Association and the Adh locus with a lethal factor (l(2) Stm) in Drosophila melanogaster

Keeping Drosophila cultures at 28 C results in elimination of all minor multiple ADH bands, thought to be due to conformational change. Thus in diploid and triploid adults heterozygous for the Adh ^F and Adh ^Salleles, relative staining intensities are found for the three bands which were in conformity with the assumption that both alleles are equally expressed. Among all polymorphic strains derived from natural Central European and Mediterranean populations, the strain ⁺Tüb is unique in that its Adh ^Fallele is closely linked to a new recessive lethal factor, named 1(2)Stm. All Adh ^F 1/Adh^F 1 pupae are unable to emerge, and die. The lethal effect is obvious 50 hr earlier by retarded eye, bristle, and body wall pigmentation. Although all pupae of the phenotype F die, Adh ^F allele frequency scarcely seems to be lowered in this natural population.     

The alcohol dehydrogenase alleloenzymes AdhS and AdhF from the fruitfly Drosophila melanogaster: An enzymatic rate assay to determine the active-site concentration

A rapid and reproducible enzymatic rate assay for the quantitative determination of the concentration of active sites is presented for the alleloenzymes Adh^S and Adh^F from Drosophila melanogaster. Using this procedure the turnover numbers as catalytic-center activities were found to be 12.2 sec^–1 for Adh^F and 3.4 sec^–1 for Adh^S with secondary alcohols. This showed a slower dissociation of the coenzyme from the binary enzyme-NADH complex with Adh^S and hence a stronger binding of NADH to this alleloenzyme. With ethanol, the catalytic-center activity was 1.4 sec^–1 for Adh^S and 2.8 sec^–1 for Adh^F, and hence the single amino acid mutation distinguishing the two alleloenzymes also affected hydride transfer.     

Selection against homozygotes in the ADH polymorphic system of Drosophila melanogaster associated with the lethal factor l(2) Stm

In the natural populations ⁺Tüb, ⁺Prov, and ⁺Rov, similar Adh ^F allele frequencies occur (q _F=0.11, 0.18, and 0.08, respectively). However, there is a discrepancy in that the Adh ^F allele in ⁺Tüb is closely linked to the lethal factor 1(2)Stm, which reduces relative fitness of the F phenotype to zero. In spite of this, polymorphism is maintained also in ⁺Tüb, because the heterozygotes are superior to the homozygous S type (relative fitness=0.88). Under laboratory culture conditions, in ⁺Tüb the relative fitness of the S genotype further decreases to 0.6. After outcrossing the lethal factor, relative fitnesses for S, FS, and F become 0.6, 1, and 0.48, respectively, implying that fitness for S remains the same. Relative values for S, FS, and F in ⁺Prov, not affected by the lethal factor, are calculated by the maximum average fitness method to be 1, 1.2, and 0.2 under the assumption that heterozygous FS are similarly superior to S as in the natural ⁺Tüb population and all allele frequencies found are stable equilibrium values.     

The use of Artemia biomass sampling to predict cyst yields in culture ponds

The possibility of using biomass volume (= mean biomass present in the pond.week ^–1) to predict the total amount of harvestable cysts (= kg wet weight collected. week ^–1) produced in a culture pond by an Artemia franciscana population using a mixed model regression was evaluated for two different sampling methods; horizontal transects and vertical point samples. For transects, the following equation was found: `log (0.01 + cyst yields) = –2.05 + 0.025<sup>*</sup>(biomass volume)' with F <sub>(1, 4.87)</sub> = 8.83 and p = 0.032. For the point samples, the regression was also significant with F <sub>(1, 55.2)</sub> = 13.62 and p = 0.0005 for following equation: `log (0.01 + cyst yield) = –3.613 + 0.021^*(biomass volume). As pond effect and interaction terms did not significantly explain a significant portion of the variance for either of the sampling methods (Transects: pond: F _{(3, 14.3)} = 2.48; p = 0.103; pond^*biomass volume: F _{(3, 3.61)} = 4.63; p = 0.0976; Point samples: pond: F _{(3, 44.5)} = 0.00; p = 0.999; pond^*biomass volume: F _{(3, 44.2)} = 0.11; p = 0.954 ), the variable pond (repeated measurement factor) was not included in the final calculations for the regression equations. Although a combination of factors influences the equation, the high significance levels of the regression indicate biomass volume can be safely used to predict production trends. The low investment requirements of this method make it especially attractive for on farm use, where correctly determining the point of cyst decline will help farmers to allocate resources where needed.     

