Fig 2: SETD1A degradation immediately decreases the expression of heme biosynthesis genes(A) Chromatin distribution of endogenous SETD1A or exogenous HA-tagged FKBP-SETD1A was analyzed from MOLM-13 cells or FKBP-SETD1A cells, respectively. Genome browser view at human chromosome 10 (chr10) is shown (top). The distribution of overlapped peaks around the TSS was analyzed with GREAT tool (bottom).(B) The overlap between endogenous SETD1A ChIP-seq peaks and exogenous FKBP-SETD1A ChIP-seq peaks is shown. The overlapped region is labeled as SETD1A ChIP peaks, and 11,013 genes were annotated by GREAT tool.(C) The overlap among SETD1A ChIP-annotated genes and the downregulated genes (log2 fold change < −0.5, adjusted p value < 0.05) at 1 and 4 days post-dTAG-13 treatment in FKBP-SETD1A cells was analyzed. 123 commonly downregulated (DR) targets and 10,734 non-regulated (NR) targets of SETD1A were identified.(D) GO analysis of DR genes was performed, and the biological network Gene Ontology map was created by Cytoscape. Heme biosynthetic process is high-lighted, and associated genes are shown in the box.(E) Motifs within the SETD1A peaks in DR genes were analyzed by Trawler. Position weight matrices (PWM), Z scores of discovered motifs, and the number of motifs in representative DR TSSs are shown.(F) Association of SETD1A peak score and log2 fold change (log2FC) from RNA-seq at 1 day post-dTAG-13 treatment is shown. Heme biosynthesis genes (COX15, HMBS, and UROS) are shown in red dots.(G) MOLM-13 cells and FKBP-SETD1A cells were treated with DMSO or dTAG-13 for 24 h, and RNA levels of heme biosynthesis genes were quantified by qRT-PCR.(H) Time course semi-quantification analysis for HMBS and COX15 expression after SETD1A degradation. RNA was extracted at the indicated time points from FKBP-SETD1A cells treated with dTAG-13. Two independent experiments with 3 biological replicates are performed. Data are represented as mean ± SD.See also Figure S2.

Fig 3: SETD1A silencing leads to a proliferative defect in the NSCLC cell line. (a) Stable knockdown of SETD1A in human lung adenocarcinoma cells was generated by transfection of lentiviral‐encoded shRNA into the A549 cell line. Western blotting of A549 treated with different shRNA was performed and the quantification results are listed in the right panel (n = 4). (b) Stable knockdown of SETD1A in the H1299 NSCLC cell line. The efficiency of knockdown tested with Western blots and quantification results are listed in the right panel (n = 4). (c),(d) Cell number of A549 (c) or H1299 (d) with shRNA‐SETD1A (shSETD1A‐1/2) and shRNA‐scrambled (control) were recorded at indicated time points (n = 5). (e),(f) Western blots of cell cycle‐related proteins were tested and the quantification results are listed in the right panel (n = 4). Control, A549/ H1299 with lentivirus packaged shRNA for the scrambled sequence. shSETD1A‐1 or 2, A549/ H1299 with lentivirus packaged shRNA for SETD1A mRNA at different binding sites. Statistical tests were analyzed by using paired t‐test after log transformation. n means independent biological sample

Fig 4: Downregulation of SETD1A in NSCLC cells suppresses cancer EMT. (a),(b), the images of A549 cells with different expression of SETD1A were captured after wound injury. Scale bar was 200 μm. The motility of lung cancer cells was measured at the indicated time points (n = 5). (c)–(f), the same number of A549 or H1299 cells (shSETD1A‐1 or control) were seeded in Transwell chambers and the cells that passed through the porous membrane were stained and captured (A549, (c) upper row; H1299, (e) upper row). In addition, the cells that passed through the Matrigel and porous membrane were also stained and captured (A549, (c) lower row; H1299, (e) lower row). The number of stained cells in five random fields was counted (A549, D; H1299, F) (n = 3). (g),(h) Western blot to detect EMT‐related proteins in SETD1A silenced A549 (g) or H1299 (h) cell lines (n = 4). Statistical tests were analyzed by using paired t test after log transformation. n means independent biological sample

Fig 5: A model for the mechanism by which the non-enzymatic role of SETD1A promotes the transcriptional elongation and heme biosynthesisTranscriptional pause release is an important effector of non-enzymatic SETD1A function through cyclin K. In proliferating cells, SETD1A is required as a molecular switch to convert cyclin K functions from transcriptional initiation to productive elongation (top). The SETD1A-cyclin K complex with CDKs (CDK9/12/13) may promote substitution of NELF to PAF1 in the RNAPII complex, activation of Ser5P phosphatase (PTase), and elongation via Ser2 phosphorylation. The disruption of SETD1A stabilizes the Ser5 phosphorylated RNAPII and NELF complex, leading to pausing of transcriptional elongation at the TSS of genes associated with DNA repair and heme biosynthesis pathways at H2H loci (bottom). The elongation of genes on the opposite strand of the H2H pair or NR genes could be maintained by SETD1A-free elongation regulators. The genomic structure or transcription of the paired gene may also contribute to pausing stabilization of SETD1A downstream genes.