Noncanonical interactions between serpin and β-amylase in barley grain improve β-amylase activity in vitro

Abstract

Serpin protease inhibitors and β-amylase starch hydrolases are very abundant seed proteins in the endosperm of grasses. β-amylase is a crucial enzyme in the beer industry providing maltose for fermenting yeast. In animals and plants, inhibitory serpins form covalent linkages that inactivate their cognate proteases. Additionally, in animals, noninhibitory functions for serpins are observed such as metabolite carriers and chaperones. The function of serpins in seeds has yet to be unveiled. In developing endosperm, serpin Z4 and β-amylase showed similar in vivo spatio-temporal accumulation properties and colocalize in the cytosol of transformed tobacco leaves. A molecular interaction between recombinant proteins of serpin Z4 and β-amylase was revealed by surface plasmon resonance and microscale thermophoresis yielding a dissociation constant of 10^-7 M. Importantly, the addition of serpin Z4 significantly changes β-amylase enzymatic properties by increasing its maximal catalytic velocity. The presence of serpin Z4 stabilizes β-amylase activity during heat treatment without affecting its critical denaturing temperature. Oxidative stress, simulated by the addition of CuCl₂, leads to the formation of high molecular weight polymers of β-amylase similar to those detected in vivo. The polymers were cross-linked through disulfide bonds, the formation of which was repressed when serpin Z4 was present. The results suggest an unprecedented function for a plant seed serpin as a β-amylase-specific chaperone-like partner that could optimize β-amylase activity upon germination. This report is the first to describe a noninhibitory function for a serpin in plants.

Keywords: Hordeum vulgare; amylase; barley; enzyme stability; oxidative stress; protein interaction; serpin.

Figure 1

Accumulation of β‐amylase and serpin Z4 during grain maturation and germination. Each Western blot was replicated at least three times using independent biological replicates. (a–c) Proteins were extracted from developing grains of barley cv. Harington with either PBS (upper blots) or PBS with 20 mM DTT (lower blots). In a and b, the proteins were fractionated on a nonreducing SDS‐PAGE gel transferred to a PVDF membrane and blotted against serpin Z4 in a or against β‐amylase in b. Lanes are days after pollination (DAP) Lanes 1–8; 8, 13, 18, 22, 28, 32, 39, dry grains. (c) β‐amylase activity was assayed on the same samples using the Betamyl‐3 method, shown is the ratio of free β‐amylase activity over total β‐amylase activity. (d–f) Proteins were extracted from 100 mg of germinating grains of barley cv. Harington with 1 ml of either PBS (upper blots) or PBS with 20 mM DTT (lower blots). Lanes are hour after imbibition (HAI) from left to right 1–6; 0, 5, 24, 30, 48, and 120 HAI. (d and e) Proteins were loaded on a nonreducing SDS‐PAGE gel transferred to a PVDF membrane and blotted against serpin Z4 in d and against β‐amylase in e. f, β‐amylase activity was assayed on the same grain samples as carried out in (c). The sizes of the proteins are indicated in kDaltons (kDa). The asterisk indicates the monomer size of serpin and β‐amylase, and the arrow indicates high molecular weight aggregates of the proteins

Figure 2

Transient expression of serpin Z4‐RFP and β‐amylase‐GFP in leaves of Nicotiana benthaniama. Confocal microscope observation of a, spongy mesophyll cells and b, epidermal cells

Figure 3

Interaction between serpin Z4 and β‐amylase established by microscale thermophoresis and by surface plasmon resonance. (a) β‐amylase was bound to a fluorescent probe, and increasing amounts of serpin Z4 were added. (b) Fluorescent serpin Z4 was probed with increasing β‐amylase as in a. The assays were conducted in PBS buffer with 0.1% Tween‐20. For estimation of KD, the results of three replicate experiments were combined. (c) Serpin Z4 was attached to a Biacore compatible sensor chip, and β‐amylase was injected at the following concentrations: 2 μM, 0.9 μM, 0.4 μM, 0.2 μM, and 0. 1 μM (1* to 5*, respectively). (d) β‐amylase was attached to the sensor chip and serpin Z4 injected at the following concentrations: 0.6 μM, 0.5 μM, 0.4 μM, 0.3 μM, and 0.2 μM (1–5, respectively). (e) Equilibrium constants of the interaction measured by surface plasmon resonance

