Morphological aspects of fruits , seeds , seedlings and in vivo and in vitro germination of species of the genus Cleome 1

The genus Cleome is widely distributed in drier areas of the tropics and subtropics. Cleome dendroides and C. rosea are Brazilian native species that occur mainly in Atlantic Forest and sandy coastal plains, respectively ecosystems negatively affected by human impacts. Cleome spinosa is frequently found in urban areas. Many Cleome species have been used in traditional medicine, as C. spinosa. In the present work, was investigated C. dendroides, C. rosea and C. spinosa germinative behavior under in vivo conditions, as well as was established suitable conditions to in vitro germination and seedling development. The in vivo germination was performed evaluating the influence of temperature, substrate and light. It was observed that only C. spinosa seeds presents physiological dormancy, which was overcome by using alternate temperatures. The substrate influenced significantly the germination of C. rosea and the seeds of C. dendroides showed the highest germination percentages in the different conditions evaluated. The post-seminal development stages under in vivo and in vitro conditions were defined. It was observed that the development was faster under in vitro than in vivo conditions. An effective methodology for in vitro germination, enabling the providing of material to experiment on plant tissue culture was established to C. dendroides and C. spinosa.


Introduction
Cleome is the largest genus from Cleomaceae family, with over 200 species distributed in drier areas of the tropics and subtropics (Iltis, 1960). It consists mainly of annual or as C. spinosa, used in the treatment of diseases related to respiratory tract, and several inflammatory disorders (Agra et al., 2007). Several species may be employed as ornamental due to their attractive inflorescences.
Biotechnological studies have been conducted with Cleome species in order to develop efficient protocols for mass propagation and to enhance the production of bioactive molecules under in vitro conditions. Such studies have been performed with C. rosea Vahl, Brazilian native species mainly found in coastal sandy plains, ecosystems negatively affected by human impacts and C. spinosa Jacq., found in urban areas and highly vulnerable to human activities and insect infestations. In vitro protocols were established for these species (Simões et al., 2004;Albarello et al., 2006;Albarello et al., 2007;Simões et al., 2009a;Simões et al., 2010a;Simões-Gurgel et al., 2012), and the medicinal potential of in vivo and in vitro materials was also evaluated Simões et al., 2009b;Simões et al., 2010b;Simões-Gurgel et al., 2012;Albarello et al., 2013).
Germination is a critical stage in plants the life cycle and tends to be highly unpredictable over space and time (Baskin and Baskin, 2001). Successful establishment of plants largely depends on successful germination. The germination process may be affected by environmental and internal seed factors and frequently, ideal conditions are species-specific and need to be determined through experimentation.
There is little information available in literature about seed germination of Cleome. These studies report low and non-uniform germination, presence of seed dormancy and a significantly variation in germination rates depending on the seed lot used (Ochuodho and Modi, 2007;Raboteaux and Anderson, 2010;Tlig et al., 2012;K'Opondo, 2011). As seeds, in general, are more resistant to disinfection treatments required to obtain aseptic material, when compared to tissues of a developed plant, physiological knowledge of in vitro germination allows the establishment of efficient methods that ensure the supply of stock plants in excellent phytosanitary conditions to be used as source of explants for plant tissue culture. There are no studies related to in vitro germination of the three species selected.
The purpose of this research was to investigate C. dendroides, C. rosea and C. spinosa germinative behavior under in vivo conditions, as well as to establish suitable conditions to in vitro germination and seedling development.
The variables evaluated for fruits were fruit length, thickness, shape and color, whereas for seeds, were seed length, thickness, shape, color, fresh and dry weight, moisture content, viability, coat permeability and mean number of seeds per fruit.
Analysis followed the Rules for Seed Testing (Brasil, 2009). Ten seed lots with one (C. spinosa and C. rosea) or five (C. dendroides) seeds each were weighed to determine the fresh weight (FW) and then subjected to drying in an oven at 105±3 ºC for 24 h to determine the dry weight (DW). The moisture content (MC) was defined as: MC = (FW-DW)/FW x 100.
