Expression of a truncated maize SPS increases source capacity in maize

In an attempt to increase source capacity, transgenic corn was generated by expressing a truncated maize sucrose phosphate synthase (ZmSPSΔ482) under two leaf mesophyll cellspecific promoters (CAB and PPDK). The endogenous and truncated SPS proteins from transgenic leaf extracts were distinguishable by protein immunoblot analysis. The expression of transgenic SPS protein across events varied from very low to very high and included several cosuppressed events. SPS activity showed a diurnal pattern in both transgenic and wild-type maize leaves. In greenhouse experiments, transgenic maize had higher leaf sucrose and lower leaf starch, suggesting a shift in carbon partitioning from starch to sucrose. Conversely, cosuppressed events had lower leaf sucrose and higher leaf starch. A field test was performed to compare sucrose and starch in positive and negative isolines of hybrid maize CAB and PPDK ZmSPSΔ482 events. In the field, many positive isolines had higher levels of both leaf sucrose and starch than the negative isolines. This suggests that in the field, with higher light intensity the shift in carbon partitioning from starch to sucrose, observed under greenhouse conditions did not occur. This in turn suggests that the environment affects the phenotype of the transgenics and that in the field, there was an overall increase in carbon assimilation. Six events from each construct were tested in a pilot multi-density yield trial but overall, no effect on yield was observed. Therefore, although the transgenic plants had more source capacity, this did not translate into higher seed yield.


Introduction
Starch and sucrose are primary stable end products of carbon assimilation in higher plants. Both fluctuate diurnally in leaves and reach their highest measurable levels late in the day. Starch is synthesized and stored in the chloroplast while sucrose is synthesized in the cytosol from triose phosphates transported from the chloroplast. While starch is considered a storage form of assimilated carbon, sucrose is the main transport carbohydrate in higher plants (Hofstra and Nelson 1969;Stitt et al. 1987;ap Rees 1987;Lunn and Furbank 1999). In maize, sucrose is primarily produced in mesophyll cells (Lunn and Hatch 1997;Downton and Hawker 1973;Usuda and Edwards, 1980;Furbank et al. 1985;Lunn and Furbank 1999). It is then transported to non-photosynthetic tissues, or sinks, for storage or immediate energy production.
In reproductive organs such as kernels, sucrose is hydrolyzed to hexose sugars which are then processed to provide the precursors for starch, lipids, and protein production.
Increasing sucrose production in leaves may increase source capacity available for sink tissues, and especially to developing kernels during high sink demand and may be a useful approach to increase grain yield.
Sucrose phosphate synthase (SPS, EC 2.4.2.14) is a key enzyme for sucrose synthesis in mature photosynthetic tissues (Stitt et al. 1987;Huber and Huber 1996 and references therein). SPS catalyzes the rate-limiting step in sucrose biosynthesis through the formation of sucrose-6-phospate from fructose 6-phosphate and UDP-glucose (Winter and Huber 2000). Sucrose 6-phosphate is further broken down to sucrose by a closely associated enzyme, sucrose 6-phosphate phosphatase (SPP, Echeverria et al. 1997). SPS activity is highly regulated in plants, both allosterically by glucose 6-phosphate and phosphate, and by reversible 1 3 phosphorylation (Lunn and Furbank 1999;Winter and Huber 2000;Takahashi et al. 2000).
Several reports have shown that plants have multiple forms of SPS. In potato, at least four SPS isozymes have been resolved by electrophoresis, and their expression shown to differ with development, tissue type and environmental signals such as low temperature (Reimholz et al. 1997). Three unique SPS genes were cloned from citrus fruit and their expression shown to be tissue specific (Komatsu et al. 1996). In rice, two isoforms of SPS that differ in both their tissue distribution and their kinetic properties have been described (Ingram et al. 1997). More recently additional SPS isoforms were identified (Lutfiyya et al. 2007, Castleden et al. 2004. In maize two isoforms SPS1 and SPS6 appear to be highly specific for leaves. SPS isoforms appear to fall into several groups which occur across phylogenetic groups which may have functional significance and Lutfiyya et al. (2007) analyzed sequence features across groups and concluded that, "the presence of regulatory sites on the proteins appears to be consistent across all members of the group leading to the speculation that they may be diagnostic features and that it might be possible to assign specific functions to entire groups of SPS genes." There are numerous reports of SPS over-expression in transgenic plants, especially in C 3 species. For example, there are reports of the over-expression of maize SPS in tomato (Galtier et al. 1993;Galtier et al. 1993;Micallef et al. 1995;Worrell et al. 1991;Laporte et al. 1997), potato (Ishimaru et al. 2008), canola and sugar beets (Van Asseche et al. 1999), tobacco (Baxter et al. 2001(Baxter et al. , 2003, wheat (Sparks et al. 2001) and Arabidopsis (Signora et al., 1998). A spinach SPS has been over-expressed in cotton (Haigler et al. 2007), and an Arabidopsis SPS has been over-expressed in tobacco (Ishimaru et al. 2008). In general, higher leaf sucrose/starch ratios, and in some cases increased photosynthetic rates have been observed as well as higher tomato and potato solids (Ishimaru et al. 2008) or cotton fiber thickness (Haigler et al. 2007). Overexpression of an Arabidopsis transcription factor (AtbZIP53) increased expression of some SPS transcripts and led to sucrose accumulation in Arabidopsis leaves (Thalor et al 2012). Anur et al., (2020) overexpressed SPS in transgenic sugarcane which like maize is a C4 monocot, and showed that overexpression of SPS increased sucrose, soluble invertase activity as well as glucose and fructose.
Studies have suggested that the overall yield of maize plants may be limited by their source capacity, that is, the amount of leaf carbohydrate produced and/or exported by phloem (Sarquis et al. 1988, Prioul andSchwebel-Dugue 1992). Other studies have suggested the contrary position, that photosynthesis is not limiting yield in major crops (Richards 2000), Maize is a C 4 plant which means that quantum yields of photosynthesis are much higher than they are for most C 3 crops at high temperatures, further complicating the question of whether source capacity is a limiting factor for grain yield in modern maize hybrids (Sage and Monson 1999). To directly test the hypothesis that source productivity limits yield, achieving higher sucrose synthesis and/or transport must first be demonstrated.
Higher plant SPS is highly regulated allosterically and by several different posttranslational modifications including phosphorylation related to dark/light, 14-3-3 binding and osmotic stress (Fig. 1, Huang and Huber 2001). Studies of S158 variants of tobacco leaf SPS including wildtype, S158A, S158T, S158E and S158; analogous to S161 in maize SPS1 demonstrated the importance of light/dark regulatory phosphorylation at this residue.
Here we study the over-expression of truncated maize SPS1 (Lutfiyya et al. 2007, ZmSPSΔ482) targeted to mesophyll cells using two mesophyll-specific promoters. Since transformation with full-length SPS1 failed to produce a phenotype, SPS1 gene was truncated to remove a phosphorylation site involved in light/dark regulation of enzyme activity by a CDPK (Fig. 1, Huang and Huber 2001) which we hoped would lower the chance that the transgenic plant would deregulate it.
The mesophyll specific promoters, pyruvate, orthophosphate dikinase (PPDK) promoter and chlorophyll a/b binding (CAB) protein promoter had previously been shown to target proteins mainly, if not exclusively, to mesophyll cells in maize and show diurnal patterns of expression (Taniguchi Fig . 1 Structure of truncated and full length maize SPS proteins. The figure shows the position of the three main regulatory phosphorylation sites on SPS with Ser162 removed in the truncated and partially deregulated enzyme (Takahashi 2000, ZmSPSΔ49) et al. 2000Bansal et al. 1992). We describe the cloning and construction of the truncated gene, the quantification of over-expressed SPS protein by immunoblot analysis, the screening of R0 events for over-expression, initial greenhouse testing to demonstrate gene efficacy by measuring carbohydrate levels (sucrose and starch) in leaves of F1 inbred events, demonstration of increased SPS protein and enzyme activity in selected transgenic events, demonstration of gene efficacy in field hybrid events, and finally, the results of a yield study with selected transgenic events.

