Intestinal Akkermansia muciniphila complements the efficacy of PD1 therapy in MAFLD-related hepatocellular carcinoma
肠道阿克曼氏菌增强 PD1 疗法对 MAFLD 相关肝细胞癌的疗效

Akk is identified in the MAFLD-promoted HCC mouse model
在 MAFLD 促进的 HCC 小鼠模型中鉴定出 Akk 菌

Recent reports have shown that gut microbiota alteration is associated with MAFLD and HCC progression.⁹ However, how the gut microbiota is altered during the development of MAFLD-HCC remains unknown. Therefore, we established an MAFLD-promoted HCC mouse model, where the mouse develops HCC in 35 weeks upon injection of diethylnitrosamine (DEN) and a high-fat diet (HFD)¹⁷ (Figure 1A). H&E staining revealed the hepatocellular ballooning of the NASH background and histopathology of HCC tumors (Figure 1B). Additionally, BODIPY and oil red O staining demonstrated the accumulation of numerous hepatic lipids upon HFD feeding (Figure 1B). During the HCC tumor development, we collected feces at five different time points and extracted corresponding fecal DNA for 16S rRNA sequencing analysis. The relative abundance of dominant genera was compared. Genera with abundance ≥0.5% were bar plotted. We found that Akk decreased approximately 40-fold from healthy to HCC in a time-dependent manner (Figures 1C and 1D). Principal coordinates analysis of the 16S amplicon data using a two-dimensional principal-component analysis (2D-PCA) plot of the experimental group displayed a wide separation among different time points (Figure 1E). qPCR analysis confirmed this observation, in which the levels of Akk were decreased across the experimental time points during the HCC development (Figure 1F). To determine whether age plays a role in the decrease of Akk levels, we examined the Akk levels in fecal samples from control mice fed a normal diet at the same time point as the MAFLD-promoted HCC mouse model using qPCR analysis (Figure S1A). Our results showed that there is no significant difference in Akk levels across the same five time points as HFD diet (Figure S1B), further confirming that the reduction in Akk is associated with the development of MAFLD-HCC.
近期研究表明，肠道菌群改变与代谢相关脂肪性肝病（MAFLD）和肝细胞癌（HCC）进展相关。然而，MAFLD-HCC 发展过程中肠道菌群如何变化尚不明确。为此，我们建立了 MAFLD 促进的 HCC 小鼠模型，通过注射二乙基亚硝胺（DEN）联合高脂饮食（HFD）喂养 35 周诱导小鼠发生 HCC（图 A）。H&E 染色显示非酒精性脂肪性肝炎（NASH）背景下的肝细胞气球样变及 HCC 肿瘤组织病理学特征（图 B）。BODIPY 和油红 O 染色进一步证实 HFD 喂养导致大量肝脏脂质蓄积（图 B）。在 HCC 肿瘤发展过程中，我们在五个不同时间点采集粪便样本并提取相应粪便 DNA 进行 16S rRNA 测序分析。比较优势菌属相对丰度后，将丰度≥0.5%的菌属绘制柱状图。研究发现，从健康状态到 HCC 阶段，阿克曼菌（Akk）丰度呈现时间依赖性下降，降幅约达 40 倍（图 C 和 1D）。 基于 16S 扩增子数据的主坐标分析显示，实验组二维主成分分析（2D-PCA）图中不同时间点呈现明显分离趋势（图 6E）。qPCR 分析验证了这一发现，在肝癌发展过程中，Akk 菌水平随实验时间点推移持续下降（图 7F）。为探究年龄因素是否影响 Akk 水平下降，我们通过 qPCR 检测了正常饮食对照组小鼠在与 MAFLD 诱发肝癌模型相同时间点的粪便样本中 Akk 水平（图 8A）。结果显示，与高脂饮食组相同的五个时间点间 Akk 水平无显著差异（图 9B），进一步证实 Akk 减少与 MAFLD-HCC 发展进程相关。

Akk ameliorates tumor progression in MAFLD-HCC mouse models
Akk 菌可改善 MAFLD-HCC 小鼠模型的肿瘤进展

