A Molecular and Embryonic Study of Apomixis in Cassava (Manihot Esculenta

Euphytica 102: 9–13, 1998. 9

c 1998 Kluwer Academic Publishers. Printed in the Netherlands.

A molecular and embryonic study of apomixis in cassava (Manihot esculenta

Crantz) ; Nagib M.A. Nassar 1, Marco Andre Vieira1, Clibas Vieira2 & Dario Grattapaglia3 1 Departamento de Agronomia, Universidade de Bras´ılia, C.P. 04477 Bras´ılia D.F. 70919; 2 Departamento de Fitotecnia, Universidade de Vicosa, Vicosa, MG 36570, Bras´ılia; 3 Laboratorio´ de Genetica´ de Plantas

CENARGEN-EMBRAPA – C.P. 02372 Bras´ılia D.F.; ( author for correspondence)

Received 10 February 1997; accepted 10 September 1997

Key words: embryo sac, RAPD, interspeciﬁc hybrid

Summary

In cassava, apomixis fixes heterosis and avoids transmission of systemic pathogens which complicate vegetative propagation of the crop. A combination of evidence from maternal inheritance of RAPD markers and the structure of the embryonic sac in large progeny sets of two distinct genotypes have further confirmed the occurrence of apomixis in cassava. We could advance further on earlier reports of the detection of apomixis in four ways: (1) we could arrive at an estimate of the rate of facultative apomixis in the range of 2%; (2) we detected the occurrence of apomixis in a second genotype, derived from a different interspecific cross; (3) apomictic behavior was demonstrated in an F1 individual and (4) parallel embryonic evidence was generated that corroborate the potential occurrence of apomixis by apospory. The fact that apomixis was detected in an F1 interspecific hybrid hints to the possibility of directly transferring genes for apomixis from a wild relative to cultivated cassava.

Introduction va genotypes could be maintained through successive generations. Apomixis in cassava was noted for the first Cassava (Manihot esculenta), yuca, mandioca, manioc time by Nassar (1980) while working on interspecific is the most important staple crop in the tropics and sub- hybridization. Later its presence was observed geneti- tropics. It is food for more than 600 million people in cally by the same author (Nassar, 1994) and confirmed the world (FAO, 1994). Cassava is propageted vegeta- by RAPD marker analysis in a single clone (Gratta- tively by stem cuttings. Vegetative propagation perpet- paglia et al., 1996). uates superior genetic combinations, but also favors The rationale behind the analysis of putative the accumulation of viral and bacterial diseases that apomixis with the RAPD assay has been described reduce productivity and may lead to the extinction of earlier (Grattapaglia et al., 1996). Our working hypoth- superior genotypes. If seeds were used to propagate esis is that a truly apomictic seedling would display an the crop, systemic pathogen contamination could be identical RAPD pattern of bands to the maternal parent avoided. However, the breakdown of genetic superior- for all the markers surveyed. Polymorphisms, i.e. non- ity due to genetic segregation in the progenyhas always maternal bands in the offspring or vice-versa would excluded this approach. reject the hypothesis of apomixis and indicate their Heterozygosity is responsible for vigor could be zygotic nature. This work extends our earlier report efficiently fixed by apomixis in cassava. This phe- of the molecular detection of apomixis in cassava by nomenon is defined as a process in which certain analyzing a larger sample of progeny individuals and plants produce seeds without fertilization. This means showing further evidence from embryological studies bypassing female meiosis and syngamy to produce for facultative apomictic behavior in two distinct cas- embryos genetically identical to the maternal par- sava genotypes. ent. Through apomictic reproduction, superior cassa-

MENNEN/zet: Pips Nr.:150696; Ordernr.:233449-wg BIO2KAP euph4374.tex; 18/05/1998; 10:48; v.7; p.1 10

Material and methods sities of the ethidium bromide stained samples to those of lambda DNA standards. For the RAPD assay, work- Plant material ing stocks of genomic DNA were diluted in water at a

concentration of 2.5 ng/l. Two putative apomictic cassava clones, identiﬁed by Arbitrary ten-base primers (kits OP-A through numbers 031 and 200 were used in this study. Clone OP-Z) were obtained from Operon Technologies Inc.

