Helicase-like transcription factor-deletion from the tumor microenvironment in a cell line-derived xenograft model of colorectal cancer reprogrammed the human transcriptome-S-nitroso-proteome to promote inflammation and redirect metastasis

Methylation of the HLTF gene in colorectal cancer (CRC) cells occurs more frequently in men than women. Progressive epigenetic silencing of HLTF in tumor cells is accompanied by negligible expression in the tumor microenvironment (TME). Cell line-derived xenografts (CDX) were established in control (Hltf+/+) and Hltf-deleted male Rag2-/-IL2rg-/- mice by direct orthotopic cell microinjection (OCMI) of HLTF+/+HCT116 Red-FLuc cells into the submucosa of the cecum. Combinatorial induction of IL6 and S100A8/A9 in the Hltf-deleted TME with ICAM-1 and IL8 in the primary tumor activated a positive feedback loop. The proinflammatory niche produced a major shift in CDX metastasis to peritoneal dissemination compared to controls. Inducible nitric oxide (iNOS) gene expression and transactivation of the iNOS-S100A8/A9 signaling complex in Hltf-deleted TME reprogrammed the human S-nitroso-proteome. POTEE, TRIM52 and UN45B were S-nitrosylated on the conserved I/L-X-C-X2-D/E motif indicative of iNOS-S100A8/A9-mediated S-nitrosylation. 2D-DIGE and protein identification by MALDI-TOF/TOF mass spectrometry authenticated S-nitrosylation of 53 individual cysteines in half-site motifs (I/L-X-C or C-X-X-D/E) in CDX tumors. POTEE in CDX tumors is both a general S-nitrosylation target and an iNOS-S100A8/A9 site-specific (Cys638) target in the Hltf-deleted TME. REL is an example of convergence of transcriptomic-S-nitroso-proteomic signaling. The gene is transcriptionally activated in CDX tumors with an Hltf-deleted TME, and REL-SNO (Cys143) was found in primary CDX tumors and all metastatic sites. Primary CDX tumors from Hltf-deleted TME shared 60% of their S-nitroso-proteome with all metastatic sites. Forty percent of SNO-proteins from primary CDX tumors were variably expressed at metastatic sites. Global S-nitrosylation of proteins in pathways related to cytoskeleton and motility was strongly implicated in the metastatic dissemination of CDX tumors. Hltf-deletion from the TME played a major role in the pathogenesis of inflammation and linked protein S-nitrosylation in primary CDX tumors with spatiotemporal continuity in metastatic progression when the tumor cells expressed HLTF.


Introduction
of the enzyme NO synthase (NOS) using L-arginine and molecular oxygen as substrates. Of the three nitric oxide synthase (NOS) isoforms, neuronal (nNOS, NOS1) and endothelial (eNOS, NOS3) are calcium-dependent isoforms that produce nanomolar concentrations of NO for seconds or minutes when activated. In contrast, the inducible (iNOS, NOS2) isoform is calcium-independent. and produces micromolar concentrations of NO for hours or days [24]. NO induces post-translational coupling of a nitroso moiety to a reactive cysteine leading to protein S-nitrosylation that affects protein-protein interactions [25]. This pathway is independent of the NO/soluble guanylate cyclase (sGC)/cGMP-dependent protein kinase (PKG) signaling pathway. The NO signal can be stored and propagated at nitrosyl adducts at specific cysteine sites of proteins. S-nitrosylation is purported to be site-selective. INOS-S100A8/A9 targets a conserved I/L-X-C-X 2 -D/E motif [26]. However, S-nitrosylation is a reversible covalent chemical reaction and therefore one of the more difficult post-translational modifications to study. Progress in this regard is owed to the three fundamental steps of the biotin-switch assay [27]: free sulfhydryls are chemically blocked, nitrosylated cysteines are selectively reduced, and a biotin adduct is switched for the NO adduct. Recent modifications substitute iodoTMT reagents [28] in the biotin-switch procedure and use mass spectrometry to identify and quantify specific S-nitrosylated residues.
To investigate the functional importance of negligible HLTF expression in fibroblasts of the TME, we developed an HCT116 cell line-derived xenograft model of metastatic CRC. Primary tumor xenografts were established in Hltf-deleted and control male mice by direct OCMI of HLTF +/+ HCT116 Red-FLuc cells [29]. The HCT116 human colon carcinoma cell line was selected because it expresses HLTF [30], displays leading edge invasion in xenograft tumor model [31], and recapitulates the multi-step dissemination process to the liver and lungs [32]. However, HLTF +/+ HCT116 cells in Hltf-deleted mice shifted their metastatic direction. Species-specific RNAseq analysis of primary CDX tumors arising from the same passage of HLTF +/+ HCT116 Red-FLuc cells in the TME of Hltf-deleted and control mice revealed striking coordination of an inflammation pathway. Transactivation of iNOS expression and the iNOS-S100A8/A9 signaling axis in Hltf-deleted TME drove the comparison of S-nitroso-proteome of primary tumor with metastatic tumors in Hltf-deleted TMEs to reveal remarkable spatiotemporal continuity in S-nitrosylation signaling.

