INTRODUCTION

The genus Ilex L. (Aquifoliaceae Bercht. & J. Presl) comprises about 500 cosmopolitan species (Loizeau et al., 2005). A major center of species diversity is found in South America, with ca. 300 species (Loizeau et al., 2005); in the Southern Cone, from eastern Paraguay, southern Brazil, northeastern Argentina to southern Uruguay, 13-15 native species have been registered (Giberti, 1979, 1994a, b, 2001, 2008, 2011; Grela, 2004; Giberti & Gurni, 2008). Among the species occurring in this region are I. affinis Gardner, I. argentina Lillo, I. brasiliensis (Spreng.) Loes., I. brevicuspis Reissek, I. chamaedryfolia Reissek, I. dumosa Reissek, I. integerrima (Vell.) Reissek, I. microdonta Reissek, I. paraguariensis A. St.-Hil., I. pseudobuxus Reissek, I. taubertiana Loes., and I. theezans Mart. ex Reissek. All species of Ilex are perennial, insect-pollinated (Ferreira et al., 1983; Carr, 1991; ^{Tsang & Corlett, 2005}) and functionally dioecious trees or shrubs, with unisexual flowers showing rudimentary reproductive organs of the opposite sex (Giberti & Gurni, 2008). Thus, the pistillate (or female) flower possesses staminodes (aborted stamens) and the staminate (or male) flower a pistillode (aborted gynoecium) (Giberti & Gurni, 2008). Among southern South American species, I. paraguariensis, known as the “yerba mate” tree, stands out by its medicinal and nutraceutical properties and major socio-economic importance (Filip et al., 2001; ^{Bastos et al., 2007}; Turner et al., 2011); their twigs and leaves are the raw materials for the popular “mate” or “tereré” infusions. Only recently, I. dumosa, commonly known as “yerba señorita’’, also became available on the market as an herbal infusion mix of low-caffeine content (^{Maiocchi et al., 2016}).

There is little information on the biology of Ilex species, particularly regarding cytogenetics. Indeed, the chromosome number has been determined in only 5% of the species for the whole genus (Greizerstein et al., 2004). Among southern South American species, I. brevicuspis, I. dumosa, I. integerrima, I. paraguariensis, I. pseudobuxus, I. taubertiana, and I. theezans have been registered having a diploid chromosome complement of 2n=2x=40, while I. argentina is the single polyploid species (2n=4x=80) so far documented in the region (^{Andrés & Saura, 1945}; ^{Barral et al., 1995}; Greizerstein et al., 2004). Information from public databases indicate that among the 39 records for worldwide accepted Ilex species, solely three are polyploids, namely the Hawaiian I. anomala Hook. & Arn. (2n=4x=80), the Asian I. pedunculosa Miq. (2n=6x=120) and the Sudamerican I. argentina; for Ilex, seven different gametic chromosome numbers are reported, being n=20 the most common (67%) (CCDB, Chromosome Count DataBase http://ccdb. tau.ac.il/). Then, a detailed karyotype analysis is currently known solely for I. brevicuspis, which has 40 small-sized chromosomes (0.93-2.63 µm), of which 36 are metacentric, 2 submetacentric and 2 subtelocentric (Daviña, 1998). In addition, the DNA content of the unreplicated, non-reduced chromosomic complement (known as the genome size or 2C-value) (Greilhuber et al., 2005) has been estimated solely for three South American species. Barral et al. (1995) determined the DNA content of I. argentina (2C=4.27 pg) and I. paraguariensis (2C=2.23 pg) using Feulgen densitometry. Gottlieb & Poggio (2014) determined the 2C-value of I. paraguariensis (2C=1.71 pg) and I. dumosa (2C=1.89 pg) by flow cytometry. On the other hand, there are records for the Eurasian taxon I. aquifolium L. (2C=1.93 pg; Loureiro et al., 2007), and for I. mucronata (L.) M. Powell, Savol. & S. Andrews (2C=2.2 pg) and I. verticillata (L.) A. Gray (2C=4.1 pg) both from North America (^{Bai et al., 2012}); among the Asian species, there is information for I. cornuta Lindl. & Paxton (2C=1.31 pg; ^{Zhang et al., 2013}), I. latifolia Thunb. (2C=1.91 pg), I. suaveolens (H. Lév.) Loes. (2C=2.24 pg), I. viridis Champ. ex Benth. (2C=2.52 pg), and I. micrococca Maxim. (2C=3.05 pg) (Su et al., 2020). Recently, the genome size of I. polyneura (Hand.-Mazz.) S.Y. Hu has been obtained (2C=1.48 pg; ^{Yao et al., 2022}).