Based on this interesting finding, we proceeded with animal testing to examine whether supplementation with Akk can reverse tumor growth in an orthotopic MAFLD-HCC mouse model. For this purpose, we fed the mice with a methionine- and choline-deficient (MCD) diet, followed by orthotopic implantation of luciferase-labeled RIL-175 cells (Figure 2A).¹⁸ The successful establishment of steatotic livers was confirmed through H&E, BODIPY, and oil red O staining. Notably, the severity of steatosis was reduced following Akk treatment (Figures 2B and 2C). To assess tumor progression, we utilized bioluminescence imaging (BLI) to track tumor development upon tumor inoculation at day 6 and day 12 (Figure S2). In day 20 with the luciferase signal, larger tumors were found in the MCD-treated groups than in mice fed a normal diet, suggesting a tumor-promoting capacity in the NASH background (Figures 2D–2F). Akk-treated group exhibited the smallest tumor burden when compared with the other two groups. To examine whether Akk retards tumor growth by suppressing the proliferation of HCC cells, we examined the expression of proliferating cell nuclear antigen (PCNA) between the PBS-control and Akk-treated groups of MCD diet. We observed a significant suppression of HCC tumor proliferation upon Akk treatment (Figure S3). In a separate MAFLD-HCC model, we administered Akk to mice on a high-fat, high-cholesterol (HFHC) diet for more than 8 months (Figure S4A). Compared to control group, Akk-treated mice exhibited a trend toward decreased tumor burden, as evidenced by reduced liver weight (p = 0.0592) (Figure S4B), which was associated with reduced liver steatosis (Figure S4C). Moreover, qPCR analysis confirmed that the levels of Akk were decreased across the experimental time points with the MAFLD development (Figure S4D). Overall, our results indicate that Akk treatment can slow down the development of MAFLD into HCC.
基于这一有趣发现，我们随即开展动物实验，探究补充 Akk 菌能否逆转原位 MAFLD-HCC 小鼠模型的肿瘤生长。为此，我们给小鼠喂食缺乏蛋氨酸和胆碱（MCD）的饲料，随后原位植入荧光素酶标记的 RIL-175 细胞（图 0A）。通过 H&E 染色、BODIPY 染色和油红 O 染色证实了脂肪肝模型的成功建立。值得注意的是，Akk 治疗组肝脏脂肪变性程度显著减轻（图 2B 和 2C）。为评估肿瘤进展，我们在接种后第 6 天和第 12 天采用生物发光成像（BLI）技术追踪肿瘤发展情况（图 3）。第 20 天荧光信号显示，MCD 饲料组比正常饮食组出现更大肿瘤，表明 NASH 微环境具有促肿瘤特性（图 4D-2F）。与其他两组相比，Akk 治疗组表现出最小的肿瘤负荷。 为探究 Akk 是否通过抑制肝癌细胞增殖来延缓肿瘤生长，我们检测了 MCD 饮食组中 PBS 对照组与 Akk 治疗组的增殖细胞核抗原（PCNA）表达水平。结果显示 Akk 治疗能显著抑制肝癌肿瘤增殖（图 5）。在另一项 MAFLD-HCC 模型中，我们对高脂高胆固醇（HFHC）饮食喂养超过 8 个月的小鼠给予 Akk 干预（图 6A）。与对照组相比，Akk 治疗组小鼠表现出肿瘤负荷降低趋势，肝脏重量减轻（p=0.0592）证实了这一结果（图 7B），且与肝脏脂肪变性程度减轻相关（图 8C）。此外，qPCR 分析证实随着 MAFLD 病程发展，各时间点 Akk 水平均呈下降趋势（图 9D）。综上，我们的研究结果表明 Akk 治疗可延缓 MAFLD 向 HCC 的进展。

Akk alleviates hepatic steatosis by suppressing cholesterol biosynthesis and bile acid metabolism
Akk 通过抑制胆固醇生物合成和胆汁酸代谢缓解肝脏脂肪变性

