Vibrational Molecular Spectroscopy as a Tool to Study Molecular Structure Features of Cool-Season Chickpeas Impacted by Varieties and Thermal Processing in Relation to Nutrient Availability in Ruminants

Simple Summary Feed molecular structure profile affects nutrient metabolism, utilization, and availability. Feed processing often induces feed internal structure change. These internal structure changes will affect animal nutrition. This study aimed to reveal the molecular structure features among chickpea varieties and detect the molecular structure changes induced by thermal processing methods. Our results show that with vibrational molecular spectroscopy, chickpea structure on a molecular basis was revealed in relation to ruminant nutrition. Abstract To our knowledge, there is no study on the relationship between molecular spectral features and nutrient availability in chickpeas. The purpose of this study was to reveal molecular structure spectral profiles among cool-season adapted CDC chickpea varieties and detect the molecular structure changes induced by thermal processing methods using vibrational Fourier-transform infrared (FTIR) spectroscopy. Three varieties of chickpea samples (CDC Alma, Cory, Frontier) were finely ground using a 0.12 mm screen. Spectral analyses were conducted using a JASCO FTIR-4200 spectroscope with Spectra Manager II software in the mid-infrared region from ca. 4000–800 cm−1 with a 4 cm−1 resolution. Data were analyzed using the “Mixed” procedure of SAS 9.4. Multiple regression was performed with PROC REG analysis for variable selection. Results showed that amide I area was higher (p = 0.038) in CDC Frontier than CDC Cory (30.85 vs. 24.64 AU). Amide I peak height (p = 0.028) was also higher in CDC Frontier and CDC Alma (0.45 AU in both) than CDC Cory (0.36 AU). Cellulosic compound (CEC) to total CHO (TCHO) area ratio was higher in CDC Frontier (0.05 AU) than the other two varieties (0.14 AU in both). As to thermal treatment impact, the results showed that total amide area was higher (p = 0.013) with autoclave and microwave heating (47.38 and 45.19 AU, respectively) than dry heating (33.06 AU). The CEC area was also higher (p < 0.001) for autoclave and microwave heating (3.74 and 3.61 AU, respectively) than dry heating (2.20 AU). Moreover, the ratio of amide I to II height was higher (p = 0.022) with microwave heating than dry heating (1.44 vs. 1.16 AU, respectively). Relationship analysis showed that the effective degraded crude protein (EDCP) and bypass dry matter (% BDM) were associated with STCHO peaks and CEC height (p < 0.05, R2 = 0.68). Also, feed milk value (FMVDVE) was associated with STC1, STC_A, and CEC_A (p < 0.05, R2 = 0.85). In conclusion, vibrational molecular spectroscopy mid-infrared FTIR was able to reveal different molecular spectral characteristics among the cool-season adapted CDC chickpea varieties and detect molecular structure changes induced by thermal processing (dry heating, autoclaving, and microwave heating).