SPS1 gene isolation
7 unique maize SPS gene sequences were identified from internal and public databases in which SPS1 was the major SPS gene expressed in leaves (Lutfiyya et al. 2007). The published sequence of the maize leaf SPS cDNA clone (Gen Bank Acc. No. 401114) was used to design the following synthetic oligonucleotide primers: 5′TAT GTG GTC GAA CTT GCA AGA3′, 5′CCG AAT TCT CAC ATG CCG CTG GAA GTC TTG GAG ACT TGC TTC AG3′, 5′TAT GTG GTC GAA CTT GCA AGA3′, 5′CCG CAT CAC TTC GGC CCA AAT TGG GGG CAT TGA3′.
Three PCR products, spanning the entire length of the SPS cDNA clone, were generated using first-strand DNA, reverse transcribed from maize leaf mRNA. The most 5′ and 3′ PCRproducts were used to screen a maize scutellum library. A nearly full-length cDNA, missing 15 nucleotides at the 5' end was isolated. The 15 nucleotides were added by inserting a synthetic DNA fragment based on the published DNA sequence. The full-length cDNA was expressed in E. coli using both lacZ and tac promoters and in plant cells and in leaf protoplasts using the e35S promoter.

Truncation of maize SPS1
Higher plant SPS enzymes are regulated by reversible phosphorylation on an N-terminal serine residue (Winter and Huber 2000). In maize SPS1 this residue is Ser161 (Fig. 1). To create a maize SPS which not be inactivated at the major site of regulatory phosphorylation (Huang and Winter), a truncated form of the gene was generated, missing the first 5′ 482 nucleotides of the SPS1 coding sequence (SPSΔ482), which translated into a protein missing the first N-terminal 162 amino acids. Synthetic DNA oligonucleotide primers were used in PCRreactions, to produce the truncated protein in Fig. 1. The truncated SPS was then expressed in E. coli. The specific activity of the truncated SPS under limiting conditions (Toroser et al. 1999) in crude bacterial extracts was 2.7-fold greater than that of the full-length SPS (data not shown), suggesting that it had been deregulated.

