Aberrant NF-kB signaling fuels tumor growth in multiple human cancer types including both hematologic and solid malignancies. Chronic elevated alternative NF-kB signaling can be modeled in transgenic mice upon activation of a conditional NF-kB-inducing kinase (NIK) allele lacking the regulatory TRAF3 binding domain (NT3). Here, we report that expression of NT3 in the mesenchymal lineage with Osterix (Osx/Sp7)-Cre or Fibroblast Specific Protein (Fsp1)-Cre caused subcutaneous, soft tissue tumors. These tumors displayed significantly shorter latency and a greater multiple incidence rate in Fsp1-Cre;NT3 compared to Osx-Cre;NT3 mice, regardless of sex. Histological assessment revealed poorly differentiated solid tumors with some spindled patterns, as well as robust RelB (Relb) immunostaining, confirming activation of alternative NF-kB. Even though NT3 expression also occurs in the osteolineage in Osx-Cre;NT3 mice, we observed no bony lesions. The staining profiles and pattern of Cre expression in the two lines pointed to a mesenchymal tumor origin. Immunohistochemistry revealed that these tumors stain strongly for SMA (Acta2), although vimentin (Vim) staining was uniform only in Osx-Cre;NT3 tumors. Negative CD45 (Ptprc) and S100 immunostains precluded hematopoietic and melanocytic origins, respectively, while positive staining for CK19 (Krt19), typically associated with epithelia, was found in subpopulations of both tumors. Principal component, differential expression, and gene ontology analyses revealed that NT3 tumors are distinct from normal mesenchymal tissues and are enriched for NF-kB related biological processes. We conclude that constitutive activation of the alternative NF-kB pathway in the mesenchymal linage drives spontaneous sarcoma and provides a novel mouse model for NF-kB related sarcomas.
Overall design: Tumors were excised to remove any gross overlaying tissue, and either stored at -20C in RNAlater (R0901; Sigma, USA) or flash frozen in liquid nitrogen, and stored at -80C. Frozen tumors were pulverized in liquid nitrogen followed by RNA extraction as previously described (PMID: 31246323). Samples were screened by quantitative real-time PCR (qPCR) for blood- (Hba1, Hba2, and Hbb) and monocyte-specific (CD68) markers to assess low peripheral blood content. Gene expression was normalized to the levels of Gapdh, a housekeeping gene. A Nanodrop device (ND-2000; ThermoFisher, USA) and other RNA quality assessments (Bioanalyzer 2100; Agilent Technologies, USA) were used to determine RNA concentration and RNA integrity number (RIN score). Samples with sufficient RNA yields as well as RIN scores and DV200 scores were prioritized for RNA-Seq analysis as previously described (PMID: 32240779 and 24905804). Final samples for analysis (n=3 each genotype, all female, Osx-Cre;NT3 = 3 trunk or FSP1-Cre;NT3 = 1 trunk and 2 perineal) had an average RIN score of 7.7 for Osx-Cre;NT3 and 8.5 for FSP1-Cre;NT3.
Less...