Trim24 and Trim33 Play a Role in Epigenetic Silencing of Retroviruses in Embryonic Stem Cells

Abstract

Embryonic stem cells (ESC) have the ability to epigenetically silence endogenous and exogenous retroviral sequences. Trim28 plays an important role in establishing this silencing, but less is known about the role other Trim proteins play. The Tif1 family is a sub-group of the Trim family, which possess histone binding ability in addition to the distinctive RING domain. Here, we have examined the interaction between three Tif1 family members, namely Trim24, Trim28 and Trim33, and their function in retroviral silencing. We identify a complex formed in ESC, comprised of these three proteins. We further show that when Trim33 is depleted, the complex collapses and silencing efficiency of both endogenous and exogenous sequences is reduced. Similar transcriptional activation takes place when Trim24 is depleted. Analysis of the H3K9me3 chromatin modification showed a decrease in this repressive mark, following both Trim24 and Trim33 depletion. As Trim28 is an identified binding partner of the H3K9 methyltransferase ESET, this further supports the involvement of Trim28 in the complex. The results presented here suggest that a complex of Tif1 family members, each of which possesses different specificity and efficiency, contributes to the silencing of retroviral sequences in ESC.

Keywords: Tif1 family; Trim24; Trim28; Trim33; embryonic stem cells (ESC); endogenous retroviruses (ERVs); epigenetic silencing; murine leukemia virus (MLV).

Figure 1

Tif1 family members are co-expressed and bind to each other in embryonic stem cells (ESCs). (a) Whole-cell protein extract co-immune precipitation (co-IP) and Western blot analysis with the indicated antibodies; (b) ESCs were grown with retinoic acid RA and no 2i/Lif for the indicated times, RNA expression levels of the indicated genes were measured by RT-qPCR and normalized to UBC control. NIH3T3 fibroblasts are used as differentiated control n = 3; (c) protein extract from ESCs treated with RA for 0, 4 or 8 days, and NIH3T3 cells were subjected to western blot analysis (top), showing downregulation of Trim24 protein levels after differentiation onset. Normalized density measurements are displayed (bottom).

Figure 2

The depletion of Trim33 disrupts the binding of Trim24 to Trim28. (a) Whole-cell protein extract was generated from control or Trim33 KD ESCs and subjected to co-IP assay using anti Trim24 antibody flowed by Western blot analysis. Binding between Trim28 and Trim24 was measured by the bend intensity of Trim28/Trim24 in the coIP lanes. (b) A statistical representation of 3 biological repeats of the experiment described in (a) Student’s t-test p < 0.001; n = 3. (c). RT-qPCR expression analysis to verify depletion in KD cells normalized to UBC control gene and WT ESC cells. No change was observed in Trim28 and the Oct4 pluripotency gene expression. The values are averages of three or more independent experiments ± SEM. ** p < 0.01; *** p < 0.001.

Figure 3

Tif1 family members contribute to the silencing of distinct ERVs subfamilies. (a) ERV expression change of KD vs. WT was measured using RT-qPCR. The three KD are ESCs expressing shRNA targeting either Trim24, Trim33 or Trim28 (n > 3). (b) H3K9me3 enrichment on ERVs was measured by chromatin immunoprecipitation (ChIP). Normalized log2 fold change of KD vs. WT is presented (n = 2) Statistical significance determined using the Holm-Sidak method, with alpha = 0.05, see also Figure S3.

Figure 4

An increase in expression and a decrease in epigenetic silencing of pro-viral sequences following Trim24 or Trim33 depletion. Flow analysis of ESCs expressing shRNA targeting either Scr, Trim24, Trim33 or none, infected by exogenous murine leukemia virus (MLV) vector containing a GFP reporter, with either PBS-P (a) or PBS-Q (b). The percentage of cells expressing the GFP reporter was measured and normalized to the expression levels in NIH3T3 cells (see Figure S4). Statistics analysis by two-tailed Student’s t-test; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; n > 5. (c) ChIP followed by RT-qPCR using H3K9me3 antibody and primers for pro-viral sequences. The ChIP was done separately on cells infected by either PBS-P (top) or PBS-Q (bottom). The location of primers on the proviral genome is illustrated in the diagram below. See Figure S3 for ChIP controls, statistical analysis was performed by two-tailed student t-test; * = p < 0.05; n = 2. (d) Bisulfite sequencing analysis of the 5′ LTR–the 40nt region-of the infecting virus was performed on WT, Trim24 KD and Trim33 KD ESC and NIH3T3 differentiated cells as control; ESC was next differentiated by 8 days RA induction and methylation status was re-examined. Percentages of methylated CpGs are shown for 10–20 cloned DNA molecules per cell (see Figure S5).

Figure 5

Model for Tif1 family members’ complex formation. A suggested interpretation of the data, showing binding of Trim24 and Trim33 to areas distal from the transcription start site and the PBS. By modifying the 3D structure of the provirus, the complex can be formed or collapse in a dynamic manner.