Abstract
Multiple myeloma (MM) is the second most common hematological malignancy. Advancements in therapy have resulted in a five-year survival of 62.4%, however MM remains incurable, and patient relapse is seen as inevitable. BCMA targeted CAR-T therapy has been proven to induce responses in heavily pretreated relapse refractory MM patients with high response rates. However, duration of response has been disappointing. To improve outcomes, we examined the anti-apoptotic mechanisms MM cells engage to survive interacting with CAR-T cells. MM resides in the bone marrow where signaling from the bone marrow microenvironment (BMM) promotes MM survival, proliferation, and drug resistance. We examined the impact of the BMM on anti-apoptotic mechanisms that protect MM cells from CAR-T therapy.
MM cell lines were exposed to BCMA directed CAR-T therapy or rCD95L for 24H with cell death being measured by Annexin V. Bone marrow stroma was modeled using the HS-5 cell line. CD3, CD38, and mCherry positivity were used to differentiate T cells, MM cells, and HS-5 cells respectively in coculture experiments. In experiments involving stromal coculture (SCC) or conditioned media (SCM), MM cells were cultured for 1H before exposure to therapy. Knock out models were generated using CRISPR-Cas9 technology. Caspase 8 activity was measured with an IETD-FMK fluorometric assay.
SCC and SCM both protected KMS12 PE and KMS18 from CAR-T induced cell death. No difference in T cell surface CD107a or TNFa production was observed suggesting protection was not due to lack of T cell activity. To understand the mechanism through which CAR-T cells kill target MM cells we ablated cell surface CD95 expression. Loss of CD95 protected 3 of 4 cell lines examined from CAR-T therapy. To determine if CAR-T cells were utilizing alternative extrinsic cell death pathways such as TRAIL or TNFα we also generated FADD KO KMS18 cells. These cells responded similar to CD95 KO KMS18. We next determined if disruption of the CD95 pathway is the mechanism of stromal-derived protection. SCC and SCM both protected MM cell lines from rCD95L induced cell death. To further our understanding of CD95L induced cell death we utilized MM cells deficient in BAK and BAX (DKO) to eliminate mitochondria-dependent apoptosis. Loss of mitochondrial apoptosis protected KMS18 from rCD95L but not KMS 12 PE suggesting that the former engages type 2 CD95-induced cell death while the later engages type 1 (mitochondria independent) CD95-induced cell death. To determine if the stromal-derived protection was due to alterations in type 2 pathway we exposed DKO cells to SCM and treated them with rCD95L. In the type 2 KMS18 cell line the loss of BAK and BAX did not significantly protect beyond SCM on its own. However, in the type 1 KMS12 PE, loss of BAX and BAX significantly enhanced protection induced by SCM. This indicates that the presence of SCM acted to convert a type 1 cell death signal to a type 2 death signal. To determine the mechanism of this protection, we measured the impact of SCM on caspase 8 activation in response to rCD95L. In all MM cell lines tested SCM inhibited the activation of caspase 8. To isolate CD95L induced cell death in a CAR-T killing assay, we overexpressed the serine protease inhibitor SERPINB9. SERPINB9 over expression protected KMS18 but not KMS12 PE from CAR-T induced cell death. A strong ablation of cellular cytotoxicity was observed when SERPINB9 was overexpressed in KMS18 cells lacking CD95 indicating that the CAR-T cells are killing primarily through degranulation and CD95 engagement. In DKO cells, overexpressing SERPINB9, SCM was able to enhance protection in the type 1 KMS12 PE but not the type 2 KMS18.
These results demonstrate that the BMM can induce MM intrinsic protection against CAR-T therapy and that this protection is due in part to the inhibition of the extrinsic apoptotic pathway. Stromal factors inhibit the extrinsic apoptotic pathway by inhibiting the activation of caspase 8. These results are consistent with a model whereby the reduction in caspase 8 activity switches type 1 to type 2 CD95-induced death. This switch would allow for the death signal to be inhibited by anti-apoptotic BCL2 family members and provides a rational for combining CAR-T therapy with other therapies that impact the intrinsic and extrinsic cell death pathways in MM.
