3.1 Overview of History and Extent of Interspecific Hybridization in Nicotiana

Interspecific hybrids may arise spontaneously in laboratories that maintain their in-house collections of different Nicotiana species. Unlike these rare events, such as the hybrid 2x (N. setchellii × N. otophora) reported by Bawolska et al. (1978), the vast majority of the known hybrids within the genus were the result of purposeful manipulation. The first experimental crossing of two different species of Nicotiana is generally credited to the German scholar J. G. Koelreuter, who produced a hybrid between N. paniculata and N. rustica while working in St. Petersburg, Russia in 1760 (Kostoff, 1943; Ternovsky, 1962; Mayr, 1986). Koelreuter’s classical work confirmed the existence of sexuality in plants and demonstrated the equivalency of maternal and paternal parents in contributing to their offspring. The first interspecific hybrid that involved N. tabacum was probably also created by Koelreuter, who backcrossed the hybrid N. rustica × N. paniculata to N. paniculata and mated the offspring with N. tabacum (Kostoff, 1943). One may speculate that the trispecific combination thus produced was probably composed of aneuploid plants possessing full haploid genomes of N. paniculata and N. tabacum plus varying numbers of chromosomes contributed by N. rustica. Koelreuter also produced the hybrid 2x (N. glutinosa x N. rustica) and created the first true amphihaploid hybrids between N. tabacum and other Nicotiana species: 2x (N. glutinosa × N. tabacum), 2x (N. paniculata x N. tabacum) and 2x (N. rustica x N. tabacum) (Kostoff, 1943; Mayr, 1986).

In the nineteenth century, several other interspecific hybrid combinations with N. tabacum were synthesized. According to the accounts by East (1928) and Kostoff (1943), Sageret crossed N. suaveolens with N. tabacum in the 1820s, Gaertner produced N. rustica × N. tabacum and N. quadrivalvis × N. tabacum in the 1840s, and in the 1850s, Naudin was the first to produce the hybrid N. tabacum × N. alata. Brongniart and Gris (1861) were the first to report the hybrid N. glauca × N. tabacum.

East (1928), one of the early cytogeneticists of Nicotiana, divided artificially made interspecific hybrids of Nicotiana into ‘pre-Mendelian’ and ‘post-Mendelian’. The former were produced by Koelreuter and by nineteenth century investigators such as Gartner, Focke, Gudron, Naudin, and Sageret. The post-Mendelian hybrids were those created by East himself and by his contemporaries. East’s list of viable interspecific hybrids includes 22 ‘pre-Mendelian’ hybrid combinations and 43 hybrids produced later, 65 interspecific hybrids altogether, not including reciprocals. Kostoff (1943) described 181 interspecific hybrids, and Goodspeed (1945) and Goodspeed and Thompson (1959) elevated their number to 243. Citing those authors, Smith (1968) estimated that more than 300 interspecific hybrids had been reported in the genus. Apparao and Ramavarma (1974) raised the estimated number to 340. The author of this volume found references to 455 interspecific hybrid combinations within the genus Nicotiana that have been produced by 2023, including several that were reported as nonsurviving seedlings or as mere viable seeds (see Chap. 7). The former, mostly included in the type II seedling death category, have been included on the grounds that hybrid seedling lethality, especially type II, was recently demonstrated as a remediable condition controlled by a simple genetic system. The hybrids, together with literature references limited mostly to the earliest ones, were compiled in Chap. 7 of this review. The supplement chapter lists hybrid combinations and the parentage of each including the reciprocal hybrid, if reported. Multispecies crosses have not been included (those that involve N. tabacum as one of the parents are treated separately in Chap. 4 and listed in Tables 4.5 and 4.6).

Table 3.1 lists Nicotiana species and the numbers of hybrid combinations with other Nicotianae reported for each of them. It also includes information on whether a particular species was successfully crossed with the cultivated species N. tabacum. Included in Table 3.1 is also information about the degree of chromosome homology between N. tabacum and the other Nicotiana species. Structural similarity between the chromosomes of N. tabacum and those of its wild relatives can be treated as a preliminary element of the assessment of how much restriction to the genetic flow exists between the species of interest and cultivated tobacco. As can be readily seen from Table 3.1, this piece of information, essential from the standpoint of a potential interspecific breeder, is either lacking or difficult to locate for nearly half of the species that have been successfully hybridized with N. tabacum.

Table 3.1 Potential of Nicotiana species to hybridize with Nicotiana tabacum and with other sister Nicotianae
Full size table

The majority of Nicotiana species (fifty-nine) listed in Table 3.1 have a record of having been hybridized with both Nicotiana tabacum and at least one other sister Nicotiana. Some (N. mutabilis, N. petunioides, N. attenuata, N. corymbosa, N. linearis, N. burbidgeae, N. thyrsiflora and N. wigandioides) were hybridized with at least one other Nicotiana but not with N. tabacum. Twenty-eight species listed in Table 3.1, the majority of which were discovered only recently (N. gandarela, N. azambujae, N. acaulis, N. ameghinoi, N. cutleri, N. paa, N. longibracteata, N. spegazzini, N. bilybara, N. candelabra, N. fatuhivensis, N. faucicola, N. gascoynica, N. heterantha, N. hoskingii, N. insecticida, N. karijini, N. monoschizocarpa, N. murchinsonica, N. nota, N. paulineana, N. pila, N. rupestris, N. salina N. stenocarpa, N. truncata, N. yandinga, N. walpa), were not found to have any hybridization record whatsoever.

3.2 Pre- and Post-fertilization Barriers to Production of Viable Hybrid Seeds in Nicotiana

3.2.1 Manifestations of Pre- and Postzygotic Barriers to Hybridization

The number of interspecific hybrids produced between various Nicotiana species, although impressive, is still only a small fraction of several thousand theoretically possible combinations. Not all hybrids are obtainable with equal ease. The process of speciation depends on the development of different blocks preventing plants of diverging taxa from crossing with each other and producing viable offspring. The barriers to hybridization can be roughly divided into prezygotic and postzygotic (He et al., 2019). Prezygotic barriers include inhibition of pollen adhesion to the stigma, inhibition of pollen germination, obstruction of pollen tube growth in the stigma and style, and failure of the pollen tube to penetrate the mycropyle. Postzygotic barriers manifest themselves as failure to produce a functional zygote, embryo abortion and failure of hybrid seeds to germinate, lethality of germinated seeds and seedlings, disturbed development of hybrid plants and their failure to develop flowers and, finally, failure of flowering hybrid plants to perpetuate themselves because of their inherent inability to produce offspring. In Nicotiana, blocks to hybridization have evolved relatively slowly, unlike gene and chromosome alterations, which have progressed more rapidly (Goodspeed, 1945). Barriers to hybridization were already encountered by early Nicotiana researchers. Numerous cases were recorded where upon crossing two different species, either no seeds were produced, the seeds were produced but they failed to germinate, the seeds germinated but the hybrid plants died, mostly at juvenile stages of growth (McCray, 1933; Kostoff, 1943) or mature flowering plants could not be perpetuated because of different forms of sterility (see Sect. 3.6).

3.2.2 Cross Incompatibility

The causes of cross incompatibility between Nicotiana species were discussed in detail by Kostoff (1943). Three major causes of the failure to obtain viable hybrid seeds were noted by him:

  • failure of the pollen tube to reach the ovary

  • failure of the sperm to fuse with the egg

  • abortion of the hybrid embryo

The first of the causes may be attributed to various factors, such as differences in the length of styles between the maternal and paternal species (Swaminathan & Murthy, 1957; Stoyanova, 1979; Lee et al., 2008), physiological inhibition of pollen tube growth or mechanical obstruction to the growth of pollen tubes (Kostoff, 1943). Growth inhibition of the pollen tube (gametophytic incompatibility) was found to be conditioned genetically (Pandey, 1977). The growth-promoting substance (GPS) diffused in pollen walls becomes deactivated in incompatible stigmata by the genetically controlled system attached to GPS and prevents its normal function and pollen growth. Gametophytic incompatibility in the hybrid N. tabacum x N. alata was demonstrated to be controlled by differences in ribonuclease activities in self-compatible (SC) N. tabacum and self-incompatible (SI) N. alata (McClure et al., 1989; McClure, 1996). RNAase linked to the compatibility locus (S) resulting in the rejection of alien pollen was also implicated in the rejection of alien pollen in other intrasectional and intersectional crosses involving N. alata (Murfett et al., 1996). Mechanical obstruction may be caused by differences in cell size since somatic chromosome number is, at least within a certain range, correlated with cell size (Kostoff, 1943). By way of example, the thick pollen tubes of N. tabacum cannot make their way through the styles of N. alata, whereas reciprocal mating is possible (Kostoff, 1943). Another studied case in point was the cross between N. tabacum and N. obtusifolia, where the pollen tubes of the former species stopped growing in the styles of the latter (Chung et al., 1996). There are many other cases on record where a hybrid was successfully produced one way but not the other. It came to be known as unilateral incompatibility. Kostoff (1938a) found the hybrid N. glauca × N. langsdorffii to be produced with ease, whereas in reciprocal mating, the pollen tube of N. glauca was frequently arrested in the style of N. langsdorffii. Kuboyama et al. (1994) found that the pollen tubes of N. repanda, N. obtusifolia, and N. rustica were arrested at different phases of their growth through the styles of N. tabacum. The ability of N. tabacum pistils to inhibit foreign pollen tube elongation gradually increased as the flowers matured. A reverse situation was observed by Liao et al. (2017) in the hybrid N. tabacum cv. ‘K326’ x N. alata. Young flowers of N. tabacum inhibited the growth of N. alata pollen tubes and prevented them from reaching the ovary, but the inhibition was removed, and seeds could be obtained when senescent flowers were pollinated.

The Australian species of the section Suaveolentes are known to produce hybrids with N. tabacum when used as maternal parents, whereas reciprocal crosses are usually unsuccessful (Kostoff, 1943; Goodspeed, 1945). The hybrid of N. tabacum by the African species N. africana is a well-known exception to that rule (Gerstel et al., 1979; Doroszewska & Berbeć, 1996; Hancock et al., 2015), although successful reciprocal mating, N. africana × N. tabacum, was also reported (Nikova & Zagorska, 1990). Unilateral incompatibility is also absent when N. fragrans is crossed with N. tabacum (Tezuka et al., 2010). There are several cases on record when using the autotetraploid variant of a Suaveolentes species as the male parent resolved the problem of unilateral incompatibility with N. tabacum (see Sect. 3.4.1). Interestingly, Zaitlin and Mundell (2006) experienced no difficulty in obtaining viable hybrids by fertilizing female N. tabacum with pollen of N. benthamiana, whereas DeVerna et al. (1987) found that the mating was completely incompatible and they had to resort to ovule culture to overcome the incongruity.

Artificial pollination of ovules cultured in vitro was an effective approach to overcome unilateral incompatibility of N. tabacum crossed as females with several Suaveolentes species. In this way, viable hybrids of N. tabacum as the maternal parent were obtained with: N. amplexicaulis (Larkina, 1980), N. benthamiana (Subhashini et al., 1986; Mihaylova-Kroumova et al., 2020), N. debneyi (Butenko et al., 1970; Larkina, 2015, 2017), N. excelsior (DeVerna et al., 1987; Tezuka et al., 2010), N. maritima (Tezuka et al., 2010), N. rosulata (Ternovsky et al., 1976; Ternovsky & Larkina, 1978a; Larkina, 2015, 2017), N. suaveolens (Marubashi & Onosato, 2002, N. velutina (Tezuka et al., 2010) (for more information, see Sect. 3.4.3). Unilateral incompatibility between N. tabacum and the species of the section Suaveolentes also has important practical implications. Many of these species carry resistance to important diseases and pests of tobacco, and cytoplasmic factors from many of them produce cytoplasmic male sterility in cultivated tobacco. Since CMS is inherited maternally, it may obstruct the breeding process when the transfer of a desired trait from the Suaveolentes species to N. tabacum starts and proceeds in a matrilineal fashion (see Chap. 5).

The hybridization process between N. tabacum and another Nicotiana species that gets stalled at some stage, either pre- or postfertilization, cannot always be accurately classified in simple binary terms: cross compatible vs. cross-incompatible. There are many instances where success depends on whether the right genotype has been chosen within the parental species to be crossed. Kostoff (1943) cites several examples, e.g., N. rustica × N. tabacum or N. tabacum × N. pauciflora, where the wild species was successfully crossed with one N. tabacum variety but not with others. The crossability of N. tabacum with N. alata, generally very poor, was found to be largely affected by the genotype of N. tabacum (Kostoff, 1943), a phenomenon also confirmed by Ternovsky (1962) and Berbeć (1987b).

The ease with which hybrids of N. tabacum × N. africana could be obtained either directly from seeds or by resorting to an in vitro culture was substantially affected by the genotype of the N. tabacum parent in the study by Depta and Doroszewska (2019).

Different accessions of N. suaveolens also vary substantially in their ability to cross with N. tabacum (He et al., 2019).

