ORIGINAL ARTICLES
Self and Cross-incompatibility of Citrus depressa Hayata (Shiikuwasha) and Its Relatives
2023 Volume 92 Issue 2 Pages 134-141
Details
2023 Volume 92 Issue 2 Pages 134-141
Self-incompatibility and compatibility of Citrus depressa (Shiikuwasha) and its related species were determined by pollen tube growth within the style. Among the 30 accessions investigated, twelve C. depressa, two C. ryukyuensis (Tanibuta), and three species derived from C. depressa, C. keraji, C. oto, and C. tarogayo, were self-incompatible, while ten C. depressa, one C. tachibana (Tachibana), and two C. tachibana relatives showed self-compatibility. Of the self-incompatible accessions, some cross-incompatible relationships were discovered. All self-incompatible C. depressa except “Ishikunibu” were cross-incompatible with each other. Two C. ryukyuensis were cross-incompatible. This species showed cross-compatibility with C. depressa. C. depressa “Ishikunibu” was cross-compatible with all self-incompatible C. depressa and C. ryukyuensis. From the above, three genotypes determining incompatibility were found in C. depressa and C. ryukyuensis. In addition, all self-incompatible C. depressa and C. ryukyuensis accessions were cross-compatible with self-incompatible C. keraji, C. oto, and C. tarogayo which are related to C. depressa. The present study suggests that S (self-incompatible) alleles of C. depressa and C. ryukyuensis are derivatives of themselves and not from C. maxima (pummelo), major source of citrus S alleles.
Self-incompatibility is a genetically controlled phenomenon preventing self-pollination fertilization, despite both male and female tissues showing fertility, and is an important trait for citrus fruit production. Without parthenocarpy, cross-pollination is required to achieve stable fruit production (Krezdorn and Robinson, 1958; Mustard et al., 1956; Reece and Register, 1961; Soost, 1956). However, coupling with parthenocarpy may produce seedless fruit (Li, 1980; Yamamoto et al., 1995; Yamamoto and Tominaga, 2002).
Thus, the study of incompatibility is important for citrus cultivation. Soost (1965, 1969) revealed that the citrus incompatibility system is gametophytic. He proposed S (self-incompatibility) genotypes for some accessions. Thereafter, Kim et al. (2010, 2011, 2020) and Zhou et al. (2018) determined the S genotype of various citrus by pollination with pollen homozygous for the S genotype. Self-incompatibility of various accessions was also determined (Ngo et al., 2019; Yamamoto et al., 2006a, 2012). These results indicate that all pummelos and many of their relatives were self-incompatible. As pummelo (Citrus maxima (Burm.) Merr.) is a basic Citrus species and ancestor of various citrus accessions (Tshering Penjor et al., 2016; Wu et al., 2018), it may be a major S allele source. On the other hand, accessions belonging to other basic species, citron (C. medica L.) and mandarin (such as C. reticulata Blanco), showed self-compatibility, although some mandarins with pummelo genes were self-incompatible.
However, C. depressa Hayata (Shiikuwasha), a small mandarin, originating from the Ryukyu Islands of Japan showed self-incompatibility (Yamamoto et al., 2006a). This species is a true mandarin, although a very small amount of pummelo admixture is observed (Wu et al., 2021). In contrast, other true mandarins, such as C. sunki (Hayata) hort. ex Tanaka (Sunki), C. reshni hort. ex Tanaka (Cleopatra), and C. tachibana (Makino) Tanaka (Tachibana) showed self-compatibility. Moreover, self-compatible C. depressa exist. These results reveal both self-incompatible and compatible accessions in C. depressa. As marked genetic variability was discovered in this species (Ishikawa et al., 2016; Yamamoto et al., 2017), these results are not surprising. However, previous reports (Yamamoto et al., 2006a) used only three C. depressa accessions. Thus, identifying of self-incompatibility in more C. depressa accessions is necessary.
This study explores local citrus grown on the Ryukyu Islands and various preserved C. depressa accessions over the last 20 years in the orchard of the Faculty of Agriculture of Kagoshima University (Yamamoto et al., 2006b, 2021b). First, self-incompatibility in these preserved C. depressa accessions and related species was investigated. Then, cross-incompatible relationships between self-incompatible accessions were determined. As incompatibility is controlled by the S allele (Soost, 1965, 1969), cross-compatibility is observed when two or more S genotypes exist in self-incompatible accessions. The elucidation of self- and cross-incompatibility in C. depressa and its related species, and the derivation of incompatible alleles, are important to progress studies on citrus incompatibility and identify important genetic resources.
