Limits...
Impaired in vivo binding of MeCP2 to chromatin in the absence of its DNA methyl-binding domain.

Stuss DP, Cheema M, Ng MK, Martinez de Paz A, Williamson B, Missiaen K, Cosman JD, McPhee D, Esteller M, Hendzel M, Delaney K, Ausió J - Nucleic Acids Res. (2013)

Bottom Line: However, the specific interactions of MeCP2 with methylated or non-methylated chromatin regions and the structural characteristics of the resulting DNA associations in vivo remain poorly understood.Although a fraction of ΔMeCP2 is found associated with nucleosomes, its interaction with chromatin is transient and weak.Thus, our results unequivocally demonstrate that in vivo the MBD of MeCP2 together with its adjacent region in the N-terminal domain are critical for the proper interaction of the protein with chromatin, which cannot be replaced by any other of its protein domains.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, University of Victoria, British Columbia, V8W 2Y2, Canada.

ABSTRACT
MeCP2 is a methyl-CpG-binding protein that is a main component of brain chromatin in vertebrates. In vitro studies have determined that in addition to its specific methyl-CpG-binding domain (MBD) MeCP2 also has several chromatin association domains. However, the specific interactions of MeCP2 with methylated or non-methylated chromatin regions and the structural characteristics of the resulting DNA associations in vivo remain poorly understood. We analysed the role of the MBD in MeCP2-chromatin associations in vivo using an MeCP2 mutant Rett syndrome mouse model (Mecp2(tm1.1Jae)) in which exon 3 deletion results in an N-terminal truncation of the protein, including most of the MBD. Our results show that in mutant mice, the truncated form of MeCP2 (ΔMeCP2) is expressed in different regions of the brain and liver, albeit at 50% of its wild-type (wt) counterpart. In contrast to the punctate nuclear distribution characteristic of wt MeCP2, ΔMeCP2 exhibits both diffuse nuclear localization and a substantial retention in the cytoplasm, suggesting a dysfunction of nuclear transport. In mutant brain tissue, neuronal nuclei are smaller, and ΔMeCP2 chromatin is digested faster by nucleases, producing a characteristic nuclease-resistant dinucleosome. Although a fraction of ΔMeCP2 is found associated with nucleosomes, its interaction with chromatin is transient and weak. Thus, our results unequivocally demonstrate that in vivo the MBD of MeCP2 together with its adjacent region in the N-terminal domain are critical for the proper interaction of the protein with chromatin, which cannot be replaced by any other of its protein domains.

Show MeSH

Related in: MedlinePlus

(A–C) ΔMeCP2 distribution is altered in Neuro-2A cells and does not concentrate in heterochromatin. (A) Live Neuro-2A cells expressing GFP-tagged wt or ΔMeCP2 from transfected plasmid constructs. Nuclei were counterstained with Hoechst 33 342. wt GFP-MeCP2 is restricted to the nucleus and co-localizes with heterochromatin puncta. Mutant GFP-ΔMeCP2 is diffuse throughout the cell but reaches its highest concentration in the nucleus. Dark patches in the nucleus co-localize with bright Hoechst puncta, indicating reduced levels in heterochromatin. Red arrows indicate end points of the line segments plotted in (B). Scale bar = 20 µm. (B) Line plots of greyscale intensities for GFP-MeCP2 (left axis) and Hoechst 33342 (right axis). Although GFP-MeCP2 peak intensities coincide with Hoechst staining and are brightest at heterochromatin puncta, this correlation is not observed with GFP-ΔMeCP2. Greyscale values were squared to highlight subtle differences in staining patterns. (C) The ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt. The coefficient of variation was obtained for both GFP and Hoechst greyscale intensity over the entire nucleus. In the wt, a ratio value >1 is observed, indicating a greater variance in the distribution of GFP-MeCP2 relative to Hoechst staining. The inverse is observed in the mutant (Mann–Whitney U-test, P < 0.0001). (D and E) Nuclear size is reduced in cortical neurons of ΔMeCP2 mutant mice. (D) Wide field fluorescence images of Hoechst 33342-stained nuclei. Mutant brain tissue contains a greater proportion of small nuclei. Scale bar = 10 µm. (E) The average nuclear area is reduced by 20.2% in 6-week-old mutant males (85.00 ± 1.04 µm2 in wt versus 67.84 ± 1.71 µm2 in mutant, t4 = 9.284, P < 0.01). (F) Western blot analysis of histone H1 and histone H3 with an IRDye fluorescent antibody detection. CE: chicken erythrocyte histones used as a histone marker. Bwt: nuclear brain from wt mice; BΔ: nuclear brain from Mecp2tm1.1Jae mutant; Lwt: nuclear liver from wt mice; LΔ: nuclear liver from Mecp2tm1.1Jae mutant. The numbers underneath the image represent the percentile of MeCP2 with respect to brain wt taken as 100% (in red) and of H1 with respect to liver wt taken as 100% (in black).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3643609&req=5

