Chromosome C-banding of the teosinte Zea nicaraguensis and comparison to other Zea species.

The Nicaraguan teosinte Zea nicaraguensis was studied cytologically to determine its chromosome number and C-banding pattern. The C-banding pattern was compared with that of the close relative Zea luxurians as well as with Zea diploperennis and cultivated maize, Zea mays ssp. mays. Karyograms were constructed for the four Zea species. It is shown that Z. nicaraguensis, like most other Zea species, is a diploid with 2n=20 chromosomes. The C-banding pattern shows that Z. nicaraguensis is very similar to Z. luxurians and more similar to Z. luxurians than to Z. diploperennis and cultivated maize. Whether or not Z. nicaraguensis and Z. luxurians should be regarded as subspecies instead of individual species is, however, not possible to conclude from this study.

The genus Zea is divided into two sections, section  Iltis & B.F. Benz (MOLINA and GARCIA 1999;ILTIS and BENZ 2000). The wild species of the genus Zea have the common name teosinte.
Both cultivated maize and the teosinte species are diploids (2n 020) with a tetraploid origin (except for Z. perennis which is a tetraploid) with the basic chromosome number X 010 (MOLINA and GARCIA 1999). The chromosome number of Z. nicaraguensis has not been reported before. Today, this species is found at only two locations in Pacific Coastal Nicaragua, but was much more common and widespread just some 20 years ago. Young plants of Z. nicaraguensis are consumed by cattle while mature plants, which can grow as high as 5 m, are used to make fences and shelters. The species occurs at 6Á15 m above sea level, and has the ability to grow in standing or slowly moving water. Because of this ability, due to a high capacity to form root aerenchyma and adventitious roots, it may be useful in wide hybridisations with maize, in order to improve maize growth in water-logged soils (ILTIS and BENZ 2000;MANO et al. 2006).
Z. nicaraguensis is to a large extent an unutilised genetic resource and its properties in terms of adaptation, disease resistance etc. have to be phenotypically and genetically characterized in order to determine its potential in maize improvement. Crosses between maize and Z. nicaraguensis have already been performed and hybrids have been obtained in an ongoing project of the authors. MANO et al. (2005) have in teosinte identified Quantitative Trait Loci (QTL) for adventitious root formation and efforts to transfer such QTLs from teosinte into maize are underway (MANO et al. 2006).
The closest relative to Nicaraguan teosinte, Z. nicaraguensis, is the Guatemalan teosinte, Z. luxurians (DOEBLEY and ILTIS 1980). The two species show close morphological similarity but the developmental behaviour differs considerably between the two, supporting a taxonomic segregation. Nicaraguan teosinte has longer and more abundant tassel branches, a larger number of spikelets per branch and a habitat differing from its Guatemalan counterpart (ILTIS and BENZ 2000).
The aims of this study was (1) to verify the chromosome number of Z. nicaraguensis, (2) to compare the C-banding pattern of Z. nicaraguensis with that of its close relative Z. luxurians as well as with Z. diploperennis and cultivated maize in order to see if similarities in banding pattern reflect their relationships.  Table 1 lists species authorities, accession name/ number and origin of the plant materials used in the study. Seeds of Z. nicaraguensis were collected at two different locations in Nicaragua, both in the Chinandega department, in the lowlands of the Gulf of Fonseca. Seeds are preserved in the gene bank of the National Agrarian University in Managua, Nicaragua. Seeds from the other wild Zea species were obtained from the Cimmyt Gene Bank in El Batan, Mexico through the Nordic Gene Bank in Alnarp, Sweden. Seeds of the French maize cultivar ''Birko'' were obtained from Svalö f Weibull AB in Svalöv, Sweden.

Chromosome preparation and C-banding
Seeds were placed in water for two days and then put on moist filter paper at 208C for two to three days. Roots were excised when they were around 10 mm long and placed in 0.02 M hydroxyquinolin for three hours, followed by ice water treatment for 21 h. After fixation in ethanol:acetic acid (3:1) the roots were kept in the freezer. Roots were also taken directly from plants in the greenhouse. Root tip chromosome squash preparations were made using the standard cellulasepectinase enzyme digestion method (SCHWARZACHER and LEITCH 1994).
The C-banding procedure followed the method described by GILL et al. (1991) and good cells were photographed with a Leica CCD digital camera. Ten individuals of each species were examined. Chromosome length, chromosome arm ratio, length of large heterochromatic regions and position of regularly appearing thin heterochromatic bands (in Zea species the large regions are often referred to as knobs) were measured on one cell from each of three different individuals per species using the freeware computer program Micromeasure 3.3 (available at http: //www.colostate.edu/depts/biology/micromeasure). These measurements were used together with visual examination of pictures from all ten individuals to construct karyograms for the four species. The presence or absence of centromeric heterochromatin was recorded for each chromosome arm but was not included when measuring heterochromatin content, as it usually was small in size and difficult to measure accurately.