Seed viability was determined through the triphenyl tetrazolium chloride (TTC) test using five lots of 20 seeds. Seeds were longitudinally sectioned and soaked in 2% TTC solution at 30 °C for 24h in the dark. The viability was assessed using a stereomicroscope (Olympus SD30).
To evaluate seed-coat permeability, eleven lots of five seeds were immersed in distilled water for 24 h at 26 ± 1 °C. During successive periods the fresh weight of each lot was measured to establish the imbibition curve.
Seeds were washed with detergent Fisoex® (10%), rinsed in tap water, and placed in sodium hypochlorite (NaOCl) at different concentrations (1.0; 2.0 or 2.5% w/v) containing Tween 80 (0.05% w/v) for different periods (5; 10 or 20 min). Seeds were rinsed thrice with sterile distilled water and were germinated in transparent plastic germination boxes (Gerbox) (11 x 11 x 3.5 cm). Different substrates were tested: vermiculite (average grain diameter 1.4 to 4.0 mm), sand and towel paper. Substrates were autoclaved at 121 °C for 45 min and moistened with sterile distilled water. The effect of constant (20 and 25 °C) and alternate (15-25 and 20-30 °C) temperatures was evaluated. The experiments were conducted in germination chamber (Eletrolab EL202), under a 16 h photoperiod (20 mmol m -2 s -1 ) and relative humidity of 80%. Another set of experiments was conducted in the absence of light using the best conditions of temperature and substrate established for each species in the previous experiment.
Based on the method proposed by Labouriau (1983), germination rate (GR), mean germination time (GT), coefficient of velocity of germination (GV) and relative frequency of germination (GF) were measured and defined as: Germination rate: GR = N/A x 100 where: N = total number of germinated seeds; A = total number of seeds; Mean germination time: GT = S n i t i / S n i where: n i = total number of germinated seeds per day; t i = incubation time (days); S n i = total number of germinated seeds.
Coefficient of velocity of germination: GV = GT -1 Relative frequency of germination related to the incubation time: GF (%) = n i .100 / S n i Evaluations were performed every two days during 50 days. When needed substrates were moistened to keep suitable conditions for germination and seedling development. Twelve treatments (three substrates x four temperatures) were considered with three replications of 20 seeds. Experiments were repeated twice.
Seeds were washed with detergent Fisoex® (10%), rinsed in tap water, and disinfected under aseptic conditions in laminar flow hood. Two protocols of decontamination were evaluated: exposure to atmosphere saturated with formaldehyde (paraformaldehyde 80%) for 1, 2 and 3 h, and immersion in a solution of NaOCl (1.0; 1.5; 2.0 and 2.5% w/v) containing Tween 80 (0.05% w/v) for 10, 15, 20 and 30 min. After that, the seeds were rinsed thrice with sterile distilled water and inoculated into flasks containing 30 mL of MS medium (Murashige and Skoog, 1962) added with 30 g.L -1 sucrose and solidified with 8 g.L -1 agar. The pH of media was adjusted to 5.8 prior to autoclaving (121 °C -15 min).
Flasks were closed with polypropylene caps and maintained in a growth chamber at 26 ± 2 °C, under a 16 h photoperiod provided by cool white fluorescent tubes (45 mmol m -2 s -1 ) or in germination chambers (Eletrolab EL202) under alternate temperatures (20-30 °C), a 16 h photoperiod (20 mmol m -2 s -1 ) and relative humidity of 80%. Four seeds were inoculated in each flask in a total of 30 flasks per treatment. Experiments were repeated twice.
The germination was carried out on MS medium with reduced salt concentrations (MS1/2, MS1/3 and MS1/4). Seeds of C. rosea were also inoculated on MS medium containing glucose (30 g.L -1 ) as source of carbohydrate and phytagel (2 g.L -1 ) as solidified agent. The influence of aeration was evaluated using aerated caps (Sigma B-3031) to close the flasks.
