Cell Type-Specific Transcriptome Profiling Reveals a Role for Thioredoxin During Tumor Initiation

Abstract

Neutrophils in the tumor microenvironment exhibit altered functions. However, the changes in neutrophil behavior during tumor initiation remain unclear. Here we used Translating Ribosomal Affinity Purification (TRAP) and RNA sequencing to identify neutrophil, macrophage and transformed epithelial cell transcriptional changes induced by oncogenic Ras^G12V in larval zebrafish. We found that transformed epithelial cells and neutrophils, but not macrophages, had significant changes in gene expression in larval zebrafish. Interestingly, neutrophils had more significantly down-regulated genes, whereas gene expression was primarily upregulated in transformed epithelial cells. The antioxidant, thioredoxin (txn), a small thiol that regulates reduction-oxidation (redox) balance, was upregulated in transformed keratinocytes and neutrophils in response to oncogenic Ras. To determine the role of thioredoxin during tumor initiation, we generated a zebrafish thioredoxin mutant. We observed an increase in wound-induced reactive oxygen species signaling and neutrophil recruitment in thioredoxin-deficient zebrafish. Transformed keratinocytes also showed increased proliferation and reduced apoptosis in thioredoxin-deficient larvae. Using live imaging, we visualized neutrophil behavior near transformed cells and found increased neutrophil recruitment and altered motility dynamics. Finally, in the absence of neutrophils, transformed keratinocytes no longer exhibited increased proliferation in thioredoxin mutants. Taken together, our findings demonstrate that tumor initiation induces changes in neutrophil gene expression and behavior that can impact proliferation of transformed cells in the early tumor microenvironment.

Keywords: gene expression; keratinocyte; migration; neutrophil; thioredoxin (txn); tumor initiation.

Figure 1

Oncogenic HRas induces differentially expressed genes in neutrophils and transformed cells in zebrafish larvae. (A) Schematic of Translating Ribosomal Affinity Purification (TRAP) procedure. Wild-type embryos were injected with the TRAP plasmid (pTol2-Krt4-EGFP-L10a) and either control (pTol2-Krt4-HRas-mcherry) or oncogenic transformation (pTol2-Krt4-HRas^G12V-mcherry) constructs. Immunoprecipitation and sequencing were conducted 3dpf on batches of ~50 larvae collected on three separate days (n=3). Tg(LyzC : EGFP-L10a) or Tg(mpeg:EGFP-L10a) larvae were injected with either control (pTol2-Krt4-HRas-mcherry) or oncogenic Ras (pTol2-Krt4-HRas^G12V-mcherry) constructs for neutrophil- or macrophage-targeted L10a expression, respectively. (B) Cell type specific genes expressed in keratinocytes (krt17, krt15, krt97, krt91, krt8, krt92, krt18, krt94), neutrophils (mpx,lyz), or macrophages (mpeg1.1, irf8, csf1rb, spi1a, mfap4, irg1, csf1ra, spi1b, ncf1, cyba) and their expression profiles in TRAP samples from keratinocytes, neutrophils, and macrophages. (C) Venn diagram showing distribution of genes altered >2fold by TRAP-seq in response to HRas^G12V expressing keratinocytes (n=3).

Figure 2

Thioredoxin expression is induced in neutrophils and keratinocytes with HRas^G12V transformation. (A) Volcano plots displaying number of significantly (adjusted p<0.05) downregulated and upregulated genes in keratinocytes and neutrophils in HRas^G12V-keratinocyte expressing zebrafish larvae, compared to wild-type HRas controls. (B) Gene Set Enrichment Analysis (GSEA) of significantly enriched Hallmark pathways in neutrophils and keratinocytes for HRas^G12V-keratinocyte expressing zebrafish. FDR q-value is (a) below 0.10 or (b) 0.10-0.25. (C) Heat maps displaying significantly differentially expressed genes (adjusted p<0.05) in keratinocytes (56 genes) and neutrophils (53 genes).

