Fig 1.
Fluorescence lifetime measurements of chromatin condensation state in human umbilical vein endothelial cell nuclei.
(A) Fluorescence intensity confocal images (top) and mean fluorescence lifetime heat maps (bottom) of chromatin is measured in endothelial cell nuclei labeled with Hoechst 33342. Cells are treated with NaN3+2-DG for chromatin hypercondensation or with TSA for chromatin decondensation. Altered fluorescence intensity with treatments show differential chromatin condensation state, with more intense fluorescence arising from highly concentrated condensed chromatin. Mean fluorescence lifetime heat maps similarly indicate spatial arrangement of local fluorophore environments for labeled chromatin consistent with varying chromatin condensation state. Treatment with NaN3+2-DG results in more punctate regions of fluorescence intensity and shorter mean fluorescence lifetime (orange) relative to untreated controls, while TSA resulted in a significant reduction in punctate regions and longer mean fluorescence lifetime (blue). Scale bar is 10 μm. (B) The mean fluorescence lifetime of segmented nuclei for the various treatment conditions was calculated using Eq 2. Treatment with NaN3+2-DG resulted in a strong reduction in the mean fluorescence lifetime relative to untreated controls. By contrast, TSA treatment resulted in a dramatic increase in the mean fluorescence lifetime relative to untreated controls which indicated an increase in chromatin condensation state homogeneity throughout the cell nucleus. Error bars indicate standard error of the mean of pixel-to-pixel mean fluorescence lifetime differences of segmented nuclei in fields of view across multiple fields of view under each treatment condition (*** p<<0.001). Histograms and standard deviations are in S2 File.
Fig 2.
Dynamic light scattering measurements of in vitro λ-DNA solutions of varying condensation state.
Measurements of PEG 6000 (gray) and λ-DNA (green) alone indicate their location within the combined solutions. As we increase PEG concentration, initially we see a negligible effect on the λ-DNA diffusivity distribution (shades of blue). At 50 mg/mL, the solution is above a threshold concentration of PEG 6000 and we observe a reduction in the λ-DNA diffusivity distribution, including a sharp decrease beyond the overlap concentration for PEG 6000 at 100 mg/mL (shades of yellow-orange). The initial reduction stems from the polymer-and-salt-induced (psi or ψ) condensation by macromolecular crowding-induced depletion forces. We show the regime over which λ-DNA is condensed and decondensed along with the location of the PEG population. Distributions are derived from 10–15 runs per individual measurements and averages of several measurements.
Fig 3.
Fluorescence lifetime measurements of in vitro λ-DNA solutions of varying condensation state.
As in the DLS experiments, we observe a dramatic reduction in the mean fluorescence lifetime above the threshold PEG 6000 concentration (~50 mg/mL; p<0.01) that is maintained at higher concentrations (shades of yellow-orange symbols). Interestingly, despite the increase in viscosity that occurs with increasing PEG concentration (including the sharp increase in trend above the overlap concentration at 100 mg/mL) we see no further statistical change in the mean fluorescence lifetime despite the dependence of the fluorescence lifetime on local viscosity. Error bars reflect standard deviation. Statistical significance based on Student’s t-test with the 0 mg/mL PEG, with **p<0.025 and ***p<0.01.
Fig 4.
Fluorescence lifetime measurements of in vitro λ-DNA solutions of varying ionic strength solutions.
The mean fluorescence lifetime of solutions of λ-DNA with varying concentration of MgCl2 shows no statistical dependence on ionic strength. Across a wide distribution of salt concentration varying over three orders of magnitude we see no statistically significant effect on the mean fluorescence lifetime, indicating it is not strongly influenced by salt concentration. Statistical comparisons made by Student’s t-test, with no statistical difference between solutions.
Fig 5.
Fluorescence lifetime measurements of in vitro λ-DNA solutions of varying viscosity.
We determine the mean fluorescence lifetime dependence of Hoechst 33342 bound to λ-DNA in solutions of varying viscosity glycerol-ethylene glycol solutions. We see a strong dependence of the mean fluorescence lifetime on viscosity over the range here. Viscosity measurements for the glycerol-ethylene solutions were determined using a Discovery Hybrid Rheometer-2. Statistical comparisons made by Student’s t-test, with *p<0.05, **p<0.025, ***p<0.01 and ****p<0.001. All statistical comparisons of the magnitudes are with the previous point unless otherwise indicated by lines. The viscosity power-law fit dependence of the mean fluorescence lifetime, based on the known phenomenological relationship between fluorescence lifetime and viscosity, was determined using ANOVA (p<<<0.001).
Fig 6.
Fluorescence lifetime spatial distribution around nucleoli in endothelial cell nuclei.
(A) An overlay of a confocal image of HUVECs labelled for chromatin with Hoechst 33342 (blue, B) and nucleoli with transfected nucleolar protein GFP-Fibrillarin (green, C). (D) The spatially resolved mean fluorescence lifetime of the same nucleus shows largely decondensed chromatin within the interior of the nucleolus (deep blue regions). Further, the nucleoli are bounded by condensed chromatin (yellow-orange regions) consistent with heterochromatin-bound nucleolar regions. Other condensed chromatin regions show correspondence with the brighter, more punctate regions of chromatin fluorescence (B) consistent with heterochromatin. Scale bar is 5 μm.
Fig 7.
Fluorescence lifetime spatial distribution around of endothelial cells labelled for heterochromatin marker H3K9me3 with immunocytochemistry.
A confocal image of a HUVEC nucleus labelled for chromatin with Hoechst 33342 (blue, A) and constitutive heterochromatin marker histone H3 tri-methylated at lysine residue 9 (H3K9me3) by immunocytochemistry (red, B) with merge (C). (D) The spatially resolved mean fluorescence lifetime of the same nucleus shows a spatially heterogeneous lifetime distribution. (E) Binning of fluorescence lifetimes rather than a continuous scale (from D) allows discrete segregation for overlay. (F) Very low lifetime regimes (red) overlaid with enhanced H3K9me3 signal (white) shows alignment of these regions. Intermediate lifetimes (green, G) show some small colocalization, but high lifetime regimes (blue, H) show no overlap with H3K9me3 staining. Scale bar is 5 μm.
Fig 8.
Changes in chromatin condensation state in the nuclear interior impact the local viscosity which strongly influence the fluorescence lifetime.
In most cells, the nucleus has regions of highly condensed chromatin (A). While the concentration of chromatin in the nuclear interior is unchanged upon decondensation, the nuclear interior becomes more viscous due to the reduction in densely packed chromatin (B). As such, the chromatin in condensed regions has a low fluorescence lifetime (orange in A) and chromatin in decondensed regions has a high fluorescence lifetime (blue in B). (C) Decondensed chromatin undergoes fluctuations influenced by frictional drag (ζ) from the surrounding environment. Regions of higher viscosity arising from chromatin decondensation have increased mean fluorescence lifetimes.