Compact engineered human mechanosensitive transactivation modules enable potent and versatile synthetic transcriptional control - PubMed (original) (raw)

. 2023 Nov;20(11):1716-1728.

doi: 10.1038/s41592-023-02036-1. Epub 2023 Oct 9.

Alan Cabrera 1, Daniel A Brenner 1, Rosa Selenia Guerra-Resendez 2, Jing Li 1, Jacob Goell 1, Kaiyuan Wang 1, Yannie Guo 1, Mario Escobar 3, Abinand Krishna Parthasarathy 1, Hailey Szadowski 2, Guy Bedford 1, Daniel R Reed 1, Sunghwan Kim 1, Isaac B Hilton 4 5 6

Affiliations

Compact engineered human mechanosensitive transactivation modules enable potent and versatile synthetic transcriptional control

Barun Mahata et al. Nat Methods. 2023 Nov.

Abstract

Engineered transactivation domains (TADs) combined with programmable DNA binding platforms have revolutionized synthetic transcriptional control. Despite recent progress in programmable CRISPR-Cas-based transactivation (CRISPRa) technologies, the TADs used in these systems often contain poorly tolerated elements and/or are prohibitively large for many applications. Here, we defined and optimized minimal TADs built from human mechanosensitive transcription factors. We used these components to construct potent and compact multipartite transactivation modules (MSN, NMS and eN3x9) and to build the CRISPR-dCas9 recruited enhanced activation module (CRISPR-DREAM) platform. We found that CRISPR-DREAM was specific and robust across mammalian cell types, and efficiently stimulated transcription from diverse regulatory loci. We also showed that MSN and NMS were portable across Type I, II and V CRISPR systems, transcription activator-like effectors and zinc finger proteins. Further, as proofs of concept, we used dCas9-NMS to efficiently reprogram human fibroblasts into induced pluripotent stem cells and demonstrated that mechanosensitive transcription factor TADs are efficacious and well tolerated in therapeutically important primary human cell types. Finally, we leveraged the compact and potent features of these engineered TADs to build dual and all-in-one CRISPRa AAV systems. Altogether, these compact human TADs, fusion modules and delivery architectures should be valuable for synthetic transcriptional control in biomedical applications.

© 2023. The Author(s).

PubMed Disclaimer

Conflict of interest statement

B.M., J.G. and I.B.H. have filed a patent related to this work. I.B.H. has filed patent applications related to other CRISPR technologies for genome engineering. The remaining authors declare no competing interests.

Figures

Fig. 1

Fig. 1. CRISPR-DREAM displays potent activation at human promoters, has high specificity and is robust across cell types.

a, dCas9, a gRNA containing two engineered MS2 stem-loops (MS2 SLs) and MCP-fused transcriptional effector proteins are schematically depicted. b, dCas9 and MCP fusion proteins are schematically depicted. Nuclease-inactivating mutations are indicated by yellow bars with dots above. c, The expression levels of dCas9 and dCas9-VP64 (top), FLAG-tagged MCP-mCherry, FLAG-tagged MCP-MSN, FLAG-tagged MCP-p65-HSF1 (middle) and β-tubulin (loading control; bottom) are shown as detected by western blotting in HEK293T cells at 72 h post-transfection. d,e, Relative expression of endogenous human genes after control, DREAM or SAM systems were targeted to their respective promoters using pools of 4 or 3 gRNAs (HBG1 and CD34, respectively; d), or using single gRNAs (ACE2 and HGF, respectively; e), as measured by qPCR. f, RNA-seq data generated after the DREAM or SAM systems were targeted to the HBG1/HBG2 promoter using 4 pooled gRNAs. mRNAs identified as significantly differentially expressed (fold change >2 or <−2 and FDR < 0.05) are shown as red dots. In the top MA plot (CRISPR-DREAM), mRNAs corresponding to _HBG1_/_HBG2_ (target genes) are highlighted in light blue. In the bottom MA plot (SAM system), mRNAs corresponding to _HBG1_/_HBG2_ (target genes) are highlighted in light gray. mRNAs encoding human components of the MSN or SAM systems shown were also significantly differentially expressed (fold change >2 and FDR < 0.05) and are shown in red. g, Six endogenous genes were activated by DREAM or SAM using a pool of gRNAs (1 gRNA per gene) in HEK293T cells. h,i, OCT4 (h) or HBG1 (i) gene activation by DREAM or SAM systems when corresponding promoters were targeted by 4 gRNAs per promoter in hTERT-MSC or PMBC cells, respectively. All qPCR and RNA-seq samples were processed at 72 h post-transfection. Data are the result of 4 biological replicates for d, e and g and 3 biological replicates for h and i. See the source data for more information. Data are presented as mean ± s.e.m. P values were determined using unpaired two-sided _t_-test. FDR, false discovery rate. Source data

