NMR Metabolome-Based Classification of Cymbopogon Species: a Prospect for Phyto-equivalency of its Different Accessions Using Chemometric Tools

Cymbopogon species are widely distributed worldwide and known for their high essential oil content with potential commercial and medicinal benefits justifying for their inclusion in food and cosmetics. Most species received scant characterization regarding their full complement of bioactive constituents necessary to explain their medicinal activities. In this study, the metabolite profiles of 5 Cymbopogon species, C. citratus, C. flexuosus, C. procerus, C. martini, and C. nardus, were characterized via NMR-based metabolomics. The results of 13 shoot accessions revealed the identification and quantification of 23 primary and secondary metabolites belonging to various compound classes. Multivariate analyses were used for species classification, though found not successful in discrimination based on geographical origin. Nevertheless, C. citratus was found particularly enriched in neral, geranial, (E)-aconitic acid, isoorientin, and caffeic acid as the major characterizing metabolites compared to other species, while an unknown apigenin derivative appeared to discriminate C. martini. The high essential oil and phenolic content in C. citratus emphasizes its strong antioxidant activity, whereas (E)-aconitic acid accounts for its traditional use as insecticide. This study affords the first insight into metabolite compositional differences among Cymbopogon species. Moreover, antimicrobial, insecticidal, antidiabetic, and antioxidant compounds were identified that can be utilized as biomarkers for species authentication.


Introduction
Medicinal plants play significant roles in public health care (Avoseh et al. 2015). With an increasing interest in the use of herbal drugs, 60% of the world population are considered to rely on medicinal plants as therapeutic agents (Balakrishnan et al. 2014). Among medicinal herbs with outstanding record in ayurvedic therapy are Cymbopogon (C.) species, e.g., C. citratus (lemongrass) has widespread use in folk medicine for the treatment of inflammation, digestive disorders, fever, and diabetes (Shah et al. 2011).
The genus Cymbopogon (Poaceae), comprising 54 species, is dispersed in the tropical and subtropical regions of the world of which C. citratus, C. pendulus, and C. flexuosus are widely distributed (Avoseh et al. 2015). The genus name originated from the Greek word "kymbe-pogon" which means boat-beard referring to its many-awned inflorescences and boat-shaped spathes (Shah et al. 2011).
Most phytochemical reports on Cymbopogon have focused on C. citratus, with less emphasis on other species within that genus. Most studies have been also related to the essential oil composition (Cerceau et al. 2020) with limited research on other secondary bioactives likely to mediate for the genus biological effects. Numerous pharmacological activities were reported for the genus including anti-inflammatory, antitumor, antioxidant, antimicrobial, cardio-protective, and antirheumatic. Effects attributed to its essential oil include antiprotozoal, insecticidal, antidiabetic, antitumor, and anti-inflammatory activities (Avoseh et al. 2015;Ekpenyong et al. 2015).
To better correlate between these biological effects and metabolite profile, a holistic approach is warranted to unveil the metabolite composition and level for further standardization and quality control measures.
For functional food analysis, several modern approaches are increasingly employed including that of large-scale metabolomics aiming at the detailed characterization of metabolites within food specimens. Metabolomics is considered a fast and reproducible method to gain insight into the metabolome in different biological materials (Chen et al. 2013;Fiehn 2002). It is extensively applied for exploring the effect of growth stage; processing method, seasonal variation, or storage conditions on the plant metabolome; detecting biomarkers of cultivars from different geographical origins; or verifying the quality and safety of new food products (Huo et al. 2017;Klockmann et al. 2017;Patiño-Rodríguez et al. 2018).
Mass spectrometry and nuclear magnetic resonance are the most common analytical techniques utilized in metabolomics. NMR spectroscopy offers a powerful tool for rapid, simple, reproducible, and non-destructive measurements, and unlike other techniques, it shows high accuracy and selectivity (Girelli et al. 2018;Yuan et al. 2017). It has been broadly used in metabolomic analysis of ginseng roots, Passiflora leaves, and date palm by-products (Farag et al. 2016;Otify et al. 2019;Yang et al. 2012). Moreover, NMR has been satisfactorily utilized for quantification in different food products dismissing the need of standard response measurements (Farag et al. 2018). Despite the high potentiality of this technique, NMR-based metabolomic investigations on Cymbopogon species are still scarce (Abdelsalam et al. 2017).
Our goal is to investigate variations in primary and secondary metabolites from different Cymbopogon species in terms of genotypes to provide the first insight into chemotaxonomic relatedness. To achieve our aim, metabolite profiling and fingerprinting, using a NMR approach, was employed for the first-time analysis of the official C. citratus shoots from 2 geographical origins as well as 4 other Cymbopogon species (a total of 13 samples, Table 1). 1 H-NMR was further utilized to determine the absolute concentrations of the identified metabolites in sample extracts for future standardization purposes. The main advantage of NMR quantification is that signal integration of the compound is directly proportional to its molar concentration, making such technique well suitable for quality control or adulteration detection.
Due to the complexity of acquired data represented by large specimen number and variables, i.e., chemical shifts as typical in case of NMR metabolomics, multivariate statistical analyses are usually performed to organize NMR datasets. The combination of multivariate analysis and NMR provides a more powerful technique for comparing the chemical profiling of different Cymbopogon species.
Such information is deemed to be of interest for the quality evaluation of Cymbopogon in the future and to determine metabolite heterogeneity among its different accessions. Additionally, quantification of the major metabolites detected using NMR is presented for standardization of its extracts.
Hexamethyldisiloxane (HMDS) and methanol-d 4 (99.80% D) were supplied from Deutero GmbH (Kastellaun, Germany). All other chemicals were obtained from Sigma Aldrich (St. Louis, MO, USA). For NMR quantification and calibration of chemical shift, HMDS was added to a final concentration of 0.94 mM.

