Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Agricola
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Related Collections
Right arrow Cotton
Right arrow Cell Biology & Molecular Genetics
Right arrow Crop Genetics
Crop Science 41:289-294 (2001)
© 2001 Crop Science Society of America

CROP BREEDING, GENETICS & CYTOLOGY

Inheritance and Genetic Relationships of the D8 and D2-2 Restorer Genes for Cotton Cytoplasmic Male Sterility

J.F. Zhang and James McD. Stewart

Dep. of Crop, Soil & Environmental Sciences, Univ. of Arkansas, Fayetteville, AR 72701

Corresponding author (jstewart{at}uark.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
In cotton, cytoplasmic male sterility conditioned by the D8 alloplasm (CMS-D8) is independently restored to fertility by its specific D8 restorer (D8R) and by the D2 restorer (D2R) that was developed for the D2 cytoplasmic male sterile alloplasm (CMS-D2). However, knowledge of the relationship between the two restorer genes is lacking. The objectives of these studies were to determine the inheritance and genetic relationship between D2R and D8R. A series of intercrosses and test crosses were used to determine the inheritance and relationships. CMS-D2 lines crossed with D8R resulted in all sterile F1 plants, confirming that the D8 restorer gene does not suppress D2 cytoplasmic dysfunction. A segregation ratio of one fertile to one sterile in three testcross combinations revealed that D2R contained one restorer gene, Rf, for both CMS-D8 and CMS-D2. The Rf gene functioned sporophytically, whereas the D8R gene is known to function gametophytically. However, Rf pollen with the D8 or D2 cytoplasm was preferentially effective in fertilization based on the low number of sterile plants obtained when alloplasmic F1 plants were self-pollinated or used as the pollen-parent in crosses with CMS-D8. In tests for allelism only 8 sterile plants out of 1921 were obtained in two testcrosses of (D8R x D2R)F1 x normal cotton. Thus, the two restorer loci are not allelic but are tightly linked with an average genetic distance of 0.93 cM. The D2 gene Rf is redesignated Rf1, and Rf2 is assigned to the D8 restorer gene.

Abbreviations: CMS, cytoplasmic male sterility • RFLP, restriction fragment length polymorphism


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
CYTOPLASMIC MALE STERILITY (CMS) occurs in at least 150 species in higher plants (Kaul, 1988). Different CMS systems have been identified based on the restorer genes required to correct the CMS condition. Most CMS systems require only one dominant restorer gene to restore fertility, although some CMS systems require either two independent or complementary dominant genes for restoration (Laughnan and Gabay-Laughnan, 1983; Tang et al., 1998; Zhang et al., 1997). Other studies demonstrated that multiple linked or unlinked restorer genes can give similar but distinct effects in the restoration of the same CMS system. For example, in common bean (Phaseolus vulgaris L.), different sources of restorers were identified and two non-allelic, but tightly linked genes, Fr and Fr2, were genetically and molecularly characterized (Jia et al., 1997). Fr effects the loss of mitochondrial sterility-associated sequences, pvs, resulting in a permanent recovery of fertility (Mackenzie, 1990). Fr2, in contrast, restores fertility by suppressing expression of pvs orf239 post-transcriptionally (Sarria et al., 1998). In maize (Zea mays L.), two non-allelic restorer genes, Rf8 and Rf* mimic Rf1 in partially restoring fertility to CMS-T in an Rf2/Rf2 background (Dill et al., 1997). Expression of the unique T-urf13 mitochondrial coding sequence was changed to restore pollen fertility to T-cytoplasm maize. In rape (Brassica napus L.), Fang and McVetty (1989) reported that two non-allelic dominant genes, Rfp1 and Rfp2, derived from two different sources, were independently capable of restoring CMS-pol. However, molecular mapping showed that Rfp1 and Rfp2 were likely allelic, as both are perfectly linked to the same restriction fragment length polymorphism (RFLP) marker on linkage group 18 (Jean et al., 1997). More interestingly, the restorer gene, Rfp, was allelic to Rfn, the restorer gene for CMS-nap, as both were mapped to the same chromosomal position (Li et al., 1998).

