Cryoprotective activities of FK20, a human genome-derived intrinsically disordered peptide against cryosensitive enzymes without a stereospecific molecular interaction

Expression and preparation of the IDP samples

Human genome-derived IDP samples and their truncation mutants were prepared by an E. coli expression system optimized for preparing IDP samples, as previously described (Goda et al., 2015a). In brief, the N-terminal autoprotease N(pro) from bovine viral diarrhea virus was selected as a fusion partner for protein expression using the pET-based N(pro) fusion protein expression system (Achmüller et al., 2007). The IDPs were expressed, N-terminal tags were removed, and finally purified by reversed phase HPLC (COSMOSIL^® 5C4-AR-300, Nacalai Tesque, ϕ4.6 mm × 250 mm) with 0.1% trifluoroacetic acid–acetonitrile solvent system. All the peptides were quantified by UV absorbance at 280 nm, lyophilized, and stored at −30 °C until use. FK20, a 20 amino acid peptide, and its all-D-enantiomeric isomer, FK20-D, were chemically synthesized (Biologica Co. Ltd. Nagoya, Aichi, Japan) with at least 80% purity, and purified by reversed phase HPLC.

Cryoprotection assay for LDH

Rabbit muscle lactate dehydrogenase (LDH) (L-2500; Sigma-Aldrich, St. Louis, MO, USA) was selected as the reporter enzyme for cryoprotection activity of shorter IDP peptides, using a slightly modified protocol based on Hughes and Graether (Hughes & Graether, 2011). The initial LDH solution contained 50 µg/ml in 10 mM sodium phosphate (pH 7.4). In a 1.5-ml microfuge tube, we mixed a 10-µl aliquot of the LDH solution with 10 µl of solution containing each of the individual protectants (IDP peptides) at concentrations ranging from 2.5 to 500 µg/ml. Five cycles of freezing in liquid nitrogen for 30 s and thawing in a water bath at 4 °C for 5 min were applied to each sample. Subsequently, LDH activity was measured using a standard NADH oxidase coupled-enzyme system according to the our previous report (Matsuo et al., 2018). For the analysis, we set the LDH activity of the untreated sample (the enzyme without freeze and thaw processes and without the addition of a cryoprotectant) as 100%. All measurements were performed in triplicate.

Cryoprotection assay for GSH

Glutathione S-transferase from Schistosoma japonicum (GST) was selected as the second reporter enzyme for cryoprotection activity of shorter IDP peptides because this enzyme was easily prepared in the laboratory and also suited for the subsequent CD spectroscopy experiments. The cryoprotection assay was performed according to our previous report (Matsuo et al., 2018). After five freeze/thaw cycles, GST activity was measured using a standard 1-chloro-2,4-dinitrobenzene (CDNB, 138630; Sigma-Aldrich, St. Louis, MO, USA) assay with a 96-well plate and 2300 EnSpire Microplate Reader (Perkin Elmer, Waltham, MA, USA). For the analysis, we set the GST activity of the untreated sample (the enzyme without freeze and thaw processes and without the addition of a cryoprotectant) as 100%. All measurements were performed three times and the standard deviations were calculated.

Circular dichroism (CD) measurements

CD spectra between 190 and 300 nm were collected on a J-805 spectropolarimeter (JASCO, Tokyo, Japan). The time constant, scan speed, bandwidth/resolution, and sensitivity of the spectropolarimeter were set at 1 s, 100 nm/min, 1 nm, and 100 mdeg, respectively. We measured a 300 μl solution of 250 µg/ml of FK20, FK20-D or 125 µg/ml each of the FK20/FK20-D mixture with a 10 mM sodium phosphate buffer (pH 7.4) in a quartz cuvette with a 1 mm light path length at 20 °C. Accordingly, we measured CD spectra of a 300 μl solution of 100 µg/ml of GST with 0, 25, 50 and 100 µg/ml each of FK20/FK20-D mixture in the same buffer. Similarly, we measured CD spectra of a 300 μl solution of 100 µg/ml of LDH with 0, 25, 50 and 100 µg/ml each of FK20/FK20-D mixture in the same buffer.

