Original Paper An inexpensive and environmentally friendly staining method for semi-permanent slides from plant material probed using anatomical and computational chemistry analyses

One of the main methods for plant anatomy study is the analysis of thin, transparent, and stained tissue sections. Synthetic dyes traditionally used in anatomical studies might be expensive and produced by specific companies. In contrast, the use of alternative industrial dyes can both represent an inexpensive substitute as well as an environmentally friendly option for conducting plant anatomy studies. In this study, a set of 22 textile dyes was evaluated. Transversal-, longitudinal, and paradermal sections of plant organs obtained using the freehand cutting technique were stained using hydroalcoholic solution (0 to 100%) of textile dyes purchased from a local market. Dyes mixed with 50% hydroalcoholic solution showed higher efficiency in tissue contrast, allowing greater solubility of dye powder and better solution interaction with the plant tissues. Most of the tested dyes showed satisfactory staining results. Cell wall, especially lignified one, showed higher staining efficiency. Computational docking analysis and molecular models of cellulose and lignin showed the probable association mechanisms and dye selectivity to cell wall constituents. Our findings suggest that the developed method can be useful in mixed practical classes of plant anatomy, chemistry, and/or biochemistry, both at high school as well as undergraduate levels.


Introduction
Plant anatomy is an important tool to allow a realistic interpretation of morphology, physiology, and phylogeny and to understand ecological processes. It has been proved the most useful in combined studies to better understand the phylogenetic relationships between different plant groups. Furthermore, knowledge of plant structure is essential to address many important everyday problems such as the identification of unknown food contaminants and forensic problems (Metcalfe 1961;Yeung 1998;Dengler 2002;Bock & Norris 2016).
One of the main methods for studying plant anatomy is the analysis of thin and transparent tissue sections. This facilitates the observation and interpretation of plants' tissues (Macedo 1997). The interpretation of any colorless plant material can be improved by the application of dyes to provide contrast between tissues, facilitating their visualization under optical microscopy and revealing their specific cell structures (Kierman 1990;Kraus & Arduim 1997).
Dyes are substances that impart color to a substrate via a process that alters any crystal structure of colored substances. Dyes are classified according to their application and chemical structure. They comprise a group of atoms known as chromophores, which are responsible for the dye color. These chromophore-containing centers have diverse functional groups such as azo, anthraquinone, methine, nitro, aril-methane, and carbonyl (Dos Santos et al. 2007;Arun & Bhaskara 2010;Bafana et al. 2011).
Staining generally occurs via physical and/ or chemical processes. Dyes can be grouped into three classes: dyes that stain the acidic and basic cell components, specialty colorants, and metal salts (Gartner & Hiatt 1999;Geneser 2003;Rohde et al. 2006). In histological staining, the most widely used dyes are the basic and acid dyes that react, respectively, with cell acidic and basic components (Walwyn et al. 2004).
Traditionally, various synthetic dyes have been used for staining plant structures, for instance, alcian blue, safranin, toluidine blue, fast green, crystal violet, iodine green, Congo red, and neutral red (Dop & Gautié 1928;Kirk 1970;Bukatsch 1972;O'Brien & McCully 1981). Most of these dyes can be expensive and produced by specific companies.
Studies have attempted to use alternative dyes in histological studies in animals (Rohde et al. 2006); however, to our knowledge, no study has reported the use of alternative dyes in plant anatomy analysis. Thus, studies on identifying alternative dyes that are affordable for students and teachers at various education levels might facilitate and widen the scope of plant anatomy studies.
Textile dyes are listed as the most hazardous industrial pollutants; they are highly visible and undesirable in effluents even in very low concentrations (Robinson et al. 2001;Hubbe et al. 2012;Vikrant et al. 2018). In fact, the complete degradation of chemicals commonly used in such dyes involves onerous and expensive processes, and water and soil pollution caused by their occasional discharge in effluents is a critical environmental problem at a global scale (Khan & Malik 2018;Vikrant et al. 2018). Thus, new alternative applications of such dyes need to be explored while considering the economic and environmental sustainability (Marcucci et al. 2001;Granato et al. 2017).
Therefore, this study aimed to develop a lowcost and environmentally promising alternative by using textile dyes that can interact with plant tissues, to create contrast between cell structures. In addition to testing the efficiency of alternative textile dyes for staining anatomical sections, we elucidated the molecular mechanisms of interaction of representative compounds commonly found in dyes to develop computational models of archetypal cell wall polymers (crystalline cellulose and lignin models) by using virtual docking methodologies (Martínez et al. 2010;Yuriev et al. 2015).
Our results corroborate the high potential of using textile dyes for plant tissue staining, as well as the possibility of using both anatomical and computational methodologies for practical and didactic purposes.

