APRIL and BCMA promote human multiple myeloma growth and immunosuppression in the bone marrow microenvironment

Abstract

Here we show that overexpression or activation of B-cell maturation antigen (BCMA) by its ligand, a proliferation-inducing ligand (APRIL), promotes human multiple myeloma (MM) progression in vivo. BCMA downregulation strongly decreases viability and MM colony formation; conversely, BCMA overexpression augments MM cell growth and survival via induction of protein kinase B (AKT), MAPK, and nuclear factor (NF)-κB signaling cascades. Importantly, BCMA promotes in vivo growth of xenografted MM cells harboring p53 mutation in mice. BCMA-overexpressing tumors exhibit significantly increased CD31/microvessel density and vascular endothelial growth factor compared with paired control tumors. These tumors also express increased transcripts crucial for osteoclast activation, adhesion, and angiogenesis/metastasis, as well as genes mediating immune inhibition including programmed death ligand 1, transforming growth factor β, and interleukin 10. These target genes are consistently induced by paracrine APRIL binding to BCMA on MM cells, which is blocked by an antagonistic anti-APRIL monoclonal antibody hAPRIL01A (01A). 01A is cytotoxic against MM cells even in the presence of protective bone marrow (BM) myeloid cells including osteoclasts, macrophages, and plasmacytoid dendritic cells. 01A further decreases APRIL-induced adhesion and migration of MM cells via blockade of canonical and noncanonical NF-κB pathways. Moreover, 01A prevents in vivo MM cell growth within implanted human bone chips in SCID mice. Finally, the effect of 01A on MM cell viability is enhanced by lenalidomide and bortezomib. Taken together, these data delineate new molecular mechanisms of in vivo MM growth and immunosuppression critically dependent on BCMA and APRIL in the BM microenvironment, further supporting targeting this prominent pathway in MM.

© 2016 by The American Society of Hematology.

Figures

Figure 1

BCMA promotes MM cell proliferation and survival. (A) MM1R cells were infected with lentiviruses expressing shBCMA or control shRNA (shCnt) followed by these assays: CellTiter-Glo for viability, Caspase-Glo 3/7 for apoptosis, and methylcellulose-based colony formation. (B) BCMA mRNA levels were measured by real time qRT-PCR (upper) and viable cell number were determined by trypan blue (lower) in BCMA-overexpressing RPMI8226 (R-BCMA) and empty control vector-expressing RPMI8226 (R-empty) cells which were infected with shBCMA or shCnt lentivirus. Cell growth and survival were determined by (C) DNA synthesis and (D) CellTiter-Glo in paired (C) R-BCMA/R-empty MM and (E) Daudi-BCMA/Daudi-empty B lymphoma cells. (D) Colony formation is determined in R-BCMA vs R-empty cells. (F) Protein lysates were examined by immunoblotting using indicated (F) Abs and (G) Kinex KAM-880 antibody microarray using Kinexus Bioinformatics. (G) Most significant 29 protein hits in R-BCMA relative to R-empty cells are presented. (H) Nuclear extracts were assayed for p65, p50, and p52 DNA binding induction to determine NF-κB activity. *P < .01, **P < .001 Shown are means and standard error of the means of ≥3 independent experiments.

Figure 2

BCMA overexpression enhances MM tumor growth in vivo via upregulation of multiple MM-related genes and immunosuppressive molecules. (A) Mice (n = 6) were injected subcutaneous with equal numbers of R-BCMA (solid line) or R-empty (control vector, dotted line) cells, and tumor development was monitored. *P < .002. Tumors were removed at day 40. (B) Immunohistochemistry for tumor sections with similar volume from additional mice were used for CD31/MVD assessment. Original magnification, ×200. (C) Fold increase in 19 mRNA levels in R-BCMA vs R-empty cells was determined by real-time qRT-PCR. GAPDH and 18S were used as internal controls. (D) Programmed death ligand 1 (PD-L1) increase in R-BCMA vs R-empty cells was further confirmed with immunofluorescence staining (left; PD-L1, pink and DAPI, blue) and immunoblotting (right). (E) RNA of R-BCMA– and R-empty–derived tumors was further assayed for IL-10 pathway-associated genes using TaqMan array human IL10 pathway. Transcripts were normalized by geomean of 4 internal controls, and fold increases in R-BCMA vs R-empty cells are shown.