Measurement of the chlorophyll fluorescence induction kinetics with a 10 µs time resolution and its application in the forest decline research

Summary Measurement methods are described which determine the initial phase of the fluorescence induction kinetics with a maximum time resolution of 10 µs simultaneously for the two fluorescence componentsF ₆₈₅(t) andF ₁₃₀(t) selected by filters at the wavelengths 685 nm and 730 nm, respectively. The excitation light provided by a He-Ne laser (632.8 nm) is switched on within 0.3 µs (maximum intensityI _e=12 mW/cm²).F _o,F _p, andF _s, the initial-, peak-, and steady-state intensity and the initial valueR _o of the ratioR(t)=F ₇₃₀(t)/F ₆₈₅(t) can accurately be determined as well as the initial time derivativeF _o ^* of the fluorescence intensity.F _o andF _o ^* are related to the quantum yield _a of the antenna and to the photochemical quantum yield _pc, respectively. Spruce, oak, birch, poplar, and soy bean show a decline ofR(t) fromR _o to a first minimumR _b at some 10 ms which has a similar value as the second minimumR _p in the time range of seconds. Furthermore, the initial valueR _o and the steady-state valueR _S ofR(t) are also very similar. Measurements on spruce with water deficiency and with varying excitation light intensityI _e show effects on the initial phase of the fluorescence induction kinetics. Further measurements on spruce of different damage classes indicate that for the current year's needles the ratioF _p/F_o, is the most sensitive parameter to differentiate between the damage classes and thatF _o/F_s andR _o/R_b are also affected. As demonstrated by measurements on leaves of soy beans, the initial decrease ofR(t) fromR _o toR _b originates from a change of the fluorescence spectrum because no change of the leaf transmission can be observed in the time range between 10 µs and 1 ms.     

Dominance Relationship between Two Self-Incompatible<Emphasis Type="Italic">Brassica campestris</Emphasis> Haplotypes in Response to CO<Subscript>2</Subscript> Gas

InBrassica, self-incompatibility (SI) can be overcome by CO₂ application, an effective method for obtaining numerous inbred lines for F, commercial seed. We previously reported two different S-alleles ofBrassica campestris, S⁷³³ and S⁷³⁴, with extremely different degrees of susceptibility to this gas. In the current study, we raised a cross-population between those two genetic lines, and analyzed their reaction level of self-incompatibility to CO₂ (RLSICO₂). Here, all 40 of our progeny from the F₁ cross-population were susceptible, maintaining high values of RLSICO₂. This suggests that the susceptible line, S⁷³⁴, is dominant to the insusceptible line, S⁷³³. We also generated an F₂ selfing-population of each crossed progeny, S⁷³³♀ S⁷³⁴♂ and S⁷³³♂ S⁷³⁴♀, to assess the RLSICO₂ of each individual. PCR-RFLP analysis was performed to determine the S-genotype of the F₂ population. The S⁷³⁴ allele segregated in a theoretical ratio of the dominant trait, and the RLSICO₂ was consistent with the dominance relationship. Therefore, we have now demonstrated that high RLSICO₂ in β.campestris is controlled by a dominant gene. Both authors contributed equally to this work     

Dissociation-recombination of intergenic sunflower alcohol dehydrogenase isozymes and relative isozyme activities