Introduction
Infrared (IR) spectroscopy has been successfully used as an analytical technique in organic chemistry since the 1940s. It has been widely applied in the analysis of chemical composition as a rapid and simple operation to determine multinutrient conformation in a nondestructive and nonpollutive manner [1]. The basic principle of IR spectroscopy is that following exposure to IR radiation, chemical functional groups exhibit specific energy absorptions at certain frequencies, which enables the detection of chemical and structural differences in a variety of samples. The typical IR absorption peaks of the relevant biopolymers have been well documented [2].
One common method based on IR analysis includes vibrational Fourier-transform infrared (FTIR) spectroscopy. This well-established experimental technique is used to study structural composition, stability, and conformational changes, such as the effects of temperature, pH, and pressure in feed materials [3]. The mathematical method of Fourier transform converts the symmetric interferogram into functions with frequency components to form continuous transmittance or absorbance spectra [4]. In comparison with conventional dispersive spectroscopy, FTIR spectroscopy exhibits more effective and powerful properties due to its excellent sensitivity, larger optical throughput, and good signal-to-noise (S/N) ratio. Some advantages of this technique are the rapid scan speed and the ease of detecting the entire IR region simultaneously [4].
Several chemical functional groups can be detected using FTIR spectroscopy. For instance, protein molecular structure is unique in its peptide bonds, which contain C=O, C-N, and N-H functional groups. Amide-related groups show energy absorbance peaks at around 1700-1500 cm −1 . The amide I band is often used for protein structure analysis, as amide II usually overlaps other bands such as lignin at ca. 1515 cm −1 [5]. On the other hand, carbohydrate's molecular structure has many OH and CO bonds. Some infrared spectra related to carbohydrates appear at ca. 1200-800 cm −1 . Cellulose is mainly characterized around ca 1170-1150, 1050, and 1030 cm −1 and hemicellulose could be found at ca. 1732 and 1240 cm −1 [5,6]. Data obtained from spectral analysis lets researchers compute spectral ratios that represent the biological component ratio intensity and its distribution in the tissue. These ratios are obtained dividing the spectral height or area under one chemical functional group band (e.g., amide I) by the height or area under another functional group band (e.g., amide II) [7].
It is well stablished that thermal treatments inactivate antinutritional factors in legume seeds and improve their nutritional value. Additionally, heat processing alters not only the physical but also the intrinsic structures of feeds by (1) disrupting the protein configuration, making it more susceptible to digestive enzymes, (2) disrupting starch crystallites and promoting starch chain interactions within amorphous and crystalline areas, and (3) affecting hydration properties of fiber and modifying its physical properties [8]. As a result, these changes directly affect the functional properties, digestion, and absorption of several feeds.
The study of molecular structure characteristics of feedstuffs is hence important to stablish precise feeding techniques for ruminant systems. In our case, there is no systematic analysis on the relationship between molecular structure features of chickpeas and nutrient utilization and availability in ruminants. There is no study on the possible molecular structure changes induced by thermal processing in cool-season adapted chickpea varieties by using advanced vibrational molecular spectroscopy. Hence, the objectives of this study were to (1) study cool-season adapted CDC chickpeas molecular structure features and molecular structure changes induced by thermal processing and (2) reveal the association of molecular profiles with nutrient utilization and availability in ruminant systems. It was hypothesized that (1) the molecular intrinsic structure related to nutrient utilization and availability could significantly differ among CDC chickpea varieties and among thermal heat processing methods, and (2) there was an association between molecular structure features of chickpeas and nutrient utilization and availability in the ruminant system.

Materials and Methods
All cows involved in the present study were cared for in accordance with the guidelines of the Canadian Council on Animal Care [9]. The Animal Use Approval Protocol (19910012) was approved by the Animal Research Ethics Board (AREB) at the University of Saskatchewan (Saskatoon, SK, Canada).

Univariate Molecular Spectral Analysis of Functional Groups Related to Carbohydrates and Proteins
The FTIR analysis was performed using a JASCO-FTIR-4200 spectroscope (JASCO Corp, Tokyo, Japan) at the SRP Feed Research Chair Lab, Department of Animal and Poultry Science at the University of Saskatchewan (Saskatoon, SK, Canada). Molecular structure features of functional groups were analyzed in the mid-infrared region of ca. 4000-800 cm −1 . Five spectra for each sample were obtained with a resolution of 4 cm −1 using the software JASCO Spectra Manager II [3,5,23]. To identify chemical functional groups for carbohydrates and proteins, OMNIC 7.3 software (Spectra Tech, Madison, WI, USA) was used [3,5,23]. For detailed spectroscopic methods, please check our publications [3,5,23].
To detect responses and sensitivity of chemical functional groups to thermal processing methods and compare among varieties, various functional groups and ratios can be tested and analyzed. These functional groups include peaks centered at ca. 1740

Association between Molecular Structure Spectral Profiles and Nutrient Metabolic Characteristics of Protein and Carbohydrates
Multiple regression was performed to study molecular spectral features that could explain a variation in chemical profiles, protein and carbohydrate fractions, energy values, in situ rumen degradation, intestinal digestion, and truly absorbable nutrient supply. The detailed methods have been reported previously [3,5,23].

