This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
INTRODUCTION
Mediterranean, arid, and semi-arid areas are important centres of biodiversity (Giovino et al., 2016) and can host unique plant species with crucial roles in such fragile ecosystems (Giovino et al., 2014). However, due to a high anthropogenic pressure and climate change, the resilience of these ecosystems decreases. This also caused by the expansion of alien, often aggressive, species (from either plants or animals), which directly or indirectly menace many native taxa, especially the endemic ones (Badalamenti, 2016; Brundu, 2013; Giovino et al., 2016). Cistus crispus L. is a western Mediterranean species typical of the thermo-Mediterranean zone. It is a medium-sized, evergreen shrub with plant height usually up to approximately 100 cm. This species has various uses including its application as an ornamental and/or medicinal plant, whose extract can have strong fungicidal/fungistatic effects (Bouyahya et al., 2018), or it can be used for xeriscaping and recolonisation of fragile environments (Batista et al., 2017; Gómez-Zotano et al., 2017), where it also plays an important role for feeding bees (Ortiz, 1994). C. crispus has been classified as endangered in the Red Data Book of plants of Italy (Conti et al., 1992). In Italy, it is presumed to be native only in Sicily (Rizzotto, 1979), where it is very rare and localised (Bartolo et al., 1994). C. crispus has also been reported as a casual xenophyte in Liguria (Gentile and Gentile, 1994). It has been shown that Cistus species have a polyphyletic genetic structure and an insular diversification (Carlier et al., 2008; Guzmán and Vargas, 2005). Indeed differences were frequently found, in various species, between genotypes from Sicily and others from North Africa despite the share of similar soils (Barrajõn-Catalán et al., 2011; Giovino et al., 2015c; Said et al., 2016). However, in the Sicilian area, the environment is subjected to strong anthropogenic pressure. Such a condition does not allow C. crispus to easily renew and can have drawback for repopulating the plant community, especially if considering that such a species is also collected for ornamental purposes. This calls for a genetic and environmental protection of this species through adoption of ad hoc conservation practices, especially if considering its rarity and that Cistus species can undergo hybridisation with non-native species from the same genus (Navarro-Cano et al., 2017). Such measures of propagation should rely on both sexuate and asexuate reproduction methods to conserve the intraspecific variability and increase the number of individuals available for ornamental purposes and to recolonise the native areas. However, orthodox seeds, such as those of Cistus (Papafotiou et al., 2000; Scuderi et al., 2010), frequently have reproduction strategies, including an asynchronous ger mination, which reduces the efficacy of sowing to rapidly recolonise native areas. This often occurs for other species whose embryo has a maturation stage after seed dispersal from the same areas (Giovino et al., 2015b, 2015d). In addition, it hampers to address the use of this species as ornamental or medicinal plant or for xeriscaping, given its aesthetic value, high resistance to stresses, and content in secondary compounds with potential industrial uses, as showed for other species (Giovino et al., 2014; Kubica et al., 2017). In addition, Cistus spp. showed own rooting from cuttings (Papafotiou et al., 2000), which hampers its uses.
In vitro propagation has important applications for both the protection of endangered species and genotype conservation (Preil, 2003; Rout and Jain, 2004). Micropropagation by direct organogenesis of endangered species is a valid method in comparison with traditional propagation practices to avoid depletion of natural populations, which further implies the disturbance of the native sites for seed collection and a high amount of nursery/greenhouse/field work for plant establishment.
The aim of this study was thus setting a micropropagation protocol for this species. In particular, we tested the response of explant proliferation of various protocols of explant sterilisation under micropropagation and rooting after and hormone (i.e., growth regulators (GR)) treatments at both the micropropagation stage and rooting stage.
MATERIALS AND METHODS
Culture initiation
The propagation material was collected from a plant from the population of Colle San Rizzo (Messina province, Sicily, Italy). Explants used were 4- to 5-mm-long apical sprout with two verticils and a bud at its base, coming from primary and secondary branches (Ruta and Morone-Fortunato, 2010) with no evident damages or pathogenic attacks. The explants were collected from actively growing upper branches during the year of collection. The micropropagation procedure was performed under horizontal laminar flow after device sterilisation under ultraviolet-C radiation at 15 W per 30 min.