Given that serum total cholesterol (TC) and triglyceride (TG) are positively correlated with liver steatosis, we further examined whether supplementation with Akk would result in alterations of these two serum proteins in an orthotopic MAFLD-HCC mouse model fed with MCD diet. Upon analysis, Akk was further demonstrated to alleviate liver steatosis, as evidenced by a marked decrease in serum TC and TG levels (Figures 3A and 3B). Similar findings were observed in an MAFLD-HCC mouse model fed with HFHC (Figures S4E and S4F). To further examine the mechanism underlying the negative regulatory role of Akk in hepatic steatosis, we performed bulk RNA sequencing to compare the genetic alterations between the Akk-treated and PBS-control groups. Over 90% (41/50) of hallmark gene sets were upregulated in the control group, while 15 gene sets were significant at a false discovery rate (FDR) <25%. Less than 20% (9/50) of hallmark gene sets were upregulated in the Akk group, while only 3 gene sets were significant. Gene set enrichment analysis (GSEA) demonstrated that Akk downregulated pathways were related to cholesterol homeostasis, Wnt/β-catenin, PI3/Akt mTOR (mammalian target of rapamycin) signaling, etc. (Figure 3C). We focused on the effect of Akk on cholesterol biosynthesis (Figure 3D), as previous reports showed the role of cholesterol biosynthesis in the promotion of steatohepatitis-related HCC.^19,20 We analyzed the genes related to cholesterol homeostasis between the Akk-treated and PBS-control groups. Most cholesterol homeostasis-related genes were differentially expressed in Akk-treated and PBS-control groups. Consistent with this finding, several genes related to cholesterol biosynthesis were found to be significantly downregulated (Figure 3E). Bile acid synthesis is a major pathway for hepatic cholesterol catabolism,²¹ and its metabolites significantly contribute to the development of steatosis.²² Likewise, the Akk-treated group showed a significant decrease in the number of bile acid metabolites, including taurocholic acid, cholic acid, deoxycholic acid, tauroursodeoxycholic acid, and chenodeoxycholic acid (Figure 3F). Taken together, these results indicate that Akk administration ameliorates hepatic steatosis possibly via the suppression of cholesterol biosynthesis.
鉴于血清总胆固醇（TC）和甘油三酯（TG）与肝脏脂肪变性呈正相关，我们进一步研究了在 MCD 饮食诱导的 MAFLD-HCC 原位小鼠模型中补充 Akk 是否会导致这两种血清蛋白水平变化。分析显示，Akk 能显著降低血清 TC 和 TG 水平（图 0#A 和 3B），证实其可缓解肝脏脂肪变性。在 HFHC 饮食诱导的 MAFLD-HCC 小鼠模型中也观察到类似结果（图 1#E 和 S4F）。为深入探究 Akk 抑制肝脏脂肪变性的分子机制，我们通过批量 RNA 测序比较了 Akk 处理组与 PBS 对照组间的基因表达差异。对照组中超过 90%（41/50）的标志性基因集表达上调，其中 15 个基因集在错误发现率（FDR）<25%时具有统计学意义；而 Akk 组仅不到 20%（9/50）的标志性基因集表达上调，其中仅 3 个基因集达到显著水平。 基因集富集分析(GSEA)显示，Akk 菌下调的通路与胆固醇稳态、Wnt/β-连环蛋白、PI3/Akt mTOR(哺乳动物雷帕霉素靶蛋白)信号传导等相关(图 2C)。我们重点关注 Akk 菌对胆固醇生物合成的影响(图 3D)，因既往研究表明胆固醇生物合成在促进脂肪性肝炎相关肝癌中起关键作用。通过对比 Akk 菌处理组与 PBS 对照组，我们发现两组间多数胆固醇稳态相关基因存在差异表达。与此一致的是，若干胆固醇生物合成相关基因呈现显著下调(图 6E)。胆汁酸合成是肝脏胆固醇分解代谢的主要途径，其代谢产物对脂肪变性发展具有重要影响。同样，Akk 菌处理组中胆汁酸代谢物(包括牛磺胆酸、胆酸、脱氧胆酸、牛磺熊去氧胆酸和鹅去氧胆酸)数量显著减少(图 9F)。 综上所述，这些结果表明 Akk 菌的施用可能通过抑制胆固醇生物合成来改善肝脏脂肪变性。

Akk improves intestinal membrane integrity
Akk 菌可改善肠道黏膜完整性

Akk is a strictly anaerobic and mucin-degrading bacterium that colonizes the gut of humans and rodents. Akk is highly abundant (1%–5%) in the outer mucus layer near the intestinal lining of the gut.²³ Akk functions to maintain and strengthen the integrity of the gut lining. Given the physiological function of Akk in the intestine, we believe that a decrease in Akk may lead to impairment of the intestinal lining. Consistently, we found that Akk repaired the intestinal lining, as evidenced by the increased expression of tight junction proteins, including ZO-1, Occludin, and Claudins 1 and 4, in colonic tissues upon treatment with Akk (Figures S5A–S5E). Upon improvement of intestinal membrane integrity, we believe that an increase in Akk may lead to improvement of the intestinal lining, which subsequently decreases the flux of lipopolysaccharide (LPS). Consistently, LPS was downregulated in the serum of Akk-treated mice in an MAFLD-HCC model fed with HFHC diet (Figure S4G) and an orthotopic MAFLD-HCC mouse model fed with MCD diet (Figure S5F).
阿克曼菌（Akk）是一种严格厌氧且能降解黏蛋白的细菌，定植于人类和啮齿动物的肠道中。该菌在肠道黏膜外层含量丰富（1%-5%），其功能是维持并增强肠道黏膜屏障的完整性。鉴于阿克曼菌在肠道中的生理功能，我们认为其数量减少可能导致肠道黏膜损伤。实验证实，经阿克曼菌处理后，结肠组织中紧密连接蛋白（包括 ZO-1、Occludin 以及 Claudin 1 和 4）的表达增加，表明该菌具有修复肠道黏膜的作用（图 1A-S5E）。随着肠道黏膜完整性的改善，我们认为阿克曼菌数量增加可能改善肠道屏障功能，从而降低脂多糖（LPS）的渗透。在喂食高脂高胆固醇（HFHC）饮食的 MAFLD-HCC 模型中（图 2G）及喂食蛋氨酸胆碱缺乏（MCD）饮食的原位 MAFLD-HCC 小鼠模型中（图 3F），阿克曼菌处理组小鼠血清中的 LPS 水平均出现下调。

m-MDSCs and M2 macrophages are decreased in tumors of Akk-treated mice
阿克曼氏菌处理小鼠的肿瘤中 m-MDSCs 和 M2 巨噬细胞数量减少