Figure 3

Thioredoxin regulates neutrophil recruitment and redox balance in damaged tissues. (A) Schematic of txn gene with annotations of gRNA target site and 1bp insertion identified in txn mutant larvae. (B) Western blot of Txn and αTubulin on whole lysates from txn wildtype and mutant larvae. (C) Schematic of tail transection assay to measure neutrophil recruitment and H₂O₂ abundance conducted on txn ^+/+ Tg(LyzC:H2B-mcherry) or txn ^-/-Tg(LyzC:H2B-mcherry) larvae 3dpf. (D) pfbs-f probe for H₂O₂ (1uM) in response to tail transection 30mpw (wt n=56, mut n=54) and 4hpw (wt n=51, mut n=51) with (E) quantification of signal density in region posterior to notochord. (F) Neutrophil recruitment in response to tail transection 2hpw (wt n=79, mut n=85) and 6hpw (wt n=87, mut n=90) with (G) quantification of neutrophil abundance in region posterior to notochord. Scale bar, 50um. Large, bolded shapes indicate average value per replicate (n), with small shapes representing data points from independent larvae. Samples were analyzed for statistical significance by Mann-Whitney U test (p < 0.001***, p < 0.0001****). ns, not significant.

Figure 4

Thioredoxin depletion in a neutrophil-like cell line increases ROS but does not intrinsically affect motility. (A) Western blot of thioredoxin confirming CRISPR/Cas9 mediated txn knockdown in three HL60 clonal lines (KD4, KD6 and KD7) compared to luciferase control, with beta-tubulin as a loading control. (B) Quantification of intracellular ROS production (Carboxy-H2DCFDA probe) over time in txn knockdown lines compared to control (n=3). (C) Quantification of extracellular H₂O₂ production (Amplex Red/HRP Probe) over time in txn knockdown lines compared to control (n=3). (D) 2D chemotactic index and (E) mean track speed for txn knockdown lines compared to control (N = 2 repeats, n = 5-12 devices). Samples were analyzed for statistical significance via generalized estimating equation (GEE) fit to the log fold change with confidence intervals set to 95% coverage (B, C) or t-test (D, E) (p < 0.001***).

Figure 5

Transformed keratinocytes exhibit increased proliferation and reduced apoptosis during tumor initiation in thioredoxin mutants. (A) EdU staining (white) of txn ^+/+ or txn ^-/- larvae with KRas^G12V-expressing basal keratinocytes (magenta) at 3dpf. (B) Quantification of the frequency of EdU-positive KRas^G12V-expressing basal cells 3dpf in txn wt (n=35) and mutant (n=35) larvae. (C) Cleaved caspase-3 immunofluorescence of txn ^+/+ or txn ^-/- larvae. (D) Quantification of relative signal intensity within transformed cells 3dpf in txn ^+/+ (n=45) and txn ^-/- (n=51) larvae. Scale bar, 10um. Large, bolded shapes indicate average value per replicate (n), with small shapes representing data points from independent larvae. Samples were analyzed for statistical significance via t-test (p < 0.001***, p < 0.0001****).

Figure 6

Thioredoxin affects neutrophil motility around KRas^G12V-transformed keratinocytes. (A) Neutrophil (blue) recruitment around KRas^G12V-expressing basal keratinocytes (magenta) in txn ^+/+ Tg(LyzC:H2B-mcherry) or txn ^-/- Tg(LyzC:H2B-mcherry) larvae 3dpf. (B) Quantification of neutrophil abundance in txn wt (n=74) and mutant (n=116) larvae. (C) Stills from live imaging of neutrophil behavior around KRas^G12V-expressing cells in txn wt and mutant larvae. Neutrophils were tracked across frames (1.5min/frame) and track color delineates instantaneous neutrophil velocity. (D) Quantification of average neutrophil speed in txn wt (n=4) and mutant (n=4) larvae. (E) TRAP-qPCR of KRas^G12V-expressing genes assessing inflammatory cytokines il1b, cxcl8a, cxcl8b.1, tnfa (pool, 30 larvae/replicate, wt n=3, mut n=3). Scale bar, 10uM. Bolded shapes indicate average values per larvae, and small shapes represent instantaneous velocities of individual neutrophils. Samples were analyzed for statistical significance via Mann-Whitney U test (B) and t-test (D, E) (p < 0.01**). ns, not significant.

Figure 7

Transformed keratinocyte proliferation is not altered in thioredoxin mutants in the absence of neutrophils. (A) EdU staining (white) of neutrophil-deficient Tg(mpx:Rac2^D57N-mcherry) txn ^+/+ and Tg(mpx:Rac2^D57N-mcherry) txn ^-/- larvae with KRas^G12V-expressing basal keratinocytes (magenta) 3dpf. (B) Quantification of frequency of EdU-positive KRas^G12V-expressing basal keratinocytes 3dpf in Rac2^D57N txn wt (n=21) and Rac2^D57N txn mutant (n=22) larvae. Scale bar, 10um. Bolded shapes indicate average value per technical replicate, with shaded values representing data points from independent larvae. Samples were analyzed for statistical significance via t-test ns, not significant.