Present survey represents a contribution to the genetic knowledge of an understudied plant family that comprises many species which are highly appreciated worldwide as ornamentals, herbal teas, or even as medicinal plants. Gaining insight on genetic biodiversity is relevant for the potential use of non-industrialized species in the improvement of economically significant ones. Thus, to contribute to the characterization of Ilex species from southern South America, in the present study we established the genome size of 10 species and determined for the first time the chromosome number of I. brasiliensis.

MATERIALS AND METHODS

Fresh plant materials for I. argentina, I. brasiliensis, I. brevicuspis, I. dumosa, I. integerrima, I. microdonta, I. paraguariensis, I. pseudobuxus, I. taubertiana and I. theezans were obtained from the Ilex Germplasm Bank held at the “Estación Experimental Agropecuaria INTA Cerro Azul” (EEA-INTA-CA; Misiones, Argentina) and the Carlos Thays Botanic Garden of Buenos Aires City (Table 1). As no reliable living plants of I. affinis and I. chamaedryfolia were available to us, these species could not be studied.

DNA content was measured by flow cytometry following the procedure of Gottlieb & Poggio (2014), with some modifications. The fresh leaves from each plant sampled were kept in darkness for at least 24 h before processing. For each species, 1-3 individuals (plants) were assayed (that is, 34 individuals in total, from 27 accessions), and 2-3 measurements per individual plant were done using different leaves. Nuclear suspensions were prepared by simultaneously chopping 1 cm² of adult leaves of the target species and the internal standard (Zea mays ssp. mays cultivar CE-777 [2C=5.43 pg; Lysák & Doležel, 1998] or cultivar B73 [2C=4.6 pg; Schnable et al., 2009]) in 2 ml of Partec buffer (Partec GmbH, Münster, Germany). This suspension was treated with 10 µl RNase (20 mg/ml), incubated for 30 min at room temperature and then filtered through a 42 µm-nylon mesh. Next, 500 µl of propidium iodide (50 mg/ml) was added and the mix was incubated in darkness. The incubation time was set to 5 min and it was extended to 10, 15 or 20 min when necessary. The time was considered adequate when the histograms of fluorescence intensity vs. event counts showed well-defined peaks, and the minimum number of events collected was above 5,000. Only coefficients of variation below 5% were considered. Measurements were carried out in a CyFlow Ploidy Analyser (Partec GmbH, Münster, Germany) at the “Instituto de Floricultura del INTA” (Castelar, Buenos Aires, Argentina). For each individual plant surveyed, raw counts were obtained by means of histograms of event counts vs. fluorescence intensity using the Flowing Software version 2.5.1 (Cell Imaging Core, Turku Center for Biotechnology, Finland).

The 2C-value was obtained by multiplying the known DNA content of the standard (in pg) by the quotient of the mean fluorescence intensity of the target species and that of the internal standard, which were deduced from the histograms (Doležel et al., 1989). The quotient between the mean 2C-value of cultivar CE-777 and that of B73 was calculated to standardize measurements made for individuals of the same species when different internal standards were used. A correction factor (=1.167) was applied to measurements where cultivar B73 was the internal standard. An analysis of variance (ANOVA) was performed, followed by post-hoc Tukey test for multiple pairwise comparisons, using the lme4 (^{Bates et al., 2015}) and emmeans packages (Length, 2022) implemented in R (R Core Team, 2018).

Chromosome counts were made from secondary roots and staminate flower buds of individuals of I. brasiliensis from accessions EEA-INTA-CA 221 and EEA-INTA-CA 226, and to further explore our findings, individuals of I. theezans from accessions EEA-INTA-CA 101 and EEA-INTA-CA 225, were also inspected. For mitotic preparations, roots were treated with 8-hydroxyquinoline (0.002 M) for 2 h, fixed in an absolute ethanol:glacial acetic acid (3:1) solution for 72 h and then preserved in 70% ethanol until processing.

Next, roots were washed in isocitrate buffer (1:10 citric acid:sodium citrate, pH 4.8) followed by incubation with an enzymatic solution of cellulase (2% w/v; Onozuka R10 Merck) and pectinase (20% v/v; Sigma P4716) for 2 h at 37 ºC. Root tips were dissected under a stereoscopic microscope and slides were prepared by the squash method, and stained with DAPI (4′,6-diamidino-2-phenylindole, 2 µg/ml) according to ^{Sumner (1990}).

For meiotic preparations, flower buds were fixed as described above. The anthers were squashed and stained with DAPI. All slides were examined and photographed using a fluorescence microscope (Leica DMLB) equipped with a digital camera Leica DFC 350 FX.