The type of cytoplasm can also affect crossability. Liao et al. (2017) failed to obtain viable seeds by crossing regular male fertile N. tabacum cv. ‘K326’ with N. alata, but the mating was successful when a cytoplasmically male sterile lineage of K326 with the cytoplasm of N. glauca was used as the female parent.

3.2.3 Mechanisms of Pre- and Post Zygotic Incongruity

When the foreign pollen tube has reached the ovary but failed to fertilize the egg, the following phenomena may follow: cell division may occur in the nucellus without embryo or endosperm formation, resulting in no seeds, diploid endosperm may develop, resulting in seeds with no embryos, and induction of cell divisions in the egg may lead to the formation of parthenogenetic haploids or, rarely, diploid embryos. The phenomenon of haploid and diploid maternal soffspring of interspecific crosses will be addressed in one of the subsequent sections.

Various interspecific matings in Nicotiana result in the actual fusion of gametes, but the resulting zygote is dysfunctional, and the embryo is aborted soon after fertilization. An example case is the hybrid N. obtusifolia x N. tabacum (Chung et al., 1996), the hybrid embryo of which aborts soon after fertilization or develops poorly, resulting in inviable, nongerminating seeds. Disturbed and collapsed development of the hybrid embryo may result from the incongruity between the maternal endosperm and the “foreign” embryo as well as from the genetic factors inherent in the embryo itself. Brink and Cooper (1941) and Bannikova (1965a) studied seed development in the F1 hybrid N. rustica × N. tabacum. The cross usually resulted in aborted seeds with less than 1% germination capability. Retarded growth of the endosperm, the failure of intergumentary cells to differentiate into conductive tissues and the overgrowth of the nucellus were the causes of the abortion of the hybrid seeds. Disturbed development of the embryo and the endosperm and deranged metabolic exchange between the embryo, endosperm and maternal tissue set in within a week after fertilization in the hybrid N. rustica × N. paniculata (Bannikova, 1965b). The development of the hybrid embryos was substantially slower compared to the embryo growth rate of self-fertilized N. rustica.

The nonviable seeds obtained from crossing N. tabacum with N. alata were void of both the embryo and the endosperm (Stoyanova, 1979). The embryo death by starvation in the hybrids N. stocktonii x N. tabacum, N. nesophila x N. tabacum and N. repanda x N. tabacum was attributed to the cessation of endosperm development (Reed & Collins, 1980a).

Recently, arrest of endosperm development leading to starvation and abortion of the embryo in the hybrid 4n N. suaveolens x 2n N. tabacum was attributed to disturbed EBN (Endosperm Balance Number) (He et al., 2020, 2022, 2023). According to the EBN hypothesis, the normal ratio of maternal vs. paternal ploidy level is approximately 2:1 due to the triploid genome of the endosperm. Disturbance of this ratio caused by different ploidy levels of parental genomes would lead to endosperm growth arrest. Specifically, excess maternal EBN in the cross 4n N. suaveolens x 2n N. tabacum led to precocious developmental transition and subsequent endosperm development failure accompanied by hypertrophy of the embryo arrested at the globular stage (He et al., 2020) or very early cessation of hybrid embryo growth, depending on the N. suaveolens accession used.

Chromosome elimination in postzygotic stages is yet another manifestation of postzygotic incongruity. Disturbed mitotic cell divisions in the hybrid embryo may lead to failure of some essential chromosomes to be included in the daughter cells and, consequently, result in embryo abortion. If uniparental chromosome elimination is the case, the affected embryos may develop into semiviable dysfunctional aneuploid hybrid plants or, in extreme cases, yield maternal or paternal haploids (see Sect. 3.5). Kramer and Reed (1988) and Hancock et al. (2015) reported the appearance of such irregular plants in the progeny of the hybrid N. tabacum x N. africana.

3.3 Seedling Lethality

3.3.1 Occurrence of Premature Plant Death in Nicotiana Hybrids

The incongruity between species of Nicotiana may extend beyond the survival of the embryos and affect the germinating seeds and newly emerged plantlets. Root necrosis, decay and death of juvenile plants are common among the hybrid offspring of different Nicotianae. In the first half of the twentieth century, East (1928) and Gentscheff (1931) listed N. tabacum × N. alata, N. tabacum × N. langsdorffii, N. tabacum × N. longiflora, and N. tabacum × N. plumbaginifolia among the hybrids that die at the cotyledonary stage or thereafter. McCray (1932) added N. tabacum × N. glutinosa, N. nudicaulis × N. tabacum and N. suaveolens × N. tabacum to the list of “weak” or prematurely dying hybrids. Kostoff (1943) also cited previously mentioned combinations and added N. tabacum × N. pauciflora. Foster (1943) was unable to raise the hybrid N. repanda x N. tabacum to maturity, and Clayton (1950) encountered that phenomenon in the hybrid N. debneyi × N. tabacum. Inviable hybrids of N. eastii × N. tabacum and N. megalosiphon × N. tabacum were reported by Chaplin and Mann (1961), and those of N. gossei × N. tabacum and N. occidentalis × N. tabacum were reported by Apparao and Ramavarma (1972) and Ternovsky et al. (1976), respectively. In later years, several other hybrid combinations were found to be lethal, especially those involving species of the section Suaveolentes and N. tabacum (Lloyd, 1975; Berbeć & Doroszewska, 1981; Nikova et al., 1991; Laskowska & Berbeć, 2012). There was one particular case in which the massive death of hybrid seedlings was advantageous. In the cross N. tabacum × N. africana, the seed set was good, and the seeds germinated well, but germination was followed by extensive root necrosis and death of the emerged seedlings (Gerstel et al., 1979). In that particular case, only ca. 1% of the plants survived beyond the cotyledonary stage, and the population of rare survivors was composed of true hybrids and occasional maternal haploids. The latter were used to produce completely homozygous lines in a one-step procedure that came to be deployed for both academic and practical purposes (Burk et al., 1979; Wernsman, 1992) as an alternative to the anther culture method (Nitsch & Nitsch, 1969).

3.3.2 Types of Hybrid Lethality

The mechanisms underlying the phenomenon of hybrid lethality remained obscure until the nineties of the last century. In-depth investigations into the causes of seedling unviability in interspecific hybrids of Nicotiana were undertaken primarily by Japanese researchers in the late 1980s. They were summed up by Tezuka et al. (2012) in an exhaustive review of the work that had been carried out by him and by his Japanese associates and colleagues.

According to the external symptoms exhibited by inviable seedlings, five types of hybrid lethality in Nicotiana have been distinguished to date:

  • Type I: browning of shoot apex and root tip

  • Type II: browning of hypocotyls and roots

  • Type III: yellowing of true leaves

  • Type IV: formation of multiple shoots

  • Type V: fading of shoot color

Type II lethality seems to be prevalent in crosses involving N. tabacum. It is shown by 19 hybrids of N. tabacum with the species of the section Suaveolentes (Tezuka, 2012). The phenomenon occurred regardless of cross direction, with the exception of N. wuttkei × N. tabacum, the reciprocal of which was not successful (Laskowska & Berbeć, 2012). Further studies identified the causative factor of type II lethality in N. debneyi that interacts with the factors allegedly on chromosome Q in N. tabacum (Tezuka, 2012, see also Sect. 3.3.4).

Outside the hybrids with N. tabacum, Type II lethality was also observed in N. paniculata × N. gossei, N. suaveolens × N. gossei and N. debneyi × N. repanda (Tezuka, 2012). The lethality of the hybrid N. tabacum × N. langsdorffii (Watanabe & Marubashi, 2004) was not assigned to any of the lethality types. Based solely on the description by these authors but also from other accounts (Burk, 1972, Berbeć, unpublished), it could fit in with type II. However, from the account of Burk (1972), who used monosomics of N. tabacum to overcome the unviability of that hybrid, it appears that neither the use of Haplo H nor Haplo Q restored viability to the hybrid; some viable plants have been obtained from Haplos A, G, L and Z (see, however, also Sect. 3.3.4).

N. occidentalis produces type V lethal hybrids when crossed with N. tabacum (Tezuka, 2012).

Lethality types II and V are also peculiar in that the symptoms are temperature dependent, i.e. they develop at temperatures of 28 °C and below, are completely suppressed when the temperature is raised to 34–36 °C and recur once the temperature drops again below 30 °C. This phenomenon was first reported by Manabe et al. (1989) in the hybrid N. suaveolens × N. tabacum.

Of the other types of seedling lethality, type I was exhibited by N. nudicaulis × N. tabacum, type III by N. repanda × N. tabacum, (Iwai et al., 1984), N. paniculata × N. nudicaulis and N. glutinosa × N. repanda, and type IV by N. paniculata × N. alata and N. paniculata × N. glutinosa (Tezuka, 2012).

According to Iizuka et al. (2012), N. benthamiana and N. fragrans produce only viable hybrids with N. tabacum. However, Krusteva et al. (2003) and Nikova et al. (2006) reported serious survival issues among the emerging seedlings of the F1 hybrid N. benthamiana x N. tabacum. Subhashini et al. (1986) and DeVerna et al. (1987) failed to obtain any seeds from mating male N. benthamiana to female N. tabacum.

3.3.3 Causes of and Phenomena Associated with Hybrid Lethality

The temperature-dependent lethality of hybrid seedlings observed in types II and V is called programmed cell death (PCD) and is accompanied by apoptotic phenomena such as condensation of chromatin, fragmentation of nuclei, disintegration of DNA and aggregation of insoluble proteins in dying cells (Ueno et al., 2019). It is one of the mechanisms of reproductive isolation and occurs in other plant genera (Tezuka & Marubashi, 2004; Hancock et al., 2015). Bomblies (2009) suggested that PCD and apoptotic hybrid lethality are caused by the hyperactivation of plant defense responses and thus may share a common mechanism with resistance to some viral pathogens (e.g., TMV or TSWV), which is also based on eliciting a similar hypersensitive reaction (HR). Hancock et al. (2015) drew attention to the fact that the gene conferring the hypersensitive necrotic response to TMV and the gene controlling apoptotic hybrid lethality in N. tabacum are both located on the same chromosome H contributed by the ancestral form of N. tomentosiformis. The idea of the common genetic background of PCD and HR plant defense response was also discussed by Chen and Lin (2016). The two authors thought it very likely that the genes that cause necrosis and death in interspecific hybrids are also involved in immune responses and that the appearance of apoptotic phenomena in wide hybrids can be related to the activation of pathogenesis-related genes. It was argued that the activation of the PCD-related processes is due to the mismatch between the delicately configured immune systems of individual species that make up an interspecific hybrid. This interpretation became more precise when the gene Nt6549g30 was identified to encode a CC-NRL protein in N. tabacum, the CC-NRL group being the largest class of plant defense proteins that recognize effectors from a pathogen (Ma, 2017; Ma et al., 2020). The same Nt6549g30 gene was also demonstrated to control seedling lethality in the hybrid N. tabacum x N. africana (Ma, 2017, Ma et al., 2020). More recently, Katsuyama et al. (2021) suggested that both NRL proteins and the associated chaperone protein complexes that aid NRLs in unfolding were involved in developing seedling lethality in the hybrid N. gossei x N. tabacum. Hence, hybrid seedling death is essentially the effect of autoimmunity-based responses caused by incorrect interactions between parent-of origin NRLs and their associated proteins. As the result of those PCD-type autolytic processes, insoluble protein progressively aggregates in dying hybrid cells, a process demonstrated for the hybrid N. suaveolens x N. tabacum (Ueno et al., 2019).

Development of teratological changes is a peculiar phenomenon that may be involved in the failure of certain hybrid combinations to grow to maturity. Apparently, in most interspecific hybrids, genetic tumors usually appear in senescing plants (see Sect. 4.5.9), although Burk (1972) reported their occurrence before flowering in a plant of N. tabacum x N. langsdorffii. Likewise, in some crosses that did not involve N. tabacum (N. rustica × N. alata and N. rustica × N. langsdorffii), tumors and teratomata were observed in juvenile plants of those hybrids and in the hybrid N. rustica × N. alata (Takanashi & Marubashi, 2017). The onset of these abnormalities was temperature dependent. The hybrid N. obtusifolia × N. tabacum developed vitrification and tumors as early as the seedling stage, and these abnormalities were the main cause of seedling death in that hybrid (Liu & Marubashi, 2014). In the latter case, the abnormalities were probably of an origin different than genetic tumors since neither parental species belonged to the “plus” group as classified by Näf (1958).

Nicotiana species that were found to produce inviable hybrids with N. tabacum are listed in Table 3.2.

