The BAM7 gene in Zea mays encodes a protein with similar structural and catalytic properties to Arabidopsis BAM2

Abstract

Starch accumulates in the plastids of green plant tissues during the day to provide carbon for metabolism at night. Starch hydrolysis is catalyzed by members of the β-amylase (BAM) family, which in Arabidopsis thaliana (At) includes nine structurally and functionally diverse members. One of these enzymes, AtBAM2, is a plastid-localized enzyme that is unique among characterized β-amylases since it is tetrameric and exhibits sigmoidal kinetics. Sequence alignments show that the BAM domains of AtBAM7, a catalytically inactive, nuclear-localized transcription factor with an N-terminal DNA-binding domain, and AtBAM2 are more closely related to each other than they are to any other AtBAM. Since the BAM2 gene is found in more ancient lineages, it was hypothesized that the BAM7 gene evolved from BAM2. However, analysis of the genomes of 48 flowering plants revealed 12 species that appear to possess a BAM7 gene but lack a BAM2 gene. Upon closer inspection, these BAM7 proteins have a greater percent identity to AtBAM2 than to AtBAM7, and they share all of the AtBAM2 functional residues that BAM7 proteins normally lack. It is hypothesized that these genes may encode BAM2-like proteins although they are currently annotated as BAM7-like genes. To test this hypothesis, a cDNA for the short form of corn BAM7 (ZmBAM7-S) was designed for expression in Escherichia coli. Small-angle X-ray scattering data indicate that ZmBAM7-S has a tetrameric solution structure that is more similar to that of AtBAM2 than to that of AtBAM1. In addition, partially purified ZmBAM7-S is catalytically active and exhibits sigmoidal kinetics. Together, these data suggest that some BAM7 genes may encode a functional BAM2. Exploring and understanding the β-amylase gene structure could have an impact on the current annotation of genes.

Keywords: Arabidopsis thaliana; BAM genes; Zea mays; gene evolution; gene structure; small-angle X-ray scattering; starch metabolism; β-amylases.

Figure 1

Conservation of key residues in BAM2 and BAM7 from flowering plants and how they compare with the corresponding residues in DF-BAM and BAM1, and a map of the corn BAM7 gene with its two predicted transcripts. (a) WebLogo illustrating the conservation of residues predicted to be part of the nuclear localization signal in the DNA-binding domain as identified by mutagenesis in A. thaliana BAM7 (Reinhold et al., 2011 ▸). The alignment included 14 species of flowering plants that contain BAM7 that were selected to represent a diversity of orders and 12 species that contain DF-BAM7 (see Supplementary Table S1 for species and accession numbers). (b) WebLogo illustrating the conservation of 15 residues in the same species as in (a) identified as forming hydrogen bonds to starch in the active site of soybean (Glycine max) BAM5 (Laederach et al., 1999 ▸). (c) WebLogo illustrating the conservation of residues as in (a) compared with the three residues identified in BAM2 as being involved in allosteric regulation of activity in BAM2 (Ser464) or starch binding to a surface binding site for starch (Gly335, Gly446 and Trp449) (Monroe et al., 2017, 2018 ▸). (d) WebLogo illustrating the conservation of three residues in the same species as in (a) that were identified by mutagenesis as being important for tetramer stabilization: Trp456 and Asp590 in interface ‘A’ and Phe238 in interface ‘B’ (Monroe et al., 2018 ▸). (e) Predicted dual-function ZmBAM7 gene model. Coding regions of exons are colored black, with the exception of a region unique to the N-terminus of ZmBAM7-S that is colored red. The locations of the two putative transcriptional start sites (TSS1 and TSS2) and their respective translational start sites (AUG) in the two predicted transcripts are indicated with arrows. The 5′ and 3′ untranslated regions (UTR) of both transcripts are colored white.

Figure 2

Purification of recombinant ZmBAM7-S. (a) Pure protein was eluted in 10 ml fractions at increasing imidazole concentrations during affinity chromatography; only one elution fraction is shown for ZmBAM7-S (lane 6). A wash fraction with less than 12.5 mM imidazole is also shown (lane 5). Size-exclusion chromatography (SEC) was also performed in preparation for small-angle X-ray scattering analysis (lane 7). Markers are present in lanes 1 and 8. (b) SEC of ZmBAM7-S. Absorbance data were normalized to the largest value. The peak elution volume for ZmBAM7-S was 50.0 ml. (c) SEC molecular-weight calibration curve. The black line with the equation y = −3.62x + 10.37 was created from nine different calibration standards (gray points). The expected ZmBAM7-S tetrameric molecular weight (black) calculated from the ZmBAM7-S sequence (232 kDa) is shown. (d) Disorder predictions from IntFOLD and IUPred2A for ZmBAM7-S. Using a disorder probability score cutoff of 0.5 (pink line), the probability of being disordered was predicted for each residue in the ZmBAM7-S homology model (black line) or from the sequence of ZmBAM7-S using IUPred2A (blue line).

Figure 3

SAXS data for ZmBAM7-S and truncated ZmBAM7-S. (a) Log of intensity versus momentum transfer for ZmBAM7-S (black) and truncated ZmBAM7-S (gray) The inset plot shows the radius of gyration (R _g) versus ZmBAM7-S concentration (mg ml⁻¹). The R _g values of ZmBAM-S at four different concentrations submitted to SAXS analysis were calculated from the Guinier plot (red data points) and from the P(r) plot (black data points). (b) Kratky plot of ZmBAM7-S (black) and truncated ZmBAM7-S (gray). (c) Pair distance distribution function plot comparing ZmBAM7-S with other BAMs. ZmBAM7-S is shown as a solid black line, truncated ZmBAM7-S is shown as a solid gray line, AtBAM1 is shown as a dotted blue line, IbBAM5 is shown as a dotted gray line and AtBAM2 is shown as a dotted black line. (d) SASREF fits of ZmBAM7-S monomer and oligomers to the SAXS data. SASREF models were combined into a single file and fitted to the SAXS data using FoXS. The χ² values were 1123.26 for P1, 258.1 for P2, 2.83 for P4, 2.25 for P222, 11.81 for P6, 7.32 for P3₂, 20.16 for P4₂ and 1150.59 for P5₂. (e) FoXS fit of the P222 model from SASREF to the SAXS data.

Figure 4

Activity of ZmBAM7-S. (a) Gel showing the purity and amount of AtBAM2 and ZmBAM7-S used in activity assays. (b) Effect of substrate concentration on ZmBAM7-S (gray) and AtBAM2 (black) activity on a per milligram of protein basis. Points are shown for each of three replicate assays. Data were fitted to the Michaelis–Menten equation for cooperative enzymes. For AtBAM2 the K _m was 59.4 mg ml⁻¹ (95% confidence interval 58–61 mg ml⁻¹) and for ZmBAM7-S the K _m was 86.9 mg ml⁻¹ (95% confidence interval 83–91 mg ml⁻¹). The fitted data and all kinetic values are shown in Supplementary Fig. S4.