Tree Genetics & Genomes

, Volume 8, Issue 6, pp 1371–1377

Genome-wide detection of genetic loci triggering uneven descending of gametes from a natural hybrid pine

Authors

  • Shuxian Li
    • Key Laboratory of Forest Genetics and BiotechnologyNanjing Forestry University
  • Zaixiang Tang
    • Department of Epidemiology and Biostatistics, School of Public HealthMedical College of Soochow University
  • Defang Zhang
    • Key Laboratory of Forest Genetics and BiotechnologyNanjing Forestry University
  • Ning Ye
    • Key Laboratory of Forest Genetics and BiotechnologyNanjing Forestry University
  • Chenwu Xu
    • Jiangsu Provincial Key Laboratory of Crop Genetics and Physiology; Key Laboratory of Plant Functional Genomics of Ministry of EducationYangzhou University
    • Key Laboratory of Forest Genetics and BiotechnologyNanjing Forestry University
Original Paper

DOI: 10.1007/s11295-012-0524-5

Cite this article as:
Li, S., Tang, Z., Zhang, D. et al. Tree Genetics & Genomes (2012) 8: 1371. doi:10.1007/s11295-012-0524-5
  • 165 Views

Abstract

Marker transmission ratio distortion (TRD) revealed in genetic mapping studies on distant crosses can be used to infer the genetic basis relating to reproductive barriers between species. Unlike measuring the degree of TRD by the overall number of segregation distorted markers in the affected genome regions, mapping the segregation distorting loci (SDL) provides reliable statistic parameters that help to confine the target genomic regions for further characterization at molecular level. Using the linkage map constructed for a natural hybrid of Pinus hwangshanensis and Pinus massoniana, we perform SDL analyses and align the established map to the loblolly pine consensus map. Altogether, six SDLs with relatively strong LOD supports are detected on four linkage groups of the established map. Since gametes inheriting different alternate chromatid blocks from the SDL affecting genome regions have uneven chance to descend to the offspring, the corresponding genome regions are supposed to play more significant roles in rendering the reproductive isolations between P. hwangshanensis and P. massoniana. This paper presents a case study on a crucial step for uncovering the hidden genetic factors that trigger the uneven descending of gametes in a natural hybrid pine.

Keywords

Transmission ratio distortionSegregation distortion lociReproductive isolationHybrid pine

Introduction

Understanding the genetic mechanism triggering reproductive isolations between closely related species is one of the most interesting topics for evolutionary biologists. To date, the underlying causes of reproductive isolation are only partially understood in a few species (Greig 2009). Hybridization between different species often has negative consequences that result in reduced fertility or viability (Hall and Willis 2005; Moyle and Graham 2006), leading to transmission ratio distortion (TRD) of gametes from the parental species. Genetic mapping method provides a unique opportunity to study the interactions of differentiated genes and genomes in a hybrid genetic background (Rieseberg et al. 2000). TRD marker clusters are rampant in contemporary of many mapping studies across a diverse range of organisms including wildflowers (e.g., Hall and Willis 2005), insects (e.g., Wang and Porter 2004), marine invertebrates (Li and Guo 2004), fish (e.g., Kocher et al. 1998), trees (e.g., Myburg et al. 2004), and crop plants (e.g., Jenczewski et al. 1997). In distant crosses, patterns of TRD alone have been used to infer the genetic basis of hybrid incompatibility among species (e.g., Li et al. 2010; Harushima et al. 2001; Myburg et al. 2004). However, the majority evidences of TRD occurring in hybrid genetic backgrounds are based on studies employing experimental hybrids, map-based studies of natural hybrids are much rarer (Rieseberg et al. 2000).