Morphological aspects were recorded since root protrusion to primary leaves expansion (Brasil, 2009). Root, hypocotyl and epicotyl length as well as the number of leaves, were measured. Seedlings were considered normal when essential structures were perfectly developed (root system well developed, hypocotyl, two cotyledons, epicotyl with apical bud and primary leaves expanded), according to Rules for Seed Testing (Brasil, 2009). The external morphology of seeds, fruits and seedlings were described and illustrated.
Graphic records were obtained from fresh material using a stereomicroscope (Olympus SD30) and photographs were taken using Sony H9 camera.
Data were subjected to analysis of variance (ANOVA). Data from germination percentage, when necessary, were transformed into arcsine prior to analysis to normalize their distribution. Means were compared by Tukey test at 5% probability (p < 0.05). The analyses were performed using the GraphPad Prism version 5.0 software.

Results and Discussion
Although the three species under study presented morphological similarities, some traits showed to be speciesspecific (Table 1). Fruits are siliqua-shaped capsules, dry and dehiscent, and in C. dendroides they are also inflated. In C. rosea the capsules are greenish ( Figure 1A), while in C. dendroides ( Figure 1B), they are greenish and turn pale brown when riped and in C. spinosa capsules are pale yellow turning brown when riped ( Figure 1C). The fruit dehiscence occurs from base to apex separating the two valves, to release the seeds. Some seeds may remain attached to the replum, a strip holding the valves edges ( Figure 1D).
Seeds of three species are small and present particular morphological traits in each species. In C. spinosa the seeds are pale yellow to dark brown with kidney-like to cochlear shape with exarillate seed coat, having open terminations and crests on the surface ( Figure 1E). Cleome rosea seeds are also pale yellow to dark brown and cochlear. They present vesicular aril and tegument with crusts and transversal ribs ( Figure  1F). Cleome dendroides has brown orbicular seeds. They are applanate, with papilose and muricate surface with fingershape extensions ( Figure 1G). Figure 1H shows C. dendroides seed internal structure. Embryos are cotyledon-type with flat, folded and overlapping cotyledons. They present slightly larger thickness than the radicle-hypocotyl axis.
All species had viability higher than 80% and moisture content ranged from 6% to 15% (Table 1). Imbibition curves showed a fast imbibition phase up to the first two hours, with seeds water content reaching around 55%-60% (data not shown), demonstrating high seed-coat permeability.
The substrate and the use of constant or alternate temperatures had no effect on GR in C. dendroides (Table 2). Lowest GT values and highest GV values were reached on sand and paper. Cleome rosea germination was significantly influenced by the different substrates. Highest values of GR were obtained in vermiculite, although no statistical differences were observed on GT and GV. The use of alternate temperatures was an essential condition to seed germination in C. spinosa. As observed in C. dendroides, lowest GT values and highest GV values were reached in sand and paper. In the three species, the germination was observed both in the presence and absence of light.
When considering all physical conditions evaluated, it was found that C. dendroides and C. spinosa seeds had the highest germination capacity, showing GR values above 90%, while GR values for C. rosea stayed below 80%. In addition, in all species, seedling development was superior in vermiculite . No abnormal seedlings were observed in all tested conditions.    The relative frequency polygons for germination, which determine the proportion of seeds germinated daily (Figure 3), showed that C. spinosa seeds presented the fastest germination when kept in alternate temperatures ( Figure 3A). In this species, radicle protrusion took place at 1 -3 days, on paper and sand. Maintenance of C. dendroides seeds in alternate temperatures of 20-30 ºC or in constant temperature of 25 ºC, accelerated radicle protrusion, which took place around the seventh day, on paper and sand ( Figure 3B). These results showed that both C. spinosa and C. dendroides have a synchronic germination process. Cleome rosea showed a delayed and intermittent germination, in some cases, the beginning of the process occurred after 40 days ( Figure 3C).