Table 3.2 Lethality of juvenile hybrid plants of Nicotiana tabacum with other Nicotiana species
Full size table

3.3.4 Location and Identification of Genetic Factors Responsible for Lethality of Hybrid Seedlings

Chromosomal Location of Heritable Factors in N. tabacum by Monosomic Analysis

The first monosomic line of N. tabacum experimentally produced was probably the one derived from backcrossing the hybrid N. sylvestris x N. tabacum to N. tabacum. It was reported by Clausen and Goodspeed (1926) and named ‘corrugated’. The original complete series of 24 lines of N. tabacum, each of them monosomic for one of the 24 chromosomes of the species, was developed by Prof. R. E. Clausen with his associates at the University of California, Berkeley, starting from the mid-twenties of the last century (Olmo, 1935; Clausen & Cameron, 1944; Lewis, 2011). Generally, four sources of monosomic lines in N. tabacum were indicated by Clausen and Cameron (1944):

  1. (a)

    Monosomic variants in segregating populations from backcrossing F1 (N. tabacum x N. tomentosaFootnote 1) and F1 (N. tabacum x N. sylvestris) to N. tabacum. The method may have taken advantage of the ‘Drosera’ scheme pairing in the gametogenesis of the initial hybrids (Goodspeed, 1954) and of other irregularities in homoeologous pairing between the 48-chromosome N. tabacum and its ancestral 24-chromosome progenitors. Thus, the backcross F1 (N. tabacum x N. sylvestris) x tabacum could occasionally yield some true N. tabacum individuals and aneuploids, including those monosomic for the T subgenome chromosomes. In a similar fashion, individuals monosomic for the S subgenome chromosomes could be picked among the survivors of the cross F1 (N. tabacum x N. tomentosa) x N. tabacum. In those early studies, several monosomics were obtained in this manner but only from backcross populations that involved N. sylvestris (Olmo, 1935). More recently, monosomic individuals of N. tabacum were also identified in hybrid populations derived from the cross F1 (N. tabacum x N. tomentosiformis) x N. tabacum using SSR markers for specific linkage groups of N. tabacum (Liu et al., 2017). These plants were monosomic for linkage groups 3 and 6 according to the N. tabacum genome map developed by Bindler et al. (2011);

  2. (b)

    Spontaneous genomic mutations in normal disomic tobacco populations;

  3. (c)

    Segregating offspring of N. tabacum lines that carried the asynaptic pale-sterile mutation discovered by Clausen (1931). Pale sterile plants produce microsporocytes with, on average, only 11 bivalents per cell, and the rest of the chromosomes remain unpaired. The cross pale sterile x normal type produces highly diversified offspring that consist of trisomics, single, double and triple monosomics and various monosomic-trisomic combinations. Monosomics for a single chromosome pair are bred into the normal type by repeated backcrossing accompanied by morphological and cytological selection;

  4. (d)

    Monosomics that arise spontaneously in the offspring of established ones due to the tendency of limited asynapsis in many monosomic types. Ultimately, those spontaneous variants proved to be the major source of monosomics developed in the genetic laboratory of the University of California in Berkeley (Clausen & Cameron, 1944).

The monosomics of N. tabacum and their morphological and developmental features in the background of the variety Red Russian were described by Prof. D. R. Cameron (Smith, 1968, 1979). At the same time, a labeling system of N. tabacum chromosomes was adopted in which chromosomes contributed by N. tomentosiformis are lettered from A through L and those by N. sylvestris originally from M through X (Olmo, 1935). The lettering was later changed from M through Z (Olmo, 1935; Smith, 1968) because the letters X and Y were traditionally assigned to the sex chromosomes.

Until recently, monosomic analysis was the only available tool with which the genes that control simply inherited traits could be assigned to particular chromosomes.

Location and Identification of the Lethality Factor in N. tabacum Responsible for Type II Hybrid Seedling Lethality

Inoue et al. (1996) studied the survival of hybrids from crossing N. suaveolens with putative progenitor species of N. tabacum, i.e., N. sylvestris and N. tomentosiformis, which contributed subgenomes S and T, respectively, to the allopolyploid cultivated species. They found the hybrid plants N. suaveolens x N. sylvestris to develop typical lethality symptoms, whereas the plants from crossing N. tomentosiformis by N. suaveolens remained viable. The authors inferred from those results that it is the subgenome S of N. tabacum that carried the lethality factor. Based on an analogous study of the hybrids N. debneyi x N. sylvestris and N. tomentosiformis x N. debneyi, Tezuka et al. (2007) likewise concluded that the subgenome S of N. tabacum houses the factor/s that cause lethality in the hybrid N. debneyi x N. tabacum (Table 3.3,Footnote 2 column 4).

Table 3.3 Chromosomal and genomic location of the N. tabacum factor controlling type II lethality in hybrid seedlings from crossing N. tabacum with several species of the section Suaveolentes on chromosome H (subgenome T) vs. chromosome Q (subgenome S) as established by different research teams
Full size table

Based on the evidence discussed in the preceding paragraph, Marubashi and Onosato (2002) studied genetic causes of lethal hybrids by crossing ten monosomic N. tabacum lines for the sylvestroid chromosomes of N. tabacum (from Haplo N to Haplo Z except Haplo P and Haplo V) with N. suaveolens. When tested against N. suaveolens as the pollen parent, all but Haplo Q consistently yielded lethal hybrid offspring. In contrast, the progeny of Haplo Q x N. suaveolens segregated for viable and lethal seedlings, and this result was also confirmed in the study of Tezuka and Marubashi (2006a). Similarly, the progeny of the monosomic Haplo Q x N. debneyi segregated for lethal and viable individuals. Q chromosome-specific markers revealed the presence of the Q chromosome in lethal plants, whereas viable seedlings were deficient for the Q chromosome (Tezuka et al., 2007). In another study conducted in that laboratory, Tezuka et al. (2010) found that in the offspring from crossing Haplo-Q monosomics of N. tabacum with N. africana, N. excelsior, N. goodspeedii, N. gossei, N. maritima, N. megalosiphon and N. velutina, the seedlings that carried chromosome Q were lethal, whereas those deficient for chromosome Q were viable. Hence, the authors concluded that the factor/s that triggered the seedling death response in those hybrid combinations must reside on chromosome Q of N. tabacum. Additionally, Tezuka et al. (2012) found that the SSR (simple sequence repeats) markers present in the inviable hybrid seedlings of N. tabacum x N. africana mapped to a linkage group that had been coded as linkage group 11 in subgenome S of N. tabacum according to Bindler et al. (2011).

Recent investigations have indicated that rare spontaneous survivors among the dying offspring of type II lethality hybrids may arise due to the loss of the distal part of the Q chromosome as a result of homoeologous translocations during spermatogenesis of the male N. tabacum parent (Nakata et al., 2021).

Controversy Regarding the Identitification of the N. tabacum Chromosome That Carries the Lethality-Controlling Factor

The results discussed in the preceding section are at odds with what was reported by Gerstel et al. (1979), Hancock et al. (2015) and Ma et al. (2020). Gerstel et al. (1979) tested the whole range of N. tabacum monosomics (from Haplo A through Haplo Z) by crossing them with N. africana. It is only in the Haplo H x N. africana progeny that he found segregation for viable and lethal individuals. The remaining combinations, including Haplo Q x N. africana, yielded predominantly lethal offspring. Their conclusion was that chromosome H of the T subgenome of N. tabacum controlled the lethality of the hybrid N. tabacum x N. africana.

The more recent report by Hancock et al. (2015) supported those old findings. By using the same or similar set of SSR markers as those used by Tezuka et al. (2012), the authors recognized the association of the seedling lethality factor with linkage group 11 of the N. tabacum genome but were firm in associating linkage group 11 with chromosome H rather than with chromosome Q. They supported their claim by citing evidence mainly from earlier studies on the inheritance of TMV resistance in N. tabacum (Lewis et al., 2005, other unpublished data) but also because one of the markers for linkage group 11 was detected on chromosome H.

In his thesis, Ma (2017) modified the position on the controversy by conceding that chromosome H originated from the S genome. In a subsequent paper (Ma et al., 2020), the authors reiterated that chromosome H (linkage group 11) of N. tabacum was contributed by the ancestral form of N. sylvestris, thereby effectively moving chromosome H from subgenome T to subgenome S. They did so in recognition of the facts (see the final paragraph of this section) but in contradiction to the existing nomenclature and classification of monosomics (Clausen & Cameron, 1944; Smith, 1968, 1979; Tezuka et al., 2010; Lewis, 2011).

Table 3.3 presents the results of the genetic control of type II hybrid lethality in hybrids involving N. tabacum reported by Japanese and American scientists.

A very intriguing and unanswered question is how this controversy may have come about. The monosomic stocks used by both Gerstel et al. (1979) and Marubashi and Onosato (2002) were bred into Red Russian. In all probability, both Japanese and American stocks descended from those developed and described by Dr. Clausen and Dr. Cameron. This was explicitly stated by Marubashi and Onosato (2002), and it is the best guess for the monosomics used in the USA. The provenance of the Haplo-Q stock used by Tezuka et al. (2007, 2010) was not specified, although in both studies, the monosomics were apparently also ultimately derived from the same source in Berkeley, California. Both chromosome Q, as reported in Japanese studies (Tezuka et al., 2012), and chromosome H, as studied by the Americans (Hancock et al., 2015), showed structural instability and were prone to breakage in the hybrid N. tabacum x N. africana. Marubashi and Onosato (2002) and Tezuka et al. (2007, 2010) reported the necessity to use Haplo-Q as the maternal parent in crosses with the Australian species because of the failure of their Haplo-Q to produce viable pollen, a feature characteristic of both Haplo-H and Haplo-Q (Smith, 1968). Although Haplo-H and Haplo Q seem to be discernible from each other by some morphological and growth features (Smith, 1968, 1979), morphology is not a wholly dependable criterion in selection for monosomic types, as these differences are strongly influenced by the environment (Liu et al., 2017). Clausen and Cameron (1944) wrote at length on how difficult it was to maintain the identity and integrity of their monosomics and that despite their efforts, doubts remained as to the validity of some of them. All these things put together, it is obvious that at least some monosomics of N. tabacum are not unlikely to be confused with one another. Hence, it is probable that Haplo-H used earlier by Professor Gerstel and more lately by the team headed by Professor Lewis and Haplo-Q used by Professor Tezuka, his associates and by other research teams from Japan in reality represented the same monosome type. It was tacitly conceded by Ma (2017) when, in support of the assertion that the hybrid lethality factor locus, N. tab_HL1, resides near the end of chromosome H, he quoted both Hancock et al. (2015) and Tezuka et al. (2012), even though the Japanese researchers had reported on chromosome Q rather than H. The fact was communicated in the very title of their study (Tezuka et al., 2012). The same interpretation of the controversy was recently also advanced by Nakata et al. (2021) and summarized by Mino et al. (2022).

With what we know, it is not wholly unjustified to assume that while Clausen and Cameron’s collection of monosomics was being developed and/or maintained in American laboratories, monosome H was at some point wrongly named and wrongly assigned to the T (tomentosoid) genome, whichever came first. A very puzzling part of the story, however, is how the “rectification” of both the misnaming (from H to Q) and the misplacing (from T to S subgenome) came about in the “Japanese chapter” of the collection. The Japanese researchers seem to have ignored or been unaware of the inconsistency of their results with those published in the US, at least in the early stages of their studies. They did not question the validity of any of the collection stocks, nor did they comment on the discrepancy or report on any corrections thereof. Other things put aside, the series of Japanese reports concerning the issue under discussion has turned out to be consistent and logical, by coincidence or otherwise.

Pivotal to the success in resoving this controversy is to establish the association of the linkage groups in the gene map of N. tabacum developed by Bindler et al. (2011) to the physical carriers of those genes i.e. the chromosomes. In Bindler’s map there are 24 linkage groups corresponding to 24 chromosomes of the amphidiploid N. tabacum. Eleven of these were assigned to the tomentosoid (T) genome, nine to the sylvestroid (S) genome, and the remaining four were mixed and consisted of groups of either T or S genetic markers. Neither Bindler (Bindler et al., 2011) nor anyone else has made any attempts to link those linkage groups to particular physical units (chromosomes). Generally, reports on associating linkage groups with chromosomes of N. tabacum have been scant as yet and limited to chromosomes A, H and Q. Vontimitta and Lewis (2012) assigned chromosome A to Bindler’s linkage group 4 but later corrected the association in favor of group 15 (Ma, 2017). It is not clear, however, if the assignment of chromosome A was changed accordingly. The assignments of chromosomes H and Q were discussed above. Mapping linkage groups to the corresponding chromosomes has barely started, and nearly all the work is still to be done. Most likely, it is not until linkage groups and chromosomes become mutually identifiable that controversies such as the one above will finally be prevented.

In a recent development that may be regarded as a sui generis sequel to the chromosome Q vs. H controversy, the transposon-tagging method combined with CRISP-Cas9 gene-editing technology was used to identify the gene in the genome of N. tabacum that controls the death of juvenile hybrid plants of N. tabacum x N. africana (Ma et al., 2020). To this end, a strategy was deployed that involved the use of a maize-derived binary transposable Ac/Ds system consisting of the autonomous immobile activator (Ac) and the Ac-controlled transposon (Ds) plus selectable markers linked to the Ac and Ds elements. As a result, the association was established between the DNA sequence identified as Nt6549g30 and the NtHL1 locus previously demonstrated to house the factor responsible for triggering the series of apoptotic events that cause the death of seedlings of that hybrid (Hancock et al., 2015, see also Sect. 3.4.3). The authors stated that its chromosomal location was on chromosome H, or linkage group 11, contributed by N. sylvestris. They found a high degree of similarity of Nt6549g30 to several homologous genes in N. sylvestris.