Figure 3

01A blocks proliferation and viability of MM cells induced by paracrine APRIL. (A) Supernatant was collected from CD14+ cells purified from MM patients (1-5) BM aspirates cultured with media (−, open column), M-CSF (M) (gray column), or M with receptor activator of nuclear factor kappa-B ligand (M+R, black column). APRIL secretion is measured by specific ELISA. APRIL mRNA was also confirmed by real-time qRT-PCR in 1 representative sample (left). (B) Daudi-BCMA/Daudi-empty cells, (C) IL-6–dependent INA6, IL-6–independent MM1S, and (D) RPMI8226, as well as (E) CD138+ cells from a relapsed MM patient, were cultured for 3 days with APRIL in the presence or absence of anti-APRIL 01A mAb (10 μg/mL), followed by (B,E) [3H]thymidine uptake and (C-D) luminescence-based viability assays. Two additional patient MM cells were subject to viability and caspase 3/7 activity (F; newly diagnosed) and annexin V/PI (G; relapsed) flow cytometry.

Figure 4

01A inhibits various APRIL-induced signaling cascades and further blocks APRIL-induced adhesion and migration in MM cells. (A) MM1R cells were serum starved, pretreated with 01A, and stimulated with APRIL (100 ng/mL). Cell lysates were made and subjected to immunoblotting with specific Abs. MM1S and H929 cells were also treated with indicated reagents for 4 hours. (B,D) Cell lysates and (C) mRNA were assayed by immunoblotting and real-time qRT-PCR, respectively. (D) NF-κB DNA binding activity assays were performed in cell lysates (right). (E) MM1S and H929 cells, (F) as well as CD138+ patient MM cells, were subjected to adhesion assays on BMSC-coated plates, in the presence or absence of 01A and APRIL. (G) H929 and INA6 MM cells were incubated with 01A in APRIL-induced migration assays.

Figure 5

01A significantly decreases many APRIL-induced target genes and blocks MM cell growth in a SCID-hu mouse model of human MM. (A) MM1S and H929 MM cells were incubated overnight with (+) or without (−) APRIL in the presence (+) or absence (−) of 01A. Levels of downstream target genes were assayed using real-time qRT-PCR, normalized to 18S or GAPDH internal controls. Fold changes to control media are shown. Two-way ANOVA test (Tukey’s multiple comparison) was done for all genes except BIRC3 and IL-8 (1-way ANOVA). Fold changes are significantly higher on APRIL induction in H929 vs MM1S cells for BIRC3 (24-fold vs 2-fold) and IL-8 (8-fold vs 1.3-fold). *P < .05; **P < .01; ***P < .001; ***P < .0001. (B) INA-6 MM cells were injected into the implanted human bone chips in SCID-hu mice. Two days later, mice were treated (5 days/week) with 01A (20 mg/kg, n = 4) or isotype IgG control (n = 4); tumor growth was then monitored by measuring shuIL-6R in mouse serum samples weekly. Two-way ANOVA test (Tukey’s multiple comparison) determined P < .0001 at day 32.

Figure 6

01A inhibits MM cell proliferation and survival induced by osteoclasts, macrophages, and pDCs. (A) MM cell lines (n = 5) and (B) patient MM cells (n = 3) were cultured with osteoclasts (OCs), with or without 01A, followed by DNA synthesis assays. MM cells were cultured with macrophages (Mϕ), in the presence or absence of 01A (20 μg/mL) with or without APRIL, for (C) 1 or (D) 21 days, followed by (C) viability and (D) colony formation assays. Macrophages do not form colonies under these culture conditions. (E) pDCs were purified from patient BM aspirates (n = 3) and incubated with APRIL in the presence of 01A for 2 days. (F) MM cells were cultured with pDCs in the presence of APRIL and 01A for 3 days.

Figure 7

01A selectively blocks APRIL-induced protection in MM cells, alone and in cocultures with OCs, an effect that is further enhanced by lenalidomide or bortezomib. (A) U266 cells were incubated with APRIL or IL-6 in the presence or absence of lenalidomide, with or without 01A (10 μg/mL), and assayed by [3H]thymidine uptake at day 3. MM cell lines were cultured for 4 days in 2% fetal calf serum growth media with OCs in the presence of 01A with either (B) lenalidomide or (C) bortezomib. *P < .01.