Sample Preparation for NMR Analysis
Sample extraction followed the protocol described in Farag et al. (2015). Briefly, 130 mg of dried shoot powder was homogenized with 5 ml 100% methanol using a Turrax mixer (11,000 RPM) for five 20-s periods, with 1-min interval to prevent warming. Extracts were then vigorously vortexed and centrifuged (3000 g for 30 min) to remove plant debris. Three milliliters were then aliquoted by a syringe, and the solvent was evaporated under nitrogen to dryness. Dried extracts were re-suspended with 700 µl 100% methanol-d 4 containing 0.94 mM hexamethyldisiloxane (HMDS) as an internal chemical shift NMR standard. The supernatant was centrifuged (13,000 g for 1 min) and transferred to a 5-mm NMR tube. All 1 H-NMR spectra for multivariate data analysis were obtained successively within a 48-h time interval with samples prepared directly before data acquisition. Repeated control experiments after 48 h showed no additional variation. Three biological replicates for each specimen were extracted and analyzed in parallel under the same conditions to assess for biological variance.

NMR Data Acquisition
All 1 H-NMR spectra were recorded on an Agilent VNMRS 600 NMR spectrometer operating at a proton NMR frequency of 599.83 MHz equipped with a 5-mm inverse detection cryoprobe, digital resolution 0.367 Hz/point (32 k complex data points), pulse width (pw) = 2.1 μs (30°), relaxation delay = 18 s, acquisition time = 2.0 s, number of transients = 160, and temperature = 297 K. Zero filling up to 128 k and an exponential window function with lb = 0.4 were used prior to Fourier transformation. 2D-NMR spectra were recorded at a frequency of 599.83 MHz using standard CHEMPACK 6.2 pulse sequences (COSY, HSQC, HMBC) implemented in standard VNMRJ 4.0A spectrometer software. The HSQC experiment was optimized for 1 J CH = 146 Hz with DEPT-like editing and 13 C-decoupling during acquisition time. The HMBC experiment was optimized for a long-range coupling of 8 Hz; a 2-step 1 J CH filter was used (130-165 Hz).

NMR Data Processing and Multivariate Data Analyses
Spectra were imported to ACD/NMR Manager lab version 10.0 software (Toronto, Canada) and automatically Fourier transformed to ESP files. The spectra were referenced to internal HMDS at 0.062 ppm for 1 H-NMR and 1.96 ppm for 13 C-NMR, respectively. Spectral intensities were reduced to integrated regions, referred to as buckets, of equal width (0.04 ppm) for all spectral (δ 0.4-11.0 ppm) and aromatic (δ 5.5-11.0 ppm) regions. The spectral regions corresponding to the residual solvent signals, δ 4.90-4.80 (water) and δ 3.33-3.28 ppm (methanol), were removed before multivariate analyses. This binning allowed to evaluate the absolute quantification of the identified metabolites. Table of buckets was imported into SIMCA-P version 13.0 (Umetrics, Umea, Sweden), and hierarchical cluster analysis (HCA), principal component analysis (PCA), and orthogonal projection to latent structure-discriminant analysis (OPLS-DA) were performed with all variables mean centered and scaled to Pareto variance. Seven-fold cross-validation method was applied for selecting the optimum number of principal components required for modeling the data, and distance to the model (DModX) test was used for verifying the presence of outliers. Validation of the developed OPLS-DA models was verified using permutation tests, receiver operating characteristic (ROC) curve, and CV-ANOVA (ANOVA of cross-validated residuals). Metabolite markers were then recognized by analyzing the S-plot, which was declared with covariance (p) and correlation (pcor) in addition to the variable influence in the projection (VIP).

