Properties of the CesT/Map84 complex in solution

To detect dynamic regions of the CesT/Map84 complex and differences in dynamics between free and effector-bound CesT, the amplitude of motions were determined via the steady-state heteronuclear ¹⁵N-¹H-NOE (see 3.5.4.3) measured for a uniformly

15N-labeled CesT/Map84 sample (Figure 6.23). ¹⁵N-¹H-NOE values suggest overall

A B

0 10 20 30 40 50 60 70 80

residue

0 20 40 60 80 100 120 140

residue -1

-0.5 0 0.5 1

NOE

Figure 6.23: ¹⁵N-¹H-NOE values are plotted against the residue number. Values for free CesT are indicated as black circles. Values for CesT (A) and Map84 (B) in complex with each other are shown as red filled triangles and are determined for a coexpressed and copurified ¹⁵N-CesT/Map84 complex sample. ¹⁵N-¹H-NOE values without assignment in (B) are indicated by empty triangles and connected by a dashed line. Secondary structure elements of the X-ray structure are indicated for CesT and proposed β-motifs for Map84 [161].

unchanged dynamics of CesT upon complex formation (Figure 6.23A). Only ¹⁵

N-1H-NOE values for residues 138-144 of CesT in complex are increased. This region includes the additional β-strand of the X-ray structure of free CesT which was not found under solution NMR conditions. However, ¹⁵N-¹H-NOE values for residues 138-144 for CesT in the complex are smaller than for the secondary structure regions and secondary chemical shifts do not indicate structure (Figure6.18). Thus, reduced flexibility but no additional secondary structure formation is indicated for residues 138-144 of CesT in complex with Map84 suggesting a conformation for Map84-bound CesT which is between the conformation of free CesT in the crystal and in solution.

A conformational change of C-terminal residues upon complex formation is also in-dicated by chemical shift changes ∼0.2 ppm for residues 151-153 (Figure 6.19).

For Map84 in complex only∼50 % of the backbone is assigned. The longest stretches with missing assignments are residues 39-58 and 10 C-terminal residues. For 12 non-assigned peaks ¹⁵N-¹H-NOE values between 0.6 and 0.8 were found. Overall

15N-¹H-NOE values indicate flexibility for most residues of Map84 in complex with CesT (Figure 6.23B). Only for residue 36 and around residue 60 ¹⁵N-¹H-NOE values

>0.5 and>0.6 are found, respectively. Therefore the ¹⁵N-¹H-NOE values without as-signment were assumed to belong to residues 39-58. Under this assumption residues 40-64 have ¹⁵N-¹H-NOE values above 0.6 and are thus suggested to directly interact with CesT. Due to homology to chaperone/effector complexes with known structures, β-motifs forming an intermolecularβ-sheet were proposed for residues 20-27 and 54-61 of Map [154-61]. ¹⁵N-¹H-NOE values ∼0.4 are found thereby pointing at restricted flexibility but not complete rigidity. Combined with the fact that theβ-motif was in-ferred from a chaperone/effector-complex structure containing only one β-motif, one β-motif is concluded as sufficient to provide the specific recognition between CesT and its effectors. This β-motif is most likely composed of residues 54-61 of Map. For Tir this β-motif is probably formed by residues 49-56 because the other proposed β-motif is only partly affected by CesT binding (Figure 6.13). Through interactions with the second β-motif of the effector, CesT may confer a hierarchy to the secre-tion of its seven known effectors via the TTSS of EPEC. To confirm this, at least for one of the CesT/effector complexes, the effector has to be assigned in complex with CesT. For Map84 in complex with CesT the assignment has to completed but

∼15-20 of the Map84 signals are weak or missing completely in the ¹⁵ N-TROSY-HSQC spectrum of Map84 after refolding the CesT/Map84 complex (Figure 6.15).

But in the ¹⁵N-TROSY-HSQC spectrum of coexpressed and copurified CesT/Map84 complex (Figure6.24) there are∼10 sharp peaks of Map84 which are weak in the spec-trum of the CesT/Map84 complex after refolding (compare Figures 6.24 and 6.15).

