Electrofusion of protoplasts of anther-derived dihaploid lines of commercial potato cultivars

Protoplasteilla tarkoitetaan soluja, joiden solunseina on poistettu entsymaattisesti. Tallaisia soluja voidaan fuusioida yhteen, jolloin kahden eri perunalinjan yhdistelmasta muodostuu solujen kasvatuksen jalkeen somaattisia hybrideja. Somaattiset hybridit ovat perunan jalostukselle tarkeita, koska tavanomainen jalostus, joka tapahtuu tetraploidien (4x) kasvien suvullisin risteytyksin, tarvitsee runsaasti risteytyksia ja suuren maaran jalkelaisia. Kayttamalla hyvaksi dihaploideja (2x) perunoita ja fuusioimalla linjoja yhteen (2x + 2x),voidaan perunan jalostusta seka nopeuttaa etta tehostaa. Tassa tyossa ponsiviljelylla tuotetuista dihaploideista perunalinjoista (peraisin lajikkeista Matilda, Nicola, Pito, Stina, Van Gogh ja White Lady) eristettiin protoplasteja, joita fuusioitiin sahkoisesti. Kahden eri dihaploidin protoplasteja fuusioimalla saatiin tuotettua somaattisia hybrideja. Saadut kuusi uutta fuusioyhdistelmaa analysoitiin kayttamalla RAPIDmerkkeja. Oletetut hybridit maaritettiin useimmiten kahdella alukkeella, jotka kumpikin tuottivat dihaploideille vanhempaislinjoille spesifisen merkin. Useimmat tuotetut somaattiset hybridit olivat tetraploideja (4x), mutta neljasta fuusioidusta yhdistelmasta muodostui myos kasveja, joilla oli joko heksaploidinen (6x), oktoploidinen (8x) tai miksoploidinen genomi. Tama tutkittiin laskemalla hybridien kromosomit tai analysoimalla kasvit virtaussytometrilla. Somaattisten hybridien DNA-pitoisuuksien oletettiin olevan yhta suuria kuin tetraploidien perunoiden DNA-pitoisuudet tai dihaploidien fuusiovanhempien DNA-pitoisuuksien summat. Kuitenkin useiden tetra- ja heksaploidien hybridien DNA-pitoisuuksien keskiarvot olivat pienempia kuin oletetut DNA-pitoisuudet.

lines cannot be considered only as a sophisticated research method, but rather a technique which can now be applied successfully in potato breed- ing (Möllers and Wenzel 1992, Schweis and  Munzert 1993, Möllers et al. 1994).
Somatic hybrids must be distinguished from unfused material or fusion products resulting from homokaryon fusions.Various selection methods have been developed.Hybrid fusion products can be selected using a micromanipulator (Waara et al. 1991), flow sorter (Puite et  al. 1988), hybrid vigour (Debnath and Wenzel 1987), intermediate morphology (Gleddie et al.  1986), mutant lines (White and Vasil 1979) or with selectable markers (Masson et al. 1989).Identification of hybridity based on molecular (Pehu et al. 1989, 1990, Baird et al. 1992) and biochemical analysis (Waara et al. 1989, Cooper-Bland et al. 1994) has also been applied.It is essential that the identification for hybridity is simple and quick due to the high number of fu- sion products to be screened.
The genetic composition of the intraspecific somatic hybrids is expected to be balanced (the hybrids are euploid tetraploids), but aneuploidy and different ploidy levels are common among hybrid regenerants (Waara et al. 1992, Rasmus- sen and Rasmussen 1995).In this study, anther- derived dihaploid potato lines derived from cvs.Matilda, Nicola, Pito, Stina, Van Gogh and White Lady were electrofused in various combinations.The somatic hybrids were characterized by chro- mosome counts and nuclear DNA content deter- minationby flow cytometry.The objective of the study was to produce and characterize new intraspecific somatic hybrids for potato breeding purposes.The aim of the study was also to get more information on potato breeding at the diploid level in order to move anther cultures and protoplast fusions as part of the practical potato breeding.