Effect of sodium hypochlorite concentration (CHC) and time of sterilisation (TS)
Explants were sterilised by washing in water and then in ethanol-water solution (70:30 v/v, Sigma-Aldrich), after which sterilisation treatments were sodium hypochlorite (5% active chlorine, w/v, Sigma-Aldrich) concentration 2.0%, 2.5%, 3.0%, or 3.5% (in water by v/v, mixed at 25°C and atmospheric pressure) where 20 drops · l−1 of Tween 20 (Sigma-Aldrich) were added. Sterilisation treatments lasted 25 or 35 min. Soon after sterilisation, explants were rinsed various times in sterile distilled water. The explants described earlier were placed in Magenta™ B-cap jars (Sigma-Aldrich) filled with 50 mL of a Murashige and Skoog (MS) macronutrient medium (Murashige and Skoog, 1962) (Sigma-Aldrich), Nitsch and Nitsch micronutrients (Nitsch and Nitsch, 1969), Fe-EDTA (30 mg · l−1), thiamine (0.4 mg · l−1), myoinositol (100 mg · l−1), agar (Bacteriological Agar, No. 1 Oxoid; 8 g · l−1), and sucrose (30 g · l−1). The experimental unit was a jar with 20 explants (Figure 1A). Each treatment was replicated four times, kept at 24°C ± 1°C under 50 μmol · m−2 · s−1 light by cool-white fluorescent tubes (Osram), and a 16 h/8 h day–night photoperiod. The experiment was arranged according to a randomised block design (RBD). After 4 weeks, percentage of sterile explants and their vitality were computed as number of vital explants above the number of sterile explants. An explant was considered vital when free of browning or necrosis.
Figure 1
(A) Experimental unit at the beginning of the experiment: apical sprout from primary and secondary branches. (B) Experimental unit at the time of establishment in vitro.
Propagation
Effect of benzylaminopurine (BA) concentration
Shoot clumps obtained during initiation (4 weeks after initiation) from the treatment with the highest percentage of vital explants were used to test the proliferation rate after increasing concentration of BA (Sigma-Aldrich) in the propagation. Explants were placed in jars filled with an MS medium as earlier and supplemented with 0.88, 1.78, 3.55, 7.45 mM BA or were untreated (referred as 0). The medium was solidified with agar as earlier. Explants were maintained as in the previous experiment. The experimental design was a RBD with four replicates. One subculture lasted for 4 weeks. After 10 subcultures, proliferation rate, (computed as number of shoots per explant), number of adventitious roots per explant, and percentage of healthy explants (referred as c.p., n.r., and h.e., respectively) were computed. An explant was considered healthy when it was free of browning or necrosis and leaves appeared turgid.
Effect of kind (BA, kinetin (Kin), dimethylallylamino purine (2iP)) and concentration of cytokinins (CYTs)
Stem explants (15 mm long, two-node sections) obtained from in vitro plantlets of Cistus multiplied on a MS medium as earlier (Sigma-Aldrich), 1.5 μM indole butyric acid (IBA; Sigma-Aldrich), and agar as earlier and pH set to 5.7 were used for the experiment. The experimental unit was a jar with 20 explants. The stock of shoot cultures was increased by sub-culturing every 4 weeks on a complete MS medium. Experimental design was a RBD replicated four times. The effect of three CYTs was investigated, BA (molecular weight (MW): 225.25), Kin (MW: 215.21), and 2iP (MW: 203.24) (Sigma-Aldrich), which were used in the following concentrations: 0.3, 0.6, and 1.2 μg · g−1, respectively, and an untreated control referred as 0.0. After 5 weeks, the coefficient of shoot proliferation was calculated.
Rooting
Effect of auxin
Stem explants (~15-mm-long, two-node sections) were obtained from in vitro plantlets of Cistus multiplied on a MS medium as earlier, 1.5 mM IBA and agar as earlier, and pH set to 5.7 in the present experiment. The experimental unit was a jar with 15 stem explants ~150-mm-long, two-node sections. Treatments were as follows: application of rooting hormone (RH; GR): indole acetic acid (IAA) or IBA (Sigma-Aldrich) and concentration (Cgr) of each of these RHs: 0.5, 1.0, and 2.0 mg of compound g−1 medium and an untreated control referred as 0.0. After 5 weeks, root number per cutting was counted.
Acclimatization
Plantlets were transferred to Jiffy® pots, and plant acclimatisation was evaluated after 4 weeks in a heated greenhouse according to Lazzara et al. (2017).
Computations and statistical analysis
Before starting the analysis, all data were checked for normality by means of the Shapiro–Wilk test (proc. Univariate, SAS/STAT 9.2, SAS Institute Inc., Cary, NC, USA) since the number of samples was lower than 2000. All data had a normal distribution, including those of ratios or percentages, and thus were analysed with no transformation.