Experimental timeline ensured mouse survival
There is a negative correlation between HLTF expression in tumor cells and survival in mice [33]. Therefore, when HLTF +/+ human HCT116 Red-FLuc cells were used to establish an orthotopic xenograft model in Hltf +/+ (control mice) and Hltf-deleted mice a timeline of 35 days was established during which primary tumor size and metastasis was assessed weekly by BLI. At necropsy, primary tumors were � 10 mm, i.e. below the maximum allowable tumor size of 20 mm tumor development/metastasis did not interfere with daily activities (eating, drinking, nest-building, locomotor). Additionally, none of the mice experienced bowel obstruction or changes in stooling or stool consistency. There was no evidence of blood associated with stooling. There was no evidence of anemia, i.e. loss of pink condition of mouse footpads. There was no evidence of compromised behavior, i.e. mice did not become lethargic and there was no incidence of piloerection, poor grooming or inability to thermoregulate. The mice experienced no sustained weight loss. Comparison survival curves for Hltf-deleted mice (n = 20) and control mice (n = 12) mice with the logrank (Mantel-Cox) test (Chi square 1.898, p = 0.1683), and the Gehan-Breslow-Wilcoxon test (Chi square 1.895, p = 0.1687), indicated Hltf-deletion had no significant effect on the mortality of the mice during the 35-day post-surgery timeline (Fig 1). However, three Hltf-deleted mice died. Thus, the hazard ratio (Mantel-Haenszel) indicated Hltf-deleted mice at any time during the treatment protocol were 5-times more likely to die than control mice. This approach determined an earlier humane endpoint of 28 days for Hltf-deleted mice in future experiments. The mice were treated alike and maintained under identical conditions.

HLTF expression is negatively associated with survival in human CRC
There is a negative correlation between HLTF expression and the progression of CRC in humans [11][12][13][14][15][16]. Previous studies have shown the progressive loss of HLTF from tumor cells correlates with negligible HLTF expression in the TME [19,20]. The CDX model is predicated on the paucity of HLTF expression in the fibroblasts of the TME as shown in a well-differentiated adenocarcinoma (Fig 2) from a male patient that resulted in hemicolectomy.

Hltf-deletion from the TME promotes metastasis in the CDX model
The goal of the first experiment was to establish primary tumor xenografts with HLTF expressing HCT116 Red-FLuc cells (easily detected by bioluminescence imaging-BLI) in control male mice by direct OCMI between the mucosa and the muscularis layers of the cecal wall ( Fig  3). Care was taken to avoid microinjection into the lymphoid nodules in the distal part of the mouse cecum. Histopathological confirmation of tumor foci in the most clinically relevant [32] metastatic sites (e.g., colon-draining lymphatics, liver, lung and peritoneum) coincided with BLI. The goal of the second experiment was to establish primary tumor xenografts via OCMI in Hltf-deleted mice. Tumor take rates were 100% for primary xenograft tumors for Hltf-deleted (n-20) and control (n = 12) mice. However, a shift in the pattern of metastases (Fig 4) to regional lymph nodes, peritoneal carcinomatosis (PC), inguinal canal in close proximity to seminal vesicles and into the scrotum occurred in Hltf-deleted TME (Fig 5). PC is a late stage manifestation of CRC and considered lethal in humans [34].

The dynamic interaction between cells of the tumor and the TME was decoded by species-specific comparative transcriptomics
Alternative splicing increases the complexity of HLTF gene expression in human and mouse cells. Because of its putative role in tumor development, alternative splicing in cells of the tumor and TME were evaluated. Two transcript variants encode the same HLTF protein in Homo sapiens [21]. Cuff.diff alternative splicing analysis identified transcript variant 1 (NM_003071.4), the longer transcript variant (5320-bp mRNA), as the only protein ( Fig 4A) encoding mRNA transcript in primary CDX tumor cells. Hltf is alternatively spliced in mouse tissues [33,[35][36][37], and Cuff.diff alternative splicing analysis quantified the usage of each exon and each possible splice junction for Hltf in RNA-seq samples from control mouse TME. A 4:1 ratio of full-length 4956 nucleotide (nt) message isoform (NM_09210) to a long non-coding (Lnc) 3643 nt transcript variant 4 (Lnc-Hltf-4; NR_105047) was identified. This implicates Lnc-Hltf-4 in CRC progression.
Cuff.diff FPKM tracking files were analyzed with iPathwayGuide (Advaita Bioinformatics)-using the q value of 0.05 for statistical significance and a log fold change of expression with an absolute value of at least 0.06-30 differentially expressed (DE) genes were identified out of a total of 15,857 genes with measured expression in the mouse TME (S1 Table). Similarly, for human tumor analysis-using the q value of 0.05 for statistical significance and a log fold change of expression with an absolute value of at least 0.06-151 DE genes were identified out of a total of 12,358 genes with measured expression in the human tumor (S2 Table).