3.3.5 Genetic Control of Type V Lethality in N. occidentalis x N. tabacum

First, it was suggested that Type V lethality in the hybrid N. occidentalis x N. tabacum was related to factors located on both subgenomes, S and T, of N. tabacum (Tezuka & Marubashi, 2012). Although type V was phenotypically observed in the hybrid, the type II genetic system must have also been functioning alongside because the hybrids of N. occidentalis with the progenitor species of N. tabacum, N. sylvestris and N. tomentosiformis showed type II and type V lethality, respectively. Using two new accessions of N. occidentalis (PI555541 and PI555690) alongside the old one (JT), Kawaguchi et al. (2021) demonstrated that the genetic control of type V hybrid lethality in the hybrid N. occidentalis x N. tabacum differed depending on the N. occidentalis accession involved: in JT x N. tabacum, it was controlled by alleles in both S and T subgenomes, whereas in the hybrids involving the new N. occidentalis accessions (PI555541 and PI555690), it was only subgenome T of N. tabacum that housed the factor/s controlling type V lethality. Both PI555541 and PI555690 yielded viable offspring when crossed with N. sylvestris, and the hybrid progeny of analogous crosses with N. tomentosiformis showed type V lethality.

3.4 Methods to Overcome Cross-Incompatibility, Embryo Abortion and Mortality of Juvenile Hybrid Plants

3.4.1 Managing Cross Incompatibility

Numerous approaches have been used to overcome cross incompatibility between species of Nicotiana. Swaminathan and Murthy (1957) bypassed the incompatibility between N. debneyi and N. tabacum that results from the difference in style length of parental species by cutting off the style of the maternal N. tabacum to a length of 2–3 mm and smearing the cut surface with a drop of sucrose agar. The pollen of N. debneyi was placed on the thus-prepared styles of the egg parent. The problem with this particular cross was more efficiently resolved by resorting to culturing excised ovules of N. tabacum fertilized in vitro with the pollen of N. debneyi (Butenko et al., 1970; Ternovsky et al., 1976). By resorting to the same technique, Marubashi et al. (1988) and Marubashi and Onosato (2002) produced seedlings of the hybrid N. tabacum x N. suaveolens. The same method was adopted to overcome cross-incompatibility between the long-styled N. tabacum used as the female parent and the relatively short-styled N. knightiana, N. rustica, N. benthamiana and N. rosulata as the pollen parent (Slusarkiewicz-Jarzina & Zenkteler, 1983, Larkina, 1980, DeVerna et al., 1987, Ternovsky et al., 1976).

The problem of unilateral incompatibility of N. gossei and N. tabacum was overcome by using the autotetraploid variant of N. tabacum as the pollen parent (Valleau, 1952). In some other hybrid combinations, the use of the autotetraploid variant of wild Nicotiana ensured success. The first attempt to cross 4n N. repanda as a female with 2n N. tabacum was only partly successful since the hybrid plants failed to reach flowering (Foster, 1943). The same cross was repeated with success by Pittarelli and Stavely (1975). Surviving hybrid plants were also obtained from crossing 4n N. palmeri with N. tabacum (Berbeć et al., 1982). However, an opposite effect was recently reported for the cross 4n N. suaveolens x 2n N. tabacum, which was incompatible due to collapsed endosperm development (He et al., 2020, 2022), whereas the regular combination (2n N. suaveolens x 2n tabacum) was compatible. For more details on manipulating ploidy levels to bypass the incompatibility of interspecific Nicotiana crosses and the underlying EBN (Endosperm Balance Number) theory, the reader is referred back to Sect. 3.2.3 of this chapter.

Senescing flowers of N. tabacum were found to be more receptive to fertilization by pollen of N. alata than freshly developed flowers (Liao et al., 2017).

In the hybrid N. gossei × N. tabacum, the incongruity between the two species was overcome by inducing structural changes in the parental chromosomes through the exposure of the paternal pollen to helium ions or gamma rays (Kitamura et al., 2003). Chromosome loss and chromosome rearrangements were observed in rare surviving hybrids, and their survival may have been the result of the deletion or disablement of the gene/s controlling apoptotic seedling death in that hybrid (see also Sect. 3.4.3).

3.4.2 Embryo Rescue by Culturing Ovules

Inviable embryos can be rescued, and hybrid plants can be regenerated by pollinating field- or greenhouse-grown plants in situ and culturing the excised fertilized ovules in vitro. Such an approach was adopted to overcome incongruity between N. tabacum and, e.g., N. plumbaginifolia, N. sanderae, N. acuminata, N. nesophila, and N. stocktonii (Nikova et al., 2006; Iwai et al., 1986; Reed & Collins, 1978).

Alternatively, in vitro cultured ovules can also be pollinated in vitro. Pollinating the cultured ovaries of N. tabacum with the pollen of N. rustica Marubashi & Nakajima, 1985 rescued the lethal hybrid N. tabacum × N. rustica. Liu et al. (2017) used this approach to obtain progeny from backcrossing the amphihaploid hybrid N. tabacum x N. tomentosiformis to N. tabacum used as the male parent.

In another successful embryo rescue experiment, the ovaries of N. repanda were X-irradiated prior to fertilization with pollen of N. tabacum (Shintaku et al., 1985). The two surviving hybrid plants were similar to N. tabacum in habit and growth type, but their flower morphology resembled that of N. repanda. Both showed chromosome deficiency (for 1 and 2 chromosomes, respectively) as judged by the expected amphihaploid chromosome number. This prompted the authors to surmise that in those plants, the lethality factor/s were eliminated with the missing chromosome/s. A similar approach based on partial chromosome elimination in the male parent through the use of irradiation was used by Shizukuda et al. (1983) to obtain “partial” or asymmetric hybrids of N. tabacum x N. rustica.

More examples of culturing fertilized ovules as a means to raise other incongruous interspecific hybrids to maturity are given in Table 3.4. Where embryo abortion is the result of a malfunctioning endosperm, the beneficial effect of aseptic culture is that it substitutes for the endosperm in providing nutrients to the embryo, thus assuring its continued development (Reed & Collins, 1980a).

Table 3.4 In vitro cultures used to overcome the death of embryos and juvenile plants in incongruous hybrids of N. tabacum with other Nicotiana species
Full size table

3.4.3 Managing Lethality of Juvenile Plants

Overcoming Hybrid Lethality Through Natural Processes

Up to the mid-sixties of the last century and in some laboratories even much later, the Nicotiana investigators and breeders tried to manage the lethality of hybrid seedlings by the simple expedient of pollinating as many plants and flowers and sowing as many seeds as possible in the hope of obtaining rare phenotypes that would be able to survive to maturity. In quite a few cases, the policy bore fruit (e.g., Tsikov, 1966; Clayton et al., 1967; Gerstel et al., 1979; Berbeć & Doroszewska, 1981; Hancock et al., 2015). Since one of the major obstacles in obtaining hybrid seeds was premature flower drop after pollination, Burk and Chaplin (1979) gave a general recommendation to apply indole acetic acid (IAA) to the pedicels of pollinated flowers to prevent abscission. To the same end, Wark (1970) applied IAA to calyx sectors of maternal flowers. Recently, He et al. (2022), in a study on the causes of hybrid lethality in Nicotiana, confirmed the effectiveness of IAA in suppressing abscission in the interspecific cross N. suaveolens x N. tabacum, although the treatment per se had no effect on the lethality of hybrid seeds.

At least in the case of N. tabacum × N. africana, the survival of rare hybrid phenotypes was found to be related to the somatic instability of the hybrid embryos. As a result of chromosome loss or fragmentation, the specific loci contributed by N. africana and/or by N. tabacum and responsible for the development of apoptotic plant death syndrome in hybrid seedlings were eliminated, and rare hybrid genotypes deficient for lethality factors could be grown to maturity (Hancock et al., 2015). In another hybrid that showed PCD syndrome, N. amplexicaulis × N. tabacum (Berbeć & Doroszewska, 1981), different chromosomal constitutions were found among rare surviving hybrids, including an aneuploid deficient for 3 chromosomes, 6 aneuploids with two to four supernumerary chromosomes 12, and apparently regular amphihaploids (2x = 42) plus one apparent amphidiploid (4x = 84).

Overcoming Hybrid Lethality Through In Vitro Cultures

The methods to overcome hybrid unviability that depended on fortuitous processes were inefficient, time and labor consuming and final success was a matter of luck. For plant breeders, another trade-off of this approach is the possibility that the surviving aneuploids may be deficient for the gene/genes of their specific interest.

Butenko and Luneva (1966) were probably the first to overcome hybrid lethality in Nicotiana by resorting to in vitro culture. They placed aseptically produced seeds of the nonsurviving hybrid N. alata × N. glauca on a nutrient medium supplemented with kinetin and produced viable hybrid plants. With time, different variants of in vitro cultures on solidified media to rescue dying hybrid seedlings became a widely used practice. A compilation of cases where interspecific hybrids involving N. tabacum were obtained by resorting to in vitro culture, regardless of method applied and cause of incongruity, is presented on a species-by-species basis in Table 3.4. By comparing Tables 3.4 and 3.10, it is notable that the majority of the hybrids listed in Table 3.4 were also obtained by earlier hybridizers in the natural way by direct pollination without the aid of in vitro techniques. Nonetheless, the gains of saved time and labor resulting from the improved seedling survival rate (Tezuka et al., 2010) caused the traditional approaches to have been all but displaced by aseptic cultures.

The role of tissue culture in overcoming lethality of Nicotiana hybrids is still not wholly understood, and the extent to which particular mechanisms are involved may differ among individual cases. The process of organ culture and plant regeneration is known to exacerbate somaclonal variation. According to Tezuka et al. (2012), the deletion of a chromosome or a chromosome fragment may lead to the elimination of the causative agent for interspecific lethality in hybrids rescued by tissue culture. The validity of that supposition was confirmed by Nakata et al. (2021). The authors demonstrated that the cultured hybrid seedlings of N. suaveolens x N. tabacum that had overcome lethality lacked the distal part of chromosome Q of N. tabacum that carries the Nt6549g30 gene responsible for the development of lethality syndrome (Ma et al., 2020). Such spontaneous chromosomal alterations that led to the cancellation of the PCD syndrome were convincingly demonstrated for the hybrid N. tabacum × N. africana by Hancock et al. (2015), see also Sect. 3.2.3). The preponderant class of lethal N. africana × N. tabacum hybrid plants obtained by Nikova and Zagorska (1990) that were rescued by tissue culture had fewer (44) mitotic chromosomes than the theoretically expected number (47). On the other hand, the N. tabacum × N. africana hybrid plants recovered from cotyledon culture by Doroszewska and Berbeć (1996) all had a regular number of 47 chromosomes. In yet another study of the same hybrid (Depta & Doroszewska, 2019), hybrid plants regenerated from cultured cotyledons were also classified as amphihaploids. In the latter case, flow cytometry was used to determine the ploidy level of viable regenerants. However, this relatively fast and convenient method lacks the resolution power to discriminate at the single chromosome level, as can be clearly seen from the data supplied by Hancock et al. (2015). Nonetheless, the chromosome counts of one of the first lethal hybrids rescued by cotyledon culture, N. suaveolens × N. tabacum (Lloyd, 1975), also revealed a regular amphihaploid number of 40.

Culture conditions and medium components may induce mutations that result in disabling the function of the lethality genes in otherwise chromosomally regular hybrid genotypes (Tezuka, 2012). What all those and some other tissue-culture rescued hybrids (Ternovsky et al., 1972, 1976; Yamada et al., 1999) had in common was that they were cultured and regenerated on media that included cytokinins in their composition. Inoue et al. (1994, 1997) demonstrated that the addition of cytokinins to the culture medium had a decisive effect on the survival of the lethal hybrid plants of N. suaveolens × N. tabacum, but the efficacy of the treatment varied with the type and concentration of the cytokinin used. Nakata et al. (2021) surmised that the increased rate of survival of N. suaveolens × N. tabacum hybrid seedlings on cytokinin-enriched medium was due to the presence of cytokinin-induced reactive oxygen species (ROS) in the studied hybrid explants. Since ROSs are known to cause chromosome breakage, they may have induced the observed fracture and loss of the distal part of chromosome Q that carried the Nt6549g30 gene essential for triggering the plant death response. This mechanism was assumed to account for the restoration of viability to at least a large part of the potentially apoptotic hybrid plants of N. suaveolens x N. tabacum. Hancock et al. (2015) also linked the rare survival of N. tabacum x N. africana seedlings grown in vivo to chromosome instability and to the deletion of the Nt6549g30 gene, but they associated the phenomenon with the innate behavior of chromosome H in that particular hybrid milieu rather than with the influence of any external factors (see, however, section 3.3.4 for comments on the identity of chromosome H and chromosome Q).

