Abstract

   Cholesterol plays a crucial role in tumor metabolism. Studies have
   shown that the serum cholesterol level of multiple myeloma (MM)
   patients significantly decreases, probably owing to the augmented
   uptake by MM cells. Despite its significance for MM, research on its
   metabolism within MM is limited. Our analysis of clinical data from 703
   newly diagnosed MM patients revealed that low serum cholesterol is
   associated with poor prognosis, and it stems from the elevated
   cholesterol consumption by MM cells. By exploring the transcriptome and
   single-cell RNA-seq data of patients with different cholesterol levels
   in our center, we identified LRP8 as a key regulator of cholesterol
   metabolism in MM, which is closely related to prognosis and disease
   stages. We verified the oncogenic role of LRP8 in vitro and in vivo.
   Knockdown of LRP8 can facilitate apoptosis and cell cycle arrest in MM
   cells. Meanwhile, we employed mouse xenograft tumor model to replicate
   the phenomenon that MM cells with high LRP8 expression consume
   cholesterol, causing low serum cholesterol. Mechanistically, high LRP8
   expression enhances cholesterol utilization and uptake by MM cells;
   LRP8 inhibition reduces cholesterol absorption, further weakening the
   activity of the cholesterol-dependent mTORC1 pathway in MM cells and
   inducing apoptosis. Concurrently, it triggers an upregulation of
   protective autophagy. Further suppression of autophagy can lead to
   extensive apoptosis of MM cells. Our study reveals that LRP8 regulates
   cholesterol metabolism in MM cells and influences the processes of cell
   apoptosis and autophagy through metabolic-related pathways. LRP8 holds
   potential as a therapeutic target for MM.

   Subject terms: Myeloma, Cancer metabolism, Oncogenes, Myeloma

Introduction

   During tumorigenesis, tumor cells undergo metabolic reprogramming to
   meet the energy and material demands. Cholesterol is a vital lipid for
   cellular function, essential for cell proliferation and
   differentiation, as well as intracellular signal transduction [[32]1].
   Cholesterol primarily circulates in the bloodstream in the form of
   low-density lipoprotein (LDL) and is internalized by cells through
   protein-mediated endocytosis [[33]2, [34]3]. Dysregulation of genes
   associated with cholesterol metabolism has been identified in various
   types of tumors, and overexpression of the low-density lipoprotein
   receptor (LDLR) and related proteins (LRP), responsible for mediating
   cholesterol uptake, has been implicated in the development and
   progression of several cancer cell types [[35]4–[36]6].

   The survival and proliferation of MM cells have been found to be
   intricately linked with cholesterol metabolism. Multiple retrospective
   clinical analyses have demonstrated that serum cholesterol levels are
   significantly reduced in MM patients, correlating with disease staging,
   tumor burden, treatment response, and prognosis [[37]7–[38]9].
   Furthermore, statins can impede the survival of MM cells by suppressing
   cholesterol biosynthesis, and enhancing patient prognosis [[39]10,
   [40]11]. The observed low serum LDL in MM patients may stem from
   heightened uptake of LDL by MM cells [[41]8]. Given the elevated energy
   demands of MM due to its own proliferation and survival requisites as
   well as the secretion of immunoglobulin and light chains by plasma
   cells, cholesterol metabolism assumes a pivotal role in supporting
   these processes. However, despite its significance, research on the
   implications of cholesterol metabolism for MM cells has been limited.
   Recently, it has been revealed that cholesterol serves as an essential
   activator factor for the master growth regulator, mTORC1 kinase
   [[42]12–[43]14]. Cholesterol facilitates mTORC1 translocation from the
   cytosol to the lysosomal membrane where it initiates downstream
   programs for biomass production while inhibiting catabolic metabolism
   [[44]15]. Moreover, cholesterol plays a crucial role in regulating
   autophagy in tumor cells [[45]12, [46]16].

   Through an extensive investigation of a large cohort of newly diagnosed
   MM (NDMM) patients in our center, we have observed a significant
   reduction in serum cholesterol and LDL levels among MM patients. The
   decrease in serum cholesterol is associated with the uptake of MM cells
   and represents a substantial prognostic factor for individuals with MM.
   Additionally, low-density lipoprotein receptor-related protein 8
   (LRP8), a member of the LDLR family, has emerged as a pivotal factor
   contributing to dysregulated cholesterol metabolism and promoting the
   progression of MM. LRP8 exerts a significant impact on the metabolism,
   proliferation and survival of MM cells via cholesterol uptake.

Results

The reduction of serum cholesterol levels is related to a poor prognosis for
MM patients

   We analyzed the prognostic value of serum lipids in 703 NDMM patients
   in our center (Baseline characteristics in Supplementary Table [47]1).
   We analyze the prognostic value of seven parameters, including serum
   total cholesterol, LDL, high-density lipoprotein (HDL), apolipoprotein
   A-I (ApoAI), apolipoprotein B (ApoB), apolipoprotein E (ApoE) and
   apolipoprotein a (Apoa). The results showed that cholesterol, LDL, HDL,
   ApoAI, and ApoB were prognostically significant (Fig. [48]1A).

Fig. 1. Serum cholesterol levels and the prognosis of patients with multiple
myeloma.

   [49]Fig. 1
   [50]Open in a new tab

   A The results of the univariate COX regression analysis regarding
   different serum cholesterol indicators and the OS of 703 newly
   diagnosed MM patients (NDMM patient data for cholesterol and prognosis
   studies were from 2014 to 2021. All received first-line therapy with
   core drugs of proteasome inhibitors and/or immune modulators. Only
   ineligible-for-transplantation patients were included. 703 patients
   were included. B, C The median values of patients’ serum cholesterol
   (B) and LDL (C) were utilized as the cut-off thresholds to stratify 703
   patients into two groups, and the Kaplan–Meier survival curves of OS
   for patients in disparate groups were presented. D–F The disparities in
   serum cholesterol indicators among patients with different DS(D),
   ISS(E) and R-ISS(F) stages. G, H The serum cholesterol(G)/LDL(H) levels
   of 355 newly diagnosed MM patients at baseline and after 4 cycles of
   treatment. (A total of 355 patients were recruited from NDMM patients
   who received continuous first-line treatment for 4 cycles from 2016 to
   2021 and had complete efficacy evaluations after the 4 cycles.) The
   results indicated that the patient’s serum cholesterol and LDL levels
   were significantly elevated after 4 cycles of treatment. I, J The serum
   cholesterol(I)/LDL(J) levels of 355 newly diagnosed MM patients with
   different treatment responses at the baseline and after 4 cycles of
   treatment. Among patients with treatment response rates of PR, VGPR,
   and CR, serum total cholesterol and LDL were significantly elevated,
   whereas no significant elevation of serum total cholesterol and LDL was
   observed in patients with treatment response rates of SD and PD).