Based on Figures S4G and S5F, we examined the effect of Akk on the regulation of innate immunity in the tumor microenvironment, as LPS is regarded as an important regulator of innate immune cells. We extracted CD45⁺ cells from tumors treated with either Akk or PBS and analyzed them by using scRNA-seq (Figure 4A). The uniform manifold approximation and projection (UMAP) plot, generated with scRNA-seq data, shows different immune cell clusters based on their marker expression (Figure 4B). In general, we observed increased populations of mast cells, plasma cells, dendritic cells (DCs), T cells, natural killer (NK) cells, and B cells in the Akk-treated group (Figure 4C). Secondary clusters were identified as Mo/Macs, m-MDSCs, M1 macrophages, M2 macrophages, and tumor-associated macrophages (TAMs) (Figure 4D). Interestingly, we found that the total populations of m-MDSCs and M2 macrophages were decreased, while total Ly6c2-high immature monocytes (“Ly6c2⁺monocytes”) were increased in the Akk-treated groups (Figure 4E). Consistently, immunoprofiling analysis found a significant decrease in m-MDSCs and M2 macrophages in Akk-treated tumors (Figure 4F). To further understand how Akk regulates the populations of m-MDSCs and M2 macrophages, we performed velocity analysis, which revealed the transcriptional dynamics of m-MDSCs and M2 macrophages. RNA velocity predicted that the differentiation and polarization of the M2 subpopulation from Ly6c⁺monocytes and m-MDSCs in tumors of the PBS group were increased. The polarization of M2 macrophages from Ly6c⁺monocytes in tumors of the Akk-treated group was suppressed (Figure 4G). To further clarify these cellular changes, we first elucidated the pathway that was altered in m-MDSCs upon Akk treatment. By GSEA analysis using the same set of scRNA-seq data, we found suppression of the nuclear factor κB (NF-κB) pathway with consistent suppression of NF-κB genes, which was further confirmed by qPCR analysis (Figures S6A and S6B). Notably, TLR2, a critical receptor of LPS, was found to be downregulated (Figure S6C). As TLR2-NF-κB activation was previously reported to play a crucial role in m-MDSC expansion,^24,25,26 we reasoned that Akk suppressed the expansion of m-MDSCs by blocking the influx of LPS to the liver via the intestinal membrane. Consistently, we found a significant decrease in the population of TLR2⁺ MDSCs, as evidenced by multiplexed immunofluorescence (IF) staining (Figure S6D). To further understand how Akk regulates the populations of these innate immune cells and, lastly, to further validate the function of m-MDSCs in M2 macrophage polarization, we performed a co-culture system in which mouse macrophage cell line (Raw264.7) was co-incubated with m-MDSCs in either PBS- or Akk-treated groups and found that m-MDSCs in Akk-treated groups significantly suppressed M2 polarization compared to the control group (Figure S6E). To confirm the role of LPS influx through the intestinal barrier in the regulation of m-MDSCs and M2 macrophages by Akk, we employed a mouse model of colitis, in which the intestinal barrier is compromised by giving 2% dextran sulfate sodium (DSS) for a week (Figure S7A). This model was successfully established as evidenced by H&E (Figure S7B) and Alcian blue staining (Figure S7C), which was accompanied by a substantial decrease in body weight (Figure S7D). Consistently, Akk mitigated the damaging effects of 2% DSS on intestinal membrane integrity and significantly reduced LPS detection (Figure S7E) and accompanied with decrease in m-MDSCs (Figures S7F and S7G) and M2 macrophages detection (Figures S7H and S7I). This finding further provides a mechanistic insight of how Akk regulates the populations of these innate immune cells.
基于 Figures S4 G 和 S5 F 的研究，我们考察了 Akk 对肿瘤微环境先天免疫调节的影响，因为 LPS 被认为是先天免疫细胞的重要调节因子。我们从经 Akk 或 PBS 处理的肿瘤中提取 CD45 ⁺ 细胞，并采用单细胞 RNA 测序（scRNA-seq）进行分析（ Figure 4 A）。通过 scRNA-seq 数据生成的均匀流形近似与投影（UMAP）图谱显示，基于标记物表达的不同免疫细胞群存在差异（ Figure 4 B）。总体而言，我们观察到 Akk 处理组中肥大细胞、浆细胞、树突状细胞（DCs）、T 细胞、自然杀伤（NK）细胞和 B 细胞数量增加（ Figure 4 C）。次级细胞群被鉴定为单核/巨噬细胞（Mo/Macs）、髓系来源的抑制性细胞（m-MDSCs）、M1 型巨噬细胞、M2 型巨噬细胞和肿瘤相关巨噬细胞（TAMs）（ Figure 4 D）。值得注意的是，Akk 处理组中 m-MDSCs 和 M2 型巨噬细胞总量减少，而 Ly6c2 高表达的不成熟单核细胞（"Ly6c2 ⁺ 单核细胞"）总量增加（ Figure 4 E）。免疫谱分析一致显示，经 Akk 处理的肿瘤中 m-MDSCs 和 M2 型巨噬细胞显著减少（ Figure 4 F）。 为深入探究 Akk 如何调控 m-MDSCs 和 M2 巨噬细胞群体，我们进行了速率分析，揭示了这两种细胞的转录动态特征。RNA 速率预测显示：PBS 组肿瘤中 Ly6c ⁺ 单核细胞向 M2 亚群的分化极化及 m-MDSCs 的 M2 极化程度增强；而 Akk 治疗组肿瘤中 Ly6c ⁺ 单核细胞向 M2 巨噬细胞的极化过程受到抑制（ Figure 4 G）。为进一步阐明这些细胞变化，我们首先解析了 Akk 干预后 m-MDSCs 中发生改变的信号通路。通过同一单细胞 RNA 测序数据的 GSEA 分析，发现核因子κB（NF-κB）通路及其相关基因表达均受到抑制，qPCR 分析进一步验证了这一结果（ Figures S6 A 和 S6B）。值得注意的是，LPS 的关键受体 TLR2 表达水平下调（ Figure S6 C）。鉴于既往研究报道 TLR2-NF-κB 激活对 m-MDSCs 扩增具有关键作用 ^{24 25 26} ，我们推测 Akk 可能通过阻断 LPS 经肠黏膜向肝脏的迁移，从而抑制 m-MDSCs 的扩增。 我们通过多重免疫荧光(IF)染色( Figure S6 D)证实，TLR2 ⁺ MDSCs 细胞群数量显著减少。为进一步探究 Akk 如何调控这些先天免疫细胞群，并最终验证 m-MDSCs 在 M2 型巨噬细胞极化中的作用，我们建立了共培养体系：将小鼠巨噬细胞系(Raw264.7)分别与 PBS 组或 Akk 处理组的 m-MDSCs 共孵育，发现相较于对照组，Akk 处理组的 m-MDSCs 显著抑制了 M2 极化( Figure S6 E)。为验证 Akk 通过肠道屏障的 LPS 内流调控 m-MDSCs 和 M2 巨噬细胞的作用机制，我们采用结肠炎小鼠模型（通过给予 2%葡聚糖硫酸钠(DSS)一周破坏肠道屏障）( Figure S7 A)。H&E( Figure S7 B)和阿利新蓝染色( Figure S7 C)显示模型构建成功，同时伴随体重显著下降( Figure S7 D)。 与之一致的是，Akk 减轻了 2% DSS 对肠道黏膜完整性的破坏作用，并显著降低了 LPS 检测水平（图 25E），同时伴随 m-MDSCs（图 26F 和 S7G）与 M2 型巨噬细胞（图 27H 和 S7I）数量的减少。该发现进一步从机制层面揭示了 Akk 如何调控这些先天免疫细胞群。