Hypoxia is an important regulator in CRC, and Hltf is a transcriptional regulator of murine hypoxia inducible factor-1α (Hif-1α) in heart [36]. However, transcriptional availability of HLTF protein in a pT3 adenocarcinoma with no regional lymph node metastasis (pN0). HLTF-immunostaining in well-differentiated (G1) cancer cells of an adenocarcinoma that arose from a tubulovillous adenoma (polyp). Tumor cells invaded through the muscularis propria into the pericolorectal tissue at the hepatic flexure. There was no evidence of lymphovascular invasion. The photomicrographs in panels A-C show the cellular architecture of the tumor tissue composed of HLTF-positive cancer cells and HLTF-negative fibroblasts. A (4X magnification), B (10X magnification) and C (20X magnification). Antibodies to the HLTF N-terminus (residues 164-300) that are common to all known human HLTF proteins [21] in nuclear and cytoplasmic locations show the positive immunostain (brown) is predominately although not exclusively cytoplasmic. HLTF-negative fibroblasts (blue arrows) are devoid of nuclear and/or cytoplasmic staining. https://doi.org/10.1371/journal.pone.0251132.g002 HIF1α in the CDX tumors was unaffected by the presence or absence of Hltf in the TME. In addition, of the 151 DE genes, only 14 of~2450 TRANSFAC predicted HIF1α targets were either up (n = 3) or down (n = 11) regulated, and no functional links between them were identified by iPathwayGuide. These findings argue Hltf-deletion from the TME had little or no effect on the oxygenated/metabolic state of CDX tumors. In fact, cytokine-receptor interaction (KEGG: 04060) was the top biological pathway in CDX tumors (p = 0.002) and mouse TME (p = 2.546e-5) with Bonferroni corrected p-values. Increased mRNA ( Fig 6A) and protein ( Fig  6B and 6C) expression of the pro-inflammatory cytokines, interleukin 8 (IL8) and intercellular adhesion molecule-1 (ICAM-1) was identified in CDX tumors with an Hltf-deleted TME. Transactivation of interleukin 6 (IL6) and all components of iNOS-S100A8/A9 signaling in the Hltf-deleted TME ( Fig 6D) coincided with dramatic increases in the mRNA abundance of haptoglobin (HP) and SERPINA3, and decreased hepatocyte growth factor (HGF) in CDX tumors In vivo BLI (one-second exposure) followed IP injection of D-luciferin (B). A 50 μl Hamilton syringe with 30GA/0.5 inch/30-degree custom needles eliminated retrograde leakage. Cell signal authenticity (C) was confirmed in vitro (cm = culture medium, luc = luciferin). BLI intensity (D) ranged from red (highest), through yellow, green, blue, and purple (lowest). Histopathological confirmation with Hematoxylin and Eosin staining of tumor foci in the most clinically relevant metastatic sites coincides with BLI (E) in control mice. BLI and histopathology showing progression in the same mouse.
https://doi.org/10.1371/journal.pone.0251132.g004 Imaging of CDX metastasis in Hltf-deleted mice at necropsy. Mice were anesthetized with isoflurane and metastatic CRC was detected by BLI (A) prior to necropsy. Immediately following euthanasia, mice were opened via a mid-sagittal incision and the primary CDX tumor (B1) was removed followed by lymph nodes near the stomach and spleen (B2). Metastatic sternal lymph nodes below (B3) and above (B4) the diaphragm were removed as was a peritoneal carcinomatosis (B5). Mice were imaged a second time (C) to guide removal of inguinal lymph nodes (D6) and metastatic tissue extending into the scrotum (D7). All tissues were evaluated for S-nitrosylation of human proteins.

