a. Relation of flower type and microspore development stages
a.1. Flower morphology of T2-12-1-2 mutant line
The flowers of A. annua used in this research, the T2-12-1-2 mutant line, were yellow with globelike flower heads. The size of the fully open flower was very small (2.00-3.00 mm). The flower was compositum surrounded by two series of lanceolate bracts known as phyllaries and having several individual small flowers (florets).
Four types of flowers based on their developmental stages are presented in Figure 1 as follows: (1) flowers still in bud form (FB stage) marked with long phyllaries positioned above the buds’ top (Fig.1. A, B); (2) flowers about to bloom (AB stage) marked with short phyllaries positioned below the buds’ top (Fig. 1. E, F); (3) flowers in early bloom (EB stage) marked with closed petal of the central disk florets and the early emergence of bifurcated stigmas of ray florets (Fig.1. I, J); and (4) fully open flower (FO) marked with the full opening of the petal of the central disk florets and longer bifurcated stigmas of ray florets (Fig.1. M, N).
The floret of T2-12-1-2 mutant line was very small in size (50-100 µm) and borne within small disk-shaped capitula (0.5-3.0 mm wide) which were organized in loose spread (Fig. 1. A, E, I). Each capitulum contained two floret types: (1) central disk florets, which were hermaphroditic (Fig.1. C, G, K), and (2) marginal ray florets, which were pistillate (Fig. 1. D, H, L). Florets were sessile and had the inferior ovary attached to a mound-shaped receptacle (red arrow in Fig. 1. C, D, G, H, K, L). Calyx was absent in the florets. Corolla comprised five united petals tubular with five recurred lobes can be seen clearly in Figure 1. (C, D, G, H, K. L). In a fully open flower, the petal was open (Fig. 1. O, P) giving the yellow color to the flower. The opening of the marginal ray florets preceded that of the central disk florets. The early emergence of the stigmatic arms of pistillate ray florets from the apical opening of the corolla could be seen in flower type AB (Fig.1. J).
After considering the insignificant size of this A. annua flower and the fact that only the disk florets had the anther containing microspores (Fig.1. C, G. K), thus opposite to previous anther culture research conducted in China by Wu and Tang (2012), we are convinced that for A. annua the isolated microspore culture (IMC) is more suitable for producing haploid and DH rather than through the isolated anther culture (IAC), because manual dissection of anthers in compositum flower may be tedious and removal of a sufficient number of anthers from the disk florets may be difficult considering all the work shall be done under the microscope. This finding is in line with a previous report that the procedure of IMC is more timesaving and labor-saving than that of IAC due to the difficulty of excision of individual anthers from flower buds of Brassica and Apiaceae species, which also have small flowers ranging from 2.0- 4.5 mm and 1.0-1.5 mm, respectively (Shymikova et al. 2016; Dong et al. 2021; Shymikova et al. 2021)).
a.2. Microspore development stages
In this T2-12-1-2 mutant line, a single flower bud or a capitulum had 22 disk florets (Fig. 2. A) and 9 ray florets (Fig.2. B). Wetzstein et al. (2014) reported that various total numbers of florets per capitulum were observed among A. annua genotypes originating from open pollination of lines from several Brazilian and Chinese germplasm, and inter crosses of high-yielding lines with values ranging from 20 to 29 florets.
Disk florets of varying maturity marked with the staggered development of florets within an inflorescence were present in a single flower bud (Fig. 2. C). This is possible because the florets of A. annua were positioned in whorls, and the outer whorl of florets developed earlier than the inner whorls (Wetzstein et al. 2014). Therefore, microspore populations isolated from such flowers will be highly heterogeneous. In this T2-12-1-2 mutant line, one disk floret contained 5 anthers tubes (Fig.2. D, E).
Under microscope evaluation, round microspores with different stages of development from uninucleate to binucleate stages can be observed in a single disk floret within the anther tubes (Figure 2. F-K). In the uninucleate stage, the microspores contained only one nucleus. The nucleus of the uninucleate microspore was located at the center and large with dense cytoplasm in the early uninucleate and mid-uninucleate with slightly noticeable vacuoles (Fig. 2. H), but the nucleus position was off the center in the late uninucleate (Fig.2. I) stages. Mid-uninucleate was slightly bigger than early uninucleate, while late uninucleate had a bigger and more distinct vacuole than mid-uninucleate, so the nucleus was pushed to the peripheral side in the late uninucleate stage. In the binucleate stage, the microspores contained two nuclei. Early binucleate microspore showed two nuclei and cytoplasm (Fig.2. J). In this study, in most binucleate microspores, the two nuclei are similar in size and shape indicating their origin by a symmetric division. Late binucleate microspores or young pollen could be observed in a fully open flower marked with bigger and denser cytoplasm than that of early binucleate (Fig.2.K). Therefore, the development of microspores was found asynchronous in a single flower of A. annua.