Table 1 Studied Ilex species, accession number, geographic origin (country, province-state, department) and herbarium number. 

a Original BACP Herbarium numbers, stated when available; these vouchers are currently located at BA and BAA; ^{Thiers (2022}).

b Plants collected in Jardín Botánico Carlos Thays, Buenos Aires, Argentina.

RESULTS

The mean DNA content among the species ranges between 1.691 and 3.600 pg (Table 2, Fig. S1); I. theezans shows the highest 2C-value, while I. pseudobuxus has the lowest DNA content. In general, the 34 individuals analyzed show a slight variation in DNA content at the intraspecific level (Table S1). However, there are significant differences in 2C-values among the nine individuals of I. theezans: eight of these (corresponding to five accessions) have a mean genome size of 3.600 ± 0.040 pg, while the single individual of accession EEA-INTA-CA 225 has a mean genome size 1.95 times lower (Table 2).

Table 2 DNA content of southern South American Ilex species. Abbreviations: IT, incubation time in propidium iodide (in min); N, number of plants; 2C, mean 2C-value ± SD; TT, Tukey’s test: different letters indicate significant differences (p < 0.05) among species´ 2C-value means. * accession EEA-INTA-CA 225 of I. theezans. 

ANOVA showed significant differences (p < 0.05) in mean genome size among species. Tukey test identified the presence of three groups: (a) I. dumosa, I. paraguariensis and I. pseudobuxus, (b) I. brevicuspis, and (c) I. argentina and I. theezans. The species I. brasiliensis, I. integerrima, I. microdonta and I. taubertiana, and the accession EEA-INTA-CA 225 of I. theezans could not be assigned either to group (a) or (b) (Table 2).

The appropriate incubation time of nuclear suspensions in propidium iodide was 5 min for most species (Table 2); longer incubation times reduced fluorescence detection. In I. theezans, best results were achieved with incubations between 5 and 10 min, while in I. brasiliensis and I. brevicuspis the appropriate time was 20 min.

Cytological analysis of I. brasiliensis indicated that this species has 2n=40 chromosomes (Fig. 1A) and regular meiotic behavior, with 20 bivalents (Fig. 1B). Given the variation in genome size among I. theezans individual plants, chromosome counts were carried out on individuals from accessions with contrasting 2C-values (i.e., EEA-INTA-CA 101 and EEA-INTA-CA 225). Even though this was accomplished using flower buds because root tips were not available, all cells inspected from individuals of I. theezans showed 20 bivalents (2n=40) (Fig. 1C-E).

DISCUSSION

Genome size plays an important evolutionary role, and its variation has been associated with cytological, morphological, and developmental features, among others (^{Bennett & Leitch, 2005}; Kron et al., 2007). Here we focused on the estimation of the genome size of most of the Ilex species from the Southern Cone of South America.

It is known that cytosolic metabolites such as caffeine and chlorogenic acids affect the accessibility of propidium iodide to nuclear DNA causing errors in the estimation of genome size, as previously observed in coffee by ^{Noirot et al. (2003}). Gottlieb & Poggio (2014) detected a strong interference in the fluorescence signal of nuclear suspensions of I. paraguariensis which was attributed to the high concentrations of the secondary compounds typically shown by the “yerba mate”. In fact, the proportion of metabolites in the leaves of I. paraguariensis is up to 81 times higher than in the other southern South American species (Filip et al., 1998, 2000, 2001), and, even further, it is the only species with significant amounts of caffeine (Kim et al., 2010). The different incubation times required here to attain adequate measurements may be attributed to the particular combination and concentrations of the cytosolic compounds that characterize each species (Filip et al., 2001; Kim et al., 2010).

This is the first report of the DNA content of I. brasiliensis, I. brevicuspis, I. integerrima, I. microdonta, I. pseudobuxus, I. taubertiana and I. theezans. The species I. brevicuspis, I. dumosa, I. integerrima, I. paraguariensis and I. taubertiana, previously documented as diploids by ^{Andrés & Saura (1945}), ^{Barral et al. (1995}) and Greizerstein et al. (2004), showed here that their DNA content ranged between 1.69 and 2.04 pg. The Euroasian I. aquifolium and the Asian I. cornuta are the only species with both verified chromosome number (de Clavijo, 1991; Xu, 1992) and known 2C-value (Loureiro et al., 2007; ^{Zhang et al., 2013}). Although both species were considered diploids, solely the former agrees with the DNA content range of our known diploid species.

