The heteromultimeric debranching enzyme involved in starch synthesis in Arabidopsis requires both isoamylase1 and isoamylase2 subunits for complex stability and activity

Abstract

Isoamylase-type debranching enzymes (ISAs) play an important role in determining starch structure. Amylopectin - a branched polymer of glucose - is the major component of starch granules and its architecture underlies the semi-crystalline nature of starch. Mutants of several species lacking the ISA1-subclass of isoamylase are impaired in amylopectin synthesis. Consequently, starch levels are decreased and an aberrant soluble glucan (phytoglycogen) with altered branch lengths and branching pattern accumulates. Here we use TAP (tandem affinity purification) tagging to provide direct evidence in Arabidopsis that ISA1 interacts with its homolog ISA2. No evidence for interaction with other starch biosynthetic enzymes was found. Analysis of the single mutants shows that each protein is destabilised in the absence of the other. Co-expression of both ISA1 and ISA2 Escherichia coli allowed the formation of the active recombinant enzyme and we show using site-directed mutagenesis that ISA1 is the catalytic subunit. The presence of the active isoamylase alters glycogen biosynthesis in E. coli, resulting in colonies that stain more starch-like with iodine. However, analysis of the glucans reveals that rather than producing an amylopectin like substance, cells expressing the active isoamylase still accumulate small amounts of glycogen together with a population of linear oligosaccharides that stain strongly with iodine. We conclude that for isoamylase to promote amylopectin synthesis it needs to act on a specific precursor (pre-amylopectin) generated by the combined actions of plant starch synthase and branching enzyme isoforms and when presented with an unsuitable substrate (i.e. E. coli glycogen) it simply degrades it.

Figure 1. ISA1 and ISA2 co-elutes in a high molecular weight complex.

A. Gel filtration chromatography of extracts of wild-type Arabidopsis leaves. Fractions from a Sephacryl HiLoad 200 prep grade column were collected, concentrated and separated by SDS-PAGE. ISA1 and ISA2 were detected by immunoblot analysis probed with specific antibodies. For comparison crude extracts of the wild type and the isa1isa2 double mutants are in the left two lanes. Molecular masses based on the elution positions of known standards are indicated. B. RNA was extracted from 20-day old plants harvested at the end of day, reverse-transcribed to cDNA and transcript levels were analysed by real-time PCR. The expression of ISA1 and ISA2 in wild-type and the different mutants are shown relative to that of the housekeeping gene PP2A. C. Soluble proteins in extracts of wild-type and isa mutant Arabidopsis leaves were analysed by SDS-PAGE followed by immunoblotting with specific antibodies. 50 µg of total protein was loaded into each lane.

Figure 2. TAP-tagging ISA2 restores the isoamylase activity.

A. Schematic overview of the structure of the TAP-tag fused to the ISA2 coding sequence (not to scale). B. The ISA2-TAP-expressing plants were analysed for isoamylase activity on amylopectin-containing native gels. C. Immunoblot analysis of SDS-PAGE gels, probed with antiserum raised against and ISA2-specific peptide. The TAP-tagged ISA2 is 20 kDa larger than native ISA2, corresponding to the predicted molecular weight of the tag.

Figure 3. TAP-tagged ISA2 complements the isa2-1 phenotype.

Leaves from individual plants of wild-type, isa2-1 mutant and ISA2-TAP lines were harvested at the end of the day and immediately frozen in liquid N₂. The insoluble and soluble glucans were extracted using perchloric acid. A. Starch and phytoglycogen accumulation was measured after enzymatic hydrolysis to glucose. Each value is the mean ± SE of five biological replicates. B. The chain-length distribution (CLD) of insoluble glucans was analysed with HPEAC-PAD. Peak areas were summed and the areas of individual peaks were calculated as a percentage of the total. Equal amounts of glucan from the insoluble material from biological replicates in (A) were pooled. The means ± SE of four technical replicate digests are shown. C. Difference plot derived by subtracting the relative percentage values of wild-type (WT) amylopectin from the ISA2-TAP line and isa2-1 mutant.

Figure 4. ISA1 co-purifies with ISA2-TAP.