gkt213-F4: (A–C) ΔMeCP2 distribution is altered in Neuro-2A cells and does not concentrate in heterochromatin. (A) Live Neuro-2A cells expressing GFP-tagged wt or ΔMeCP2 from transfected plasmid constructs. Nuclei were counterstained with Hoechst 33 342. wt GFP-MeCP2 is restricted to the nucleus and co-localizes with heterochromatin puncta. Mutant GFP-ΔMeCP2 is diffuse throughout the cell but reaches its highest concentration in the nucleus. Dark patches in the nucleus co-localize with bright Hoechst puncta, indicating reduced levels in heterochromatin. Red arrows indicate end points of the line segments plotted in (B). Scale bar = 20 µm. (B) Line plots of greyscale intensities for GFP-MeCP2 (left axis) and Hoechst 33342 (right axis). Although GFP-MeCP2 peak intensities coincide with Hoechst staining and are brightest at heterochromatin puncta, this correlation is not observed with GFP-ΔMeCP2. Greyscale values were squared to highlight subtle differences in staining patterns. (C) The ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt. The coefficient of variation was obtained for both GFP and Hoechst greyscale intensity over the entire nucleus. In the wt, a ratio value >1 is observed, indicating a greater variance in the distribution of GFP-MeCP2 relative to Hoechst staining. The inverse is observed in the mutant (Mann–Whitney U-test, P < 0.0001). (D and E) Nuclear size is reduced in cortical neurons of ΔMeCP2 mutant mice. (D) Wide field fluorescence images of Hoechst 33342-stained nuclei. Mutant brain tissue contains a greater proportion of small nuclei. Scale bar = 10 µm. (E) The average nuclear area is reduced by 20.2% in 6-week-old mutant males (85.00 ± 1.04 µm2 in wt versus 67.84 ± 1.71 µm2 in mutant, t4 = 9.284, P < 0.01). (F) Western blot analysis of histone H1 and histone H3 with an IRDye fluorescent antibody detection. CE: chicken erythrocyte histones used as a histone marker. Bwt: nuclear brain from wt mice; BΔ: nuclear brain from Mecp2tm1.1Jae mutant; Lwt: nuclear liver from wt mice; LΔ: nuclear liver from Mecp2tm1.1Jae mutant. The numbers underneath the image represent the percentile of MeCP2 with respect to brain wt taken as 100% (in red) and of H1 with respect to liver wt taken as 100% (in black).

Mentions: To look in more detail at the nuclear distribution of ΔMeCP2 with respect to wt MeCP2 in live cells, particularly in light of potential fixation artifacts, Neuro-2A cells were transfected with GFP-MeCP2 and GFP-ΔMeCP2 plasmid expression constructs. As seen in Figure 4A and B, the wt GFP-MeCP2 exhibits enrichment in the pericentric heterochomatin. By contrast, GFP-ΔMeCP2 exhibits pan-cellular distribution and seems to be homogeneous throughout the nucleoplasm unlike either wt protein or the distribution of DNA as revealed by Hoechst staining (Figure 4A). This is similar to what was observed immunohistochemically with the native ΔMeCP2 in brain tissue sections (Figure 3A). The absence of correlation between DNA distribution (Hoechst staining) and GFP-ΔMeCP2 is illustrated quantitatively in the line-scan (Figure 4B). Furthermore, the ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt (Figure 4C). Moreover, in agreement with what was observed in cortical neurons in Figure 3, a substantial component of GFP-ΔMeCP2 is distributed in the cytoplasm.Figure 4.


Impaired in vivo binding of MeCP2 to chromatin in the absence of its DNA methyl-binding domain.

