Abstract Background The increasing use of metal nanoparticles in industry and biomedicine raises the risk for unintentional exposure. The ability of metal nanoparticles to penetrate the skin ranges from stopping at the stratum corneum to passing below the dermis and entering the systemic circulation. Despite the potential health risks associated with skin exposure to metal nanoparticles, the mechanisms underlying the toxicity of metal nanoparticles on skin keratinocytes remain unclear. In this study, we proposed that exposure of human epidermal keratinocytes (HaCaT) to metal nanoparticles, such as nickel nanoparticles, dysregulates tight-junction associated proteins by interacting with the HIF-1α/miR-29b/MMPs axis. Methods We performed dose-response and time-response studies in HaCaT cells to observe the effects of Nano-Ni or Nano-TiO2 on the expression and activity of MMP-2 and MMP-9, and on the expression of tight junction-associated proteins, TIMP-1, TIMP-2, miR-29b, and HIF-1α. In the dose-response studies, cells were exposed to 0, 10, or 20 μg/mL of Nano-Ni or Nano-TiO2 for 24 h. In the time-response studies, cells were exposed to 20 μg/mL of Nano-Ni for 12, 24, 48, or 72 h. After treatment, cells were collected to either assess the expression of mRNAs and miR-29b by real-time PCR or to determine the expression of tight junction-associated proteins and HIF-1α nuclear accumulation by Western blot and/or immunofluorescent staining; the conditioned media were collected to evaluate the MMP-2 and MMP-9 activities by gelatin zymography assay. To further investigate the mechanisms underlying Nano-Ni-induced dysregulation of tight junction-associated proteins, we employed a HIF-1α inhibitor, CAY10585, to perturb HIF-1α accumulation in one experiment, and transfected a miR-29b-3p mimic into the HaCaT cells before Nano-Ni exposure in another experiment. Cells and conditioned media were collected, and the expression and activities of MMPs and the expression of tight junction-associated proteins were determined as described above. Results Exposure of HaCaT cells to Nano-Ni resulted in a dose-dependent increase in the expression of MMP-2, MMP-9, TIMP-1, and TIMP-2 and the activities of MMP-2 and MMP-9. However, exposure of cells to Nano-TiO2 did not cause these effects. Nano-Ni caused a dose-dependent decrease in the expression of miR-29b and tight junction-associated proteins, such as ZO-1, occludin, and claudin-1, while Nano-TiO2 did not. Nano-Ni also caused a dose-dependent increase in HIF-1α nuclear accumulation. The time-response studies showed that Nano-Ni caused significantly increased expressions of MMP-2 at 24 h, MMP-9 at 12, 24, and 48 h, TIMP-1 from 24 to 72 h, and TIMP-2 from 12 to 72 h post-exposure. The expression of miR-29b and tight junction-associated proteins such as ZO-1, occludin, and claudin-1 decreased as early as 12 h post-exposure, and their levels declined gradually over time. Pretreatment of cells with a HIF-1α inhibitor, CAY10585, abolished Nano-Ni-induced miR-29b down-regulation and MMP-2/9 up-regulation. Introduction of a miR-29b-3p mimic into HaCaT cells by transfection before Nano-Ni exposure ameliorated Nano-Ni-induced increased expression and activity of MMP-2 and MMP-9 and restored Nano-Ni-induced down-regulation of tight junction-associated proteins. Conclusion Our study herein demonstrated that exposure of human epidermal keratinocytes to Nano-Ni caused increased HIF-1α nuclear accumulation and increased transcription and activity of MMP-2 and MMP-9 and down-regulation of miR-29b and tight junction-associated proteins. Nano-Ni-induced miR-29b down-regulation was through Nano-Ni-induced HIF-1α nuclear accumulation. Restoration of miR-29b level by miR-29b-3p mimic transfection abolished Nano-Ni-induced MMP-2 and MMP-9 activation and down-regulation of tight junction-associated proteins. In summary, our results demonstrated that Nano-Ni-induced dysregulation of tight junction-associated proteins in skin keratinocytes was via HIF-1α/miR-29b/MMPs pathway.