Where Does the Heme Go? Unraveling Heme and Porphyrin Metabolism in Healthy and Oncogenic Human Livers

The liver, a pivotal organ in human metabolism, serves as a primary site for heme biosynthesis, critical for detoxification and drug metabolism. Maintaining precise control over heme production is paramount in healthy livers to meet high metabolic demands while averting potential toxicity from intermediate metabolites, notably protoporphyrin IX. Intriguingly, our recent research uncovers a disrupted heme biosynthesis process termed 'Porphyrin Overdrive' in cancers, fostering the accumulation of heme intermediates, potentially bolstering tumor survival. Here, we investigate heme and porphyrin metabolism in both healthy and oncogenic human livers, utilizing primary human liver transcriptomics and single-cell RNA sequencing (scRNAseq). Our investigations unveil robust gene expression patterns in heme biosynthesis in healthy livers, supporting electron transport chain (ETC) and cytochrome P450 function, devoid of intermediate accumulation. Conversely, liver cancers exhibit impaired heme biosynthesis and massive downregulation of cytochrome P450 expression. Notably, despite diminished drug metabolism, heme supply to the ETC remains largely unaltered or even elevated with cancer progression, suggesting a metabolic priority shift. Liver cancers selectively accumulate intermediates, absent in normal tissues, implicating their role in disease advancement as inferred by expression. Furthermore, our findings establish a link between diminished drug metabolism, augmented ETC function, porphyrin accumulation, and inferior overall survival in aggressive cancers, indicating potential targets for clinical therapy development. Overall design: Single-cell RNA sequencing (scRNAseq) was conducted on primary human liver cell populations obtained from the healthy donor UBV. The mixture of hepatic cell populations were carefully washed in Dulbecco's phosphate-buffered saline (DPBS, 1X; Corning, Cat No. 21-031-CV) and resuspended at a concentration of 10^6 cells/mL to prevent cell aggregates. The scRNAseq protocol utilized the 10x Genomics Chromium controller and the Chromium single-cell 3' library and gel bead kit (10x Genomics, Cat No. PN-1000075) following the standard manufacturer's protocols. Briefly, gel beads-in-emulsion (GEMs) were generated by combining barcoded single-cell 3' v3 gel beads, a master mix containing cells, and partitioning oil onto Chromium chip B. Cells were delivered at a limiting dilution to achieve single-cell resolution, with the majority (~ 90–99%) of generated GEMs containing no cells. Between 2,000–21,000 live cells were loaded onto the Chromium controller to recover between 1,500 - 15,000 cells for library preparation and sequencing. Following GEM generation, the gel beads were dissolved, primers were released, and any co-partitioned cells were lysed. An Illumina TruSeq Read 1, 16 nt 10x barcode, 12 nt unique molecular identifier (UMI), and 30 nt poly(dT) sequence were mixed with the cell lysate and a master mix containing reverse transcription (RT) reagents. Incubation of the GEMs produced barcoded, full-length cDNAs from poly-adenylated mRNAs. Subsequently, GEMs were broken, and cDNA was amplified and quantified using an Agilent high sensitivity DNA screentape. SPRIselect magnetic beads were used to purify the first-strand cDNA, followed by PCR amplification to generate sufficient mass for library construction. Enzymatic fragmentation and size selection optimized the cDNA amplicon size, with TruSeq Read 1 added during GEM incubation. P5, P7, a sample index, and TruSeq Read 2 were added via end repair, A-tailing, adaptor ligation, and PCR. Library quality was assessed using an Agilent Bioanalyzer high sensitivity chip. Sequencing was performed on the Illumina NextSeq 550 with a target of 150,000 reads/cell. Data processing, sample demultiplexing, and gene expression quantification were performed using the Cell Ranger Single-Cell software Suite (10x Genomics). Genes with more than one unique molecular identifier (UMI) count were considered for analysis. The top 1000 most variably expressed genes were used for clustering, and t-Distributed Stochastic Neighbor Embedding (t-SNE) analysis was employed to reduce data dimensionality. K-means clustering identified cell populations, with cell classification based on Spearman's correlation of mean expression profiles.