iNOS-S100A8/A9 targets in CDX tumors
We tested the hypothesis that iNOS derived nitric oxide (NO) production in the Hltf-deleted TME promoted metastasis in the CDX model via S-nitrosylation. A snapshot of the entire CDX tumor S-nitroso-proteome in the presence/absence of Hltf in the TME was obtained when iodoTMT enriched S-nitrosylated human proteins were interrogated by nanoLC-MS/ MS. NanoLC-MS/MS-based protein identification provided a comprehensive collection of 136 SNO-proteins in CDX tumors with an Hltf-deleted TME compared to 178 SNO-proteins in CDX tumors with control TME (S1 File). Protein-protein interaction analysis with PANTHER designated cytoskeletal regulation by the Rho GTPase (P00016) the major pathway for CDX tumors from both Hltf-deleted and control TME. However, a secondary pathway of inflammation-mediated by chemokine and cytokine signaling (P00031) increased from 11 protein candidates to 18 protein candidates in CDX tumors with the Hltf-deleted TME. The emPAI value (0.28) for POTEE, the primate-specific POTE (prostate, ovary, testis and embryo expressed) Upregulated genes are shown in red and down regulated genes are shown in blue. Box and whisker plots on the left of each histogram summarize the distribution of all DE genes in a specific pathway or annotated to a GO term out of all of the target genes. The box represents the first quartile, the median and the third quartile. Increased mRNA (A) and protein expression of pro-inflammatory cytokines, IL8 (B) and ICAM-1 (C), occurs in human tumors with an Hltf-deleted TME (40X magnification). There is no IL8 staining in the TME because IL8 is unique to the human genome, i.e. there is no IL8 homologue in the mouse. The log fold increase in message for IL6 and all members of the iNOS-S100A8/A9 signaling axis was documented in the mouse Hltf-deleted TME (D). Log fold changes in message for HP (increase, p = 0.01), SERPINA3 (increase, p = 0.01) and HGF (decrease, p = 0.043) in CDX tumors from Hltf-deleted TME are highlighted with yellow dots in a volcano plot (E). The horizontal axis is the log fold change, and the vertical axis is the negative base-10 logarithm of the p-value. The red-dotted lines represent the threshold. The upper regulated genes (positive long fold change) are shown in red, while the down-regulated genes are blue. Unaffected genes are black.
https://doi.org/10.1371/journal.pone.0251132.g006 ankyrin domain family member E that promotes CRC growth [38], indicates this protein is a general target of S-nitrosylation in CDX tumors from Hltf-deleted and control TME.
Differential expression of TMT-labeled proteins was studied by 2DE with Cy5 label followed by 2D-DIGE (S2 File) and MALDI-TOF/TOF (S3 File). IodoTMT-labeled proteins from Hltf-deleted and control mice were covalently tagged with either Cy3 (Hltf-deleted) or Cy2 (control) for 2D-DIGE (Fig 8). A gel with three fluorescent images (Cy5/Cy2/Cy3) was processed for spot-codetection and protein identification by MALDI-TOF/TOF with a standard database search for human proteins with MASCOT. Proteins-peptides with a total ion C.I.% = 100 are presented in the protein ID summary.
2D-DIGE and protein identification by MALDI-TOF/TOF led to the authentication of iNOS-S100A8/A9 consensus site-specific S-nitrosylated proteins in CDX tumors from Hltfdeleted TME. As shown in Fig 8, spot 21 that is unique to CDX from Hltf-deleted TME has authentic S-nitrosylation of Cys 638 in the consensus sequence motif (I/L-X-C-X-X-D/E) of three members (E/F/I) of the primate-specific POTE gene family comprised of 14 members and subdivided into four groups. Group 3 POTE-actin genes (POTEs E/F/I/J/KP) encode proteins that are actin chimeras with a full-sized long inverted repeat, seven ankyrin repeats and a C-terminal coiled-coil domain. The consensus iNOS-S100A8/A9 target sequence (aa 636-641) immediately precedes the start of the coiled-coil domain (aa 642-698). POTEE-SNO in CDX tumors is both a general S-nitrosylation target and an iNOS-S100A8/A9 site-specific target in the Hltf-deleted TME. When additional very weak spots (30)(31)(32)(33)(34)(35) were added to the analysis, Immunohistochemistry for the iNOS-S100A8/A9 signaling axis. Increased S100A8/A9 was exclusive to the Hltfdeleted mouse TME. S100A8 and S100A9 typically form heterodimers (S100A8/A9, calprotectin) and homodimers are not generally detectable. Increased S100A8/A9 expression coincides with increased iNOS protein levels in the Hltf-deleted mouse TME. Increased endogenous iNOS expression, a major mediator of inflammation, correlates with poor patient survival in CRC due to increased metastasis.
https://doi.org/10.1371/journal.pone.0251132.g007 authentic S-nitrosylation of Cys 238 in the consensus sequence motif (I/L-X-C-X-X-D/E) was identified in the tripartite motif (TRIM) family member TRIM52. Dysregulation of TRIM family members characterized by a tripartite motif-RING domain, one or two B-box domains, and a coiled-coil domain-has been implicated in CRC cell proliferation [39]. This newly identified iNOS-S100A8/A9 target sequence (aa 236-241) is located in the B-box domain. In contrast, all members of the ERM family-ezrin, radixin, and moesin-were found in spot 31 as general targets of S-nitrosylation (Cys 284 ). There was no evidence of iNOS-S100A8/A9 sitespecific S-nitrosylation (Cys 117 ). 2D-DIGE and protein identification by MALDI-TOF/TOF led to the authentication of S-nitrosylation of 53 individual cysteines in half-site motifs of either I/L-X-C or C-X-X-D/E in proteins from CDX tumors from both Hltf-deleted and control TME (Table 1).  Table 1. S-nitrosylated cysteine residues (yellow) in half-site motifs of either I/L-X-C or C-X-X-D/E from CDX tumors from Hltf-deleted (white) and control (green) TME.

