We determined the radial positions of three representative chromosomes in interphase nuclei of HGPS cells; chromosomes 10, 18 and X. Chromosome 10 is found in different nuclear positions in proliferating, quiescent and senescent nuclei 42 43 . Chromosome 18 moves from the nuclear periphery to the interior when cells transit from proliferation to a non-proliferative state and is found in the nuclear interior in proliferating laminopathy cells, including an HGPS cell line 41 . The X chromosome remains at the nuclear periphery in all cell cycle states and is located at the periphery in all laminopathy cells analyzed 41 and as such is used as a negative control for chromosome repositioning.

To position chromosomes by fluorescence in situ hybridization (FISH) in interphase nuclei, we fixed cells in methanol:acetic acid (3:1) to produce flattened cytoplasm-free nuclei followed by two-dimensional FISH with specific chromosome paints. More than 50 digital images were then used in an erosion analysis that creates five concentric shells of equal area across the nucleus and the amount of DNA signal (DAPI) and chromosome paint signal were measured in each shell 38 39 . To normalize the data, fluorescence intensity of the chromosome signal was divided by the intensity of the DNA signal and the data were plotted as histograms, with the nuclear periphery represented by shell 1 and the nuclear interior by shell 5. The proliferative status of the cells is determined by indirect immunofluorescence using antibodies to the proliferative marker Ki-67 52 . Positive signal indicates that the cells are in proliferative interphase whereas cells negative for Ki-67 in cultures kept in high serum denote senescent cells 53 . Young quiescent cells, that is, serum starved or cells that have reached confluency, are also negative for anti-Ki-67.

Figure 1a, d confirms that chromosome 10 occupies an intermediate location in proliferating control nuclei (as determined by pKi-67 staining) and a peripheral location in control quiescent nuclei (Figure 1g, j). Figure 1p, v, a'' reveals that chromosome 10 is located at or towards the nuclear periphery in proliferating HGPS nuclei. Chromosome 18 is located towards the nuclear periphery in proliferating control cells (Figure 1e) but is then interior in control quiescent cells (Figure 1k), and in all three HGPS cell lines (Figure 1q, w, a'''). Chromosome × is found at the nuclear periphery in control proliferating (Figure 1f) and quiescent cells (Figure 1l), as well as in all three HGPS cell lines (Figure 1r, x, a''''). These relative positions for chromosomes 10 and × have been confirmed using three-dimensional fixation, laser scanning confocal microscopy, optical image reconstruction and measurement in three-dimensions (Figure S1 in Additional file 1).

We have recently shown that chromosomes relocate very rapidly to new nuclear locations in control proliferating fibroblasts placed into low serum 42 . When proliferating control fibroblasts (Figure 2a) are placed in low serum, chromosome 10 moves towards the nuclear periphery within 15 minutes (Figure 2I:d), chromosome 18 repositions from the nuclear periphery in proliferating fibroblasts (Figure 2I:g) to the nuclear interior, again within 15 minutes of incubation in low serum medium (Figure 2I:j), and chromosome × remains at the nuclear periphery from 0 minutes to 7 days (Figure 2I:m-r). When HGPS cells (AG11498) are placed in low serum there is no significant change in chromosome location over 7 days; that is, chromosome 10 remains near the nuclear periphery (Figure 2II:a-f), chromosome 18 remains in the nuclear interior (Figure 2II:g-l) and chromosome × remains at the nuclear periphery (Figure 2II:m-r).

