Report of chromosome number and karyotype asymmetry in Schizopetalon Sims. and related genera from the South American tribe Schizopetalae (Brassicaceae)

ABSTRACT

Schizopetalae is one of the representative South American tribes of Brassicaceae, which has remained poorly studied for cytological data. With a single report existing, no additional cytological information is known for the rest of the species and genera of this tribe (Atacama, Mathewsia and Schizopetalon). The present study reports the chromosome number and karyotype asymmetry for five species of Schizopetalon, two for Mathewsia and the only species of the monotypic genus Atacama. While a constant 2n = 18 is confirmed in all analysed samples, larger interchromosomal asymmetry levels were found in Schizopetalon than Atacama and Mathewsia. Both results suggest different processes of fixation and recombination in chromosome architecture among genera, probably associated to life history strategies maintaining genetic diversity under extreme and unpredicted environmental conditions. The present study represents the first report of chromosome number in any South American Brassicaceae in the last 30 years, revealing the need of more cytological studies to contextualize the diversification of Schizopetalae and other closely related tribes.

KEYWORDS:

• Schizopetalae
• Brassicaceae
• Atacama Desert
• chromosome number
• interchromosomal asymmetry

Introduction

Schizopetalae is a South American tribe of Brassicaceae of three genera (Atacama O. Toro, Mort & Al-Shehbaz, Mathewsia Hook. & Arn. and Schizopetalon Sims; Figure 1), occurring in the drylands and Mediterranean zone of Chile and the neighbouring areas of Argentina and Peru (Toro-Núñez et al. 2013). Most of their species are typically found in coastal and mountain habitats of the Atacama Desert in Chile, which is considered one of the driest deserts in the world (Houston and Hartley 2003; Houston 2006a, 2006b). The taxa of Schizopetalae are characterized by their distinctive flowers with putative moth-mediated pollinator attributes (e.g. aromatic flowers, pollen grains with rough surfaces and negative phototropism behaviour in their petals), making them a distinctive group of plants with interest for ornamental purposes (Al-Shehbaz 1989; Toro-Núñez et al. 2013).

While patterns of diversification in Schizopetalae have been well characterized with the use of molecular and morphological data (Toro-Núñez et al. 2013; Salariato et al. 2016), other characters have been explored to a limited extent (Al-Shehbaz 1989). Among the available options, cytological data represent a potential valuable source of evolutionary and taxonomic information for this group, as they can provide important evidence of reproductive dynamics across different groups of angiosperms (Stebbins 1971; Grant 1981; Levin 2002). Cytological reports in Schizopetalae exist only in S. walkeri, with a diploid number of 2n = 18 (Manton 1932) or a haploid number of n = 10 (Al-Shehbaz 1989). Besides this record, no additional information exists about ploidy number or karyotype variation for its remaining taxa.

Given that cytological attributes tend to exhibit dynamic changes of chromosome number and genome size in Brassicaceae (e.g. Johnston et al. 2005; Lysak and Lexer 2006; Warwick and Al-Shehbaz 2006; Lysak et al. 2009), it remains speculative whether cytological data can provide useful evidence for the systematics and taxonomy of Schizopetalae. Hence, the present study aims to report the chromosome number and chromosome asymmetry patterns from several species of Schizopetalon and their sister genera Mathewsia and Atacama.

Materials and methods

Plant material

Seeds from all three genera were collected during 2010 and 2015 field trips. While seeds of Schizopetalon corresponded to the species from the Atacama area (S. biseriatum, S. tenuifolium, S. rupestre and S. arcuatum) and a closely Mediterranean related species (S. brachycarpum; Figure 2), the sister genera Atacama and Mathewsia were sampled for one (A. nivea) and two species (M. auriculata, M. foliosa) respectively (Figure 2). Representative voucher specimens were deposited in the McGregor Herbarium at the University of Kansas (KANU) and the Universidad de Concepcion Herbarium (CONC; Table 1).

Table 1. List of populations, voucher information, number of sampled individuals, chromosome numbers and asymmetry indexes recorded in this study.