Two unlinked genes, Adh ₁ and Adh ₂, control the production of alcohol dehydrogenase (ADH) in seeds of the annual sunflower (Helianthus annuus). Each gene is polymorphic, having F and S alleles. Starch gel electrophoretic zymograms of the four possible double homozygotes have three bands, representing two homodimers and an intermediately migrating intergenic isozyme. Zymograms of double heterozygotes consist of nine bands produced by ten isozymes: six intragenics and four intergenics, two of which are coincident. Results of dissociation-recombination (D-R) experiments are reported which demonstrate the subunit composition of the intergenic isozymes, thus supporting the relationships suggested by genetic studies. Densitometric tracings of the zymogram of a cleared gel and measurements of activities of homodimer isozymes eluted from gels following D-R of an intergenic isozyme showed that the Adh ₂ isozymes were more than twice as active as those of Adh ₁. Measurements of activities of crude extracts from the four possible double homozygous genotypes indicated that the seeds of the genotype Adh ₁ ^F /Adh ₁ ^F , Adh ₂ ^S /Adh ₂ ^S produced more activity than the other three. This genotype is the most common one found in wild and cultivated stocks. Isozymes eluted following electrophoresis of the same extracts had averages of 19%, 70%, and 11% of total activity contributed by the Adh ₁, Adh₂, and intergenic isozymes, respectively. A simple but efficient method of isozyme elution from starch gels is described which resulted in nearly full expected recovery (approximately 46%) of the ADH activity in the applied sample.Supported by Graduate School and BioMed grants and by NSF Grant GB35853.     

Inheritance of body color in the brown planthopper, Nilaparvata lugens

The segregation frequencies of color phenotypes of F₁, F₂, and backcross progenies revealed that the body coloration of the brown planthopper, Nilaparavata lugens(Stål), exhibited multiple allelism. Monogene for color morphism exists in three allelic forms >-b ⁺ (brown or wild type), b ^o (orange) and b ^b (black). The dominance relationship was b ^b>b ⁺>b ^o. The relationship between phenotype and genotype in N. lugens was not fixed. The six possible genotypes produced eight phenotypes.

---

Résumé Les fréquences de ségrégation des différentes colorations de F₁, F₂ et des croisements en retour ont montré que la coloration du corps de Nilaparvata lugens est due à un multi-allèlisme. Le monogène pour la couleur existe sous trois formes allèliques: b⁺ (brun ou sauvage), b^o (orange) et b^b (noir). Les relations de dominance étaient: b^b>b⁺>b^o. On n'a pas établi les relatios entre phénotypes et génotypes chez N. lugens. Les 6 génotypes possibles produisent 8 phénotypes.

     

The genetics of self-incompatibility in Brassica campestris L. ssp. Oleifera Metzg. I. Characteristics of S-locus. Control of self-incompatibility

Self-incompatibility in Brassica campestris c.v. Arlo is controlled by a single locus sporophytic system. The identity and expression of the S alleles were determined in eight inbred and two hybrid families. It was found that co-dominance of alleles is more frequent in the stigma, whereas dominance relations between pairs of alleles predominate in the pollen. A linear order of dominance was established between six S alleles and alleles high, intermediate and low in the dominance series were recognized.In considering the variation in the expression of compatibility and the segregation ratios in inbred, F₁, F₂ and backcross progenies, the presence of a specific S allele conditioning self-fertility, or a single dominant self-compatibility factor independent of the S locus could not be established. Instead, self-compatibility in this cultivar was ascribed to the segregation of a polygenic complex which is capable of modifying the incompatibility reaction to the point of self-fertility, or to a reduction in the strength of the reaction due to the presence of S alleles low in the dominance series.     