The GST and LDH denaturation temperatures (T_m) were determined by CD spectra at varying temperatures from 25 °C to 75 °C or 85 °C by monitoring the ellipticity at 222 nm at the speed of heating by 1 °C/min. Every 10 °C, CD spectra between 200 and 300 nm were automatically collected. For LDH, CD spectra of LDH were measured at 25 °C and 85 °C. A moving average of CD values of each five temperature points were plotted, and three straight lines corresponding to the baseline, the plateau, and the slope, were indicated. T_m was determined as the temperature of 50% denatured state.

Cryoprotective activity of shorter IDPs toward LDH and GST

To determine the minimal length of IDPs showing a substantial cryoprotective activity toward LDH and GST, we constructed several recombinant plasmids containing coding regions of the human genome-derived IDPs C1, D10, E1, and a series of their C-terminal truncated mutants. Amino acid sequences with schematic diagrams of IDPs and their truncated mutants are shown in Fig. 1. Human genome-derived IDPs C1, D10, and E1 consisted of 36, 37 and 39 amino acids, respectively. These peptides were proven as IDPs by both CD spectra and ¹H-¹⁵N 2D-NMR spectra in our previous studies (Goda et al., 2015b; Matsuo et al., 2018). We used the PONDR server (http://pondr.com) (Obradovic et al., 2005) to predict whether the series of C-terminal truncated peptides were also disordered. Accordingly, the charge-hydropathy plots (Uversky plots) were shown in Fig. S1. Cryoprotective activity of all peptides toward LDH at concentrations of 5, 10, 25, 50 and 500 μg/ml were measured. All peptides showed stronger cryoprotective activity than BSA in a concentration-dependent manner, with none showing less than 90% cryoprotection at the highest concentration (500 μg/ml) (data not shown). We compared these cryoprotective effects at the intermediate concentration (50 μg/ml) (Fig. 2A) and found that the cryoprotective activity of most of IDPs was independent of their amino acid lengths, except the three shorter peptides, C1²⁰, D10¹⁵ and E1²⁰. These peptides showed decreased cryoprotection activity of less than 60%. As a result, D10²⁰ was the shortest peptide with practical cryoprotective activity toward LDH at 50 μg/ml.

Subsequently, we examined the cryoprotective activity of these peptides toward GST (Fig. 2B) and the results were similar to those found with LDH. Also, D10²⁰ was again the shortest peptide with reasonably sufficient cryoprotective activity toward GST at 50 μg/ml. We further analyzed the normalized cryoprotective activity by amino acid lengths and calculated isoelectric points, pI (Figs. 2C and 2D, respectively). In order to compare the cryoprotective activities for the different reporter enzymes, both the maximum preserved enzymatic activities with the cryoprotective peptide, LDH with E1³¹ and GST with D10²⁰, were set to 100%, respectively. Peptides shorter than 20 amino acids were less cryoprotective, which is consistent with the proposed cryoprotection mechanism of DHNs as a molecular shield effect, where cryoprotective activity of various DHNs were roughly proportional to R_H logarithms of the cryoprotectants (Cuevas-Velazquez, Rendon-Luna & Covarrubias, 2014; Ferreira et al., 2018). Since we only examined cryoprotective IDPs smaller than 42 residues, only the lower limit of the correlation has been observed. In addition, we found that the IDP peptides with either extremely high or low pI were less cryoprotective, whereas the peptides with neutral pI were more potent.

Hereafter, our shortest cryoprotective IDP-derived peptide D10²⁰ is referred to as FK20.