According to García et al. (2016), microspores within the anther would gradually uptake the substances provided by the tapetum, according to their needs. The eventual metabolic differences created among microspores because of developmental asynchronies might influence their competitiveness, i.e., some microspores would be at a more advanced stage than others, possibly taking the resources first. Therefore, the selection of a flower containing microspores at the optimal stage in the donor plant is an important factor in increasing rates of microspore embryogenesis.
Generally, anthers of which microspores were at the uninucleate stage were favorable for anther culture because anther tissues could nurture the development of microspores at early stages within the anther by providing nutrients and protection against stress. In this study, similar to previous reports by Binarova et al (1997), Shariathpanahi et al (2006), and Salas et al (2012), the microspore culture represents an added physical stress during microspore isolation compared to that of anther isolation in anther culture, the late stages of pollen development might be more effective. Several studies also reported that in microspore culture of the genus Brassica, Raphanus, and Daucus, microspores during the late uninucleate to early binucleate stage had the potential to alter the microspores’ developmental pathway from gametophytic to sporophytic (Han et al. 2014; Kozar et al. 2020; Dong et al. 2021; Shymikova et al. 2021; Romanova et al. 2023).
The proportion of uninucleate to binucleate microspores in different growth stages is presented in Table 1. From this microscope observation, it was expected that the uninucleate microspores in the FB would be dominated by the early uninucleate stage, while in AB and EB flowers would be dominated by microspores in the mid to late uninucleate stages. Thus, for binucleate stages in AB and EB flowers, the microspore population would be dominated by early binucleate microspores, while in FO flowers would be dominated by late binucleate microspores or young pollen.
Table 1 The proportion of uninucleate to binucleate microspore in the flower of T2-12-1-2 mutant line of Artemisia annua (L.) used for isolated microspore culture.
Flower type
|
The developmental stage of microspores
|
Uninucleate Stage (%)
|
Binucleate Stage (%)
|
FB
|
96.77 ± 8.84
|
3.13 ± 8.84
|
AB
|
61.22 ± 15.82
|
35.71 ± 15.82
|
EB
|
41.86 ± 16.62
|
58.14 ± 16.62
|
FO
|
20.98 ± 6.18
|
79.02 ± 6.18
|
FB= Flower bud; AB= Flower about to bloom; EB= Flower in early bloom; FO: Fully open flower
According to Mineykina et al. (2021), to select a more efficient population of microspores for the induction of embryogenesis, the individual flower containing predominantly microspores at the late uninucleate vacuolated stage and that at the early binucleate stage must be selected for each genotype because the embryoid yield would be determined by the interaction of these two factors. Generally, in anther culture research using normal and big-size flowers, the researchers used the correlation of flower bud length to microspore developmental stages to induce embryogenesis (Kumar et al. 2019; Barroso et al. 2015). In this study, for inducing microspore embryogenesis, we concluded that the only morphological marker to be used in A. annua with a very small composite flower (< 3.0 mm) is by using a certain flower type to obtain a suitable proportion of microspores at certain developmental stage. Therefore, AB and EB flowers (Table 1) were selected for further microspore culture of A. annua, due to their proportion of uninucleate and binucleate microspore stages. These types of flowers (AB and EB) are visible to the naked eye and, thus can be easily differentiated (without using a microscope) from the FB and FO flower types (Fig. 1).
b. Microspore-derived embryoid (MDE)
b.1. Microspore embryogenesis in microspore culture medium
Microspore embryogenesis (ME) represents a unique system of single cell reprogramming in which a highly specialized cell, the microspore, is induced to switch its fate from gametophyte to embryo development during in vitro culture in plants (Prem et al. 2012; Sorriano et al. 2013; Testillano 2019). In our study, we did not sequentially track the formation of each cell structure individually. The results were obtained and analyzed by comparing the morphology of different structures recorded during the culture of microspores for 2 months, relying on literature data. The mode of development of the MDE was studied through (1). The growth of embryogenic microspores (Fig.3. A-E), (2). Development of two different forms of embryoid (Fig. 3. F-H), and (3). Microspore reprogramming (Fig. 3. I-L).
At the first embryogenic division, the microspore still exhibited the exine wall (Fig 3. A, B). The changes in the microspore population could be observed at 2 weeks after culture (WAC). The microspores divided and produced multicellular structures persisting within the microspore wall (exine) and the microspores continued to swell and burst, releasing the contents (Fig. 3. C). Under the microscope, MCA media had more visible swelled microspores than MCAD media. Then, these wall less cells were enlarged (Fig. 3. D), and some would continue to divide and form multicellular embryoid structures or primordium (Fig. 3. E). Chiancone et al. (2015) stated that when the microspores in vitro underwent a symmetrical division, first they formed the two-celled structures, then continued to multicellular microspores, and to large multicellular structures or proembryos. This evolution indicated that the reprogramming of the microspore and the first steps of the embryogenic pathway were achieved.
