Abstract
Gut microbiome influences tumorigenesis and tumor progression through regulating the tumor microenvironment (TME) and modifying blood metabolites. However, the mechanisms by which gut microbiome and blood metabolites regulate the TME in multiple myeloma (MM) remain unclear. By employing16S rRNA gene sequencing coupled with metagenomics and ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry, we find that Lachnospiraceae are high and phosphatidylcholine (PC) are low in MM patients. We further show that Lachnospiraceae inhibits PC production from MM cells and enhances cytotoxic CD8 T cell function. Mechanistically, PC promotes Sb9 mRNA maturation in MM cells by LIN28A/B via lysophosphatidic acid, thus enhances exosamal Sb9 production. Exosamal Sb9 then reduces GZMB expression by suppressing tumor protein p53 (TP53) UFMylation via the competitive binding of TP53 with the ubiquitin-fold modifier conjugating enzyme 1 in CD8 T cells. We thus show that Lachnospiraceae and PC may be potential therapeutic targets for MM treatment.
Similar content being viewed by others
Introduction
Multiple myeloma (MM) is the second most common hematological malignancy, accounting for 10% of all hematologic malignancies1,2. It is characterized by the infiltration of malignant plasma cells secreting monoclonal immunoglobulins into the bone marrow, leading to renal damage, bone lesions, hypercalcemia, and anemia3,4. Over the last two decades, significant therapeutic advances have improved the outcomes of patients with MM by introducing novel agents, including proteasome inhibitors and immunomodulatory drugs4,5,6. However, MM remains incurable in most cases4. Therefore, the mechanisms regulating MM development and progression must be elucidated to develop effective strategies for MM treatment.
Gut microbiome plays a critical role in the tumorigenesis and progression of various neoplasms by regulating the tumor microenvironment (TME) by modulating metabolism and immunity4,7,8,9. Gut microbiome also contributes to the development and progression of MM by controlling TME8. For instance, Prevotella heparinolytica promotes MM progression by inducing Th17 cell differentiation10. Modulating the TME by targeting the gut microbiome and thereby enhancing antitumor immunity is a promising strategy for MM treatment. However, the mechanisms by which the gut microbiome regulates the TME in MM remain unclear.
The gut microbiome exerts considerable effects on blood metabolites11, such as lipids12. Phosphatidylcholine (PC)-derived lipid mediators contribute to the immunosuppressive TME by modulating immune cells13. Lysophosphatidic acid (LPA), which is derived from the turnover of PC14, impedes antitumor immunity by suppressing CD8 T-cell infiltration and cytotoxic function15,16. The ratio of granzyme B (GZMB)+CD8 T cells decreases in patients with MM17. GZMB + CD8 T cells kill cancer cells via cytotoxic mediators regulated by GZMB18,19. However, the effects of PC and LPA on GZMB + CD8 T cells in MM have not yet been elucidated.
SerpinB9 (Sb9) is an inhibitor of GZMB, and its suppression has been demonstrated to enhance the efficacy of GZMB-based tumor immunotherapy20,21. Additionally, Sb9 is upregulated in MM cells and positively correlated with the resistance of proteasome inhibitors22. However, the roles of PC and LPA in the Sb9 expression of MM cells and the function of Sb9 derived from MM cells in GZMB + CD8 T cells remain largely unknown.
Therefore, this study aims to determine whether gut microbiota-reprogrammed PC and LPA suppress the cytotoxic effects of GZMB + CD8 T cells by regulating MM cell-derived Sb9. Results indicate that Lachnospiraceae-inactivated PC from MM cells inhibits cytotoxic CD8 T cell function by promoting Sb9 mRNA maturation and exosamal Sb9 production via LPA in MM cells, and exosamal Sb9 reduces GZMB expression by suppressing tumor protein p53 (TP53) UFMylation in CD8 T cells. These findings provide potential novel targets and strategies for MM treatment.
Results
Gut microbiome taxonomic indicators of MM using 16S rRNA gene sequencing
To investigate the differences in the gut bacterial community composition between patients with MM (n = 20, MM group) and healthy controls (n = 38, HC group), we performed 16S rRNA gene sequencing. A total of 554,596 high-quality sequences were obtained after filtration. In addition, 823 operational taxonomic units (OTUs) were identified. Alpha diversity analysis indicated significant differences in the observed OTUs and Chao1 index but not in the Shannon and Simpson indices between the MM and HC groups (Fig. 1a). These results indicated that the microbiota abundance in HC was higher than that in MM patients, yet there was no difference in microbiota diversity between HC and MM patients. Principal component analysis (PCA) results revealed a significant difference in the overall composition between the two groups (Fig. 1b). Additionally, a cladogram (Fig. 1c) and histogram (Fig. 1d) generated by linear discriminant analysis combined with effect size (LEfSe) analysis demonstrated that in the MM group, the abundances of genera belonging to Enterobacteriales, Streptococcaceae, Pseudomonadaceae, Fusobacteriaceae, Xanthomonadaceae, Caulobacteraceae, and Atopobiaceae increased, whereas the abundances of genera belonging to Ruminococcaceae, Bacteroidaceae, Lachnospiraceae, Christensenellaceae, Rikenellaceae, Muribaculaceae, and Deferribacteraceae decreased (Supplementary Data 1). A heatmap of the relative abundances of genera between the two groups is presented in Supplementary Fig. 1.
a Differences in alpha diversity between the HC and MM groups based on the observed species, Chao1, Shannon, and Simpson indices. b PCA performed using log-cumulative-sum scaling at the sequence variant level with read counts transformed. c Cladogram showing the phylogenetic distribution of gut microbiota associated with the HC or MM group. d Histogram of the LDA scores indicating the effective size and ranking of each differentially abundant taxon. MM patients with multiple myeloma, HC healthy controls, PCA principal component analysis, LDA linear discriminant analysis. Source data are provided as a Source Data file.
Gut microbiome taxonomic indicators of MM using metagenomics
Stool samples of five patients from the MM group and five individuals from the HC group included in 16S rRNA gene sequencing were further analyzed using metagenomic sequencing to improve the resolution of the findings obtained by 16S rRNA gene sequencing. A total of 189,036 metagenome-assembled genomes (MAGs) were recovered. The cladograms (Supplementary Fig. 2a) and histograms (Supplementary Fig. 2b) generated by LEfSe analysis indicated enrichment of Bacteroidaceae and Odoribacteraceae in the MM group, while Ruminococcaceae, Lachnospiraceae, Bifidobacteriaceae, Coriobacteriaceae, Eggerthellaceae, Erysipelotrichaceae, and Potyviridae were depleted at the genus level (Supplementary Data 2). A heatmap of the relative abundances of these genera is shown in Supplementary Fig. 3. These findings revealed that Ruminococcaceae and Lachnospiraceae were more prevalent in the HC group than in the MM group, which is consistent with the data from 16S rRNA gene sequencing. However, Bacteroidaceae were more prevalent in the MM group than in the HC group, contrary to the 16S rRNA gene sequencing data. These results indicate that Ruminococcaceae and Lachnospiraceae are key differentiators between the MM and HC groups.
Functional analysis of the MM gut microbiome
Metagenomic reads were annotated using predicted functions based on alignment against available databases (Uniprot, Kyoto Encyclopedia of Genes and Genomes [KEGG], eggNOG, and Carbohydrate-Active Enzyme [CAZy]). KEGG analysis revealed numerous distinct annotated functions between the two groups. The majority of the predicted functions encoded in the genomes of the gut microbiome in the MM group included phosphonate and phosphinate metabolism, ubiquinone and other terpenoid quinone biosynthesis, the PPAR signaling pathway, lipopolysaccharide biosynthesis, and linoleic acid metabolism, whereas those in the HC group were enriched in proteasome, selenoprotein metabolism; arginine biosynthesis; valine, leucine, and isoleucine biosynthesis; and pantothenate and CoA biosynthesis (Supplementary Fig. 4 and Supplementary Data 3).
Overall blood metabolome of the MM and HC groups
The gut microbiome substantially influences blood metabolites11. The overall blood metabolomes of the MM and HC groups were identified using ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (UPLC/Q-TOF MS). The screening of blood metabolites and gut microbes came from paired individuals. A total of 17180 and 8787 peaks were detected in the serum of the two groups in the positive ion mode (ESI+) and negative ion mode (ESI−), respectively. Orthogonal partial least squares-discriminant analysis (OPLS-DA) demonstrated that the samples from the MM and HC groups displayed notably explicit separation on opposite sides under the ESI+ and ESI− modes (Supplementary Fig. 5a), suggesting remarkable differences in metabolic profiles between the two groups. The interpretative ability of the OPLS-DA model was validated using a permutation test (n = 200), which demonstrated that all Q2 values were over 0.88 and all R2Y values were no <0.93 (Supplementary Fig. 5b). These results indicate that the OPLS-DA models under the ESI+ and ESI− modes are valid.
Functional analysis of differential blood metabolites
Under the ESI+ mode, 46 metabolites were higher, and 10 metabolites were lower in the MM group than in the HC group (Supplementary Data 4). Under the ESI− mode, 22 metabolites were higher, and 26 metabolites were lower in the MM group than in the HC group (Supplementary Data 4). The heatmaps represent all the differential and identified blood metabolites from the MM group compared with those from the HC group under the ESI+ (Supplementary Fig. 6) and ESI− (Supplementary Fig. 7) modes, respectively. The most abundant blood metabolites in the MM group included protoporphyrinogen IX, netilmici, flurandrenolide, cholylglutamic acid, tyrosylleucine, porphyrinogen, 17-hydroxyprogesterone, PC (O-14:0/16:0), alliospiroside C, ginsenoside Rb1, and dehydroepiandrosterone sulfate. In contrast, the most abundant blood metabolites in the HC group included galabiosylceramide (d18:1/26:0), lysoPC (20:4(5Z,8Z,11Z,14Z)/0:0; 16:0/0:0; 15:0), 1-palmitoylphosphatidylcholine, 1-heneicosanoyl-glycero-3-phosphoserine, 1-lyso-2-arachidonoyl-phosphatidate, lysoPA (18:2(9Z,12Z)/0:0; 16:0/0:0), and 2-deoxybrassinolide.