Pinus hwangshanensis and Pinus massoniana are native conifers in southeast China. The distribution ranges of these two species are frequently found to be adjacent to each other and overlap with a narrow contact zone. P. hwangshanensis normally distributes above 700 m, while P. massoniana distributes at lower elevations. The contact zone roughly spans a vertical range from 700 to 900 m (Wu 1980). These two species are different in morphological, cytological, and timber anatomical characteristics and show clear environmental and spatial separation (Wu 1980; Xing et al. 1992). In the contact zone, spontaneous hybrids are frequently observed, whose morphological characteristics are intermediate that of P. hwangshanensis and P. massoniana (Xing et al. 1992; Luo and Zou 2001). Previous phylogenetic studies of Pinus suggest that these two species are closely related, being placed in the same subgenus, section, and subsection (Wang et al. 1999; Gernandt et al. 2005). Within their distribution ranges, no other pine species from the same subgenus are present, suggesting that present-day hybridization with other pine species does not occur (Zhou et al. 2010). The overlapping distributions and inter-fertility lend these two species as an ideal system to study reproductive isolations between closely related species (Xing et al. 1992; Luo and Zou 2001; Li et al. 2010; Zhou et al. 2010).

In a previous study, we genotyped 192 megagametophytes of seeds from a natural hybrid of these two species and built a genetic map for this mapping individual. Intensive TRD clusters were revealed to distribute along some of the linkage groups (Li et al. 2010). There are many other studies showing that the TRD clusters non-randomly distribute along linkage maps of many different organisms (e.g., Li et al. 2010; Harushima et al. 2001; Myburg et al. 2004; Fishman et al. 2001; Hall and Willis 2005; Moyle and Graham 2006; and references therein), suggesting segregation distorting loci (SDLs) underlying this pattern. SDL refers to chromosomal regions that cause distorted segregation ratios in early life stages (Vogl and Xu 2000). These distortions are caused either by differential representation of SDL genotypes in gametes before fertilization or by viability differences of SDL genotypes after fertilization. In both cases, the observable phenotype is a distortion of marker locus in chromosomal regions close to the SDL (Vogl and Xu 2000). SDLs are hidden but carry an important function in evolution (Xu 2008). In genetic mapping studies, the change of marker allelic frequencies is reflected by the departure from the Mendelian expectation. The non-Mendelian segregation of markers can be used to map SDL along the genome. Based on the genotypic frequencies of observed markers, Luo et al. (2005) proposed a novel liability model for mapping SDL by assuming a continuous liability controlling the viability of individuals. Mapping SDL can then be formulated as a problem of general QTL mapping, only using the information of allelic frequencies of markers along the linkage groups (Vogl and Xu 2000; Carr and Dudash 2003; Luo and Xu 2003; Luo et al. 2005; Zhu and Zhang 2007). Mapping SDLs is an essential step to confine the target genomic regions for further characterization at molecular level. However, few such studies have been carried out with real-life data due to the lack of user friendly software. Recently, the SDL mapping module was integrated in PROC QTL software, making it friendly to use for the common audiences in versatile disciplines (Hu and Xu 2009).

For molecular characterization of the genetic causes of distorted segregation ratios, mapping the underlying SDLs is desirable (Vogl and Xu 2000). Meanwhile, the previous genetic map is constructed with anonymous amplified fragment length polymorphism (AFLP) markers (Li et al. 2010); connecting this map to the pine consensus map is essential for future characterization of the target genome regions. Thus, the perspectives of this paper are to (1) align the established genetic map to the pine consensus map with sequence tagging markers and (2) map the SDLs triggering the TRD marker clusters uncovered in the genome of the mapping hybrid.