   Then, we stratified the patients into two groups based on the median
   values of cholesterol and LDL. The survival analysis revealed that
   patients with high serum cholesterol (P < 0.001) and LDL (P < 0.001)
   exhibited a longer overall survival (OS) (Fig. [51]1B, C).

   We posited two potential causes for the decreased serum cholesterol in
   MM patients: the insufficient nutrient intake of patients and the
   cholesterol consumption resulting from tumor growth. However,
   correlation analysis between low serum cholesterol and low BMI
   ( < 18.5) showed no significant association (p = 0.895), ruling out
   malnutrition as a primary cause.

   As the DS staging primarily evaluates the tumor burden of myeloma
   patients, we analyzed the cholesterol levels of patients at different
   DS stages. The findings demonstrated a close association between
   cholesterol levels and DS stage (Fig. [52]1D). We also calculated the
   relationship between cholesterol level and ISS stage (Fig. [53]1E) and
   R-ISS stage (Fig. [54]1F), finding that serum cholesterol level
   decreased as the stage increased.

   The relationship between treatment response and serum cholesterol was
   also examined. Through the analysis of patients after four treatment
   cycles, it was revealed that serum cholesterol significantly increased
   in patients with therapeutic efficacies of PR (p < 0.001), VGPR
   (p < 0.001), and CR (p < 0.001), while no significant changes were
   observed in patients with SD (p = 0.066) or PD (p = 0.127). Similar
   trends were observed for LDL levels (Fig. [55]1G–J).

LRP8 is a crucial regulatory gene for cholesterol metabolism in MM cells and
influences the prognosis of patients

   Using serum cholesterol levels from 703 patients, we established an
   optimal cholesterol reduction cutoff of < 4.06 mmol/L (Fig. [56]2A). To
   explore key regulatory genes in hypercholesterolemic metabolism, bone
   marrow samples from 20 newly diagnosed multiple myeloma (NDMM) patients
   with low serum cholesterol (< 3.06 mmol/L, > 1 mmol/L below the cutoff)
   and 20 with high serum cholesterol (> 5.06 mmol/L, > 1 mmol/L above the
   cutoff) were analyzed. RNA sequencing after CD138+ magnetic bead
   sorting identified 319 differentially expressed genes (via DESeq2)
   (|logFC | >1, FDR < 0.05) (Fig. [57]2B).

Fig. 2. LRP8 serves as the critical gene modulating cholesterol metabolism in
MM cells and is associated with the prognosis, disease stage, and staging of
patients.

   [58]Fig. 2
   [59]Open in a new tab

   A The determination of the cutoff value of serum cholesterol level
   (Determined via the “cutpoint” function of the “Survival” package). B
   The volcano plot of differentially expressed genes in the plasma cells
   from 20 MM patients with conspicuously reduced serum cholesterol and 20
   MM patients with normal serum cholesterol (|logFC | > 1, FDR < 0.05). C
   Univariate Cox analysis identified genes related to cellular
   cholesterol uptake and metabolism that had prognostic value in the TT2
   cohort. D A total of 25 genes related to cholesterol metabolism and
   with prognostic significance were identified in the TT2 cohort. A total
   of 319 differentially expressed genes in the plasma cells of 40
   patients with diverse serum cholesterol levels were determined. The
   intersection gene was LRP8. E, F The KM survival curves of OS(E) and
   PFS(F) for patients in our center with different expression levels of
   LRP8. G, H The serum cholesterol(G) and LDL(H) levels of patients
   presenting with different expression level of LRP8. I, J The KM
   survival curves of OS(I) and EFS(J) for patients in TT2 cohort with
   different expression levels of LRP8. K, L The KM survival curves of
   OS(K) and EFS(L) for patients in TT3 cohort with different expression
   levels of LRP8. M The expression level of LRP8 in plasma cell neoplasms
   at diverse stages in the [60]GSE6477 cohort. LRP8 shows a gradually
   ascending trend in line with the disease stage progression. N The KM
   survival curves of OS for patients in MMRF cohort with different
   expression levels of LRP8. O The expression level of LRP8 in plasma
   cell neoplasms at different ISS stages in the MMRF cohort. LRP8 shows a
   steadily rising trend as tumor stages progress).

   From MSigDB [[61]17], 138 cholesterol metabolism-related genes were
   screened for prognostic value in the TT2 ([62]GSE24080) cohort,
   yielding 25 significant genes (Fig. [63]2C). Subsequently, the
   intersection of the two screening methods was taken, and it was found
   that LRP8 was a gene involved in cholesterol metabolism and had
   significant prognostic value (Fig. [64]2D). Then, further expanded
   validation was carried out in the specimens of a cohort of 108 NDMM
   patients in our center (Baseline characteristics shown in Supplementary
   Table [65]2). Taking the median value of LRP8 expression, the patients
   were classified into two groups. It was discovered that patients in
   high-LRP8-expression group had significantly shorter OS (p = 0.002)
   (Fig. [66]2E) and progression-free survival (PFS) (p = 0.007) (Fig.
   [67]2F). Meanwhile, the serum cholesterol/LDL level of patients in the
   high-LRP8-expression group was lower (Fig. [68]2G, H). Then, separate
   validations were conducted in the TT2 and TT3 cohorts. The findings
   indicated that a high expression level of LRP8 was associated with
   shorter event-free survival (EFS) and OS (Fig. [69]2I–L).
   Simultaneously, the expression level of LRP8 at distinct stages of
   plasma cell neoplasms was verified ([70]GSE6477). The results disclosed
   that during the evolutionary process of monoclonal gammopathy of
   undetermined significance (MGUS), smoldering multiple myeloma (SMM),
   NDMM, and relapsed/refractory multiple myeloma (RRMM), LRP8 exhibited a
   continuously ascending tendency (Fig. [71]2M). Concurrently, the
   prognostic value of LRP8 was validated in the MMRF database (Fig.
   [72]2N), and it was discovered that the expression level of LRP8
   gradually escalated with the increase in the ISS stage (Fig. [73]2O).

Single-cell transcriptome confirmed that serum cholesterol reduction
originated from high cholesterol metabolism of MM, with LRP8 identified as a
key differentially expressed gene

   We conducted single-cell transcriptome sequencing (scRNA-seq) on 4 NDMM
   patients with high serum cholesterol (MM01) and 4 NDMM patients with
   low serum cholesterol (MM02) (with the screening criteria remaining
   consistent as before). We implemented dimensionality reduction,
   clustering and cell identification on the cells (Fig. [74]3A–D) and
   screened out the plasma cell subpopulations (Fig. [75]3E–G). The
   activities of cholesterol metabolism pathways in patients of different
   groups were analyzed using irGSEA. The results revealed that the levels
   of cholesterol endocytosis, cholesterol metabolism, cholesterol
   homeostasis and cellular responses to cholesterol in the low serum
   cholesterol group (MM02) were significantly elevated (Fig. [76]3H).
   This further substantiated that the consumption attributed to the high
   cholesterol metabolism of MM cells was the reason for low serum
   cholesterol. Concurrently, we also ascertained that the expression
   level of LRP8 in MM cells in the MM02 group was significantly higher
   than that in the MM01 group (P < 0.001) (Fig. [77]3I–K).

