Genome size estimation of three endemic plant taxa from Malta

Agius, D. R.,


Introduction
The Mediterranean basin is a hotspot of plant biodiversity (Myers & al. 2000).Sixty percent of the native plant species are endemics to the region with most being restricted to a single well-defined geographical area.It has also been observed that in the Mediterranean region, endemic species tend to occur on rocky habitats, on steeper slopes and in open vegetation rather than woodland (Thompson 2020).The Maltese archipelago, being central to this biodiversity hotspot, has a flora rich in rare endemic and relict species.This is also attributed to its geomorphology, edaphic factors, climatic fluctuations over the past eons and past geological activity amongst others (Cassar & al. 2008; Brullo & al. 2020).
Malta is one of the most densely populated countries.This creates strong pressures that are reshaping the distribution of the endemic taxa and changing the natural landscape.Conflicting land uses are resulting in loss of habitat for the endemics, with most of them becoming extremely rare and critically endangered.Many of these species are relevant to phylogenomic studies of the major taxa (Rokas & Abbot 2009; Brullo & al. 2020).A restricted and fragmented habitat is also resulting in loss of biodiversity within these taxa.This genetic diversity is currently of scientific interest since it confers adaptability to the extreme climatic conditions found locally (Scheben & al. 2016).It can provide insight into these characteristics in view of the worldwide need of crops adapted to such climates, which are becoming more prevalent in wider geographical areas due to climate change (Rajpal & al. 2023).Genomic studies will address these issues including the conservation of this diversity (Theissinger & al. 2023).Of special interest are the three endemic and threatened taxa Cheirolophus crassifolius (Bertol.)Susanna, the Maltese rock-centaury, which is the islands' own national plant, Sedum album L. subsp.rupimelitense Mifsud, Stephenson & Thiede and Anthyllis hermanniae L. subsp.melitensis Brullo & Giusso del Galdo.
Cheirolophus crassifolius (Asteraceae: Cardueae: Centaureinae), a rupestral endemic of the Maltese islands, is considered a relict species of the preglacial circum-Mediterranean distribution of this genus (Susanna & al. 1999).Genome size has been recorded for twenty-four out of the twenty-seven species in this genus (Garnatje & al. 2007; Hidalgo & al. 2017).Garnatje & al. (2007) reported a statistically significant difference between mean 2C nuclear DNA contents of continental (1.58 pg) and insular (1.38 pg) Cheirolophus species.The smaller genome size of the insular species was attributed to selection pressures linked to speciation in restricted space or due to founder effects.The genome size of C. crassifolius was not reported in this study.Both chloroplast and nuclear DNA sequences have shown that C. crassifolius forms the most basal lineage in this genus (Hidalgo & al. 2017).This Maltese endemic is designated as critically endangered (Stevens & Lanfranco 2006).
Sedum album subsp.rupimelitense is a rupestral endemic, which is also critically endangered (Mifsud & al. 2015).S. album reverts to CAM photosynthesis under drought conditions, making it of prime interest in agricultural productivity research (Wai & al. 2019).Cytologically it is heterogenous with a base chromosome number of 17 from which a polyploid series is derived.The documented genome size for a diploid S. album is 0.15 pg ('t Hart 1991).Wai & al. (2019) reported genome size of 0.62 pg for a tetraploid individual.The Maltese subspecies, which reproduces asexually, has 2n = 8x = 136 but undocumented genome size (Mifsud & al. 2015).
Anthyllis hermanniae L. species complex is divided in several disjunct and mostly isolated populations in Asian countries and the north-eastern Mediterranean.This complex was recently revised, to include several subspecies which are considered schizoendemics.A. hermanniae subsp.melitensis, has been consequently identified as a Maltese endemic.It is given the status of endangered in the IUCN Red List (Brullo & Giusso del Galdo 2006).The members of this genus are typically diploid with 2n = 2x = 12, 14 (Goldblatt 2007).The genome size of A. hermanniae and its Maltese subspecies have not been estimated yet.
In the past fifty years, genome size estimation has enabled a multitude of investigations into biological patterns and processes (Kron & al. 2007).The genome size is the amount of DNA (C-value) in the haploid gametic nucleus (Doležel 2005), quantified in picograms (pg) or megabase pairs (Mbp), where 1 pg of nuclear DNA amounts to a length of 978 Mbp (Doležel & al. 2003).Genome size estimation by flow cytometry (FCM) is the gold standard of techniques for this metric (Al-Qurainy & al. 2021).FCM is not destructive which is critical when studying endangered species.Genome size estimation is one of the initial steps in genomic studies (Rhie & al. 2021).Its importance is evident in the fact that online databases documenting C-values have seen a steady input of data in the past decades.
Bioinformatics based methods for genome size estimation are becoming increasingly popular, although their level of accuracy, especially in highly repetitive plant genomes, is still to be investigated (Pellicer & Leitch 2020).
The main purpose of this work is to estimate the genome size of the three endemic, and threatened taxa from the islands of Malta, C. crassifolius, S. album subsp.rupimelitense and A. hermanniae subsp.melitensis using FCM.

