Here we present the first MCB-based study on three-dimensionally preserved interphase nuclei derived from B-lymphocytes and sperm (Figs. 1, 2 and 3). Previously, comparable MCB-studies were performed on flattened nuclei with the known disadvantages of possible artifacts due to transformation of a spherical into a pancake-like object. Nonetheless, also important findings on decondensation of chromosomes during interphase could be obtained 45. While it was initially not possible to apply MCB probe sets 12 in S-FISH 15, this problem was now successfully solved 16. However, the comprehensive evaluation of one single interphase nucleus, including image acquisition, processing and analysis lasts about 4–5 hours using Cell-P software (Olympus). Thus, the number of investigated nuclei had to be restricted to 30 per tissue and/or species in this study.

The evaluation of the molecular cytogenetic results was done as described in the Method-part (see below). Position and distance of homologous chromosomes in the interphase nucleus was determined as peripheral (P) and central (C). Thus, either both homologues were located in the center (CC), in the periphery (PP) or both, in periphery and center (PC). The localization of the homologue chromosomes to each other, when located in the periphery was estimated as 'close together' (t), 'near by each other' (n) or 'on the opposite sides of the nucleus' (o) – see Figure 1.

The obtained results for the position of HSA chromosomes #18, #19, #21 and #22 in B-lymphocytes of HSA and of their corresponding homologous regions in GGO and HLA are summarized in Tables 1 and 2. Below the results for each of the four studied chromosomes are reported and discussed. Overall, there were no significant differences in the localization of the studied chromosomes in the three different species for B-lymphocytes.

According to our results (Tables 1 and 2) chromosome #18 is located in the periphery of the interphase nucleus in all three studied species: 85.0 ± 4.6% up to 88.3 ± 4.2% of studied #18 were located marginal. The localization of #19 was determined to be central in HSA, GGO and HLA with 81.7 ± 5.0% to 88.3 ± 4.2%. Thus, the position of #18 and #19 in nuclear architecture significantly differs from each other (for HSA t = 11.77; for GGO t = 11.26; for HLA t = 10.72; for all compared groups df = 118, p = 0.001). Hence, our data is in concordance with the results of 8 who suggested as a reason for that difference the divergence in gene density in these two chromosomes. The gene-density correlated radial arrangements of #18 and #19 were conserved during primate evolution, as previously shown 9.

For the analysis of position of homologous to each other only the data from the peripheral part of nucleus can be taken into account – thus, for chromosomes located in the center of the nucleus, like #19 (and #22) no such analysis was performed. For positions of homologous #18 to each other (Table 2) there was no significant differences in localization "PPt", "PPn" and "PPo". Thus, arrangement of human 18 homologue chromosomes to each other has random way in all three studied species (ANOVA-test for – HSA: F = 0.171, p = 0.843; – GGO: F = 0.054, p = 0.946; – HLA: F = 0.171, p = 0.843).

According to 67 chromosome #21 preferentially locates close to the center of the nucleus. The authors there postulate that this is due to the fact that acrocentric chromosomes carry nucleolar organizer regions on their short arms and the nucleolus is generally located in the inner nuclear space. Bolzer et al. (2005) 6 demonstrated that the distance between chromosomal territories of homologues acrocentric chromosomes is significantly smaller than the mean distance for five largest chromosomes. In contrast, they showed as well that the distance between chromosomal territories of homologues acrocentric chromosomes was not significantly different from the mean distances for the other small chromosomes.

In the present study the arrangement of #21 in interphase nuclei is non-random. In contrast with previous studies 67, human homologues chromosomes #21 mostly localized in the periphery of the nuclei in 70.0 ± 5.9% up to 81.7 ± 5.0% (χ² = 55.6, df = 1, p = 0.001). Thus, the preferentially localization of #21 in the peripheral part of nucleus could be explained by gene-density correlated arrangement: at a size of 33.5 Mb chromosome #21 contains about 225 genes, which is two times less than in chromosome #22 of approximate the same size. Moreover, in HLA the homologous region to HSA #21 is not located on an acrocentric chromosome.

The analysis of homologous #21 localized in peripheral part of the nucleus to each other showed that PPt and PPn compared to PPo is not significantly different for HSA (t = 0.816, df = 58, p = 0.418) and HLA (t = 1.348, df = 58, p = 0.1839)but for GGO (t = 3.928, df = 58, p = 0.001). Thus, only GGO #21 behaves like postulated for an acrocentric chromosomes, as homologous ones mostly localized 'together' and 'nearby'.

