Research Article: The X chromosome is necessary for ovule production in Silene latifolia

Date Published: May 23, 2019

Publisher: Public Library of Science

Author(s): Paris Veltsos, Lynda F. Delph, Randall P. Niedz.


Sex chromosomes stop recombining and accumulate differences over time. In particular, genes on the chromosome restricted to the heterogametic sex degenerate and become non-functional. Here, we investigated whether or not the degeneration of a plant Y chromosome was sufficient to cause ovules containing a Y to fail to develop, thereby eliminating the possibility of YY individuals. We used two genotypic assays to determine the genotype—XX, XY, or YY—of offspring from a single fruit of an otherwise normal male XY plant of Silene latifolia. The fruit contained fewer ovules than normal pistillate flowers, produced an equal offspring sex ratio, and generated no YY offspring. The results indicate that ovaries must contain an X chromosome to develop properly. While haploid selection has slowed down Y-chromosome degeneration in S. latifolia, we find that it has progressed sufficiently to prevent the proper development of ovules, and hence prevent the presence of YY individuals.

Partial Text

Sex chromosomes evolve differently than the rest of the genome, because they are present in one sex more often than the other. The first plant sex chromosomes were identified in Silene latifolia almost a century ago [1,2], and have influenced early theoretical work on sex chromosomes. In fact, the canonical model of sex-chromosome evolution is based on plants, and describes the evolution of separate sexes from a hermaphrodite ancestor [3–5]. The model assumes the spread of a male-sterility mutation, resulting in a population of hermaphrodites and females. After the frequency of females increases, a second mutation—for female sterility—will spread in the population, as long as the resulting males can outcompete the existing hermaphrodites in male function, which is facilitated by existing inbreeding depression [3]. The presence of two separate sterility mutations, one for each sex, selects for a restriction of recombination between them, so that only male- and female-producing haplotypes occur, as they are the most fit possible haplotypes (the other possible haplotypes resulting in hermaphrodite, or completely sterile plants). Recombination is expected to gradually stop over more of the length of those proto-sex chromosomes, as they capture male- and female-beneficial mutations. This is because linkage with the sex-determining genes is advantageous for alleles with sexually antagonistic effects [6,7], i.e. alleles conferring a benefit to the fitness of one sex, while being deleterious for the other.

The plant used in this study came from a cross between individuals from a population in Virginia (37°21’36”N, 80°40’54”W, 500 m above sea level). The study was carried out on private land, the owner gave permission to conduct the study on this site and the study did not involve endangered or protected species. The natural field population had a sex ratio of 54.5% females (697 females, 582 males) in 2018. The seed was germinated in a greenhouse at Indiana University, and following a 2-week growth period in the greenhouse, the resulting seedling was transplanted into a 2L pot that was placed outside in Bloomington, IN for a 3-month period (from 30 May to 30 Aug 2018). The focal plant produced 1150 normal staminate (pollen-producing) flowers during its 85 days of flowering, and 1 pistillate (ovule-producing) flower that developed into a fruit. This individual was otherwise indistinguishable from a normal male. For example, the mean calyx width (a very sexually dimorphic character) of other males from this population used in this planting was 7.33 (± 0.14 SE) and the mean calyx width for this plant was 7.85. In contrast, females from this planting had calyces that averaged 9.35 (± 0.19 SE). The focal plant was surrounded by other S. latifolia plants, allowing the flower that produced the fruit to be open pollinated. Hence, all offspring from the fruit are expected to be the product of a XY x XY cross, which would produce a genotype ratio of 1 XX: 2 XY: 1YY if Y ovules were produced and YY individuals were viable. The chisquare tests were conducted in R v 3.5.3 [22] using the chisq function.

To test whether an X chromosome is required to produce viable seed in the dioecious species S. latifolia, which has heteromorphic sex chromosomes, we took advantage of seeds produced from a single ovule-bearing flower from a plant that was otherwise a normal XY male individual. The complete absence of YY offspring from the fruit that developed from this flower suggests that the X contains genes not present on the Y that are necessary for megasporophyte development, such that no ovules with a Y develop. The lack of development of ovules containing a Y is further supported by the fact that the XY fruit contained abnormally few ovules.