Materials and Methods
Plant material collection.-Leaves from five different individuals of each taxon were used in this investigation.Table 1 lists the sites of specimen collection.In the case of Anthyllis hermanniae subsp.melitensis and Cheirolophus crassifolius, one sample was from a plant cultivated on the premises of the local Plant Protection Directorate at Lija, Malta.Collection and shipment of samples were done according to permit EP 1288/22 issued by the Environmental and Resource Authority of Malta.Photographic records were kept of the plants, together with their geographic coordinates.Each plant was tagged (with a reference number attached to the stem of the plant using a PVC wrap) for future reference.
Sample preparation and cytometric measurements.-The FCM was carried out at the Centre of Plant Structural and Functional Genomics (Institute of Experimental Botany, Czech Republic) using a Sysmex CyFlow ® Space flow cytometer (Partec GmbH, Goẗtingen, Germany) installed with the software FloMax and equipped with a 532 nm green high-grade solid-state laser.
All leaves chosen for this analysis were young and visibly free from infection by parasites.The internal standard and isolation buffer were chosen so that the coefficient of variation (CV) was kept to a minimum (Loureiro & al. 2007).Mechanical isolation of the plants' nuclei was carried out according to Galbraith & al. (1983).This involved chopping the sandwiched leaf material from the studied plant and internal reference standard together, in isolation buffer, using sharp razor blades in glass Petri dishes.The homogenate was filtered through a double layer of 50 µm nylon mesh (Silk & Progress, CR) and stained with the fluorochrome propidium iodide (PI) at a concentration of 50 mg/mL and supplemented with 50 mg/mL RNase.

Anthyllis hermanniae melitensis
Table 1.Site of specimen collection for each species (N/A: not applicable).
Woody plant buffer (Loureiro & al. 2007) was found to be the optimal buffer for A. hermanniae subsp.melitensis.In this case the internal standard used was Solanum lycopersicum L. cv.Stupické with a 2C DNA content of 1.96 pg (Doležel & al. 2007).
CyStain PI OxProtect Kit (Sysmex, United Kingdom), used according to manufacturer's instructions, was found to be the optimal system for the genome size estimation of C. crassifolius and S. album subsp.rupimelitense.The internal standards used were Glycine max Merr.cv.Polanka with a 2C DNA content of 2.50 pg and Zea mays L. cv.CE-777 with a 2C genome size of 5.43 pg respectively (Doležel & al. 2007).A minimum of five thousand events were analysed for each run.To improve accuracy and precision, the genome sizes were determined for each species as the mean of five individual specimens and several technical replicates (Table 1) to enable the standard error of the mean to be calculated.The technical replicates were spread out on different days.
The obtained histograms were visualised using FlowJo software (Version 10.8.2, Treestar, Ashland, OR, United States).The sample 2C DNA content was calculated according to the formula: Where MFI is the mean fluorescence intensity, the standards used were as specified above for each sample plant and the 2C DNA content is given in picograms (pg).
A one-way Analysis of Variance (ANOVA) was performed to compare the effect of each individual replicate (per taxon) on estimated genome size measured using FCM.The ANOVA test was implemented in R (version 4.3.0).