Additionally, the related pathways for differential blood metabolites were analyzed using MetaboAnalyst. These pathways included glycerophospholipid metabolism, linoleic acid metabolism, steroid hormone biosynthesis, alpha-linolenic acid metabolism, glycosylphosphatidylinositol-anchor biosynthesis, glycerolipid metabolism, sphingolipid metabolism, phosphatidylinositol signaling system, porphyrin and chlorophyll metabolism, arachidonic acid metabolism, and primary bile acid biosynthesis (Supplementary Fig. 8 and Supplementary Table 1).
Integrated correlation analysis of microbes and blood metabolites in MM
A joint analysis of differential gut microbiome taxonomic indicators and differential blood metabolites was performed. Shared pathways between differential gut microbiome taxonomic indicators identified by metagenomics and differential blood metabolites were linoleic acid and alpha-linolenic acid metabolism. PC was involved in both linoleic acid and alpha-linolenic acid metabolism, the levels of which were increased in the MM group.
The correlation between differential genera identified by metagenomics and differential blood metabolites was analyzed using corr.test in R v3.6.1 (Fig. 2a, b). PC (ID 714.5367096), identified by the ESI+ mode, was negatively associated with the abundances of Lachnospiraceae and Prevotellaceae (P < 0.05; Fig. 2a), both of which displayed higher levels in the HC group than in the MM group (Supplementary Data 2). Therefore, Lachnospiraceae and PC are critical differentiators between the MM and HC groups. Additionally, co-culture with Lachnospiraceae reduced the PC level derived from KAS-6/1 and U266 cells (Supplementary Fig. 9). As PC could originate from tumor cell membranes23, above results suggest that the decrease in Lachnospiraceae should enhance the PC production from MM cells to increase abundance of blood PC in MM patients.
a Heatmap of the correlation analysis of differential genera identified by metagenomics and differential blood metabolites identified under the ESI+ mode. b Heatmap of the correlation analysis of differential genera identified by metagenomics and differential blood metabolites identified under the ESI− mode. The enrichment is indicated by colored bars on the right and top of the plot. Red indicates a positive correlation, while blue demonstrates a negative correlation. Significant correlation regions are denoted by black stars (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). MM multiple myeloma, ESI+ positive ion mode, ESI− negative ion mode. Source data are provided as a Source Data file.
Clinical correlations of Lachnospiraceae and PC in MM patients
Then the correlation between clinicopathological features and Lachnospiraceae abundance or blood PC level in MM patients was analyzed using the multivariate logistic regression analysis. Results showed that Lachnospiraceae abundance was positively correlated with hemoglobin level, while blood PC level was negatively correlated with leukocyte and platelet levels in MM patients (Supplementary Fig. 10a and Supplementary Table 2). However, no association between Lachnospiraceae abundance or blood PC level and overall survival (OS) of MM patients was found (Supplementary Fig. 10b).
PC elevates Sb9 expression in MM cells to reduce the cytotoxic effect of CD8 T cells on MM cells by suppressing GZMB expression via LPA
Above results have revealed that PC is a critical differentiator between the MM and HC groups. PC-derived lipid mediators contribute to the suppressive TME by modulating immune cells13. Thus, we explored the mechanisms by which PC and PC-derived lipid mediators modulate immune cells to trigger a suppressive TME. LPA is derived from the turnover of PC14. Enzyme-linked immunosorbent assay (ELISA) revealed that PC treatment dose-dependently increased LPA levels in KAS-6/1 and U266 cells (Fig. 3a), indicating that LPA was derived from PC in MM cells. As MM cells produced most LPA (220 ng/mL) at 24 h post 200 μM PC treatment, MM cells would be treated with 200 μM PC or 220 ng/mL LPA in subsequent in vitro experiments. It was worth noting that the concentration of LPA in vitro is lower than that in the blood of MM patients. Moreover, the Cell Counting Kit-8 (CCK-8) assay indicated that PC exerted no toxic effects on KAS-6/1 and U266 cells (Supplementary Fig. 11).
a LPA level of KAS-6/1 and U266 cells treated with or without PC. Mean± SD, n = 3 biological replicates, data represent two independent experiments. b The correlation heatmap between Lachnospiraceae abundance or blood PC and PC levels and the rate of GZMB + CD8 T cells in MM patients. c Sb9 mRNA level in KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. d Sb9 protein level in KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. e Representative images of the flow cytometric analysis and quantification of the rate of GZMB + CD8 T cells in CD8 T cells co-cultured with WT or Sb9 KO KAS-6/1 and U266 cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. f Released GZMB level in CD8 T cells co-cultured with WT or Sb9 KO KAS-6/1 and U266 cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. g Representative images of the flow cytometric analysis and quantification of the rate of RFP-labeled WT or Sb9 KO KAS-6/1 and U266 cells co-cultured with CD8 T cells plus PC or LPA treatment. Mean± SD, n = 3 biological replicates, data represent two independent experiments. Sb9 SerpinB9, GZMB granzyme B, KO knockout, PC phosphatidylcholine, LPA lysophosphatidic acid, WT wild type. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post hoc Tukey’s test except for (b). Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01; ****P < 0.0001. Source data are provided as a Source Data file.
PC-derived lipid mediators contribute to the immunosuppressive TME by modulating immune cells13. In addition, the ratio of GZMB + CD8 T cells decreases in patients with MM17. Besides, the multivariate logistic regression analysis revealed that Lachnospiraceae abundance was positively correlated with the ratio of GZMB + CD8 T cells, whereas blood PC and LPA levels were negatively correlated with the ratio of GZMB + CD8 T cells in MM patients (Fig. 3b). Therefore, the effects of PC and LPA treatments on CD8 T cells derived from HC were determined. The CCK-8 assay showed that PC and LPA treatments did not affect CD8 T cell viability (Supplementary Fig. 12a). Moreover, flow cytometric analysis revealed that PC and LPA treatments exerted no significant effect on the ratio of GZMB + CD8 T cells (Supplementary Figs. 12b and 13a). ELISA results further revealed that PC and LPA treatments did not reduce the level of GZMB released from CD8 T cells (Supplementary Fig. 12c), suggesting that PC did not directly affect GZMB expression in CD8 T cells via LPA. Thus, PC and LPA may regulate GZMB expression in CD8 T cells by modulating MM cells.
Sb9 is an inhibitor of GZMB, and suppressing its expression could enhance the efficacy of GZMB-based tumor immunotherapy20,21. In the present study, we determined the regulatory effects of PC and LPA on Sb9 expression in MM cells. Quantitative reverse transcription-PCR (qRT-PCR) and western blot (WB) analyses demonstrated that PC and LPA treatments increased Sb9 mRNA and protein levels in KAS-6/1 and U266 cells (Fig. 3c, d), indicating that PC upregulated Sb9 expression in MM cells via LPA.
Sb9 knockout (Sb9 KO) was performed in KAS-6/1 and U266 cells using clustered regularly short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) technology (Supplementary Fig. 14a, b). Subsequently, CD8 T cells derived from the HC group were co-cultured with Sb9 KO KAS-6/1 and U266 cells. Flow cytometric analysis revealed that co-culture of KAS-6/1 and U266 cells reduced the number of GZMB+ CD8 T cells, whereas Sb9 KO in KAS-6/1 and U266 cells abolished these effects (Fig. 3e and Supplementary Fig. 13a). ELISA further confirmed that co-culture of KAS-6/1 and U266 cells decreased the level of GZMB released from CD8 T cells, whereas Sb9 KO in KAS-6/1 and U266 cells reversed these effects (Fig. 3f). These data suggest that PC and LPA reduce GZMB expression in CD8 T cells by elevating Sb9 expression in MM cells.
Additionally, the cytotoxic effects of CD8 T cells on MM cells were detected. First, red fluorescent protein (RFP) expression vectors were transfected into KAS-6/1 and U266 cells to label MM cells. Then, RFP-labeled KAS-6/1 and U266 cells were co-cultured with CD8 T cells. Flow cytometric analysis showed that co-culture of CD8 T cells reduced the number of RFP-labeled KAS-6/1 and U266 cells (Fig. 3g and Supplementary Fig. 13b), but both PC and LPA treatments increased these numbers. However, Sb9 KO in KAS-6/1 and U266 cells reversed the effects of PA and LPA treatments on the number of RFP-labeled KAS-6/1 and U266 cells co-cultured with CD8 T cells (Fig. 3g and Supplementary Fig. 13b). These data indicate that PC increases Sb9 expression in MM cells to attenuate the cytotoxic effects of CD8 T cells on MM cells by inhibiting GZMB expression via LPA.
PC facilitates Sb9 mRNA maturation through lin-28 homolog A (LIN28A) and LIN28B via LPA in MM cells
We investigated the mechanisms by which PC and LPA treatments modify Sb9 expression in MM cells. Both PC and LPA treatments increased Sb9 mRNA levels in KAS-6/1 and U266 cells (Fig. 3b). LPA has been demonstrated to augment microRNA-30c-2 maturation24. Further, qRT-PCR analysis demonstrated that PC and LPA treatments improved Sb9 mRNA levels but decreased Sb9 pre-mRNA levels in KAS-6/1 and U266 cells (Fig. 4a). Moreover, PC and LPA treatments enhanced splicing of introns between exons in Sb9 pre-mRNA of KAS-6/1 and U266 cells, especially intron 2 between exon 2 and 3 (E2/I2) (Fig. 4b). Touchdown PCR further identified that both PC and LPA treatments facilitated the splicing of intron 2 between exon 2 and 3 (E2/E3) in Sb9 pre-mRNA of KAS-6/1 and U266 cells (Fig. 4c), suggesting that PC promoted Sb9 mRNA maturation by enhancing Sb9 pre-mRNA splicing.
a Sb9 pre-mRNA level in KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. b Levels of junctions of exon-intron in Sb9 pre-mRNA detected by qRT-PCR in KAS-6/1 and U266 cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments c Fragments of exon 2-intron 2-exon 3 or exon 2-exon 3 detected by touchdown PCR in KAS-6/1 and U266 cells. d The GGAGA motif was found in exon 2 and intron 2 of Sb9 pre-mRNA. e Quantification of Sb9 pre-mRNA by qRT-PCR in KAS-6/1 and U266 cells treated with or without PC and LPA following RIP using LIN28A and LIN28B antibodies. Mean± SD, n = 3 biological replicates, data represent two independent experiments. f Sb9 mRNA level in KAS-6/1 and U266 cells transfected with or without LIN28A and LIN28B shRNAs. Mean± SD, n = 3 biological replicates, data represent two independent experiments. g Splicing analysis of Sb9 minigene and indicated deletion mutation of the GGAGA motif in KAS-6/1 and U266 cells. h Splicing of Sb9 pre-mRNA was analyzed by a relative luciferase reporter activity assay in scrambled and LIN28A/B shRNAs-transfected KAS-6/1 and U266 cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. Sb9 SerpinB9, PC phosphatidylcholine, LPA lysophosphatidic acid, E exon, I intron, NC negative control, RIP RNA immunoprecipitation. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post-hoc Tukey’s test for a, b, e, and f, which was performed utilizing unpaired Student’s t test for h. Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Source data are provided as a Source Data file.