Temperature may be another factor related to overcoming hybrid lethality by tissue culture. High temperature-dependent suppression of hybrid lethality of types II and V is known to become ineffective once the hybrid plant is returned to normal growth conditions (Tezuka, 2012). However, in many cases, the temperatures prevailing inside culture containers are likely to be above the critical threshold of ca. 30 °C and thus may facilitate the operation of other, not temperature-related, factors in the in vitro environment. Incidentally, the survival in vivo of hybrid seedlings of N. amplexicaulis × N. tabacum was substantially improved when seed germination and seedling culture were moved from the greenhouse to a growth chamber fitted with fluorescent tubes that kept the ambient temperatures ca. 30 °C, and the process was conducted in covered glass containers (Berbeć & Doroszewska, 1981).

Shiragaki et al. (2020) reported that treatment of juvenile plants of the hybrid N. suaveolens x N. tabacum with L-2-aminooxy-3-phenylpropionic acid (AOPP) suppressed the development of type II lethality by inhibiting the phenylalanine ammonia-lyase (PAL) responsible for the production of phenolic compounds involved in seedling death.

Overcoming Hybrid Lethality Through the Use of N. tabacum Monosomics

Burk (1972) was the first to resort to a monosomic parent to overcome hybrid lethality in an interspecific hybrid of Nicotiana. By crossing N. langsdorffii as the male parent to each of the 24 monosomies of N. tabacum as females, he was able to produce several viable hybrid plants, most of which had been obtained from Haplo A plus a few single survivors from Haplos G, L, and Z. A few years later, Gerstel et al. (1979) demonstrated the involvement of the N. tabacum H chromosome in the lethality of the hybrid N. africana x N. tabacum by producing viable and inviable plants in the progeny of N. africana x Haplo H. Tezuka et al. (2010) repeated the success of Gerstel et al. (1979) with the hybrid N. africana x N. tabacum and with hybrids of N. tabacum with several other species of the section Suaveolentes, reporting Haplo Q as the monosomic N. tabacum parent (see, however, Sect. 6.4.3 on the identity of chromosomes H and Q).

Even though the monosomic method seems to be easy and dependable for the production of viable hybrids, especially those exhibiting type II lethality, few, if any, researchers, apart from those cited above, have tried this approach. Most of those who did had goals in mind other than bypassing lethality obstacles. The most likely reason is that the monosomic stocks are relatively hard to access from external sources, and their in-house development and perpetuation are both costly and troublesome.

Overcoming Hybrid Lethality Through Gene Editing

Ma et al. (2020) identified the gene of N. tabacum that controls the seedling death of N. tabacum x N. africana hybrid plants (see also Sect. 3.3.3). As the confirmation part of the study, the authors used CRISPR–Cas9 gene-editing technology to insert frameshift mutations within the Nt6549g30 region. A transformant homozygous for a mutation of the Nt6549g30 gene was found that conditioned nearly 100% survival of the hybrid seedlings that resulted from pollinating the engineered N. tabacum by N. africana. Thus, it was demonstrated that the mutation disabled the lethality-conferring function of Nt6549g30, thereby reconfirming the identity of Nt6549g30 as the gene controlling hybrid lethality in N. tabacum x N. africana.

In yet another part of the same study (Ma et al., 2020), several crosses were made of the mutant line of N. tabacum with some other species of the section Suaveolentes (N. amplexicaulis, N. debneyi, N. excelsior, N. gossei, N. megalosiphon, N. occidentalis, N. simulans, N. umbratica, N. velutina) to determine whether the same factor was responsible for seedling death of those hybrids that previously had been known to exhibit type II seedling lethality (Tezuka, 2012). The viability of all but one of those hybrids was restored. The exception was the hybrid from crossing N. occidentalis with the Nt6549g30 mutant of N. tabacum. The results were consistent with lethality types exhibited by the studied hybrids. While all the surviving hybrids represented type II lethality, the hybrid N. occidentalis x N. tabacum that failed to respond to the Nt6549g30-mutant gene had been classified as type V (Tezuka, 2012). Obviously, a different mechanism of hybrid lethality is involved in the latter hybrid. The authors of the study (Ma et al., 2020) pointed out that their novel approach may expand the gene pool available for tobacco breeding. The current practical disadvantage of the method is that the CRISPR–Cas technology involves what in many countries is termed genetic modification and as such is subject to legal restrictions.

3.5 Maternal and Paternal Plants in the Offspring of Interspecific Crosses in Nicotiana

3.5.1 Gynogenic and Androgenetic Haploids

Both gynogenic and androgenetic haploids were reported in the offspring of interspecific crosses in Nicotiana. Gynogenic (maternal) haploids of several Nicotiana species (N. alata, N. rustica, N. paniculata, N quadrivalvis) stimulated by the pollen of an alien Nicotiana species were probably first reported by Wellington (1913). Single haploid plants of N. rustica were found among the progeny from crossing N. tabacum as the pollen parent with the former species as the female (Savelli cited by Kostoff, 1943) and with N. gossei (Apparao & Ramavarma, 1972). Clausen and Mann (1924) discovered a haploid plant of N. tabacum among the offspring of the cross N. tabacum × N. sylvestris. Haploids of N. tabacum were also reported from crossing N. tabacum with N. alata and N. quadrivalvis (Wellington, 1913), N. glutinosa (McCray, 1932), and N. longiflora and N. glauca (Ternovsky, 1936a, 1936b). Haploid plants were induced in the offspring of N. tabacum plants whose stigmata were dusted with X-irradiated pollen of N. alata (Tanaka & Kurihara, 1968). The thus obtained gynogenic haploids were compared with those from anther culture (Kumashiro & Oinuma, 1985).

Androgenetic haploids among the offspring of interspecific crosses seem to occur much more rarely than their gynogenic counterparts. A supposedly androgenetic haploid of N. tabacum was found in the progeny from crossing maternal amphidiploid 4x (N. glutinosa × N. tabacum) with male N. tabacum (Clausen & Goodspeed, 1925; Clausen & Lammerts, 1929). A single androgenetic haploid of N. africana was detected among the offspring from mating N. tabacum with genetically transformed N. africana (Hancock et al., 2015). The divisions of the sperm of N. tabacum in the nucleus of N. eastii followed by chromosome doubling in the developing embryo may have accounted for the appearance of a cytoplasmically male sterile N. tabacum plant as a result of crossing maternal N. eastii with N. tabacum as the pollen parent (Berbeć & Berbeć, 1992). A case that bears a strong resemblance to the story of cms eastii was reported by Nikova et al. (1997). The authors attempted to transfer cms factors from N. excelsior to N. tabacum. Following the F1 generation, the transfer seemed to proceed smoothly, but the very nature of the F1 plants cannot be readily explained based on what the authors themselves reported. In the account by Nikova and her associates, their F1 plants were similar to the male N. tabacum save for flower malformations, and they could be backcrossed with ease to the male recurrent parent. Based solely on morphological and fertility evidence supplied by the authors, it is verging on impossible that those F1’s were regular amphihaploid hybrids and were very unlikely to be spontaneous amphidiploids. Based on what the investigators themselves reported, it can be hypothesized that the male sterile N. tabacum-like phenotypes in the F1 offspring may have originated as a result of unilateral elimination of N. excelsior chromosomes from the hybrid zygotes followed by chromosome doubling (compare the reports on gynogenic haploids by Chimoyo and Pupert (1988) and by Hancock et al. (2015) referred to in the paragraph next but one below.

Spontaneous androgenetic haploids in N. tabacum were proposed to be a convenient vehicle to transfer cytoplasmic male sterility from one variety to another in a single step by using a rootless mutation to discriminate between rare androgenetic haploids and true intervarietal hybrids (Horlow et al., 1993).

One of the interspecific crosses, N. tabacum × N. africana, produces maternal haploids very regularly, albeit normally at a very low rate. The phenomenon was deployed to develop a method by which haploid plants of N. tabacum can be generated for experimental and breeding purposes (Burk et al., 1979, Nielsen & Collins, 1989, see also Sect. 3.3.1). Gerstel and Wernsman (1979) assumed that gynogenic haploids arise spontaneously during the reproductive process without direct involvement of the pollinating species. The massive death of true interspecific hybrids at the cotyledonary stage gives a selective advantage to rare gynogenic haploids that are otherwise difficult to discern and pick up among regular progeny. One must note, however, that not necessarily all the haploids found in the progenies of mating N. tabacum to N. africana are purely gynogenic. At least some of them may have arisen as products of selective elimination of N. africana chromosomes from originally hybrid embryos (Chimoyo & Pupert, 1988; Hancock et al., 2015).

3.5.2 Maternal Diploids

Various Cases of Maternal Diploid Induction in Nicotiana

Maternal diploid plants in the offspring of interspecific matings in Nicotiana have long been a controversial topic. At first, Goodspeed (1915) dismissed diploid maternals, apomictic or otherwise, as products of experimental errors in disagreement with East (1930) but later abandoned his former view, citing N. paniculata as the prime example of maternal diploidy following hybridization with other species (Goodspeed, 1954). Maternal plants continued to be reported as byproducts or even as sole products of mating maternal N. tabacum with N. rustica (Lehmann, 1936), N. repanda (Pittarelli & Stavely, 1975), N. amplexicaulis (Berbeć & Doroszewska, 1981), N. wuttkei (Laskowska et al., 2015), N. africana (Hancock et al., 2015), N. sylvestris (Eghis 1930) and various other interspecific matings (Murthy & Subbarao, 2004). Alleged apomictics were also reported in the progeny of N. alata and N. forgetiana as induced by pollination with X ray-irradiated pollen of N. langsdorffii (Pandey, 1974). Maternal plants were reported as the preponderant category of offspring when N. tabacum was mated to the amphidiploid 4x (N. wuttkei x N. tabacum) (Laskowska et al., 2015).

A singular case was reported by Kostoff (1935, 1938a), who described a parthenogenetically produced amphidiploid 4x (N. glauca × N. langsdorffii) obtained by fertilizing an amphihaploid plant 2x (N. glauca × N. langsdorffii) with pollen of N. langsdorffii. Kostoff explained the phenomenon by the presence of an unreduced monad cell with the doubled chromosome complement of the maternal hybrid that had been stimulated by pollen of N. langsdorffii to develop parthenogenetically into an amphidiploid embryo (Kostoff, 1935). Another case of that type that also involved N. langsdorffii was the production of amphidiploid seeds by the amphihaploid N. knightiana x N. tabacum upon pollination with pollen of F1 N. alata x N. langsdorffii (Berbeć et al., 1982). In the latter case, along with maternal origin, another plausible course of events is the fusion of the unreduced egg cell produced by the amphihaploid with the sperm of N. alata x N. langsdorffii followed by selective elimination of the alatoid chromosomes during embryogenesis or/and embryo development. The plausibility of such an explanation is further supported by the fact that the same amphihaploid 2n (N. knightiana x N. tabacum) was prone to spontaneous seed setting (Berbeć et al., 1982). In a related phenomenon, Apparao et al. (1980) observed a massive selective loss of N. gossei chromosomes from the hybrid 4n N. tabacum × 2n N. gossei, resulting in F1 progeny composed of plants phenotypically close to N. tabacum, each of which contained a full diploid complement of the maternal parent and two chromosomes from N. gossei.

Induction of Maternal Plants by Pollen of N. alata

The mating of N. tabacum with N. alata is noteworthy, as it seems to have yielded maternal phenotypes more frequently than any other interspecific cross that involved N. tabacum as the female parent. Both maternal haploids and diploids of N. tabacum were induced by X-ray-irradiated pollen of N. alata (Tanaka & Kurihara, 1968; Pandey & Phung, 1982; Kumashiro & Oinuma, 1985). Spontaneous diploid maternal phenotypes were observed in the offspring of the cross N. tabacum × N. alata by Stoyanova (1979), Sarychev (1987), Berbeć (1987b), Naumenko (2012), and Liao et al. (2017). In the study of Stoyanova (1979), diploid maternals were practically the only class of progeny obtained from mating an unspecified Virginia variety of N. tabacum to N. alata, whereas crosses between other varieties of N. tabacum and N. alata yielded no maternal plants. In the latter instance, the author herself suspected experimental error, but she did not exclude other causes.

In one of his early experiments with interspecific hybrids, the author of this review studied the hybrid N. tabacum x N. benavidesii (Berbeć, 1978). In the course of his study, he obtained a number of subsesquidiploids (TTB) with chromosome numbers ranging from 55 to 58. Under open self-pollination, these sesquidiploids did not produce seeds by selfing but could be easily backcrossed to N. tabacum. A surprising part of the study was that they could also set seeds when their flowers were pollinated with pollen of N. alata. Generally, the offspring from regular backcrosses were morphologically similar to those resulting from pollinating the sesquidiploid with N. alata, with no traces of introgression from the pollinator species. Likewise, both classes of offspring showed a similar behavior in meiosis with approximately 24 bivalents and varying numbers of univalents, but the backcross products involving N. alata, univalents were generally higher in number. All these observations pointed to induction of parthenogenetic development of unreduced female gametes of the BC1 plants by pollen of N. alata without gametic fusion having actually taken place. Alternatively, fusion of gametes may have taken place, but the N. alata chromosomes were selectively eliminated from the developing embryo. However, another plausible but, under these circumstances, not very likely explanation is an experimental error, i.e. inadvertent spontaneous self-pollination in BC1 plants. Interestingly, in the previously mentioned study on the hybrid N. tabacum x N. alata by the same author (Berbeć, 1987b), N. tabacum (TT) plants well protected against uncontrolled pollen contamination were mated to the sesquidiploid (TTA) as pollen parent. The mating resulted in 36 surviving plants, of which 29 closely resembled the maternal plants and the remaining seven, while departing from the maternal phenotype, did not show any traces of introgression from N. alata.

