Skip to main content
Dryad

A dominant-negative SOX18 mutant disrupts multiple regulatory layers essential to transcription factor activity

Cite this dataset

Lou, Jieqiong et al. (2021). A dominant-negative SOX18 mutant disrupts multiple regulatory layers essential to transcription factor activity [Dataset]. Dryad. https://doi.org/10.5061/dryad.h9w0vt4gn

Abstract

Few genetically dominant mutations involved in human disease have been fully explained at the molecular level. In cases where the mutant gene encodes a transcription factor, the dominant-negative mode of action of the mutant protein is particularly poorly understood. Here, we studied the genome-wide mechanism underlying a dominant-negative form of the SOX18 transcription factor (SOX18RaOp) responsible for both the classical mouse mutant Ragged Opossum and the human genetic disorder Hypotrichosis-Lymphedema-Telangiectasia-Renal Syndrome. Combining three single-molecule imaging assays in living cells together with genomics and proteomics analysis, we found that SOX18RaOp disrupts the system through an accumulation of molecular interferences which impair several functional properties of the wild-type SOX18 protein, including its target gene selection process. The dominant-negative effect is further amplified by poisoning the interactome of its wild-type counterpart, which perturbs regulatory nodes such as SOX7 and MEF2C. Our findings explain in unprecedented detail the multi-layered process that underpins the molecular aetiology of dominant-negative transcription factor function.

Methods

All microscopy measurements were performed on an Olympus FV3000 laser scanning microscope coupled to an ISS A320 Fast FLIM box for fluorescence fluctuation data acquisition. A 60 X water immersion objective 1.2 NA was used for all experiments and live HeLa cells were imaged at 37 oC in 5 % CO2. For single color fluorescence fluctuation spectroscopy experiments (e.g., N&B) SOX7 (Fig. S9A and D), SOX18 (Fig. 5C and D, and Fig. S9E), SOX18 DNA-binding and homodimerization mutants (SOX18AH1 and SOX18DIM; Fig. S11A-D) and SOX18 dominant-negative mutant SOX18RaOp (Fig. 5C and D) were labelled 15 minutes prior to imaging via direct addition of 1 µM of JF549 Halo-tag dye, where JF549 was excited by a solid-state laser diode operating at 561 nm. The fluorescence signal was then directed through a 405/488/561 dichroic mirror to remove laser light and the JF549 emission collected through a 550 nm long pass filter by an external photomultiplier detector (H7422P-40 of Hamamatsu) fitted with a 620/50 nm bandwidth filter. A 100-frame scan acquisition of the JF549 signal was then collected by selecting a region of interest within a HeLa cell nucleus at zoom 20, which for a 256 x 256-pixel frame size resulted in a pixel size of 41 nm. The pixel dwell time was set to 12.5 µs, which resulted in a line time of 4.313 ms and a frame time of 1.108 s.

For the two-color experiments (e.g., cRICS) (Fig. 5E-H, Fig. S9B and C, and Fig. S11E-I) SOX7, SOX18, SOX18AH1, SOX18DIM and SOX18RaOp were co-labelled with 500 nM of JF549 as well as 500 nM of JF646 and these two dyes were excited by solid-state laser diodes operating at 561 nm and 640 nm (respectively). For the two-color experiment interrogating SNAP-SOX18/HALO-SOX18RaOp heterodimers, 500 nM of snap646 and 500 nM of JF549 was used (Fig. 5G and 5H). The fluorescence signal was then directed through a 405/488/561/640 dichroic mirror to remove laser light and the JF549 versus JF646 emission was detected by two internal GaAsp photomultiplier detectors set to between the following bandwidths: JF549 570-620 nm, JF646 650-750 nm. A two channel 100-frame scan acquisition of the J549 and JF646 signal was then collected simultaneously employing the same settings described above for the single channel experiment.

Number and brightness (N&B)

Number and brightness (N&B) analysis of single-channel frame scan acquisition was performed using a moment-based analysis as described in previously published papers (38–40). Briefly, in each pixel of the frame scan there is an intensity fluctuation that has an average intensity (first moment) and a variance (second moment). The ratio of these two properties describes the apparent brightness (B) of the molecules that give rise to the intensity fluctuation. In the case of a photon counting detector, the true molecular brightness (ɛ) of the molecules is related to the measured apparent brightness (B) by Equation 5:

 

B=ε+1

(5)

where 1 is the brightness contribution from the photon-counting detector (38). Calibration of the monomeric brightness of JF549, via measurement of monomeric SOX7 tagged with this fluorophore (Fig. S9A and D), enabled extrapolation of JF549 tagged SOX18 dimers and oligomers (Fig. S9E), as well as quantitation of the fraction of pixels within a given N&B acquisition that contain each of these species (e.g., fraction of dimer in Fig. 5D). Artefacts due to cell movement or cell bleaching were subtracted from acquired intensity fluctuations via use of a moving average algorithm (N = 10 frames). All brightness calculations were carried out from the NB page in SimFCS from the Laboratory for Fluorescence Dynamics (www.lfd.uci.edu). A robust regression and outlier removal (ROUT) outlier test using default settings (Q = 1 %) in GraphPad Prism was performed to identify and remove outliers. Statistical significance was assessed for data with two categories using a using a Mann-Whitney U-test (Figure 5D), whereas data with more than two categories was assessed by ANOVA and was log transformed in order to meet the assumptions of homoscedasticity and normality of residuals (Figure S11D) required for ANOVA analysis. Pairwise comparisons were performed using a Tukey post-hoc test.

Raster image correlation spectroscopy (RICS) and cross-RICS (cRICS)

Raster image correlation spectroscopy (RICS) as cross-RICS (cRICS) analysis of two-channel frame scan acquisitions was performed via use of the RICS and cRICS spatiotemporal correlation functions described in previously published papers (41–43). Briefly, the fluorescence intensity recorded in each frame (N = 100), within each channel (CH1 and CH2) was spatially correlated via application of the RICS function, and then spatially cross-correlated between channels (CC) via application of the cRICS function, alongside a moving average algorithm (N = 10 frames). In the case of SOX18, the three-dimensional (3D) RICS correlation profile in CH1 and CH2 fit to a 2-component 3D diffusion model (Fig. S11H), while the 3D cRICS cross-correlation profile CC fit a 1-component 3D diffusion model (Fig. S11I). In each case, the recovered amplitudes (G) as well as diffusion coefficients (D) were recorded to enable the fraction of SOX18 molecules in a hetero-dimeric or homo-dimeric complex to be calculated, and in either case, this complex’s mobility quantified. A robust regression and outlier removal (ROUT) outlier test using default settings (Q = 1 %) in GraphPad Prism was performed to identify and remove outliers. Statistical significance for data with two categories was assessed using a Mann-Whitney U-test (Figure 5F), data with more than two categories not involving comparison to an internal reference control were assessed using a Kruskal-Wallis test (Figures 5G and S11I), and data with more than two categories involving comparison to an internal reference control (e.g. SOX7 is monomeric and does not form homodimers) were assessed using a one-tailed t-test with a mu of 0 (expected SOX7 homodimer % = 0) (Figures 4Ci and 5H).

Funding

National Health and Medical Research Council, Award: GNT1120381, GNT1164000, GNT1104461

Australian Research Council, Award: DP180101387

National Health and Medical Research Council, Award: GNT1111169, GNT1124762

Jacob Haimson Beverly Mecklenburg Lectureship

Australian Research Council, Award: LE130100078