Data from each experiment were subjected to analysis of variance by means of the GLIMMIX procedure in SAS/STAT 9.2 environment, according to the experimental design. Differences among mean values of treatments with more than two statistical levels and interactions were compared by applying Tukey's “t-grouping” at the 5% probability level to the LSMEANS estimate. See the supplementary material in Saia et al. (2020) for both a description of the procedure and the SAS package model applied.
RESULTS
Culture initiation
Percentage of sterile explants (Figure 2) increased by 32% and 20% when increasing CHC from 2.0% to 2.5% and 2.5% to 3.0%, respectively, with no differences between 3.0% and 3.5%. Similarly, increasing TS raised the sterile explants by 14%, with no CHC × TS interaction.
Figure 2
Percentage of sterile (left panel) and vital explants (right panel) of Cistus crispus at increasing sodium hypochlorite concentration (CHC) and time of sterilisation (TS). Data are values of mean ± standard error. For sterile explants, CHC: F = 31.2, p < 0.001; TS: F = 423.6, p < 0.001; and CHC × TS: F = 2.5, p = 0.088. For vital explants, CHC: F = 135.2, p < 0.001; TS: F = 276.7, p < 0.001; and CHC × TS: F = 4.0, p = 0.022. When CHC × TS was significant, treatments were separated by t-grouping of the LSMEANS estimate. Treatments with a letter in common are not different at t0.05-grouping.
Percentage of vital explants (Figure 2) decreased by 38%, on average, when TS increased from 25 to 35 min, and such a difference mostly occurred at CHC higher than 2.0%, although with a scarce interaction. Varying CHC resulted in a variation of the percentage of vital explants, with a boost from 2.0% to 2.5% and then a constant decrease while further increasing CHC to values close to 0.
Propagation
Effect of BA concentration
Coefficient of proliferation (c.p.), number of roots (n.r.), and percentage of healthy explants (h.e.) peaked at 1.78 mM BA concentration and decreased at higher BA concentrations (Figures 3 and 1B, respectively). In particular, h.e. variation was particularly evident at increasing BA from 0.88 to 1.78 μM, whereas they were scarce when adding 0.88 μM BA compared to the untreated control. The decrease in these variables was almost constant when increasing BA from 1.78 to 7.45 μM.
Figure 3
Coefficient of proliferation (c.p.), number of root (n.r.), and percentage of healthy explants (h.e.) of Cistus crispus cuttings at increasing benzylaminopurine (BA) concentration. Data are values of mean ± standard error. c.p.: F = 652.3, p < 0.001; n.r.: F = 15.1, p < 0.001; and h.e.: F = 314.3, p < 0.001. Within each variable, treatments with a letter in common are not different at p > 0.05 according to the Tukey's test applied to the LSMEANS estimates’ differences.
Effect of kind (BA, Kin, 2iP) and concentration of CYTs
Application of GR increased by 5.1% to 192.3% axillary shoot proliferation. When applying to 0.3 μg · g−1 of 2iP and Kin, axillary shoot proliferation increased compared to the untreated control, whereas no differences were found for BA (Figure 4).
Figure 4
Coefficient of axillary shoot proliferation of Cistus crispus microcuttings at increasing concentration (CCYT) of cytokinins (CYTs): benzylaminopurine (BA), kinetin (Kin), or dimethylallylamino purine (2iP). Data are values of mean ± standard error. CYT: F = 67.3, p < 0.001; CCYT: F = 75.2, p < 0.001; and CCYT (CYT): F = 20.6, p < 0.001. DFnum and DFden of CCYT (CYT) were 4 and 27, respectively. Treatments with a letter in common are not different at p > 0.05 according to the Tukey's test applied to the LSMEANS estimates’ differences.
A further increase from 0.3 to 0.6 μg · g−1 GR increased axillary shoot proliferation of C. crispus in BA and Kin, but not in 2iP. An additional increase in the GR concentration from 0.6 to 1.2 μg · g−1 did not affect axillary shoot proliferation in 2iP and BA and decreased it in Kin, which showed similar coefficients compared to the other GR.
Rooting
Effect of auxin
Application of IBA promoted root emission 26.5% less than IAA (Figure 5) and resulted in no promotion at 2.0 μg · g−1. RH increased root number especially at 0.5 μg · g−1 (+100% and +50% in IAA and IBA, respectively, compared to the untreated control).
Figure 5
Root number of Cistus crispus microcuttings at increasing concentration (CAUX) of auxin (AUX): indole acetic acid (IAA) or indole butyric acid (IBA). Data are values of mean ± standard error. AUX: F = 11.4, p = 0.003; CAUX: F = 8.6, p = 0.002; CAUX (AUX): F = 0.007, p = 0.935. DFnum and DFden of CAUX (AUX) were 2 and 18, respectively.