Bioinformatics analysis of CLIP-seq data in the ENCORI database demonstrated that LIN28A/B can bind to Sb9 mRNA in cancer cells (Supplementary Fig. 15). LIN28A/B is an important splicing factor that regulates mRNA maturation25. LIN28B is also associated with adverse outcomes in MM26. It has been identified a GGAGA motif enriched within LIN28 binding sites in mRNAs27, which was found in exon 2 and intron 2 of Sb9 pre-mRNA (Fig. 4d). RNA immunoprecipitation (RIP) performed using LIN28A and LIN28B antibodies, followed by qRT-PCR analysis, showed that LIN28A/B binds to Sb9 pre-mRNA, whereas PC and LPA treatments enhanced the association between LIN28A/B and Sb9 pre-mRNA in KAS-6/1 and U266 cells (Fig. 4e). Moreover, LIN28A or LIN28B silencing using two short hairpin RNAs (shRNAs) (Supplementary Fig. 16a, b) decreased Sb9 mRNA levels in KAS-6/1 and U266 cells, and LIN28A/B double silencing further reduced SERPINB9 mRNA levels in KAS-6/1 and U266 cells (Fig. 4f), indicating that LIN28A/B might facilitate Sb9 pre-mRNA splicing in MM cells.
To further determine whether Lin28A/B regulates Sb9 pre-mRNA splicing in MM cells, a minigene reporter plasmid (Sb9-FL) composed of a genomic DNA fragment of Sb9 exons 2–3 and that (Sb9-Mut) containing deletion mutation of GGAGA motif based on Sb9-FL were constructed. Then minigene reporter plasmids were transfected into KAS-6/1 and U266 cells while the splicing pattern was analyzed using RT-PCR. The minigene could mimic the endogenous splicing pattern of Sb9, whereas deletion mutation of GGAGA motif suppressed the exclusion of intron 2 (Fig. 4g). Additionally, intron 2 of Sb9 was inserted into the firefly luciferase gene in the pGL3 vector (pGL3-Sb9 intron 2) to further identify the effects of LIN28A/B on Sb9 pre-mRNA splicing. Subsequently, intron 2 of Sb9-containing pGL3 vectors was co-transfected with pRL-TK Renilla luciferase vectors into KAS-6/1 and U266 cells, and luciferase expression was determined using a dual luciferase reporter assay. The luciferase activity could be restored when intron 2 of Sb9 was spliced out (Fig. 4h). However, LIN28A/B double silencing decreased the luciferase activity in KAS-6/1 and U266 cells transfected with intron 2 of Sb9 gene-containing pGL3 vectors (Fig. 4h). Moreover, deletion mutation of GGAGA motif the splicing of intron 2 of Sb9, and LIN28A/B double silencing had no effect on the luciferase activity in KAS-6/1 and U266 cells transfected with pGL3-Sb9 intron 2 containing deletion mutation of GGAGA motif (Fig. 4h). These data suggest that PC promotes Sb9 mRNA maturation through LIN28A/B via LPA in MM cells.
PC hinders the cytotoxic effect of CD8 T cells on MM cells by decreasing GZMB expression through exosomal Sb9 derived from MM cells via LPA
The mechanisms by which PC and LPA treatments reduce GZMB expression in CD8 T cells were investigated. Exosomes are small vesicles that deliver their cargo after internalization28,29. Importantly, exosomes play a critical role in the immunosuppressive TME of MM28,30. Results of the present study showed that PC suppressed GZMB expression in CD8 T cells by increasing Sb9 expression in MM cells via LPA. Therefore, the levels of exosomal Sb9 derived from MM cells were measured after PC and LPA treatments. Exosomes derived from KAS-6/1 and U266 cells treated with or without PC and LPA were isolated. The MM cell-derived exosomes were 30–150 nm in size (Supplementary Fig. 17a). WB analysis of the exosomal markers CD9 and CD63 confirmed that the isolated small particles were exosomes (Supplementary Fig. 17b). Subsequently, WB results revealed that PC and LPA treatments elevated the levels of exosomal Sb9 derived from KAS-6/1 and U266 cells (Fig. 5a).
a Level of exosomal Sb9 derived from KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. b Representative images of the flow cytometric analysis for CD8 T cells cultured with PKH67-labeled exosomes (green) derived from WT or Sb9 KO KAS-6/1 and U266 cells treated with or without PC and LPA. c Sb9 protein levels in CD8 T cells cultured with PKH67-labeled exosomes (green) derived from WT or Sb9 KO KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. d GZMB mRNA levels in CD8 T cells cultured with exosomes derived from WT or Sb9 KO KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. e Released GZMB level in CD8 T cells cultured with exosomes derived from WT or Sb9 KO KAS-6/1 and U266 cells treated with or without PC and LPA. Mean± SD, n = 3 biological replicates, data represent two independent experiments. f Representative images of the flow cytometric analysis and quantification of the rate of RFP-labeled KAS-6/1 and U266 cells co-cultured with WT or Sb9-overexpressed CD8 T cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. Sb9 SerpinB9, PC phosphatidylcholine, LPA lysophosphatidic acid, GZMB granzyme B, KO knockout, OE overexpression, WT wild type. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post-hoc Tukey’s test except for (b). Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Source data are provided as a Source Data file.
CD8 T cells were cultured with exosomes derived from KAS-6/1 and U266 cells treated with or without PC and LPA. Exosomes from KAS-6/1 and U266 cells were labeled with PKH67 to verify their uptake by CD8 T cells, and fluorescence microscopy was used to determine the uptake of PKH67-labeled exosomes by CD8 T cells (Fig. 5b and Supplementary Fig. 13c). The expression levels of Sb9 and GZMB in CD8 T cells were then detected. WB blot results revealed that exosome co-culture increased Sb9 levels in CD8 T cells, and co-culture of exosomes derived from MM cells treated with PC and LPA further elevated Sb9 levels in CD8 T cells (Fig. 5c). However, Sb9 KO in MM cells abolished the effect of co-culture of exosomes derived from MM cells treated with PC and LPA on Sb9 levels in CD8 T cells (Fig. 5c). In contrast, qRT-PCR and ELISA analyses indicated that exosome co-culture decreased GZMB levels in CD8 T cells, while co-culture of exosomes derived from MM cells treated with PC and LPA further reduced GZMB levels in CD8 T cells (Fig. 5d, e). Moreover, Sb9 KO in MM cells abolished this effect of co-culture of exosomes derived from MM cells treated with PC and LPA on GZMB levels in CD8 T cells (Fig. 5d, e). These results suggest that PC decreases GZMB expression in CD8 T cells by increasing the levels of exosomal Sb9 derived from MM cells via LPA.
Additionally, the cytotoxic effects of CD8 T cells on MM cells were detected after overexpressing Sb9 in CD8 T cells (Supplementary Fig. 18a, b). RFP expression vectors were transfected into KAS-6/1 and U266 cells to label the MM cells. RFP-labeled KAS-6/1 and U266 cells were co-cultured with Sb9-overexpressed CD8 T cells. Subsequent flow cytometric analysis demonstrated that co-culture with CD8 T cells decreased the number of RFP-labeled KAS-6/1 and U266 cells, whereas Sb9 overexpression (OE) in CD8 T cells neutralized the cytotoxic effects of CD8 T cells on MM cells (Fig. 5f and Supplementary Fig. 13b). Together, these data reveal that PC inhibits the cytotoxic effects of CD8 T cells on MM cells by reducing GZMB expression in CD8 T cells through exosomal Sb9 derived from MM cells via LPA.
Sb9 suppresses GZMB expression by reducing TP53 expression in CD8 T cells
The mechanisms by which Sb9 suppresses GZMB expression were also investigated. Bioinformatics analysis using the JASPAR website (https://jaspar.genereg.net/) revealed potential binding sites for TP53 at the GZMB promoter (Supplementary Fig. 19). TP53 is downregulated and plays a suppressive role in MM31. However, the effect of TP53 on CD8 T cells in MM remains unclear. Chromatin immunoprecipitation (ChIP) performed using the TP53 antibody revealed that TP53 is bound to the GZMB promoter in CD8 T cells (Fig. 6a). Additionally, an electrophoretic mobility shift assay (EMSA) utilizing a probe corresponding to the wild-type (WT) GZMB promoter together with the TP53 protein detected a DNA–protein complex band, whereas adding the TP53 antibody resulted in a super-shift of this band (Fig. 6b). However, mutation of the TP53-binding site (MUT) on the probe corresponding to the GZMB promoter blocked the association between the GZMB promoter and TP53 and the super-shift of the DNA–protein complex band (Fig. 6b). These data indicate the DNA-binding ability of TP53 to the GZMB promoter in CD8 T cells.
a PCR and qRT-PCR results for the GZMB promoter in CD8 T cells following ChIP using TP53 antibody. Mean± SD, n = 3 biological replicates, data represent two independent experiments. b Representative image of EMSA using probes corresponding to the WT or MUT GZMB promoter together with TP53 protein and TP53 antibody. EMSA was repeated independently three times with similar results. c GZMB mRNA level in CD8 T cells transfected with or without TP53 shRNAs. Mean± SD, n = 3 biological replicates, data represent two independent experiments. d Transcription activity of GZMB analyzed by relative luciferase reporter activity assay in scrambled and TP53 shRNAs-transfected CD8 T cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. e TP53 mRNA level in CD8 T cells transfected with or without the Sb9 expression vector. f TP53 protein level in CD8 T cells transfected with or without Sb9 expression vector. Mean± SD, n = 3 biological replicates, data represent two independent experiments. Sb9 SerpinB9, GZMB granzyme B, WT wild type, MUT mutation, OE overexpression, EMSA, electrophoretic mobility shift assay, PCR polymerase chain reaction, qRT-PCR quantitative reverse transcription-PCR. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post-hoc Tukey’s test except for (b). Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01; ****P < 0.0001. Source data are provided as a Source Data file.