Species	Voucher	Locality	Coordinates	N (n)	2n	THL	CVCL	MCA
A. nivea	Toro & Fernandez 60 (KANU)	CHILE. Atacama. Road from El Salvador to Potrerillos.	26°17.263S, 69°30.159 W	4 (4)	18	1.65 ± 0.26	16.71 ± 2.69	11.75 ± 3.07
M. auriculata	Toro & Fernandez 95 (KANU)	CHILE. Atacama. Road to Paso San Francisco.	27°1.201S, 69°39.811W	3 (8)	18	1.86 ± 0.21	18.66 ± 5.69	12.63 ± 1.41
M. foliosa	Toro 120 (CONC)	CHILE. Valparaiso. Playa Tunquen.	33°17.743S, 71°39.224W	1 (1)	18	2.19	14.29	8.87
S. arcuatum	Toro & Lira 43 (KANU)	CHILE. Atacama. Northern beaches close to Huasco.	28°29.92S, 71°14.532W	4 (8)	18	1.62 ± 0.24	14.67 ± 6.56	16.23 ± 4.18
	Toro & Lira 49 (KANU)	CHILE. Atacama. Road to Chañar de Aceituno and Caleta Carrizalillo.	28°56.844S, 71°17.161W
S. biseriatum	Toro & Lira 12 (KANU)	CHILE. Atacama. Road Ruta 5 Norte, close to Copiapo.	27°16.906S, 70°37.247W	3 (4)	18	1.77 ± 0.28	20.91 ± 8.01	17.98 ± 5.60
	Toro & Lira 21 (KANU)	CHILE. Atacama. Road to Barranquilla, Caldera.	27°28.127S, 70°49.671W
	Toro & Lira 39 (KANU)	CHILE. Atacama. Road to Carrizal Bajo and Caleta Totoral.	27°54.343S, 71°5.498W
S. brachycarpum	Toro 57 (KANU)	CHILE. Region Metropolitana. Close to Represa el Yeso, San Jose de Maipo.	33°38.893S, 70°2.239W	3 (5)	18	2.12 ± 0.65	15.51 ± 3.93	16.23 ± 4.18
S. tenuifolium	Toro & Lira 42 (KANU)	CHILE. Atacama. Road to Canto del Agua, heading to Carrizal Alto.	28°9.058S, 70°52.125W	6 (9)	18	1.75 ± 0.23	19.76 ± 5.83	16.54 ± 3.83
	Toro & Lira 41 (KANU)	CHILE. Atacama. Road Ruta 5 Norte, between Vallenar and Copiapo.	28°13.913S, 70°41.34W
	Toro & Lira 33 (KANU)	CHILE. Atacama. Road to El Totoral, from Ruta 1.	28°20.575S, 71°9.24W
S. rupestre	Toro & Fernandez 61 (KANU)	CHILE. Atacama. Road El Salvador – Potrerillos.	26°17.136S, 69°30.344W	9 (20)	18	1.96 ± 0.42	14.31 ± 5.74	19.24 ± 3.40
	Toro & Fernandez 87 (KANU)	CHILE. Atacama. Road to mine Nueva Esperanza.	26°39.676S, 69°8.943W
	Toro & Fernandez 112 (KANU)	CHILE. Coquimbo. Road to mine El Indio.	29°54.031S, 70°3.011W
	Toro 128 (CONC)	CHILE. Atacama. Road to Nevado Tres Cruces National Park.	27°4.094S, 69°14.889W

Notes: N (n), number of individuals and mitotic plates (in parenthesis) per species; TLC, total chromosome length; CV_CL, intrachromosomal asymmetry; M_CA, interchromosomal asymmetry.

Figure 1. Habit and flowers of Atacama (A. nivea), Mathewsia (M. incana), and Schizopetalon (S. tenuifolium).

Figure 2. Distribution of analyzed species of Atacama, Mathewsia and Schizopetalon.

Figures in gray represent known localities from herbarium records. Figures in black represent sampled populations for the present study. Scale bars represent distances in km.

Sampling of karyotypes

Root tips were collected from seeds germinated in Petri dishes, kept in the dark for 48–56 h until radicles were elongated enough (2–4 cm length). Cut radicles were pre-treated in 8-hydroquinoline (2 mM) for 24 h at 4°C. Subsequently, samples were fixed with a fresh solution of ethanol/acetic acid (3:1) at 4°C until ready for use. Squash preparations were conducted using an acidic hydrolysis pre-treatment with HCL 0.5 N for 17 min at 45°C. The material was immediately stained with an orcein solution at 1%. Metaphase chromosome plates were photographed using a Zeiss Axioskop microscope system (Oberkochen, West Germany) with a mounted digital camera. Photographs were digitally edited for contrast, light, and sharpness with ImageJ v. 1.54k (Abramoff et al. 2004). Additional tuning was conducted to correct focal displacement, by using several serial photographs and obtaining a consensus image with the ImageJ plugging extended-depth-of-field (Forster et al. 2004).

Chromosome measurement consisted of recording arm lengths and their respective ratios (Levan et al. 1964). Intrachromosomal asymmetry (M_CA) and interchromosomal asymmetry (CV_CL) indexes were analysed for each population, following Peruzzi and Eroğlu (2013). Additionally, the total haploid length (THL) was included, as a coarse surrogate of genome size (Peruzzi et al. 2009). Measurements were obtained using the software KaryoType v2 (Altinordu et al. 2016).

