The Morphological and Anatomical Traits of the Leaf in Representative Vinca Species Observed on Indoor- and Outdoor-Grown Plants

Morphological and anatomical traits of the Vinca leaf were examined using microscopy techniques. Outdoor Vinca minor and V. herbacea plants and greenhouse cultivated V. major and V. major var. variegata plants had interspecific variations. All Vinca species leaves are hypostomatic. However, except for V. minor leaf, few stomata were also present on the upper epidermis. V. minor leaf had the highest stomatal index and V. major had the lowest, while the distribution of trichomes on the upper epidermis was species-specific. Differentiated palisade and spongy parenchyma tissues were present in all Vinca species’ leaves. However, V. minor and V. herbacea leaves had a more organized anatomical aspect, compared to V. major and V. major var. variegata leaves. Additionally, as a novelty, the cellular to intercellular space ratio of the Vinca leaf’s mesophyll was revealed herein with the help of computational analysis. Lipid droplets of different sizes and aspects were localized in the spongy parenchyma cells. Ultrastructural characteristics of the cuticle and its epicuticular waxes were described for the first time. Moreover, thick layers of cutin seemed to be characteristic of the outdoor plants only. This could be an adaptation to the unpredictable environmental conditions, but nevertheless, it might influence the chemical composition of plants.

Vinca plants are intensely studied for their medical properties [3][4][5] due to the rich alkaloid content [6][7][8][9]. These natural products are produced and stored in the aerial parts of Vinca plants [6,10,11], serving as protection against herbivores and pathogens. [12,13]. V. minor is one of the most important medicinal species [14] and the sole source of vincamine in nature, one of the few alkaloids with beneficial effects on cells [3,15]. Quantitative and qualitative differences of the chemical composition in Vinca extracts are dependent on species and the high content of alkaloids, flavonoids, and phenolic compounds, are in correlation with different pharmacological effects [3,[16][17][18][19].  On the abaxial leaf side, all Vinca species presented typical puzzle-shaped pavement cells, while stomata were randomly distributed on the entire surface, except for the midvein (Figure 2).

Vinca Trichomes and Stomata Characteristics
Unicellular conical trichomes were observed on the margins of the adaxial side of V. major, V. major var. variegata, and V. herbacea leaves (Figure 3), while V. minor had trichomes

Vinca Trichomes and Stomata Characteristics
Unicellular conical trichomes were observed on the margins of the adaxial side of V. major, V. major var. variegata, and V. herbacea leaves (Figure 3), while V. minor had trichomes only on the midvein. The trichomes had the same orientation in all Vinca species, i.e., from the petiole to the apex (from back to front). The morphological parameters of the stomata and trichomes, and their distribution were calculated for each side of the leaves (Table 1). side, Ab = abaxial (lower) side, SD = Stomatal density, SI (%) = stomatal index, SL = Stomatal pore length, SPI (%) = Stomatal pore index, TD = Trichomes' density; the criteria marked with * are measured in a one mm 2 surface area; SL was measured three times on a minimum of six independent stomata ± standard error of the mean (s.e.m.).
Trichomes have protective roles in plants and their number is usually regulated by the same factors that affect the number of stomata [32]. As the Vinca trichomes are nonglandular, their main role remains as protectors against water loss, and their density is inversely correlated (Figure 4) with the stomatal index (r = −0.7). The morphological parameters of the stomata and trichomes, and their distribution were calculated for each side of the leaves (Table 1). Ad = adaxial (upper) side, Ab = abaxial (lower) side, SD = Stomatal density, SI (%) = stomatal index, SL = Stomatal pore length, SPI (%) = Stomatal pore index, TD = Trichomes' density; the criteria marked with * are measured in a one mm 2 surface area; SL was measured three times on a minimum of six independent stomata ± standard error of the mean (s.e.m.).
Trichomes have protective roles in plants and their number is usually regulated by the same factors that affect the number of stomata [32]. As the Vinca trichomes are nonglandular, their main role remains as protectors against water loss, and their density is inversely correlated (Figure 4) with the stomatal index (r = −0.7).

