Standard solid-phase peptide synthesis methods were used for the synthesis of the truncated granulin peptides. The peptide chains were assembled on 2-chlorotrityl chloride resin (Peptides International, USA, product code RCT-1056-PI) using manual procedures as previously described [13]. Each peptide contained two cysteine residues, and oxidation of these cysteine residues into a disulfide bond was carried out by incubating the peptides in 100 mmol/L ammonium bicarbonate (pH 8.2) for 24 h at room temperature. The pH was checked after dissolution to confirm the solutions were still at pH 8.0. Preparative reversed-phase high-performance liquid chromatography (RP-HPLC) was used to purify the peptides following the oxidation reaction (Phenomenex Jupiter C18, 10 μm, 300 Å, 250 mm × 21.2 mm), and the purity assessed using analytical RP-HPLC (Phenomenex Jupiter 4 μm Proteo column C12, 10 μm, 90 Å, 150 mm × 2 mm). A 1% gradient was used to monitor the purity (0–60% solvent B over 60 min with a flow rate of 0.4 mL/min) and the absorbance was monitored at 214 nm and 280 nm. A 5800 MALDI TOF/TOF mass spectrometer (SCIEX, Framingham, MA, USA) and an LCMS-2020 (Shimadzu, Japan) were used to analyze the masses of the fractions.

The three-dimensional structures of the peptides were analyzed using nuclear magnetic resonance (NMR) spectroscopy data and calculations based on torsion angle dynamics. Peptides were dissolved at a concentration of ~ 0.2 mmol/L in 90% H₂O/10% D₂O (v/v), pH 4.5, and 4,4-dimethyl-4-silapentane-1-sulfonic acid (DSS; Cambridge Isotope Laboratories) used as an internal chemical shift reference. A 600-MHz AVANCE III NMR spectrometer (Bruker, Karlsruhe, Germany) was used to record spectra which included TOCSY (80 ms mixing time), NOESY (200–300 ms mixing time), and DQF-COSY experiments. The relaxation delay for the experiments was set to 1 s and the temperature was set to 290 Kelvin (K). TOPSPIN software (Bruker, Billerica, MA, USA) was used to process the data, and CCPNMR [14] was used to analyze and assign the spectra based on established protocols [15]. The program CYANA was used to calculate the three-dimensional structures, using a macro for automated assignment of the non-intra-residue NOEs [16]. Restraints for dihedral angles were predicted by TALOS-N and incorporated into the structure calculations [17]. Distance restraints for the disulfide-bonds were included in the calculations. Structures were visualized using MOLMOL, a molecular graphics program for displaying and analyzing the three-dimensional structures of molecules [18].

The cell line used for the cell proliferation experiments (human skin normal fibroblast cell line, 1BR.3.GN), was obtained from the European Collection of Authenticated Cell Cultures (ECACC, product code 90020509). The conditions used to culture and maintain the cells are those previously described [13] (i.e. complete media: Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12), 1 × GlutaMAX (Thermofisher, product code 10565042) and 1 × antibiotic/antimycotic (penicillin, streptomycin, and amphotericin B; Thermofisher, product code 15240062), 10% fetal bovine serum (FBS; Thermofisher, product code 12664025) at 37°C and 5% CO₂). A 21-residue peptide was used as the negative control. Cell proliferation assays were done in the presence of low nutrient media: DMEM/F12 as above with 0.5% FBS.

To monitor the cell proliferation, 1BR.3.GN cells seeded at 5,000 cells per well were grown overnight at 37°C in 150 μL of complete media. Agilent E-plates (Agilent) were used for the assays. The cells were monitored using an xCELLigence SP system (Agilent). This system uses gold microelectrodes, which are integrated into the base of tissue culture plates, to measure electrical impedance. Low-nutrient media was used to wash the cells (three times) and incubate the cells (minimum of 6 h) prior to treatment with peptides. Peptides were quantified based on absorbance measurements at 214 nm and predicted extinction coefficients based on the amino acid compositions [19, 20]. Treatments were prepared at 8.5 × target concentration and added to each well in 20 μL, for a final total volume of 170 μL and 1 × concentration. The pH was checked after dissolution to confirm the solutions were still at pH 7. The cell proliferation experiments were carried out over a period of 5–6 days following treatment with the peptides. The xCELLigence system was used to record cell indexes at intervals of 1 h over this time period. To determine the cell proliferation ratios cell index readings were normalized before peptide treatment.