Methods

Alignment of the linkage groups

Alignment of the established linkage groups to the pine consensus map was conducted by adopting the expressed sequence tag (EST) reported in the paper of Temesgen et al. (2001). In that study, the authors developed a large number of EST markers from the cDNA libraries constructed from mRNAs expressed in the needles of loblolly pine seedlings or in the xylem from young trees and succeeded in positioning 56 EST markers on a loblolly pine linkage map (Sewell et al. 1999). In this study, all the 56 EST markers mapped on the loblolly pine consensus map (Temesgen et al. 2001) were selected to test segregation in our mapping population (Supplemental Table 1). The corresponding primer pairs were synthesized by Jerry Bio-Technology LTD, Shanghai, China. DNAs were residues from the previous study (Li et al. 2010). Since only limited quantity of DNA was extracted from megagametophyte tissue of each seed, 48 DNAs in relatively high quantity were reserved for population genotyping and the other DNAs were used for segregation testing. During the screening process, each primer pair was amplified with eight DNAs. The primer pairs generated segregating loci were further genotyped with the 48 reserved DNAs to build a new map for this mapping pedigree. Segregation of EST marker was examined using single-stranded DNA conformation polymorphism (SSCP) on the polymerase chain reaction (PCR) product of each primer pair. PCR and SSCP protocols were as described in Yin et al. (2004a, b). The new EST markers were integrated into the previous established map using JoinMap 3.0 (Van Ooijen and Voorrips 2001), and map graph was drawn by MapChart 2.2 (Voorrips 2006). The nomenclature for EST markers and linkage group identities followed the designations as that reported in the paper of Temesgen et al. (2001).

SDLs detection

Genotyping data used for SDL mapping were described in Li et al. (2010). This data set contained 321 segregating AFLP markers that were genotyped with 192 megagametophytes, which represented 192 meiosis events of the mapping individual. The established map consisted of 14 major linkage groups with 311 AFLP markers. The total map length was 1,615.6 cM. PROC QTL Version 2.0 was employed to perform SDL detection (Hu and Xu 2009). The following codes were used to invoke PROC QTL to detect the underlying SDL with a step length of 1 cM along each linkage group:
  • proc qtl data=tree map=map out=tree.result

  • method='ml'/distortion step=1.0;

  • model trt=;

  • matingtype 'BC';

  • genotype A1A1='A' A1A2='H' ;

  • estimate 'A'=-1 1;

  • run;

For descriptions of the above statements, see PROC QTL Version 2.0 manual (http://statgen.ucr.edu/download/software). In the output result table, threshold to call a significant SDL was set at LOD score equaling to 2.5.

Results

Map alignment

The selected primer pairs (Supplemental Table 1) showed highly successful rate in PCR amplification across taxa of pine. Among the 56 primer pairs, only 5 of them failed in amplification. Thus, the amplification successful rate of these primer pairs was about 91 %. The subsequent segregation test revealed that 15 of the successful primer pairs generated segregating loci. Since megagametophytes are haploid tissues, normal Mendelian segregation loci are expected to segregated following 1:1 ratio. With genotypes of eight megagametophytes, the probability to detect polymorphism for any normal Mendelian segregating loci is \( \left[ {1 - {{\left( {\frac{1}{2}} \right)}^8}} \right] \times 100\% = 99.6\% \). The previous map was estimated to cover about 95.83~98.62 % of the pine genome at a 10-cM marker (Li et al. 2010). In this study, all the 15 segregating markers were mapped into different linkage groups. This result validated the good coverage of the previous genetic map. Out of the 14 linkage groups, 9 of them were aligned to the loblolly pine consensus map (Temesgen et al. 2001). Subsequently, the corresponding linkage groups were assigned with the same identities as those of the loblolly pine consensus map (see Fig. 3 in the paper of Temesgen et al. 2001). They were named as chrom consensus-01, chrom consensus-02, chrom consensus-04, chrom consensus-05, chrom consensus-06, chrom consensus-07, chrom consensus-08, chrom consensus-10, and chrom consensus-16 based on the shared markers which were mapped both on our map and on Temesgen et al. (2001). Nomenclatures of the remaining five linkage groups without integrated EST markers were LG-03, LG-09, LG-11, LG-12, and LG-13, and these linkage groups were not aligned to the loblolly pine consensus map. The linkage group identities were displayed in Supplemental Fig. 1. One of the linkage group, chrom consensus-06, contained multiple EST markers, and these EST markers were in complete colinearity on our map and on the loblolly pine consensus map. With these mapping efforts, two out of the four linkage groups with intensive TRD clusters were aligned to the loblolly pine consensus map. They were chrom consensus-02 and chrom consensus-04. The other two linkage groups containing intensive TRD clusters, LG-03 and LG-12, remained as pedigree specific (Supplemental Fig. 1).

