Halogenation dictates the architecture of amyloid peptide nanostructures

Upon changing the position, nature and number of the halogen atoms, the same amyloidogenic peptide self-assembles into different nanostructures.

Rheology experiments were performed using a TA instrument ARG2 Rheometer. A 20 mm Stainless steel, parallel-plate geometry was used with a gap distance of 1000 µm.
Oscillatory frequency sweep studies were performed for a range of 0.1-100 rad/s, using a 0.5% strain. Oscillatory amplitude sweep studies were conducted from 0.01 to 100% strain with an angular frequency of 1 rad/s. The ring cast method was used for hydrogel preparation at a peptide concentration of 15 mM. The peptide solution was sonicated for 20 s in a sealed glass vial before heating to 90 °C to afford complete dissolution of the peptide.
After cooling the solutions were transferred into ring-casts of 22 mm diameter and placed in tightly sealed tissue-culture dishes for 48 hours. All measurements were repeated a minimum of three times. Figure S1. Reological characterization of the gels by frequency sweep studies whereby the storage modulus (G') was recorded as a function of angular frequency (ω). Amplitude sweep studies of peptide hydrogels showing G'' as a function of oscillation strain (γ). Figure S2. Polarized Optical Microscope images of a 48h aged KLVF(Br)F(Br) 15 mM solution, showing some birefringent spherulites. These images were registered with a Leica DM4500P POM system that was equipped with a Canon EOS 60D camera.

Transmission Electron Microscopy (TEM)
TEM bright field images were acquired using a Philips CM200 electron microscope operating at 200 kV equipped with a Field Emission Gun filament. A Gatan US 1000 CCD camera was used and 2048x2048 pixels images with 256 grey levels were recorded. The suspension were dropped onto a 200 mesh carbon-coated copper grid and air dried for several hours before analysis. No negative staining was used.

Dynamic Light Scattering measurements were performed on an ALV/CGS-3 Platform-based
Goniometer System equipped with an ALV-7004 correlator and an ALV / CGS-3 goniometer.
The signal was detected by an ALV-Static and Dynamic Enhancer detection unit. The light source was the second harmonic of a diode-pumped Coherent Innova Nd:YAG laser (λ = 532 nm), linearly polarized in the vertical direction. Measurements were performed at 25 °C.
Approximately 1 mL of sample solution was transferred into the cylindrical Hellma scattering cell.
The dynamic information on particles present in the peptide solutions were derived from the normalized autocorrelation function g2(q, τ) of the scattered intensity, which is measured according to where q is the scattering vector, τ is the relaxation time and I is the scattered intensity. Data analysis has been performed with two different methods: in the first one, the autocorrelation functions have been analyzed through Laplace inversion (CONTIN algorithm), which resulted in a double distribution of the decay rates associated to two distinct populations. In the second method, the same autocorrelation functions have been analyzed using a double exponential decay model, yielding two distinct decay rates. For each sample, at least three measurements were performed at four different angles (70°, 90°,110°,130°), corresponding to four different scattering vectors (q): where n is the refractive index of the medium and θ is the scattering angle. The extracted decay rates were plotted versus the square scattering vectors (q 2 ) showing a linear dependence typical of Brownian motion. The slope of this curve represents the averaged translational diffusion coefficient (DT), from which, through the Stokes-Einstein equation (kB is the Boltzmann constant, T is the temperature and η is the viscosity) it is possible to calculate the averaged hydrodynamic radius (RH) associated to each population. This analysis, although neglecting polydispersity of the two populations, is a good control on the Laplace inversion method (CONTIN), which can produce artefacts if the autocorrelation functions are significantly noisy, with low amplitude to baseline ratio.   The autocorrelation functions associated to four different angles (70°, 90°, 110°, 130°) were fitted with a double exponential model, leading to two different decay times, associated to two distinct populations. c) Evolution of the autocorrelation function at 90° for KLVF(Br)F(Br) 5 mM solution.

