Fourier Transform Infrared Spectroscopy Reveals Molecular Changes in Blood Vessels of Rats Treated with Pentadecapeptide BPC 157

Recently, it was found that when confronted with major vessel occlusion and vascular failure, stable gastric pentadecapeptide BPC 157 therapy might rapidly functionally improve minor vessels to take over the function of disabled major vessels, reorganize blood flow, and compensate failed vessel function. We focused on the BPC 157 therapy effect obtained by giving 10 ng/kg ip to rats 5 min before sacrifice on the rat thoracic aorta, which we assessed with Fourier transform infrared spectroscopy (FTIR) 90 min thereafter. We applied a principal component analysis (PCA). The PCA model showed, with a clear distinction being mostly due to the PC1 score, differences between the spectra of BPC 157- and saline-treated rats. The comparison of the averaged spectra of these two groups with their differential spectrum and PC loadings allowed us to identify the parts of the FTIR spectra that contributed the most to the spectral separation of the two observed groups. The PC1 loadings and the differential spectrum showed that the main bands affecting the separation were the amid I band around 1650 cm−1, the amid II band around 1540 cm−1, and the vibrational band around 1744 cm−1. Fitting the spectral range between 1450 and 1800 cm−1 showed changes in protein conformation and confirmed the appearance of the vibrational band at 1744 cm−1. Controls had a substantially more intense vibrational band at 1744 cm−1. These spectral results showed the cells from saline-treated (control) rats to be in the early stage of cell death, while the samples from BPC 157-rats were protected. Thus, BPC 157 therapy changed the lipid contents and protein secondary structure conformation, with a rapid effect on vessels, within a short time upon application.

Likely, this therapeutic effect [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26] might present as an effect on the whole blood vessel wall. Conceptually, this functional improvement effect on blood vessels might be achieved immediately upon administration; then, it might be sustained to ascertain adequate blood vessel function to prevent and/or reverse the otherwise inevitable adverse effects that might be created by upcoming noxious events [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]. Noteworthily, the cytoprotection concept holds that the direct effect on blood vessels occurs as an essential cytoprotective effect that further maintains tissue integrity even under ex vivo conditions [4,5]. Thus, to substantiate this essential, rapid vascular effect, we used FTIR to discriminate particular cell components; therefore, the assessment was performed using thoracic aortas obtained from rats that received BPC 157 or saline therapy shortly before sacrifice. As a particular novelty point, it appears that the changes in aortic tissues under the influence of any drug with infrared spectroscopy have not been investigated so far. FTIR spectroscopy [35] involves the measurement of many spectra of a sample, each containing the chemical signatures (absorption bands related to vibrational modes of specific chemical bonds) of compounds at a specific location within the specimen. We recorded the whole mid-range infrared spectrum (600-3900 cm −1 ). To show the spectral differences, we used a principal component analysis (PCA). Moreover, the differences were the most pronounced in the lipid part of the spectrum.

Tissue Samples
We compared the spectra of abdominal aortic tissue samples from 10 Wistar Albino male rats, 200 g in body weight, either treated with BPC 157 (10 ng/kg ip) or treated with saline (5 mL/kg ip) 5 min before sacrifice. Then, abdominal aortic tissues were embedded in distilled water before further procedures (for 5 min). From each rat (5 rats per group), 20 samples were cut, and the tissue samples were put in silicon windows. Then, the samples were dehydrated (removing excess water from tissue, not from cells) under vacuum, since FTIR spectra are sensitive to water. Subsequently, the FTIR spectra of the samples were recorded. Ninety minutes was the time elapsed between the sacrifice and FTIR spectrum recording. Note that all sections were 10 to 15 µm thick. Each of the 20 tissue sections from every rat became a sample for FTIR spectroscopy.

FTIR Spectroscopy
The vibrational spectra of the samples were recorded with a Perkin-Elmer Spectrum GX spectrometer equipped with a liquid N 2 refrigerated Mercury Cadmium Telluride (MCT) detector. Optical-grade silicon windows were used for acquiring the 1000-scan background, which was automatically subtracted from the tissue spectra. One hundred scans were recorded for each tissue section to obtain the vibrational spectra, which lasted around 5 min for each sample. Data were acquired in the 450-4000 cm −1 spectral range, in transmission mode, with a resolution of 4 cm −1 . A sample area with a diameter of approximately 1 cm was recorded. The resulting spectrum was a sum of the contributions of all the tissues in the sample area with characteristics depending on the relative contributions of those tissues (i.e., different relative concentrations of standard molecules existing in every biological material).

