Molecular cytogenetic studies in Chenopodium quinoa and Amaranthus caudatus

Chenopodium quinoa Wild. and Amaranthus caudatus L., two plant species from South America, have small and numerous chromosomes. Looking for chromosome markers to distinguish pairs of homologous chromosomes double fluorescence staining, in situ hybridization with 45S rDNA and silver staining were applied. Fluorescent in situ hybridization with 45S rDNA has shown two sites of hybridization occurring on one pair of chromosomes in qunion genre (lines PQ-1, PQ-8). The number of RDA loci in Amaranth's caudate L. genre depends on the accession. Kiwicha 3 line has one pair of chromosomes with signals and Kiwicha Molinera cultivar two pairs. All observed rDNA loci were active. After chromomycin/DAPI staining in all cases, except Kiwicha Molinera cultivar, the CMA 3 positive bands co-localized with signals of in situ hybridization with rDNA. In Kiwicha Molinera the number of CMA + bands was higher than the number of 45S rDNA signals after FISH.


INTRODUCTION
Quinoa (Chenopodium quinoa Wild.) and kiwicha (Ama ranthus caudatus L.) were cultivated as important prc-Colombian "grains". Both of these species are annual, self-pollinat ing C4 plants derived from South America. Until recently they have been cultivated only in the highlands of Argentina, Bolivia, Chile. Colombia, Ecuador and Peru. Today they are becoming more and more popular as crops in different regions. For this reason, they have started to attract scientific attention. Seeds of both species have a high nutritional value and a better amino acid balance than the proteins in most cereals. Their pro teins arc rich in the essential amino acids such as lysine, methio nine, and cysteine, which make them complementary to both cereals and leguminous. They have a wide adaptability to dif ferent ecological niches. Several cultivars have been selected for their tolerance to heat and cold as well as for resistance to disease (Popcnoe et al. 1989).
Chenopodium quinoa Wild. (Chcnopodiaccac) is a tetrapioid species with the chromosome number 2n = 4x = 36 (Wang el al. 1993). There is little information about the genome and kary otype of quinoa. The small size and great number of quinoa chromosomes make cytogenetic analysis difficult. Amaranthus caudatus L. (Amaranthaceae) is also a tctraploid species. The chromosome number for this species varies. Usually it was described as 2n = 4x = 32 and occasionally as 2n = 4 x = .34 (Pal and Khoshoo 1973). It has been suggested that the gametic chromosome number n = 17 originated from the n = 16 through primary trisomy and the basic chromosome number is x = 8 (Pal et al. 1982). The chromosome number and karyotype of many varieties of these species should be estab lished for better understanding of their origin and relationships, which occurred during evolution and plant breeding programs.
Today molecular cytogenetics offers excellent methods for analysis of plant genomes, even with small chromosomes (Osuji ct al. 1997;Hasterok and Maluszynska 2000). One of the most useful methods is fluorescence in situ hybridization (FISH), which allows localisation of different genes and noncoding DNA sequences on chromosomes and in interphase nuclei. rDNA loci are widely used as chromosome markers for kary otyping and studying the evolutionary relationships within gen era and families (Lee et al. 1999). In the presented study chro mosomes of quinoa and kiwicha were examined using in situ hybridization and differential chromosome staining. The num ber and localization of rDNA loci and their activity for the two species are presented.

Plant material
Two lines of Chenopodium quinoa: PIQ-1 and PIQ-8 and Amaranthus caudatus: cultivar Kiwicha Molinerc and line Kiwicha 3 were used. Plants were grown in pots in the growth chamber. For cytogenetic analysis young leaves were collected and pretreated with 8-hydroxyquinoline (2 hours at room temp, 2 hours at 4°C) prior to fixation. The material was fixed in ethanol: glacial acetic acid (3:1) and stored at -20°C until use.
For chromosome preparation, the material was washed with 0.01 sodium citrate buffer (pH 4.8) for 15 min and digested with enzyme mixture: cellulase (4% w/v, Onozuka) and pectinase (20% v/v,Sigma) for 2 hours at 37°C. After digestion, the mate rial was washed again with sodium citrate buffer for 30 min. Squash preparations were made in a drop of 60% acetic acid. Alter freezing the slide and removal of coverslip, preparations were dried overnight and stained with 2 ug/ml 4,6-diamidiono-2-phcnylindolc (DAP1) to check the quality of the chromosome preparations.