Plant material
The samples (Tab. 1) were prepared using the common techniques used in plant anatomy studies. Fresh stems, leaves, and roots were sectioned using disposable razors and polystyrene as a support (Ruzin 1999). Wood fragments were macerated using Jeffrey solution (Johansen 1940). Leaf diaphanization was performed according to the technique of Kraus & Arduim (1997).
Sections were cleared in sodium hypochlorite (5%) until they became translucent. Subsequently, they were thoroughly washed with distilled water. Some sections were rinsed in distilled water containing drops of acetic acid, as indicated by Macedo (1997), to check for possible improvement in the results.

Preparation of staining solutions
Commercial textile dyes (powder; Tab. 2) were obtained from local markets. The material safety data sheet (MSDS) from the manufacturer suggested that the dyes used in this study were a mixture of pigments containing azo/stilbene groups, without specifying the chemical structure of the dye components. About 10% (weight/volume) of each textile dye in a series of dilutions of hydroalcoholic solution (0%−100%) was prepared at room temperature and heated at 100 o C. The pH was adjusted to acidic (pH = 4.0) and basic (pH = 11.0). The efficiency of the use of mordants in the interactions of dyes with plant tissues, as proposed by Johansen (1940), was analyzed by adding 0.2 g sodium acetate + 0.4 mL of formaldehyde to the dye solutions. Mordants were also prepared in 3% acetic acid solution.

Staining procedure
The sections were dehydrated in hydroalcoholic solution until the required concentration of dye was achieved. The material was stained for 05−60 min and washed with hydroalcoholic solution at the same concentration of the dye solution with varying pH (4.0 and 11.0; Macedo 1997). Preliminary tests were performed using a mixture of two dyes containing 50% hydroalcoholic solution. The macerated plant material was stored in a dye solution for six days.

Treatment after staining
The dye manufacturers also provide a "fixative" (cross-linked amide resin) that, according to them, "improves the fixation of natural dyeing." This fixative was prepared in 50% hydroalcoholic solution (1:1) and applied for 1 min after staining.

Mounting medium
All preparations were mounted in glycerinated gelatin according to Kaiser (1880) with the following modifications: (i) gelatin (colorless and unflavored) and glycerin were obtained from local markets; (ii) phenol was substituted by five commercial Indian cloves buds (Syzygium aromaticum (L.) Merr. & L.M. Perry), and incubation was performed overnight.
Optical analysis and photographic documentation were performed using a microscope Primo Star Zeiss coupled to a camera (Axiocam ERc MRc 5s) and software Axiovision documentation.

Computational chemistry
Computational chemistry based on structural modeling and virtual docking was used to obtain insights about the fixing mechanisms of the major constituents (i.e., canonical stilbene/azo pigments) of the dyes used in this study to representative cell wall constituents.

Microcrystalline cellulose and racemic lignin models
The microcrystalline cellulose I-β model (Matthews et al. 2006) was generated using the Cellulose Builder package (Gomes & Skaf 2012); it consisted of three (1:1:1) monolayers of cellulose fibrils with approximate dimensions of 8 × 45 × 60 Å. Cellulose Builder is an optimized software to generate models of cellulose crystallites of arbitrary size in XYZ coordinates and PDB formats (Gomes & Skaf 2012).
For lignin, three different models were generated to determine the significant representativeness of different subconformations and chiral configurations for this amorphous and racemic polymer. Three preliminary and racemically distinct models following the canonical softwood lignin depiction (Sakakibara 1980;Petridis et al. 2011) were generated using software Avogadro (Hanwell et al. 2012). Since no "specialized" software is available to generate lignin models, Avogadro was chosen owing to its versatility both in the generation as well as optimization of complex molecular models at a relatively low computational cost (Hanwell et al. 2012). As described in the literature (Sakakibara 1980;Petridis et al. 2011), these models presented a random distribution of chirality at the bonds connecting the three equally canonical monomers (paracoumaryl, coniferyl, and sinapyl alcohols). These initial models were first subjected to a in vacuum energy minimization protocol by using a conjugated gradient (in order to eliminate local clashes and/or distortions) with the UFF force field (Rappé et al. 1992), followed by geometrical optimization by using the same software by using software default rotamer search algorithm. Three higher models were also generated using random crosslinking of the C-O-C and C-C extremities of the abovementioned initial units, according to the canonical softwood model (Sakakibara 1980). The approximate C:H:O ratio and molecular formula for these higher models were 3:3:1 and (C 31 H 34 O 11 ) 26 , respectively. Therefore, a solvation layer, containing 25.000 water molecules within a 50 Å radius, was added to the models by using PACKMOL software (Martínez et al. 2009). PACKMOL was used because it could generate reliable models of heterogeneous molecular systems respecting spatial constraints that ensure no repulsive interactions (Martínez et al. 2009). A final energy minimization and conformational optimization were performed for these solvated models, by using Avogadro and the UFF, resulting in three final and optimized lignin.