Fig. 3. The findings of bone marrow single-cell RNA-seq in newly diagnosed
multiple myeloma patients presenting with distinct serum cholesterol levels.

   [78]Fig. 3
   [79]Open in a new tab

   The samples encompassed bone marrow specimens from 4 patients with
   conspicuously decreased serum cholesterol (less than 3.06 mmol/L) and 4
   patients with normal serum cholesterol (greater than 5.06 mmol/L).
   Herein, MM01 was the subgroup featuring normal serum cholesterol, while
   the MM02 cohort belonged to the subgroup with significantly lowered
   serum cholesterol. MM_H01-MM_H04 were assigned to the MM01 subgroup
   (normal cholesterol), and MM_L01-MM_L04 to the MM02 subgroup (low
   cholesterol). A Cell clustering results from single-cell RNA-seq in
   eight newly diagnosed MM patients. B The cell subpopulation profiles of
   bone marrow samples from eight patients with NDMM. C The distribution
   of bone marrow samples from eight NDMM patients. D The distribution of
   bone marrow samples from different group. E The plasma cells in the
   bone marrow specimens of patients from different groups. F, G The
   plasma cells in the bone marrow specimens of patients from MM01
   subgroup(F) and MM02 subgroup(G). H The activation of
   cholesterol-related pathways in the plasma cells of patients in
   different subgroups. The cholesterol metabolism in the plasma cells of
   patients in the low serum cholesterol group (MM02) was significantly
   enhanced, suggesting that the low serum cholesterol resulted from the
   high consumption by MM cells. I, J The expression level of LRP8 in the
   plasma cells of patients within the MM01(I) and MM02(J) subgroup. K The
   expression level of LRP8 in the plasma cells of patients in the MM02
   subgroup was significantly elevated. L The activation states of the
   related pathways of the hallmark gene set in the plasma cells of
   patients in different subgroups. A significant upregulation of the
   mTORC1 pathway and the cholesterol uptake pathway was observed in
   MM02).

The expression level of LRP8 influences the cholesterol uptake capacity of MM
cells

   In the CCLE database, LRP8 manifested a relatively high expression
   level in the majority of MM cell lines (Fig. [80]4A). Subsequently,
   qPCR confirmed LRP8 expression in KMS-11, MM1-S, RPMI-8226, H929, and
   AMO1 cell lines, with the highest levels observed in AMO1 and H929
   (Fig. [81]4B). These two lines were selected for subsequent studies.
   LRP8-overexpressing (LRP8-OE) lentivirus and normal control (NC)
   lentivirus, as well as LRP8-konckdown (shLRP8) lentivirus and NC
   lentivirus were employed to infect the AMO1 and H929 cell lines. We
   adopted qRT-PCR and WB to detect the expression level of LRP8 in
   different cell lines (Fig. [82]4C–F). Cholesterol levels in cell lines
   were measured by ELISA (Fig. [83]4G). The results indicated that LRP8
   expression levels were positively correlated with cholesterol uptake
   capacity in MM cells.

Fig. 4. LRP8 significantly enhances the proliferation, invasion and migration
of myeloma cells.

   [84]Fig. 4
   [85]Open in a new tab

   A The expression level of LRP8 in different MM cell lines in the CCLE
   database. B The expression level of LRP8 in various MM cell lines was
   assessed by RT-PCR. C, D RT-PCR and WB were adopted to assess the
   expression level of LRP8 in the normal control (NC) and
   LRP8-overexpression (OE) AMO1(C)/H929(D) cell lines. E, F RT-PCR and WB
   were adopted to assess the expression level of LRP8 in the NC and
   LRP8-knockdown (shLRP8) AMO1(E)/H929(F) cell lines. G ELISA was
   utilized to determine the cholesterol content in the NC and
   LRP8-knockdown (shLRP8) AMO1/H929 cell lines. The downregulation of the
   expression level of LRP8 conspicuously diminished the cholesterol
   uptake capability of MM cells. H, I The proliferation capacity of NC
   and LRP8-OE AMO1(H)/ H929(I) cell lines (2000 cells/well) were assessed
   via the CCK8 assay. J, K. Transwell assays were conducted to assess the
   migratory potential of NC and LRP8-OE AMO1 (J) and H929 (K) cell lines,
   demonstrating a significant enhancement in migration ability in the
   LRP8-OE group. L, M The invasion ability of NC and LRP8-OE
   AMO1(L)/H929(M) cell lines were assessed via the Transwell assay. The
   invasion ability of MM cells in LRP8-OE group was significantly
   strengthened. N–Q T Subcutaneous xenograft mouse model: Tumor growth
   (N, O), weight (P), and volume (Q) were significantly increased in
   LRP8-OE groups compared to controls).

   CCK-8 assays revealed that LRP8 overexpression significantly enhanced
   MM cell proliferation (Fig. [86]4H, I). Transwell assays demonstrated
   increased migration and invasion capacities in LRP8-OE cells (Fig.
   [87]4J–M). Then, we utilized the LRP8-OE and NC AMO1 cell lines to
   construct the subcutaneous xenograft tumor model in BNGD mice through
   subcutaneous injection. Subsequent analyses indicated that the tumor
   growth in the LRP8-OE group mice was faster (Fig. [88]4N–Q).

   In an attempt to investigate the impact of LRP8 expression levels on
   the biological behavior of myeloma cells, we utilized the LRP8-OE and
   NC AMO1 cell lines to construct a BNGD mouse xenograft tumor model via
   tail vein injection. We employed ELISA detection to quantify serum M
   protein level and serum cholesterol level of mice. We found that,
   compared with the mice in the NC group, mice in the LRP8-OE group were
   lighter (p = 0.014) (Fig. [89]5A, B). The serum M protein level of mice
   showed an overall upward trend over time, and the increase was
   significantly more prominent in LRP8-OE group than in NC group (Fig.
   [90]5C). Concurrently, the results of ELISA assay revealed that serum
   cholesterol of mice exhibited an overall downward trend over time, and
   the decline was significantly faster in LRP8-OE group than in NC group
   (Fig. [91]5D). Histological analyses (HE staining) of femurs confirmed
   MM cell infiltration, with faster growth in LRP8-OE mice (Fig. [92]5E).
   Micro-CT imaging demonstrated exacerbated bone destruction in LRP8-OE
   mice, with significantly lower bone mineral density (BMD, p = 0.005)
   and trabecular thickness (Tr.Th, p = 0.002) compared to controls (Fig.
   [93]5G).