The materials used in the present study were 22 C. depressa, two C. ryukyuensis sp. nov. (a synonym: C. tachibana (Makino) Tanaka var. attenuata) (Tanibuta), one C. tachibana (Makino) Tanaka (Tachibana), two C. tachibana relatives, and three accessions, “Kabuchii” (C. keraji hort. ex Tanaka), “Oto” (C. oto hort. ex Y. Tanaka), and “Tarogayo” (C. tarogayo hort. ex Y. Tanaka) derived from C. depressa (Yamamoto et al., 2021a, 2022) (Table 1; Fig. 1). C. ryukyuensis (Tanibuta) is a new species named Tanibuta as designated by Wu et al. (2021). Although it was considered a type of C. tachibana, its morphological traits differ from regular Tachibana (Inafuku-Teramoto et al., 2010). In the present study, C. ryukyuensis is used as the scientific name for Tanibuta to avoid confusion between Tanibuta and regular Tachibana (C. tachibana). All accessions except “Shiikuwasha (Okitsu)” and “Tachibana” were collected previously (Yamamoto et al., 2006b, 2008, 2021b). When accessions were introduced, their scions were collected on Koshikijima, Kuroshima, Yakushima, and Tokunoshima. The remaining accessions were introduced by seed due to the spread of Huanglongbing (HLB or citrus greening) disease on some islands (Iwanami, 2022). In those accessions, preserved trees have a nucellar seedling origin due to their polyembryony. However, as C. ryukyuensis showed monoembryony, the seed-derived plants were not identical to the original tree on Okinawajima. All accessions were grafted on trifoliate orange (Poncirus trifoliata (L.) Raf.) and preserved in the Toso Orchard of the Faculty of Agriculture, Kagoshima University (Kagoshima, Japan, ca. N 31.34°, E 130.32°, at an elevation of 65 m).
Self-incompatibility/compatibility and pollen fertility of Citrus depressa and its relatives used in this study.
Map of the Ryukyu Islands. Osumi and Amami Islands: Kagoshima Prefecture. Okinawa and Sakishima Islands: Okinawa Prefecture.
All experiments were conducted during the flowering season (late April to early May) in 2015 to 2022. Self- and cross-incompatibility/compatibility were determined by pollen tube growth within the style (Fig. 2). The hand-pollinated flowers were emasculated and enclosed in bags to prevent open pollination. Three to seven days after pollination, flowers were collected and fixed in methanol-acetic acid (3:1, v/v) and stored at −20°C until use. Squash preparations of pistils were stained with decolorized anilin blue (Martin, 1958) for fluorescence microscopy. Thirty accessions were used to determine self-incompatibility and compatibility (Table 1). To determine cross-incompatibility or compatibility in self-incompatible accessions, 63 cross-pollination combinations were conducted (Tables 2 and 4). The number of pollen tubes in the top, middle, and base of the style were counted.
Growth of pollen tubes within style. C: self-compatible “Damannin”, I: self-incompatible “Ishikunibu”.
Cross-incompatibility/compatibility of Citrus depressa and C. ryukyuensis.
Pollen fertility was evaluated by staining with acetocarmine according to Ueno (1986). Five hundred to 600 pollen grains of each accession were stained with 1% acetocarmine.
Pollen fertility was rated before self-incompatibility status determination. In all accessions except “Shokakukei”, pollen fertility was over 80% (Table 1). The pollen fertility of “Shokakukei” was the lowest at 13.9%. Despite the low pollen fertility, this accession was considered able to produce sufficient viable pollen to determine incompatibility or compatibility. This is supported by the previous determination of self-incompatibility and compatibility for the ‘Marsh’ grapefruit, which has a low fertile pollen count of 12.0% (Yamamoto et al., 2006a).