SDLs detection

Since the newly mapped EST markers were only genotyped with a subset of 48 megagametophytes, these markers were excluded in the SDL detection analyses. The actual population size for SDL detection was 192. Altogether, six SDLs were detected among four of the linkage groups (Table 1). The strongest SDL was detected on chrom consensus-02, with a LOD support of 6.02. This SDL caused a maximum TRD with a chi-square value equaling to 26.39 and rendered the TRDs of a genome region spanning 59.2 cM (block I, see Table 1). This region encompassed 11 AFLPs with significant TRDs. Since AFLPs are dominant markers, for a particular locus, a gamete either inherits the visible allele or the invisible allele. Linkage phase examination demonstrated that the visible alleles of all AFLP markers in this region were in coupling phase. In order to distinguish the counterpart of the homologous haploid block in this region, we named the chromatid block that mainly consisted of the visible alleles as the “+” haploid block, and its counterpart on the alternate chromatid that mainly consisted of the invisible alleles as the “−” haploid block. Distortion direction showed that gametes inheriting the “+” haploid block I (see Table 1) were dominant in the offspring. Another SDL was detected on chrom consensus-04, rendering the TRDs of a genome region encompassing 35.9 cM. This genome region contained nine AFLPs, and the visible alleles of eight AFLPs in this region were in coupling phase. In contrast to haploid block I, distortion direction revealed that gametes inheriting the “−” haploid block II (see Table 1) were predominant in the progeny. It was notable that two SDLs on LG-12 jointly caused the TRDs in a continuous genome region spanning 46.1 cM, and gametes inheriting the through “+” haploid block III (see Table 1) were over abundant in the offspring. There were also two SDLs detected on LG-03. However, distortion direction caused by the two SDLs was opposite to each other. Gametes simultaneously inheriting the “+” haploid block IV in the proximal end and the “−” haploid block V in the distal end of this linkage group (see Table 1) were predominant in the viable seeds. The plotting charts of the significant SDLs detected along linkage groups were shown in Fig. 1.
Table 1

SDLs detected along the established linkage groups

SDL

Linkage group

Position (cM)

LOD score

2-LOD confidence interval (cM)

Maximum chi-square value

Effect span (cM)

Affected block

Dominant haploid block

1

consensus-02

159.6

6.02

[151.65, 169.8]

26.39

59.2

I

“+”

2

consensus-04

71.7

4.14

[66.7, 77.6]

17.89

35.9

II

“–”

3

LG-03

73.5

4.66

[62.4, 79.1]

20.89

14.9

IV

“+”

4

LG-03

129.6

3.63

[121.9, 129.6]

16.51

21.3

V

“–”

5

LG-12

2.8

4.27

[0, 19]

17.71

46.1

III

“–”

6

LG-12

46.1

5.02

[43.3, 49.1]

22.35

   

“+” represents the haploid block that mainly consists of visible alleles of TRD markers mapped in the affected genome region, and “–” indicates the homologous counterpart in opposition to the “+” haploid block. 2-LOD confidence interval indicates the left and right boundaries that cover 2 LOD decreases from that of the highest peak. Effect span indicates the genetic length encompassed by the corresponding TRD marker cluster

https://static-content.springer.com/image/art%3A10.1007%2Fs11295-012-0524-5/MediaObjects/11295_2012_524_Fig1_HTML.gif
Fig. 1

Plotting charts of the significant SDLs detected along linkage groups. Note: the Y-axis indicates the LOD scores for the corresponding genetic positions on x-axis. LOD scores are calculated at a step length of 1 cM along each linkage group

Discussion

Pine megagametophyte is a maternally derived haploid tissue, which houses the majority of the storage reserves of seed. In the sexual reproduction process of conifer, a diploid precursor cell of the nucellus enlarges dramatically and undergoes meiosis to produce four megaspores. Three of them die, and the fourth enters a protracted phase of mitotic divisions and enlargement to give rise to the megagametophyte and the female egg. Thus, megagametophyte is a haploid tissue and has the same DNA as the female egg that is fertilized by pollen to form embryo in each seed. As a result, marker distortion revealed by megagametophyte genotyping reflects the uneven transmission of genetic materials to the offspring from the maternal tree (Li et al. 2010).

