Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jul 2;11(1):419.
doi: 10.1186/s13104-018-3516-1.

Dynamic bimodal changes in CpG and non-CpG methylation genome-wide upon CGGBP1 loss-of-function

Divyesh Patel  1 , Manthan Patel  1 , Bengt Westermark  2 , Umashankar Singh  3
Affiliations

Dynamic bimodal changes in CpG and non-CpG methylation genome-wide upon CGGBP1 loss-of-function

Divyesh Patel et al. BMC Res Notes. .

Abstract

Objectives: Although CpG methylation is well studied, mechanisms of non-CpG methylation in mammals remains elusive. Studying proteins with non-CpG cytosine methylation-sensitive DNA-binding, such as human CGGBP1, can unveil cytosine methylation regulatory mechanisms. Here we have resequenced a published genome-wide bisulfite sequencing library and analyzed it at base level resolution. CpG, CHG and CHH (where H is any nucleotide other than G) methylation states in non-targeting or CGGBP1-targeting shmiR lentivirus-transduced cells have been analyzed to identify how CGGBP1 regulates CpG and non-CpG methylation.

Results: We report that CGGBP1 acts as a dynamic bimodal balancer of methylation. Both gain and loss of methylation observed upon CGGBP1 depletion were spatially overlapping at annotated functional regions and not identifiable with any sequence motifs but clearly associated with GC-skew. CGGBP1 depletion caused clustered methylation changes in cis, upstream of R-loop forming promoters. This was complemented by clustered occurrences of methylation changes in proximity of transcription start sites of known cytosine methylation regulatory genes, altered expression of which can regulate cytosine methylation in trans. Despite low coverage, our data provide reliable estimates of the spectrum of methylation changes regulated by CGGBP1 in all cytosine contexts genome-wide through a combination of cis and trans-acting mechanisms.

Keywords: CGGBP1; Cytosine methylation; GC-skew; Genome-wide bisulfite sequencing.

