*4.5. CD*

The far-UV CD spectra were collected on a Jasco J815 spectropolarimeter (Jasco, Tokyo, Japan) fitted with a thermostated cell holder, and interfaced with a Peltier unit, at 25 ◦C. The instrument was periodically calibrated with (+)-10-camphorsulphonic acid. Several protein concentrations of LrtA (4.5, 9.8 and 19.6 μM, in protomer units) were used to test for concentration-dependent changes in the shape and intensity of the steady-state spectrum at two different pH values; differences were not observed at these concentrations (9.8 and 19.6 μM) at pH 8.0. However, protein-concentration-dependent changes (in the concentration range of 4.5 and 9.8 μM, in protomer units) were observed in the shape and ellipticity at pH 4.5, suggesting that at this pH the protein showed a larger tendency to self-associate (see above, Section 2.1, Figure S4) than at pH 8.0. Molar ellipticity was calculated as described [44,45,50].

Steady-State Spectra—Experiments were acquired with the same experimental set-up described previously [50]. Typically, spectra were acquired at a scan speed of 50 nm/min with a response time of 2 s and averaged over six scans with a bandwidth of 1 nm. Spectra were corrected by subtracting the baseline in all cases. Protein concentrations were 9.8 μM (in protomer units) for pHand chemical-denaturation experiments (for both unfolding and refolding). The refolding samples of LrtA were prepared as described above for fluorescence.

Thermal Denaturations—Thermal denaturations were carried out with the same experimental set-up described previously [50] and a protein concentration of 9.8 μM. Briefly, thermal denaturations were performed at a constant heating rate of 60 ◦C/h, a response time of 8 s, a band width of 1 nm, and acquired every 0.2 ◦C. Thermal scans were collected in the far-UV region at 222 nm from 25 to 85 ◦C.

### *4.6. NMR Spectroscopy*

The NMR experiments were acquired at 20 ◦C or 15 ◦C, when stated, on a Bruker Avance DRX-500 spectrometer equipped with a triple resonance probe and z-pulse field gradients.

1D 1H-NMR experiments—Homonuclear 1D-1H-NMR experiments were performed with LrtA at a concentration of 30 μM (in protomer units) in 0.5 mL, in 100 mM phosphate buffer (pH 8.0 or pH 6.9) and 500 mM NaCl in H2O/D2O (90%/10%, *v*/*v*), uncorrected for deuterium isotope effects. The 1D-1H spectra under these conditions were acquired with 16 K data points, with 2 K scans and a 6000 Hz spectral width (12 ppm), with the WATERGATE sequence [53]. Baseline correction and zero-filling were applied before processing. We also acquired a spectrum at pH 4.5 (100 mM acetate) with 64 K scans, while the other experimental parameters were as above. During buffer exchange in Amicon centrifugal devices the sample precipitated and we had to increase the number of scans to have a good signal-to-noise ratio. All spectra were processed and analyzed by using TopSpin 2.1 (Bruker GmbH, Karlsruhe, Germany). TSP was used as the external chemical shift reference, taking into account the pH-dependence of its resonance [15].

Translational Diffusion Measurements*—*The DOSY experiments at pH 8.0 and 4.5 were performed with the pulsed-gradient spin-echo sequence, as described previously [50], with 16 gradient strengths ranging linearly from 2 to 95% of the total power of the gradient unit. At both pH values, the duration of the pulse gradient was 2.5 ms, and the time between the two gradients was 150 ms. Samples were exchanged in D2O buffer (100 mM phosphate buffer (pH 8.0, not corrected by isotope effects) or 100 mM deuterated acetate buffer (pH 4.5, not corrected by isotope effects) with 500 mM NaCl at both pH values) by using Amicon centrifugal devices during 4 to 6 h. Precipitation was observed after the buffer exchange (from pH 8.0 to pH 4.5), and the sample was centrifuged at 13,000 rpm for five minutes, before putting it into the NMR tube. Further and more severe precipitation was observed during the D2O buffer exchange at pH 4.5 and attempts to acquire a DOSY by increasing the number of scans (until 8 K) at each pulse field gradient strength failed. The methyl region was used for intensity measurements in the experiment acquired at pH 8.0.