FTIs have been used to correct several cellular aberrations in HGPS cells and in whole organisms. It has been suggested that by blocking farnesylation, certain proteins can be alternatively modified by geranylgeranylation. Thus, we employed FTI-277 both separately and simultaneously with GGTI-2147 to determine if we could restore chromosome position to normal in HGPS cells. An HGPS cell line (AG11498) was treated with 2.5 μM FTI-277 (Figure 3I:c, g, k) and with 2.5 μM each of FTI-277 and GGTI together (Figure 3I:d, h, l). The small amount of DMSO that was used to dissolve the inhibitors was used as a control (Figure 3I:b, f, j). As expected, the X chromosome did not change nuclear position with any of the treatments. However, with FTI-277 alone and together with GGTI-2147, chromosome 10 became located in an intermediate radial location in nuclei (Figure 3I:c, d). Chromosome 18 was also repositioned after treatment with FTI-277 alone and together with GGTI-2147 from an internal location to a peripheral one (Figure 3I:g, h). Chromosome × was not repositioned after FTI-277 treatment alone nor with FTI-277 and GGTI-2147 together (Figure 3I:k, l). DMSO alone had no significant effect on chromosome repositioning (Figure 3I:b, f, j).

After the 48 hour treatments the inhibitors were removed and the cells permitted to go through two more passages before chromosome positioning was analyzed again (Figure 3II). The newly corrected chromosome positions in the HGPS cells were maintained for treatment with FTI-277 alone and FTI-277 together with GGTI-2147.

After the HGPS cells had been treated with FTI-277 we wished to see if the rapid active chromosome repositioning after serum removal 42 was restored. Indeed, for all three chromosomes the starting location in proliferating nuclei was similar to that in the control cells and movement from an intermediate location to the periphery and the periphery to interior for chromosomes 10 and 18, respectively, was apparent after just 15 minutes (Figure 4II:b, f) with no change in the nuclear position of the X chromosome (Figure 4II:j). The shape of any aberrant, herniated, invaginated nuclei was also restored to more smoothened ellipsoid shapes after the 48-hour treatment with FTI-277 (data not shown).

There is evidence that rapid chromosome repositioning using the serum removal assay is elicited through nuclear motor activity, probably involving NM1β 42 . We used an antibody to NM1β that we and others have employed previously 42 , and analyzed the nuclear distribution of this protein in HGPS cells. In control fibroblasts, NM1β is distributed homogenously or as fine punctuate foci throughout the nucleoplasm with a concentration at the nuclear periphery and the nucleolus (Figure 5a) 42 . The distribution of NM1β in HGPS nuclei is very different to that found in control proliferating cells, being much more like the distribution observed in non-proliferating control cells 42 ; there is some nucleolar anti-NM1β staining but, in addition, NM1β is localized in large aggregates towards the interior of the nucleus, without any localization at the nuclear periphery and some weak staining in the nucleoplasm (Figure 5b). Most of the proliferating HGPS cells displayed NM1β as aggregates (85.1%; Figure 5a, Table 1) but when they were treated with FTI-277 for 48 hours, 73.7% of them displayed a normal distribution of NM1β (Figure 5c, Table 1), while only 18.6% of treated HGPS cells displayed NM1β aggregates (Table 1).

