How 'arm-twisting' by the inducer triggers activation of the MalT transcription factor, a typical signal transduction ATPase with numerous domains (STAND)

Abstract

Signal transduction ATPases with numerous domains (STAND) get activated through inducer-dependent assembly into multimeric platforms. This switch relies on the conversion of their nucleotide-binding oligomerization domain (NOD) from a closed, ADP-bound form to an open, ATP-bound form. The NOD closed form is stabilized by contacts with the arm, a domain that connects the NOD to the inducer-binding domain called the sensor. How the inducer triggers NOD opening remains unclear. Here, I pinpointed the NOD-arm interface of the MalT STAND transcription factor, and I generated a MalT variant in which this interface can be covalently locked on demand, thereby trapping the NOD in the closed state. By characterizing this locked variant, I found that the inducer is recognized in two steps: it first binds to the sole sensor with low affinity, which then triggers the recruitment of the arm to form a high-affinity arm-sensor inducer-binding site. Strikingly, this high-affinity binding step was incompatible with arm-NOD contacts maintaining the NOD closed. Through this toggling between two mutually exclusive states reminiscent of a single-pole double-throw switch, the arm couples inducer binding to NOD opening, shown here to precede nucleotide exchange. This scenario likely holds for other STANDs like mammalian NLR innate immunity receptors.

Figure 1.

Residue 70 of the MalT NBD can be cross-linked with residues of the arm domain. (A) Schematic representation of the MalT primary sequence. NBD, nucleotide binding domain; HD helical domain; WHD, winged-helix domain; DBD, DNA-binding effector domain. In the upper part of the cartoon, the location of residue Q70 and the limits of the peptides (black bars and dotted lines) that were found cross-linked to AET-derivatized C70 after irradiation of HMalT^C-,Q70C in the resting form, followed by endoproteinase LysC or trypsin digestion (see text), are indicated. NODh and ASh indicate the proteinase K sensitive hinges (34). (B) The STAND activation scheme of MalT (same color coding). The transition from the resting to the multimerization competent form involves three events: inducer binding to the sensor, NOD isomerization, nucleotide exchange. The precise sequence of these events is not known, and the role of the sensor and the arm in this process is unclear (indicated by dotted outlines). (C) Model of the MalT structure from aa 1 to 417. Color coding is as in (A) except for Q70 and E395, which are represented in green. ADP is colored according to its atoms. (D) Analysis of endoproteinase LysC proteolysis products of AET-cross-linked HMalT^C-,Q70C (See also Fig. S1B, C, D). Endoproteinase LysC digests of 4 µg HMalT^C-,Q70C were analyzed by SDS-PAGE in the presence (+) or absence (-) of DTT. Molecular weight of markers are indicated in kDa, 18 kDa and 14 kDa fragments are highlighted. (E) HMalT^Q70C,E395C disulfide bond formation detected by SDS-PAGE. Proteins HMalT, HMalT^Q70C, HMalT^E395C and HMalT^Q70C,E395C were analyzed by SDS-PAGE in the presence and absence of DTT. The + signs indicate the subsitution(s) harboured by the protein (Q70C, E395C), whether the protein was pretreated with DTT as a first step (DTT pretreatment, see Supplementary Materials and Methods), whether the sample was treated with N-ethylmaleimide (NEM) and whether DTT was present during SDS-PAGE analysis (DTT).

Figure 2.