Vinca Mesophyll
The semithin sections ( Figure 5) helped to determine the anatomical features of the Vinca leaves ( Table 2). The mesophyll of V. minor and V. herbacea had a similar structure, with obvious delimitation between the palisade and spongy parenchyma (Figure 5a,d). In V. major and V. major var. variegata, the palisade parenchyma had a lax aspect and large intercellular spaces (Figure 5b,c). 3 LT = Leaf thickness, PTT = Palisade tissue thickness, STT = Spongy tissue thickness, CTR (%) = Cell tense ratio, SR (%) = Spongy tissue ratio; at least three independent measurements were conducted for each parameter measured on the leaf; for LT and STT n = 3, ± s.e.m.; for PTT n = 6, ± s.e.m.
A computational estimation of the intercellular space in the mesophyll was conducted ( Figure 6) revealing a high percentage (over 60%) of airspaces for the studied Vinca species.

Vinca Mesophyll
The semithin sections ( Figure 5) helped to determine the anatomical features of the Vinca leaves ( Table 2). The mesophyll of V. minor and V. herbacea had a similar structure, with obvious delimitation between the palisade and spongy parenchyma (Figure 5a,d). In V. major and V. major var. variegata, the palisade parenchyma had a lax aspect and large intercellular spaces (Figure 5b,c). LT = Leaf thickness, PTT = Palisade tissue thickness, STT = Spongy tissue thickness, CTR (%) = Cell tense ratio, SR (%) = Spongy tissue ratio; at least three independent measurements were conducted for each parameter measured on the leaf; for LT and STT n = 3, ± s.e.m.; for PTT n = 6, ± s.e.m.
A computational estimation of the intercellular space in the mesophyll was conducted ( Figure 6) revealing a high percentage (over 60%) of airspaces for the studied Vinca species.

Vinca Epidermal Cell Walls and Cuticle Layers
For a thorough investigation of the anatomical features of Vinca leaves, transmission electron microscopy (TEM) analyses were performed. The external pavement cell walls in the upper and lower epidermises (Supplementary Figure S1) were investigated ( Figure 7). It was observed that except for V. major (Figure 7c,d), all leaves had a thin layer of epicuticular waxes present above the cutin layers.

Vinca Epidermal Cell Walls and Cuticle Layers
For a thorough investigation of the anatomical features of Vinca leaves, transmission electron microscopy (TEM) analyses were performed. The external pavement cell walls in the upper and lower epidermises (Supplementary Figure S1) were investigated (Figure 7). It was observed that except for V. major (Figure 7c,d), all leaves had a thin layer of epicuticular waxes present above the cutin layers.   The thickness of the upper and lower epidermises was measured with the help of TEM images. The results showed that the upper epidermis is significantly thicker than the lower epidermis for V. minor (p = 0.005), V. major var. variegata (p = 0.03), and V. herbacea (p = 0.0002), and of all species, V. major had the thinnest cell walls on both epidermises ( Figure 8a). Also, the cuticle present on the upper epidermis was thicker than the cuticle on the lower epidermis for V. minor and V. herbacea (Figure 8b). The thickness of the upper and lower epidermises was measured with the help of TEM images. The results showed that the upper epidermis is significantly thicker than the lower epidermis for V. minor (p = 0.005), V. major var. variegata (p = 0.03), and V. herbacea (p = 0.0002), and of all species, V. major had the thinnest cell walls on both epidermises ( Figure 8a). Also, the cuticle present on the upper epidermis was thicker than the cuticle on the lower epidermis for V. minor and V. herbacea (Figure 8b).   The thickness of the upper and lower epidermises was measured with the help of TEM images. The results showed that the upper epidermis is significantly thicker than the lower epidermis for V. minor (p = 0.005), V. major var. variegata (p = 0.03), and V. herbacea (p = 0.0002), and of all species, V. major had the thinnest cell walls on both epidermises (Figure 8a). Also, the cuticle present on the upper epidermis was thicker than the cuticle on the lower epidermis for V. minor and V. herbacea (Figure 8b).