Therefore, we investigated the effect of TP53 on GZMB expression in CD8 T cells. TP53 silencing using two shRNAs (Supplementary Fig. 20a, b) decreased GZMB mRNA levels in KAS-6/1 and U266 cells (Fig. 6c). A dual luciferase reporter assay showed that TP53 silencing suppressed GZMB transcription in CD8 T cells, and mutation of TP53 binding site in GZMB promoter decreased GZMB transcription and TP53 silencing had no effect on the transcription of GZMB containing the mutation of TP53 binding site in CD8 T cells (Fig. 6d). These results suggest that TP53 enhances GZMB transcription in CD8 T cells. However, Sb9 OE decreased TP53 protein levels but not mRNA levels in CD8 T cells (Fig. 6e, f). Thus, these results reveal that Sb9 reduces GZMB expression by inhibiting TP53 expression in CD8 T cells.
Sb9 decreases TP53 expression by suppressing TP53 UFMylation via the competitive binding of TP53 with ubiquitin-fold modifier conjugating enzyme 1 (UFC1) in CD8 T cells
The above data showed that Sb9 OE decreased TP53 protein levels but not mRNA levels in CD8 T cells (Fig. 6e, f), suggesting that Sb9 regulated TP53 at the post-translational level. UFMylation is known to improve TP53 stability by antagonizing ubiquitination32. Coimmunoprecipitation (Co-IP) performed using the ubiquitin-fold modifier 1 (UFM1) antibody, which modifies the UFMylation of target proteins32, revealed the interaction of TP53 with UFM1, but Sb9 OE reduced their interaction in CD8 T cells (Fig. 7a, Supplementary Fig. 21), suggesting that Sb9 suppressed the UFMylation of TP53 in CD8 T cells. In addition, UFM1 silencing using two shRNAs (Supplementary Fig. 22a, b) decreased TP53 levels in CD8 T cells (Fig. 7b). Then, the effect of UFMylation on TP53 stability in CD8 T cells was determined after exposure to 50 μg/mL cycloheximide for increasing times. TP53 began to decrease at 3 h after the exposure of cycloheximide, and there was 40% TP53 remained in control CD8 T cells, but only 20% TP53 remained in UFM1-silenced CD8 T cells at 12 h after cycloheximide treatment (Fig. 7c). Thus, Sb9 diminished TP53 expression by reducing stability via suppressing UFMylation in CD8 T cells.
a Representative image of Co-IP using a UFM1 antibody in CD8 T cells. Rabbit IgG was used as a negative control. Mean± SD, n = 3 biological replicates, data represent two independent experiments. b TP53 protein level in CD8 T cells transfected with or without UFM1 shRNAs. Mean± SD, n = 3 biological replicates, data represent two independent experiments. c TP53 protein level detected in CD8 T cells transfected with or without UFM1 shRNAs after cycloheximide treatment. Mean± SD, n = 3 biological replicates, data represent two independent experiments. d UFM1 protein level in CD8 T cells transfected with or without the Sb9 expression vector. e Representative images of Co-IP using TP53 antibody in scrambled and Sb9 expression vector (0.5, 1, and 2 μg)-transfected CD8 T cells. Rabbit IgG was used as a negative control. Mean± SD, n = 3 biological replicates, data represent two independent experiments. f TP53 protein level in CD8 T cells transfected with or without UFC1 shRNAs. Mean± SD, n = 3 biological replicates, data represent two independent experiments. g Representative image of Co-IP using a UFM1 antibody in CD8 T cells. Mean± SD, n = 3 biological replicates, data represent two independent experiments. Rabbit IgG was used as a negative control. Sb9 SerpinB9, NC negative control, OE overexpression, UFC1, ubiquitin-fold modifier conjugating enzyme 1. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post-hoc Tukey’s test for (b, c, e, and f), which was performed utilizing unpaired Student’s t test for a, d, and g. Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001. Source data are provided as a Source Data file.
The mechanisms by which Sb9 inhibits TP53 UFMylation in CD8 T cells were determined. WB analysis showed that Sb9 OE had no effect on UFM1 levels in CD8 T cells (Fig. 7d). Further competitive Co-IP using TP53 antibody after transfection of different dosages of Sb9 expression vectors (0.5, 1, and 2 μg) indicated that Sb9 blocked the interaction between TP53 and UFC1, the UFM1 ligase32, via the competitive binding of TP53 with UFC1 in CD8 T cells (Fig. 7e, Supplementary Fig. 23). Moreover, UFM1 silencing using two shRNAs (Supplementary Fig. 24) declined TP53 level in CD8 T cells (Fig. 7f). Additionally, UFM1 silencing decreased the interaction of UFM1 and TP53 in CD8 T cells, suggesting that Sb9 facilitated the UFMylation of TP53 in CD8 T cells (Fig. 7g, Supplementary Fig. 25). These data indicate that Sb9 reduces TP53 expression by inhibiting TP53 UFMylation via the competitive binding of UFC1 to Sb9 in CD8 T cells.
Lachnospiraceae enhances the in vivo cytotoxic effect of CD8 T cells on MM cells
Then, the in vivo role of Lachnospiraceae in the cytotoxic effect of CD8 T cells on MM cells was explored. First, Lachnospiraceae was transplanted into NOD/SCID mice by gavage for 14 days. Results of qRT-PCR showed that the amount of 16 s rRNA gene of Lachnospiraceae was significantly elevated in stools from NOD/SCID mice transplanted with Lachnospiraceae (Supplementary Fig. 26), indicating the abundance of Lachnospiraceae was increased in guts of NOD/SCID mice transplanted with of Lachnospiraceae. The above data suggest that Lachnospiraceae is successfully colocalized after transplantation.
Next, 2 × 106 RFP-labeled WT KAS-6/1 and U266 cells were given to NOD/SCID mice using tail vein infusion. Mice died around day 24 after injection of WT KAS-6/1 and U266 cells, yet infusion of human CD8 T cells expanded the life span of mice to an average of 33 days (Supplementary Fig. 27a). Moreover, transplantation of Lachnospiraceae promoted the effect of injection of human CD8 T cells (Supplementary Fig. 27a).
Subsequently, the number of RFP-labeled WT KAS-6/1 and U266 cells in mouse blood on day 14 after injection was verified by flow cytometric analysis to further determine the role of Lachnospiraceae in the cytotoxic effect of CD8 T cells on MM cells in vivo. Results indicated that infusion of human CD8 T cells reduced the number of WT KAS-6/1 and U266 cells in mouse blood (Supplementary Fig. 27b and Supplementary Fig. 13c), and transplantation of Lachnospiraceae enhanced the effect of injection of human CD8 T cells (Supplementary Fig. 27b).
Furthermore, the effects of Lachnospiraceae and CD8 T cells on tumor growth were identified. 2 × 106 bioluminescent WT KAS-6/1 and U266 cells were injected into the right dorsal flank of NOD-SCID mice and monitored via an in vivo imaging system. After initial tumor engraftment on day 9 post-subcutaneous injection, 1 × 107 human CD8 T cells from HC were intraperitoneally injected into NOD-SCID mice. At day 25 post the subcutaneous injection of MM cells, the size and weight of tumors developed from WT KAS-6/1 and U266 cells in the mice injected with human CD8 T cells showed a reduced size and weight than those originating from WT KAS-6/1 and U266 cells in the control mice. Transplantation of Lachnospiraceae strengthened the effect of injection of human CD8 T cells on the size and weight of tumors developed from WT KAS-6/1 and U266 cells (Supplementary Figs. 27c–e and 28). Therefore, these data demonstrate that Lachnospiraceae enhances the cytotoxic effect of CD8 T cells on MM cells in vivo.
PC reduces the in vivo cytotoxic effect of CD8 T cells on MM cells by Sb9
To determine the role of PC in the cytotoxic effect of CD8 T cells on MM cells in vivo, 2 × 106 RFP-labeled WT or Sb9 KO KAS-6/1 and U266 cells were given into NOD/SCID mice by tail vein infusion combined with 200 μM/kg PC and 1 × 107 human CD8 T cells from HC. Mice died around day 25 after WT KAS-6/1 and U266 cells injection, whereas infusion of human CD8 T cells expanded the life span of mice to an average of 35 days (Supplementary Fig. 30a). However, administration of PC reversed the effect of injection of human CD8 T cells (Supplementary Fig. 29a). Moreover, Sb9 KO in MM cells had no significant effect on the life span of mice, but PC did not abolish the expanded effect of human CD8 T cell infusion on the life span of mice injected with Sb9 KO KAS-6/1 and U266 cells (Supplementary Fig. 29a). Furthermore, the injection of Sb9-overexpressed human CD8 T cells could not expand the life span of mice injected with WT KAS-6/1 and U266 cells (Supplementary Fig. 29a).
To further identify the role of PC in the cytotoxic effect of CD8 T cells on MM cells in vivo, flow cytometric analysis was utilized to verify the number of RFP-labeled KAS-6/1 and U266 cells in mouse blood at day 14 after injection. It was found that infusion of human CD8 T cells reduced the number of KAS-6/1 and U266 cells in mouse blood (Supplementary Figs. 29b and 13 d), whereas administration of PC neutralized the effect of injection of human CD8 T cells (Supplementary Figs. 29b and 13d). Sb9 KO in MM cells had no significant effect on the number of KAS-6/1 and U266 cells in mouse blood, but PC did not abolish the reduced effect of human CD8 T cell infusion on the number of KAS-6/1 and U266 cells in mice injected with Sb9 KO KAS-6/1 and U266 cells (Supplementary Fig. 29b). Furthermore, the injection of Sb9-overexpressed human CD8 T cells could not decline the number of KAS-6/1 and U266 cells in mice injected with WT KAS-6/1 and U266 cells (Supplementary Fig. 29b).