Germination under in vitro conditions was successfully established to C. dendroides and C. spinosa, but it was not efficient to C. rosea. The exposure of seeds to formaldehyde saturated atmosphere was more efficient protocol of decontamination when compared to the immersion in sodium hypochlorite solution. To C. dendroides, three hours exposure to formaldehyde resulted in an efficient decontamination level (100%) and in GR of 100%. The use of sodium hypochlorite (2.5% for 20 min) allowed low contamination (10%), seeds showed lower GR (90-95%) and delayed the start of germination process. Similar results were observed in C. spinosa. Decontamination in formaldehyde atmosphere for 1 h resulted in GR above 80%, while the best results obtained with sodium hypochlorite (1.5%, 20 min) were 48%. To C. rosea the highest GR (85%) was reached with sodium hypochlorite at 1% for 10 min. After decontamination, seeds where inoculated on MS medium in regular or with reduced salt concentration (MS1/2 and MS1/4). Cleome dendroides seedlings showed hyperhydricity on MS medium, but developed normally on MS1/4 ( Figure 2D). Germination occurred both in constant (26 ± 2 ºC) and alternate (20-30 ºC) temperatures. As for C. spinosa, salt concentration had no influence on seedling development ( Figure 2E) and alternate temperatures were required for germination.
In vitro germination of C. rosea resulted in seedlings presenting an intense callusing in folioles, cotyledons and stem ( Figure 2F). To establish healthy seedlings new assays were performed with modifications on culture conditions, however the new conditions did not prevent callus formation and, in addition, glucose caused a decrease in GR (60%) and seedlings size.
Cleome dendroides in vivo germination started after around seven days, with radicle protrusion, which breaks the tegument in the hilum ( Figure 4B). After nine days, primary root was white, cylindrical, thin, and reached 8 mm ( Figure   4C). On the day 11, hypocotyl growth was faster than the root, showing cylindrical shape and bright green color. In this stage, the cotyledons are carried above the soil surface by the elongating hypocotyl ( Figure 4D) characterizing the epigeal germination. Around day 12, cotyledons were free from the tegument, which is typical of phanerocotyledonous germination. The cotyledons were orbicularis (diameter 0.5 cm), opposite, glabrous, dark green on the adaxial surface and bright green on the abaxial surface, entire margin and 1 cm petiole. Cotyledon expansion was observed on the day 13 ( Figure 4E). At this stage, hypocotyl elongation ceased (1.5 -2.0 cm) and it was observed the development of a root system of the axial type, with cylindrical, thin, hairy and white secondary roots ( Figure 4F). Epicotyl was recorded around day 25. It was cylindrical and bright green. Around day 30, the pair of primary leaves was formed. Cotyledons abscission occurred after about 35 days. The young plant, after 40 ( Figure 4G), presented the pair of opposite primary leaves, trifoliate, membranous, and oblong-elliptical leaflets with attenuated base, obtusely acuminate at apex, smooth edges, pubescent adaxially and slightly pubescent on the abaxial surface with printed rib on the adaxial and prominent on the abaxial surface, dark green on the adaxial and bright green on the abaxial surface, but more clear than the cotyledons. Central leaflets were larger than lateral ones. The cylindrical petioles had about 0.8 cm, leaflets were sessile and the axial root system become fully established. Cleome spinosa in vivo germination took place around day 3. On day 5, primary root reached 0.5 cm and hypocotyl with the same dimensions. Hypocotyl was bright green and cylindrical. During its development, elevated foliar cotyledons already released the tegument, characterizing the epigeal germination. Around the day 15, the epicotyl (0.2 cm) was formed. On day 20, the first pair of membranous and trifoliate leaves was established and after 30 days the second pair of leaves with alternate phyllotaxy was formed. The compound leaves presented the central leaflets larger than the lateral ones.
Cleome rosea in vivo germination was observed 7 -9 days after sowing. From day 10 hypocotyl entered a period of active growth, stretching and elevating the cotyledons above the substrate characterizing the epigeal germination. On day 15 the early development of the epicotyl and first pair of leaves were observed. Between 18 and 20 day, first pair of trifoliate leaves was formed and after 40 days they possessed a brightgreen, glabrous stem axis consisting of the hypocotyl (3.5 cm) and epicotyl (3.8 cm) and two pairs of trifoliate leaves. At day 60, plants had an axial root system, well developed, with many secondary roots and three pairs of compound leaves, arranged in alternate phyllotaxy, with membranous and hairy leaflets, with entire margin and bright green color. Central leaflets were larger than the laterals.