Statistical Analyses
Carbohydrate-and protein-related molecular structure spectral data were analyzed using the "Mixed" procedure of SAS 9.4 (SAS Institute, Inc., Cary, NC, USA). The model used for analysis was Y ijk = µ + T i + S(T i ) + e ijk , where Y ijk is the observation of the dependent variable ijk; µ represents the population mean of the variable; T i means the treatment effect as a fixed effect; and S(T i ) represents the subsample nested within treatments. In this case, five scans were made per sample, with e ijk representing random errors associated with the observation ijk. The model assumption was tested using the univariate procedure for residual analysis with normal and plot options. All significant analyses were declared at p < 0.05 and trends at 0.05 < p ≤ 0.10. The treatments were compared by using Tukey's comparison method.
The associations between molecular structure spectral parameters and chemical composition, rumen degradation, intestinal digestion, and truly absorbable nutrient supply were analyzed using SAS 9.4. Multiple regression study was carried out using the PROC REG procedure. The following model was used: Using STEPWISE for selection criteria: "SLENTRY = 0.05, SLSTAY = 0.05." Collinearity tests were performed using VIF to eliminate the influence of correlated dependent variables. All variables kept in as predictors were significant at the alpha 0.05 level, and models with an R 2 greater than 0.65 were selected to report in this study. Univariate procedures were used for residual analysis with normal and plot options.

Univariate Analysis of Molecular Structure Spectral Profiles in Different Varieties of CDC Chickpeas Grown in Western Canada
In this study, significant differences (p < 0.05) were observed in the protein-related spectral profiles among three CDC chickpea varieties: CDC Alma, CDC Cory, and CDC Frontier (Table 1). CDC Frontier and CDC Alma showed higher amide I and II peak heights than CDC Cory (p < 0.05). No differences (p < 0.05) among the three varieties were observed for cellulosic compounds (CEC) or total carbohydrates (TCHO) related spectral profiles. Peak ratios of STCHO: TCHO area were higher (p < 0.001) in CDC Alma and CDC Frontier (0.32 and 0.34 AU, respectively) than CDC Cory (0.29 AU). The CEC: TCHO area ratio was higher (p < 0.001) in CDC Frontier (0.05 AU) than CDC Alma and CDC Cory (0.04 AU for both). In a study from Sun et al. (2018) [23], it was reported that Kabuli and Desi varieties had higher amide I and peak height than barley grain, but no difference in many of the other measured spectral characteristics.

Univariate Analysis of Protein and Carbohydrate Related Molecular Structure Spectral Profiles Using Different Processing Methods
Molecular structure spectral characteristics of CDC chickpeas among thermal processing methods are presented in Table 2. Total amide area was larger (p = 0.013) when using autoclave and microwave heating (47.38 and 45.19 AU, respectively) than dry heating (33.06 AU). Amide I peak height was higher (p = 0.028) in CDC Alma and CDC Frontier (0.45 AU for both) than CDC Cory (0.36 AU). This indicates that thermal processing induced changes in protein-related spectral profiles, but the sensitivity and response to each processing method differed among treatments. Moreover, the absorbance for STCHO area was higher (p = 0.014) with dry heat and microwave treatments. Conversely, CEC area was lower (p < 0.01) with dry heat (2.20 AU) than autoclave and microwave heating (3.74 and 3.61 AU, respectively). These results also indicate that thermal processing induced changes in CHO-related molecular spectral profiles, but the sensitivity and response to each processing method differed among treatments.  Results related to peak ratios showed that the height peak ratio of amide I to II was higher (p = 0.022) with microwave than dry heat treatment (1.44 vs. 1.16 AU). No significant difference was observed in the other spectral peak ratios analyzed (p < 0.05).

Relationship between Protein and Carbohydrates Related Molecular Structure Features and Nutritional and Metabolic Characteristics of Protein and Carbohydrates
The spectral studies showed that several important nutritional parameters could be predicted using protein-and carbohydrate-related molecular spectral features. In rumen degradation kinetics study, the results (Table 3) showed that EDCP and % BDM were associated (p < 0.05) with molecular spectral variables related to STCHO peaks and CEC height (R2 = 0.68). In the chemical profile study (Table 4), ADF NDF , hemicellulose, and starch were highly associated (p < 0.05) with CEC_STC ratio (R 2 > 0.63), followed by TC-related peak heights and STC1. In the chemical and nutrient profile and predicted protein supply studies (Tables 5-7), the nutrient supply values were associated (p < 0.05) with spectral features of STC1, STC_A, and CEC_A (R 2 between 0.65 and 0.85).    [23], where the authors stated that FTIR molecular spectroscopy, a nondestructive bioanalytical technique, can be used to evaluate true nutrient supply using feed-inherent molecular spectral features for large quantities of feeds in a very short time.