RESULTS AND DISCUSSION
The measurements of the chromosomes of the four species are summarized in Tables 2 and 3 and can be seen in Fig. 2. The numbering of the chromosomes is following the C-banding pattern described by MOLINA (1982). The banding patterns among individuals within species were the same except for a few minor bands, which were not seen in all cells. This is probably more due to a more or less successful C-banding procedure than to actual differences. Fig. 1a-d show pictures of the chromosomes of the four species after C-banding.
The chromosomes of Z. mays are the smallest with an average length of 11.2 mm. The wild species have larger chromosomes with the largest in Z. nicaraguensis with an average length of 19.6mm. The chromosome lengths are not in agreement with the per cent of heterochromatin found in the present study. Differences in chromosome contraction or differences in nuclear DNA content not related to heterochromatin could explain the observed differences in chromosome size. Nucleolus organizing regions (NOR) could be localized in the two species with the longer chromosomes, Z. nicaraguensis and Z. luxurians but not in the shorter chromosomes of Z. diploperennis and Z. mays. The average arm ratios are similar between the four species and individual chromosomes vary from 1.0 to 2.5 (Table 2).
In cultivated maize, Zea mays ssp. mays, the number of heterochromatic knobs has been found to vary in different maize populations and lines. The Cbanding pattern hence differs between studies depending on material (HADLACZKY and KÁ LMÁ N 1975;MOLINA 1982;TITO et al. 1991). However, in all cases, it has mostly subterminal knobs and only a few terminal ones, which makes it different from the wild teosinte species in section Luxuriantes. The maize cultivar used in this study has terminal knobs on both arms of chromosome 1, terminal knobs on one of the arms of chromosomes 5 and 7 and subterminal knobs on one of the arms of chromosomes 2, 4, 6, 8 and 9.  Chromosomes 3 and 10 are lacking heterochromatic knobs. Chromosomes 1, 2, 3, 4, 6 and 9 have centromeric heterochromatin and chromosomes 5, 7 and 10 regularly show a thin intercalary heterochromatic band (Fig. 2a, Table 3). The amount of heterochromatin in a chromosome set, measured as a mean value from the three individuals, is 19.9% (Table 2). Also in Z. diploperennis, the size and position of heterochromatic knobs have been reported to differ between populations (KATO and LOPEZ 1990). It has, however, only terminal heterochromatic knobs making it different from cultivated maize. In the population used in this study chromosomes 1 and 2 have terminal knobs on both arms, chromosomes 4, 5, 6, 8 and 9 have terminal knobs on one of the chromosome arms and chromosomes 3, 7 and 10 are lacking heterochromatic knobs. Chromosomes 2, 3 and 4 show centromeric heterochromatin and chromosomes 1, 7, 8 and 10 regularly show a thin intercalary heterochromatic band (Fig. 2b, Table 3). The C-banding pattern is very similar to the one reported by MOLINA (1982). The amount of heterochromatin in the chromosome set is lower than in cultivated maize, only 14.0% (Table 2).
Zea luxurians has only terminal heterochromatic knobs. Chromosomes 1, 2 and 3 have terminal knobs at both ends while the remaining chromosomes have terminal knobs on one of the arms. Chromosomes 1, 2, 3, 4, 6, 8 and 9 have centromeric heterochromatin and chromosomes 1, 5, 6, 7 and 10 regularly show a thin intercalary heterochromatic band (Fig. 2c, Table  3). There is no pair lacking heterochromatic knobs and the mean amount of heterochromatin in a chromosome set is as high as 20.9% (Table 2). TITO et al. (1991) reported about a relationship between the number of C-bands and the DNA content where Z. luxurians showed a higher DNA content and more C-bands than both Z. diploperennis and Z. mays. Also Z. diploperennis, (c) Z. luxurians, (d) Z. nicaraguensis. Scale bar represents 10 mm. Hereditas 144 (2007) in this study, the number of C-bands on the chromosomes is higher in Z. luxurians (13 terminal knobs) than in Z. diploperennis (9 terminal knobs) and Z. mays (9 terminal or subterminal knobs).
Zea nicaraguensis, like most other Zea species, is a diploid with 2n 020 chromosomes. It has a C-banding pattern similar to Z. luxurians, and the two species are therefore probably closely related. Chromosomes 1, 2 and 3 have terminal knobs on both arms, chromosomes 4-9 have terminal knobs on one of the arms while chromosome 10 is lacking heterochromatic knobs. Chromosomes 1, 2, 3, 4, 6 and 9 have centromeric heterochromatin and chromosomes 1, 5, 7 and 10 regularly show a thin intercalary heterochromatic band (Fig. 2d, Table 3). The biggest difference compared to Z. luxurians is that chromosome 10 is knobless and this is the case in all 10 individuals studied. This leads to a lower heterochromatin content compared to Z. luxurians, a mean value of 16.4% (Table 2).
From the C-banding pattern it is clear that Z. nicaraguensis is more similar to Z. luxurians than to Z. diploperennis and cultivated maize. The similarity in C-banding pattern most likely reflects a closer relationship between the two species. The small differences in banding pattern that exist between them neither contradict nor support regarding them as subspecies as the C-banding pattern can vary considerably even between populations of the same species (as in maize and Z. diploperennis ). Whether they should be regarded as subspecies or separate species is thus not possible to conclude from this study.