Effector memory CD4⁺/CD8⁺ T cells are increased in tumors of Akk-treated mice
Akk 处理小鼠肿瘤中效应记忆 CD4 ⁺ /CD8 ⁺ T 细胞数量增加

Next, we examined the effect of Akk on T cells by analyzing various T cell clusters (Figure 5A). In parallel with the decrease in m-MDSCs and M2 macrophages in the Akk-treated groups, as shown in Figure 4, we found increases in effector memory CD4⁺ cells, naive and effector memory CD8⁺ T cells, and CD4⁻CD8⁻ double-negative T cells in the tumors of the Akk-treated groups (Figure 5B). We also showed an increase in populations of exhausted CD4⁺ T cells (CD4⁺PD1⁺ and CD4⁺LAG3⁺) and exhausted CD8⁺ T cells (CD8α⁺PD1⁺) (Figure 5C) using flow cytometry analysis. Consistently, we showed increased infiltration of CD4⁺PD1⁺ and CD8α⁺PD1⁺ cells in the tumors of Akk-treated groups by multiplexed immunohistochemistry (IHC) analysis (Figure 5D). Concurrently, we found that Akk alters the immune-suppressive microenvironment, favoring T cell function by suppressing immune-suppressive cytokines, including Cxcl1, Ccl2, and Cxcl10, by qPCR analysis (Figure S8).
接着，我们通过分析各类 T 细胞亚群来考察 Akk 对 T 细胞的影响（图 0#A）。与 Akk 处理组中 m-MDSCs 和 M2 型巨噬细胞减少的趋势一致（如图 1#所示），我们发现该组肿瘤组织中效应记忆 CD4#细胞、初始及效应记忆 CD8#T 细胞、以及 CD4#CD8#双阴性 T 细胞均有所增加（图 6#B）。流式细胞术分析还显示，耗竭型 CD4#T 细胞（CD4#PD1#和 CD4#LAG3#）与耗竭型 CD8#T 细胞（CD8α#PD1#）的群体也有所增多（图 15#C）。多重免疫组化（IHC）分析同样证实，Akk 处理组肿瘤组织中 CD4#PD1#和 CD8α#PD1#细胞的浸润程度增强（图 20#D）。同时，qPCR 分析表明 Akk 通过抑制 Cxcl1、Ccl2 和 Cxcl10 等免疫抑制性细胞因子，改变了免疫抑制微环境，从而有利于 T 细胞功能发挥（图 21#）。

Combination treatment with Akk and PD1 suppressed the growth of the MAFLD-HCC mouse models
Akk 与 PD1 联合治疗抑制了 MAFLD-HCC 小鼠模型的肿瘤生长