In cotton, several different sources of CMS have been reported including CMS-D2-2 (Meyer, 1975), CMS-hir (Jia, 1990), CMS-D8 (Stewart, 1992), CMS-D4 (Meshram et al., 1994), and CMS-C1 (Zhang and Stewart, 1999). Fertility can be restored to CMS-D2-2 by the D2 restorer in which the restorer factor(s) was introduced from the genome of G. harknessii Brandegee (D2-2). In her preliminary study, Meyer (1975) first suggested two restorer genes were responsible for fertility restoration. Others (da Silva et al., 1981; Maranhao et al., 1984) reported three restorer genes, and some modifying genes based on monosomic analysis. One of the restorer genes was located on chromosome 18. But, genetic data from Kohel et al. (1984) and Weaver and Weaver (1977) indicated that only one restorer gene, Rf, was involved in the restoration of CMS-D2-2. No linkage between Rf and 13 morphological markers (including v1 on chromosome 18) was identified. The CMS-D8 and D8 restorer system was developed by introducing G. trilobum (DC.) Skovst. (D8) cytoplasm and nuclear restorer factor into cotton (Stewart, 1992). The male sterile plants have smaller flowers and anthers without pollen, similar to CMS-D2-2. Microsporogenesis in both CMS systems aborts during the premeiotic stage (Black and Stewart, 1995; Black, 1997). One dominant restorer gene from the D8 restorer was identified to restore fertility of CMS-D8 (Stewart, 1995; Stewart and Zhang, 1996; Stewart et al., 1996; Zhang and Stewart, 2001). Interestingly, the D2 restorer for CMS-D2-2 also was reported to restore the fertility of CMS-D8 (Stewart, 1995), CMS-hir (Jia, 1990), and CMS-C1 (Zhang and Stewart, 1999). Wang et al. (1996) showed that the restoration of CMS-hir (derived from interspecific hybridization between upland cotton and G. barbadense L.) by the D2 restorer is controlled by a single restorer gene. Similar results were obtained for the restoration of CMS-C1 by the D2 restorer (Zhang and Stewart, 1999).

The objective of the present study was to determine the inheritance of restoration to fertility of CMS-D8 by the D2 restorer and the relationship between the D2 and D8 genes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
All plants were grown in Fayetteville, AR, under greenhouse or field conditions. The greenhouse was maintained as nearly as possible to 30°C maximum day and 20°C minimum night with natural lighting. Plants were grown in a commercial potting mix in 15-cm diameter pots that were watered daily as needed and fertilized with Peters 10-30-20 or 15-16-17 water-soluble fertilizer (Scotts-Sierra Horticultural Products, Marysville, OH) once a week. Different field plots were used each year, but in each the soil was Captina silt loam (fine-silty, siliceous, active, mesic Typic Fragiudult). Seeds were usually germinated in the greenhouse in 2- by 2- by 4-inch peat pots and transplanted to the field when the plants were approximately 15 cm tall. Row spacing was approximately 1 m and interplant distance approximately 25 cm. The plots were managed according to standard recommended practices.

D2-2 restorer lines (D2R), B411R, B416R, and B418R (Cook and Namken, 1995), were grown in 1995, 1996, and 1997. The D2R was used as a pollen source to cross with CMS-D8-8518, a D8 cytoplasmic male sterile line of cotton with ARK8518 nuclear background, an elite upland cotton genotype known not to have a restorer factor (Zhang and Stewart, 2001). The resulting F1 plants were grown in the greenhouse in the winter of 1995 to 1996 to generate F2 seed and the first backcross generation (BC1) onto CMS-D8-8518. The F1 of CMS-D8-8518/D2R was also pollinated with ARK8518. Fertility segregation in the two crosses was scored during the 1996 and 1997 growing seasons. In 1996, F1 field plants from a cross between TM-1 (genetic standard of G. hirsutum) as female and D2R were crossed as the male parent with CMS-D8-8518. Fertility was evaluated in the field in 1997. The differences between the sterile and fertile flowers were evident in that the former had smaller flowering buds and flowers with no pollen produced. About 2 wk after the appearance of flowering buds, the sterile plants could be scored without ambiguity by observing the flower bud size and shape and staminal column development, and by squeezing the anthers to observe presence or absence of developing pollen.