Stuss DP, Cheema M, Ng MK, Martinez de Paz A, Williamson B, Missiaen K, Cosman JD, McPhee D, Esteller M, Hendzel M, Delaney K, Ausió J - Nucleic Acids Res. (2013)

(A–C) ΔMeCP2 distribution is altered in Neuro-2A cells and does not concentrate in heterochromatin. (A) Live Neuro-2A cells expressing GFP-tagged wt or ΔMeCP2 from transfected plasmid constructs. Nuclei were counterstained with Hoechst 33 342. wt GFP-MeCP2 is restricted to the nucleus and co-localizes with heterochromatin puncta. Mutant GFP-ΔMeCP2 is diffuse throughout the cell but reaches its highest concentration in the nucleus. Dark patches in the nucleus co-localize with bright Hoechst puncta, indicating reduced levels in heterochromatin. Red arrows indicate end points of the line segments plotted in (B). Scale bar = 20 µm. (B) Line plots of greyscale intensities for GFP-MeCP2 (left axis) and Hoechst 33342 (right axis). Although GFP-MeCP2 peak intensities coincide with Hoechst staining and are brightest at heterochromatin puncta, this correlation is not observed with GFP-ΔMeCP2. Greyscale values were squared to highlight subtle differences in staining patterns. (C) The ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt. The coefficient of variation was obtained for both GFP and Hoechst greyscale intensity over the entire nucleus. In the wt, a ratio value >1 is observed, indicating a greater variance in the distribution of GFP-MeCP2 relative to Hoechst staining. The inverse is observed in the mutant (Mann–Whitney U-test, P < 0.0001). (D and E) Nuclear size is reduced in cortical neurons of ΔMeCP2 mutant mice. (D) Wide field fluorescence images of Hoechst 33342-stained nuclei. Mutant brain tissue contains a greater proportion of small nuclei. Scale bar = 10 µm. (E) The average nuclear area is reduced by 20.2% in 6-week-old mutant males (85.00 ± 1.04 µm2 in wt versus 67.84 ± 1.71 µm2 in mutant, t4 = 9.284, P < 0.01). (F) Western blot analysis of histone H1 and histone H3 with an IRDye fluorescent antibody detection. CE: chicken erythrocyte histones used as a histone marker. Bwt: nuclear brain from wt mice; BΔ: nuclear brain from Mecp2tm1.1Jae mutant; Lwt: nuclear liver from wt mice; LΔ: nuclear liver from Mecp2tm1.1Jae mutant. The numbers underneath the image represent the percentile of MeCP2 with respect to brain wt taken as 100% (in red) and of H1 with respect to liver wt taken as 100% (in black).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3643609&req=5