Given the observation showing the role of Akk in the promotion of T cell infiltration and activity in Figure 5, we next examined the therapeutic efficacy of Akk in combination with PD1 in three MAFLD-HCC models. Firstly, we evaluated the effect of Akk supplementation in potentiating the effect of PD1 treatment in an orthotopic MAFLD-HCC mouse model fed with MCD diet (Figure 6A). Treatment with Akk and PD1 led to the maximal suppression of tumor growth, indicating that Akk treatment synergizes with PD1 treatment and is effective against liver tumors in vivo (Figures 6B–6E). We observed an increase in CD4⁺ and CD8⁺ T cells in both the Akk-treated and combination treatment groups (Figure 6F). To rule out the possibility of a tumor-specific effect, we utilized another mouse HCC cell line, Hep55.1c, which was injected directly into the left liver lobe of mice fed with MCD diet (Figure S9A). We observed that the combination of Akk and PD1 treatment exerted the maximal suppression of HCC tumor growth (Figures S9B–S9D). Furthermore, we further rule out the possibility of diet-specific effect by evaluating the combination effect of Akk and PD1 in the same orthotopic model in mice fed with HFHC diet (Figure S10A).²⁷ Consistent with the MCD diet model, we also showed that Akk in combination with PD1 treatment exerted maximal growth-suppressive effect in HCC tumor growth in the HFHC diet model (Figure S10B). Akk consistently delayed the progression of MAFLD (Figure S10C) with decrease in m-MDSCs (Figure S10D) and M2 macrophages (Figure S10E) but increase in infiltration of CD4⁺ and CD8⁺ cells (Figure S11) in the combination treatment group. Lastly, we examine whether the enhanced efficacy of PD1 by Akk is MAFLD specific by employing the orthotopic HCC model of mice fed with normal diet. As shown in Figures S1C and S1D, our findings revealed that Akk demonstrated a tendency to suppress HCC tumors and improve the effectiveness of PD1 therapy, although the results did not reach statistical significance. This suggests that Akk supplementation may be more effective in restraining HCC tumor growth and complementing PD1 treatment in HCC with an MAFLD background.
鉴于观察到 Akk 在促进 Figure 5 中 T 细胞浸润和活性方面的作用，我们接下来在三种 MAFLD-HCC 模型中检测了 Akk 联合 PD1 的治疗效果。首先，我们在 MCD 饮食诱导的原位 MAFLD-HCC 小鼠模型中评估了补充 Akk 对增强 PD1 治疗效果的影响（ Figure 6 A）。Akk 与 PD1 联合治疗实现了最大程度的肿瘤生长抑制，表明 Akk 治疗与 PD1 治疗具有协同作用，并在体内对肝脏肿瘤有效（ Figures 6 B–6E）。我们观察到 Akk 治疗组和联合治疗组中 CD4 ⁺ 和 CD8 ⁺ T 细胞均有所增加（ Figure 6 F）。为排除肿瘤特异性效应的可能性，我们使用了另一种小鼠 HCC 细胞系 Hep55.1c，将其直接注射至 MCD 饮食喂养小鼠的左肝叶（ Figure S9 A）。研究发现 Akk 与 PD1 联合治疗对 HCC 肿瘤生长产生了最大程度的抑制效果（ Figures S9 B–S9D）。 此外，我们通过评估高脂高胆固醇饮食（HFHC）小鼠原位模型中 Akk 与 PD1 的联合作用，进一步排除了饮食特异性效应的可能性（图 8A）。与 MCD 饮食模型一致，我们在 HFHC 饮食模型中也发现 Akk 联合 PD1 治疗对肝癌肿瘤生长具有最大抑制效果（图 10B）。联合治疗组中，Akk 持续延缓了 MAFLD 进展（图 11C），表现为 m-MDSCs（图 12D）和 M2 巨噬细胞（图 13E）减少，同时 CD4（图 14）和 CD8（图 15）细胞浸润增加（图 16）。最后，我们通过正常饮食小鼠的原位肝癌模型验证 Akk 增强 PD1 疗效是否具有 MAFLD 特异性。如图 17C 和 S1D 所示，尽管结果未达到统计学显著性，但 Akk 显示出抑制肝癌肿瘤并提升 PD1 治疗效果的倾向。这表明在 MAFLD 背景下，补充 Akk 可能更有效抑制肝癌肿瘤生长并增强 PD1 治疗效果。

Akk as a crucial bacterium in the regulation of PD1 treatment sensitivity
作为调节 PD1 治疗敏感性的关键细菌