For comparison purposes, the D2R with D2-2 cytoplasm was used as the female parent to cross with ARK8518 in 1997, and the resulting F1 plants were selfed or backcrossed as the female parent with ARK8518. The F2 and BC1F1 populations were grown and evaluated for male fertility in the 1999 growing season.

To test whether the D8 restorer could restore fertility to CMS-D2-2, two male sterile lines, CMS-D2-2-57-4 and CMS-D2-2-81-4442 (both with G. barbadense nuclear background), were pollinated by the D8 restorer lines (D8R) in 1996. Fertility of the F1 plants was assessed in the field at Fayetteville in 1997.

To conduct the allelic test between the two restorers, the F1 between D8R as female and D2R, made in the field in 1995 and grown in the winter greenhouse, was selfed to generate F2 seed. The F1 plants were either used as the male parent to cross with CMS-D8 or as the female parent to cross with TM-1 or ARK8518. The F2 and the testcrosses were grown in 1996, 1997, and 1998. Most of the plants from (D8R x D2R)F1 x TM-1 were grown in the winter of 1997 in the greenhouse to evaluate for fertility. The plants not previously evaluated were transplanted to the field in 1998 for fertility scoring. Since only six plants for the testcross (D8R x D2R)F1 x ARK8518 from the above source were evaluated, D8R was again crossed as the female parent to D2R in 1997 and individual plant lineages subsequently maintained. The resulting F1 plants were grown in 1998 and testcrossed as the female parent with ARK8518. Male fertility was evaluated in the 1999 growing season. The recombination fraction r between the two restorer genes was calculated by the formula,

In 1997, the fertile heterozygous plants from (D8R x TM-1)F1 x TM-1 were crossed as the female parent with the fertile heterozygous plants from the crosses (CMS-D8 x D2R)F1 x ARK8518 and CMS-D8 x (TM-1 x D2R)F1. Male fertility was scored during the 1998 growing season. In the inheritance and allelic tests, chi-square was used to test if fertility segregation agreed with the expected ratios.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Inheritance
The F1 plants of CMS-D8 x D2R were all fertile and had normal-sized flowers and pollen shed, indicating that restoration of CMS-D8 by D2R was complete. Very few sterile plants occurred in the F2 generation of CMS-D8 x D2R (Table 1). In 1996, only 9 sterile plants segregated among 152 F2 plants. In 1997, 18 sterile plants were obtained from 365 F2 plants. The F2 data indicated that more than one dominant restorer gene was involved in fertility restoration. The chi-square test suggested two dominant duplicate genes in the restoration of D2R to CMS-D8. Testcrosses of the F1 with CMS-D8 differed among tests (Table 1). If there were two restorer genes, the testcross would be expected to have a three fertile to one sterile segregation ratio. In 1996, a ratio of 15:1 was obtained, while in 1997 the ratio was 7:1, corresponding to four- and three-gene-locus models, respectively. Thus, two, three, and four restorer genes were predicted from the results when the heterozygous restored F1 plants in D8 cytoplasm were used as the pollen source. The testcross populations in 1996 and 1997 were each minimal for accurate judgment. Even though the two-year pooled data allowed a more robust conclusion, no significant deviation from 7:1 or 15:1 was detected. The heterogeneity test also indicated the homogeneity for both ratios from the two-year's data (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Number of male fertile and sterile cotton plants segregated in crosses involving the D2 restorer

 
On the other hand, when the F1 was used as the female parent to cross with a maintainer line, ARK8518, the testcross population had 51 fertile and 52 sterile plants (Table 1), which was a near-perfect fit to a 1:1 ratio ({chi}2 = 0.01). The result was confirmed by another testcross, in which 77 fertile and 102 sterile plants were observed, when heterozygous F1 plants (TM-1 x D2R) on AD1 cytoplasm were pollinated onto CMS-D8. The deviation of the segregation from 1:1 was not significant ({chi}2 = 3.49). The D2 restorer gene, Rf (Weaver and Weaver, 1977; Kohel et al., 1984), presumably is responsible for fertility restoration of CMS-D8.