Maize transformation with ZmSPSΔ482
Two transformation vectors were constructed to target expression to maize leaf mesophyll cells. The vector constructs, herein called Zeke and Pat, employed the CAB and PPDK promoters, respectively. Each construct contained the truncated SPS Δ 482 gene engineered with an ATG start codon at the 5' end. The full length SPS1 gene, from which the ZmSPSΔ482 gene was made, was 93% identical at the DNA level and 100% identical at the protein level to the maize SPS1 gene, which had previously been tested in plants (Van Asseche et al. 1999).
The Zeke construct contained the gene of interest cassette (CAB:Hsp70:ZmSPSΔ482:Nos3′) and a selectable marker cassette (Lox-35 s:Hsp70:NptII:Nos3′-Lox). The Lox sites allow for the eventual removal of the selectable marker, NptII, for commercial events.
In the first step of vector construction, a "shuttle" vector was created. An HpaI fragment containing the CAB promoter and the Hsp70 intron from pMON47112 were joined with an HpaI fragment containing the SPSΔ482 gene and Nos3'. In the final step, a NotI fragment from pMON17704 containing the gene of interest cassette (CAB:Hsp70:ZmSPSΔ482:Nos3′), was joined with a NotI linearized fragment of pMON36176, which contained the NptII selectable marker flanked by Lox sites and the backbone necessary for Agrobacterium transformation.
The Pat construct contained the gene of interest cassette (PPDK:Hsp70:ZmSPSΔ482:Nos3′) and a selectable marker cassette . This vector was created by ligating an HpaI fragment, which contained the PPDK promoter and the 5' portion of the Hsp70 intron, to an HpaI fragment of the Zeke vector, which contained the 3' end of the Hsp70 intron, the SPSΔ482 gene, the Nos3' element, along with the Lox-flanked selectable marker (NPTII) and the backbone necessary for Agrobacterium transformation.
The Pat and Zeke vectors were screened for the correct orientation of the promoter/intron fragments and were subsequently used to stably transform an elite inbred maize variety.

SPS event screening
Gene insert copy number was determined by NPTII TaqMan assay at R0. Only events with 1 or 2 insert copies were advanced.
Selections of advanced events in which zygosity had previously been determined (e.g., homozygous positive or negative parents) were routinely screened by gene specific PCR. DNA from leaf tips of individual plants was extracted using a Sigma DNA isolation kit. After DNA extraction, the presence of the SPSΔ482 gene was determined by PCR using an Hsp70 intron forward primer (hsp 727-5′) and a SPSΔ482 reverse primer (11,710-8528-3′) which yielded a 1.1 kb fragment. DNA integrity was confirmed by amplifying the endogenous ADH gene.
Additional event selections were screened by an NPTII TaqMan assay, primarily to determine segregation ratios and zygosity. At several points during event selection, the gene-specific PCR assays were confirmed by SPS proteinimmunoblot analysis.

SPS protein extraction
All greenhouse and field efficacy plants were screened for expression by immunoblot (western) analysis. Typically, a small leaf portion was removed from a V4 to V10 plant, frozen on dry ice and stored at − 80 °C. The frozen leaf sample was ground to a coarse frozen powder in an Eppendorf tube on dry ice and extracted into 300 µL of PBST containing 1% (v/v) Protease Inhibitor Cocktail for plant cell and tissue extracts (Sigma Catalog No. P 9599). After thawing on wet ice, tubes containing extract were vortexed and centrifuged at 14,000 rpm for 4 min. An 80 µL portion of the supernatant was transferred to a tube containing 20 µL of 6X SDSSample Buffer (Laemmli 1970) and boiled for 10 min. A second 50 µL portion of supernatant was used for protein concentration by Bradford analysis, using Sigma bovine serum albumin as a protein standard.

Protein immunoblot analysis
Denatured extract samples (16 µg total protein) in SDS Sample Buffer were loaded onto 10% SDS-PAGE gels (Bio-Rad Catalog No. 161-1119). The gels were electrophoresed in Tris-Glycine-SDS Running Buffer (25 mM Tris, 192 mM glycine, 0.1% SDS, pH 8.3) for 60 min at 100 V plus 30 min at 150 V. Gel proteins were transferred to Immunoblot PVDF membranes (Bio-Rad Catalog No. 162-0177) in Transfer Buffer (25 mM Tris, 192 mM glycine, 20% MeOH) for 100 min at 250 mAmps. Membranes were blocked overnight in Tris Buffered Saline with Tween (TBST) + 5% milk at 4 °C. Blocked membranes were washed 3X with TBST and incubated for 90 min at room temperature in TBST + 5% milk containing SPS primary antibodies from rabbit . After incubation with primary antibody, membranes were washed 3X with TBST and incubated for 60 min at RT in TBST containing anti-rabbit secondary antibody (1:10,000; Anti-Rabbit IgG-Alkaline Phosphatase, Sigma). Membranes were washed 3X in TBST and incubated in AP substrate (SigmaFast BCIP/NBT, Sigma Catalog No. B 5655) until protein bands were stained purple.