To further investigate the role of Akk in determining PD1 sensitivity, 6-week-old C57BL/6 male mice were used for fecal microbiota transfer (FMT) using antibiotic cocktail for 2 weeks. After 2 weeks of treatment, we conducted an FMT from PD1 responder patients with high abundance of Akk in an orthotopic MAFLD-HCC mouse model (Figure S12A). The treatment FMT from PD1 responder patients together with PD1 significantly impacted HCC tumor growth and showed a tendency to enhance the effectiveness of PD1 therapy (Figure S12B). Strikingly, co-administration of Akk with FMT showed the significant tumor-suppressive effect in combination with PD1 treatment (Figure S12B). This study highlights the importance of Akk as a key bacterial species among the gut microbiota of PD1 responder patients in regulating the sensitivity of PD1 treatment.
为深入探究 Akk 在决定 PD1 敏感性中的作用，我们采用 6 周龄 C57BL/6 雄性小鼠，使用抗生素混合物进行为期 2 周的粪便微生物群移植（FMT）。治疗两周后，我们在原位 MAFLD-HCC 小鼠模型中实施了来自 Akk 高丰度 PD1 应答患者的 FMT（图 0#A）。来自 PD1 应答患者的治疗性 FMT 联合 PD1 显著抑制了 HCC 肿瘤生长，并显示出增强 PD1 治疗效果的倾向（图 1#B）。值得注意的是，Akk 与 FMT 联合给药时，与 PD1 治疗协同产生了显著的肿瘤抑制效应（图 2#B）。本研究揭示了 Akk 作为 PD1 应答患者肠道菌群中关键菌种，在调控 PD1 治疗敏感性方面的重要性。

Clinical relevance of Akk in MAFLD-HCC and PD1 sensitivity
Akk 在 MAFLD-HCC 与 PD1 敏感性中的临床相关性

To examine whether the role of Akk in MAFLD-HCC is clinically relevant, we analyzed fecal samples from HCC patients at the Fifth Medical Center of PLA General Hospital, Beijing, China. Patient characteristics are summarized in Table S1. Akk was found to be more abundant in the normal control group than in MAFLD-HCC patients (Figure 7A) by qPCR analysis. Interestingly, we found that there is no significant difference in the relative expression of Akk between HCC patients with non-hepatitis virus and HBV-induced HCC patients in a publicly available dataset (NCBI: PRJNA428932)²⁸ (Figure 7B). This result suggests that hepatitis-induced HCC also did not affect Akk levels. Consistent with the in vivo findings shown in Figure S5F, increased detection of LPS was observed in the plasma of MAFLD-HCC patients with lower expression of Akk (Figure 7C). Next, we examined the role of Akk in the treatment response to ICI therapy and the outcomes of patients with HCC. A total of 53 PD1-treated patients were screened (Table S2); 18 patients were enrolled in the PD1 responder group, and 35 patients were enrolled in the PD1 non-responder group. Strikingly, the analysis from metagenomic sequencing of Akk showed that the PD1 responder group patients had a higher Akk relative abundance than the PD1 non-responder group patients (Figure 7D). We further explored whether Akk was associated with patient survival by dividing HCC patients into three groups according to the relative abundance of Akk levels: Akk^High, Akk^Medium, and Akk^Low. Based on the relative abundance of Akk in patients, we divided the patients at a relative abundance greater than the 50th percentile (3.8 × 10⁻⁴) into the Akk^High group. Then, the next half of patients was grouped into Akk^Medium and Akk^Low in the 25th percentile (1.0 × 10⁻⁴) (Figure 7E) like the study of Derosa L. et al.²⁹ The percentage of patients with complete response (CR) or partial response (PR) was higher in the Akk^High group than in the Akk^Low group (41% vs. 7%, Figure 7F). Furthermore, Akk^High and Akk^Medium patients showed a better progression-free survival than their Akk^Low counterparts (Figure 7G). Our findings suggest that the enrichment of Akk might enhance the clinical response to PD1 treatment and prolong patient survival. Akk can be a predictive factor for the survival benefit of HCC patients receiving immunotherapy.
为探究 Akk 在 MAFLD-HCC 中的作用是否具有临床相关性，我们分析了中国人民解放军总医院第五医学中心肝癌患者的粪便样本。患者特征总结见 Table S1 。通过 qPCR 分析发现，正常对照组中 Akk 的丰度高于 MAFLD-HCC 患者（ Figure 7 A）。值得注意的是，我们在公开数据集（NCBI: PRJNA428932）中发现，非肝炎病毒导致的 HCC 患者与 HBV 诱导的 HCC 患者之间 Akk 相对表达量无显著差异（ ²⁸ ， Figure 7 B），这一结果表明肝炎诱发的 HCC 同样不影响 Akk 水平。与 Figure S5 F 所示的体内实验结果一致，在 Akk 低表达的 MAFLD-HCC 患者血浆中观察到 LPS 检测值升高（ Figure 7 C）。随后我们考察了 Akk 对 ICI 治疗反应及 HCC 患者预后的影响，共筛选出 53 例接受 PD1 治疗的患者（ Table S2 ），其中 18 例纳入 PD1 应答组，35 例纳入 PD1 无应答组。 值得注意的是，通过对 Akk 菌的宏基因组测序分析发现，PD1 应答组患者的 Akk 菌相对丰度高于非应答组患者（图 7D）。我们进一步探究 Akk 菌是否与患者生存期相关，将肝癌患者按 Akk 菌相对丰度水平分为三组：Akk 高组、Akk 中组和 Akk 低组。根据患者体内 Akk 菌相对丰度，我们将丰度高于 50 百分位数（3.8×10^5）的患者归入 Akk 高组，随后将剩余患者按 25 百分位数（1.0×10^5）分为 Akk 中组和 Akk 低组（图 16E），该方法参照了 Derosa L.等学者的研究方案。数据显示，Akk 高组患者获得完全缓解（CR）或部分缓解（PR）的比例显著高于 Akk 低组（41% vs. 7%，图 20F）。此外，Akk 高组和中组患者的无进展生存期较 Akk 低组患者更优（图 24G）。本研究结果表明，Akk 菌的富集可能增强 PD1 治疗的临床响应并延长患者生存期。 阿克曼菌可作为预测肝癌患者免疫治疗生存获益的指标。

