Background: Natural Killer (NK) cells play a critical role in host defense. Studying human NK immunobiology is mainly focused on using in vitro assays with limited NK cells from peripheral blood. It is challenging to study human NK cell biology in vivo due to potential ethical issues in human study and the lack of suitable animal models. Developing a suitable animal model to study human NK cell biology in vivo is critical to support NK-based clinical immunotherapy. Results: Here, we develop a novel method to study human NK cells in vivo by using hu-BLT (humanized bone marrow-liver-thymus) mice that constitutively express human IL-15 (henceforth, hu-BLT-IL15). We also compare human NK cells between hu-BLT-IL15 and hu-BLT mice without IL-15 expression by a newly developed approach for the rapid propagating of primary human NK cells from various sources (including peripheral blood, spleen, and bone marrow). NK cells from hu-BLT-IL15 show superior number, purity, and cytotoxicity (including natural cytotoxicity and antibody-dependent cellular cytotoxicity [ADCC]), compared with NK cells from hu-BLT. Unexpectedly, we also identify a significantly increased percentage of NK-like T cells (CD3+ CD16+ CD56+) from hu-BLT-IL15, indicating that IL-15 signaling enhances both NK and NKT cell development. Conclusions: A better understanding of the immunobiology of the NK-like T cells in the hu-BLT-IL15 mouse model may provide critical information for determining the clinical value of these cells in predicting disease progression. Thus, we propose that the hu-BLT-IL15 mouse model in combination with the 721.221-mIL21 feeder cell expansion system can serve as a superior model to study human NK and NK-like T cells in comparison with hu-BLT.

Hepatocellular carcinoma (HCC) is the most prevalent form of primary liver cancer and ranks as the second leading cause of cancer-related mortality globally. Despite advances in current HCC treatment, it remains a malignancy with poor prognosis. Therefore, developing novel treatment options for patients with HCC is urgently needed. Chimeric antigen receptor (CAR)-modified natural killer (NK) cells have shown potent anti-tumor effects, making them as a promising immunotherapy strategy for cancer treatment. Glypican-3 (GPC3), a cell surface oncofetal glycoprotein, is highly expressed in most HCC tissues, but not in normal tissues, and functions as a key driver of carcinogenesis. Given its high expression level on the cell surface, GPC3 is considered as an attractive immunotherapy target for HCC. In this study, two GPC3-specific CAR-NK cells, NK92MI/NH3 and NK92MI/HS20, were established using NK92MI cells, a modified IL-2-independent NK cell line. These cell lines were engineered with third generation GPC3-specific CAR, and their activities were subsequently evaluated in the treatment of HCC. We found that NK92MI/NH3 cells, rather than NK92MI/HS20 cells, exhibited a significant cytotoxicity effect against GPC3 + HepG2 cells in vitro and efficiently suppressed tumor growth in a xenograft model using NSG mice. In addition, irradiated NK92MI/NH3 cells displayed similar anti-tumor efficacy to unirradiated NK92MI/NH3 cells. Furthermore, we observed that NK92MI/NH3 cells showed higher killing activity against the GPC3 isoform 2 overexpression cell line (SK-Hep1-v2) than those with GPC3 isoform 1 overexpression cell line (SK-Hep1-v1) both in vitro and in vivo. This suggest that the presence of different GPC3 isoforms in HCC may impact the cytotoxicity activity of NK92MI/NH3 cells and potentially influence therapeutic outcomes. These findings highlight the effective anti-HCC effects of NK92MI/HN3 cells, as well as the underlying therapy resistance, suggesting their potential as a promising therapy for HCC.