The disulfide bond of oxidized HMalT^Q70C,E395C traps the protein in a long-lived complex with ADP. (A) ADP is sequestered by oxidized HMalT^Q70C,E395C irrespective of the presence of the inducer. Dissociation of the complexes formed by [α-³²P] ADP and HMalT or HMalT^Q70C,E395C was monitored in the presence and absence of maltotriose by a filter-binding assay, under reducing and non-reducing conditions. Data points and error bars represent the average and standard deviation of measurements taken at the same time point in three (HMalT^Q70C,E395) or two (HMalT) independent experiments. Dotted lines correspond to the best fit of a pseudo first-order equation to the data. Note that the + and − triose y-axis scales differ. (B) Half-lives of the complexes in the different conditions assayed. Half-lives ± standard errors were derived from the (A) plots by non-linear least-squares fitting (see above). (C) Residual ADP-binding in the absence of inducer is abolished by the disulfide bond of oxidized HMalT^Q70C,E395C. [α-³²P] ADP binding to HMalT (6.2 μM) or HMalT^Q70C,E395C (5.2 μM) in the presence or absence of DTT was measured by a filter binding assay. Data points and error bars represent the average and standard deviation of measurements taken at the same time point in two independent experiments.

Figure 3.

The HMalT^Q70C,E395C C70-C395 disulfide bond prevents inducer-dependent multimerization and inducer-triggered dissociation of the complex of HMalT^Q70C,E395C-MalY (MalY is a protein inhibitor of MalT which specifically binds MalT in its resting form). Purified proteins (10 μM in monomers) were incubated for 15 min at 20°C and filtered through a Superdex 200 column in the absence of a reducing agent. CC stands for HMalT^Q70C,E395C. Maltotriose was present both in the incubation and in the running buffer at the indicated concentration. Absorbance at 390 nm (multiplied by 15, represented by hairline traces in the 1.08–1.6 ml range) is indicated for cofiltration experiments to monitor the presence of MalY (11). Activated MalT generates polydisperse head-to-tail multimers in fast equilibrium with each other, hence the typical trailing shape of the HMalT curves in the presence of maltotriose (center and right panel, top).

Figure 4.

Characterization of the preactivated intermediate. (A) Inducer-bound non-reduced HMalT^Q70C,E395C adopts an intermediate conformation with protease-resistant NOD and AS hinges. Limited proteolysis of HMalT and HMalT^Q70C,E395C by proteinase K with or without inducer, under reducing or non-reducing conditions. For each set of conditions, increasing proteinase K/HMalT w/w ratios were used: 0, 1:1067, 1:533, 1:267, 1:133. The position of protein markers of 100, 70, 55, 35 and 25 kDa are indicated by black dashes on the left of the figure. Location of proteolysis products corresponding to AS and NOD hinge cleavage are indicated by open arrows on the right (ASh and NODh, respectively). (B) Non-reduced HMalT^Q70C,E395C binds maltotriose with low affinity. Maltotriose saturation curves for HMalT (wt, open circles) and HMalT^Q70C,E395C (CC, solid circles) obtained from quantification of the bands corresponding to AS hinge proteolysis by proteinase K in the absence of reducing agent. Data points and error bars represent the average and standard deviation of three independent experiments. K_D of 2.6 ± 0.5 mM for HMalT^Q70C,E395C and 131 ± 12 μM for HMalT were extracted from these data by non linear least-squares fitting (solid lines).

Figure 5.

Affinity of the arm-sensor polypeptide HMalT^336–806 for maltotriose. Representative maltotriose (filled circles) or maltose (open circles) binding curves of HMalT^336–806 (1 μM) obtained by monitoring the fluorescence emission change at 323 nm, after excitation at 295 nm, are presented. The fitting curve used to extract the dissociation constant for maltotriose binding is shown (continuous line). The K_D indicated represents the average of three independent experiments.

Figure 6.

Transition from the resting form (a) to the multimerization competent form (d) in the MalT STAND protein. The two inducer binding steps and the preactivated intermediate (b) are presented. High affinity inducer binding involves both the sensor and the arm, and is incompatible with arm-NBD contacts, so that unlocking of the NOD ensues (c). Note that this incompatibility can have two origins: either a competition between the liganded sensor and the NBD for the same site on the arm (as illustrated here), or the toggling of the arm between two conformations favoring interactions either with the NBD or with the liganded sensor. In the end, only when NOD is open can nucleotide exchange occur, leading to the multimerization competent form.