Quantification of Major Metabolites via 1H-NMR
For the quantification of metabolites listed in Table S2 using NMR spectroscopy, the peak area of selected proton signals belonging to the target compounds and the peak area of the internal standard (HMDS) were integrated manually for all the samples. The following equation was applied for the calculations:

Statistical Analysis
NMR quantification data were analyzed using the Co-Stat version 8 software (Monterey, CA, USA). Data are expressed as mean ± SD of the groups. One-way ANOVA followed by Student-Newman-Keuls tests were used to determine significant differences among Cymbopogon sample groups, with 95% confidence limit. Differences were considered statistically significant when p ≤ 0.05.

Phenolic Acids
Phenolic acids represent a common class of plant secondary metabolites, and like flavonoids, they contribute mainly to the antioxidant activity of food materials aside for their taste. NMR analysis revealed the presence of 2 phenolic acids (N19 and N20). In detail, caffeic acid (N19) was identified from the characteristic doublets of the olefinic protons H-2 and H-3 exerting distinct cross-peaks in 1 H-1 H COSY experiment and appearing at δ 6.29 and 6.57 (each d, J = 16.0 Hz) ppm indicative of E-oriented olefinic protons correlated with the 13 C signals at δ 115.8 and 147.7 ppm, respectively (Table S1). Moreover, the 1 H-NMR spectrum also revealed an ABX-ring system from signals resonating at δ 7.05 (d, J = 2.0 Hz), 6.78 (overlapped), and 6.95 (dd, J = 8.0, 2.0 Hz) ppm corresponding to H-2′, H-5′, and H-6′, respectively (Fig. 2b). A distinct aromatic singlet at δ 6.99 ppm (H-2′/H-6′) showing HSQC cross peak with 13 C at δ 117.1 ppm and HMBC long correlation with carboxylate carbon at δ 166.0 ppm (Figs. S6 and S7) identified compound N20 as gallic acid. Assignments of all phenolic acids (N19 and N20) were confirmed via HMBC spectral data (Fig. S7).

Alkaloids
Trigonelline (N21) was identified in the down-field region (δ 8.0-9.5 ppm) showing 4 aromatic signals: a doublet of doublet at δ 8.05 (overlapped) ppm, two doublets at δ 8.86 and 8.90 (J = 6.3 Hz) ppm, and a singlet at δ 9.18 ppm (Fig. S11). A final singlet resonating at δ 4.43 ppm ascribed to the N-methyl group confirmed the structure of trigonelline. All values were supported via COSY and HMBC correlations and were consistent with those reported in the literature (Campos et al. 2014) (Figs. S12 and S13).

Monoterpenes
The key aroma compound of Cymbopogon was detected in nearly all shoots identified as citral (a mixture of neral and geranial). In the region from δ 9.5 to 10.0 ppm and from 5.0 to 6.0 ppm, signals were assigned to neral (N22) and geranial (N23), two unsaturated aldehydic monoterpenes characteristic of Cymbopogon species (Mannina et al. 2012). The two compounds could be readily distinguished by chemical shifts of the aldehyde protons at δ 9.84 (d, J = 9.9 Hz) and 9.92 (d, J = 8.1 Hz) ppm, respectively, and the methyl groups in cis or trans position at δ 2.01 and 2.20 ppm, respectively (Table S1 and Figs. S11, S13, and S14). Compared to previous NMR studies directed mainly to essential oil analysis, quality, or detection of its of adulteration, this study provides the first comprehensive NMR metabolites fingerprinting of 5 Cymbopogon species. The employed NMR technique revealed the identification of 23 metabolites of important nutritional, flavoring, and biological significance and belonging to various classes, i.e., fatty, amino, and organic acids, sugars, triterpenes and sterols, phenolics compounds, alkaloids, and volatile monoterpenes.