To test whether the same result would be obtained in the restoration of CMS-D2 by D2R, male fertility was scored in two populations, that is, (D2R x ARK8518)F2 and [(D2R x ARK8518)F1 x ARK8518] BC1F1. If the D2 restorer allele (Rf/rf) in the CMS-D2 cytoplasm segregated normally in the pollen, a three fertile to one sterile ratio would be expected in the F2 population. However, out of 492 plants, only 36 sterile plants were observed in the F2 population. The chi-square test indicated that two duplicate dominant restorer genes existed in the D2 restorer (3:1, {chi}2 = 0.31). When the F1 of D2R x ARK8518 with the CMS-D2 cytoplasm was backcrossed as the female parent to ARK8518, 62 fertile and 56 sterile plants were obtained, as expected for the one restorer allele model in the D2R (1:1, {chi}2 = 0.95).

The non-restorer allele (rf) from tetraploid cottons apparently is not transmitted normally through pollen when it is in either D8 or D2 cytoplasm. Some of the pollen grains with the non-restorer allele may fail in fertilization because of inviability or reduced competitiveness relative to the pollen grains with the restorer allele. Since some of the rf pollen did participate in fertilization, restoration of CMS-D8 and CMS-D2 by the Rf gene from the D2-2 restorer is sporophytic. In the normal sporophytic situation, 1/2 Rf pollen and 1/2 rf pollen both participate in fertilization. In restoration of fertility to CMS-D8 or CMS-D2 by D2R, presumably both Rf and rf pollen grains are viable, but the rf pollen is less competitive than the Rf pollen. One explanation for the results is as follows. In the F1, the two female gametes Rf and rf participate in fertilization in a 1:1 manner, while the Rf and rf pollen grains produce zygotes in a 7:1 ratio. As a result, segregation in the F2 appeared to be a 15 fertile to 1 sterile ratio (Table 1). The testcross of the F1 onto CMS-D8 or CMS-D2 confirmed that the proportion of the rf pollen successful in fertilization was only 1/7 of the Rf pollen. The F2 and the testcross data demonstrate that, in both the D8 and the D2 cytoplasm, the Rf gametes are preferred to the rf gametes in fertilization in a 7:1 ratio.

Clearly rf pollen is much less competitive than the Rf pollen in the alloplasms. Possibly rf pollen grains could become completely noncompetitive under some environmental or developmental conditions. In 1998, no sterile plants were observed in more than 200 plants from the F2 population of CMS-D8 x D2R. The results indicate that the rf pollen grains produced in the F1 plants in the field in 1996 did not participate in fertilization. Further studies are required to determine factors that affect the viability of rf pollen. In comparison, as indicated previously (Stewart, 1995; Zhang and Stewart, 2001), rf2 pollen grains, that is, male gametes with the non-restorer allele (rf2) of D8R in heterozygous plants, are completely nonfunctional in the D8 cytoplasm, indicating gametophytic restoration of fertility.