Protoplast isolation
Leaf material of 4 to 6-week-old plants was cut into small sections and placed in 10-20 ml of preplasmolysis solution (0.5 M mannitol) for I h.
The material was then transferred into 10 ml of enzyme solution (Rokka et al. 1994).
After 16-18 h enzyme treatment in the dark at 24°C, the protoplast suspension was filtered through a 48 pm nylon sieve.The filtrate was centrifuged at 80 g for 5 min.The protoplast pellet was resuspended in wash solution contain- ing the major salts of CPW medium with man- nitol (Jones et al. 1989).Viable protoplasts were separated from dead protoplasts by centrifugation at 120-160 g for 5 min on 30% (v/v) Per- coll (Pharmacia Fine Chem.AB).The layer of viable protoplasts on the surface of the Percoll solution was collected and washed with the wash solution followed by two further washes in the fusion solution (0.5 M mannitol, 0.2 mM CaCl 2 ).
Electrofusion and culture of the protoplasts For the fusion experiments the protoplasts of the fusion parents were mixed in a 1:1 ratio.The protoplast mixture, adjusted to a density of 2 x 10 5 ml 1 with the fusion solution, was transferred into a lamellar chamber.The protoplasts were aligned and fused according to Rokka et al.  (1994).Following fusion, the protoplasts in man- nitol solution (500 pi) were pipetted into 3.5 cm diameter Petri dishes to which double strength V-KM culture medium (Bokelmann and Roest  1983) was added in a ratio of 1:1.The proto- plasts were embedded adding 0.9% (w/v) low- gelling-temperature agarose (Type VII, Sigma) and cultured in the dark at 24°C.After 10-14 days the cultures were resuspended with 3-7 ml per plate normal strength V-KM medium and transferred to dim light.When colonies devel- oped (after 3-5 weeks), the cultures were con- tinued as described by Rokka et al. (1994), ex- cept that STS (2 mg I 1) was added to the media D and SP (Creissen and Karp 1985).
Analysis of hybridit/ by RAPD patterns DNA extraction was carried out according to Rokka et al. (1994).The RAPD primers were synthesized either on an Applied Biosystems 372 DNA/RNA Synthesizer or purchased from Operon Technologies (Alameda, USA).Different 10or 11-mer primer sequences were used to estab- lish polymorphisms between the dihaploid lines.Primers producing unique amplification products in both parental lines, or preferably, sequential- ly used primers producing genotype specific bands, were used in identification of somatic hybrids.The PCR (polymerase chain reaction) was carried out as described by Rokka et al.   (1995).
Chromosome counts and nuclear DNA content determination Chromosome numbers were counted from root tip cells of in vitro (MS2O + 0.05 mg 1 1 NAA) cultured plants according to Tiainen (1992).
Nuclear DNA content (2C values) was measured from the dihaploid parental lines, the tetraploid original cultivars and the somatic hybrids using flow cytometry as described by Rokka et al.  (1995).1000-4000 nuclei were analysed in each sample.