Next, the effects of PC and CD8 T cells on tumor growth were investigated. 2 × 106 bioluminescent WT or Sb9 KO KAS-6/1 and U266 cells were injected into the right dorsal flank of NOD-SCID mice and monitored using an in vivo imaging system to determine whether PC reduces the cytotoxic effects of CD8 T cells on MM cells by Sb9 in vivo (Fig. 8a). After initial tumor engraftment on day 9 post-subcutaneous injection, 200 μM/kg PC and 1 × 107 human CD8 T cells from HC were intraperitoneally injected into NOD-SCID mice. At day 25 after the subcutaneous injection of MM cells, the size and weight of tumors developed from WT KAS-6/1 and U266 cells in the mice injected with human CD8 T cells exhibited a reduced size and weight than those originating from WT KAS-6/1 and U266 cells in the control mice. However, administration of PC reversed the effect of injection of human CD8 T cells on the size and weight of tumors developed from WT KAS-6/1 and U266 cells (Fig. 8a–c, Supplementary Fig. 30). Moreover, Sb9 KO in MM cells had no significant effect on the size and weight of tumors originating from KAS-6/1 and U266 cells; however, PC did not abolish the effect of human CD8 T cell injection on the size and weight of tumors developed from Sb9 KO KAS-6/1 and U266 cells (Fig. 8a–c, Supplementary Fig. 30). Furthermore, the injection of Sb9-overexpressed human CD8 T cells did not reduce the size and weight of tumors developed from WT KAS-6/1 and U266 cells (Fig. 8a–c, Supplementary Fig. 30). In sum, the above data reveal that PC attenuates the cytotoxic effect of CD8 T cells on MM cells induced by Sb9 in vivo.
a Representative image of tumors developing from bioluminescent WT or Sb9 KO KAS-6/1 and U266 cells in NOD-SCID mice co-injected with PC and WT or Sb9-overexpressed human CD8 T cells at days 9 and 25 postinjection of MM cells. b Quantification of the volume of xenograft tumors derived from WT or Sb9 KO KAS-6/1 and U266 cells in the presence and absence of PC and WT or Sb9-overexpressed human CD8 T cells in NOD-SCID mice. Mean± SD, n = 5 biologically independent mice. c Images of xenograft tumors derived from WT or Sb9 KO KAS-6/1 and U266 cells in the presence and absence of PC and WT or Sb9-overexpressed human CD8 T cells in NOD-SCID mice. Histograms indicate the quantification of the weight of tumors. Mean± SD, n = 5 biologically independent mice. Sb9 SerpinB9, MM multiple myeloma, WT wild type, PC phosphatidylcholine, KO knockout, OE overexpression. All statistical tests were two-sided. Statistical significance was assessed using one-way ANOVA followed by post-hoc Tukey’s test for (b) and (c). Adjustments were made for multiple comparisons. *P < 0.05; **P < 0.01. Source data are provided as a Source Data file.
Lachnospiraceae strengthens the in vivo cytotoxic effect of CD8 T cells on MM cells by reducing PC production to activate Sb9-inhibited GZMB via modulating TP53 UFMylation
Next, the effects of Lachnospiraceae and PC on blood markers in NOD/SCID mice were also determined. Lachnospiraceae was transplanted into NOD/SCID mice by gavage for 14 days, and then 2 × 106 WT or Sb9 KO KAS-6/1 and U266 cells were given into NOD/SCID mice by tail vein infusion combined with 200 μM/kg PC and 1 × 107 human CD8 T cells from HC. It was found that infusion of WT KAS-6/1 and U266 cells increased blood PC level in NOD/SCID mice (Control) compared to that in NOD/SCID mice without infusion of WT KAS-6/1 and U266 cells (NC) (Supplementary Fig. 30a), suggesting that PC could origin from MM cells. Besides, infusion of CD8 T cells had no effect on blood PC and Sb9 levels (Supplementary Fig. 31a, b) whereas elevated blood GZMB (Supplementary Fig. 31c) and TP53 UFMylation levels combined with infusion of CD8 T cells (Supplementary Fig. 31d). However, Sb9 overexpression in CD8 T cells induced blood Sb9 level (Supplementary Fig. 31b) and reversed the effects of infusion of WT CD8 T cells on blood GZMB (Supplementary Fig. 31c) and TP53 UFMylation levels (Supplementary Fig. 31d). Moreover, transplantation of Lachnospiraceae declined blood PC and Sb9 levels (Supplementary Fig. 31a, b) but elevated blood GZMB (Supplementary Fig. 31c) and TP53 UFMylation levels combined with infusion of CD8 T cells (Supplementary Fig. 31d). Furthermore, infusion of PC improved blood PC and Sb9 levels (Supplementary Fig. 31a, b) yet decreased blood GZMB (Supplementary Fig. 31c) and TP53 UFMylation levels (Supplementary Fig. 31d) combined with infusion of CD8 T cells. Additionally, Sb9 KO in MM cells had no effect on blood PC level but reduced blood Sb9 level (Supplementary Fig. 31a, b) and increased blood GZMB (Supplementary Fig. 31c) and TP53 UFMylation levels combined with infusion of CD8 T cells (Supplementary Fig. 31d). Above data indicate that Lachnospiraceae should strengthen the in vivo cytotoxic effect of CD8 T cells on MM cells by declining PC production to activate Sb9-inhibited GZMB via modifying TP53 UFMylation.
Discussion
The current study reveals that Lachnospiraceae and PC are critical differentiators between patients with MM and healthy individuals. Besides, Lachnospiraceae inhibits PC production from MM cells. In addition, Lachnospiraceae enhances whereas PC hinders the cytotoxic effect of CD8 T cells on MM cells by Sb9. Mechanistically, PC promotes Sb9 mRNA maturation and expression in MM cells through LIN28A/B and LPA. Moreover, PC inhibits the cytotoxic effects of CD8 T cells on MM cells by reducing GZMB expression in CD8 T cells via exosomal Sb9 derived from MM cells via LPA. Furthermore, Sb9 reduces GZMB expression by inhibiting TP53 expression and suppressing TP53 UFMylation via the competitive binding of TP53 to UFC1 in CD8 T cells (Fig. 9).
Lachnospiraceae and PC are critical differentiators between MM patients and healthy individuals. Besides, Lachnospiraceae inhibits PC production from MM cells. In addition, Lachnospiraceae enhances whereas PC hinders the cytotoxic effect of CD8 T cells on MM cells via Sb9. Mechanistically, PC promotes Sb9 mRNA maturation and expression in MM cells through LIN28A/B via LPA. Moreover, PC inhibits the cytotoxic effect of CD8 T cells on MM cells by reducing GZMB expression in CD8 T cells by exosomal Sb9 derived from MM cells via LPA. Furthermore, Sb9 reduces GZMB expression by inhibiting TP53 expression by suppressing TP53 UFMylation via competitive binding of TP53 with UFC1 in CD8 T cells. Sb9 SerpinB9, PC phosphatidylcholine, LPA lysophosphatidic acid, MMmultiple myeloma, GZMB granzyme B, UFC1 ubiquitin-fold modifier conjugating enzyme 1.
Lachnospiraceae plays a protective role against carcinogenesis. For instance, Lachnospiraceae protects against colorectal cancer (CRC) progression by enhancing the activation of CD8 T cells to promote tumor immune surveillance33. Moreover, the depletion of Lachnospiraceae contributes to CRC development34, whereas a high abundance of oral Lachnospiraceae protects against CRC35. Moreover, patients with gastrointestinal cancer and a high abundance of Lachnospiraceae show a preferred response to anti-PD-1/PD-L1 immunotherapy36. Furthermore, Lachnospiraceae is less prevalent in patients with leukemia37. However, the role of Lachnospiraceae in MM has not been reported. The present study is the first to uncover the protective role of Lachnospiraceae against MM.
Our findings indicate that PC levels are negatively associated with Lachnospiraceae abundance in MM, and Lachnospiraceae inhibits PC production from MM cells. However, the mechanism by which Lachnospiraceae regulates PC levels in MM has not been investigated. Lachnospiraceae activates the adenosine monophosphate-activated protein kinase (AMPK) pathway by producing butyrate38. Activation of the AMPK pathway can reduce PC synthesis39, and this pathway exerts an anti-cancer effect on MM40,41. Thus, Lachnospiraceae may reduce PC production by activating the AMPK pathway in MM.
LPA has been demonstrated to augment microRNA-30c-2 maturation24. However, the role of PC in mRNA maturation has not been reported. The present study showed that PC promoted Sb9 mRNA maturation in MM cells through LIN28A/B via LPA. Thus, this study is the first to reveal the effects of PC and PC-derived lipid mediators on mRNA maturation.
Additionally, our data show that Sb9 expression is detected in CD8 T cells, whereas PC and LPA treatments do not directly regulate GZMB expression in CD8 T cells, suggesting that PC may not promote Sb9 mRNA maturation through LIN28A/B by LPA in CD8 T cells. LIN28 phosphorylation facilitates the association between LIN28 and direct mRNA targets42. Thus, phosphorylation of LIN28A/B may be suppressed, resulting in a decrease in the interaction between LIN28A/B and Sb9 pre-mRNA in CD8 T cells. In addition, PC and LPA may have no effect on LIN28A/B phosphorylation; therefore, PC and LPA treatments do not directly reduce GZMB expression by enhancing Sb9 mRNA maturation in CD8 T cells.
Only one study has demonstrated that primary cytomegalovirus infection could increase circulating Sb9 levels in the blood after renal transplantation43. However, no study has indicated the role of exosomal Sb9. The current study reveals that PC inhibits the cytotoxic effect of CD8 T cells on MM cells by reducing GZMB expression in CD8 T cells through exosomal Sb9 derived from MM cells via LPA. This study is the first to demonstrate the role of exosomal Sb9.
As a novel post-translational modification, UFMylation is involved in cancer development and progression. For example, UFMylation of solute carrier family 7 member 11 inhibits ferroptosis in breast cancer44. Additionally, UFMylation of ribosomal protein L10 facilitates pancreatic adenocarcinoma development45. Moreover, UFMylation of placenta-associated 8 promotes breast cancer cell proliferation and hinders antitumor immune response46. In contrast, UFMylation improves TP53 stability by antagonizing its ubiquitination and suppressing renal cell carcinoma progression32. However, the role of UFMylation in MM and immune cells remains unclear. The findings of the present study reveal that Sb9 reduces GZMB expression by inhibiting TP53 expression and suppressing TP53 UFMylation in CD8 T cells. This finding expands our knowledge of the role of UFMylation in regulating MM and immune cells.
