Citrus Polyembryony

Abstract

Polyembryony is a type of sporophytic apomixis common in citrus species. Previous studies discovered that embryo traits relate to their sexual or asexual origin. Smaller embryos at the micropyle end are considered zygotic embryos, whereas larger embryos are nucellar. Early embryogenesis in the ovule of apomictic citrus promotes the development of nucellar embryos. The chalaza region inhibits the early development of the zygotic and nucellar embryos; thus, both embryos must grow at the micropyle end. Numerous researchers agree that highly polyembryonic cultivars produce nucellar seedlings more often as the zygotic embryos cannot survive field conditions. Thus, the selection of polyembryonic genotypes facilitates clonal propagation. This chapter analyzes the factors that affect polyembryony in citrus.

1. Introduction

Angiosperms have two reproduction routes: gametophytic (sexual) and apomictic (asexual). The first form of reproduction leads to seeds with an embryo from the union of the egg cell nucleus with one of the generative nuclei from the male gamete producing a plant with genetic characteristics different from the female parent. This process promotes genetic diversity through hybridization and the adaptation of plants that allows the conservation of species [1]. We will start this chapter by describing the apomictic process and the evolutionary mechanism of alternate pathways that have allowed cloning plants using seeds [2], including most citrus species.

Various mechanisms intervene to generate asexual embryos, eluding the fundamental aspects of sexual reproduction: meiosis and fertilization [3]. According to the mechanism of embryogenesis, apomixis comprises gametophytic apomixis (apospory and diplospory) and adventitious embryony. In these, apomictic embryos have maternal inheritance, but each mechanism is associated with a different probability of producing sexual offspring, varied selection pressure to maintain male fertility, and expected levels of genetic diversity within populations [4]. In apospory, the initial cells form unreduced embryo sacs from direct mitosis and can coincide with sexual embryogenesis: if endosperm forms, the process is pseudogamic, or autonomous if it does not [5]. This second route, which occurs in less than 1% of angiosperms, comprises the development of embryos from somatic cells.

Similarly, in diplospory, the megaspore does not undergo meiosis and through mitosis forms an unreduced embryo sac with cells arranged as in the Polygonum type (sexual reproduction); but the sexual process is wholly compromised [5]. Whereas in adventitial apomixis, the development of the sexual embryo sac seems to develop normally, and after fertilization and development of the sexual embryo, somatic embryos develop from nucellar or integumental cells [6, 7]. In citrus, adventitious embryonic development does not occur without zygote formation (apomixis is facultative), depends on endosperm formation, and is simultaneous with the development of sexual embryos [6, 8, 9]. Therefore, when the embryo sac expands, the embryogenic cells of the nucellus enter the endosperm, competing for space and nutrients with the zygotic embryos that may or may not complete their development [10].

Adventitial embryony or sporophytic apomixis frequently occurs in Rutaceae, Celastraceae, and Orchidaceae; furthermore, it is common in tropical and subtropical trees and shrubs [4]. Polyembryony in citrus is a type of sporophytic apomixis that occurs in most of its species and hybrids (Table 1). For example, in 12 genotypes (Sikkim, Thorny, Kinnow and Cleopatra mandarin, Calamondin, Nova×Orlando hybrid, Minneola tangelo, Thornton, Parramatta, Parson Brown, Smooth Flat Seville, and rough lemon rootstock) rates of polyembryonic seeds ranged from 69.8% to 91.4%, with up to 14 embryos per seed [8]. The following genotypes do not show polyembryony, thus classified as non-apomictic: Citrus hongheensis, C. macropera, C. medica, C. maxima, C. clementina, Chirita mangshanensis, and some mandarin hybrids [16, 17]. Although polyembryony is not a common biological phenomenon among angiosperms, it is a form of reproduction that has been studied for 14 decades. According to Cook [18], the oldest record of polyembryony in orange seeds (Citrus sinensis L. Osbeck) occurred in 1719 by Leeuwenhoek. Later, Strasburger [19] worked with different species and established that in sour orange (Citrus aurantium), all the embryos not derived by the cross originated from nucellar cells, and he called them “adventitious embryos.” Since then, polyembryony in citrus has become a form of vegetative propagation, with the problem that for genetic improvement, seedlings from hybrid embryos cannot be differentiated from seedlings from adventitious embryos. Then, hybrid embryos can be differentiated by molecular methods or by crossing unifoliate materials with trifoliate orange (Poncirus trifoliata) pollen.