Protoplast isolation and culture
The protoplast yields varied considerably be- tween dihaploids, and in many cases one of the parental lines of the fusion combination had more burst and collapsed protoplasts than the other (data not shown).Embedding in 0.9% aga- rose decreased the burst of the protoplasts dur- ing the first days of culture.Compared to cul- turing of the protoplasts in liquid medium, more divisions and colonies occured in embedded medium.The resuspension of cultures by liquid V-KM medium after 11-14 days enhanced strongly the rate of divisions, and the dilutions also prevented browning of the growing colo- Table 1.Nucleotide sequences of the RAPD primers, which were used in identification of hybridity.nies.The time required from protoplast isolation and fusion to shoot formation ranged from 5 to 8 months depending on the fusion combination.
(+) Matilda 1.dh.536.6'two primers out of 14 primers produced different patterns for the pa-rental lines.In these two combinations, a single primer allowed successful identification of hybridity (Table 1).The somatic hybrids contained the combined pattern of the parental lines, where- as unfused material and regenerants derived from homokaryon fusions had the pattern of only one of the parental dihaploid lines.Two distinct parental-line-specific primers were used in the identification of the hybrids in the other four fusion combinations ('Pito   35.dh.7.4.1.(+) Van Gogh 7.dh.12.2.1.','Pito  30.dh. 16,1.1. (+) Van Gogh 13.dh. 1 1.3.1.','Pito   12.dh.57.3  1).Thus, 90% (44/49) of the primers gen- erated polymorphism specific to one of the parental lines.Using two such parental-linespe- cific primers required twice as much resources, but were more reliable in the hybridity verifica- tion than the use of a single primer.An example of the identification of somatic hybrids is shown in Figure 1.The frequency of somatic hybrids was estimated for six different fusion combinations (Table 2).In total, 215 hybrid shoots were recovered from a total of 33 calli (Table 2). (+) Stina 4.dh.161.15' were counted.All of the hybrids were tetraploid, chromosome numbers ranging from 45 to 50.Ploidies of all of the somatic hybrids derived from seven different calli of the fusion combination 'Pito 35.dh.57.3.1. (+)  Nicola 2.dh.2.1.1.'were also tetraploid.The oth- er fusion combinations produced both tetraploid, hexaploid, octoploid and mixoploid hybrids (Ta- ble 3).The total number of dihaploid-dihaploid fusion calli, which produced somatic hybrid re- generants of different ploidy levels is shown in Table 4. Yet, the tetraploid level was the most common.
Table 3. Ploidy level of the dihaploid-dihaploid somatic hybrids of potato derived from six fusion combinations.The ploidy levels were determined with chromosome counts and/or flow cytometric nuclear DNA analysis.* mixoploid shoots at 6x and 8x (6x-8x) levels or 4x and 6x (4x-6x) levels between 1.65 and 1.73 pg.The mean 2C values of the original tetraploid cultivars were generally two times higher (3.32 pg in cv.Pito, 3.44 pg in cv.White Lady, 3.47 pg in cv.Nicola and 3.48 pg in cv.Van Gogh) compared to the corresponding dihaploid lines.The mean 2C values of all of the tetraploid somatic hybrids (3.18-3.29 pg) derived from three fusion combinations, were lower than the expected 2C values (i.e. the 2C values of the original tetraploid cultivars or the sum of the 2C values of the dihaploid parents) (Table 5).Also in hexaploid somatic hybrids the * mean of three flow cytometric measurements and standard deviation mean 2C values were lower (4.56-4.77pg) than the expected 2C values (4.97-5.18pg) (Table 5).However, in each fusion combination there were a number ofregenerants having the same or higher 2C values than the original cultivars.The hybrids having higher 2C values were probably hypertetraploids or hyperhexaploids.Examples of the flow cytometric DNA content determina- tion of the potato material are shown in Figure 2.