Cryogenic Transmission Electron Microscopy (cryo-TEM).
The cryo-TEM images were collected using JEM 3200FSC field emission microscope (JEOL) operated at 300 kV in bright field mode with Omega-type Zero-loss energy filter. The images were acquired with Gatan digital micrograph software while the specimen temperature was maintained at -187 o C. The Cryo-TEM samples were prepared by placing 3 µL aqueous dispersion of nanoparticles/clusters on a 200 mesh copper grid with holey carbon support film (CF-Quantifoil) and plunge freezed using vitrobot with 2s blotting time under 100% humidity. No negative staining was used.     included in the fit. There is a broad peak in the SAXS data near q =0.04 nm-1 which corresponds to a structure factor peak due to inter-fibril correlations which was not included in the form factor fitting. performed at the X-ray diffraction beamline (XRD1) of the Elettra Synchrotron, Trieste (Italy). [4S] The crystals were dipped in perfluoropolyether vacuum oil (Fomblin) and mounted on the goniometer head with a nylon loop. Complete datasets were collected at 100 K (nitrogen stream supplied through an Oxford Cryostream 700) through the rotating crystal method. Data were acquired using a monochromatic wavelength of 0.850 Å for KLVF(I)F(I) and 0.700 Å for KLVF(Br)F(Br) and KLVF(Cl)F(Cl) on a Pilatus 2M hybrid-pixel area detector. The diffraction data were indexed and integrated using XDS. [5S] Scaling have been done using CCP4-Aimless code. [6S,7S] Crystals appear as very thin needles prone to radiation damage, as previously reported for other halogenated molecules. [8S,9S] For the brominated peptide we managed to collect a complete dataset from a unique crystal; for the iodine and chlorine derivatives, diffraction decayed even after small doses so four different datasets had to be merged for KLVF(I)F(I) and three datasets for KLVF(Cl)F(Cl) (collected from different crystals randomly oriented). Semi-empirical absorption correction and scaling was performed for the KLVF(Br)F(Br) dataset, exploiting multiple measures of symmetryrelated reflections, using SADABS program. [10S] The structures were solved by the dual space algorithm implemented in the SHELXT code. [11S] Fourier analysis and refinement were performed by the full-matrix least-squares methods based on F2 implemented in SHELXL-2014. [12S] The Coot program was used for modeling. [13S] KLVF(Cl)F(Cl) peptide crystallized in a monoclinic unit cell (P 21 space group). The model has been fully refined anisotropic as a 2-component non-merohedral twin. Crystal showed two domains related by a 180° rotation around the c* reciprocal lattice direction (twin fraction refined to 17%). One peptide and four water molecules have been found in the asymmetric unit.
KLVF(Br)F(Br) and KLVF(I)F(I) peptides crystallized in equivalent conditions and showed the same P 212121 orthorhombic crystalline form. The cell volume is slightly bigger for the iodinated peptide, as expected from comparison of the halogen atomic radius. None of the crystals tested diffracted better than 1.1 Å, and considering radiation damage, the overall dataset resolution is not better than ~1.25 Å, for both the compounds. The number of data for model fitting was therefore limited and, to avoid over-refinement, anisotropic thermal motion modeling has been applied only to halogen atoms of the peptide (the heaviest atoms in the structures). Geometric restraints on bond lengths and angles (DFIX, DANG) have been used for all the residues and thermal motion parameters restrains (SIMU) have been applied on disordered and poorly defined fragments. Hydrogen atoms were included at calculated positions with isotropic Ufactors = 1.2 Ueq or Ufactors = 1.5 Ueq for methyl and hydroxyl groups (Ueq being the equivalent isotropic thermal factor of the bonded non-hydrogen atom). R1(free) [14S] values have been calculated for the brominated and iodinated models, omitting 5% of reflections (randomly selected) from refinement cycles. Reasonable agreement of final R1(free) with R1 values (Table S1) exclude over-refinement issues, despite poor data/parameters ratios. A final refined Flack parameter 0.005(32) [15S] for KLVF(Br)F(Br) confirms the reliability of the stereochemical configuration shown. The flack parameters for KLVF(I)F(I) and KLVF(Cl)F(Cl) are not reliable as a consequence of dataset merging in presence of significant radiation damage (Flack parameter 0.521(54) and 0.38 (18)).
Nevertheless, R1 increases significantly inverting the structure (almost doubles) suggesting that the stereochemical configuration is the same as KLVF(Br)F(Br) (as expected from synthetic pathway). Pictures were prepared using Mercury [16S] and Pymol software. [17S] Essential crystal and refinement data (Table S4)

Infrared spectroscopy
Infrared spectra were recorded at room temperature using a Nicolet iS50 FT-IR spectrometer equipped with a DTGS detector. Peptides were analyzed as solutions (after heating at 100 °C in order to break any pre-formed fibrils) or gels at 15 mM in D2O. Spectra represent an average of 64 scans recorded in a single beam mode with a 4 cm -1 resolution and corrected for the background. The second derivative analyses of the spectra were performed using the Nicolet FTIR software, Omnic 9.0®, with a 13-point and 3 rd polynomial order Savitzky and Golay function. Second derivative spectra generated negative bands as compared with the original spectra, thus for comparison all the second-derivative spectra were multiplied by -1. Figure S23. FTIR spectroscopy of 15 mM peptides gels/solutions after standing for 48 hours at r.t.