Citrus genotypes	Polyembryony (%)	Range of embryos/seed	Author
Amblicarpa mandarin	82.1	2-15	[11]
Amblicarpa mandarin	91.5	4.9	[12]
C-35 citrange	86.5	3.95	[12]
Cipó sweet orange	98	28.8	[13]
Cleopatra mandarin	84.7	1-26	[11]
Cleopatra mandarin	97.1	8.01	[14]
Cravo Santa Cruz Rangpur lime	58.5	1.97 ± 1.08	[15]
Flying dragon trifoliate orange	40	1.70 ± 1.11	[15]
Indio citradarin	100	13.52 ± 4.40	[15]
Minneola tangelo	77.5	3.45	[12]
Morton citrange	86.8	5.87	[14]
Rangpur lime	43.3	1.87	[14]
Riverside citradarin	100	12.97 ± 3.86	[15]
Rough lemon	96.2	4.89	[14]
Rusk citrange	92.3	5.05	[14]
San Diego citradarin	97.5	6.29 ± 3.19	[15]
Smooth Flat Seville sour orange	29	22.7	[13]
Sunki mandarin	66.2	1.8	[14]
Sunki Tropical mandarin	100	8.97 ± 2.76	[15]
Swingle citrumelo	64	2.48 ± 1.47	[15]
Swingle citrumelo	65	2.96	[14]
Troyer citrange	95.1	6.95	[14]
Valencia orange	95.5	4.5	[12]
Volkamerian lemon	37.8	2.50	[11]
Volkamerian lemon	85.5	3.15	[12]
Volkamerian lemon	52.2	2.57	[14]
Yuma citrange	21	1.36	[14]
Yuma citrange	31	4	[13]

Table 1.

Percentage of polyembryony and embryos per seed in apomictic genotypes of citrus.

2. Morphology of seeds, embryos, and polyembryony

The study of polyembryony is receiving increased attention from both industrial and scientific sectors as it provides a form of cloning through seed that avoids the typical complications of sexual reproduction (for example, incompatibility barriers) and vegetative propagation (replication of viruses and other diseases) [20]. Another possible benefit of asexual reproduction by seed implies fixing the hybrid vigor and allowing the propagation of hybrids through many generations of seeds [21]. However, even in 2021, in [22] indicated that fixed hybrid vigor has not been possible because hybrid seeds cannot produce offspring with the same qualities.

Over the years, attempts to link the morphological characteristics of fruits, seeds, and embryos with polyembryony or with the characteristics of nucellar or zygotic embryos to predict the sexual or asexual origin of seedlings have taken place. Some researchers have studied the relationship between fruit weight and polyembryony in Swingle citrumelo, Volkamerian lemon, Cleopatra mandarin, and Amblicarpa [11, 23]. Other studies have tried to select characteristics of polyembryonic seeds as a possible method to anticipate the origin, zygotic or nucellar, of seedlings. For example, the size and shape classification of seeds has been associated with the production of zygotic seedlings on rootstocks [24, 25]. Also, logistic regression models that predict sexual seedlings have been developed [13].

Another variable that has been considered in the possible prediction of predominant sexual or asexual reproduction is the relationship between the percentage of polyembryony in citrus genotypes and the production of nucellar seedlings. Different studies have shown that species with a higher percentage of polyembryony are less likely to develop zygotic seedlings [21, 26, 27] because only one zygotic embryo possesses competitive disadvantage against all nucellar embryos: zygotic embryos tend to be small and do not survive in field conditions. In contrast, more numerous apomictic embryos tend to be large and produce more vigorous seedlings. One case is Citrus reshni: apomictic reproduction occurs based on the degree of polyembryony of the seeds (70–90%) and the small size (≤3 mm) of the sexual embryos [28]. However, this characteristic varies between genotypes; for example, in [12] reported six genotypes with a high percentage of polyembryonic seeds (65–98%), where only Valencia orange and Amblicarpa mandarin have the possibility of apomictic reproduction. Therefore, the presence and degree of polyembryony in the seed is a genetic characteristic dependent on the interaction with the environment, particularly with the weather, the stage of development of the plant and its organs, or the physiological conditions [29, 30]. Therefore, using the phenotypic characteristics of seeds, embryos, and polyembryony to identify the seedlings’ sexual or clonal origin often leads to the omission of zygotic seedlings in segregating populations, which is unfavorable in breeding programs. This trait will be discussed later in the topic of molecular markers.