Results and Discussion
The results obtained from flow cytometric analysis of PI-stained nuclei are summarised in Table 2. Figure 1 illustrates representative histograms of the genome size investigations.All three genomes in this study are on the smaller range in the plant kingdom (Pellicer & Leitch 2020).In our design we collected five individual plants for each of the three species: C. crassifolius, S. album subsp.rupimelitense and A. hermanniae subsp.melitensis.The number of technical replicates per individual specimen ranges from three to six (Table 1).The ANOVA test reported no statistically significant difference between the individual specimens for each species using an α-threshold of 0.05.F-values, P-values and degrees of freedom for each species are reported in Table 3.
We report that the 2C-value for Cheirolophus crassifolius is 1.95 pg, equivalent to a genome size of 0.98 pg (954 Mbp).This value is higher than a previously reported one of 0.9 pg, for the same species, by Hidalgo & al. (2017) which was estimated using a single specimen from the Orto Botanico of the Università degli Studi di Palermo in Italy.The value reported in the present study is the highest recorded for this genus in the Genome Size in Asteraceae Database (Release 3.0) (Garnatje & al. 2011).It also does not fall within the range of genome size variation documented for insular members of this genus by Garnatje & al. (2007).It is even higher than the mean value for continental members of this genus reported in the same work, hence it does not fit the hypothesis, based on statistically significant results, that insular members have a smaller genome size than continental members in this genus.It also implies that there is an incremental increase in genome size in C. crassifolius when compared to the Mediterranean ancestral species C. uliginosus.This also does not fit the finding that there is a smaller genome size in the derived species within Cheirolophus (Garnatje & al. 2007).This might be interpreted as confirmation of basal position of this species within this genus which has a trend of genome size reduction in the derived species as showed by Hidalgo & al. (2017).Further genomic studies within this genus are required to shed light on the reasons for this larger genome size.
Sedum album subsp.rupimelitense was found to have an unreduced nucleus with a DNA content of 2.11 pg, amounting to a holoploid genome size of 1.05 pg (1,027 Mbp).The monoploid genome size (1Cx) is 0.26 pg equivalent to 254 Mbp of DNA.The 1Cx value, is a quantitative measure of the amount of DNA in the basic set of chromosomes.It is calculated by dividing the 2C-value by the ploidy level (Leitch & Bennett 2004).The value of 1Cx for S. album documented here is higher than the 1Cx value documented by 't Hart (1991) for a diploid S. album.This is counter to the literature finding that 1Cx values tend to decrease with increased ploidy (Leitch & Bennett 2004). However, 't Hart (1991) does not specify the method by which genome size was estimated (van Prooijen-Knegt & al. 1980; Doležel & al. 1998).This general trend, showing a decrease in 1Cx values with increasing ploidy is attributed to the predominance of deletion over insertion mutations, gene fractionation and unequal recombination which eliminates retrotransposon sequences (Bennetzen 2002).Wai & al (2019) reported a holoploid genome size of 0.62 pg and a monoploid genome size of 0.31 pg for a tetraploid S. album while k-mer analysis gave a monoploid genome size of 256 Mbp (equivalent to 0.26 pg of DNA) for the tetraploid.Our results are more concordant with the results published by Wai & al. (2019) for S. album.Further studies into the genome size of this taxon are required.The second minor peak in the histogram of S. album subsp.rupimelitense (Fig. 1B) represents the G2 nuclei of this species and as expected its position shows doubling in the DNA amount of the G1 nuclei.
Our investigations shows that the nuclear DNA content of Anthyllis hermanniae subsp.melitensis is 1.04 pg amounting to a genome size of 0.52 pg (509 Mbp).This is the third Flora Mediterranea 33 -2023 219 Table 2. Flow cytometry results for the three endemic taxa.documented C-value within this genus of 23 species.It falls well within the range for this genus of 0.50 -0.67 pg DNA (Pellicer & Leitch 2020).The chromosome number within the sporophytes of this genus varies from 12 to 14 chromosomes (Goldblatt 2007).This finding supports the indication that genome size within this genus is on the smaller side within the plant kingdom.This is the first study of nuclear genome size of Maltese endemic plant taxa.Herein we report the genome size of C. crassifolius, the Maltese rock-centaury, and two recently described taxa S. album subsp.rupimelitense and A. hermanniae subsp.melitensis.We determined the genome size by means of FCM.All three species inhabit Special Areas of Conservation.The small geographical area they inhabit, and anthropogenic factors make these taxa endangered, some of them critically so.
The genera Cheirolophus, Anthyllis and Sedum have been studied extensively ('t Hart 1991; Susanna & al. 1999; Brullo & Giusso del Galdo 2006; Garnatje & al. 2007; Pellicer & Leitch 2020).However, most of these works do not include the Maltese endemics making the available knowledge incomplete.The results presented here shed some additional light into the evolutionary history of these genera and can also be used in further studies including studies into the characteristics that give them adaptability to the extreme local hot and dry environment, which is becoming more prevalent worldwide.

Fig. 1 .
Fig. 1.Representative histograms of flow cytometry data showing 2C DNA content of 3 taxa endemic to Malta.The x-axis shows channel as a function of relative fluorescence intensity, while the y-axis represents the number of nuclei.(A) C. crassifolius (left peak) and G. max Merr.cv.Polanka (right peak); (B) S. album subsp.rupimelitense (left peak) and Z. mays L. 'CE-777' (right peak).The second minor peak in the histogram, at position 173, represents the G2 nuclei of this species.(C) A. hermanniae subsp.melitensis (left peak) S. lycopersicum L. cv.Stupické (right peak).(CV: coefficient of variation).

Table 3 .
Results of ANOVA tests between individual specimens for the three species.