Results
The dyes' interactions with plant tissues are shown in Tables 3 and 4 and Figures 1-5. The best staining time was 40 min for Azul 17, Maravilha 08, Marinho 16, Vermelho 09, Vinho 61, and Violeta 25 and 60 min for the others. Variations in color and contrast between tissues, related to alcohol solution concentration, exposure time, and dye interaction with plant material, were observed. Therefore, adjusting according to the material and the desired staining intensity was necessary, as observed in Figure 1.
The 50% hydroalcoholic solution remarkably improved the solubility of the dye powder and facilitated better interaction of the solution with the plant tissue. The pH alteration had no influence on the results. Similarly, the temperature of the solutions did not influence the powder dyes' solubility and the final staining results.
Freehand sections that were washed with distilled water and drops of acetic acid after the clarification process, as suggested by Macedo (1997), did not show improved staining. Similarly, the variation in pH (4.0 and 11.0) of the wash solutions after staining did not have significant influences on the staining results.
The post-staining treatment with the "fixative" was found to be effective to prevent the migration of the dyes to the glycerinated gelatin.
Intense staining in the lignified secondary walls was noted. The staining was less intense in the collenchyma. In the thin primary cell walls of parenchyma, less interaction of the dye was noted. The dyes Preto 31, Turquesa 13, Pink 13, and Verde Primavera 26 did not interact with the plant tissues.
The staining durability of the dyes on microscope slides prepared using glycerinated gelatin was a minimum of seven days and a maximum of three years.    The preliminary test with the mixture of two dyes in 50% hydroalcoholic solutions revealed promising results. The final staining showed good contrast between tissues (Figs. 2i-k, 6), indicating the possibility of using different combinations of dyes for double-staining by using this technique.
The modified glycerinated gelatin was found to be a suitable mounting medium, preserving the vegetal material for more than three years, without any damage due to the action of fungi, bacteria, or insects.
The staining profiles of different plant tissues described above showed high affinity of the dyes' pigments to cell wall polymers, especially to lignified tissues. To obtain insights about the molecular basis of such effects (as well as the possibilities of how to optimize the process of selective staining by using textile dyes in future studies), we performed virtual docking experiments for representative stilbene/azo pigments by using structural models of crystalline cellulose surface and lignin fragments.
In all the cases, structural analysis of the binding modes recovered during the virtual docking experiments suggested that the interactions of the dye molecules with the polymer surfaces were mainly attributed to π interactions and hydrogen bonds (in the case of polar-substituted ligands; (Figs. 6-7). A considerable range of ligands' energetic docking score values were noted for both polymers, ranging from -2.70 (less favorable score in cellulose) to -8.90 kcal mol -1 (more favorable score in lignin; Fig. 8). Consistent with the binding modes shown in Figures 6 and 7, two respective attributes showed higher significance to differentiate the ligandpolymer affinity: the ligand polarity (represented by the dipole moment in Debye units) and an arbitrary and discrete scale of "geometric factor." This last geometric factor (GF) scale was then idealized in order to increase according to the respective enhancement in the ligand surface-accessible area that can establish simultaneous π interactions with the polymer surface and/or cavities (in the case of lignin). In particular, for the stilbene/azo compounds investigated in this study, the arbitrary GF defined is the function of both geometric adjustments (linear configurations being more favored that extended ones) as well as the elongation of the ligand major axis and/or increasing of the hyperconjugation degree. Highlighting that this arbitrary/discrete GF scale is only valid for the "canonical" geometries of the stilbene and azo classes, respectively, is necessary.