Continuity between primary CDX tumors and metastatic sites
Comparative assessment of S-nitrosylation in primary CDX tumors and five metastatic sites was undertaken with iodoTMT-switch labeling and nanoLC-MS/MS. As shown in Fig 5, BLI was used to carefully isolate tumors and lymph nodes, and proteins were selectively labeled with iodoTMT isobaric label reagents as follows: iodoTMT-126, lymph nodes near stomach and spleen; iodoTMT-127, primary tumors; iodoTMT-128, peritoneal carcinomatosis; iodoTMT-129, tumors at the sternum (below and above the diaphragm); iodo-TMT130 lymph nodes in the inguinal region; and iodoTMT-131, lymph nodes extending into the scrotum. As shown in S4 File, 331 proteins were identified in primary CDX tumors and distant metastatic sites. Of that total, 60% were identified in all distant metastatic sites with different ratios of expression compared to primary CDX tumors. In contrast, 40% of the proteins from primary CDX tumors were found in some but not all sites. Protein-protein interaction analysis amongst the shared SNO-proteins with PANTHER designated cytoskeletal regulation by the Rho GTPase (P00016) the major pathway for continuity between CDX tumors and metastatic tumors/lymph nodes in Hltf-deleted TME. The previously identified secondary pathway of inflammation-mediated by chemokine and cytokine signaling (P00031) now includes SNOmyosins and provides continuity between the primary CDX tumor and all distant locations. Functional enrichment assessment with STRING included actin filament-based process, movement, and binding in conjunction with muscle filament sliding and contraction (Fig 9). These findings strongly implicate S-nitrosylation of actins and myosins under physiological conditions in the metastatic dissemination of CDX tumors in an Hltf-deleted TME. Additionally, S-nitrosylation of Cys 813 in UNC45 myosin chaperone B protein was identified in an iNOS-S1008A/A9-specific motif. UNC45B, a co-chaperone for HSP90 required for folding and accumulation of type II myosins [40] has three tetratricopeptide repeats followed by three armadillo (ARM) repeats. ARM 3 at position 751-790 ends prior to the solitary SNO site at Cys 813 . This SNO-protein is present in all distant metastatic sites except the inguinal region. The NF-kB subunit REL is an example of convergence of transcriptomic-S-nitroso-proteomic signaling. REL is transcriptionally activated (log fold increase 1.768, p = 0.01) in primary CDX tumors from an Hltf-deleted TME, and the encoded protein is an S-nitrosylation target (Cys 143 ) in all metastatic sites. However, expression is increased (14-fold) in tissue from the inguinal region compared with the primary CDX tumor. Vinculin-a general adhesion protein-dramatically distinguished the metastatic peritoneal carcinomatosis (39-fold) and the lymph nodes extending into the scrotum (168-fold) from the other metastatic sites. However,  PLOS ONE vinculin appears to be a co-immunoprecipitate because we were unable to authenticate an Snitrosylated cysteine residue.

Discussion
The CDX model-male HCT-116 cells in male mice-is predicated on the fact that male gender is a risk factor in metastatic CRC [41]. General risk factors include age, family history of CRC or inflammatory bowel disease, alcoholism, smoking, obesity, and diabetes. Regardless of lifestyle-specific risks, at any age, the incidence of CRC is higher in men [42]. Hormones in the form of oral contraception and hormone replacement therapy protect women from CRC [43]. There is gender disparity in the anatomical location for CRC. Men have a higher risk of developing left-sided (distal to splenic flexure) CRC compared to women who have a higher risk of developing right-sided (proximal to splenic flexure) CRC [44]. Epigenetic silencing of HLTF occurs more often in men [18]. Experimental models such as this CDX model are invaluable as a CRC metastasis model due to the small sample size and sample amount of relevant human tissue, and the limited availability of matched primary and metastatic samples from individual patients. Not to mention the paucity of normal control tissue. RNAseq and transcriptomic analyses have identified LncRNAs [45,46] that are gaining importance in cancer genomics studies [47]. LncRNAs-nonprotein coding transcripts >200 nucleotides-are evolutionarily conserved and have been implicated in regulatory specificity through interactions with proteins and/or RNA. LNCipedia (v 5.2) lists 10 Lnc-HLTF transcripts of varying lengths. Lnc-HLTF-1, Lnc-HLTF-2 and Lnc-HLTF-3 are cancer-related although no function is currently assigned to them. In contrast, Lnc-HLTF-5 is elevated in thoracic aorta tissue of patients with hypertension, and positively correlated with expanding ascending aortic diameter [48]. This study is the first to report murine Lnc-Hltf-4 in the Hltf +/+ TME of a CDX tumor model. It remains to be determined if silencing Lnc-Hltf-4 expression in the Hltf-deleted TME actively promoted metastasis.
