BACKGROUND CONTEXT: Modic changes (MCs) are magnetic resonance imaging (MRI) evidence of inflammatory and fibrotic vertebral bone marrow lesions that associate with adjacent disc degeneration and end plate damage. Although MC etiology is uncertain, historical data suggest a linkage to an autoimmune response of bone marrow triggered by the nucleus pulposus (NP). PURPOSE: The aim of this study was to test whether bone marrow has an autoimmune response to NP cells that is amplified by an inflammatory milieu and ultimately leads to MC development in vivo. We hypothesized that an inflammatory co-stimulus is required for bone marrow/NP crosstalk to stimulate MC. STUDY DESIGN: This is an in-vitro cell co-culture study plus in-vivo experiments in rat caudal vertebrae. METHODS: In in-vitro study, bone marrow mononuclear cells (BMNCs) and NP cells (NPCs) from rats were co-cultured with and without interleukin (IL)-1α stimulation. Cell viability (n=3) of BMNCs and NPCs and gene expression (n=7) were analyzed. In in-vivo study, proinflammatory lipopolysaccharide (LPS) and control disc nucleus surrogates (NP micromass pellets) were generated in vitro from rat NPCs and implanted into rat tail vertebrae, and the response was compared with sham surgery (n=12 each). Tissue changes were investigated with T1w and T2w MRI (7T), histology, and immunohistochemistry (tumor necrosis factor, CD3) 1 (n=6) and 2 weeks (n=6) after implantation. RESULTS: BMNC/NPC co-culture significantly increased lymphocyte viability (42%-69%, p<.05) and reduced NPC viability (96%-88%, p<.001), indicating immunogenicity of NPC. However, IL-1α was required to cause significant transcriptional upregulation of IL-1, IL-6, IL-10, and tropomyosin receptor kinase A. Therefore, an inflammatory activation is required to amplify the immune response. Immunogenicity of the NP was corroborated in vivo by CD3 cell accumulation around LPS and control disc surrogates at Day 7. However, only the LPS disc surrogate group demonstrated infiltration of CD3 cells at Day 14. Furthermore, end plate defects (p<.05, LPS: n=4/6, Ctrl: n=0/6, sham: n=0/6) and MC1-like MRI changes (T2w hyperintensity, p<.05) were only seen with LPS disc surrogates. CONCLUSIONS: NPCs are immunogenic but cannot trigger MC without an additional proinflammatory stimulus. Our data suggest that MC requires end plate defects that allow marrow/NPC co-mingling plus an adjacent inflammatory "MC disc" that can amplify the immune response.