Maternal transmission of cytoplasmic DNA in interspecific hybrids of peat mosses, Sphagnum (Bryophyta)


Rayna Natcheva, Institute of Botany, Bulgarian Academy of Sciences, 23 Acad. G. Bonchev str., 1113 Sofia, Bulgaria.
Tel.: +359 2 979 37 64; fax: +359 2 871 90 32; e-mail:


The progeny of spontaneous interspecific hybrid sporophytes of Sphagnum were used to analyse the inheritance of cytoplasmic DNA. The analysis showed that only the female parent donated chloroplasts and mitochondria in Sphagnum hybrids. Thus, this is the first study demonstrating maternal cytoplasmic inheritance in a nonvascular land plant. This finding has important implications for phylogenetic reconstructions utilizing chloroplast and mitochondrial DNA sequences as well as for the evolution of cytoplasmic inheritance in relation to the life cycle of land plants.


Organelle inheritance, extensively studied in vascular plants, is predominantly uniparental via the maternal line in angiosperms (Hagemann, 1992; Mogensen, 1996; Birky, 2001), but mainly paternal in gymnosperms (e.g. Dong & Wagner, 1994; Hippkins et al., 1994). However, there is a strong bias in the study of organelle transmission towards seed plants (Birky, 1995). Much less is known for the plastid transmission in the nonseed plants occupying more basal position in the phylogeny of land plants. Hitherto, the question of plastid and mitochondrial transmission has been studied in horsetails (Gullion & Raquin, 2000), ferns (Gastony & Yatskievych, 1992; Vogel et al., 1998) and allopolyploid bryophytes (Wyatt et al., 1988; Pacak & Szweykowska-Kulinska, 2003; Jankowiak & Szweykowska-Kulinska, 2004; Jankowiak et al., 2005) but not yet in homoploid members of any of the bryophyte clades (mosses, liverworts and hornworts) that diverged during the early history of land colonization (Goffinet, 2000).

In bryophytes, the long-lived stage of the life cycle is the haploid gametophyte. After fusion of a male and a female gamete a diploid sporophyte develops on the female gametophyte. Haploid spores are subsequently formed in the sporophyte after meiosis. Our studies of hybridization between the peat mosses Sphagnum capillifolium (Ehrh.) Hedw. and Sphagnum quinquefarium (Braithw.) Warnst. (Natcheva & Cronberg, 2007) offered an opportunity to study cytoplasmic inheritance, as we were able to detect interspecific differences in DNA sequences from both chloroplasts and mitochondria.

Materials and methods


Sphagnum capillifolium and S. quinquefarium were sampled in the vicinity of Skinnskatteberg, Central Sweden, from a selectively felled spruce forest. We chose a mixed colony where fertile clones of the two species grew in intimate contact. Ten shoots ofS. quinquefarium with intact spore capsules from the zone of contact were sampled. As there were no sporulating S. capillifolium shoots growing close to S. quinquefarium shoots (all shoots were male fertile), the nearest fertile shoots with sporophytes were collected ca. 20 cm away from any male fertile S. quinquefarium. The sporulating samples were placed in plastic bags, sprayed with water and transported the next day to the laboratory. Five additional fertile single-species colonies of each S. capillifolium and S. quinquefarium were selected and five sporulating shoots were sampled as well as five male fertile shoots of S. capillifolium (S. quinquefarium is always bisexual) from the respective colony.

Spore germination

Ten intact sporophytes from S. quinquefarium shoots from the contact zone and five from the farthest end of the patch as well as five sporophytes from S. capillifolium and S. quinquefarium from the single-species colonies were selected. The maternal shoots were dried and preserved for DNA extraction. The capsules were surface sterilized with 2% chlorine bleach for 5 min. The spores from each sporophyte were released in 1 ml of sterilized distilled water and inoculated in Petri dishes onto agar solidified sterile nutrient solution as described by Cronberg (1996). The Petri dishes were incubated at room temperature in a north-facing window. Young gametophytes were harvested after 4 months, dried and used for DNA extraction.

DNA analysis

DNA extraction was performed according to the standard CTAB protocol (Doyle & Doyle, 1987). The chloroplast haplotypes were assessed by PCR-RFLP of the tRNALeu (UAA) intron (hereafter trnL). The trnL intron was amplified using primers c and d of Taberlet et al. (1991). The PCR products were digested by Hinf I endonuclease (Roche, Mannheim, Germany) according to the manufacturer's manual. The fragments were separated on 2% agarose gel and visualized by staining with ethidium bromide. To test the sensitivity of the method, amplifications were performed on total DNA extracts of S. capillifolium and S. quinquefarium mixed reciprocally at proportions 1 : 1, 1 : 10, 1 : 15 and 1 : 20. Additionally, sequences were obtained from the trnL intron, as well as from the NADH protein coding subunits 5 and 7 of the mitochondrial genome (hereafter, nad5 and nad7 respectively). The latter two regions were sequenced according to the protocol of Shaw et al. (2003) and aligned manually. The hybrid origin of sporophytes was verified by ISSR fingerprinting (reported in Natcheva & Cronberg, 2007).