Statistical analyses

The stability of asymmetry measurements was tested, in order to detect possible differences taxonomically predictable at genus levels. For this purpose, a Kruskal–Wallis test was conducted in pooled measurements of M_CA and CV_CL, followed with a Dunn’s test for multiple comparisons. This analysis was implemented with the dunn.test package in the statistical platform R v3.4 (Dinno 2017; R Development Core Team 2017).

Results

All analysed populations of Atacama, Mathewsia and Schizopetalon revealed 2n = 2x = 18 chromosomes (Figure 3). Levels of variation in asymmetry patterns were too variable to distinguish a specific chromosome formula in Schizopetalon, which ranged from five to seven metacentric pairs and from four to two submetacentric pairs (5m + 4sm, 6m + 3sm, and 7m + 2sm; Figure 3). On some occasions, a subtelocentric pair was possible to distinguish accompanying other observed formulas (5m + 3sm + 1st or 6m + 2sm + 1st; Figures 3 and 4). Despite its potential for cytotaxonomic discrimination, no stability was possible to associate any of these formulas to a particular species or population of Schizopetalon. In Mathewsia, diploid formulas were more stable with a predominance of metacentric chromosomes, with eight metacentric pairs and one submetacentric pair in both M. auriculata and M. foliosa (8m + 1sm; Figures 3 and 4). In the case of Atacama, all plates revealed consistent metacentric chromosomes (9m; Figures 3 and 4). While stable in their recognition, it was not possible to unequivocally determine the individual asymmetry for each chromosome pair (Figure 4).

Figure 3. Metaphasic plates.

Mitotic plates observed in each analyzed taxa of this study. A: A. nivea, B: M. auriculata, C: M. foliosa, D: S. arcuatum, E: S. brachycarpum, F: S. biseriatum, G: S. tenuifolium, and H: S. rupestre. Black arrows indicate submetacentric chromosomes in Schizopetalon.

Figure 4. Asymmetry scatterplots.

Idiograms calculated using the average from the total number of mitotic plates obtained from each species analyzed of Atacama, Mathewsia and Schizopetalon. Bars in each arm represent standard deviation.

Levels of asymmetry revealed differences among the analysed samples, specifically when interchromosome asymmetry was compared at genera levels. M_CA revealed statistically significant larger values in Schizopetalon than Mathewsia and Atacama (x²_MCA = 16.641, df = 2, p < 0.001; Table 1, Figure 5). Instead, neither CV_CL nor THL indicated significant differences among genera (x²_CVCL = 0.86434, df = 2, p = 0.6491; x²_THL = 1.7993, df = 2, p = 0.4067; Table 1, Figure 5). These observations were confirmed with the Dunn’s test, which revealed significant differences in observed values in the M_CA values of Schizopetalon (Table 2).

Table 2. Dunn’s test multiple comparisons for THL, CVCL and MCA indexes. All p-values are adjusted by Bonferroni correction. z:Dunn’s statistic.

	THL		CVCL		MCA
Comparisons	z	p-value	z	p-value	z	p-value
Atacama – Mathewsia	−1.33936	0.2707	−0.24761	1	−0.09689	1
Atacama – Schizopetalon	−1.11301	0.3986	0.342358	1	−2.55672	0.0158
Mathewsia – Schizopetalon	0.616384	0.8065	0.897863	0.5539	−3.49687	0.0007

Bold numbers represent statistical significant comparisons (p < 0.05).

Figure 5. Asymmetry indexes.

Discussion

Cytology has played an important role in the study of angiosperm evolution, as cytological data provides important evidence to evaluate the impact of genetic divergence among species and populations (Guerra 2008). In Brassicaceae, cytology holds a particular place in the study of evolutionary patterns, especially providing understanding for idiosyncratic family-wise patterns of genomic duplication and polyploidization (Lysak and Lexer 2006; Barker et al. 2009; Franzke et al. 2011). Interestingly, while these patterns have been routinely studied in north hemisphere lineages (Europe and North America), other groups have remained mostly unstudied even for the most basic cytological information. Among South American tribes, the very few existing reports are only related to basic chromosome number (Manton 1932; Warwick and Al-Shehbaz 2006), for which not even the 5% of their total recognized taxa have been scrutinized in the three most diversified and representative tribes (Cremolobeae, Eudemeae and Schizopetalae as consulted in Brassibase by 8/24/2017; Koch et al. 2012; Kiefer et al. 2014).