SPS activity assay
Leaf samples were crushed to a frozen powder with liquid N 2 . 1.2 mL of SPS Extraction Buffer (50 mM Tris-HCl, pH 7.5, 400 mM sucrose, 10% (v/v) glycerol, 1 mM phenylmethane sulfonyl fluoride 5 mM dithiothreiotol, 1 mM trans-epoxysuccinyl-L-leucylamido-(4-guanidino) butane (E64), 10 µg/mL leupeptin, and 1% (v/v) Plant Protease Inhibitor Cocktail (Sigma Catalog No. P 9599) was added to the sample. The buffer was ground into the mixture with care taken to make sure that the mixture did not thaw. More liquid N 2 was added to completely homogenize the powder. A portion of this homogenized mixture was then added to a 1.5 mL microcentrifuge tube on dry ice. The tube was removed from the dry ice and allowed to thaw. Immediately upon thawing the tube was centrifuged at 14,000 rpm in a microcentrifuge for 4 min at 4 °C.
Immediately prior to assay, a portion of the supernatant was desalted on a G-25 Spin Column (Quick Spin Protein Columns, Roche Diagnostics Corp.), which had been previously equilibrated with SPS Extraction Buffer.
To assay SPS activity, 65 µL of desalted extract was added to 65 μL of assay buffer containing 50 mM HEPES (pH 7.5), 10 mM MgCl 2 , 10 mM UDP-glucose, 10 mM fructose 6-phosphate, 36 mM glucose 6-phosphate and 10 mM NaF. Substrate-negative controls were run with the same buffer without UDP glucose, glucose 6-phosphate or fructose 6-phosphate. The mixture was vortexed and incubated at 25 °C for 15-30 min. The reaction was terminated by the addition of 65 µL of 30% (w/v) KOH and the remaining fructose 6-phosphate was destroyed by boiling the sample at 100 °C for 20 min. This treatment also converts sucrose-6 phosphate to sucrose which can then be detected by the anthrone method (Yemm and Willis 1954). To detect the product sucrose, 800 µL of 0.3% anthrone in 27N H 2 SO 4 was added to the sample, which was then vortexed, centrifuged at 14,000 rpm for 2 min and then incubated at 37 °C for 20 min. The absorbance was read at 625 nm using a spectraphotometer. Activity was calculated by using a sucrose standard curve. One unit of enzyme is the amount which can produce 1 µM sucrose/min at 25 °C.

Greenhouse efficacy tests of inbred F1 events
R1 seed was generated by crossing R0 maize plants with themselves. F1 seed was generated from R0 events by backcrossing expressing (and co-suppressed) plants to wild-type inbred plants. Since the R0 plant was hemizygous, F1 individuals of single copy R0 parents would be expected to be segregating 1:1. This process removed negative phenotypic effects due to regeneration and created positive and negative isolines to conduct efficacy experiments. Events in which F1 individuals had aberrant segregation were eliminated.
Three separate greenhouse experiments were conducted in order to test for the effect of the expression of SPS. Each experiment was similarly designed in a similar manner. The greenhouse was programmed for 14 h of daylight at 30 °C and 10 h dark at 24 °C, 60% humidity. To minimize temperature and humidity gradients in the greenhouse wildtype and transgenic plants were always planted next to each other. Since the leaf levels of both starch and sucrose show a diurnal pattern (low levels early in the day accumulating to high levels late in the day), leaves were harvested at several time points. SPS expression in each individual plant was determined by protein immunoblot analysis blot at V4. Positive and negative plants were then randomized within an event block. Each block was bordered by a row of wild type plants and the aisle or outside rows were bordered by 2 rows of plants to eliminate border effects. Since comparisons are always made against the negative isoline control we have elected to combine the three experiment together for analysis; however, they are described individually below.

Leaf carbohydrate analysis of PPDK:ZmSPSΔ482 (Pat) events
A first experiment was conducted in a greenhouse with events selected to have moderate to high expression of SPSΔ482 (Table SI and S2). These included two moderate expressing events (Pat87 and Pat94), 3 high expressing events (Pat47, Pat63, and Pat81) and 1 co-suppressed event (Pat20), along with a block of control plants. On the day of leaf harvest, light intensity peaked at approximately 950 µE.
At V6, the upper most-recently expanded leaf was removed from the plant, folded into an envelope and immediately frozen on dry ice. Leaf samples were stored at − 80 °C until extraction for carbohydrates. Leaves were harvested at 10 a.m., 1 p.m., 6 p.m., and 11 p.m. A single leaf was sampled from each of four separate plants within an event block. Each plant was sampled only once.