This study has several limitations. First, the factors underlying the differences in Lachnospiraceae between patients with MM and HC have not been explored. Second, the mechanisms by which Lachnospiraceae regulate PC levels in MM have not been investigated. Third, the effect of exosomal Sb9 from MM cells on the cytotoxicity of CD8 T cells in vivo has not been identified. These issues require further investigation. Additionally, whether Lachnospiraceae abundance and PC level fluctuate following antibiotic treatment in MM patients and the effects of Lachnospiraceae and PC on other cancer cell types beyond MM cells also are future research directions.
In summary, the current study demonstrates that Lachnospiraceae and PC are critical differentiators between patients with MM and healthy individuals. Besides, Lachnospiraceae inhibits PC production from MM cells. In addition, Lachnospiraceae promotes whereas PC hinders the cytotoxic effects of CD8 T cells on MM cells, and PC reduces GZMB expression in CD8 T cells through exosomal Sb9 derived from MM cells via LPA. Moreover, Sb9 reduces GZMB expression by inhibiting TP53 expression via suppression of TP53 UFMylation in CD8 T cells. This study offers novel targets and strategies for MM treatment.
Methods
Sample collection
The diagnosis of MM patients was made according to the 2014 International MM Working Group criteria. Then stool samples from patients with MM (n = 20) and healthy controls (n = 38) were collected at the Hematology Department of Shengjing Hospital. Serum samples aspirated from 65 patients with MM and 60 healthy controls were collected at the Hematology Department of Shengjing Hospital. The stool samples were placed in sterile frozen pipes and stored at −80 °C. Blood samples were collected in coagulant tubes and then centrifuged at 3000 rpm for 10 min at 25 °C. Then serum (supernatant) was collected into 1.5 mL frozen tubes and stored at −80 °C for further analysis. All human experimental procedures were approved by the Ethics Committee of Shengjing Hospital, China Medical University (#2023PS941K). Written informed consent was obtained from all participants enrolled in the study.
Microbial DNA extraction in fecal matter
The genomic bacterial DNA was extracted from stools by the E.Z.N.A. Stool DNA Kit (M4015, Omega, Norcross, GA, USA) in accordance with the manufacturer’s guidelines. Then 1% agarose gel electrophoresis was performed to analyze the integrity and fragment size of the extracted DNA, while NanoDrop 2000 was utilized to quantify DNA (Boston, MA, USA).
High-throughput 16 s ribosomal RNA gene sequencing
High-throughput 16 s ribosomal RNA gene sequencing was carried out as previously described11. Sequencing was performed by a NovaSeq PE250 platform (Illumina, San Diego, CA, USA).
Analysis of 16 s ribosomal RNA gene sequencing data
Paired-end reads were assigned to samples based on unique barcodes and merged using FLASH. Then FqTrim (v. 0.94) was used to clean raw reads while Vsearch (v2.3.4) was utilized to filter chimeric sequences. Next, DADA2 was used for dereplication to obtain a feature table and a sequence, followed by the calculation of alpha and beta diversities utilizing QIIME2. Subsequently, R version 3.5.2 was used to construct the corresponding phylogenetic tree, and the Basic Local Alignment Search Tool (BLAST) was utilized for sequence alignment and species annotation, whereas sequence alignment was performed using SILVA and NT-16S.
Metagenomic sequence processing and recovery of MAGs
BWA v0.7.1297 was used to identify contaminated human reads by mapping against the human (Homo sapiens) genome (GRCh38) followed by adaptor removal and read trimming by Trimmomatic v0.3689. Then Spades v3.12.098 was utilized to assemble each sample with a meta-flag independently. Next, BamM v1.7.3 was used for mapping reads to each resulting assembly, while bins with a minimum contig length of 1500 bases were produced by Metabat v2.12.199. CheckM v1.0.11100 was used to assess the contamination and completeness of the bins, and dRep v2.05101 was utilized to retain and de-replicate bins with completeness >80% and contamination <7% with default settings (99% identity). The taxonomic affiliations of the recovered MAGs were identified by the Genome Taxonomy Database (GTDB) Releases 03-RS86 and 04-R89102 utilizing GTDB-Tk v0.3.0103.
Metagenomic functional profiling
For read-based analysis, Prodigal v2.6.3107 was used to predict protein fragments in raw reads, followed by the alignment to the hidden Markov model databases dbCAN CAZy v6109, Pfam r31110, and TIGRFAM v15111 utilizing HMMER v3.1b2108. Then BLAST v2.8.192 alignment to the UniProt UniRef100 database was performed to determine KEGG orthology and extract subsequent associated KO terms. DESeq2 v1.20.030 was used to compare group functional profiles based on counts per sample. Moreover, EnrichM v0.5.0 was utilized for genome-level analysis of KEGG orthology terms and module completeness, while Fisher’s exact test was used to assess significance with a Benjamini–Hochberg adjustment. Furthermore, module completeness was compared in R by the Wilcoxon rank-sum test with Benjamini–Hochberg adjustment. Additionally, BLAST was used to determine the presence of genes of interest in enriched genomes.
Processing and analysis of serum samples
To extract metabolites from the serum samples, 100 μL of sample was incubated with 300 μL of precooled 50% methanol at 25 °C for 10 min followed by thereafter at −20 °C overnight. The next day samples were centrifugated at 4000 × g for 20 min, and 200 μL supernatants were transferred to new vitals. Then the metabolites were stored at −80 °C prior to the subsequent analysis.
UPLC/Q-TOF MS analysis
Chromatographic separation was performed using a UPLC system (ACQUITY UPLCTM; Waters Corp., Milford, MA, USA). Reversed-phase separation was carried out utilizing an ACQUITY UPLC T3 column (100 mm × 2.1 mm, 1.8 mm) (186008756, Waters Corp.). Then eluted metabolites were tested and quantified by Xevo G2-XS Q-Tof and QI software (Waters MS Technologies, Manchester, UK). For the positive and negative ion modes, the ion spray floating voltages were set at 3 and −2.5 kV, respectively. The MS data were acquired in centroid mode using Masslynx NT 4.1. The mass range was 50–1000 m/z.
Metabolomics analysis
The raw data spectra collected from UPLC-Q/TOF MS were subjected to ion pair extraction, peak alignment, peak matching, and peak intensity correction utilizing QI software (Waters Corp.). The main parameters were set as follows: quality range, 100–1000 amu; quality tolerance, 0.02; minimum intensity, 1%; quality window, 0.02; retention time window, 0.20; and noise cancellation level, 6. After employing an 80% filtering rule to remove zero and missing values from the obtained data table, the data were normalized to the total ion intensity of each chromatogram and saved as new data. The results comprised the three-dimensional matrix consisting of a specified peak sequence number (corresponding to the retention time and mass-to-charge ratio), observation points (sample number), and normalized peak intensity.
Data analysis and identification of differential metabolites
Multivariate data were subjected to PCA and partial least squares-discriminant analysis (PLS-DA) using SIMCA-P 11.5. Numerous metabolites were identified in the PCA and PLS-DA loading plots. Potential biomarkers were explored using OPLS-DA-produced variable importance for projection (VIP) values, whereas variables with VIP > 1 were further processed for FC and Student’s t-tests. Moreover, variables with VIP > 1, P value < 0.05, and FC ≥ 2 were considered to be significantly differential metabolites. The metabolites were determined by the Human Metabolome Database. In addition, coverage signals used to represent the heatmap of differential metabolites were generated using pheatmap.
KEGG pathway enrichment was analyzed using MetaboAnalyst 5.0 to explore the potential functions of the differential metabolites.
Gut microbiome taxonomic and blood metabolic data integration
To investigate the association between differential gut microbiome taxonomic indicators and differential blood metabolites, the correlation between differential genera identified by metagenomics and differential blood metabolites was analyzed using corr.test by R v3.6.1.
Cell culture and treatments
KAS-6/1 (ZKCC61732) and U266 (ZKCC62024) cells were purchased from Beijing Zhongke Quality Inspection Biotechnology (Beijing, China). Besides, peripheral blood mononuclear cells (PBMCs) were isolated from the whole blood of HC utilizing ficoll gradient centrifugation, and then CD8 T cells were isolated from PBMCs by positive selection utilizing Human CD8 Microbeads (130-045-201, Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. All cells were cultured in RPMI 1640 medium (SH30809.01, HyClone, Logan, USA) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin in a humidified atmosphere of 5% CO2 at 37 °C. Besides, CD8 T cells were activated by incubating with interleukin (IL)-2 (200-02-1MG, Thermo Fisher Scientific, Waltham, MA, USA) and IL-12 (200-12-10UG, Thermo Fisher Scientific).
KAS-6/1, U266, and CD8 T cells were treated with 50, 100, and 200 μM PC (S4788, Selleck Chemical, Houston, TX, USA) or 220 ng/mL LPA (22556-62-3, Merck KGaA, Darmstadt, Germany) to identify the effects of PC and LPA. Transfections were performed using Lipofectamine 2000 (11668019, Thermo Fisher Scientific). Besides, KAS-6/1 and U266 cells were co-cultured with Lachnospiraceae (YS-JZ10537, YuShao Biology, Shanghai, China) to determine the source of PC.
Detection of PC
The level of PC derived from KAS-6/1 and U266 cells was identified by PC Colorimetric Assay Kit (E-BC-K796-M, Elabscience, Wuhan, Hubei, China) according to the manufacturer’s instruction.
ELISA
The LPA levels in KAS-6/1 and U266 cells and the released GZMB levels in CD8 T cells were measured using the human LPA ELISA kit (CSB-EQ028005HU, CUSABIO, Wuhan, Hubei, China) and the human GZMB ELISA kit (CSB-E08718h, CUSABIO) in accordance with the manufacturer’s protocol. The absorbance at 450 nm was measured using a Multiskan MK3 system (Thermo Fisher Scientific).
CCK-8 assay
To detect cell viability or proliferation rate, CCK-8 assay was performed as previously described47.