Acclimatisation in the greenhouse during 4 weeks was complete (100%) irrespective of the treatments applied in Experiment 2. Data were thus not statistically processed.
DISCUSSION
Micropropagation for conservation purposes of other endangered Cistus species has been proposed with contrasting results (Aregui et al., 1997; Gatti et al., 2004; López-Orenes et al., 2013; M’Kada et al., 1991). In the present experiment, we obtained a coefficient of proliferation of 1.3–2.0 when no GR were applied, which is similar or higher than the values found in Cistus heterophyllus (López-Orenes et al., 2013). Such a rate increased by 2.9–4.2 fold when the most suitable concentration of GR was applied. These results were better than those observed in C. × purpureus (M’Kada et al., 1991), Cistus clusii (Ruta and Morone-Fortunato, 2010), or Cistus ladanifer (Boukili et al., 2017) and similar to those found in Cistus salvifolius (Louro et al., 2017), where higher hypochlorite concentration was used than that in the present study, but lower than that in other Cistus species obtained by micropropagated plantlets from callus (Madesis et al., 2011) or seeds (Zygomala et al., 2003). Zygomala et al. (2003) also found a higher percentage of vital explants than the present study. Differences in our results compared to those of Madesis et al. (2011) and Zygomala et al. (2003) can be due to both the different genotypes tested and ability of the treatment to sterilise the explant without harming its vitality. In particular, the explant vitality may also depend on the explant antioxidant response to the treatments applied and kind of pathogens in the explant (Ortuño et al., 2018), age of the starting material or position in the mother plant, relative response to nutrients and light (Martínez-Estrada et al., 2016; Mills, 2009; Yoo and Lee, 2017), and higher amount of Na+ used in Zygomala et al. (2003) compared to the present study. Na+ concentration could have affected the K+/Na+ and Ca2+/Na+ on the plant tissues and thus increased their antioxidant capacity. This latter trait can determine the outcome of the micropropagation (López-Orenes et al., 2013). Application of sodium hypochlorite to the explant resulted in a higher percentage of vital explants at the 2.5% concentration and at the time of exposure of 25 min compared to 35 min. Interestingly, a longer time (35 min) of exposure with the same concentration did not consist of an increase in the percentage of vital explants. This implies that sterilisation of the explants also damaged some explant or bud traits. Coefficient of propagation was relatively high even when no CYTs or auxin was applied; thus, these changes could be due to the hormones’ ratios in the explant, as suggested by the low differences in root numbers at increasing BA or coefficient of proliferation at increasing BA or 2ip, whose activity compared to Kin reported to be lower (Ortuño et al., 2018). In this work, the application of low concentration of CYT regulators increased the root number less than auxin (1.26-fold and 2.00-fold, respectively). This implies that concentration of auxin was likely impairing the micropropagation process in these explants. The present results agree with those found in C. clusii (Ruta and Morone-Fortunato, 2010), which also found a complete acclimatisation with two substrates when no mist was applied. Similarly, Pela et al. (2000) found that RHs reduced explant proliferation. Nonetheless, the present results could have also depended on the age of the mother plants and culture age, as shown in other species (Öncel and Erişen, 2017; Park et al., 2017) and on the specific response to any of the auxin used (Öncel and Erişen, 2017). For example, it was found that IBA can induce more roots than IAA or other plant hormones depending on the species (Jana et al., 2017; Ou et al., 2015) and that differences between IBA and IAA can depend on the substrate (Sedlák and Paprštein, 2011). Lastly, it was shown that genotypic response to IBA and other hormones can vary even within species and depending on the micropropagation conditions (Jamwal et al., 2016; Venkatasalam et al., 2015).
The increase in the number of roots found in the present work has implication for both the production of Cistus as an ornamental species and for conservation in the native site, if considering that C. crispus has a root system less expanded than other Cistus species and a lower growth rate (Silva et al., 2003; Silva et al., 2002; Latorre et al., 2007).
CONCLUSIONS
The study of the propagation of endangered plants is oriented toward methods, such as micropropagation, that can limit further impoverishment of rare plant material (Fay, 1992; Giovino et al., 2015a; Manokari and Shekhawat, 2018). In conclusion, the protocol used here showed very high performance in number of plants obtained per unit time and could be used for various aims, including its mass production with low costs or the increase in the presence of C. crispus in the natural environment. This could also reduce the genetic introgression from other similar species that can hybridise with C. crispus. This is particularly important when considering that diversification in Cistus depends on the geographical isolation between the species (Fernández-Mazuecos and Vargas, 2011; Navarro-Cano et al., 2017).
and 