The observed position of #22 in nucleus of B-lymphocytes is different from position of #21. While the #21 mostly localized in the peripheral part of nucleus, #22 allocated more equally in the territory of nucleus. Differences in position of these chromosomes in nucleus are significant (χ² = 5.97, p < 0.015). In the present study #22 was localized to about 50% in the peripheral and to about 50% in the central part of the nucleus for all three species (Table 1). This data is in discordance to 6 demonstrating that acrocentric chromosomes preferably locate close to nuclear center. Regarding the orientation of homologous #22 to each other, in all three species they tend to be co-localized in 87% to 93% of the cases. Comparing PPt and PPn with PPo for significant differences were observed (HSA: t = 3.62, df = 58, p = 0,001;GGO: t = 4.477, df = 58, p = 0.001; HLA: t = 3.849, df = 58, p = 0.001. These results are similar with orientation of #21 in GGO (Table 2).

As shown in Table 3 localization of chromosomes #18, #19 and #21 in human B-lymphocytes and sperm is similar – however, this is not the case for chromosome #22. In sperm #22 is located in the center of the nucleus, i.e. according to its gene-density. It can be speculated that this is due to the fact that in genetically inactive sperm no nucleolus is formed. Thus, it could be postulated, that the chromosomes can here be arrange only according to gene density. In genetically active cells this primary order would then be disrupted by the fact that acrocentric chromosomes are attached to the more peripherally located nucleolus by their short arm carrying the nucleolus organizing regions. However, further studies have to be preformed to prove this suggestion.

Lymphoblastoid cell lines from human (Homo sapiens – HSA), gorilla (Gorilla gorilla gorilla – GGO) 17 and white-handed gibbon (Hylobates lar – HLA) were cultivated and cytogenetically prepared as previously reported 1718. All cell lines were karyotypically normal; GGO and HSA were female, HLA was a male.

Human sperm sample was collected in a sterile container after 3 days of sexual abstinence from a fertile, 28 year-old man with normal seminal parameters and a normal karyotype. After liquefaction at room temperature, the sample was washed three times in 1 × phosphate-buffered saline (PBS) by centrifugation (5 min at 2000 rpm) and fixed in fresh fixative (1:3 glacial acetic acid: methanol) 19.

S-FISH on interphase cells prepared according to standard procedures 18 was done as previously reported 15 with some modifications. In short, the entire FISH procedure is performed on cell suspension and the interphase nuclei are placed on a polished concave slide as the final step of the procedure, just before the evaluation. It was shown before that by S-FISH it is possible to do 3D analyses on totally spherical interphase nuclei 15.

The main steps of S-FISH technique included: pepsin treatment (i.e. 475 μl H₂O, 25 μl 0.2 N HCl, 0.005% pepsin), denaturation of DNA in interphase cells at 95°C, application of prepared DNA-probe containing 20 μg of COT1-DNA, hybridization over night at 37°C, washing in 0.4 SSC and 4 × SSC/0,2% Tween, blocking, detection and counterstaining in 0.5% DAPI-Vectashield (Vectashield; Vector, Burlingame, CA) 16. As probes multicolor banding (MCB) sets for HSA #18, HSA #19, HSA #21 and HSA #22 were applied 14. Images were captured on a Zeiss Axioplan microscope. Thirty interphase nuclei per chromosome were evaluated using software Cell-P (Olympus), rendering three-dimensional images and iso-surfaces that can be rotated freely and animated as well (Figs. 1, 2).

The aforementioned three-dimensional images and iso-surfaces were evaluated concerning 3-dimensional measurements, such as position and distance of homologous chromosomes in Cell-P software. The interphase nucleus was zoned into two spheres, i.e. periphery (P) and center (C); 50% of the nucleus radius was defined as 'center'. Thus, either both homologues were located in the center (CC), in the periphery (PP) or both, in periphery and center (PC). Chromosomes located on the borderline between the compartments were classified as C or P in relation to where the majority of the chromsome body size was located. The localization of the homologue chromosomes to each other was estimated as 'close together' (t), 'near by each other' (n) or 'on the opposite sides of the nucleus' (o) – see Figure 1.

Statistical analysis was performed using Student's t – test, One Way ANOVA (Analysis of Variance) and χ² – test to determine significant differences of chromosome's arrangement in nucleus. Statistical significance was defined as p < 0.05.