Volume 151, Issue 3, Pages 547-558 (October 2012) Activation of Innate Immunity Is Required for Efficient Nuclear Reprogramming  Jieun Lee, Nazish Sayed, Arwen Hunter, Kin Fai Au, Wing H. Wong, Edward S. Mocarski, Renee Reijo Pera, Eduard Yakubov, John P. Cooke  Cell  Volume 151, Issue 3, Pages 547-558 (October 2012) DOI: 10.1016/j.cell.2012.09.034 Copyright © 2012 Elsevier Inc. Terms and Conditions

Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 1 Irrelevant Retroviral Vector Accelerates CPP-Induced Gene Expression Changes in expression of pluripotency genes were assessed after exposure of human fibroblasts to the retroviral vectors encoding Sox2 or Oct4 or the cell permeant peptides of Sox2 or Oct4. The addition of an irrelevant retroviral vector encoding GFP (pMX-GFP) enhanced the response to the cell permeant peptides. (A) Relative fold change in gene expression levels of Nanog after exposure to pMX-Sox2 (red line), CPP-SOX2 (blue line), or pMX-GFP + CPP-SOX2 (green line). Relative fold change in gene expression was determined following treatment with 200 nM CPP-SOX2 daily or after a single pMX-Sox2 infection on day 0. All data represented as mean ± S.D., n = 3, ∗p < 0.005. (B) Summary figure showing the average fold-change in the selected genes (i.e., Jarid2, Zic2, bMyb, Oct4, Sox2, and Nanog) over time for each condition. When Sox2 is given in the form of a CPP, activation of the downstream target genes is delayed. However, in the presence of an irrelevant retroviral vector, target gene expression increases rapidly and mimics that of pMX-Sox2. (C) Relative fold change in gene expression levels of Nanog following pMX-Oct4, CPP-OCT4, or pMX-GFP + CPP-OCT4 treatments. (D) Summary figure showing the average fold-change in the selected genes (i.e., Tcf4, Gap43, Nanog, Sox2, and Oct4) over time for each condition. All data represented as mean ± S.D., n = 3, ∗p < 0.005 (see also Figures S1 and S2). Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 2 Knockdown of TLR3 Pathway Inhibits Action of Retroviral Vector Encoding Oct4 (A) Gene expression of Oct4 following retroviral-Oct4 (pMX-Oct4) infection is reduced in fibroblasts treated with the TRIF-inhibitory peptide (TRIF-inh). (B) Gene expression of Oct4 following retroviral-Oct4 (pMX-Oct4) infection is reduced in TRIF shRNA-knockdown fibroblasts. Lower: summary diagram of the average fold-changes over time in the selected pluripotency genes (Oct4, Sox2, and Nanog) in the four conditions. (C) Gene expression of Oct4 following retroviral-Oct4 (pMX-Oct4) infection is reduced in TLR3 shRNA-knockdown fibroblasts. (D–F) Summary figures showing the average fold-changes in the selected genes (i.e., Tcf4, Gap43, Nanog, Sox2 and Oct4) over time for each condition. All data represented as mean ± S.D., n = 3, ∗p < 0.005 (see also Figure S2). Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 3 TLR3 or TRIF Knockdown Fibroblasts Exhibit Impaired Nuclear Reprogramming (A) Protocol for iPSC generation using the reprogramming factors, delivered as retroviral vectors. (B) Representative images of iPSCs on day 30 after initiation of retroviral nuclear reprogramming for scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblasts. In fibroblasts where the TLR3 signaling pathway was knocked down (i.e., TRIF shRNA or TLR3 shRNA cell lines), the development of human iPSC colonies was markedly delayed. By contrast, in fibroblasts where the adaptor for all other TLRs was knocked down (MyD88 shRNA), no delay in hiPSC development was noted. (C) Total number of hiPSC colonies on day 30 in scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblast cell lines transduced by the reprogramming factors delivered by retroviral transfection. The yield of hiPSC colonies was reduced by knocking down the TLR3 signaling pathway. All data represented as mean ± S.D., ∗p < 0.05; scramble compared to TRIF or TLR3 shRNA knockdown fibroblasts. (D) Fold change in Oct4 gene expression in scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblasts at day 35. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 4 Nuclear Reprogramming Using mmRNA Is Inhibited by TLR3 or TRIF Knockdown (A) Protocol for iPSC generation using the reprogramming factors, delivered as mmRNA. (B) Representative images of human iPSCs on day 13 after initiation of mmRNA nuclear reprogramming in the scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblasts. In fibroblasts where the TLR3 signaling pathway was knocked down (i.e., TRIF shRNA or TLR3 shRNA cell lines), the development of human iPSC colonies was markedly delayed. By contrast, in fibroblasts where the adaptor for other TLRs was knocked down (MyD88 shRNA), no delay in hiPSC development was noted. (C) Total number of hiPSC colonies on day 13 in scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblast cell lines transduced by the mmRNA reprogramming factors. The yield of hiPSC colonies was reduced by knocking down the TLR3 signaling pathway. All data represented as mean ± S.D., ∗p < 0.05; scramble compared to TRIF or TLR3 shRNA knockdown fibroblasts. (D) Fold change in Oct4 gene expression in individual colonies derived from scramble, MyD88, TRIF, and TLR3 shRNA knockdown fibroblasts at day 13 of their mmRNA reprogramming. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 5 Poly I:C Accelerates CPP-Induced Target Gene Expression (A) Relative fold change in gene expression levels of Jarid2 following pMX-Sox2 (red line), CPP-SOX2 (blue line), or poly I:C + CPP-SOX2 (green line) treatments. (B) Summary figure of these selected genes (i.e., Jarid2, Zic2, bMyb, Oct4, Sox2, and Nanog) exhibiting the temporal pattern of average gene expression following each treatment. Poly I:C markedly enhances the expression of downstream genes by CPP-SOX2. (C) Relative fold change in gene expression levels of Nanog following pMX-Oct4, CPP-OCT4, or poly I:C + CPP-OCT4 treatments. (D) Summary figure of these selected genes (i.e., Tcf4, Gap43, Nanog, Sox2, and Oct4) exhibiting the temporal pattern of gene expression following each treatment. Poly I:C markedly enhances the expression of downstream genes by CPP-OCT4. All data represented as mean ± S.D., n = 3, ∗p < 0.005 (see also Figures S4, S5, S6, and S7). Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 6 TLR3 Activation Enhances Reprogramming in a Dox-Inducible System MEFS containing the dox-inducible polycistronic transgene construct encoding the four reprogramming factors were stimulated by dox, in the absence or presence of TLR3 activation with poly I:C or the irrelevant retroviral vector pMX-GFP. (A) Gene expression of Oct4 and Sox2 was accelerated by coadministration of poly I:C, or a retroviral construct encoding GFP. (B) Histogram showing SSEA-1+ colonies at 2 and 3 weeks in primary plates. Coadministration of poly I:C, or a retroviral construct encoding GFP, markedly increased the yield of dox-induced reprogramming. (C) SSEA-1 live staining showing iPSC colonies derived from dox-inducible MEFs 4 weeks after exposure to dox. In some wells, MEFs were also exposed to a retroviral construct encoding GFP, or to poly I:C (see also Figures S6 and S7). Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure 7 TLR3 Activation Stimulates CPP-Induced Reprogramming of Human Fibroblasts (A) Protocol for human iPSC generation using four CPP-TFs (OCT4-R11, SOX2-R11, KLF4-R11, and cMYC-R11). (B) Gene expression of Oct4 was increased by coadministration of poly I:C, by day 30-45. (C) TRA-1-81 positive colonies counted at day 30 and day 40 in the presence and absence of poly I:C. Coadministration of poly I:C markedly increased the yield of CPP-induced reprogramming. All data represented as mean ± S.D., n = 3, ∗p < 0.01. (D) ES-like colony formation at day 32 of CPP-induced transactivation (10×) (see also Figures S3, S4, S5, S6, and S7 and Table S1). Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S1 Irrelevant Retroviral Vector or Nonintegrating pMX-GFP Accelerates CPP-Induced Gene Expression, Related to Figure 1 (A and B) Relative fold change in gene expression levels of Jarid2 and Zic2 following pMX-Sox2 (red line), CPP-SOX2 (blue line) or pMX-GFP + CPP-SOX2 (green line) treatments. All data represented as mean ± S.D., n = 3. (C) Summary figure showing the average fold-change in the selected genes over time for each condition. (D and E) Relative fold change in gene expression levels of Tcf4 and GAP43 following pMX-Oct4, CPP-OCT4 or pMX-GFP + CPP-OCT4 treatments. All data represented as mean ± S.D., n = 3. (F) Immunofluorescent images of BJ fibroblasts infected with either pMX-GFP mutant (top) or pMX-GFP wt (bottom). (G and J) Relative fold change in gene expression levels of Oct4, Sox2, Nanog and Tlr3 following pMX-GFP (red line), pMX-GFP + CPP-OCT4 (blue line), pMX-GFP mutant (green line) or pMX-GFP mutant + CPP-OCT4 (purple line) treatments. All data represented as mean ± S.D., n = 3. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S2 Gene Expression Induced by Individual Reprogramming Factors Expressed from Viral Vectors or Delivered as Cell-Permeant Peptides, Related to Figures 1 and 2 (A–C) Fold-expression of Nanog (A), Oct4 (B) and Sox2 (C) following infection with retroviral expression vector (blue line) or cell permeant peptide (red line) at later time points (up to day 20). By comparison to the viral vector pMX-Oct4, the cell permeant CPP-OCT4 causes a modest increase in pluripotency gene expression at later time points. Relative fold change in gene expression was determined following treatment with 200 nM CPP-OCT4 daily or after a single pMX-Oct4 infection on day 0. All data represented as mean ± S.D from technical replicates. (D–F) Knockdown of TLR3 signaling decreases pluripotency gene expression. (D) Gene expression of Tcf4, NF-κB or TLR3 following retroviral-Oct4 (pMX-Oct4) infection is reduced in fibroblasts treated with the TRIF-inhibitory peptide (Pepinh-TRIF); (E) Gene expression of Tcf4, NF-κB or TLR3 following pMX-Oct4 infection is reduced in TRIF shRNA-knockdown fibroblasts; (F) Gene expression of Tcf4, NF-κB or TLR3 following pMX-Oct4 infection is reduced in TLR3 shRNA-knockdown fibroblasts. All data represented as mean ± S.D., n = 3. (G and H) Inhibition of MyD88 does not affect pMX-Oct4-induced gene expression. (G) Gene expression of Oct4, Tcf4, NF-κB or TLR3 following retroviral-Oct4 (pMX-Oct4) infection remains unaltered in fibroblasts treated with the MyD88-inhibitory peptide (Pepinh-MyD88); (H) Gene expression of Oct4, Tcf4, NF-κB or Tlr3 following pMX-Oct4 infection remains unaltered in MyD88 shRNA-knockdown fibroblasts. All data represented as mean ± S.D., n = 3. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S3 Characterization of Human piPSCs and ppiPSCs, Related to Figure 7 The iPSCs induced by reprogramming factors in the form of CPPs, in the absence of (piPSCs) or in the presence of poly I:C (ppiPSCS) express features of pluripotency. (A) Immunostaining for pluripotency markers in human piPSCs and ppiPSC. (B) Gene expression for markers of pluripotency and for the three germlines in the embryonic bodies (EB) derived from the reprogramming factors delivered as retroviral vectors (viPSC) or as CPPs (piPSC and ppiPSC). Gene expression is expressed as a ratio between that in the embryoid bodies and that in the pluripotent cells from which they were derived. (C–F) Hematoxylin and eosin staining of representative ppiPS teratoma. These slides reveal (C) a blood vessel; (D) seminiferous tubules; (E) smooth muscle; (F) liver. (G–J) Imunohistologic staining of representative ppiPS teratoma. (G) Keratin staining reveals a cluster of tumor cells forming a gland-like structure; (H) smooth muscle actin (SMA) highlights vascular smooth muscle cells; staining for S100 highlights; (I) Sertoli cells in the walls of seminiferous tubules and (J) nerve twigs stained with S100. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S4 Poly I:C/Viral Particles Promote Early Epigenetic Modification, Related to Figures 5 and 7 (A) ChIP analysis to assess H3K4me3 of the Oct4 promoters, on day 2 of exposure to the various treatment conditions. Stimulation of TLR3 with pMX-GFP or with poly I:C alone had no effect, nor did CPP-SOX2. However, in combination with pMX-GFP or with poly I:C, the cell permeant peptide CPP-SOX2 mimicked the effects of retroviral Sox2 (pMX-Sox2). (B) ChIP analysis to assess H3K9me3 of the Oct4 promoters, on day 2 of exposure to the various treatment conditions, Data represented as mean ± S.D, n = 2 (∗p < 0.005, ∗∗p < 0.02). (C and D) Immunoblot showing HDAC 1 protein expression confirms that poly I:C or pMX-GFP induces an early (day 2) and sustained (to day 6) inhibition of HDAC1 expression. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S5 Poly I:C/Viral Particles Promote Epigenetic Modification, Related to Figures 5 and 7 (A) ChIP analysis to assess H3K4me3 of the Sox2 promoters, on day 2 of exposure to the various treatment conditions. Stimulation of TLR3 with pMX-GFP or with poly I:C alone had no effect, nor did CPP-SOX2. However, in combination with pMX-GFP or with poly I:C, the cell permeant peptide CPP-SOX2 mimicked the effects of retroviral Sox2 (pMX-Sox2). Data represented as mean ± S.D, n = 2. (B) ChIP analysis to assess H3K9me3 of the Sox2 promoters, on day 2 of exposure to the various treatment conditions. Poly I:C or pMX-GFP enhanced the effect of CPPSox2 to reduce the suppressive epigenetic marking on the Sox2 promoter. Data represented as mean ± S.D, n = 2. (C) Knockdown of TRIF or TLR3 abolishes the effect of poly I:C to reduce HDAC1 expression. Immunoblot showing HDAC1 protein expression following poly I:C treatment in Scramble, MyD88, TRIF and TLR3 shRNA knockdown fibroblasts. The data indicate that knockdown of TLR3 pathway abolishes the effect of poly I:C on HDAC. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S6 Poly I:C Activates NF-κB via TLR3-TRIF Signaling and NF-κB, IKKε or IRF3 Inhibition Impairs Dox-Inducible Reprogramming, Related to Figures 5, 6, and 7 (A) Luciferase assay for NF-κB activity reveals that poly I:C (but not CPP-SOX2) substantially increases NF-κB activity, an effect that is inhibited by knocking down the TLR3 signaling pathway. (B–D) MEFS containing the dox-inducible polycistronic transgene construct encoding the four reprogramming factors were stimulated by dox, to assess the effect on reprogramming of NF-κB, IKKε or IRF3 inhibition. (B) Nuclear localization of NF-κB was inhibited by the treatment of p65 decoy (p65i); (C) IKKε or IRF3 gene expression was reduced in IKKε- or IRF3-knockdown MEFs by using shRNA inhibition. Data represented as mean ± S.D; (D) Number of SSEA-1 positive iPSC colonies was decreased after NF-κB, IKKε or IRF3 inhibition during dox-induced reprogramming, two independent experiments done in duplicate. Data represented as mean ± S.D. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions

Figure S7 NF-κB, IKKε or IRF3 Inhibition Blocks the Poly I:C-Induced Epigenetic Modification in Dox-MEF, Related to Figures 5, 6, and 7 (A and B) ChIP analysis to assess H3K4me3 of the Oct4 promoters, on day 2 of exposure to the NF-κB, IKKε or IRF3 inhibiting conditions. Stimulation of TLR3 with poly I:C during dox-induced reprogramming increases the activating H3K4 trimethylation mark on the Oct4 promoter. However, this effect of poly I:C is abolished by inhibition of NF-κB, IKKε or IRF3; (B) ChIP analysis to assess H3K27me3 of the Oct4 promoters, on day 2 of exposure to the various inhibition conditions. Stimulation of TLR3 with poly I:C during dox-induced reprogramming further reduces the suppressive H3K27 trimethylation mark on the Oct4 promoter. This effect of poly I:C is abolished by inhibition of NF-κB or IKKε, and attenuated by inhibition of IRF3. (C and D) NF-κB inhibition reverses effects of poly I:C on gene expression of chromatin modification factors in dox activated MEF. Gene expression array for chromatin modification factors after treatment with poly I:C (C) and inhibition of NF-κB (D) in dox activated MEF (24 hr). Scatter plot revealed upregulated genes (Red box) and downregulated genes (Green box) with 2-fold change results from Ct values. Cell 2012 151, 547-558DOI: (10.1016/j.cell.2012.09.034) Copyright © 2012 Elsevier Inc. Terms and Conditions