Figure 4

Activity of β‐amylase in the presence of serpin Z4. (a) The activity of 2 pmol of β‐amylase was assayed using the Betamyl substrate from Megazyme in the presence of different concentrations of serpin. The graphs are representative of three independent experiments. Graph (a) is presented as change in O.D. (b) As in (a) using the reducing sugar DNSA method. The enzymatic activity of β‐amylase in the presence of serpin beginning at the asterisk is significantly different from the no serpin control with a p‐value < .03. (c) Kinetic properties of β‐amylase with serpin Z4 and DTT. The maximal velocity (V _max) and the substrate affinity (Km) are derived from the Lineweaver–Burk model. V _max is expressed in nmol maltose per minute, and Km is in mg/ml of starch. * The V _max of β‐amylase + serpin Z4 is significantly different from the Vmax of the β‐amylase control with a p‐value = .001. ** The V _max of β‐amylase+ Z4 + DTT 5 mM is significantly different from the Vmax of the β‐amylase + DTT 5 mM control with a p‐value = .001. The kinetic properties of β‐amylase were assayed in three independent experiments each one comprising three technical replicates

Figure 5

The effect of temperature on β‐amylase activity. (a) β‐amylase (2.5 pmol) was assayed by the DNSA method at different temperatures with or without serpin Z4 at a molar ratio of 1:5. The activity is presented in nmol of maltose released per minute of reaction. From 37°C to 53.3°C, the enzymatic activity of β‐amylase in the presence of serpin is significantly different from the control with a p‐value ≤ .03. The experiment was repeated three times with three technical replicates. (b) The thermal stability of β‐amylase with and without serpin Z4 was measured by thermal shift assay using sypro‐orange dye. The arrow indicates the melting temperature of β‐amylase. The graph is representative of three independent replicates

Figure 6

Enzymatic activity and immuno‐detection of recombinant β‐amylase and serpin under oxidizing conditions. (a) The activity of 2.5 pmol β‐amylase was assayed using the DNSA method with or without the addition of serpin Z4 at a 1:5 ratio upon increasing CuCl₂ concentrations. (b–d, e) β‐amylase and serpin Z4 were mixed at the indicated ratios and treated with 5, 10, and 20 μM CuCl₂ for 10 min. The proteins were fractionated in 10% SDS‐PAGE gel with (b, c) or without the addition of reducing agent, β‐mercaptoethanol (d, e) and blotted with antibody against β‐amylase (b, d) or against serpin Z4 (c, e). The asterisk indicates β‐amylase monomer size and the arrow polymeric forms. Western blots and enzymatic assays were replicated at least three times

Figure 7

Enzymatic activity and immuno‐detection of wild type and mutant forms of β‐amylase in the presence of serpin Z4. (a) Activity of 2 pmol of β‐amylase with increasing concentrations of serpin Z4 assayed by the DNSA method. Assays were on WT, C‐terminus truncated β‐amylase (Trunc β‐amylase) and the C115R amino acid replacement. The X axis represents the ratio of serpin Z4 to the β‐amylases. The Y axis expressed the fold change in the β‐amylase activity with serpin over the activity of the β‐amylase without serpin. Serpin Z4: β‐amylase ratios range between 20:1 and 0.04:1. The arrow indicates a ratio of 1:5. (b) Immunoblot analysis of recombinant β‐amylase wild type and mutants as in a. Proteins (0.6 μg) were treated with 5, 10, and 20 μM of CuCl₂, for 10 min and fractionated on a 10% SDS nonreducing PAGE gel and immunobloted with β‐amylase antibody. (c) The residual activity of β‐amylase compared to no treatment after addition of CuCl₂ as in b assessed by the DNSA method. Capital letters indicate significant differences in β‐amylase WT‐treated enzyme, lower cases letters, significant differences in Trunc β‐amylase‐treated enzyme, no statistical differences were found in the C115R β‐amylase‐treated enzyme. The results were analyzed by performing Tukey‐Kramer HSD tests. The asterisk indicates the monomeric forms of the three β‐amylases and the arrow, the polymers of β‐amylase WT and Trunc β‐amylase. Western blots and enzymatic assays were replicated at least three times