A fitness function for optimal life history in relation to density effects,competition, predation and stability of the environment

A fitness function (function maximized under natural selection) is studied in a population model in which the growth of a population is suppressed by crowding, density-independent continuous mortality (by euryphagous predators) and periodic disturbances. The dynamics of the population density between occurrence of disturbance can be expressed as,dN/dt=(F(N/K)−D)N, whereN is the population density,K is the carrying capacity,D is the density-independent continuous mortality, andF is the growth regulation factor described as a function of crowding (N/K). The period of disturbance isS. The survival rate under disturbance isu. It is concluded that the fitness function is (approximately) a product of competitive ability (C), carrying capacity, and degree of saturation, and is given byCKF ⁻¹(D−(lnu)/S). The degree of saturation is the inverse function of regulation factor (F) at the death rate due to predators and disturbance. I assume a population in which density is regulated only through survival. In this case, a low survival rate at the critical age-group means a high value ofCKF ⁻¹(D−(lnu)/S). Therefore, the reciprocal of the density-dependent survival rate at critical age-group is a measure of the fitness function. Using this measure, I predict the optimal age (body size) at first reproduction of a species of salamander. I also found that fitness calculated from observed values ofl(x) andm(x) includes a tautology. When the concept of fitness function is compared with the ESS method, the latter is more flexible. However, there is a possibility that an ESS is at the minimum of fitness function.     

Identification of the 7,8-didemethyl-8-hydroxy-5-deazariboflavin synthase required for coenzyme F<Subscript>420</Subscript> biosynthesis

The hydride carrier coenzyme F₄₂₀ contains the unusual chromophore 7,8-didemethyl-8-hydroxy-5-deazariboflavin (FO). Microbes that generate F₄₂₀ produce this FO moiety using a pyrimidine intermediate from riboflavin biosynthesis and the 4-hydroxyphenylpyruvate precursor of tyrosine. The fbiC gene, cloned from Mycobacterium smegmatis, encodes the bifunctional FO synthase. Expression of this protein in Escherichia coli caused the host cells to produce FO during growth, and activated cell-free extracts catalyze FO biosynthesis in vitro. FO synthase in the methanogenic euryarchaeon Methanocaldococcus jannaschii comprises two proteins encoded by cofG (MJ0446) and cofH (MJ1431). Both subunits were required for FO biosynthesis in vivo and in vitro. Cyanobacterial genomes encode homologs of both genes, which are used to produce the coenzyme for FO-dependent DNA photolyases. A molecular phylogeny of the paralogous cofG and cofH genes is consistent with the genes being vertically inherited within the euryarchaeal, cyanobacterial, and actinomycetal lineages. Ancestors of the cyanobacteria and actinomycetes must have acquired the two genes, which subsequently fused in actinomycetes. Both CofG and CofH have putative radical S-adenosylmethionine binding motifs, and pre-incubation with S-adenosylmethionine, Fe²⁺, sulfide, and dithionite stimulates FO production. Therefore a radical reaction mechanism is proposed for the biosynthesis of FO.Abbreviations AdoMet (SAM) S-adenosyl-l-methionine - Compound 6 5-Amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione - FO 7,8-Didemethyl-8-hydroxy-5-deazariboflavin - HPP 4-Hydroxyphenylpyruvate     

The use of the S haplotype-specific F-box protein gene,SFB, as a molecular marker for S-haplotypes and self-compatibility in Japanese apricot (Prunus mume)

Japanese apricot (Prunus mume) exhibits the S-RNase-based gametophytic self-incompatibility system as do other self-incompatible Prunus species. This report identifies the S haplotype-specific F-box protein gene (SFB), a candidate gene for pollen-S, of Japanese apricot, which leads to the development of a molecular typing system for S-haplotype in this fruit species. Both 5- and 3-RACE (rapid amplification of cDNA ends) were performed with SFB gene-specific oligonucleotide primers to clone Pm-SFB¹ and Pm-SFB⁷ of 'Nanko (S¹S⁷)'. As in the case of SFB of other Prunus species, Pm-SFB¹ and Pm-SFB⁷ showed a high level of S-haplotype-specific sequence polymorphism and their expression was specific to pollen. Genomic DNA-blot analyses of 11 Japanese apricot cultivars with the Pm-SFB probes under low stringency conditions yielded RFLP bands specific to the S¹- to S⁸-haplotypes as well as a self-compatible S^f-haplotype. A practical usage of SFB as a molecular marker for S-haplotypes and self-compatibility in Japanese apricot is discussed.Communicated by H.F. LinskensThe nucleotide sequences reported in this paper have been submitted to the EMBL/GenBank/DDBJ database under accession numbers, AB101440 and AB101441, for SFB¹ and SFB⁷, respectively     