Flow cytometric analysis
The ratio of GZMB + CD8 T cells was analyzed by FACSARIA flow cytometry (BD Biosciences, SanJose, CA, USA) using FITC-labeled CD8 antibody (1:200, 130-110-677, Miltenyi Biotec) and PE-labeled GZMB antibody (1:200, 130-116-486, Miltenyi Biotec). Besides, the number of KAS-6/1 and U266 cells transfected with RFP expression vectors was also determined by FACSARIA flow cytometry (BD Biosciences).
qRT-PCR
In this study, qRT-PCR was performed as previously described47. The primers used for qRT-PCR were listed in Supplementary Table 3. To quantify the abundance of Lachnospiraceae, the genomic bacterial DNA was extracted from NOD/SCID mice stools by the E.Z.N.A. Stool DNA Kit (M4015-02, Omega, Norcross, GA, USA) according to the manufacturer’s guidelines, then qRT-PCR was performed using TB Green Premix Ex Taq (Tli RNaseH Plus) (RR420A, Takara Biotechnology, Daliang, Liaoning, China) and specific primers for 16S ribosomal RNA (16 s rRNA) gene of Lachnospiraceae previously described38. RNA polymerase subunit beta (rpoB) was used as the endogenous control. Additionally, the relative amount of 16 s rRNA gene of Lachnospiraceae was normalized to that of rpoB and estimated by the ΔΔCT.
Knockout of Sb9 in MM cells
Upon transfection of CRISPR-Cas9 nucleases targeting Sb9, KAS-6/1, and U266 cells were cultured for three weeks to determine the extent of Sb9 KO.
WB
WB was carried out as previously described47. The primary antibodies used for WB included Sb9 antibody (1:1000, ab150400, Abcam, Cambridge, UK), CD9 antibody (1:500, bs-6934R, Bioss, Beijing, China) and CD63 antibody (1:500, bs-1523R, Bioss), GZMB antibody (1:1000, ab255598, Abcam), TP53 antibody (1:1000, 9282, Cell Signaling Technology, Danvers, MA, USA), UFM1 antibody (1:2000, ab109305, Abcam), and GAPDH antibody (1:10,000, KC-5G5, Aksomicks, Shanghai, China).
Touchdown PCR
Total RNA was extracted from KAS-6/1 and U266 cells by TRIzol (15596026, Invitrogen, Carlsbad, CA, USA), and cDNA was synthesized by PrimeScript RT Reagent Kit (RR037A, Takara Biotechnology). Then touchdown PCR was performed in a volume of 50 μL Phanta Max Buffer (P505, Vazyme, Nanjing, Jiangsu, China) as follows: 94 °C for 3 min, 94 °C for 15 sec, 74 °C for 60 sec for 5 cycles; 94 °C for 15 sec, 72 °C for 60 sec for 5 cycles; 94 °C for 15 sec, 70 °C for 60 sec for 5 cycles; 94 °C for 15 sec, 68 °C for 60 sec for 25 cycles followed by 68 °C for 5 min.
RIP
KAS-6/1 and U266 cells were collected and lysed followed by the interruption of nucleic acid using ultrasound. Then cell lysate was incubated with LIN28A antibody (1:30, ab279647, Abcam) or LIN28B antibody (1:50, ab191881, Abcam) at 4 °C overnight. The next day immunoprecipitated RNA fragments were captured using avidin magnetic beads and analyzed by qRT-PCR.
Minigene reporter assay
The Sb9-FL minigene was constructed by cloning a sequence consisting of exons 2–4 into the pCDNA3.1 vector. Then minigene containing deletion mutation of GGAGA motif (Sb9-Mut) were established based on the Sb9-FL minigene plasmid. Subsequently, minigene plasmids were transfected into KAS-6/1 and U266 cells, followed by the detection of Sb9 pre-mRNA splicing using PCR. The sequences of primers used for PCR were listed as follows: minigene F: 5’-GCACTGTCTTTAAACACAGAGGA-3’; minigene R: 5’-TTGAAGACAGGATTCCTTAAACG-3’.
Dual-luciferase reporter assay
To identify the effects of LIN28A/B on Sb9 pre-mRNA splicing, intron 2 of Sb9 was inserted into the firefly luciferase gene in the pGL3 vector (pGL3-Sb9 intron 2). Besides, the GZMB promoter was also inserted into pGL3 vector to determine the effect of TP53 on GZMB transcription. Subsequently, intron 2 of Sb9-containing pGL3 vectors were co-transfected with pRL-TK Renilla luciferase vectors into KAS-6/1 and U266 cells, while GZMB promoter-containing pGL3 vectors were co-transfected with pRL-TK Renilla luciferase vectors into CD8 T cells, then luciferase activity was detected by Dual Luciferase Reporter Assay System (E1910, Promega, Madison, WI, USA) at 48 h post transfection.
Isolation and analysis of exosomes
Exosomes were isolated from cell culture medium using the exoEasy Maxi Kit (76064, QIAGEN, Hilden, Germany), then the size of exosomes was detected by nanoparticle tracking analyzer N30E (NanoFCM, Nottingham, UK) and the markers of exosomes were detected by WB using CD9 antibody (1:500, bs-6934R, Bioss, Beijing, China) and CD63 antibody (1:500, bs-1523R, Bioss). Besides, exosomes from KAS-6/1 and U266 cells were labeled with PKH67 (D0031, Solarbio, Beijing, China) to verify their uptake by CD8 T cells by flow cytometric analysis using FACSARIA flow cytometry (BD Biosciences).
ChIP
CD8 T cells were cross-linked with 1% formaldehyde for 15 min at 25 °C stopped by 1.375 M glycine. Subsequently, CD8 T cells were suspended in a lysis buffer to shear the DNA by sonication. Then 450 mL dilution buffer was added in a 30 mg DNA chromatin sample to achieve a total volume of 500 mL. Subsequently, DNA chromatin samples were incubated with 1 μg of TP53 antibody (ab246550, Abcam) or anti-IgG antibody (ab218427, Abcam) followed by protein A/G magnetic beads at 4 °C overnight. The next day magnetic beads were collected by a magnetic separation device and cleaned. Then immunoprecipitated DNAs were eluted with 100 μL of elution buffer at 62 °C for 2 h and analyzed using PCR and qRT-PCR.
EMSA
The TP53 protein (ab237007, Abcam) was incubated with TP53 antibody (ab246550, Abcam) for 30 min at 25 °C and then with 2 ng of biotin-labeled WT or TP53 binding site-mutated (MUT) GZMB promoter fragment for 20 min at 25 °C. Besides, linear pUC19 was used for competition experiments. Next, the mixtures were loaded onto a 5% polyacrylamide gel and run at 10 V/cm followed by the transfer onto nylon membranes. Membranes were incubated with streptavidin-conjugated horseradish peroxidase and then with ECL Plus Reagent (A10016L, Abmart, Shanghai, China) following the development using X-ray irradiation.
Co-IP
Co-IP was performed using UFM1 antibody (ab109305, Abcam) or TP53 antibody (9282, Cell Signaling Technology) as previously described47. For competitive Co-IP, 0.5 μg, 1 μg, and 2 μg Sb9 expression vector was transfected into CD8 T cells and then the association between TP53 and Sb9 or UFC1 was determined by Co-IP using TP53 antibody (9282, Cell Signaling Technology).
Cycloheximide treatment
When grown to ~80% confluence, CD8 T cells were treated with 50 μg/mL cycloheximide (IC0720, Solarbio) for 3 h, 6 h, and 12 h. Then, CD8 T cells were collected, followed by the detection of TP53 levels using WB.
In vivo assays
Animal protocols were approved by the Institutional Animal Care and Use Committee of Shengjing Hospital, China Medical University (#2023PS958K). NOD/SCID mice used for in vivo assays were obtained from the Shanghai Model Organisms Center (SM-019, Shanghai, China) and maintained under specific pathogen-free, while experimental and control mice were co-housed. First, 100 μL Lachnospiraceae (YS-JZ10537, YuShao Biology) suspension (107 colony-forming unit/mL) was transplanted into an eight-week-old female NOD/SCID mouse by gavage for 14 consecutive days. To determine the role of Lachnospiraceae and PC in the cytotoxic effect of CD8 T cells on MM cells in vivo, 2 × 106 RFP-labeled WT or Sb9 KO KAS-6/1 and U266 cells were given into each NOD/SCID mouse by tail vein infusion combined with 200 μM/kg PC and 1 × 107 human CD8 T cells from HC. Besides, flow cytometric analysis was utilized to verify the number of RFP-labeled KAS-6/1 and U266 cells in mouse blood at day 14 after injection to further identify the role of Lachnospiraceae and PC in the cytotoxic effect of CD8 T cells.
Next, the effects of Lachnospiraceae and PC and CD8 T cells on tumor growth were investigated. 2 × 106 WT or Sb9 KO KAS-6/1 and U266 cells bearing a pcDNA3.1-luciferase reporter were injected into the right dorsal flank of each NOD-SCID mouse. After initial tumor engraftment on day 9 post-subcutaneous injection, 200 μM/kg PC and 1 × 107 human CD8 T cells from HC were intraperitoneally injected into each NOD-SCID mouse. Then, mice received intraperitoneal injection of luciferin (150 mg/kg) (P1042, Promega), and luminescent tumor images were imaged with Xenogen IVIS 200 (Caliper LifeSciences, Hopkinton, MA, USA) on day 9 and 25. Moreover, mice were monitored daily, and the size of xenograft tumors was measured and calculated as (width × length) every 5 days. The maximum diameter of xenograft tumors allowed by our institutional ethical board is 15 mm in mice, and we have adhered to the size limit in our experiments. Finally, mice were anesthetized with 50 mg/kg of pentobarbital and euthanized using 30% v/min CO2 at day 25 post-subcutaneous injection to measure the size and weight of xenograft tumors. If the diameter of the xenograft tumor exceeds 15 mm, in vivo, assays would be terminated early, and mice would be anesthetized with 50 mg/kg of pentobarbital and euthanized using 30% v/min CO2.