Fig. 5. LRP8 promotes the occurrence of hypocholesterolemia in mouse
xenograft tumor, and downregulation of LRP8 induces cell cycle arrest and
apoptosis in MM cells.

   [94]Fig. 5
   [95]Open in a new tab

   A, B Tail vein-injected mouse xenograft tumor model established with
   the LRP8-OE and NC AMO1 cell lines revealed that the body weight of
   mice in the LRP8-OE group decreased significantly. C M protein levels
   in the mouse xenograft model were measured by ELISA. Results showed
   that serum M protein levels increased more rapidly in the LRP8-OE
   group, indicating faster tumor growth and a higher tumor burden. D
   ELISA assays performed on mouse xenograft tumor models disclosed that
   with the proliferation of MM cells, mice presented a continuous
   lowering of serum cholesterol levels, and the serum cholesterol levels
   in the LRP8-OE group mice were even lower. E The HE staining images of
   femoral sections in the mouse xenograft tumor model disclosed that MM
   cells successfully homed and colonized in the bone marrow, and more
   prominent tumor colonization and a greater tumor burden were noted in
   the LRP8-OE group. F, G The micro-CT assessment of the femoral bone in
   the mouse xenograft tumor model demonstrated that the bone mineral
   density and trabecular thickness of the femur in the LRP8-OE group were
   significantly diminished. H The MTT assay (24 h) indicated that the
   apoptosis of MM cell lines was significantly enhanced subsequent to
   knockdown of LRP8. I–L Flow cytometry demonstrated that knockdown of
   LRP8 in AMO1(I, J) and H929 (K, L) cell lines gave rise to cell cycle
   arrest, accompanied by a marked reduction in the proportion of cells in
   the G2/M phase. M–P Flow cytometry indicated that knockdown of LRP8 in
   AMO1 (M, N) and H929 (O, P) cell lines led to a substantial increase in
   apoptotic cells).

Downregulation of LRP8 induces growth suppression, cell cycle arrest, and
apoptosis in MM cells

   Since LRP8 is generally highly expressed in myeloma, we thus aim to
   further observe the influence that interference with LRP8 exerts on
   myeloma. MTT assays found that the cell viability of the shLRP8 cell
   lines decreased significantly (Fig. [96]5H). Then, the cell cycle of
   shLRP8 cells was examined by flow cytometry. The results revealed that
   knockdown of LRP8 led to abnormal cell cycle arrest. Notably, the G2/M
   phase cells of AMO1 and H929 cells decreased significantly (Fig.
   [97]5I–L). We further detected the cell apoptosis by cytometry flow,
   and the results indicated that suppression of LRP8 induced stronger
   cell apoptosis (Fig. [98]5M–P).

Suppression of LRP8 inhibits the mTORC1 signaling pathway and increases the
level of autophagy in MM cells

   We conducted RNA-seq on the NC and shLRP8 cell lines (Fig. [99]6A). The
   KEGG pathway analysis (Fig. [100]6B) indicated that knockdown of LRP8
   could notably down-regulate the mTORC1 signaling pathway, the E2F
   pathway, and the G2M checkpoint pathway, etc. In addition, there are
   significant differences between the NC and shLRP8 cell lines in the
   mTORC1 pathway, cholesterol metabolism pathways and regulation of
   autophagy pathway (Fig. [101]6C–E). Based on the modifications of the
   metabolism-based signaling pathways, we propose that LRP8 impacts the
   activation of the cholesterol-related mTORC1 signaling pathway through
   regulating cholesterol uptake, and affects autophagy, thereby altering
   the proliferation and apoptosis features of cells (Fig. [102]6F).

Fig. 6. The mTORC1 pathway and regulation of autophagy pathways were
significantly activated in shLRP8 MM cell lines.

   [103]Fig. 6
   [104]Open in a new tab

   A The volcano plot of differentially expressed genes derived from
   RNA-seq detection comparing the NC AMO1 cell line with the shLRP8 AMO1
   cell line (|logFC | > 1, p < 0.05). B KEGG pathway enrichment analysis
   of differentially expressed genes in the NC AMO1 cell line compared
   with the shLRP8 AMO1 cell line. C–E The GSEA analysis results of mTORC1
   (C), cholesterol homeostasis (D) and regulation of autophagy (E) and
   pathways in the NC and shLRP8 AMO1 cell lines. Notably, the mTORC1
   pathways, regulation of autophagy pathways and cholesterol homeostasis
   pathways in the AMO1 cell lines of shLRP8 underwent significant
   alterations. In the cell lines of the NC group, both the mTORC1 pathway
   and the cholesterol homeostasis pathway were conspicuously upregulated
   in comparison with those of the shLRP8 group, whereas the regulation of
   autophagy pathway was significantly downregulated. F Hypothetical
   mechanism diagram of LRP8 influencing intracellular signaling pathways
   in MM cells. LRP8 affects mTORC1 pathway via cholesterol uptake, and
   subsequently impacts the autophagy level and apoptosis of cells. G In
   the WB assay of the AMO1 and H929 cell lines, it was verified that the
   MTORC1 pathway was conspicuously inhibited in the shLRP8 cell lines. In
   contrast to the NC cell lines, the protein quantities of 4EBP, mTOR,
   and p70S6K were analogous in the shLRP8 cell lines, while the protein
   expression levels of p-4EBP, p-mTOR, and p-p70S6K were significantly
   decreased. H Immunohistochemical examination of mouse xenograft tumors
   indicated that the MTORC1 pathway in the tumors of the LRP8-OE group of
   mice was significantly activated. I Immunofluorescence analysis of
   autophagic flux indicated that the autophagy level was significantly
   elevated in shLRP8 AMO1 and H929 cell lines compared with that in the
   NC AMO1 and H929 cell lines. J WB assays revealed that the protein
   level of P62 in the shLRP8 AMO1 and H929 cell lines declined
   significantly, while the protein level of LC3BII/I increased
   significantly, indicating that the level of autophagy was upregulated).

   Then, we detected the expression levels of marker proteins in the
   mTORC1 signaling pathway, including mTOR, 4EBP, and p70S6K, as well as
   phosphorylated mTOR (p-mTOR), phosphorylated 4EBP (p-4EBP), and
   phosphorylated p70S6K (p-p70S6K). The results showed that the
   expression levels of p-mTOR, p-4EBP, and p-p70S6K in shLRP8 cell lines
   were significantly decreased (Fig. [105]6G). Meanwhile, we conducted
   IHC on mouse xenograft tumors and found that the expression level of
   the mTORC1 signaling pathway significantly increased in LRP8-OE group
   (Fig. [106]6H). Furthermore, in scRNA-seq data, a significant
   upregulation of the mTORC1 pathway and the cholesterol uptake pathway
   was also observed in the high LRP8-expression group (MM02) (Fig.
   [107]3L).