Leaf carbohydrate analysis of CAB:ZmSPSΔ482 (Zeke) events
A second experiment was conducted in a greenhouse with Zeke (CAB:SPSΔ482) events which were selected in order to test a range of expression of SPS, as measured at R0 (Table S1 and S2). These included a low expressing event (Zeke122), 2 moderate expressing events (Zeke83 and Zeke113), 2 high expressing events (Zeke9 and Zeke108) and 1 co-suppressed event (Zeke10). At V6, leaves were harvested for carbohydrate analysis at 4 time points: 10 a.m., 1 p.m., 4 p.m., and 6 p.m. Daytime light intensities reached approximately 600 µE. There were up to 10 replicate leaves harvested for each treatment and each plant was sampled only once. Leaves were stored at − 80 °C until extraction for carbohydrates.
Leaf carbohydrate analysis of combined study of PPDK:ZmSPSΔ482 (Pat) and CAB:ZmSPSΔ482 (Zeke) events A final greenhouse efficacy experiment was conducted in mid spring with a combination of F1 PPDK:ZmSPSΔ482 (Pat) events and F1 CAB:SPSΔ482 (Zeke) events. As in the two previous greenhouse experiments, events were chosen so that a range of expression levels, based on R0 expression levels, could be compared. Gene expression was determined by PCRand events were blocked, randomizing positive and negative plants within each event block. There were 2 low expressing Pat events (Pat87 and Pat95), 2 moderate expressing Pat events (Pat18 and Pat89), and 1 very high expressing Pat event (Pat47). There were 3 low expressing Zeke events (Zeke17, Zeke112, and Zeke113), 1 moderate expressing Zeke event (Zeke64), and 1 high expressing Zeke event (Zeke69). A very high expressing Zeke event, Zeke11 turned out to have no plants expressing SPS. There were no plants for the co-suppressed event, Zeke58. Zeke113 was subsequently determined to be segregating.
Because of the limited number of plants, the numbers and time points for sampling was variable. Up to 10 plants were sampled at 1:30 p.m., 4 p.m., and 6 p.m. Plants were at V8 to V9 at the time of sampling. Greenhouse light intensities approached 1100 E during the middle of the day. The upper most recently expanded leaf was harvested into a paper envelope and frozen on dry ice and stored at − 80 °C until extraction for carbohydrates.

Field efficacy test of pat and Zeke hybrid events
Several field tests were established to measure the efficacy of SPS overexpression in hybrid maize leaves. One test was a comparison of positive and negative isolines of Pat and Zeke events. Positive and negative isoline plants were planted side-by-side in 2-row plots. This test was replicated in the field twice. A similarly designed test, replicated once, compared positive and negative isoline plants of Zeke cosuppressed events.
In total there were 2 co-suppressed Zeke events (Zeke 10 and Zeke 58), 6 expressing Pat events (Pat18, Pat 66, Pat85, Pat87, Pat89, and Pat95), and 6 expressing Zeke events (Zeke11, Zeke17, Zeke19, Zeke64, Zeke69, and Zeke112). Zeke112 was represented twice in the Day 1 Harvest plots. Leaves were harvested on 2 consecutive days designated Day 1 and Day 2. On Day 1, leaves were harvested at 9 a.m., 1 p.m., 3 p.m., and 5 p.m. On Day 2, leaves were harvested at 1 p.m., 6 p.m., and 7 p.m. Light intensities approached at least 1700 E at midday on both Day 1 and Day 2. Plants were approximately at V8 at harvest. After harvest, leaves were frozen on dry ice and stored at − 80 °C until extraction for carbohydrates. The results from Day 1 and Day 2 harvests are similar but for the sake of brevity we will only show analyzed data from the Day 2 harvest in this study.

Yield test and yield components test of pat and Zeke hybrid events
A yield component field trial was performed to compare positive and negative hybrids of 6 expressing Pat and 6 expressing Zeke events. The tests were planted at 6 locations and at two densities (25 and 37 plants/row).