Fluorescence in situ hybridization (FISH)
Fluorescence in situ hybridization was applied according to the method described by Maluszynska and Heslop-Harrison (1991) with some minor modifications. 18S-25S rDNA isolated from Atubidopsis tlialiana was used as a probe for in situ hybridization. The DNA probe was directly labelled with Cy3-11-dUTP by nick-translation (Amersham Life Sciences) according to the manufacturer's instruction.
Prior to. FISH, the slides were pretreated with 100 jjl/ml RNase for 1 h at 37°C and dehydrated in 70% and 90% ethanol. The chromosomes and hybridization mixture were denatured separately. The chromosomes were denatured in 70% for mamide for 5 minutes at 78°C on a hot plate (Hybaid Thermal Cycler PCRin situ).
The hybridization mixture, consisting of 3 ng/pl of labelled probe 60% (v/v) formamide, 10% (w/v) dextran sulphate and 0,1 mg/pl salmon testes DNA in 2 x SSC (saline sodium cit rate). was denatured at 95°C for 10 minutes and immediately placed on ice for a few minutes. 15 pl of the hybrydization mix ture were added to each chromosome preparation and covered with a plastic coverslip. Hybridization was carried out at 37°C for about 72 hours. Stringent washing was performed at 40°C in 2 x SSC. 0.1 x SSC and 2 x SSC for 5 min each. Chromosomes were counter-stained with DAPI (2 mg/ml) and mounted in antifadc medium CITIFLUOR (Pelco).

Silver staining
Silver staining was performed following the method of Hizumc (1980). Slides were treated with a borate buffer (pH 9.2) for 5 min and air-dried. 50 pl of freshly prepared 50% sil ver nitrate in distillate water were applied to each preparation. Slides were covered with nylon mesh and incubated in a humid chamber for 20 min at 42°C, washed in distillate water, and air dried. Preparations were mounted in DPX medium. The number of nucleoli per nucleus was analysed in about 6000 cells for each type of plant.

Fluorescent banding
Fluorescent staining followed the method described by Schweizer (1976). Briefly, material was incubated with 0.5 mg/ml chromomycin Ai (CMA3) and DAPI (2 mg/ml) for 60 min each with subsequent washing in distilled waler. Preparations were mounted in a 1:1 mixture of glycerol and Mcllvaine buffer with 2.5 mM MgClj. Slides were analysed after two weeks maturation at 37°C and then stored at + 4°C.

Chromosome analysis
Preparations were examined with an OLYMPUS PROV1S epifluorescent microscope using the proper filter set. Photos were taken on KODAK 400 or 100 ASA film and processed with an AnalySIS program (OLYMPUS). Measurements of chromosome length were done using the same AnalySIS pro gramme for 6 metaphase plates, for each line.

RESULTS
The small and numerous chromosomes of quinoa and kiwicha do not facilitate cytogenetic analysis. Additionally, they are dif ficult to obtain from root tips grown in soil due to root cap resis tance to enzyme digestion. Therefore, in the present study young leaves were used for chromosome preparation. The mitotic index in this tissue is relatively high and both metaphase and prometaphase chromosomes can be obtained. The chromo somes of all analysed species arc relatively small and very poor ly differentiated in their morphology. The ccntromer position cannot be easily determined. Chromosomes with secondary constriction (NOR-chromosomes) can only be recognised in quinoa chromosome complement.
Both analysed lines of quinoa have .36 small chromosomes (Figs la. 2a). Their length falls within the range of 1.0-3.3 pm. The chromosome number of the analysed kiwicha plants is dif ferent. Kiwicha 3 has 34 chromosomes (Fig. 3a) and Kiwicha Molinere possesses only 32 chromosomes (Fig. 4a). The size of chromosomes ranges from 0.9 to 2.4 pm.
Double fluorescence staining was applied to metaphase chro mosomes and interphase nuclei of all analysed lines. Two CMA-positive (CMA+) bands were detected at the satellited chromosomes of both quinoa lines (Figs le, 2e). In interphase nuclei two separate signals were also seen (Figs if. 2f). Similarly, in Kiwicha .3 CMA+ bands were observed al the dis tal region of two chromosomes (Fig. 3c). In Kiwicha Molinere six chromosomes possess CMA+ bands. Four of them were major (Fig. 4e) and seen in all cells, but they differed in size. Two minor bands could be observed after prolonged (for six weeks) maturation of chromosome preparations stained with CMA. In the interphase nuclei the maximum number of CMA+ signals observed was equal to the number of bands in (he chro mosome complement of all analysed plants (Figs 3f. 4f).
After fluorescence in situ hybridization with 45S rDNA. sig nals at the distal region of two chromosomes were observed in Chenopodium quinoa lines (Figs lc. 2c). In the interphase nuclei, two separate sites of hybridization were seen (Figs Id. 2d).
In the Amaranthus caudatus chromosomes of Kiwicha 3 line, two rDNA loci were observed, while in the Kiwicha Molinera cultivar, four signals of in situ hybridization were present in all metaphase plates and interphase nuclei (Figs 3c. d. 4c. d). The signals in the Kiwicha Molinera cells differ in their size, which may indicate that one pair of loci possesses more copies of rRNA genes than the other.
Silver staining, an indicator of the transcriptional activity of rDNA. revealed the presence of two distally located silver-pos itive bands in one pair of quinoa chromosomes of both lines   (Figs lg, 2g). In the interphase nuclei one (Figs lh, 2h) or two nucleoli were observed. However, most frequently one nucleo lus was present (Fig. 5).
In Amaranthus caudatus in the investigated line and cultivar all loci of rRNA genes were transcriptionally active. Two Agpositive bands and one or two nucleoli in the interphase nuclei were observed in the Kiwicha 3 cells (Figs 3g, h). The Kiwicha Molinera cultivar had four Ag-NORs and one to four nucleoli in the interphase nuclei (Figs 4g, h). Similarly as after in situ hybridization Ag-NORs were different in size, two small and two larger.