When emasculated but unprotected flowers of N. tabacum were fertilized with pollen of N. alata (Berbeć, 1987b), some genotypes regularly yielded viable and fertile N. tabacum-like phenotypes as the preponderant class of offspring, the rest being inviable or poorly viable hybrid seedlings. When carried out alongside and under the same conditions, interspecific matings of female N. tabacum with male N. langsdorffii and N. benavidesii only occasionally produced surviving maternal plants (unpublished observations of the author of this volume).

A phenomenon very similar to if not identical with those described in the preceding two paragraphs was reported by Naumenko (2010, 2012). Diploid maternals described as ‘pseudogamic’ and resulting from fertilizing intraspecific F1 hybrids of N. tabacum with pollen of N. alata could be generated regularly and were not much different from the regular selfed progeny of those plants. Apomictic populations from crossing F1 intervarietal hybrid plants showed segregation for plant height, leaf number and leaf size that was characteristic of the F2 pulations obtained from the same plants by selfing. According to Naumenko, the difference between the alleged apomictics and the regular hybrids was that the former became stabilized already in the F2 generation whereas the regular hybrids kept segregating in subsequent generations. This description recalls the behaviour of populations derived from gynogenic or androgenetic doubled haploid hybrid plants. In this particular instance, N. alata sperms may have induced chromosome doubling and embryo development in the maternal plants without actual gamete fusion. Alternatively, gamete fusion may have been followed by massive elimination of N. alata chromosomes and chromosome doubling, whichever is more likely. This way or the other, the most puzzling feature of Naumenko’s account is the massive and regular incidence of the described phenomenon, the fact also reported earlier by Sarychev (1987).

Liao et al. (2017) recovered self-fertile maternal phenotypes along with apparently regular interspecific hybrids by fertilizing senescent flowers of cytoplasmically male sterile N. tabacum with pollen of N. alata. The puzzling part of the latter report was that the N. tabacum-like plants showed restored normal stamen morphology and self-fertility, and according to flow cytometry measurements, their genome was considerably smaller than that of their maternal parent. According to Liao et al. (2017), the maternal phenotypes in the offspring were actually hybrids rather than true maternals since N. alata-specific fragments, including the putative male fertility-restoring genes, were amplified in them with selected SSR probes. The two accounts, by Naumenko (2012) and by Liao et al. (2017), bear intriguing similarities, but Naumenko obviously used male fertile maternal plants in her study since she propagated her apomictics by selfing.

Some of the results described in this section are loosely reminiscent of the experiments reported by Pandey on what he called egg transformation without gametic fusion (Pandey, 1974, 1975, Pandey & Phung, 1982, see Sect. 4.4.4).

Overall, due to its elusive and inconsistent nature as well as different manifestations, the phenomenon of maternal genotypes in the progeny of some interspecific crosses is not readily amenable to systematic study and, therefore, is habitually dismissed as resulting from inadequate protection against inadvertent pollen contamination. This notwithstanding, maybe it deserves more attention than it has hitherto received.

3.6 Sterility of Interspecific Hybrids

3.6.1 Causes of Sterility in Amphihaploid Hybrids of Nicotiana

In Nicotiana, once all previously discussed pre- and postfertilization barriers have been successfully overcome or circumvented and the hybrid plant has been brought to flowering, it can be reasonably expected to be completely sterile. Actually, the very first interspecific Nicotiana hybrid ever reported, N. paniculata x N. rustica, was described by its creator J. G. Koelreuter as sterile, which was in stark contrast to the full fertility of the parental forms (Mayr, 1986).

In tobacco, just as in other organisms that perpetuate themselves by sexual reproduction, the fusion of two viable gametes, male and female, is a crucial event in the reproductive process. An orderly and undisturbed reductional division in meiocytes whereby the number of somatic chromosomes of an individual is reduced by half is the key element in assuring the formation of viable and functional gametes. In interspecific hybrids, this regular process can be upset because the chromosomes contributed by the two parental species are structurally different, which prevents their normal pairing, and/or the numbers of chromosomes are different, which leaves some of the chromosomes without a partner with which to pair.

The genus Nicotiana is unique among other plant genera in that its species show a high degree of cross-compatibility (approximately 450 hybrid combinations have been produced thus far (see Chap. 7 of this review), and at the same time, the vast majority of those hybrids are practically self- and cross-sterile, i.e., they are not capable of producing seeds either if pollinated by their own pollen or by the pollen of another species. In other genera, approximately 75% of successful interspecific hybrids show at least some degree of self-fertility (Stebbins, 1950). According to Goodspeed (1954), in the genus Nicotiana, chromosomal rearrangements and other structural changes accumulated faster than barriers to hybridization. This leads to deranged micro- and macrosporogenesis in hybrids, resulting in chromosomally imbalanced and thus inviable or dysfunctional gametes and, consequently, aborted or inviable ovules and pollen. Some intrasectional hybrids, e.g., in the sections Alatae, Trigonophyllae, Tomentosae and Suaveolentes, with equal numbers of chromosomes and chromosome homology high enough to ensure normal pairing regularly produce selfed offspring and can be backcrossed to their parental species. A puzzling exception to this rule was the intersectional hybrid between N. alata (section Alatae, chromosome number n = 9) and N. amplexicaulis (section Suaveolentes, chromosome number n = 18) reported by Gopinath et al. (1970). Although difficult to produce, once obtained, the amphihaploid hybrid (27 somatic chromosomes) was reportedly self- and cross-fertile. Unfortunately, no one else is known to have reproduced and studied that hybrid.

All known interspecific hybrids involving the cultivated tobacco N. tabacum with their number approaching 60 are sterile, although those originating from crossing it with its direct ancestors or their close relatives occasionally yield viable offspring when backcrossed to the N. tabacum parent (Clausen & Cameron, 1944, 1957). Upon backcrossing to N. tabacum, the hybrid N. tabacum × N. tomentosiformis produced offspring among which fertile 48-chromosome N. tabacum-like segregants were found (Brieger (1928). This is not surprising bearing in mind the ‘Drosera’ chromosome pairing in the hybrid parent of the cross (compare Sect. 3.3.4). Other exceptions and reservations concerning the sterility of interspecific hybrids with N. tabacum are indicated in subsequent sections. The sterility of an interspecific hybrid makes it of little use, especially if further generations are considered, e.g. for breeding purposes.

3.6.2 Bioconfinement

In some cases, hybrid sterility may be of potential advantage. Chambers et al. (2011), Ling et al. (2012), and Rice et al. (2013) considered the applicability of the hybrid N. tabacum × N. glauca for transgene-controlled production of pharmaceuticals because of the hybrid’s capacity for biomass production and its other merits. They pointed to the sterility of the hybrid as the safeguard against unintended escape of the transgene to the environment, although they were aware that the sterility of the hybrid was not complete (Chambers et al., 2011; Rice et al., 2013), an issue that will be briefly discussed in the next paragraph of this section. The bioconfinement effect of hybrid sterility was also established for two closely related species, Nicotiana tabacum and N. sylvestris (Ahl-Ahmad et al., 2006). The hybrid of the two species was reported to produce no offspring either by selfing or by backcrossing to N. sylvestris. The authors concluded that the cultivation of a transgene-carrying N. tabacum in close proximity to native or ornamental N. sylvestris poses no significant risk of the unintended release of some novel genetically modified genes to the environment. Because of the high degree of self- and cross sterility plus the added benefits of perennial growth and the ease of clonal propagation, Lim et al. (2006) envisaged the usefulness of genetically modified first-generation ‘synthetic tobacco’ (4x (N. sylvestris × N. tomentosiformis)) for biopharmacy. However, the latter hybrid, although practically sterile, may not be entirely gene escape-proof (see Chap. 2), especially when commercial tobacco is grown nearby. Cost-effective production of seeding material may also be an issue.

3.6.3 Conversion of Sterile Amphihaploids to Fertile Alloploids by Making Use of Natural Processes in Hybrid Plants

Partial or Occasional Self- and Cross-Fertility of Otherwise Sterile Allohaploid Hybrids

In most intended or actually implemented practical uses of interspecific Nicotiana hybrids, their sterility is a liability rather than an asset. The desired goal has been a hybrid that can be sexually perpetuated by selfing and, preferably, one that is also able to produce offspring upon backcrossing to the cultivated species. In their efforts to restore fertility to their interspecific hybrids, the early breeders were, consciously or otherwise, heavily dependent on the production of restitution gametes by their experimental materials. They thus made use of the fact that during the aberrant events of gametogenesis, one or both reductional divisions may fail, resulting in the formation of unreduced gametes that are usually chromosomally balanced and viable, as they retain the genomic integrity of their parental plants. The rate of formation of such restitution gametes may vary from negligible to quite substantial depending on parental genotypes, environment, plant age, etc. (Kostoff, 1943, Goodspeed, 1954, Doroszewska & Berbeć, 1996).

Kostoff (1936, 1938a) heavily relied on a high rate of restitution gametes in his attempt to develop an amphidiploid N. sylvestris x N. tomentosiformis (SSTT) by crossing the sterile F1 hybrid (ST) first to N. sylvestris and the resulting SST hybrid to N. tomentosiformis, the form that came to be known as “Kostoff’s hybrid” (see also Sect. 2.2.2). Kostoff (1943) observed that the production of restitution nuclei by N. sylvestris × N. tomentosiformis was further intensified when the hybrid was exposed to high temperatures accompanied by restricted water supply. Similarly, Rybin (1927) and Eghis (1927) crossed a tetraploid variant of N. tabacum (TTTT) with diploid N. rustica (RR). Upon backcrossing the resulting sesquidiploid (TTR) to the diploid parent (RR), they obtained occasional plants that had the amphidiploid (TTRR) genomic constitution. The above procedure did not always prove successful. In an attempt to produce amphidiploid 4x (N. tabacum x N. benavidesii), the author of this review backcrossed the sesquidiploid N. tabacum-N. tabacum-N. benavidesii to N. benavidesii. All but one of the offspring thus obtained were poorly viable amphihaploids with extra univalents, probably from N. benavidesii, and the remaining single plant was an unstable 68-chromosome subamphidiploid (Berbeć, 1978).

In a sterile amphihaploid, if an unreduced female gamete fuses with its unreduced male counterpart, a fertile amphidiploid may be produced by selfing. The best illustration of this fortuitous process is the case of the first artificial fertile Nicotiana amphidiploid ever made. After many years of unsuccessful attempts to self-pollinate the sterile hybrid 2n (N. glutinosa × N. tabacum), Clausen and Goodspeed (1925) obtained three seeds, out of which one germinated and grew to an amphidiploid. A few years later, that amphidiploid was used by Holmes (1938) to transfer resistance to tobacco mosaic virus (TMV) from N. glutinosa to N. tabacum (see Sect. 4.1). That resistance continues to be deployed in contemporary cultivars and thus, at least in several cases, its origin can be traced back to those remote but fateful events recorded by Clausen and Holmes many decades ago. There are several cases on record in which otherwise sterile hybrids of different Nicotiana species with N. tabacum yielded viable offspring through assiduous effort of selfing the amphihaploid plants. The key to success consisted in self-pollinating on a scale large enough to increase the small odds of an unreduced male and female gamete fusing together into an allopolyploid zygote. In exceptional cases, amphidiploids or sesquidiploids can be produced directly from mating two diploid species. Such a chance fusion of two restitution gametes produced by two parental species may give rise to a direct allopolyploid hybrid. Ternovsky (1962) reported spontaneous amphidiploids directly from the crosses N. glutinosa × N. tabacum, N. tabacum × N. glauca and N. tabacum × N. sylvestris. Another case of this kind is the appearance of the sesquidiploid plant in the hybrid progeny of the cross N. tabacum × N. sylvestris that appears to have arisen from the union of diploid restitution gametes of N. tabacum with a normal haploid gamete of N. sylvestris (Webber, 1930). If an amphihaploid was backcrossed to the N. tabacum parent, partly fertile sesquidiploids or near sesquidiploid plants were the usual outcome. Clayton (1954) observed a high rate of restitution gametes in the hybrid N. debneyi × N. tabacum, resulting in partially restored self-fertility and the production of allopolyploid progeny.

Low temperatures and excess moisture led to partially restored fertility in the hybrids N. debneyi × N. tabacum and N. tabacum × N. glauca (communicated to the author of this review by J. Berbeć). The production of restitution nuclei and viable pollen grains increased in the hybrid 2x (N. tabacum × N. africana) as the plants grew older (Doroszewska & Berbeć, 1996). In the latter case, the hybrid plants failed to yield any spontaneous seeds, notwirhstanding.