Statistical analyses
SPSS software (version 21.0; IBM Corp., Armonk, NY, USA) was used for statistical analyses. All statistical tests were two-sided, and the standard deviation of the mean is represented as error bars. For in vitro and in vivo studies, unpaired Student’s t test was used for the statistical analysis between two groups, while one-way ANOVA followed by post-hoc Tukey’s test was utilized for the statistics among multiple groups. Adjustments were made for multiple comparisons. A multivariate logistic regression analysis was utilized for analyzing the correlation between clinicopathological features and Lachnospiraceae abundance or blood PC level in MM patients. Cox analysis was performed to analyze the association between the survival rate and Lachnospiraceae abundance or blood PC level in MM patients. Statistical significance was set at P < 0.05.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data are included in the Supplementary Information or available from the authors, as are unique reagents used in this Article. The raw numbers for charts and graphs are available in the Source Data file whenever possible. 16 s ribosomal RNA gene sequencing and metagenomic sequencing data have been deposited in SRA under the PRJNA1106227, https://www.ncbi.nlm.nih.gov/sra/?term=PRJNA1106227. FACS files have been deposited in Figshare, https://figshare.com/s/4c04c2fb25f63134e23c. Source data are provided with this paper.
References
Siegel, R. L., Miller, K. D., Fuchs, H. E. & Jemal, A. Cancer statistics, 2021. CA Cancer J. Clin. 71, 7–33 (2021).
Rajkumar, S. V. Multiple myeloma: every year a new standard? Hematol. Oncol. 37, 62–65 (2019).
Vrábel, D., Pour, L. & Ševčíková, S. The impact of NF-κB signaling on pathogenesis and current treatment strategies in multiple myeloma. Blood Rev. 34, 56–66 (2019).
Jian, X. et al. Alterations of gut microbiome accelerate multiple myeloma progression by increasing the relative abundances of nitrogen-recycling bacteria. Microbiome 8, 74 (2020).
Pawlyn, C. & Davies, F. E. Toward personalized treatment in multiple myeloma based on molecular characteristics. Blood 133, 660–675 (2019).
Fricker, L. D. Proteasome inhibitor drugs. Annu. Rev. Pharmacol. Toxicol. 60, 457–476 (2020).
Ponziani, F. R. et al. Hepatocellular carcinoma is associated with gut microbiota profile and inflammation in nonalcoholic fatty liver disease. Hepatology 69, 107–120 (2019).
Zhang, L., Xiang, Y., Li, Y. & Zhang, J. Gut microbiome in multiple myeloma: mechanisms of progression and clinical applications. Front. Immunol. 13, 1058272 (2022).
Pei, B., Ma, X., Wu, N., Wang, C. & Yang, W. Effect of microbiome group on immune checkpoint inhibitors in treatment of gastrointestinal tumors. Chin. J. Cancer Res. 35, 252–265 (2023).
Calcinotto, A. et al. Microbiota-driven interleukin-17-producing cells and eosinophils synergize to accelerate multiple myeloma progression. Nat. Commun. 9, 4832 (2018).
Zhao, F., An, R., Wang, L., Shan, J. & Wang, X. Specific gut microbiome and serum metabolome changes in lung cancer patients. Front. Cell Infect. Microbiol. 11, 725284 (2021).
Wang, Y. et al. The gut microbiota reprograms intestinal lipid metabolism through long noncoding RNA Snhg9. Science 381, 851–857 (2023).
Saito, R. F., Andrade, L. N. S., Bustos, S. O. & Chammas, R. Phosphatidylcholine-derived lipid mediators: the crosstalk between cancer cells and immune cells. Front. Immunol. 13, 768606 (2022).
Law S. H. et al. An updated review of lysophosphatidylcholine metabolism in human diseases. Int. J. Mol. Sci. 20, 1149 (2019).
Mathew, D. et al. LPA(5) is an inhibitory receptor that suppresses CD8 T-cell cytotoxic function via disruption of early TCR signaling. Front. Immunol. 10, 1159 (2019).
Matas-Rico, E. et al. Autotaxin impedes anti-tumor immunity by suppressing chemotaxis and tumor infiltration of CD8(+) T cells. Cell Rep. 37, 110013 (2021).
Botta, C. et al. Large T cell clones expressing immune checkpoints increase during multiple myeloma evolution and predict treatment resistance. Nat. Commun. 14, 5825 (2023).
Cao, X. et al. Granzyme B and perforin are important for regulatory T cell-mediated suppression of tumor clearance. Immunity 27, 635–646 (2007).
Bassez, A. et al. A single-cell map of intratumoral changes during anti-PD1 treatment of patients with breast cancer. Nat. Med. 27, 820–832 (2021).
Han, R. et al. Inhibition of SerpinB9 to enhance granzyme B-based tumor therapy by using a modified biomimetic nanoplatform with a cascade strategy. Biomaterials 288, 121723 (2022).
Jiang, L. et al. Direct tumor killing and immunotherapy through anti-SerpinB9 therapy. Cell 183, 1219–1233.e1218 (2020).
Chen, Y. et al. Proteomics-based approach reveals the involvement of SERPINB9 in recurrent and relapsed multiple myeloma. J. Proteome Res. 20, 2673–2686 (2021).
Szlasa, W., Zendran, I., Zalesińska, A., Tarek, M. & Kulbacka, J. Lipid composition of the cancer cell membrane. J. Bioenerg. Biomembr. 52, 321–342 (2020).
Jia, W., Eneh, J. O., Ratnaparkhe, S., Altman, M. K. & Murph, M. M. MicroRNA-30c-2* expressed in ovarian cancer cells suppresses growth factor-induced cellular proliferation and downregulates the oncogene BCL9. Mol. Cancer Res. 9, 1732–1745 (2011).
Zhang, Y. et al. Lin28 enhances de novo fatty acid synthesis to promote cancer progression via SREBP-1. EMBO Rep. 20, e48115 (2019).
Manier, S. et al. The LIN28B/let-7 axis is a novel therapeutic pathway in multiple myeloma. Leukemia 31, 853–860 (2017).
Wilbert, M. L. et al. LIN28 binds messenger RNAs at GGAGA motifs and regulates splicing factor abundance. Mol. Cell 48, 195–206 (2012).
Menu, E. & Vanderkerken, K. Exosomes in multiple myeloma: from bench to bedside. Blood 140, 2429–2442 (2022).
Xu, Y. et al. Exosomes and their derivatives as biomarkers and therapeutic delivery agents for cardiovascular diseases: situations and challenges. J. Transl. Int. Med. 11, 341–354 (2023).
Giannotta, C., Autino, F. & Massaia, M. The immune suppressive tumor microenvironment in multiple myeloma: the contribution of myeloid-derived suppressor cells. Front. Immunol. 13, 1102471 (2022).
Jovanović, K. K. et al. Deregulation and targeting of TP53 pathway in multiple myeloma. Front. Oncol. 8, 665 (2018).
Liu, J. et al. UFMylation maintains tumour suppressor p53 stability by antagonizing its ubiquitination. Nat. Cell Biol. 22, 1056–1063 (2020).
Zhang, X. et al. Tissue-resident Lachnospiraceae family bacteria protect against colorectal carcinogenesis by promoting tumor immune surveillance. Cell Host Microbe 31, 418–432.e418 (2023).
Shi, F. et al. Altered gut microbiome composition by appendectomy contributes to colorectal cancer. Oncogene 42, 530–540 (2023).
Flemer, B. et al. The oral microbiota in colorectal cancer is distinctive and predictive. Gut 67, 1454–1463 (2018).
Peng, Z. et al. The gut microbiome is associated with clinical response to anti-PD-1/PD-L1 immunotherapy in gastrointestinal cancer. Cancer Immunol. Res. 8, 1251–1261 (2020).
Guevara-Ramírez, P. et al. Role of the gut microbiota in hematologic cancer. Front. Microbiol. 14, 1185787 (2023).
Yang, C. J. et al. Oral fecal transplantation enriches Lachnospiraceae and butyrate to mitigate acute liver injury. Cell Rep. 43, 113591 (2024).
Houweling, M., Klein, W. & Geelen, M. J. Regulation of phosphatidylcholine and phosphatidylethanolamine synthesis in rat hepatocytes by 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR). Biochem. J. 362, 97–104 (2002).
Zhong, Y. et al. FTY720 induces ferroptosis and autophagy via PP2A/AMPK pathway in multiple myeloma cells. Life Sci. 260, 118077 (2020).
Podhorecka, M., Ibanez, B. & Dmoszyńska, A. Metformin - its potential anti-cancer and anti-aging effects. Postepy Hig. Med. Dosw. 71, 170–175 (2017).
Tsanov, K. M. et al. LIN28 phosphorylation by MAPK/ERK couples signalling to the post-transcriptional control of pluripotency. Nat. Cell Biol. 19, 60–67 (2017).
Rowshani, A. T. et al. The granzyme B inhibitor SERPINB9 (protease inhibitor 9) circulates in blood and increases on primary cytomegalovirus infection after renal transplantation. J. Infect. Dis. 192, 1908–1911 (2005).
Yang, J. et al. Metformin induces ferroptosis by inhibiting UFMylation of SLC7A11 in breast cancer. J. Exp. Clin. Cancer Res. 40, 206 (2021).
Wang, K. et al. The ufmylation modification of ribosomal protein L10 in the development of pancreatic adenocarcinoma. Cell Death Dis. 14, 350 (2023).
Mao M. et al. Modification of PLAC8 by UFM1 affects tumorous proliferation and immune response by impacting PD-L1 levels in triple-negative breast cancer. J. Immunother. Cancer 10, e005668 (2022).
Wang, X. et al. MUC20 regulated by extrachromosomal circular DNA attenuates proteasome inhibitor resistance of multiple myeloma by modulating cuproptosis. J. Exp. Clin. Cancer Res. 43, 68 (2024).
Acknowledgements
This work was supported by the Natural Science Foundation of Liaoning Province (grant no. 2022-YGJC-61; 2022-MS-219), the 345 Talent Project of the Shengjing Hospital of China Medical University, and the Applied Basic Research Program of Liaoning Provincial Department of Science and Technology (grant no. 2023JH2/101300036).
Author information
Authors and Affiliations
Contributions
H.W., R.Z., X.T.W., and X.B.W. designed the study and wrote and edited the manuscript; W.Y., X.S., Y.Z., X.L., X.J., L.G., J.Y., A.L., and H.Y. performed the experiments and analyzed the data. All authors read and approved the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Ulrike Seifert and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Yan, W., Shi, X., Zhao, Y. et al. Microbiota-reprogrammed phosphatidylcholine inactivates cytotoxic CD8 T cells through UFMylation via exosomal SerpinB9 in multiple myeloma. Nat Commun 16, 2863 (2025). https://doi.org/10.1038/s41467-025-57966-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-57966-5