Development of polarity in the ovules of the F2-progeny of theOenothera hybridalbicans · hhookeri

In the F₂-progeny of hybrids from crosses betweenOenothera biennis orsuaveolens andOe. hookeri with theRenner-complexesalbicans and^hhookeri, the development of callose pattern in meiocytes and megaspore tetrads is the same as in the F₁ and the parentOe. hookeri. During the development of the megaspore tetrads and the embryo sacs primary and secondary heteropolarity as well as homopolarity is observed. Estimates for the initial frequency of homo- and heteropolar tetrads at the end of the degeneration of megaspores in the tetrads immediately before the start of embryo sac development could be calculated. The F₂-plants can be arranged in three groups, distinguished by the frequency of the two polarity types. One of these groups behaves similar to the parentOe. hookeri, the two others have more homopolar tetrads. The segregation can be interpreted as recombination of genes, which influence the development of the polarity in the ovules. This is possible by crossing-over of genes between the twoRenner-complexes of the hybrid.     

Sex determination in<Emphasis Type="Italic"> Drosophila melanogaster</Emphasis> and<Emphasis Type="Italic"> Musca domestica</Emphasis> converges at the level of the terminal regulator<Emphasis Type="Italic"> doublesex</Emphasis>

Sex-determining cascades are supposed to have evolved in a retrograde manner from bottom to top. Wilkins 1995 hypothesis finds support from our comparative studies in Drosophila melanogaster and Musca domestica, two dipteran species that separated some 120 million years ago. The sex-determining cascades in these flies differ at the level of the primary sex-determining signal and their targets, Sxl in Drosophila and F in Musca. Here we present evidence that they converge at the level of the terminal regulator, doublesex (dsx), which conveys the selected sexual fate to the differentiation genes. The dsx homologue in Musca, Md-dsx, encodes male-specific (MdDSX^M) and female-specific (MdDSX^F) protein variants which correspond in structure to those in Drosophila. Sex-specific regulation of Md-dsx is controlled by the switch gene F via a splicing mechanism that is similar but in some relevant aspects different from that in Drosophila. MdDSX^F expression can activate the vitellogenin genes in Drosophila and Musca males, and MdDSX^M expression in Drosophila females can cause male-like pigmentation of posterior tergites, suggesting that these Musca dsx variants are conserved not only in structure but also in function. Furthermore, downregulation of Md-dsx activity in Musca by injecting dsRNA into embryos leads to intersexual differentiation of the gonads. These results strongly support a role of Md-dsx as the final regulatory gene in the sex-determining hierarchy of the housefly.Edited by D. Tautz     

Characterization of acid phosphatase-1 null activity mutants of Drosophila melanogaster

A complementation analysis was performed on 15 Acph-1 ⁿ alleles. Only three of these alleles proved to be nonleaky and to exhibit no evidence of complementation. These were then tested immunologically to determine their level of antigenically cross-reacting material (CRM). Their CRM levels were virtually zero, as compared to low, but positive, levels for two other homozygous Acph-1 ⁿ mutants and considerably higher levels for heteroallelic combinations of some of the leaky alleles. Acid phosphatase-1 enzyme subunits, formed by dissociating native enzyme, have close to 100% CRM activity in tests with antibodies elicited by native enzyme.Submitted by John Bell in partial fulfillment of the requirements for a Doctor of Philosophy degree from Cornell University.     