PubMed Disclaimer

Figures

Fig. 1
Fig. 1
CGGBP1 regulates cytosine methylation in a GC-content and cytosine context dependent manner. a Chromosome-wise distribution of cytosines exhibiting methylation changes (GoM and LoM) or no methylation changes (RoM and RuN). The similar distribution of the four methylation change states on all the chromosomes showed no major chromosomal preference for CGGBP1-regulation of cytosine methylation. b, c Measurement of intra-chromosomal variabilities in methylation states in S1 and S2 shows preference for GoM in G-rich R-bands (Giemsa-negative) and LoM in G-bands (Giemsa positive 100). In R-bands, the paired GoM and LoM events had closely related values like RoM and RuN events but GoM was significantly more than LoM (paired t-test p value = 2.239e−012) (b). However, at G-bands the methylation state was reversed and LoM was higher than GoM (p value = 0) (c). Similarly, while RuN was significantly higher in R-bands (paired t-test p value = 0) (b), RoM was higher in G-bands (paired t-test p value = 0.000636). For paired t-test, n = 780 R bands in B and n = 81 G bands in c. d–k The methylated and unmethylated fractions of cytosines in all three contexts have differential susceptibility to methylation change in absence of CGGBP1 function. d–f The context-wise distribution of methylated cytosines in S1 (d) underwent LoM wherein the proportion of CpG was lower and that of CHG and CHH were more than expected (e). From the same pool of cytosines methylated in S1 (d) that retained methylation upon CGGBP1 loss-of-function had a highly enriched CpG fraction and lower CHG and CHH fractions (f). g–i The context-wise distribution of unmethylated cytosines in S1 (g) that underwent GoM (h) also displayed an unexpected and disproportionate increase in CpG context. The context distribution amongst the cytosines that remained unmethylated upon CGGBP1 loss-of-function displayed a reduction in CpG context (i). A comparison of d and g clearly shows that the major fraction of CpG context was methylated in presence of CGGBP1 whereas CHH and CHG together comprised the most of unmethylated fractions. Comparison of e and h show that despite differences in the absolute numbers as well as relative abundance of the three contexts in methylated and unmethylated pools in S1 (d, g), the GoM (e) and LoM (h) cytosines were unexpectedly similar in magnitude with near identical context composition. If LoM and GoM were occurring randomly in the methylated and unmethylated pools of cytosines, then the magnitude and context distributions observed in d and g were proportionately expected in e and h respectively. Obs/Exp analyses of e and h against d and g revealed a highly significant unexpected composition of e and h (refer to Additional file 1: Table S6). j Plotting of the number of cytosines sequenced in S1 that underwent methylation change upon CGGBP1 depletion shows a disproportionate change in methylation states such that the GoM and LoM are quantitatively coincidental. k Conversely to j, the number of cytosines sequenced in S1 that resisted methylation change upon CGGBP1 depletion are disproportionately different and non-coincidental. All graphs are generated using GraphPad Prism7
Fig. 2
Fig. 2
Genomic regions dependent on CGGBP1 for stability of cytosine methylation have inter-strand G/C asymmetry. a A frequency plot of GC-skew calculated as {(G−C)/(G+C)} for all GoM (red solid line) and all LoM (blue solid line) regions showed a clear clustering of GoM and LoM regions into two groups; one peaking near − 0.5 and other around + 0.5. The distribution of the data could not be fitted with a single Gaussian curve, but with a sum of two Gaussian curves with very high confidence (Additional file 1: Table S10). When these datasets were split into L1-LINEs (nearly 20% of the GoM and LoM sets; dashed broken lines) and non-L1 regions (dotted lines), we observed a clear difference between the L1 sequences versus the rest with the GoM-L1 and LoM-L1 sequences exhibiting lesser GC-skew than the non-L1 GoM and LoM sequences. However, all of these could be fitted only with a sum of two Gaussian curves. b The GC-skew observed with the GoM-L1 and LoM-L1 sequences was unexpected as the L1 sequences from Repbase and NCBI L1 consensus showed an absolutely Gaussian distribution of GC-skew centred near zero. c, d The RoM and RuN regions did not display the GC-skew as seen for GoM and LoM regions. The GC-skew frequency for RoM and RuN was centred around zero in a binomial fashion. e–h GC-skew regions genome-wide are prone to methylation gain upon CGGBP1 depletion. The distribution of methylated cytosines centred at the middle of GC-skew regions displayed a binomial increase in methylation on both the strands in the absence of CGGBP1 function. This increase in methylation is highly specific and restricted to less than 1 kb flanks of the GC-skew regions genome-wide [6] with mean length of 747 ± 482 bp. e Negative GC-skew, methylation on bottom strand. f Negative GC-skew, methylation on top strand. G: Positive GC-skew, methylation on bottom strand. h Positive GC-skew, methylation on top strand. Red line = S2, blue line = S1. X axis represents genomic location from the centre of GC-skew regions. Y axis represents methylated cytosine counts in bins with sizes as indicated. Plots were generated using deepTools [8]
Fig. 3
Fig. 3
CGGBP1 regulates methylation at TSSs of the cytosine methylation regulatory genes. Cytosine methylation levels in 1 kb flank from TSSs of various transcripts the cytosine methylation regulatory genes (DNMT1, DNMT3A, TET2, AICDA, TDG, NEIL1, MBD4, APOBEC3H and APOBEC3G) was plotted for both the samples S1 and S2 and for both the strands (top and bottom). The cluster of cytosines exhibiting methylation change is highlighted by boxes with dashed lines. For each gene, direction of transcription is marked by arrowheads along-with ENSEMBL transcript ID. All transcripts in the regions are not shown as in Additional file 1: Fig. S7. Plots were generated using deepTools [8] and compiled in Keynote (Apple)
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Agarwal P, Collier P, Fritz MH-Y, Benes V, Wiklund HJ, Westermark B, et al. CGGBP1 mitigates cytosine methylation at repetitive DNA sequences. BMC Genomics. 2015;16:390. doi: 10.1186/s12864-015-1593-2. - DOI - PMC - PubMed
    1. Singh U, Bongcam-Rudloff E, Westermark B. A DNA sequence directed mutual transcription regulation of HSF1 and NFIX involves novel heat sensitive protein interactions. PLoS ONE. 2009;4:e5050. doi: 10.1371/journal.pone.0005050. - DOI - PMC - PubMed
    1. Singh U, Westermark B. CGGBP1–an indispensable protein with ubiquitous cytoprotective functions. Ups J Med Sci. 2015;120:219–232. doi: 10.3109/03009734.2015.1086451. - DOI - PMC - PubMed
    1. Agarwal P, Enroth S, Teichmann M, Jernberg Wiklund H, Smit A, Westermark B, et al. Growth signals employ CGGBP1 to suppress transcription of Alu-SINEs. Cell Cycle. 2016;15:1558–1571. doi: 10.4161/15384101.2014.967094. - DOI - PMC - PubMed
    1. Schmidt D, Schwalie PC, Wilson MD, Ballester B, Gonçalves A, Kutter C, et al. Waves of retrotransposon expansion remodel genome organization and CTCF binding in multiple mammalian lineages. Cell. 2012;148:335–348. doi: 10.1016/j.cell.2011.11.058. - DOI - PMC - PubMed

LinkOut - more resources

-