Other features of the embryos, such as size and location, have been studied to define the position and origin of nucellar and zygotic embryos. Zygotes often appear as more diminutive in seed size, with slower growth than adventitious embryos and growth prevailing at the micropylar end [31, 32]. However, during nucellar embryogenesis associated with the fertilized embryo sacs, there seems to be an inhibitory effect in the chalaza region [10, 33]. Consequently, both embryos preferentially grow towards the micropyle. These arguments are helpful to generate reproduction models in polyembryonic citrus; however, it should be noted that these studies focused on the initial stages of embryo and seed formation. Therefore, it is essential to develop innovative methods of identifying nucellar and zygotic embryos in mature seeds and differentiating the seedlings that originate in the early stages of development. We will cite the work of Andrade-Rodríguez et al. [28], who, using RAPD (Random Amplification of Polymorphic DNA) markers, in vitro culture, and embryo morphology, concluded that not all small embryos located in the micropyle produce zygotic seedlings in Volkamerian lemon seeds. This contribution generated various questions about the reproduction models proposed for citrus and motivated us to pursue our current research focus.

Our laboratory has studied polyembryony in citrus from 2000 to date (as well as in other fruit trees not mentioned in this chapter). We have used molecular markers, in vitro culture of embryos separated by size, and grafting of the plants obtained in the greenhouse. These techniques have produced plants from embryos one, two, three, four, five, six, and seven (Example: E1, E2, E3, E4, E5, E6 and E7 in Figure 1) and propose new approaches to study the relationship between size and sexual, asexual, or different origin of the plants obtained from each of the embryos. Among the works published in citrus are references [7, 11, 12, 28, 34].

Figure 1.

Embryos classified by size in a seed of Amblicarpa mandarin (Citrus amblycarpa (Hassk) Ochse). E1 corresponds to the largest embryo in length and weight. E7 corresponds to the smallest. For each embryo, its location towards the micropyle or the chalaza and radicle orientation were recorded.

3. Identification of nucellar embryos and possible zygotes

Successfully harnessing apomixis for citrus propagation requires understanding its facultative nature; the formation of nucellar and zygotic embryos varies among environments and times of the year. Additionally, each species genetic makeup acts as an activation “switch” that relies on environmental changes. All these variables lead to variations in the degree of polyembryony, number of embryos per seed, and even the type of reproduction that will prevail, asexual or sexual [3, 30, 35]. Understanding the apomictic phenomenon in citrus is central to various studies to benefit from its dual characteristics: clonal propagation from seed and hybrids production. We seek to generate models that relate the characteristics of polyembryony, size, and location of the embryos in the seed, with their sexual or asexual origin. However, the selection of the hybrids that result from the cross with polyembryonic parents is complex if they do not express a reproducible dominant trait under several environmental conditions. In addition, the percentage of zygotic progenies in various citrus hybrids has been found to vary based on the seed parent used [36], the pollen origin [37], and environmental factors [38]. For this reason, researchers used various morphological, biochemical, and molecular markers to identify seedlings in the early stages of development.

Plant morphology is the first visible marker for hybrid selection in plant breeding. The first morphological trait used to differentiate nucellar plants from zygotic ones relied on linking vigor with asexual origin. For example, in 1949 in Ref. [39] published that nucellar plants are those that develop juvenile characteristics, such as vigorous growth, presence of thorns, or slow fruiting. Another morphological marker, which is also a dominant phenotypic trait, is leaf morphology [26]. In Figure 2, Amblicarpa nucellar plant (Figure 2a) and the hybrid that exhibits more than one leaflet (Figure 2b) are compared; both plants have thorns, which according to Cameron and Johnston [39] are juvenile characteristics. It has been used in nurseries since the first decades of the 20th century to identify hybrids by crossing trifoliate genotypes. The method is still valid and is used to obtain trifoliate rootstocks such as citrandarins (Cleopatra mandarin× P. trifoliata). However, both the presence of juvenile traits and multifoliate leaves are markers that vary depending on environmental conditions and plant development, so they are unreliable [40].