A Brown-Forsythe test was used to determine variance homogeneity. It was not possible to determine if the data displayed a normal distribution because of the limited number of data points. A one-way ANOVA test with Holm-Sidak’s multiple comparison corrections was used to compare the cell proliferation rates between treatments at each concentration and the nearest peptide control concentration in GraphPad Prism 9.0.

Three Ov-GRN_{12–35_3s} fragments (GRN-L1, -L2, and -L3) were chemically synthesized on a 0.1 mmole scale using fluorenylmethyloxycarbonyl (Fmoc) solid-phase peptide synthesis. In the truncated analogue GRN-L3, Cys III and Cys V were replaced with alanine residues to prevent disulfide bond formation between these residues and enable the formation of the Cys IV to Cys VI disulfide bond present in Ov-GRN_{12–35_3s}. The folding yields for GRN-L1, GRN-L2, and GRN-L3 were 25%, 33%, and 30% respectively. The sequences of GRN-L1, -L2, and -L3 are shown in Figure 1. In addition, the locations of the loops related to these three fragments are highlighted on the three-dimensional structure of Ov-GRN_{12–35_3s}. Following purification of the peptides with RP-HPLC, the correct masses of the peptides were confirmed using MALDI TOF/TOF and electrospray mass spectrometry (Figure S1). Purified GRN-L1 was very broad despite only displaying one mass. This appears to be related to the additional, minor conformations present, as evidenced in the NMR spectra.

Analysis of the one-dimensional (Figure 2) and two-dimensional NMR spectra of the truncated granulin peptides allowed the assignment of the majority of the proton resonances and determination of the secondary shifts (Figure 3A). This analysis indicated that the GRN-L1 and -L3 peptides had no consecutive positive or negative secondary chemical shifts, consistent with the peptides being unstructured in solution. The chemical shifts for the Ov-GRN-1 fragments are given in the Table S1. Although GRN-L2 had consecutive positive secondary shifts for residues 4–7, which are consistent with the presence of β-sheet structure, there was a relatively low number of non-intra-residue or sequential NOEs in the NOESY spectrum, which prevented the determination of a well-defined structure. Three-dimensional structures were calculated with CYANA and confirmed the peptides generally had poorly defined structures, albeit constrained by the single disulfide bond as shown in Figure 3B. Structure statistics are given in Table S2. GRN-L2 appears to have the most well-defined structure as a result of having the smallest ring size.

The influence of the Ov-GRN_{12–35_3s} fragments on the proliferation of a human normal fibroblast cell line in real time was assessed using xCELLigence technology. GRN-L2 resulted in a 13.6% and 13.7% increase (P < 0.01) in cell growth compared to the negative control peptide when applied to cells at concentrations of 200 nmol/L and 1 μmol/L respectively. A 21-residue tropomyosin fragment was used as a negative control; tropomyosin fragments have previously been shown to have no cell proliferation effects [12, 13, 22]. Importantly, this peptide was produced in the laboratory using the same conditions as the Ov-GRN_{12-35_3s} fragments. GRN-L1 and GRN-L3 peptides tested at concentrations of 200 nmol/L and 1 μmol/L did not display statistically significant cell proliferation compared to the peptide control (Figure 4). GRN-L1 and GRN-L2 were also tested at lower concentrations but the cell proliferation was not statistically significant compared to the control peptide. Data from days 1 to 3 are given in Figure S2. The real-time response of the Ov-GRN-1 fragments at 1,000 nmol/L (Figure S3) highlights the enhanced potency of the GRN-L2 compared to the two other fragments.