Fig. 1 Mitotic and meiotic chromosomes of Ilex brasiliensis and meiotic chromosomes of I. theezans. A, mitotic metaphases of I. brasiliensis EEA-INTA-CA 221 with 40 chromosomes. B, diakinesis of I. brasiliensis EEA-INTA-CA 221 showing 20 bivalents. C-D, diakinesis of I. theezans with 20 bivalents (C, accession EEA-INTA-CA 101; D, accession EEA-INTA-CA 225). Scale bars: 5 µm. 

Prior to the present study, the genome size of southern South American Ilex species was known for I. argentina, I. paraguariensis and I. dumosa (^{Barral et al., 1995}; Gottlieb & Poggio, 2014). The 2C-values of I. argentina and I. paraguariensis recorded herein were ca. 20% lower than the data obtained by Barral et al. (1995), yet the relationship between the 2C-values of both species remains similar between studies (1.94 and 1.91, respectively). These differences could be explained by the methodological approaches employed. Indeed, our measurements on I. paraguariensis and I. dumosa agree with those reported by Gottlieb & Poggio (2014), who also used flow cytometry.

Regarding I. theezans, our results indicate that the mean genome size obtained in eight out of the nine individuals analyzed (from six accessions) was 1.6 to 2.2 times higher than that of the known diploids, and even 8% higher than that of I. argentina. It should be noted that the leaf morphology of I. theezans and I. argentina is quite distinct (Fig. S2), making species misidentification unlikely (Giberti, 1979, 1990, 1994). In contrast, the DNA content of the individual of EEA-INTA-CA 225 -from Paraguay- was similar to that of other Ilex species recognized as diploids (Table 2). Morphologically, I. theezans, I. brasiliensis and I. integerrima resemble and the literature consider they are closely related (^{Ricco et al., 2013}), but the latter species does not occur in Paraguay (Giberti, 1989, 2001). Still, it cannot be completely ruled out that EEA-INTA-CA 225 had been inadvertently misidentified. Surprisingly, the individuals of I. theezans analyzed so far, and those previously reported (^{Andrés & Saura, 1945}; Greizerstein et al., 2004), exhibited 20 bivalents, being thus diploids (2n=2x=40). In this context, the genomic sizes estimated for most accessions of I. theezans are unexpectedly high. It should be noted that, unfortunately, the materials available for meiotic analyses were a pool of individuals -from the same accession-, but not necessarily the same individuals used for genome size estimation. The existence of cytotypes due to intraspecific ploidy-level variation is known to occur in different plant species, including olive, elm, and poplar trees (^{Blonder et al., 2021}). It is well established that the cytotypes could be distributed in different regions and may coexist in sympatry (Kolář et al., 2017). Assuming that the taxonomic identification of the plants held at the Germplasm Bank is correct, the DNA content variation observed in I. theezans could be attributed to the occurrence of cytotypes within Brazil, Argentina and Paraguay. There are other discordant results within the very few chromosome counts known in the genus. For instance, I. verticillata was established as diploid (2n=36) by Jensen (1944) and as a polyploid (2n=72) by Faasen & Nadeau (1976), suggesting the occurrence of diploid and polyploid cytotypes as well. On the other hand, the intraspecific genome size variation detected here in I. theezans may be explained by differences in heterochromatin content, as was documented in maize (^{Realini et al., 2016}, González & Poggio, 2021). Chromosome bandings would aid in elucidating if highly repetitive sequences are involved in this variation. Southern South American species of Ilex yielded different groupings when scrutinized through AFLP genotyping (Gottlieb et al., 2005), metabolomics (Kim et al., 2010) and molecular phylogenetics (Gottlieb et al., 2005; Cascales et al., 2017). In this survey, we could not establish any association among the groupings based on genome sizes and those obtained before with other data type. Gottlieb et al. (2005) associated the greatest number of AFLP bands shown by I. argentina with its polyploidy. Yet, in that study, all the individuals of I. theezans showed banding patterns compatible with diploids. Thus, the evaluation of more individuals from diverse origins is required to further test the existence of cytotypes for I. theezans.

The chromosome number of I. brasiliensis (2n=2x=40) which is reported here for the first time, coincides with that of other diploid co-generic species, and agrees with its determination as a diploid through the 2C-value estimated here. Regarding I. microdonta, although neither roots nor flower buds could be obtained for determining its chromosome number, the estimated 2C-value may indicate that it is a diploid species as well.

CONCLUSIONS

Herein, the exploration of different incubation times in propidium iodide allowed obtaining, in a simple manner, adequate flow cytometry measurements; this could be a recommendable practice when dealing with non-model plants suspecting of producing secondary metabolites, as several Ilex species do.

Our study presented the genome sizes of most of the southern South American species of Ilex for the first time, and revealed, as well, an intraspecific variation of the DNA content in I. theezans. This leads us to propose the existence of cytotypes in this species.