   Then, we infected cells with RFP-GFP-LC3 lentivirus and employed
   confocal fluorescence microscopy to detect the autophagic flux. The
   results indicated that the numbers of autophagosomes and autolysosomes
   significantly increased in shLRP8 AMO1 (p < 0.001) and shLRP8 H929
   (p = 0.024) (Fig. [108]6I) cells. Concurrently, we discerned that in
   the shLRP8 cell lines, the expression of LC3BII/I was significantly
   enhanced, while P62 was significantly decreased, inferring that the
   autophagy level was significantly elevated (Fig. [109]6J).

Suppression of LRP8 inhibits the mTORC1 signaling pathway via cholesterol
depletion, thereby resulting in an increase in autophagy

   Firstly, NC AMO1 and H929 cells were treated with methyl-β-cyclodextrin
   (MBCD) for 2 h to deplete cholesterol, followed by 24 h culture with
   varying cholesterol concentrations. Cholesterol depletion significantly
   downregulated mTORC1 pathway proteins, while cholesterol
   supplementation restored p-p70S6K and p-4EBP1 levels in a
   dose-dependent manner (Fig. [110]7A).

Fig. 7. In MM cells, depletion of LRP8 affects cholesterol uptake,
subsequently down-regulating the mTORC1 pathway and up-regulating the
autophagy level.

   [111]Fig. 7
   [112]Open in a new tab

   A NC AMO1 and H929 cells were treated with 50 μM MBCD for 2 h to
   deplete cholesterol in cells, followed by culturing the cells with
   different concentrations of cholesterol for 24 h. The dosages of
   cholesterol supplementation are 0 μM, 40 μM, 60 μM, 80 μM, and 100 μM,
   respectively, in the sequence from left to right. WB analysis revealed
   that the activation of the mTORC1 pathway in the AMO1 and H929 cell
   lines was significantly associated with the intake of cholesterol. B
   shLRP8 AMO1 and H929 cells were processed as follows: untreated group,
   cholesterol group (treated with 100 μM cholesterol for 24 h), MBCD
   group (treated with 50 μM cholesterol for 2 h), and MBCD + cholesterol
   group (treated with 50 μM cholesterol for 2 h followed by 100 μM
   cholesterol for 24 h). NC cells were used as the control. WB detection
   results showed that compared with the NC group, the expression level of
   mTORC1 pathway in the “untreated group” of shLRP8 cell lines was
   significantly decreased. In the shLRP8 cell lines, the expression level
   of mTORC1 signaling in the “cholesterol group” was significantly higher
   than that in the “untreated group”. The expression level of mTORC1
   signaling pathway in the “MBCD group” was the lowest. Additionally, the
   addition of cholesterol reversed the down-regulation of the mTORC1
   signaling pathway in the MBCD group. C The expression ratio of LC3B
   II/I in the NC group cells was significantly lower than that in the
   “untreated group” shLRP8 cells. In shLRP8 cell lines, compared with
   “untreated group”, expression of LC3B II/I in “cholesterol group” was
   significantly decreased, with an increase in the protein expression of
   P62; the expression of LC3B II/I in the “MBCD group” was significantly
   higher than that in the “untreated group”, indicating that cholesterol
   depletion further enhanced cell autophagy; the expression of LC3B II/I
   in the “MBCD + cholesterol group” was significantly lower than that in
   the “MBCD group”, and the protein expression of P62 was increased,
   suggesting that the addition of cholesterol could reverse the promotion
   effect of MBCD on autophagy. D Autophagic flux was assessed via
   immunofluorescence. The outcomes manifested that, in contrast to the NC
   group, the counts of autophagosomes and autolysosomes in shLRP8 cells
   within distinct cholesterol treatment groups augmented. Upon addition
   of cholesterol to shLRP8 cells, the quantities of autophagosomes and
   autolysosomes declined. When cholesterol was depleted by MBCD, the
   numbers of autophagosomes and autolysosomes escalated. The “MBCD +
   cholesterol” treatment reversed the autophagy upregulation witnessed in
   the cells treated with MBCD. E In shLRP8 cell lines, after stimulation
   with MHY1485, the expression levels of p-mTOR, p-p70s6k and p-4ebp were
   significantly enhanced. F Upon treatment of cells with MHY1485, the
   expression ratio of LC3B-II/I was decreased, and the expression level
   of P62 protein was increased. This result implies that the autophagy
   process of the cells is obstructed. G Autophagic flux was determined by
   immunofluorescence, showing that the numbers of autophagosomes and
   autolysosomes within cells were conspicuously decreased following
   MHY1485 stimulation).

   Then, the shLRP8 AMO1 and H929 cells were processed as follows and
   divided into four groups: untreated group, cholesterol group (100 μM
   cholesterol for 24 h), MBCD group (50 μM MBCD for 2 h), MBCD +
   cholesterol group (50 μM MBCD for 2 h followed by addition of 100 μM
   cholesterol for 24 h); untreated NC cells were utilized as controls.
   The results indicated that compared with the NC group, the expression
   level of mTORC1-related proteins in the “untreated group” of shLRP8
   cell lines was significantly decreased. In shLRP8 cell lines, the
   expression level of mTORC1-related proteins in “cholesterol group” was
   significantly increased compared to “untreated group”. Concurrently,
   the expression level of mTORC1-related proteins in the “MBCD group” was
   the lowest. Furthermore, addition of cholesterol reversed the
   downregulation of the mTORC1 signaling pathway in “MBCD group” (Fig.
   [113]7B). This indicated that cholesterol mediates the up-regulation of
   the mTORC1 signaling pathway via LRP8.

   Autophagy-related proteins were then analyzed. LC3B-II/I expression was
   higher and p62 expression lower in “untreated group” shLRP8 cells
   compared to NC controls. In shLRP8 AMO1 and NCI-H929 cell lines,
   compared with “untreated group”, the expression of LC3B II/I
   significantly decreased in “cholesterol group”, while the protein
   expression of P62 increased, suggesting that the addition of
   cholesterol could reverse the promoting effect of LRP8 knockdown on
   autophagy; the expression of LC3B II/I in “MBCD group” was
   significantly higher than that in “untreated group”, indicating that
   cholesterol depletion further enhanced cellular autophagy; the
   expression of LC3B II/I in “MBCD + cholesterol group” was significantly
   lower than that in “MBCD group”, and the protein expression of P62
   increased, suggesting that the addition of cholesterol could reverse
   the promoting effect of MBCD on autophagy (Fig. [114]7C).
   Immunofluorescence further confirmed these results. Compared to NC
   cells, shLRP8 cells displayed an increase in autophagosomes and
   autolysosomes. Upon adding cholesterol to shLRP8 cells, the numbers of
   autophagosomes and autolysosomes decreased. When cholesterol was
   depleted by MBCD, the amounts of autophagosomes and autolysosomes
   increased. The cells under the “MBCD + cholesterol” treatment reversed
   the upregulation of autophagy observed in cells treated with MBCD (Fig.
   [115]7D). These findings suggest that cholesterol mediates the
   regulatory effect of LRP8 on autophagy.