Discussion
There are quantitative and qualitative differences regarding the chemical composition of Vinca extracts, and previous studies showed that the morphological and anatomical traits of the leaves (shape and surface, cuticle, stomata, trichome, epidermis, etc.) are responsible for the chemical composition variability in the extracts [33][34][35][36]. These morphological and anatomical aspects of the Vinca leaf had been investigated in this study as well. The leaf surface area and dry leaf mass are directly correlated for vines species (Vinca major, Trachelospermum jasminoides, Hedera nepalensis var. sinensis) [37] and for Arabidopsis thaliana [38]. This aspect is important in extract preparation for maximal exploitation of the plants.
Additionally, the morphological characters of all plant organs are used for accurate identification of species [14,[33][34][35]. The leaf shape varies within species and populations, and even within the same individual plants; therefore, it affects photosynthesis, water balance, temperature control, and the interactions with other organisms [39].
On the same surface of the Vinca leaf within different species, the epidermal cells are distinct, and here was shown that they could vary on both sides even for the same species (Figures 1-3). The Vinca leaf epidermis has a single layer of cells composed of pavement cells, stomatal complexes, and trichomes (Figures 1-5), that serves as a protective barrier to environmental factors [40]. On the adaxial side, Vinca leaves had puzzle-shaped epidermal cells, except for V. herbacea that had papillary epidermal cells. Similar results were previously reported for V. minor and V. herbacea only [41], and there are several theories that try to explain this type of organization of the pavement cells [42][43][44].
These specialized epidermal cells differ intra-and interspecific, depending on various extrinsic and intrinsic parameters, e.g., the leaf age, dimension, insertion position on the stem, combined with CO 2 atmospheric concentration, irradiance, humidity, and other environmental factors (pollution, pathogens, etc.) [44][45][46]. Therefore, the current study was conducted under minimum variation of these parameters. The outdoor plants (V. minor and V. herbacea) grew in a shaded area and had the same light exposition (46 • 45 51" N; 23 • 34 47" E). The indoor plants (V. major and V. major var. variegata) were exposed only to sunlight with the same circadian cycles as the outdoor plants and the humidity and CO 2 levels were regulated so that they could resemble the natural conditions. All the above-mentioned criteria play vital roles for the medicinal taxa [47,48].
Also, a connection between the cell walls of the leaves' epidermis and the natural products in plants had been previously established [6]. In most plant species, the epidermis has specialized roles in the biosynthesis and accumulation of a wide range of natural products, including alkaloids [40], terpenes, and flavonoids [40,49], indicating that the variances observed here could be indicative of the chemical variation of the plant extracts.
The Vinca leaf epidermal cells are covered by a cuticle (Figures 5, 7 and 8) with complex chemical composition and have different thicknesses on the upper and lower epidermises as previously indicated [50][51][52]. In terms of electron-dense cutin layers, the V. minor, V. major var. variegata and V. herbacea leaves are similar to the leaves of Pyrus communis and Populus bolleana [50], but were never described for the Vinca species. The cuticular waxes play significant roles in Vinca species due to their involvement in the defense mechanisms against pathogens, temperature variations, salinity, or excessive ultraviolet radiations [52].
Glandular and secretory trichomes are responsible for biosynthesis, secretion and/or accumulation of phytochemicals useful for defense [40]. In the leaves of Vinca species analyzed herein, only tector (protective) trichomes had been identified, with the main role in defense mechanisms against water loss. The number of trichomes increases while the stomatal index decreases, and this way, the transpiration process is balanced [53], explaining our results where the stomatal index was found to be inversely proportional to trichomes' density. For V. minor, the trichomes were placed only on the midvein, while the other species had trichomes on the margins and on secondary veins. Similar results had been previously reported as well for V. minor and V. herbacea [41,54].
Stomata are important structures involved in exchange processes of the plant with the environment, balancing the CO 2 influx during photosynthesis and water vapor efflux during transpiration [28,33,35,40,55]. Stomata exhibit a diverse range of shapes, sizes, and numbers across different plant species [53]. Leaf stomata in plants from the same genus or from different plants of the same species are variable [56,57]. Significant negative correlation was documented between stomatal density and stomatal shape parameters (stomatal area, stomatal perimeter, stomatal long axis, stomatal short axis) in other plant species as well [58].
The stomatal index is the recommended parameter to be measured since it reports the stomata density in relation to the number and size of the pavement cells, in each area [28,46]. The stomatal index for the adaxial side (around 4%), is evidently lower than the one registered for the abaxial side (20-30%); therefore, the studied Vinca species are all hypostomatic.