Cleome dendroides in vitro germination took place after 7-8 days of culture, by the rupture of the tegument in the hilum and root protrusion. After 8 days, hypocotyl differentiation and primary root elongation occurred. After 9 days, the tegument remained adhered to the cotyledons and hypocotyl elongation occurred. The emergence and expansion of cotyledons was observed after 9 -10 days. The pair of primary leaves emerged around day 17 and epicotyl and secondary roots around the day 20. After 25 days, the seedling had all the essential structures: roots, hypocotyls, cotyledons and epicotyl with the pair of primary leaves expanded.
The start of C. spinosa in vitro germination took place after 1 -2 days after inoculation. Hypocotyl and primary root developed after three days. The elevation of foliaceous cotyledons occurred after 5 days and epicotyl was formed after 7 days. The first pair of membranous and trifoliolate leaves was observed after 12 days and the visualization of the second pair of leaves occurred after 18 days.
Mature fruits of both C. rosea and C. spinosa contained seeds of varying color and size, prevailing dark brown seeds with larger dimensions, higher dry weight, and lower moisture. Considering C. dendroides seeds they are only brown. The analyzed lots showed moisture content ranging from 6 to 15%. Tegument color, along with these other characteristics, pl hp epi soil rt may provide data on the quality of the seed lot (Ochuodho and Modi, 2010). Physiological maturity of seeds is associated with low moisture content and at this stage, seeds reach maximum germination (Black et al., 2006). Thus, dark brown seeds of all species evaluated were considered fully mature.
The topographical tetrazolium test enables fast and efficient determination of the percentage of viable seeds, especially those that present dormancy, the recalcitrant and species that require long time to germinate in laboratory (Brasil, 2009). This test is routinely used in quality control programs for various species (Carvalho et al., 2013). The TTC test results showed seed viability ranging from 80 to 100% and the high viability of the lots was confirmed in germination tests.
Considering the tegument permeability, the results indicated that Cleome seeds do not exhibit dormancy in the tegument (physical dormancy) since it presented no resistance to water entry. Physical dormant seeds restrict imbibition by different mechanisms, such as the presence of one or more layers of lignified palisade cells on the tegument, making it impermeable (Baskin and Baskin, 2001).
The influence of substrate and temperature on the germination process showed that the optimum conditions varies among the species evaluated. The substrate influenced most significantly the germination in C. rosea, while for C. spinosa the alternating temperature was an essential condition. On the other hand, C. dendroides reached high germination rates in all conditions tested. Some species of the genus Cleome present no dormancy and germinate without the need of pretreatment. Cleome amblyocarpa germinates in the absence of light, and 25 °C was considered the ideal temperature, with germination percentage of 81% (Tlig et al., 2012). However, several species of Cleome exhibit physiological dormancy. Ekpong (2009) studying C. gynandra observed germination rate of 17.3% in freshly harvested seeds, at alternate temperatures (20 -30 °C), without any pre-germination treatment. According to the author, the species requires a post-maturation period of three months, at environment temperature in order to break dormancy, after what germination rates were at 90%. Physiological dormancy was also reported for C. lutea (Cane, 2008), C. serrulata (Cane, 2008) andC. hassleriana (Raboteaux andAnderson, 2010).
In the present study the germination process was not influenced by light, characterizing the Cleome seeds as neutral photoblastic. However, experiments carried out to evaluate the influence of temperature and light on germination of C. gynandra from Africa (Ochuodho and Modi, 2007) concluded that the species showed photoinhibition. The authors recommend that germination should be performed in the absence of light and/or alternate temperature of 20-30 ºC.