   We further subjected cells to MHY1485 (mTOR agonist) at a concentration
   of 10 μM for a duration of 4 hours. The outcomes of the WB assays
   manifested that in the shLRP8 group of cells upon the stimulation of
   MHY1485, the protein expression magnitudes of p-mTOR, p-p70s6k and
   p-4ebp exhibited a significant increase (Fig. [116]7E). This suggests
   that MHY1485 counteracts the inhibitory impact of LRP8 knockdown on the
   mTORC1. Furthermore, we found that after treating the cells with
   MHY1485, the expression level of LC3B-II/I decreased, the expression
   level of P62 increased (Fig. [117]7F), and the number of autophagosomes
   and autolysosomes decreased (Fig. [118]7G). This suggest that the
   treatment of cells with MHY1485 blocked autophagy, reversed the
   promoting effect of LRP8 knockdown on autophagy in cells.

LRP8 knockdown-induced apoptosis could be aggravated by autophagy inhibitor
3-methyladenine in MM cells

   To further clarify whether the enhanced autophagy induced by LRP8
   knockout in MM cells promotes or inhibits apoptosis, we utilized the
   autophagy inhibitor 3-methyladenine (3MA) to investigate its impact on
   MM cell. Flow cytometry revealed that 3-MA treatment (24/48 hours)
   further suppressed the cell cycle and increased apoptosis in MM cells
   (Fig. [119]8A–H). MTT assays demonstrated significantly reduced cell
   survival upon 3-MA addition (Fig. [120]8I, J). Then, we explored the
   alterations of apoptosis-related proteins. Western blot analysis showed
   that LRP8 knockdown upregulated cleaved caspase-3 and cleaved PARP
   (Fig. [121]8K, L). Moreover, when autophagy was inhibited, cleaved
   caspase 3 and cleaved PARP were further enhanced. These findings
   suggest that the elevated autophagy level resulting from the
   downregulation of LRP8 acts as a protective mechanism employed by MM
   cells under the condition of cholesterol deficiency.

Fig. 8. The decreased cholesterol uptake due to the suppression of LRP8
expression leads to an augmented autophagy level in MM cells, and this
autophagy acts as a protective mechanism under the circumstances of nutrient
deficiency.

   [122]Fig. 8
   [123]Open in a new tab

   A, B Flow cytometry revealed that in the shLRP8 AMO1 cell line, 24 h
   (A)/48 h (B) treatment with 3MA led to a further increase in cell cycle
   arrest, and the proportion of cells in the G2/M phase was even lower
   than before 3MA addition. C, D Flow cytometry revealed that in the
   shLRP8 H929 cell line, 24-hour (C)/48-hour (D) treatment with 3MA led
   to a further increase in cell cycle arrest, and the proportion of cells
   in the G2/M phase was even lower than before 3MA addition. E, F Flow
   cytometry indicated that in the shLRP8 AMO1 cell line, exposure to 3MA
   for 24 h (E)/48 h (F) would cause a further elevation in the proportion
   of apoptotic cells. G, H Flow cytometry indicated that in the shLRP8
   H929 cell line, exposure to 3MA for 24 h (G)/48 h (H) would cause a
   further elevation in the proportion of apoptotic cells. I, J Influence
   of 3-MA on the viability of AMO1 and H929 cell lines via MTT analysis.
   Upon the addition of autophagy inhibitors, the survival rate of MM
   cells decreased. K, L The alterations in the expression levels of
   apoptosis-related proteins in AMO1 (K) and H929 (L) cell lines treated
   with 3-MA were investigated. Compared with NC cells, following the
   downregulation of LRP8 in MM cells, cleaved caspase 3 and PARP were
   significantly upregulated. After the addition of 3-MA, the expression
   levels of cleaved caspase 3 and PARP were further elevated).

Pan-cancer dependency analysis of LRP8

   To further observe whether LRP8 possesses a distinctive prognostic
   value solely in myeloma, we conducted a pan-cancer dependency analysis
   of LRP8 via the DepMap database and calculated the dependency scores of
   LRP8 across different cancer lineages. Through evaluating the
   CRISPR-Cas9 gene effect analysis (CERES score) in DepMap, we discovered
   that LRP8 exhibited an extremely high dependency in lymphoid and
   myeloid tumor cells. Specifically, the highest dependency was observed
   in mature B-cell lymphoma, which once again accentuates the significant
   value of LRP8 in myeloma. Secondly, the second-highest LRP8 dependency
   score was found in B-cell lymphoma. In T-cell lymphoma, LRP8 also
   manifested a high dependency, although it was marginally lower than
   that in B-cell lymphoma. Among solid tumors, some solid tumors
   displayed varying degrees of dependency. LRP8 demonstrated a relatively
   high dependency in colon cancer and ovarian cancer, and also exhibited
   a certain degree of dependency in central nervous system tumors and
   pancreatic tumors. We presented the dependency scores of different
   tumors on LRP8 in Fig.[124]S1 (Fig.[125]S1A). We also displayed the
   expression of LRP8 in different tumor cell lines (Fig.[126]S1B).
   Subsequently, we further explored the prognostic value of LRP8 for
   different tumors. We analyzed the influence of LRP8 on patient OS in
   the TCGA and GTEx databases. We found that LRP8 had significant
   prognostic value in mixed kidney cancer, papillary renal cell
   carcinoma, clear cell renal cell carcinoma, and acute myeloid leukemia,
   etc. (Fig.[127]S2).

Discussion

   Here, we determined the significance and prognostic value of
   cholesterol metabolism for MM patients through large-scale samples. We
   confirmed that the reduction of serum cholesterol was related to the
   tumor burden and cholesterol uptake in MM. Through further exploration,
   we discovered the important role and prognostic value of LRP8 in the
   cholesterol metabolism of MM cells. Our research revealed that LRP8, by
   enhancing the utilization and uptake of cholesterol in MM cells,
   anomalously activates the cholesterol-related mTORC1 pathway within MM
   cells, lowers the autophagy level, and promotes the proliferation and
   cellular activity of MM cells. Inhibiting LRP8 leads to a decrease in
   cholesterol uptake by MM cells, augments cell apoptosis, but induces
   the occurrence of protective autophagy in the cells.

   Cellular metabolism has always been a research hotspot in the field of
   tumors [[128]18]. In MM, although we frequently discover the prognostic
   value of cholesterol and also confirm the inhibitory function of
   statins on the proliferation of MM cells, it has not received
   sufficient attention. The proliferation, survival and biological
   activity of MM tumor cells, coupled with the continuous secretion of
   immunoglobulins and light chains by MM cells, confer greater
   significance on cell metabolism in MM. Our study is the first to deeply
   explore the significance and key regulators of cholesterol metabolism
   in MM and explain the inherent meaning and mechanism of the decreased
   serum cholesterol levels in MM patients.