Hybrid leaf carbohydrate analysis
Whole frozen leaves were ground to a powder by crushing and grinding on stainless-steel trays kept on dry ice using a rolling pin. Liquid N 2 was used to keep the tissue brittle. Large veins were discarded. For the greenhouse efficacy experiments, frozen leaf powder was extracted for carbohydrate analysis and powder was stored at -80 °C until extraction. For the field efficacy experiment, leaf powder was lyophilized, and dry leaf powder was extracted for carbohydrate analysis.
Approximately 50 mg of frozen leaf powder, or approximately 20 mg of dried leaf powder, was transferred to a tared, 1.4 mL Falcon tube and weighed.
Leaf powder was extracted into warm 80% (v/v) ethanol with three successive extractions which extracted over 95% of the soluble sugars. First, 800 µL of 80% ethanol was dispensed to extraction tubes containing powder samples using a multi-channel pipettor. The tubes were sealed with a Beckman cap mat and incubated at 55 °C for 15 min, shaking the rack two-or three-times during incubation. The plates were centrifuged at 1800 rpm and 500 µL of the supernatant was transferred to a Nalgene 96-well plate with 2 mL wells. Then 500 µL of 80% ethanol was added to the extraction tubes and the extraction steps were repeated, removing 500 µL supernatant to the previous extracts. A final 300 µL of 80% ethanol was added to all tubes, repeating the extraction steps, and transferring a final 400 µL of supernatant to the Nalgene plate, resulting in a final extraction volume of 1.4 mL.
The green ethanolic extract was clarified by adding 30 mg of activated charcoal to each well. The plate wells were sealed with a cap mat and the plate was inverted several times over 15 min, returning the plate to wet ice. The plates were centrifuged at 1800 rpm and 800 µL of clear supernatant was transferred to a 96-well strip tube plate. The supernatant was dried in a SpeedVac and resuspended in 200 µL MQ-H 2 O and stored at − 20 °C for sugar analysis.
For starch analysis, the extract pellet was washed 4 times with absolute ethanol to completely remove pigments and oven dried at 65 °C for several days. Starch in the pellet was gelatinized by adding 400 µL MQ-H 2 O to the tubes, sealing with a cap mat, and boiling in a water bath for 15 min using a "rig" designed to keep the tube contents from boiling out of the tube. After boiling, the tubes were cooled on wet ice.
The gelatinized starch was enzymatically digested to glucose. To digest the starch, 500 μL of 100 mM acetate buffer, pH 4.5, containing 5 U/500 µL amyloglucosidase (Sigma Catalog No. A 7420) was added to the tubes using a multichannel pipettor and the tubes were sealed with a cap mat and incubated for 4 h. at 55 °C. The tubes were inverted several times during incubation. After 4 h., the tubes were boiled for 10 min and stored at − 20 °C until assaying for glucose.
Extract sugars, i.e., glucose, fructose and sucrose, and starch-glucose, were measured enzymatically in microtiter plates by coupling substrate conversion with the reduction of NADP and measuring absorbance increase at 340 nm. Each assay plate contained appropriate standards and extract sugar concentrations were derived from the linear curve(s) of that plate.
For the glucose/fructose assay , 20 µL of aqueous resuspended extract was assayed in a final well volume of 300 µL containing 100 mM Triethanolamine buffer, pH 7.6, 10 mM MgCl 2 , 1.1 mM ATP, and 0.8 mM NADP. After reading the background absorbance at 340 nm using a microtiter plate reader, 5 µL of an enzyme mix containing glucose 6-phosphate dehydrogenase (0.2U/well) and hexokinase (0.4 U/well) was added, and the OD was read at 340 nm after 10 min. The OD change was due to glucose. Then.phosphoglucoisomerase (1U/well) was added and the OD change was read after 15 min. The OD change was due to fructose.
For the sucrose assay , 10 µL of aqueous resuspended extracts were transferred to microtiter plate wells containing 100 µL of 100 mM acetate, pH 4.65 containing invertase (1U/well) to invert sucrose to glucose (plus fructose). The resulting glucose was measured by adding buffer, substrates, and enzymes as described above.
For the assay of glucose in the starch digest, 10 µL of digest was transferred to microtiter plate wells and assayed as described above. Glucose, fructose, and sucrose were expressed as µmol/gFW (or µmol/gDW), and starch was expressed as glucose-equivalents, µmol/gFW (or µmol/ gDW).

Variable levels of higher expression of SPSΔ482-SPS in PPDK:SPSΔ482 and CAB:SPSΔ482 maize events
A total of 85 PPDK:SPSΔ482 (Pat) events and 105 CAB:SPSΔ482 (Zeke) events were tested for expression at R0 by immunoblot analysis. Based on the intensity of the 101 kDa SPSΔ482 band, event expression levels were comparatively assigned as very low, low, moderate, high, very high, or negative. Co-suppressed events were also identified in which both the endogenous upper band and transgenic lower band were reduced or absent. Figure 2 depicts protein immunoblots for 12 of the events followed in this study. They include events over the entire range of expression. There were no apparent differences in levels of expression between PPDK:ZmSPS1Δ482 and CAB:ZmSPS1Δ482 events. Events were advanced for multiple reasons including seed return and expression. Tables S1 and S2 summarize the expression of all Pat and Zeke events generated. In 2 Pat events and 6 Zeke events both the endogenous SPS and SPSΔ482 band were reduced or missing, indicating that these events were co-suppressed in these events. One of these events (Zeke 10) is included in this study. No visual phenotypes were observed in plants expressing SPSΔ482.