Fig. 2

Fig. 2. CRISPR-DREAM efficiently activates transcription from diverse human regulatory elements.

a–c, CRISPR-DREAM and the SAM system activated downstream mRNA expression from OCT4 (a); HBE, HBG and HBD (b); and SOCS1 (c), when targeted to the OCT4 distal enhancer (DE), HS2 enhancer or one of two intragenic SOCS1 enhancers, using pools of 3 (OCT4 DE), 4 (HS2), 3 (SOCS1 + 15 kb) or 2 (SOCS1 + 50 kb) gRNAs, respectively. d, CRISPR-DREAM and the SAM system activated sense eRNA expression when targeted to the NET1 enhancer using 2 gRNAs. e,f, CRISPR-DREAM and the SAM system bidirectionally activated eRNA expression when targeted to the KLK3 (e) or TFF1 (f) enhancers using pools of 4 or 3 gRNAs, respectively. g,h, CRISPR-DREAM and the SAM system activated the expression of lncRNA when targeted to the CCAT1 (g) or GRASLND (h) promoters using pools of 4 gRNAs, respectively. i, CRISPR-DREAM and the SAM system activated the expression of pre- and mature miR-146a when targeted to the miR-146a promoter using a pool of 4 gRNAs. All samples were processed for qPCR at 72 h post-transfection. Data are the result of 5 or 6 biological replicates for b and 4 biological replicates for a and c–i. See the source data for more information. Data are presented as mean ± s.e.m. P values were determined using unpaired two-sided _t_-test. E1, Enhancer 1; E2, Enhancer 2. Source data

Fig. 3

Fig. 3. CRISPR-DREAM is portable to orthogonal dCas9 proteins and amenable to miniaturization.

a, The SadCas9-DREAM system is schematically depicted. Nuclease-inactivating mutations are indicated by yellow bars with dots above. b, HBG1 or TTN gene activation using the SadCas9-DREAM or SadCas9-SAM system, when recruited using pools of 4 promoter-targeting gRNAs. c, HBG1 or TTN gene activation using the SadCas9-DREAM or SadCas9-VPR system, when recruited using pools of 4 MS2-modifed (SadCas9-DREAM) or standard promoter-targeting gRNAs (SadCas9-VPR), respectively. d, The CjdCas9-DREAM system is schematically depicted. Nuclease-inactivating mutations are indicated by yellow bars with dots above. e, HBG1 or TTN gene activation using the CjdCas9-DREAM or CjdCas9-SAM system, when recruited using pools of 3 MS2-modified promoter-targeting gRNAs. f, HBG1 or TTN gene activation using the CjdCas9-DREAM or CjdCas9-VPR system, when recruited using pools of 3 MS2-modifed (SadCas9-DREAM) or standard promoter-targeting gRNAs (CjdCas9-VPR), respectively. g, A 3x 9aa TAD derived from MYOCD and MRTF-B TADs is schematically depicted; GS, glycine-serine linker. h, HBG1 or TTN gene activation when the 3x 9aa TAD was fused to MCP and recruited using dCas9 and a pool of 4 MS2-modified promoter-targeting gRNAs. i, The mini-DREAM system is schematically depicted. MCP-eN3x9 is a fusion protein consisting of MCP, eNRF2 and the 3x 9aa TAD from g. j, HBG1 or TTN gene activation when either the mini-DREAM or CRISPR-DREAM system was recruited using a pool of 4 MS2-modified promoter-targeting gRNAs. k, A simplified biosynthetic pathway for progesterone production is schematically depicted. l, The workflow to build progesterone-producing HEK293T cell factories using the mini-DREAM platform and corresponding gRNA array is shown. m, STAR, CYP11A1 and HSD3B2 gene activation after mini-DREAM-transduced HEK293T cells were transfected with the indicated gRNA array or a nontargeting gRNA control plasmid. n, Secreted progesterone levels after mini-DREAM-transduced HEK293T cells were transfected with the indicated gRNA array or a nontargeting gRNA control plasmid. All samples were processed for qPCR or ELISA at 72 h post-transfection. Data are the result of 4 biological replicates for b, c, e, f, j, m and n, and 3 or 4 biological replicates for h. See the source data for more information. Data are presented as mean ± s.e.m. P values were determined using unpaired two-sided _t_-test. BH, bridge helix; eN, engineered NRF2; M, MRTF-A; PI, PAM-interacting domain; REC, recognition lobe; S, STAT1. Source data