Results and discussion

Hybrid sporophytes had particularly low fertility with few spores germinating and even fewer surviving to produce gametophytes. Thirty-four hybrid gametophytes and 10 gametophytes of each pure S. capillifolium and S. quinquefarium were screened. A single chloroplast and mitochondrial haplotype was observed in every accession indicating uniparental inheritance. A total of two PCR-RFLP chloroplast haplotypes were found in the material studied (Fig. 1). The progeny of all sporophytes possessed the haplotype that was present in the maternal shoot. Haplotype A was fixed in the progeny of sporophytes borne on S. capillifolium shoots and haplotype B was fixed in the progeny of S. quinquefarium. The hybrid gametophytes, from sporophytes developed on S. quinquefarium shoots, also possessed haplotype B.

Figure 1.

 Chloroplast haplotypes found in Sphagnum capillifolium (Sc) and Sphagnum quinquefarium (Sq), and their hybrid progeny (Hy) developed on shoots of S. quinquefarium using PCR-RFLP of trnL digested with Hinf I.

Two partial 420 bp trnL sequences were observed in S. capillifolium differing in one substitution. One of the sequences was present in one single-species colony, the other was found in the remaining accessions. A single 409 bp sequence was present in all plants of S. quinquefarium that was identical to the trnL sequence found in the hybrid gametophytes. The differences between the sequences are shown in Table 1. The nad7 region yielded a partial sequence of 602 bp. There was no difference between the sequences of the two species. The nad5 region (partial sequence of 340 bp) differed in one substitution and one 2-bp deletion. The maternal S. quinquefarium sequence was present in the accessions of hybrid gametophytes (Table 1).

Table 1.   Aligned trnL and nad5 sequences found in Sphagnum capillifolium, Sphagnum quinquefarium and the gametophytic progeny of their spontaneous hybrid sporophytes.
  1. Only the informative regions are shown. The dots (.) indicate identical sequences, dashes (–) denote deletions, an asterisk (*) indicates the sequences found in the mixed colony.

Sphagnum capillifolium..g......----..a..tctaccattttatgttatgtcggag..
S. capillifolium*..g......----..g..tctaccattttatgttatgtcggag..
Sphagnum quinquefarium*..-......ataa..g..-------------------------..
nad 5
S. capillifolium*
S. quinquefarium*........a................–..................

The sensitivity test of mixed amplifications was able to detect the minority haplotype at proportion as low as 1 : 15. Thus, the male parent must have donated less than 6.7% of the chloroplast DNA to remain undetected by our method. We therefore conclude that at least the chloroplast transmission in Sphagnum is strictly maternal.

Maternal inheritance in bryophytes has been suggested based on ultrastructural features of the spermatogenesis. Mature sperms of all bryophytes contain one (rarely two) chloroplast and two mitochondria (Renzaglia & Garbary, 2001). However, in Sphagnum (Manton, 1957), some other mosses (Duckett & Renzaglia, 1988) and some liverworts (Renzaglia & Duckett, 1987) they seem to be lost soon after the sperms are released from the antheridium before reaching the egg, suggesting that the paternal plant delivers no organelles to the zygote. In contrast, the hybrid gametophytes inherit the larger part of their nuclear genome from the paternal species, S. capillifolium (Natcheva & Cronberg, 2007).

In the majority of vascular plants plastids and mitochondria are exclusively or predominantly uniparentally inherited from the female parent, although there are exceptions within all studied groups (Birky, 1995, 2001; Mogensen, 1996). Of the few nonseed plants studied uniparental maternal organellar inheritance has been found in horsetails (Gullion & Raquin, 2000) and ferns (Kuroiwa, 1991; Vogel et al., 1998). Likewise, in allopolyploid mosses uniparental cytoplasmic inheritance proved to be the rule both among mosses (Wyatt et al., 1988; Jankowiak et al., 2005) and liverworts (Pacak & Szweykowska-Kulinska, 2003; Jankowiak & Szweykowska-Kulinska, 2004). The scarcity of data on the more basal groups of land plants coupled with the high incidence of homoplasy in the groups studied so far have not allowed an unequivocal reconstruction of the evolutionary history of cytoplasmic inheritance (Birky, 1995). The phylogenetic pattern is relatively clear for mitochondrial gene inheritance with uniparental maternal transmission being predominant in the basal lineages (Birky, 1995). The ancestral state for chloroplast transmission, however, has remained unresolved.

This is the first study to demonstrate that maternal cytoplasmic inheritance is present in homoploid mosses, one the oldest lineages of land plants, and adds weight to the general pattern of the predominance of uniparental maternal cytoplasmic inheritance. If maternal inheritance of chloroplasts and mitochondria proves to be a general pattern in bryophytes it would strongly suggest an ancestral state and may have important implications for the evolution of the life cycle and sexual reproduction in land plants.


This study was financed by Ove Almborn's fund. We are grateful to two anonymous reviewers for the critical comments.