Figure 2.

Mandarin Amblicarpa seedling (Citrus amblycarpa (Hassk) Ochse), obtained from embryos in vitro cultured, grafted on Volkamerian lemon (50 days after grafting). 2a nucellar plant and 2b sexual plant, both plants show “vigor” in height, leaf development, and thorn size; however, the sexual plant develops leaflets.

Not all morphological markers can reliably distinguish between zygotic and nucellar seedlings [40]. The expression of “trifoliate leaf” is a dominant trait over the recessive unifoliate leaf trait makes it easy to identify the first-generation hybrid seedlings in crosses between unifoliate citrus and trifoliate orange male parents [41]. Thus, hybrid seedlings with multifoliate leaves would be expected to appear when crosses of trifoliate male parents (P. trifoliata) are made. However, when Poncirus hybrids (such as citranges or citrumelos) are backcrossed to Citrus, seedlings with bi- or trifoliate leaves may vary considerably [37]. Furthermore, it is particularly difficult to remove off-type seedlings based on seedling morphology when pummelo has been used as a seed parent because the morphological characteristics associated with pummelo dominate the seedling phenotype [29]. An alternative to morphological markers is molecular markers, which can unequivocally discriminate the zygotic seedlings [28, 29, 30, 42, 43].

Biochemical tests have also been used as genetic markers in discriminating zygotic seedlings. Such is the case of colorimetry [44], chromatography [45], polyphenol darkening [46], spectroscopy [47], and isozyme analysis techniques [48]; however, these techniques fail to identify all clones and ignore some hybrids. For example, although they are codominant markers with simple methods, isoenzymes express few polymorphic loci to differentiate F1 seedlings from the female parent. Additionally, their expression level is qualitatively and quantitatively affected by environmental factors, plant development stage, or physiological conditions [29, 30].

Molecular biology expanded the tools available for identifying seedling sexual or asexual origin through molecular markers based on PCR (polymerase chain reaction). Among the most used markers are RAPD (Random Amplification of Polymorphic DNA), ISSR (Inter Simple Sequence Repeat), SSR (Simple Sequence Repeat), and SNP (single-nucleotide polymorphism) [30, 34, 42, 43]. Various studies compare the efficiency of genetic markers in selecting hybrids, in Ref. [29] evaluated RAPD and EST-SSR (Expressed Sequence Tag-SSR), in [43] worked with morphological markers and SSR, and in [45] compared morphological markers and SNPs. RAPDs (dominant marker) and EST-SSRs (codominant marker) efficiently and accurately identified nucellar plants of mandarin (C. reticulata Blanco) and pumelo (C. maxima Meer.); both markers were highly related (p = 0.001) [29]. The expression of trifoliate leaves with molecular markers was complemented in Ref.s [43, 49]; their research showed the viability of early selection of hybrid seedlings based on the morphological marker and subsequent analysis with molecular markers in the seedlings that did not express multifoliate leaves. This procedure reduces the cost and time of the analysis.