Alleles at a locus that deviate from the expected Mendelian segregation ratio is a common phenomenon that has been described in many different organisms (e.g., Alheit et al. 2011; Hall and Willis 2005; Kulmunia et al. 2010; Moyle and Graham 2006; Myburg et al. 2004; Rieseberg et al. 2000; and references therein). Several potential causes may trigger these departures from Mendelian expectations. Sample size and genotyping errors are some non-biological factors that can contribute to segregation distortion (Alheit et al. 2011), leading to random distribution of TRD markers along genetic maps. Biologically, active processes that influence the segregation of alleles immediately prior to, during, and immediately following meiosis can produce distorted ratios at the affected loci and other loci linked to it (Moyle and Graham 2006). Besides meiotic drive, zygotic selection against dysfunctional allelic combinations, chromosomal rearrangements, and genic interactions are also some important reasons which cause the biological deviations from Mendelian expectations (Jiang et al. 2000; Moyle and Graham 2006; Rieseberg et al. 1995). In contrast to the non-biological factors, the biological processes will always affect a cluster of markers within the chromosomal region surrounding the underlying genetic loci. Thus TRD markers are most commonly observed to occur in clusters on linkage maps of different organisms (Myburg et al. 2004; Fishman et al. 2001; Hall and Willis 2005; Moyle and Graham 2006; and references therein). Since TRD clusters are caused by biological factors, mapping the underlying SDLs is a crucial step for uncovering the hidden genetic factors. Our analyses detected six SDLs triggering the uneven descending of gametes to the offspring from a natural hybrid of P. hwangshanensis and P. massoniana. The affected genome regions were chromatid blocks I, II, III, IV, and V (Table 1). Further examination revealed that a super gamete was supposed to simultaneously inherit the “+” haploid blocks I and IV, and the “−” haploid blocks II, III, and V. All the SDLs detected in this study were with strong LOD supports statistically, suggesting major genes underlying marker TRDs observed in this hybrid pine.

Empirical studies have provided evidence for fewer distorted markers in intraspecific crosses relative to interspecific crosses in agricultural plants (Zamir and Tadmore 1986; Causse et al. 1994; Jenczewski et al. 1997), suggesting a positive correlation between the degree of TRD and the level of genomic divergence (Harushima et al. 2001; Hall and Willis 2005; Taylor and Ingvarsson 2003). Hybrid incompatibility is known to be frequently due to disrupted genetic interactions among loci that have diverged in distant parental lineages (Orr and Turelli 2001). When hybrids between distant lineages are created, these negative interactions can cause inviability and/or sterility in particular recombinant genotypes, such that they are removed by natural selection from hybrid populations (Moyle and Graham 2006). This non-random elimination of specific allelic combinations can be used to explain the non-random distribution of TRD clusters observed on linkage maps of different mapping studies.

Mapping QTLs governing the morphological traits that directly associated with reproductive isolation is ideal for uncovering the specific genetic factors. However, plant species are typically isolated by a large number of different pre- and post-zygotic barriers, and their potentially complex interactions. Phenotypic data are normally unavailable for those hidden reproductive barriers. SDL mapping is independent from the phenotypic data; thus, it supplies a useful analytical tool for evolutionary biologist who is interested in mapping genetic loci underlying fitness and hybridization incompatibility. Moyle and Graham (2006) reported a case study on genome-wide associations between hybrid sterility QTL and TRD marker clusters, and they found evidence that TRD regions were chromosomally collocated with hybrid incompatibility loci more frequently than was expected by chance alone. Since SDLs alone are not directly associated with particular morphological traits, combining the QTL and SDL approaches will give us a better opportunity to evaluate and understand the genetic basis and complexity of loci governing reproductive barriers between closely related species.