Furthermore, there are ∼10 additional peaks for the coexpressed CesT/Map84 com-plex sample (Figure6.24) which are not found in the spectra of complex samples after refolding. This finding is in agreement with the fact that the coexpressed sample is stable for more than a year while the refolded sample shows degradation after two

10 9 8 7 130

125 120 115 110

15 N (ppm)

1H (ppm)

Figure 6.24: ¹⁵N-TROSY-HSQC spectrum of uniformly ¹⁵N-labeled CesT/Map84 complex. Red circles indicate signals from Map84 which are weak in the HSQC spec-trum of the refolded complex (Figure 6.15) and green circles indicate peaks which are absent in both HSQC spectra with one of the components of the refolded complex labeled.

weaks. Thus, TROSY-based triple resonance spectra measured for a coexpressed and copurified,²H(75 %),¹³C,¹⁵N-labeled CesT/Map84 sample should be most suitable to complete the assignment of Map84 in complex with CesT.

As yet, ∼30 % of the expected Tir108 signals could not be detected in complex with CesT (Figure 6.13) and a stable coexpressed and copurified CesT/Tir108 complex could not be prepared. Since NMR-titration experiments with TirN and CesT sug-gested a role of residues outside the N-terminal 100 residues of Tir on CesT-binding (Figure 6.16), a coexpressed and copurified CesT/TirN complex should be prepared and may turn out to be best for NMR experiments and assignment of Tir in complex with CesT.

6.3.13 Conclusions

In this chapter, the role of the chaperone CesT and its chaperone/effector complexes for the TTSS mechanism were investigated using mainly NMR spectroscopy. Struc-tural data about TTSS chaperones and their complexes from solution NMR spec-troscopy are not yet available while many X-ray structures have already been solved.

In the first part of this chapter, the solution structure of CesT was revealed to be the non-swapped model structure based on experimental and charge-shape predicted RDCs. Results presented in the second part of this chapter demonstrate the relevance of models and conclusions derived for homologous chaperones and chaperone/effector complexes for CesT and its complexes. Firstly, the wrapping of the effector around the chaperone dimer is indicated by a shift of almost all CesT signals in the ¹⁵N-HSQC upon effector binding whereas no peaks are doubled due to a broken homodimer symmetry thus hinting at a pseudo-symmetric chaperone/effector complex in which effector residues with similar properties mediate the interaction with both monomers (Figure 6.17). Secondly, an extended, non-globular N-terminal structure of the effec-tor in the chaperone/effeceffec-tor complex is consistent with the small number of signals low field to 8.5 ppm for Map84 in complex with CesT (Figure 6.15). Thirdly, the in-solubility of Map in the absence of CesT agrees with the suggested chaperone function of preventing inappropriate interactions. The solubility of Tir without the chaperone agrees with experiments showing a CesT-independent secretion of Tir [29] thereby im-plying a pre-formed secretion signal for free Tir. Finally, the hydrophobic properties of the residues of CesT with the strongest chemical shift perturbations upon complex formation reveal the CesT/effector complex to be mediated by many hydrophobic interactions (Figure6.19).

However, the absence of an effector core protected by the chaperone and the addi-tional flexible ∼20 C-terminal residues point at differences in the effector binding mode with CesT compared to the homologous chaperones and the function for the TTSS mechanism in general. The depicted results agree with and extend the current view of the role of the chaperone in the effector targeting to the TTSS. Figure 6.25

summarizes the overall extended effector targeting model, including additional chap-erone/effector interactions proposed based on the results presented in this chapter.

(I) CesT recognizes an effector in the bacterial cytoplasm via hydrophobic surface

N ~100

Figure 6.25: Model of effector targeting to the TTSS through CesT in pathogenic E. coli. CesT is shown as a red ribbon diagram with the C-terminus added as a black line and the effector in blue with a globular C-terminal and extended or unfolded∼100 N-terminal residues. In the bacterial cytoplasm, the effector is recognized by CesT (I) and the CesT/effector complex is formed via β-motifs (II). (III) Mediated by the unfolded C-termini of the CesT dimer and the unfolded N-terminal ∼15 N-terminal residues of the effector the CesT/effector complex interacts with the ATPase EscN of the TTSS. (IV) Conformational change catalyzed by EscN leads to CesT dissociation and translocation of the effector into the eukaryotic cell.

patches. (II) Upon formation of manifold hydrophobic CesT/effector interactions, a part of the N-terminal∼100 residues wraps around the chaperone. In the complex, the most important new tertiary structure is formed between two β-motif regions of the tertiary structure lacking effector and identical regions of the CesT monomers. (III) The two unfolded C-terminal 20 residues of CesT and the∼15 N-terminal residues of the effector [170] direct the complex to the TTSS and mediate the association with

the ATPase EscN. (IV) Using the unfolded regions as starting points, EscN catalyses conformational changes which lead to dissociation of CesT and translocation of the unfolded effector into the eukaryotic cell.

7