Limitations of the study
本研究的局限性

Due to the substantial cost associated with single-cell sequencing, the sample size for this analysis is currently limited. To enhance the study’s validity, it would be advantageous to expand the sample size more. Besides that, we need to validate the infiltration and activations of T cells by examining the quantity of effector CD4⁺/CD8⁺ T cells in different mice models via flow cytometry analysis. For the FMT experiment, the current study is limited by the number of stool samples from MAFLD-HCC patients. In the future, fecal donors should be stratified into Akk^High and Akk^Low groups to verify Akk as key microbiota for determining PD1 sensitivity. Our in vivo animal experiments demonstrated that Akk preferentially promotes PD1 therapy in treating MAFLD-HCC. However, our clinical correlation between Akk level and PD1 sensitivity predominantly involves HBV-positive HCC patients, which may contradict the findings of our animal experiments. Future research should focus on conducting more comprehensive animal studies to evaluate and contrast the PD1 sensitivity in animal models of HBV-induced HCC and MAFLD-HCC. In the future, it is necessary to recruit stool samples from patients with MAFLD-HCC to verify the role of Akk in relation to PD1 responsiveness.
由于单细胞测序成本高昂，当前分析样本量有限。为提高研究可靠性，扩大样本规模将更具优势。此外，我们需要通过流式细胞术检测不同小鼠模型中效应性 CD4 ⁺ /CD8 ⁺ T 细胞的数量，以验证 T 细胞的浸润与活化状态。在粪菌移植实验中，当前研究受限于 MAFLD-HCC 患者粪便样本数量。未来应将粪便供体分为 Akk ^High 与 Akk ^Low 组别，以验证 Akk 菌作为决定 PD1 疗效关键菌群的作用。我们的体内动物实验证实 Akk 菌能优先增强 PD1 疗法对 MAFLD-HCC 的疗效，但临床数据中 Akk 水平与 PD1 敏感性的相关性主要来自 HBV 阳性 HCC 患者，这可能与动物实验结果存在矛盾。后续研究应开展更系统的动物实验，对比评估 HBV 诱发 HCC 与 MAFLD-HCC 动物模型中 PD1 疗法的敏感性差异。 未来有必要收集 MAFLD-HCC 患者的粪便样本，以验证 Akk 菌与 PD1 治疗反应之间的关联作用。

Lead contact  主要联系人

More information and inquiries about resources and reagents should be directed to Prof. Terence Kin Wah Lee (terence.kw.lee@polyu.edu.hk), who will address your requests.
有关资源和试剂的更多信息及查询，请直接联系李建华教授（terence.kw.lee@polyu.edu.hk），他将处理您的请求。

Materials availability  材料可用性声明

This study did not generate new unique reagents.
本研究未产生新型独特试剂。

Data and code availability
数据与代码可用性声明

The bulk RNA sequencing, the 16S RNA sequencing, and the scRNA-seq datasets have been submitted to the NCBI Gene Expression Omnibus (GEO) database repository and can be accessed using the accession numbers GSE242467, GSE242468, and GSE242469. All data were accessible to the public. There is no original code reported in this paper. For any further information needed to reanalyze the data, please contact the lead author, Prof. Terence Kin Wah Lee.
批量 RNA 测序数据、16S RNA 测序数据及单细胞 RNA 测序数据集已提交至 NCBI 基因表达综合数据库(GEO)，可通过登录号 GSE242467 、 GSE242468 和 GSE242469 获取。所有数据均向公众开放。本文未报告原始代码。如需重新分析数据的其他信息，请联系通讯作者 Terence Kin Wah Lee 教授。

Key resources table  关键资源表

Experimental model and study participant details

Method details

Quantification and statistical analysis

GraphPad Prism 8 (GraphPad Software) was used to perform statistical analyses. The two-tailed unpaired t test, wherever appropriate, were used for comparisons between experimental groups. Data in graphs are shown as the means and standard deviations, and p values less than 0.05 were considered statistically significant unless stated otherwise. There was no estimate of variation within each group of data. The variance was similar between the groups that were being statistically compared. Kaplan‒Meier survival analysis was used to analyze disease-free survival, and a log rank test was used to determine the statistical significance.