Base on SDL analysis, even greater insight can be obtained in species with complete genome sequences because gene identity, gene number, and other features of interest can be assessed for the target genomic regions (Payseur and Nachman 2005). With the progress in life technology, it is feasible to detect SNPs context by resequencing a genome region of interest (Albert et al. 2007). Using these SNPs, we can assess the patterns of gene flow between P. hwangshanensis and P. massoniana. The combined tool package will enable us to resolve the SDLs into particular genetic factors. Like P. hwangshanensis and P. massoniana, a similar plant system was observed between Fremont (Populus fremontii) and narrowleaf (Populus angustifolia) cottonwood (Martinsen et al. 2001). The two species can naturally hybridize with each other. Meanwhile, their distributions overlap with a narrow hybrid zone, with narrowleaf cottonwood at high elevations whereas Fremont cottonwood at low elevations (Martinsen et al. 2001). Although the above analytical toolkit is not mature to be applied in pines due to the unavailability of genome sequences, it is feasible to be applied in the plants system like Fremont and narrowleaf cottonwood since the poplar genome has been completely sequenced (Tuskan et al. 2006). Nevertheless, SDL detection is a crucial link in the analytical toolkit for uncovering the hidden genetic factors underlying reproductive isolation between species.

Conclusion

In this study, we performed SDL detection using the linkage map constructed for a natural hybrid of P. hwangshanensis and P. massoniana and detected six SDLs triggering the uneven descending of gametes from the mapping hybrid. Since the affected genome regions are supposed to play more significant roles in rendering the reproductive isolations for these two species, this study has dedicated a clear picture for displaying the genome regions and for further uncovering the hidden genetic factors that act as reproductive barriers between P. hwangshanensis and P. massoniana. As a conclusion, the analytical tool demonstrated in this study is useful to explore the genetic mechanisms underlying reproductive isolation between closely related species, and it may draw broad interest of the evolutionary biologists.

Acknowledgments

We thank Dr J. Armento in Oak Ridge, Tennessee, USA for his comments and editing for this manuscript, and special thanks go to the editor and anonymous reviewers for their help in formulating the revisions. Funding for this work was provided by 948 project (2012-4-41), 973 project (2012CB114505), and the Natural Science Foundation of China (31125008, 31070543). It is also funded by the Doctorate Fellowship Foundation and PAPD program at Nanjing Forestry University.

Supplementary material

11295_2012_524_MOESM1_ESM.doc (109 kb)
Supplemental Table 1Pine EST primer pairs adopted from a paper of Temesgen et al. (2001) for map alignment in this study (DOC 109 kb)
11295_2012_524_MOESM2_ESM.ppt (255 kb)
Supplemental Fig. 1The linkage map integrated with EST markers for a natural hybrid of P. massoniana and P. hwangshanensis. In this figure, the linkage group identity is listed on top of each linkage group. This map is aligned to the pine consensus map by integrating EST markers selected from the loblolly pine consensus map (Temesgen et al. 2001). The linkage group identities started with “chrom” are named after the designations as that reported in the paper of Temesgen et al. (2001), whereas linkage group identities started with “LG” represent groups that were not aligned to the loblolly pine consensus map. EST markers are in italic bold fonts. Marker with name ending with an “r” is in repulsion linkage phase. “+” and “++” following a marker name show that segregation distortion skews to more visible allele of the corresponding marker at significance level of p ≤ 0.05 and p ≤ 0.01. Whereas “−” and “−−” show that segregation distortion skews to more invisible allele of the corresponding marker at significance level of p ≤ 0.05 and p ≤ 0.01 (PPT 255 kb)

Copyright information

© Springer-Verlag 2012