In this study, the proinflammatory mediator S100A8/A9 in combination with species-specific expression of proinflammatory cytokines IL6 (mouse) and IL8 (human) established a pro-metastatic primary tumor niche in the Hltf-deleted TME. Changes in CDX gene expression further promoted metastasis as ICAM-1 increases cancer cell invasion/intravasation into the microvasculature [49] and mediates peritoneal carcinomatosis [50] HP promotes colorectal cancer cell motility [51] and SERPINA3 promotes tumor cell migration and invasion [52]. Decreased HGF may be partly responsible for reduced liver metastasis in this model [53]. Combinatorial induction of IL6 and S100A8/A9 in the TME and ICAM-1 and IL8 in the CDX tumor comprise a positive feedback loop that drives inflammation [54]. These findings support the development of nano-therapeutic tumor-specific and TME-targeted anti-inflammatory therapy for CRC [55,56].
Increased HP known to preserve vascular NO signaling [57] occurred in conjunction with increased iNOS in the Hltf-deleted TME. INOS contributed to the metastatic phenotype via Snitrosylation of cytoskeletal target proteins (actin, myosin). One of the newest components of the TME under investigation is the mechanical microenvironment [58] with an emphasis on actin [59] especially in invadopodia formation [60] where S-nitrosylation likely plays a regulatory role. The metastatic phenotype was promoted via iNOS-S100A8/A9 site-specific S-nitrosylation of UNC45 myosin chaperone B protein that engages in non-muscle myosin (myosin II) assembly [40]. Most intriguing is the enrichment of vinculin in the two sites of peritoneal metastasis [34]-metastatic peritoneal carcinomatosis [61,62] and testicular/scrotal metastasis [63]-considered rare and life-threatening. Vinculin, a coimmunoprecipitate-an intracellular F-actin-binding protein and a mechanotransducer in cell-cell and cell-matrix adhesions [64,65]-has been implicated in tumor progression [66].
Functional protein posttranslational modifications include phosphorylation, ubiquitination and S-nitrosylation. Despite the ongoing search for S-nitrosylation motifs [67], iNOS-S100A8/ A9 signaling is the only known process of S-nitrosylation that is site-selective [26]. There are more than 100 putative targets of iNOS-S100A8/A9 nitrosylation. Of those targets, the five with experimentally validated consensus motifs are Annexin A5, Ezrin, GAPDH, Moesin and Vimentin [26]. As a result of this study, POTEE, TRIM52 and UN45B can be added to the list of proteins with authentic SNO sites in the conserved I/L-X-C-X 2 -D/E motif indicative of iNOS-S100A8/A9-mediated S-nitrosylation. Moreover, it appears there is iNOS-S100A8/A9 half-site selective S-nitrosylation. Jai et al. [26] authenticated the iNOS-S100A8/A9 consensus site with in vitro mutagenesis experiments in which the target cysteine was changed, or knockdown experiments (siRNA) in which individual components of the heteroduplex (S100A8/A9) were eliminated. However, in this in vivo situation, in the presence of both S100A8 and S100A9 proteins, half-site S-nitrosylation was authenticated. Global S-nitrosylation has a role in the regulation of gene transcription [68]. Zinc-fingers motifs are labile to NO where the thio-ligands are vulnerable to S-nitrosylation that can result in zinc release and collapse of the ring-motif resulting in loss of function [69]. HLTF is a zinc finger protein with a highly conserved C3HC4 ring motif (aa760-801); however, we found no evidence of human HLTF-SNO in primary tumors from Hltf +/+ TME. Collectively, these findings support ongoing efforts to find selective iNOS inhibitors as chemo-preventive agents against CRC [23].

Conclusions
In this study, we show for the first time that Hltf-deletion from the TME promotes inflammation in the shared TME-primary tumor niche. Induction of iNOS in the TME produced general and iNOS-S100A8/A9 site-specific S-nitrosylation of previously unidentified human tumor proteins. We establish continuity between global S-nitrosylation of proteins in pathways related to cytoskeleton and motility under physiological conditions in the metastatic dissemination of CDX tumors in an Hltf-deleted TME. We provide the first evidence of cross-talk between increased gene transcription and S-nitrosylation of the encoded protein. This new role for Hltf-deletion in NO-mediated protein S-nitrosylation to promote metastasis extends our understanding of Hltf as a tumor suppressor.
Immune-deficient mice were housed with a 12:12 light/dark cycle with access to food and water ad libitum and bedding was changed 2-3 times/week. Routine testing of sentinel mice ensured the colony was disease free. All studies and the anticipated mortality were conducted in accord with the NIH Guidelines for the Care and Use of Laboratory Animals, as reviewed and approved by the Animal Care and Use Committee at Texas Tech University Health Sciences Center (NIH Assurance of Compliance A3056-01; USDA Certification 74-R-0050, Customer 1481, S1 Checklist). TTUHSC's IACUC (# 02009) specifically approved this study.