Fig. 4

Fig. 4. The MSN and NMS effector domains are portable to diverse DNA binding platforms and enable superior multiplexing when fused to dCas12a.

a, Synthetic TALE proteins harboring indicated effector domains were designed to target the human IL1RN promoter. RVD, repeat variable di-residue. Relative IL1RN expression (bottom) at 72 h after indicated TALE fusion protein-encoding plasmids were transfected. b, Synthetic zinc finger (ZF) proteins harboring indicated effector domains were designed to target the human ICAM1 promoter. Relative ICAM1 expression (bottom) at 72 h after indicated ZF fusion protein-encoding plasmids were transfected. c, The Type I CRISPR system derived from E. Coli K-12 (Eco-Cascade) is schematically depicted along with an effector fused to the Cas6 protein subunit. d, HBG1 gene activation when the MSN, NMS or p300 effector domains were fused to Cas6 and the respective engineered Eco-Cascade complexes were targeted to the HBG1 promoter using a single crRNA. e, Multiplexed activation of 6 endogenous genes at 72 h after co-transfection of Eco-Cascade complexes when MSN was fused to Cas6 and targeted using a single crRNA array expression plasmid (1 crRNA per promoter). f, The dCas12a protein and indicated fusions are schematically depicted along with the E993A DNase-inactivating mutation indicated by a yellow bar with a dot above. g,h, IL1B (g) or TTN (h) gene activation using the indicated dCas12a fusion proteins when targeted to each corresponding promoter using a pool of 2 crRNAs (for IL1B) or a single array encoding 3 crRNAs (TTN), respectively. i, Multiplexed activation of 16 indicated endogenous genes at 72 h after co-transfection of dCas12a-NMS and a single crRNA array expression plasmid encoding 20 crRNAs. All samples were processed for qPCR at 72 h post-transfection in HEK293T cells. See the source data for more information. Data are the result of 4 biological replicates for a, b, d, e, g, h and i. Data are presented as mean ± s.e.m. P values were determined using unpaired two-sided _t_-test. NLS, nuclear localization signal. Source data

Fig. 5

Fig. 5. dCas9-NMS permits efficient in vitro reprogramming of human fibroblasts.