Chromosome numbers
Both lines of Chenopodium quinoa examined in the present work have chromosome number 2n = 36 and according to Wang et al. (1993) they are tetrapioids. This was also reported 2n = 32 for C. quinoa (Kawatani and Ohno 1950). Among other Chenopodium species different ploidy levels with the basic chromosome number x = 9 have also been described, e.g. diploid (2n = 18) for C. glaucum and C. album, tetrapioid (2n =  36) for C. acuminatum and (2n = 32) for Chenopodium ambro sioides and hexapioid (2n = 54) for C. album (Tanaka and Tanaka 1980). Some authors also described the frequent occur rence of mixoploidy within the Chenopodium genus and related genera. Extensive analysis of this phenomenon has been con ducted in Spinacia oleracea (Lorz 1937). Wang et al. (1993) have also observed mixoploidy in root tip cells of C. neomexicanum. C. palmed, C. berlandieri. both, in seedlings and mature plants. Mixoploidy, described also as ..aneusomaty", occurs in natural and cultivated plant populations quite frequently (D'Amato. 1995). In the present study no mixoploidy has been observed in leaves of C. quinoa.
The chromosome number of Amaranlhus caudatus is usually 2n = 32 and occasionally 2n = 34 (Popenoe et al, 1989). Plants of cultivar and lines analysed in the present work represent both karyotypes. Some authors reported both numbers for the same species in genus Amaranlhus (Khoshoo and Pal 1972), but oth ers showed only 2n = 32 for all examined individuals (Greizerslein and Poggio 1994). Other Amaranlhus species have a similar chromosome number. 2n = 32 (A. hypochondriacus and A. mantegazzianus) or 2n = 34 (A. ententes) (Greizerstein and Poggio 1994).

Chromosomes carrying nucleolar regions (NORs)
In letraploid species C. acuminatum, two pairs of chromo somes with secondary construction were distinguished . Other Chenopodium species examined by those authors, such as C. ambrosioides (2n = 32) and C. glaucum (2n = 18), had one pair and hexapioid C. album had two pairs of chromosomes with secondary construction.
In situ hybridization shows that there is only one pair of chro mosomes possesses 45S rDNA loci in both Chenopodium quinoa lines. More than one pair of NOR chromosomes could be expected because of tetraploidy of the species. This suggests that at least one locus must have been lost. The different chro mosome number and loci number of rDNA in the examined Amaranlhus cultivar and line may be due to their different ori gin or chromosome rearrangements during evolution or the breeding process. Kiwicha Molinera possesses two pairs, but Kiwicha 3 only one pair of NOR chromosomes. The presence of two different sites of rDNA on separate chromosomes of Kiwicha Molinera may indicate the origin by hybridization of two diploid ancestors. One pair of NOR in Kiwicha 3 karyotype suggests the loss of one pair of rDNA loci and that this cultivar being under some chromosomal rearrangements is evolutional-ly older. The reduction of loci number in polyploid species has been described for several species (Maluszynska and Ileslop--Harrison 1993;Leitch et al. 1998).
All rDNA loci of C. quinoa and A. caudatus detected by in situ hybridization arc transcriptionally active. In other Amaranlhus species (A. ententes. A. hypochondriacus and A. mantagazianus) only one pair of NOR chromosomes was described (Greizerstein and Poggio 1994). The number of nucleoli in the interphase nuclei observed in this study indicates a tendency toward fusion in cells of young leaves. One or two nucleoli per nucleus in A. caudatus cells were reported in a pre vious work (Greizerstein and Poggio 1994). The lower number of nucleoli than number of NOR-chromosomes was observed in many plant genera such as llypochaeris (Ccrboh ct al. 1998). Brassica (Hasterok andMaluszynska 2000) and Arabidopsis (Weiss and Maluszynska 1998).
In the investigated plants the 45S rDNA sites co-localisc with CMA+ bands. In Kiwicha Molinera the number of CM A+ bands was higher than the number of 45S rDNA signals after FISH. In this genotype CMA+ bands differed in the strength of fluores cence, which may imply that the blocks of GC-rich heterochro matin are of a different size, similarly as the signal after FISH with 45S rDNA and silver staining. Only the two strongest pairs of CMA+ coincide with rDNA sites. This is a common feature among a wide variety of plants. The coincidence of CMA-band and rDNA loci was described for Vicia faba (Huizumc 1992). Clivia (Ran et al. 1999) and Brassica species (Hasterok and Maluszynska 2000).
The knowledge of the genome structure of Chenopodium and Amarahlus species is still limited, but recently developed genet ic studies should bring new information. Genetic polymorphism has been documented using allozyme analysis (Wilson 1988) and DNA markers (Ruas el al. 1999).