Pollen tube growth was markedly inhibited in 17 accessions with no pollen tubes observed at the base of the style. On the other hand, remaining 13 accessions produced between 17 to 100 pollen tubes (Table 1). The former and latter show self-incompatibility and -compatibility, respectively. Accessions were classified as follows (Table 1):
Self-incompatibility: “Kozumikan”, “Shiikuwasha #15”, “Shiikuwasha #11”, “Shiikunin (Kara)”, “Shiikuribu (Amata)”, “Shiikuribu (Kamishiro)”, “Shiikuribu (Yakomo)”, “Kinkan #4”, “Kinkan #15”, “Kinkan #17”, “Ishikunibu”, and “Shiikuwasha (Okitsu)” (C. depressa), “Tanibuta #1” and “Tanibuta #2” (C. ryukyuensis), “Kabuchi” (C. keraji), “Oto” (C. oto), and “Tarogayo (C. tarogayo)”;
Self-compatibility: “Yamatate” and “Kozu A” (C. tachibana relatives), “Kozu B”, “Shiikuwasha #14”, “Shiikunin (Ama)”, “Kinkan #13”, “Ogimi Kuganii”, “Izumi Kuganii”, “Shokakukei”, “Shiikuwasha #9”, “Shiikuwasha #10”, and “Damannin” (C. depressa), and “Tachibana” (C. tachibana).
Cross-incompatibility/compatibilityCross-incompatibility or compatibility relationships among self-incompatible C. depressa and C. ryukyuensis were first investigated. The results are shown in Table 2.
“Ishikunibu” was cross-compatible with “Kozumikan”, “Shiikuwasha #15”, “Shiikuwasha #11”, “Shiikunin (Kara)”, “Shiikuribu (Amata)”, “Kinkan #4”, and “Tanibuta #1”. No cross-incompatible accessions with “Ishikunibu” were observed. “Tanibuta #1” and “Tanibuta #2” were cross-incompatible. Both were cross-compatible with C. depressa in eleven cross combinations. All cross combinations using self-incompatible C. depressa except “Ishikunibu” as both parents showed cross-incompatibility. Based on the above results, the genotypes for incompatibility were considered as follows: type I: “Ishikunibu”, type S: “Kozumikan”, “Shiikuwasha #15”, “Shiikuwasha #11”, “Shiikunin (Kara)”, “Shiikuribu (Amata)”, “Shiikuribu (Kamishiro)”, “Shiikuribu (Yakomo)”, “Kinkan #4”, “Kinkan #15”, “Kinkan #17”, and “Shiikuwasha (Okitsu)”, and type T: “Tanibuta #1” and “Tanibuta #2” (Table 3).
Type of genotype on incompatibility in Citrus depressa and C. ryukyuensis used in the present study.
Next, reciprocal crosses between accessions belonging to types I, S, or T and “Kabuchii”, “Oto”, or “Tarogayo” were conducted (Table 4). The genotypes of “Kabuchii”, “Oto”, and “Tarogayo” were different as they were cross-compatible with each other. All 18 cross combinations showed cross-compatibility (Table 4).
Cross-incompatibility/compatibility of Citrus depressa and C. ryukyuensis and their related species.
Twelve accessions were self-incompatible and ten were self-compatible, in the C. depressa accessions studied. Self-incompatible C. depressa was distributed on Kuroshima, Amami Oshima, Kakeromajima, Tokunoshima, Okinoerabujima, Yoronjima, and Okinawajima. Self-compatible C. depressa was found on Yakushima, Amami Oshima, Tokunoshima, Yoronjima, Okinawajima, Kohamajima, and Yonagunijima. Both self-incompatible and compatible C. depressa were grown on Amami Oshima, Tokunoshima, Yoronjima, and Okinawajima. Self-incompatible accessions were more common in Kagoshima than the Okinawa Prefecture. Moreover, all self-incompatible C. depressa accessions grown on the islands of Kagoshima Prefecture have the same S genotype as they were cross-incompatible with each other. A previous study (Yamamoto et al., 2017) reported the close relation between these accessions. In the dendrogram, all self-incompatible accessions of C. depressa were included in subcluster A1 and distinguished from self-compatible ones belonging to subcluster A2. This suggests that the same accession types are distributed on islands located in Kagoshima Prefecture. “Shiikuwasha (Okitsu)” originally introduced from Okinawajima showed the same S genotype as the above-mentioned self-incompatible accessions. It was cross-compatible with 22 self-incompatible accessions (Yamamoto et al., 2012). These results indicate that these accessions possess a characteristic S genotype.
“Ishikunibu” was also self-incompatible C. depressa. It was cross-compatible with all self-incompatible accessions used in the present study which indicates a characteristic S genotype. DNA analysis revealed that it was distinct from other C. depressa accessions (Yamamoto et al., 2017) yet Zhou et al. (2018) reported the self-compatibility of “Ishikunibu”. The previous and present study’s results are inconsistent, and as local citrus grown on the Ryukyu Islands are occasionally called by the same name (Yamamoto et al., 2022), thus these two “Ishikunibu” are considered different.