Not only can genetic markers complement each other, but in vitro germination can also be used in embryo identification to increase the development of small embryos (which do not germinate under in vivo conditions). Similarly, faster acclimatization and growth of plants resulting from these embryos result from grafting on a rootstock. Three studies show the advantages of combining in vitro culture and identifying hybrids with SSR markers. It is worth mentioning that SSRs have been widely used to discriminate embryos according to their origin because they favor the selection of plants obtained by self-pollination and cross-pollination. Embryos from F1 seeds from a cross between ‘Shiranuhi’ mandarin and ‘Shiranuhi’ orange were cultivated in vitro to determine the percentage of zygotic embryos detected with SSRs depending on the days after pollination (90, 105, 125, 145 and 180 days, DAP) [50]. Growth in an artificial medium allowed them to maintain constant humidity and nutrient supplementation and achieved 36–75% germination depending on DAP, allowing identification of zygote embryos: 12% at 90 days, 8% at 105 days, 7% at 125 days, 1% at 145 days and 4% at 180 days after pollination. On the other hand, in vitro culture, SSR markers, and morphological markers have been used [49]; they first selected using a trifoliate leaf morphological marker and then, with SSRs, selected 41% of hybrid seedlings of rough lemon and citrandarin X-639c, and 46% of rough lemon and Swingle citrumelo hybrids. Likewise, these authors state that not all zygotic embryos survived until fruit maturation as the ratio of hybrid seedlings decreases with advancing stages of embryonic development after 95 days after pollination. Finally, in vitro culture, seedling grafting, and SSR markers have been used to identify nucellar embryos in citrange C-35, Amblicarpa mandarin, and Volkamerian lemon rootstocks, and Valencia orange and Minneola tangelo cultivars [12]. Also, this chapter includes the identification of nucellar embryos in Parson Brown orange, not published in the 2021 paper. This study differs in its approach as the authors only studied the largest embryo per seed, finding that the weight of the seed correlates with the weight of the largest embryo (0.76–0.96, p ≥ 0.05). They used in vitro germination until they achieved 3–4 mm long seedlings and grafted them on Volkamerian lemon to promote their growth for 5 months until foliar collection. This procedure allowed for enough leaf material for DNA extraction (1.8–2.0 at 260/280 nm absorbance) without damage to the grafted plant, which developed under simulated nursery management. The SSR markers identified 70% nucellar seedlings in C-35, 65% in Volkamerian and Minneola, 85% in Amblicarpa and Valencia, and 100% in Parson Brown (Figure 3). Focus on the embryos with the most significant capacity to germinate in each seed allowed linking size (8–10 mm in length, 6–19 mg in weight) and the probability of generating nucellar plants; only Amblicarpa, Valencia, and Parson Brown showed a tendency to clonal propagation.

Figure 3.

Dendrograms showing the Genetic Similarity Indices (GSI) between nucellar seedlings and the female parent, obtained from the largest embryo per seed in six polyembryonic citrus cultivars (A, Citrange C-35; B, Amblicarpa Mandarin; C, Volkamerian lemon; D, Parson Brown orange; E, Tangelo Minneola; F, Valencia orange). The GSI was calculated with 30 SSR markers.

Various works that use molecular markers to classify seedlings according to their sexual or asexual origin state that those with total similarity to the female parent are nucellar. In [29] consider seedlings of Swingle citrumelo (Citrus paradisi Macf. × Poncirus trifoliata (L.) Raf.) and sour orange (C. aurantium L.) as “true nucellar seedlings” when they were identical to the female parent at all loci (100%), using EST-SSR and Euclidean distance with UPGMA cluster analysis. However, in [12] seedlings of six citrus genotypes with a Genetic Similarity Index [51] of 0.95 (95%) were considered “nucellar” compared to the female parent (Figure 3). Both studies show the discrepancy between the genetic similarity values needed for a plant to be considered “true nucellar” or “possible zygotic” to the female parent. As any marker has the disadvantage of characterizing only a tiny part of the genome, a discussion on the number of molecular markers needed for these studies is required. As an example, in [29] evaluated 12 EST-SSR markers, of which eight primer pairs revealed polymorphism, whereas in [12] examined 30 SSR markers, but only 17 were polymorphic enough to differentiate hybrids (TAA 41 and F4 being the most informative).