Multivariate Data Analysis of NMR Dataset
A total of 39 samples was employed in this study (13 accessions; each has 3 biological replicates); this warranted for the employment of multivariate data analysis for detecting the trends among the specimens. PCA and HCA were conducted to obtain a global overview of all investigated Cymbopogon samples and to distinguish between their corresponding metabolic profiles. PCA was performed considering NMR datasets in two attempts, from all spectral width (δ 0.4-11.0 ppm) and from the aromatic region only (δ 5.5-11.0 ppm) to eventually focus on secondary metabolites in Cymbopogon, i.e., phenolic acids, alkaloids, terpenoids, and flavonoids (Farag et al. 2012).
In the first PCA model (δ 0.4-11.0 ppm), several Cymbopogon samples showed considerable overlap in NMR score plot with PC1 and PC2 components accounting for 45% of the total variance (Fig. S15). Such model showed weak discrimination power with even biological replicates within the same sample failing to group together. Hence, a second PCA model considering only the aromatic region (δ 5.5-11.0 ppm) was applied yielding a model prescribed by two components (PC1 and PC2) and explaining 54% of the total variance (Fig. 3). No clear variation could also be seen with respect to the sample cultivars as evident from PCA score plot (Fig. 3a), with C. citratus samples from 2 different origins (Germany and Egypt) failing to cluster together and spreading along PC1. Similar results were also observed from HCA dendrogram (Fig. 3c). Such models confirmed the unsupervised models' inability to determine growing habitat effect on Cymbopogon chemical composition. Nevertheless, the Egyptian C. citratus appeared most distant from all other species with negative PC1 score values. Additionally, a distinct cluster of C. martini shoots was also observed as an outlier negatively to both PC1 and PC2 (Fig. 3a).
Such partial segregation observed in the second PCA model scores could be attributed to some discriminatory NMR signals as evident from the loading plot (Fig. 3b). In detail, high levels of neral, geranial (δ 5.84 ppm) and (E)-aconitic acid (δ 6.87 ppm) levels were observed in most Cymbopogon species, whereas high isoorientin (δ 6.50, 6.55, and 7.38 ppm) and caffeic acid (δ 6.78 ppm) content was detected in Egyptian C. citratus specimens. The signals discriminating the C. martini samples could not conclude a confirmed structure, yet, these signals seem to belong to an unknown apigenin derivative, most probably a C-glycoside, a characteristic class in genus Cymbopogon (Avoseh et al. 2015). This is evident from the singlet resonating at δ 6.62 ppm (H-3), the 2 ortho-coupled doublets at δ 6.94 (H-3′, H-5′) and 7.95 (H-3′, H-5′) ppm of ring B, and the absence of H-6 and H-8 protons suggesting their substitution by two sugar moieties (Fig. S16) Table 1 for origin of Cymbopogon shoot samples. b Load-ing plot for PC1 and PC2 showing signals of isoorientin and caffeic acid as the major signals contributing to sample discrimination. c HCA of Cymbopogon species, the model is colored according to marked groups of sugars and the exact structure need further confirmation using other spectral tools and post isolation,to enrich its levels.
OPLS-DA models were also attempted to help in sample discrimination that failed during unsupervised analysis, with special emphasis on modeling C. citratus versus other Cymbopogon samples. OPLS-DA is a supervised classification technique having a great potential in finding the maximum separation among overlapping sample groups and identifying differential biomarkers (Bylesjö et al., 2006). The first OPLS-DA model was performed for the whole NMR range (δ 0.4-11.0 ppm) with all C. citratus samples modeled against other C. species (Figs. S17 and S18). Interestingly, a clear separation of C. citratus from other Cymbopogon samples was observed along the predictive component (Fig. S17a). The metabolites that influence such pattern can be verified from the loadings S-plot (Fig. S17b) and VIP plot (Fig. S17c), where C. citratus samples were found to be particularly enriched in (E)-aconitic acid and betaine as the major characterizing metabolites compared to other Cymbopogon species. Besides, ROC curve was also performed to demonstrate the model classification ability. The area under the ROC curve (AUC) was considered for C. citratus group as a validation criterion for its classification and was found to be 0.996 (Fig. S17d), indicating an effective classification model. Other validation procedures include permutation tests that showed negative Q2 intercept value (Fig. S18c) and CV-ANOVA with p value below 0.05 (Fig. S18d) typical for valid models. This predictive model also has a fitness of data R2(X) of 76% and a cumulative sum of variation R2(Y) of 89% with predictive ability Q2 of 78% (Fig. S18a). However, one of the major pitfalls regarding this model is the large number of principal components, i.e., one predictive and 5 orthogonal components required to obtain such validity based on the cross-validation results with root mean square error of cross-validation (RMSECV) of 0.23 (Fig. S18b).
A second OPLS-DA model for C. citratus versus other Cymbopogon samples was also developed considering only the aromatic region (δ 5.5-11.0 ppm) and focusing on the plant secondary metabolites (Figs. S19 and S20). Interestingly, it was found that one predictive and two orthogonal components were sufficient to model the data based on the cross-validation results with R2(X) of 60%, R2(Y) of 81%, Q2 of 72%, and RMSECV of 0.26 (Figs. S20a and S20b). Increasing the number of orthogonal components to three has led to overfitting of the developed model with decreasing in the Q2 and increasing in the RMSECV values (Figs. S20a and S20b) ensuring that one predictive and two orthogonal components were sufficient for data modeling. The score plot derived from this OPLS-DA model showed a clear segregation between both sample groups, along the predictive component (Fig. S19a), while the corresponding S-plot loadings (Fig. S19b) and VIP plot (Fig. S19c) highlighted for the enrichment of C. citratus samples in neral, geranial, (E)-aconitic acid, isoorientin, and caffeic acid compared to other Cymbopogon species. The high essential oil and phenolic content in C. citratus highlights its valuable antioxidant activity and rationalizes for its commercial and medicinal use among Cymbopogon genus. Besides, ROC curve obtained for this model passes through the upper left corner (100% selectivity, 100% sensitivity) with AUC value for C. citratus group of 1 indicative for better classification model (Fig. S19d). Permutation test with 200 times and CV-ANOVA were conducted for evaluating whether the model is over fitted (Figs. S20c and S20d), with negative Q2 intercept value and p value below 0.05 proving the model validity.

Quantification of Major Metabolites via 1 H-NMR
To ensure quality control of different Cymbopogon products, precise metabolites measurement is warranted. 1 H-NMR was further used to measure the absolute levels of identified metabolites in Cymbopogon species via integration of their well-determined corresponding peaks in the NMR spectra (see experimental section). The concentration of metabolites was calculated as μg/mg dry powder in different samples as shown in Table S2.
Sugars amounted to the major metabolites in all species with maximal levels found in C. citratus extracts (104.1 ± 4.3 μg/mg total sugars) and with glucose (in its αand β-forms) representing the major sugar in most samples. The high sugar content adds to the nutritional value and palatable taste of C. citratus.
The abundance of the unsaturated fatty acids in all Cymbopogon species (up to 26.8 ± 2.2 and 25.3 ± 3.1 μg/mg in C. martini and C. nardus, respectively) poses the genus as a healthy functional food product due to their potential to decrease serum LDL cholesterol (Liang and Liao 1992;O'Brien 2009).
Concerning non-essential amino acids, alanine was almost equal in all shoots (1.4 ± 0.2-2.9 ± 0.8 μg/mg), while asparagine was found at comparable levels in the examined species (8.1 ± 0.2-23.3 ± 3.2 μg/mg). Phenylalanine, an essential amino acid, was also quantified at nearly the same level in the different samples (1.1 ± 0.1-1.5 ± 0.4 μg/mg). However, valine could not be quantified in most examined specimens owing to signals overlap.
Total choline and betaine level was quantified in all sample extracts reaching ca. 6.3 ± 3.7 μg/mg in C. citratus, albeit found at its lowest concentration in C. flexuosus (0.9 ± 0.1 μg/mg). Both compounds have anti-inflammatory and antidiabetic actions and their high levels in C. citratus rationalize for its use in inflammation and diabetes (Chung et al. 2015;Gao et al. 2017;Zhao et al. 2018).
Regarding the phenolic metabolites, contributing mainly to the antioxidant potential of food, gallic acid was found with nearly equal content in all species (1.0 ± 0.1-2.4 ± 0.1 μg/mg). However, isoorientin and caffeic acid could not be quantified in all samples due to signals overlap.
The key aroma compounds, neral and geranial, responsible for the lemony flavor in Cymbopogon species were quantified in nearly all shoots presenting them as important agents in food and flavor industry (Carlson et al. 2001). A previous study reported that the potential antibacterial action of C. citratus essential oil resides mainly in these two components (Onawunmi et al. 1984). Both volatiles were found at maximal levels (8.6 ± 1.0 and 10.6 ± 1.0 μg/mg, respectively) in C. citratus, yet absent in C. martini and Egyptian C. citratus cultivar, justifying their appearance most distant from all other Cymbopogon species in foregoing multivariate analysis results (Fig. 3a).
Cymbopogonol, as an important antimicrobial nonvolatile terpenoid, was quantified in all samples reaching its highest level in C. martini and C. citratus (20.3 ± 0.8 and 19.2 ± 5.0 μg/mg, respectively). This high cymbopogonol content together with other essential oil constituents may explain the strong antimicrobial activity of C. martini although found to be poor in neral and geranial content (Gemeda et al. 2018;Ragasa et al. 2008).
Trigonelline was found at comparable levels (1.0 ± 0.9-3.1 ± 0.8 μg/mg) in nearly all examined species. Research studies revealed that trigonelline exhibited significant hypoglycemic activity in both animal models (50 mg/kg) and humans (500 mg oral daily dose) (Mishkinsky et al. 1967;Sharma et al. 1990;Zhou et al. 2012). Moreover, the antidiabetic action of fenugreek seed, well-known for its high trigonelline content (0.13-0.36%), has been attributed to its trigonelline as a major hypoglycemic agent (Barnes et al. 2002;Mishkinsky et al. 1974;O'Neil 2001). Trigonelline levels quantified in Cymbopogon species were close to those found in fenugreek seed and thus strongly suggesting that trigonelline contributes mainly to the antidiabetic activity reported for many Cymbopogon species (Kamble et al. 2020;Kouame et al. 2016).
To the best of our knowledge and compared to previous NMR studies directed mainly to essential oil analysis, this study provides the first comprehensive NMR metabolite fingerprinting and standardization of 5 Cymbopogon species from 2 different biological origins.

Conclusions
This study reports a comprehensive NMR metabolic profiling study of 5 Cymbopogon species used extensively in folk medicine, yet with limited phytochemical characterization or standardization. NMR coupled with multivariate data analyses was further employed for sample classification and suggests that geographical origin cannot be revealed from the NMR dataset in any of the examined Cymbopogon specimens. Only differential metabolites of C. citratus and C. martini could be identified with neral, geranial, (E)-aconitic acid, isoorientin, and caffeic acid as the major discriminating metabolites for C. citratus, while an unknown apigenin derivative appeared to distinguish C. martini. The enrichment of C. citratus in essential oil, phenolic compounds, and (E)-aconitic acid justified its preferred commercial use in food and cosmetics and further highlighted its medicinal potential as antioxidant and insecticidal agent in comparison to other Cymbopogon species.
Although the results here represent a preliminary report in terms of finding potential specific biomarkers related to Cymbopogon species, insights gained by the present work, as the first one based on NMR metabolic profiling of genus Cymbopogon, are useful for further species or cultivars as well as employing other analytical investigations. Further research can be done by enlarging the sample size and by covering more cultivation areas to validate these exploratory results. In addition, an extended approach utilizing liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry can be applied to pinpoint differences in bioactive secondary metabolite profiles among Cymbopogon accessions and to rationalize for more of its flavor composition and health effects in more depth. Quantification of key metabolites using NMR provides the first basis for Cymbopogon shoot standardization which can be used further for their estimation when present in complex drug mixtures. Moreover, the effect of seasonal variation, drying method, and storage conditions of Cymbopogon shoots on their metabolite composition have yet to be addressed using this platform.
Funding Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB).
Data Availability Materials are not available, and data is available from authors upon request.
Code Availability Not applicable.

Declarations
Ethical Approval This article does not contain any studies with human participants or animals performed by any of the authors.

Conflict of Interest
Asmaa M. Otify declares that she has no conflict of interest. Ahmed Serag declares that he has no conflict of interest. Andrea Porzel declares that she has no conflict of interest. Ludger A. Wessjohann declares that he has no conflict of interest. Mohamed A. Farag declares that he has no conflict of interest.

Informed Consent Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.