Glucan, water dikinase activity stimulates breakdown of starch granules by plastidial beta-amylases

Abstract

Glucan phosphorylating enzymes are required for normal mobilization of starch in leaves of Arabidopsis (Arabidopsis thaliana) and potato (Solanum tuberosum), but mechanisms underlying this dependency are unknown. Using two different activity assays, we aimed to identify starch degrading enzymes from Arabidopsis, whose activity is affected by glucan phosphorylation. Breakdown of granular starch by a protein fraction purified from leaf extracts increased approximately 2-fold if the granules were simultaneously phosphorylated by recombinant potato glucan, water dikinase (GWD). Using matrix-assisted laser-desorption ionization mass spectrometry several putative starch-related enzymes were identified in this fraction, among them beta-AMYLASE1 (BAM1; At3g23920) and ISOAMYLASE3 (ISA3; At4g09020). Experiments using purified recombinant enzymes showed that BAM1 activity with granules similarly increased under conditions of simultaneous starch phosphorylation. Purified recombinant potato ISA3 (StISA3) did not attack the granular starch significantly with or without glucan phosphorylation. However, starch breakdown by a mixture of BAM1 and StISA3 was 2 times higher than that by BAM1 alone and was further enhanced in the presence of GWD and ATP. Similar to BAM1, maltose release from granular starch by purified recombinant BAM3 (At4g17090), another plastid-localized beta-amylase isoform, increased 2- to 3-fold if the granules were simultaneously phosphorylated by GWD. BAM activity in turn strongly stimulated the GWD-catalyzed phosphorylation. The interdependence between the activities of GWD and BAMs offers an explanation for the severe starch excess phenotype of GWD-deficient mutants.

Figure 1.

Breakdown of granular starch by Arabidopsis proteins is stimulated by simultaneous glucan phosphorylation. A, Starch degrading activity of the protein fraction obtained from Arabidopsis leaves using four purification steps. Sex1-3 starch granules (2.5 mg) were incubated with 60 μL fraction 18 (MonoQ) or 60 μL buffer (control) with (black bars) or without (white bars) 0.5 mm ATP in the presence of 1.6 μg recombinant potato GWD. Final volume of the assay: 120 μL. Following incubation at 25°C for 90 min the starch was sedimented by centrifugation. Starch breakdown products present in the supernatant were hydrolyzed with acid and subsequently Glc was quantified. The slightly increased Glc content in the ATP-containing control was not consistently observed in independent experiments. B, SDS-PAGE of fraction 18 and protein identification using MALDI-MS. Proteins were separated by SDS-PAGE (10% acrylamide in the separation gel). Following staining with Coomassie Blue bands were excized, digested with trypsin, and the peptides were analyzed by MALDI-MS and database search. The following proteins were identified: 1, 2, 3 = SBE3 (At2g36390); 4 = ISA3 (At4g09020) + SBE3; 5 = putative Phosphatase (At3g01510); 6 = unknown protein (At3g55760); 7 = BAM1 (At3g23920); 8 = DPE1 (At5g64860). Proteins ≤55 kD were also analyzed but were not related to carbohydrate metabolism.

Figure 2.

Glucan phosphorylation by GWD stimulates degradation of granular starch by BAM1. Sex1-3 starch granules were incubated in buffer with (black bars) or without (white bars) 0.25 mm ATP for 90 min. The following amounts of recombinant enzymes were added: BAM1 (7.5 μg, GST-fusion protein), StGWD (4.5 μg), StISA3 (0.45 μg). A, The starch degradation products released into the soluble phase were hydrolyzed with acid and Glc was quantified. The Glc present in control samples that lacked any recombinant protein was subtracted from all other samples. B, HPAEC-PAD analyses of the starch degradation products without acid hydrolysis. All samples contained ATP and the following recombinant proteins: GST-BAM1 (1), GST-BAM1 and StGWD (2), GST-BAM1 and StISA3 (3), GST-BAM1, StISA3, and StGWD (4). The peak labeled with an asterisk (*) comprises HEPES. Maltose (G2) or maltotriose (G3) were not detectable in samples containing either no recombinant protein or only StISA3 and/or StGWD (data not shown).

Figure 3.

Release of radioactive products from ³³P-labeled starch by BAM1 and/or StISA3. ³³P-labeled starch granules (equivalent to 50,000 cpm) were incubated with 4 μg GST-BAM1, 0.75 μg StISA3, or a mixture of both for 60 min. The release of label into the soluble phase was quantified. In a control lacking any recombinant protein 230 cpm were present in the supernatant. This value was subtracted from all other samples. The experiment was repeated under similar conditions with essentially the same result.

Figure 4.

Glucan phosphorylation by StGWD stimulates degradation of granular starch by BAM3. Sex1-3 starch granules were incubated in buffer with 0.25 mm ATP for 45 min. The following amounts of recombinant enzymes were added: GST-BAM3 (2 μg), StGWD (3 μg), AtPWD (3 μg), and StISA3 (0.45 μg). The starch degradation products released into the soluble phase were hydrolyzed with acid and Glc was quanitified. The Glc present in control samples that lacked any recombinant protein was subtracted from all other samples. Increased starch breakdown in the presence of StGWD was ATP dependent and was also observed if BAM3 devoid of the GST tag was used instead of GST-BAM3 (data not shown).

Figure 5.

An inactive Arabidopsis mutant GWD cannot stimulate β-amylolysis of granular starch. Sex1-3 starch granules were incubated in buffer containing 0.25 mm ATP for 60 min. The following amounts of recombinant enzymes were added: GST-BAM3 (8 μg), AtGWD (3 μg), and AtGWD(H1004A) (3 μg).

Figure 6.

β-Amylolytic attack leads to increased phosphorylation of granular starch by GWD. sex1-3 starch granules (4 mg) were phosphorylated with 0.5 μg StGWD and 50 μm ATP containing 1 μCi [β-³³P]ATP in a final volume of 0.4 mL in the presence (black circles) or absence (white circles) of 1 μg GST-BAM3 for the times indicated. In another sample 5 mg starch was pretreated with 1 μg GST-BAM3 for 40 min. GST-BAM3 was then removed by washing the starch in 2% SDS and water. Subsequently, 4 mg starch was phosphorylated with GWD only (squares).

Figure 7.

Hypothetical interplay of GWD, BAM, and ISA3 during breakdown of starch granules. A, Densely packed amylopectin (double) helices. Six glucosyl residues are required for a full helix turn. Chains as depicted here have a degree of polymerization of approximately 15. The exact lengths were arbitrarily chosen but they are in a realistic order of magnitude for chains within the crystalline lamellae of starch (Smith, 2001). Gray double helices represent neighboring clusters of glucan chains. B, A single chain is degraded by BAM. The space thereby provided allows GWD attack on a neighboring double helix. C, GWD unwinds a double helix and phosphorylates one strand. The phosphate residue (red dot) stabilizes the open chain conformation. D, BAM attacks the single chains but cannot go beyond the phosphorylated residue and the α-1,6 linkages. E, ISA3 releases the remaining stubs and also phosphodextrins. GWD can attack another double helix and a new cycle of phosphorylation and degradation starts.