Higher activity of SPS in leaves of PPKK:SPSΔ482 and CAB:SPSΔ482 maize events
Endogenous SPS activity in leaves is known to vary diurnally (Kerr et al. 1985, andKalt-Torres et al. 1987), due to both circadian and light/dark regulation (Winter and Huber 2000). Total SPS activity was measured in leaf tissue collected diurnally from two expressing PPDK:SPSΔ482 events and one co-suppressed PPDK:SPSΔ482 event and wild type SPS. Activity in both over-expressing events was significantly higher than wild-type and showed a similar diurnal pattern (Fig. 3.). Activity levels were approximately twofold higher at 1 p.m. and 6 p.m. SPS activity of the co-suppressed Activity is expressed as total sucrose produced/30 min/mg total protein event was reduced by approximately 85 to 90% and did not show a diurnal pattern although it did increase between 10 a.m. and 12 p.m.
Although the SPSΔ482 protein was not expected to undergo diurnal deregulation, the diurnal expression pattern of total SPS activity in the transgenic maize plants might be due to the upregulation of the PPDK and CAB during the day and/or a carryover of the diurnal expression of the native SPS protein with the transgenic over the entire diurnal cycle.
SPS activity in leaves harvested at 1 p.m. of wild-type and additional SPS events is shown in Table 1. Activity was increased by approximately two-to three-fold. Since advancements had already been made using expression from protein immunoblots (Fig. 2), only one of these events (Pat 18) ended up advanced for further study.

The expression of truncated SPS modulates leaf carbohydrate levels in inbred maize
In three separate experiments sucrose and starch levels were measured in inbred greenhouse grown transgenic maize and their isoline controls. The sucrose and starch data is summarized in Table 2. Sucrose and starch generally followed a diurnal pattern which peaked at 6 p.m.. In all events (except the co-suppressed event) and at all time points, sucrose trended higher in the transgenic plants compared to their isoline controls. In most events and at most timepoints (except the co-suppressed events) starch trended lower in transgenic maize compared to their isoline control. In 6 cases (event x timepoint) sucrose was significantly higher in transgenic plants and in 11 cases starch was significantly lower that the isoline control. In no cases was starch significantly higher than the isoline control.
In the co-suppressed event, Zeke 10, the opposite pattern was observed. Sucrose trended or was significantly lower in transgenic plants than their isoline control and starch trended or was significantly higher in transgenic maize than the isoline control. Figure 4 shows a comparison of sucrose and starch values of leaves in positive and negative isolines of Pat and Zeke hybrid events grown in the field.

The expression of truncated SPS modulates leaf carbohydrate levels in hybrid field grown maize
In Zeke over-expressing events (Fig. 4a), a strong trend toward higher starch is observed across most cases. Overall, 14 out of 18 cases (a case = an event x a time point) trended positive with 9 cases being significantly positive. A strong trend toward higher sucrose is also observed with 17 cases trending positive with 5 that were significantly positive. Interestingly, there were three cases (Zeke11 at 6 p.m., Zeke19, 1 p.m., Zeke112, 6 p.m.) in which both sucrose and starch was significantly higher than the isoline control.
In Pat events (Fig. 4b), starch was mixed, and sucrose trended positive in all Pat cases and was significantly positive in 3 cases.
The two co-suppressed events (Zeke10 and Zeke58) tested in the field showed a different pattern (Fig. 4c). Starch levels were trended or were significantly higher in transgenic plants and sucrose levels were generally lower.

The expression of truncated SPS does not change seed yield in field grown hybrid maize
Six Pat and six Zeke events were tested for yield. Positive and negative hybrid isolines from each event were planted at two densities (25 and 37 plants/row) at six field locations. None of the events had a significant difference in yield between the positive and negative isoline (Table 3). In addition, the overall yield for all Pat and Zeke events was almost identical in the positive and negative isolines.