Self-compatible C. depressa was widely distributed, from Yakushima to Yonagunijima. These self-compatible accessions were genetically distinct from self-incompatible ones, showing diversity based on DNA analysis (Yamamoto et al., 2017). Zhou et al. (2018) revealed self-compatibility in eight C. depressa accessions. There appears to be various self-compatible accessions in this species. On the other hand, as compatibility is dominant over incompatibility, an S allele was identified not only in self-incompatible accessions but also in self-compatible ones such as satsuma mandarin (C. unshiu Marcow.), grapefruit (C. paradisi Macfad.), and ‘Dancy’ (C. tangerina hort. ex Tanaka) (Soost, 1965, 1969; Vardi et al., 2000). Thus, self-compatible C. depressa accessions may have an S allele. Elucidation of the S allele distribution in C. depressa is an important future issue.
Herein, the self-incompatibility of C. ryukyuensis was identified with the two accessions cross-incompatible. This is the first report of self-incompatibility in this species. As C. ryukyuensis is monoembryonic, the genotypes of these two accessions arose from seedlings different from the original tree on Okinawajima. However, at least one S allele of C. ryukyuensis studied was derived from the original genotype. Hence, the S allele was found distributed in C. ryukyuensis. This species is a pure mandarin without pummelo admixture (Wu et al., 2021). These results suggest that the S allele of C. ryukyuensis is not pummelo-derived but original in this species, which is a new finding in citrus S allele distribution. C. ryukyuensis is reported to be a parental species of C. depressa (Wu et al., 2021). These results suggest that the S allele of C. depressa was derived from C. ryukyuensis.
C. ryukyuensis is reported to show the lowest heterozygosity compared to other citrus species (Wu et al., 2021). This may result from self-hybridization or intraspecific hybridization. Self-compatibility is necessary for the former hybridization. However, if multiple S alleles in C. ryukyuensis are assumed, hybrid seedlings may be obtained by cross-pollination within this species even if self-incompatibility is displayed. Thus, further elucidation of S allele distribution and/or self-compatible genes in this species is necessary.
The Citrus tachibana used in the present study showed self-compatibility. Although three types of C. tachibana are reported (Shimizu et al., 2016), only one accession was used. The self-incompatibility or compatibility of all three types is necessary to ascertain. Ngo et al. (2011) reported the existence of the S allele in self-compatible C. tachibana. As with C. depressa, C. ryukyuensis is considered a parent of C. tachibana (Wu et al., 2021). Therefore, the latter’s S allele may be derived from the former.
Self-incompatible hybrids with a known pedigree are sometimes useful for estimating the S genotype. C. kabuchii, C. oto, and C. tarogayo were considered hybrids of self-incompatible C. nobilis Lour. (Kunenbo) and C. depressa (Yamamoto et al., 2006a, 2021a). They revealed the existence of at least one S allele in the parental C. depressa. When both parents share the same S allele, half of the hybrids are cross-compatible with the pollen parent. This is useful in determining the S allele of given accessions (Yamamoto et al., 2006a, 2010). However, these three species were all cross-compatible with three types of self-incompatible C. depressa and C. ryukyuensis. Thus, the S allele determination for these species was impossible.
In conclusion, the self and cross-incompatibility of C. depressa and C. ryukyuensis was successfully demonstrated. The S alleles of both species are considered to have originated from themselves, not from the pummelo. Citrus maxima is a major source of the S allele (Kim et al., 2010, 2011, 2020; Ngo et al., 2019; Yamamoto et al., 2006a, 2012; Zhou et al., 2018) and, in the present study, C. depressa and/or C. ryukyuensis were revealed to also be S allele sources. These findings contribute to studies on citrus incompatibility and genetic resources. Further studies elucidating self-incompatibility and distribution of the S allele using more accessions of C. depressa, C. ryukyuensis, C. tachibana, and their related species are necessary. Recently, S allele determination in citrus was achieved by S-RNase analysis (Honsho et al., 2021; Liang et al., 2020). This method is a powerful tool to detect the S allele. Thus, this method will contribute to the elucidation of S allele distribution in citrus.
The author would like to thank H. Fukudome, J. Hirose, Y. Nishizawa, and S. Kawaguchi for managing the citrus orchard.
