Correlation of the amino-acid sequence and the 3D structure of the functional domain of EmaA from Aggregatibacter actinomycetemcomitans

Abstract

Adhesion to collagen is an important virulence determinant for the periodontal pathogen Aggregatibacter actinomycetemcomitans. Binding to collagen is mediated by the extracellular-matrix protein adhesin-A (EmaA). EmaA is a homotrimeric autotransporter protein that forms flexible antenna-like appendages on the bacterium surface. An ellipsoidal structure at the distal end of the appendage, composed of three subdomains, contains the functional domain of the molecule. A correlation between amino-acid sequence and subdomain structure (SI and SII) was proposed based on an analysis of the volume/molecular weight ratio. EmaA from three mutant strains (deletions of amino-acids 70-206 and 70-386 and a substitution mutation G162S) has been studied by electron microscopy to test this hypothesis. 3D structures were analyzed using single-axis tilt tomography of negatively stained preparations of bacteria combined with subvolume averaging. Additionally, a large number of 2D images of the apical domain of the adhesins from the mutants were extracted from micrographs of the bacterial surface, aligned and classified. The combined data showed that amino-acids 70-206 localize to subdomain SI and 70-386 comprise subdomains SI and SII. Moreover, we showed that the substitution mutation G162S, which abolishes collagen binding activity, does not affect the overall structural integrity of the functional domain. However, the structure of subdomain SI in this mutant is slightly altered with respect to the wild-type strain. These data also have allowed us to interpret the architectural features of each subdomain of EmaA in more detail and to correlate the 3D structure of the functional domain of EmaA with the amino-acid sequence.

Figure 1

(A) Schematic diagram of the protein sequence of EmaA. The EmaA domain organization indicated is based on the information previously described (Mintz, 2004; Yu et al., 2009). SP, signal peptide; head domain (dark gray); neck sequence (NS); stalk domain (light gray); MAD, membrane anchor domain. (B) Schematic diagram of the amino terminus (aminoacids 1–627) of the wild-type EmaA; in-frame deletion mutants and substitution mutant G162S. The asterisks show the location of amino-acid 162.

Figure 2

Transmission electron micrographs of A. actinomycetemcomitans strains stained with Nano-W. Below each micrograph are boxed out images of the apical domains of EmaA for each corresponding strain. (A) wild-type strain. (B) Δ70–206 mutant, EmaA mutant with a deletion corresponding to amino-acids 70 to 206. (C) Δ70–386 mutant, EmaA mutant with a deletion corresponding to amino-acids 70 to 386. (D) G162S mutant, EmaA mutant with a substitution mutation G162S. EmaA appendages on the bacterial surface are marked by arrows in the micrograph. Bends or kinks in the EmaA structure are marked with short arrows in the windowed images. Scale bars, 100 nm (micrographs) and 10nm (boxed images).

Figure 3

3D reconstruction of the functional domain of the EmaA appendages obtained by averaging EmaA subvolumes extracted from the tomograms. Left) Surface representation of the 3D structure of the mutant (blue) fitted in the wild-type volume (white mesh) and Right) Z-sections through the volume of (A) Δ70–206 deletion mutant, (B) Δ70–386 deletion mutant, and (C) G162S substitution mutant. Scale bar, 10 nm.

Figure 4

Classification of the Δ70–206 EmaA deletion mutant. (A) Class averages after correspondence analysis and hierarchical ascendant classification, (B) Corners of a 4×4 self-organizing map of the members of classes 5 (left) and 7 (right). The C_# in A denotes the class number and the numbers in B indicate the number of images contributing to the node. Scale bar, 10 nm.

Figure 5

Classification of the Δ70–386 EmaA deletion mutant. (A) Class averages after correspondence analysis and hierarchical ascendant classification, (B) Corners of a 4×4 self-organizing map of the members of classes 1 (left) and 3 (right). The C_# in A denotes the class number and the numbers in B indicate the number of images contributing to the node. Scale bar, 10 nm.

Figure 6

Classification of the G162S EmaA substitution mutant. (A) Class averages after correspondence analysis and hierarchical ascendant classification, (B) Corners of a 4×4 self-organizing map of the members of classes 3 (left) and 5 (right). The C_# in A denotes the class number and the numbers in B indicate the number of images contributing to the node. Scale bar, 10 nm.

Figure 7

Revised schematic model of the EmaA molecule. SP, signal peptide; functional domain (dark gray) including subdomains I, II and III (SI, SII, SIII); neck sequences (N1 to N7; location of necks 4–7 is based on daTAA server prediction); stalk domain (light gray); MAD, membrane anchor domain. Two of the bends observed in the EmaA structure correlate with N1 (10.3 nm) and N3 (29.4 nm).