Discussion
In this study, six new dihaploid potato line com- binations were produced by electrofusion of pro- toplasts.The genetic material of the anther-de- rived dihaploid parents originated from Scandi- navian (cvs.Matilda, Pito and Stina), Dutch (cv.Van Gogh) and Hungarian (cv.White Lady) po- tato cultivars.Furthermore, the results of this study demonstrate the applicability of RAPD analysis in identification of intraspecific somat- Fig. 2. Flow cytometry of leaf nuclei of a dihaploid potato 'Nicola 2.dh.2.1.1.' (Fig. ), a tetraploid somatic hybrid 0406 between two dihaploid lines ('Pito 12.dh.57.3.1. (+) Nicola 2.dh.2.1.1.')(Fig. c.) and a hexaploid somatic hybrid 0402 between two dihaploids ('Van Gogh 19.dh.37.1.1.(+) White Lady 4,dh.2.3.2.') (Fig. d.).The histograms were generated by propidium-iodide stained leaf nuclei and chicken red blood cell (CRBC) controls using linear scale of fluorescence intensity (FL2-H).CRBC were added as an internal standard to the plant nuclei samples.The signal threshold was adjusted to eliminate most debris from analysis.The nuclear DNA content (2C value) was calculated by direct comparison of the modal position of the plant peaks to the modal position of the CRBC peak (DNA content = 2.33 pg).2C is defined as the DNA contentof the plant in the Gl phase of the cell cycle and 4C in G 2 phase (C is the DNA content of a haploid cell).
ic hybrids of potato, and flow cytometric nucle- ar DNA content analysis of the hybrids.
In previous research works, anther-derived dihaploids have seldom been used in intraspecific somatic hybridizations, because anther cul- ture of S. tuberosum has been considered ineffective in the production of dihaploids.Our re- cent results, however, in androgenesis of agronomically important potato cultivars, have been promising (Rokka et al. 1996).The advantage of producing dihaploid lines through anthercul- ture is that the anther-derived dihaploids do not contain any other genetic material than that of the anther culture source plant.Dihaploids produced by S. phureja pollinations may contain S. phureja DNA or variable chromosome num- bers (Clulow et al. 1993).The application of electrofusion rather than chemical fusion increases also the final number of somatic hybrids (Tempelaar and Jones 1985).The embedding of fused protoplasts with agarose followed by di- lution steps, enhanced first divisions of the cultured protoplasts.However, the whole culturing process of 5. tuberosum protoplasts is still quite limiting, if a wide range of genotypes is to be included in a protoplast fusion programme.
Generally, the methods for identification and selection of hybrid plants have also been a bot- tle-neck in protoplast fusion.However, RAPDs offer an opportunity to confirm the hybridity during very early stages of cultures.Because simple DNA extraction method can be applied for RAPD analysis, it is possible to screen a large number of regenerated plants in a short time (Rokka et al. 1994).Compared with isozyme (Waara et al. 1989, Möllers and Wenzel 1992)  and RFLP (restriction fragment length polymorphism) analysis (Pehu et al. 1989), the RAPD method is fast.Isozyme analysis and RAPDs have given similar results in verification of so- matic hybridity (Rasmussen and Rasmussen  1995).In the case that the hybrids should con- tain the combination of the bands of the both parental lines (using a single primer), competition for amplification sites in the target DNA may result in the absence of line-specific bands in the true somatic hybrids.Thus, the reproduc-ibility ofparental-specific bands would be more reliable using two primers, when each produce parental-specific bands rather than a single primer.However, one parental band may also be miss- ing as noted by Rasmussen and Rasmussen  (1995), who suggested this to be due to the lack of specific chromosomes carrying the corre- sponding primer sequence in the hybrids.In the present work, all the strongly diagnostic bands gave consistent results in the identification of hybridity.Primers can also amplify sequences of mitochondrial and chloroplast origin (Lorenz  et al. 1994).Complete chloroplast segregation is normal in intraspecific somatic hybrids (Lössi  et al. 1994), but mitochondria can produce rear- rangements (Xu et al. 1993, Lössi et al. 1994).
In rare cases nuclear hybrids may be identified as non-hybrids, if the primer amplifies chloro- plast DNA of the other parent.
In this study, two fusion combinations produced only tetraploid somatic hybrids, but in four combinations either tetraploid, hexaploid, octo- ploid or mixoploid hybrids were regenerated.In some cases the same callus regenerated into shoots which differed in ploidy levels from each other, which may be due to genetic rearrangements during the callus stage and shootregener- ation or the calli were derived from aggregated protoplasts or cell colonies (Waara et al. 1992).Rasmussen and Rasmussen (1995) noticed that in one fusion combination only few of the hybrids were tetraploid.Other than the expected tetraploid levels can also be explained by fusion of more than two individual protoplasts.Chime- ras are also possible to occur after compaction of protoplasts or because of the grafted groups of cells as described by Binding et al. (1988).The mean nuclear DNA content of the intraspecific somatic hybrids was lower than the expected DNA content.Valkonen et al. (1994) found a high correlation between 2C values and chromo- some numbers in diploid, tetraploid and hexa- ploid Solanum species.The low 2C value of most of the somatic hybrids produced in the present experiment could be due to aneuploidy.One to three individual chromosomes may be missed in many regenerants.There were, however, also some individual hybrids that had the expected or higher DNA content than the original culti- vars.Also Rasmussen and Rasmussen (1995)  noticed that hypoploidy was more general than hyperploidy among somatic hybrids.However, the association between the number of chromosomes and the phenotype of the plant is unclear.Lössi et al. (1994) have found no association, but Karp et al. ( 1989) have noticed that some aneuploids expressed phenotypic differences.
This paper reports successful production of somatic hybrids of anther-derived dihaploid po- tato lines.Further experiments are underway to characterize the disease resistance traits of the hybrids and to fuse new dihaploid lines having superior agronomic traits.
Chromosome counts and flow cytometric determination of nuclear DNA content Chromosomes of 24 shoots regenerated from one callus of a fusion combination 'Pito 35.dh.7.4.1.

Table 4 .
Number of dihaploid-dihaploid fusion calli, which produced somatic hybrid regenerants of different ploidy * shoots derived from the same callus, but having different ploidy levels

Table 5 .
DNA content(2C values)in leaf nuclei of dihaploid parental lines and the corresponding somatic hybrids in three fusion combinations.