gkt213-F4: (A–C) ΔMeCP2 distribution is altered in Neuro-2A cells and does not concentrate in heterochromatin. (A) Live Neuro-2A cells expressing GFP-tagged wt or ΔMeCP2 from transfected plasmid constructs. Nuclei were counterstained with Hoechst 33 342. wt GFP-MeCP2 is restricted to the nucleus and co-localizes with heterochromatin puncta. Mutant GFP-ΔMeCP2 is diffuse throughout the cell but reaches its highest concentration in the nucleus. Dark patches in the nucleus co-localize with bright Hoechst puncta, indicating reduced levels in heterochromatin. Red arrows indicate end points of the line segments plotted in (B). Scale bar = 20 µm. (B) Line plots of greyscale intensities for GFP-MeCP2 (left axis) and Hoechst 33342 (right axis). Although GFP-MeCP2 peak intensities coincide with Hoechst staining and are brightest at heterochromatin puncta, this correlation is not observed with GFP-ΔMeCP2. Greyscale values were squared to highlight subtle differences in staining patterns. (C) The ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt. The coefficient of variation was obtained for both GFP and Hoechst greyscale intensity over the entire nucleus. In the wt, a ratio value >1 is observed, indicating a greater variance in the distribution of GFP-MeCP2 relative to Hoechst staining. The inverse is observed in the mutant (Mann–Whitney U-test, P < 0.0001). (D and E) Nuclear size is reduced in cortical neurons of ΔMeCP2 mutant mice. (D) Wide field fluorescence images of Hoechst 33342-stained nuclei. Mutant brain tissue contains a greater proportion of small nuclei. Scale bar = 10 µm. (E) The average nuclear area is reduced by 20.2% in 6-week-old mutant males (85.00 ± 1.04 µm2 in wt versus 67.84 ± 1.71 µm2 in mutant, t4 = 9.284, P < 0.01). (F) Western blot analysis of histone H1 and histone H3 with an IRDye fluorescent antibody detection. CE: chicken erythrocyte histones used as a histone marker. Bwt: nuclear brain from wt mice; BΔ: nuclear brain from Mecp2tm1.1Jae mutant; Lwt: nuclear liver from wt mice; LΔ: nuclear liver from Mecp2tm1.1Jae mutant. The numbers underneath the image represent the percentile of MeCP2 with respect to brain wt taken as 100% (in red) and of H1 with respect to liver wt taken as 100% (in black).
Mentions: To look in more detail at the nuclear distribution of ΔMeCP2 with respect to wt MeCP2 in live cells, particularly in light of potential fixation artifacts, Neuro-2A cells were transfected with GFP-MeCP2 and GFP-ΔMeCP2 plasmid expression constructs. As seen in Figure 4A and B, the wt GFP-MeCP2 exhibits enrichment in the pericentric heterochomatin. By contrast, GFP-ΔMeCP2 exhibits pan-cellular distribution and seems to be homogeneous throughout the nucleoplasm unlike either wt protein or the distribution of DNA as revealed by Hoechst staining (Figure 4A). This is similar to what was observed immunohistochemically with the native ΔMeCP2 in brain tissue sections (Figure 3A). The absence of correlation between DNA distribution (Hoechst staining) and GFP-ΔMeCP2 is illustrated quantitatively in the line-scan (Figure 4B). Furthermore, the ratio of the variability of GFP-ΔMeCP2 staining relative to Hoechst staining is the inverse of that observed in the wt (Figure 4C). Moreover, in agreement with what was observed in cortical neurons in Figure 3, a substantial component of GFP-ΔMeCP2 is distributed in the cytoplasm.Figure 4.

Bottom Line: However, the specific interactions of MeCP2 with methylated or non-methylated chromatin regions and the structural characteristics of the resulting DNA associations in vivo remain poorly understood.Although a fraction of ΔMeCP2 is found associated with nucleosomes, its interaction with chromatin is transient and weak.Thus, our results unequivocally demonstrate that in vivo the MBD of MeCP2 together with its adjacent region in the N-terminal domain are critical for the proper interaction of the protein with chromatin, which cannot be replaced by any other of its protein domains.

View Article: PubMed Central - PubMed

Affiliation: Department of Biology, University of Victoria, British Columbia, V8W 2Y2, Canada.

ABSTRACT
MeCP2 is a methyl-CpG-binding protein that is a main component of brain chromatin in vertebrates. In vitro studies have determined that in addition to its specific methyl-CpG-binding domain (MBD) MeCP2 also has several chromatin association domains. However, the specific interactions of MeCP2 with methylated or non-methylated chromatin regions and the structural characteristics of the resulting DNA associations in vivo remain poorly understood. We analysed the role of the MBD in MeCP2-chromatin associations in vivo using an MeCP2 mutant Rett syndrome mouse model (Mecp2(tm1.1Jae)) in which exon 3 deletion results in an N-terminal truncation of the protein, including most of the MBD. Our results show that in mutant mice, the truncated form of MeCP2 (ΔMeCP2) is expressed in different regions of the brain and liver, albeit at 50% of its wild-type (wt) counterpart. In contrast to the punctate nuclear distribution characteristic of wt MeCP2, ΔMeCP2 exhibits both diffuse nuclear localization and a substantial retention in the cytoplasm, suggesting a dysfunction of nuclear transport. In mutant brain tissue, neuronal nuclei are smaller, and ΔMeCP2 chromatin is digested faster by nucleases, producing a characteristic nuclease-resistant dinucleosome. Although a fraction of ΔMeCP2 is found associated with nucleosomes, its interaction with chromatin is transient and weak. Thus, our results unequivocally demonstrate that in vivo the MBD of MeCP2 together with its adjacent region in the N-terminal domain are critical for the proper interaction of the protein with chromatin, which cannot be replaced by any other of its protein domains.

Show MeSH
Related in: MedlinePlus