Cryoprotective activity of FK20 and its all-D-enantiomeric isomer FK20-D

We further characterized cryoprotective activity and its FK20 mechanism by using its all-D-enantiomeric isomer, FK20-D. Based on this study and our previous study, the cryoprotective activity of IDP-derived peptides is likely not amino acid sequence specific. For example, we measured and compared ¹H-¹⁵N HSQC spectra of the target molecule Aβ(1-42) in the absence and the presence of the cryoprotective IDPs in our previous study, and we observed almost no spectra change (Ikeda et al., 2020). Although it is difficult to completely prove the absence of any specific interaction between the IDPs and the reporter enzymes, we challenged to provide indirect evidence to support this assumption as many as possible. In this study, we examined the cryoprotective activity of FK20-D and compared to that of the parent FK20. The cryoprotective activities of FK20, FK20-D, and the same total concentration of the racemic equimolar mixture of FK20 and FK20-D (FK20-LD) toward LDH and GST at varying concentrations were examined (Figs. 3A and 3B, respectively). FK20-D showed similar concentration-dependent cryoprotective profiles toward the reporter enzymes. The results suggested that specific molecular interaction between the reporter enzymes and either the FK20 or FK20-D peptides was absent, because the contribution of any residue-specific interaction of FK20 is not reproducible with FK20-D. We assumed that the effect of FK20 and FK20-D toward LDH and GST were more “environmental”. Indeed, FK20-LD showed similar concentration-dependent cryoprotective profiles (Figs. 3A and 3B). Note that the additive property of different cryoprotective IDPs was also observed between full length C1 and D10 (Fig. S2). A 1:1 mixture (in weight) of C1 and D10 showed a concentration-dependent cryoprotective profile similar to that of C1.

Effect of cryoprotective IDP peptide toward structure and thermal stability of GST and LDH

Next we tried to show that the cryoprotective activity was not due to either structural change or thermal stabilization of GST, but rather that reporter enzyme cryoprotection can be explained by avoiding denaturation of the enzymes by structural or thermal stabilization through either specific or non-specific IDP interactions. In our the other previous study, we succeeded in demonstrating the amyloid formation inhibitory activity of the same human genome-derived IDP peptides against Aβ(1-42) (Ikeda et al., 2020). At that time, we employed solution NMR techniques to monitor existence of a specific molecular interaction between ¹⁵N-labelled Aβ(1-42) in the presence of non-labelled IDPs. However, in this study, the molecular weight of the reporter enzyme (for example, GST) seems not suitable for solution NMR. Therefore, we employed CD spectroscopy, since FK20-LD, the equimolar mixture of FK20 and FK20-D, is silent in CD measurement. Figure 4A presents the experimental evidence of this unique feature of FK20-LD (grey), showing no significant CD band. FK-20 (black) showed the typical CD spectrum of a disordered state, whereas FK20-D (black, dashed line) showed the exact same CD spectrum with an inverted sign, as expected.

Figure 4B shows the CD spectra of GST in the absence (dashed line) and presence (solid line) of FK20-LD. There was no significant change in CD spectra, indicating that GST did not change its three-dimensional structure. Thus, FK20-LD cryoprotective activity was not a result of GST structural change. Figures 4C and 4E show a series of GST CD spectra at increasing temperatures from 25 °C to 75 °C with and without FK20-LD. Figure 4D and 4F show GST thermal denaturation plots taken from Figs. 4C and 4E, respectively. The GST thermal denaturation temperature (T_m) was 56.5 ± 1.2 °C for GST alone, and 58.2 ± 1.1 °C in the presence of 0.2 mg/ml of FK20-LD. Similarly, Fig. 5A shows the CD spectra of LDH in the absence (dashed line) and presence (solid line) of FK20-LD. Figures 5B and 5D show initial and final LDH CD spectra at increasing temperatures from 25 °C to 75 °C with and without FK20-LD. Figures 5C and 5E show LDH thermal denaturation plots taken from Figs. 5B and 5D s, respectively. The LDH thermal denaturation temperature (T_m) was 61.2 °C for LDH alone, and 62.0 °C in the presence of 0.2 mg/ml of FK20-LD. The results suggest that the thermal stability gains of GST and LDH by FK20-LD are small. Thus, we concluded that the FK20-LD cryoprotective activity was likely not a result of a thermal stabilization effect on the reporter enzymes.