A. Amylopectin-containing native gels revealing the isoamylase activity in eluates from the affinity resins during tandem-affinity purification of ISA2-TAP and the wild type (WT). The first affinity step is designated ‘TEV-eluate’, and the second ‘Elution Calmodulin’ (see materials and methods).B. Proteins eluting from the calmodulin affinity resin during purification of ISA2-TAP and wild-type were concentrated, separated on SDS-PAGE and visualised by silver staining. ISA2-TAP is indicated with the black arrow, ISA1 with the white arrow.

Figure 5. Molecular structures the affinity purified ISA2–TAP complex.

Purified ISA2-TAP protein was spotted onto carbon grids, negatively stained with uranyl acetate and analysed by transmission electron microscopy. The most abundant particles were rod-like, or in some cases dumbbell-like (upper images). In additional globular particles were observed, consistent with the presence of some residual RUBISCO contamination (lower images). In each case the field of view is 80 nm × 80 nm.

Figure 6. Activity of the ISA1-ISA2 isoamylase against different branched glucan substrates.

Potato amylopectin, oyster glycogen or maize β-limit dextrin (200 µg) was digested with partially purified ISA1–ISA2-TAP at pH 7.2 and 30°C. At different time points, aliquots were withdrawn and adjusted to 0.1 M NaOH to stop the reaction. Liberated glucan chains were separated and detected by HPAEC-PAD. Note that digestion of the β-limit dextrin gives rise to additional peaks that represent branched malto-oligosaccharides.

Figure 7. Recombinant expression of the ISA1–ISA2 isoamylase.

A. Soluble protein lysates from E. coli cells co-expressing different expression vector combinations of ISA1 and ISA2 were analysed by native PAGE in amylopectin-containing gels. The vector combinations are: 1, p0GWA-ISA1+ pET29a-ISA2 and 2, pET29a-ISA1+ p0GWA-ISA2. An extract of wild-type Arabidopsis leaves is shown for comparison (WT). The black arrow marks ISA1-ISA2 activity and the white arrow marks a background activity from E. coli lysate. B. Gel filtration chromatography of extracts of E. coli cells co-expressing ISA1 and ISA2. Fractions from a Sephacryl HiLoad 200 prep grade column were collected, concentrated and separated by SDS-PAGE. ISA1 and ISA2 were detected by immunoblot analysis probed with specific antibodies. The molecular masses based on the elution positions of known standards are indicated. A crude extract of E. coli expressing the recombinant isoamylase is shown on the left for comparison (rISA1–ISA2).

Figure 8. Expression of the recombinant ISA1–ISA2 isoamylase alters glycogen biosynthesis in E. coli.

A. Glucan content per gram wet weight (WW) of untransformed E. coli cells and transformed cells expressing either the active or inactive isoamylase complex. Black bars, insoluble glucan; Grey bars, methanol-precipitable soluble glucan; white bars, non-precipitable soluble glucan. Values are the means ± the standard error from measurements from duplicate experiments. B. E. coli cells expressing either the active (1) or inactive (2) isoamylase complex were grown on solid medium (top) and stained with iodine. Note the dark staining of cells expressing the active enzyme. Similar staining is observed with cells grown in liquid medium (bottom). C. Absorption spectra of the glucan-iodine complexes from untransformed E. coli cells (solid line) and transformed cells expressing either the active (dashed line) or inactive (dotted line) isoamylase complex. Top panel: methanol-precipitated glucans from the soluble fraction. Middle panel: glucans retained after ultrafiltration using a 50 kDa molecular weight cut-off filter. Lower panel: filtrate from the 50 kDa molecular weight cut-off filter.

Figure 9. Chain-length analysis of glucans synthesised in E. coli in the presence or absence of the ISA1–ISA2 isoamylase.

A. Methanol-precipitated glucans from the soluble fraction from of untransformed E. coli cells and transformed cells expressing either the active or inactive isoamylase complex were analysed by HPAEC-PAD, with or without a preceding debranching step. Note the presence of glucans between d.p. 20 and 50 in the non-debranched fraction of cells expressing the active isoamylase but not of the other lines (inset). Peaks marked with a red asterisk derive from E. coli itself and are not α-1,4−/α-1,6-linked glucans, judged by their resistance to digestion by amyloglucosidase (not shown). B. Methanol-precipitated glucans as in [A] were subject to ultrafiltration using a 50 kDa molecular weight cut-off filter. The glucans retained by the filter were analysed as in [A]. Note that the glucans between d.p. 20 and 50 that were present in the non-debranched fraction of cells expressing the active isoamylase are now absent, having passed through the filter.