they are differentially expressed between sweet and grain sorghum leaf and stem tissues. In mature leaves, SbTST1 and SbTS2 showed ∼3.5-fold and ∼7.4-fold higher expression levels in sweet sorghum relative to grain sorghum (p ≤ 0.05; Fig. 3A, B). Within stem tissues at anthesis, SbTST1 and SbTST2 showed significantly higher expression levels in sweet sorghum compared to grain sorg-hum (∼2.6- and ∼4.4-fold, respectively) (Fig. 3A, B). SbSWEET13A showed reduced expression in sweet compared to grain sorghum leaves and comparable expression in stem tissues of both genotypes (Fig. 3C). These data indicate SbTST1 and SbTST2 are significantly more highly expressed in leaves and stem tissues of sweet sorghum than in grain sorghum, and that SbSWEET13A expression was reduced in sweet sorghum leaves compared to grain sorghum but not differently expressed in stem tissues. Thus, these data suggest that differential expression of SbTST1 and SbTST2 genes, but not SbSWEET13A may play an important role in sugar accumulation in sweet sorghum stems. To our knowledge, no previous reports have shown the differential expression of SbTSTs associated with sugar accumulation in the stems of sweet vs. grain sorghum.

Figure 3.

Expression levels of SbTST1, SbTST2, and SbSWEET13A in leaves and stems of sweet sorghum relative to grain sorghum. Expression levels are shown for SbTST1 (A), SbTST2 (B), and SbSWEET13A (C). An asterisk indicates significantly different means between the 2 lines at p ≤ 0.05 of N = 5 plants.

From our expression studies, we developed a model of the various sucrose transporter protein functions to explain the basis of sugar accumulation within sorghum leaf and stem tissues and to stimulate new directions in research (Fig. 4). Within leaves, SbSUT2 and SbSUT4, but not SbSUT1, were more highly expressed in sweet sorghum than in grain sorghum, suggesting that SbSUT4 may function to import sucrose into cells, and SbSUT2 may function to export transitory stored sucrose from the vacuole.²¹ SbSUT1 function is likely conserved between grain and sweet sorghum, and based on orthology with the maize (Zea mays) ZmSUT1 gene, it likely functions in sucrose phloem loading in leaves.^34-36 SbSWEET13A showed reduced expression in sweet compared with grain sorghum leaves, whereas SbTST1 and SbTST2 were both more highly expressed in sweet sorghum leaves, suggesting that they may function to import sugars into the vacuole for temporary storage during daylight. In stem tissues, none of the SbSUT or SbSWEET13A genes were differentially expressed, suggesting they do not account for the differences in sugar accumulation. However, both SbTST1 and SbTST2 were highly significantly expressed in sweet sorghum stems, suggesting these genes function to import sucrose for storage in the vacuole within stem parenchyma cells. Based on these results, we hypothesize the ∼24-fold increase in total stem solutes observed in sweet sorghum compared with grain sorghum is predominantly due to the significantly higher expression of SbTST1 and SbTST2 in sweet sorghum tissues.

Figure 4.

A model illustrates the roles for different sucrose transport proteins in sucrose movement across cellular membranes in sorghum leaf (green shaded background) and stem (blue shaded) tissues in grain sorghum (left) vs. sweet sorghum (right). The vacuole is shown in gray. SPC = stem parenchyma cell. SUT proteins are shown by a blue circle, with an arrow indicating the direction of sucrose movement, and the numbers correspond to SbSUT1, SbSUT2, or SbSUT4. Purple diamond with an arrow refers to a TST protein located on the tonoplast, and the numbers represent SbTST1 or SbTST2. The green boxes labeled 13A correspond to SbSWEET13A. The increased size of the shapes indicates increased expression of the corresponding gene in sweet (UNL 71-2011) vs. grain (UNL 3016) sorghum tissue.

In summary, based on both our previous and current results, we determined SbTST1 and SbTST2, but probably not SbSUTs or SbSWEETs, are likely responsible for the substantial sugar accumulation in sweet sorghum stems. Testing of this hypothesis will potentially require characterizing loss-of-function mutations in both genes, since they were found to be partially functionally redundant in Arabidopsis thaliana.²⁹ These efforts are currently underway. Furthermore, these data suggest TSTs have been the target of selection for sugar accumulation in both the sweet sorghum stem and the sugar beet taproot. It will be interesting to determine whether TSTs have similarly been selected within sugarcane stem tissues. If so, it would indicate TSTs have been convergently selected during domestication of the world's 3 major sucrose storage crops. Our findings tantalizingly suggest SbTST1 and SbTST2 are candidate genes for the control of sucrose accumulation in sweet sorghum stems. Hence, modifying the expression or function of TSTs through genetic engineering or selective breeding, could potentially achieve greater sucrose accumulation and therefore enhancement of crop yields in sugar-storing organs, which would lead to direct benefits for food and fuel production.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed

Acknowledgments

We thank R. Frank Baker for comments on the manuscript. We wish to clarify UNL71-2011 and UNL3016 are the same lines referred to as cultivars Wray and Macia, respectively, in our previous publication.²¹ No conflicts of interest declared. This research was supported by the US DOE Office of Science, Office of Biological and Environmental Research (BER), grant no. DE-SC0006810, and by the US National Science Foundation Plant Genome Research Program, grant no. IOS–1025976, to DMB.