DIFFERENCES OF SIZE AND SHAPE OF ACTIVE AND INACTIVE X-CHROMOSOME DOMAINS IN HUMAN AMNIOTIC-FLUID CELL-NUCLEI

It is a widely held belief that the inactive X-chromosome (Xi) in female cell nuclei is strongly condensed as compared to the largely decondensed active X-chromosome (Xa). We have reconsidered this problem and painted X-chromosome domains in nuclei of subconfluent, female and male human amniotic fluid cell cultures (46,XX and 46,XY) by chromosomal in situ suppression (CISS) hybridization with biotinylated human X-chromosome specific library DNA. FITC-conjugated avidin was used for probe detection and nuclei were counterstained with propidium iodide (PI). The shape of these nuclei resembling flat ellipsoids or elliptical cylinders makes them suitable for both two-dimensional (2D) and three-dimensional (3D) analyses. 2D analyses of Xi- and Xa-domains were performed in 34 female cell nuclei by outlining of the painted domains using a camera lucida. Identification of the sex chromatin body in DAPI-stained nuclei prior to CISS-hybridization was confirmed by its colocalization with one of the two painted X-domains. In 31 of the 34 nuclei the area A(Xi) for the inactive X-domain was smaller than the area A(Xa) for the active domain (mean ratio A(Xa)/A(Xi) = 1.9 +/- 0.8 SD, range 1.0-4.3). The signed rank test showed a highly significant (P < .0001) difference both between A(Xa) and A(Xi) and between the ratios r(Xa) and r(Xi), calculated by dividing the maximum length L of each X-domain by its maximum width W. In most nuclei (26/34) we found r(Xa)>r(Xi) demonstrating a generally more elongated structure of Xa. For 3D analysis a confocal scanning laser fluorescence microscope (CSLFM) was used. Ten to 20 light optical sections (PI-image, FITC-image) were registered with equal spacings (approx. 0.4 mum). A thresholding procedure was applied to determine the PI-labeled nuclear and FITC-labeled X-domain areas in each section. Estimated slice volumes were used to compute total nuclear and X-domain volumes. In a series of 35 female nuclei most domains extended from the top to the bottom nuclear sections. The larger of the two X-chromosome domains comprised (3.7 +/- 1.7 S.D.)% of the nuclear volume. A mean ratio of 1.2 +/- 0.2 SD (range 1.1-2.3) was found for the volumes of the larger and the smaller X-domains in these female nuclei. In a series of 27 male amniotic fluid cell nuclei the relative X-chromosome domain volume comprised (4.0 +/- 2.6 S.D.)%. These findings indicate that differences in the 3D expansion of active and inactive X-chromosome domains are less pronounced than previously thought. A current model suggests that chromosome domains consist of a compact core surrounded by loosely coiled outer chromatin fiber loops. The latter fraction may be considerably larger in Xa- as compared to Xi-domains. We suggest that the interactive outlining procedure used in the 2D analyses included the loosely structured domain periphery more accurately, while the threshold algorithm applied to light optic-al sections delineated the more compact core of the domains, leading to smaller and more similar volume estimates of Xa and Xi. Present limitations of nuclear and chromosome domain volume measurements using confocal laser scanning microscopy are discussed.