In wheat (Triticum aestivum L.), more than 25 different cytoplasmic lines from T. timopheevi Zhuk. and Aegilops spp. were developed and many of them cause male sterility (Tsunewaki, 1996). Restoration to pollen fertility was determined to be sporophytic in all wheat genotype-alien cytoplasm combinations tested because pollen grains not carrying the Rf genes participated in fertilization. However, except for Rf3 of Splt in G cytoplasm from T. timopheevi, the pollen with Rf was functionally favored to different degrees in fertilization over rf pollen in the presence of most male-sterile cytoplasms. For example, in Mt cytoplasm, Rf pollen was favored in fertilization three to one over rf pollen (Tsunewaki, 1993). In Ae. crassa Boiss. cytoplasm, pollen grains carrying the recessive gene rfd1 had about a 10% disadvantage to Rfd1 pollen in fertilization (Murai and Tsunewaki, 1994). The Rfc1 gene from Ae. caudata L. leads to preferential fertilization, even in common wheat cytoplasm. Not only exotic cytoplasms, but also some alien chromosomes or genes, can have negative effects on the transmission of certain genes from cultivated species. These chromosomes or genes, called gametocidal chromosomes or genes on alien chromosomes (mainly homoeologous chromosomes 2, 3, and 4) from different Aegilops species were preferentially transmitted in wheat and even caused chromosomal breakage (Endo, 1990; Sano, 1990; Tsujimoto, 1995; Nasuda et al., 1998). A similar situation has been described in other plant species including maize (Schawartz, 1950); tobacco, Nicotiana tabacum L. (Cameron and Mov, 1957); tomato, Lycopersicon esculentum Mill (Rick, 1966; Alexander, 1973); lima beans, Phaseolus lunatus L. (Allard, 1963); barley, Hordeum vulgare L. (Konishi et al., 1990); and rice, Oryza sativa L. (Yabuno, 1990; Li et al., 1997). In CMS-S of maize, a gametophytic system, Kamps and Chase (1997) found a few unexpected male-sterile plants in four crosses when the expectation was all heterozygous fertile plants. Based on RFLP mapping information, Kamps and Chase concluded that Rf3rf3 plants with CMS-S cytoplasm may aberrantly transmit the non-restoring allele, rf3, through the male gametophyte. In D8 or D2 cytoplasm of cotton, DNA fragments tightly linked or co-segregating with the Rf gene from D2R will also be preferentially transmitted to the next generation.

Many studies have shown that male fertility is influenced by genetic background, the environment, and their interaction with CMS cytoplasms (Kaul, 1988). In cotton, expression of male sterility in the partially cytoplasmic male sterile alloplasmic lines conditioned by A2 and B1 cytoplasms was strongly affected by temperature (Meyer, 1969; Marshall et al., 1974). It is not surprising that different ratios of fertile and sterile plants were obtained in different populations and in the same population in different years (Table 1). Genetic studies on restorer factors can potentially be confounded significantly by the direction of cross and environmental conditions. This may explain the variable results among earlier reports on D2 restoration (Meyer, 1975; Weaver and Weaver, 1977; da Silva et al., 1981; Kohel et al., 1984). The relationship between male fertility and environmental conditions (such as temperature, relative humidity, and light intensity and duration) needs to be investigated.

Interestingly, in contrast to the D2 restorer, which can restore fertility to both CMS-D8 and CMS-hir, the D8 restorer cannot restore fertility to CMS-D2. Two hybrids, CMS-D2-57-4 x D8R and CMS-D2-81-4442 x D8R were all sterile, indicating that the D8 restorer gene is unable to suppress the D2 cytoplasmic dysfunction. A similar situation was observed for CMS-C1, a cytoplasmic male sterile line based on G. sturtianum Willis (C1) cytoplasm, that is, D2R restored fertility but D8R did not (Zhang and Stewart, 1999).

Allelism
The allelic test was conducted by making crosses between D8R as the female parent and D2R (Table 2). No sterile plants were observed in the F2 population with a total of 249 plants in 3 yr, most likely because the non-restoring allele of D8R in D8 cytoplasm is not transmitted through male gametes. Therefore, whether the restorer genes are allelic or not, the functional male gametes from the F1 plants contained at least one of the restorer alleles. As expected, crossing the F1 onto CMS-D8 also gave all fertile plants.