Prem et al (2012) reported that in B. napus under 18ºC treatment, the reprogrammed microspores followed the development of two different forms of embryoid, the major pathway involving the formation of suspensor-like structures, and the minor pathway producing multicellular embryos without suspensor. In this study, although there was no temperature pre-treatment applied, the early development of embryoids could be observed under a microscope as early as 2 WAC. At 4 WAC, some microspores stopped growing after a few divisions, some formed suspensor-like structures (Fig.3.F), and the rest formed many loosely callus-like structures (Fig. 3. G).
The suspensor-like structures lagged in development, while the callus-like structures or nascent embryoids continued to grow (Fig 3. H). This situation had previously been noted for Chinese cabbage (Shumillina et al. 2015), and broccoli (Domblides et al. 2018). The size of some embryoids continued to increase, and after approximately 2 months, the microspore reprogramming and their sporophytic development were observed by the presence of multinucleated calli (Fig. 3. I, J) together with the appearance of globular embryos. The globular embryos were pearly white, transparent, and round (Fig. 3. K). Until the end of 2.5 months of observation, the round embryos continued to grow to be elongated or were slightly oval (Fig.3. L).
b.2. Effect of high sucrose and PGRs on MDE formation
The effect of the sucrose concentration and PGRs on embryoid formation in microspore culture of T2-12-1-2 mutant lines is presented in Figure 4. Microspore-derived embryoids (MDEs) in the form of callus-like structures were visible to the naked eye in MCA medium at 6 WAC (Fig. 4. A, B, E, F) and in MCAD medium (Fig. 4. C, D, G, H) at 8 WAC. In carrots, the MDEs were visible over some time from 3-5 WAC to 6 months (Li et al. 2013; Górecka et al. 2010).
In this experiment, microspore culture medium with 13% sucrose, MCA13 (Fig.4. A, E) and MCAD13 (Fig. 4. C, G), gave rise to abundant micro calli than those of MCA17 (Fig.4. B, F) and MCAD17 (Fig.4. D, H). In an earlier study of isolated microspores from B. napus, Lichter (1982) found the best yields of embryoids obtained on a medium containing 12% sucrose, and some success was achieved with concentrations up to 17%. The sucrose content in the medium may be of greater importance than the concentration of macronutrients and it seems to be species-specific. A later study by Dunwell and Thurling (Dunwell and Thurling 1985) indicated that a high level of sucrose (17%) was beneficial for the induction of microspore embryogenesis of B. napus because the medium generates an osmotic potential similar to that of the anther homogenate, while a lower level of sucrose (13%) was important for a sustained level of microspore division. This is in line with Shymikova et al. (2021), where sucrose at relatively high concentrations might act as an osmotic stress as well as be required as an energy source for the formation of embryoid bodies.
The MCA13 and MCA17 media, an NLN-based media with 13% or 17% sucrose, supplemented with NAA and BA showed 2 weeks earlier embryogenic microspore development than the other two media containing 2,4-D and Kinetin (Fig.4. A, B, E, F). ME efficiency is affected by cell physiological status, and one component includes homeostasis of PGRs, especially a combination of auxin and cytokinin, known as key signaling molecules that control cell fate, proliferation, and differentiation (Perianez-Rodriguez et al., 2014; Zur et al. 2015). In several cases, the use of BA and NAA at low concentrations increases the effectiveness of embryoid production in microspore cultures of Brassica spp. (Dong et al. 2021; Niu et al. 2019; Na et al. 2011).
In the experiment with broccoli (B. oleracea var, italica), in the treatment with NAA, MDE formation decreased as the BA concentration increased above 0.5 mg L-1. The MDE formation rate was much improved when 0.05 mg L-1 NAA and 0.01 mg L-1 BA were added to a half-strength NLN liquid culture medium (Na et al. 2011). Ewes et al. (2023) reported that in anther culture of pomegranate (Punica gratum L.) using anther with uninucleate pollen grains at the tetrad stage, higher concentration of NAA (1 mg L-1) and BA (2 mg L-1) added to modified MS medium for 30 days gave the highest value of percentage of callus induction. It was known that NAA alone would not affect embryo yields, but BAP alone increased the MDEs formation in cabbage (B. rapa) as the BAP concentration increased (0, 0.01, and 0.05 mg L-1) and then decreased at BAP concentration above 0.5 mg L-1 (Zhang et al. 2012).
The addition of 2,4-D in MCAD media gave rise to MDEs covered and mixed with something like a white-gluey structure making them difficult to observe (Fig. 4. C, D, G, H). It was suspected that 2,4-D was the cause of that phenomenon due to the higher sensitivity of microspores to 2,4-D than that of somatic cells. According to Ardebilli et al. (2011), the induction of embryogenesis in microspores needed a shorter treatment time of 2,4-D (less than one hour) compared to that of somatic cells (more than one day). Furthermore, Rodríguez-Sanz et al. (2015) stated that while microspore culture does not require exogenous 2,4-D, endogenously the phytohormone may participate in microspore reprogramming in Brassica napus and in vitro embryo formation.