Our results showed that the overall thickness of the Vinca leaf differs with the species and this influences the photosynthetic rates as well [59]. Besides this, the stomatal index has different values for each Vinca species, and the chemical composition could vary as well. This is because a higher specific stomatal index is related to a higher net rate of photosynthesis, and consequently, the plant could enhance the carbon fixation [28,32]. For example, several sugars were shown to affect the growth and development of V. minor and C. roseus leaves [14]. As carbon is present in many sugars and alkaloids, this could also influence the chemical composition in the long run [28].
Vinca leaf mesophyll has different thicknesses, a parameter which is species depended [36]. The Vinca leaves contain different numbers of columnar palisade cells. It was confirmed that sunlight-grown plants have more columnar palisade cells than those of shade-grown plants [29,34,35]. Similar to our results where the indoor plants had a less differentiated palisade tissue than the outdoor plants, other studies showed that the palisade tissue cells in V. major var. variegata leaves were single-layered for the plants located at roof level and double-layered for ground-level plants [35]. Another study demonstrated that depending on light exposition, the leaf of V. minor had a double-layered palisade tissue and V. herbacea leaf, a single-layered one [41]. The shape of palisade cells and the movement of chloroplast in accordance with the light conditions, are essential for efficient leaf photosynthesis, by cell development regulation with the help of phototropin [29]. The number and shape of palisade cells, which compose the photosynthesis unit area, are different for each Vinca leaf analyzed herein, therefore the photosynthesis process varies along with the anatomical traits [29] and this might influence the chemical composition as well if the plants are to be considered for extract preparation.
The thickness of the spongy parenchyma in the Vinca leaf varies with the species and contains several layers of ovoid, oblong, or circular shaped cells and the mesophyll airspace [34,35,41]. The cells of the spongy parenchyma store important nutrients for the plant [60]; therefore, its thickness would be directly related to the abundance of natural compounds [61]. Other studies described 7 to 8 layers of cells in the spongy tissues of V. minor and V. herbacea [34,41], while for V. major var. variegata, the number of layers decreased to 6-7 for plants grown at ground level [35]. The results presented herein showed that the plants grown outdoor had 7 to 8 layers, while the plants grown indoor, had 5 to 6 layers of cells in the spongy parenchyma-results which are consistent with other findings-and this indicates that growth conditions are important parameters that should be considered before extract preparation.
The programmatic method used for computation helped to determine the ratio between intercellular spaces and cells in the mesophyll of the Vinca species. First, it must be stressed that the main objective was the accuracy of the mathematical computations with respect to the ground truth. Initially, various methods were employed for the semi-automatic segmentation of the regions in the semithin sections, using state-of-the-art methods such as those based on computer vision [62,63] or machine learning approaches [64,65]. The employed methods obtained the same accuracy levels as those found in the literature.
However, the visual assessment of the results leads to the conclusion that since the end goal was to compute a percentage as close to reality as possible through an automated method of visual segmentation and labeling of the cellular and inter-cellular regions, significant errors in the computational process occurred. In other words, an accuracy level that is acceptable from a computer science perspective may, however, damage the biological scientific accuracy of the research. V. minor had the lowest mesophyll airspace (62.8%), followed by V. major (63.8%) and very close to each other were V. major var. variegata and V. herbacea (67.3% and 67.2%, respectively). According to recent studies, there is a connection between the mesophyll airspace and functional stomata [36,66], implying that larger airspaces are connected to a higher stomatal index. Since the Vinca species are hypostomatic, the high percentage for intercellular spaces is justified in relation to the high stomatal index observed.
The ultrastructural analysis of the leaves offered a more detailed view of the anatomy and confirmed some of the hypotheses regarding the spongy parenchyma. Lipids are used as substitutes for energy production and their presence as big droplets in the V. major and V. major var. variegata leaf could indicate a compensatory mechanism adopted by the plant because of the greenhouse storage conditions [67]. As indicated in other studies, alkaloids are harmful to the plant and are therefore stored or synthesized in small amounts [61] and only when needed, and the chemical composition variability of Vinca plant extracts could be reflected by different amounts of lipid droplets identified in leaf spongy parenchyma [67]. Overall, the methods used to determine the morphological and anatomical aspects of the Vinca leaves are correlated with each other and can be used to analyze the leaves of other species as well. This is an important step since Vinca species are recognized for their medicinal properties, and a brief morphological and anatomical description of a batch could help determine if the plants are suited for extraction.