One of the objectives of breeding programs is to select plants based on agronomic and economic traits, such as the expression of polyembryony. However, it is relevant to consider the advantage of using molecular markers. These have the possibility of generating QTL (Quantitative Trait Loci) by associating the polymorphism to a phenotypic trait. The QTL Apo1, Apo2, Apo3, Apo4, and Apo6 associated molecular markers (RAPD, SSR, and CAP -Cleaved Amplified Polymorphic Sequence) with apomixis in the progeny of C. volkameriana× P. trifoliata cv ‘Rubidoux.’ Apo2 is also associated with the activation of the type of embryo (mono-embryonic or poly-embryonic) in apomictic genotypes [52]. Further information has been obtained from AFLPs (Amplified Fragment Length Polymorphism) as marker loci have identified a genomic region associated with the percentage of polyembryonic seeds in Citrus× Poncirus hybrids [53]. The genetic control of adventitious embryology was also documented in C. reticulata [2, 22]; it has been proposed that the CiRKD1 gene regulates the somatic embryogenesis that occurs with two alleles that originate polyembryony and embryogenesis [54, 55]. So, the genetic control of apomictic reproduction and polyembryony is “more than a single switch” to activate it; it involves various genome regions in citrus.

Despite the effectiveness of molecular markers to discriminate between hybrids and nucellar plants, ongoing efforts try to relate the expression of morphological features of the embryo and its location in the seed with sexual origin. The purpose is to separate hybrid plants from clonal plants from seed or in the early stages of development without a laboratory. Various studies have related polyembryonic traits and origin; however, we will only analyze the Citrus volkameriana Pasq rootstock. The sexual origin was related to the embryo’s size employing the trifoliate leaf marker expressed by the cross between the Volkamerian lemon and P. trifoliata [26]. They report large embryos (>5 mm) generate up to 82.7% of hybrid plants, while small embryos (1–2.9 mm) produce only 5.8%. Meanwhile, 25.9% of zygotic plants were identified with RAPDs in mature polyembryonic seeds, of which small embryos produced less than 43% (2–3 mm) and near the micropyle [28]. On the other hand, in Volkameriano seeds, when the largest embryo was separated from each polyembryonic seed and studied its location; all older embryos are in the chalaza and 45% of it produced non-clonal plants [12]. It is relevant to mention that the relationships between embryo morphology and sexual origin vary among genotypes [12, 26], strengthening the idea that embryos with a greater capacity to germinate in vivo in a polyembryonic genotype do not imply clonal reproduction. Thus, the competition caused by the asynchronous development between embryos during seed formation [7, 56] does not always favor the growth of nucellar embryos.

Therefore, molecular markers and complementary techniques (in vitro culture and grafting) are adequate to identify sexual and asexual seedlings with greater certainty than the exclusive use of morphological markers. Nevertheless, the characteristics of these techniques need to be assessed based on the objective and possibilities of the investigation. For example, although RAPDs and ESS-SSRs show a similar capacity to differentiate zygotic and nucellar seedlings [29], RAPDs are simpler and cheaper than EST-SSRs. However, the latter shows greater reproducibility among laboratories and detects all alleles of a locus (codominance). In the case of SNPs, their limitation is the cost and availability of the necessary equipment in the laboratory. Instead, in vitro culture is an excellent technique to be applied in research where germination of embryos is required for subsequent identification, but it is expensive in commercial citrus propagation schemes.

4. Conclusions

Polyembryony in citrus is complex and affected by genetic, physiological, and environmental factors. Therefore, a better understanding of the development of the embryos in the seed requires further studies. We propose combining complementary techniques to molecular markers to identify the sexual or asexual origin, the individual follow-up of each embryo in the seed, and the subsequent evaluation of morphological and production characteristics of the plant. This process in citrus requires approximately 3 and 4 years to start fruiting. However, it is necessary to answer a few questions: If a seedling identified with molecular markers as “different or possibly zygotic” (with similarity indices less than 0.95) expresses these polymorphisms phenotypically? Do the embryos, based on their size and position in mature seeds, tend to be nucellar or possibly zygotic? Is it convenient to continue considering the percentage of polyembryony in citrus as a form of clonal propagation in nurseries? Are the plants obtained from the larger embryo asexual? These and so many more questions need further research.

Acknowledgments

The authors thank Postgraduate College for the facillities made available during these studies. We also thank Vivero Cazones for providing the plant material and plant care help. This work was supported by the Mexican Council of Science and Technology (CONACyT).

The authorship criteria are listed in our Authorship Policy: https://www.intechopen.com/page/authorship-policy.

This section of your manuscript may also include funding information.

Conflict of interest

The authors declare no conflict of interest.