View this table:
[in this window]
[in a new window]
  		Table 2. Number of male fertile and sterile cotton plants segregated in crosses involving both D2 restorer and D8 restorer in the restoration of CMS-D8

 
When the F1 as the female parent was pollinated by maintainer B lines (TM-1 and ARK8518), consistent results in fertility segregation were obtained in the two testcrosses. In Testcross 1 [(D8R x D2R)F1 x TM-1], only one sterile plant (0.52%) was observed in 191 plants. To rule out the possibility that this sterile plant was not from seed contamination, a second large testcross population (D8R x D2R)F1 x ARK8518 (Testcross 2) was evaluated in the field in 1999. Out of 1730 plants, only seven sterile plants (0.41%) were observed (Table 2). Therefore, the results from the two testcross populations were consistent with a total of 8 sterile plants obtained among 1921 plants. Theoretically, if the two restorer genes were independently assorted, 1/4 of the plants would be sterile. If they were allelic, no sterile plants would be expected, and if they were on the same linkage group, the fraction of sterile plants would be between 0 and 1/2, depending on the recombination frequency (Table 3). The data revealed that the D8 restorer gene and the D2 restorer gene are tightly linked with an average recombination fraction of 0.93%, ranging from 0.81 (Testcross 2) to 1.05% (Testcross 1). The average standard error for the estimator of r is 0.48% (Liu, 1998).


View this table:
[in this window]
[in a new window]
  		Table 3. Expected cotton genotypes and their frequencies in the testcross (D8R x D2R)F1 x TM-1 under different assumptions

 
From the above result, we propose to change the designation of the D2 restorer gene from Rf to Rf1, and assign Rf2 to the D8 restorer gene. While doing so, we must also note the results from two other crosses that cannot be completely discounted (Table 2). Two types of female gametes, rf1Rf2 and rf1rf2, in a 1:1 ratio, were expected from the fertile plants in the population of [(D8R x AD1) x AD1]F1, while two pollen types, Rf1rf2 and rf1rf2, were also expected from fertile plants of either [(CMS-D8 x D2R)F1 x ARK8518]F1 or [CMS-D8 x (TM-1 x D2R)]F1. Normally, the two pollen types would occur in a 1:1 ratio, resulting in a three fertile to one sterile progeny ratio after pollinating onto the female parent. If Rf1rf2 and rf1rf2 in D8 cytoplasm were in a 7:1 ratio, as suggested previously, the two crosses would be segregating 1 sterile plant to 15 fertile plants. However, the results did not support either model since no sterile plants were observed in two populations with 47 and 62 plants each in 1998. It seemed that the rf1rf2 pollen in the D8 cytoplasm from Rf1rf1rf2rf2 plants was not functional and could not be transmitted. Supporting evidence was that almost no sterile plants were observed in an F2 population of CMS-D8 x D2R in the same year. Even though sterile F2 progeny were obtained in 1996 and 1997 from F1 plants (Rf1rf1rf2rf2) heterozygous at locus Rf1, indicating that the rf1rf2 pollen was viable, the chance for rf1rf2 pollen to participate in fertilization was much lower than for the Rf1rf2 pollen.

Since G. harknessii (D2-2) and G. trilobum (D8) are both American wild species belonging to the same genome, they share several characteristics, including similarities at the DNA level (Wendel and Albert, 1992). Molecular comparative mapping in grasses and some other families demonstrated that related species have very similar gene organization in that the same DNA markers, isozymes, and functional cDNA genes are arranged in one linkage group across species (Paterson, 1996). Thus, it is not surprising to see that the restorer genes from the two species are on the same chromosome and tightly linked, or even allelic to one another. Rf1 and Rf2 may exhibit synteny in their original D2 and D8 genomes, but unequal crossing-over in the F1's when crossed with upland cotton may have resulted in their localization on neighboring loci of the same linkage group. However, the relationship and interaction between Rf1 and Rf2 are not entirely clear at present and need additional study. In the F1 between D8R and D2R, the double heterozygous plants theoretically can produce four types of gametes: rf1Rf2, Rf1rf2, Rf1Rf2, and rf1rf2. The first two are parental types and last two are recombinant types. Several questions arise concerning the two new male gametes. Is there any epistatic effect between Rf1 and Rf2? What is the restoration mechanism (sporophytic or gametophytic) when Rf1 and Rf2 come together? Is the male gamete rf1rf2 viable? The pivotal question is the viability of rf1rf2 pollen. If Rf1 were epistatic to Rf2, the function of Rf2 would be inhibited by Rf1, and rf1rf2 pollen would be functional. Whereas, if Rf2 were epistatic to Rf1 or if the two restorer genes did not interact, then rf1rf2 pollen would not be functional. Therefore, whether sterile plants can be found or not in the F2 or testcross [CMS-D8 x (D8R x D2R)F1] should provide insight into the answer.