Despite the fact that no significant cytotaxonomic evidence was observed for taxonomic discrimination, two important elements could reveal useful information to contextualize pattern diversification in Schizopetalae. First, compared to other tribes, the basic number of chromosomes remains largely stable in Schizopetalae, corroborating the observations of ploidy number made by Manton (1932) in Schizopetalon species and expanding this to Atacama and Mathewsia. This result is surprising, considering the rare condition of constant chromosome number among most Brassicaceae tribes (Manton 1932; Warwick and Al-Shehbaz 2006). Changes in chromosome number are usually expected in this family, owing to the effects that recurrent events of allopolyploidization and reduction by aneuploidism have represented for the retention of different ploidy numbers of chromosomes within lineages (Lysak 2009). While the results of our study might be suggestive of a lack of hybridization or polyploidism in Schizopetalae, further studies are needed given the inherent dynamics controlling chromosome changes in Brassicaeae. For example, mechanisms that preserve small variable genome sizes are typically found in Brassicaceae (Johnston et al. 2005; Lysak et al. 2009), making questionable the inference of hybridization using referential ploidy numbers as a sole source of evidence (Lysak 2009). Another limitation to this hypothesis is the effect that hybrid homoploid speciation presents retaining stable diploid numbers after hybridization events (Soltis and Soltis 2009); a process that cannot not be discarded for Brassicaceae (e.g. Cheung et al. 2015)

Second, higher levels of interchromosomal asymmetry found in the karyotype of Schizopetalon, in comparison to Atacama and Mathewsia (M_CA; Figure 5), could be linked to the effect that life history presents on different rates of recombination and fixation. Given the effect that habitat stress presents on plant reproduction systems, high levels of self-pollination should be found in annual plants – like Schizopetalon – subject to hyper aridity in the Atacama Desert area (González and Pérez 2010). This type of reproduction should produce high frequency of chromosome recombination, resulting from an adaptive strategy to preserve genetic diversity in the presence of high levels of selfing (e.g. Charlesworth 2003; Wright et al. 2008). As a result, high levels of karyotype diversity should be rapidly fixed within and among populations, as seed banks promote the retention of individuals with new karyotypes in local populations (Levin 2002). The fixation of new karyotypes should be also enhanced from the effects of uneven seedling recruitment under unpredictable environments, a pattern recurrently found in community guilds from the Atacama Desert flora (Vidiella et al. 1999). This reproductive adaptation, while also influential, could not represent a determinant factor for the fixation of karyotypes in Atacama and Mathewsia. Instead, their perennial habit with longer periods of survival and more stable generational recruitment should promote more flexibility on selfing reproduction; likely decreasing the chances of positive selection for intense chromosome recombination and fixation.

While our results did not fully support the use of cytological data for taxonomic discrimination in Schizopetalae, differences in asymmetry levels could be also seen as a promising element for – at least – genera discrimination. For example, different karyotype formulas support the separation of Atacama nivea from the rest of Mathewsia, resulting from consistent low levels of asymmetry differences detected among samples of both taxa (9m vs 8m + 1sm; Figure 3). This observation is in line with reports of predominant metacentric chromosome organization and limited asymmetry variation in Brassicaceae, which have served useful for the taxonomic characterization of several species and genera (e.g. Yue et al. 2004; Zhou et al. 2008; Martin et al. 2016). Nevertheless, despite this promising use, prospects of using karyotype formula differences for taxonomic discrimination in Atacama and Mathewsia should be considered only tentative, as further scrutiny is necessary to control the effect of asymmetry differences in very small chromosome sizes (< 2µm) obtained from the observation of a limited number of species and populations. Yet, this limitation should not affect our interpretations of overall differences in asymmetry levels, given the consistent statistical differences observed in M_CA values of Schizopetalon compared to those of Atacama and Mathewsia (Figure 5).

More refined approaches will certainly improve the evidence to support or reject the hypotheses made on chromosome changes of this study. Experience with techniques like FISH-GISH staining demonstrates the potential of molecular cytotaxonomy for studying patterns of differentiation in Brassicaceae (Lysak 2009). Also, as molecular resources become more extrapolatable from model systems like Arabidopsis thaliana, cytological studies can provide further insight on the study of recombination patterns into other non-model Brassicaceae taxa (Schranz et al. 2006; Franzke et al. 2011). Keeping those promising approaches in mind, we expect that this study encourages the conduction of more cytological studies in Schizopetalae and other South American lineages, which still remain a largely understudied group of the highly diversified Brassicaceae family (i.e. Expanded Lineage II; Franzke et al. 2011).

Acknowledgements

The authors thank the National Forest Corporation of Chile (CONAF), especially to the Atacama Administrative Division and the personal from the Llanos del Challe National Park and Nevado Tres Cruces National Park, for their logisticalassistance and collection permits.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by CONICYT N° 3160453 (Chile); The University of Kansas (US) Fratcher Botany Fellowship Scholarship; Sigma Xi (US) Grants-in-Aid of Research Awards.