In some other reports (Berbeć & Opoka, 1966; Berbeć, 1971), the hybrid N. tabacum × N. glauca and its reciprocal retained some vestigial self-fertility, and the two hybrids could also be used both as pollen and egg parents to produce backcross offspring with N. tabacum from which further selfed generations were obtained. This simple expedient to overcome the sterility barrier by backcrossing the sterile hybrid to one of the parents was probably fairly frequently resorted to and in several cases proved successful (East, 1928; Ternovsky, 1936a, 1936b; Burk, 1967; Wichert-Kobus, 1967, 1971; Berbeć, 1980; Berbeć et al., 1982; Nikova et al., 1997).

Aneuploid plants having from 28 to 34 chromosomes obtained by culturing the anthers of the sterile hybrid N. tabacum × N. sylvestris probably also arose from the restitution gametes produced by that hybrid (Takahashi, 1973). Another plausible mechanism was the production of partly functional generative nuclei by the hybrid as a result of the ‘Drosera’ pairing process.

The citations of cases where spontaneous amphidiploids or sesquidiploids involving N. tabacum were produced are given in Table 3.5 on a species-by-species basis. The allopolyploids were obtained either directly from crossing diploid parental species or by selfing or cross-pollination of amphihaploid hybrids.

Table 3.5 Spontaneous production of allopolyploid offspring by direct crosses of N. tabacum with an alien species or by amphihaploid hybrids involving N. tabacum as a result of self-fertilization (S) or cross-pollination (CP) with the N. tabacum parent
Full size table

Use of Autotetraploid Forms of One or Both Parental Species

Sterility barriers can also be circumvented if one or both parental species are used in the autotetraploid form. Clayton (1947) crossed autotetraploid N. tabacum with autotetraploid N. longiflora and obtained a partially fertile amphidiploid that was used in the interspecific transfer of a disease resistance factor. Similarly, an autotetraploid accession of N. knightiana was mated as the female to an autotetraploid variant of N. tabacum (4n = 96) to produce a fertile amphidiploid 4x (N. knightiana × N. tabacum) (Berbeć & Doroszewska, 1992). After a series of successful backcrosses to N. tabacum as the recurrent male parent, an alloplasmic lineage N. tabacum cms knightiana was developed that showed vestigial self-fertility. The negative aspect of this “tetraploid-tetraploid” approach is that autotetraploids usually show a certain percentage of imbalanced gametes due to frequently occurring polyvalent associations in meiosis. The resulting offspring of such tetra-tetra crosses are thus very likely to contain aneuploids along with regular amphidiploids and may happen to be deficient for a desired genetic factor from the wild species.

More frequently, autotetraploid variants of N. tabacum were crossed with diploid Nicotiana species to produce sesquidiploids. Since sesquidiploids contain, at least in theory, a full diploid chromosome complement of the tetraploid parent, they are able to produce some chromosomally balanced, viable gametes and are usually at least partially fertile. The first sesquidiploids obtained in this manner were produced by crossing autotetraploid N. tabacum with N. gossei (Valleau, 1952) and with N. plumbaginifolia (Chaplin, 1954; Ar-Rushdi, 1957). Discussion on the use of autotetraploid forms of Nicotiana as components of interspecific hybrids is further expanded in part 4.4.3 of this discourse.

Known interspecific combinations involving autotetraploid N. tabacum are listed in Table 3.6.

Table 3.6 Instances of sesquidiploid hybrids obtained by direct crossing autotetraploid forms of N. tabacum with other Nicotiana species
Full size table

3.6.4 Conversion of Sterile Hybrid Plants to Fertility by Means of External Agents

Use of Diverse Chemical or Physical Agents

Since conversion to fertility that relies on the vagaries of nature was both time consuming and highly unreliable, external agents that might induce the chromosome doubling process were tried. Eghis (1930) applied chloroform to obtain the allopolyploid 4x (N. tabacum × N. sylvestris).Footnote 3 In another early attempt, Ternovsky (1939) restored fertility to the F1 hybrid N. tabacum × N. sylvestris by exposing it to high temperature. Another approach to polyploidization was attempted by Kostoff (1937), who centrifuged the allohaploid germinating seeds of N. tabacum x N. rustica and, as a result, a fertile branch with the doubled chromosome complement was obtained.

Use of Antimitotic Drugs

The search for efficient methods to induce chromosome doubling resulted in the identification of substances that act as antimitotic drugs. Such substances interfere with the formation of the spindle during cell division. The chromosomes lag at the equatorial plate of the dividing cell, and the newly synthetized nuclear membrane surrounds the chromosomes, which are now double in number. If the anti-mitotic agent is removed in a timely manner, the cells continue to divide, thus giving rise to polyploid tissues. Acenaphtene was probably the first antimitotic drug used by some Nicotiana investigators for its chromosome doubling effects. Using that agent, Ternovsky (1962) obtained the amphidiploid 4x (N. sylvestris × N. tomentosiformis). Bolsunov used acenaphtene to produce amphidiploids from F1 (N. rustica x N. tabacum) (Bolsunov, 1963) and from F1 (N. rustica × N. exigua) (Bolsunov, 1970).

More recently, oryzalin, an herbicidal substance but also a very powerful antimitotic, was used to induce chromosome doubling in the amphihaploid N. sylvestris × N. tomentosiformis (Lim et al., 2006).

Starting with the first experiments by Warmke and Blakeslee (1939) that involved the hybrid N. tabacum × N. glutinosa, colchicine became the antimitotic drug of choice to restore fertility to sterile Nicotiana hybrids. Colchicine is used in various formulations, e.g., as water solutions of different strengths, water solutions with agar, mixtures with lanolin, etc. It was applied both in situ, on field- or greenhouse-grown plants and in vitro in various types of aseptic cultures. Chromosome doubling is induced at various growth stages and in different plant parts, in cultured embryos, germinating seeds, seedlings and growing plants. The dividing meristems to which the drug is applied include apices and ancillary buds. Some examples are given in Table 3.7.

Table 3.7 Induction of alloploidy in some interspecific hybrids involving N. tabacum by using colchicine
Full size table

3.6.5 Regeneration of Hybrid Plants from In Vivo and In Vitro Cultures

Regeneration from Callus In Situ

This oldest and probably long-forgotten method made use of polyploid cells already present in the plant by inducing the growth of calli in vivo. Polyploid cells originate during the tissue differentiation process, mostly through endomitosis, which involves the mitotic division of chromosomes within an intact nuclear membrane. Using this approach, Protassenya (1935) obtained an allopolyploid hybrid from 2x (N. rustica × N. tabacum and Greenleaf (1938) doubled the chromosome complements of 2x (N. sylvestris × N. tomentosiformis), 2x (N. sylvestris × N. setchellii), 2x (N. sylvestris × N. tomentosa) and 2x (N. glutinosa × N. sylvestris). The callus growth in the amphidiploids produced by Greenleaf was induced by the application of heteroauxin (IAA).

Regeneration from Calli In Vitro

As discussed in a previous section, aseptic cultures were found to be helpful in overcoming incongruity of certain interspecific combinations that resulted in premature death of hybrid seedlings. Since the regeneration of viable plants from the explants of lethal hybrids passes through the callus phase and sometimes requires several passages of culture to take effect, it also offers an opportunity for the preexisting endomitotic cells to develop into allopolyploid along with amphihaploid shoots. The reader is referred back to Table 3.4 for rare instances of hybrids involving N. tabacum, the culture of which resulted in restoration of both viability and fertility. More frequently, chromosome number was doubled, and fertility was restored by taking explants from sterile hybrid plants that had grown past the stage critical for survival (Table 3.8).

Table 3.8 Instances of restoring fertility to amphihaploid hybrids involving N. tabacum by means of culturing explants from viable hybrid plants at advanced growth stages
Full size table

3.6.6 ‘Synthetic Species’

Synthetized amphidiploids in Nicotiana are known to differ from one another in the extent of variation they exhibit in successive selfed generations. Some lineages of the new amphidiploid 4x (N. wuttkei × N. tabacum) (Laskowska et al., 2015) were stable enough to deserve the name of a ‘synthetic species’. Such stable, self-perpetuating ‘synthetic species’ were previously developed within the section Suaveolentes by Krishnamurthy and Gopinath (1969): 4x (N. velutina × N. amplexicaulis), 4x (N. occidentalis × N. amplexicaulis), including the nullisomic lineage 4n = 76 of the latter amphidiploid. Each of those artificial amphidiploids was given a regular botanical description and treated, as the authors put it, “equal in rank with the existing Nicotiana species”. Some of such ‘synthetic species’ were even given specific names: N. × obtusiata for 4x (N. obtusifolia × N. attenuata) (Anssour et al., 2009; Krügel, 2010; McCarthy et al., 2015) and N. × mierata for 4x (N. miersii x N. attenuata) (Pearse et al., 2006; Krügel, 2010), N. edwardsonii for 4x (N. glutinosa × N. clevelandii) (Christie, 1969), N. vavilovii for 4x (N. glauca × N. langsdorffii) (Kostoff, 1938b; Kostoff, 1939a, 1939b), N. × diruex for 4x (N rustica × N. exigua) (Bolsunov, 1970), N. × didebta for 4x (N. debneyi × N. tabacum) (Clayton et al., 1967, He et al., 2019, misnamed by the latter authors ‘N. x didepta), N. × digluta for 4x (N. glutinosa × N. tabacum) (Clausen & Goodspeed, 1925; Clausen, 1928); N. x disualovii for N. suaveolens x N. quadrivalvis (bigelovii) (Modilevsky, 1939); N. x flindersiensis for 4x (N. suaveolens x N. glauca) (Smith & Abashian, 1963); N. × ditagla for 4x (N. tabacum × N. glauca) (Ternovsky, 1934; Modilevsky, 1936). Other lineages of the latter amphidiploid were found to be highly unstable (Szilagyi, 1975). Unstable amphidiploids will be discussed in Sect. 4.5.4.

The relatively recently produced synthetic species N. excelsiana for 4x (N. excelsior × N. benthamiana) (Fitzmaurice, 2002) shows that there is still a large unexploited potential, both academic and commercial, in interspecific combinations in Nicotiana. Due to its biological properties, N. excelsiana gained the status of a “proprietary species” protected by patent rights as a convenient tool in a newly developed protein production technology. In a recent study, N. x excelsiana was demonstrated as a valuable and agronomically exploitable source of griffithsin, an anti-HIV drug (Eapen et al., 2020) and was also studied for other secondary metabolites (Mihaylova-Kroumova et al., 2020).

3.7 Parasexual Hybrids

3.7.1 Parasexual Hybrids by Fusion of Isolated Somatic Protoplasts

The idea of circumventing prefertilization barriers to crossability that had evolved at the gametic level prompted the attempts to fuse the somatic cells that lacked such obstructions. It was also envisaged that fusion of unreduced somatic cells would directly produce fertile amphidiploids, thereby bypassing the sterility of sexually produced hybrids. In its basics, the procedure has remained unchanged from its inception and is divided into three stages:

  • isolation of protoplasts by enzymatic degradation of cell walls

  • inducing the naked protoplasts to fuse by the presence of chemical agents or by electric fields

  • selective culture of fused protoplasts based on their physical properties, the presence of fluorescent markers or genetic complementation, e.g., resistance to antibiotics

Since spontaneous fusion of naked protoplasts is a rare phenomenon, several agents, both chemical and physical, were tested for fusion-inducing action. Among the physical agents, the electric-field mediated method, also called electrofusion, is most frequently used. In this method, the protoplasts are brought into close contact by the application of an alternating electric field (AC) followed by exposure to direct current (DC) pulses (Davey, 2017).

Despite the high efficiency of electrofusion, polyethylene glycol (PEG), a chemical agent, has been most popular in creating interspecific somatic hybrids in Nicotiana, although the compound is toxic to plant cells. Another frequently used approach is the combination of high pH and high concentration of Ca2+ cations (Ilcheva & San, 1997; Davey, 2017).

After the fusion treatment, the medium contains both fused heterokaryons and unfused parental protoplasts. Different methods were applied to discriminate against the latter in the selection process. They included simple visual identification (Bates, 1985; Nagao, 1978; Hamill et al., 1984) and the use of various genetic markers, such as chlorophyll mutations (Evans et al., 1981, 1982, 1983; Aviv & Galun, 1986, 1987), resistance to antibiotics, both spontaneous (Medgyesy et al., 1980) and transgenic (Bates, 1990; Pental et al., 1988, 1989; Lu & Yang, 1996). Transgenic resistance to two different antibiotics provided double complementary selective systems under which only heterokaryons could survive in a growth medium containing both markers (Sproule et al., 1991; Donaldson et al., 1993, 1995; Ilcheva et al., 2000, 2001).