For cellulose, a polar, non-aromatic polymer, the major binding modes recovered in this study resided in two major subtypes of interactions. Structural analysis suggested that the nonpolar ligands interact by C-H-π stacking with the cellulose-CH groups (Fig. 6b-e). However, in the case of polar-substituted dyes, the ability of the same to position the polar groups in the grooves between the fibrils is crucial to the establishment of strong hydrogen interactions with the sugar -OH groups, as well as with the oxygen molecules from the rings and glycosidic bonds (Fig. 6c,e). This similar dependence of ligand affinity on its ability to establish both π interactions as well as hydrogen bonds with the cellulose surface reflects in the more negative docking scores (Fig. 8a) for ligands that also present high geometric factors and high dipole moments.
In contrast, lignin seems to favor the association of most of the binders compared to cellulose, returning energetic docking scores significantly more negative than the last (Fig. 8b). This is in considerable concurrence with the higher ability of the majority of the dyes experimentally investigated in this study to stain more lignified tissues compared to the less lignified ones (as related at the last topic). This relative higher average affinity seems to be related to the higher ability of the lignin polymer -amorphous, rich in hydrophobic cavities, and (overall) with polyaromatic constitution (Sakakibara 1980;Petridis et al. 2011) -to establish π aromatic interactions with the ligands (Fig. 7bd). This can be verified both by the considerably higher direct dependence of the docking score for this polymer on the geometric factor comparing the same dependence on cellulose, as well as by the lower dependence of the lignin scores on the intensity of the ligand dipole moment (Fig. 8b). For instance, the trans-stilbene derivative with GF 7 (Fig. 7d) presents the highest geometric factor index, but lower dipole moment. This ligand presents, in concurrence with its GF value, a more negative (favorable) average docking score in lignin (-8.9 kcal mol -1 ), independently of the minor polarity (Fig. 8b). The three sulfonated stilbenes computationally investigated in this study (one of them with its binding mode in lignin depicted in Fig. 7c) presented the highest dipole moments, but intermediate geometric factors (one with factor 4 and two with factor 6). These ligands present relatively less negative docking score values compared to those with factor 7 (Fig. 8b). The scores for these ligands are comparable with those of their respective less polar counterparts with similar geometric factors. This confirms the minor participation of polar interactions and the higher participation of multiple π interactions in ligandlignin associations for the dye classes analyzed in this study. Figure 6 -a-e. General view of the virtual docking experiment and major interaction modes of representative ligand poses for microcrystalline cellulose model -a. top -docking box involving all the microcrystalline cellulose I-b model; a. bottom -poses of three representative stilbene derivatives (numbered 1, 2 and 3) docked over the same model at the same docking box; b-d. interaction modes of the best poses of the respective three ligands 1 (StilCis1, ZINC8585877), 2 (AzoStTr2, ZINC6844860) and 3 (StilTr5, ZINC02556773) are showed. Receptor hydrogens involved in C-H-----π interactions are shown as Van der Waals spheres marked with a red asterisk. Hydrogen bonds between respective polar groups of the ligand and the receptor are depicted as dashed lines in c. and d.; e. angular and polar azo derivative (Azo9, ZINC19014778) showing the hydrogen bonds intercalated between the fibers (dashed lines) that this class is able to stablish specifically at the cellulose interaction, beyond the single C-H----π interaction (hydrogen in Van der Waals sphere and with red asterisk). Rodriguésia 71: e01662018. 2020