   The expression level of LRP8 is significantly elevated in types of
   tumors [[129]19–[130]21]. In breast cancer, LRP8 could promote the
   proliferation and migration of tumor cells by activating the
   Wnt/β-Catenin pathway [[131]6, [132]22]. In addition, LRP8 can induce
   ERK1/2 phosphorylation and activate the cell cycle [[133]23].

   Our research has confirmed that in MM, LRP8 activates the mTORC1
   pathway through a cholesterol-dependent route. The mTORC1 pathway is
   one of the most important pathways in the pathogenesis of MM [[134]24,
   [135]25]. We have found that LRP8 activates the mTORC1 pathway by
   highly uptaking cholesterol in MM and further downregulates the
   autophagy level of MM cells in the context of abundant cholesterol
   supply. mTORC1, as an important autophagy regulator [[136]26],
   undergoes multiple metabolism-related changes after receiving the
   cholesterol signal transmitted by LRP8.

   Different from the definite oncogenic role of the mTORC1 in MM, the
   influence of autophagy presenting a dual role [[137]27]. In our
   research, regarding the LRP8-mTORC1-autophagy response, the alterations
   of mTORC1 and autophagy constitute the overall metabolic changes in MM
   cells following the uptake of cholesterol via LRP8. After the
   expression level of LRP8 is upregulated and the uptake of cholesterol
   by MM cells increases, the autophagy level of the cells declines. We
   suggest that this might be a form of energy metabolism regulation under
   a state of sufficient nutrition. When the expression level of LRP8 is
   attenuated, the cholesterol-uptaking capacity of MM cells is
   compromised, the mTORC1 pathway is repressed, facilitating the
   apoptotic process of MM cells and engendering an augmented level of
   protective autophagy within MM cells. Further repression of autophagy
   can elicit a more conspicuous apoptotic phenomenon in MM cells. Our
   discoveries have verified the paramount role of cholesterol metabolism
   in MM and insinuate that LRP8 may potentially serve as a metabolic
   target for the therapeutics of multiple myeloma.

Methods and Materials

Patient characteristics

   The clinical and clinicopathological data of patients were derived from
   a cohort of 703 newly diagnosed multiple myeloma (MM) patients between
   2014 and 2021. These patients received first-line treatments primarily
   involving proteasome inhibitors and/or immunomodulatory drugs, with
   regimens including at least two therapeutic agents.

   For the analysis of LRP8 expression levels and prognosis, bone marrow
   samples were collected from 108 newly diagnosed MM patients between
   2016 and 2020. Only patients who had not undergone autologous
   hematopoietic stem cell transplantation (ASCT) were included. The
   initial treatment regimen for these patients consisted of a three-drug
   combination: a protease inhibitor, lenalidomide, and dexamethasone.
   Bone marrow samples were isolated using CD138 magnetic beads for
   subsequent analyses.

Single cell mRNA sequencing

   Bone marrow samples were obtained from Zhongshan Hospital, Fudan
   University. Mononuclear cells were isolated and used for single-cell
   RNA sequencing (scRNAseq) using Oebiotech’s services in China.
   Single-cell libraries were generated from unsorted bone marrow
   mononuclear cells (BMMCs) using the 10× Genomics Chromium Single Cell
   5′ Library Kit and sequenced on an Illumina platform. FASTQ files were
   aligned to the GRCh38 reference genome using Cell Ranger (v5.0.1).
   Seurat v4.0.1 was employed for clustering and dimensionality reduction.
   Pathway activation analysis was conducted using the irGSEA package and
   AUCell method based on gene sets from the MSigDB database.

Bulk RNA sequencing

   Cells were processed into granules, and RNA was extracted for library
   preparation using Oebiotech’s services. Transcriptomic RNA sequencing
   was performed on the Illumina NovaSeq 6000 platform, generating
   paired-end reads of 150 bp length. Differentially expressed genes were
   identified using the limma software, and Gene Set Enrichment Analysis
   (GSEA) [[138]28] was conducted on these genes.

Cell culture

   The AMO1 and H929 cell lines were sourced from the Institute of Basic
   Medical Sciences, Chinese Academy of Medical Sciences. Cells were
   cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum
   (FBS) under conditions of 37 °C, saturated humidity, and 5% CO[2].

Lentiviral gene transduction

   LRP8-overexpression and LRP8-knockdown lentiviruses were constructed
   using the Ubi-MCS-3FLAG-CBh-gcGFP-IRES-puromycin vector (Hanbio,
   China). Based on viral titer and MOI, the appropriate volume of
   lentivirus was added to the cell suspension. Stable transfectants were
   selected by adding puromycin to achieve a final concentration of
   5 μg/mL. Successfully transformed cell lines were prepared for further
   experiments.

Flow cytometry

   1 × 10^6/ cells were treated with 5 mL of cold ethanol at −20 °C
   overnight for fixation. The cells were resuspended in PBS, supplemented
   with a 10 μl RNase A stock solution, and incubated at room temperature
   for 1 h. Subsequently, the cells were stained with 40 μg/mL propidium
   iodide (PI) solution, and the cell cycle was analyzed using a Flowsight
   flow cytometer (Merck Millipore, Germany). For the cell apoptosis
   experiment, cells treated with 2 μg/mL for 48 h at a density of
   2 × 10^5 cells were collected. The collected cells were then stained
   under dark conditions using Annexin-V-Allophycocyanin (BioLegend, USA)
   and PI for 15 min. Finally, the apoptosis of the cells was detected
   using a flow cytometer.

Quantitative RT-PCR analysis

   RNA was extracted from cells using the EZ Bioscience RNA extraction kit
   according to the manufacturer’s protocol. cDNA was synthesized using
   the Takara cDNA Synthesis Kit. Real-time PCR was performed using the
   Takara TB Green Premix Ex Taq (Tli RNaseH Plus) quantitative PCR kit.
   Primer sequences are provided in Supplementary Table [139]3. Data were
   normalized to means with standard deviations.

Western blotting

   Cells were lysed in RIPA buffer (Thermo Fisher, USA) containing a
   protease inhibitor cocktail (Yeasen, China). Protein extracts were
   subjected to SDS-PAGE electrophoresis and transferred onto PVDF
   membranes (Millipore, Germany). Membranes were blocked with 5% skimmed
   milk and incubated with primary antibodies overnight at 4 °C. Detection
   was performed using HRP-conjugated secondary antibodies and visualized
   with Super ECL detection reagent (ECL BioSharp, China).