Physiological effects of an allozyme polymorphism: Glutamate-pyruvate transaminase and response to hyperosmotic stress in the copepod Tigriopus californicus

In order to regulate cell volume during hyperosmotic stress, the intertidal copepod Tigriopus californicus, like other aquatic crustaceans, rapidly accumulates high levels of intracellular alanine, proline, and glycine. Glutamate-pyruvate transaminase (GPT; EC 2.6.1.2), which catalyzes the final step of alanine synthesis, is genetically polymorphic in T. californicus populations at Santa Cruz, California. Spectrophotometric studies of homogenates derived from a homozygous isofemale line of each of the two common GPT alleles indicated that the GPT^F allozyme has a significantly higher specific activity than the GPT^S allozyme. Under conditions of hyperosmotic stress, individual adult copepods of GPT^F and GPT^F/S genotypes accumulated alanine, but not glycine or proline, more rapidly than GPT^S homozygotes. When young larvae were subjected to the same hyperosmotic conditions, GPT^S larvae suffered a significantly higher mortality than GPT^F or GPT^F/S larvae. These results suggest that the biochemical differences among GPT allozymes result in specific physiological variation among GPT genotypes and that this physiological variation is manifested in differential genotypic survivorships under some naturally occurring environmental conditions.This work was supported in part by a grant from the Lerner Fund for Marine Research of the American Museum of Natural History, an NIH Training Grant in Integrative Biology, and NIH Grants GM 28016 and GM 10452.     

Genetics of complement C4. Two homoduplication haplotypes C4S C4S and C4F C4F in a family

Summary A family in which two homoduplicated C4 haplotypes (or supergenes) segregate is described. One haplotype C4F ^* 3 C4F ^*2.2 is composed of two C4F alleles and the other C4S ^* 5.1 C4S ^*1 of two C4S alleles. The C4F duplication haplotype is a partial inhibitor of the Rodgers antigen, and judged from our family and population material, it seems to be rather frequent and associated with HLAB ^*35, Bf ^* F, and HLAD/DR ^*1. The C4S duplication haplotype is Rg(a-) and is not identified in individuals without another S, Ch(a+) variant.This work was supported by grant No 12-1727 from the Danish Medical Research Council     

Procedures for identifying S-allele genotypes of Brassica

Summary Procedures are described for efficient selection of: (1) homozygous and heterozygous S-allele genotypes; (2) homozygous inbreds with the strong self- and sib-incompatibility required for effective seed production of single-cross F₁ hybrids; (3) heterozygous genotypes with the high self- and sib-incompatibility required for effective seed production of 3- and 4-way hybrids.From reciprocal crosses between two first generation inbred (I₁) plants there are three potential results: both crosses are incompatible; one is incompatible and the other compatible; and both are compatible. Incompatibility of both crosses is useful information only when combined with data from other reciprocal crosses. Each compatible cross, depending on whether its reciprocal is incompatible or compatible, dictates alternative reasoning and additional reciprocal crosses for efficiently and simultaneously identifying: (A) the S-allele genotype of all individual I₁ plants, and (B) the expressions of dominance or codominance in pollen and stigma (sexual organs) of an S-allele heterozygous genotype. Reciprocal crosses provide the only efficient means of identifying S-allele genotypes and also the sexual-organ x S-allele-interaction types.Fluorescent microscope assay of pollen tube penetration into the style facilitates quantitation within 24–48 hours of incompatibility and compatibility of the reciprocal crosses. A procedure for quantitating the reciprocal difference is described that maximizes informational content of the data about interactions between S alleles in pollen and stigma of the S-allele-heterozygous genotype.Use of the non-inbred I_o generation parent as a known heterozygous S-allele genotype in crosses with its first generation selfed (I₁) progeny usually reduces at least 7 fold the effort required for achieving objectives 1, 2, and 3, compared to the method of making reciprocal crosses only among I₁ plants.Identifying the heterozygous and both homozygous S-allele genotypes during the I₁ generation facilitates, during subsequent inbred generations, strong selection for or against modifier genes that influence the intensity of self- and sib-incompatibility. Selection for strong self and sib incompatibility can be effective for both homozygous inbreds and also for the S-allele heterozygote, thus facilitating production of single-cross F₁ hybrids and also of 3-and 4-way hybrids.Department of Plant Breeding and Biometry paper No. 690     