Presented results show that C. spinosa needs the breaking of physiological dormancy in order to germinate, and this can be achieved with the use of alternate temperatures. This is in contrast to what was observed in C. dendroides and C. rosea that showed no dormancy. Alternate temperatures may favor dormancy breaking, by modulating hormone synthesis, altering endogenous levels and influencing seed germination (Black et al., 2006).
Germination was synchronic in C. spinosa and C. dendroides and asynchronous in C. rosea. Asynchronous germination leads to a wider distribution in seed germination time, which enhances the species' chances of survival under natural conditions by favoring the formation of seed banks (Baskin and Baskin, 2001). Among the studied species, C. rosea is the only that occurs in restinga, coastal ecosystems characterized by elevated temperatures and saline sandy soils. These abiotic features may require longer germination periods to ensure the species' establishment success. Asynchrony may be attributed to environmental oscillations during seed formation, as well as between individual intrinsic genetic variability within populations. Zamith and Scarano (2004), analyzing seedling production in species from Rio de Janeiro's restingas, observed large amplitude in germination time requirement.
In vitro conditions promote continuous physiological stress, which sometimes leads to hyperhydricity. Hyperhydric plants or buds present turgid, thick, wrinkled, twisted, translucent, with glass appearance, pale green stems and leaves, rigid and easily breakable. Shoots with shortened internodes and this morphological appearance suggest that water is in excess. The type of sealing may reduce or control the water on the atmosphere of culture flasks (Vasconcelos et al., 2012). In this present study, hyperhydricity was observed in plants of C. dendroides kept on MS medium, and the type of sealing showed no influence on its prevention. In vitro germination of C. dendroides was achieved on MS¼ medium with high GR yields. Besides the relative humidity, the concentration and type of gelling agent, and growth regulators, must be controlled to avoid hyperhydricity (Vasconcelos et al., 2012). The decrease in salt concentration, in particular of ammonium ions, may also be effective in preventing hyperhydricity (Guerra and Nodari, 2006), as seen in C. dendroides. The germination process under in vitro conditions of C. rosea seeds was not efficient. Despite the high percentage (85%) of seed germination in MS medium, plants presented low development of stem axis when compared to plants germinated in vivo. Callus formation was observed over the entire surface of the leaves.
The use of glucose as carbon source reduced germination efficiency in comparison to sucrose. Carbohydrates play an important role in maintaining proper osmolarity on culture medium, as well as serving as nutrient source. The low germination percentage of C. rosea seeds inoculated in medium supplemented with glucose could be the result of an inadequate osmolarity to allow proper water exchange between the medium and seeds.
Calli observed in C. rosea leaves may be related to the presence of ethylene in the culture flasks. Under in vitro conditions ethylene may be produced after the buckling as a byproduct of combustion accumulating in vials. George (2008) reported the formation of callus-like structures on leaves and stems of intact plants kept on in vitro conditions, when they were maintained in ethylene enriched atmosphere for a prolonged time. The stimulation of cell division observed in C. rosea leaves and consequent callus formation may also have been due to auxin accumulation in their synthesis sites or in close regions, caused by the antagonistic action of ethylene on polar auxin transport (George, 2008). No reduction in leaf callogenesis was observed when flasks sealed with lids that allow greater gas exchange. According to Armstrong et al. (1997), only a forced ventilation system was able to significantly reduce ethylene levels inside culture flasks. Considering the possibility that C. rosea has a high sensitivity to ethylene, new methods to decrease its concentration in culture flasks must be employed in order to develop an effective protocol for germination the species under in vitro conditions. During post-seminal development of C. dendroides and C. spinosa, it was observed faster development in vitro in comparison to in vivo conditions, which may be due to carbon supply (sucrose) and other nutrients present in culture medium, in addition to natural seed reserves.

Conclusions
The present study shows that optimal conditions for the germination are different for each of the species. Cleome spinosa has physiological dormancy, while C. dendroides and C. rosea do not exhibit dormancy under in vivo conditions. The different stages of in vivo post-seminal development of the three species are defined. In addition, an effective methodology for in vitro germination, enabling the providing of material to experiment on plant tissue culture is established to C. dendroides and C. spinosa.