Data Analysis
Matlab10 (Mathworks) was used for processing the spectra. First, all recorded spectra were baseline-corrected and normalized. Baseline correction was performed using the SNIP (statistics-sensitive nonlinear iterative peak-clipping) algorithm. After that, the vector normalization of the spectra was conducted to compensate for spectral differences caused by small variations in the recording conditions. Many physical and chemical factors, such as sample size, humidity, molecular interactions, etc., can affect the recording of spectra. Preprocessing techniques compensate for those deviations and intensify the relationship between the spectral signals and the observed characteristics of tissues.
Matlab10 and PLS Toolbox (Eigenvector Research) were used to perform the principal component analysis (PCA). The PCA is an unsupervised statistical method that reduces the multidimensional experimental data set to a much smaller number of uncorrelated variables called principal components [23]. Usually, only the first two principal components, PC1 and PC2, which account for most of the variance presented in the experimental data, are utilized in the majority of applications. Using PCA, we made a quantitative model, which was used to identify the differences between the control and the treated tissue samples according to their FTIR spectra.
Origin Pro 2019b (OriginLab Corporation, Northampton, MA, USA) was used for the fitting of the vibrational bands.

Results
The PCA separated the BPC 157-treated and control samples on a PCA score-score graph. Using 100 spectra of both BPC-treated and control samples, the PCA model showed the scatterplot of the first two principal components, with a clear distinction, mostly due to the PC1 score, between the spectra of the BPC 157-treated group (green) and those of the control group (red) (Figure 1). The comparison of the averaged spectra of these two groups with their differe spectrum and PC loadings allowed us to identify the parts of the FTIR spectra that tributed the most to the spectral separation between BPC 157-treated rats and the co group (Figure 2). The two spectra (red and black) in the middle of the figure represen The comparison of the averaged spectra of these two groups with their differential spectrum and PC loadings allowed us to identify the parts of the FTIR spectra that contributed the most to the spectral separation between BPC 157-treated rats and the control group ( Figure 2). The two spectra (red and black) in the middle of the figure represent the average spectra of the BPC 157-treated and control groups, respectively. The differential spectrum of the average spectra (BPC-treated spectra minus control spectra) is presented at the bottom of the figure and the PC1 loadings at the top. Only the first principal component was used, because it exhibited the strongest difference between the two observed groups in the analysis of the PCA scores. The samples of the raw FTIR spectra are presented in Figure S1 in the supplementary materials.
Biomedicines 2022, 10, x FOR PEER REVIEW Figure 2. The average FTIR spectra of the control group (red line) and BPC 157-treated (bla group were obtained in the spectral range between 4000 and 500 cm -1 . The differential spec the average BPC 157-treated spectrum minus the control spectrum (grey line) and corresp PC1 loadings (blue line) are shown.
The PC1 loadings and the differential spectrum showed that the main bands af the separation of the BPC-treated and control sample spectra were the amid I band 1650 cm −1 , the amid II band around 1540 cm −1 , and the vibrational band around 174 There was also a small but observable difference in intensity in the spectral range CH vibrations of lipids between 2800 and 3000 cm −1 . Table 1 lists the assignment brational bands relevant to this paper. β-sheet of amide I ∼1641 Random coil of amide I Figure 2. The average FTIR spectra of the control group (red line) and BPC 157-treated (black line) group were obtained in the spectral range between 4000 and 500 cm −1 . The differential spectrum of the average BPC 157-treated spectrum minus the control spectrum (grey line) and corresponding PC1 loadings (blue line) are shown.
The PC1 loadings and the differential spectrum showed that the main bands affecting the separation of the BPC-treated and control sample spectra were the amid I band around 1650 cm −1 , the amid II band around 1540 cm −1 , and the vibrational band around 1744 cm −1 . There was also a small but observable difference in intensity in the spectral range of the CH vibrations of lipids between 2800 and 3000 cm −1 . Table 1 lists the assignments of vibrational bands relevant to this paper.
Then, to see the changes in the protein secondary structure, we fitted the spectral range between 1450 and 1800 cm −1 with three Gaussians. The amid I band consisted of four vibrational bands related to the secondary protein structure, β-sheet, Random coil, α-helix, and turns (Table 1). However, because of the lower signal-to-noise ratio in the case of the control spectra, we chose an approximate approach with three Gaussians (one each for amide I, amide II, and the band at 1744 cm −1 ) for the BPC 157-treated spectra and four Gaussians for the control spectra ( Figure 3). We used the fourth Gaussian for fitting the amid I band of the control spectra. Because the amide I band could not be successfully fitted, the fit did not converge; so, instead of one Gaussian, we had to apply two. The fitting parameters are presented in Table 2.   The fitting results presented differences in the BPC 157-treated and control group spectra ( Table 3). The vibrational band at 1744 cm −1 was substantially more intense in the control group. Changes in the protein secondary structure were also discovered. The average position of the amid II band in the control group was shifted toward lower wavenumbers, and the amid I vibrational band had to be fitted with two Gaussians instead of one, as in the case of the BPC 157-treated samples. The intensity values of the amid I and amid II vibrational bands were much lower in the control group spectra. Those changes in the amid I and amid II bands indicated changes in the protein secondary structure. All these changes are visible in the array of spectra presented in Figure S2 in the supplementary materials.  