031 was selected based on vigor in an F2 population (Alameda CA). Ampliﬁcation reactions (13 l) were resulting from a cross between an interspeciﬁc hybrid carried out according to Williams et al. (1990) with

(M. dichotoma M. esculenta) and cultivated clone the following modifications: 0.4 mM ten-base primer, Branca Santa Catarina. Clone 200 is an F1 individual 10 g/ l Non-acetylated Bovine Serum Albumin (New from hybridization between cultivated cassava and M. England Biolabs), 5 to 10 ng of genomic DNA and glaziovii. Progenies from both clones were obtained by 1 unit of Taq DNA Polymerase. Amplifications were open pollination. Thirty seven progeny plants of each performed in 96-well microwell plates using a MJ clone were used in the anatomic and molecular studies. Research PT-100 thermal controller. RAPD products were analyzed by electrophoresis in 1.5% or 2.0% Embryo sac analysis agarose gels containing 0.2 mg/ml ethidium bromide. Gel images were captured and digitalized with an Morpho-structural development of embryo sacs were Eagle-Eye II system (Stratagene CA). Gel scoring was studied histologically. Unpollinated pistillate buds at performed directly from the gel images on a computer presumably 1 day pre-anthesis and pollinated pistils screen and images stored electronically on a laser CD. at post-anthesis were sampled and immediately fixed A set of 24 arbitrary primers selected earlier for in Farmer’s fixative (1:3 = glacial acetic acid:95% high multiplex content and discrimination power in ethanol) in the field between 7.30 and 12.00 mid day. cassava (Grattapaglia et al., 1996) were used. The Fixed pistils were dissected under a dissection micro- presence and absence of RAPD fragments was scored

scope (magniﬁcation 40, transmitted light). Dissect- by visual inspection of the gel images. Informative ed ovules were dehydratedin ethanol series and cleared RAPD markers were identiﬁed as described previ- overnight in the benzyl-benzoate-four-and-a half (BB- ously (Grattapaglia & Sederoff, 1994). Two replicate

41=2) fluid (lactic acid:chloral hydrate:phenol:clove RAPD analysis experiments including DNA extrac- oil:xylene:benzyl benzoate = 2:2:2:2:1:1, w/w devised tions, RAPD assays and marker scoring were carried by Herr (1982),treated in a modified Herr’s fluid as pre- out with the set of selected primers on the putatively viously reported by Ogburia & Adachi (1994). Trans- apomictic individuals to confirm the patterns of bands parent ovules were then observed using an Olympus observed. BX5O microscope equipped with Nomarski’s differential interference contrast (DIC) optics and a 100-W high pressure mercury lamp with appropriate filters Results and discussion for optimal viewing and photography. Both megaspo- rangia and megagametophytes components were pho- The two sets of maternal parent and 67 offspring indi- tographed and printed by a Sony color video printer viduals were genotyped with 24 selected arbitrary ten- Mavigraph UP-1200. One hundred and twenty seven base primers. Each selected primer amplified an aver- ovules from clone 031, and one hundred and thirty age of 8.25 clearly interpretable RAPD fragments with four ovules of clone 200 were used in the embryo sac a range of 5 to 14 fragments. A total of 198 clear- analysis. ly interpretable and reproducible RAPD markers were surveyed in this study. This number of markers was DNA extraction and RAPD assay considered to provide a representative genome cover- age for the objective of this study. On the average, Total genomic DNA was isolated from 200 mg of fresh 81% of the scored RAPD bands were monomorphic leaf tissue ground in liquid nitrogen using the CTAB between maternal parent and progeny set in the two protocol of Doyle & Doyle (1987), modified by the families. The remaining bands were polymorphic and addition of 1% PVP and 1% 2-mercaptoethanol to the thus useful for testing the hypothesis proposed herein. isolation buffer. DNA concentration was estimated by Figure 1 is an example of a RAPD profile generated gel electrophoresis comparing the fluorescence inten- using primer OPG-5. The arrows indicate diagnostic

euph4374.tex; 18/05/1998; 10:48; v.7; p.2 11

Figure 1. RAPD patterns for maternal clone 200 and 24 progeny individuals generated using primer OPG-5. From left to right the ﬁrst lane is 1 Kb ladder (Gibco-BRL) fragment size marker followed by the maternal parent and offspring individuals. The arrow indicates a polymorphic RAPD marker of estimated size 1000 bp absent in 10 individuals used to demonstrate the non-apomictic nature of these off springs.