The orthotopic HLTF +/+ HCT116 xenograft model was established as follows: randomly selected six-to eight-week old Hltf-deleted (n = 20) and HLTF +/+ (n = 12) male Rag2 -/-IL2rg -/mice received direct orthotopic cell microinjections (OCMI) of HLTF +/+ HCT116 Red-Fluc cells (2x10 6 cells/10 μl) between the mucosa and the muscularis layers of the cecal wall. Hereafter the mice were designated Hltf-deleted and control. All surgery was performed with isoflurane (Isothesia) and the SomnoSuite 1 Low-Flow anesthesia system (Kent Scientific) with far infrared warming pads during surgery and recovery. Additional efforts to minimize suffering included an IP injection of Buprenorphine (Buprenex, 0.1 mg/kg) prior to surgery to manage incisional pain followed by a second dose 4-8 hours later. The cecum was exteriorized via a small midline laparotomy on the vertical linea alba to eliminate bleeding. Non-invasive bioluminescence imaging (BLI) with an IVIS Spectrum In Vivo Imaging System was used to validate the quality and accuracy of the injection, and to track and quantify tumor growth and metastasis. Histopathology at necropsy confirmed placement of the inoculum. Mouse behavior and well-being were monitored daily. Tumor growth/metastasis was monitored weekly with BLI. Immunohistochemistry. Tissue blocks were serially sectioned (3-4 μm). Two sections were placed on each slide and deparaffinized prior to staining. Beginning with the first slide, sections on every fifth slide were stained with hematoxylin and eosin (H&E) for evaluation by light microscopy. Sections on alternate slides were processed for immunohistochemistry with heat-induced epitope retrieval. Two tissue sections per slide facilitated the use of one section for positive immunostaining, and the serial section for negative (minus primary antibody) control staining. All primary antibodies (1:50) were paired with an appropriate HRP-conjugated secondary antibody (1:200) depending upon the species in which the primary antibody was generated. Nuclei were counterstained (blue) with hematoxylin.

Techniques
Immunoprecipitation and Western blotting. Immunoprecipitation and Western analysis were performed as previously reported [33,35]. Briefly, whole cell lysates from two T150 flasks of HCT116 Red-FLuc cells (80% confluent) were immunoprecipitated with rabbit polyclonal anti-HLTF (ab17984) to human HLTF aa 600-700 at a concentration of 5 ug/ml. Western blotting was achieved with the same primary antibody (1:5000) followed by HRPconjugated mouse anti-rabbit (1:5000). Signal was detected by chemiluminescence with the Clarity Western ECL Substrate Kit.

Tumor transcriptome analysis (RNA-seq)
Primary tumor xenografts (1 per individual mouse x 3 biological replicates for Hltf-deleted and control male mice = 6 total samples) were flash frozen and sent to Otogenetics Corp. (Norcross, GA) for RNA-seq assays as previously described [33,[35][36][37]. Briefly, total RNA was isolated, and evaluated for its integrity and purity with an Agilent Bioanalyzer (Table 2). RNA samples were rRNA-depleted prior to random-primed cDNA preparation/QC, Illumina

Tumor S-nitroso-proteome analysis
Step-wise tumor S-nitroso-proteomic analysis leading to iNOS-S100A8/A9 site-specific analysis was performed with iodoTMT-switch labeling [28], affinity enrichment and high-resolution LC-MS/MS analysis by Applied Biomics, Inc (Hayward, CA). Briefly, primary tumor xenografts from Hltf-deleted (n = 2 mice) and control (n = 2 mice) male mice were homogenized/sonicated (Polytron) in 4 volumes HENS buffer (100mM HEPES, pH 7.8, 1 mM EDTA, 0.1 mM Neocuproine, 1% SDS). Protein concentrations were determined (OD 280 ) with Nano-Drop One (ThermoFisher), and adjusted to a final concentration of 2 μg/μl with HENS buffer. Two hundred microgram samples were incubated for 30 minutes at room temperature with 20 mM methyl methanethiosulfonate (MMTS) in dimethylformamide (DMF) to block free cysteine thiols. Proteins were precipitated with 6 volumes pre-chilled (-20˚C) acetone for a minimum of 60 minutes to remove MMTS, pelleted by centrifugation (10,000 x g) for 10 minutes at 4˚C, and dried for 10 minutes. Following sample resuspension in HENS buffer (100 μl), Snitrosylated cysteines were selectively reduced with ascorbate (protected from light) and irreversibly labeled with iodoTMTzero reagent for 2 hours at room temperature. Proteins were precipitated with 6 volumes pre-chilled (-20˚C) acetone for a minimum of 60 minutes, pelleted by centrifugation (10,000 x g) for 10 minutes at 4˚C, and dried for 10 minutes. Samples were resuspended in HENS buffer (100 μl), added to anti-TMT resin and incubated overnight with end-over-end mixing at 4˚C. Resin was washed 3X with 1XTBS, 3X with water, and eluted with 4 volumes TMT elution buffer. Eluates were frozen and lyophilized to dryness. In experiment 1, global S-nitrosylation of CDX tumors from TMT-tagged Hltf-deleted and control samples was analyzed by NanoLC-MS/MS with a data base search for human genes in Swis-sProt using MASCOT. Exponentially Modified Protein Abundance Index (emPAI defined as 10 PAI -1) values, where PAI (Protein Abundance Index) is the ratio of observed to observable peptides, is approximately proportional to the logarithm of protein concentration and indicates absolute protein abundance.