The F2 data seemingly supported the hypothesis that rf1rf2 pollen is not functional and unable to participate in fertilization. The implication from this would be that when the sporophytic Rf1 and gametophytic Rf2 come together in the F1, restoration of male fertility to CMS-D8 expresses gametophytically. However, the following calculation shows that a larger number of F2 plants would be required to verify the hypothesis. In heterozygous F1 plants (Rf1rf1Rf2rf2) carrying the D8 cytoplasm, gametes rf1rf2 have a frequency of 1/2 r (Table 3). In the absence of gamete selection, the sterile plants in F2 would be (1/2 r)2, that is, 1/40 000 in the case of r = 1% between Rf1 and Rf2. Since the rf1rf2 pollen is at least seven times less favored in fertilization (1/8 x 1/2 r) than pollen with one or two of the two restorer genes (7/8 x 1/2 r), the frequency of sterile F2 plants is 1/8 x (1/2 r)2, that is, 1/32 r2, meaning that sterile plants would account for only 1/320 000 of the F2 population.

Since we scored only 249 F2 plants, all of which were fertile, that number is far less than necessary to test which hypothesis is valid. It is not practical to consider an F2 population in this case. However, a testcross using the F1 as pollen source to cross onto CMS-D8 might be a feasible alternative. The frequency of sterile plants in the testcross would be 1/2 r, that is, 1/200, if pollen competition did not occur. Since the rf1rf2 pollen is less competitive, the expected frequency of sterile plants would be 1/8 x 1/2 r, that is, 1/1600. To ensure that at least one sterile plant would be obtained, about 4800 plants from the testcross should be scored.

Further cytological and genetic studies will shed additional light on the mechanisms of restoration to CMS-D8 by the two restorer genes. Fine mapping at the molecular and physical levels will provide detailed information on Rf1 and Rf2 localization and their physical relationship on their specific chromosome. The efforts will eventually lead to the isolation of the two restorer genes. Investigation of functional gene expression will provide better insight into the direct and indirect products of Rf1 and Rf2, the interaction between them, and their interactions with male sterile cytoplasms such as CMS-D8, CMS-D2, CMS-C1, CMS-D4, and CMS-hir.


   ACKNOWLEDGMENTS
 
The authors thank G.G. Coyle for her excellent assistance in the field and greenhouse work.

Received for publication March 3, 2000.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
H. V. Tang, J. F. Pedersen, C. D. Chase, and D. R. Pring
Fertility Restoration of the Sorghum A3 Male-Sterile Cytoplasm through a Sporophytic Mechanism Derived from Sudangrass
Crop Sci., May 31, 2007; 47(3): 943 - 950.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
J. Zhang, J. M. Stewart, and T. Wang
Linkage Analysis between Gametophytic Restorer Rf2 Gene and Genetic Markers in Cotton
Crop Sci., January 1, 2005; 45(1): 147 - 156.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
J. Zhang and J. McD. Stewart
Identification of Molecular Markers Linked to the Fertility Restorer Genes for CMS-D8 in Cotton
Crop Sci., July 1, 2004; 44(4): 1209 - 1217.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
J.F. Zhang and J.McD. Stewart
CMS-D8 Restoration in Cotton Is Conditioned by One Dominant Gene
Crop Sci., March 1, 2001; 41(2): 283 - 288.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Agricola
Right arrow Articles by Zhang, J.F.
Right arrow Articles by Stewart, J. McD.
Related Collections
Right arrow Cotton
Right arrow Cell Biology & Molecular Genetics
Right arrow Crop Genetics

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome