*2.8. LC-MS/MS Analysis*

Fractions obtained from nine different semi-preparative GPC runs (three of CTRL, three of BSPH and three of PSPH hydrolysates) were brought to dryness and reconstituted in 0.2% formic acid. The peptide mixture concentration was estimated by measuring absorbance at 280 nm with a NanoDrop 2000 spectrophotometer (Thermo Scientific, San Jose, CA, USA), using dilutions of the MassPREP *E. coli* Digest Standard (Waters, Milford, MA, USA) to generate a calibration curve. Peptide concentration was adjusted to 1 μg μL−1. Two technical replicates for each sample were performed.

LC-MS/MS analyses were carried out using a Q Exactive mass spectrometer (Thermo Scientific) interfaced with an UltiMate 3000 RSLCnano LC system (Thermo Scientific). After loading, peptide mixtures (4 μg per run) were concentrated and desalted on a trapping precolumn (Acclaim PepMap C18, 75 μm × 2 cm nanoViper, 3 μm, 100 Å, Thermo Scientific), using 0.2% formic acid at a flow rate of 5 μL min−1. The peptide separation was performed at 35 ◦C using a C18 column (EASY-Spray column, 50 cm × 75 μm ID, PepMap C18, 3 μm, Thermo Scientific), using a linear gradient of 245 min from 5% to 37.5% of eluent B (0.1% formic acid in 80% acetonitrile) in eluent A (0.1% formic acid), at a flow rate of 250 nL min−1. MS data were acquired using a data-dependent Top12 method dynamically choosing the most abundant precursor ions from the survey scan, under direct control of the Xcalibur software (version 1.0.2.65 SP2, Thermo Fisher Scientific), where a full-scan spectrum (from 300 to 1700 *m*/*z*) was followed by tandem mass spectra (MS/MS). The instrument was operated in positive mode with a spray voltage of 1.8 kV and a capillary temperature of 275 ◦C. Survey and MS/MS scans were performed in the Orbitrap with resolution of 70,000 and 17,500 at 200 *<sup>m</sup>*/*<sup>z</sup>*, respectively. The automatic gain control was set to 1,000,000 ions and the lock mass option was enabled on a protonated polydimethylcyclosiloxane background ion as an internal recalibration for accurate mass measurements. The dynamic exclusion was set to 30 s. Higher Energy Collisional Dissociation (HCD), performed at the far side of the C-trap, was used as fragmentation method by applying a 25 eV value for normalized collision energy. An isolation width of *m*/*z* 2.0. Nitrogen was used as the collision gas.

Peptide identification was performed using Proteome Discoverer (version 1.4; Thermo Scientific), with Sequest-HT as the search engine for protein identification, according to the following criteria: Database: custom, obtained by merging *Bos Taurus* and *Ovis aries* downloaded from UniprotKB, (release 2019\_01); Precursor mass tolerance: 10 ppm; Fragment mass tolerance: 0.02 Da; Dynamic modification: methionine oxidation, Asparagine/Glutamine, Arginine deamidation, Serine/Threonine/Tyrosine phosphorylation), and Percolator for peptide validation (FDR < 1% based on peptide q-value). Results were filtered in order to keep only Rank 1 peptides, and protein grouping was allowed according to the maximum parsimony principle. Protein abundance was expressed by means of the normalized spectral abundance factor (NSAF). NSAF was calculated to evaluate the relative abundance of each protein and peptide, and "log ratio (log R)" values (log2 NSAF group a/NSAF group b) were obtained to estimate the fold changes of peptides between

experimental groups expressed as base 2 on a logarithmic scale [42,43]. In this approach, the spectral counts of each peptide were divided by its length and normalized to the average number of spectral counts in a given analysis. In order to eliminate discontinuity due to SpC = 0, a correction factor, set to 0.01, was used.

Peptides showing log ratio >1.5 or < −1.5 were considered as differentially abundant between groups. A two-tailed *t*-test was applied, using in house software to evaluate the statistical significance of differences between groups. The differentially abundant peptides were then evaluated using the "profiles of potential biological activity" analysis, available on BIOPEP [44] to find within them any sequence with known DPP-IV inhibitory, antioxidant and antibacterial activity.

### **3. Results and Discussion**

### *3.1. DPP-IV Inhibitory, Antioxidant Activity and GPC Profile of the Selected Hydrolysates*

The DPP-IV inhibitory and ABTS activity of the obtained hydrolysates and of the control samples are showed in Table 2. The P32/98 positive control showed a IC50 of 1.395 ± 0.007 × 10−<sup>3</sup> mg mL−1. In contrast, the retentate control samples (CTRL) did not show a measurable inhibition. BSPH had a significantly higher DPP-IV inhibitory activity compared to PSPH (*p* = 0.0381). The obtained DPP-IV inhibitory activity was lower than that previously described for milk and whey hydrolysates obtained from WPC, WPI of different species [45–50]. However, the observed activity was just about one order of magnitude less than that measured on hydrolysates obtained from pure β-lactoglobulin [41]. In addition, a DPP-IV inhibitory activity on hydrolysates from scotta has never been measured before, leading to the consideration of this matrix as a candidate substrate for the industrial production of DPP-IV inhibitory peptides.

**Table 2.** DPP-IV and antioxidant activity of hydrolysates, and control 1.


1 Values are mean ± standard deviation (*n* = 3). Within rows, values with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05). n.d.: absence of inhibition.

Further, the obtained data confirmed that enzymatic hydrolysis is a suitable way to increase the radical scavenging ability of sheep milk by-products [51]. In fact, BSPH showed a higher antioxidant activity compared to the control. Despite that, the hydrolysates did not differ regarding this property. A similar pattern was found by Monari et al. [15] for bovine scotta hydrolysates, which showed that the antioxidant activity of hydrolysates obtained using bromelain and pancreatin enzymes, did not differ significantly, even using different E:S ratios, both in unconcentrated bovine scotta and retentates.

Figure 1 shows the peptide distribution in terms of relative abundance according to the molecular weight obtained by gel permeation chromatography (GPC), and the comparison among BSPH, PSPH and CTRL.

As expected, most of BSPH and PSPH components were low molecular weight peptides (<1 kDa), whilst the high and medium molecular weight peptides (>10 kDa, 5–10 kDa, and 1–5 kDa) were more abundant in the control samples, which conversely showed a very low contribution of components with MW < 1 kDa (2.82 ± 0.21%). The amount of the 1–5 kDa fraction, even lower than the control, was significantly higher in BSPH (31 ± 0.9%) than in PSPH (23.56 ± 0.02%). Conversely, pancreatin in our system was more effective in producing peptides with low MW (74 ± 1%), compared to bromelain (66 ± 1%). Since pancreatin contains a mixture of proteases including chymotrypsin, trypsin and elastase [52], we suppose that in our system it exerted a more generalized proteolytic behavior than bromelain, which conversely has been reported to be less effective in producing free amino acids [15]. The specificity of bromelain may be responsible of the higher DPP-IV inhibitory activity measured in our hydrolysates.

**Figure 1.** Distribution of the relative abundance (%) according to molecular weight obtained by gel permeation chromatography (GPC), and the comparison among BSPH, PSPH and CTRL. Values (*n* = 3) with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05).

### *3.2. Antibacterial Activities of Hydrolysates*

The enzymatic hydrolysates obtained from ovine scotta tested did not show a complete inhibitory or bactericidal effect on the target microorganisms at the concentration tested (100 mg mL−1). However, the hydrolysates showed a variable but slightly inhibitory effect depending on the species or strain of bacteria tested. In particular, PSPH decreased significantly (*p* < 0.001), the maximum specific growth rate (μmax) respect to control being half the values in all bacteria tested except for *Salmonella bongori* 13,772 DSMZ strain, where the difference of its growth rate respect to control was not significant (see Figure 2A–F). The influence of BSPH was strain-dependent, decreasing significantly the μmax of *Listeria monocytogenes* B, *Listeria monocytogenes* D and *Staphylococcus aureus* 20,231 DSMZ, whereas the BSPH did not influence the μmax of *L. monocytogenes* 20,600 DSMZ, *L. monocytogenes* B and *Salmonella bongori* 13,772 DSMZ. An intriguing result was obtained with the CTRL, that decreased the μmax of *Listeria monocytogenes* B, *L. monocytogenes* C and *L. monocytogenes* E. All three strains were isolated from ovine ricotta cheese. Regarding the effect of scotta hydrolysates tested, no effect was observed on the lag time (λ) of *L. monocytogenes* C, *L. monocytogenes* E, *S. aureus* 20,231 DSMZ and *S. bongori* 13,772 DSMZ. An opposite effect of PSPH on lag time with respect to μmax was observed on *L. monocytogenes* 20,600 DSMZ (λ: 3.96 h) and *L. monocytogenes* B (λ: 1.42 h) strains, with a lag time for each bacterial strains that did not differ significantly from the control (3.83 h and 3.39 h for *L. monocytogenes* 20,600 DSMZ and *L. monocytogenes* B control respectively), but was significantly different from the other two treatments (BSPH and CTRL). Overall, all treatments reduced the bacterial density, confirming that scotta hydrolysates negatively influenced the growth of the bacteria tested. Contrasting with our results, Lestari et al. [36] revealed a strong antimicrobial activity of goa<sup>t</sup> milk protein hydrolysate by using bromelain as a hydrolyzing agent. Indeed, the minimum inhibitory concentrations of these hydrolysates against *S. aureus* and *Escherichia coli* were below 100 ppm. Bovine β-LG and α-LA were previously treated with pancreatin, and the resulting hydrolysates were shown to possess antimicrobial activity [53].

**Figure 2.** Effect of the different scotta-hydrolysates at concentration of 100 mg mL−<sup>1</sup> on the maximum growth rate (μmax) of bacteria strains target. BHI-WH, Brain Heart infusion broth medium without hydrolysates; BSPH, Bromelain filter sterilized hydrolysate; PSPH, Pancreatin filter sterilized hydrolysate; CTRL, Scotta not hydrolysate filter sterilized). (Panel (**A**–**F**): *Listeria monocytogenes* B (**A**); *L. monocytogenes* C (**B**); *L. monocytogenes* 20,600 DSMZ (**C**); *L. monocytogenes* E (**D**); *Staphylococcus aureus* 20,231 DSMZ (**E**), *Salmonella bongori* 13,772 DSMZ (**F**)). Different lowercase letters above the bar indicate statistically significant differences between different treatments (*p <* 0.001).

### *3.3. LC-MS/MS Analysis of Scotta Hydrolysates*

The LC-MS/MS analysis of the GPC fraction of the hydrolysates (BSPH, PSPH and CTRL) described in Section 2.8 allowed acquisition of 12,000 spectra from each run. A total of 58 ± 7 proteins and 547 ± 24 peptides were identified in BSPH, while 75 ± 6 proteins and 559 ± 33 peptides were identified in PSPH and further 83 ± 15 proteins and 1506 ± 381 peptides were identified in the CTRL samples (see Supplementary Materials, Sheet S1).

Considering BSPH vs. CTRL, a total of 29 proteins showed significant differences (*p* ≤ 0.05). Twenty-one of them were more abundant in BSPH, while eight were more abundant in CTRL (see Supplementary Materials, Sheet S2). The differential analysis of BSPH vs. CTRL highlighted 751 significant peptides (*p* ≤ 0.05). Of these, 388 were more abundant in BSPH samples, whilst 363 were more abundant in CTRL samples (see Supplementary Materials, Sheet S3).

Considering the available literature, differential peptides in BSPH were investigated to find sequences with reported DPP-IV inhibition, and antioxidant and antibacterial activity. The candidate peptides were further evaluated using the tool "profiles of potential biological activity" analysis, available on BIOPEP [44].

This approach highlighted 97 differential peptides containing at least one of the following sequences with known biological activity: LPQNI, VLGP, VLVLDTDYK, IPAVF, IPA, LKPTPEG, YPVEPF, YQEPVLGPVR, YVEEL, LDTDYKK, IDALNENK, KVAGT, AAS-DISLLDAGSAPLR, and ALK (see Table 3). All these peptides were attributable to βlactoglobulin protein (P67976), except for LPQNI, VLGP, YPVEPF, YQEPVLGPVR derived from β-casein protein (P11839), and the tripeptide IPA originating from k-casein (P02669). In the following text and in the tables the active sequences contained in the identified peptides will be highlighted with bold characters.

The differential peptides within the DPP-IV sequence showed a length ranging from eight to twenty-eight amino acid residues. The shortest peptide was **LDTDYKK**Y from β-lactoglobulin with an estimated MW of 1044.52 Da (Log R = 2.38), whilst the longest was AIPPKKDQDKTE**IPA**INTIASAEPTVHS released from k-casein with an estimated MW of 3051.55 Da (Log R = 2.78).

Considering PSPH vs. CTRL, a total of 32 proteins showed a significant difference (*p* ≤ 0.05). Twenty-five of them were more abundant in PSPH samples, while seven were more abundant in CTRL samples (see Supplementary Materials, Sheet S4). The peptide differential analysis of PSPH vs. CTRL highlighted 667 significant peptides (*p* ≤ 0.05). Of these, 294 were differential in PSPH, and 373 were more abundant in the CTRL profile (see Supplementary Materials, Sheet S5).

The "profiles of potential biological activity" analysis on BIOPEP revealed 75 differential peptides containing at least one of the sequences previously observed in BSPH (see Table 4). The sequences with DPP-IV inhibitory activity were encrypted in peptides of 9 to 22 amino acid residues. In detail, K**IDALNENK** and **ALK**ALPMHI appeared the shortest peptides originating from β-lactoglobulin with an estimated MW of 1044.56 Da (Log R = 2.84) and MW of 992.59 Da (Log R = 2.03) respectively. Furthermore, the longer peptide identified was KDQDKTE**IPA**INTIASAEPTVH deriving from k-casein, with an estimated MW of 2884.55 Da (Log R =5.07).

Venn diagrams were used to evaluate the number of differential peptides, more abundant in BSPH vs. CRTL and PSPH vs. CRTL (Figure 3A,B, respectively), with potential biological activities.

**Figure 3.** Distribution of the differential peptides more abundant in BSPH vs. CRTL (**A**) and PSPH vs. CRTL (**B**), according to their putative biological activities (DPP-IV inhibition, antioxidative and antibacterial properties).

Venn diagrams highlighted that none of the peptides contained sequences with only antibacterial or antioxidant known activity (Figure 3A,B). Interestingly, the 72% of the peptides in PSPH compared to the CTRL, contained sequences associated with antioxidant and DPP-IV inhibitory activity (Figure 3B). Moreover, a 5.6-fold higher number of peptides

containing sequences with only DPP-IV inhibitory activity was found in BSPH vs. CTRL compared to PSPH vs. CTRL.

Furthermore, the differential analysis of BSPH vs. PSPH showed 82 proteins in total and 29 differentials (*p* ≤ 0.05). Among them, 21 were more abundant in BSPH, while eight were more abundant in PSPH (see Supplementary Materials, Sheet S6). The peptide analysis indicated 1181 peptides, and 752 were significantly differential (*p* ≤ 0.05). Among them, 388 peptides were more abundant in BSPH, while 364 were more abundant in PSPH (see Supplementary Materials, Sheet S7).

A total of 208 differential peptides contained sequences with known DPP-IV inhibitory activity (see Table 5).

Figure 4 groups differential peptides (BSPH vs. PSPH) generated by the same protein and dividing the components according to reported biological activity. Histograms show, for each protein (β-casein, k-casein, and β-lactoglobulin), the peptides more abundant in BSPH or in PSPH, respectively.

**Figure 4.** Number of differential peptides (BSPH vs. PSPH) grouped by the proteins and biological activities.

Sixty-four differential peptides were derived from β-casein, 52 of which were more abundant in PSPH, and 12 were more abundant in BSPH (Figure 4). Fifty-eight differential peptides were originated from k-casein 24, of which were more abundant in BSPH and 34 were more abundant in PSPH.

Interestingly, the number of differential peptides derived from β-lactoglobulin was differently distributed between BSPH and PSPH. In fact, a total of 86 differential peptides derived from β-lactoglobulin were identified, 80 of them more abundant in BSPH and six in PSPH. This result could help to interpret the higher DPP-IV inhibitory activity showed in vitro by BSPH compared to PSPH. Moreover, none of the identified peptides derived from α-lactalbumin. Power et al. [50], found in silico a three-fold higher content of peptide sequences with potential DPP-IV inhibitory activity in β-lactoglobulin compared to αlactalbumin. Furthermore, Tulipano et al. [54] found by in silico analysis that bovine β-lactoglobulin was a better source of DPP-IV inhibitory peptides compared with αlactalbumin after treatment with digestive proteases.


**Table 3.** Analysis of differential peptides of BSPH vs. CTRL (Log R ≥ 1.5 and Log R ≤ −1.5).


**Table 3.** *Cont.*

The active sequences contained in longer peptides are highlighted with bold characters.


**Table 4.** Analysis of differential peptides of PSPH vs. CTRL (Log R ≥ 1.5 and Log R ≤ –1.5).


**Table 4.** *Cont.*

The active sequences contained in longer peptides are highlighted with bold characters.

**Table 5.** Analysis of differential peptides of BSPH vs. PSPH (Log R ≥ 1.5 and Log R ≤ −1.5).



**Table 5.** *Cont.*

−3.42

PPKKDQDKTE**IPA**INTIASAEPTV129–153




**Table 5.** *Cont.*

The active sequences contained in longer peptides are highlighted with bold characters.