Discussion
We have successfully over-expressed a truncated maize SPS gene (ZmSPSΔ482) in maize. We have shown that leaves from transgenic plants over-expressing SPS have increased SPS activity.
Starting with a hypothesis that improving source capacity could increase maize yield, we devised a method to successfully over-express a gene (SPS) by targeting expression to a specific cell type (mesophyll leaf cells of maize). We then screened events based first on copy number and RNA expression. Next, we tested the activity and protein levels of the transgenic protein in the tissue and classified overall expression levels as co-suppressed, none, low, medium and high. We then tested the metabolic efficacy of the gene by determining the steady-state sucrose and starch levels in the transgenic plants, first using inbred plants in the greenhouse, and finally using hybrid plants in the field. Based  Blue, no significant difference in sucrose or starch level between positive and negative isoline; green, positive isoline has sta-tistically higher sucrose or starch than negative line (p < 0.5); red, positive isoline has statistically lower sucrose or starch than positive isoline. Units are in µmol/gram dry weight. a Pat non-co-suppressed lines. b Zeke non-co-suppressed lines. c Co-suppressed lines on protein expression and inbred sucrose levels we picked several events and tested them in a multi-location yield trial to determine yield efficacy.
Since the SPS protein used in this study has many regulatory features (Lutfiyya et al. 2007), it is possible that a higher plant SPS might be necessary to improve source capacity in higher plants. Alternatively, isozymes from different phylogenetic groups (Lutfiyya et al. 2007) might have different yield effects. More events will need to be tested to determine the correlations between SPS expression level and yield as well as sucrose/starch levels and yield. Also, it is possible that higher levels of sucrose actually inhibit certain processes and it may be necessary to stack SPS with a leaf sucrose transporter to ensure that the additional sucrose is transported out of the leaf to the sink tissues. Finally, other SPS stacks with other genes in the same or different metabolic pathways in the same or different tissues, including sucrose phosphorylase (Narimatsu et al 2004), fructose 1 6-bisphosphate aldolase, sucrose phosphate phosphatase, glutamine synthetase, and glucose dehydrogenase could be considered to test various hypotheses.
Earlier attempts to over-express SPS in maize using an e35S promoter had been unsuccessful. Leaf tissue expression experiments showed that the e35S promoter targets expression to bundle sheath cells. In C4 plants having Kranz anatomy such as maize, CO2 carboxylation and chloroplast starch accumulation are compartmentalized in bundle sheath cells, while CO2 reduction by Rubisco and sucrose biosynthesis are compartmentalized in mesophyll cells. In this study, two different promoters were employed to target over-expression of SPS specifically to leaf mesophyll cells, the PPDK promoter and the CAB promoter. These two promoters also drive diurnal expression (Taniguchi et al. 2000;Bansal et al. 1992). The results of this study indicate that targeting of SPS to leaf mesophyll cells is necessary for expression in maize.
In this experiment the expression of transgenic SPS protein in maize leaves varied from very low to very high, based on immunoblot intensity. We also identified individual transgenic events in which both endogenous and transgenic SPS proteins were co-suppressed. The effect of transgenic SPS expression was measured by comparing sucrose and starch levels in leaves of positive and negative isoline plants grown side by side in three separate greenhouse studies ( Table 2).
The sucrose data from the greenhouse tests was strikingly consistent. Every event overexpressing SPS had trended to higher sucrose than the control at every time point. Many of these were significant and more were almost significant. The one exception which proves the rule is Zeke10 (cosuppressed) which trended lower or was significantly lower at every time point. The starch data, almost as strikingly, had the opposite trend. Most events at most time points had lower starch. Zeke10 (co-suppressed) showed exactly the opposite trend to higher starch and lower sucrose at every time point. The inverse correlation trends between sucrose and starch across transgenic vs. isoline and across higher expression vs. co-suppression suggests there may be a limitation either due to the light or in inbred maize compared to hybrid maize (Bellasio and Griffiths 2014). Although we did not measure photosynthesis directly, it is likely that photosynthesis was not inhibited but rather a reportioning of carbon between sucrose and starch similar to the relationship between them in endosperm (Salerno 1986). This is significant since there are several reports that sucrose accumulation inhibits photosynthesis [Stitt et al. 2010, McCormick et al. 2009, Inman-Bamber et al. 2011, and exogenous sucrose supply strongly reduces the net CO 2 assimilation in sugarcane (Lobo et al 2015). Overall, there appears to be room to increase source capacity in inbred maize.
It is noteworthy that the expression of SPS did not change the overall diurnal pattern of sucrose/starch leaf accumulation nor was sucrose accumulation in SPS (over)expressors correlated with expression based on western. This suggests that SPS may be near limiting in maize leaves or alternatively, higher levels of sucrose synthesis may be inhibitory. In either case, improvement of source capacity will require fine-tuning of SPS expression.
In field-grown hybrid maize, expression of SPS clearly increases leaf sucrose levels, although the expression of PPDK-SPS had no clear effect on leaf starch levels (possibly a slight decrease). In these PPDK-SPS plants there may be a slight inverse trend between sucrose and starch level. CAB-SPS plants clearly have significant increases in starch and sucrose often simultaneously. In these events SPS is clearly enhancing source capacity in field-grown hybrid maize. The difference between PPDK-SPS and CAB-SPS efficacy further suggests that fine-tuning of the level and location of SPS expression will be necessary to maximize source capacity. Overall, these results further support the conclusion that SPS is a pinch point in leaf metabolism.
Results from other transgenic experiments in maize in which another sucrose synthesis pathway enzyme (fructose 1,6-bisphosphate aldolase, FDA) did not show a consistent sucrose phenotype (data not shown) also reinforce the conclusion since only the rate-limiting enzyme would be expected to have a large and consistent effect on sucrose.
Yield improvements in the beginning of the last century have come about by increases in kernel number per hectare (Egli, 2015(Egli, , 2019 requiring higher plant populations and through a concomitant increase in both source and sink strength suggesting that increasing source capacity alone with no increase yield. Although the activity of SPS can satisfy demand for carbon assimilate (Prioul and Schwebel-Dugue 1992), cause and effect has not been established with the possibility that sink and source work together to determine yield (Egli 2015). Since there was no increase in seed yield it might be expected that vegetative mass would be higher in the transgenic plants; however, we could not confirm this since we did not measure biomass.
Therefore, it is perhaps not surprising that we did not see an effect on yield. In the multi-location study of six PPDK:SPSΔ482 events and six CAB:SPSΔ482events, there was no significant effect on yield (p > 0.05) in any event at either planting density. Only one event, Pat 95, had a yield increase of 625 kg/hectare over both densities which was nearly significant (p = 0.066). Since maize source and sink are thought to be tightly controlled (Seebauer et al. 2010) this result might be expected and suggests that to increase maize yield both source and sink (and perhaps other factors) will have to be considered.