KRAB-Induced Heterochromatin Effectively Silences PLOD2 Gene Expression in Somatic Cells and Is Resilient to TGFβ1 Activation

Epigenetic editing, an emerging technique used for the modulation of gene expression in mammalian cells, is a promising strategy to correct disease-related gene expression. Although epigenetic reprogramming results in sustained transcriptional modulation in several in vivo models, further studies are needed to develop this approach into a straightforward technology for effective and specific interventions. Important goals of current research efforts are understanding the context-dependency of successful epigenetic editing and finding the most effective epigenetic effector(s) for specific tasks. Here we tested whether the fibrosis- and cancer-associated PLOD2 gene can be repressed by the DNA methyltransferase M.SssI, or by the non-catalytic Krüppel associated box (KRAB) repressor directed to the PLOD2 promoter via zinc finger- or CRISPR-dCas9-mediated targeting. M.SssI fusions induced de novo DNA methylation, changed histone modifications in a context-dependent manner, and led to 50%–70% reduction in PLOD2 expression in fibrotic fibroblasts and in MDA-MB-231 cancer cells. Targeting KRAB to PLOD2 resulted in the deposition of repressive histone modifications without DNA methylation and in almost complete PLOD2 silencing. Interestingly, both long-term TGFβ1-induced, as well as unstimulated PLOD2 expression, was completely repressed by KRAB, while M.SssI only prevented the TGFβ1-induced PLOD2 expression. Targeting transiently expressed dCas9-KRAB resulted in sustained PLOD2 repression in HEK293T and MCF-7 cells. Together, these findings point to KRAB outperforming DNA methylation as a small potent targeting epigenetic effector for silencing TGFβ1-induced and uninduced PLOD2 expression.


SUPPLEMENTAL MATERIALS
MCF7 cells were engineered to contain TET ON ZF-ED transgenes, as described for the MDA-MB-231 cells, or to constitutively express dCas9-fusion proteins as described for HEK293 cells.
MDA-MB-231 cells, engineered to contain TET ON ZF-ED transgenes, cultured in DMEM and treated with doxycycline as described, were frozen after 20 days subculturing in 10%FBS complete medium after doxycycline removal (Figure 5,day 20). HDF cells, engineered to contain TET-ON ZF-ED transgenes and cultured in EMEM as described, were frozen (ca passage 8) before doxycycline treatment was performed. Upon thawing from liquid nitrogen, cells were either subcultured in culture medium supplemented with 10% regular FBS or in culture medium supplemented with 10% charcoal stripped FBS (Sigma-Aldrich). For HDF cells, FBS was heat inactivated as described before. MDA-MB-231 cells were seeded at a concentration of 150,000 cells per well and HDF cells were seeded at a concentration of 100,000 cells per well. Cells were collected at different time points as indicated, or seeded in a 6-well plate for further experiments. Depending on the research question, cells were treated with +/-doxycycline and +/-TGFβ1 as described (at day 25 for the transgenic MDA-MB-231 cells, at day 10 for HDF cells engineered to express dCas9-SKD and at day 21 for HDF cells engineered to express dCas9-MSssI). Expression of the ZF-EDs and PLOD2 was measured with quantitative RT-PCR analysis, as described in the main manuscript.

Leakiness of TET ON system explored
Although a degree of leakiness is often observed for the TET ON system, many reports have been published to demonstrate the transient nature of KRAB-induced gene repression in somatic cells (Groner et al., 2012;Amabile et al., 2016), also using TET ON transgenic cells  without obvious leaky effects . However, since in our experimental system, PLOD2 repression was apparent with and without doxycycline induction  Figure 7A, day 20) and cultured these cells for an additional 25 days either in 10% regular FBS medium or in 10% certified charcoal stripped medium (csFBS) which is depleted of lipophilic agents like doxycycline. After the total of 45 days subculturing after the first doxyxyxline treatment, a second doxyxycline treatment could still induce expression of the ZF-fusions (Supplemental Figure 7B). For all in between time points, the expression of ZF-EDs was not different in csFBS versus regular FBS (Supplementary Figure S8A). The similar expression levels in the two media did not support the notion that the observed leaky expression was caused by traces of doxycycline in regular FBS.
For either conditions, and at all time points, the repression of PLOD2 by ZF-SKD was very effective, leading to a complete silencing of PLOD2, whilst ZF-M.SssI was able to repress PLOD2 expression for 70%, independently of the growth conditions (Supplementary Figure S8B).
Intriguingly, the leaky background expression of ZF-SKD was, at least on the mRNA level, about 10-fold higher than the background expression of ZF-M.SssI, which may explain the more effective repression by SKD. However, as each cell only has two PLOD2 alleles, theoretically only two ZFprotein molecules can bind at any given time point. Since ZF-fusion proteins likely are present in excess after longer time points of culturing, it seems that the induced DNA methylation just cannot further repress the PLOD2 expression, but this point requires more in depth investigation.
As we also saw doxycycline-independent PLOD2 silencing in the HDFs stably engineered to    Figure S1, related to figure 1. PLOD2 ZF coding sequences. Synthesized DNA coding for the eight engineered zinc finger proteins (ZF1-8) that target PLOD2. Depicted in red are the cloning sites and in blue the six individual fingers that make up the ZF.         Figure S10, related to figure 8. PLOD2 expression modulation using the CRISPR-dCas9 platform. (A) sgRNA expression in HEK293T cells transiently transfected or HEK293T-SKD stable cells measured after 2 or 12 days of transfection. Expression is presented relative to GAPDH. (B) PLOD2 mRNA expression levels after 2 days of different cell lines constitutively expressing G9A or G9Amutant, transiently transfected with sgPLOD2 1-4 (mean ± SEM; n = 3, unpaired two-way Student's t-test *=P < 0.05). (C) PLOD2 and SPDEF mRNA expression levels, 2 and 14 days after transient transfection with sgRNAs in MCF-7 cells engineered to constitutively express dCas9-SKD cells, normalized to expression in cells transfected with empty vector (n=2). (D) Stable HEK293T-SKD cells transiently transfected with different groups of sgRNAs (mean ± SEM; n = 3, unpaired two-way Student's t-test *=P < 0.05) Table S1. Primers used in this study for mRNA qPCR, bisulfite sequencing, pyrosequencing and qChIP.