RAPD markers that revealed the non-apomictic nature Given the number of markers surveyed, the probability of some offspring. that complete uniformity in RAPD markers between Progenies of clones 031 and 200 displayed a high the maternal parent and their respective progeny indi- uniformity of DNA ﬁngerprints.However, it was possi- viduals happened due to chance alone was equal or less ble to ﬁnd markers that readily showed that individuals than 103 in both clones. Therefore putative apomixis were not derived from apomixis, except for one indi- was detected in progenies of both clones, 031 and 200, vidual in each progeny. In the progeny of clone 031 at a rate of 3.13 and 2.70, respectively. These results individual 4 (Figure 2) showed a pattern of RAPD clearly indicate that the type of apomixis detected in bands identical to the maternal one for all primers this study is facultative and occurs at very low frequen- examined. The same identical pattern was observed cy in cassava. in individual 5 in the progeny of clone 200. A total of 261 ovules were analyzed histological- Although it would seem very unlikely that the ly. In both clones studied we found the presence of maternal parent and a zygotic progeny individual could aposporic sacs inside the sexual embryo sacs (Figure have an identical combination of over 100 RAPD 3). The presence of 2 embryo sacs in an ovule is the fragments, it could be argued that it is possible. As indicator of apospory nature of apomixis. It is sup- described in our previous report (Grattapaglia et al., posed that one of them is derived from somatic cells 1996), to exclude this possibility we used the statistical at its location in the ovule, while the second embryo procedure described by Novy et al. (1994) to show that sac is derived from a normal megaspore mother cell. the complete similarity between the two samples is not At a certain stage, before the complete maturation of an artifact resulting from a limited number of RAPD sexual embryo sac, it may abort and is replaced by markers surveyed and rather has a biological basis. the developing aposporous sac, otherwise, it contin-

euph4374.tex; 18/05/1998; 10:48; v.7; p.3 12

Figure 2. RAPD fingerprints sets of the maternal parent (left) and apomictic progeny individual (right) of clone 031 amplified with 16 arbitrary primers. Lanes 1, 18 and 25 are 1 kb ladder (Gibco-BRL) fragment size marker. From left to right sets of RAPD fingerprints obtained with arbitrary primers OPAA1, OPK19, OPK16, OPY16, OPH7, OPY20, OPM4, OPR10, OPX15, OPX1, OPY17, OPK1, OPY15, OPL17, OPZ18 and OPK4.

ues developing giving rise to two embryo sacs and two embryos in the ovule consequently. This abnormality was verified in 2.36% and 1.49% of the ovules of clones 031 and 200, respectively. Sim- ilar results using the same histological clarification technique were reported in Cenchrus ciliaris (Young et al., 1979). By the help of differential interfase contrast (DIC) microscope, it was possible to see in the cleared pistils details of cassava embryo sac. The normal sacs showed an egg, two polar nuclei, and three antipodals. Synergids were occasionally seen. The egg was often incospicuous. The antipodals were distin- guished by the swallen, tear-drop shape, dense cyto- plasm, chalazal position, and absence of a wall sepa- rating them from the cavity of the sac. The aposporous sacs lacked antipodals and had only one nuclei per sac. Sometimes there was a single polar nucleus and an egg. These results strongly suggest that the mechanism responsible for apomixis in cassava is apospory (development of aposporic embryo sacs). Apospory in the angiosperms is the most common mechanism responsible for apomixis. In this type of apomixis the apospory embryo sac originates from one or more somatic cells of the ovule followed by its enlargement Figure 3. DIC micrograph of a cleared ovule showing two embryo and vaculation (Asker, 1980). sacs; aposporic and sexual (instead of one sexual embryo in normal This type of apomixis explains the multiple plants cases). Identification of the aposporous and sexual embryos can be made on the basis of its location. In the ovule, the sexual member is per seed found in clone 031 by Nassar (1994). The located contrary in the micropylar region. presence of apomixis in clones 031 and 200 shows