Transwell migration and invasion assays

   Transwell chamber with 8.0 μm pore size was used to measure cell
   migration, and Matrigel transwell chamber was used for cell invasion
   tests (Corning, USA). Cells were initially inoculated onto
   top-perforated membranes and incubated for 24 h (migration experiment)
   or 48 h (invasion experiment). After incubation, cells in the upper
   part of the insert were removed and cells adhering to the lower part
   were fixed, stained with crystal violet, photographed at three
   locations, and counted using Image J (NIH, US).

MTT assays

   Cells were quantified by Trypan blue staining and a hemocytometer and
   subsequently seeded into 96-well plates at a density ranging from 3000
   to 4000 cells per well. After the treatments, the cells were subjected
   to MTT for 4 h. Subsequently, the supernatant was withdrawn and the
   formazan crystals were dissolved in 200 μL of dimethyl sulfoxide
   (DMSO). Finally, the optical density was determined at 570 nm.

Immunohistochemistry (IHC) analysis

   Mouse xenograft tumor tissues were paraffin-embedded and sectioned. IHC
   was performed following the manufacturer’s protocol for the IHC kit
   (Sevier, China). Primary antibodies are listed in Supplementary Table
   [140]4.

Detection of autophagy flow

   The mRFP-GFP-LC3 adenovirus (Wanleibio, China) was utilized for
   real-time surveillance of autophagy. Target cells were infected with
   lentivirus. The cells were gathered through centrifugation and counted.
   Five hundred thousand cells were inoculated into each well of the
   6-well plates (1 ml/well), with the virus added at a multiplicity of
   infection (MOI) of 20. Six hours after infection, the cells were
   re-collected by centrifugation, the supernatant was removed, the cells
   were re-suspended in fresh medium, and re-planted back into the
   corresponding wells for continuous culture. 48 hours following virus
   infection, the cells were obtained via centrifugation. The autophagy
   state of the cells was observed and captured by a laser confocal
   microscope (1000X).

Subcutaneous xenograft tumor mouse model

   During logarithmic growth phase, cells were harvested, washed with PBS,
   and resuspended to a density of approximately 5 × 10^6 cells/100 μL.
   These cells were subcutaneously injected into B-NDG mice. Five randomly
   assigned 5-week-old female mice were divided into control (NC) and
   LRP8-overexpression (LRP8-OE) groups.

Xenograft tumor mouse model through intravenous injection

   Cells were harvested during logarithmic growth, washed with PBS, and
   resuspended at a density of 10^7 cells/100 μL. Within 2.5 h, cells were
   intravenously injected into the tail veins of B-NDG mice. Each group of
   five 5-week-old female mice was randomly assigned to either the NC or
   LRP8-OE group. Tumor development was monitored consistently.

Image acquisition and quantitative analysis

   All Western blot membranes were imaged using a chemiluminescence
   imaging system under non-saturating exposure conditions. Protein band
   quantification was conducted using ImageJ software (version 1.54k).
   Specifically, regions of interest (ROIs) were manually outlined around
   the target bands, followed by background subtraction using the “Rolling
   Ball” algorithm with a radius of 50 pixels. Integrated optical density
   (IOD) values were normalized to corresponding loading controls (e.g.,
   β-actin or GAPDH) to account for potential variations in protein
   loading and transfer efficiency. Prior to hypothesis testing, normality
   was verified using the Shapiro-Wilk test, and homogeneity of variance
   was assessed via Levene’s test. Intergroup differences were analyzed
   using two-tailed unpaired Student’s t-tests performed in GraphPad Prism
   version 8.0 (GraphPad Software). Statistical significance was defined
   as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.
   The outcomes are presented in Fig.[141]S3.

Statistical analysis

   Statistical analyses were conducted using R version 4.0.3 and Stata
   statistical software, version 14.0. Two-sided P-values were used with a
   significance threshold of P < 0.05. Data were presented as means ± SD
   and analyzed using GraphPad Prism. Group differences were evaluated
   using Student’s t-test or Wilcoxon test, while survival data were
   visualized with Kaplan–Meier curves and analyzed via log-rank tests.

Supplementary information

   [142]Supplementary Table 1^ (18.5KB, docx)
   [143]Supplementary Table 2^ (18.1KB, docx)
   [144]Supplementary Table 3^ (16.3KB, docx)
   [145]Supplementary Table 4^ (18.4KB, docx)
   [146]Supplementary Figure S1^ (877.7KB, tif)
   [147]Supplementary Figure S2^ (1.6MB, tif)
   [148]Supplementary Figure S3^ (20.5MB, pdf)
   [149]Original Data File^ (7.3MB, pdf)

Author contributions

   The process of conceptualization is undertaken by PL. The research
   approaches were devised and executed by YW and TL. The role of
   assistant is fulfilled by CZ. The task of data collection is
   accomplished by YW and TL. The original drafting is executed by YW and
   TL. The activities of writing review and editing are conducted by PL.

Funding

   Science and Technology Innovation Plan of Shanghai Science and
   Technology Commission (21YF1406300). Natural Science Foundation of
   Shanghai Municipality (22ZR1411400).

Data availabilty

   The single-cell RNA-seq processed gene expression data reported in this
   paper have been deposited into the CNGB Sequence Archive (CNSA)
   [[150]29] of the China National GeneBank DataBase (CNGBdb) [[151]30]
   with accession number CNP0005613(CSE0000446), and raw sequencing data
   have been deposited in the OMIX, China National Center for
   Bioinformation, Chinese Academy of Sciences (accession no. OMIX007547)
   [[152]31]. To request access to raw sequencing data, please apply at
   the Human Genetic Resources Service System of the Ministry of Science
   and Technology ([153]http://apply.hgrg.net/) in accordance with the
   Regulations on the Management of Human Genetic Resources of China. The
   multiple myeloma RNA-seq data [154]GSE24080 were downloaded from the
   GEO database. Analysis code described in the manuscript will be made
   available upon request to the corresponding author pending approval.

Competing interests

   The author declares no competing interests.

Ethics approval and consent to participate

   The research received approval from the Ethics Committee of Fudan
   University Zhongshan Hospital (under the Observation Research Cohort
   for Plasma Cell Diseases, Registration Number: B2017-031R) and adhered
   to the principles outlined in the Helsinki Declaration. Bone marrow
   specimens were obtained from all participants after obtaining their
   informed consent and were stored in the tissue repository at Fudan
   University Zhongshan Hospital. Additionally, the animal studies were
   conducted with the approval of the Animal Ethics Committee of Fudan
   University Zhongshan Hospital.

Footnotes

   Edited by Alessandro Finazzi-Agrò

   Publisher’s note Springer Nature remains neutral with regard to
   jurisdictional claims in published maps and institutional affiliations.

   These authors contributed equally: Yue Wang, Tianwei Lan.

Supplementary information

   The online version contains supplementary material available at
   10.1038/s41419-025-07625-w.

References