Discussion
The interaction of textile dyes with cellulose and lignin in plant tissues occurred satisfactorily, as expected, since they are largely used in the textile industries to stain cotton and other natural or synthetic fibers (Villalobos et al. 2016;Ajmal et al. 2016;Rayati & Sheybanifard 2016).  1, 2 and 3) docked over the same model at the same docking box; b-d. interaction modes of the best poses of the respective three ligands 1 (StilCis1, ZINC8585877), 2 (AzoStTr2, ZINC6844860) and 3 (StilTr5, ZINC02556773) are showed. It can be noted the more significant entrapment of the ligand by the lignin cavities compared to the planar microcrystalline cellulose (Fig. 6) and the higher contribution of the aromatic, hydrophobic and C-H-----π interactions (the hydrogens involved in these last showed as Van der Waals spheres marked with a red asterisk). The MSDS from the manufacturer informed that the dyes used in this study are a mixture of pigments from the azo/stilbene groups, without specifying the chemical structure of the dye components. However, information about the common chemical properties of azo/stilbene derivatives (Kierman 1990;Yousefi et al. 2013), as well as our experimental and computational data, allowed the inference of some probable mechanisms in which these commercial dyes could selectively stain plant tissues. For instance, the best solubility experimentally found for the dyes mixed with 50% hydroalcoholic solution might be attributed to the chemical structural groups usually found in these compounds. In fact, the majority of the 37 representative compounds of these classes randomly selected in this study from the zinc database (Irwin et al. 2012) commonly showed both hydrophobic (aromatic) regions and polar groups. This intermediary polar/apolar nature could have been responsible for the recovered interaction modes in the computational molecular docking assays for microcrystalline cellulose. Thus, not surprisingly, the dyes compounded with mixtures of pigments of these classes showed maximum solubility in solutions with intermediary polarity (as is the case of 50% hydroalcoholic solution). However, information on the solubility of textiles dyes is limited. According to Leal (2011), although considerable studies have been performed on dyes, surprisingly, a fundamental thermodynamic property -solubility -has been little investigated. The azo compounds carry the azo group (-N = N-) as a chromophore, which connects two aromatic systems, and form a staining class having many practical applications (Kierman 1990;Yousefi et al. 2013). They provide bright, strong color shades varying from red to green and blue (Yazdanbakhsh et al. 2010;Geng et al. 2011); are stable in light; resistant to chemical and microbial attacks; and have superior fixing quality (Adedayo et al. 2004;Forss & Welander 2011). We found that this set of characteristics can be favorably used to provide contrast in optical microscopy analyses of plant tissue sections. Several studies have hypothesized the interactions between cellulose and these and other dyes. Ladchumananandasivam & Miles (1994) suggested that dyes could adsorb onto the surfaces of crystalline nano-domains in cotton. Kreze et al. (2002) suspected that dyes, depending on their molecular size and solubility characteristics, might permeate into the amorphous regions of cellulose, and Timofei et al. (2002) concluded that both crystalline and amorphous cellulose sites are likely to bind with dye molecules. In our computational analysis, a binding mechanism based on surface Rodriguésia 71: e01662018. 2020 -CH-π interactions and hydrogen bonds with the polar groups between the fibrils corroborates the hypothesis of Ladchumananandasivam & Miles (1994) and Timofei et al. (2002). However, the greater surface contact of the ligands in the virtual assays on the cavities of the amorphous lignin models (and the higher dissociation energies involved), as well as the higher anatomical staining experimentally verified for tissues richer in this amorphous polymer are also considerably consistent with the hypothesis of Kreze et al. (2002). Future computational studies with amorphous cellulose models, as well as more refined experiments in which the selectivity of different isolated pigments to tissues richer in microcrystalline or amorphous cellulose needs to be verified could help in the better elucidation of this issue. All the textile dyes assessed in these studies could stain the tissues without the help of a mordant (substance used to bind the dye to the substrate) and hence are classified as "direct" dyes. The most common direct dyes are usually aromatic substances containing side groups that tend to ionize. Many direct dyes tend to carry a negative charge, and this ability has already been shown to have positive results in the staining of biological tissues (Isenmann 2013). In fact, in our computational models, the negatively charged sulfonated azo-stilbenes had higher affinity to the microcrystalline cellulose surface, corroborating this experimental observation. Even our findings suggest that a crucial factor for the interaction of polar and charged dyes with microcrystalline cellulose is the ability to intercalate their polar/ charged substituents between the fibrils, in order to establish hydrogen bonds with the hidden hydroxyl groups. Hence, a strong participation of ligand geometry at the affinity and selectivity of such polar/charged direct dyes (i.e., ligands with respective geometries that allow more adequate positioning of the polar groups between fibrils could, in principle, provide more significant staining) would be expected. A similar behavior was observed in other virtual docking studies for the affinity of oligolignol-derived adhesives to microcrystalline cellulose (Martínez et al. 2010). This suggests that these are the common determinants for the affinity of small ligands to crystalline cellulose, which, in turn, can be used in future studies for optimizing the intensity and