The fitting results presented differences in the BPC 157-treated and control group spectra ( Table 3). The vibrational band at 1744 cm −1 was substantially more intense in the control group. Changes in the protein secondary structure were also discovered. The average position of the amid II band in the control group was shifted toward lower wavenumbers, and the amid I vibrational band had to be fitted with two Gaussians instead of one, as in the case of the BPC 157-treated samples. The intensity values of the amid I and amid II vibrational bands were much lower in the control group spectra. Those changes in the amid I and amid II bands indicated changes in the protein secondary structure. All these changes are visible in the array of spectra presented in Figure S2 in the supplementary materials.
As an approach not employed before, we used the FTIR study of rat thoracic aortas, the changes in aortic tissues, and infrared spectroscopy to reveal the particular influence of BPC 157 on vascular tissue components. It might be that these timely in vivo and ex vivo findings correlate with the evidenced functional improvement of vessels. In general, this might indicate that particular changes might lead functionally improved minor vessels to take over the function of disabled major vessels to compensate otherwise irreparable vascular failure, i.e., vascular occlusion [12][13][14][15][16][17][18][19][20]26], compression [21], bile duct occlusion [25], or drug-induced vascular damage [22][23][24]. Based on this FTIR study, this therapeutic effect appeared to be complex. Thus, it might involve both changes in the lipid contents and in protein secondary structure conformation.
In general, the FTIR results showing significant differences between the BPC 157-treated and control group samples in the spectral region between 1450 and 1800 cm −1 (part of the spectrum below 1400 cm −1 was not reliable because it was under the strong influence of baseline correction) might be highly reliable. At this point, we used a PCA assessment to overlap the FTIR spectra of the biological samples, as they are often extremely similar and usually only advanced chemometric methods can reveal the possible distinctions between different pathological states.
We found that the amid I band around 1650 cm −1 , the amid II band around 1540 cm −1 , and the vibrational band around 1744 cm −1 were the main bands affecting the separation of the BPC 157-treated and control sample spectra. The amide II and amide I bands originated from the vibrations of the amide groups (CO-NH) of proteins [39], and the vibrational band at 1744 cm −1 corresponded to the non-hydrogen-bonded ester carbonyl C=O stretching mode within phospholipids. There were also smaller changes in the range between 2840 and 3090 cm −1 . This spectral range belonged to the symmetric C-H and asymmetric C-H, C-H 2 , and C-H 3 vibrations in lipids [36] ( Table 2).
The vibrational band at 1744 cm −1 , much more intense in the control group spectra, could be related to dying cells [40,41] (note that the period between sacrifice and FTIR assessment was 90 min). Moreover, we did not observe any significant change in other lipid vibrational bands, so the increase in the band at 1744 cm −1 was not caused by a simple increase in the number of lipid molecules [40]. The fact that the peak at 1744 cm −1 was significantly more intense than the peak at 1725 cm −1 could also have been related to cell death [40], as it implied that the C=O ester carbonyl groups of lipids in the cell had become predominantly non-hydrogen-bonded, suggesting that oxidative damage had occurred [40]. Therefore, the recorded spectra suggested that the tissue we observed was in an early stage of cell death.
Along with the above evidence (i.e., cells in control samples were in an early stage of cell death), there were also changes in protein conformation. The average position of the amid II band in the control group was shifted toward lower wavenumbers, and the amid I vibrational band had to be fitted with two Gaussians instead of one, as in the case of BPC 157-treated samples. The intensity values of the amid I and amid II vibrational bands were much lower in the control group spectra. Thus, the cells in BPC 157-treated samples were still protected.
Of note, in line with these specific findings on the vascular wall components, to resolve this interconnected entirety, it has been proposed that BPC 157 therapy might modulate vasomotor tone [5,[42][43][44][45]. Obviously, it might exert the described particular effects, given the relaxation noted in the aorta with endothelium as well as the relaxation noted in the aorta without endothelium ex vivo [44]. Possibly, these findings together might suggest additional particular points along with the evidenced release of NO on its own [5,[42][43][44][45], such as the activated phosphorilazation of eNOS [44], the counteraction of the adverse effect of the NOS blockade (i.e., L-NAME hypertension and pro-thrombotic effect) [8,42], and the counteraction of the adverse effect of NOS over-stimulation (i.e., L-arginine hypertension and anti-thrombotic effect) [8,42]. The VEGFR2-Akt-eNOS signaling pathway might be activated without the need for other known ligands or shear stress, controlling vasomotor tone and the activation of the Src-Caveolin-1-eNOS pathway [44,45]. These might also be perceived as BPC 157-specific mechanisms, in addition to direct NO stimulation.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/biomedicines10123130/s1, Figure S1: Example of two raw spectra: BPC treated and control group; Figure S2: Spectra of BPC treated and control samples after preprocessing.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.