Although Zheng et al. (2018) wrote that the cytoplasmically male sterile line of N. tabacum ‘cms-sua’ used in their study originated from the somatic fusion between N. suaveolens and N. tabacum made in the 1950s (sic!) the first well-documented interspecific somatic hybrid in Nicotiana was reported by Carlson et al. (1972). The hybrid N. langsdorffii + N. glauca was a regular fertile amphidiploid (4x = 42) and thus seemed to confirm the hopes attached to interspecific hybridization at the somatic level. Soon thereafter, however, it became apparent that the method had serious constraints. Over the years, the number of interspecific somatic combinations, mostly those involving N. tabacum, that yielded genetically stable, fertile amphidiploids was disappointingly small and limited to hybrids of N. tabacum with N. glauca, N. nesophila, N. debneyi, N. megalosiphon, and N. otophora (Table 3.9). The other hybrids synthesized by protoplast fusion that involved N. tabacum showed high variability in external morphology and mostly aneuploid chromosome numbers due to chromosome elimination, either random or preferential, resulting in asymmetric hybrids, i.e., those with predominance of genetic material from one parental species (e.g., Donaldson et al., 1995; Ilcheva et al., 1997, 2000). Nonetheless, the irregularity and imbalance of interspecific protoplast fusion products may carry inherent advantages, which is explained in the paragraph to follow.

Table 3.9 Interspecific parasexual hybrids involving N. tabacum
Full size table

Asymmetric hybrids have also been synthetized consciously using chemical agents (iodoacetate) or irradiation (gamma or X-rays) to inactivate all or part of the nuclear genome of one parent prior to fusion (e.g., Bates, 1990). Highly asymmetric hybrids to which the cytoplasmic DNA was contributed almost exclusively by the irradiated parent whereas the other parent donated both cytoplasmic and nuclear DNA (the so-called cybrids obtained by the donor-recipient method) proved to be a useful tool to produce interspecific mitochondrial recombinants and provided a fast method to transfer whole plasmons or selected cytoplasmic traits from one species to another (see also Sect. 5.3.1). Cytoplasmic recombinants are practically impossible to obtain by sexual hybridization since in Nicotiana, as in most other genera, the cytoplasmic DNA is inherited unilaterally through maternal lineage, save for some rare exceptions (Medgyesy et al., 1985; Horlow et al., 1990; Svab & Maliga, 2007). In this context, one may also note that such unilaterally incomplete or asymmetric hybrids can also be obtained from sexual matings using irradiated pollen (see Sect. 3.4.2). An extreme case of that latter approach was the highly contested “egg transformation” (see Sect. 4.4.4).

3.7.2 Gametosomatic Hybrids

A fusion of diploid mesophyll species of N. tabacum with haploid gametophyte protoplasts of an alien species was devised with an expectation to obtain sesquidiploid plants as a starting material for interspecific gene transfer equivalent to sesquidiploids from sexual matings (Davey et al., 1996, see also Sects. 3.6.3, 4.5.6). The method was experimentally tested by Pirrie and Power (1986) and by Giddings and Rees (1992). Their gametosomatic hybrids 2n N. tabacum + 1n N. glutinosa actually mimicked sesquidiploids (pentaploids) from sexual matings. In those experiments, haploid protoplasts were isolated at the tetrad stage. Alternatively, protoplasts isolated from mature pollen grains can be used for gametosomatic fusion (Desprez et al., 1992, Lu & Yang, 1996, Ping et al., 1996). Another benefit of gametosomatic fusion is that haploid gametophyte protoplasts fail to divide and do not form colonies in culture, which simplifies the selection of heterokaryons (Davey et al., 1996).

However, not all products of gametosomatic fusion represent true pentaploids. Gametosomatic regenerants (1n N. tabacum + 2n N. plumbaginifolia) made by Desprez et al. (1992) represented an array of sterile aneuploid forms that also showed mixoploidy plus a 48-chromosome asymmetric cybrid. According to those investigators, androgenetic regenerants obtained by asymmetric gametosomatic fusion can be used to transfer mitochondrial genomes separately from chloroplast genomes.

Despite their potential value, few interspecific gametosomatic hybrids have been reported. They include, apart from the previously mentioned 2n N. tabacum + 1n N. glutinosa, 1n N. tabacum + 2n N. plumbaginifolia (Desprez et al., 1992) and 1n N. tabacum + 2n N rustica (Pental et al., 1988, 1989; Mukhopadhyay et al., 1991; Lu & Yang, 1996; Ping et al., 1996).

A concise and informative review of parasexual hybridization by protoplast fusion in Nicotiana was prepared by Ilcheva and San (1997).

3.7.3 Graft Hybrids and Horizontal Gene Transfer

Grafting occurs when segments of two different plants come into close contact and fuse together into one separate and independent plant organism. In such a union of two plants or plant pieces, one of them provides the root of the new plant and is called rootstock or simply stock, while the other serves as the shoot with leaves and is referred to as ‘scion’. Grafting as a fusion of two or more separate plant organisms occurs spontaneously in nature but has also been practiced by agriculturists.

In fruit farming practice, grafts have been used for two different purposes: for better growth and desirable performance and for genetic effects (Ohta, 1991).

The former function of grafting has been widely practiced by horticulturists from early antiquity, and its major objective is to improve the growth and performance of fruit plants or ornamentals by combining the superior qualities of the stock with those of the scion, e.g., winter hardiness with palatable fruit or other edible parts. The latter, much less popular and more controversial, was about to impose heritable changes in the scion by transferring genetic information from the stock. In the 1940s and 1950s, the concept gave rise to the so-called Michurinist genetics developed by the Soviet horticulturist I. V. Michurin, and the idea was officially proclaimed in the Soviet Union as an essential part of the so-called “revolutionary agrobiology”, the movement advanced by T. Lysenko and his followers throughout the Eastern bloc in deliberate and blatant opposition to the Mendelian principles of heredity (Goldschmidt, 2014; Zhou & Liu, 2015). The ‘graft hybrids’ theory was deeply mistrusted by scientists in the West and repeatedly proven to lack a scientific basis (Goldschmidt, 2014). Ultimately, it was abandoned both in the Soviet union and in other Eastern bloc countries. However, the recent decades have witnessed a revived interest in epigenetic and hereditary effects of grafting (Ohta, 1991; Goldschmidt, 2014).

Nicotiana species are easily amenable to grafting both among themselves or with other related solanaceous plants, but the technique has, until very recently, attracted little attention from Mendelian tobacco geneticists and breeders. It is only at the beginning of this century that R. Bock and his team from the Max Planck Institute came up with evidence that the entire chloroplast genomes could be transferred through the graft junction from N. tabacum to two other Nicotiana species: N. glauca and N. benthamiana (Stegemann et al., 2012). In their experiment, the authors demonstrated that the transfer was restricted entirely to chloroplast DNA and did not involve any nuclear DNA fragments.

However, in another experiment reported 2 years later (Fuentes et al., 2014), the same team demonstrated the interspecific fusion of the whole genomes of two grafting partners, N. glauca and N. tabacum. After the fusion of stock and scion had taken place, fragments of tissue of the fusion zone were excised and cultured in vitro. Callus culture and plant regeneration were performed by following the genetic complementation protocol based on double selectable markers, a technique routinely applied in somatic hybridization by protoplast fusion. As a result, a fully fertile, regular 72-chromosome amphidiploid 4x (N. glauca × N. tabacum) was obtained and given the name ‘Nicotiana tabauca’, a new species that arose by natural fusion of somatic cells.

A few observations can be made in connection with this unusual finding. The authors’ argument that their discovery supports the likelihood of spontaneous asexual hybridization taking place in nature seems to be essentially valid, although in their experiments they had to resort to artificial tools, unknown to nature, to make it happen. This notwithstanding, when put in their long-term evolutionary perspective, even extremely rare and least likely events can and most likely do occur. The authors also presented their discovery as a new tool for crop improvement mostly because, as they argued, grafting is technically less demanding than protoplast fusion. Indeed, it is, but the sexual method is even less demanding. Precisely the allopolyploids N. tabacum × N. glauca and their reciprocals have been repeatedly obtained by conventional crossing for nearly a century, and even the name for that ‘artificial species’ was invented (‘Nicotiana ditagla’) preceding ‘N. tabauca’ coined by the authors of the report by several decades (see Sect. 3.6.6). This notwithstanding, the significance of that discovery cannot be overestimated and fully deserves to be followed by other experiments involving other Nicotiana species, including those that show a high degree of mutual incompatibility. Regrettably, no new reports in that area, theoretical or practical, seem to have emerged thus far.

The demonstration of the feasibility of obtaining an interspecific hybrid by grafting is obviously reminiscent of the graft hybrids in the former Soviet Union mentioned in the introductory remarks to this section. The recent report by scientists from the Max Planck Institute may shed somewhat different light on the work of Mitschurin in Russia but also similar horticultural experiments by Burbank in the United States. The historical context to the achievement of Ignacia Fuentes and her colleagues was recalled by Zhou and Liu (2015).

One should also add that an analogous interspecific horizontal transfer through grafting was recently reported for mitochondrial genes from N. sylvestris responsible for restoring male fertility to the alloplasmic line of N. tabacum with the mitochondrial genome of N. undulata (Gurdon et al., 2016, see Sect. 5.3.2 on restoring male fertility to cms lines).

3.8 Ending Notes on Sexual and Asexual Interspecific Hybrids Involving N. tabacum

Somatic and gametosomatic hybrids that involve N. tabacum are listed in Table 3.9. A compilation of all interspecific hybrids involving cultivated tobacco, regardless of the method by which they were obtained, is presented in Table 3.10. Information was found on a total of 59 hybrids involving N. tabacum, and reciprocals were not included in the count.

Table 3.10 All interspecific hybrids involving the cultivated species Nicotiana tabacum (after Berbeć and Doroszewska 2020) with minor additions and modifications)
Full size table

In spite of various barriers to crossability discussed in the previous sections, the vast majority of those hybrids could be obtained by conventional crossing. Actually, only very few of those combinations may be considered to have been made possible owing to the use of advanced technologies. N. nesophila x N. tabacum (Reed & Collins, 1978), N. occidentalis x N. tabacum (Butenko et al., 1970), N. rosulata x N. tabacum (Ternovsky et al., 1976), N. stocktonii x N. tabacum (Reed & Collins, 1978) are known only from reports where hybridization was aided by tissue culture. N. rotundifolia + N. tabacum was reported as a somatic hybrid only (Ilcheva et al., 2001). Genetic engineering was deployed to obtain viable hybrids of N. simulans x N. tabacum and N. umbratica x N. tabacum (Ma et al., 2020). However, the latter two hybrids were also reported or hinted at by other authors (Kubo, 1985; Murthy et al., 2014), and N. stocktonii x N. tabacum was reported by Wong (1975). The reports on the latter three hybrids lacked details on how they were produced. Of two reports on N. tabacum x N. bonariensis, no details are known on the hybrid reported by Busconi et al. (2010).

Thirty-six Nicotiana amphihaploids that involved N. tabacum amphidiploids or near amphidiploids were produced by using various approaches. Most of those diploidized F1 hybrids showed at least partially restored self-fertility: one was a female sterile amphidiploid (N. tabacum × N. setchellii), and the other was a nearly female sterile amphidiploid (N. obtusifolia × N. tabacum). Female sterility was also observed in the trigenomic allohexaploid 6x (N. tabacum × (N. setchellii × N. otophora) (Berbeć et al., 1982).

The author of this book found information on 26 Nicotiana species that were hybridized with N. tabacum by protoplast fusion, fewer than half the number of reported sexual hybrids (compare Tables 3.9 and 3.10). Among those 26 somatic hybrids, only a handful represented those in which whole genomes of both parents became united (N. glauca + N. tabacum, including a graft hybrid, N. nesophila + N. tabacum, N. rustica + N. tabacum, N. debneyi + N. tabacum, N. otophora + N. tabacum, N. glutinosa + N. tabacum). To the authors’ knowledge, only two somatic hybrids with N. tabacum obtained by somatic fusion, N. rotundifolia + N. tabacum and possibly also N. arentsii + N. tabacum, have not been obtained by conventional sexual methods.

It appears that while the parasexual approach circumvents the prefertilization barriers that separate species from one another, its role in alleviating the incongruities existing between the fused genomes is of far less importance. The experience with somatic hybrids has confirmed an early observation by Zenkteler and Melchers (1978) that protoplast fusion contributes but little to expanding the crossability of different species within a genus. Chromosome loss, genome instability and other manifestations of intergenomic incongruities are among the common consequences of alloploidization and have been documented for both natural and synthetic allopolyploids that have arisen through the sexual process, a subject discussed in one of the previous sections. From numerous accounts, it appears that parasexual hybridization seems to exacerbate rather than mitigate those inherent incongruities since they are obviously far more numerous and more intense in somatic hybrids than those encountered in analogous hybrids synthesized via the sexual process. This said, asymmetry, a frequent phenomenon in most asexually produced hybrids, may actually facilitate gene flow between the fused genomes by bypassing the so-called bottle-necks of sexual introgression and by eliminating many deleterious linkage or epistatic effects that plague the sexual routes of gene transfer.

Last but not least, from the perspective of practical breeding issues to be discussed in the subsequent chapters, parasexual methods share the same disadvantage with genetic transformations at the molecular level. To wit, lawmakers of at least some countries have listed protoplast fusion been among the technologies used to generate genetically modified organisms (GMOs) thus effectively banning it in the development of commercially exploitable cultivars.