a, Primary HFFs were nucleofected with plasmids encoding 15 multiplexed gRNAs targeting the OCT4, SOX2, KLF4, c_-MYC_ and LIN28A promoter and EEA motifs (as previously reported), and either dCas9-NMS (middle row) or dCas9-VP192 (bottom row). HFF morphology was analyzed 8 and 16 d later (white scale bars, 100 μm). b, Relative expression of pluripotency-associated genes OCT4 (left) and SOX2 (right) in representative iPSC colonies (C1 or C2) approximately 40 d after nucleofection of either dCas9-NMS (blue) or dCas9-VP192 (gray) and multiplexed gRNAs compared with untreated HFF controls. n = 2 independent measurements from independent subclones per colony. c, Relative expression of mesenchymal-associated genes THY1 (left) and ZEB1 (right) in representative iPSC colonies (C1 or C2) approximately 40 d after nucleofection of either dCas9-NMS (blue) or dCas9-VP192 (gray) and multiplexed gRNAs compared with untreated HFF controls. n = 2 independent measurements from independent subclones per colony. d,e, Immunofluorescence microscopy of HFFs approximately 40 d after nucleofection of either dCas9-NMS or dCas9-VP192 and multiplexed gRNAs compared with untreated HFF controls (white scale bars, 100 μm). Cells were stained for the expression of pluripotency-associated cell surface markers SSEA4 (d, green) or TRA-1-81 (e, green). All cells were counterstained with DAPI for nuclear visualization. Data presented in a is a representative of 3 independent experiments. Data are the result of 2 biological independent measurements from independent subclones per colony for b and c. Data presented in d and e are representative of 2 independent experiments. Data are presented as mean ± s.e.m. Source data

Fig. 6

Fig. 6. CRISPR-DREAM components are well tolerated in primary cells and compatible with viral delivery methods.

a,b, Immunofluorescence microscopy showing mCherry/EGFP expression levels in MSCs (a) and human T cells (b) at 72 h after co-transduction of dCas9 in combination with either MCP-mCherry (control), MCP-eN3x9-T2A-EGFP, MCP-MSN-T2A-EGFP, MCP-NMS-T2A-EGFP or MCP-VPR-T2A-EGFP, respectively (white scale bars, 250 μm for MSCs; 100 μm for T cells). MCP-fusion vectors also contain a U6-driven gRNA expression cassette and either a TTN (MSCs) or CARD9 (T cells). c,d, Relative expression of TTN (c) or CARD9 (d) in MSCs and T cells, respectively, 3 d after lentiviral co-transduction using indicated components. e, AAV constructs used for dual-delivery of CRISPR-DREAM components are schematically depicted. The EFGP control vector is shown (top) along with the hSyn promoter-driven SpdCas9 vector (middle), which consists of a modified WPRE/polyA sequence (W3SL). The U6 promoter-driven gRNA expressing vector (bottom) is also shown and also encodes MCP fused to MSN, which is driven by the hSyn promoter. f, Agrp gene activation in mouse primary cortical neurons using the dual AAV8 transduced CRISPR-DREAM system described (in e) at 5 d post-transduction. g, AIO SadCas9-based AAV vectors are schematically depicted. AIO vectors consist of M11 promoter-driven gRNA cassettes and either SCP1 (top) or EFS (bottom) promoter-driven NMS-SadCas9. A modified WPRE/polyA sequence (CW3SA) was used in the AIO vectors. h, Agrp gene activation in mouse primary cortical neurons transduced with AIO AAV vectors (in h) at 5 d post-transduction. Data are the result of 3 biological replicates for c, d and h and 3 biological replicates for f. See the source data for more information. Data are presented as mean ± s.e.m. P values were determined using unpaired two-sided _t_-test. N/A, not applicable. Source data

Similar articles

Cited by

References

    1. Qi LS, et al. Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression. Cell. 2013;152:1173–1183. - PMC - PubMed
    1. Gilbert LA, et al. CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes. Cell. 2013;154:442–451. - PMC - PubMed
    1. Perez-Pinera P, et al. RNA-guided gene activation by CRISPR-Cas9-based transcription factors. Nat. Methods. 2013;10:973–976. - PMC - PubMed
    1. Thakore PI, Black JB, Hilton IB, Gersbach CA. Editing the epigenome: technologies for programmable transcription and epigenetic modulation. Nat. Methods. 2016;13:127–137. - PMC - PubMed
    1. Liao HK, et al. In vivo target gene activation via CRISPR/Cas9-mediated trans-epigenetic modulation. Cell. 2017;171:1495–1507.e15. - PMC - PubMed

MeSH terms

Substances

LinkOut - more resources