To determine if the NM1β distribution was different in HGPS cells that had been made quiescent, serum starved HGPS fibroblasts were also analyzed using indirect immunofluorescence with the anti-NM1β antibody. The distribution of NM1β in quiescent HGPS cells was similar to that in control cells made quiescent by serum starvation (Figure 6a), but also to that in proliferating HGPS cells (Figure 6c), which also had some aggregates of NM1β staining. In control cells that have been made quiescent and re-stimulated, the NM1β distribution returned to a proliferating-type distribution only 24 to 36 hours after the re-addition of serum 42 . Serum-starved HGPS cells (7 days) were re-stimulated with serum and samples taken at 24, 36 and 48 hours (Figure 6). In HGPS cells there was no significant difference in NM1β distribution (aggregates) in proliferating cells (85.1%; Figure 6a), in quiescent cells (77.3%; Figure 6c) or in cells re-stimulated with serum and fixed after 24 hours (71.9%; Figure 6e), 36 hours (83.3%; Figure 6g) and 48 hours (76.4%; Figure 6i). These data are given in Table 1 and demonstrate that the cells were not responding to growth factor cues with respect to NM1β, as occurs in control cells. However, if HGPS cells are treated with FTI and GGTI together for 48 hours, the distribution of NM1β becomes very similar to control cells, with more staining throughout the nucleoplasm and a concentration at the nuclear periphery and nucleolus (Figure 7a, Table 2). When the treated HGPS cells are made quiescent for 7 days, the distribution of NM1β in these cells is typical for a non-proliferating control culture, with large aggregates of NM1β. When quiescent cultures of HGPS cells treated with FTIs were re-stimulated by the re-addition of serum, the cells showed a more normal distribution of NM1β, with nucleoplasmic, nucleolar and nuclear rim staining. We observed increases in the normal distribution of NM1β from 2.1% in quiescent HGPS cells to 35% at 24 hours after re-stimulation (Figure 7e, Table 2), 51.6% at 36 hours (Figure 7g, Table 2) and 64% by 48 hours (Figure 7i, Table 2). This implies that the cells were able to respond to growth factors after the FTI treatment and re-position chromosomes using a nuclear motor activity that, in a further experiment, was blocked by the nuclear myosin inhibitor BDM (2,3-butanedione-2-monoxime; Figure S2 in Additional file 1), showing that we restored functional motor activity in HGPS cells for chromosome relocation.

Control human HDFs, 2DD 62 , and HDFs derived from three classic HGPS patients (cell lines AG11513, AG01972C and AG11498, Coricell Repositories) were cultured in 15% FBS in DMEM with passaging twice every week. The proliferative status of the cell cultures was assessed by the presence of pKi-67 in cells 53 using indirect immunofluorescence. In 2DD HDFs the pKi-67 fraction of cells ranged from 40% to 20%. For HGPS cells the range was 70% to 2% over time in culture, demonstrating hyperproliferation in the HGPS cells, as has been determined before 63 . To elicit a chromosome movement response, cells were grown in 15% FBS for 2 days and then placed in 0.5% FBS in DMEM for 5 minutes, 10 minutes, 15 minutes, 30 minutes or 7 days. For serum restoration experiments the cells were cultured in 15% FBS in DMEM for 2 days, then placed in 0.5% FBS in DMEM for 7 days, which was replaced with 15% FBS in DMEM for 8 hours, 24 hours, 32 hours and 36 hours.

Inhibitors of farnesylation and prenylation used in this study were FTI-277 (Calbiochem-Novabiochem, La Jolla, CA, USA) and GGTI-2147 (Calbiochem-Novabiochem). Both inhibitors were dissolved in DMSO and stored at -20°C. HGPS HDFs were seeded at 2 × 105 cells in 10 cm² tissue culture dishes and then allowed to grow for at least 2 days in 15% FBS in DMEM. Cells were incubated with 2.5 μM final concentration of FTI-277 and 2.5 μM of GGTI-2147 in 15% FBS in DMEM for 48 hours.

Myosin polymerization was inhibited by treating cells with 10 mM BDM (Calbiochem) for 15 minutes 42 .

For the two-dimensional FISH assay, HDFs were harvested and placed in hypotonic buffer (0.075 M KCl, w/v) for 15 minutes at room temperature and spun at 400 g. The cells were fixed in 3:1 (v/v) methanol:acetic acid between five and seven times before being dropped onto humidified glass microscope slides. After dehydration in an ethanol row the cells were denatured in 70% formamide, 2× sodium saline citrate buffer (SSC), pH 7, at 70°C for 2 minutes. Chromosome paints for chromosomes 10, 18 and × were amplified from flow-sorted whole chromosome templates and labeled with biotin-16-dUTP by Degenerate OligoPrimer-PCR 64 . We used 200 to 400 μg chromosome paint, 7 μg C0t-1 DNA and 3 μg herring sperm per slide. Hybridization was performed in a humidified chamber for 18 to 24 hours at 37°C. The slides were washed in 50% formamide, 2× SSC, pH 7, at 45°C for 15 minutes, followed by 0.1× SSC prewarmed to 60°C for 15 minutes at 45°C. Labeled hybridized probes were detected with streptavidin-cyanine 3 (Amersham Life Science Ltd, GE Healthcare UK Ltd, Little Chalfont, Buckinghamshire, UK).