euph4374.tex; 18/05/1998; 10:48; v.7; p.4 13 the potential of the utilization of wild species as a References source of genes for apomixis in cassava, since clone Asker, S., 1979. Progress in apomixis research. Hereditas 91: 231– 031 represents the generation F2 of a cross between cassava and M. dichotoma, and clone 200 is an F of 240. 1 Asker, S., 1980. Gametophytic apomixis:elements and genetic reg- cassava and M. glaziovii. Sources of genes controlling ulation. Hereditas 93: 277–293. apomixis in wild species that are relatives to corn,sugar Doyle, J.J. & J.L. Doyle, 1987. Isolation of plant DNA from fresh beet, wheat, and several forage grass have already been tissues. Focus 12: 13–15. FAO, 1994. FAO Yearbook. Rome. reported (Asker, 1979). Grattapaglia, D. & R.R. Sederoff, 1994. Genetic linkage maps of In conclusion this report further validates the occur- Eucalyptus grandis and E. urophylla using a pseudo-testcross rence of apomixis in cassava by presenting a combina- strategy and RAPD markers. Genetics 137: 1121–1137, 1994 tion of molecular and embryonic evidences from two Grattapaglia, D., C. Costa e Silva & N.M.A. Nassar, 1996. Strict maternal inheritance of RAPD fingerprints confirm apomixis in putative apomictic clones. By analyzing a larger set cassava (Manihot esculenta Crantz). Can J Plant Sci 76: 379–382. of progeny we could advance over the earlier report of Herr, J.M., Jr., 1982. An analysis of methods for permanently mount- detection of apomixis in clone 031 in four fundamental ing ovules cleared in four-and-a-half type clearing fluids. Stain ways: (1) we could arrive to an estimate of facultative Technol 57: 161–169. Nassar, N.M.A., 1980. Attempts to hybridize wild Manihot species apomixis in the range of 2%; (2) we detected the occur- with cassava. Econ Bot 34: 13–15. rence of apomixis in a second genotype, clone 200, Nassar, N.M.A., 1994. Development and selection of apomixis in derived from a different interspecific cross; (3) apomic- cassava, Manihot esculenta Crantz. Can J Plant Sci 74: 857–858. tic behavior was demonstrated in an F individual and Novy, R.G., C. Kobak, J. Gofreda & N. Vorsa, 1994. RAPDs identify 1 varietal misclassification and regional divergence in cranberry (4) parallel embryonic evidences were generated that (Vaccinium macrocarpon (Ait.) Pursh). Theor Appl Genet 88: corroborate the potential occurrence of apomixis in the 1004–1010. two genotypes investigated. Furthermore, the fact that Ogburia, M.N. & T. Adachi, 1994. An improved cleared-pistil tech- clone 200 is an F interspecific hybrid hints to the pos- nique for rapid in toto observation of embryo sac malformation 1 in cassava (Manihot esculenta Crantz). Proceedings of the Sec- sibility of directly transferring genes for apomixis from ond International Scientific Meeting Cassava Biotechnology Net- a wild relative to cultivated cassava. work, Indonesia, pp. 117–127. Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski & S.V. Tingey, 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucl Acids Res 18: 6531– Acknowledgements 6535. Young, B.A., R.T. Sherwood & E.C. Bashaw, 1979. Cleared-pistil This work was partially supported by the National and thicksectioning techniques for detecting aposporous apomixis in grasses. Can J Bot 57: 1668–1672. Council for Scientific and Technological Development (CNPq) to N.M.A.N., and a research grant to D.G. from PADCT/SBIO-FINEP. M.A.R.V. had a Msc. fellow- ship from CNPq. Special thanks go to the International Development Research Center (IDRC) Ottawa for the support in establishing the living Manihot collection at the Universidade de Bras´ılia.

euph4374.tex; 18/05/1998; 10:48; v.7; p.5