In experiment 2, Hltf-deleted and control TMT-tagged samples were resolved by twodimensional gel electrophoresis (2DE) with mouse anti-TMT as the primary antibody (1 μg/ ml final concentration) and Cy5-labeled donkey anti-mouse IgG (1:2000) as the second antibody. In experiment 3, TMT-labeled proteins were co-labeled with different color fluorescent dyes, i.e. Cy3 for Hltf-deleted and Cy2 for control for two-dimensional difference gel electrophoresis (2D-DIGE), which detected as little as 1.0 fmol of protein in each sample. Spots were excised and proteins identified by matrix-assisted laser-desorption ionization time-of-flight (MALDI-TOF/TOF). Database search was for human proteins in SWISSProt using MASCOT.
In experiment 4, unmodified cysteines in proteins (1 mg/ml) from primary tumor xenografts (n = 4 mice) and metastatic tumors (n = 2 mice) from Hltf-deleted mice were blocked with MMTS, selectively reduced with ascorbate (protected from light), and individually tagged with iodoTMTsixplex isobaric reagent such that a unique mass reporter (126-131 Da) in the low-mass region of the MS/MS spectrum was generated for samples from six locations. Equal amounts of six different samples (1 mg each) were combined into a single sample for enrichment with anti-TMT resin and NanoLC-MS/MS analysis. Relative expression for each protein fragment, i.e. the average ratio(s) for the protein, together with the number of peptide ratios that contributed (N) and the geometric standard deviation (SDgeo) were calculated.
In all experiments, peptides labeled with iodoTMT were quantified. Off-target (non-cysteine) labeling was <5%, and authentic cysteine nitrosylation as well as authentic iNOS/S100A8/ A9 consensus sequences [26] were confirmed by peptide sequence. Gene symbols for differentially expressed proteins were input to the Protein ANalysis THrough Evolutionary Relationships (PANTHER) database [70] for functional classification and pathway identification. Protein-protein networks (cluster analysis) constructed with STRING [71] with EMPAI values identified direct (physical) and indirect (functional) interactions.
Supporting information S1 Checklist. ARRIVE guidelines 2.0 (fillable). (PDF) S1 Table. Met-analysis gene summary identified 30 DE genes in mouse TME. Cuffdiff, part of Cufflinks, uses RPKM values to calculate changes in gene expression. Cuffdiff data were input into iPathwayGuide. Met-analysis calculated long fold change (logFC) and an adjusted p-value (adjpv) for each comparison. The minus sign in the column for logFC shows 12 genes were downregulated and 18 genes were upregulated. (XLSX) S2 Table. Met-analysis gene summary identified 151 DE genes in CDX tumor cells. Cuffdiff, part of Cufflinks, uses RPKM values to calculate changes in gene expression. Cuffdiff data were input into iPathwayGuide. Met-analysis calculated long fold change (logFC) and an adjusted p-value (adjpv) for each comparison. The minus sign in the column for logFC shows 136 genes were downregulated and 15 genes were upregulated. (CSV) S1 File. Excel files for control, test (Hltf-deleted TME) and summary of the first experiment in tumor S-nitroso-proteome analysis. Control and test files each contain three sections: section 1 is the legend (how to read the data sheets); section 2 is the protein peptide summary with peptide sequence, and section 3 is the list of significant hits (95% confidence). The summary file contains a comparative list of proteins with emPAI values for control or test or both. (ZIP)

S2 File. Data from the second and third experiment in tumor S-nitroso-proteome analysis.
Western blot analysis of test and control samples, 2D-DIGE analysis, and spot identification. Spots were excised and proteins identified by matrix-assisted laser-desorption ionization timeof-flight (MALDI-TOF/TOF). (ZIP) S3 File. Excel files for spot analysis by MALDI-TOF/TOF. Analysis was conducted in two phases in which spots with a unique or strong signal were analyzed first followed by analysis of spots with a very weak signal Two files each have a section on how to read MS report, a protein peptide summary and a protein ID summary. Only high confidence matches are reported. (ZIP) S4 File. Excel files for proteins shared by primary and metastatic tumors. There is a legend showing how to read the file, a peptide list and a protein list. The protein list contains ratio data show the relationship of protein expression in the secondary metastatic sites to the primary CDX tumor in the Hltf-deleted TME.