Plant Material and Growth Conditions
The Vinca plants ( Figure 10) were collected in April-May from Alexandru Borza Botanical Garden, of Babeş-Bolyai University in Cluj-Napoca. A voucher specimen for each species was deposited at the Herbarium of Babeş-Bolyai University (CL 665977 for V. minor, CL 668019 for V. major, CL 668018 for V. major L. var. variegata 'Louden', and CL 668021 for V. herbacea).

Scanning Electron Microscopy
An optimized and adapted method [68] was used for electron microscopy analyses. Shortly after harvest, the leaves were placed in 2.7% glutaraldehyde (GT) solution in 0.1 M PBS (phosphate buffer saline) at pH 7.2. After 1 h, GT was changed and left for an Fully developed leaves from adult plants were collected from approximately the mid part of the stem. V. minor and V. herbacea grew in the shade within the grounds of the Botanical Garden, while V. major and V. major var. variegata were potted plants grown in the greenhouse. Before harvest, the selected leaves were thoroughly washed with distilled water.

Scanning Electron Microscopy
An optimized and adapted method [68] was used for electron microscopy analyses. Shortly after harvest, the leaves were placed in 2.7% glutaraldehyde (GT) solution in 0.1 M PBS (phosphate buffer saline) at pH 7.2. After 1 h, GT was changed and left for an additional hour, after which the samples were thoroughly washed four times with PBS. For SEM, the leaves were cut in half and further dehydrated using increasing concentrations of acetone as follows: 30 min at 30%, 30 min at 50%, 30 min at 70%, 1 h at 80%, 1 h at 90%, 4 h at 100%, changing the solution every 1 h. Subsequently, the samples were placed in hexamethyldisilazane:acetone solution (1:2, 1:1, 2:1, 1:0 for 1 h each). All steps were conducted at 4 • C and the reagents were acquired from Sigma Aldrich (Merck, Bucharest, Romania). For examination, SEM HITACHI SU8230 (Hitachi, Tokyo, Japan) was used at an acceleration voltage of 30 kV after the samples were covered with a 9-nm-thick layer of gold, using the Quorum Q150T ES turbomolecular pumped coater (Quorum Technologies, London, UK). Each parameter was measured three times and form one leaf. At least six independent measurements were taken in six different spots on the leaf.

Leaf Stomatal Traits
The obtained SEM images were also used to calculate the stomatal index (SI) according to Equation (1), and the stomatal pore index (SPI), according to Equation (2), for each side of the leaves [45,69]. The density of the trichomes per mm 2 was determined as well. SI (%) = SD/(SD + E) × 100 (1) where SD is the stomatal density (number of stomata/mm 2 leaf surface) and E is the epidermal cell density in the same surface area.
where SL is the stomatal pore length calculated for six random stomata present on the surfaces, and the mean was calculated.