For the three-dimensional FISH assay, fibroblasts were washed in PBS and then fixed in 4% paraformaldehyde (w/v) in PBS for 10 minutes. A permeabilization step was performed with 0.5% Triton-X100 (v/v) and 0.5% saponin (w/v) in PBS for 20 minutes. The cells were then incubated in 20% glycerol in PBS for 30 minutes prior to being snap-frozen in liquid nitrogen. The cells were repeatedly frozen and thawed up to five times. After the freeze-thaw cycles, the cells were washed in PBS for at least 30 minutes and then incubated in 0.1 N HCl for 10 minutes. The cells were then washed in 2× SSC for 15 minutes and incubated in 50% formamide, 2× SSC, pH 7.0, overnight. For denaturation, cells were incubated at 73 to 76°C in 70% formamide, 2× SSC, pH 7, solution for 3 minutes and then were immediately transferred to 50% formamide, 2× SSC, pH 7, solution for 1 minute at the same temperature. All the subsequent steps were as for two-dimensional FISH.

To reveal proliferating cells, rabbit anti-Ki-67 antibody (1:1,500; Dako, Ely, Cambridgeshire, UK). or mouse anti-pKi67 were incubated with the fixed cells. Secondary antibodies employed were swine anti-rabbit conjugated to either fluorescein isothiocynate (FITC; 1:30; Dako) or tetrarhodamine isothiocynate (TRITC; 1:30; Dako) and donkey anti-mouse conjugated to FITC. Rabbit anti-NM1β (1:200; Sigma-Aldrich, Poole, Dorset, UK) was used to reveal NM1β distribution with swine anti-rabbit conjugated to TRITC (Dako) as the secondary antibody.

For two-dimensional FISH analyses, digital grey-scale images of random nuclei were captured using a Photometrics cooled CCD camera on a Leica fluorescence microscope (Leitz DMRB) using a Plan Fluotar 100× oil immersion lens and Digital Scientific Smart Capture software. The images were run through a simple erosion script in IPLab spectrum software as described in 38 . DAPI images of the nucleus were outlined and divided into 5 concentric shells of equal area, the first shell being most peripheral and the innermost denoting the interior of the nucleus. The script measures the pixel intensity of DAPI and the chromosome probe in these five shells. The probe signal was normalized by dividing the percentage of the probe by the percentage of DAPI signal in each shell. Histograms were plotted with standard error bars representing standard error of the mean. Simple statistical analyses were performed using the unpaired two-tailed student's t-test using Microsoft excel.

For three-dimensional FISH analyses, images of nuclei were captured using a Nikon confocal laser scanning microscope (TE2000-S) equipped with a 60×/1.49 Nikon Apo oil immersion objective. The microscope was controlled by Nikon confocal microscope C1 (EZ-C1) software version 3.00. Stacks of optical sections with an axial distance of 0.2 μm were collected from 20 random nuclei. Stacks of 8-bit gray-scale two-dimensional images were obtained with a pixel dwell of 4.56 and eight averages were taken for each optical image. The positioning of chromosomes in relation to the nuclear periphery was assessed by performing measurements using Imaris Software (Bitplane Scientific Solutions, Zurich, CH-8048, Switzerland) whereby the distance in micrometers between the geometric center of each chromosome territory and the nearest nuclear periphery, as determined by DAPI staining, was measured in three dimensions. These data were not normalized for size but when the data were normalized by dividing by the length of the major axis plus the length of the minor axis divided by 2 or the length of the major axis alone, the relative positions of the individual chromosomes in frequency distributions did not change. Frequency distribution curves were plotted with the distance between the geometric center of chromosome territory and the nearest nuclear periphery on the x-axis in actual micrometers and the frequency on the y-axis.