Circular Dichroism (CD) Spectroscopy
All the circular dichroism experiments were carried out in deionized water (18.2 MΩ . cm) in detachable quartz cuvettes, using a JASCO J-815 CD spectrometer. Acquisitions were performed between 190 and 250 nm with a 0.1 nm data pitch, 1 nm bandwidth, 100 nm min -1 scanning speed and 1 s response time. All the spectra are an average of 10 scans and were corrected from a reference solution, comprised of deionized water (18.2 MΩ . cm) alone. Raw data (θ, in mdeg) were subsequently converted to mean residue ellipticity ([θ] in deg . cm 2. dmol -1 ) for the sake of comparison, in accordance with the following formulae: [18S] [ ] = 10 * * * ( − 1) where θ is the observed ellipticity in mdeg, c is the concentration of the sample in mol . L -1 , (n-1) is the number of peptide bonds, and l is the pathlength of the cuvette in cm. Figure S24. Circular Dichroism spectra of the forming gel-peptides. All the spectra were recorded at 15 mM concentrations. It is well known that the presence of aromatic groups in a peptide sequence can perturb CD signals related to secondary structure since n-π* and π-π* transitions between aromatic groups also absorb in the same region. It can, therefore, be not straightforward to draw definitive conclusions related to peptide secondary structure motifs from CD. Figure S25. Circular Dichroism spectra of the KLVFF halogenated derivatives recorded in deionized water at 400 µM concentration.

Confocal Microscopy
Hydrogels were imaged using a Zeiss LSM 710 microscope with a He/Ne laser (λex= 543 nm). The fluorescent dye, Rhodamine B, was incorporated into an aged hydrogel (48 h) scaffold by addition of 10 µl of the dye solution (0.1% w/v). Following complete absorption of the dye, the sample was excited at 543 nm and emitted light recorded using the E570LP emission filter.

Congo Red Staining
All samples were monitored for green birefringence using an Olympus BX50 polarizing microscope with a SensiCam PCO camera used to display and enhance images.
An 80% ethanol solution saturated with NaCl and Congo Red was freshly prepared before each measurement. A piece of each peptide hydrogel was placed on a glass microscope, allowed to air dry and then stained with Congo Red solution. Subsequently, excess Congo Red solution was blotted off the slide and the samples were analyzed using both bright and polarized light.

CTC Resin loading
CTC resin (400mg, 1.6 mmol/g loading) was swollen in CH2Cl2 for 30 min and then washed with DMF (3 × 5 mL). A solution of the entering amino acid (200 DCM (4 mL) was added and the resin shaken at rt for 4 h. The resin was washed with DMF (2 × 3 mL) and capping was performed by treatment with a methanol/DIEA solution in DCM (1 x 30 min). The resin was then washed with DMF (2 × 4 mL), CH2Cl2 (2 × 4 mL), and DMF (2 × 4 mL). The resin was subsequently submitted to manual iterative peptide assembly (Fmoc-SPPS).

Peptide Assembly via Iterative manual SPPS
Peptides were assembled by stepwise manual Fmoc-SPPS. Activation of entering Fmoc-

Cleavage from the Resin
Resin-bound peptide was treated with an ice-cold TFA, TIS, water, thioanisole mixture (90:5:2.5:2.5 v/v/v/v, 6mL). After gently shaking the resin for 2 hours at room temperature, the resin was filtered and washed with neat TFA (2 x 4 mL). The combined cleavage solutions were worked-up as indicated below.

Work-up and Purification
Cleavage mixture was concentrated under nitrogen stream and then added dropwise to icecold diethyl ether (40 mL) to precipitate the crude peptide. The crude peptide was collected by centrifugation and washed with further cold diethyl ether to remove scavengers. Peptide was then dissolved in 0.1% TFA aqueous buffer (with minimal addition of ACN to aid dissolution, if necessary). Residual diethyl ether was removed by a gentle nitrogen stream and the crude peptide was purified by RP-HPLC.

RP-HPLC analysis and purification
Analytical and semi-preparative reversed phase high performance liquid chromatography (RP-HPLC) were carried out on a Tri Rotar-VI HPLC system equipped with a MD-910 multichannel detector for analytical purposes or with a Uvidec-100-VI variable UV detector for preparative purpose (all from JASCO, Tokyo, Japan). A Phenomenex Jupiter 5µ C18 90Å column (150 x 4.6 mm) was used for analytical runs and a Phenomenex Jupiter 10µ C18 90Å (250 x 21.2 mm) for peptide purification. Data were recorded and processed with Borwin software. UV detection was recorded in the 220-320 nm range. Pure RP-HPLC fractions (>97%) were combined and lyophilized.

Electro-spray ionisation mass spectrometry (ESI-MS)
Electro-spray ionization mass spectrometry (ESI-MS) was performed using a Bruker Esquire 3000+ instrument equipped with an electro-spray ionization source and a quadrupole ion trap detector (QITD).Samples were dissolved at a concentration of 0.1mg/ml in 0.1% formic acid (aq) and injected.