Discussion
There is very limited published research on spectral features of chickpeas with which to compare our current findings. There is no systematic study on the relationship between molecular structure features of chickpeas and nutrient utilization and availability in ruminant system. There is no study on molecular structure changes induced by thermal processing in cool-season adapted CDC chickpea varieties revealed by advanced vibrational molecular spectroscopy either. Hence, this study provided an insight on specific molecular spectral features that might be associated with nutritional and digestive characteristics of CDC chickpeas.
In this study, heating treatments did impact both nutritional and molecular structure profiles of CDC chickpeas. The response and sensitivity to each thermal processing differed among dry heating, autoclaving, and microwave treatments. This agrees with the effects observed when moisture, pressure, or dry heating is applied to improve the nutritive value of feeds by physical and molecular modifications. Rodriguez-Espinosa et al. [27] indicated that vibrational spectroscopy can also be used for determining the possible alteration of structure during processing.
In this study, rumen degradation kinetics, intestinal digestibility and true nutrient supply to dairy cows were highly associated with molecular structure features in chickpeas. For example, feed milk value (FMV DVE ) was associated with three spectral variables of STC1, STC_A, and CEC_A (R 2 = 0.85, p < 0.01), and absorbed rumen undegraded protein (ARUP) was associated with spectral variables of STC1, STC_A, and CEC_A (R 2 = 0.72, RSD = 22.53, p < 0.01). Total true protein supply (DVE) to dairy cows and total metabolizable protein (MP) were associated with three spectral variables of STC1, STC_A and CEC_A (R 2 > 0.70, RSD = 21, p < 0.01). The studies conducted by Xin et al. [28] (2014) and Xin and Yu [29] showed that the spectral features are correlated with nutrient values. However, in studies conducted by Xin and Yu [30,31] to compare the alteration of spectral profiles of canola and Brassica carinata during microbial digestion, chemical profiles were used to correlate with structural change, which cannot reflect the utilization and digestion condition of the feed.
Molecular spectral analysis is a useful method to associate with nutrient values. The spectral variable analysis showed that with molecular spectroscopic technique-FTIR, the true nutrient supply to dairy cows was associated with a few specific molecular spectral parameters in the vibrational mid-infrared region-ca. 4000-800 cm −1 .
These results show us the potential of the molecular spectroscopic technique of FTIR in using spectral parameters in the vibrational mid-infrared region to predict the nutrient supply in ruminant systems. The implication of this study is that feed nutritional value is not only related to total chemical composition but also inherent molecular structure. Future NRC models or any other feed evaluation systems could consider feed molecular structure features as an important part of developing modern and improved nutrition models.

Conclusions
In conclusion, CDC Frontier and CDC Alma showed higher peak heights in proteinrelated profiles. However, CDC Alma, CDC Cory, and CDC Frontier tended to differ in peak area of cellulosic compound profiles. Area of protein-related molecular spectral profiles differed among dry heat, autoclave, and microwave treatments in amide I and amide II. STCHO peak heights were higher in the 1st and 3rd peaks for dry heat treatment. Greater values were observed in the CEC spectral area in autoclave and microwave treatments (average 3.65 AU) and lower absorbance in dry heat treatment with 2.20 AU. The peak height ratio in the amide I to II area was higher than in the microwave treatment.
Vibrational molecular mid-infrared FTIR spectroscopy was able to reveal different molecular spectral characteristics among the cool-season adapted CDC chickpea varieties and detect molecular structure changes induced by thermal processing (dry heating, autoclaving, and microwave heating). There is an association between molecular structure spectral features of chickpeas and nutrient availability in ruminant systems.