Light Microscopy
For light microscopy (LM) and transmission electron microscopy (TEM) analyses, the samples were cut into~3 mm × 1 mm × 0.5 mm pieces and placed in osmium tetroxide solution (2% OsO 4 in 0.15 M PBS of 7.4 pH) for 3 h. Afterward, the samples were washed in PBS and then dehydrated with increasing concentrations of acetone, as follows: 15 min at 30%, 15 min at 50%, 15 min at 70%, 30 min at 80 and 90%, 2 h at 100%, changing the solution once every 30 min. Samples were then infiltrated with Epon 812:acetone solution (1:2, 1:1, 2:1, 1:0, 1 h each step) and then left overnight in freshly prepared Epon 812 epoxy resin. The samples were encapsulated in a horizontal silicone matrix at 60 • C for three days. Using glass knives, a Diatome diamond knife (DiATOME, Hatfield, PA, USA), and Leica UC7 ultramicrotome (Leica Microsystems, Wetzlar, Germany), semithin and ultrathin sections were obtained. The semithin sections (300-400 nm thickness) were placed on glass slides and stained with Epoxy tissue stain, after which they were examined using the Olympus BX51 light microscope coupled with a CCD camera (Olympus, Hamburg, Germany).

Leaf Anatomical Traits
Using the semithin sections, the cell tense ration (CTR) and spongy tissue ratio (SR) were calculated, according to Equations (3) and (4) [69]: where PT is the palisade tissue thickness, and LT is the leaf thickness, both parameters measured along the leaves' sections in three different spots, and the mean was calculated.
where STT is the spongy tissue thickness. The parameter was calculated along the leaves' sections in three different spots, and the mean was calculated.

Transmission Electron Microscopy
The ultrathin sections (80-150 nm thickness) were placed on 200 mesh copper grids and double-stained with uranyl acetate (2.6 g in 20 mL of 50% ethanol in ultrapure water, for 6 min) and lead citrate (1.41 g in 42 mL MQ water and 8 mL NaOH 1 N, pH 12, for 3 min). The samples were examined using the TEM Jeol JEM 1010 (JEOL, Tokyo, Japan), operated at 80 kV, and coupled with a Mega View III digital camera.

Computational Estimation of the Intercellular Space in the Mesophyll
Using the semithin sections obtained, the cellular and intercellular regions were calculated. The determination and labeling were conducted manually, by use of a professional image processing tool (Photoshop, version 13.0; Adobe Systems Inc., San Jose, CA, USA) for the manual segmentation of the regions. Each region of interest was colored differently as follows: the cellular region was labeled green, the intercellular region was labeled grey, the epidermises were labeled blue, and unimportant regions were labeled yellow. The epidermises were excluded from the calculation, as they are not part of the mesophyll.
Further, the images were used as input for a straightforward Python script (version 3.8.6; available online) that merely quantifies the pixels of different colors and later on computes the required surface ratios, based on simple pixel quantization as the ratio between the number of pixels labeled grey over the number of both grey and green pixels.

Statistical Analyses
One-way ANOVA and Tukey statistical analyses were performed using OriginPro 2016 software (OriginLab Corporation, Northampton, MA, USA). At least three independent measurements were conducted for each parameter, and the mean was calculated ± standard error of the mean. The obtained values at significance levels of p ≤ 0.05 were considered statistically significant.

Conclusions
A scan through the leaf morphological and anatomical traits of Vinca minor, V. herbacea, V. major and V. major var. variegata species revealed complex details. The stomatal index is related to trichome density, cuticle, spongy and palisade parenchyma thickness, and probably to the lipid and alkaloid content as well. As a significant novelty, the epicuticular waxes and cuticles of Vinca leaf epidermises were described for the first time, along with a mesophyll airspace determination method. These results provide a new perspective for other plant species as the parameters analyzed herein can directly influence the photosynthetic rates. Because the leaves of various medicinal plants are often used in extract preparation, a thorough and beforehand examination of these morphological and anatomical parameters could help determine if those plants are suited for pharmacological assessments. The investigated traits justify the differences in the chemical composition of Vinca extracts as previously observed in other studies as well. Since photosynthesis is responsible for carbon fixation and plant nutrition overall, the morpho-anatomical data of the Vinca species complete the literature data while it brings novelty to this field.