In vivo analysis of chlorophyll <Emphasis Type="Italic">a</Emphasis> fluorescence induction

Quantitative characteristics of photosynthetic electron transport were evaluated in vivo on the basis of the multi-exponential analysis of OJIP fluorescence transients induced by saturating actinic light. The OJIP fluorescence curve F(t), measured in Chlamydomonas reinhardtii cells, was transformed into the (1 − F _O/F(t)) × (F _V /F _M)⁻¹ transient, which is shown to relate to PS 2 closure. We assumed that kinetics of PS 2 closure during OJIP rise reflects time-separated processes related to the establishment of redox equilibrium at the PS 2 acceptor side (OJ), PQ pool (JI), and beyond Cyt b/f (IP). Three-exponential fitting was applied to (1 − F _O/F(t)) × (F _V /F _M)⁻¹ transient to obtain lifetimes and amplitudes of the OJ, JI, and IP components of PS 2 closure, which were used to calculate overall rates of reduction and re-oxidation of the PS 2 acceptor side, PQ pool, and intermediates beyond Cyt b/f complex. The results, obtained in the presence of inhibitors, oxidative reagents, and under different stress conditions prove the suggested model and characterize the introduced parameters as useful indicators of photosynthetic function.     

Assessing hybrid sterility in <Emphasis Type="Italic">Oryza glaberrima</Emphasis> × <Emphasis Type="Italic">O. sativa</Emphasis> hybrid progenies by PCR marker analysis and crossing with wide compatibility varieties

Interspecific crossing of the African indigenous rice Oryza glaberrima with Oryza sativa cultivars is hindered by crossing barriers causing 100% spikelet sterility in F₁ hybrids. Since hybrids are partially female fertile, fertility can be restored by back crossing (BC) to a recurrent male parent. Distinct genetic models on spikelet sterility have been developed predicting, e.g., the existence of a gamete eliminator and/or a pollen killer. Linkage of sterility to the waxy starch synthase gene and the chromogen gene C, both located on chromosome 6, have been demonstrated. We selected a segregating BC₂F₃ population of semi-sterile O. glaberrima × O. sativa indica hybrid progenies for analyses with PCR markers located at the respective chromosome-6 region. These analyses revealed that semi-sterile plants were heterozygous for a marker (OSR25) located in the waxy promoter, whereas fertile progenies were homozygous for the O. glaberrima allele. Adjacent markers showed no linkage to spikelet sterility. Semi-sterility of hybrid progenies was maintained at least until the F₄ progeny generation, suggesting the existence of a pollen killer in this plant material. Monitoring of reproductive plant development showed that spikelet sterility was at least partially due to an arrest of pollen development at the microspore stage. In order to address the question whether genes responsible for F₁ sterility in intraspecific hybrids (O. sativa indica × japonica) also cause spikelet sterility in interspecific hybrids, crossings with wide compatibility varieties (WCV) were performed. WCV accessions possess "neutral" S-loci (Sⁿ) improving fertility in intraspecific hybrids. This experiment showed that the tested Sⁿ-loci had no fertility restoring effect in F₁ interspecific hybrids. Pollen development was completely arrested at the microspore stage and grains were never obtained after selfing. This suggests that distinct or additional S-loci are responsible for sterility of O. glaberrima × O. sativa hybrids.Communicated by H.C. Becker