selectivity of staining. The intense staining of lignin observed for most of the dyes was in accordance with the findings of Hubbe et al. (2012), who suggested that the relatively hydrophobic character of lignin and the many aromatic groups can be useful in the sorption of dyes and a set of other relatively hydrophobic molecules. This is significantly corroborated by our findings of virtual molecular docking analyses. These findings suggest that not only the hydrophobicity and poly-aromaticity of lignin, but also its amorphous configuration, which is rich in cavities with considerable contact surface, are the determinants for the significantly higher affinity for the dyes investigated in this study, compared to cellulose. Drnovsek & Perdih (2005) suggested the reasons for the affinity of dyes; for lignin, some of the lignin phenolic groups placed at the fiber-water interface ionize in contact with cationic dyes to enable a strong ionic interaction. However, in the same study, several, if not most, of the lignin phenolic groups were shown to interact with non-ionized acidic groups, i.e., by donating hydrogen to form hydrogen bonds with electron-rich groups in the dyes. This non-intrinsic dependence of ionization processes at the lignin-dye affinity was confirmed in our staining experiments by the verified poor influence of pH on staining by the majority of the mixtures. Further, in the computational docking analyses, considerable participation of aromatic interactions verified for the pigments' association with lignin models suggested that polar-polar interactions might not be involved in the association of these non-ionized groups. Further, considering polar-π and/or cation-π interactions are important in the interaction of the polar or ionized dyes, given the significance of π-conjugated groups recovered for the interactions by using the lignin models in our computational analysis and the known importance of polar/ion-π interactions in biological systems (Martínez & Iverson 2012;Kumar et al. 2018).
The application of the commercial "fixative" during the post-staining treatment did not cause the migration of the dyes to the gelatin glycerin. In the staining processes for textile fibers, fixatives are substances used to create a bridge between the dye molecules and fibers in order to prevent the migration of the dye from the interior of the fiber to the external environment (Peruzzo 2003). The MSDS from the manufacturer provided little information about the compounds of the fixative. It only informed that the dyes were a crosslinked resin amide. According to Costa (2016), the components of a resin used to provide fabric abrasion resistance properties are simply derived from a combination of formaldehyde and urea derivatives forming bifunctional species containing reactive methylol groups. In the presence of a suitable catalyst, these species react with cellulosic materials to form stable and relatively strong entanglements ("crosslinks") between cellulose polymer chains.
Regarding durability, although synthetic dyes are superior in terms of color availability, permanency, and light stability (Douissa et al. 2014;Chowdhury & Saha 2010), in some plant materials tested, the staining lasted only for a few days. Conversely, the duration was more than three years in other vegetal materials. Thus, the alternative dyes tested might be classified as other dyes used in plant anatomy both as semi-permanents (lasting weeks to months) and permanents (very durable, fading resistant; Kraus & Arduim 1997).
The dye solutions need to be stored in amber bottles in a refrigerator and must be filtered on a filter paper or glass wool periodically, as recommended by Kraus & Arduim (1997).
The application of dyes to plant material in closed containers (e.g., penicillin bottles) is strongly recommended to avoid the formation of precipitate that can compromise the staining result.
Taking together, our findings suggest that commercial textile dye mixtures based on azo and stilbene pigments showed satisfactory results in staining and microscopy contrast compared to dyes traditionally used in plant anatomy techniques. Preliminary tests performed using double-staining technique suggest that these dyes are a promising alternative to commercially available stains for tissues. They can be used in studies that require manual cutting of fresh plant material, as well as can be successfully used for preparing material for studying plant anatomy. The mixtures investigated in this study have shown special staining efficiency to cell wall regions (consistent with their ability to stain vegetal fibers in textile materials), with a relatively greater affinity for lignified tissues. Although the MSDS from manufactures do not unequivocally identify what molecules could bind to the cell wall components and the specific binding mechanisms, a more general computational docking approximation has shown similar molecular basis of interactions for a set of representative compounds. The results of computational analyses corroborated the higher affinity of the dyes for lignin than for cellulose and are in agreement with a set of technical nuances; they provide helpful insights for future studies for the optimization of the staining intensity and selectivity. Finally, the integrated plant anatomy/computational chemistry-biochemistry approach used in this study can be didactically reproduced in interdisciplinary classes both at high school level as well as for undergraduate programs of science and engineering (in courses in the fields of biology, agronomy, forestry, and environmental engineering). Such interdisciplinary classes have been proven to be promising tools for the teachinglearning paradigm (Lima et al. 2009;Haynes & Leonard 2010). Hence, we expect that these results provide insights for the rational planning of lowcost and selective histological ligands and dyes for scientific and educational uses, as well as allow an environmentally sustainable alternative for using textile dyes.